<%BANNER%>

Demand Shifts in Outlet Selection in the United States Market for Fresh Flowers

xml version 1.0 encoding UTF-8
REPORT xmlns http:www.fcla.edudlsmddaitss xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.fcla.edudlsmddaitssdaitssReport.xsd
INGEST IEID E20110318_AAAAMQ INGEST_TIME 2011-03-18T23:11:15Z PACKAGE UFE0011838_00001
AGREEMENT_INFO ACCOUNT UF PROJECT UFDC
FILES
FILE SIZE 6416 DFID F20110318_AACMJY ORIGIN DEPOSITOR PATH iniguez_c_Page_084thm.jpg GLOBAL false PRESERVATION BIT MESSAGE_DIGEST ALGORITHM MD5
1a1247451aef18e851f5bd50f1037ab7
SHA-1
3e99d794f3bac71bc908499b571c92238c7dc48b
420668 F20110318_AACMLB iniguez_c_Page_098.jp2
352a671467b806adffb6fd05a688f8dd
8c395150330c80137e7bbab2336c2d9783d6770e
908297 F20110318_AACMKN iniguez_c_Page_109.jp2
e6e466d4e705a34308544ee73b371c0a
219c171bf6502f528f0cc71305897e2e8e57b1df
22216 F20110318_AACMJZ iniguez_c_Page_099.pro
2bc92d2a9bbc96d7af4f7a6457c653d1
db17c3a1211d9e92c8224b41eaea31cc1f8fdae0
25271604 F20110318_AACMLC iniguez_c_Page_103.tif
5a8094957b3a6c55b3cbaf6b13b0703b
916654573b7504dbc590ced2a8ac99e9aca2d1b4
46465 F20110318_AACMKO iniguez_c_Page_053.pro
1f40d52a03d9bf06e55cf3a324052b32
125f87f009edeb9a57551083cdd162dc0f8f83cd
6103 F20110318_AACMLD iniguez_c_Page_086thm.jpg
7fb29a4a5b7d5043b25b462ecab3f6f1
ac45360bc7828956632d3517b255a804786c01da
44218 F20110318_AACMLE iniguez_c_Page_054.pro
66b357ec8264e8a3e29a939786e87f9c
22ade5f1d232f1e0a3451ed7a2ee25b612486543
4644 F20110318_AACMKP iniguez_c_Page_137thm.jpg
387b56a3d427aa8f40ff56d6abc7148b
66983792fa954699de98fdee2a86d234ebb5f15a
27073 F20110318_AACMLF iniguez_c_Page_083.QC.jpg
a7d62bbbc6932d5b4aa0ee1d1b1df1dc
2fd1cef92a66d5cc0100be9022d1ea9514937525
26124 F20110318_AACMKQ iniguez_c_Page_078.QC.jpg
a257228d708087e6b577094d6d66fcf0
16cd3f45c03b18775e7a83f9e274944d9f497140
83765 F20110318_AACMLG iniguez_c_Page_016.jpg
0cc622d1f25c496acb377b9f8ad6d7e0
5a89bed03fdd297a1f196d109fb766d3fc7f1fee
40298 F20110318_AACMKR iniguez_c_Page_013.pro
6e478bbb2fd1d960e4b46cbe7b901dec
db1e0799d93e87ceb92f4ad637971a9c0116ecde
106726 F20110318_AACMLH iniguez_c_Page_024.jp2
a729f6823dbb904a1024742a6a1a34de
b69178926f45b706fc2d22b97767426484d61c8b
15546 F20110318_AACMKS iniguez_c_Page_059.QC.jpg
31e11bde2cd914caddb3d845027e547a
21c11f3c15a8607a6f2dc0f9720af0c3a638381e
3442 F20110318_AACMLI iniguez_c_Page_062thm.jpg
b6e12c23c8e67a9881bbaa5417a2c707
f13ac5fafcf1d5f8d20d2af9878e95b363775320
28277 F20110318_AACMKT iniguez_c_Page_137.pro
452b007066702931103a26768c5a6ddc
c9ee97c4ade717764032a07761ee1ac0a19e203a
1507 F20110318_AACMLJ iniguez_c_Page_082.txt
93af882a95aff6928ef1c41fc4d13928
d8e09105e911eae51719cdc74b77a93cfa0baf23
725012 F20110318_AACMKU iniguez_c_Page_118.jp2
c2d5da4ee2933a381811ab8c705aba2e
27eb6c3236a391e6c587f591e38402be817a40df
74923 F20110318_AACMLK iniguez_c_Page_041.jpg
c0ae0cf7649e3baf87707126a18e0f1d
c2e670d174a862c587c38de886d136c012bda78f
29836 F20110318_AACMKV iniguez_c_Page_130.pro
cdb53346a975c3a35be1c223d37ffa37
da1731f17a18b7a15bba6ff2dc9a0c2f82909bfe
65632 F20110318_AACMLL iniguez_c_Page_008.pro
4b0c9ed49b4b166305b34a5793f8b7c5
25ad5395469bf8cd97e110cf33cb37aea18c732b
1051985 F20110318_AACMKW iniguez_c_Page_044.jp2
7bf651b8696fdcbfcde3af823480936f
0434c0490b9c03c85dccce69426d1f078f4ee954
2673 F20110318_AACMMA iniguez_c_Page_008.txt
d405d2a5930b961ba046ae73527536af
1c67d92acce5cb551a854531ddf9ccdc5f8bfcd7
81079 F20110318_AACMLM iniguez_c_Page_018.jpg
d5ce2c4331dc26a34d3447577c8246ed
27033c73d9f3f894d58586320bd3efa4281a2567
14527 F20110318_AACMKX iniguez_c_Page_062.QC.jpg
512946182971cf8604bcb210c1b47085
4aa7f016cacdf98d22c9f654a6482aba732b136c
60207 F20110318_AACMMB iniguez_c_Page_004.jpg
cb46460120bfd3f578afe818209e9d57
a57c44c2a3d02d2da03c1995470f030c9cc8ab53
F20110318_AACMLN iniguez_c_Page_099.tif
854c37a8e0f8f2aca4874b93d7c1cfa6
3008446711df3995a988c9c09362b991c13bfd70
79003 F20110318_AACMKY iniguez_c_Page_027.jpg
b69d969bf6dfb3e4cd16df1d441a0427
f6b5565042d77d0e73784635b0e5eb9e07d33876
84918 F20110318_AACMMC iniguez_c_Page_135.jp2
9d2ba3950ba5c2450c56f95d558bbe4e
80a2ad3d2e12e34669fe12c9393f880e3c7d00b8
6466 F20110318_AACMLO iniguez_c_Page_113thm.jpg
b85dd8a5cc4a779815d8b9007f72f12b
29e424eea65d35a35a4c58df8ec12f2bfd76fc35
4237 F20110318_AACMKZ iniguez_c_Page_136thm.jpg
3c04d3f7aa60cd7a926a097ed0a2583d
e42e0671e98a7c97681fb31344c48a4bd692dad1
41810 F20110318_AACMMD iniguez_c_Page_117.pro
950097202832cb316a46b2da89982ac0
7eb1f3e043a87715c31d01ced50d2e787aeace10
1053954 F20110318_AACMLP iniguez_c_Page_058.tif
9295d083b4a994f31cb11c1f7cdf6e41
59a36963f34393166921bb0b5f64ff9b203c87d0
6273 F20110318_AACMME iniguez_c_Page_040thm.jpg
25051075d6c453204ffcccf297f8bbb2
5cfe0637810c873bfeb05e44513271022d557eba
38767 F20110318_AACMMF iniguez_c_Page_051.pro
fb62fc17d4a2099594b151f0c5db13cd
e43b2e7912dbfbece6e541d09aeaeee3d97fedc9
55063 F20110318_AACMLQ iniguez_c_Page_062.pro
ecce7f811fd8b5fca7683446233ce6df
c012449c57be32b18290b35015b1e0f093b2b7f9
1881 F20110318_AACMMG iniguez_c_Page_107.txt
3c6b57f8eb7783b30bcb1fac2562b429
516601538570ec11a1007d9809ff80314d745b61
5723 F20110318_AACMLR iniguez_c_Page_050thm.jpg
fc05a169172b5d4ac1dba4abd355ed6a
e879b5d343fac2d0e02318550df6f2e3294c97fb
6414 F20110318_AACMMH iniguez_c_Page_121thm.jpg
46fc0b44864803fca46598a2aa7a361e
3847bf2e4ba4eba47ae54d03ba550102fe6c0da5
5688 F20110318_AACMLS iniguez_c_Page_117thm.jpg
aa1ac69b52862352d79f3093beb0cf5c
59b73bcd07183d642aa0ff390f443ab475841c64
F20110318_AACMMI iniguez_c_Page_049.tif
b68dc346a61918165922497ee99fd5fc
d479ff31c3cf6b30c68a07069546c00428f8c5c2
1458 F20110318_AACMLT iniguez_c_Page_066.txt
177c2c67e81ce3cf45ec77bed04ea286
eea495e8be6ae417c8ae8f003e5dfbd104a8f6c1
5309 F20110318_AACMMJ iniguez_c_Page_022thm.jpg
f78fef77dbcb4599be663336f23ad476
c9c78b7880bde78082830fe97bc889122437e647
48297 F20110318_AACMLU iniguez_c_Page_090.pro
fc197e52f68f1fa09b236ec98e479940
1bb654293c805b7ea72dd668ea60c87d01161e8c
1092 F20110318_AACMMK iniguez_c_Page_037.txt
9b6b5a56490b3a3fab1134c0813dbaec
c7ed082b63475f5f187a07c61169b18d5d46edef
75937 F20110318_AACMLV iniguez_c_Page_053.jpg
1d2b928ae95098b711cc19769de67045
edece2bfdea4ba103012ea56bfe13cd629b7d151
5078 F20110318_AACMML iniguez_c_Page_134thm.jpg
d240ad0f83bef6125f0cb2b80aad7eec
6451360c61a8f6379b161d822f9cee9a86850581
2167 F20110318_AACMLW iniguez_c_Page_118.txt
b7dceda3b51f67dadf1058d6ba975ac3
4a0e9c0298001c034ef8736d54caeedc861cb768
87503 F20110318_AACMMM iniguez_c_Page_134.jp2
6cf2151a8de2cdafddcd10de916e13f9
552b303531005d21e3193b0ad97fd6f70d02b2e0
19674 F20110318_AACMLX iniguez_c_Page_014.pro
217f772f9b6b8bdac6c7cdf2bc166261
4086f1bb8d26004fbb872a6e57b0893f233b7217
27169 F20110318_AACMNA iniguez_c_Page_095.QC.jpg
f07274a31b241a1f7c96641c1f7f72fb
e37005633ce0c0b2287f06df9208ec974bba51e9
48576 F20110318_AACMMN iniguez_c_Page_150.jp2
5dfc7e4d905d63e369ba07b46e92c2b4
5842da4021a0073208975234a3277b3ff210aa77
24851 F20110318_AACMLY iniguez_c_Page_027.QC.jpg
c156a2c8feb295630b342bd1076a86fa
d9aaf87201a03120aed6814b7e0cf0d0ac157e8b
F20110318_AACMNB iniguez_c_Page_052.tif
329298fc96217320374ed3b6d83d990f
64c93821b35bafeef98e933e6f2756d3cc836d85
23677 F20110318_AACMMO iniguez_c_Page_053.QC.jpg
f76f26b3df6579dab7fab0b0dd1105ee
85cc1dc94a84f8e7ac51483cd7aa77d8fb783647
1999 F20110318_AACMLZ iniguez_c_Page_065.txt
7d220efda57c66110d5a331c78ccdf45
8b9560c271e4ded65cc116312b35376387dac9e2
1751 F20110318_AACMNC iniguez_c_Page_075.txt
2096741edf5f5d559d5eeeb85c1b7b06
6a31f7d29255a60c4f86b7b874a7c39526a80bbc
63932 F20110318_AACMMP iniguez_c_Page_103.jpg
825537d6b26499e927f120aef95e785f
342a0505e161f0c6617393877ffebdceb591f15e
29156 F20110318_AACMND iniguez_c_Page_087.pro
b673f003bb03d1c31e2cc6bc339276b1
14018aaec33008c8e0c2b18455ec4ff635685bb1
60841 F20110318_AACMMQ iniguez_c_Page_116.jpg
875aebb1657266e913b2d0ddcce66125
258d6cbdb5322e820546d03b79ffe4054903a98a
5387 F20110318_AACMNE iniguez_c_Page_007.QC.jpg
da1fcf33d2909a1e44693c449feb73b3
135aa459d0fd6d4485dfd374b0b3e8d53ded251d
5911 F20110318_AACMNF iniguez_c_Page_044thm.jpg
2c4706485711774788211225610efdc8
a246053e7212e6a947e7dc47dbd27e9605618c8e
4397 F20110318_AACMMR iniguez_c_Page_056thm.jpg
c902f21c37791aff591a757b25044671
55421a118f2f439f422636ce500746a51d67e0e9
47190 F20110318_AACMNG iniguez_c_Page_094.pro
7884783d4e59535273d3a675de61e75a
5ce8b541636d8bcb2d42ef7fbc99d0dc523cfaa6
27396 F20110318_AACMMS iniguez_c_Page_110.pro
bdbb03564d1202c014350f0131582c01
de560826be221c032f0c4e01c23c1382e195257e
5497 F20110318_AACMNH iniguez_c_Page_087thm.jpg
40be0ff81c927557e92b82f3b0b7cee2
039e360789570458bdd095fff839b58c8452b102
F20110318_AACMMT iniguez_c_Page_083.jp2
ea26421059c70100b25f2b02b5bee11a
62c402ec029b3f45d3f3a85776b2810d8c323bd2
20656 F20110318_AACMNI iniguez_c_Page_051.QC.jpg
523b786d7c3a70b247721cb4fb65b2bc
9da276eb03e2fdb4968e3261ea935e2e617209cd
F20110318_AACMNJ iniguez_c_Page_094.tif
504a06994a87df0cc10484298b729a12
eff849552029a2cf5d8ec641f2c43b972dc01b02
33160 F20110318_AACMMU iniguez_c_Page_135.pro
538f8f297767aa334ecd5d91182a6f29
20487bc27eaba288d41c6ebc21ef304f9902c71f
894627 F20110318_AACMNK iniguez_c_Page_111.jp2
057e747ba7810b4e71676855e69e3ad2
c0a7e831bd42de7aa453e0e291f730c278610511
25812 F20110318_AACMMV iniguez_c_Page_025.QC.jpg
708f20e8eab75fd7701ace9afb67a78c
c7eb8725423d3cd126dfd1b82116f348eb00694a
1333 F20110318_AACMNL iniguez_c_Page_081.txt
38240017e94f8c48c2d3fcab5c9e5c42
a3793f076dcaf637e8463e64fe6e068a578ed305
64173 F20110318_AACMMW iniguez_c_Page_136.jpg
d2d91e9444c734f745cf79e31b9a3fc0
a85d182616e131eea28bb9129943f30df7e829e5
41737 F20110318_AACMOA iniguez_c_Page_044.pro
a0c8876de85a3a7d9f892eeae1352522
ca7d776efbacdd46b20741f672b6167b90e120e7
10153 F20110318_AACMNM iniguez_c_Page_098.QC.jpg
56f578b656712e4de9b31095b4929273
073c403bd57f74ef0639b100e3422c53d6390dd9
30582 F20110318_AACMMX iniguez_c_Page_011.QC.jpg
3467eee19baac1a86acaf6695dbb3fd0
13350b7562f08d1ca9330cb9903ebf0fbcd58771
3770 F20110318_AACMOB iniguez_c_Page_005thm.jpg
206054dfcacde635b75857a20885e470
cb5f57b4452f269702b042b3e676d2b8f536e07b
113803 F20110318_AACMNN iniguez_c_Page_084.jp2
03789a36a8079042dc6f9bfbde943c13
1dbf9a03cce25a9f81d9d02d085aba7f21272843
F20110318_AACMMY iniguez_c_Page_140.tif
a235bcce829d666f2a5ef36ad77f359e
7a3e7d8c4c36b92219e233f51081bc744ef25144
61553 F20110318_AACMOC iniguez_c_Page_132.jpg
eba74f4023034bda0deb4c4095b0ef71
7a5f0ec4311932414aea10d95a6cd244a02ac0a2
4238 F20110318_AACMNO iniguez_c_Page_141thm.jpg
59064b2e64ad8aac61e80304461308c9
0875d7afe7265cac0760ecda15a4e166ae0f375e
43842 F20110318_AACMMZ iniguez_c_Page_057.jpg
dbb7fc639d84bbf905e9ef3fd0551e7c
2df7f9f82f9a9e0eff2e0b896e149be3c06e2c94
4428 F20110318_AACMOD iniguez_c_Page_130thm.jpg
f3ca88c994036c4b7b295d63d64e7b96
ca9abaa97127ab867ff4c395f5629b461cb93513
27417 F20110318_AACMNP iniguez_c_Page_077.pro
f0bfcc8c3b0327d6947ea67bb2d13781
9fa0a1a0291558c89b798209725ae42eabf7577b
F20110318_AACMOE iniguez_c_Page_040.tif
08c82f00094ae542d73f754f4830c1b8
36113c610a1ca912a1e9a5d1a9c283026ef1a82a
64454 F20110318_AACMNQ iniguez_c_Page_104.jpg
852e7f0b689eb1a77c13167572809425
fce239c3d7565ca13afed1f2606581f0273a1a82
1025 F20110318_AACMOF iniguez_c_Page_126.txt
6963d0adf34c73d77dd0900dc5b57320
220c7810cdb78692852554caae1b14acaab71dab
35824 F20110318_AACMNR iniguez_c_Page_128.pro
a3424ac4b97a976af5f23eafe90bf0fc
0fa7749ab1222f2125054fb20cf489965955da8d
2033 F20110318_AACMOG iniguez_c_Page_030.txt
f131a62962b1aa91de91828d216b3b6c
09aa34194f377abe73a974624c5ebb6f2c6525d0
2717 F20110318_AACMOH iniguez_c_Page_111.txt
0ada5b710e7f4ae17a83247f4ae1eb34
ebb55ab7d8ecab48a54733f9775f0e29eada61f1
1466 F20110318_AACMNS iniguez_c_Page_149.txt
2bc40609d1c8f500bbd39a4df6d3c80f
704c2247620d33107c5a70f19d6c060d8fd21e52
1341 F20110318_AACMOI iniguez_c_Page_139.txt
24e2344a10cbdbdf276de4baa2f197f0
0118de567e8faf6b93a3d5a0895820b189b48ca5
2406 F20110318_AACMNT iniguez_c_Page_063.txt
6c0ace5e9367818f32d8da7875637ab3
ffc42bf6ce63d68e223c4bcdcbaca6a07f835cd5
19914 F20110318_AACMOJ iniguez_c_Page_038.pro
14cc57cbc5c15d566324370c8df4815d
e76bfae055075d84d45d73063c7315b50b10a589
F20110318_AACMNU iniguez_c_Page_124.tif
32d881ad6fe27993dad3b6f71777ee9b
f88ebbf0cc3515ee558177a53d6071de45a10d37
816231 F20110318_AACMOK iniguez_c_Page_037.jp2
3eec6a5a30a47ea638d1060fe43cc012
434cb827301044ca93ac6a698ec4675a77920512
5931 F20110318_AACMNV iniguez_c_Page_054thm.jpg
aad9b4b0be37e4859a9d2c9b0b5e2282
dc876eaaf26d2b3d209d25cbd3625c30503587de
6571 F20110318_AACMOL iniguez_c_Page_109thm.jpg
6bd7299b3f8718f0c01d8026c0474c4d
0abc1a2237f89329b2eb6b043f5a177f1166995a
59970 F20110318_AACMNW iniguez_c_Page_032.jpg
a9118e7d28351d4a4eaf7df96438fb32
93cd1cfe6453d1dccf18a2e3dc6be14066dc61e7
19008 F20110318_AACMOM iniguez_c_Page_100.QC.jpg
1d071ca36502a24b1794e1cee0e72cd8
4f7c25e98ac74be8d2270866e284a77afa358ece
95887 F20110318_AACMNX iniguez_c_Page_070.jp2
f295e066094b30dad6c5227c0989d641
118dc7cc6b6951171e1508370b6b0754d1109b59
114840 F20110318_AACMPA iniguez_c_Page_073.jp2
d11f910c55e375e41b103ec4537999c1
3ff36a5d3b4fc629772b44a3f45c601d2ceb9d7c
1545 F20110318_AACMON iniguez_c_Page_142.txt
3d9f0f9fb79b4a3124eefb3ea4b2ac39
19f1223eedca74728dc9a2f2f4cb76277e527142
18670 F20110318_AACMNY iniguez_c_Page_101.QC.jpg
cb712efaf44f34db8eb01b64c0bd6556
f41bbd33c2f9efd0a60a981709d4e6133e3c6a95
21694 F20110318_AACMPB iniguez_c_Page_049.QC.jpg
a204079b26728ccb17b26d66588b5e2c
62e804e2352d4d70e53bc438a95b44007e0b7d72
888494 F20110318_AACMOO iniguez_c_Page_081.jp2
31a1220046377eafc22be4ab998c8329
9705e06a0aee4529b9ee530b40764e10027ee442
1264 F20110318_AACMNZ iniguez_c_Page_130.txt
0d76937a85de89ced2245f74728d4d2c
7df23935a19fb02929053f5d856199701592b74c
22626 F20110318_AACMPC iniguez_c_Page_041.QC.jpg
b5c37c165631bf38e0cc31c1c4833b5a
3bc95f89d51f1c253d694faa116995a7721f8e4d
5464 F20110318_AACMOP iniguez_c_Page_042thm.jpg
291b9cb1590a3bd1cc883c4e83e50411
80dc6c83218a7ea938aa8bebb5e6a2b58b134144
F20110318_AACMPD iniguez_c_Page_148.tif
669bb614f8bbd93e4506b3087fa2afd5
e7f85409db917a545f033ffaa5a5820faa482cec
2586 F20110318_AACMOQ iniguez_c_Page_120.txt
a209049ceb34d38d81f9ea6e54085755
4bfa0a94aec445a451eb15b9cf1902834dbe7f4f
715124 F20110318_AACMPE iniguez_c_Page_086.jp2
71a2072b99f9a533d4ac96310d1f8dd7
9d86d375b9da1b22232efcdd51f7c4935f906b98
6441 F20110318_AACMOR iniguez_c_Page_038thm.jpg
90b2bd7e5e34fefdccf2e479ed73b71e
bbeb529c541b4b7f788c48f0fde08aa89fcadd3a
F20110318_AACMPF iniguez_c_Page_002.tif
ca695abbdb6f8a9a33db521f42e5c142
661499aabdaa06d2296d60753b5f2d756fd66427
1875 F20110318_AACMOS iniguez_c_Page_089.txt
a195425a3a574e6500d3fb5fbd17044f
dc41d848c95fd36a9a8a4f8cde6c70eafa72a69c
52808 F20110318_AACMPG iniguez_c_Page_096.pro
cc13876dd60bba276275a1d8f59cd505
5e29aa71a37dfe3a2d916f575e23b945a10ba927
34726 F20110318_AACMPH iniguez_c_Page_004.pro
86a5e3c22c92d4acdd610d4dfdac5c49
cda9574b4a09dc9cc2143307b0eb072a80b9574e
36372 F20110318_AACMOT iniguez_c_Page_109.pro
9b4ba4d794a9d67fedf2db21c1b40bd2
f3fbb3ca8c57732bf2155c1fa78840af89136bd7
71890 F20110318_AACMPI iniguez_c_Page_081.jpg
f8675fdb84e659897e36d9c7e4d3a83f
09b7a6c4a1da1e4e8ea9fbc5d033f3692403f113
F20110318_AACMOU iniguez_c_Page_041.tif
ef41aef74d95474c32a91848c9b00671
6a0e72039c5121bb5043db554cc0d8dddf1a946e
5016 F20110318_AACMPJ iniguez_c_Page_103thm.jpg
dc5427027b32d84fbb54680b66560c60
1125d3f57bd25fe3b8c2c6d9339776b74458d84a
28205 F20110318_AACMOV iniguez_c_Page_144.pro
4e786012b92dd9e927932d633758b473
8f13f5716ec1a5d984688f35616128bed1073959
74956 F20110318_AACMPK iniguez_c_Page_113.jpg
ef3a91837bfefba4b6c8f1433c788bde
a47522386b9989a677963549bd2b81a3e70ada5f
4626 F20110318_AACMOW iniguez_c_Page_032thm.jpg
b9660d2e491d475f37645c1a1e002e22
f2b53c3dd4bc08d70ec6accd268a812f23a64aee
18769 F20110318_AACMPL iniguez_c_Page_143.QC.jpg
4ade5cd803a3f4a30bc025036e4b5486
b2f25a6f31ae8a96a5b4301ea717fad924e499bc
1054428 F20110318_AACMOX iniguez_c_Page_063.tif
65958e31df1a3890929c121336b9aa47
8135cecad2323697f81cde6c3c864922adf76558
F20110318_AACMQA iniguez_c_Page_085.tif
03dbe6d1073086dbcb7ab73116826898
cc64b1deeae31ad404bd6aa25db1756794bb0009
800956 F20110318_AACMPM iniguez_c_Page_115.jp2
1e3ea8d3476b3a457bc0c124c536101d
5308885f75035a8ac380ce3a78276cb08b901aa2
2426 F20110318_AACMOY iniguez_c_Page_062.txt
0782ffb7aa706ba6f6a8017284414a99
0cc53cbc3b3cbff3f385cd725a5ac29304bc862b
50123 F20110318_AACMQB iniguez_c_Page_078.pro
3f19912d843fac15eb69abaab96c3fee
a41dde79e1d7c38f4f4aa03364f4c8e7d547b719
2495 F20110318_AACMPN iniguez_c_Page_117.txt
013de337215c8d2c020a043d55065a19
12fdfb65bbc61c0f8fed356be90fa0f08a040858
2617 F20110318_AACMOZ iniguez_c_Page_009.txt
0a0a9e35a8faff107ee5c7bdf614a3a9
340e0f723f112b5296583b718ba6f5486aea3348
82683 F20110318_AACMQC iniguez_c_Page_078.jpg
cccf1365d1c6d2e4e84b011f52db9a93
31ecf1b398abce7fa5c3b8dc12dc3d750d40f841
5648 F20110318_AACMPO iniguez_c_Page_066thm.jpg
103b0f7eb271f42f18911871886337eb
be00c98dc230bec671f486369f098d228914f188
51740 F20110318_AACMQD iniguez_c_Page_029.pro
9a641bf1ed889406ffe93592b50c4af9
4c5e9268f2f672fc60f0eab4671c5802b60b308e
27703 F20110318_AACMPP iniguez_c_Page_140.pro
d09978c441b11acd1832c6efa22dfe25
327e6535c1597f656c84ad7b9d82a9d9dcd138e5
32972 F20110318_AACMQE iniguez_c_Page_100.pro
c5eca6ba2a1457bd02c1056ab2dc179d
fe790db8a99a94f255ad7cf33e2207ccc7109627
1130 F20110318_AACMPQ iniguez_c_Page_136.txt
caeae7fdcc3a6428c679668cb8c332f4
3242c94fcd191019df0c11bb7f55e57567fc4422
19291 F20110318_AACMQF iniguez_c_Page_120.QC.jpg
3fc9becb5b8a4123100e3ce1b214f336
c4a2c0372985c3f3ceb2c31da8354f85bb29e90e
64679 F20110318_AACMPR iniguez_c_Page_143.jpg
964d014cc5cbf734a0061f44d0f530d8
d6c9cc220bd5d21e1eaaf99a022acbd470b1af1a
6550 F20110318_AACMQG iniguez_c_Page_073thm.jpg
76c4f7e3857125fba3146005ebe2ebac
ff262abf39ce87db0239dada1bc7fcc39d9c6469
933134 F20110318_AACMPS iniguez_c_Page_102.jp2
bca18440c2c1829872481ffa428eabcd
9add99da076474e079fa8ee3981873f6fb050f7a
71530 F20110318_AACMQH iniguez_c_Page_145.jp2
75d8bf5d14cb40e6235210204b064f52
407e529d16c70ef00fca0f9913ef736d1a5461f2
F20110318_AACMPT iniguez_c_Page_010.tif
cbbde2e7ed0f720de8f432f3bf641941
08fef25a2971daa2ca4efded6b0c47691e5440bc
4541 F20110318_AACMQI iniguez_c_Page_127thm.jpg
3ac8dea7068348cb7b4f17e8f67c32bf
7af430d1fa596c2709633978499afa3787bb9e8c
52140 F20110318_AACMQJ iniguez_c_Page_040.pro
f9349e8aa9601b6ffe98ace1bd034323
4ee1838c7074282f29bf8daeff7dd2647b77bd75
16849 F20110318_AACMPU iniguez_c_Page_056.QC.jpg
be8a947d779b574d9c3f7f6b92d2f4b1
dd8a3dc15673a3cdb48f69554524a0eba71dcd59
2006 F20110318_AACMQK iniguez_c_Page_025.txt
340dc3ebf5bc13f4b560e1d437190308
01ae5c9a91e0e613e80df9f5c282492797b2903d
2099 F20110318_AACMPV iniguez_c_Page_096.txt
6aeb298d2976cd3966704d909ce7d459
59b6557c6c1e5b5b65d746247cb450fbc9ef24f6
66539 F20110318_AACMQL iniguez_c_Page_133.jpg
a1d15dada237279eae84269b3bfc23f1
814ebc8fad295c3384a70639485e4aaccc90daed
5856 F20110318_AACMPW iniguez_c_Page_118thm.jpg
cfff3a76b7fe8826d47dd04009d002cf
6a3d348c24987cd497e588d728b9ac5605171ed5
114375 F20110318_AACMRA iniguez_c_Page_036.jp2
efb3a9076cd3be029b1eacd10ea098a7
a40e54eb4784a35df919fde48ee77d1a3d6cf93f
69056 F20110318_AACMQM iniguez_c_Page_043.jpg
9ebee0aea09adaa3cbd8eba4654dea0a
707407073991a9b98952eb8559f3df2afbb3b5b8
19900 F20110318_AACMPX iniguez_c_Page_039.pro
4beec10ab30a486f555040a42fc78834
c2128610727897b1340e094fed42848b90d44b7b
2965 F20110318_AACMRB iniguez_c_Page_014thm.jpg
8629bf4f816e78d9434868dd9c911a10
d90e85f74f687257236d4dbbbe1cc8cca8cee532
20266 F20110318_AACMQN iniguez_c_Page_116.QC.jpg
05aaf87480d1b55a51070a95dc8691e0
a71c21d458afd0b69377010cb54584678b133ef3
28015 F20110318_AACMPY iniguez_c_Page_133.pro
9fca2c56f324011aacd40678c32c4d8c
8fd06fd16b13872af16c5d409e0094c767d1f0a7
1756 F20110318_AACMRC iniguez_c_Page_061.txt
fec1ac7d7da2b40513b9d4083d404cfa
13fb81c216cbce394b10f552609ec07d9a2fdbd6
6175 F20110318_AACMQO iniguez_c_Page_078thm.jpg
8455d8e3fc38780fdf4287427b2c2a31
29d89ea57aa2558dcaf822533c7fd9d2bab06b4d
75945 F20110318_AACMPZ iniguez_c_Page_133.jp2
afd8848fff914cf5aa33a9d972376c10
7738bf6ed9bfc26692c455ed8c72e9dbacac387e
83948 F20110318_AACMRD iniguez_c_Page_035.jpg
dd85cdc96c129f87d5ac1653f56b9833
4de31e117e8c1dcad102999f81c3cc44f2cf57e2
795331 F20110318_AACMQP iniguez_c_Page_087.jp2
c65c282832fde903e848ed4d468ecc2c
0b667901635723a4c1deef0347ee35079caa6690
20231 F20110318_AACMRE iniguez_c_Page_139.QC.jpg
097f1bd136e0ab563e681ead8e08e7df
398d30081c37711b85ce49a72cfd1ccf7e9871bc
19448 F20110318_AACMQQ iniguez_c_Page_130.QC.jpg
8cfef5b282898cbd3ad165028b972f5d
81f3419cd8d06f979ddf34a6e04067c70693a19d
18453 F20110318_AACMRF iniguez_c_Page_038.QC.jpg
2dcd41c5bef35537c27ec79cf9bacb9c
4406e969ddd56ed1c78d304a903dc3cda38172a3
1239 F20110318_AACMQR iniguez_c_Page_123.txt
e8f4763a25c807fd2c35847076adb645
114ac492b94847f0118f89cfc57a4f82c138512d
6190 F20110318_AACMRG iniguez_c_Page_105thm.jpg
86386fcc357823d2312911f78631cd24
b8d496b7a9ad5e17ceec36f929a1b5364eb919ef
1953 F20110318_AACMQS iniguez_c_Page_041.txt
e38f58bf4861f855997ce2ca5616df9f
4aa939f5cab43099d62a8e34c56d2331e05ab99b
76113 F20110318_AACMRH iniguez_c_Page_044.jpg
c60d285eaaefa1ea6b3c40f0f5b2ae5d
6ae7726a18d7f13836fb4ca096047499364d9099
F20110318_AACMQT iniguez_c_Page_026.tif
d9110ee72308fa746940336e0fbf9da8
fafa5fc64b317621d9bc157e720f22c2aced5f1b
1474 F20110318_AACMRI iniguez_c_Page_131.txt
77a015a91952a06fd6c658f11f6ebbf1
96bd9d746923bccecf7a29740692258abd6ff28a
F20110318_AACMQU iniguez_c_Page_060.tif
2b6254c20f18f1a537d29e22ab1e4633
e9b9b361f4eb620613256c0d7fe9f75f5b74e4ee
F20110318_AACMRJ iniguez_c_Page_033.tif
24cf6888b281b106bf2b1a8457456b46
062674f8912b285288059d64ba5f9fb3014bf487
2003 F20110318_AACMRK iniguez_c_Page_028.txt
1d3c18e2f813d60418237466d530ace2
329923cfd4c4aa895550e1e5a9b3daadf430a391
87546 F20110318_AACMQV iniguez_c_Page_125.jp2
f9571c3498dba4945f41cce37ab9bff7
b730c9971aade5571f135b2e2d2c7b147130579b
F20110318_AACMRL iniguez_c_Page_139.tif
9dbc9d4eda526388bf178fffd3c794a1
8e76ec912afa972a8af1f22ce2263585c7b3eed1
5152 F20110318_AACMQW iniguez_c_Page_124thm.jpg
0f9c006798d1c33d9bf56c6db23a1ed1
223c93012b7d2aa525371b688d209f19e2f4a692
59752 F20110318_AACMRM iniguez_c_Page_087.jpg
b50de11fb68769aa29a6a84e55a8b288
93073ff8e6451bc8805fa454ef7786458ca53e7c
4403 F20110318_AACMQX iniguez_c_Page_059thm.jpg
02dd2a5bf8fef8a7bb305d1b0000bc01
581c29079ed156b0021624bf39700632583f9ee9
47399 F20110318_AACMSA iniguez_c_Page_059.jpg
ce379532cc18b0fa8af36a1f31e8a19a
7261928cbab09d6394aca1f5979824429cb34bff
87198 F20110318_AACMRN iniguez_c_Page_084.jpg
fd6e1ed1641d41d7be74cc74870ac208
ca0c267d9336d1a9b3bd416354ee793c4833ae5f
2815 F20110318_AACMQY iniguez_c_Page_116.txt
3e5d45a96c0cf0794217eb29682d8b8d
a3e74ba241874340687d7e4adad9cc974d4d8641
5118 F20110318_AACMSB iniguez_c_Page_058thm.jpg
430e2c178ee2268c4666893461e575a6
e825eb4f1db388dfaf8a73e41b26d0800c512c66
66712 F20110318_AACMRO iniguez_c_Page_077.jpg
fc5f62d4ca135c79aad1407d1fdbe5ba
8b19e94ac57ab5d41e8362540778908af0449f7b
18033 F20110318_AACMQZ iniguez_c_Page_061.QC.jpg
9c7e99c646ee186ecf8cfbdc63654a43
906e5e10840c437d050366b930588f78d8381a6a
23871 F20110318_AACMSC iniguez_c_Page_089.QC.jpg
4998fd0342fe0d5b85f873c7ff74a219
c3f21fe43573ff5447a460877882c2a4ddc1fb8e
22528 F20110318_AACMRP iniguez_c_Page_138.QC.jpg
ab29b78c138d094039aaae2e103dd5e1
97d5837a2393da3cd3490154dd388b4aa9ad64d7
6193 F20110318_AACMSD iniguez_c_Page_016thm.jpg
3efed41fba17e22d8f3da8bd303807d4
b99dc9dbdd0e6ef6b3cdd723cd152732530ed2d6
5262 F20110318_AACMRQ iniguez_c_Page_013thm.jpg
4fb28162c0ba7f7bac78dc5c1d25dcb2
004e4ce6c9c022e34e41ad4697bd9962e4d42f2f
64293 F20110318_AACMSE iniguez_c_Page_126.jp2
c71640d026ca2e590b3d761bfad17537
6f748a5e2d74cf568f03a0e864fceafbaf9d2c8a
F20110318_AACMRR iniguez_c_Page_038.tif
e38864f0290758a3f05817506d05c259
0d3673801eb8cf4a0ea8267bd03b4b3d2c56ccc8
6200 F20110318_AACMSF iniguez_c_Page_019thm.jpg
3bbe72861eaa2f4ba80680edc666d8f3
830a5e04f80151a3b2083a333bbb08f847cadd4b
32514 F20110318_AACMSG iniguez_c_Page_103.pro
55a4a885ec4f458f7b25becef2399381
7321614ccb471986ef930018218fe91b45dad6e3
F20110318_AACMRS iniguez_c_Page_001.tif
b226451b9c3b46b5c03d70552cda520c
93356d302995a7e04dc238048a59c20f1dbd5d82
66250 F20110318_AACMSH iniguez_c_Page_144.jpg
1e441a950ef8974e289ba6382b74c9ef
bca3121e7f9a617c4923a7e73921f94d304c4797
6251 F20110318_AACMRT iniguez_c_Page_082thm.jpg
3fc118964e6c8dba54ff87a1519dbb23
3cb286382b907db1fd286b42471c0344128493cf
4083 F20110318_AACMSI iniguez_c_Page_057thm.jpg
24e8e27a25698ddefe95505527823d1b
862b7916cacf38b9169a1b5cd9928bbe14d4e737
1212 F20110318_AACMRU iniguez_c_Page_143.txt
fa5dc4a5f74d4df3814a77b00b6ef3f3
229ca2c782799239b27458492beef987506c0040
77742 F20110318_AACMSJ iniguez_c_Page_094.jpg
525c682c663472f1a694be10d0d45127
f3b07c2e3932d4b8f8870c1b3cc114fb55b3fccf
F20110318_AACMRV iniguez_c_Page_113.tif
4e0cc752ca0a567411a0e662872baae3
5653614ba2f87b288e17ca5f6a47c43004d61a80
29609 F20110318_AACMSK iniguez_c_Page_071.pro
541e0299609386f320bdbcaf37e13d91
1c5f19ce66113e0b3114324b33697481aba6f4d3
28230 F20110318_AACMSL iniguez_c_Page_104.pro
91b2844ab290bbc4150a794270776da5
e042a9c148d3ace9994af016b7b8c5c1f94a7c16
66625 F20110318_AACMRW iniguez_c_Page_140.jpg
83bcdeaa7b6f79a2386c86985a8cd47a
1d5ca32c728e25a8162281a1f4862d5915f57f02
1140 F20110318_AACMTA iniguez_c_Page_012.txt
52e5d3a3a4377b20ce8a7d87e8dd847d
73a355cfe4a775e699f5000f7cdf7fa7accceac5
F20110318_AACMSM iniguez_c_Page_129.tif
2b7e266dd051e2343dda03872e9d3aaf
36cd9c964c1832193209dfb760a80b22c2c4e458
2289 F20110318_AACMRX iniguez_c_Page_147.txt
e72a489db73cc07072e4a91a0e34a1e6
9596e7e7fd6cec5a94765b4e38bb51fe7714ead1
119 F20110318_AACMTB iniguez_c_Page_002.txt
b28178b71ee6d2fe7181fe046fab5a42
2cfd864bed6ad3a99eccd7204cc515273eaead2e
1051958 F20110318_AACMSN iniguez_c_Page_011.jp2
4fa378022b59eeeb6c37bd15299cada3
0b5ec247b1f93035d17cb0a7335cf7e05b1d70f5
65610 F20110318_AACMRY iniguez_c_Page_127.jpg
b6997c757cf3c8decbc8f3319e51a5a6
b922fa5121e8ceaaf0e97674de13c405d22e21c7
903 F20110318_AACMTC iniguez_c_Page_098.txt
6a606064741374132fd2626d2782dc8b
13072f8eb6d448e482d61300d983f5297e20ad6c
68008 F20110318_AACMSO iniguez_c_Page_021.jpg
c2d6608f1584af69428de2e6519c4e95
e123025096f9addf72f802e24022c62b599084db
F20110318_AACMRZ iniguez_c_Page_064.tif
d0d2783ab89431dc2cf9b99b3a559c07
330ca9897e0ab97ba668a9c0c5cf03adc88522be
3067 F20110318_AACMTD iniguez_c_Page_010.txt
662c34ef5d45599d32591ca66cad9359
4e9aceb7121b09b845b5bbd9246d6b5bb3959339
31545 F20110318_AACMSP iniguez_c_Page_139.pro
8de44dc9ecc1628af9c51e6892187faf
a2904f42c853b20def7d4e23066543c7aecdfc05
76712 F20110318_AACMTE iniguez_c_Page_048.jpg
486e2b2d74324190e4381cfbee14516e
cd6a768197d7aad4bfec08d1533aef2a67907ec8
F20110318_AACMSQ iniguez_c_Page_011.tif
51fe8ae790baf07394bdd934746d1172
2426f762e62b922cf9a1c60b082749c17b146549
F20110318_AACMTF iniguez_c_Page_115.tif
d1f3c123fd531444b146a88fbbbd9102
d8c0902788a63f15501f8abbb0acd631d187703d
19212 F20110318_AACMSR iniguez_c_Page_140.QC.jpg
b5bd12b38e7863d54ce37e01ecceab55
c50db852a6620d346890177533e8461742133d89
20865 F20110318_AACMTG iniguez_c_Page_001.jpg
4d02c4712cd4e98da8a61e59a6373310
d4df886c00db8f8e6fe3cc216fd1f72cc3082332
6079 F20110318_AACMSS iniguez_c_Page_049thm.jpg
0ec11dddd64aa175e26c3e61fd1412a3
a44cb67205999f782202f5eb179ce43aa3cc9558
5880 F20110318_AACMTH iniguez_c_Page_116thm.jpg
4b8c4f42968ce1194152b792705b65e6
73047e1ecd6100eb1839116beb95864e6df1aa00
6125 F20110318_AACMST iniguez_c_Page_081thm.jpg
c1762760a3f79f7a55f70c4ecb38cf1a
e7e1185be5b6a9d8bad18b64faae684b83ab724f
104865 F20110318_AACMTI iniguez_c_Page_052.jp2
dce37754d9ed6f80b8a5d6bec98fc9e9
18a529a027a13667c66de7493f2be4c69846b765
47749 F20110318_AACMSU iniguez_c_Page_129.pro
ac61fa375846103bfef7eab8c90c99b5
63bf61123b80f68f61ba459ae022be1cfe1d9360
95316 F20110318_AACMTJ iniguez_c_Page_054.jp2
92ae1e4f8fee1b8dae6b38f121c6ecbf
301183caa7a958059a67d0f9b5511615af126bd4
F20110318_AACMSV iniguez_c_Page_104.tif
56c55dae84059160a8e23c7241f3ee8f
a78cfd0df3010cf4b2b4b827a80fab22fdc487ec
1748 F20110318_AACMTK iniguez_c_Page_050.txt
2cbb6e7bb2d0ab4247fce489d94cb342
aaaa42c2b62e5adf37594119aa75b2e351943079
899609 F20110318_AACMSW iniguez_c_Page_046.jp2
a63573eaff3d0363a2240922532b854e
bdd3f5bbb39fd24f31b71fe09f91efbb62afe41b
F20110318_AACMTL iniguez_c_Page_056.tif
9e97c72f8f33379f36fd59de762ed46b
c3d363ca4b13e417abb72ccd44b57abad7ffa08b
85554 F20110318_AACMTM iniguez_c_Page_040.jpg
3e5e7addc4d775e70719974a381103da
d0065ec1f27dc9d013cdef6d53b391c6701cd809
F20110318_AACMSX iniguez_c_Page_121.tif
44b05c3db33489a6f3750ed07f406322
ba96d513dcb6c66b1b40adaa08c56aa1951a2ec6
F20110318_AACMUA iniguez_c_Page_100.tif
99aec36e7db3c2a9fc63cfa24dfe8bf7
41f00e515659810d305453b30ee7ff798cab3342
F20110318_AACMTN iniguez_c_Page_045.tif
2129181541a71396878822b05fc1b9cb
9f576409292bc61d97ceb9b0896727b0988bbc3f
481 F20110318_AACMSY iniguez_c_Page_007.txt
43c6818b592d8b1b101833de797b7c7b
0ec1a8067bd6299aab90cc592200aff6ce53f6b4
73235 F20110318_AACMUB iniguez_c_Page_011.pro
bcc0e111dbadf4cf1874be3570b67f2f
d98aaae6953e0f1bc94faff53d5a648ee6c19f73
47942 F20110318_AACMTO iniguez_c_Page_052.pro
f3a2dfa9f113795682e78c82c65ad2ec
6e754bf39f1e2b7a34508e9761b29833c986db49
F20110318_AACMSZ iniguez_c_Page_081.tif
a43fa36e98954107b7ade7137908c33b
d380b29910a9db567118d31804b28fdce51a0775
26729 F20110318_AACMUC iniguez_c_Page_074.QC.jpg
82c0fc79462c97a25cc7f6ee25fb713d
5a4428cea1ef83eb124e57d2d0ba6116cd780a9f
22179 F20110318_AACMTP iniguez_c_Page_125.QC.jpg
35dcf1c6bcddcfaa1c30ed9145459f2c
542b849ade451221aa809c2447989039e51749fb
F20110318_AACMUD iniguez_c_Page_075.tif
a89201cc1a84cc64882be5730c2e7e6e
e51c53a51f0f6d296e81d5142ef1e3dbc073a250
25051 F20110318_AACMTQ iniguez_c_Page_017.QC.jpg
b53fdf45bcb7e75123aedb0051dcf1e8
b2fbbc447648b72b66a03005531c531475bed2b8
1407 F20110318_AACMUE iniguez_c_Page_056.txt
b093aead6266756a99e5bd4fd854c73c
6b92665e07839e7aca37e8b1dc48c418c1a53215
F20110318_AACMTR iniguez_c_Page_061.tif
b150178836dc9e32e55c67084ac391f6
30a764457cadd3ac229e36436854d861d1892225
F20110318_AACNAA iniguez_c_Page_022.tif
11f74c970ce47eb675fab65d3a85cbfa
2a4f90b1d3d12bdd9d5fb316d205b8c8ca48169e
6319 F20110318_AACMUF iniguez_c_Page_029thm.jpg
f5b35abc7a68fe31d78db2169530f693
825cb8cee135090acdbd32a240a0219121c5444a
1869 F20110318_AACMTS iniguez_c_Page_033.txt
029408270f759df36db345c92a356e71
584127ec4e6641e934e2cc357928146ab9ca1717
22809 F20110318_AACNAB iniguez_c_Page_037.pro
043a934a8b2372d92c11f124adb8e31f
6d0bfb6df1e303843946f859eca91adbfd0cb142
1173 F20110318_AACMUG iniguez_c_Page_048.txt
3904b83478f450c757726c1e14472f5f
84a51d6730bde2a913f32105e808e2a4cfe7b270
3107 F20110318_AACMTT iniguez_c_Page_012thm.jpg
4349048b91dff024b25b28de33b1e030
74e68bd69b0e694210758618ee5993bed20ec251
5325 F20110318_AACNAC iniguez_c_Page_088thm.jpg
696f2e56a8d5a82bcc37e3a9d9580ec1
ecc71bc38fef7630e67c90e89dd2ae461ed94b9a
F20110318_AACMUH iniguez_c_Page_083.tif
f6873284c3cf2b0df3783101d85c6b00
8207fb5ab4bc6bc97ab1ca9c7293069f73a2067b
1854 F20110318_AACMTU iniguez_c_Page_015.txt
83b6d4556bb8ca7c0612a6fcef26c015
a18a83ed8bc16b6fa74c19a9b0bc09f45977bc31
2228 F20110318_AACNAD iniguez_c_Page_097thm.jpg
232cebe3727cae190680859044ad5e65
d760942e461232f12163a94a2f03bb506e8f3f04
F20110318_AACMUI iniguez_c_Page_046.tif
e0d366c62865ac3787959f2f0b126f0e
2e54cf5aa168cc6c23e717df9b89c714296f9fb3
1992 F20110318_AACMTV iniguez_c_Page_027.txt
712eb98a99dea674f751b71b35fe0e77
30737dab51157a50d0541cae47981bffee484197
34814 F20110318_AACMUJ iniguez_c_Page_050.pro
90babf7e190e1587732b335675d8bc1e
19335089603ab20de70e3a72f932516730a82eaf
27250 F20110318_AACMTW iniguez_c_Page_147.QC.jpg
9704e565561234c0c41fc7dfed3b9cca
904cd74b9ab54c74d25b13cd39bdb7a964998d9f
2079 F20110318_AACNAE iniguez_c_Page_040.txt
440bccf76439531da351e7d55d44b203
46c5daf066906ec0eeff786382c091e63ad1c69a
859567 F20110318_AACMUK iniguez_c_Page_099.jp2
dd2e8144b9f3c58d89c225f917f31495
5d385334c942dd7cf3224e90b21eff4adc92bbbb
79576 F20110318_AACMTX iniguez_c_Page_090.jpg
a4b2a879c6f916d4395d73b6d43ae8ca
0384d35fcf7482c892040c1aaba37715d7c70f82
59700 F20110318_AACNAF iniguez_c_Page_071.jpg
8af5ee96e0ee6ca81741002f84745357
dcafad6ff16f7ff4e75a41404c78de3da8a46570
1511590 F20110318_AACMUL iniguez_c.pdf
c533142c4ad16fef4960d831f5851556
f48b05b96f14754031a558a0ca21e3766a3da9ad
110644 F20110318_AACNAG iniguez_c_Page_029.jp2
f11d47752387ab01428915a348e337fc
c8212a7d82e261b9c84b65a6d4f3b5da8a3a2e19
49187 F20110318_AACMVA iniguez_c_Page_116.pro
66317948ca7539a64ffb08440efe7750
0f12f0db21a9513194f5bf4d52c10e43e5764036
85145 F20110318_AACMUM iniguez_c_Page_091.jpg
7ddceb363d1e8cb7a9306a8e196ae303
4a0ce43ac52472a9d887d9b6e6fe88aba8c7c643
5313 F20110318_AACMTY iniguez_c_Page_125thm.jpg
fc426b73729bfc4af823123ed1900c43
8f84b2fae3215d470f5317faae6e9435459c1f4f
7530 F20110318_AACNAH iniguez_c_Page_048thm.jpg
f60bc295de2ff43b4772fa18ae4c79ca
9e6851b8b230128d4b0ed9cbbdf0a48f5ce0479c
26618 F20110318_AACMVB iniguez_c_Page_029.QC.jpg
d2089c1c9b88661d3a823d215f0be5c3
d12176e81f5ede3f32514a2561cb25b46bfeb424
94728 F20110318_AACMUN iniguez_c_Page_022.jp2
305da49218a3ccfbb1fac49dfeb9fb36
ecf0891e08f1314a737e405910af40b99cfdaf3a
6130 F20110318_AACMTZ iniguez_c_Page_024thm.jpg
3e55ee82a9314a0cf7f423480ecf85aa
8993b94e7aee4b79f517f491e9fefe89b7b9acd4
65979 F20110318_AACNAI iniguez_c_Page_137.jpg
e0037620b752c024704f4510872c50b1
406429e0dd7dc535720b228e17ff05ce855efef9
21504 F20110318_AACMVC iniguez_c_Page_134.QC.jpg
2bc17050aad9ae6756a6d220c3d9529c
66d1dec76e4c0ff46ae692ed9bbae73670e52601
F20110318_AACMUO iniguez_c_Page_114.tif
d070c815d3e0681c251be815d269719d
ced3080ca9fdc397370df838ab1f829209e34ecf
5137 F20110318_AACNAJ iniguez_c_Page_051thm.jpg
411a1a83d0606d6b7e13129404672f1e
64692ee19ce22aef79871f73855e09a10c5f634d
4996 F20110318_AACMVD iniguez_c_Page_021thm.jpg
cce2e18a1a3a17aee5393a2a7467b74d
84deead847983f16874a22dc7e1a33e5a5cd79f4
36200 F20110318_AACMUP iniguez_c_Page_014.jpg
5dce11860b050ad10a491da711b184f1
4236f8c313a1395aca395f307b6bbd960d618303
5391 F20110318_AACNAK iniguez_c_Page_043thm.jpg
2fa1c4889865ab399830c924efc0d10e
51c219b6a184ac6316f5f18b693f13d3b59d8f03
F20110318_AACMVE iniguez_c_Page_141.tif
e1455dd0295b18ec619719e908d585a4
fa860dde2dd013886345b899b8fc44d9bfe622ce
901923 F20110318_AACMUQ iniguez_c_Page_110.jp2
46de24f11ed039869855a80718281ab8
87ab033438ffce2fc798165974c33802de433c43
27335 F20110318_AACNAL iniguez_c_Page_091.QC.jpg
4364e752b33e475d280a255aeb6cc698
9d33217377470849f59d1966b316ea90379a2d8e
F20110318_AACMVF iniguez_c_Page_078.tif
c3b8140b25ccf31b558abca7e220bdee
0838c54ffead28e006413b44b1696e94e968ba45
103388 F20110318_AACMUR iniguez_c_Page_089.jp2
ae9a5a6dcc0060efe8c387383d05a634
635122525230be879dccbd169668daf8d7747dbf
88328 F20110318_AACNBA iniguez_c_Page_013.jp2
14af4628b326b3e7410c46a0a8771063
6ba6de527b756f836baec6ad5715edd4035885b0
64055 F20110318_AACNAM iniguez_c_Page_059.jp2
12fce446298b0d2af4db1353f39bb401
ae5e94f657ddb254fa20151dd605ce6cf289d166
77592 F20110318_AACMVG iniguez_c_Page_076.jpg
79cea55f70e3342c039fb55cf68eac8a
81981a45cf71dbba5775915af0acf7c6f11bf0da
F20110318_AACMUS iniguez_c_Page_136.tif
f2134a61bf73e9f51cac7a826192f952
86739373ea30db14d67f2d1b1d54016a8d3e5176
65480 F20110318_AACNBB iniguez_c_Page_067.jpg
9c3532d85c47d670959324865cebe895
9e9da584fbbd5b423bbf876cd1420d235e2efdf3
1051916 F20110318_AACNAN iniguez_c_Page_009.jp2
842b03a15c641063e9dc5da813fbe625
9ca533514626ab06cbfab6f03ea94c2659ac3ae8
106598 F20110318_AACMVH iniguez_c_Page_017.jp2
f59039c9e806ab5c3de1a95e39137af9
aa8ca7cabda1678a33ad51dfb63188bec362bd2a
30566 F20110318_AACMUT iniguez_c_Page_034.pro
7069f9bd941c3761f7bf9f4cda64e9d7
2083c850a1a7ebbf39bd16b40a0fb4ff6243ff15
F20110318_AACNBC iniguez_c_Page_096.tif
2312ce9f427c9444656f57512bd98b18
8cf57df6ca3db9fabd51c11c172085dd3d4b99e3
32626 F20110318_AACNAO iniguez_c_Page_098.jpg
57241c61219fcfbfb30d092540ede4af
05d01acf72242462e75a25e3997179b4395e6a17
49759 F20110318_AACMVI iniguez_c_Page_023.pro
b2c4f654a0a0df0fe36f539c4618cc13
dc6362c7aca6f53f8af8019aaa0c0810cd2c61cb
2005 F20110318_AACMUU iniguez_c_Page_031.txt
b8e59d7d4786a5aa4a53de93cbb72739
26f12706fe062544591b3eefc456b5d662c9654b
53018 F20110318_AACNBD iniguez_c_Page_036.pro
ca89f528cf80bd4237c587d196b11cf3
820f6c099eea47689315440b8d8d817d36dc4cd2
74631 F20110318_AACNAP iniguez_c_Page_061.jp2
66d9cf5f0e195e60436c743a3d2cbcc9
3108be86bd83a39e18d8a9cc1fc5319dc66707b8
5763 F20110318_AACMVJ iniguez_c_Page_026thm.jpg
91c6ac1057dd2c9e4a1df97e495c21d6
b43c1a6aac8545bbe730333b3e5a2921abf06803
59894 F20110318_AACMUV iniguez_c_Page_120.jpg
5b1c51b8be9b6e421897e5bd74654398
6b4b9b9d2e197bc28dd9e5092745add9366c0456
24832 F20110318_AACNBE iniguez_c_Page_076.QC.jpg
62c7dcf4541451173823db29453ca606
d60414057dfc59b97875bd3ccbd119584e49708f
89379 F20110318_AACNAQ iniguez_c_Page_131.jp2
102634dd2ad8a59f7ce79f7be4c5a990
4942af63e85135c56d792008e2ab71f093d744c0
F20110318_AACMVK iniguez_c_Page_009.tif
84ba9c5d4d11e324c076d7ad8278e752
e003879948de98ad469b0e5ac0226cc133a40b4e
61759 F20110318_AACMUW iniguez_c_Page_061.jpg
2031055a9cd4e0cca88975cbff4fdbe0
43535e343f25cd2a8ed742536eb732e44957f370
22717 F20110318_AACNAR iniguez_c_Page_088.QC.jpg
b022d119bbd9293438b275a4ec78e84e
f21b9acb75b37f9a4d3e69cfa616738f917d0513
64977 F20110318_AACMVL iniguez_c_Page_101.jpg
d2e58ae6cf9a564bd0e032015440f46b
744c7833a46a21943a07b8362034ef0363984c14
62628 F20110318_AACMUX iniguez_c_Page_102.jpg
422319706d22819ed967653953c0476e
76670c91dc21ff792f6428ed6e43b25a606708b1
65090 F20110318_AACNBF iniguez_c_Page_100.jpg
b3c39d4e49d83a60d1200837bf147706
bdbc9b0356e7fdf48a44eb9a3800a09dce46ad18
1051970 F20110318_AACMWA iniguez_c_Page_020.jp2
7bc27c0ef1a369d73e23bdb6f59c4833
dad305479832407c14cf70c9abcfe2716653d531
2383 F20110318_AACNAS iniguez_c_Page_148.txt
1c24418b2bade1618c068487f2e3c4ee
ea6a12525704d203ad94fc8c96ba3abd6e514d33
6163 F20110318_AACMVM iniguez_c_Page_094thm.jpg
a99776bd48cb336cecaa70989773b63c
058a0740bd12e648fb1d93d75e657d9b5ec9b3be
53362 F20110318_AACMUY iniguez_c_Page_073.pro
044e1b121b15176f2615e29c306007e9
e5bbceae234997682e107edeb3a793623ca3816d
F20110318_AACNBG iniguez_c_Page_034.tif
36d920d83802f816dfec76b4d51f84b3
59161f091bd290cb2b959b946053ebb1d8fc1af6
F20110318_AACMWB iniguez_c_Page_118.tif
f62485856e7f1f62a87f115ef7d3b247
956f6f5b0d87883837ed1b99320fb87fb1798e55
20817 F20110318_AACNAT iniguez_c_Page_128.QC.jpg
110a8bb7b4218e8ff9aed9bc4c5e59d5
865e70f6765bc0f1566a97f5d283f144b49fce69
23673 F20110318_AACMVN iniguez_c_Page_106.QC.jpg
33a1b2a01f91a0fed9d2bd7ac87cc62a
bd1541292aa43076d7eb35491dd6cb129906750e
29054 F20110318_AACNBH iniguez_c_Page_010.QC.jpg
a092d979dedab26e653e9efef4de9bdb
560195087916d8cff54d9259319a5e7a142b2113
F20110318_AACMWC iniguez_c_Page_018.tif
b857d6472d3fd5515c3d063620c62991
9e4739c30d7ea2fc87d4726e04a1ebd5c9ff53d3
20418 F20110318_AACNAU iniguez_c_Page_104.QC.jpg
2362373ce74208d56f1a3b05cc64b030
b37ce7e6b8a143cd8613efedbf91e7da7f011e3f
23647 F20110318_AACMVO iniguez_c_Page_105.pro
609ef31f3db8480271b025eb419defd6
20dd5990a4d682c0579ae9eaace0043af39702ec
40549 F20110318_AACMUZ iniguez_c_Page_114.jpg
9e076a51ecdd553c0382ad006ea71c3c
042aa52f6d7a93e1478b89eb60e5a44b7fed55e8
1677 F20110318_AACNBI iniguez_c_Page_051.txt
3bfb02ee0f3ecb7f1a20b3de571ad58a
4b89d2810d4baf2153fffea19a0c35e827e2c00a
F20110318_AACMWD iniguez_c_Page_020.tif
49c53fb45e77127713af42a8f1ada847
2f3ae8c55b1f443e9f6a849d40303a7c4c908837
85665 F20110318_AACNAV iniguez_c_Page_096.jpg
4d25e64c71dde7bccb10c0188517066e
7ca63b10d77e7d11b853fb50312bf5c93ed67cef
6528 F20110318_AACMVP iniguez_c_Page_080thm.jpg
74f0d14939d9eeeeae1fcbd2208f7e97
83dad1b37a1b5ddde4ced0f6166d2df1ac9495ac
74229 F20110318_AACNBJ iniguez_c_Page_045.jpg
37affcc3fae5cc0dd729b998af3ab72b
e994b5c9ad12b61b41d6da5b387725fd747cbae6
75527 F20110318_AACMWE iniguez_c_Page_080.jpg
92f5a4c53c82875f19dabea354dc2ead
9a854b6f7825e0c53bb88f419dfc8a7758fed420
44198 F20110318_AACNAW iniguez_c_Page_146.pro
0e12a68a77e12dfe2f1b3eb416f6522f
210b41a4cd7bdfdcc3d04f77f7338d5c35ef4d72
49856 F20110318_AACMVQ iniguez_c_Page_018.pro
2c6284639acbab69b328ba905bc5ff93
a084864736d2f6bdae88f3accc8f38815e5909c5
16571 F20110318_AACNBK iniguez_c_Page_123.QC.jpg
a42479000f2aec2bd788fa19bc7f1378
ef98221e27db9ff58802d0137c40b0f445c054e7
2022 F20110318_AACMWF iniguez_c_Page_074.txt
4b9e44dbf78676f2b8c13b2ce021bb7d
c0b43130f0eaa96ba052e9e6bb3fb00412f5b999
F20110318_AACNAX iniguez_c_Page_059.tif
33ee33a18bad5e952c6fe61a16406300
c7379f68785dc7f500501c515476ddabd8de0260
1858 F20110318_AACMVR iniguez_c_Page_088.txt
8a8dda0ea64cec5534146c5479f6cf96
508cb6cf4e20944a2ec71861708759df234c7769
46788 F20110318_AACNCA iniguez_c_Page_111.pro
753cf3d8cbbe29a4a3b67c9db1b04998
01fd83b61ff84a664ea23acbb863cb8153108d9e
F20110318_AACNBL iniguez_c_Page_076.tif
3a4701c806e4e0158902d24c9bb4b492
90758e1684b539e433c97f2be37fffd3f3927a76
49782 F20110318_AACMWG iniguez_c_Page_092.pro
5c1956f6ddcc057a3a70f59acf3f4aa2
6e7ab6327ff6ffefbdb6a9ed8fa5a1b16826bf96
F20110318_AACNAY iniguez_c_Page_073.tif
84cb5a5a23ba6e0e0464d7cda0214396
ce6629abbbdf7206a967d8ce3516cc2c0233c63a
F20110318_AACMVS iniguez_c_Page_146.tif
879903bb7212b35776eb6dcbda2d1202
1fd4a740ef2953b5732e85ba6c42858e6556acd0
3674 F20110318_AACNCB iniguez_c_Page_114thm.jpg
030c7b8ae3ffa82c7d9445e7252c359b
fcdd4517f181cac69e33121c86c6b9b366fd314d
1997 F20110318_AACNBM iniguez_c_Page_016.txt
b0fc6cdeafe6dbc7872303064a57290a
6abbf090a1d61b6feb4cef70e45c038b2debd022
1575 F20110318_AACMWH iniguez_c_Page_067.txt
f3cd74116729bd6d4f6637f27df0a424
d1a3138d1dff41c015670f5646b631e53ea57437
76805 F20110318_AACNAZ iniguez_c_Page_146.jpg
b127d05bb1be23b08a012ca06416cd60
00bbbd0cfb61aefa9af93d94c846cf9e64a6080d
74049 F20110318_AACMVT iniguez_c_Page_070.jpg
4cff9b4e215a87d0362199142fba9483
54899bde09803c66cf3fdf5baaae9e0fb094488f
30793 F20110318_AACNCC iniguez_c_Page_101.pro
3dc81b8cd1b0864c7837af15615b2484
f5588d972a557d39699aea1bb705d369fdf6b336
82828 F20110318_AACNBN iniguez_c_Page_047.jpg
d405a701b2c58655823b02c37ed89af7
271c0996815610a0172a7fa0987c52d0f98509ad
3534 F20110318_AACMWI iniguez_c_Page_069thm.jpg
2f3fb07446270497fe5cb85941a5ca7e
12b04ef6eb7d8f522c9dd789423f4d9a47133de6
9235 F20110318_AACMVU iniguez_c_Page_097.QC.jpg
85643ea955dbcaef1e603bfe0d45f5e0
8b97e696c653460891d51796365679db08090da5
F20110318_AACNCD iniguez_c_Page_050.tif
48d7bfb13d626e7822263b7542e71732
cb8bd5fb4717fc5c0d156358e9c24e5ada614704
80 F20110318_AACNBO iniguez_c_Page_003.txt
f80d81d834e5e866d6711587b6d0bd30
b14b2a2b7899e9f58a84a76d9e90c2b1cb52c171
5809 F20110318_AACMWJ iniguez_c_Page_008thm.jpg
6c89ee31b8bcc4375e39caa0296acfb2
87d33f9b8af00faca0d44d4e76d442490071185b
12245 F20110318_AACMVV iniguez_c_Page_114.QC.jpg
257017c7347749ea3d6895530011d135
e11cbed0a887c7d7701050744a8eb0028799ab6f
F20110318_AACNCE iniguez_c_Page_084.tif
5f757b8add8b6fa6c0973a19a5ced1bc
276ed0df47852dcda21c5bfc68e143e3d275f7d5
32592 F20110318_AACNBP iniguez_c_Page_060.pro
f350f763680e098f2bff175f95522778
286c27b7b6039f512ac74cbb7dfabd1ac5005676
32641 F20110318_AACMWK iniguez_c_Page_067.pro
91248e1c42121c8a586a30138d9974a8
a76c6632bd24fae3b54c770b5289965ca64fb6a5
37226 F20110318_AACMVW iniguez_c_Page_142.pro
6218705480be47dd4890ff272034d779
8a935160ad0963e495bfea6668a4e12312864704
242970 F20110318_AACNCF UFE0011838_00001.xml FULL
bb6fa6f42e0d77adee530b4e067f8c6d
58c670438a570cb30c82ee8c337b698840711df5
51772 F20110318_AACNBQ iniguez_c_Page_030.pro
348521c0c78dd973f0448efb2e767c99
38172ca240ed09610fa87a9fe8da8f090fb2de62
19991 F20110318_AACMWL iniguez_c_Page_086.QC.jpg
76103c51e6b77faad0abc259f0afc964
2905c27373d035a1db3bc05ee5264e5ceba66658
23160 F20110318_AACMVX iniguez_c_Page_070.QC.jpg
834eaa475bb4959e49bd427acb9ec859
c152556acd925d06c6d56e020a62651ecf19c951
2201 F20110318_AACNBR iniguez_c_Page_044.txt
06703b4609a9d84f424159d4bf95c77e
86201ebdacd07ce57be03eba7e49d03fdb4fc4bc
26637 F20110318_AACMWM iniguez_c_Page_096.QC.jpg
48344b9df6d5da3e9d5321c03f3591d6
2b17126649647472d9edf9d007c67aa79df90896
76961 F20110318_AACMVY iniguez_c_Page_004.jp2
906de807e14175b4d96f59c42111b1ac
175a34adac31499521b336f1740e01f0974e01f2
23888 F20110318_AACMXA iniguez_c_Page_068.QC.jpg
145ea865ee9984ded4198a62edd0341c
f543870bd242f05e54738433e9edf42736248784
F20110318_AACNBS iniguez_c_Page_126.tif
a4fc95058bafb0dc9d159849b67f2235
2935dcc94a6daf8abf94ee7441f099455d16ec31
F20110318_AACMWN iniguez_c_Page_145.tif
13c19baeb7b012a43f9fb04589bb6b2e
ae19bcd32184fe57cb1eef3fe1cc978f65c66085
6861 F20110318_AACMVZ iniguez_c_Page_106thm.jpg
aa14dc11b00beb8185c19607bd18d781
d8f3cf8f907bdeaa22e2b9c27b78e370bbaaba54
19643 F20110318_AACMXB iniguez_c_Page_117.QC.jpg
55c154339cbdda8a3255154edad9c1d8
2a06ce84fe3892973d477090363767e52a8d8596
78025 F20110318_AACNBT iniguez_c_Page_026.jpg
bd293a43b1beaa8726b6a43253b569c6
878ac8b747de4bd59a30735c0221ba249d8ac7cc
21530 F20110318_AACMWO iniguez_c_Page_135.QC.jpg
4d0be24c84f5a013afdf92858c8bf810
9af0625e24f9f3ee8d5c28e7b95c4d807d08395c
F20110318_AACNCI iniguez_c_Page_005.tif
a57d1378c69fc60fec2c564be573c0c6
e5aedf3d1179f004531deb6d3c761712c30c1be3
6102 F20110318_AACMXC iniguez_c_Page_023thm.jpg
c4a06ac5bc86a274fb0560ccf41fe068
50cdd957195c7343b276408147b61255adf771c3
859079 F20110318_AACNBU iniguez_c_Page_049.jp2
9e76a3d1663b31f88e02223a3fa5ace9
1ecb5b0436f9f1312bafc9e26ba7ffd5157635c9
F20110318_AACNCJ iniguez_c_Page_007.tif
fb5cbf582c0f02de3667a5fbb1b78486
8c1f1111cd443469ecb4f6d91c2cd355c692c083
83463 F20110318_AACMXD iniguez_c_Page_085.jpg
2b3e7976abe784cfdc6ca56b9cc95625
70ada634d3fc7f0caf2f014fa047bb4324300e6d
46570 F20110318_AACNBV iniguez_c_Page_068.pro
63149a670bf70432dcd3cbab1981cc5b
8c948fa35d63cc4a858f1c614d9a298b2161f4b7
52119 F20110318_AACMWP iniguez_c_Page_062.jpg
e06a32b82b8e7414034adf11e3b3d45d
7ec9bd8ac8501c071c81914aa007379c000b3e9b
F20110318_AACNCK iniguez_c_Page_008.tif
b852f07ca8c3689633100791bace885f
0c60bfda991a67702c19556612f41a2209bc26f0
51479 F20110318_AACMXE iniguez_c_Page_093.pro
72bfc7744ef8281774c2283d6fd81532
657cddbdcb3fa989aafeaf96cfbc052f918f54fb
25909 F20110318_AACNBW iniguez_c_Page_093.QC.jpg
4c31bf5c0ed57ed744657f243bf7bb2c
4285bf8e249fdcd602b6f9d9e0321a688a7c8bb4
61692 F20110318_AACMWQ iniguez_c_Page_121.jpg
19aa00db9b0392b1ff5782a41ecc45f5
bdb9beb9870a00dab05b571666ee0ba630c3bb8f
F20110318_AACNDA iniguez_c_Page_086.tif
4f474b80c4d495806781c9b7752c5572
590a4ac5f738c295e710f8be6a77bf0d959bd126
F20110318_AACNCL iniguez_c_Page_012.tif
f339bb052175e1a958374e20384e9879
eb1931696a7891cd961323f50cedbba833e9f704
66001 F20110318_AACMXF iniguez_c_Page_009.pro
7beb79bb85aa6950188a89d2796689d0
d01c30edc51e43ef86749704b579bb13b655abe5
2015 F20110318_AACNBX iniguez_c_Page_035.txt
b79b343afa6976d88ebd88ac106e2a44
17d498ef369dbdfe89d1691ae11e80cae487b511
6811 F20110318_AACMWR iniguez_c_Page_060thm.jpg
b24d131a1435a9bfb406a4d928f12aa5
3b27ec7410c335d74497e64164b696307566e86c
F20110318_AACNDB iniguez_c_Page_093.tif
61e71f207a94cea5b03489ecce44617e
37b9fe41f263636e118e70d137c92362b4069002
F20110318_AACNCM iniguez_c_Page_015.tif
d48e16a08728d090c30440f2b82467c4
6bbd765dc3b80e6e468b4ba81666c8d1e26080dd
84160 F20110318_AACMXG iniguez_c_Page_028.jpg
c59630403a37e07e463e4db884c38493
f46295f7e4f55baac8c46c6cefa48988a3d0d917
F20110318_AACNBY iniguez_c_Page_036.tif
6f12f443c53cecfd03cf7e9bed113eae
02ab72befa549e1f31b5e6d80a1413aa30b7d3de
66213 F20110318_AACMWS iniguez_c_Page_056.jp2
520eb4c7978dfcd08989ac382298e032
8778c828075eb42aebdceb1a54ff55ea09e865d6
F20110318_AACNDC iniguez_c_Page_095.tif
ecb10cfb9863bae7b4a4c627772c4789
9700f580f7add47b597856dd06b913b503a149d9
24898 F20110318_AACMAA iniguez_c_Page_122.pro
20c90ec50bca56168d190e1fa56b34af
032da8aab085e385b136a49688b033d21703d44a
F20110318_AACNCN iniguez_c_Page_017.tif
59e5153548269c27cfa8cbabf36fd4f7
8a4dfccad63ae602593f5a4fd01c26ecd3f5b7b4
F20110318_AACMXH iniguez_c_Page_054.tif
cc548238a7817590873970b24c8e6e06
0e9264de74a5ab419aeb7711e3980eafc6568946
12575 F20110318_AACNBZ iniguez_c_Page_012.QC.jpg
e1890c6051e68d5e5d84cb6338d8c4e3
a3f7c74ae9052875e4c6051fa9db8b85a9364d31
89124 F20110318_AACMWT iniguez_c_Page_021.jp2
5479c500c5d13bfb848b2d5070898149
65e3663d95a28db080d5f6d113335b62cd8aa640
F20110318_AACNDD iniguez_c_Page_098.tif
49055b5ce54bf32789bbfda613acf255
1b182d7b93acc895577f95790ebb3a13b8d6c84a
F20110318_AACMAB iniguez_c_Page_144.tif
66ed1b2d6e4628a7e032fb11325f1c6b
bb42d7678e5749dfe205934c454133353fa96fe4
F20110318_AACNCO iniguez_c_Page_030.tif
907edbf5017f8ff26d32d4c75c148509
d6bc1597e48aa470ac69a7a8986ce4df80aec398
65746 F20110318_AACMXI iniguez_c_Page_141.jpg
056413855b09e8f773260064b003f8e4
e86ff4d68fcfeabf204be58c00348270aaf9a7cb
F20110318_AACMWU iniguez_c_Page_080.tif
31564b2c65f5c449e083a97b07d410aa
83ac434a69fcae46ccb5997ad6eaece9a232b6b1
F20110318_AACNDE iniguez_c_Page_108.tif
d4cc0ccbd50e1d4ff453b7962b953fe4
e5e4b24e91c5c69ff6f0bab30fd9c4b0ac35173a
F20110318_AACNCP iniguez_c_Page_037.tif
cb8c2ee98f4e3c68022062a3269a7529
760007a98a93c3280e564fbe8e11fb08e862e034
3322 F20110318_AACMXJ iniguez_c_Page_063thm.jpg
b0098973c3776fb0ec8375ebe348e34f
c9bcc1e17c7fa8e9e7e9ba87ae3e837a3593e232
108582 F20110318_AACMWV iniguez_c_Page_025.jp2
6a2ff585f7f506f7b1317bcf2f32b4d2
754ea5cc167b6f95ae882233a7e9924d4242852b
109802 F20110318_AACMAC iniguez_c_Page_085.jp2
a2dc1394297ed379fc26c0c00a6cbb94
58db90c06402cb502452f4709f682c2e9d22472d
F20110318_AACNDF iniguez_c_Page_109.tif
bff6446f6351ff10c2b36b96fd22ba20
8a756e555ec1c8899aaba23bf29c2f0f08fa734d
F20110318_AACNCQ iniguez_c_Page_039.tif
181c5a8706ed46a9b5dd745ec89b4f33
b4f20dc64495b6d878cbab8bd5dbb04258d1ec10
26375 F20110318_AACMXK iniguez_c_Page_031.QC.jpg
cad2b6fb30388f8d504bb4d56fea0651
e01affbee5fabe90662dc9ce799cd6e63664d4c2
28718 F20110318_AACMWW iniguez_c_Page_055.pro
0e9bf449fa9e53e31ba8c8f474e6a61a
01f6d9472b36181350c2931bf07b443145fec2e6
60808 F20110318_AACMAD iniguez_c_Page_099.jpg
3def2af8464377420b8650962f911c54
7959c7f2d4bc121882e119377988acc957d614de
F20110318_AACNDG iniguez_c_Page_110.tif
250929f86c8b650cb526c7fa1aad991e
259bed05a22ff0ca142408f41751c51358624f67
F20110318_AACNCR iniguez_c_Page_044.tif
9a57d06adab630965339df156eb95364
5f3fe9a145f852fb37a6b953dbe0707218bce7e4
6195 F20110318_AACMXL iniguez_c_Page_047thm.jpg
20f38f236201a3b7d7f68c40ed2ee9df
422b6c74156ba7235dd6000c83def6e824f616db
111652 F20110318_AACMWX iniguez_c_Page_030.jp2
0a7fd115a3ec40226936ecd0c376e238
78afdbd3062f4223bb3b0fe77f9acb638f17bd19
11981 F20110318_AACMAE iniguez_c_Page_150.QC.jpg
c59e6eb576c4af991c9dd408e1dcbc6b
42fdb19034b6790d2a308858bf41272e084ef2f9
1051986 F20110318_AACMYA iniguez_c_Page_010.jp2
b44c1ed67449850845b21b933086353e
3fec6a97ed9d65215de47beb4f19034039d5c8b8
F20110318_AACNCS iniguez_c_Page_048.tif
740dedb61b2ca92f7a55be4f0600d3bd
0d3bc7826e2eeb7158792523c865dd0a735ca142
2176 F20110318_AACMXM iniguez_c_Page_020.txt
f2f7edc16f0762c78905a4109f47907b
538c67330d9c2bf2b696fc9cb3c8357877a65303
2086 F20110318_AACMWY iniguez_c_Page_103.txt
83b96e0bc5e9003cae58f06d855e00e6
6f8366707b84ad415b8573de1587806a6ba60aa6
F20110318_AACNDH iniguez_c_Page_117.tif
aabd695c15f4b86f544345d04c4729eb
f732be0b10a7c50cff631184ae7ca46b47d2f053
20529 F20110318_AACMYB iniguez_c_Page_048.pro
a15c24043da577265b7fd956d2eeb3c6
bf8cfde00d03cd045bdd73b7ade3fe8d1d59f157
F20110318_AACNCT iniguez_c_Page_051.tif
e36b877cf62e14b21759f38a3cb4ebb3
c856361e317bad4683aa1c9da7484797977fab4b
105855 F20110318_AACMXN iniguez_c_Page_090.jp2
0a70d82618fa090d7e0698c4310c7ffa
05b4b41f11386f98fe9aefde662b7a51cd568255
59334 F20110318_AACMWZ iniguez_c_Page_123.jpg
f234fb30d95eb32e651c96c55ef87a99
4b817f4b8246918672125f68f84de7c1ae271b3c
112740 F20110318_AACMAF iniguez_c_Page_040.jp2
7f8061500766d6d0d849f610da43fb67
3b06c5d0b0655762c2e6e09139a31a26ccbae7b6
F20110318_AACNDI iniguez_c_Page_120.tif
c900e37a425e7850af1d1f60f4969f3f
ffd1acc87b4171606dc9ec1c6594a02834533c63
23745 F20110318_AACMYC iniguez_c_Page_057.pro
d4e6bd209c7fb6c14e2ca4029401a4a9
6d66fe5ce0d93cbd62d055d2bb4c75bee7748e99
F20110318_AACNCU iniguez_c_Page_066.tif
7d8a4cb4cc8dfd6ed2121c8a99cdf79a
8f74cf9e3d76d6cb84a6f157f755fe3411cf184c
919797 F20110318_AACMXO iniguez_c_Page_103.jp2
75755a1e108afbc888d44367ad2029eb
70c6f21c6469e62eab3355e248f0c8b7fdc0bbb7
1754 F20110318_AACMAG iniguez_c_Page_001thm.jpg
369d4c7eacebc9890c4b735d82d4b84e
0ffb8fcb6eebf3fec53d3473c76f8ea3d4db3133
F20110318_AACNDJ iniguez_c_Page_123.tif
939cebea5b74935468a618ee0913a474
36f99d46a48fef0a7cff2c5b9621c7be979d5cbb
F20110318_AACNCV iniguez_c_Page_067.tif
f9cd2a102a19191f8d102d9b462afecf
bc85127d4382365e53e9c08d2d2eaeaadeb8fc36
50963 F20110318_AACMYD iniguez_c_Page_031.pro
c6c2173af3fc58bafe9a6d595a46128f
5e3f7f91fd84365253321fde9ace516124aa80fc
79719 F20110318_AACMXP iniguez_c_Page_032.jp2
3faa844031205429f0a5958e5ad7dc90
be6aadd9d5fe89316642254c0981d94dd6dfb35f
F20110318_AACMAH iniguez_c_Page_077thm.jpg
0b06ef2bec4a96cea30e3b460dd2d67b
43773912840566becc919388584133c1a6f802fc
F20110318_AACNDK iniguez_c_Page_128.tif
68d68bc3d0f3367748b91839d2879afd
451b144bd269d32e1b5ef2b9089a816978290711
F20110318_AACNCW iniguez_c_Page_069.tif
2573e6f8de786ba91c7e2d1ee1b4035e
f1a4fbedbb4dd2cc78cd3fceeb18d2412cd1154e
107858 F20110318_AACMYE iniguez_c_Page_065.jp2
0f1ef9dabf61d653598c2a2270de5e7e
ddfdcdceed81ba1f75f423c56028306e058670d5
50164 F20110318_AACMXQ iniguez_c_Page_025.pro
b6ef0b97cf60578ff22785ba9b0ebd61
85feacdf4cb865d8aff219d48c34c804997cce39
F20110318_AACMAI iniguez_c_Page_107.tif
be2e1b50697bddfac40714940cdfc5af
720be7ca3fc3f1199d999e46fecc383387c50c4d
2224 F20110318_AACNEA iniguez_c_Page_046.txt
199315e58f82e84c630419e8a7043b85
dddccade6135ea7b03921a424dae8e897d88ea07
F20110318_AACNDL iniguez_c_Page_130.tif
725903a4cca1ba3ab88dbd6728cd12f1
eaf49cc10b77295e38a018d2e2ab7cee6d305d95
F20110318_AACNCX iniguez_c_Page_071.tif
0d3088d251acd6ab5ddca96fd50d3d5a
a10d7e37fbe4f6665459e7bbf9b3be5879f0e3bd
37280 F20110318_AACMYF iniguez_c_Page_041.pro
45c2e190a2d87376801bce8ddb69c263
1a3de1aece9b4badced95e28d061b40b2a24cde1
F20110318_AACMXR iniguez_c_Page_035.tif
62c96b18dbe0fe3a02f4fa60eccb2df5
325317ebecd8f35ac8b6f5f10b8d4083e8e30d18
101460 F20110318_AACMAJ iniguez_c_Page_026.jp2
109e5f97cdb186d3dec95cd407e9bdee
6bad9037f0e602d64b1adb3ffa89b111ccb0b9b5
1985 F20110318_AACNEB iniguez_c_Page_047.txt
a6d49d7bd6b9a42a4a560f6b1c098754
46767b91bce49212557c35cddd9a286f538bc514
F20110318_AACNDM iniguez_c_Page_131.tif
6cf931287c85afb97998b36380500d11
b8a08cafb57f723095e7fc81d70c76b9fe7ee13c
F20110318_AACNCY iniguez_c_Page_074.tif
9f0995b0f7f261476a8543ec4a757b89
70ca08561de08b1d0f74a8b10fb969272becd083
81366 F20110318_AACMYG iniguez_c_Page_139.jp2
3c70ecf19b125d434b15223a7d5f78ad
ab3e1b5f900fddb08ff31a6427d24820462a0124
18013 F20110318_AACMXS iniguez_c_Page_136.QC.jpg
61e2721caf68093b53b840391f861469
4489f1f2774483a2bf0e80c425520852e36f069d
6597 F20110318_AACMAK iniguez_c_Page_020thm.jpg
b7adc5f40f9a2b358af92a3838810e6d
59cda6a1bd8f08a0ce313f0886af38527a728edc
1902 F20110318_AACNEC iniguez_c_Page_053.txt
cfb1121d2afbf54a5e134a254daf097a
35ca348ac4c0ff79b7821eb3d9c872925a8f02a4
F20110318_AACNDN iniguez_c_Page_138.tif
7426cd203fb8587ed71903fcd30e55cb
0b7b99315785b8131ace7be5c43441dafa78a7ea
F20110318_AACNCZ iniguez_c_Page_082.tif
9c5509f7e323ba1d3e8c97d0688f7ca3
9bd96767a7b6043d1b18815034bea9d059921bec
19188 F20110318_AACMYH iniguez_c_Page_032.QC.jpg
bc5dc8d63f6a3b5e869d18b5c81a223e
8ef3c9488b56f3c68dbd6c00a0c0572d9b457953
35910 F20110318_AACMXT iniguez_c_Page_043.pro
0854047afdbcb937d22a4877bd42a98b
65b96880607120f88b5c42dcfab1c06d717c15b6
98440 F20110318_AACMBA iniguez_c_Page_033.jp2
3c28261baef8c643b5901fa1d979bf9b
9c948246c84be4083f595e7e87720077c64939e3
21095 F20110318_AACMAL iniguez_c_Page_042.QC.jpg
4d80af9e6972a26810a384ce25a50e1f
a320aa8765e479fc52e18c5f2221814646aded9f
1171 F20110318_AACNED iniguez_c_Page_057.txt
57ac46d1b230b1e5f9ed12519178dac9
d68559f6685b53921239ef3c859b61d48b63086d
F20110318_AACNDO iniguez_c_Page_142.tif
3ea5a479c75eeeef5eed183247c63436
552ecee5a399fc856884c5795612395e54b94c35
81621 F20110318_AACMYI iniguez_c_Page_065.jpg
82242c993f7267995c7712a284747018
17dec3571a56ff0b436686f6d44df3609da8b13a
2039 F20110318_AACMXU iniguez_c_Page_018.txt
eb3e225dd80605a747b97ed5488fdc3a
a58924c997f30afe6b7327f4e9a0ab85aefa8e93
2997 F20110318_AACMBB iniguez_c_Page_003.jpg
27f52210c63acfcacba8ac440fc41842
843d8f83a21aa0f28f83c758782cb00dd03b4218
47422 F20110318_AACMAM iniguez_c_Page_089.pro
adb52ff2c01b0c6431ce66dfb063cb69
236b9f63186430af7af3313f66b1bc47be649c54
1855 F20110318_AACNEE iniguez_c_Page_058.txt
26e40753703f8af12445f3bf2441cf34
200b727af0676d8db106a33038b95aa283b8d312
F20110318_AACNDP iniguez_c_Page_147.tif
c96ab9efd611e0bfe6dbfad425efb0fe
a3a423923b53660237931edfac99dd2a262bfc0a
4410 F20110318_AACMYJ iniguez_c_Page_002.jpg
6d4ae2a90e2e379ac8a295892f3bc31d
1e1deb9cfd3b26ab105c43a11dff4e6347f13d73
56870 F20110318_AACMXV iniguez_c_Page_039.jpg
b27ef6e8ee236474d96fc685baf5e6a8
b025100d7c71f6045c186181ae9c89233c06dd69
F20110318_AACMBC iniguez_c_Page_105.tif
7f901ac0e0a4658b717d3ab9ebbc117e
f2b82618bdfe448ff59dc8e5b3234d361d88f87c
84302 F20110318_AACMAN iniguez_c_Page_029.jpg
72ae036fcae8b97510c728c14904f8c2
8ce7a9d5c2a35b8e8e4f53adf359a6423766e262
1561 F20110318_AACNEF iniguez_c_Page_059.txt
4a7439965f49772f4468da0f0aae7258
370598406b41d1e3ff27c874b3dda5902a5d9900
F20110318_AACNDQ iniguez_c_Page_150.tif
23bfc1d9758c874c0918d84a68cd526e
934ff12c0ce838bc153e62f8635ec40467046542
84272 F20110318_AACMYK iniguez_c_Page_051.jp2
1ab1fc07acbf1550e64a635323ff7ba3
b043d273e44f6dafd1920ac25fb6b5a5734cbbf9
897772 F20110318_AACMXW iniguez_c_Page_112.jp2
2c0b9869a57c696434c83125d08e1152
a7196d41504cf843aef0bd938f59836498b022c2
F20110318_AACMBD iniguez_c_Page_085.txt
48c6b3e092f1d2d03deb98a6342c96e1
a0761547300ac4f3de69253dc0016ded0fadcc12
1971 F20110318_AACMAO iniguez_c_Page_078.txt
e4fb80b9f9b89470db7b6c4026cc5c5b
e53956d7a7e95ef93548ce849632752f69d974ab
1848 F20110318_AACNEG iniguez_c_Page_068.txt
6f45a4d9f131146d35e5b39f7f835371
7e9c5848cf91b647cb37cb76c26a24aaac4840b0
1428 F20110318_AACNDR iniguez_c_Page_004.txt
056497957ece8d5c5bc7944e4e1675a3
bf3cb932359ae8db3a93bc9ae4796cdf5bd0ff93
20241 F20110318_AACMYL iniguez_c_Page_050.QC.jpg
d2c64749e529a301befd30f3c63624b3
6a678d32f7b13f7b5041927cfe28bf66f7957cd8
58596 F20110318_AACMXX iniguez_c_Page_117.jpg
98edddfd8ab1fda04d17e2edccb7ca1c
0a7bac3aafba2c9d9f7eff94880487add8f76d8f
1094 F20110318_AACMBE iniguez_c_Page_069.txt
373f36b75286ac7de57e18f304dd54c7
fb16369c9733d5746b384909dbbf2b04faf33169
F20110318_AACMAP iniguez_c_Page_097.tif
2610c6ed03f2ede06c978878bc29e5d3
d0db0dc9a04a611613bdb6b0842e13f87614fa32
F20110318_AACNEH iniguez_c_Page_070.txt
22cc9a265ed63065acd92c3d298ac98d
d3f7cc6f3d120766b3b1ac32e2e766a2f4ccc027
782 F20110318_AACNDS iniguez_c_Page_014.txt
57d1082292922e70bc152253613ecedf
9d904c57d0f97486ca45131d88faff968b8fe08f
3023 F20110318_AACMYM iniguez_c_Page_006.txt
ea5fd9de0f154ee81fb54b126e34bd6a
f271da58c7ef9c1b7c3f06b926508e2d9b1c021f
78623 F20110318_AACMXY iniguez_c_Page_089.jpg
41249315ef95e3b7f006cd75d7b91b61
21bf95fafd10157c48861dd8d12f2611adaa260c
20029 F20110318_AACMBF iniguez_c_Page_121.QC.jpg
4632499f87c1db197e288df362112c25
da4e5c7f667e74cb4ce4de8f67367b0f97ec1c64
467 F20110318_AACMAQ iniguez_c_Page_001.txt
33c4b0127f5b52652ecd4ddebcece13b
b46bb8a9e86bec377211a0b625cedb7c56ff32f3
27611 F20110318_AACMZA iniguez_c_Page_145.pro
cb787e491f310809797f8dabb94eb087
64f15cd7a89cbcaf248af0351968eb57b56508d9
2014 F20110318_AACNDT iniguez_c_Page_019.txt
d739d18f7b406b1ff5f8e330dafde513
79396e31b1e37decf16ea21ee6e9bb1d55b626aa
23348 F20110318_AACMYN iniguez_c_Page_045.QC.jpg
a3e8432768acce3e30158a87ecc743fb
9b7e748dca0523fe5357b13f13ad27e8f2f4d423
F20110318_AACMXZ iniguez_c_Page_057.tif
de2ef475b9adbf7fe687753292d5b90d
5a213e5f9cbfc863421ece4097b60b54c95dc8d9
946991 F20110318_AACMAR iniguez_c_Page_100.jp2
f5cdf950fe61e013f39c23f1c0011024
6301d9bfee101f184e32a3f3382580818153d3e5
4054 F20110318_AACMZB iniguez_c_Page_123thm.jpg
68be8f8aec63e029a357fc4fdc35d7d4
1f3c380ef73a44901abc03c4e5089ba2dd36b0ce
1440 F20110318_AACNEI iniguez_c_Page_071.txt
471845c512ae2bc7bf613f23af0c526b
f8d34806ae4d7e1fd2e6e32f1e9e8f8bdfdbbcff
1960 F20110318_AACNDU iniguez_c_Page_023.txt
6f0f7c49590e870850eadf5a33a74db0
b68e54435a20e92cba2816ea2768b201335238f2
20721 F20110318_AACMYO iniguez_c_Page_077.QC.jpg
ec1cf79d49bc1b4c49fc2b74f2b9687c
9e81c43a6423a58c07ce1a799ec593782c9e9e62
9952 F20110318_AACMBG iniguez_c_Page_007.pro
2db93ee0c876e56f8fee1b1e4da8bd27
c74581ee3a879784e326068bb6fe177a99233960
82121 F20110318_AACMAS iniguez_c_Page_024.jpg
52983300003d65b6ecc03e564e594a16
b63fd517daac3cbc53306145d9f7e2fab458a5a7
20909 F20110318_AACMZC iniguez_c_Page_021.QC.jpg
e473b344f71abd7ec1b76382b165eac0
2d073cd543f9d252bed289dab08c70a38d244326
1915 F20110318_AACNEJ iniguez_c_Page_072.txt
6873b8d904f53aeb1ce41ad014b8958b
bdc549c402573022e3182e1492c08c11b327a7aa
1946 F20110318_AACNDV iniguez_c_Page_024.txt
37669942f8429178c1733e2d2597ae86
7cb9674b371bc63827beeccbf254a78d67585899
5512 F20110318_AACMYP iniguez_c_Page_034thm.jpg
7f3c9b4b5d42e4ef2dc94961c7bc5f2c
f6ffcf9aaf40c3f46d0d86b80c4f8584438fd9f4
46191 F20110318_AACMBH iniguez_c_Page_055.jpg
faeaea6f8d105a5f0d18dd62f19eb8bf
427275aee54bb5e027e700194855761cafdf81a2
85942 F20110318_AACMAT iniguez_c_Page_008.jpg
1a57cd17036426e8d8d763f73a214d36
81a005412c08321db95f12dffe97fbe1bf744710
33191 F20110318_AACMZD iniguez_c_Page_122.jpg
6e95b03976c732c2549a61ba8d53de91
c5b94651ea6f53f43e157c73a3fa8b4136029db8
1423 F20110318_AACNEK iniguez_c_Page_077.txt
48583b5f439d6a5dd96ab298e3fa3084
9409e4bcab1b345ff46643f1fdf091def3be8467
1416 F20110318_AACNDW iniguez_c_Page_032.txt
7d7b38e59db4c298acd74b08be46245b
a528dd49141803a80c9520ffa64f7a9b19c960a6
1306 F20110318_AACMYQ iniguez_c_Page_002.pro
55ae2ac6fea247853151bae210c55ede
07d925ae3c408fbacbfb1a8922eba458e6879606
5676 F20110318_AACMBI iniguez_c_Page_041thm.jpg
1d9db100e2e0f1310593827024dd9b69
0ed70f81484bc9cab3404ffa13387c10bc1481f7
F20110318_AACMAU iniguez_c_Page_019.tif
526211b1e85c45ffb851d764d4df68d8
db66a074d0c4ea9cc12de86cb41aa614ceddfa37
83684 F20110318_AACMZE iniguez_c_Page_142.jpg
87af963d88420b9079ab3ade00c3de22
0942edc46cb89c6d1cc523a32eed36aea05791d1
1326 F20110318_AACNFA iniguez_c_Page_125.txt
a1633d31657b47e893ae4900383d2fce
de261f5a74660b7f833e55189621fcc65348ab46
1593 F20110318_AACNEL iniguez_c_Page_080.txt
3a34d8d9d52c8a3bb12605da87636732
8281c8bf9858740ccc32d072515e930404473479
1107 F20110318_AACNDX iniguez_c_Page_039.txt
1b19357e9eb704c7e1fbf1a427f0d61f
01503b2cca3680a7159cbd1f9ac9b8f132ab9bf2
19951 F20110318_AACMYR iniguez_c_Page_115.QC.jpg
b6d266c30504ff0b0f94d33aef731bb8
4f1d31e148bd3f0c3d296310f820bb3ea90855f1
72014 F20110318_AACMBJ iniguez_c_Page_141.jp2
9ad20fb59579c60d1717aff167bf7a0d
ebce103d05511723380afdf76b4ed27d7ac90049
F20110318_AACMAV iniguez_c_Page_134.tif
f4260696abc14fac2fbd626735ea4709
86017aaebf978f773324b390c276b357c93eb174
1190 F20110318_AACMZF iniguez_c_Page_133.txt
7a8915855b5f7e08361499640881d001
3c3c2a8c5c0e81b1986543b5992b33aa7dc5bfce
1189 F20110318_AACNFB iniguez_c_Page_127.txt
80df7078851d0a914c8e3182a69f6b18
8f73c898c5ddb6c7a37f929aa523c178bcdeb5ad
F20110318_AACNEM iniguez_c_Page_084.txt
a1e316ba42025c8d79432652ee83a28e
27fb1a5536992f72afa4f79a0fa75a6b21ae4e6f
1949 F20110318_AACNDY iniguez_c_Page_043.txt
d7b82ebc0f98d97b9150cb3a6873ace6
9d697ccae493ff2ac80837e51ce18378959c656e
23166 F20110318_AACMYS iniguez_c_Page_008.QC.jpg
4869860f0bd3e8273421cdd7bfe294b4
a57b884be798d2f63d8703365cc3026aec18fd0f
4436 F20110318_AACMBK iniguez_c_Page_055thm.jpg
623db805cddd1435e7a2596c37e350e4
d4fb7d628a6b16c481f3c840804e4d171ab92f88
F20110318_AACMAW iniguez_c_Page_064.txt
d22366eda3858c862733b953ebc0de53
704479a3ba960b202c9f6a581bc9103eb00f33f7
26478 F20110318_AACMZG iniguez_c_Page_040.QC.jpg
de2d0bf1cf5d1805bdaf8e5f5ed01869
a91b00431a40152829a921efa7ec932c0927eaec
1424 F20110318_AACNFC iniguez_c_Page_128.txt
0859594ff10109922f5914543c1558d5
e566b9d16a68db56645119db84de40b66806c6cb
1959 F20110318_AACNEN iniguez_c_Page_092.txt
f590f21901c88961e2a30532f8781c8e
e79884deae5803efe224bab487e0c4553606c89b
1791 F20110318_AACNDZ iniguez_c_Page_045.txt
ced831c17cc77a5bd3e70225e97b933d
640cd29cea005602887d627b44880d4e9a21f31b
17704 F20110318_AACMYT iniguez_c_Page_132.QC.jpg
5f22eae506d43c501bd874572e7a051a
5e396f4f1c0276db45647a9204554a763cec73f7
F20110318_AACMCA iniguez_c_Page_112.tif
96500fdff653aab38550ccc56162bed4
4ad892e9334154edf2e88d29ba5547f7af0cf98d
F20110318_AACMBL iniguez_c_Page_102.tif
30a48c16753df7b668fb48fb818b7c3f
f643bb5527174d87190d6a0b547ada6098580150
F20110318_AACMAX iniguez_c_Page_004.tif
4047628c29336f9f5b23b5fe22069bd9
5f18d9f70423b1390f9b3541755088b24b0af58a
78512 F20110318_AACMZH iniguez_c_Page_127.jp2
2bfd3db55211bafac3ab9bace83191c9
b5b898caa4873073de5776aaa69dc4e012d5d388
F20110318_AACNFD iniguez_c_Page_129.txt
93a0d419e5bd397f4ba83dbb83e3d1aa
45ba1d340395ff5b528b2dccb27760085d340552
1867 F20110318_AACNEO iniguez_c_Page_094.txt
78a8c83b4a732342cf99d0df13a58fca
defcfce95db5f6f0c71e7295e09f9b708216be1d
49646 F20110318_AACMYU iniguez_c_Page_017.pro
ee81ff2b37e152cf1c24149878b72792
44e2c656048db02784271344a652473a7c3dabf8
F20110318_AACMCB iniguez_c_Page_137.tif
274acb8f1f1999f8f4ca1aece23b84ce
f16adb35d0fcb59462ab2e577b0dde14f0a21cbf
939494 F20110318_AACMBM iniguez_c_Page_042.jp2
0ca2a0292d1327363d230ee528fe24d2
d846b997a10521e45f73adf8447d5c7e6d3e6158
23720 F20110318_AACMAY iniguez_c_Page_142.QC.jpg
96a9a75501b6e5c7aca48baf7bc084cc
791da03fd2f4d9e3650c07b8854bf4a3557a5fb7
109182 F20110318_AACMZI iniguez_c_Page_078.jp2
8eec6251fa0be4fb70ecc8e979af8328
fead9bc446e80d8933f580ca62f171cab2004545
1152 F20110318_AACNFE iniguez_c_Page_132.txt
5eadf55b565491d4785aaa0199ac5cca
9dac79bd6065d419589babe484d1551e5f3819a3
2104 F20110318_AACNEP iniguez_c_Page_095.txt
72f87c68352910d1202e5956797825c4
3fbadd84b1882aa9af7a77af443ba549c89dad6b
70269 F20110318_AACMYV iniguez_c_Page_013.jpg
782b1a41893b19a53a7cc69468bd5406
a264f9a13d3c23a721ad663f94f7e74f049c1606
58748 F20110318_AACMCC iniguez_c_Page_119.jpg
f94d6b1e0c9a88da32fd7069ad952128
eea7149abe961b0268090032f6994002b263a687
F20110318_AACMBN iniguez_c_Page_053.tif
a4aa7fa077b7567fb0f9a928262d9ff5
8b1e4d843a1ce73a9b4091185f6598e4cf9ae582
903652 F20110318_AACMAZ iniguez_c_Page_043.jp2
e392ab3534d898690110012716d3b974
727d82b419117607307331178dfdbc83de8684a0
F20110318_AACMZJ iniguez_c_Page_116.tif
5870134fefc97615613bf94f97921149
8d5e80a1d00d710ddc25d3d1cf0148ae06459f31
1406 F20110318_AACNFF iniguez_c_Page_135.txt
35e4266ce0236791ea625a2f41c5229b
642fd680140ceeb150b9339cab2d0df4f204d42d
2469 F20110318_AACNEQ iniguez_c_Page_100.txt
238b7b56a646a760b1caef7f620699cf
efe8d7a98bcd602e638b4057f14adcabd8e421ed
75934 F20110318_AACMYW iniguez_c_Page_033.jpg
bfb913bc2d4f7547da82d3ded20461db
6cc8133ff1611bc9aa867c34b3e12726aba682d2
70142 F20110318_AACMCD iniguez_c_Page_005.pro
151f9efaffb214ebf39582983b67f111
4d7759c832393574bc4012febf84ad90619bf64a
2035 F20110318_AACMBO iniguez_c_Page_113.txt
8ac01c1f2f57675f7662f7bbe8e2475a
ff937eaa8a74a33b24eff694e87903735caeb38b
867793 F20110318_AACMZK iniguez_c_Page_075.jp2
6be0bdbc1369af6a4ef14cf07b86dee4
4bf243c305c0d4441a34814aa0f15f0d4d4c34bf
1419 F20110318_AACNFG iniguez_c_Page_138.txt
0905a8b0bd5da885802cc2810ed813c1
16343ad9cc005310ec55ab6e324f0b564433ce36
2134 F20110318_AACNER iniguez_c_Page_101.txt
ae17805dfb03283f9007411da0446304
a92ef793916650295bd90fc2ddf7fc40b70cb560
F20110318_AACMYX iniguez_c_Page_089.tif
36c9c93f8d5a2aafeb1540efef4be9c4
b65ee0acba396e2a5b13593bbe55b15880aa1f9a
F20110318_AACMCE iniguez_c_Page_065.tif
8ca9210b689bd2ababb284b427266b44
25fa0b7b947a04ead7045a68d8aa58423d5e14e0
46111 F20110318_AACMBP iniguez_c_Page_124.pro
c5d8d6df8e4a03c4fcb3e6eae87ac428
2a903c440d6a3461f7f6374e4ff67ce6577920d6
740635 F20110318_AACMZL iniguez_c_Page_119.jp2
bdf553f00947261d543d31fb5e2d96fb
adff2415f5461033be66a89282c1e1df710e93bd
F20110318_AACNFH iniguez_c_Page_145.txt
95c65330354f6eca98ae998d803d62a9
b492cb7ccd73cc571fe1097111ed012902b91f3d
1290 F20110318_AACNES iniguez_c_Page_105.txt
c2b2d92ae8316ed122c4db267f8b76c1
e772229fc335fc44221a0bc7be4c1b7129161755
1017190 F20110318_AACMYY iniguez_c_Page_041.jp2
cc0a0adf32e061b4e53c11ed633591a6
c63cb43ec904ee7accf7ed8acb1f9e1e5c7ec637
102815 F20110318_AACMCF iniguez_c_Page_027.jp2
faf339ef8250677918f1c31d0a2c34fb
b8b19df32a07dde94f7fcd099ee8504a0d2a795b
43392 F20110318_AACMBQ iniguez_c_Page_118.pro
75626b0ba7b28b38566d74edfade9155
43e7959eb0f6e2342275c15e534e84202ff9cbc0
24983 F20110318_AACMZM iniguez_c_Page_024.QC.jpg
cf3fe5dfa0cf58992286e80ef5fa53bb
235ce1c2704edd16371302aee781986936af16cb
1813 F20110318_AACNFI iniguez_c_Page_146.txt
76f8af9c85726e5a7ba566bd21431782
cbb662adcc2198da5214cd73df565f6932333746
877 F20110318_AACNET iniguez_c_Page_108.txt
911d295b055a26c7a6325af1e30ace8c
9b108fbfb7d320c72a4d562cf4e59a31b9934d5d
42878 F20110318_AACMYZ iniguez_c_Page_120.pro
85f8d38d5ff0f6a4070a51b357a272b8
fb6316c3a825fbddbf66cefd8d242e990715f74d
2864 F20110318_AACMCG iniguez_c_Page_005.txt
5c74df08606c401304081d356685d628
c22c32621d4374bd3cb641e72f83c6430a996fb1
6145 F20110318_AACMBR iniguez_c_Page_092thm.jpg
d75e7aae487ea3fd3944e017d41973de
eccea798e567637a1c0e05ea8439051af47af65e
58506 F20110318_AACMZN iniguez_c_Page_034.jpg
81bac309c520a5c69c0e9da212acfa65
95db3ef066ce7f4fca3d9dd2433db5cf11ca9ade
2023 F20110318_AACNEU iniguez_c_Page_109.txt
3a38d029c3550a874f2365a2c48e5dbc
89ca03201d8f7b24e26d66ce789e8cf57745996b
2071 F20110318_AACMBS iniguez_c_Page_091.txt
56939d0b65d7ce9bf759f83036633042
655ee66e901dd0e5f90c06c450c6de8e1fb3113b
456391 F20110318_AACMZO iniguez_c_Page_114.jp2
d0f1cc59b613bee65527418b1ccb14d3
f13e5cd6f037c6cb4decc9ca65c216f7b9d09375
860 F20110318_AACNFJ iniguez_c_Page_150.txt
4831065148195be43d1b6400efd492e3
f5f38e5a4e764c48ba3875e5c3c519a93114f009
2643 F20110318_AACNEV iniguez_c_Page_112.txt
7f6c663f50e8075497302a2819d85347
de95572b6d517aa2241bdf39647dcd94fada9662
108193 F20110318_AACMCH iniguez_c_Page_023.jp2
fa92c191d3f7965dbe1d776a45b3a105
fb77ed77e81c86ce61cb3a0629188b85106e56a6
5470 F20110318_AACMBT iniguez_c_Page_070thm.jpg
20057c2a8d83dbbec95e3f0e70935fe0
6ad130e1e16991dc4ac0dd378c8e7d2b48cd2bb4
F20110318_AACMZP iniguez_c_Page_070.tif
5f81dca659dbc104353fbd4e39e15e50
477655d52415cb43c6a98b42b400727e94062c13
8195 F20110318_AACNFK iniguez_c_Page_001.pro
71fab3d278c8c8cc44b2c31a0d682e93
4c7bafb5a2b495a123404294a15ef8dd46ced9eb
2284 F20110318_AACNEW iniguez_c_Page_115.txt
879a22f4e213dbf1573b46f103fc4f85
a0dc77d166ad968f6f759c04591e5ee060a01176
F20110318_AACMCI iniguez_c_Page_106.tif
f7a09f81688a28d7b20653599f3a2217
5955759e206521cdbdbc76e183299ba6ed2a8c04
32258 F20110318_AACMBU iniguez_c_Page_082.pro
46ba4d49b19a6f1f886e8bef27be1d3a
80634696e8fe0680d6f0d9dc12ba9c8545a6a62c
74782 F20110318_AACMZQ iniguez_c_Page_006.pro
9e2d60aa57625c7c9ba4db1982147bad
a64676707c970f275405efa66f11b3892c4b5412
27442 F20110318_AACNGA iniguez_c_Page_069.pro
febb0f90e881e2c5f53edd5df2ccad04
1fbd943c1176754b72515dccfa99c996d0dab725
43950 F20110318_AACNFL iniguez_c_Page_015.pro
ef760c7af7ca27a45c3b2a5554de9911
eeb9eda226e99e7f878baaf02d9a6d7b372d35c3
2668 F20110318_AACNEX iniguez_c_Page_119.txt
92e138574bf711024e862d95de000972
b01a498600366874700e94a7f507dcce78f65fce
F20110318_AACMCJ iniguez_c_Page_043.tif
67f714c28da3b5b439ff19c85602a1ac
1b3ef934c927790a9ed9b86275a63c00f393b0e9
27830 F20110318_AACMBV iniguez_c_Page_148.QC.jpg
b3edceeaf34290435ec859a862748497
49d5c2010276b6684f395c16f1d907c304741096
25057 F20110318_AACMZR iniguez_c_Page_019.QC.jpg
756865f9a312462804f4b708094eef98
321b1efd241866de7369858d3f0a03c332cd3141
51418 F20110318_AACNGB iniguez_c_Page_074.pro
59700ea0fef722be4f5000b702f25e22
73a8d35465d741814096ce29e4146560a726370c
55296 F20110318_AACNFM iniguez_c_Page_020.pro
f76dcd602e422ffd48153d07b2882c0d
e49165c1b229910bd34778cdb4c4a45bf9d562f8
1337 F20110318_AACNEY iniguez_c_Page_122.txt
eed3560b888652a7edd4c5d80c4a11c8
2807a96f019bfb10dcbfa53e17f8d70675ac4bb1
1795 F20110318_AACMCK iniguez_c_Page_022.txt
18f89d155bce23458588569c4ef054a5
2798c467b008e7a7d3a54ddede8bf37ffc227e74
6217 F20110318_AACMBW iniguez_c_Page_065thm.jpg
284b80ed44244201b1126286f72d1cdd
dc0877d4bafc557e07f07d093e9ecfc1ebc153b4
46381 F20110318_AACMZS iniguez_c_Page_014.jp2
932ba5b0f181d8600d076e7aa554714e
91f1691bd794b4fd0ba859dcf11a4e2f23fc6e7d
38831 F20110318_AACNGC iniguez_c_Page_075.pro
ceebe3b22213d23a3f5e320361b54f40
fdf9b7d193d0fbd92d9c6024ad5ef390a2c542b8
40001 F20110318_AACNFN iniguez_c_Page_021.pro
83cd0a09f3d70e5e6ddd841387aeb96f
b6a1c926d6766768042cf8b88e5f8f5c28655cc6
1818 F20110318_AACNEZ iniguez_c_Page_124.txt
e08b4d0e43079ed90782b719814514a7
12c98393b0da2377477f3ac27ceacf8bcc64bf3e
61363 F20110318_AACMCL iniguez_c_Page_058.jpg
b8c6bea0c478da5fe6ea3b48d312e3a7
4df27acc51dec763b2d63c43917f7883e8e78cf4
40276 F20110318_AACMBX iniguez_c_Page_061.pro
7c879da861c0afb2f907f4fb708cae7d
ab4c3c988bcd909a91759222b98a5967a4353e74
127432 F20110318_AACMZT iniguez_c_Page_147.jp2
b6eccae57d25065f6fa1e95ecfc58e03
ea1102e28d35156270294683af281606727b08cb
1661 F20110318_AACMDA iniguez_c_Page_076.txt
3de5c8dc4472bf244d1687d7e9ad4af4
f9bdbca58c7467d0683b52afeea5bc9c2f3a54ec
39419 F20110318_AACNGD iniguez_c_Page_076.pro
55643eda5a47f4c084784bf2288c0e42
86ba8c30a18ce0001513d22239ae97bcc8f04ab0
49465 F20110318_AACNFO iniguez_c_Page_024.pro
9bef86556196c5ccc23aeb577d5e39c1
b8cd9d1ec093ca5bcfc6957fd1ec7bc791586502
21185 F20110318_AACMCM iniguez_c_Page_067.QC.jpg
c78811422805d419c137255ae0b4deee
5409eae2a934b898c0118955257625594561cedb
76605 F20110318_AACMBY iniguez_c_Page_109.jpg
42af62b500a8f2af6e874d10d3667516
2ac0ccc0ba48f165e80a23571c57b24234182d64
5798 F20110318_AACMZU iniguez_c_Page_053thm.jpg
06b6bacf943922e2c4c831af44d8078e
8dae3846541d7bd39a763911615c9b31cb0f93d1
33515 F20110318_AACMDB iniguez_c_Page_125.pro
9b89e676d7ee1628e6fddaf3580f6b80
5e5c16641dae582c8f528665e12bdcd890df3e63
35072 F20110318_AACNGE iniguez_c_Page_079.pro
697d0a872d8b99f35fd95e2d52dcc6ef
212859d874719d32e3cb00e7c2bd388d35d8f97a
46880 F20110318_AACNFP iniguez_c_Page_026.pro
22e48d2d6196cac7929c426c8250986e
58054efc95f18c3422d62043e556a17999e7607a
19370 F20110318_AACMCN iniguez_c_Page_102.QC.jpg
67603d9d42160ec10c22ac466d7f5187
b5f10c933383c2d3339a42e99b5187e96b14ac7d
24765 F20110318_AACMBZ iniguez_c_Page_090.QC.jpg
c067a7d96e945fd2ec1e21e8f705d6dc
129c12598c3cf9fd6be33f583f2854580c334af8
27960 F20110318_AACMZV iniguez_c_Page_097.jpg
8e34258013554951708c4f3e69416be4
9b096d51077d3d43a92b0c34ece03593b9a9387f
25748 F20110318_AACMDC iniguez_c_Page_028.QC.jpg
a93eb673d901e541c9f315c1de27c1ba
4ff07715f3c321fbf0df2e90615a1dd3f2b4f114
29535 F20110318_AACNGF iniguez_c_Page_081.pro
e5ca4ffe7d53e209983685c3d7b23744
8b9960495a8e4a25ad6fe0a24ff64557cbd77933
51045 F20110318_AACNFQ iniguez_c_Page_028.pro
9f914db0616f924b82a792a25b03c7cc
acd3b78ab374b0f5fd7771f9aaac74a0fccd0f65
4769 F20110318_AACMCO iniguez_c_Page_004thm.jpg
d214920223e0aa007f82fd8e07eab539
939cbf669e0b65bc35793649fd6a8e1751d1ed99
2961 F20110318_AACMZW iniguez_c_Page_011.txt
ade7d259338b10221490d9461d11a206
f3fcadf1f355f400350eaea280ebc4d9d97bba33
4481 F20110318_AACMDD iniguez_c_Page_061thm.jpg
07bb505b563e980755f5b883f9a12711
bc222a192e5eb4d4c0828448aae6a7db34991bc9
44850 F20110318_AACNGG iniguez_c_Page_083.pro
516029f1476b6537bd6dc0213eade87d
bbdf8ef1ef213046caa6c833e24ba8a62a7fa4c0
35499 F20110318_AACNFR iniguez_c_Page_032.pro
503d90e5e8a9578bd58b6eaf61b172df
ad4426d1689f096d44ac1cd659fab3f4ded59aca
F20110318_AACMCP iniguez_c_Page_032.tif
b13336c7ba135f51113322a36f5206a9
1a774d1b9b62d955260dd3f790702c37724d6b49
61686 F20110318_AACMZX iniguez_c_Page_115.jpg
47dec682ba88320fa56364e667f51d83
d4ead92345750c2554fc3533ffcbcff99a7814f7
32609 F20110318_AACMDE iniguez_c_Page_134.pro
d2ef3c9ec882bf9919ef927101a78348
2d430af5f66b4e0a6ac861de33a1a158317ffb6a
50734 F20110318_AACNGH iniguez_c_Page_085.pro
765b68500f60cbab58d53b40c2cbd5ad
89e419a2e65ee7832ecfae843d5b74b0b791fec6
44995 F20110318_AACNFS iniguez_c_Page_033.pro
0738c9de5ee265f247d503a0b69a39e8
00644f66d2b9b6e0a89efcddce36e894d3cd3f11
43673 F20110318_AACMCQ iniguez_c_Page_115.pro
b4afa10c496e6a0104315af8228e5694
f903b518912f9557a62f13933463e90c298b13b2
108034 F20110318_AACMZY iniguez_c_Page_018.jp2
14f6c7b6a17b0f56e48c85d681b51d00
ae6389b46be368046f2fe2b2df6973f034fa7d85
18715 F20110318_AACMDF iniguez_c_Page_087.QC.jpg
ea386ab21873b90e1fdd2f1cbc378ef0
bfee14b24e3ed8b5737f54db7427de170f2f7cd3
37705 F20110318_AACNGI iniguez_c_Page_086.pro
35e5ed06e2ffe60edb83c89f0f58fb51
357a30739348aa88106e11aa1b2ae1ea18ba59bb
50494 F20110318_AACNFT iniguez_c_Page_047.pro
5694cae3568c280c7f0ddf4106a0c477
40d8066fa5feabddb0d974d17d3ae30a740952d8
18956 F20110318_AACMCR iniguez_c_Page_034.QC.jpg
ce9536ecbc5f4d3bfa5bde17d2eafeed
18ffc82aee011259aa3d69d4546f78456ed0fb98
2065 F20110318_AACMZZ iniguez_c_Page_029.txt
5d7500d37873517105cf9407edbe85c0
7c6dfdac012395978067db92f31d99e6229629fe
6526 F20110318_AACMDG iniguez_c_Page_001.QC.jpg
b51ea57883c61f09c7d03b118919a7bf
a509f1e93c6718472c1cf61b933c7480cc324570
43540 F20110318_AACNGJ iniguez_c_Page_088.pro
b1c46edb2a4e7e321210b3ad6346fc4d
c1027b1f56839abb9c108bb2ce7d7a899a9c750b
36441 F20110318_AACNFU iniguez_c_Page_049.pro
72a67e0da95f495a4a053da019df0bd3
63df3b6274795e4a64542072a2fc8281b46b8b62
15759 F20110318_AACMCS iniguez_c_Page_126.QC.jpg
28f358151e778d85c1604cf8400fed45
e71fcee8950723ee9bfe86730aa84cf48971be67
61819 F20110318_AACMDH iniguez_c_Page_066.jpg
5da13b4b06628c241b02652deff95b31
55cac4333f7bef2b38834960e077820591ba2f25
30959 F20110318_AACNFV iniguez_c_Page_056.pro
ec7a40b28f3d500be92fd49b7a3ca2a1
b3d4e07b7899371bda82a6609d01200697c7c3e1
1911 F20110318_AACMCT iniguez_c_Page_099.txt
e52bd14a30d626c3f271ccb2ef06fff4
e4cf5fbdf26206ebcdef1e0ff7fa105b74c4a086
52676 F20110318_AACNGK iniguez_c_Page_095.pro
4bdea2953305d287b07dedde767203cc
b8a9dfa18f2d746f8642a43087874f46970db9e2
31386 F20110318_AACNFW iniguez_c_Page_059.pro
97df54715dc8ec538c0b5fefe5a5a20f
01ca6db626154f42795bb88b90c1018d333a5df5
20052 F20110318_AACMCU iniguez_c_Page_071.QC.jpg
7d61767d3480758391df1604692d503b
c0a42e0bbe6c037c6ed94fc8c4b5c023f293d1b0
109962 F20110318_AACMDI iniguez_c_Page_016.jp2
81a407d68e7e82f4b75af4401f4e61d6
83ad9eaff627c31ae465f083c54be06d847acd99
15403 F20110318_AACNHA iniguez_c_Page_005.QC.jpg
21d4855efb94f58ffe7388da8e3c234e
8453eaacfe83d88fc7f867219108eeea773a1063
12179 F20110318_AACNGL iniguez_c_Page_098.pro
c273fce21fef5fefc8a74ed49cee3182
6ea79d98665352a76e6b7d6b4596869ae477b2b1
51689 F20110318_AACNFX iniguez_c_Page_064.pro
6cc6c9468af15b7d25583c1e8e721db1
ea9bbd4d7879512faab2b9c59f48fbc8aa0d6950
25348 F20110318_AACMCV iniguez_c_Page_126.pro
8bc9cb1ff338d0400050722ce67c7fd9
d6f0b00875c91aeef8554cf8bad14c40de425b6d
27157 F20110318_AACMDJ iniguez_c_Page_036.QC.jpg
b56316db5c506e7c8378b008a2f82b9e
14ca4fdd3d70de78993b73d73987d400faf103c1
78772 F20110318_AACNHB iniguez_c_Page_006.jpg
4f793619857584dd703d9f8e0847efe6
5aa30c4e3906a400e5c723e517f7f9ad35eb320f
23907 F20110318_AACNGM iniguez_c_Page_102.pro
bb6c7ec2c02bae9d47edbdbbecdb624c
da43c9e0a0a7ac01582ae8b333be1cee7ece5065
50027 F20110318_AACNFY iniguez_c_Page_065.pro
faa660019f93ad0c518512a31076ac75
9e5c9b46da66d1d616d17fb8e0c78580db5b410b
F20110318_AACMCW iniguez_c_Page_101.tif
6aa15cb008be9508898dd3a7aaae93a1
d50281f1465da454c28f67e4d36315b5077bfd2b
73446 F20110318_AACMDK iniguez_c_Page_111.jpg
a62c3405ac17a50f6cf51b0b366817d7
99914b99e430218cd241bfe2a373389407c2c9f4
17586 F20110318_AACNHC iniguez_c_Page_007.jpg
6041dc218719d505b923fe9a719ac123
db9ba6e063b9b0b3c6d36fb695727383b7e3503f
47730 F20110318_AACNGN iniguez_c_Page_112.pro
3bdc4dcbc2a96943d246c153db6c3794
75607faffa162cdf99ec11d0d51d485b8e34e714
30274 F20110318_AACNFZ iniguez_c_Page_066.pro
8ec6cecf07d86a004290c0ebb1f17646
82173a6d06625734c20f13f068c566a05927c28f
14077 F20110318_AACMCX iniguez_c_Page_063.QC.jpg
f8ff2146edfb19e9b0b27330c2d5e705
971d805f34ede379005828a5d79b6c02f4a78d3b
F20110318_AACMEA iniguez_c_Page_060.txt
b73f6e2380a1e510c25eb8032f24b0c8
511ad35e4d686dd0fb8213941ed884a3f1a8d621
F20110318_AACMDL iniguez_c_Page_087.tif
8ada6aa42053c9473bbdb0c0a642a445
5222d536031bd7e234958df9cec5457abb27133e
117053 F20110318_AACNHD iniguez_c_Page_011.jpg
763e53ce31499f5465db20826d9291f0
943c3c4fab559d0076c0b703288d4d7ba35cc95a
37224 F20110318_AACNGO iniguez_c_Page_113.pro
02515d7e673c56823bccdf74804283ee
44ad6411ebd812dbd419a10aa65bddda5eba22e8
26467 F20110318_AACMCY iniguez_c_Page_046.pro
a5e40eff1441aaf0e88dc8f4d956a52f
ddc33005bf3a74b695e103bbbaa5c874dda9c855
28821 F20110318_AACMEB iniguez_c_Page_012.pro
b96f2408b1ca7df70c9a686a5644a1a2
a6b5d02182d66f111f7926c3918caea6f887e00f
4025 F20110318_AACMDM iniguez_c_Page_006thm.jpg
455982e001f82ed2de9bf0b2cb099c79
be95a81c86808037e44b7b8e565f2b9e42f69200
43773 F20110318_AACNHE iniguez_c_Page_012.jpg
1a081c58335be62df28e550bd2818a06
d642c77e40387f561f2ecf2621d0fae10ab6fecf
13031 F20110318_AACNGP iniguez_c_Page_114.pro
bec516152051d47b053ed376259351a1
4165a984263a763e1d54d094a031054c2019d714
63215 F20110318_AACMCZ iniguez_c_Page_050.jpg
7402063efe303a5c7ec6ade8bc5e58c0
2ca6b42569da378fddecb59a78c6f9c6cd2cc352
5474 F20110318_AACMEC iniguez_c_Page_146thm.jpg
c55e250b1a005c67a612425a69a8550f
6c23b4053ad95aac1b74961e10edbb28b1a9bc36
F20110318_AACMDN iniguez_c_Page_132.tif
f6dc732cabcbcaacaaf1795b9f027e6a
ec157752612acd5171d415359953e47ec096a53b
11318 F20110318_AACNHF iniguez_c_Page_014.QC.jpg
d8f5dbc04935e7b0efe80b2159f54dea
40a6624165b97ae2c92523fa68f013cd38841a05
45214 F20110318_AACNGQ iniguez_c_Page_119.pro
f68b9af8aeb447d9bc01490a3336322f
3a58079d7d9905f5b6166b180a1b32479df899a9
24512 F20110318_AACMED iniguez_c_Page_124.QC.jpg
be7a72eb125bf1c4f3d200cecf875a24
0f08017969035494d34942f6c0bed94e81060eac
109945 F20110318_AACMDO iniguez_c_Page_093.jp2
2c1f4eb6e0f27f15fedd831728e57817
a3fc98245fd8a2718a079fae31a77aef236a941c
74003 F20110318_AACNHG iniguez_c_Page_015.jpg
837d60492161aa8e5f16bb3a5eb6b9e2
60f29c2466cdb7eddcc1d80e40eff705da3a9513
30037 F20110318_AACNGR iniguez_c_Page_123.pro
af61ad19541285702b8153bbf7ee4a83
7dfcb67bf9c369798467cd515f03edc766f61d29
61997 F20110318_AACMEE iniguez_c_Page_149.jpg
4acb4f6eb28339649a4f714cf1ccde66
15f7cb519795a4b662636586c5ea99f2fec1bbc7
44629 F20110318_AACMDP iniguez_c_Page_045.pro
b39a6037d5c0f9a98a314fe24d7d35e5
e9c09c6e158f5506bf2fad31010a3c5327bb3d90
22340 F20110318_AACNHH iniguez_c_Page_015.QC.jpg
19dfc0a2291fbef47a5a2568881fcced
a318dac241426b2508a63d3726059c7895f97508
27117 F20110318_AACNGS iniguez_c_Page_132.pro
93580d3ac06d93165790184426be538e
9a47a7b966537a0278d457b0f91d299f24f4da0e
4051 F20110318_AACMEF iniguez_c_Page_145thm.jpg
97308acfc784c2756124b584a90a2222
47141d77299e2c9f23c9cbca85b57780481c3add
1742 F20110318_AACMDQ iniguez_c_Page_049.txt
cf0dd1aa6f1967ba9c22fed8f26dd3c5
21ef8a3817769ca1b8d7a16bcb510202790b801e
81219 F20110318_AACNHI iniguez_c_Page_017.jpg
eeb3f24616513b6271883e8192b13703
525fc5f7e450db164638b4a6efe1a54934ce790f
33988 F20110318_AACNGT iniguez_c_Page_138.pro
850664932c5f64b8eb35f5c022fca1c3
f7ff00172f7c1d154f0bf9a140aab37ae47af2c3
2085 F20110318_AACMEG iniguez_c_Page_083.txt
9b4fd0194de5df54c9fa76276e670a81
c727c16a9e3dbace54b6d11ad539772696a28ca3
15388 F20110318_AACMDR iniguez_c_Page_097.pro
a66532eb0a2467f738a636015cf487ac
ab3b1fc611ecd56745586dab22322d7e0a4ad2a9
24771 F20110318_AACNHJ iniguez_c_Page_018.QC.jpg
393fb92a6e4cafda1782913c0638982f
fdd3d8c88f055781f51f58d5a5d890338cad0de5
26589 F20110318_AACNGU iniguez_c_Page_141.pro
3b9a3e5d070ae6aef0e814458c1be897
35a1dde95096f68d66a52970f228c1913f4a16c1
21468 F20110318_AACMEH iniguez_c_Page_107.QC.jpg
d8e6ed9754530d5f53a1bf8950798513
2a244106c6da855b61cf116876a27b4971e03a50
F20110318_AACLZA iniguez_c_Page_003.tif
0a4755ea2b4de32e26fe4d950205b3f6
e9279d9c9edff9716bcc8c9686f346019cf68304
1051981 F20110318_AACMDS iniguez_c_Page_012.jp2
071b02c0cd091b4cd6995260d74bd753
38e046d05d5e9fb387c206b11e65ea96cb37af0a
80249 F20110318_AACNHK iniguez_c_Page_019.jpg
a3ba6389a92c7e04fd48195fe20defbe
8df4e5e6c10421b0511e6d647e93623e02a11f5f
56464 F20110318_AACNGV iniguez_c_Page_147.pro
dcbe30d8e38828c1f014b49cc34ccd36
50c99aff2eb7c4ee13267f96b71688406d68947e
6574 F20110318_AACMEI iniguez_c_Page_107thm.jpg
fc8e8781d078d63cba7d6848296fcff5
318093d2bfd3389144feeb5402e53be802e1a435
6123 F20110318_AACLZB iniguez_c_Page_035thm.jpg
42ad8f14339f8ff6122d212debfe1cf4
0c9060a7cf467a5a4bd9747f74a4beca840ffa13
26390 F20110318_AACMDT iniguez_c_Page_085.QC.jpg
ead694b18d6ab8d446c082e1d7e2840b
743cdeec89b090cb87d58827a7cf7de721df17c7
59075 F20110318_AACNGW iniguez_c_Page_148.pro
458462fef8bc391ae38ce4d1e5496dec
f6dadc8d9c491f36232773fb0af6653acde82336
62528 F20110318_AACLZC iniguez_c_Page_086.jpg
c8609f3a8581bc7fc286bb2d20e957c6
7a661112e7149386a3b26e9f300505abc2f9b032
4220 F20110318_AACMDU iniguez_c_Page_133thm.jpg
20f35ae50c46108ba4554310c7e155af
561f78e2846ca48a970b735fd9076e2c38eb74ee
15075 F20110318_AACNIA iniguez_c_Page_055.QC.jpg
29eb30551c4dd319bec7478b0b383e45
9c14e1fedb9fdfdeecce51e4ec8402b541995292
25922 F20110318_AACNHL iniguez_c_Page_020.QC.jpg
f9a8af28a97cf625bb1ba7df427c94c9
29fb14bc680f4c493814db91b0c0f304f1ca1e85
20496 F20110318_AACNGX iniguez_c_Page_150.pro
ab82fa4c3e55d0b66bbb2f4dd13bb5ba
f82e881cfd00f6de02f1b949dd79c7d58f5f8a72
52672 F20110318_AACMEJ iniguez_c_Page_084.pro
c47476da35ad1e75aee17782ac50db74
e14a8d1580099117eb4364d21d83ec274b44e61e
5620 F20110318_AACLZD iniguez_c_Page_045thm.jpg
75737762f159127256e706c83869674a
bd17f95fe8c02c250588a977a4a40b8457f298d6
F20110318_AACMDV iniguez_c_Page_149.tif
89a5cb69f41f0096839d3c2deb954026
7a76795ad32c36aa264a8fe6478fc507063a431d
29771 F20110318_AACLYP iniguez_c_Page_127.pro
10ee0f16d07edd31ef3b39b4f834ab60
ebeca5596844d395c54cb23641a004833d9e8849
50924 F20110318_AACNIB iniguez_c_Page_056.jpg
2ab1446215cc6a78a53a370e6c46e05c
bd2cfce6c6a0dfec79d52aa19c412ccf17a6c6d1
72488 F20110318_AACNHM iniguez_c_Page_022.jpg
f915b0594b995b6e6abd66c0eb50ad14
c5aaa0749d4822752a9de8ceacd6ea0c8318bbcc
1446 F20110318_AACNGY iniguez_c_Page_002.QC.jpg
9ec6f58699317a1ccaf4350e9a478bed
ba33ce3b494b601dc7a526a2632691a8b3d40743
F20110318_AACMEK iniguez_c_Page_021.txt
9930d95551de4cb854ebba03c52388fe
58bac6bf2c4d0d895128a8f6dbf5ce5ba9e9bb36
17089 F20110318_AACLZE iniguez_c_Page_145.QC.jpg
88277cac3d1b502d5c7842304510280d
eba9d7cfbf0a35df5b5bbef9632210ec4280e40a
27647 F20110318_AACMDW iniguez_c_Page_073.QC.jpg
9d32730951a27b9a023fa91c34102eec
25fd9439f67ce71b57eca7d7a021732008c2eb40
4876 F20110318_AACLYQ iniguez_c_Page_102thm.jpg
c271525010b6a20b2e488f12e8509e02
256bc6470dd178b7de73e3b44fd2c3318a37e8a8
73609 F20110318_AACNIC iniguez_c_Page_060.jpg
ef0278b9c5bfc4ff93d7e3b852270413
9f2a5742abca7d464b4ace9b07760a89c81e8c30
22338 F20110318_AACNHN iniguez_c_Page_022.QC.jpg
bcf6bbbfad5b12435285f77497e388c5
00d305661492fd7946de4cce2d97ce6a59b232b0
18966 F20110318_AACNGZ iniguez_c_Page_004.QC.jpg
f4a14edf55d229908adc34992bcc4c7a
60eba7ee0f23e89dca122e76732c835735270539
6372 F20110318_AACMEL iniguez_c_Page_036thm.jpg
dd070278726ac6b71d1a65fe5fc72553
15e4c6f4632ed17303bde685ecf6c79878e53926
626 F20110318_AACLZF iniguez_c_Page_003.pro
9aaf068d467d2f5f20ba44e792cceab9
0738e640e44d87e824301f6c16f38781f90ece4c
39255 F20110318_AACMDX iniguez_c_Page_058.pro
07394234a886595f876d6c3b220f658e
8c480c566934a3fe430a257758fd98143fb86d03
5237 F20110318_AACLYR iniguez_c_Page_142thm.jpg
28be29700ef2923c3c325743ffc8f7a1
733b1a3903e27f521aa3bc04b03dcb81062006d5
35859 F20110318_AACMFA iniguez_c_Page_042.pro
3a86bc3dc1b90be504aa216283120ecc
4671f3a503001876269e40950abbd374c4387aef
51797 F20110318_AACNID iniguez_c_Page_063.jpg
5129cedd5a9079e347c32d2234196d5d
8c2f1be4da67d8600781c1e8bdd977e4e9475649
82352 F20110318_AACNHO iniguez_c_Page_023.jpg
abf57ab78ae7951fc62f94ad590895a7
e8b1d733ffe7daf2bb67810989180434d0bbfe86
18198 F20110318_AACMEM iniguez_c_Page_141.QC.jpg
7dd19c97d52fbb35bd793c09b38f932b
6fdcd00125e4340daa294156cac3e5caee9b78b1
48904 F20110318_AACLZG iniguez_c_Page_027.pro
9e6b772a212d747bc310dca6ff753dc6
419db795bd60058652c2c3e8eb3ca880ec40829c
4736 F20110318_AACMDY iniguez_c_Page_139thm.jpg
03f107cf2ad14d30aceaac25acca096c
2483af9cb234f1741abf503bc55671524b5ad0cd
20685 F20110318_AACLYS iniguez_c_Page_066.QC.jpg
3f41ecb51c29c540b83f319423a45825
9e943027a963f4f2f30fc4f17f8ef60e4ba6a840
71835 F20110318_AACMFB iniguez_c_Page_054.jpg
c739c498adfbd0a19fedf5767b2f6a47
7aa0c9d5492df9674e1a7c83008e46127427f31c
84065 F20110318_AACNIE iniguez_c_Page_064.jpg
7a3718e9dbbc2c7ea20fdd044d573f9e
304ce7ec89b4c1c817d456540b5f2b3020ea3a3a
85508 F20110318_AACNHP iniguez_c_Page_030.jpg
a38a81a2326e1c6d4d7195cad82fbb03
c7643f3e850e0395384e396e26a8f5feb9ce555c
1796 F20110318_AACMEN iniguez_c_Page_054.txt
4f37327bc5537b5ef0f3004ee17fffcf
a943e7753969428831333618a10ee7fbc220fb2c
886 F20110318_AACLZH iniguez_c_Page_003.QC.jpg
a8d247660fcc892cc923c12dfa7e1548
2ede269c51086f244ee8b5a8d66740b0b957b0b1
F20110318_AACMDZ iniguez_c_Page_133.tif
f123db33251a9640f6e59812bfd40fb5
90ea5e59148aad19f64ae6a881bbb7f4f8e29eff
2110 F20110318_AACLYT iniguez_c_Page_073.txt
cf478519c2766430e99a7ecb18cdcb21
75029c95d493f5622e7c4697bd7ae93c53544516
100898 F20110318_AACMFC iniguez_c_Page_068.jp2
79bdc6b432ac6d364065fd69b986e062
06c480d2ea8142029d6b55f76bffe994a0f78686
26519 F20110318_AACNIF iniguez_c_Page_064.QC.jpg
b769ebe87a14fd42bf03301a447b42d9
a154e640135076d3f6c17c217f17523505c3ba86
26290 F20110318_AACNHQ iniguez_c_Page_030.QC.jpg
f5af882fcb17473f6f784d3cf158ffb2
4702cbd2678d8dc770399c688633ff2e193ea8bc
F20110318_AACMEO iniguez_c_Page_111.tif
89aa1acba83edddc009f9cfa8b67a26c
bbd9ef4208e586650b6d1f9cbfa7dc5c172e4b66
18818 F20110318_AACLZI iniguez_c_Page_058.QC.jpg
fdf621d66177663bbea229bcc6d0de18
2e9fabf76fdd76095e97014ec63211ad05ade266
94177 F20110318_AACLYU iniguez_c_Page_124.jpg
816fa952a9d85e993f8bbecddc2601cf
0796a9848c885a0560bef3a4e0e53980ab85963d
6656 F20110318_AACMFD iniguez_c_Page_079thm.jpg
e5d8c9789f7c894a091af80702854fdf
d610afe47abf2a19d299c24590609304bbd998a8
25249 F20110318_AACNIG iniguez_c_Page_065.QC.jpg
febc6f25039500dd8b4a4c60bcd13be2
03d780bdaa97a4952b14259d4e612b6ac811a207
23033 F20110318_AACNHR iniguez_c_Page_033.QC.jpg
b415db99e285c48ae2b74197b23ac798
95bd4f5d1004f287a3465f5dbe2e4dc0b5db2393
6841 F20110318_AACMEP iniguez_c_Page_083thm.jpg
81c8477ea5901b9915e581871c499d69
3057715f995a02689357de7e9d07677f2579928b
78189 F20110318_AACLZJ iniguez_c_Page_131.jpg
2edbe7854528141004e2e020fc2ab6b3
5044070a254abb7a1e28245970840a113db076af
110431 F20110318_AACLYV iniguez_c_Page_010.jpg
166362b5d9c8ddb3676c298d87b8b5b8
6e750fb5ff42db05fdef7f8d61f34783097d8ec6
23520 F20110318_AACMFE iniguez_c_Page_079.QC.jpg
3d78c3b7bcbd6c213af96e3bc58c3690
c99833155aac101395cad3f944d07e8986aa4afd
14863 F20110318_AACNIH iniguez_c_Page_069.QC.jpg
c31936843f70ac18420c20e36f4a973f
fb46471478caf2ef056461e1898a5dd0220f6c6f
18990 F20110318_AACNHS iniguez_c_Page_037.QC.jpg
2598a49f17309423cdbbe43a0f51c0ce
e3b470dd622dc9da8d969d8d9c4e1f3ac61ae581
782521 F20110318_AACMEQ iniguez_c_Page_108.jp2
ed8146cb7e252e309b3d6f231267f7f2
ad98a3a8e013bfaf927bb065cb0b7284d64a77b7
F20110318_AACLZK iniguez_c_Page_135.tif
41f8741fe70fc5e634ae139cbf02f91c
b6885483417b990cf0ab74a2ae843676dabd6867
24355 F20110318_AACLYW iniguez_c_Page_094.QC.jpg
92648e3f7e05a52e08a47ff57603ecc4
07162f7281ac8489e17aeda2e0407adf51ad375f
28364 F20110318_AACMFF iniguez_c_Page_143.pro
86d177ece71e145d8e8dbeec13c1a660
a7806003025d2f87e43dfb54def6afc302279200
70543 F20110318_AACNII iniguez_c_Page_072.jpg
1777c9c233ba3b488d3adf4323c4aed5
7e9f992e435289e9396760a94f27286c52b23230
53547 F20110318_AACNHT iniguez_c_Page_038.jpg
e507a9e3f7c3f205f647825c3106bfc0
f60dcc168ce57dd9ab2f79e28708b59463c51217
83490 F20110318_AACMER iniguez_c_Page_074.jpg
aaafbe2b88b15bf4b59b93bb061d6a5c
5a1e53d708f0420ba65a880829e0b44b02068b78
19848 F20110318_AACLZL iniguez_c_Page_039.QC.jpg
dc08a987689009e283eb6d525f776fad
2e5c57d83c72da124375cd7a29e431ddde9667a0
23161 F20110318_AACLYX iniguez_c_Page_081.QC.jpg
77048a00cc5439dbf248bfd2c2c61e33
a3219467983e48fb7c5cb7de55342aea4520461f
F20110318_AACMFG iniguez_c_Page_079.tif
410e7685b0f2c3aa2871076ac5e72486
850aecf698d9e3fcee4380b643d9114507cf9888
21449 F20110318_AACNIJ iniguez_c_Page_072.QC.jpg
81ebd48aae46584928b7a0ed3cfdcba7
65abad8ca2f0b4dae44744711a3912797059d468
21518 F20110318_AACNHU iniguez_c_Page_043.QC.jpg
e77ff96865655ff976b1d3a411b820b4
22c3195f8cad8ea510656385c795cb6094e466e8
106824 F20110318_AACMES iniguez_c_Page_092.jp2
45151860007f20dd1535820fe40c7ecf
76deef928153156327da650089d2ee476eaaceb1
2025 F20110318_AACLZM iniguez_c_Page_093.txt
51385bc99e85bbe845a73d317006d1b5
5f235db7bb251197bb3a58a1808950c238ec22fe
1202 F20110318_AACLYY iniguez_c_Page_144.txt
4d5b43c0b951cb76d7bc1baa0420eb9e
f2b2c36a92988355be03bde55d41b81a6cb9e40f
F20110318_AACMFH iniguez_c_Page_092.tif
b553c4391fd83b4758d71054cd66ccd5
ce4ca7228f861a0f5b31d34cc0884f0d624d8b31
87730 F20110318_AACNIK iniguez_c_Page_073.jpg
691ea16bef6258bbb3869e52008344e7
3a4fe77c829dff5cb5ee7c6bf7ab4e3e9b6eb5ac
23218 F20110318_AACNHV iniguez_c_Page_044.QC.jpg
98003167651042e01ba931feff874598
e760ac0608dfef966b7c9bfa8280bb79d6a3a2f7
111617 F20110318_AACMET iniguez_c_Page_096.jp2
d102a0be45393e6ac0a4b58962881c3a
803e093ef0c95fe21c80a9d89d8d2f132f9e5df4
76941 F20110318_AACLZN iniguez_c_Page_068.jpg
d613844415f07bc3ec59f5ba20ce0e17
d7ce2b6bce8383d0e7bda1edc911bf53c1f3b6d3
1597 F20110318_AACLYZ iniguez_c_Page_110.txt
6249dd9d1370a83f8feb3f1c7c45ab1e
e9d908cfaf63b5ad92339842a9423d73d7aaca8a
76228 F20110318_AACMFI iniguez_c_Page_010.pro
913bb4535dafce91559414f2b3f7af40
7f1ea9f26e7e48fcd1495b1edac85e9ed78b3b81
70684 F20110318_AACNIL iniguez_c_Page_075.jpg
da91fca241503ff3aa991bba7761c0fc
e5418207c57d9ec0bab0aac4a1563a62bec14e21
67678 F20110318_AACNHW iniguez_c_Page_049.jpg
797e2ac4e0de71990b3481fcbc2024b3
09cf063731cf2aaa4d742f97f675692d22a83a76
6303 F20110318_AACMEU iniguez_c_Page_028thm.jpg
92fe6b4d753ded7a244d87381211cb0f
84777a1ae294e8b13a19174a497210d844397a6a
F20110318_AACLZO iniguez_c_Page_125.tif
3ec1e37fe54e4b29310dcd114ef43965
e2e9b8ff700e2d5369603948fb47d40da4353e35
5679 F20110318_AACMFJ iniguez_c_Page_120thm.jpg
a830859c1179128a4bfc0702bae5e069
bf8561091fbe1f3a3f5d7d0cdfac43786adfd857
75643 F20110318_AACNJA iniguez_c_Page_110.jpg
9ec64d708b270f774a4f133c695e437d
d278974e6f92c1d45115af9a1dcac94769e8332a
65042 F20110318_AACNHX iniguez_c_Page_051.jpg
188fd4b0e5cb925eb2f7a321f8b69c1b
4fa8961abf2ee766adefe0eba027e2499426b7c3
2106 F20110318_AACMEV iniguez_c_Page_036.txt
b260c9753eba6ba8a9d597d468d43b74
fd8062b6acb4003f598b5c6d64aa81f1bcd344cc
F20110318_AACLZP iniguez_c_Page_062.tif
83b17ee2d9623384a742498e4a29fc16
09f770a29724e5104d0624c53c38ea413e364513
23467 F20110318_AACNJB iniguez_c_Page_110.QC.jpg
0350e86048f867b0ac152e63c4659139
d234a0ebf360b69f03d43d2726dcfdf8a1366621
73487 F20110318_AACNIM iniguez_c_Page_079.jpg
c18377dc073820e3d94258a2a2e83c39
f86bab6f8a7789510805096a65a1343d65d61d4c
77509 F20110318_AACNHY iniguez_c_Page_052.jpg
d450a711ee05adf575fcde03068d5d54
4f3564443fe172d23c5137cd80ebeeec4b1ce306
936772 F20110318_AACMEW iniguez_c_Page_060.jp2
0547a32e2dcc910d3e58625b2f3200e8
ead28d46b4effc6a4ddc5cc7421ee0f7edc7282f
100058 F20110318_AACLZQ iniguez_c_Page_146.jp2
4f37d1b855b321f4fd67a2a4dd4cbe35
8b9cfe8bdbf981ab3465f1108a5a710afb15068a
F20110318_AACMFK iniguez_c_Page_072.tif
3b374eca2ae615207cf47366faedafcb
28f65191210029ce6101a5b54b2563ca6e40073d
74905 F20110318_AACNJC iniguez_c_Page_112.jpg
1d3971b3b79b117e8ae92a51a7ca1a3d
6e8216d161b6f5a797ef584d9a718b6dee09d707
24653 F20110318_AACNIN iniguez_c_Page_080.QC.jpg
65b5527bcddbcc7dc35b6d0725563327
24c58553af6b837b7e00c48c97b45ef23c6e7c1a
24928 F20110318_AACNHZ iniguez_c_Page_052.QC.jpg
b81544391a0c3da10aa7fc6860ad0d25
2524f25355023a70fb4fbc7d12843e40eae8cabd
741492 F20110318_AACMEX iniguez_c_Page_107.jp2
b132f90e14f4774566c9566918ab82bb
53a1a23955839ef57fef2782579f3d43b70ba98a
50454 F20110318_AACLZR iniguez_c_Page_035.pro
8054423852b468bb1fb13bbfb6a1a9cc
735c889babed32bea38f415c7468b1b0492ff42c
52131 F20110318_AACMGA iniguez_c_Page_091.pro
2895107c98764fa8fcf1b14ce6211075
6dedee4580e5411668da0410172d46af4b4f76a5
6028 F20110318_AACMFL iniguez_c_Page_052thm.jpg
a8562c64ea4ff86e7881618fff8b8265
03a22f50064c345ed016cde45b01be991a8ae0c8
53561 F20110318_AACNJD iniguez_c_Page_118.jpg
896c10cb5f074b9ca7ae8577bf22034b
84f07d1067f5ec7db97666a302d1e96a0e94bd88
83667 F20110318_AACNIO iniguez_c_Page_082.jpg
af3ddb5c6a227ebe3feba65ed65e3379
3f81572aea372d84dfef4171f4b8266af0382e02
5654 F20110318_AACMEY iniguez_c_Page_033thm.jpg
7662bf2258a93d3fd53a61a4677c0905
ab3f7c46e1bec4f8a7abf70c4d2600b9290da5a6
81822 F20110318_AACLZS iniguez_c_Page_092.jpg
06a8248435795660decc428773c6ab8f
83079501c877d0ac675dbebc7e11615e3cda9d46
17791 F20110318_AACMGB iniguez_c_Page_006.QC.jpg
3b5fdcfabfad3acaec4c012182be88d4
ba6b6cf215c0b192c665e2a03644013777ab7ab5
65664 F20110318_AACMFM iniguez_c_Page_046.jpg
44b5f099e0cbabc27ea1a09ca041512f
8016d58457ba0da5439a46013bf8a502c7c7aa9a
18997 F20110318_AACNJE iniguez_c_Page_118.QC.jpg
b4216595614b3358f291b776c3fbb2db
1556c70db64c01b1f60867ef54ae129bb7093046
88656 F20110318_AACNIP iniguez_c_Page_083.jpg
259d0cd25066b043d9b5f477e9eab3b4
72a05cb99d658f3ecbb8cf98ccfc2cbc6f705789
78234 F20110318_AACMEZ iniguez_c_Page_125.jpg
5429b8c10b72f9930adffc32fe075d9e
b313e76d8be891666d238e4b95bb65a1efcbf79b
4599 F20110318_AACLZT iniguez_c_Page_101thm.jpg
872db437318bd35bcaf46c4cc4185800
ab5a3c3aa18704e3f1d9b4835408b626c6969407
1958 F20110318_AACMGC iniguez_c_Page_017.txt
872f3b63f220b6040347255179498267
84114309433dd158d95b0fc9f35c500d7e55349f
6353 F20110318_AACMFN iniguez_c_Page_091thm.jpg
4e50a07346f9d22023b5c0c7ab123055
b7b22ff394e4fea293a9d7bb0760bae5c0b34661
18984 F20110318_AACNJF iniguez_c_Page_119.QC.jpg
79cf4005c7b572eb8f77f87c0f714aa5
40b744df25e68b8971aa9def701ab0d58ebd2a9e
27278 F20110318_AACNIQ iniguez_c_Page_084.QC.jpg
3151501bda1a956b216739b3a219108b
0bb259195d02a9085e36f041936f62abe129bb67
35115 F20110318_AACLZU iniguez_c_Page_080.pro
b95723f91f00359b3034e218f765ca2f
2ef666393a29affd4ec397f12a381e5eaac5d7da
50903 F20110318_AACMGD iniguez_c_Page_016.pro
cf04ab90a17b7baba789e2e5a7b6781d
6f60c6ec63b49c9e5abe8c5bcbea224f6f8c99f2
F20110318_AACMFO iniguez_c_Page_027.tif
02081cada852123a3edbd601ede5049e
509d6173c7867bda93505508245f95a9eaff0e78
10650 F20110318_AACNJG iniguez_c_Page_122.QC.jpg
3e645fddd01af3f84415f6afe971ef4b
dd7c7ace2525ee0cf9b4591863816715b31323a9
73755 F20110318_AACNIR iniguez_c_Page_088.jpg
28c11cf6d2983de6ba4f0d40a84a5f17
bd742af2227cccd52b767db3d88c907097857a50
5252 F20110318_AACLZV iniguez_c_Page_129thm.jpg
3281c98b76f377ea9416cda2cbace9dd
f380a8013e75bfd02cf2a6022134c2e49bef0c8d
707 F20110318_AACMGE iniguez_c_Page_114.txt
11e74fcd4a76f846bb81a56bf31b607f
cbd3b951fa93a50e989f08ecc661102d5a43ed35
67776 F20110318_AACMFP iniguez_c_Page_105.jpg
c9f004402b3e691d4c15015552a02712
95a763ae5aa6c4a4230474cd1cd6ffde03e7ec7c
56814 F20110318_AACNJH iniguez_c_Page_126.jpg
c35acb8f25548339b56b1dd2bd65e118
29c65bc08d014da7ffd6886ae49c8ee0e5690ca0
17999 F20110318_AACNIS iniguez_c_Page_099.QC.jpg
393e9d3bd69dbba9a745f186a9d8693e
eee1be15391423a116f60ba8a2b9d84c218675cc
109326 F20110318_AACLZW iniguez_c_Page_063.jp2
bc1f78083a5ac185dfec13d2abce394e
536cc9dfb06fc24c7cab2dc53dd1ef42b78a29ec
85773 F20110318_AACMGF iniguez_c_Page_095.jpg
cca8290eb597cb6f1f618aa432ac1458
8e985458e4e11fbf62a7f4fb011df5715c75d40f
52283 F20110318_AACMFQ iniguez_c_Page_063.pro
62f71ff058ae407adf496cd48facac5e
486665c24aacf6d9c7d878137fb119a9e67b3185
72887 F20110318_AACNJI iniguez_c_Page_128.jpg
34b4b9ac2f7891b4c5b9b4555f3d7398
b8649c3a77598b6426a2b3955a1b11a01531b5d6
19855 F20110318_AACNIT iniguez_c_Page_103.QC.jpg
bf811ae52483524d092f7a12b633bd1a
7286b9bf8f7eedbcdb011d0b649635d354ba60b2
F20110318_AACLZX iniguez_c_Page_122.tif
1387297b543a10cb5e99b1d8c5a4253f
7bf5329f96b0818895d5079ba1aef67c9f9738b0
F20110318_AACMGG iniguez_c_Page_077.tif
eebe62f16677713ad73a27309232e420
650d9e594e011fab733159ea6b86a8a435c36148
F20110318_AACMFR iniguez_c_Page_025.tif
f3a499cdc4debd3589f326fb76cf7b47
e98660d4255e5fdbfdb3c9150a6b740ac9b5ca5f
23736 F20110318_AACNJJ iniguez_c_Page_129.QC.jpg
6dac46079028a9b9b33b421bfb55f295
78d98edb904588e44542ffd70a3ae2590384bfe8
22312 F20110318_AACNIU iniguez_c_Page_105.QC.jpg
4683f4de71e45dd58e3d5cbc6cd0a95b
1bcf134366c4a14bec336c2721cf7dd7c38d2d78
6267 F20110318_AACLZY iniguez_c_Page_112thm.jpg
1e2639b0f335b99aee07266a840d8acc
79f198fcbe953c97171888da78f8297151055979
21624 F20110318_AACMGH iniguez_c_Page_146.QC.jpg
a761a05536dea9d6f086a6eafcf5bc60
b0f05a7753fafab45002ddb20efe1612130c9b96
84332 F20110318_AACMFS iniguez_c_Page_020.jpg
46b22ad48d049bce58c7c0eb6a71e281
9999942134d5803c691ab55e8d788ba65114fdbf
68587 F20110318_AACNJK iniguez_c_Page_130.jpg
6bbe37d23721489c1883f8801bf2a8ad
73d2d2a5d1479f63d15d243312b98a30af29b656
70311 F20110318_AACNIV iniguez_c_Page_106.jpg
a91419d06c0c790db4507c47608b20c8
1560bb1ce3bee5a00ffe0a0fcf3ae33faa486ba1
2083 F20110318_AACLZZ iniguez_c_Page_104.txt
7b835922da8df4668d36e6f12d4c8d7d
8693caf0c6470c6b56d62a324246fba41d536b26
F20110318_AACMGI iniguez_c_Page_014.tif
466ed4ec4bc2970e9f37585bb53de1d9
a4ec1a19b6998c3a57e630b08f3e907d67b1d273
F20110318_AACMFT iniguez_c_Page_023.tif
3cb6643f30a5e0aed5ca9d4f9364b022
f6501e91e1a32d967617ab986c274071ab966e21
18993 F20110318_AACNJL iniguez_c_Page_133.QC.jpg
0c583c6e5124e04efd89af80e89a0750
a26c5582d709014c02e673ff8dd0b0cd48f54f8a
63418 F20110318_AACNIW iniguez_c_Page_107.jpg
86a1fd9a54f4bc891866697f17e9e934
bc1c82db55c6c46e6c2271d9730c657f830fe93d
F20110318_AACMGJ iniguez_c_Page_013.tif
4e9f3a10e53d00886260418e4a55e8db
c9dd133d491a62800327f79ec4de0ec58779b2b8
F20110318_AACMFU iniguez_c_Page_021.tif
31a96386c38c76ecadf1ab45b548226e
56f34c63fb445eeb2693ce2f0c74c8bde0e9eb45
773735 F20110318_AACNKA iniguez_c_Page_038.jp2
5cea8ac48f041a3c29ed74fd08d3747c
396e22346360467a4cf108c9f32f539d715122b2
78540 F20110318_AACNJM iniguez_c_Page_134.jpg
44f76d94706c89f53791e151f5c622a7
73e2f0e44b2abe1f5f801ec774707db68a62ac33
66564 F20110318_AACNIX iniguez_c_Page_108.jpg
ac2d4473fd1fb25b93b7906fff0eba46
3cfb0bbed27a100dda35f961966e1b137b5fc7d9
26398 F20110318_AACMGK iniguez_c_Page_136.pro
725aa36c69e07b195f3fa456caef4637
d01d43cc05abbfd151ef688f7a32eac9fff106a5
24972 F20110318_AACMFV iniguez_c_Page_009.QC.jpg
97c42784a7898fc39d90b9936d64a393
3890da619d151aa01d19c821bf64ce4b3ff517b8
895035 F20110318_AACNKB iniguez_c_Page_039.jp2
bbf389d65d5499f1c081026959b93a16
df38ba54a0d1169a24f42085b3248eba127c341a
21556 F20110318_AACNIY iniguez_c_Page_108.QC.jpg
7be5079aa419b41fa1f32a2612754aae
231393b4041f43911d6a977ceb3d3701a79c7a6c
771762 F20110318_AACMFW iniguez_c_Page_077.jp2
fdd2f5ac40bd6d1997306302097e445b
1127d0691d7edefb834dbf88c383383b286565d4
108172 F20110318_AACNKC iniguez_c_Page_047.jp2
3a18f4f2c20e27998544fe8666327b70
3d6646d1c949600ca829bf5e0390ffb3cf0eb316
79475 F20110318_AACNJN iniguez_c_Page_138.jpg
f4cbd9d5ae4876a76aac3119645078e3
da330a48750117f487b5a6f4b2e1fa2da8e80f33
23671 F20110318_AACNIZ iniguez_c_Page_109.QC.jpg
b67040d7c35190b0b11838c09383b0fe
960ae2dfdbe621520bcc828c6fd384708a1f9217
23493 F20110318_AACMHA iniguez_c_Page_112.QC.jpg
cf649a3eb15c56cf82f7cb81317fc33f
6ec2b9b0396d254e5a80f7138ba11c05cb62e4cb
652 F20110318_AACMGL iniguez_c_Page_097.txt
3532aae18259cb2f351395037e17aa9c
109348beee813e66eb21bde4c8061be88a348f6d
97920 F20110318_AACMFX iniguez_c_Page_045.jp2
b5826e3820006ecd7c3e678d3a1fabb8
5b7dffb0a4bd2648854643eb6da415cfd06c3066
922433 F20110318_AACNKD iniguez_c_Page_048.jp2
9149a6e0edc73c8b714bf6051203e9e8
af31da59b7614be268444e16ca80472594bd8fce
71933 F20110318_AACNJO iniguez_c_Page_139.jpg
f3c365e373d72275af08f411dc02b1ca
8ff2a503dadd66601790cd7bccb7a951bde32a05
6070 F20110318_AACMHB iniguez_c_Page_095thm.jpg
482eede6438c9fc2663015660155f839
78af7d79af79a26f05e5c57eac4790fd79b25bbd
1781 F20110318_AACMGM iniguez_c_Page_013.txt
c767fb4f02b0009fd6eddcc73a8a2124
3d94731bdb20ebc5342c316a16dba0b7b5d10a28
1134 F20110318_AACMFY iniguez_c_Page_038.txt
ae8fe9051d5e9f52e715a8a57df6cfe7
b88098ad3f57002951ed4fdd1e686605c1657ad4
795372 F20110318_AACNKE iniguez_c_Page_050.jp2
e2ff88e6b5bdead8b57a398cae10101e
8036a1b977db108fb730fb14063bf6109a4de83e
99701 F20110318_AACNJP iniguez_c_Page_148.jpg
688cf90aa791a965e6af565178aa92ef
4a9f44d718b190356d67ce2c8e7117b2613bd171
77938 F20110318_AACMHC iniguez_c_Page_130.jp2
152e3d7950ec9e699bc387e4141ab846
6bda8c29a24423d20970e997a0ebb32533254163
110642 F20110318_AACMGN iniguez_c_Page_124.jp2
3342b60492ab5396a8ee6975f7e0c78c
1b0826dbdf54971e66797f0bfef6b37258c3613f
1551 F20110318_AACMFZ iniguez_c_Page_079.txt
cb0e6848f182e39c674874b95e9e42a3
ae04763d917a6b7deb5f91e5b6f2daa752c82767
54044 F20110318_AACNKF iniguez_c_Page_057.jp2
50365dddbb49040d0c68e4074781cdcb
ab7bd0e5581dd6f90d0be22de90bb363cdd3405b
38334 F20110318_AACNJQ iniguez_c_Page_150.jpg
7f9720516997ddb1441e7d0e9c20d134
2b80c99a3f8ce22756a5fb29332dccef47476550
6652 F20110318_AACMHD iniguez_c_Page_111thm.jpg
e03f15765796ee518ff81d3e5c91cdf4
e266aa60191e75785b76d11b696078273da185cc
110458 F20110318_AACMGO iniguez_c_Page_035.jp2
95993dd5f6c2d284af6d5806af29449b
0c66d8eeb3b77e6f8926be01853d89073af47bec
80154 F20110318_AACNKG iniguez_c_Page_058.jp2
2edace74f39a9273d9e3f2041b1a4792
42894baec43b01a6cbd8d179330f92cc07bbfa29
23763 F20110318_AACNJR iniguez_c_Page_001.jp2
c50d8023511c1a4e32e8456d67b83480
4e6b30740ef80a8d3aea201de0b0086e1dd69a4c
6329 F20110318_AACMHE iniguez_c_Page_031thm.jpg
13883c1c189a5454e618a1a6c532dcd6
6cd098aced78d7fc9b5687c0bd4b305bb9f9b754
67883 F20110318_AACMGP iniguez_c_Page_005.jpg
82971c8158b4272e07117a718dd6f63d
151f8d76139a81badbc6566b8ccfb5711156057d
111351 F20110318_AACNKH iniguez_c_Page_064.jp2
c941b31ca3754ed5045f4b833d466fa6
f63adb834a259904cbeb5569426559c211ceff5c
F20110318_AACNJS iniguez_c_Page_005.jp2
8dbeb3107fa9f8c8c1bee8f6a7cdc26c
f90e83877d311cb63f28f3cc26086f67e8544514
1952 F20110318_AACMHF iniguez_c_Page_052.txt
3371c52a7c1ef6cab275b5d0f377c7d3
be230b301d6b9c97a9b3a4914cc7d6893b3ff94c
42947 F20110318_AACMGQ iniguez_c_Page_022.pro
e5cf75fb0c94d8171da0cacdfed11e98
bc18053d32ffc33b9e202ac8b7998ee96bd49186
828099 F20110318_AACNKI iniguez_c_Page_066.jp2
4cd28fd9ba12be12ba077965064f714b
2576f68918729c33b9a5b390fd37a1286a3ac10e
1051977 F20110318_AACNJT iniguez_c_Page_006.jp2
a5b601d216e0a1ac0571ce881b104768
fa305e5f511b723e7d9b7abecaeb0ff1ffe4e211
861950 F20110318_AACMHG iniguez_c_Page_104.jp2
47fd0171a3660670221061a01b7ed4aa
0873c55f95a509cd8dcb58932a6c983fe4d1b6fe
6172 F20110318_AACMGR iniguez_c_Page_074thm.jpg
1c794c8d68b21407017cf191b932e6b6
2cd349dbdb118017b7bfea3834de3274e6b490a8
894588 F20110318_AACNKJ iniguez_c_Page_067.jp2
b26695241d326c69157d8e5cda9d9f2d
496fd28f98eebb7d1e9c52e81cb4737952c643dc
1051869 F20110318_AACNJU iniguez_c_Page_008.jp2
ee9c28749dd8cad7d93dcc864efaa278
a816ff0c2bd773bc0cc6f5c80e67d8bcbefdf668
110248 F20110318_AACMHH iniguez_c_Page_062.jp2
24eb30daeff4f3b806552254f005cca0
63b3ab68600f16a231f3fee37bc856876c473d5b
89400 F20110318_AACMGS iniguez_c_Page_138.jp2
a59660f2d0e83bb8502cbd4530c4c92e
ff3a4ca73f56fac31ecbd57fe7fb7decfed4c0e8
61451 F20110318_AACNKK iniguez_c_Page_069.jp2
760082195a04cd4b553ef4f10e5198d1
e665e3f234724cc99612e1a0559a5363be446461
95762 F20110318_AACNJV iniguez_c_Page_015.jp2
96aeeb56b5b4cc5267f38d56d0246541
6457483d7c45ca367f7ea2120765879674b7239c
69883 F20110318_AACMHI iniguez_c_Page_042.jpg
c23d7e4a82c82ff9dfd40e87de2b717b
4d4895be0aecb8ad812c184d718fbfbf5bf1a455
25723 F20110318_AACMGT iniguez_c_Page_035.QC.jpg
237a026eb8aabe0cd16a5ad49568382a
82467d1cc8c441625185cbf2c33db26434038e7c
761619 F20110318_AACNKL iniguez_c_Page_071.jp2
546ccc1e2cc50be91313ef476891036f
a1b982967bf63e198df6e49b462c1b498666fb62
106351 F20110318_AACNJW iniguez_c_Page_019.jp2
98d883a846a2af7d8bf2e31edf258b56
200a2d7348b97b6982ebb8b8718517e7e2917263
F20110318_AACMHJ iniguez_c_Page_016.tif
05c08a344949841a7a28588eb3346f66
d7d8129cacc7667eec3d8d308c2621104bdfea7f
46936 F20110318_AACMGU iniguez_c_Page_069.jpg
cbd2d3955d8b599287e7b7de19cfc7b8
7d78c88fe86ff5f7f560c741475c8863f5788f16
764981 F20110318_AACNLA iniguez_c_Page_117.jp2
d4a36932ae27d76bd09363ad7274a69f
21c9b27781afc1406e34745cdd59df128e6368f7
877596 F20110318_AACNKM iniguez_c_Page_072.jp2
868292a9b94f1599ff46fac2aa5454f9
b6634a768d35686871646f10c470de887c3fc396
112029 F20110318_AACNJX iniguez_c_Page_028.jp2
73c8a971c041b40f0bde585a680720b2
0f8e7faee053ab47a90be116ca3ba991e2a4413d
F20110318_AACMHK iniguez_c_Page_088.tif
93d5177bd38e1f533a3d373b05e03910
10a36356eefdc145a8a886be94238ddbdc262a86
6034 F20110318_AACMGV iniguez_c_Page_018thm.jpg
0d768f222fd18ce6aea6b60ef71d836a
e9f0580ed801a80d181c7901862a5db08f86d336
778255 F20110318_AACNLB iniguez_c_Page_120.jp2
ca5d3bc0dcf84bde40e0d5ff33a290e1
805f6e0d5348bbc95364c2bf9939fa872abc8e11
110239 F20110318_AACNKN iniguez_c_Page_074.jp2
eb2e2e65681e6e178b855a0703a774d0
9ab750795a6b1089ed0fe0e48f345da17c22fc33
111217 F20110318_AACNJY iniguez_c_Page_031.jp2
03475466e0d08de03c16602616ad0eb1
c287b35bdbaeebe5942f701fe3c9d148febe66ac
788001 F20110318_AACMHL iniguez_c_Page_116.jp2
ef26ac5d21bbaa7c34491db38c5d7397
1bc72e7e0f9fe2d88943db6153c49e7d3a995336
1383 F20110318_AACMGW iniguez_c_Page_087.txt
c79c608df64f2e4cfc98f42c9e98dbbe
271efa0e862a53f04a0f30d41c154ffdf9e25061
415970 F20110318_AACNLC iniguez_c_Page_122.jp2
97899a3b46d70194e18716f117ac9da7
2dced429f2de7316462d306a03d388c512ebd427
828696 F20110318_AACNJZ iniguez_c_Page_034.jp2
4994536d8aa503c38c49ac020937510a
9325d85b53d0a8385286377c7da26d08c8f7e409
23581 F20110318_AACMGX iniguez_c_Page_054.QC.jpg
cfd7194c221852bad19b184fdedffa27
dee7fbf5f27068216aa194db7f70ec2dbbb0c1be
1948 F20110318_AACMIA iniguez_c_Page_086.txt
075fdd86e4ab22e6d5b7f10ee6b8fdb7
8a0e2f018a96d248ee277234d41e74c4ac873b48
71107 F20110318_AACNLD iniguez_c_Page_123.jp2
56b8cdefab85bf5515166c160b22f90a
111c3a908d4e69fc70821de78b5cccd2b110b32b
949331 F20110318_AACNKO iniguez_c_Page_076.jp2
e5761dc084fc7870f0778711513bd705
ffcea2b2a49c064f51ed3d1f3e6bc62979347189
17769 F20110318_AACMHM iniguez_c_Page_149.QC.jpg
2e7ecece35d4fd5b66f20f31635d0768
710fa0b8d0cba6372f3924e5784c32185ea890ee
F20110318_AACMGY iniguez_c_Page_090.tif
d6b070f036ef0d87844d32bcf458d39f
cb37103126e473ffd8f5676dff2b12f8622bc03c
4070 F20110318_AACMIB iniguez_c_Page_003.jp2
f45957badb4f69061d056ab0e09376a0
511e4ce618e95f1024425db9cc999bdcd7b7143a
83261 F20110318_AACNLE iniguez_c_Page_128.jp2
3b8d3098b13e969c1e90b7d1003db115
e326eec26eacb997861b84c6629d75f67dd2eeec
918074 F20110318_AACNKP iniguez_c_Page_079.jp2
826ce421103a73eea4932eb7c4efa9f1
fdf5ac51bb805101d3679c4a939e8e19d0fe31d9
25698 F20110318_AACMHN iniguez_c_Page_047.QC.jpg
ba192749ffc938c46ed7ea57c10bc775
dcc4d0f1fb3e55abf32ccf5fd5f921a1641facf5
1412 F20110318_AACMGZ iniguez_c_Page_055.txt
fbb1818af2e758d01a74f61fe8ed6d19
f8f4e887740c5dcd2704852ed71aab84fafa0f48
99047 F20110318_AACMIC iniguez_c_Page_053.jp2
904296f085f7e010862e2d1c007db840
be79ed0b2e0586d4cc545f37a5d26b6a8274390f
103225 F20110318_AACNLF iniguez_c_Page_129.jp2
882208762eebc7e4fb08d516c8fa9194
4c290fc04aaa0f66f3a766d6f8c37d77477adf65
954870 F20110318_AACNKQ iniguez_c_Page_080.jp2
e1dd143ffc8f744edc812ffe2ac0ec52
367b6efb5e149b4696bef0d495ad04def9d7673f
1925 F20110318_AACMHO iniguez_c_Page_042.txt
822b3ff543e834681f5c4c9868887554
fbe948486dadf882d60e535a7348a2bd846f92f5
22173 F20110318_AACMID iniguez_c_Page_075.QC.jpg
0db459c6afb431ac41392e6b9812e574
0ee0a30426211786d8e170a1c710c0e94fc9034d
70404 F20110318_AACNLG iniguez_c_Page_132.jp2
3fe1551a941002052d85f07924ae85b1
6aaca9675b64327258ca21994949dec3c3e5cb38
1014626 F20110318_AACNKR iniguez_c_Page_082.jp2
2cbc3e8cb8e999c89fb931cd0f64e92c
1ee57fbb50fbd94eeef345ef2b38bdb72ef051b1
F20110318_AACMHP iniguez_c_Page_068.tif
4a46a02dde677b3a708db8315182cd2c
90a465bc75723a8c81680a5815bac635fb5e0c74
1362 F20110318_AACMIE iniguez_c_Page_134.txt
19b1dcc272e9cbcdf9594e315fd8e2a5
62128909f9fb63c474a27ea07daeefe5be4fc850
72208 F20110318_AACNLH iniguez_c_Page_136.jp2
470f83645304ac5208a963fe66757d50
229177d2f8fe82dab9a2f04bb7025f664e2e287b
112832 F20110318_AACNKS iniguez_c_Page_091.jp2
542ddb1445c9474c406fe11a4ff6bb01
7855960b6b18dc3172dde680366c94a157ee388d
19500 F20110318_AACMHQ iniguez_c_Page_127.QC.jpg
f1a587fcb4e226f7509350c5e4e9d32f
c360b32b6f094fa9cb137606f6691a504c7a2f56
24274 F20110318_AACMIF iniguez_c_Page_026.QC.jpg
780573b513d30964704930138c852ace
77443262af87aeaa5ee78fa1fdea2f69af15ef36
74900 F20110318_AACNLI iniguez_c_Page_137.jp2
f174031280ec1da26cc98cfd5058ea14
bf6aee5a6ef5ea0143a371a8b93c116bb454ca30
103690 F20110318_AACNKT iniguez_c_Page_094.jp2
b0b6c8ea3bbf6fd5f712953f8d77c037
a1b3812a26ef48db7de68b03d9ded8dff205ef24
1657 F20110318_AACMHR iniguez_c_Page_106.txt
af8e6b8ee5d989df7ad78d7d82dcfbbd
6294d76b054912f3f0cfd0559d7dbace3c151b69
1914 F20110318_AACMIG iniguez_c_Page_090.txt
b069178e592d04a43e579ee682941627
24ba4f22e1421d0ad2aff5ca763dbc6cd475c715
74575 F20110318_AACNLJ iniguez_c_Page_140.jp2
41ab08fb73b8157c08f428bbde7a53ad
4f9e2a3274977ca43181597cdd01739dbad69efa
111545 F20110318_AACNKU iniguez_c_Page_095.jp2
8939b806bc88792398279424fe71b9d9
33b356945c31fc7ade218848a0bafb60f5af45d1
F20110318_AACMHS iniguez_c_Page_024.tif
9f53d43c558887425dcfa8627acf0d48
13f56d35add7b07136a0f1158e8857a92869c09a
40487 F20110318_AACMIH iniguez_c_Page_072.pro
54a23076ad1875c8ebc590b4d4ea4958
dc7d434d4a26d31466823f2d8c4b1379365250ec
95279 F20110318_AACNLK iniguez_c_Page_142.jp2
ee6b764de859579489d70a61266c1eec
cf957269450e778155ac9a9c711d5015541bbb62
36433 F20110318_AACNKV iniguez_c_Page_097.jp2
82b4ac6319d412b4d4bebccad46e55c4
094479775a1bd16835d24c7da016c7b9f9ad6353
17024 F20110318_AACMHT iniguez_c_Page_108.pro
57938e5f28fd271631caa6ce27af0974
51b3f2eba11086dcd6da80c53d9f49913df9b888
23272 F20110318_AACMII iniguez_c_Page_111.QC.jpg
eee438448547555c89e49dcd17bd642d
1c1721fb452126075a723f1bf9edb94d543f35a5
76719 F20110318_AACNLL iniguez_c_Page_144.jp2
ff89ba6555ede1a2534dbb992e88e415
acdf6fd8974e0a2b8f9f38446b72734a3000db89
956010 F20110318_AACNKW iniguez_c_Page_101.jp2
ab4edc9348dc864c51116666a2dba3dc
b8f2c29ec1b2678bfdd98e25e49cdf1d44396f4c
82923 F20110318_AACMHU iniguez_c_Page_025.jpg
2056db5fae6886f9b67ded31810b4e90
d9d25a9ec83dd1d708877a15ff10874a2a04ca5d
55232 F20110318_AACMIJ iniguez_c_Page_037.jpg
4378120549f70e2f99332358421967cf
373b46c48e6a72cc19a9b9fd179c37683baff135
5384 F20110318_AACNMA iniguez_c_Page_046thm.jpg
d67bdc72060fa30cdf4cddafc1f0bc24
97f3a16318ba4aa083cd0465fff0ab2e56befd9e
131048 F20110318_AACNLM iniguez_c_Page_148.jp2
27f85fe21fe05a19a7a83c20cb7b7800
b965443bf66ecc1beba2d538557d2f5256ec09c4
835414 F20110318_AACNKX iniguez_c_Page_105.jp2
124b2be742970c0595a1aeb1e2d6e46b
5ae6abc1be55102fafa2af285d805650e400cfb0
19416 F20110318_AACMHV iniguez_c_Page_137.QC.jpg
cfdaefb048093205f545c519f8e11a1b
872f327cab782029ac493faf4e63a12ba9c85739
F20110318_AACMIK iniguez_c_Page_042.tif
5d580e99dc4111df85d925126be82639
90a31627db73da26767748af20b33922a42d5201
6501 F20110318_AACNMB iniguez_c_Page_064thm.jpg
a53b157315eb6a071bd18969b7ff82b2
1b5d109eedf68764d2624126991f2763fc3e64d9
80775 F20110318_AACNLN iniguez_c_Page_149.jp2
3f26a32bbc5b25550a91c414ced8f94d
55d966ac02e269ebf0ea9839ff7b3b0683fa75d4
841478 F20110318_AACNKY iniguez_c_Page_106.jp2
10f71db119d49d62740a8486e1cce858
0ba000f8f168dd039e95214edeecc4aad9ad71d5
35057 F20110318_AACMHW iniguez_c_Page_131.pro
61d5066fbbdbfa0d5e2f842080c35754
18bae6bd486d2b30c3d3f23c2e7f9c1d91d9eabd
59928 F20110318_AACMIL iniguez_c_Page_055.jp2
5fcc6a5d4527a667861d403cdf3a6baa
c7ea57586107577592785b28c5da7dfdefda2294
5788 F20110318_AACNMC iniguez_c_Page_067thm.jpg
f0a56e25ece4326416d40e87f798342e
ff1d98f07519c123aa18c1bce6c4da337356a25f
540 F20110318_AACNLO iniguez_c_Page_002thm.jpg
805dcb98417e6c191bc10153ac989927
50706e514baf724e6a433dd62e8b7f2d28d7a14d
882599 F20110318_AACNKZ iniguez_c_Page_113.jp2
26fd75fb566c067852f500b3ec83e5fe
586a54f48a9211cc3e7c103375414a24654fc21c
45090 F20110318_AACMJA iniguez_c_Page_070.pro
0c1df3ee9fe75104356a67d2390db8dc
6bdcdb6a84ae0531a1051b308b804bc00c1c60be
22106 F20110318_AACMIM iniguez_c_Page_131.QC.jpg
37bfc62a205895ec1b92665f24cd0e7d
33fcea440470701a6b3745f4b0ed8dbaefa35a3d
5859 F20110318_AACMHX iniguez_c_Page_002.jp2
3d84a7cbc9d2fcde76a86f0d688b258c
0cb0080507044ebd754e18f9903121da26da1524
5860 F20110318_AACNMD iniguez_c_Page_068thm.jpg
ff0f1f2754df1a85c3b9b0bfc36d229c
0fec3e43f6457895bbf2edd815551754310c5712
25298 F20110318_AACMJB iniguez_c_Page_023.QC.jpg
cd6870a46bbc6068a84f95fcfbb6f3a7
dcc7861b472c0380b5dddad429251980008aad32
F20110318_AACMHY iniguez_c_Page_028.tif
4dba03692a52b2dad7e97ab51af3ca41
c06790b060681821d2b5a4262325a96fc5bb2b3a
5397 F20110318_AACNME iniguez_c_Page_071thm.jpg
68e84ce9ef39ae63d2693692ff1968b3
8ed98a91021d77b45c03d118b2e41ff88fbabbb1
373 F20110318_AACNLP iniguez_c_Page_003thm.jpg
e2fa7810cf7d95b22e76ccf450955a32
49f718beefa2809cf340b8bac4bd4b30d690bccb
86458 F20110318_AACMJC iniguez_c_Page_036.jpg
9bdf8baa107bf31b162874c055b3f053
ae21357ebacdc311e2ec087c8b2f7e0dc4aa816b
1857 F20110318_AACMIN iniguez_c_Page_026.txt
64bf57747d96d9992c93bf74a5c406bf
69b4e9371abf4a3c759a50203b1c004ecd413a6a
25391 F20110318_AACMHZ iniguez_c_Page_092.QC.jpg
57e65bf2f2faf7f4c3fce98871a1b6fe
2e6edcc48392026d308c3853461d86c325336aaf
5764 F20110318_AACNMF iniguez_c_Page_072thm.jpg
f419a8245ef5e4c5b70498dfbbada6f7
adbe514428e2327b7f0a84147c357bff1e5a79cc
1663 F20110318_AACNLQ iniguez_c_Page_007thm.jpg
f74c69ba669c407bec1029df30477fc9
681572253504f78e959a240ab7c17647ae0a0518
3355 F20110318_AACMJD iniguez_c_Page_122thm.jpg
26a16c73b9171a79f5e95d0dd052b35e
c99ab1103e829fad23153891f012d96441babef9
20990 F20110318_AACMIO iniguez_c_Page_013.QC.jpg
8df03ac27c3a967e158914704f5c89dd
b54005f99f14b35f8c3b749d8906e82267f13d99
6138 F20110318_AACNMG iniguez_c_Page_076thm.jpg
9923655ac3b6d2beab7f75734f0fa8ea
72bb59ffb75c67b79d570710e195dbd8d136b56a
6158 F20110318_AACNLR iniguez_c_Page_009thm.jpg
ff140ee3de58d5213e7e3f145d3b718b
777a8737975cc82eab5a5ebbe07478a68fbf1b78
F20110318_AACMJE iniguez_c_Page_055.tif
e18bdd8d51e7067c7d9d7ac04d466fda
50459071e8d2581c4977ea3b4f570c87ad00291d
73888 F20110318_AACMIP iniguez_c_Page_143.jp2
5a7b0d74698fb5d79fdef4135ba8d885
082186b74e9b5a5f4d42c59c88c5ac9d64955e01
5858 F20110318_AACNMH iniguez_c_Page_089thm.jpg
b27deb88ee9554d4284e063cb109bf9c
812f2736beee1752d27fa911802f9b4dd55e4250
6920 F20110318_AACNLS iniguez_c_Page_010thm.jpg
d0efc6efdc8d0bbc9292390bc156a938
4bbc2e675b678c77135f491ff7c8dee3da400913
18895 F20110318_AACMJF iniguez_c_Page_144.QC.jpg
802a6d6b00ecf4f16a8fc13dfaad71e1
560679a9df18df30731437d18592a25ade7c832b
3654 F20110318_AACMIQ iniguez_c_Page_121.txt
e126d4acddfe5a5cca6d326bf38761f2
4bd4835a9d5e97f71da4e96d27b3031beb73d46a
F20110318_AACNMI iniguez_c_Page_090thm.jpg
d8b34a00042db8986d60dc7ab9016f96
7b81a758773edc51e376dfa67dd03ecf4125f62f
7126 F20110318_AACNLT iniguez_c_Page_011thm.jpg
429c138171adb4a771fd627326b3c5fc
5c27bd2c17f11c51c647a9e58e471e3a37121870
101421 F20110318_AACMJG iniguez_c_Page_147.jpg
81169a107ce2a8a330f9bb8299ef28e9
bd374782bef04b172b1f7bb661a5e099dcb07651
1198 F20110318_AACMIR iniguez_c_Page_137.txt
bb7bc12f481b7716ec45b379db14a1dc
b0d0397fde506a026719c691f2d7ab5cdff6bc1d
6515 F20110318_AACNMJ iniguez_c_Page_093thm.jpg
e5cc5ce784dd6cea705fe86dba786872
24260fea2ff6a3798a430dc0ed7ca29149abda11
5458 F20110318_AACNLU iniguez_c_Page_015thm.jpg
776b50361939ffcd8c1caedd4f5cf1ce
e6f6f633cce051f6dcf9178bdb16fdd90072eba2
63198 F20110318_AACMJH iniguez_c_Page_121.pro
d51a3b67a53fad3df73d844f65410b8d
8b9dc5d5174cabf720d3c1879cf1eab8ad751cf5
F20110318_AACMIS iniguez_c_Page_047.tif
e6cab2f41b61084fa952fd4deb808120
f09aec5240d9350267e31b4419893f0c10e82944
2726 F20110318_AACNMK iniguez_c_Page_098thm.jpg
a02858feb84c9c77ec32fdcd0ef084be
a93dbcfe7b4fbfa7403fc1f3ae83b3ac6d0837f2
6004 F20110318_AACNLV iniguez_c_Page_017thm.jpg
b211a1630ac7a1cf77b5594e8e224318
b4ec2245eea1bbb0d6540b248338087cbb5e2daf
86427 F20110318_AACMJI iniguez_c_Page_129.jpg
81aaaf7a31c886616823daf013418532
d245357ae8655cc2bbb209421d47ce4d6d24fa3c
23670 F20110318_AACMIT iniguez_c_Page_082.QC.jpg
250966b5449a0331e123dd58872d7db5
d9a113a459935689193bc382099fd005be480d7c
4606 F20110318_AACNML iniguez_c_Page_099thm.jpg
013d3cfce6d18c5dd767a1355e0fd032
1e96d93e7479dbc6ae72281d28a528edc16abf40
5807 F20110318_AACNLW iniguez_c_Page_027thm.jpg
21b245043b9a6e1a9be320168a9cdccd
8f61b678d977bbfdcc2a92cd45078018c8aa8fec
1448 F20110318_AACMJJ iniguez_c_Page_102.txt
e43a85503b3eba8049cd77a57c4e1669
33bda967b2d240dc2886b48365985ba06e5078e8
49238 F20110318_AACMIU iniguez_c_Page_019.pro
3203581cbec9f74edeb592e18ec5c362
7616f0ba32324f96a038f8633fc63bffd9f0244a
6714 F20110318_AACNNA iniguez_c_Page_148thm.jpg
fcf1003f53b1343bcc6395f378422168
dc43fde80b676aadfe71bd8a06eb12db9908d899
5104 F20110318_AACNMM iniguez_c_Page_100thm.jpg
3ae2269a1e7ea5c5879f59951b5f1877
927fcd3904db25edadb1485ce239e99439234651
6397 F20110318_AACNLX iniguez_c_Page_030thm.jpg
d8e24c62a843ccb41559d3eef5027d42
2ad4872be3086b48674520fcfe18847667359a03
1131 F20110318_AACMJK iniguez_c_Page_141.txt
7d56a5ed31368fd22b074a5a6170fff9
4839272df62ea58071661db7c154a81c8a16367e
F20110318_AACMIV iniguez_c_Page_025thm.jpg
a58991d778c6a3a59c846a462780c138
4906948cff487458c174c4289880569ed69c578b
4251 F20110318_AACNNB iniguez_c_Page_149thm.jpg
a73f22ee58a094e2943f0816c33359c1
457844fe048363ff332262452c01500dd1888c30
4787 F20110318_AACNMN iniguez_c_Page_104thm.jpg
4016428d4140fb278b836652f3964e00
db84a504229fa4c71ca84b97cdebbbb4bf7df9cf
6461 F20110318_AACNLY iniguez_c_Page_037thm.jpg
8a536b5fe1724a60d781363026071cba
424d7e2176223bb8bcf8b8be1d13085f70435a05
84826 F20110318_AACMJL iniguez_c_Page_031.jpg
b6156efe134fd8a2e29be644a3260e5e
a4c53ded1177e1ee25aeb1d473ed43a6e3d0102f
60804 F20110318_AACMIW iniguez_c_Page_145.jpg
de0c5d3d5f5c1d8b5f1ec7011e75107b
7dce0efcb10892fc3e2046b67a8815afeaa002d3
3030 F20110318_AACNNC iniguez_c_Page_150thm.jpg
e950ccbecc10841c0e53dc0e1575dae4
591ff1aa751cc5d2a7bf29c782dfcca9c293518f
6691 F20110318_AACNMO iniguez_c_Page_108thm.jpg
d08586f6318047c52e15b1e26f740b9f
7884afaf41b5161634f62c4b97f8d3e43cc30c2b
6749 F20110318_AACNLZ iniguez_c_Page_039thm.jpg
731c7fa72d8d4e2aa9a1bc0c2d7d2003
bdccbb743cc3102deb46d106ddb7fdb8d6f3c332
1185 F20110318_AACMJM iniguez_c_Page_140.txt
9948db671c11d929a9160cc3c5ceeb9e
edc9fe906a20ff539a857eb941a6036937cd0661
6250 F20110318_AACMIX iniguez_c_Page_085thm.jpg
843014818459fc4b5ae77d8c0d95b322
60bb5b0fe3b7940bfa36803bfb509a5217e84025
F20110318_AACMKA iniguez_c_Page_143.tif
a854a654209e88af039092a8925ca9a9
734c7b269aa64d289ba78e4f5f10144748fd5f90
176578 F20110318_AACNND UFE0011838_00001.mets
84c04858e63da2e4acd416707245e2be
fb6deb60c2e7bc22a6997e1e24fe463fcd26241b
6532 F20110318_AACNMP iniguez_c_Page_110thm.jpg
97025f59e5b9ce49b8fab4a01d40ed84
18309031995880c7c0ad3ea94cc1ec13724e0485
F20110318_AACMJN iniguez_c_Page_091.tif
b252458b8014b2036f148a41c04bb71c
f90481030dd6c570a37ba69c3800413c1acfcf95
5099 F20110318_AACMIY iniguez_c_Page_135thm.jpg
dc7d845035fb8504dd352c80b3fcb1ce
f4fc80fe7a44c5a945aacee7ee6f2bf102618add
F20110318_AACMKB iniguez_c_Page_006.tif
0666a5334b567d13d7da83dce00e0093
cbb1311b11167acb616920e395f1ae28bb5ae8e1
25702 F20110318_AACMIZ iniguez_c_Page_016.QC.jpg
7b7db09d5b25a162cb74dbb729b41145
3834c546511394c8e423eb1cb1177e2102a6a59b
20332 F20110318_AACMKC iniguez_c_Page_046.QC.jpg
944662875d03a0e7ade8e92e447ff530
273a872244d3be2e21fdbf0675b6b049d013e7e3
5656 F20110318_AACNMQ iniguez_c_Page_115thm.jpg
c4899de1b63873c7b8f763dcf1697b7b
11815af012c4d0e8adb68d782fdfd8cfac877cda
F20110318_AACMJO iniguez_c_Page_031.tif
1fc29cf82b4bbee49663d848cbdd9dff
1f8c1f1df1b5fe5e750e91391b6378d092faf3b2
76017 F20110318_AACMKD iniguez_c_Page_135.jpg
08951b04863966d06d665083199e2b54
2e0141168adf72eb71001d0c4de58f24677b2e2a
3987 F20110318_AACNMR iniguez_c_Page_126thm.jpg
9689e232bc34453a610cdb7f3a307095
0cff6f756837975bfd2a0772cb132ec331b6c0fc
94065 F20110318_AACMJP iniguez_c_Page_088.jp2
ef6a71a6889464641565dbfdebf1e67c
96de761edc7eeda762451f983a1a63e830fe8daa
F20110318_AACMKE iniguez_c_Page_119.tif
91f783c75e47ba9fb791e04631cbb167
5696ce24430953a15ee0e7cb336ade6f1b9cd8be
4484 F20110318_AACNMS iniguez_c_Page_128thm.jpg
5bfc42f5d6f1a7ddf7944f65f963d4b8
bf7a460e3b9710d8e45633f24b72df119aa0a168
330678 F20110318_AACMJQ iniguez_c_Page_007.jp2
708200e52f60af9b1879248e8e209c90
ed58e2d8b665302295528b473492f2276dc921a7
F20110318_AACMKF iniguez_c_Page_029.tif
0fd883ade077a5746775f0b4e1c34b1f
05434d2682833bcab35318e5d90a6c2a0e781a98
4919 F20110318_AACNMT iniguez_c_Page_131thm.jpg
7f903b0294069ad8c6fca01bf8e78719
ce483c85e5319442867b757b3d008477c041b94a
24929 F20110318_AACMJR iniguez_c_Page_048.QC.jpg
570d19c204cca1f654cabd16acf3d8bb
6496d1dfb867d9c9481fc9caf4e0337f6de5e5a3
27551 F20110318_AACMKG iniguez_c_Page_106.pro
aeb6304f45925617e00dd90d718885c3
33bd5d4bcf7a454265fd6f17d6185a50b8b917ee
4046 F20110318_AACNMU iniguez_c_Page_132thm.jpg
7c01952842082bd51d1a7e4bb2cadc28
3941011b8759365ffb213d2662d9bc8e1c0e88ea
35618 F20110318_AACMJS iniguez_c_Page_149.pro
d198245ddb35ed7bc0fd89bf626b6649
1c1b510505d4450e3106866034a38751ef7e83d3
27360 F20110318_AACMKH iniguez_c_Page_107.pro
6f4f653495bb154edcba565f8cb8806d
4ec2e833e9eeac58297285388b662233d9248da9
5294 F20110318_AACNMV iniguez_c_Page_138thm.jpg
9f8c1192b655dee46d1a7cc3d3f293f1
8c104875ef4ed5194ce2613684197fcf69a80fa0
92311 F20110318_AACMJT iniguez_c_Page_009.jpg
0eb56f4191f250f36b7a72d6821dbc2e
adc71655f0314c3a931ef47280feeed7be177ff6
14401 F20110318_AACMKI iniguez_c_Page_057.QC.jpg
fb22de8cdd82aae94b40cc82b52280da
24325c64a5b7b1ea00c27f1b7f3cbabc798e96ac
F20110318_AACNMW iniguez_c_Page_140thm.jpg
ef30db4c1b8478231036887fd476cca8
e135ca3a546e1be61b28fae2faea3a4140aa3441
5899 F20110318_AACMJU iniguez_c_Page_075thm.jpg
986022f5f02ff809270121c5396f2ff2
1c0b2cf24962aeb12addc1fbc7ba588de4f9513d
84773 F20110318_AACMKJ iniguez_c_Page_093.jpg
5a4d679d5840348885378178e3f97b98
d4fd02d8af79a4773d90b51732761af15e4be649
4269 F20110318_AACNMX iniguez_c_Page_143thm.jpg
64a431aa613c48130e9ba13f4c640ead
05b4b6b2e22c69f7cc9875df9437d6e7bf3b0d01
F20110318_AACMJV iniguez_c_Page_127.tif
4096b1baa91d5c7d4969158cc464b6dd
b688c0ba5d9b6c8cc213956cc7ac2c94219b0043
24911 F20110318_AACMKK iniguez_c_Page_060.QC.jpg
1dfaa8e08f9e1a8c95c50fd4423edab5
93d0a439af38eb4f77dd4dcd6420a86a074be85a
4472 F20110318_AACNMY iniguez_c_Page_144thm.jpg
1b73f259679fe03704e588a1062c88ac
c69910f6b76cb0ae4e62d6176a4dc56e01c87218
22307 F20110318_AACMJW iniguez_c_Page_113.QC.jpg
869057233bb78cec49296450029c131f
f1c1c6b01e7822f4ea625ca51d74292c8018a42a
6348 F20110318_AACMKL iniguez_c_Page_096thm.jpg
a2c597161474f96613727f2612622793
e0507e2faec09cf4fddcad13385861813c664c28
6371 F20110318_AACNMZ iniguez_c_Page_147thm.jpg
d6bccf20d9018f282007a87b19fd60ca
8705700584ea3a6edb1f11d4b4df7fbe7171cc93
811732 F20110318_AACMJX iniguez_c_Page_121.jp2
2b793f932f9bf25542b7da0b1214c610
f25f458d33bf32e1bfdf3a2e9f7e9f70acd74bfc
5488 F20110318_AACMLA iniguez_c_Page_119thm.jpg
31e9170c5671e0709f99e810649759f1
462f3a4a8d7110bbe47b86af99d6a1250d14bf1e
1780 F20110318_AACMKM iniguez_c_Page_034.txt
fbd58202ce1003ca7835370994e452dc
0031100cf10a95f3f5ef94c7a90c59d2fd319034



PAGE 1

DEMAND SHIFTS IN OUTLET SELECTI ON IN THE UNITED STATES MARKET FOR FRESH FLOWERS By CHRISTIAN R. INIGUEZ A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE UNIVERSITY OF FLORIDA 2005

PAGE 2

Copyright 2005 by Christian R. Iniguez

PAGE 3

To Mami Lety.

PAGE 4

ACKNOWLEDGMENTS This thesis would not have done without the help of Dr. Ronald W. Ward, chair of my supervisory committee. Over these two years Dr. Ward has been an excellent teacher, and inspired researcher, academic counselor and a friend. I would like to thank him for all his support and guidance throughout the duration of this thesis. I would also like to thank Dr. Ramon Espinel for encouraging me to study at the Food and Resource Economics Department. His support and advice during these past two years have guided my academic beliefs. I also would like to thank Dr. Jeffrey Burkhardt, Dr. James Sterns, Dr. Peter Hildebrand and Dr. Gary Fairchild for their approachable way of teaching which, combined with their deep knowledge, helped create an enthusiastic learning experience in the classroom. I would specially like to thank Jessica Herman, our program supervisor, for all her help during my degree program. Finally, I would like to thank my parents Roberto and Teresita, my sister Daniela, and my wife Maria Fernanda. Their support was the only thing that kept me going when things did not go as planned. At the beginning of the program I lost one of the most important person in my life, my grandmother Lety, but her memory and kind words were always there with me. She will always be in my heart. iv

PAGE 5

TABLE OF CONTENTS page ACKNOWLEDGMENTS.................................................................................................iv LIST OF TABLES............................................................................................................vii LIST OF FIGURES.........................................................................................................viii ABSTRACT.....................................................................................................................xiii CHAPTER 1 INTRODUCTION........................................................................................................1 Overview.......................................................................................................................1 Objective.......................................................................................................................4 Hypotheses....................................................................................................................4 Problem Statement........................................................................................................5 Scope.............................................................................................................................5 Types of Stores Selected.......................................................................................6 Nature of the Data.................................................................................................6 2 LITERATURE REVIEW.............................................................................................8 Consumer Demand and Preferences.............................................................................8 Market Share Concepts...............................................................................................11 Censored Models........................................................................................................15 3 US FRESH FLOWER DEMAND..............................................................................19 Indoor Flower Categories...........................................................................................19 Flower Retail Outlets..................................................................................................21 Market Shares by Product Form.................................................................................22 Share Distribution over Time.....................................................................................26 Market Shares across Variables..................................................................................31 4 THEORETICAL MODEL AND MODEL SPECIFICATION..................................37 Consumer Demand for Flowers..................................................................................37 Details on the Data......................................................................................................38 v

PAGE 6

Censored Data Model.................................................................................................38 Tobit Model for Outlet Selection................................................................................44 First-Step Results: Decision to Become a Buyer........................................................46 Second-Step Results: Intensity of Buying..................................................................51 5 MODEL SIMULATIONS..........................................................................................56 Expected Share for the Average Conditions...............................................................57 Expected Outlet Shares by Demographics.................................................................59 Outlet Shares by Household Age Groups............................................................59 Outlet Share over Gender....................................................................................60 Buying Purpose Impact on Outlet Shares............................................................60 Outlet Shares Across Incomes.............................................................................61 Outlet Shares over Flower Forms........................................................................64 Combined Effect of Purpose and Form...............................................................64 Simulations by Seasons.......................................................................................66 Rankings Factors Impacting the Outlet Shares...........................................................70 Dynamics in the Outlet Share Coefficients................................................................71 6 CONCLUSION...........................................................................................................74 Introduction.................................................................................................................74 Overview of Outlet Analyses......................................................................................74 Major Outlet Selection Conclusions...........................................................................77 Limitations..................................................................................................................81 Recommendations.......................................................................................................82 APPENDIX A INDUSTRY OVERVIEW..........................................................................................84 B TIME RECURSIVE COEFFICIENTS.....................................................................101 C TSP CODE................................................................................................................109 LIST OF REFERENCES.................................................................................................132 BIOGRAPHICAL SKETCH...........................................................................................136 vi

PAGE 7

LIST OF TABLES Table page 4-1 Description for the Variables in the Heckman model..............................................47 4-2 Estimated Probit and Tobit Coefficients for Expenditures......................................48 4-3 Estimated Probit and Tobit Coefficients for Transactions.......................................49 vii

PAGE 8

LIST OF FIGURES Figure page 3-1 Percent of household market shares expenditures and transactions on indoor flowers....................................................................................................................20 3-2 Percent of specialty market shares on expenditures and transactions for cut-flowers....................................................................................................................23 3-3 Percent of mass merchandising market shares on expenditures and transactions for cut-flowers........................................................................................................23 3-4 Percent of specialty market shares on expenditures and transactions for flowering/green house plants.................................................................................24 3-5 Percent of mass merchandising market shares on expenditures and transactions for flowering/green house plants...........................................................................24 3-6 Percent of household market shares for florists by product form..........................25 3-7 Percent of household market shares for supermarkets by product form................25 3-8 Percent of yearly market shares in cut-flowers for outlet groups by expenditures and transactions................................................................................27 3-9 Percent of yearly market shares in flowering/green house plants for outlet groups by expenditures and transactions...............................................................28 3-10 Percent of yearly market shares in cut-flowers for florists and supermarkets by expenditures and transactions................................................................................29 3-11 Percent of yearly market shares in flowering/green house plants for florists and supermarkets by expenditures and transactions.....................................................30 3-12 Percent of yearly market shares in arrangements for florists and supermarkets by expenditures and transactions...........................................................................32 3-13 Percent of yearly market shares in non-arrangements for florists and supermarkets by expenditures and transactions.....................................................32 3-14 Percent market shares of cut-flowers expenditures and transactions by demographics.........................................................................................................34 viii

PAGE 9

3-15 Percent market shares of flowering/green house plants expenditures and transactions by demographics................................................................................34 3-16 Percent of cut-flowers market shares expenditures and transactions by outlets....35 3-17 Percent of flowering/green house plants market shares expenditures and transactions by outlets............................................................................................36 4-1 Distribution of values in the first stage probit model............................................46 4-2 Distribution of values in the second stage tobit model for florists........................52 5-1 Distribution of shares.............................................................................................57 5-2 Average outlet shares the fresh flower market for florists and supermarkets........58 5-3 Florist and supermarket probabilities over age......................................................61 5-4 Florist and supermarket probabilities over gender.................................................62 5-5 Florist and supermarket probabilities over purpose...............................................63 5-6 Florist and supermarket probabilities over income................................................63 5-7 Florist and supermarket probabilities over form....................................................65 5-8 Florist probabilities over purpose and form...........................................................66 5-9 Supermarket probabilities over purpose and form.................................................67 5-10 Florist and supermarket probabilities over months................................................68 5-11 Florist probabilities over income and months by expenditures.............................68 5-12 Supermarket probabilities over income and months by expenditures...................69 5-13 Variable rankings for florist...................................................................................72 5-14 Variable rankings for supermarket.........................................................................72 5-15 Time Varying Coefficients for the Average Consumer.........................................73 A-1 Percent of yearly market shares for specialty based in cut flowers by expenditures and transactions................................................................................84 A-2 Percent of yearly market shares for specialty based in flowering/green house plants by expenditures and transactions.................................................................85 ix

PAGE 10

A-3 Percent of yearly market shares for mass merchandising based in cut flowers by expenditures and transactions...........................................................................85 A-4 Percent of yearly market shares for mass merchandising based in flowering/green house plants by expenditures and transactions............................86 A-5 Percent of monthly market shares based in cut flowers by expenditures and transactions............................................................................................................86 A-6 Percent of monthly market shares based in flowering/green house plants by expenditures and transactions................................................................................87 A-7 Percent of monthly specialty market shares based in cut flowers by expenditures and transactions................................................................................87 A-17 Percent of monthly specialty market shares based in flowering/green house plants by expenditures and transactions.................................................................88 A-18 Percent of monthly mass merchandising market shares based in cut flowers by expenditures and transactions................................................................................88 A-19 Percent of monthly mass merchandising market shares based in flowering/green house plants cut flowers by expenditures and transactions.........89 A-20 Percent of monthly market shares in cut flowers for florists and supermarkets by expenditures and transactions...........................................................................89 A-21 Percent of monthly market shares in flowering/green house plants for florists and supermarkets by expenditures and transactions..............................................90 A-22 Percent of monthly market shares in arrangements for florists and supermarkets by expenditures and transactions.....................................................90 A-23 Percent of monthly market shares in non-arrangements for florists and supermarkets by expenditures and transactions.....................................................91 A-24 Distribution of market shares based in cut flowers arrangements by expenditures and transactions................................................................................91 A-25 Distribution of market shares based in cut flowers non-arrangements by expenditures and transactions................................................................................92 A-26 Distribution of market shares based on age by expenditures and transactions......92 A-27 Distribution of market shares based on income by expenditures and transactions............................................................................................................93 A-28 Distribution of market shares based on purpose by expenditures and transactions............................................................................................................93 x

PAGE 11

A-29 Distribution of market shares based on gender by expenditures and transactions............................................................................................................94 A-30 Distribution of market shares for specific outlets in cut flowers based on age (first and second group) by expenditures and transactions....................................94 A-31 Distribution of market shares for specific outlets in flowering/green house plants based on age (first and second group) by expenditures and transactions....95 A-32 Distribution of market shares for specific outlets in cut flowers based on age (third and fourth group) by expenditures and transactions....................................95 A-33 Distribution of market shares for specific outlets in flowering/green house plants based on age (third and fourth group) by expenditures and transactions....96 A-34 Distribution of market shares for specific outlets in cut flowers based on gender by expenditures and transactions...............................................................96 A-35 Distribution of market shares for specific outlets in flowering/green house plants based on gender by expenditures and transactions......................................97 A-36 Distribution of market shares for specific outlets in cut flowers based on income (first and second group) by expenditures and transactions.......................97 A-37 Distribution of market shares for specific outlets in flowering/green house plants based on income (first and second group) by expenditures and transactions............................................................................................................98 A-38 Distribution of market shares for specific outlets in cut flowers based on income (third and fourth group) by expenditures and transactions.......................98 A-39 Distribution of market shares for specific outlets in flowering/green house plants based on income (third and fourth group) by expenditures and transactions............................................................................................................99 A-40 Distribution of market shares for specific outlets in cut flowers based on purpose by expenditures and transactions..............................................................99 A-41 Distribution of market shares for specific outlets in flowering/green house plants based on purpose by expenditures and transactions..................................100 B-1 Time recursive parameters for the under 25 years of age group..........................101 B-2 Time recursive parameters for the 25 to 39 years of age group...........................101 B-3 Time recursive parameters for the 40 to 54 years of age group...........................102 B-4 Time recursive parameters for the 55 and more years of age group....................102 xi

PAGE 12

B-5 Time recursive parameters for females................................................................103 B-6 Time recursive parameters for males...................................................................103 B-7 Time recursive parameters for gift.......................................................................104 B-8 Time recursive parameters for self......................................................................104 B-9 Time recursive parameters for the under $25,000 income group........................105 B-10 Time recursive parameters for the $25,000 to $49,999 income group................105 B-11 Time recursive parameters for the $50,000 to $74,999 income group................106 B-12 Time recursive parameters for the $75,000 and more income group..................106 B-13 Time recursive parameters for arrangements.......................................................107 B-14 Time recursive parameters for non-arrangements...............................................107 B-15 Time recursive parameters for flowering/green house plants..............................108 xii

PAGE 13

Abstract of Thesis Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Master of Science DEMAND SHIFTS IN OUTLET SELECTION IN THE UNITED STATES MARKET FOR FRESH FLOWERS By Christian R. Iniguez August 2005 Chair: Dr. Ronald W. Ward Major Department: Food and Resource Economics Over the past decade consumers outlet selection in the fresh flower industry has shifted among outlet types, and primarily with changes in using florists and supermarkets. This study focuses on demand shifts in outlet selection in the fresh flower industry at the retail level. Florists and supermarkets were chosen because of their high market share levels seen throughout the industry for these two outlet types. Historically, florists have attained higher levels of market shares expenditures justified in part by the creative value added to their products, particularly in the arrangement sector of the industry. In contrast, supermarkets have focused on the non-arrangement sector which has higher levels of market share transactions. Over time, supermarkets have increased their market shares while florists have experienced a decline. If this trend continues, the industry will experience significant restructuring at the retail level. To quantitatively measure these outlet shares, an initial sample of households of approximately 189,000 observations was collected by a professional survey agency. Over a period ten years the agency recorded xiii

PAGE 14

the purchasing pattern of approximately 9,000 households using consumer diaries. This paper used socioeconomic and demographic variables to describe outlet selection in the fresh flower industry. A two-stage estimation model was used to describe the decision process faced by potential flower consumers with the first stage estimating the entry of buyers and the second the level of purchases among buyers. Simulations were used to forecast consumers intensity of buying. It is the goal of this study to facilitate the understanding of factors influencing outlet selection when purchasing fresh flowers. This knowledge then is important to understanding structural change and for designing policies that could alter the structure if needed. xiv

PAGE 15

CHAPTER 1 INTRODUCTION Overview The present research paper addresses issues of consumers outlet selection in the fresh flower industry. Flowers fall within the category of goods that are usually purchased for their perceived aesthetic value. Aesthetic characteristics include decorative, emotional, environmental, and visual needs of the customer at any given time. For several years now the flower industry has been experiencing major structural changes. Some of this change may be attributed to outlet changes, import dependency, direct marketing practices, packaging, and promotions. This study focuses on outlet changes and how they may affect market shares distribution in the fresh flower industry. Specifically, changes in florists and supermarkets shares are of particular interest. Many factors influence consumer preferences for using a particular outlet type. Price differences usually reflect the type of product that consumers buy with high value products usually associated with florists. Florists are usually able to charge a higher price because of the creative valued added to the flowers (i.e., arrangements). Supermarkets, which usually sell lower priced products, specialize in selling larger quantities of fresh flowers with less value added to the flowers. Up to a certain extent, this change in product preference mix is changing the industrys market structure. Current demand for cut-flowers among different value added products goes beyond pricing considerations. Packaging, purpose, occasions, and product range also influence consumer buying decisions. Historically, supermarkets and florists have differed in the 1

PAGE 16

2 range of flower products and services provided. As we may see later in this study, florists have experienced a decline in their market share, thus suggesting underlying preference changes or alternative sources for the same products. In contrast, supermarkets have experienced increases both in transactions and expenditures on fresh flowers. Changing market shares have significant implications to the flower industry in terms of the competitive buying structure and the types and volumes of purchases by a single outlet. Historically, florists are quite small and have minimal buying power. Whereas, considerable concentration among the major retail food chains suggest that more pressure from the buyers could occur. Yet, growth through supermarkets has the potential for widely expanding the exposure to fresh flowers simply because of the high traffic through most grocery stores. Buyer pressure versus expanded consumer exposure to fresh flowers is a real potential tradeoff that must be considered. Ultimately, for both of these outlets one must establish the degree of share change and measure what is driving the changes. Retail outlets can easily be grouped into four categories defined as: specialty shops, mass merchandisers, internet, and others. Specialty shops are further divided in florists and in others subcategories. Florists have historically had a major share of the arrangement section of the market, providing value added to the retail flowers. Their business is highly seasonal and influenced by calendar occasions such as Mothers Day, Valentines, and Christmas, etc. Florists have experienced declining market shares over the past decade, with supermarkets capturing most of this loss. It is particularly important to know whether this decrease in overall sales has been the consequences of major supply changes throughout the vertical market system (e.g. higher costs, low prices).

PAGE 17

3 Mass merchandisers stores are also subcategorized in supermarkets and others. Supermarkets have focused mainly in the non-arrangement section of the market providing low cost products for a larger share of the consumers base. Their main business are repetitive purchases for non-calendar occasions although they also experience high peak sales in the above mentioned calendar occasions. Overall, supermarkets have increased their share of the market relatively to florists. Many factors could be affecting this trend in the industry, including better inventory practices (economies of scale applied on highly perishable goods such as flowers), a lower cost structure, and a bigger target audience. Other reasons that may influence consumers preferences are purpose and convenience. For the present study purpose is divided into two categories; self and gift. Self often comprises fresh flowers with little value added characteristics, such as non-arrangements. Gifts are mostly purchased for special or calendar occasions throughout the years. More recently, internet sales have grown as a visible outlet for fresh flowers. Internet sales are probably more convenient for the average customer. It might be particularly important to determine whether the appearance of these sales undermine florists or supermarkets market share in the long run. Note, however, that internet sales still usually require delivery of a perishable product that normally requires local services and particularly florists. Hence, there may be both competition and complementarities among some of the outlets. In order to avoid data duplication, internet sales are catalogued as such if they take place on internet retail stores. Internet purchases have been steadily growing in the past years; however, current data are not sufficient to be included extensively in the present study.

PAGE 18

4 Objective This study focuses on expenditures levels and number of transactions at the retail level in the fresh flower industry using quarterly data over the 1992 through 2004 period. In general terms, florists consistently have a larger share of household expenditures relative to supermarkets and other retail outlets. Whereas, supermarkets account for a larger share of household transactions on fresh flowers. The trend shows that florists are loosing market share to mass merchandising. A primary objective of this study to evaluate market share changes in the industry based on household demographic variables, purpose, flower forms, and occasions. Hypotheses Demand for flowers and plants as ornamentals or as environmental investments, as well as, the emotional needs should depend on the household discretionary incomes. That is, rising incomes should have a positive effect on demand. Since flower purchases are somewhat discretionary, sales of floral products may be more responsive to income changes than more essential goods such as food. However, the expenditure response to income is expected to reach a point where continual higher incomes would generate marginal declining response rates. Identifying that particular level would be of particular importance when developing marketing strategies in the flower industry. It is expected that upon completion of this study the reader will have a better understanding of the current and future changes in the outlet structure for the U.S. fresh flower industry at the retail level. From preliminary data and current trends it is expected that the increase in expenditures levels and transactions can be explained by identifying demographics changes, reasons, and product offerings. Beyond those variables expected to influence the share changes, it is possible that linkages between the shares and the

PAGE 19

5 outlet identifiable variables have also changed. For example, has income become more or less important as an outlet demand driver? Hence, a major hypothesis is that beyond the measurable variables driving share changes, there is an underlying shift in the coefficients linking the shares to the causal variables. In this case, we test if the variables are time varying. Problem Statement Market share changes at the retail level in the flower industry can be illustrated by determining the probability of selecting these outlets based on household expenditures or transactions. For the present study we assume a relationship between the characteristics of the buyer, the product, and the reasons for purchasing among other variables. Demographics are measured with income levels, gender, and buyer age. Econometric analysis will capture the importance of each of the variables included in the market share models. By definition all outlets shares must sum to one if the list of outlet selection is exhaustive. Yet if one looks at a subset of outlets such as florists or supermarkets, the share models can possibly be considered separately. Furthermore, depending on the subclasses for expressing the shares, it is feasible that within some combination of subclasses that an outlet shares is zero or even one-hundred percent. That is, the shares may be censored from above and below. Thus the problem becomes one of measuring market share and their drivers (i.e., causal variables) while dealing with the censored values. We will see that this is a classic doubled-censored Tobit model where market shares are estimated while dealing with these upper and lower limits to these shares. Scope Reports are compiled from information reported by a panel of around 9,000 nationally representative households who maintain purchasing diaries (Ipsos-NPD).

PAGE 20

6 Ipsos is one of the worlds leading market research organizations with one of their major products being the collection of household purchasing data. Specifically, the household purchasing data for flowers is from Ipsos-NPD and organized and funded by the American Flower Endowment through an ongoing consumer tracking study. Types of Stores Selected In this study, flowers are grouped into three categories: cut-flowers (arrangements and non-arrangements), flowering and green house plants (flowering plants and foliage), and dry/artificial. The retail stores are also grouped into categories such as: specialty, mass merchandisers, florists shops, supermarkets, warehouses/price clubs, internet retailers, and others. According to Ward (2003), most specialty sales are from traditional florists while supermarkets account for most of the mass merchandising sales. The demand for floral products, and especially cut-flowers, is highly seasonal. Sales are normally highest from February through May and drop precipitously in the fall. Sales of cut-flowers peak during holidays such as Valentines Day and Mothers Day. Cut-flowers and foliage plants, however, are increasingly popular throughout the year as indoor home and workplace decorations.1Aspects of the models will measure these seasonal effects. Nature of the Data The data collecting were funded by American Flower Endowment.2 The database as organized for this thesis includes 82,232 observations at the household level and only comprises non-commercial purchases. The observations were the result of a professional 1 Economic Research Service. United States Department of Agriculture. The Economics of Food, Farming, Natural Resources, and Rural America. Floriculture Crops: Background. http://www.ers.usda.gov/Briefing/floriculture/Background.htm 2 The American Flower Endowment (AFE) is the leading not-for-profit, non-governmental source for floriculture/environmental horticulture research and development funding in the US. For more information see: http://www.endowment.org/pressrelease/general/spcrpt2001.htm

PAGE 21

7 firm investigates in a method called waves. Within each wave, information is compiled every two week period. Consumers were given a very detailed consumer diary and they reported their purchasing habits. The wave dairies can be matched to specific months and years, thus giving a continuous time series of data. One important consideration is that the information filled in the diaries is not a recall but the actual purchases made by the household consumer. Approximately 9,000 demographically balanced households are included in the survey. From the database major divisions can be obtained: we know the population, the number of households, the amount spent, number of transactions or making a purchasing event, and the quantity of the flowers bought. Note that the quantities have less meaning because of the diversity of product purchases such as a bunch, arrangement or single stem. Although the average of weeks that a household remains in the program is about 3 weeks some studies by Ward (2003) have shown that there is no negative influence to the integrity of the data. From the database we can calculate market penetration (buyers over households), and frequency (transactions over buyers). Two major limitations of the database are that it leaves out commercial purchases and only the retail level of the vertical market system is presented. The present study demonstrates and assesses the impact of marketing strategies, spending levels, and consumer behavior through the use of econometric simulations.

PAGE 22

CHAPTER 2 LITERATURE REVIEW The three sections included in this chapter will provide the reader with a basic understanding of the major topics covered in this study. The first section introduces consumer choice theory and the effect of preferences in modeling demand. Included in this section are concepts like utility maximization, product acceptability, and the decision making process faced by consumers. Market share analysis concepts and alternative methodologies are discussed in the second section. This section also discusses model constraints, variable specification, and data aggregation considerations. Finally, the third section presents econometric models used for censored data. Particular emphasis is given to models that use a two-step decision process approach to estimate future purchases. Consumer Demand and Preferences It is widely accepted that marketing effort can be more successful if it is based on knowledge regarding consumer preferences. Taking an economics point of view, Rhodes (1955) presents a general approach to the preference determination problem in consumer demand. According to him, preference is manifested if the consumer chooses the most desirable product available to him. However, this presents a problem to the researcher when the most desirable product is not consumed. It is then important to recognize the difference between product preference and product acceptability or actual purchases. He concludes distinguishing that while acceptance of products among consumers is an absolute definition preference over products is often hard to record. The problem then 8

PAGE 23

9 becomes one of measuring consumers preferences and finding appropriative methodologies that capture the effect on actual consumption. Traditional economics depicts the consumer as a logical or rational thinker that maximizes his utility function based on a given budget constraint. Following that line, Basmann (1956) formulates a theory of demand linking preferences changes and ordinal utility functions faced by the consumer. However, Hollbrook and Hirschman (1982) argue that the study of consumer behavior should look beyond the information processing model of the logical thinker. For him, consumer behavior is also influenced by what he called experiential views which incorporates aesthetic values, enjoyment, sensory pleasures, and emotional responses. Acknowledging such influence on consumers could help understand the demand for products that are usually purchased for their aesthetic and emotional attributes. It is widely accepted in consumer theory that the analysis of product attributes linked to expected preferences allows a better forecast of the consumers future choices. In their article, Blin and Dodson (1980) analyze the underlying relationships between attributes, preferences, and choice. In addition, they present two traditional marketing theories to model consumer choice. First, he describes the multi-attribute expectancy value model in which attributes are identified, measured, and evaluated based on stated preference. Second, he presents the stochastic choice model which tries to explain the complexity of the choice process by using consumer panel diaries to record the purchasing pattern of consumers over time. One of the shortcomings of the multi-attribute expectancy model is that it assumes that the consumer will always choose its preferred brand and that may not always be the case. Furthermore, the model follows a

PAGE 24

10 one-time purchase approach when in fact most questions in consumer theory are oriented to the frequency of purchases. The strength of the stochastic model relies on the ability to predict a quantifiable likelihood of choice. However, one of the drawbacks of the model appears when no product has a higher probability over the other in which case the model becomes one of attribute differentiation. Frequency of buying sometimes influences the decision making process of consumers. In his article, Hoyer (1984) states in that there is a variance associated with estimating consumer choices over time that could be attributed to the intensity of buying. Furthermore, he argues that repetitive purchases may reflect not optimal but satisfactory purchasing decisions in an attempt to minimize consumers effort and time. Contrasting the traditional view that assumes that an evaluation is done each time the consumer makes a choice, he indicates that an evaluation may occur after the product is purchased. If the evaluation is satisfactory then it will guide future consumption if not then more refined choices are made by consumers. For products that experience repetitive purchases throughout the year the issue becomes one of recognizing between product loyalty and habitual purchases. Brand loyalty purchases usually reflect strong reasons for buying while habitual purchases are generally done for convenience. Gilboa and Schmeidler (1997) argue that consumers whose income greatly exceeds the cost of the product will less likely follow a budget constraint. He based his remarks on the high level of expenditures seen in higher income groups in the consumption of non-essential products. In addition, he proposed that repetitive purchases denote small choices where the consumer can afford not to calculate how much they will have left after the purchase.

PAGE 25

11 This would seem to explain why the difference in the demand for nonessential products can only by partly explained by income disparities. Finally, Feinberg et al. (1992) presents the long term market share implications of changes in variety-seeking consumers. He argues that variety seeking is an important determinant of consumer behavior and should be accounted for in consumer choice models. Regarding the choices faced by consumers Walsh (1995) argues that the consumer chooses the alternative most appropriate to him using a cost-benefit ratio. He concludes that seasonal sales for products that maintain the same attributes or meaningless differentiation over time can be partly be explained by the occasion of buying. Carpenter et al. (1994) state that consumers apparently value these differentiating attributes even though they are irrelevant. According to him, meaningless product differentiation can stimulate demand by changing consumers preferences in the long run. Market Share Concepts Bothwell et al. (1984) argue that since economics is a non-experimental science, restrictions should be imposed in models that generate observations. However, his main concerns are the validity of the restrictions and the statistical models used to analyze data. It is therefore the role of the research to avoid uncertainty in the model specification since it could lead to inconsistencies in the estimated parameters of primary interest. In order to do this, meaningful variables should be selected to conduct verifiable empirical research that yields meaningful results. Clodius and Miller (1961) provide and interesting framework for understanding market structure in agricultural products. In his study, he points out several topics ranging from problems in hypotheses testing, to what the end goal of market shares studies should be and concluding with some shortcomings of

PAGE 26

12 market share theory. He points out that the degree of product homogeneity should be considered when choosing the theoretical framework of the study. At this point it is necessary to define the subject matter of this study by understanding the difference between industry and market. Two of the clearest concepts of industry and market were provided by Smith and Dahl (1965), according to them industry is usually defined, in practice, as a group of firms that produce a like output using similar inputs and production processes. A market, on the other hand, involves two groups of firms buyers and sellers representing the forces of supply and demand in a state of interaction (p. 466). His paper also provides interesting concepts of the assumption of perfect competition, the effect of technical innovation and capital accumulation. Ghosh (1966) also provides definitions for both industry and market preferring the latter for empirical studies. He argues that that since market studies are more comprehensive in nature since they comprise both the buyer and seller side. In this study there are some restrictions imposed in the parameters of the regression model. Whenever that is the case the validity of the models estimations could be compromised by the restrictions imposed. Since models are evaluated in terms of the parameter validity and predictive accuracy, it is important to determine if the restrictions imposed in the model compromise outlet market shares estimations. However, Ghosh et al. (1984) argue that constraining parameters values improves the predictive performance of market share models. He uses the functional form, the error distribution assumption, and the parameters description to compare the performance of market shares models.

PAGE 27

13 The most used market share modeling approaches are the linear additive, multiplicative and the attraction models. The functional form of the linear additive model to estimate outlet market is given by MSjt = j + (2-1) jtjktktkjkX where MS denotes market shares, is a vector of the parameters of the regression, X is a vector of explanatory variables, and is the error associated with the regression. The attraction model is based on Kotlers market share theorem where the market share of a firm is given by the firms marketing effort divided by the marketing effort of the rest of the firms in the market (Kotler 1984). Using the profit impact of market strategies (PIMS) database over a period of nine years, Buzzel et al. (1981) utilized a cross-sectional regression analysis to prove the consistency of the attraction model over the linear additive model. He argues that linear additive models generally have limited data, defining the product of study requires too much effort, and suggestions are often not relevant for managers. Naert and Bultez (1973) also criticize the linear additive model on the grounds that it assumes that shares fall between zero and one and sum to one. They conclude that for a market share function to be logically consistent the functional form should be non-linear. However, estimating parameters in non-linear functions is less straightforward than in linear case. More importantly, the statistical properties of non-linear parameters are generally weaker than linear parameters and thus the predictions are not necessarily better. Furthermore, as stated by Ghosh et al. (1984) if we consider both parameter validity and forecast accuracy, linear and multiplicative models performs at least as well as the attraction model (p. 208). For him, the market share model

PAGE 28

14 specification should depend on the purpose of the analysis and the type of data available to the researcher. The nature of the data also presents issues to the validity of the model estimates. Moriarty (1975) criticizes large databases because the variance across different products and regions is usually lost. He argues that disaggregated models offer particular advantages in policy formulation as they are able to target specific groups of interest. Any source of variation from pooled data can then be eliminated by the dummy variable technique. Grover and Srinivasan (1989) also state that data aggregation can comprise the integrity of the data. Beyond a mere critique to the aggregation problem they propose a different approach to the problem by dividing the data into within-switching and brand-loyal segments. In doing so, they attempt to capture some of the variation in the model while allowing some degree of data aggregation. In addition, they assume that a market segment is a group of homogeneous consumers with equal probabilities of selecting the goods of a product category. Two possible shortcomings in their analysis are that they assume that the size of the segments remains the same for all periods and that homogenous consumers have homogenous utility functions. Their assumptions however may not always hold in panel data thus comprising their initial argument against data aggregation. Regarding panel data, Ahl (1970) acknowledges the use consumer diaries to record the cumulative growth in product class volume, as well as, the rate of repetitive purchases. He stresses the importance that the sample obtained from consumer diaries should be demographically balanced, demographic variables should capture consumption differences, and seasonal patterns properly acknowledged. He concludes stating that predictions based on this sampling method have proven to be highly accurate even in

PAGE 29

15 products that show seasonality patterns. This is generally the case unless major upsets are experienced in the market. Chauvin and Hirschey (1997) argue that high market shares do not appear to be a clear advantage for a firms ability to expand in the future successfully. Chauvin challenge the belief that bigger firms are more profitable by saying that high market shares do not necessarily give rise to Ricardian rents. Furthermore, he points out that market shares is simply a measure of the size distribution of competitors, and a useful dimension of the competitive environment faced by the firm and does not influence profit levels (p. 248). A similar argument was presented by Bradburd and Ross (1989) in which they state that smaller firms may be able to find niches from which they can diminish or reverse the profit advantage of larger firms (p. 258). In other words, in market niches oriented firms the performance (and service) usually equals or exceeds that of larger firms. Large firms generally exploit economies of scale by offering large quantities of products that require a small degree of specialization. To the contrary, small firms concentrate in sectors where customer support and one-to-one service is important. Finally, both authors agree that the intensity of research, development and product promotion are among the few factors that influence market shares levels in the long run. Censored Models When the observations follow a cumulative logistic function with zero and one hundred percent probabilities the data could be censored in order to calculate more accurately future demand expectations. As described by Chay and Powell (2001) a regression model is censored when the recorded data on the dependent variable cuts off a certain range with multiple observations at the end points of that range (p. 29). Tobin (1958) was among the first ones to recognize the censoring problem by taking into

PAGE 30

16 account the concentration of observations on the limiting points when trying to estimate the effect of several variables on the limited dependant variable of the relationship. Tobin argued that one should not discard the limiting values of data in order to fit a multiple regression model. Instead, he suggested incorporating such values in a model which has the characteristics of a probit and a multiple regression model. Such a model could be used, particularly on consumer purchases data, when one can not incorporate into the model the probability of events if the event does in fact occur. Tobins analysis assumes that the decision to consume is the same as how much of the good to consume (Haines et al., 1988). However, this may not always be the case for several consumer goods. Furthermore, Tobin also assumes that when corner solutions are present, changes in prices and income can make such solutions disappear. Cragg (1971) analyses in more dept the implications on censoring data and its effect in the limited dependable variable. He validates Tobins arguments about multiple occurrence of the dependent variable on regression models. However, Cragg makes a distinction in studying consumer behavior when no purchase is made by the consumer. He proposed an alternative model to simulate the two-step decision process that consumers typically face when buying goods. In his double hurdle model, Cragg use a probit model to calculate the probability of the event take place (e.g. the decision to purchase the good) and then a standard regression model estimates the magnitude of the change (e.g. how much of the good to purchase). The strength of Craggs model relies on the truncation of the probabilities of the values while accounting for the values that were closer to zero or hundred percent. In fact, Craggs the two-step decision model has been widely used by economists to estimate demand for agricultural commodities.

PAGE 31

17 Blisard and Blaylock (1993) used the two-step decision process as a market participatory model to estimate the demand for butter. The article shows the distinction between households that never consume and those who consume butter infrequently. The article proposes a purchase infrequency model because the butter unlike most agricultural commodities has storage capabilities. Being that flowers are a highly perishable product this model was not considered in the present study. In a previous article, Blisard et al. (1992) used the double hurdle model on cigarette consumption to test the validity of Tobins corner solutions. By using a set of demographic variables to show how low income womens consumption of cigarettes was affected, he was able to conclude that change in income and prices may not necessarily have a proportional change in consumption. Thus showing that consumer preference structure is not homogeneous and that they respond to different utility maximization functions which model their purchasing behavior. Haines et al. (1988) also compared the Tobit model proposed by Tobin and Craggs double hurdle as analytical models to estimate the dietary needs of approximately 15,000 households over one year. He concluded that the Tobit model underestimated consumption responses. Gould (1992) reached the same conclusions when modeling the purchase frequency of cheese. Heckman (1979) stated that Craggs model can suffer from a sample selection bias as a specification error. According to Heckman (1979), the bias that results from using non-randomly selected samples to estimate behavioral relationships is see to arise from the ordinary problem of omitted variables (p. 155). Because of the bias he proposed an alternative model which links the two-step decision by introducing the Inverse Mills Ratio calculated in the first stage as a regressor in the second stage. The Inverse Mills

PAGE 32

18 ratio is a decreasing function of the probability that an observation is selected into the sample (Heckman 1979). Byrne et al. (1996) used Heckmans model to model the two-step decision process for consumption of food-away-from-home. In his article, he used demographic variables such as education, age, and ethnicity to account for the consumer preferences on actual purchases. They stressed the importance of demographic factors to exploit the marketing potential in the consumer-driven food industry. Also, by understanding consumer trends one may more accurately forecast future household demand. Chay and Powell (2001) argue that even though a censored sample can compromise the integrity of the regressors, that is efficiency is lost, the model still yields consistent results. Amemiya (1973) provides a thorough explanation of truncation particularly to the left of zero. He proposes a different model that Heckman in which all observations are considered for the second step. In all, the strength of Heckmans model relies on differentiating between the propensity to consume and the quantity demanded among existing consumers linked through the Inverse Mills Ratio. Since non-consumers have no influence on demand they should be accounted out of the model regression but properly accounted for in order to avoid bias in the estimation.

PAGE 33

CHAPTER 3 US FRESH FLOWER DEMAND This chapter is primarily oriented to understand the composition of the sample used in this study, as well as, to understand the underlying reasons behind the decision to focus on florists and supermarkets as the subject of study. The chapter presents an overview of the US fresh flower demand divided in to cut-flower and flowering/green house plants market share distributions. Expenditure and transaction levels were considered as measurements to compare market share changes among both fresh flower categories. Later on, the chapter covers the relative change in florists and supermarkets market shares over the time. In addition, to understand the distribution of the sample market shares are presented in terms of the demographic and socioeconomic variables described in Chapter 1. Finally, florists and supermarkets market shares of cut-flowers and flowering/green house plants are compared to the rest of the outlets in the specialty and mass merchandising categories, as well as, to internet retail and others. Indoor Flower Categories Data on U.S. fresh flower consumption from 1992:7 to 2004:4 were obtained from the American Floral Endowment (AFE) and Ipsos-NPD group. The data were obtained from consumer diaries of approximately 9,000 demographically balanced households that recorded their flower purchases every two weeks. Indoor flowers were grouped into three subcategories: cut-flower, flowering and green house plants, and dry and artificial. In addition, fresh flowers were grouped into four main types of retail outlet stores: specialty, mass merchandising, internet retail, and others. Furthermore, specialty was divided in 19

PAGE 34

20 florists and others while mass merchandising was divided in supermarkets, warehouse/price club stores, and others. Figure 3-1 shows the market share distribution for indoor flowers by expenditures and transactions. The graph shows that fresh flowers comprise more than 80 percent of the indoor flower market. More specifically, cut-flowers accounted for 57.4 percent of total household expenditures and 44.7 percent transactions; flowering/green house plants accounted for 29.5 percent in terms of expenditures and 37.3 percent in terms of transactions; and finally dry/artificial accounted for 13.2 percent in expenditures and 18.0 percent in transactions. Cut-flowers and flowering/green house plants were separated throughout the chapter see the relative difference in the distribution of market shares over outlet categories. Cut Flowers57.4%Flwg/Green House Plan29.5%Dry and Artificial13.2% Cut Flowers44.7%Flwg/Green House Plan37.3%Dry and Artificial18.0%ExpendituresTransactions PlantsPlants Figure 3-1 Percent of household market shares expenditures and transactions on indoor flowers. Source: AFE and Ipsos-NPD group.

PAGE 35

21 Flower Retail Outlets Figure 3-2 presents the distribution across different outlet groups. Specialty and mass merchandising accounted for over 90 percent of cut-flower market shares in terms of expenditures and transactions during the 1992 to 2004 period. For expenditures, specialty accounted for 68.7 percent of market shares followed by mass merchandising with 25.8 percent. For the transactions, however, mass merchandising with 53.6 percent of the market showed higher market shares than florists which accounted for 41.3 percent. Within the cut-flower category, florists accounted for the largest component with more than 80 percent in both expenditures and transactions. Alternatively, other outlets in the specialty category accounted for 12.2 percent in expenditures and 21.9 percent in transactions. Figure 3-3 presents the distribution across different outlet groups focusing on mass merchandising. The outlets in mass merchandising show similar market share levels in expenditures and transactions. Clearly, supermarkets compromise most of the mass merchandising outlets with approximately 80 percent of the market followed by warehouses/price club outlets with 5 percent and other outlets with close to 12 percent. Both Figure 3-2 and 3-3 provide insight as to the relative importance of using in florists and supermarkets changes as proxies to forecast changes in specialty and mass merchandising outlet groups respectively. Differences in the two measurements are apparent with the specialty group accounting for the majority of expenditures shares and mass merchandising group the majority in transactions. Figure 3-4 shows that flowering/green house plants expenditures and transactions shares follow a different distribution among outlet groups. Unlike cut-flowers, flowering/green house plants expenditures levels are more equally distributed among specialty and mass

PAGE 36

22 merchandising. More specifically, specialty accounted for 48.5 percent of expenditures shares while mass merchandising accounted for 44.3 percent. Transactions share distribution was similar to that of cut-flowers with mass merchandising accounting for the majority of the market shares. With a market share of 64.4 percent, mass merchandising more that double specialty shares which accounted for 29.3 percent of the market. The graph also shows that in the specialty category other outlets account for 54.4 percent while florists account for 45.6 percent of the total specialty group in terms of expenditures. The difference is greater in terms of transactions where other outlets accounted for 73 percent of the market while florists only accounted for 27 percent. Figure 3-5 shows the outlet division for the mass merchandising group. The graph shows that supermarkets and warehouse/price club shares combined accounted for less than the rest of the outlets in the same category. In this case, other outlets accounted for approximately 55 percent of the market while supermarkets and warehouses/price club accounted for 43 and 2 percent respectively in both measurements. Market Shares by Product Form Figure 3-6 presents the distribution of cut-flowers market shares based on specific outlet types. In addition, florists product form is presented to show the distribution of arrangements and non-arrangements on both outlets. The graph shows that for florists the flower arrangements accounted for 70.5 percent while non-arrangements accounted for 29.5 percent in terms of expenditures. The distribution is more evenly distributed in terms of transactions with both forms accounting for approximately 50 percent of the market each. Figure 3-7 shows that the difference in flower form was considerable in supermarkets where non-arrangements accounted for the 82.1 percent and arrangements accounted for 17.9 percent based on expenditures. Alternatively, in terms of transactions

PAGE 37

23 the difference was greater with non-arrangements accounting for 91.8 percent and arrangements 8.2 percent. Florists87.8%Other12.2% Specialty68.7%Mass Merchandising25.8%Retail Internet2.1%Other3.4% Florists78.1%Other21.9% Specialty41.3%Mass Merchandising53.6%Retail Internet0.6%Other4.4% TransactionsExpenditures Figure 3-2 Percent of specialty market shares on expenditures and transactions for cut-flowers. Source: AFE and Ipsos-NPD group. Supermarkets82.0%Other11.9%Warehouses/Price Club6.0% Specialty68.7%Mass Merchandising25.8%Retail Internet2.1%Other3.4% Supermarkets83.8%Other12.1%Warehouses/Price Club4.1% Specialty41.3%Mass Merchandising53.6%Retail Internet0.6%Other4.4% TransactionsExpenditures Figure 3-3 Percent of mass merchandising market shares on expenditures and transactions for cut-flowers. Source: AFE and Ipsos-NPD group.

PAGE 38

24 Other54.4%Florists45.6% Specialty48.5%Mass Merchandising44.3%Retail Internet0.8%Other6.4% Other73.0%Florists27.0% Specialty29.3%Mass Merchandising64.4%Retail Internet0.2%Other6.2% TransactionsExpenditures Figure 3-4 Percent of specialty market shares on expenditures and transactions for flowering/green house plants. Source: AFE and Ipsos-NPD group. Supermarkets42.8%Other55.2%Warehouses/Price Club2.0% Specialty48.5%Mass Merchandising44.3%Retail Internet0.8%Other6.4% Supermarkets43.3%Other55.5%Warehouses/Price Club1.2% Specialty29.3%Mass Merchandising64.4%Retail Internet0.2%Other6.2% TransactionsExpenditures Figure 3-5 Percent of mass merchandising market shares on expenditures and transactions for flowering/green house plants. Source: AFE and Ipsos-NPD group.

PAGE 39

25 Flower Arrangements 70.5%Non-Arrangements 29.5% Florists60.3%Supermarkets21.2%Warehouses/Price Club1.6%Other Specialty8.4%Other Mass Merchand.3.1%Retail Internet2.1%Other3.4% Flower Arrangements 49.5%Non-Arrangements 50.5% Florists32.3%Supermarkets44.9%Warehouses/Price Club2.2%Other Specialty9.0%Other Mass Merchand.6.5%Retail Internet0.6%Other4.4% TransactionsExpenditures Figure 3-6 Percent of household market shares for florists by product form. Source: AFE and Ipsos-NPD group. Flower Arrangements 17.9%Non-Arrangements 82.1% Florists60.3%Warehouses/Price Club1.6%Supermarkets21.2%Other Specialty8.4%Other Mass Merchand.3.1%Retail Internet2.1%Other3.4% Expenditures Flower Arrangements 8.2%Non-Arrangements 91.8% Florists32.3%Supermarkets44.9%Warehouses/Price Club2.2%Other Specialty9.0%Other Mass Merchand.6.5%Retail Internet0.6%Other4.4% Transactions Figure 3-7 Percent of household market shares for supermarkets by product form. Source: AFE and Ipsos-NPD group.

PAGE 40

26 Share Distribution over Time Yearly percent of cut-flowers market shares expenditures and transactions on the different outlet groups are presented in Figure 3-8. To facilitate the discussion only whole years where considered for graphing purposes. Over time mass merchandising market share levels tended to increase during the 1993-2003 period. In terms of expenditures, mass merchandising increased from approximately 27 to 40 percent and in transactions from 50 to 70 percent. To the contrary, specialty decreased over the ten years showing a slight increase in 1998 but then falling back again in the last three years. Overall, specialty decreased from 60 to 50 percent in expenditures and from 35 to 29 percent in transactions. Since data on internet retail purchases was only available from 2000 the shares increase from that period up until 2003. Other outlets experienced reasonably stable market share levels up until the 1997 to 1998 period where a sharp decrease was seen in expenditures and transactions. Other outlets decreased from approximately 13 to 5 percent. In spite of all the changes in market shares, specialty continued to dominate in cut-flowers expenditures although the gap could reduce if the current trend continues. The graph also shows that the difference between mass merchandising and specialty is widening with the former gaining more shares over the latter throughout time. Figure 3-9 shows the yearly trends for flowering/green house plants outlet groups. Over the 10 year period specialty shares reduced from 53 to 44 percent in expenditures and from 32 to 25 in transactions. To the contrary, mass merchandising outlets experienced an increase in market shares from 40 to 53 percent in expenditures and from 60 to 71 percent in transactions. The market share levels for internet retailers and other outlets for flowering/green house plants follows the same trend as the cut-flowers distribution previously presented.

PAGE 41

27 0.000.200.400.600.801.00Market shares Expenditures SpecialtyMass MerchandisingRetail InternetOther 19931994199519961997199819992000200120022003Years0.000.200.400.600.801.00 Transactions SpecialtyMass MerchandisingRetail InternetOther Figure 3-8 Percent of yearly market shares in cut-flowers for outlet groups by expenditures and transactions. Source: AFE and Ipsos-NPD group. Unlike cut-flowers, mass merchandising outlets dominate the market in both measurements. More specifically, in 2001 the mass merchandising shares of expenditures superseded florists shares with the gap widening in the last years. The difference is even greater in terms of transactions with mass merchandising capturing more than the rest of the outlets combined. While the two previous graphs presented the relative changes in market share levels over time, the next two graphs show the variation that florists and supermarkets over the 10 year period. Figure 3-10 shows the yearly trends in cut-flowers expenditures and transactions for florists and supermarkets. In terms of expenditures, florists show a steady decline in market share levels particularly from the year 2000 onward. To the contrary, supermarkets market share levels have increased through time by nearly 10 percent.

PAGE 42

28 0.000.200.400.600.801.00Market shares Expenditures SpecialtyMass MerchandisingRetail InternetOther 19931994199519961997199819992000200120022003Years0.000.200.400.600.801.00 Transactions SpecialtyMass MerchandisingRetail InternetOther Figure 3-9 Percent of yearly market shares in flowering/green house plants for outlet groups by expenditures and transactions. Source: AFE and Ipsos-NPD group. It is important to notice that the difference in supermarkets share gains is not the same as that of the loss of florist which might suggest that other outlets are capturing part of the market as well. This could be particularly true for internet purchases, which as we mentioned earlier started to capture market shares in the year 2000. In terms of transactions, we see that both outlets start out at approximately the same level at 40 percent of the market each. However, as time passed a gap between the two outlets develops with supermarkets increasing its market share levels to nearly 52 percent with florists reaching 20 percent of the market. Clearly, there has been a decline in the florists purchases in cut-flowers over the last decade and the gap is increasing over time.

PAGE 43

29 0.000.200.400.600.801.00Market Shares Expenditures FloristsSupermarkets 19931994199519961997199819992000200120022003Years0.000.200.400.600.801.00 Transactions FloristsSupermarkets Figure 3-10 Percent of yearly market shares in cut-flowers for florists and supermarkets by expenditures and transactions. Source: AFE and Ipsos-NPD group. Figure 3-11 shows the trend in market share levels for flowering/green house plants for the same two outlets. Unlike the cut-flowers graph that showed a clear disparity in the market share levels of florists and supermarkets over time, flowering/green house plants levels do not seem to vary as much over time. Except for a few instances, both expenditures and transactions levels remain fairly stable over the years without showing major disturbances. It is important to notice however that florists expenditures levels started at 30 percent in 1993 and then declined to the same level as supermarkets at approximately 20 percent by 2003. In terms of transactions, both outlets maintain the same markets share level over time with 30 and 10 percent levels for supermarkets and florists respectively.

PAGE 44

30 0.000.200.400.600.801.00Market Shares Expenditures FloristsSupermarkets 19931994199519961997199819992000200120022003Years0.000.200.400.600.801.00 Transactions FloristsSupermarkets Figure 3-11 Percent of yearly market shares in flowering/green house plants for florists and supermarkets by expenditures and transactions. Source: AFE and Ipsos-NPD group. Figure 3-12 shows the percent of market shares for arrangements through both florists and supermarkets. As expected, in terms of expenditures florists dominate the market with approximately 77 percent by 2003 suffering a 10 percent loss since 1993. The low level of market share of supermarkets in this category (approximately 7 percent) was expected as well as the relatively small (3 percent) gain at the end of the period. The transactions graph shows that the market share gap between the two outlets is decreasing over time particularly since 2000. The graph not only shows that in general people tend to purchase through florists when it comes to buying flower arrangements. Figure 3-13 shows the percent of market shares in non-arrangements for florists and supermarkets over time. At the beginning of the period florists had a greater market share in terms of expenditures than supermarkets, however florists with 50 percent of the

PAGE 45

31 market lost the initial 10 percent advantage over supermarkets by the middle of 1998. From that year on we see a steady decline in florists market share reaching nearly 22 percent of the market while supermarkets ended with 53 percent. In other words, during the 10 year period florists and supermarkets switched positions in the industry. The situation in terms of transactions was also expected with supermarkets increasing their market share levels over time. This can be partly justified by the bigger base of potential consumer that supermarkets have over florists. By differentiating between arrangements and non-arrangements we can appreciate that most of the florists decline in market share over time can be explained by the great loss in the non-arrangement sectior of the market. Here again, the marketing and product mix seems to have a greater effect in supermarkets than in florists. Market Shares across Variables Figure 3-14 shows cut-flowers market share distribution by expenditures and transactions across the variables considered in the present study. The graph shows that the distribution of expenditures and transactions follows the same pattern in all the variables except purpose. In purpose, gift buying has a greater market share in expenditures which consequently is the highest level of market concentration when compared to the rest of the variables. Gender also denotes a marked difference in the distribution with female buying having more market shares than males. The first three age groups denote an increasing market share distribution while the last one shows a slight decrease. Consequently the first age group has the least percentage of market

PAGE 46

32 0.000.200.400.600.801.00Market Shares Expenditures FloristsSupermarkets 19931994199519961997199819992000200120022003Years0.000.200.400.600.801.00 Transactions FloristsSupermarkets Figure 3-12 Percent of yearly market shares in arrangements for florists and supermarkets by expenditures and transactions. Source: AFE and Ipsos-NPD group. 0.000.200.400.600.801.00Market Shares Expenditures FloristsSupermarkets 19931994199519961997199819992000200120022003Years0.000.200.400.600.801.00 Transactions FloristsSupermarkets Figure 3-13 Percent of yearly market shares in non-arrangements for florists and supermarkets by expenditures and transactions. Source: AFE and Ipsos-NPD group.

PAGE 47

33 shares of all the divisions within the variables. The fact that market share concentration decreases by nearly 10 percent in the fourth group tells us that people tend to buy cut-flowers up to a certain age and then on consumption drops to a level similar to people that fall in the second category. Unlike age, cut-flower consumption does not increase as income rises. The four income groups follow a more erratic distribution with the first and third groups having approximately the same market share level at 20 percent. The same is true for the second and fourth groups with approximately 30 percent of the market each. Surprisingly, the third income group does not seem to be buying to a level proportional to their purchasing power. In general, the graph shows that major combinations of variables peak demand in cut-flowers such as gift, female, of approximately 41 to 54 years of age and with an income of either the second or fourth income group. The same is true for the combinations that show a decrease in market share percent relative to the average consumer such as a male buying for self who is under 25 years and with an income in either the first or third group. Note that these are simple percentages without any constraints on the other variables when calculating the distribution. It was expected that the distribution of the variables differ when comparing cut-flowers and flowering/green house plants. As presented by the variables for flowering/green house plants in Figure 3-15, this difference is more noticeable when comparing purpose and gender. When it comes to the purpose, the distribution is equally distributed among the two divisions in expenditures. However, in terms of transactions self buying market shares, at 64 percent, nearly doubles the gift percentage of the market. In gender, females have approximately 80 percent of the market in both expenditures and

PAGE 48

34 0.630.37 0.860.14 0.180.280.220.32 0.070.280.370.27 0.000.200.400.600.801.00Market Share Expenditures AgeIncomePurposeGender 0.660.34 0.680.32 0.220.290.200.29 0.080.270.360.29 Under 25 years25 to 40 years41 to 54 years55 years and overUnder $25,000$25,000 to $49,999$50,000 to $74,999$75,000 and overGiftSelfFemaleMaleDemographics0.000.200.400.600.801.00 Transactions AgeIncomePurposeGender Figure 3-14 Percent market shares of cut-flowers expenditures and transactions by demographics. Source: AFE and Ipsos-NPD group. 0.780.22 0.500.50 0.220.280.190.31 0.050.260.360.33 0.000.200.400.600.801.00Market Share Expenditures AgeIncomePurposeGender 0.820.18 0.360.64 0.280.310.180.23 0.060.270.340.33 Under 25 years25 to 40 years41 to 54 years55 years and overUnder $25,000$25,000 to $49,999$50,000 to $74,999$75,000 and overGiftSelfFemaleMaleDemographics0.000.200.400.600.801.00 Transactions AgeIncomePurposeGender Figure 3-15 Percent market shares of flowering/green house plants expenditures and transactions by demographics. Source: AFE and Ipsos-NPD group.

PAGE 49

35 transactions. The rest of the variables follow the same distribution showed in the previous graph. Figure 3-16 shows that in terms of cut-flowers florists and supermarkets dominate the specialty and mass merchandising sections of the market. Florists have approximately 60 percent of the market in terms of expenditures followed by supermarkets with 21 percent. To the contrary, supermarkets have the highest market share concentration with approximately 45 percent followed by florists with 32 percent. In both cases, the combined market share of both outlets exceeds by more than 80 percent the total market share of the cut-flower section of the market. 0.600.210.020.080.030.020.03 FloristsSupermarketsWarehouses/Price ClubOther SpecialtyOther Mass MerchandisingRetail InternetOther0.000.200.400.600.801.00 Expenditures Cut Flowers 0.320.450.020.090.060.010.04 0.000.200.400.600.801.00 Transactions Cut Flowers Market SharesOutlets Figure 3-16 Percent of cut-flowers market shares expenditures and transactions by outlets. Source: AFE and Ipsos-NPD group. When considering flowering/green house plants the dominating outlets also fall within the specialty category and the difference to the mass merchandising shares is of

PAGE 50

36 only 3 percent. Figure 3-17 shows that even though specialty and mass merchandising stores have the greatest market shares in the industry; neither florists or supermarkets capture the majority of the percentage in their respective groups. The graph shows that the other specialty and mass merchandising stores grouped together have more than florists and supermarkets market shares. In the flowering/green house plants the tendency to buy through the internet is low as described by the low market share percentage both in expenditures and transactions. Both Figure 3-16 and 3-17 show the importance of florists and supermarkets when selecting fresh flowers, particularly in the cut-flower section of the market. 0.220.190.010.260.240.010.06 FloristsSupermarketsWarehouses/Price ClubOther SpecialtyOther Mass MerchandisingRetail InternetOther0.000.200.400.600.801.00 Expenditures Flwg/Green House Plants 0.080.280.010.210.360.000.06 0.000.200.400.600.801.00 Transactions Flwg/Green House Plants Market SharesOutlets Figure 3-17 Percent of flowering/green house plants market shares expenditures and transactions by outlets. Source: AFE and Ipsos-NPD group.

PAGE 51

CHAPTER 4 THEORETICAL MODEL AND MODEL SPECIFICATION This chapter presents a general consumer demand concept to illustrate the buying decision by outlet among U.S. households Later, Heckmans censored model is developed along with the model specification for the present study. Socioeconomic and demographic variables incorporated into outlet market share models used in this study to are presented and discussed. Finally, outlet market share estimates are presented in the form of econometric parameters and their statistical properties. Consumer Demand for Flowers Consumer demand theory analyses the behavior of consumers as they purchase a set of goods to satisfy personal needs given a specific utility function. Generally, consumers have a budget constraint that limits the choices that they make in their purchasing behavior. The theory assumes that the consumer acts as a rational economic agent, maximizing his utility function given his own budget constraint. In the general form, maximization of the utility function as described by Girapunthong and Ward (2003) is given by: Max u = u(q1,...,qn) (4-1) pjqj = m nj1 where pj and qj are the price and the quantity of the jth good, respectively, and m is the total expenditures or income on all n goods. In this study, flowers are divided in outlet market shares for cut-flowers, flowering/green house plants and dry/artificial. As shown in a previous chapter cut37

PAGE 52

38 flowers and flowering/green comprise most of the flowers industry in the U.S. The analysis further divides cut-flowers in arrangements and non-arrangements due to the importance as a source of business for the two major outlets of this study, florists and supermarkets. Details on the Data Outlet market shares in the cut flower industry can be modeled by changes in consumer preferences. A set of socioeconomic and demographic data were defined such as age, income, gender, and purpose for buying to determine their influence on consumer outlet selection. In models that use information from panel data the question on the sample representative from the population is always present. Ahl (1970) states that if, as occasionally happens, the test panel is not perfectly balanced to the test market along one or more key demographic characteristics, it may be necessary to take these imbalances to account in the prediction. If that is the case, then the predictions for each demographic group would have to be weighted according to the percent of the population it represents. That however is not the case for the data used in this study as it is collected by Ipsos, a private organization that specializes in retrieving balanced consumer information from targeted population. That is, the sample is demographically representative of the population. Censored Data Model The data of the study follow a function form with a substantial number of observations having zero/near-zero and one hundred percent values. This tendency is often seen in data that models consumer demand of certain commodities where zero is a possibility. When the data follow this functional form, the problem of censoring of the dependent variable arises. The researcher may be tempted to erase such occurrences from

PAGE 53

39 the dataset and work with the rest of the sample. However, by doing so, the integrity of the data, as well as the validity of the findings and policy suggestions are seriously compromised. The significant portion of observations on cut-flowers and flowering/green plants taking a zero or one hundred percent values (insert figures) provides justification for considering censored regression models as an appropriate framework for conducting the present investigation. Tobin proposed an estimation method for data with truncation problems later called the Tobit model (Tobin, 1958). According to Tobin, probability and multiple regression models fail to present thoroughly information about the dependent variable because of the probability of limit and non-limit responses. The general formulation of the Tobit double-censored model is yi* = Xi + i i ~ N(0, 2) (4-2) 0.11,0.10,00iiyifyifyyifyiii where yi* is defined as the latent variable and yi is the dependent variable (Greene, 2003). However, Tobin assumes that the decision to consume is the same as the decision of how much to consume of the good and this is not always the case (Haines et al., 1988). In cases in which the decision to consume and the amount of the good consumed differ, the Tobit model understates the actual magnitude of the dependent variable. Therefore, it is necessary to redefine the concept of the Tobit model to account for what Cragg called a double hurdle model (Cragg 1971). According to Cragg, there is a clear distinction between the propensity to consume and what is actually purchased. In his double hurdle model, he utilizes a probit model to calculate the probability of buying the good (first

PAGE 54

40 hurdle) and a standard regression model (second hurdle) for the amount of the good purchased. In this research the analogy is to have selected outlet, you first had to be a buyer. Hence, the first hurdle is being a buyer. Then the Tobit is part of the second equation. The nature of the data makes it necessary to censor the data in the present study and agreeing with Long (1997) that if such procedure occurs Ordinary Least Squares (OLS) is inconsistent, an alternative to the Tobit model was considered for the present analysis. This papers uses the two-step decision proposed by Heckman instead of the double hurdle presented by Cragg and discussed earlier mainly due to the fact that the portion of the residual that arises from the use of an estimated value of i, in place of the actual value of i is not orthogonal to the xi data vector (Heckman, 1979). Heckman proposed a model similar to Cragg by acknowledging the two-decision approach in purchasing behavior. However, Heckman used an inverse Mills ratio to link both processes. As described by Heckman and further explained by Long (1997) the sample selected in the probit model is given by yi*= Xi + i assuming that i ~ N(0, 2) (4-3) and .01,0011yifyifyi where Xi is a 1Kj explanatory vector of socioeconomic or demographics variables. Also, j is a Kj1 vector of parameters with j=1,2,,n. and is a residual that captures unobserved influences in the dependent variable. The magnitude of the parameters reflects the impact of changes in the x vector on the probability. The subscript 1and 2 denote the probit and Tobit model specification respectively.

PAGE 55

41 To simplify the formulas that follow, let 1 = Xi (4-4) and so the decision to become a fresh flower buyer follows a standard normal probability function given by f( y1*| 1, 1) = 111121111121exp21yy (4-5) with the cumulative distribution function of y* given by F(y1= 0 | 1, 1) = (4.6) )0*Pr(),|(1ydzzfxb and by default Pr(y1= 1) = 1 F(y1* | 1, 1) (4-7) which can rewritten as: Pr(y1 0) = 111y = F(y1= 0 | 1, 1) (4-8) Pr(y1 > 0) = 111y = 1 F(y1* | 1, 1) since we are dealing with a symmetric standard normal distribution. In order to calculate the probability distribution function of the censored part of the distribution the original distribution is divided by the region to the right of zero (positive purchases). The function is given by f(yb | y1* > 0, 1, 1) = )1 yPr(),|(1111yf (4-9) and since the data has been censored to the left of zero, E(y | y1* > 0) > E(y1*) = 1. Then, the expectation of becoming a buyer is given by

PAGE 56

42 E(yb | y1* >0) = 1 +1 110 (4-10) where 1= 0011 (4-11) is a monotone decreasing function of the probability that an observation is selected into the sample known as the Inverse Mills ratio. The value of 1 is saved and used as a regressor in the second stage estimation. Heckmans two-stage method generally utilizes a probit for participation (probability of purchasing) and an ordinary least squares (OLS) procedure for the actual consumer purchases. Both decisions are independent and thus no corner solution is observed. However, in this study corner solutions can be seen in the second stage and would imply that a consumer did not buy in either two outlets. Instead of an OLS procedure a Tobit was used to model intensity of buying among florists and supermarkets. The formulation for the Tobit model is set as 0.11,0.10,002222yifyifyyifyi (4-12) The second stage Tobit is a function of both the variables considered in the probit and the Inverse Mills Ratio estimated by the model. The notation for the Tobit is as follows: Z = f(X, 1) (4-13)

PAGE 57

43 and to simplify the formulas that follow let 2 = Z (4-14) and since we are dealing with a standard normal distribution the probability that an observation is censored from below is given by prob(y = 0 | Zi) = 2Z 22 (4-15) and censored from above by prob(y = 1.0 | Zi) = 1 21Z 1 221 221 (4-16) with the uncensored portion given by prob(0 < y <1.0 | Zi) = 2Zy (4-17) and for facilitation purposes let 0= 20Z 1= 21Z (4-18) 0 = 20Z 1 = 21Z so notation for the expected market share for the two outlets is given by E(y) = (0) + (1) + (1 0) 01102Z (4-19) 1 + Z(1 0) + 2 10 which is presented in the Tobit model. Next, the model specification is presented for the probit and Tobit models in terms of the variables of this study.

PAGE 58

44 Tobit Model for Outlet Selection Several socioeconomic and demographic variables are included in Heckmans model: the respondents age, income level, gender, and purpose for buying, flower form, and months to account for seasonality. Furthermore, the dependent variable can be expressed in terms of transactions or expenditures levels. First, dummy variables are defined for the right-hand-side were the sum of the parameters for each set of dummy variables are restricted to zero. And so the function for the first step of the Heckman estimation process using the probit model for positive expenditures and transactions is given by X = 0 + + + + (4-20) iiiDAge41 iiiDGnd214 iiiDPur216 iiiDInc418 + + iiiDMt12112 iiiDForm3124 Since each dummy class is exhaustive and mutually exclusive, inclusion of all discrete values within a class immediately creates the well known dummy variable trap. One convenient solution to this problem is to restrict the sum of the coefficients to zero where, for example, then 4 = 1 2 3. Substituting for 4 (for the other appropriate coefficients) then gives equation (4-22) where the dummy variable trap no longer exists. The rest of the coefficients are described by 041ii 4 = 1 2 3 (4-21) 6 = 5 8 = 7 12 = 9 10 11 24 = 13 14 15 16 17 18 19 20 21 22 23

PAGE 59

45 27 = 25 26 and defining XAgej = DAgej DAge4 (4-22) XGndj = DGndj DGnd2 XPurj = DPurj DPur2 XIncj = DIncj DInc4 XMtj = DMtj DMt12 XFormj = DFormj DForm3 The probit equation variables now corrected for the dummy trap are X = 0 + + 5XGnd1 + 7XPur1 ++ (4-23) iiiXAge31 iiiXInc318 iiiXMt11112 + iiiXForm2124 defined as the dummy variables in the probit model. The impacts of the variables are then compared relative to the average household measured with 0. The second stage is depicted by a Tobit model with the same variables of the probit model including the Inverse Mills Ratio as a regressor in the equation. The function is given by Z = 0 + + 5XGnd1 + 7XPur1 + + (4-24) iiiXAge31 iiiXInc318 iiiXMt11112 + + 28 Mills1 iiiXForm2124 The explanatory variables used in the model are described in Table 4.1. Tables 4.2 and 4.3 present the estimated Probit and Tobit coefficients and t-values. The coefficients reflect the sign of the relationship to the dependant variable, in this case market shares,

PAGE 60

46 while the t-values reflect the significance of the relationship. Next, the Probit and Tobit coefficients are discussed outlining the most statistically significant variables in the model. First-Step Results: Decision to Become a Buyer The decision to become a buyer was estimated by a probit model using expenditures and transactions. The parameters in Tables 4.2 and 4.3 denote the effect on the probability of buying flowers from an initial sample of 27,072 household observations with that 8,040 households not becoming buyers of fresh flowers. An econometric programming code using TSP are included in Appendix C. The results indicated that there was little distinction between the estimated coefficients and t-values in terms of expenditures and transactions as should be the case since one cannot have expenditures without transactions. Thus in the second stage Tobit we could have equally included the Inverse Mills Ratio from each model when establishing both the second stage expenditures and transaction models. 19921993199419951996199719981999200020012002200320040.000.200.400.600.801.00 Zero PurchasesPositive Purchases Figure 4-1 Distribution of values in the first stage probit model.

PAGE 61

47 Table 4-1 Description for the Variables in the Heckman model VarDinit iables ef ion 0 Intercept Age 1 = u Age 1 = 2 Age 1 = 4 Age 1 if = 55 and m Gnd 1 if = fe Gnd 1 = m. Pur 1 = g Pur 1 if = self, 0 otherwise. Inc 1 if = under $25,000 dollars, 0 otherwise. Inc2 1 if = $25,000 to $49,999 dollars, 0 otherwise. Inc3 1 = $ Inc4 1 = $ Mt 1 = J Mt 1 = F Mt 1 if = March, 0 otherwise. Mt4 1 if = April, 0 otherwise. Mt 1 = M Mt 1 if = June, otherwise. Mt7 1 if = July, 0 otherwise. Mt 1 = A Mt 1 = S Mt 1 if = October, 0 otherwise. Mt 1 if = November, 0 otherwise. Mt 1 if = December, 0 otherwise. For1 1 = Atherwise. For 1 = N Form 1 if = Flowering, 0 otherwise. Mills Monotong function of the probability that an observat 1 if nder 25 years of age, 0 otherwise. 2 if 5 to 39 years of age, 0 otherwise. 3 if 0 to 54 years of age, 0 otherwise. 4 ore years of age, 0 otherwise. male, 0 otherwise. 12 if ale, 0 otherwise 1 if ift, 0 otherwise. 2 1 if 50,000 to $74,999 dollars, 0 otherwise. if 75,000 dollars and more, 0 otherwise. 1 if anuary, 0 otherwise. 2 if ebruary, 0 otherwise. 3 5 if ay, 0 otherwise. 6 8 if ugust, 0 otherwise. 9 if eptember, 0 otherwise. 10 11 12m if rrangements, 0 o m2 if on-arrangements, otherwise. 3 ne decreasi i selected in Sig Fuctioof using each outlet among buers. to the sample n that describes the probability ma n y SouE and Ios-N rce: AF ps PD group

PAGE 62

48Table 4-2 Estimated Probit and Tobit Coefficients for Expenditures Buyers based on Expenditures Expenditures Shares Mode ls Florists Share Variables Probit CoefficientsT-valuesTobitCoefficients T-valuesvalu SupermarketsCoefficients Share Tes 0 0.8869374.693400 0.11087 8.309780.115.78032 7627 XAge1 1 -1.16447-63.360201 -0.13624 -9.90005-0.0-4.60476XAge2 2 0.2165711.983002 0.01543 2.07795-0.0-1.71780XAge3 3 0.5107426.562303 0.08026 9.998900.02.47201XGnd1 5 0.5125146.351805 0.05111 9.248860.05XPur1 7 0.7749365.636707 0.32441 44.13356-0.0XInc1 9 0.111556.237839 -0.00742 -1.046460.0XInc2 10 0.3263217.4951010 -0.00173 -0.240440.0XInc3 11 -0.16054-9.1293311 0.01908 2.63217-0.038XMt1 13 -0.19032-5.7986613 0.00485 0.349610.03256XMt2 14 0.163114.7035214 -0.00526 -0.403290.05487XMt3 15 0.087012.5520015 -0.00669 -0.505710.00736XMt4 16 0.263657.1594816 0.01653 1.224990.00871XMt5 17 0.3967010.5204017 -0.00998 -0.74081-0.046521143XMt6 18 -0.07695-2.2123818 -0.03266 -2.29710-0.021438230XMt7 19 -0.20605-6.2946619 0.00705 0.50564-0.023490252XMt8 20 -0.17209-5.2200020 0.01311 0.94799-0.039784427XMt9 21 -0.20298-6.2145121 0.00200 0.14352-0.023920617XMt10 22 -0.03584-1.0739222 0.01908 1.419800.007296524XMt11 23 -0.02847-0.8499323 -0.03811 -2.833410.039035081XForm1 25 -0.87114-57.6185025 0.34996 38.25374-0.212435013XForm2 26 0.4006526.1502026 -0.04366 -6.708990.258606436Mills1 28 0.14068 4.956420.002150889Sigma2 29 0.50350 122.350260.440756705 539010651673542945340330468605 11.-7.0.7.-6.2.5.0.0.-4.-1.-2.-3.-2.0.3.-26.46.0.139. 34665185618790026598326080867067743 27266369914075802588 Number of observations = 27072 Number of observations = 19042 Number of positive observations = 19042 R2 = 0.462364

PAGE 63

Table 4-3 Estimated Probit and Tobit Coefficients for Transactions Buyers based on Transactions Transaction Shares Models C Florists Share T Supermarkets Share Variables Probit Coefficients T-values Tobit Coefficients -values oefficients T-values 0 1 0.88724 74.70430 0 0.13627 1.61261 0 .20188 18.22941 XAge1 1 --1 -2 121 -1.16425 -63.34360 1 -0.04337-3.58532 -0.10275 -8.84949 XAge2 2 0.21645 11.97540 2 0.00168 0.25718 0.000330 0.05444 XAge3 3 0.51070 26.55770 3 0.03642 5.15586 .03830 5.70522 XGnd1 5 0.51278 46.36870 5 0.01399 2.87573 0.07338 15.70986 XPur1 7 0.77521 65.64910 7 0.23679 36.56254-0.01069-0 -1.81563 XInc1 XInc 9 0.111390.32619 6.2281817.48720 9 -0.00684-0.01935 1.09601-3.05088 .006200.05371 -1.062788.99824 2XInc 10 10 3XMt 11 -0.16076 -9.14078 11 0.02446 3.83456 00.03923 .03832 -6.36662 1XMt2 1314 -0.19054 -5.80531 1314 0.00713 0.58409 3.44324 0.16294 4.69842 -0.02147 1.86785 0.06568 6.12739 XMt3 15 0.08682 2.54620 15 -0.01278 -1.09623 0.01680 1.54508 XMt4 16 0.26350 7.15493 16 0.00287 0.24169 0.00937 0.83857 XMt5 17 0.39657 10.51640 17 -0.02293 1.93285 0.04283 -3.81800 XMt6 18 -0.07716 -2.21815 18 -0.01844 -1.47370 -0.02892 -2.48029 XMt7 19 -0.20629 -6.30172 19 0.01329 1.08297 -0.02450 -2.13158 XMt8 20 -0.17233 -5.22679 20 0.02113 1.73545 -0.04379 -3.82450 XMt9 21 -0.20322 -6.22130 21 0.00993 0.80951 -0.02705 -2.35192 XMt10 22 -0.03606 -1.08034 22 0.01822 1.54068 0.00745 0.67294 XMt11 23 -0.02868 -0.85625 23 0.02865 2.41960 0.02818 2.55180 XForm 25 -0.87143 57.63000 25 0.39821 49.61143 0.23908 30.13249 XForm2 26 0.40047 26.13730 26 0.07376 12.88202 0.27534 50.09437 Mills1 28 0.02866 1.15040 0.03889 1.620003 Sigma 29 0.44281 2.35598 0.43836 8.51323 Number of oeration= 27072 of s = 1904 19043 bs v s Number observation 2 49 Number of positive observations = R2= .462495

PAGE 64

50 This study assumes a significance level of ninety five percent also denoted by an absolute t-value of 1.96 or greater. Figure 4-1 presents the distribution of the values for the fi becoming a buyer of flowers. The graph shows that approximsahough ars. The ph also shows the importance of using a probit model for the first stage as a way to usll the ence on thnd -57pact on the decision to purchase fresh flowers. Clearly, the probability of becoming a buyer increarrangements had a negative and significant impact on the probability to become a buyer. Thuseast likely oto a lesser degree wering and green plants. olds dion to om br. However, the 5 years ofative ients icating a negative relationship to the average buyer. In other words, people who fall in this che cnts indicated that the probability of becoming a buyer increases as the consumincreases peaking in the 40 to 54 years age group and then a rst s tag e o f t he mo de l co ns ide rin g e ith er pos itia ve tely on average 70 percent of the or neg ati ve res po nse s to mpall le u, th sede p cherc osent e to of be po cosit meive a re buspo yernse ofs h anas y t in ypecre ofase fld b owy 3 ers p inerc anent y o ov utler et. the Alt 1 smgrafocva 0 yeat aflu64 at imcoeoef agthe onles ac w tuaere l fl sig ownif er cica onnt suin methe rs. m Thod e mels. ag Th nite udvar e oiabl f thes e t th -vaat h luad es i th nde g icareates ted tht in riabe pr.63 ope res nspe ityctiv to ely bec. I omn p e aurp bos uyee, r wgift er bu e pyin urpg oshad e a th nde g fore rm ate wist p th tos -vaitiv lue s es oign f 6ifi 5.can ases w hen th e r eas on for bu yi ng is not fo r se lf con su mp tio n. Fo r fo rm s, sugse a gerra stinnge g tme hatnts if pr a cefe onrri sumng er ch deoos cidin esg in to ste buad y fr no esnh flarra owng erem s heen wts a ill lnd choflobecind The a e a gesuye of th e h ous eh old haund veer 2 a s ign if ica nt i ag mpe g acrou t op n thhad e h a ouneg seh ecisfficficiee la ateg or y a re l ess in cli ned to bec om e a b uye r t han th e a ve rag e a ge T ers sli gh t de clin e in st

PAGE 65

51 groupt ince e er gree owed insignificant to the decision to become a buyer and that is consishares, rner solutions in this particular case could mean that a consumer is a buyer but did not buy Recall that the last group equals the negative sum of the parameters for each dummy class. As for the gender variable, females had the greatest effect on the decision to become a buyer. This result seems to be consistent with previous studies that argue thathe majority of purchases of products available at supermarkets are done by females sthey do most of the grocery shopping within the household. Compared to the rest of the demographic variables, income is less significant to thdecision of becoming a buyer. The first two groups, under $25,000 and $25 to $49,999showed a positive relationship to the decision to become a buyer while the third level $50,000 to $74,999 showed a negative one. Clearly the probability of becoming a buydoes not consistently increase as income rise. Apparently seasonality influenced the decision to become a buyer to a lesser dethan the rest of the variables. It is important to notice that the coefficients shpositive signs from January through May and negative the rest of the year. This may not be surprising since the first period captures some of the most important calendar occasions like as Valentines and Mothers Day. Throughout the year, only the months of October and November were stent with the well known demand problems in the fall months Second-Step Results: Intensity of Buying The parameters in equation 4-19 reflect the decision of much to consume in either florists or supermarkets once the decision of buying has been taken. Because this study only considers florists and supermarkets purchases to model their expected market the parameters could yield zero responses also known as corner solutions. The co

PAGE 66

52 through either two mentioned outlets. Unlike the probit estimates the coefficients andvalues estimate t-d for the second-stage Tobit differ considerably in terms of expenditures and trrobit values in the second stage Tobitt purchases is considerable and had to be accou ansactions as would be expected. It is important to recognize that while the pmodel dealt with either positive or negative answers, the Tobit model focuses on consumers that chose not to become buyers, buy some, or buy all the time in either florists or supermarkets. Figure 4-2 shows the distribution of model with the three possible outcomes in the distribution in for florists. The Tobit estimates focus on the part of the responses that fall within the middle category in the graph, which represent 61.7 percent. In Figure 4-3 the same portion for accounted for 49.3 percent of the distribution. In both graphs, the percent of values of the distribution that either reported no purchases or 100 percen nted for in the by using a censored model. Zero 29.7% Middle Figure 4-2 Distribution of values in the second stage tobit model for florists. 61.7%One8.6%

PAGE 67

53 Zero 39.7%Middle One11.0% 49.3%Figure 4-3 Distribution of values in the second stage tobit model for supermarkets. The probit results indicated that purpose and form were the variables that had the greatest impact in becoming a buyer. In accordance, the Tobit estimates showed that the exact opposite was the true for supermarkets coefficients. In all, the magnitude of the significance was slightly greater in terms of transactions. This shows that arrangements same variables had the greatest impact in selecting outlets. In terms of expenditures, the coefficients for gift buying were positive for florists and negative for supermarkets and statistically significant based on their t-values. This shows that when it comes to buying fresh flowers as a gift the consumer prefers florists than supermarkets. In terms of transactions, the same relationship was seen in florists with supermarkets (t-value of -1.81) having no significance from the average. Form was statically significant for both arrangements and non-arrangements for the two measurements. The relationship was clear with florists having positive and negative coefficients in terms of arrangements and non-arrangements respectively. The

PAGE 68

54 are bought more in florists than in supermarkets. The opposite is true for non-arrangements and flowering/green plants. Therefore, product type tends to dictate the type of outlet from where to purchase. The previous results indicated that females were more inclined than males to become buyers. The Tobit estimates showed that females were also more predisposed to choose supermarkets than florists for their respective purchases. This intensity of buying is considerably greater in terms of transactions. Income groups describe different intensities of buying between the two outlets. People under 25 years had negative and insignificant impact on outlet choice. The 24 to 40 years of age group showed negative and significant impact for florists and positive for supermarkets. The group clearly prefers to buy in supermarkets once a decision to buy was made. The third group, 40 to 54 years of age presents exact opposites parameters for florists and supermarkets preferring the former outlet. From this relationship, people with a less amount of income primarily buy in supermarkets until their income increases to a point in which they switch to florists. Seasonality impacts on outlet selection were more heterogeneous than was initially expected. The months of March, April, June (only in florists), and October showed insignificant coefficients for the two outlets. Thus, showing that in those months there is no clear distinction of preference among buyers in the two outlets. The coefficients for florists were insignificant except in the months of June and November when they have negative values. To the contrary, supermarkets show significant coefficients with positive values in January, February and November and negative in the rest.

PAGE 69

55 As stated before, the Inverse Mills Ratio is a decreasing function of the pselected into the sample. The Inverse Mills Ratio coefficients were insignificant excepterms of florists expenditures. This me robability t in ans that except for florists expenditures a regular Tobit. The sumers. The sigma values represent the decisit model could have been estimated using just these households that were buyerspositive and significant coefficient in florists expenditures shows a more complicated decision making process. A Tobit estimation in this case could have erroneously oversimplified the decision process faced by con on to become buyers in either florists or supermarkets. In this case the significanimpact of the sigma coefficients, which represents the Inverse Mills Ratio for outlet selection, means that by running the model only among consumers of a particular outlet will create a sample selection bias. Clearly, the coefficients show that something influenced consumers outlet selection once a decision to buy has been taken.

PAGE 70

CHAPTER 5 MODEL SIMULATIONS This chapter uses the estimates from the previous chapter to show simulated changes in the model variables. Simulations were conducted over each of the econand socioeconomic variables in the model. Two simulations in particular combine the variables that had the greatest impact on outlet selection. Then, a ranking of the variablethat showed the largest effect of the coefficients and range is discussed. Finally, timerecursive methods are estimated to analyze the variation of the coefficient for the averageconsumer. For any give combination of variables included in the models from Chap omic s ter 4, one can calculate the estimated market share either for those shares when the share lies between zero and one or for estimated shares across the full range of values including the limits. Since, each variable influences both the continuous and limits in the shares as well as the probability of being a buyer (i.e., as captured with the Inverse Mills ratio in the model), the following simulations will be based on the expected shares including the full range of share possibilities. For example, in Figure 5-1 the distribution of the shares clearly reflects the need for the Tobit estimates as presented in Chapter 4. Give the estimated model then the expected share can be for the non-limited share or for the full range as suggested with the three arrows combined in the upper expectation in Figure 5-1. Since most marketing policies intended to influence the outlet shares are likely over the full range of the independent variables suggest with Figure 5-1, the more useful information would be to have the expected shares over the full range. 56

PAGE 71

57 Figurclude d by e 5-1 Distribution of shares Hence, in all of the following simulations or sensitivity analyses, the expectations inthe upper and lower limits alone with the continuous portion of the estimates. The exact procedures are demonstrated in the Appendix C. Expected Share for the Average Conditions Recall from Chapter 4 that each dummy variable was estimated by imposing the restriction that the sum of the coefficients for each dummy was set to zero. What this means is that each estimated coefficient is a deviation from the average household anthe intercept in the model represents the average. Hence, as a reference point of the subsequent expectations across a number of variables, the expected shares using just the intercept for the outlet (florist and supermarket) and measurement (expenditures and transactions) equations provide those average shares. Later simulations are measured

PAGE 72

58 adjusting the variables relative to these means. Note also that these means are over the full range as discussed with Figure 5-1. Figure5-2 shows the expected market shares for florists and supermarkets based on both expenditures and transactions. As discussed earlier, it is possible that someone may be a buyer but chose not to use either a florist (or supermarket) in a given period. That is the zero share is feasible with the right set of circumstances. Similarly, the circumstances could be that one type outlet is always used, thus giving the share of 100 percent. In this figure, on average florists and supermarkets account for approximately 58 percent of the total fresh flower expenditures and 55 percent of the transactions for the average household. Between the two outlets the distributions are quite similar especially for the expenditures. Note that supermarkets, on average, capture a slightly larger share of the transactions as would be expected given the volume of traffic through supermarkets versus florists. 0.280.27 0.000.200.250.300.350.55 0.400.450.500.60Outlet probability by expenditures 0.050.100.15 0.25 0.30 Floriupermark0.000.050.100.200.250.300.350.400.450.550.60Outlet probability by transactionsFigure 5-2 Average outlet shares the fresh flower market for florists and supermarkets. stsSets0.150.50

PAGE 73

59 Expected Outlet Shares by Demographics Outlet Shares by Household Age Grou ps of rket ilities showed the expected upward slope as age increases while florists that the range of impact of age somewhat greater for supermarkets than for florists. This Figure5-3 represents the outlet expected shares of the market for fresh flowers with the probabilities (shares) based on the four age groups defined in Chapter 4 and shown onthe bottom axis of Figure 5-3. Each bar in Figure 5-3 shows the deviation from the average (see Figure 5-2) with bars below the average reflecting negative impacts and values above the mean indicating positive impacts. In addition to the direction, the magnitude of the change is directly comparable since all units are now in terms of marketshares. Hence, one can draw meaning conclusions about the direction and degree impact across the four age groups. For the most part in the graph, expenditures and transactions levels displayed similar distributions with each type outlet. Between outletsthe impact of age is considerably different both in magnitude and direction. Supermaselection probab probabilities show a more erratic increase up to the 40 to 54 age group but then decreased among those 55 and older. The first and second age groups consistently showed probabilities either below or close to the means for both outlets. The third and fourth groups present opposite outlet selection inclinations with the 40 to 54 years old group more inclined to purchase in florists and the 55 years and more age group showing a higher probability of selecting supermarkets. Recall from the previous chapter that the 25 to 40 years age group showed the least impact on the decision to become a buyer in either outlet. In this case a less insignificant impact on the decision to become buyers could mean that people from that category are indifferent as to which outlet to choose for their purchases, thus showing probabilities closer to that of the average consumer. Note also

PAGE 74

60 is particularly true for n the use of outlets d magnitude for supermarkets. Outlehare nd ility ery the e clear ts rs the transactions. Overall, the impact of aging o is far more consistent both in the direction an t Share over Gender Gender was one of the variables showing considerable impact on outlet selection as initially suggested with the Tobit responses from Table 4-2. Again using expected srelative to the average shares, Figure5-4 shows the outlet probabilities between male afemale buyers. As expected, the simulations show that females have a higher probabthan males of selecting the two outlets for fresh flower purchases. This difference is greater in terms of supermarkets than in florists and could possibly be related to grocshopping periods where females are more likely buying the groceries. Going back toestimates in the previous chapter the first Inverse Mills Ratio showed that the propensity to become a buyer was significant only when considering florists. This indicated that something influenced the consumers decision to buy in florists. In this case, a more complicated outlet selection decision is reflected with a probability that is closer to the model means of selecting that particular outlet. From a marketing perspective, thimplication from Figure 5-4 is that florist targeting gender would likely have far less impact than expected with supermarket programs. At least the potential gains for florisare substantially less than for supermarkets as evident from the relative sizes of the bain Figure 5-4. Buying Purpose Impact on Outlet Shares The model estimates showed that purpose had the most significant impact amongthe variables expected to influence outlet selection. The simulations in Figure5-5 show major variations between outlet selections with gift buying depicting a higher probabilityfor florists than for supermarkets. In addition, supermarkets showed a higher probability

PAGE 75

61 of selection when considering buying for personal use (self). Also, the range of shares forsupermarkets according to purpose was considerably less that the range for flo rists. As tlet probabilities when considering purpose suggests that consu expected, the difference in ou mers tend to choose florists for gifts and to lesser degree supermarkets for self use. The range difference in florists could probably be attributed to customary behavior, where flowers are generally purchased for gifts on either calendar or non-calendar occasions. 0.220.32 0.240.260.280.300.340.360.38Florist probability by expenditures und25yrs25 40 yrs40 54 yrs55us yrsSupermarket probability by transactions er to to pl un 25yrs25 40 yrs40 54 yrs55us yrs0.220.240.260.280.300.340.360.38Supermarket probability by expenditures derto to pl0.32 Florist probability by transactionsFigure 5-3 Flor ist and supermarket probabilities over age. Outlet Shares Across Incomes Outlet selection probabilities in the simulations in Figure5-6 show that the two lower income groups are more inclined to shop in supermarkets while the two upper income groups show a higher probability of buying in florists. This behavior suggests a

PAGE 76

62 shift from supermarkets to florists as income rises and likely reflects with increased purchasing power buyers move to the more valued added goods found more in florists 0.22 0.240.260.300.320.340.360.38Florist probability by expenditures 0.28 FemaleMaleSupermarket probability by transactions FemaleMale0.220.240.260.280.300.320.340.360.38Supermarket probability by expenditures Florist probability by transactions Figure 5-4 Florist and supermarket probabilities over gender. Recall from Chapter 1 that florists usually charge a premium for the creative value added for their products while supermarkets offer lower priced products. The graphs also show that the first and last income groups have probabilities closer to the mean of the average consumer. In the case of the under $25,000 income groups, a probability slightly above ditures and below the mean The second income group, $25,000 to $49,999, showed proba the group, $75,000 and above, shows small increases above the mean for florists and below the mean is seen only when considering supermarket expen for the rest of the measurements. bilities above average for supermarkets and below average for florists suggestingthat supermarkets are generally preferred. The third group, $50,000 to $74,999 showsopposite behavior, preferring florists for their fresh flower purchases. Finally, the last

PAGE 77

63 for supermarket. The distribution of probabilities in the four income groups suggests in the first and fourth group the demand for fresh flowers is more inelastic with respect that to income than the third and fourth groups. 0.000.100.200.300.400.500.60Florist probability by expenditures GiftSelfSupermarket probability by transactions GiftSelf0.000.100.200.300.400.500.60Supermarket probability by expenditures Florist probability by transactions Figure 5-5 Florist and supermarket probabilities over purpose. 0.220.240.280.300.320.340.38Florist probability by expenditures 0.260.36 unde25,000$25 t49,999$50 t74,99975 plus r $o $o $$Supermarket probability by transactions unde25,000$25 t49,999$50 t74,99975 plus0.220.240.260.300.320.360.38 r $o $o $$0.280.34Supermarket probability by expenditures Florist probability by transactionsFigure 5-6 Florist and supermarket probabilities over income.

PAGE 78

64 Outlet Shares over Flower Forms One of the hypotheses in Chapter 1 was that outlet selection decision was closely related to the form of the product. This wa s validated by the significance of the estimates t comes to this product form. Probably more than any other figure, the response to forms capture the e graph shows that the in Chapter 4. In addition, Chapter 3 showed that florists main source of business comprise the arrangement sector of the market while supermarkets focused on the non-arrangement sector. Figure5-7 presents the probabilities of selecting each outlet based on the product form with the forms being arrangements, non-arrangement, and flowering plants. As expected, when it comes to arrangements florists showed a higher probability (more than fifty percent) of being selected with supermarkets showing approximately the same probability for non-arrangements purchases. The probability distribution suggests a clear outlet selection decision for arrangements and non-arrangements. However, flowering/green plants probabilities are below the mean of the average consumer suggesting that the two outlets are less probable choices when i fundamentally differences between florists and supermarkets in terms of the product offerings. Combined Effect of Purpose and Form Going back to the estimates in Chapter 4, purpose and form were the two variables that had the greatest impact on both the decision to become a buyer and on the outlet selection. Showing the combined effects of these two important variables give additional insight into the expected outlet shares. In Figure5-8 the horizontal dotted lines represent the probability of outlet selection for the average consumer as initially estimated in Figure 5-2. The graph illustrates changes in the probabilities of outlet selection as different combinations of form and purpose are considered. Th

PAGE 79

65 combinations for gift buying have a h igher probability of selection than the combinations for se lf. Also, gift combinations were the only ones that showed above average probabilities of selecting florists, thus proving that gift buying weights considerably on the decision to choose florists. Furthermore, the arrangements sector was the onlycategory that showed probabilities above the mean for the three flower forms. 0.000.200.400.500.60 0.100.30Florist probability by expenditures ArrtsNon ArrtsFngSupermarket probability by transactions angemenangemenlorweri Angemenngemenlorwerin0.000.100.200.400.500.60 rratsNon ArratsFg0.30Supermarket probability by expenditures Florist probability by transactions ilities over form. Flowehe of rket Figure 5-7 Florist and supermarket probab ring/green houseplants showed probabilities well below the mean suggesting that such fresh flower form is less probable of being bought at florists when compared to tother two. Note that the patterns are almost identical when based on expenditures and transactions. Figure5-9 shows the probabilities of supermarket selection for the combinationpurpose and form. Unlike florists, the distributions of the probabilities for supermaselection show are more evenly distributed when considering purpose. Recall from

PAGE 80

66 Chapter 3 that non-arrangements were the biggest component of supermarket fresh flower demand. As expected, the consumers probability of selection was larger arrangements than for the other two product forms. Unlike florists, some combinations did not show a large increase in the probability of selection in terms of purpose. In words, self and gift displayed similar probabilities when compared against arra for non-other ngements, non-arrangements, and flowering/green plants. As expected, arrangements had the lowest probability of selection among the three product forms with supermarkets. 0.000.200.400.600.801.00Florist probability by expenditures AvgAvg AvArrangeArrangeArrangNonArgNonArgNonArFlwingFlwingFlwinFigure 5-8 Florist probabilities over purpose and form. Simulations by Seasons __Gg___e___g__Avg_Giftg_Self0.200.400.600.801.00Florist probability by transactions Seasonality was expected to influence outlet selection in fresh flowers consumption to the extent that calendar occasions could influence where one buyers things for those special occasions. However, the model estimates showed that seasonality was only significant when considering supermarkets. This is consistent with the distribution of Avgift SelfAvgGiftSelfAvgGiftSelf0.00

PAGE 81

67 0.000.400.601.00 0.200.80Supermarket probability by expenditures Avg _AvgAvg _Gift Avg_SelfArrange_AvgArrange_GiftArrange_SelfNonArg_AvgNonArg_GiftNonArg_SelfFlwing_AvgFlwing_GiftFlwing_Self0.000.200.400.6080 Figure 5-9 Supermarket probabilities over purpose and form. probabilities in Figure 5-10 where supermarkets show a higher variance relative to the means of the model. The graph also shows that supermarkets have a higher probability of outlet selection in the first four months of the year while florists experience the same in the fall months. The graphs show that supermarkets are more probable to be chosen for ll from Chapter 3 that demand fofresh flowers peaked in calendaralentines (February) and Moth 1.00Supermarket probability by transactions 0. Valentines Day relative to the average consumer. Reca r occasions particularly in V ers Day (May). The remaining calendar occasions do not affect the probability of outlet selection for florists and supermarkets. In fact, one could argue that the above-average probabilities seen in florists during Fall could be attributed to non-calendar occasions.

PAGE 82

68 0.220.240.260.280.300.320.340.360.38Florist probability by expenditures JanFebMarAprMayJunJulAugSepOctNovDecSupermarket probability by transactions JanFebMarAprMayJunJulAugSepOctNovDec0.220.240.260.280.300.320.340.360.38Supermarket probability by expenditures Florist probability by transactions Figure 5-10 Florist and supermarket probabilities over months. 0.220.240.28 0.260.300.340.360.38Under $25,000 dollars 0.32 JanFebMarAprMayJunJulAugSepOctNovDec$75,000 dollars and more JanFebMarAprMayJunJulAugSepOctNovDec0.220.240.280.300.320.340.38$50,000 to $74,999 dollars 0.260.36 $25,000 to $49,999 dollars n in lities of selection as income rises. The variation in transactions and Figure 5-11 Florist probabilities over income and months by expenditures. The last simulation combines income and months to show the monthly variatiothe probabi

PAGE 83

69 expenditures levels is minimal and so they provide essentially the same information. In this case expenditures levels were selected to discuss changes between the two variables. From a previous simulation it was clear that outlet selection shifts from supermarkets to florists as income rises. Figure5-11 shows the same behavior among income groups throughout the year. The first two income groups showed probabilities either below or near the mean. For these groups the lowest probabilities of selection coincided with the February, May and June months showing that these groups are less inclined to purchase in florists in calendar occasions. The probability of selecting florists increases in the third income group especially in non-calendar months and decreases in the fourth group. In fact, the fourth group behaves more closely to the average consumer depicting a relative homogeneous probability of selecting florists. 0.220.240.260.280.300.320.340.360.38Under $25,000 dollars JanFebMarAprMayJunJulAugSepOctNovDec$75,000 dollars and more JanFebMarAprMayJunJulAugSepOctNovDec0.220.240.260.280.300.320.340.360.38$50,000 to $74,999 dollars $25,000 to $49,999 dollars Figure 5-12 Supermarket probabilities over income and months by expenditures. Figure5-12 shows the combined probabilities of income and months to account for the seasonality effect of different income groups. Unlike the previous graph the variation

PAGE 84

70 in expenditures and transactions is clear. The first income group shows a probability of distribution above the mean from the January to April period then below the mean from May to September and then peaks back above the mean in the last three months. This distribution clear shows a higher probability in February coinciding with Valentinesand the lowest in May. The second income group shows above the mean probabilities except in May. The graph shows that summer months show the least probability of selecting supermarkets but still above to the probability for the average consumer. The third income group shows a different behavior showing above the mean probabilities for February and below the mean the rest of the year. The last group shows above probabilities on January, February and November and below the mean in the rest of the year. The graphs groups show that in the summ Day er months the probability of selecting the coefficients and by the difference between the variable that showed the least and the supermarkets decreases. It was expected that calendar occasions influence the probability of selecting supermarkets. However, only Valentines appears to influence the probability of selecting supermarkets. Rankings Factors Impacting the Outlet Shares In the previous figures a range of probabilities were shown and since the probabilities are comparable, they can be ranked in terms for the magnitude and direction of change. Figure5-13 shows the ranking of both the range of the variables as well as the absolute high and low expected market shares for florists. Recall from the model estimates that form and purpose were the variables that had the greatest impact on outlet selection decisions. In the florists case, the same variables that showed the biggest range within each variable categories. The range difference is influenced by the magnitude of variable that showed the greatest probability of outlet selection within each demographic

PAGE 85

71 and socioeconomic variable category. Form showed 43 percent outlet selection probability difference in expenditures between arrangements (highest probability) and flowering/green houseplants (lowest probability). In addition, purpose showed a 32 percent outlet selection probability in expenditures between gift (highest) and self (lowest). The ranking in terms of transactions was similar to that of expenditures. The rest of the variables showed probabilities differences of less than 10 percent denotinmore homogenous distribution between each of the divisions within those variables. What is most apparent in Figure 5-13 is the relative low level of importance of demographics relative to form and purpose when considering what truly impactlower probabilities of using florists, i.e., product form and purpose of buying. Figure 5-14 shows the variable ranking for supermarkets following the same criteria for florists except that the magnitude of change is somewhat less for supermarkets. Purpose was also the variable that showed the biggest range in outlet selection probabilities. Unli g a s the ke florists, gender and age were the variables that followed purpose in the varia in each of them. The g ble ranking with approximately 8 percent difference raph showed that purpose had the least probability range in the case of supermarkets. For supermarkets, form is the dominate variable impacting the likelihood for using supermarkets. Dynamics in the Outlet Share Coefficients Figure 5-15 shows the variation in the coefficient for the average consumer from 1993 to 2004. Recall from the previous chapter that 0 is the coefficient for the average consumer after the dummy trap was corrected in the model. The coefficient was calculated recursively from 1993 to find out if there were any other variables that would influence the model above and beyond the socioeconomic and demographic ones

PAGE 86

72 FormsAgeMonthIncomeGenderExpenditures Purpose 0.420.320.060.030.030.440.270.040.030.01 0.040.03 FormsPurposeMonthAgeGender0.000.100.200.300.400.500.60 IncomeTransactions Figur e 5-13 Variable rankings for florist. FormsGenderAgeIncomeExpenditures MonthPurpose 0.290.070.060.050.320.09 0.060.060.090.070.060.03 FormsAgeMonthTransactions GenderIncomePurpose 0.000.100.200.300.400.500.60Figure 5-14 Variable rankings for supermarket. considered in the study. The graph shows that when choosing an outlet for flowerpurchases, florists as an outlet choice is becoming less important for consumers. To the

PAGE 87

73 contrary, supermarkets are becoming more important to the average consumer as an outlet selection choice. With the exception of the age coefficients, the rest of the coefficients showed no significant variation over the period considered and further discussion was not deemed necessary. Furthermore, any changes seen in the coefficients seemed to affect florists and not supermarkets which validate the information presented in the previous graph. 1993-19961993-19971993-19981993-19991993-20001993-20011993-20021993-20031993-2004 0.000.100.200.30-0.20Florist Expenditure Coefficient -0.10 Intercept 1993-19961993-19971993-19981993-19991993-20001993-20011993-20021993-20031993-20040.000.100.200.30-0.10-0.20Florist Transactions Coefficient Intercept 1993-19961993-19971993-19981993-19991993-20001993-20011993-20021993-20031993-20040.000.100.200.30-0.10-0.20Supermarket Expenditure Coefficient Intercept 1993-19961993-19971993-19981993-19991993-20001993-20011993-20021993-20031993-20040.000.100.200.30-0.10-0.20Supermarket Transactions Coefficient Intercept Figure 5-15 Time Varying Coefficients for the Average Consumer

PAGE 88

CHAPTER 6 CONCLUSION Introduction This chapter prese nts the summary, findings and recommendations of the study. A the variables that had the greatest impact on the decision to choose either florists or supermarkets once they decided to buy fresh flowers. Estimates from the Tobit model were used to simulate the probability of selecting these two outlets. Finally, the parameters were recursively estimated to test if there were changes in the parameters over time. Parameter changes could suggest structural and/or preference changes not initially captured in the original Tobit estimates. Also, often there are not specific variables to brief summary of the previous chapters is presented outlining the major findings. Then, the conclusions of the chapter are discussed with a particular emphasis placed on whether the hypotheses from Chapter 1 were either validated or refuted. Finally, the implications and limitations of the study and the recommendations for further research are presented. Overview of Outlet Analyses The main objective of this study was to analyze outlet market share changes given a change in demographic and socioeconomic variables associated with fresh flower consumers. By focusing in the fresh flower section of the market, the study covers approximately 90 percent of all indoor flowers. A two-step estimation model was used to describe the outlet selection process faced by buyers in the fresh flower industry. In the first stage decision, the model used a probit model to differentiate between buyers and non-buyers of flowers. In the second stage, the analyses used a Tobit model to estimate 74

PAGE 89

75 measure these changes and the use of time-varying parameters is an indirect way to test for such changes. The flower industry in the United Statesflowers, flowering and green house plants, and artificial. This study focuses on the fresh flower portion of the market since it c 90 per cent of the flower industry. Basedoutlet share models. Florists a the study for the relative importance in their resp was grouped into three categories: cutomprises nearly on the major outlets used by consumers, the outlets were divided into 4 categories: mass merchandising, specialty, internet retail, and others. The data showed that supermarkets comprised the majority of mass merchandising purchases with florists capturing most of the purchases in the specialty category. Data for the retail internet purchases were not available until the year 2000 and, hence, were not included in the nd supermarkets were chosen in ective categories but also because in the last decade major structural changes have occurred in the industry as described by market share changes particularly in these two outlets. Chapter 1 introduced the problematic situation and the major hypotheses of this study focused on the changes in market shares for both outlets. One of the strengths of the study is that the sample used was drawn from an extensive database which covered many aspects of the flower demand in the US. The database used is maintained by Ipsos, a private company who along with the National Panel Diary (NPD) collected the information from approximately 9,000 demographically balanced households every two weeks through the use of consumer diaries. The diaries included comprehensive information on actual purchases recording flower type, outlet selection, number of transactions, and occasions.

PAGE 90

76 Chapter 2 presented a literature review that was divided in three parts covering thearticles on choice t heory and consumer preferences, market share theories, and econoboth nd ted s alternatives for estimation. Unlike the original Heck the range of variable values. Purpose and form were combined metric models for censored data. Chapter 3 presented an overview of the fresh flower industry for the years from 1992 through 2004. The chapter showed the relative change in market shares among outlets using expenditures and transaction market share levels based on cut-flowers and flowering/green house plants. The chapter covered seasonal and yearly trends over florists and supermarkets, as well, as changes in arrangements and non-arrangements. The chapter showed the importance of florists asupermarkets in the flower industry and the relative changes in market shares over thestudy period. Chapter 4 explained the theoretical framework and the model specification to model consumer behavior in the fresh flower industry. The model specification starwith the neo-classical utility maximization theory and then explained the nature and problems of censored data with it man two-stage decision process which used a probit and a ordinary least squares (OLS) procedure, this study used a tobit model to account for the corner solutions previously discussed. In the first stage probit model, which assumed a significance levelof 95 percent, the estimates showed that purpose and form were the two variables that had the greatest impact on the decision to become a buyer. The same variables hadgreatest impact on the second stage decision to choose either florists or supermarkets for their flower purchases. Chapter 5 presented simulations drawn from the model estimates to simulate expected outlet shares over a

PAGE 91

77 to shod ral tatistically significant with the significance test being measured against the averamost on me 9 eatest impact on the decision to buy in either florists or super Recall from Chapter 3 that florists and supermarkets specialized in different sectors of the w the effect of both variables on the outlet use. A ranking of the variables from the largest to smallest impacts on the probability of outlet selection was presenteFinally, a time recursive model was used to determine if there were underlying structuchanges taking place within the outlets. Major Outlet Selection Conclusions Estimates from the two-stage model showed that purpose and form were the variables having the greatest impact on both the decision to become a buyer and on thedecision to choose either florists or supermarkets for their purchases. Assuming a significance level of 95 percent, almost all variables included in the probit and tobit models were s ge consumer. The probit estimates showed purpose, gender, and form were the importance factors impacting the likelihood of becoming a buyer. The combination of females buying non-arrangements for gifts had the largest impact on market penetratior attracting buyers. Among the variables that showed the least probability to becobuyers were people of less than 25 years of age having an income of $50,000 to $74,99and buying flower arrangements. In general, the majority of the variables were statistically significant given the mentioned confidence level. Seasonality showed the typical decline in the likelihood of buying flowers in the later half of each year, thus again highlighting the seasonal problem found throughout the flower industry. The second stage tobit model estimates showed that purpose and form were the variables that had the gr markets. The estimates showed a marked difference between the two outlets interms of the combination of the variables that increased their probability of selection.

PAGE 92

78 industry as described by their market share levels. The estimated coefficients for the tobmodel described the same level of market segmentation between the two outlets. Foexample, in the florists case the variables that had the greatest impact on outlet decision such as the second age group from 40 to 5 it r 4 years old, females, gift, and arrangements are the same variables that haose supermarkets excepthird uced s ults hat nd even a switce dicators s the least impact on the decision to cho t for females who are important on both outlets. In the supermarkets case, the age group from 40 to 54 years of age, non-arrangements and buying for self use had the greatest positive impact on consumer outlet selection decision. The result from theestimates reflected the market share conditions explained in Chapter 3 but also introdsome of the variables that have influenced fresh flower demand in the last few years. The demographic and socioeconomic variables used in the models therefore explained part of the difference in market shares between the two outlets. Starting from Chapter 3 it is apparent that the fresh flower industry was experiencing major restructuring at the retail level. Furthermore, when analyzing floristand supermarkets alone the models show that over the past decade florists have lost market shares while supermarkets have improved their market position. The resindicated that florists main source of business falls in the arrangement sector of the market while supermarkets are in the non-arrangement sector. The results also show t1998 was a turning point for both outlets either widening the gap among them a hing of market share position relative to other outlets. The statistically significancof the estimates show that the variables considered for the study were accurate inof what the flower industry has been experiencing in recent years.

PAGE 93

79 The perceived characteristics of the product play a major role in fresh flower demand as described by the purpose and form variables. The fact that both outlets specialize in different segments of the market with statistically significant coefficients associated with them implies that a degree of concentration in a sector could be achieved by product differentiation. Recall from Chapter 1 that florists charge a premium for value added to their products in the form of arrangements and supermarkets focus on large quantities of non-arrangements. Since florists share of the market remain essentiallythe same in the arrangement section, the systematic market share loss in the non-arrangement section could be the reason why florists market share levels have declined over time. In addition, the increasing market share in the non-arrangement section of the market could be the reason why supermarkets have experienced a steady increasing in total market participation. However, it is important to clar the ify that the results of this study do noRecall that er nd supermarkets based on both expenditures and transactions. The probability means for florists were 28 t imply that florists share loss has been totally captured by supermarkets. while the study focused on two outlets in particular all the outlets, some of the declines inmarket shares could be reflected in the smaller outlets not included in the modeling efforts. Important to the overall analysis was the hypotheses that rising incomes and othdemographics could impact the types of outlets used. Drawing from the outlet share model estimates, simulations where used to explicitly show the range of probabilities of selecting florists or supermarkets as each variable was considered. In Chapter 5, each simulation was completed by adjusting the variables relative to the mean of all the other variables with the expected market shares being shown for florists a

PAGE 94

80 percefor ntly n et, the gender effect on supermarkets is more than twice as great for superd re nt in expenditures and 25 percent for transactions while the probability means supermarkets were 27 percent in expenditures and 28 percent in transactions. As ageneral rule variations in income had a greater impact on using supermarkets than for using florists and generally the florists share gains while the supermarket shares decline as income increases. When considering age, the simulations showed that the probability of selecting florists increased until the third age group 40 to 54 years of age and then dropped for people of 55 years of age or older. In direct contrast, supermarket shares consisteincreased over the age groups. Purpose of buying showed marked differences for the two outlets. The fact that florists main source of business comes from gift buying and supermarkets from self buying were apparent from the distributions in Chapter 3. However, in the simulation over purpose and interesting fact arises in that supermarkets probability of selection is approximately the same as of the average consumer. When it comes to gender, the simulations show that females have a higher probability of choosing both outlets thamales. Y markets compared with florists. The variable rankings in Chapter 5 showed that purpose and form were the two factors showing the biggest variation within their divisions for florists with approximately 40 and 30 percent, respectively. The ranking shows the variables that hathe biggest range from the least and most probable selection. Interestingly, the variables form with 30 percent and gender to a lesser degree with approximately 10 percent wethe ones that showed the biggest variation when considering supermarkets.

PAGE 95

81 In general, the statistically significance of the estimates provided important information to modeling the fresh flower outlet demand given a set of socioeconomicdemographic variables. However, as it is sometimes the case in econometric analysis some variables are not considered in the study because of data limitations among others. To account for this, a and time recursive models were estimated to see if the variables consi is was fairly stable. In fact, Chapter 5 only presents the time vt e flowers. y of studies to compare the results of this study. Howeh vegetable better its position within the industry, efforts were made to present the information in the most dered in the study became more or less important in modeling fresh flower outletdemand. The results indicated that the coefficients for the majority of the variables considered did not vary appreciably over time suggesting that the importance of the variables included in the analys arying coefficients for the average consumer for both florists and supermarkets. Iis safe to conclude that when considering the two outlets only florists appear to becomless important to the outlet selection decision faced by consumers. Limitations While many studies focus on pricing considerations and volume of sales as anapproach to estimate market share changes in an industry, this study assumes a relationship between the characteristics of the buyer and the demand for fresh Unfortunately, there is not a wide arra ver, the methodology and the results are similar to several studies in agricultural commodities such as away-from-home food consumption, cigarette, and fresconsumption. Because a two-stage process was used to model first the decision to become a buyer and then the propensity to buy in either florists or supermarkets, the results from this study could be of interest from an outlet category or an industry point ofview. While any member of the flower industry could use this information to

PAGE 96

82 imparshare stry. forts fit producers and consumers and may facilitate the decis policy makers. It is important to inves y the e r tial manner possible. Therefore, this study does not provide any particular suggestions for improvements to any particular outlet category and only analyses and tries to explain the current trend in market share levels within the flower industry. A limitation of the study was that the simulations chapter only dealt with the combinations of variables that had the greatest impact on outlet selection yielding higher market levels. One interesting aspect for further research could be to combine the variables thatwould represent the least market share level and analyze its implications to the induThe negative impacts highlight targets for potential promotions or other marketing efto offset negative influences. Recommendations Information of the impact of the relative impact of various socioeconomic factors on the consumption of fresh flowers discussed in this study can bene ion making of tigate the effect of socioeconomic and demographic variables on the decision toconsume as a proxy to future market shares changes. The lost in market shares bflorists sector continues to be a troubling factor for that portion of the industry and anytime that helps them to counter the loss in shares would be beneficial. The negativefactors impacting florists provide areas for marketing and promotion efforts that need tobe explored in more detail. Also, it is important to remember that this study deals only with household data and that commercial transactions were not incorporated into thmodel. It is likely that florists capture a larger share of the commercial market but ouanalysis does not specifically show that. Therefore, incorporating commercial data into the analysis would be beneficial. Realistically, however, getting the commercial data isvery difficult and often impossible. One of the strengths of the study was its ability to

PAGE 97

83 pinpoint the divisions within each variable, for example the low probability of sein the first age group, in which promotions or advertisement are needed to stimulate demand. Another is that by understanding what products comprise the main source ofbusiness of both outlets we can understand the effect that generic or brand promotions would have in the flower industry. An important extension would be to specifically design the marketing strategies using the targets suggested with this research. That has not been done since it was beyond the scope of the study. lection

PAGE 98

APPENDIX A INDUSTRY OVERVIEW 0.000.200.80 0.400.601.00Market shares Expenditures Florists Other 19931994199519961997199819992000200120022003Years0.000.200.400.600.801.00 Transactions FloristsOther Figure A-1 Percent of yearly market shares for specialty based in cut flowers by expenditures and transactions. Source: AFE and Ipsos-NPD group 84

PAGE 99

85 0.000.20 0.400.600.801.00Market shares Expenditures FloristsOther 19931994199519961997199819992000200120022003Years0.000.200.400.600.801.00 Transactions FloristsOther Figure A-2 Percent of yearly market shares for specialty based in flowering/green house plants by expenditures and transactions. Source: AFE and Ipsos-NPD group 0.000.200.400.601.00Market shares Expenditures Supermarkets 0.80 Warehouses/Price ClubOther 19931994199519961997199819992000200120022003Years0.000.200.400.600.801.00 Transactions SupermarketsWarehouses/Price ClubOther Figure A-3 Percent of yearly market shares for mass merchandising based in cut flowers by expenditures and transactions. Source: AFE and Ipsos-NPD group.

PAGE 100

86 0.000.200.400.600.80 1.00Market shares Expenditures SupermarketsWarehouses/Price ClubOther 19931994199519961997199819992000200120022003Years0.000.200.400.600.801.00 Transactions SupermarketsWarehouses/Price ClubOther F igure A-4 Percent of yearly market shares for mass merchandising based in flowering/green house plants by expenditures and transactions. Source: AFE and Ipsos-NPD group. 0.000.200.400.600.801.00Market shares Expenditures SpecialtyMass MerchandisingRetail InternetOther JanFebMarAprMayJunJulAugSepOctNovDec 0.000.200.400.600.801.00 Transactions SpecialtyMass MerchandisingRetail InternetOther Months Percent of monthly market shares based in cut flowers by expenditures Figure A-5 and transactions. Source: AFE and Ipsos-NPD group.

PAGE 101

87 0.000.200.400.600.801.00Market shares Expenditures SpecialtyMass MerchandisingRetail InternetOther JanFebMarAprMayJunJulAugSepOctNovDec 0.000.200.400.600.801.00 Transactions SpecialtyMass MerchandisingRetail InternetOther Months Percent of monthly market shares based in flowering/green house plants by expenditures and transa Figure A-6ctions. Source: AFE and Ipsos-NPD group. 0.000.200.400.600.801.00Market shares Expenditures FloristsOther JanFebMarAprMayJunJulAugSepOctNovDec Months 0.000.200.400.600.801.00 Transactions FloristsOther Figure A-7in cut flowers by expenditures and transactions. Source: AFE and Ipsos-NPD group. Percent of monthly specialty market shares based

PAGE 102

88 0.000.200.400.600.801.00Market shares Expenditures FloristsOther JanFebMarAprMayJunJulAugSepOctNovDec Months 0.000.200.400.600.801.00 Transactions FloristsOther Figure A-4n house plants by expenditures and transactions. Source: AFE and Ipsos-NPD group. Percent of monthly specialty market shares based in flowering/gree 0.000.200.400.600.801.00Market shares Expenditures SupermarketsWarehouses/Price ClubOther JanFebMarAprMayJunJulAugSepOctNovDecMonths0.000.200.400.600.801.00 Transactions SupermarketsWarehouses/Price ClubOther Figure A-5 Percent of monthly mass merchandising market shares based in cut fby expenditures and transactions. Source: AFE and Ipsos-NPD grou lowers p.

PAGE 103

89 0.000.200.400.600.801.00Market shares Expenditures SupermarketsWarehouses/Price ClubOther JanFebMarAprMayJunJulAugSepOctNovDecMonths0.000.200.400.600.801.00 Transactions SupermarketsWarehouses/Price ClubOther Figure A-6 Percent of monthly mass merchandising market shares based in flowering/green house plants cut flowers by expenditures and transactions. Source: AFE and Ipsos-NPD group. 0.000.200.400.600.801.00Market Shares Expenditures FloristsSupermarkets JanFebMarAprMayJunJulAugSepOctNovDecMonths0.000.200.400.600.801.00 Transactions FloristsSupermarkets Figure A-7 Percent of monthly market shares in cut flowers for florists and supermarkets by expenditures and transactions. Source: AFE and Ipsos-NPD group.

PAGE 104

90 0.000.200.400.600.801.00Market Shares Expenditures FloristsSupermarkets JanFebMarAprMayJunJulAugSepOctNovDecMonths0.000.200.400.600.801.00 Transactions FloristsSupermarkets Figure A-8 Percent of monthly market shares in flowering/green house plantand supermarkets by expenditures and transactions. Source: AFE and Ipsos-NPD group. s for florists 0.000.200.400.600.801.00Market Shares Expenditures FloristsSupermarkets JanFebMarAprMayJunJulAugSepOctNovDec 0.000.200.400.600.801.00 Transactions FloristsSupermarkets Months Percent of monthly market shares in arrangements for florists and Figure A-9supermarkets by expenditures and transactions. Source: AFE and Ipsos-NPD group.

PAGE 105

91 0.000.200.400.600.801.00Market Shares Expenditures FloristsSupermarkets JanFebMarAprMayJunJulAugSepOctNovDecMonths0.000.200.400.600.801.00 Transactions FloristsSupermarkets Figure A-10 Percent of monthly market shares in non-arrangements for florists and supermarkets by expenditures and transactions. Source: AFE and Ipsos-NPDgroup. 0.650.35 0.970.03 0.170.290.220.32 0.070.280.370.28 0.000.200.400.600.801.00Market Shares Expenditures AgeIncomePurposeGender 0.650.35 0.940.06 0.210.320.210.27 0.080.300.360.26 Under 25 years25 to 40 years41 to 54 years55 years and overUnder $25,000$25,000 to $49,999$50,000 to $74,999$75,000 and overGiftSelfFemaleMaleDemographics0.000.200.400.600.801.00 Transactions AgeIncomePurposeGender Figure A-11 Distribution of market shares based in cut flowers arrangements bexpenditures and transactions. Source: AFE and Ipsos-NPD group. y

PAGE 106

92 0.600.40 0.730.27 0.190.270.220.32 0.090.270.370.27 0.000.200.400.600.801.00Market Share Expenditures AgeIncomePurposeGender 0.670.33 0.610.39 0.220.290.200.29 0.080.260.370.29 Under 25 years25 to 40 years41 to 54 years55 years and overUnder $25,000$25,000 to $49,999$50,000 to $74,999$75,000 and overGiftSelfFemaleMale 0.000.200.400.600.801.00 Transactions AgeIncomePurposeGender Demographics Figure A-12 Distribution of market shares based in cut flowers non-arrangements by expenditures and transactions. Source: AFE and Ipsos-NPD group. 0.090.270.370.27 0.070.280.370.28 0.070.280.370.27 0.000.200.400.600.801.00Market Shares Expenditures Cut FlowersArrangementsNon-arrangements 0.080.260.370.29 0.080.300.360.26 0.080.270.360.29 Under 25 years25 to 40 years41 to 54 years55 years and overUnder 25 years25 to 40 years41 to 54 years55 years and overUnder 25 years25 to 40 years41 to 54 years55 years and overAge0.000.200.400.600.801.00 Transactions Cut FlowersArrangementsNon-arrangements Figure A-13 Distribution of market shares based on age by expenditures and transactions. Source: AFE and Ipsos-NPD group.

PAGE 107

93 0.190.270.220.32 0.170.290.220.32 0.180.280.220.32 0.000.200.400.600.801.00Market Shares Expenditures Cut FlowersArrangementsNon-arrangements 0.220.290.200.29 0.210.320.210.27 0.220.290.200.29 Under $25,000$25,000 to $49,999$50,000 to $74,999$75,000 and overUnder $25,000$25,000 to $49,999$50,000 to $74,999$75,000 and overUnder $25,000$25,000 to $49,999$50,000 to $74,999$75,000 and overIncome 0.000.200.400.600.801.00 Transactions Cut FlowersArrangementsNon-arrangements Figure A-14 Distribution of market shares based on income by expenditures and transactions. Source: AFE and Ipsos-NPD group. 0.730.27 0.970.03 0.860.14 0.000.200.400.600.801.00Market Shares Expenditures Cut FlowersArrangementsNon-arrangements 0.610.39 0.940.06 0.680.32 GiftSelfGiftSelfGiftSelfPurpose0.000.200.400.600.801.00 Transactions Cut FlowersArrangementsNon-arrangements Figure A-15 Distribution of market shares based on purpose by expenditures and transactions. Source: AFE and Ipsos-NPD group.

PAGE 108

94 0.600.40 0.650.35 0.630.37 0.000.200.400.600.801.00Market Shares Expenditures Cut FlowersArrangementsNon-arrangements 0.670.33 0.650.35 0.660.34 FemaleMaleFemaleMaleFemaleMale Gender 0.000.200.400.600.801.00 Transactions Cut FlowersArrangementsNon-arrangements Figure A-1 expenditures and transactions. Source: AFE and Ipsos-NPD group. 6 Distribution of market shares based on gender by 0.410.370.010.070.090.010.04 0.610.200.010.080.040.020.03 0.000.200.400.600.801.00Market Shares Under 25 years ExpendituresTransactions 0.360.410.020.090.070.010.04 0.610.190.010.100.030.030.03 FloristsSupermarketsWarehouse/Price clubAll Retail InternetOtherOther SpecialityOther MassFloristsSupermarketsWarehouse/Price clubAll Retail InternetOtherOther SpecialityOther Mass Age 0.000.200.400.600.801.00 25 to 40 years ExpendituresTransactions Figure A-1in cut flowers based on age (first and second group) by expenditures and transactions. Source: AFE and Ipsos-NPD group. 7 Distribution of market shares for specific outlets

PAGE 109

95 0.060.250.010.210.390.000.06 0.160.200.010.270.300.020.05 0.000.200.400.600.801.00Market Shares Under 25 years ExpendituresTransactions 0.070.250.000.210.410.000.06 0.190.160.010.270.280.010.07 FloristsSupermarketsWarehouses/Price ClubOther SpecialtyOther Mass MerchandisingRetail InternetOtherFloristsSupermarketsWarehouses/Price ClubOther SpecialtyOther Mass MerchandisingRetail InternetOtherAge0.000.200.400.600.801.00 25 to 40 years ExpendituresTransactions Figure A-18 Distribution of market shares for specific outlets in flowering/green house plants based on age (first and second group) by expenditures and transSource: AFE and Ipsos-NPD group. actions 0.320.450.030.100.060.010.05 0.600.210.020.090.030.020.04 0.000.200.400.600.801.00Market Shares 41 to 54 years ExpendituresTransactions 0.280.510.020.090.060.000.05 0.600.240.010.070.030.010.04 FloristsSupermarketsWarehouses/Price ClubOther SpecialtyOther Mass MerchandisingRetail InternetOtherFloristsSupermarketsWarehouses/Price ClubOther SpecialtyOther Mass MerchandisingRetail InternetOtherAge0.000.200.400.600.801.00 55 years and over ExpendituresTransactions Figure A-19 Distribution of market shares for specific outlets in cut flowers based on age (third and fourth group) by expenditures and transactions. Source: AFE and Ipsos-NPD group.

PAGE 110

96 0.080.280.010.210.350.000.06 0.240.190.010.260.230.010.07 0.000.200.400.600.801.00Market Shares 41 to 54 years ExpendituresTransactions 0.080.310.010.220.320.000.06 0.230.210.010.260.220.010.06 FloristsSupermarketsWarehouses/Price ClubOther SpecialtyOther Mass MerchandisingRetail InternetOtherFloristsSupermarketsWarehouses/Price ClubOther SpecialtyOther Mass MerchandisingRetail InternetOtherAge0.000.200.400.600.801.00 55 years and over ExpendituresTransactions Figure A-20 Distribution of market shares for specific outlets in flowering/green house plants based on age (third and fourth group) by expenditures and transactions. Source: AFE and Ipsos-NPD group. 0.290.480.020.080.070.010.05 0.590.220.020.080.030.020.04 0.000.200.400.600.801.00Market Shares Female ExpendituresTransactions 0.390.380.020.100.060.010.03 0.630.200.010.090.030.020.02 FloristsSupermarketsWarehouses/Price ClubOther SpecialtyOther Mass MerchandisingRetail InternetOtherFloristsSupermarketsWarehouses/Price ClubOther SpecialtyOther Mass MerchandisingRetail InternetOther 0.000.200.400.600.801.00 Male ExpendituresTransactions Gender Figure A-21 Distribution of market shares for specific outlets in cut flowers based on gender by expenditures and transactions. Source: AFE and Ipsos-NPD group.

PAGE 111

97 0.080.280.010.210.360.000.06 0.220.190.010.250.250.010.07 0.000.200.400.600.801.00Market Shares Female ExpendituresTransactions 0.100.260.010.240.330.000.05 0.220.170.010.300.240.010.06 FloristsSupermarketsWarehouses/Price ClubOther SpecialtyOther Mass MerchandisingRetail InternetOtherFloristsSupermarketsWarehouses/Price ClubOther SpecialtyOther Mass MerchandisingRetail InternetOtherGender0.000.200.400.600.801.00 Male ExpendituresTransactions Figure A-22 Distribution of market shares for specific outlets in flowering/green house plants based on gender by expenditures and transactions. Source: AFE and Ipsos-NPD group. 0.340.420.010.070.100.000.05 0.620.210.010.070.050.010.03 0.000.200.400.600.801.00Market Shares Under $25,000 ExpendituresTransactions 0.330.470.010.080.070.000.04 0.600.230.010.080.030.020.03 FloristsSupermarketsWarehouse/Price clubAll Retail InternetOtherOther SpecialityOther MassFloristsSupermarketsWarehouse/Price clubAll Retail InternetOtherOther SpecialityOther MassIncome0.000.200.400.600.801.00 $25,000 to $49,999 ExpendituresTransactions Figure A-23 Distribution of market shares for specific outlets in cut flowers based on income (first and second group) by expenditures and transactions. Source: AFE and Ipsos-NPD group.

PAGE 112

98 0.070.290.000.200.380.000.06 0.210.210.000.240.270.010.05 0.000.200.400.600.801.00Market Shares Under $25,000 ExpendituresTransactions 0.070.270.010.200.390.000.06 0.210.200.010.240.270.000.07 FloristsSupermarketsWarehouses/Price ClubOther SpecialtyOther Mass MerchandisingRetail InternetOtherFloristsSupermarketsWarehouses/Price ClubOther SpecialtyOther Mass MerchandisingRetail InternetOtherIncome0.000.200.400.600.801.00 $25,000 to $49,999 ExpendituresTransactions Figure A-24 Distribution of market shares for specific outlets in flowering/green house plants based on income (first and second group) by expenditures and transactions. Source: AFE and Ipsos-NPD group. 0.340.430.030.090.050.010.05 0.640.190.020.070.020.020.03 0.000.200.400.600.801.00 M arket Shares $50,000 to $74,999 ExpendituresTransactions 0.290.470.030.120.040.010.04 0.570.210.020.100.020.030.04 FloristsSupermarketsWarehouses/Price ClubOther SpecialtyOther Mass MerchandisingRetail InternetOtherFloristsSupermarketsWarehouses/Price ClubOther SpecialtyOther Mass MerchandisingRetail InternetOtherIncome0.000.200.400.600.801.00 $75,000 and over ExpendituresTransactions Figure A-25 Distribution of market shares for specific outlets in cut flowers based on income (third and fourth group) by expenditures and transactions. Source: AFE and Ipsos-NPD group.

PAGE 113

99 0.080.270.010.220.340.000.07 0.220.180.010.280.230.010.07 0.000.200.400.600.801.00Market Shares $50,000 to $74,999 ExpendituresTransactions 0.090.280.010.240.300.000.06 0.240.170.010.290.210.010.07 FloristsSupermarketsWarehouse/Price clubAll Retail InternetOtherOther SpecialityOther MassFloristsSupermarketsWarehouse/Price clubAll Retail InternetOtherOther SpecialityOther MassIncome0.000.200.400.600.801.00 $75,000 and over ExpendituresTransactions Figure A-26 Distribution of market shares for specific outlets in flowering/green house plants based on income (third and fourth group) by expenditures and transactions. Source: AFE and Ipsos-NPD group. 0.430.360.020.080.060.010.04 0.670.160.010.080.030.020.03 0.000.200.400.600.801.00Market Shares Gift ExpendituresTransactions 0.100.630.030.110.070.000.06 0.200.530.040.120.060.000.05 FloristsSupermarketsWarehouses/Price ClubOther SpecialtyOther Mass MerchandisingRetail InternetOtherFloristsSupermarketsWarehouses/Price ClubOther SpecialtyOther Mass MerchandisingRetail InternetOtherPurpose0.000.200.400.600.801.00 Self ExpendituresTransactions Figure A-27 Distribution of market shares for specific outlets in cut flowers based on purpose by expenditures and transactions. Source: AFE and Ipsos-NPD gro up

PAGE 114

100 0.170.330.010.170.260.000.06 0.390.200.010.190.140.010.05 0.000.200.400.600.801.00Market Shares Gift ExpendituresTransactions 0.030.250.010.240.410.000.06 0.050.180.010.340.340.000.07 FloristsSupermarketsWarehouses/Price ClubOther SpecialtyOther Mass MerchandisingRetail InternetOtherFloristsSupermarketsWarehouses/Price ClubOther SpecialtyOther Mass MerchandisingRetail InternetOtherPurpose0.000.200.400.600.801.00 Self ExpendituresTransactions Figure A-28 Distribution of market shares for specific outlets in flowering/green house plants based on purpose by expenditures and transactions. Source: AFIpsos-NPD group. E and

PAGE 115

APPENDIX B TIME RECURSIVE COEFFICIENTS 1993-19961993-19971993-19981993-19991993-20001993-20011993-20021993-20031993-20040.000.100.200.30-0.10-0.20Florist Expenditure Coefficient Under 25 years 1993-19961993-19971993-19981993-19991993-20001993-20011993-20021993-20031993-20040.000.100.200.30-0.10-0.20Florist Transactions Coefficient Under 25 years 1993-1996 1993-19971993-19981993-19991993-20001993-20011993-20021993-20031993-20040.000.200.30-0.10-0.20Supermarket Expenditure Coefficient Under 25 years 1993-19961993-19971993-19981993-19991993-20001993-20011993-20021993-20031993-20040.200.30-0.10-0.20Supermarket Transactions Coefficient 0.10 0.000.10 Under 25 years Figure B-1 Time recursive parameters for the under 25 years of age group. 1993-19961993-19971993-19981993-19991993-20001993-20011993-20021993-20031993-20040.000.100.200.30-0.10-0.20Florist Expenditure Coefficient 25 to 39 years 1993-19961993-19971993-19981993-19991993-20001993-20011993-20021993-20031993-20040.000.100.200.30-0.10-0.20Florist Transactions Coefficient 25 to 39 years 1993-19961993-19971993-19981993-19991993-20001993-20011993-20021993-20031993-20040.000.100.200.30-0.10-0.20Supermarket Expenditure Coefficient 25 to 39 years 1993-19961993-19971993-19981993-19991993-20001993-20011993-20021993-20031993-20040.000.100.200.30-0.10-0.20Supermarket Transactions Coefficient 25 to 39 years Figure B-2 Time recursive parameters for the 25 to 39 years of age group. 101

PAGE 116

102 1993-19961993-19971993-19981993-19991993-20001993-20011993-20021993-20031993-2004-0.10-0.20 0.000.100.200.30Florist Expenditure Coefficient 40 to 54 years 1993-19961993-19971993-19981993-19991993-20001993-20011993-20021993-20031993-2004 0.20 0.000.10 0.30-0.10-0.20Florist Transactions Coefficient 40 to 54 years 1993-19961993-19971993-19981993-19991993-20001993-20011993-20021993-20031993-20040.000.100.200.30-0.10-0.20Supermarket Expenditure Coefficient 40 to 54 years 1993-19961993-19971993-19981993-19991993-20001993-20011993-20021993-20031993-20040.000.100.200.30-0.10-0.20Supermarket Transactions Coefficient 40 to 54 years Figure B-3 Time recursive parameters for the 40 to 54 years of age group. 1993-19961993-19971993-19981993-19991993-20001993-20011993-20021993-20031993-20040.000.100.200.30-0.10-0.20 Florist Expenditure Coefficient 55 and more years 1993-19961993-19971993-19981993-19991993-20001993-20011993-20021993-20031993-20040.000.100.200.30-0.10-0.20 Florist Transactions Coefficient 55 and more years 1993-19961993-19971993-19981993-19991993-20001993-20011993-20021993-20031993-20040.000.100.200.30-0.10-0.20Supermarket Expenditure Coefficient 55 and more years 1993-19961993-19971993-19981993-19991993-20001993-20011993-20021993-20031993-20040.000.100.200.30-0.10-0.20Supermarket Transactions Coefficient 55 and more years Figure B-4 Time recursive parameters for the 55 and more years of age gro up.

PAGE 117

103 1993-19961993-19971993-19981993-19991993-20001993-20011993-20021993-20031993-20040.000.100.200.30-0.10-0.20Florist Expenditure Coefficient Females 1993-19961993-19971993-19981993-19991993-20001993-20011993-20021993-20031993-20040.000.100.200.30-0.10-0.20Florist Transactions Coefficient Females 1993-19961993-19971993-19981993-19991993-20001993-20011993-20021993-20031993-20040.000.100.200.30-0.10-0.20Supermarket Expenditure Coefficient Females 1993-19961993-19971993-19981993-19991993-20001993-20011993-20021993-20031993-20040.000.100.200.30-0.10-0.20Supermarket Transactions Coefficient Females Figure B-5 Time recursive parameters for females. 1993-19961993-19971993-19981993-19991993-20001993-20011993-20021993-20031993-20040.000.100.200.30-0.10-0.20Florist Expenditure Coefficient Males 1993-19961993-19971993-19981993-19991993-20001993-20011993-20021993-20031993-20040.000.100.200.30-0.10-0.20Florist Transactions Coefficient Males 1993-19961993-19971993-19981993-19991993-20001993-20011993-20021993-20031993-20040.000.100.200.30-0.10-0.20Supermarket Expenditure Coefficient Males 1993-19961993-19971993-19981993-19991993-20001993-20011993-20021993-20031993-20040.000.100.200.30-0.10-0.20Supermarket Transactions Coefficient Males Figure B-6 Time recursive parameters for males.

PAGE 118

104 1993-19961993-19971993-19981993-19991993-20001993-20011993-20021993-20031993-20040.000.200.40-0.20-0.40Florist Expenditure Coefficient Gift 1993-19961993-19971993-19981993-19991993-20001993-20011993-20021993-20031993-20040.000.200.40-0.20-0.40Florist Transactions Coefficient Gift 1993-19961993-19971993-19981993-19991993-20001993-20011993-20021993-20031993-20040.000.200.40-0.20-0.40Supermarket Expenditure Coefficient Gift 1993-19961993-19971993-19981993-19991993-20001993-20011993-20021993-20031993-20040.000.200.40-0.20-0.40Supermarket Transactions Coefficient Gift Figure B-7 Time recursive parameters for gift. 1993-19961993-19971993-19981993-19991993-20001993-20011993-20021993-20031993-20040.000.200.40-0.20-0.40Florist Expenditure Coefficient Self 1993-19961993-19971993-19981993-19991993-20001993-20011993-20021993-20031993-20040.000.200.40-0.20-0.40Florist Transactions Coefficient Self 1993-19961993-19971993-19981993-19991993-20001993-20011993-20021993-20031993-20040.000.200.40-0.20-0.40Supermarket Expenditure Coefficient Self 1993-19961993-19971993-19981993-19991993-20001993-20011993-20021993-20031993-20040.000.200.40-0.20-0.40Supermarket Transactions Coefficient Self Figure B-8 Time recursive parameters for self.

PAGE 119

105 1993-19961993-19971993-19981993-19991993-20001993-20011993-20021993-20031993-20040.000.100.200.30-0.10-0.20Florist Expenditure Coefficient Under $25,000 1993-19961993-19971993-19981993-19991993-20001993-20011993-20021993-20031993-20040.000.100.200.30-0.10-0.20Florist Transactions Coefficient Under $25,000 1993-19961993-19971993-19981993-19991993-20001993-20011993-20021993-20031993-20040.000.100.200.30-0.10-0.20Supermarket Expenditure Coefficient Under $25,000 1993-19961993-19971993-19981993-19991993-20001993-20011993-20021993-20031993-20040.000.100.200.30-0.10-0.20Supermarket Transactions Coefficient Under $25,000 Figure B-9 Time recursive parameters for the under $25,000 income group. 1993-19961993-19971993-19981993-19991993-20001993-20011993-20021993-20031993-20040.000.100.200.30-0.10-0.20Florist Expenditure Coefficient $25,000 to $49,999 1993-19961993-19971993-19981993-19991993-20001993-20011993-20021993-20031993-20040.000.100.200.30-0.10-0.20Florist Transactions Coefficient $25,000 to $49,999 1993-19961993-19971993-19981993-19991993-20001993-20011993-20021993-20031993-20040.000.100.200.30-0.10-0.20Supermarket Expenditure Coefficient $25,000 to $49,999 1993-19961993-19971993-19981993-19991993-20001993-20011993-20021993-20031993-20040.000.100.200.30-0.10-0.20Supermarket Transactions Coefficient $25,000 to $49,999 Figure B-10 Time recursive parameters for the $25,000 to $49,999 income group.

PAGE 120

106 1993-19961993-19971993-19981993-19991993-20001993-20011993-20021993-20031993-20040.000.100.200.30-0.10-0.20Florist Expenditure Coefficient $50,000 to $74,999 1993-19961993-19971993-19981993-19991993-20001993-20011993-20021993-20031993-20040.000.100.200.30-0.10-0.20Florist Transactions Coefficient $50,000 to $74,999 1993-19961993-19971993-19981993-19991993-20001993-20011993-20021993-20031993-20040.000.100.200.30-0.10-0.20Supermarket Expenditure Coefficient $50,000 to $74,999 1993-19961993-19971993-19981993-19991993-20001993-20011993-20021993-20031993-20040.000.100.200.30-0.10-0.20Supermarket Transactions Coefficient $50,000 to $74,999 Figure B-11 Time recursive parameters for the $50,000 to $74,999 income g roup. 1993-19961993-19971993-19981993-19991993-20001993-20011993-20021993-20031993-20040.000.100.200.30-0.10-0.20Florist Expenditure Coefficient $75,000 and more 1993-19961993-19971993-19981993-19991993-20001993-20011993-20021993-20031993-20040.000.100.200.30-0.10-0.20Florist Transactions Coefficient $75,000 and more 1993-19961993-19971993-19981993-19991993-20001993-20011993-20021993-20031993-20040.000.100.200.30-0.10-0.20Supermarket Expenditure Coefficient $75,000 and more 1993-19961993-19971993-19981993-19991993-20001993-20011993-20021993-20031993-20040.000.100.200.30-0.10-0.20Supermarket Transactions Coefficient $75,000 and more Figure B-12 Time recursive parameters for the $75,000 and more income group.

PAGE 121

107 1993-19961993-19971993-19981993-19991993-20001993-20011993-20021993-20031993-20040.000.200.40-0.20-0.40Florist Expenditure Coefficient Arrangements 1993-19961993-19971993-19981993-19991993-20001993-20011993-20021993-20031993-20040.000.200.40-0.20-0.40Florist Transactions Coefficient Arrangements 1993-19961993-19971993-19981993-19991993-20001993-20011993-20021993-20031993-20040.000.200.40-0.20-0.40Supermarket Expenditure Coefficient Arrangements 1993-19961993-19971993-19981993-19991993-20001993-20011993-20021993-20031993-20040.000.200.40-0.20-0.40Supermarket Transactions Coefficient Arrangements Figure B-13 Time recursive parameters for arrangements. 1993-19961993-19971993-19981993-19991993-20001993-20011993-20021993-20031993-20040.000.200.40-0.20-0.40Florist Expenditure Coefficient Non-arrangements 1993-19961993-19971993-19981993-19991993-20001993-20011993-20021993-20031993-20040.000.200.40-0.20-0.40Florist Transactions Coefficient Non-arrangements 1993-19961993-19971993-19981993-19991993-20001993-20011993-20021993-20031993-20040.000.200.40-0.20-0.40Supermarket Expenditure Coefficient Non-arrangements 1993-19961993-19971993-19981993-19991993-20001993-20011993-20021993-20031993-20040.000.200.40-0.20-0.40Supermarket Transactions Coefficient Non-arrangements Figure B-14 Time recursive parameters for non-arrangements.

PAGE 122

108 1993-19961993-19971993-19981993-19991993-20001993-20011993-20021993-20031993-20040.000.200.40-0.20-0.40Florist Expenditure Coefficient Flowering/greenhouse plants 1993-19961993-19971993-19981993-19991993-20001993-20011993-20021993-20031993-20040.000.200.40-0.20-0.40Florist Transactions Coefficient Flowering/greenhouse plants 1993-19961993-19971993-19981993-19991993-20001993-20011993-20021993-20031993-20040.000.200.40-0.20-0.40Supermarket Expenditure Coefficient Flowering/greenhouse plants 1993-19961993-19971993-19981993-19991993-20001993-20011993-20021993-20031993-20040.000.200.40-0.20-0.40Supermarket Transactions Coefficient Flowering/greenhouse plants Figure B-15 Time recursive parameters for flowering/green house plants.

PAGE 123

APPENDIX C TSP CODE Options memory=500; Title 'Outlet Shares for Fresh Flowers'; IN 'C:\zstudent\Christian_Iniguez\OUTFORM'; PRINT @NOB; LIST ZVAR1 IDD MT YR AGE GND INC PUR HWD; ? 0 ALL INDOOR FLOWERS ? 1 INDOOR FLOWERS ? 2 CUT FLOWERS ? 3 SUB(2) FLOWER ARRANGEMENT ? 4 SUB(2) NON-ARRANGEMENTS ? 5 FLWRG & GREEN HOUSE PLNTS ? 6 SUB(5) FLOWERING PLANTS ? 7 SUB(5) FOLIAGE ? 0 ALL GROUPS ? 1 TOTAL SPECIALTY ? 2 TOTAL MASS MERCHANDISERS ? 3 FLORIST SHOP SUB(1) ? 4 SUPERMARKET SUB(2) ? 5 WAREHOUSE/PRICE CLUB SUB(2) ? 6 INTERNET RETAILER ? 7 OTHER ? 8 ALL OTHER SPECIALITY (NEW) ? 9 ALL OTHER MERCHANDISERS (NEW) ? 10 TOTAL OUTLETS (1)+(2) + (6) + 7) (NEW) ? IDD=1 THEN THE DATA IS ONLY FOR OBSERVATIONS BY DEMOGRAPHICS AND FORMS. LIST ZZOUTE EXP_S0 EXP_S0 EXP_S1 EXP_S2 EXP_S3 EXP_S4 EXP_S5 EXP_S6 EXP_S7 EXP_S8 EXP_S9 EXP_S10; LIST ZZOUTT TRN_S0 TRN_S0 TRN_S1 TRN_S2 TRN_S3 TRN_S4 TRN_S5 TRN_S6 TRN_S7 TRN_S8 TRN_S9 TRN_S10; ?OC HWD 'NUMBER OF HOUSEHOLDS'; ?OC IDD 'IDENDTIFER FOR MONTHS VS TOTALS'; ?OC YR 'YEARS 1993 TO 2004'; ?OC MT 'MONTHS 1-12'; ?OC AGE 'AGE 1=UNDER 25, 2=25-40YRS, 3=40-54YRS, 4=55+'; 109

PAGE 124

110 ?OC GND 'FEMALE=0 AND MALE=1'; ?OC INC 'INCOME UNDER $25=1, $25/$50=2, $50/75=3, $75+=4'; ?OC PUR 'REASON FOR BUYING SELF=1, GIFT=0'; ?OC FORM 'FLOWER FORM TOTAL=0,INDOOR=1,CUT=2,ARRANG=3,NONARR=4,PLANTS=5,FLWING=6,FOLIAGE=7'; ?OC EXP_S0 'EXPENDITURES THROULETS'; ?OC EXP_S1 'EXPENDITURES THROUGH ALL S'; ITURES THROUGH ALL MASS MERCHANDISING'; UGH ALL FLORIST'; UGH ALL SUPERMARKETS'; LL WARESHOUSES/PRICE'; PENDITURES THROUGH ALL RETAIL INTERNET'; TY'; OC EXP_S9 'EXPENDITURES THROUGH OTHER MASS MERCHANDISING'; S THROUGH TOTAL OUTLETS'; THROUGH ALL OUTLETS'; S THROUGH ALL SPECIALTY'; ALL MASS MERCHANDISING'; L FLORIST'; H ALL SUPERMARKETS'; GH ALL WARESHOUSES/PRICE'; HROUGH ALL RETAIL INTERNET'; ROUGH ALL OTHER'; ACTIONS THROUGH OTHER SPECIALTY'; OUGH OTHER MASS MERCHANDISING'; TOTAL OUTLETS'; ====================================POSTITIVE TOTAL EXPENDITURES; ============================================DD=1; ====================================================================NTS, NONARRANGEMENTS, AND ========================================== GH ALL OUTPECIALTY ?OC EXP_S2 'EXPEND ?OC EXP_S3 'EXPENDITURES THROOC EXP_S4 'EXPENDITURES THRO ? ?OC EXP_S5 'EXPENDITURES THROUGH A ?OC EXP_S6 'EX ?OC EXP_S7 'EXPENDITURES THROUGH ALL OTHER'; OC EXP_S8 'EXPENDITURES THROUGH OTHER SPECIAL ? ? ?OC EXP_S10 'EXPENDITURE ?OC TRN_S0 'TRANSACTIONS ?OC TRN_S1 'TRANSACTION ?OC TRN_S2 'TRANSACTIONS THROUGH ?OC TRN_S3 'TRANSACTIONS THROUGH ALG ?OC TRN_S4 'TRANSACTIONS THROU ?OC TRN_S5 'TRANSACTIONS THROU ?OC TRN_S6 'TRANSACTIONS TOC TRN_S7 'TRANSACTIONS TH ? ?OC EXP_S8 'TRANS ?OC EXP_S9 'TRANSACTIONS THR ?OC EXP_S10 'TRANSACTIONS THROUGH ?================================ ====; RO AND ? CREATING THE CONTROL FOR ZE ?======================== ====; TM = INT(YR*100 + MT); MSD TM; DD=(TM<200404) & IDD=1; HIST(DISCRETE) FORM; ELECT S DOT EXP TRN; X._S0=(._S0>0); HIST(DISCRETE) X._S0; ENDDOT; ? =======; ? CREATING DUMMIES FOR FORMS ARRANGEME FLOWERING; ====== ?==================== =======; FZ=(FORM=3 | FORM=4 | FORM=5);

PAGE 125

111 SELECT 1; HIST(DISCRETE) FZ; SELECT DD=1; ?========================================================================; ? CREATING OUR DUMMY VARIABLES FOR THE RIGHT-H AND-SIDE VARIABLES; ====================ALE (0 & 1); REE FORMS; ERING; ============ITURES; ====================================================================T XEXP_S0 C DAGE1-DAGE3 DGND1 DPUR1 DINC1-DINC3 DMT1-DMT11 ILLSEXP=NORM(@COEF(1))/CNORM(@COEF(1)); NCOEF -1); MAEXP MILLSEXP; 1-DAGE3 DGND1 DPUR1 DINC1-DINC3 DMT1-DMT11 2; COEF(1))/CNORM(@COEF(1)); TRN = @SSR/(@NCOEF -1); RINT SIGMATRN; AT COEF_TRN = @COEF; ==========================================================1 & FZ=1 & ( XEXP_S0>0) & (XTRN_S0>0); ?================================================ ====; DOT MT AGE GND INC PUR ; DUMMY .; ENDDOT; ? MONTHS; DOT 1-11; DMT.=MT.-MT12; ENDDOT; DOT 1-3; DAGE.= AGE. AGE4; ENDDOT; DOT 1 ; DGND. = GND. GND2; ENDDOT; ? FEMALE MINUS M DOT 1 ; DPUR. = PUR. PUR2; ENDDOT; ? GIFT MINUS SELF (0 & 1); DOT 1-3; DINC. = INC. INC4; ENDDOT; SELECT DD=1 & FZ=1; ? SELECTION ON TIME, ALL DEMOG, AND TH PRINT @NOB; HIST(DISCRETE) FZ FORM PUR; DUMMY FORM; ? 1=ARRANGEMENTS 2=NON-ARRANGEMENTS 3=FLOW DOT 1-2; DFORM.=FORM. FORM3; ENDDOT; HIST(DISCRETE) FORM1 FORM2 FORM3; ?======================================================== =======; ? PROBIT MODELS TO GET THE PROBABILITIES OF POSITIVE EXPEND ? =======; PROBI DFORM1-DFORM2; MILLEXP = @MILLS; SET M SET SIGMAEXP = @SSR/(@ PRINT SIG MAT COEF_EXP = @COEF; PROBIT XTRN_S0 C DAGE DFORM1-DFORM MILLTRN = @MILLS; SET MILLSTRN=NORM(@ SET SIGMA P M ?===========================================================================; ? NOW SELECTING FOR POSITIVE EXPENDITURES OR TRANSACTIONS; ?=========== =======; SELECT DD

PAGE 126

112 ? ====================================================================TING THE OUTLET SHARES; =9; HEXP_S. = EXP_S. / EXP_S0; _S.; ; E SHARE OF THE MARKET; ; =====================================================; R LIMIT TOBIT MODEL; ET 2 LIMIT PROCEDURE; ================; ) @COEF=0; COL=1) @T=0; DYMID XBEQ XB EOS EOSG SIGMA SIGI TOBIT2 TOBIT2G; ND UPPER LIMITS LIM); ); 0 + B1*DAGE1 + B2*DAGE2 + B3*DAGE3 + B4*DGND1 + B5*DPUR1 + =======; ? CREA ?=================================================================== =======; DOT 1S SHTRN_S. = TRN_S. / TRN_S0; HIST SHEXP HIST SHTRN_S.; ENDDOT; CONST LOWLIM 0; ? LOWER LIMIT TO THE SHARE OF THE MAREKT CONST UPLIM 1.0; ? UPPER LIMIT TO TH LIST XS C DAGE1-DAGE3 DGND1 DPUR1 DINC1-DINC3 DMT1-DMT11 DFORM1DFORM2 MILLS ?======== ? WE WILL PASS YDEP AND MILL INTO THE PROCEDURE; ? UPPER AND LOWE ? STARTING THE TOB ?============================================= MFORM(TYPE=GEN,NROW=1,NCOL=1 MFORM(TYPE=GEN,NROW=1,N PROC TOB2LIM ; LOCAL DYLOW DYUP ? CREATE DUMMY VARIABLES FOR LOWER, MIDDLE, A DYLOW = (YDEP=LOWLIM); DYUP = (YDEP=UP DYMID = 1 (DYLOW+DYUP MSD(TERSE) DYLOW DYMID DYUP; FORM(VARPREF=B) XBEQ XB XS; FRML XB B B6*DINC1 + B7*DINC2 + B8*DINC3

PAGE 127

113 + B9*DMT1 + B10*DMT2 + B11*DMT3 + B12*DMT4 + B13*DMT5 + B14*DMT6 + 15*DMT7 + B16*DMT8 + B17*DMT9 B18*DMT10 + B19*DMT11 + B20*DFORM1 + B21*DFORM2 + B22*MILLS; ((YDEP XB)/SIGMA) ; ? RESIDUAL/SIGMA POSITIVE STARTING VALUE FOR SIGI RAL LOG LIKELIHOOD BIT2 LOGL = DYLOW*LCNORM(EOS) + DYMID*(LNORM(EOS) S); OS XB; =N,HCOV=N) TOBIT2; RINT @COEF; =================================; ==========================================================================; CREAT A LOGISTIC FORM FOR THE SHARES RECOGNIZING THAT SOME ZERO AN OCCUR; 3= FLORISTS 4=SUPERMAREKTS 5=WAREHOUSES 6=INTERNET; ==========================================================================; FORM(TYPE=GEN, NROW=33,NCOL=1) ZMATZ=0; OB2LIM; ; 0 0 0 0 0; AT MM = @T; MAKE(VERT) MN MM MS; ENDDOT; B + FRML EOS PARAM SIGMA 1 B0-B22; ?, SIGI,1; ? ? STRUCTU FRML TO LOG(SIGMA)) + DYUP*LCNORM(-EO EQSUB TOBIT2 E ML(HITER P ENDPROC; ?======================================================== = ? = ? C ? ? = M SELECT DD=1 & FZ=1 & ( XEXP_S0>0) & (XTRN_S0>0); DOT(CHAR=#) EXP TRN; DOT(CHAR=%) 3 4; ? 5 6; ? FLORISTS AND SUPERMARKETS ONLY; YDEP = SH.#_S.%; MILLS = MILL.#; T PRINT @COEF @T MMAKE COEF.#_S.% @COEF; ? SAVING THE COEFFICIENTS FOR USE IN THE SIMULATIONS; MAT MM = @COEF; MMAKE(VERT) MS 0 0 0 0 MMAKE(VERT) MN MM MS; MMAKE ZMATZ ZMATZ MN; M MMAKE(VERT) MS 0 0 0 0 0 0 0 0 0; M MMAKE ZMATZ ZMATZ MN; ENDDOT;

PAGE 128

114 ? WRITE(FORMAT=EXCEL,FILE='C:\zstudent\Christian_Iniguez\TOBITCOEFB.xls') ZMATZ; ? DEFINE THE SIMULATED VARIABLES AS X AND THE COEFFICIENTS AS B; DEFINE L = THE LOWER VALUE; ============================================; ; ; SIMULATOR FOR CALCUATING THE AVERAGE SHARE FOR EACH TYPE AND UTLET ; THE FOLLOWING MUST BE DEFINED: SIMNUM, SIMTYPE, SIMOUTL BIT COEF_PROBIT SIG_TOBIT SIG_PROBIT ; =========================================================AR SIM_AGE SIM_GENDER SIM_PURPOSE SIM_INCOME SIM_MONTH OT SIMVAR; SET .=0; ENDDOT; ROC INIT; ? INTIIALIZING THE SIMULATION VARIABLES TO ZERO; DDOT; SHZ; ET ZAGE1 = (SIM_AGE=1) -(SIM_AGE=4); M_AGE=3) -(SIM_AGE=4); ENDER=1) -(SIM_GENDER=2); ? FEMALE MINUS MALE; _PURPOSE=2); ? GIFT MINUS SELF; ) -(SIM_INCOME=4); -(SIM_INCOME=4); _INCOME=3) -(SIM_INCOME=4); _MONTH=12); -(SIM_MONTH=12); -(SIM_MONTH=12); = (SIM_MONTH=4) -(SIM_MONTH=12); = (SIM_MONTH=5) -(SIM_MONTH=12); ? ? DEFINE U = THE UPPER VALUE; MFORM(TYPE=GEN,NROW=500,NCOL=8) MSHAREM=0; ?========================== ============= ? ? STARTING THE SIMULATION SECTION ; ? ? O ? ; ? COEF_TO ? ; ?=========== ===============; SET INTCP = 1; LIST SIMV SIM_FORM; D P DOT SIMVAR; SET .=0; EN ENDPROC INIT; PROC ZSIM SET I=I+1; S SET ZAGE2 = (SIM_AGE=2) -(SIM_AGE=4); SET ZAGE3 = (SI SET ZGND1 = (SIM_G SET ZPUR1 = (SIM_PURPOSE=1) -(SIM SET ZINC1 = (SIM_INCOME=1 SET ZINC2 = (SIM_INCOME=2) SET ZINC3 = (SIM SET ZMT1 = (SIM_MONTH=1) -(SIM SET ZMT2 = (SIM_MONTH=2) SET ZMT3 = (SIM_MONTH=3) SET ZMT4 SET ZMT5

PAGE 129

115 SET ZMT6 = (SIM_MONTH=6) -(SIM_MONTH=12); SET ZMT 7 = (SIM_MONTH=7) -(SIM_MONTH=12); ET ZMT8 = (SIM_MONTH=8) -(SIM_MONTH=12); ET ZMT9 = (SIM_MONTH=9) -(SIM_MONTH=12); ET ZMT10 = (SIM_MONTH=10) -(SIM_MONTH=12); _FORM=5); M_FORM=5); AT U = 1.0; AT NR = NROW(COEF_TOBIT); EF_TOBIT(1)-COEF_TOBIT(J); YPE=1)*(SIGMAEXP) + (SIMTYPE=2)*(SIGMATRN); MUST CREATE THE MILLS VALUE FROM THE SIMULATED X VALUES; ZINC1-ZINC3 ZMT1-ZMT11 ZFORM1 W(Z); MAT NNC=NCOL(Z); PROBIT VARIABLES AND COEFFICIENTS; (ZA); RINT ZMILLS; T11 ZFORM1 OR THE TOBIT MODEL; ORM[ ( L(1) XB(1) ) / ((SIGT(1) ) ) ]; _L = CNORM[ ( L(1) XB(1) ) / ((SIGT(1) ) ) ]; T(1) ) ) ]; IGT(1) ) ) ]; ORM_U)/(CNORM_U CNORM_L) ] }; ; ) AND TRANSACTIONS (2); PERMARKETS (4) ...; ES FOR EACH DUMMY; S S S SET ZMT11 = (SIM_MONTH=11) -(SIM_MONTH=12); SET ZFORM1= (SIM_FORM=3) -(SIM SET ZFORM2= (SIM_FORM=4) -(SI MAT L = 0; M M SET J=NR-1; MMAKE(VERT) B CO MAT A = COEF_PROBIT; MAT SIGT = COEF_TOBIT(NR); MAT SIGP = (SIMT ? SET INTCP =1; MMAKE Z INTCP ZAGE1-ZAGE3 ZGND1 ZPUR1 ZFORM2; MAT NNR=NRO ? VARIABLES FOR THE PROBIT MODEL; MAT ZA = Z'A; ? SET ZMILLS = NORM(ZA) /CNORM P MMAKE X INTCP ZAGE1-ZAGE3 ZGND1 ZPUR1 ZINC1-ZINC3 ZMT1-ZM ZFORM2 ZMILLS; ? VARIABLES F MAT XB = X'B; SET NORM_L = N SET CNORM SET NORM_U = NORM[ ( U(1) XB(1) ) / ((SIG SET CNORM_U = CNORM[ ( U(1) XB(1) ) / ((S SET MID = { XB(1) + ( SIGT(1) )*[ (NORM_L N SET EY = L(1)*CNORM_L + U(1)*[ 1 CNORM_U] + MID*[ CNORM_U CNORM_L ]; ? EXPECTED VALUES ACROSS THE FULL RANGE OF SHARES SET MSHAREM(I,1)= SIMNUM; SET MSHAREM(I,2)= SIMTYPE; ? EXPENDITURES (1 SET MSHAREM(I,3)= SIMOUTL; ? FLORISTS (3) SU SET MSHAREM(I,4)= K; ? THE VARIABLE VALU SET MSHAREM(I,5)= CNORM_L; SET MSHAREM(I,6)= 1CNORM_U; SET MSHAREM(I,7)= MID*[ CNORM_U CNORM_L ]; SET MSHAREM(I,8)= EY;

PAGE 130

116 ENDPROC ZSIMSHZ;; SET I=0; ?========= ===================================================; N =1 ; AVERAGE HOUSEHOLD AND EXPENDITURES / FLORIST ; ============================================================; = 1; TRANSACTIONS (2); ? FLORISTS (3) SUPERMARKETS (4); _S3*(SIMTYPE=1 & SIMOUTL=3)+ OEFEXP_S4*(SIMTYPE=1 & SIMOUTL=4)+ COEFTRN_S3*(SIMTYPE=2 & SIMOUTL=3)+ COEFTRN_S4*(SIMTYPE=2 & ? TOBIT COEFFICIENTS FOR EXPENDITURES THROUGH FLORISTS; ); ? PROBIT COEFFICIENTS FOR EXPENDITURES ON ALL OUTLETS; =========; ; HOUSEHOLDS AND EXPENDITURES / FLORIST ; IT = COEFEXP_S3*(SIMTYPE=1 & SIMOUTL=3)+ OEFEXP_S4*(SIMTYPE=1 & SIMOUTL=4)+ COEFTRN_S3*(SIMTYPE=2 & SIMOUTL=3)+ COEFTRN_S4*(SIMTYPE=2 & URES THROUGH FLORISTS; _EXP*(SIMTYPE=1); URES ON ALL OUTLETS; ===============================; SEHOLDS AND EXPENDITURES / FLORIST ; ? SIMULATIO ? ? SET K =1; SET SIMNUM SET SIMTYPE = 1; ? EXPENDITURES (1) AND SET SIMOUTL = 3; MAT COEF_TOBIT = COEFEXP C SIMOUTL=4); MAT COEF_PROBIT = COEF_TRN*(SIMTYPE=2) +COEF_EXP*(SIMTYPE=1 INIT; ZSIMSHZ; ?=================================================== ? SIMULATION =2 ? OVER AGES OF ?============================================================; SET SIMNUM = 2; MAT COEF_TOB C SIMOUTL=4); ? TOBIT COEFFICIENTS FOR EXPENDIT MAT COEF_PROBIT = COEF_TRN*(SIMTYPE=2) +COEF ? PROBIT COEFFICIENTS FOR EXPENDIT DO K=0 TO 4; INIT; SET SIM_AGE=K; ZSIMSHZ; ENDDO; ?============================= ? SIMULATION =3 ; ? OVER GENDER OF HOU

PAGE 131

117 ?================== ==========================================; ET SIMNUM = 3; AT COEF_TOBIT = COEFEXP_S3*(SIMTYPE=1 & SIMOUTL=3)+ OEFEXP_S4*(SIMTYPE=1 & SIMOUTL=4)+ COEFTRN_S3*(SIMTYPE=2 & SIMOUTL=3)+ COEFTRN_S4*(SIMTYPE=2 & IMOUTL=4); ? TOBIT COEFFICIENTS FOR EXPENDITURES THROUGH FLORISTS; F_PROBIT = COEF_TRN*(SIMTYPE=2) +COEF_EXP*(SIMTYPE=1); TLETS; ENDER=K; =4 ; OEF_TOBIT = COEFEXP_S3*(SIMTYPE=1 & SIMOUTL=3)+ S4*(SIMTYPE=1 & SIMOUTL=4)+ COEFTRN_S3*(SIMTYPE=2 & SIMOUTL=3)+ COEFTRN_S4*(SIMTYPE=2 & IMOUTL=4); ? TOBIT COEFFICIENTS FOR EXPENDITURES THROUGH FLORISTS; R EXPENDITURES ON ALL OUTLETS; SIMULATION =5 ; E OF HOUSEHOLDS AND EXPENDITURES / FLORIST ; ========================================================; ; _TOBIT = COEFEXP_S3*(SIMTYPE=1 & SIMOUTL=3)+ _S4*(SIMTYPE=1 & SIMOUTL=4)+ COEFTRN_S3*(SIMTYPE=2 & SIMOUTL=3)+ COEFTRN_S4*(SIMTYPE=2 & IMOUTL=4); ORISTS; =2) +COEF_EXP*(SIMTYPE=1); UTLETS; S M C S MAT COE ? PROBIT COEFFICIENTS FOR EXPENDITURES ON ALL OU DO K=0 TO 2; INIT; SET SIM_G ZSIMSHZ; ENDDO; ?============================================================; ? SIMULATION ? OVER PURPOSE OF HOUSEHOLDS AND EXPENDITURES / FLORIST ; ?============================================================; SET SIMNUM = 4; MAT C COEFEXP_ S MAT COEF_PROBIT = COEF_TRN*(SIMTYPE=2) +COEF_EXP*(SIMTYPE=1); ? PROBIT COEFFICIENTS FO DO K=0 TO 2; INIT; SET SIM_PURPOSE=K; ZSIMSHZ; ENDDO; ?============================================================; ? ? OVER INCOM ?==== SET SIMNUM = 5 MAT COEF COEFEXP S ? TOBIT COEFFICIENTS FOR EXPENDITURES THROUGH FL MAT COEF_PROBIT = COEF_TRN*(SIMTYPE ? PROBIT COEFFICIENTS FOR EXPENDITURES ON ALL O

PAGE 132

118 DO K=0 TO 4; INIT; SET SIM_INCOME=K; ZSIMSHZ; ENDDO; ?=========== =================================================; LATION =6 ; OUSEHOLDS AND EXPENDITURES / FLORIST ; ====================================================; UM = 6; AT COEF_TOBIT = COEFEXP_S3*(SIMTYPE=1 & SIMOUTL=3)+ OEFEXP_S4*(SIMTYPE=1 & SIMOUTL=4)+ COEFTRN_S3*(SIMTYPE=2 & SIMOUTL=3)+ COEFTRN_S4*(SIMTYPE=2 & EXPENDITURES THROUGH FLORISTS; ); LETS; =================================================; LATION =7 ; SEHOLDS AND EXPENDITURES / FLORIST ; ====================================================; UM = 7; AT COEF_TOBIT = COEFEXP_S3*(SIMTYPE=1 & SIMOUTL=3)+ OEFEXP_S4*(SIMTYPE=1 & SIMOUTL=4)+ COEFTRN_S3*(SIMTYPE=2 & SIMOUTL=3)+ COEFTRN_S4*(SIMTYPE=2 & EXPENDITURES THROUGH FLORISTS; 1); TLETS; ? SIMU ? OVER SEASONS OF H ?======== SET SIMN M C SIMOUTL=4); ? TOBIT COEFFICIENTS FOR MAT COEF_PROBIT = COEF_TRN*(SIMTYPE=2) +COEF_EXP*(SIMTYPE=1 ? PROBIT COEFFICIENTS FOR EXPENDITURES ON ALL OUT DO K=0 TO 12; INIT; SET SIM_MONTH=K; ZSIMSHZ; ENDDO; ?=========== ? SIMU ? OVER FORMS OF HOU ?======== SET SIMN M C SIMOUTL=4); ? TOBIT COEFFICIENTS FOR MAT COEF_PROBIT = COEF_TRN*(SIMTYPE=2) +COEF_EXP*(SIMTYPE= ? PROBIT COEFFICIENTS FOR EXPENDITURES ON ALL OU DOT(VALUE=K) 0 3 4 5; INIT; SET SIM_FORM=K; ZSIMSHZ; ENDDOT;

PAGE 133

119 ?==== ========================================================; ; RMS AND PURPOSE OF HOUSEHOLDS AND EXPENDITURES / FLORIST ============================================================; ET SIMNUM = 8; AT COEF_TOBIT = COEFEXP_S3*(SIMTYPE=1 & SIMOUTL=3)+ OEFEXP_S4*(SIMTYPE=1 & SIMOUTL=4)+ PE=2 & FLORISTS; OBIT COEFFICIENTS FOR EXPENDITURES ON ALL OUTLETS; ====================================================; TION =9 ; OVER INCOME OF HOUSEHOLDS AND EXPENDITURES / FLORIST ; ============================================================; ET SIMNUM = 9; IMTYPE=2 & BIT COEFFICIENTS FOR EXPENDITURES THROUGH FLORISTS; YPE=1); R EXPENDITURES ON ALL OUTLETS; SIMSHZ; O; =================================================================================================; ? SIMULATION =8 ? OVER FO ; ? S M C COEFTRN_S3*(SIMTYPE=2 & SIMOUTL=3)+ COEFTRN_S4*(SIMTY SIMOUTL=4); ? TOBIT COEFFICIENTS FOR EXPENDITURES THROUGH MAT COEF_PROBIT = COEF_TRN*(SIMTYPE=2) +COEF_EXP*(SIMTYPE=1); ? PR DOT(VALUE=K) 0 3 4 5; ? FORMS; DO G=0 TO 2; INIT; SET SIM_FORM=K; SET SIM_PURPOSE=G; ZSIMSHZ; ENDDO; ENDDOT; ?======== ? SIMULA ? ? S MAT COEF_TOBIT = COEFEXP_S3*(SIMTYPE=1 & SIMOUTL=3)+ COEFEXP_S4*(SIMTYPE=1 & SIMOUTL=4)+ COEFTRN_S3*(SIMTYPE=2 & SIMOUTL=3)+ COEFTRN_S4*(S SIMOUTL=4); ? TO MAT COEF_PROBIT = COEF_TRN*(SIMTYPE=2) +COEF_EXP*(SIMT ? PROBIT COEFFICIENTS FO DO K=1 TO 4; DO D=1 TO 12; INIT; SET SIM_INCOME=K; SET SIM_MONTH=D; Z ENDDO; ENDD ? =

PAGE 134

120 ? NOW TYPE =2 AND OUTL=3; ============================================================================================; HOUSEHOLD AND EXPENDITURES / FLORIST ; (1) AND TRANSACTIONS (2); BIT = COEFEXP_S3*(SIMTYPE=1 & SIMOUTL=3)+ =2 & ? TOBIT COEFFICIENTS FOR EXPENDITURES THROUGH FLORISTS; (SIMTYPE=2) +COEF_EXP*(SIMTYPE=1); ? PROBIT COEFFICIENTS FOR EXPENDITURES ON ALL OUTLETS; ============================================================; SIMULATION =2 ; OVER AGES OF HOUSEHOLDS AND EXPENDITURES / FLORIST ; 3*(SIMTYPE=2 & SIMOUTL=3)+ COEFTRN_S4*(SIMTYPE=2 & R EXPENDITURES THROUGH FLORISTS; ? PROBIT COEFFICIENTS FOR EXPENDITURES ON ALL OUTLETS; ET SIM_AGE=K; ====================================================; TION =3 ; ENDER OF HOUSEHOLDS AND EXPENDITURES / FLORIST ; ============================================================; ET SIMNUM = 3; AT COEF_TOBIT = COEFEXP_S3*(SIMTYPE=1 & SIMOUTL=3)+ ? ==============================; ?==================================== ? SIMULATION =1 ; ? AVERAGE ?============================================================; SET K =1; SET SIMNUM = 1; SET SIMTYPE = 2; ? EXPENDITURES SET SIMOUTL = 3; ? FLORISTS (3) SUPERMARKETS (4); MAT COEF_TO COEFEXP_S4*(SIMTYPE=1 & SIMOUTL=4)+ COEFTRN_S3*(SIMTYPE=2 & SIMOUTL=3)+ COEFTRN_S4*(SIMTYPE SIMOUTL=4); MAT COEF_PROBIT = COEF_TRN* INIT; ZSIMSHZ; ? ? ? ?============================================================; SET SIMNUM = 2; MAT COEF_TOBIT = COEFEXP_S3*(SIMTYPE=1 & SIMOUTL=3)+ COEFEXP_S4*(SIMTYPE=1 & SIMOUTL=4)+ COEFTRN_S SIMOUTL=4); ? TOBIT COEFFICIENTS FO MAT COEF_PROBIT = COEF_TRN*(SIMTYPE=2) +COEF_EXP*(SIMTYPE=1); DO K=0 TO 4; INIT; S ZSIMSHZ; ENDDO; ?======== ? SIMULA ? OVER G ? S M COEFEXP_S4*(SIMTYPE=1 & SIMOUTL=4)+

PAGE 135

121 COEFTRN_S3*(SIMTY PE=2 & SIMOUTL=3)+ COEFTRN_S4*(SIMTYPE=2 & TS FOR EXPENDITURES THROUGH FLORISTS; R EXPENDITURES ON ALL OUTLETS; =================================================; MNUM = 4; _TOBIT = COEFEXP_S3*(SIMTYPE=1 & SIMOUTL=3)+ OEFEXP_S4*(SIMTYPE=1 & SIMOUTL=4)+ COEFTRN_S3*(SIMTYPE=2 & SIMOUTL=3)+ COEFTRN_S4*(SIMTYPE=2 & IMOUTL=4); ORISTS; =2) +COEF_EXP*(SIMTYPE=1); L OUTLETS; ============================================================; =5 ; R INCOME OF HOUSEHOLDS AND EXPENDITURES / FLORIST ; =============================================; M = 5; F_TOBIT = COEFEXP_S3*(SIMTYPE=1 & SIMOUTL=3)+ OEFEXP_S4*(SIMTYPE=1 & SIMOUTL=4)+ COEFTRN_S3*(SIMTYPE=2 & SIMOUTL=3)+ COEFTRN_S4*(SIMTYPE=2 & EXPENDITURES THROUGH FLORISTS; ); TLETS; SET SIM_INCOME=K; SIMOUTL=4); ? TOBIT COEFFICIEN MAT COEF_PROBIT = COEF_TRN*(SIMTYPE=2) +COEF_EXP*(SIMTYPE=1); ? PROBIT COEFFICIENTS FO DO K=0 TO 2; INIT; SET SIM_GENDER=K; ZSIMSHZ; ENDDO; ?=========== ? SIMULATION =4 ; ? OVER PURPOSE OF HOUSEHOLDS AND EXPENDITURES / FLORIST ; ?============================================================; SET SI MAT COEF C S ? TOBIT COEFFICIENTS FOR EXPENDITURES THROUGH FL MAT COEF_PROBIT = COEF_TRN*(SIMTYPE ? PROBIT COEFFICIENTS FOR EXPENDITURES ON AL DO K=0 TO 2; INIT; SET SIM_PURPOSE=K; ZSIMSHZ; ENDDO; ? ? SIMULATION ? OVE ?=============== SET SIMNU MAT COE C SIMOUTL=4); ? TOBIT COEFFICIENTS FOR MAT COEF_PROBIT = COEF_TRN*(SIMTYPE=2) +COEF_EXP*(SIMTYPE=1 ? PROBIT COEFFICIENTS FOR EXPENDITURES ON ALL OU DO K=0 TO 4; INIT;

PAGE 136

122 ZSIMSHZ; ENDDO; ?=========== =================================================; LATION =6 ; OUSEHOLDS AND EXPENDITURES / FLORIST ; ====================================================; UM = 6; AT COEF_TOBIT = COEFEXP_S3*(SIMTYPE=1 & SIMOUTL=3)+ OEFEXP_S4*(SIMTYPE=1 & SIMOUTL=4)+ COEFTRN_S3*(SIMTYPE=2 & SIMOUTL=3)+ COEFTRN_S4*(SIMTYPE=2 & EXPENDITURES THROUGH FLORISTS; ); TLETS; =================================================; LATION =7 ; SEHOLDS AND EXPENDITURES / FLORIST ; ====================================================; UM = 7; AT COEF_TOBIT = COEFEXP_S3*(SIMTYPE=1 & SIMOUTL=3)+ OEFEXP_S4*(SIMTYPE=1 & SIMOUTL=4)+ COEFTRN_S3*(SIMTYPE=2 & SIMOUTL=3)+ COEFTRN_S4*(SIMTYPE=2 & EXPENDITURES THROUGH FLORISTS; 1); TLETS; ========================================================; ; ? SIMU ? OVER SEASONS OF H ?======== SET SIMN M C SIMOUTL=4); ? TOBIT COEFFICIENTS FOR MAT COEF_PROBIT = COEF_TRN*(SIMTYPE=2) +COEF_EXP*(SIMTYPE=1 ? PROBIT COEFFICIENTS FOR EXPENDITURES ON ALL OU DO K=0 TO 12; INIT; SET SIM_MONTH=K; ZSIMSHZ; ENDDO; ?=========== ? SIMU ? OVER FORMS OF HOU ?======== SET SIMN M C SIMOUTL=4); ? TOBIT COEFFICIENTS FOR MAT COEF_PROBIT = COEF_TRN*(SIMTYPE=2) +COEF_EXP*(SIMTYPE= ? PROBIT COEFFICIENTS FOR EXPENDITURES ON ALL OU DOT(VALUE=K) 0 3 4 5; INIT; SET SIM_FORM=K; ZSIMSHZ; ENDDOT; ?==== ? SIMULATION =8

PAGE 137

123 ? OVER FO RMS AND PURPOSE OF HOUSEHOLDS AND EXPENDITURES / FLORIST ============================================================; ET SIMNUM = 8; AT COEF_TOBIT = COEFEXP_S3*(SIMTYPE=1 & SIMOUTL=3)+ OEFEXP_S4*(SIMTYPE=1 & SIMOUTL=4)+ PE=2 & FLORISTS; OBIT COEFFICIENTS FOR EXPENDITURES ON ALL OUTLETS; ====================================================; TION =9 ; OVER INCOME OF HOUSEHOLDS AND EXPENDITURES / FLORIST ; ============================================================; ET SIMNUM = 9; IMTYPE=2 & BIT COEFFICIENTS FOR EXPENDITURES THROUGH FLORISTS; YPE=1); R EXPENDITURES ON ALL OUTLETS; SIMSHZ; O; ==================================================================== ; ? S M C COEFTRN_S3*(SIMTYPE=2 & SIMOUTL=3)+ COEFTRN_S4*(SIMTY SIMOUTL=4); ? TOBIT COEFFICIENTS FOR EXPENDITURES THROUGH MAT COEF_PROBIT = COEF_TRN*(SIMTYPE=2) +COEF_EXP*(SIMTYPE=1); ? PR DOT(VALUE=K) 0 3 4 5; ? FORMS; DO G=0 TO 2; INIT; SET SIM_FORM=K; SET SIM_PURPOSE=G; ZSIMSHZ; ENDDO; ENDDOT; ?======== ? SIMULA ? ? S MAT COEF_TOBIT = COEFEXP_S3*(SIMTYPE=1 & SIMOUTL=3)+ COEFEXP_S4*(SIMTYPE=1 & SIMOUTL=4)+ COEFTRN_S3*(SIMTYPE=2 & SIMOUTL=3)+ COEFTRN_S4*(S SIMOUTL=4); ? TO MAT COEF_PROBIT = COEF_TRN*(SIMTYPE=2) +COEF_EXP*(SIMT ? PROBIT COEFFICIENTS FO DO K=1 TO 4; DO D=1 TO 12; INIT; SET SIM_INCOME=K; SET SIM_MONTH=D; Z ENDDO; ENDD ? ==============================; ? NOW TYPE =1 AND OUTL=4;

PAGE 138

124 ?====================================================================== ============================; ; ========================; = 1; OEFEXP_S4*(SIMTYPE=1 & SIMOUTL=4)+ =2 & SIMOUTL=3)+ COEFTRN_S4*(SIMTYPE=2 & ? TOBIT COEFFICIENTS FOR EXPENDITURES THROUGH FLORISTS; E=2) +COEF_EXP*(SIMTYPE=1); ? PROBIT COEFFICIENTS FOR EXPENDITURES ON ALL OUTLETS; SIMSHZ; M_AGE=K; ============================================================; SIMULATION =3 ; OVER GENDER OF HOUSEHOLDS AND EXPENDITURES / FLORIST ; ============================================================; ET SIMNUM = 3; AT COEF_TOBIT = COEFEXP_S3*(SIMTYPE=1 & SIMOUTL=3)+ SIMOUTL=3)+ COEFTRN_S4*(SIMTYPE=2 & ?============================================================; ? SIMULATION =1 ? AVERAGE HOUSEHOLD AND EXPENDITURES / FLORIST ; ?==================================== SET K =1; SET SIMNUM SET SIMTYPE = 1; ? EXPENDITURES (1) AND TRANSACTIONS (2); SET SIMOUTL = 4; ? FLORISTS (3) SUPERMARKETS (4); MAT COEF_TOBIT = COEFEXP_S3*(SIMTYPE=1 & SIMOUTL=3)+ C COEFTRN_S3*(SIMTYPE SIMOUTL=4); MAT COEF_PROBIT = COEF_TRN*(SIMTYP INIT; Z ?============================================================; ? SIMULATION =2 ; ? OVER AGES OF HOUSEHOLDS AND EXPENDITURES / FLORIST ; ?============================================================; SET SIMNUM = 2; MAT COEF_TOBIT = COEFEXP_S3*(SIMTYPE=1 & SIMOUTL=3)+ COEFEXP_S4*(SIMTYPE=1 & SIMOUTL=4)+ COEFTRN_S3*(SIMTYPE=2 & SIMOUTL=3)+ COEFTRN_S4*(SIMTYPE=2 & SIMOUTL=4); ? TOBIT COEFFICIENTS FOR EXPENDITURES THROUGH FLORISTS; MAT COEF_PROBIT = COEF_TRN*(SIMTYPE=2) +COEF_EXP*(SIMTYPE=1); ? PROBIT COEFFICIENTS FOR EXPENDITURES ON ALL OUTLETS; DO K=0 TO 4; INIT; SET SI ZSIMSHZ; ENDDO; ? ? ? ? S M COEFEXP_S4*(SIMTYPE=1 & SIMOUTL=4)+ COEFTRN_S3*(SIMTYPE=2 & SIMOUTL=4);

PAGE 139

125 ? TOBIT COEFFICIENTS FOR EXPENDITURES THROUGH FLORISTS;MAT COEF_PROBIT = COEF_TRN*(SIM TYPE=2) +COEF_EXP*(SIMTYPE=1); TLETS; ENDER=K; =4 ; OEF_TOBIT = COEFEXP_S3*(SIMTYPE=1 & SIMOUTL=3)+ S4*(SIMTYPE=1 & SIMOUTL=4)+ COEFTRN_S3*(SIMTYPE=2 & SIMOUTL=3)+ COEFTRN_S4*(SIMTYPE=2 & IMOUTL=4); ? TOBIT COEFFICIENTS FOR EXPENDITURES THROUGH FLORISTS; R EXPENDITURES ON ALL OUTLETS; SIMULATION =5 ; E OF HOUSEHOLDS AND EXPENDITURES / FLORIST ; ========================================================; ; _TOBIT = COEFEXP_S3*(SIMTYPE=1 & SIMOUTL=3)+ _S4*(SIMTYPE=1 & SIMOUTL=4)+ COEFTRN_S3*(SIMTYPE=2 & SIMOUTL=3)+ COEFTRN_S4*(SIMTYPE=2 & IMOUTL=4); ORISTS; =2) +COEF_EXP*(SIMTYPE=1); UTLETS; ? PROBIT COEFFICIENTS FOR EXPENDITURES ON ALL OU DO K=0 TO 2; INIT; SET SIM_G ZSIMSHZ; ENDDO; ?============================================================; ? SIMULATION ? OVER PURPOSE OF HOUSEHOLDS AND EXPENDITURES / FLORIST ; ?============================================================; SET SIMNUM = 4; MAT C COEFEXP_ S MAT COEF_PROBIT = COEF_TRN*(SIMTYPE=2) +COEF_EXP*(SIMTYPE=1); ? PROBIT COEFFICIENTS FO DO K=0 TO 2; INIT; SET SIM_PURPOSE=K; ZSIMSHZ; ENDDO; ?============================================================; ? ? OVER INCOM ?==== SET SIMNUM = 5 MAT COEF COEFEXP S ? TOBIT COEFFICIENTS FOR EXPENDITURES THROUGH FL MAT COEF_PROBIT = COEF_TRN*(SIMTYPE ? PROBIT COEFFICIENTS FOR EXPENDITURES ON ALL O DO K=0 TO 4; INIT; SET SIM_INCOME=K; ZSIMSHZ; ENDDO;

PAGE 140

126 ?=========== =================================================; LATION =6 ; OUSEHOLDS AND EXPENDITURES / FLORIST ; ====================================================; UM = 6; AT COEF_TOBIT = COEFEXP_S3*(SIMTYPE=1 & SIMOUTL=3)+ OEFEXP_S4*(SIMTYPE=1 & SIMOUTL=4)+ COEFTRN_S3*(SIMTYPE=2 & SIMOUTL=3)+ COEFTRN_S4*(SIMTYPE=2 & EXPENDITURES THROUGH FLORISTS; ); TLETS; =================================================; LATION =7 ; SEHOLDS AND EXPENDITURES / FLORIST ; ====================================================; UM = 7; AT COEF_TOBIT = COEFEXP_S3*(SIMTYPE=1 & SIMOUTL=3)+ OEFEXP_S4*(SIMTYPE=1 & SIMOUTL=4)+ COEFTRN_S3*(SIMTYPE=2 & SIMOUTL=3)+ COEFTRN_S4*(SIMTYPE=2 & EXPENDITURES THROUGH FLORISTS; 1); TLETS; ========================================================; ; RMS AND PURPOSE OF HOUSEHOLDS AND EXPENDITURES / FLORIST ? SIMU ? OVER SEASONS OF H ?======== SET SIMN M C SIMOUTL=4); ? TOBIT COEFFICIENTS FOR MAT COEF_PROBIT = COEF_TRN*(SIMTYPE=2) +COEF_EXP*(SIMTYPE=1 ? PROBIT COEFFICIENTS FOR EXPENDITURES ON ALL OU DO K=0 TO 12; INIT; SET SIM_MONTH=K; ZSIMSHZ; ENDDO; ?=========== ? SIMU ? OVER FORMS OF HOU ?======== SET SIMN M C SIMOUTL=4); ? TOBIT COEFFICIENTS FOR MAT COEF_PROBIT = COEF_TRN*(SIMTYPE=2) +COEF_EXP*(SIMTYPE= ? PROBIT COEFFICIENTS FOR EXPENDITURES ON ALL OU DOT(VALUE=K) 0 3 4 5; INIT; SET SIM_FORM=K; ZSIMSHZ; ENDDOT; ?==== ? SIMULATION =8 ? OVER FO ;

PAGE 141

127 ? ============================================================; ET SIMNUM = 8; AT COEF_TOBIT = COEFEXP_S3*(SIMTYPE=1 & SIMOUTL=3)+ OEFEXP_S4*(SIMTYPE=1 & SIMOUTL=4)+ PE=2 & FLORISTS; OBIT COEFFICIENTS FOR EXPENDITURES ON ALL OUTLETS; ====================================================; TION =9 ; OVER INCOME OF HOUSEHOLDS AND EXPENDITURES / FLORIST ; ============================================================; ET SIMNUM = 9; IMTYPE=2 & BIT COEFFICIENTS FOR EXPENDITURES THROUGH FLORISTS; YPE=1); R EXPENDITURES ON ALL OUTLETS; SIMSHZ; O; ================================================================================================; S M C COEFTRN_S3*(SIMTYPE=2 & SIMOUTL=3)+ COEFTRN_S4*(SIMTY SIMOUTL=4); ? TOBIT COEFFICIENTS FOR EXPENDITURES THROUGH MAT COEF_PROBIT = COEF_TRN*(SIMTYPE=2) +COEF_EXP*(SIMTYPE=1); ? PR DOT(VALUE=K) 0 3 4 5; ? FORMS; DO G=0 TO 2; INIT; SET SIM_FORM=K; SET SIM_PURPOSE=G; ZSIMSHZ; ENDDO; ENDDOT; ?======== ? SIMULA ? ? S MAT COEF_TOBIT = COEFEXP_S3*(SIMTYPE=1 & SIMOUTL=3)+ COEFEXP_S4*(SIMTYPE=1 & SIMOUTL=4)+ COEFTRN_S3*(SIMTYPE=2 & SIMOUTL=3)+ COEFTRN_S4*(S SIMOUTL=4); ? TO MAT COEF_PROBIT = COEF_TRN*(SIMTYPE=2) +COEF_EXP*(SIMT ? PROBIT COEFFICIENTS FO DO K=1 TO 4; DO D=1 TO 12; INIT; SET SIM_INCOME=K; SET SIM_MONTH=D; Z ENDDO; ENDD ? ==============================; ? NOW TYPE =2 AND OUTL=4; ?======================================================================

PAGE 142

128 ?============================================================; ? SIMULATION =1 ; ========================; = 1; OEFEXP_S4*(SIMTYPE=1 & SIMOUTL=4)+ =2 & SIMOUTL=3)+ COEFTRN_S4*(SIMTYPE=2 & ? TOBIT COEFFICIENTS FOR EXPENDITURES THROUGH FLORISTS; E=2) +COEF_EXP*(SIMTYPE=1); ? PROBIT COEFFICIENTS FOR EXPENDITURES ON ALL OUTLETS; SIMSHZ; M_AGE=K; ============================================================; SIMULATION =3 ; OVER GENDER OF HOUSEHOLDS AND EXPENDITURES / FLORIST ; ============================================================; ET SIMNUM = 3; AT COEF_TOBIT = COEFEXP_S3*(SIMTYPE=1 & SIMOUTL=3)+ SIMOUTL=3)+ COEFTRN_S4*(SIMTYPE=2 & TYPE=2) +COEF_EXP*(SIMTYPE=1); ? AVERAGE HOUSEHOLD AND EXPENDITURES / FLORIST ; ?==================================== SET K =1; SET SIMNUM SET SIMTYPE = 2; ? EXPENDITURES (1) AND TRANSACTIONS (2); SET SIMOUTL = 4; ? FLORISTS (3) SUPERMARKETS (4); MAT COEF_TOBIT = COEFEXP_S3*(SIMTYPE=1 & SIMOUTL=3)+ C COEFTRN_S3*(SIMTYPE SIMOUTL=4); MAT COEF_PROBIT = COEF_TRN*(SIMTYP INIT; Z ?============================================================; ? SIMULATION =2 ; ? OVER AGES OF HOUSEHOLDS AND EXPENDITURES / FLORIST ; ?============================================================; SET SIMNUM = 2; MAT COEF_TOBIT = COEFEXP_S3*(SIMTYPE=1 & SIMOUTL=3)+ COEFEXP_S4*(SIMTYPE=1 & SIMOUTL=4)+ COEFTRN_S3*(SIMTYPE=2 & SIMOUTL=3)+ COEFTRN_S4*(SIMTYPE=2 & SIMOUTL=4); ? TOBIT COEFFICIENTS FOR EXPENDITURES THROUGH FLORISTS; MAT COEF_PROBIT = COEF_TRN*(SIMTYPE=2) +COEF_EXP*(SIMTYPE=1); ? PROBIT COEFFICIENTS FOR EXPENDITURES ON ALL OUTLETS; DO K=0 TO 4; INIT; SET SI ZSIMSHZ; ENDDO; ? ? ? ? S M COEFEXP_S4*(SIMTYPE=1 & SIMOUTL=4)+ COEFTRN_S3*(SIMTYPE=2 & SIMOUTL=4); ? TOBIT COEFFICIENTS FOR EXPENDITURES THROUGH FLORISTS;MAT COEF_PROBIT = COEF_TRN*(SIM

PAGE 143

129 ? PROBIT COEFFICIENTS FOR EXPENDITURES ON ALL OUT LETS; ENDER=K; =4 ; OEF_TOBIT = COEFEXP_S3*(SIMTYPE=1 & SIMOUTL=3)+ S4*(SIMTYPE=1 & SIMOUTL=4)+ COEFTRN_S3*(SIMTYPE=2 & SIMOUTL=3)+ COEFTRN_S4*(SIMTYPE=2 & IMOUTL=4); ? TOBIT COEFFICIENTS FOR EXPENDITURES THROUGH FLORISTS; R EXPENDITURES ON ALL OUTLETS; SIMULATION =5 ; E OF HOUSEHOLDS AND EXPENDITURES / FLORIST ; ========================================================; ; _TOBIT = COEFEXP_S3*(SIMTYPE=1 & SIMOUTL=3)+ _S4*(SIMTYPE=1 & SIMOUTL=4)+ COEFTRN_S3*(SIMTYPE=2 & SIMOUTL=3)+ COEFTRN_S4*(SIMTYPE=2 & IMOUTL=4); ORISTS; =2) +COEF_EXP*(SIMTYPE=1); UTLETS; DO K=0 TO 2; INIT; SET SIM_G ZSIMSHZ; ENDDO; ?============================================================; ? SIMULATION ? OVER PURPOSE OF HOUSEHOLDS AND EXPENDITURES / FLORIST ; ?============================================================; SET SIMNUM = 4; MAT C COEFEXP_ S MAT COEF_PROBIT = COEF_TRN*(SIMTYPE=2) +COEF_EXP*(SIMTYPE=1); ? PROBIT COEFFICIENTS FO DO K=0 TO 2; INIT; SET SIM_PURPOSE=K; ZSIMSHZ; ENDDO; ?============================================================; ? ? OVER INCOM ?==== SET SIMNUM = 5 MAT COEF COEFEXP S ? TOBIT COEFFICIENTS FOR EXPENDITURES THROUGH FL MAT COEF_PROBIT = COEF_TRN*(SIMTYPE ? PROBIT COEFFICIENTS FOR EXPENDITURES ON ALL O DO K=0 TO 4; INIT; SET SIM_INCOME=K; ZSIMSHZ; ENDDO;

PAGE 144

130 ?=========== =================================================; LATION =6 ; OUSEHOLDS AND EXPENDITURES / FLORIST ; ====================================================; UM = 6; AT COEF_TOBIT = COEFEXP_S3*(SIMTYPE=1 & SIMOUTL=3)+ OEFEXP_S4*(SIMTYPE=1 & SIMOUTL=4)+ COEFTRN_S3*(SIMTYPE=2 & SIMOUTL=3)+ COEFTRN_S4*(SIMTYPE=2 & EXPENDITURES THROUGH FLORISTS; ); LETS; =================================================; LATION =7 ; SEHOLDS AND EXPENDITURES / FLORIST ; ====================================================; UM = 7; AT COEF_TOBIT = COEFEXP_S3*(SIMTYPE=1 & SIMOUTL=3)+ OEFEXP_S4*(SIMTYPE=1 & SIMOUTL=4)+ COEFTRN_S3*(SIMTYPE=2 & SIMOUTL=3)+ COEFTRN_S4*(SIMTYPE=2 & EXPENDITURES THROUGH FLORISTS; 1); LETS; ========================================================; ; RMS AND PURPOSE OF HOUSEHOLDS AND EXPENDITURES / FLORIST ============================================================; ET SIMNUM = 8; ? SIMU ? OVER SEASONS OF H ?======== SET SIMN M C SIMOUTL=4); ? TOBIT COEFFICIENTS FOR MAT COEF_PROBIT = COEF_TRN*(SIMTYPE=2) +COEF_EXP*(SIMTYPE=1 ? PROBIT COEFFICIENTS FOR EXPENDITURES ON ALL OUT DO K=0 TO 12; INIT; SET SIM_MONTH=K; ZSIMSHZ; ENDDO; ?=========== ? SIMU ? OVER FORMS OF HOU ?======== SET SIMN M C SIMOUTL=4); ? TOBIT COEFFICIENTS FOR MAT COEF_PROBIT = COEF_TRN*(SIMTYPE=2) +COEF_EXP*(SIMTYPE= ? PROBIT COEFFICIENTS FOR EXPENDITURES ON ALL OUT DOT(VALUE=K) 0 3 4 5; INIT; SET SIM_FORM=K; ZSIMSHZ; ENDDOT; ?==== ? SIMULATION =8 ? OVER FO ; ? S

PAGE 145

131 M AT COEF_TOBIT = COEFEXP_S3*(SIMTYPE=1 & SIMOUTL=3)+ OEFEXP_S4*(SIMTYPE=1 & SIMOUTL=4)+ PE=2 & FLORISTS; OBIT COEFFICIENTS FOR EXPENDITURES ON ALL OUTLETS; ====================================================; TION =9 ; OVER INCOME OF HOUSEHOLDS AND EXPENDITURES / FLORIST ; ============================================================; ET SIMNUM = 9; IMTYPE=2 & BIT COEFFICIENTS FOR EXPENDITURES THROUGH FLORISTS; YPE=1); R EXPENDITURES ON ALL OUTLETS; SIMSHZ; O; RMAT=EXCEL,FILE='C:\zstudent\Christian_Iniguez\SIMPROB#3.XLS') SHAREM; ND; C COEFTRN_S3*(SIMTYPE=2 & SIMOUTL=3)+ COEFTRN_S4*(SIMTY SIMOUTL=4); ? TOBIT COEFFICIENTS FOR EXPENDITURES THROUGH MAT COEF_PROBIT = COEF_TRN*(SIMTYPE=2) +COEF_EXP*(SIMTYPE=1); ? PR DOT(VALUE=K) 0 3 4 5; ? FORMS; DO G=0 TO 2; INIT; SET SIM_FORM=K; SET SIM_PURPOSE=G; ZSIMSHZ; ENDDO; ENDDOT; ?======== ? SIMULA ? ? S MAT COEF_TOBIT = COEFEXP_S3*(SIMTYPE=1 & SIMOUTL=3)+ COEFEXP_S4*(SIMTYPE=1 & SIMOUTL=4)+ COEFTRN_S3*(SIMTYPE=2 & SIMOUTL=3)+ COEFTRN_S4*(S SIMOUTL=4); ? TO MAT COEF_PROBIT = COEF_TRN*(SIMTYPE=2) +COEF_EXP*(SIMT ? PROBIT COEFFICIENTS FO DO K=1 TO 4; DO D=1 TO 12; INIT; SET SIM_INCOME=K; SET SIM_MONTH=D; Z ENDDO; ENDD WRITE(FO M E

PAGE 146

LIST OF REFERENCES Ahl, David H., New Product Foreca sting Using Consumer Panels, Journal of g Research, May 1970, 7, 160-167. hen the Dependent Variable Is Truncated ometrica, November 1973, 41, 997-1016. merican Floral Endowment. American Flower Endowment Tracking Study, 1992-2004 ain, Joe S., Relation of Profit-Rate to Industry Concentration: American 1, 65, nuary 1965, 24, 47-58. tionship between Attributes, Brand d May 1993, 75, 314-320. James Blaylock, U.S. Cigarette Consumption: The Case of Low-gricultural Economics, August 1992, 74, James L., Thomas F. Cooley, & Thomas E. Hall, A New View of the market ance Debate, The Journal of Industrial Economics, June 1984, 32, 397-417. lph M., & David R. Ross, Can Small Firms Find and Defend Strategic Niches? A Test of the Porter Hypothesis, The Review of Economics and Statistics, 71, 258-262. Buzzell, Robert D., Bradley T. Gale, & Ralph G.M. Sultan [1975], Market-Share A Key to Profitability, Harvard Business Review, January, 97-106. Marketin Amemiya, Takeshi, Regression Analysis W Normal, Econ A B Manufacturing, 1936-1940, Quarterly Journal of Economics, August 195 293-324. Basmann, R.L., A Theory of Demand with Variable Consumer Preferences, Econometrica, Ja Blin, Jean-Marie, & Joe A. Dodson, The Rela Preferences, and Choice: A Stochastic View, Management Science, June 1980, 26606-619. Blisard, Noel, & James Blaylock, Distinguishing between Market Participation an Infrequency of Purchase Models of Butter Demand, American Journal of Agricultural Economics Blisar d, Noel, & Income Women, American Journal of A05. 698-7 Bothwell,StructurePerform Bradburd, Ra May 1989, 132

PAGE 147

133 Byrne, Patrick J., Oral Capps, Jr., & Atanu Saha, Analysis of Food-Away-from-Home Expenditures Patterns for U.S. Households, 1982-89, American Journal of Agricultural Economics, August 1996, 78, 614-627. Carpenter, Gregory S., Rashi Glazer, & Kent Nakamoto, Meaningful Brands from Meaningless Differentiation: The Dependence on Irrelevant Attributes, Journal of Marketing Research, Augth, 97, 18, 247-254. Cooper, Lee G. & Masao Nakanishi Market-Share Analysis: Evaluating Competitive Craggimited Dependent Variables with Dems, Industry Structure, Market Rivalry, and Public Policy, Journal of Feinbesearch, May 1992, 29, 227-237. Gale, Bradley T., Market Share and Rate of Return, The Review of Economics and Ghosh, Arabinda, Market Structure Research: Comment, Journal of Farm Economics, August 1966, 48, 751-753. Ghosrocedures, Journal of Marketing Research, May 1984, 21, 202-210. Gilboory, International Economic Review, November 1997, 38, 737-761. ust 1994, 31, 339-350. Chauvin, Keith W., & Mark Hirschey, Market Structure and the Value of GrowManagerial and Decision Economics, May 19 Chay, Kenneth Y., & James L. Powell, Semiparametric Censored Regression Models,The Journal of Economic Perspectives, Autumn 2001, 15, 29-42. Clodious, Robert L., & Willard F. Mueller, Market Structure as an Orientation for Research in Agricultural Economics, Journal of Farm Economics, August 1961, 43, 515-543. Marketing Effectiveness, Norwell, MA. Kluwer, 1988. John G., Some Statistical Models for L Application to the Demand for Durable Goods, Econometrica, September 1971, 39, 829 844. etz, Harold Law and Economics, October 1973, 16, 1-9. erg, Fred M., Barbara E. Kahn, & Leigh McAlister, Market Share When Consumers Seek Variety, Journal of Marketing R Fossum, M. Truman, Horticultural Specialties: A Neglected Segment of U.S. Agriculture, Journal of Farms Economics, November 1953, 35, 622-628. Statistics, November 1972, 54, 412-423. h, Avijit, Scott Neslin, & Robert Shoemaker, A Comparison of Market Share Models and Estimation P a, Itzhak, & David Schmeidler, Cumulative Utility Consumer The

PAGE 148

134 Girapunthong, Napaporn and Ronald, Ward, Demand Drivers for Fresh-cut Flowers andTheir Substitutes: An Application of Household Expenditure Allocation ModJournal of American Agricultural Economics, July 20 els, 03. l, Greene, William H., Econometric Analysis, Fourth Edition, Englewood Cliffs, NJ: Grover, Rajiv, & V. Srinivasan, An Approach for Tracking Within-Segments Shifts in Haines, Pamela S., David K Guilkey, & Barry M. Popkin, Modeling Food Consumption cs, Heckman, James J., Sample Selection Bias as a Specification Error, Econometrica, Holbrook, Morris B., & Elizabeth C. Hirschman, The Experiential Aspects of sumer Hoyer, Wayne D., An Examination of Consumer Decision Making for a Common Jacobson, Robert, Distinguishing among Competing Theories of the Market Share Kotler, Philip, Marketing management: Analysis, Planning, and Control, Fifth Edition, Long, J. Scott Long, Regression Models for Categorical and Limited Dependent Naert, Philippe A. & Alain Bultez, Logically Consistent Market Share Models, Journal Peltzman, Sam, The Gains and Losses from Industrial Concentration, Journal of Law and Economics, October 1977, 20, 229-63. Gould, Brian W., At-Home Consumption of Cheese: A Purchase-Infrequency ModeAmerican Journal of Agricultural Economics, May 1992, 74, 453-459. Prentice-Hall, Inc., 2003. Market Shares, Journal of marketing Research, May 1989, 26, 230-236. Decisions as a Two-Step Process, American Journal of Agricultural EconomiAugust 1988, 70, 543-552. January 1979, 47, 153-162. Consumption: Consumer Fantasies, Feelings, and Fun, The Journal of ConResearch, September 1982, 9, 132-140. Repeat Purchase Product, The Journal of Consumer Research, December 1984,11, 822-829. Ipsos-NPD Group. National Panel Diary Group, Inc., Chicago Illinois, 2004 Effect, Journal of Marketing, October 1988, 52, 68-80. Englewood Cliffs, NJ: Prentice-Hall, Inc., 1984. Variables. Advanced Quantitative Techniques in the Social Sciences Thousand Oaks, CA: Sage, 1997. Moriarty, Mark, Cross-Sectional, Time-Series Issues in the Analysis of marketing Decision Variables, Journal of Marketing Research, May 1975, 12, 142-150. of Marketing Research, August 1973, 10, 334-340

PAGE 149

135 Reynolds, Anderson, Analyzing Fresh Vegetable Consumption from Household Survey Data, Southern Journal of Agricultural Economics, December 1990, 31-38. Rhodes, James V., The Measurement of Consumer Preferences, Journal of Farm Smirlock, Michael, Thomas Gilligan,& William Marshall, Tobins q and the Structure-U.S. International Trade Commission, Industry & Trade Summary: Cut Flowers, Van den Broek, Luciano, John J.Haydu, Alan W. Hodges, & Evaristo M. Neves, tes and Ward, Ronald W., National Report on Household Purchases of Fresh Flowers 1993 to Carbon, IL, April 2003. Economics, November 1955, 37, 638-651. Performance Relationship, American Economic Review, December 1984, 74, 1050-60. Smith, Frank J., Jr. & Dale C. Dahl, Market Structure Research. How and for What? Journal of Farm Economics, May 1965, 47, 465-467. Tobin, James, Estimation of Relationships for Limited Dependent Variables, Econometrica, January 1958, 26, 24-36. USITC Publication 3580, February 2003. Production, Marketing and Distribution of Cut Flowers in the United StaBrazil, IFAS Extension, Food & Resource Economics Department, University of Florida, 2002. Walsh, John, Flexibility in Consumer Purchasing for Uncertain Future Tastes, Marketing Science, 1995, 14, 148-165. 2002, Consumer Tracking Study for the American Floral Endowment, Glen

PAGE 150

niguez was born on October 15, 1980, in Cuenca, Ecuador. After il, Ecuador, he was awarded ESPOLs International Scholarship to continue his Bachelor of ScienWPI) in Massachusetts, United States. to the MasteUnive with a full research assistantship. He began his masters program in eting. BIOGRAPHICAL SKETCH Christian I studying two years at the Escuela Superior Politecnica del Litoral (ESPOL) in Guayaqu ce at the Worcester Polytechnic Institute ( After graduation he returned to his home country to teach advanced courses in management at ESPOLs Department of Economics. In 2002, he was accepted inr of Science program of the Food and Resource Economics Department at the rsity of Florida August of 2003 specializing in the fields of industrial organization and food mark 136


Permanent Link: http://ufdc.ufl.edu/UFE0011838/00001

Material Information

Title: Demand Shifts in Outlet Selection in the United States Market for Fresh Flowers
Physical Description: Mixed Material
Copyright Date: 2008

Record Information

Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
System ID: UFE0011838:00001

Permanent Link: http://ufdc.ufl.edu/UFE0011838/00001

Material Information

Title: Demand Shifts in Outlet Selection in the United States Market for Fresh Flowers
Physical Description: Mixed Material
Copyright Date: 2008

Record Information

Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
System ID: UFE0011838:00001


This item has the following downloads:


Full Text












DEMAND SHIFTS IN OUTLET SELECTION IN THE UNITED STATES MARKET
FOR FRESH FLOWERS















By

CHRISTIAN R. INIGUEZ


A THESIS PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
MASTER OF SCIENCE

UNIVERSITY OF FLORIDA


2005

































Copyright 2005

by

Christian R. Iniguez


































To Mami Lety.















ACKNOWLEDGMENTS

This thesis would not have done without the help of Dr. Ronald W. Ward, chair of

my supervisory committee. Over these two years Dr. Ward has been an excellent teacher,

and inspired researcher, academic counselor and a friend. I would like to thank him for

all his support and guidance throughout the duration of this thesis. I would also like to

thank Dr. Ramon Espinel for encouraging me to study at the Food and Resource

Economics Department. His support and advice during these past two years have guided

my academic beliefs. I also would like to thank Dr. Jeffrey Burkhardt, Dr. James Sterns,

Dr. Peter Hildebrand and Dr. Gary Fairchild for their approachable way of teaching

which, combined with their deep knowledge, helped create an enthusiastic learning

experience in the classroom. I would specially like to thank Jessica Herman, our program

supervisor, for all her help during my degree program.

Finally, I would like to thank my parents Roberto and Teresita, my sister Daniela,

and my wife Maria Fernanda. Their support was the only thing that kept me going when

things did not go as planned. At the beginning of the program I lost one of the most

important person in my life, my grandmother Lety, but her memory and kind words were

always there with me. She will always be in my heart.
















TABLE OF CONTENTS



A C K N O W L E D G M E N T S ................................................................................................. iv

LIST OF TABLES ........................................................... .... .... ........ .... vii

LIST OF FIGURES ...................................................... .. .... ........... viii

A B S T R A C T ......... .................................. ...................................................x iii

CHAPTER

1 IN TR O D U C T IO N ............................................................. .. ......... ...... .....

O overview .................................................................................1
O bj ectiv e ...................................................................................... . 4
H ypotheses ................................................ 4
P rob lem Statem en t .................................................................................. 5
S cop e ............................................................... . 5
Types of Stores Selected ........................................................ 6
N ature of the D ata ................................................................ ..........6

2 LITERA TU RE REV IEW ....................................................... 8

Consum er D em and and Preferences ........................................................................
M ark et Sh are C on cepts ................................................................................. 11
C ensored M models ............................................................................................... ....... 15

3 US FRESH FLOWER DEMAND ................................. ...........................19

Indoor Flow er Categories .............................................................................19
Flower Retail Outlets .................................. ........................... ... ....... 21
M market Shares by Product Form ........................................................... .... ........... 22
Share Distribution over Time .................................................26
M market Shares across V ariables.................................................. 31

4 THEORETICAL MODEL AND MODEL SPECIFICATION ............... ...............37

C onsum er D em and for Flow ers ............................................................................. 37
D details o n th e D ata ................................................................................................. 3 8



v









C ensored D ata M odel .............................. ......................... ... ...... .... ..... ...... 38
Tobit M odel for O utlet Selection........................................... .......... ............... 44
First-Step Results: Decision to Become a Buyer............................................ 46
Second-Step Results: Intensity of Buying .............. ............................................51

5 M ODEL SIM ULATION S .......................................................... ............... 56

Expected Share for the Average Conditions.................................... ...............57
Expected Outlet Shares by Demographics ...................................... ............... 59
Outlet Shares by Household Age Groups .........................................................59
O utlet Share over G ender ............................................................... ............... 60
Buying Purpose Impact on Outlet Shares.........................................................60
Outlet Shares Across Incomes.............. ....... ..................................... 61
Outlet Shares over Flow er Form s................................... ........................ 64
Combined Effect of Purpose and Form........................................... 64
Sim ulations by Seasons ................................................ ........................... 66
Rankings Factors Impacting the Outlet Shares................................ ...............70
Dynamics in the Outlet Share Coefficients ..................................... .................71

6 C O N C L U SIO N ......... ......................................................................... ........ .. ..... .. 74

Intro du action .................................................................................. ..................... 74
O verview of O utlet A nalyses........................................................... ............... 74
Major Outlet Selection Conclusions................ ...... .............. ... ............... 77
L im stations ....................................................................................... 8 1
R ecom m endations............... ................................ .... .... ......... ........ 82

APPENDIX

A IND U STRY OVERVIEW .............................................. .............................. 84

B TIME RECURSIVE COEFFICIENTS ................ .............................................101

C T SP C O D E ......................................................................... 109

L IST O F R E FE R E N C E S ........................................................................ ................... 132

BIOGRAPHICAL SKETCH ............................................................. ............... 136
















LIST OF TABLES

Table p

4-1 Description for the Variables in the Heckman model ...........................................47

4-2 Estimated Probit and Tobit Coefficients for Expenditures ......................................48

4-3 Estimated Probit and Tobit Coefficients for Transactions .................................49
















LIST OF FIGURES


Figure page

3-1 Percent of household market shares expenditures and transactions on indoor
flow ers ............................................................... .... ..... ......... 20

3-2 Percent of specialty market shares on expenditures and transactions for cut-
flow ers ............................................................... .... ..... ......... 23

3-3 Percent of mass merchandising market shares on expenditures and transactions
for cu t-flow ers............................. .................................................. ............... 2 3

3-4 Percent of specialty market shares on expenditures and transactions for
flow ering/green house plants. ...... ....................................................................24

3-5 Percent of mass merchandising market shares on expenditures and transactions
for flow ering/green house plants.. ...................... ............................................. 24

3-6 Percent of household market shares for florists by product form........................25

3-7 Percent of household market shares for supermarkets by product form ...............25

3-8 Percent of yearly market shares in cut-flowers for outlet groups by
expenditures and transactions. ........................................ ......................... 27

3-9 Percent of yearly market shares in flowering/green house plants for outlet
groups by expenditures and transactions.. .................................... ...............28

3-10 Percent of yearly market shares in cut-flowers for florists and supermarkets by
expenditures and transactions. ........................................ ......................... 29

3-11 Percent of yearly market shares in flowering/green house plants for florists and
supermarkets by expenditures and transactions..................................................30

3-12 Percent of yearly market shares in arrangements for florists and supermarkets
by expenditures and transactions.. ...................... ............................................. 32

3-13 Percent of yearly market shares in non-arrangements for florists and
supermarkets by expenditures and transactions................................................32

3-14 Percent market shares of cut-flowers expenditures and transactions by
dem graphics. ...................................................................... 34









3-15 Percent market shares of flowering/green house plants expenditures and
transactions by dem graphics. ........................................ .......................... 34

3-16 Percent of cut-flowers market shares expenditures and transactions by outlets....35

3-17 Percent of flowering/green house plants market shares expenditures and
transactions by outlets....................................... ... .. ........ .. ........ .... 36

4-1 Distribution of values in the first stage probit model. .....................................46

4-2 Distribution of values in the second stage tobit model for florists ...................52

5-1 Distribution of shares ........... ... ................ ..... ... .... .... ........... 57

5-2 Average outlet shares the fresh flower market for florists and supermarkets........58

5-3 Florist and supermarket probabilities over age.................................... ............... 61

5-4 Florist and supermarket probabilities over gender.................... ................62

5-5 Florist and supermarket probabilities over purpose ...........................................63

5-6 Florist and supermarket probabilities over income ............................................63

5-7 Florist and supermarket probabilities over form.................................................65

5-8 Florist probabilities over purpose and form ................................ ............... 66

5-9 Supermarket probabilities over purpose and form........................ .................67

5-10 Florist and supermarket probabilities over months ............................................68

5-11 Florist probabilities over income and months by expenditures ...........................68

5-12 Supermarket probabilities over income and months by expenditures .................69

5-13 Variable rankings for florist ................................................ 72

5-14 Variable rankings for supermarket.......................................... ...............72

5-15 Time Varying Coefficients for the Average Consumer...................................73

A-i Percent of yearly market shares for specialty based in cut flowers by
expenditures and transactions. ........................................ ......................... 84

A-2 Percent of yearly market shares for specialty based in flowering/green house
plants by expenditures and transactions ............... ....................... ............... 85









A-3 Percent of yearly market shares for mass merchandising based in cut flowers
by expenditures and transactions. ........................................ ....... ............... 85

A-4 Percent of yearly market shares for mass merchandising based in
flowering/green house plants by expenditures and transactions............................86

A-5 Percent of monthly market shares based in cut flowers by expenditures and
tran action s. ...................................................... ................. 86

A-6 Percent of monthly market shares based in flowering/green house plants by
expenditures and transactions. ........................................ ......................... 87

A-7 Percent of monthly specialty market shares based in cut flowers by
expenditures and transactions. ........................................ ......................... 87

A-17 Percent of monthly specialty market shares based in flowering/green house
plants by expenditures and transactions............... ............................ 88

A-18 Percent of monthly mass merchandising market shares based in cut flowers by
expenditures and transactions. ........................................ ......................... 88

A-19 Percent of monthly mass merchandising market shares based in
flowering/green house plants cut flowers by expenditures and transactions.........89

A-20 Percent of monthly market shares in cut flowers for florists and supermarkets
by expenditures and transactions. ........................................ ....... ............... 89

A-21 Percent of monthly market shares in flowering/green house plants for florists
and supermarkets by expenditures and transactions...........................................90

A-22 Percent of monthly market shares in arrangements for florists and
supermarkets by expenditures and transactions................................................90

A-23 Percent of monthly market shares in non-arrangements for florists and
supermarkets by expenditures and transactions................................................91

A-24 Distribution of market shares based in cut flowers arrangements by
expenditures and transactions. ........................................ ......................... 91

A-25 Distribution of market shares based in cut flowers non-arrangements by
expenditures and transactions. ........................................ ......................... 92

A-26 Distribution of market shares based on age by expenditures and transactions ......92

A-27 Distribution of market shares based on income by expenditures and
tran action s. ...................................................... ................. 93

A-28 Distribution of market shares based on purpose by expenditures and
tran action s. ...................................................... ................. 93









A-29 Distribution of market shares based on gender by expenditures and
transactions...................................................................... ..........94

A-30 Distribution of market shares for specific outlets in cut flowers based on age
(first and second group) by expenditures and transactions ..................................94

A-31 Distribution of market shares for specific outlets in flowering/green house
plants based on age (first and second group) by expenditures and transactions....95

A-32 Distribution of market shares for specific outlets in cut flowers based on age
(third and fourth group) by expenditures and transactions.................................95

A-33 Distribution of market shares for specific outlets in flowering/green house
plants based on age (third and fourth group) by expenditures and transactions....96

A-34 Distribution of market shares for specific outlets in cut flowers based on
gender by expenditures and transactions. ................................... ............... 96

A-35 Distribution of market shares for specific outlets in flowering/green house
plants based on gender by expenditures and transactions............... .................97

A-36 Distribution of market shares for specific outlets in cut flowers based on
income (first and second group) by expenditures and transactions .....................97

A-37 Distribution of market shares for specific outlets in flowering/green house
plants based on income (first and second group) by expenditures and
tran action s. ...................................................... ................. 9 8

A-38 Distribution of market shares for specific outlets in cut flowers based on
income (third and fourth group) by expenditures and transactions .....................98

A-39 Distribution of market shares for specific outlets in flowering/green house
plants based on income (third and fourth group) by expenditures and
tran action s. ...................................................... ................. 99

A-40 Distribution of market shares for specific outlets in cut flowers based on
purpose by expenditures and transactions...................... ..................... 99

A-41 Distribution of market shares for specific outlets in flowering/green house
plants based on purpose by expenditures and transactions ..............................100

B-1 Time recursive parameters for the under 25 years of age group.........................101

B-2 Time recursive parameters for the 25 to 39 years of age group.........................101

B-3 Time recursive parameters for the 40 to 54 years of age group......................... 102

B-4 Time recursive parameters for the 55 and more years of age group....................102









B-5 Time recursive parameters for females ...... ......... ...................................... 103

B-6 Tim e recursive param eters for m ales ............................................................. 103

B-7 Time recursive parameters for gift............................................. ...............104

B-8 Tim e recursive parameters for self. ........................................ ............... 104

B-9 Time recursive parameters for the under $25,000 income group ......................105

B-10 Time recursive parameters for the $25,000 to $49,999 income group..............105

B-11 Time recursive parameters for the $50,000 to $74,999 income group..............106

B-12 Time recursive parameters for the $75,000 and more income group ................106

B-13 Time recursive parameters for arrangements.................... .................. ................107

B-14 Time recursive parameters for non-arrangements. ............................................107

B-15 Time recursive parameters for flowering/green house plants..............................108















Abstract of Thesis Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Master of Science

DEMAND SHIFTS IN OUTLET SELECTION IN THE UNITED STATES MARKET
FOR FRESH FLOWERS

By

Christian R. Iniguez

August 2005

Chair: Dr. Ronald W. Ward
Major Department: Food and Resource Economics

Over the past decade consumers' outlet selection in the fresh flower industry has

shifted among outlet types, and primarily with changes in using florists and supermarkets.

This study focuses on demand shifts in outlet selection in the fresh flower industry at the

retail level. Florists and supermarkets were chosen because of their high market share

levels seen throughout the industry for these two outlet types. Historically, florists have

attained higher levels of market shares expenditures justified in part by the creative value

added to their products, particularly in the arrangement sector of the industry. In contrast,

supermarkets have focused on the non-arrangement sector which has higher levels of

market share transactions. Over time, supermarkets have increased their market shares

while florists have experienced a decline. If this trend continues, the industry will

experience significant restructuring at the retail level. To quantitatively measure these

outlet shares, an initial sample of households of approximately 189,000 observations was

collected by a professional survey agency. Over a period ten years the agency recorded









the purchasing pattern of approximately 9,000 households using consumer diaries. This

paper used socioeconomic and demographic variables to describe outlet selection in the

fresh flower industry. A two-stage estimation model was used to describe the decision

process faced by potential flower consumers with the first stage estimating the entry of

buyers and the second the level of purchases among buyers. Simulations were used to

forecast consumers' intensity of buying. It is the goal of this study to facilitate the

understanding of factors influencing outlet selection when purchasing fresh flowers. This

knowledge then is important to understanding structural change and for designing

policies that could alter the structure if needed.














CHAPTER 1
INTRODUCTION

Overview

The present research paper addresses issues of consumers' outlet selection in the

fresh flower industry. Flowers fall within the category of goods that are usually

purchased for their perceived aesthetic value. Aesthetic characteristics include decorative,

emotional, environmental, and visual needs of the customer at any given time. For several

years now the flower industry has been experiencing major structural changes. Some of

this change may be attributed to outlet changes, import dependency, direct marketing

practices, packaging, and promotions. This study focuses on outlet changes and how they

may affect market shares distribution in the fresh flower industry. Specifically, changes

in florists and supermarkets shares are of particular interest.

Many factors influence consumer preferences for using a particular outlet type.

Price differences usually reflect the type of product that consumers buy with high value

products usually associated with florists. Florists are usually able to charge a higher price

because of the creative valued added to the flowers (i.e., arrangements). Supermarkets,

which usually sell lower priced products, specialize in selling larger quantities of fresh

flowers with less value added to the flowers. Up to a certain extent, this change in

product preference mix is changing the industry's market structure.

Current demand for cut-flowers among different value added products goes beyond

pricing considerations. Packaging, purpose, occasions, and product range also influence

consumer buying decisions. Historically, supermarkets and florists have differed in the









range of flower products and services provided. As we may see later in this study, florists

have experienced a decline in their market share, thus suggesting underlying preference

changes or alternative sources for the same products. In contrast, supermarkets have

experienced increases both in transactions and expenditures on fresh flowers. Changing

market shares have significant implications to the flower industry in terms of the

competitive buying structure and the types and volumes of purchases by a single outlet.

Historically, florists are quite small and have minimal buying power. Whereas,

considerable concentration among the major retail food chains suggest that more pressure

from the buyers could occur. Yet, growth through supermarkets has the potential for

widely expanding the exposure to fresh flowers simply because of the high traffic

through most grocery stores. Buyer pressure versus expanded consumer exposure to fresh

flowers is a real potential tradeoff that must be considered. Ultimately, for both of these

outlets one must establish the degree of share change and measure what is driving the

changes.

Retail outlets can easily be grouped into four categories defined as: specialty shops,

mass merchandisers, internet, and others. Specialty shops are further divided in florists

and in others subcategories. Florists have historically had a major share of the

arrangement section of the market, providing value added to the retail flowers. Their

business is highly seasonal and influenced by calendar occasions such as Mothers Day,

Valentines, and Christmas, etc. Florists have experienced declining market shares over

the past decade, with supermarkets capturing most of this loss. It is particularly important

to know whether this decrease in overall sales has been the consequences of major supply

changes throughout the vertical market system (e.g. higher costs, low prices).









Mass merchandisers stores are also subcategorized in supermarkets and others.

Supermarkets have focused mainly in the non-arrangement section of the market

providing low cost products for a larger share of the consumers' base. Their main

business are repetitive purchases for non-calendar occasions although they also

experience high peak sales in the above mentioned calendar occasions. Overall,

supermarkets have increased their share of the market relatively to florists. Many factors

could be affecting this trend in the industry, including better inventory practices

(economies of scale applied on highly perishable goods such as flowers), a lower cost

structure, and a bigger target audience.

Other reasons that may influence consumers' preferences are purpose and

convenience. For the present study purpose is divided into two categories; self and gift.

Self often comprises fresh flowers with little value added characteristics, such as non-

arrangements. Gifts are mostly purchased for special or calendar occasions throughout

the years. More recently, internet sales have grown as a visible outlet for fresh flowers.

Internet sales are probably more convenient for the average customer. It might be

particularly important to determine whether the appearance of these sales undermine

florists or supermarkets market share in the long run. Note, however, that internet sales

still usually require delivery of a perishable product that normally requires local services

and particularly florists. Hence, there may be both competition and complementarities

among some of the outlets. In order to avoid data duplication, internet sales are

catalogued as such if they take place on internet retail stores. Internet purchases have

been steadily growing in the past years; however, current data are not sufficient to be

included extensively in the present study.









Objective

This study focuses on expenditures levels and number of transactions at the retail

level in the fresh flower industry using quarterly data over the 1992 through 2004 period.

In general terms, florists consistently have a larger share of household expenditures

relative to supermarkets and other retail outlets. Whereas, supermarkets account for a

larger share of household transactions on fresh flowers. The trend shows that florists are

loosing market share to mass merchandising. A primary objective of this study to

evaluate market share changes in the industry based on household demographic variables,

purpose, flower forms, and occasions.

Hypotheses

Demand for flowers and plants as ornamentals or as environmental investments, as

well as, the emotional needs should depend on the household discretionary incomes. That

is, rising incomes should have a positive effect on demand. Since flower purchases are

somewhat discretionary, sales of floral products may be more responsive to income

changes than more essential goods such as food. However, the expenditure response to

income is expected to reach a point where continual higher incomes would generate

marginal declining response rates. Identifying that particular level would be of particular

importance when developing marketing strategies in the flower industry.

It is expected that upon completion of this study the reader will have a better

understanding of the current and future changes in the outlet structure for the U.S. fresh

flower industry at the retail level. From preliminary data and current trends it is expected

that the increase in expenditures levels and transactions can be explained by identifying

demographics changes, reasons, and product offerings. Beyond those variables expected

to influence the share changes, it is possible that linkages between the shares and the









outlet identifiable variables have also changed. For example, has income become more or

less important as an outlet demand driver? Hence, a major hypothesis is that beyond the

measurable variables driving share changes, there is an underlying shift in the

coefficients linking the shares to the causal variables. In this case, we test if the variables

are time varying.

Problem Statement

Market share changes at the retail level in the flower industry can be illustrated by

determining the probability of selecting these outlets based on household expenditures or

transactions. For the present study we assume a relationship between the characteristics

of the buyer, the product, and the reasons for purchasing among other variables.

Demographics are measured with income levels, gender, and buyer age. Econometric

analysis will capture the importance of each of the variables included in the market share

models. By definition all outlets shares must sum to one if the list of outlet selection is

exhaustive. Yet if one looks at a subset of outlets such as florists or supermarkets, the

share models can possibly be considered separately. Furthermore, depending on the

subclasses for expressing the shares, it is feasible that within some combination of

subclasses that an outlet shares is zero or even one-hundred percent. That is, the shares

may be censored from above and below. Thus the problem becomes one of measuring

market share and their drivers (i.e., causal variables) while dealing with the censored

values. We will see that this is a classic doubled-censored Tobit model where market

shares are estimated while dealing with these upper and lower limits to these shares.

Scope

Reports are compiled from information reported by a panel of around 9,000

nationally representative households who maintain purchasing diaries (Ipsos-NPD).










Ipsos is one of the world's leading market research organizations with one of their major

products being the collection of household purchasing data. Specifically, the household

purchasing data for flowers is from Ipsos-NPD and organized and funded by the

American Flower Endowment through an ongoing consumer tracking study.

Types of Stores Selected

In this study, flowers are grouped into three categories: cut-flowers (arrangements

and non-arrangements), flowering and green house plants (flowering plants and foliage),

and dry/artificial. The retail stores are also grouped into categories such as: specialty,

mass merchandisers, florists' shops, supermarkets, warehouses/price clubs, internet

retailers, and others. According to Ward (2003), most specialty sales are from traditional

florists while supermarkets account for most of the mass merchandising sales. The

demand for floral products, and especially cut-flowers, is highly seasonal. Sales are

normally highest from February through May and drop precipitously in the fall. Sales of

cut-flowers peak during holidays such as Valentine's Day and Mother's Day. Cut-flowers

and foliage plants, however, are increasingly popular throughout the year as indoor home

and workplace decorations.'Aspects of the models will measure these seasonal effects.

Nature of the Data

The data collecting were funded by American Flower Endowment.2 The database

as organized for this thesis includes 82,232 observations at the household level and only

comprises non-commercial purchases. The observations were the result of a professional


1 Economic Research Service. United States Department of Agriculture. The Economics of Food, Farming, Natural Resources, and
Rural America. Floriculture Crops: Background. http://www.ers.usda.gov/Briefing/floriculture/Background.htm
2 The American Flower Endowment (AFE) is the leading not-for-profit, non-governmental source for floriculture/environmental
horticulture research and development funding in the US. For more information see:
http://www.endowment.org/pressrelease/general/spcrpt2001.htm









firm investigates in a method called waves. Within each wave, information is compiled

every two week period. Consumers were given a very detailed consumer diary and they

reported their purchasing habits. The wave dairies can be matched to specific months and

years, thus giving a continuous time series of data. One important consideration is that

the information filled in the diaries is not a recall but the actual purchases made by the

household consumer. Approximately 9,000 demographically balanced households are

included in the survey.

From the database major divisions can be obtained: we know the population, the

number of households, the amount spent, number of transactions or making a purchasing

event, and the quantity of the flowers bought. Note that the quantities have less meaning

because of the diversity of product purchases such as a bunch, arrangement or single

stem. Although the average of weeks that a household remains in the program is about 3

weeks some studies by Ward (2003) have shown that there is no negative influence to the

integrity of the data. From the database we can calculate market penetration (buyers over

households), and frequency (transactions over buyers). Two major limitations of the

database are that it leaves out commercial purchases and only the retail level of the

vertical market system is presented. The present study demonstrates and assesses the

impact of marketing strategies, spending levels, and consumer behavior through the use

of econometric simulations.














CHAPTER 2
LITERATURE REVIEW

The three sections included in this chapter will provide the reader with a basic

understanding of the major topics covered in this study. The first section introduces

consumer choice theory and the effect of preferences in modeling demand. Included in

this section are concepts like utility maximization, product acceptability, and the decision

making process faced by consumers. Market share analysis concepts and alternative

methodologies are discussed in the second section. This section also discusses model

constraints, variable specification, and data aggregation considerations. Finally, the third

section presents econometric models used for censored data. Particular emphasis is given

to models that use a two-step decision process approach to estimate future purchases.

Consumer Demand and Preferences

It is widely accepted that marketing effort can be more successful if it is based on

knowledge regarding consumer preferences. Taking an economics point of view, Rhodes

(1955) presents a general approach to the preference determination problem in consumer

demand. According to him, preference is manifested if the consumer chooses the most

desirable product available to him. However, this presents a problem to the researcher

when the most desirable product is not consumed. It is then important to recognize the

difference between product preference and product acceptability or actual purchases. He

concludes distinguishing that while acceptance of products among consumers is an

absolute definition preference over products is often hard to record. The problem then









becomes one of measuring consumers' preferences and finding appropriative

methodologies that capture the effect on actual consumption.

Traditional economics depicts the consumer as a logical or rational thinker that

maximizes his utility function based on a given budget constraint. Following that line,

Basmann (1956) formulates a theory of demand linking preferences changes and ordinal

utility functions faced by the consumer. However, Hollbrook and Hirschman (1982)

argue that the study of consumer behavior should look beyond the information processing

model of the logical thinker. For him, consumer behavior is also influenced by what he

called "experiential views" which incorporates aesthetic values, enjoyment, sensory

pleasures, and emotional responses. Acknowledging such influence on consumers could

help understand the demand for products that are usually purchased for their aesthetic and

emotional attributes.

It is widely accepted in consumer theory that the analysis of product attributes

linked to expected preferences allows a better forecast of the consumers' future choices.

In their article, Blin and Dodson (1980) analyze the underlying relationships between

attributes, preferences, and choice. In addition, they present two traditional marketing

theories to model consumer choice. First, he describes the multi-attribute expectancy

value model in which attributes are identified, measured, and evaluated based on stated

preference. Second, he presents the stochastic choice model which tries to explain the

complexity of the choice process by using consumer panel diaries to record the

purchasing pattern of consumers over time. One of the shortcomings of the multi-

attribute expectancy model is that it assumes that the consumer will always choose its

preferred brand and that may not always be the case. Furthermore, the model follows a









one-time purchase approach when in fact most questions in consumer theory are oriented

to the frequency of purchases. The strength of the stochastic model relies on the ability to

predict a quantifiable likelihood of choice. However, one of the drawbacks of the model

appears when no product has a higher probability over the other in which case the model

becomes one of attribute differentiation.

Frequency of buying sometimes influences the decision making process of

consumers. In his article, Hoyer (1984) states in that there is a variance associated with

estimating consumer choices over time that could be attributed to the intensity of buying.

Furthermore, he argues that repetitive purchases may reflect not optimal but satisfactory

purchasing decisions in an attempt to minimize consumers' effort and time. Contrasting

the traditional view that assumes that an evaluation is done each time the consumer

makes a choice, he indicates that an evaluation may occur after the product is purchased.

If the evaluation is satisfactory then it will guide future consumption if not then more

refined choices are made by consumers. For products that experience repetitive purchases

throughout the year the issue becomes one of recognizing between product loyalty and

habitual purchases. Brand loyalty purchases usually reflect strong reasons for buying

while habitual purchases are generally done for convenience. Gilboa and Schmeidler

(1997) argue that consumers whose income greatly exceeds the cost of the product will

less likely follow a budget constraint. He based his remarks on the high level of

expenditures seen in higher income groups in the consumption of non-essential products.

In addition, he proposed that repetitive purchases denote "small" choices where the

consumer can afford not to calculate how much they will have left after the purchase.









This would seem to explain why the difference in the demand for nonessential products

can only by partly explained by income disparities.

Finally, Feinberg et al. (1992) presents the long term market share implications of

changes in variety-seeking consumers. He argues that variety seeking is an important

determinant of consumer behavior and should be accounted for in consumer choice

models. Regarding the choices faced by consumers Walsh (1995) argues that the

consumer chooses the alternative most appropriate to him using a cost-benefit ratio. He

concludes that seasonal sales for products that maintain the same attributes or

meaningless differentiation over time can be partly be explained by the occasion of

buying. Carpenter et al. (1994) state that consumers apparently value these differentiating

attributes even though they are irrelevant. According to him, meaningless product

differentiation can stimulate demand by changing consumers' preferences in the long run.

Market Share Concepts

Bothwell et al. (1984) argue that since economics is a non-experimental science,

restrictions should be imposed in models that generate observations. However, his main

concerns are the validity of the restrictions and the statistical models used to analyze data.

It is therefore the role of the research to avoid uncertainty in the model specification since

it could lead to inconsistencies in the estimated parameters of primary interest. In order to

do this, meaningful variables should be selected to conduct verifiable empirical research

that yields meaningful results. Clodius and Miller (1961) provide and interesting

framework for understanding market structure in agricultural products. In his study, he

points out several topics ranging from problems in hypotheses testing, to what the end

goal of market shares studies should be and concluding with some shortcomings of









market share theory. He points out that the degree of product homogeneity should be

considered when choosing the theoretical framework of the study.

At this point it is necessary to define the subject matter of this study by

understanding the difference between industry and market. Two of the clearest concepts

of industry and market were provided by Smith and Dahl (1965), according to them"

industry is usually defined, in practice, as a group of firms that produce a like output

using similar inputs and production processes. A market, on the other hand, involves two

groups of firms buyers and sellers representing the forces of supply and demand in a

state of interaction" (p. 466). His paper also provides interesting concepts of the

assumption of perfect competition, the effect of technical innovation and capital

accumulation. Ghosh (1966) also provides definitions for both industry and market

preferring the latter for empirical studies. He argues that that since market studies are

more comprehensive in nature since they comprise both the buyer and seller side.

In this study there are some restrictions imposed in the parameters of the regression

model. Whenever that is the case the validity of the model's estimations could be

compromised by the restrictions imposed. Since models are evaluated in terms of the

parameter validity and predictive accuracy, it is important to determine if the restrictions

imposed in the model compromise outlet market shares estimations. However, Ghosh et

al. (1984) argue that constraining parameters values improves the predictive performance

of market share models. He uses the functional form, the error distribution assumption,

and the parameters description to compare the performance of market shares models.









The most used market share modeling approaches are the linear additive,

multiplicative and the attraction models. The functional form of the linear additive model

to estimate outlet market is given by

k
MSJ, 5 /jkXJkt + jt (2-1)
k=t

where MS denotes market shares, /f is a vector of the parameters of the regression, Xis a

vector of explanatory variables, and e is the error associated with the regression. The

attraction model is based on Kotlers' market share theorem where the market share of a

firm is given by the firm's marketing effort divided by the marketing effort of the rest of

the firms in the market (Kotler 1984). Using the profit impact of market strategies

(PIMS) database over a period of nine years, Buzzel et al. (1981) utilized a cross-

sectional regression analysis to prove the consistency of the attraction model over the

linear additive model. He argues that linear additive models generally have limited data,

defining the product of study requires too much effort, and suggestions are often not

relevant for managers. Naert and Bultez (1973) also criticize the linear additive model on

the grounds that it assumes that shares fall between zero and one and sum to one. They

conclude that for a market share function to be logically consistent the functional form

should be non-linear. However, estimating parameters in non-linear functions is less

straightforward than in linear case. More importantly, the statistical properties of non-

linear parameters are generally weaker than linear parameters and thus the predictions are

not necessarily better. Furthermore, as stated by Ghosh et al. (1984) "if we consider both

parameter validity and forecast accuracy, linear and multiplicative models performs at

least as well as the attraction model" (p. 208). For him, the market share model









specification should depend on the purpose of the analysis and the type of data available

to the researcher.

The nature of the data also presents issues to the validity of the model estimates.

Moriarty (1975) criticizes large databases because the variance across different products

and regions is usually lost. He argues that disaggregated models offer particular

advantages in policy formulation as they are able to target specific groups of interest.

Any source of variation from pooled data can then be eliminated by the dummy variable

technique. Grover and Srinivasan (1989) also state that data aggregation can comprise the

integrity of the data. Beyond a mere critique to the aggregation problem they propose a

different approach to the problem by dividing the data into within-switching and brand-

loyal segments. In doing so, they attempt to capture some of the variation in the model

while allowing some degree of data aggregation. In addition, they assume that a market

segment is a group of homogeneous consumers with equal probabilities of selecting the

goods of a product category. Two possible shortcomings in their analysis are that they

assume that the size of the segments remains the same for all periods and that

homogenous consumers have homogenous utility functions. Their assumptions however

may not always hold in panel data thus comprising their initial argument against data

aggregation. Regarding panel data, Ahl (1970) acknowledges the use consumer diaries to

record the cumulative growth in product class volume, as well as, the rate of repetitive

purchases. He stresses the importance that the sample obtained from consumer diaries

should be demographically balanced, demographic variables should capture consumption

differences, and seasonal patterns properly acknowledged. He concludes stating that

predictions based on this sampling method have proven to be highly accurate even in









products that show seasonality patterns. This is generally the case unless major upsets are

experienced in the market.

Chauvin and Hirschey (1997) argue that high market shares do not appear to be a

clear advantage for a firm's ability to expand in the future successfully. Chauvin

challenge the belief that bigger firms are more profitable by saying that high market

shares do not necessarily give rise to Ricardian rents. Furthermore, he points out that

market shares is simply "a measure of the size distribution of competitors, and a useful

dimension of the competitive environment faced by the firm" and does not influence

profit levels (p. 248). A similar argument was presented by Bradburd and Ross (1989) in

which they state that smaller firms may be able to find niches from which they can

"diminish or reverse the profit advantage of larger firms" (p. 258). In other words, in

market niches oriented firms the performance (and service) usually equals or exceeds that

of larger firms. Large firms generally exploit economies of scale by offering large

quantities of products that require a small degree of specialization. To the contrary, small

firms concentrate in sectors where customer support and one-to-one service is important.

Finally, both authors agree that the intensity of research, development and product

promotion are among the few factors that influence market shares levels in the long run.

Censored Models

When the observations follow a cumulative logistic function with zero and one

hundred percent probabilities the data could be censored in order to calculate more

accurately future demand expectations. As described by Chay and Powell (2001) "a

regression model is censored when the recorded data on the dependent variable cuts off a

certain range with multiple observations at the end points of that range" (p. 29). Tobin

(1958) was among the first ones to recognize the censoring problem by taking into









account the concentration of observations on the limiting points when trying to estimate

the effect of several variables on the limited dependant variable of the relationship. Tobin

argued that one should not discard the limiting values of data in order to fit a multiple

regression model. Instead, he suggested incorporating such values in a model which has

the characteristics of a probit and a multiple regression model. Such a model could be

used, particularly on consumer purchases data, when one can not incorporate into the

model the probability of events if the event does in fact occur.

Tobin's analysis assumes that the decision to consume is the same as how much of

the good to consume (Haines et al., 1988). However, this may not always be the case for

several consumer goods. Furthermore, Tobin also assumes that when "corner solutions"

are present, changes in prices and income can make such solutions disappear. Cragg

(1971) analyses in more dept the implications on censoring data and its effect in the

limited dependable variable. He validates Tobin's arguments about multiple occurrence

of the dependent variable on regression models. However, Cragg makes a distinction in

studying consumer behavior when no purchase is made by the consumer. He proposed an

alternative model to simulate the two-step decision process that consumers typically face

when buying goods. In his "double hurdle" model, Cragg use a probit model to calculate

the probability of the event take place (e.g. the decision to purchase the good) and then a

standard regression model estimates the magnitude of the change (e.g. how much of the

good to purchase). The strength of Cragg's model relies on the truncation of the

probabilities of the values while accounting for the values that were closer to zero or

hundred percent. In fact, Cragg's the two-step decision model has been widely used by

economists to estimate demand for agricultural commodities.









Blisard and Blaylock (1993) used the two-step decision process as a market

participatory model to estimate the demand for butter. The article shows the distinction

between households that never consume and those who consume butter infrequently. The

article proposes a purchase infrequency model because the butter unlike most agricultural

commodities has storage capabilities. Being that flowers are a highly perishable product

this model was not considered in the present study. In a previous article, Blisard et al.

(1992) used the double hurdle model on cigarette consumption to test the validity of

Tobin's corner solutions. By using a set of demographic variables to show how low

income women's consumption of cigarettes was affected, he was able to conclude that

change in income and prices may not necessarily have a proportional change in

consumption. Thus showing that consumer preference structure is not homogeneous and

that they respond to different utility maximization functions which model their

purchasing behavior. Haines et al. (1988) also compared the Tobit model proposed by

Tobin and Cragg's double hurdle as analytical models to estimate the dietary needs of

approximately 15,000 households over one year. He concluded that the Tobit model

underestimated consumption responses. Gould (1992) reached the same conclusions

when modeling the purchase frequency of cheese.

Heckman (1979) stated that Cragg's model can suffer from a sample selection bias

as a specification error. According to Heckman (1979), "the bias that results from using

non-randomly selected samples to estimate behavioral relationships is see to arise from

the ordinary problem of omitted variables" (p. 155). Because of the bias he proposed an

alternative model which links the two-step decision by introducing the Inverse Mills

Ratio calculated in the first stage as a regressor in the second stage. The Inverse Mills









ratio is a decreasing function of the probability that an observation is selected into the

sample (Heckman 1979). Byrne et al. (1996) used Heckman's model to model the two-

step decision process for consumption of food-away-from-home. In his article, he used

demographic variables such as education, age, and ethnicity to account for the consumer

preferences on actual purchases. They stressed the importance of demographic factors to

exploit the marketing potential in the consumer-driven food industry. Also, by

understanding consumer trends one may more accurately forecast future household

demand. Chay and Powell (2001) argue that even though a censored sample can

compromise the integrity of the regressors, that is efficiency is lost, the model still yields

consistent results. Amemiya (1973) provides a thorough explanation of truncation

particularly to the left of zero. He proposes a different model that Heckman in which all

observations are considered for the second step.

In all, the strength of Heckman's model relies on differentiating between the

propensity to consume and the quantity demanded among existing consumers linked

through the Inverse Mills Ratio. Since non-consumers have no influence on demand they

should be accounted out of the model regression but properly accounted for in order to

avoid bias in the estimation.














CHAPTER 3
US FRESH FLOWER DEMAND

This chapter is primarily oriented to understand the composition of the sample used

in this study, as well as, to understand the underlying reasons behind the decision to focus

on florists and supermarkets as the subject of study. The chapter presents an overview of

the US fresh flower demand divided in to cut-flower and flowering/green house plants

market share distributions. Expenditure and transaction levels were considered as

measurements to compare market share changes among both fresh flower categories.

Later on, the chapter covers the relative change in florists and supermarkets market

shares over the time. In addition, to understand the distribution of the sample market

shares are presented in terms of the demographic and socioeconomic variables described

in Chapter 1. Finally, florists and supermarkets market shares of cut-flowers and

flowering/green house plants are compared to the rest of the outlets in the specialty and

mass merchandising categories, as well as, to internet retail and others.

Indoor Flower Categories

Data on U.S. fresh flower consumption from 1992:7 to 2004:4 were obtained from

the American Floral Endowment (AFE) and Ipsos-NPD group. The data were obtained

from consumer diaries of approximately 9,000 demographically balanced households that

recorded their flower purchases every two weeks. Indoor flowers were grouped into three

subcategories: cut-flower, flowering and green house plants, and dry and artificial. In

addition, fresh flowers were grouped into four main types of retail outlet stores: specialty,

mass merchandising, internet retail, and others. Furthermore, specialty was divided in











florists and others while mass merchandising was divided in supermarkets,

warehouse/price club stores, and others.

Figure 3-1 shows the market share distribution for indoor flowers by expenditures

and transactions. The graph shows that fresh flowers comprise more than 80 percent of

the indoor flower market. More specifically, cut-flowers accounted for 57.4 percent of

total household expenditures and 44.7 percent transactions; flowering/green house plants

accounted for 29.5 percent in terms of expenditures and 37.3 percent in terms of

transactions; and finally dry/artificial accounted for 13.2 percent in expenditures and 18.0

percent in transactions. Cut-flowers and flowering/green house plants were separated

throughout the chapter see the relative difference in the distribution of market shares over

outlet categories.

Cut Flowers
57.4%






Dry and Artificial
13.2%
Cut Flowers
Flwg/Green 44.7%
House Plants
29.5%3
Expenditures


Dry and Artificial
18.0%



Flwg/Green
House Plants
37.3%
Transactions

Figure 3-1 Percent of household market shares expenditures and transactions on indoor
flowers. Source: AFE and Ipsos-NPD group.









Flower Retail Outlets

Figure 3-2 presents the distribution across different outlet groups. Specialty and

mass merchandising accounted for over 90 percent of cut-flower market shares in terms

of expenditures and transactions during the 1992 to 2004 period. For expenditures,

specialty accounted for 68.7 percent of market shares followed by mass merchandising

with 25.8 percent. For the transactions, however, mass merchandising with 53.6 percent

of the market showed higher market shares than florists which accounted for 41.3

percent. Within the cut-flower category, florists accounted for the largest component with

more than 80 percent in both expenditures and transactions. Alternatively, other outlets in

the specialty category accounted for 12.2 percent in expenditures and 21.9 percent in

transactions.

Figure 3-3 presents the distribution across different outlet groups focusing on mass

merchandising. The outlets in mass merchandising show similar market share levels in

expenditures and transactions. Clearly, supermarkets compromise most of the mass

merchandising outlets with approximately 80 percent of the market followed by

warehouses/price club outlets with 5 percent and other outlets with close to 12 percent.

Both Figure 3-2 and 3-3 provide insight as to the relative importance of using in florists

and supermarkets changes as proxies to forecast changes in specialty and mass

merchandising outlet groups respectively. Differences in the two measurements are

apparent with the specialty group accounting for the majority of expenditures shares and

mass merchandising group the majority in transactions. Figure 3-4 shows that

flowering/green house plants expenditures and transactions shares follow a different

distribution among outlet groups. Unlike cut-flowers, flowering/green house plants

expenditures levels are more equally distributed among specialty and mass









merchandising. More specifically, specialty accounted for 48.5 percent of expenditures

shares while mass merchandising accounted for 44.3 percent. Transactions share

distribution was similar to that of cut-flowers with mass merchandising accounting for

the majority of the market shares. With a market share of 64.4 percent, mass

merchandising more that double specialty shares which accounted for 29.3 percent of the

market. The graph also shows that in the specialty category other outlets account for 54.4

percent while florists account for 45.6 percent of the total specialty group in terms of

expenditures. The difference is greater in terms of transactions where other outlets

accounted for 73 percent of the market while florists only accounted for 27 percent.

Figure 3-5 shows the outlet division for the mass merchandising group. The graph

shows that supermarkets and warehouse/price club shares combined accounted for less

than the rest of the outlets in the same category. In this case, other outlets accounted for

approximately 55 percent of the market while supermarkets and warehouses/price club

accounted for 43 and 2 percent respectively in both measurements.

Market Shares by Product Form

Figure 3-6 presents the distribution of cut-flowers market shares based on specific

outlet types. In addition, florists' product form is presented to show the distribution of

arrangements and non-arrangements on both outlets. The graph shows that for florists the

flower arrangements accounted for 70.5 percent while non-arrangements accounted for

29.5 percent in terms of expenditures. The distribution is more evenly distributed in terms

of transactions with both forms accounting for approximately 50 percent of the market

each. Figure 3-7 shows that the difference in flower form was considerable in

supermarkets where non-arrangements accounted for the 82.1 percent and arrangements

accounted for 17.9 percent based on expenditures. Alternatively, in terms of transactions













the difference was greater with non-arrangements accounting for 91.8 percent and


arrangements 8.2 percent.


Mass Merchandising
25.8%



Retail Internet
2.1%







Mass Merchandising
53.6%


Specialty
68.7%


Florists
87.8%


3.4%


Expenditures


S---Other
7 --121.9%


Specialty
41.3%


Other
4.4%
Retail Internet
0.6%


Florists
78.1%


Transactions


Figure 3-2 Percent of specialty market shares on expenditures and transactions for cut-
flowers. Source: AFE and Ipsos-NPD group.


Warehouses/Price Club
Other 6.0%
11.9%




Supermarkets
82.0%





Warehouses/Price Club
Other 4.1%
12.1%




Supermarkets
83.8%


Transactions


Figure 3-3 Percent of mass merchandising market shares on expenditures and
transactions for cut-flowers. Source: AFE and Ipsos-NPD group.


Specialty
68.7%












Specialty
41.3%


Retail Internet
0.6%
Other
4.4%


Other
12.2%



















Florists
45.6%


Mass Merchandising
44.3%







Retail Internet
0.8%




Mass Merchandising%
64.4%


Specialty
48.5%
Other
54.4%



Expenditures


Florists
27.0%


Other
73.0%


Other
6.2%
Retail Internet
0.2% Transactions


Figure 3-4 Percent of specialty market shares on expenditures and transactions for

flowering/green house plants. Source: AFE and Ipsos-NPD group.


Retail Internet
0.8%


Expenditures


Warehouses/Price Club
2.0%

Supermarkets
42.8%




Other
55.2%




Warehouses/Price Club
1.2%

Supermarkets
43.3%




Other
55.5%


Transactions


Figure 3-5 Percent of mass merchandising market shares on expenditures and

transactions for flowering/green house plants. Source: AFE and Ipsos-NPD

group.


Specialty
48.5%












Specialty
29.3%


















Supermarkets
21.2%


3.1%
Retail Internet
2.1%


Warehouses/Price Club
2.2%


Other Specialty
9.0%
Other Mass Merchand.
6.5%


SFlorists
32.3%
Other
4.4%

Retail Internet Transactions
0.6%


Figure 3-6 Percent of household market shares for florists by product form. Source: AFE

and Ipsos-NPD group.


Other
3.4%


Other Mass
Merchand.


Florists
60.3%


Expenditures


Other Specialty
9.0%


Other Mass Merchand.
6.5%
Retail Internet
0.6% Other
4.4%


1.6%


Florists 8.2%
32.3% Transactions

Figure 3-7 Percent of household market shares for supermarkets by product form. Source:

AFE and Ipsos-NPD group.


Non-
Arrangements
29.5%


Florists
60.3%


Expenditures


Flower
Arrangements
70.5%









Share Distribution over Time

Yearly percent of cut-flowers market shares expenditures and transactions on the

different outlet groups are presented in Figure 3-8. To facilitate the discussion only whole

years where considered for graphing purposes. Over time mass merchandising market

share levels tended to increase during the 1993-2003 period. In terms of expenditures,

mass merchandising increased from approximately 27 to 40 percent and in transactions

from 50 to 70 percent. To the contrary, specialty decreased over the ten years showing a

slight increase in 1998 but then falling back again in the last three years. Overall,

specialty decreased from 60 to 50 percent in expenditures and from 35 to 29 percent in

transactions. Since data on internet retail purchases was only available from 2000 the

shares increase from that period up until 2003. Other outlets experienced reasonably

stable market share levels up until the 1997 to 1998 period where a sharp decrease was

seen in expenditures and transactions. Other outlets decreased from approximately 13 to

5 percent. In spite of all the changes in market shares, specialty continued to dominate in

cut-flowers expenditures although the gap could reduce if the current trend continues.

The graph also shows that the difference between mass merchandising and specialty is

widening with the former gaining more shares over the latter throughout time.

Figure 3-9 shows the yearly trends for flowering/green house plants outlet groups.

Over the 10 year period specialty shares reduced from 53 to 44 percent in expenditures

and from 32 to 25 in transactions. To the contrary, mass merchandising outlets

experienced an increase in market shares from 40 to 53 percent in expenditures and from

60 to 71 percent in transactions. The market share levels for internet retailers and other

outlets for flowering/green house plants follows the same trend as the cut-flowers

distribution previously presented.














Market shares
1.00
Expenditures
-Specialty Mass Merchandising
0.80 -- -- Retail Internet Other

0.60-



0.20 -



0.00
Transactions
-Specialty Mass Merchandising
0.80 - - -Retal Internet Other

0.60 -



0.20

0.00 mis
1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003

Years

Figure 3-8 Percent of yearly market shares in cut-flowers for outlet groups by
expenditures and transactions. Source: AFE and Ipsos-NPD group.

Unlike cut-flowers, mass merchandising outlets dominate the market in both


measurements. More specifically, in 2001 the mass merchandising shares of expenditures


superseded florists' shares with the gap widening in the last years. The difference is even


greater in terms of transactions with mass merchandising capturing more than the rest of


the outlets combined.


While the two previous graphs presented the relative changes in market share levels


over time, the next two graphs show the variation that florists and supermarkets over the


10 year period. Figure 3-10 shows the yearly trends in cut-flowers expenditures and


transactions for florists and supermarkets. In terms of expenditures, florists show a steady


decline in market share levels particularly from the year 2000 onward. To the contrary,


supermarkets' market share levels have increased through time by nearly 10 percent.















Market shares
1.00
Expenditures
-Specialty Mass Merchandising
0.80 - -- -Retail Interet Other

0.60 -

0.40

0.20

0.00 -
1.00
STransactions
-Specialty Mass Merchandising
0.80 - - -Retail Interet Other

0.60- ---

0.40

0.20

0.00
1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003

Years

Figure 3-9 Percent of yearly market shares in flowering/green house plants for outlet
groups by expenditures and transactions. Source: AFE and Ipsos-NPD group.

It is important to notice that the difference in supermarkets share gains is not the same as


that of the loss of florist which might suggest that other outlets are capturing part of the


market as well. This could be particularly true for internet purchases, which as we


mentioned earlier started to capture market shares in the year 2000. In terms of


transactions, we see that both outlets start out at approximately the same level at 40


percent of the market each. However, as time passed a gap between the two outlets


develops with supermarkets increasing its market share levels to nearly 52 percent with


florists reaching 20 percent of the market. Clearly, there has been a decline in the florists'


purchases in cut-flowers over the last decade and the gap is increasing over time.














1.00

0.80

0.60

0.40

0.20

0.00
1.00

0.80

0.60

0.40

0.20

0.00


Market Shares

I Expenditures
SH Flonsts
-Supermarkets

1 - - - - T7- - -









Transactions
Flonsts
-Supermarkets




--- - - - -- -- -- -- -- -- -- -- -- -- -- -- -- ---
-c
-I
-


1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003

Years

Figure 3-10 Percent of yearly market shares in cut-flowers for florists and supermarkets
by expenditures and transactions. Source: AFE and Ipsos-NPD group.


Figure 3-11 shows the trend in market share levels for flowering/green house plants


for the same two outlets. Unlike the cut-flowers' graph that showed a clear disparity in


the market share levels of florists and supermarkets over time, flowering/green house


plants levels do not seem to vary as much over time. Except for a few instances, both


expenditures and transactions levels remain fairly stable over the years without showing


major disturbances. It is important to notice however that florists' expenditures levels


started at 30 percent in 1993 and then declined to the same level as supermarkets at


approximately 20 percent by 2003. In terms of transactions, both outlets maintain the


same markets share level over time with 30 and 10 percent levels for supermarkets and


florists respectively.











Market Shares
1.00
SExpenditures
S0.0 Flonsts
0.80 -------T-- --- ---------------- r----- --------- - -
tSupermarkets
0.60

0.40

0.20- --

0.00
1.00
Transactions
0.80 - lonsts
Supermarkets
0.60

0.40 --
0.oo- - I i --
0.20

0.00
1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003

Years

Figure 3-11 Percent of yearly market shares in flowering/green house plants for florists
and supermarkets by expenditures and transactions. Source: AFE and Ipsos-
NPD group.

Figure 3-12 shows the percent of market shares for arrangements through both


florists and supermarkets. As expected, in terms of expenditures florists dominate the


market with approximately 77 percent by 2003 suffering a 10 percent loss since 1993.


The low level of market share of supermarkets in this category (approximately 7 percent)


was expected as well as the relatively small (3 percent) gain at the end of the period. The


transactions graph shows that the market share gap between the two outlets is decreasing


over time particularly since 2000. The graph not only shows that in general people tend to


purchase through florists when it comes to buying flower arrangements.


Figure 3-13 shows the percent of market shares in non-arrangements for florists


and supermarkets over time. At the beginning of the period florists had a greater market


share in terms of expenditures than supermarkets, however florists with 50 percent of the









market lost the initial 10 percent advantage over supermarkets by the middle of 1998.

From that year on we see a steady decline in florists' market share reaching nearly 22

percent of the market while supermarkets ended with 53 percent. In other words, during

the 10 year period florists and supermarkets switched positions in the industry. The

situation in terms of transactions was also expected with supermarkets increasing their

market share levels over time. This can be partly justified by the bigger base of potential

consumer that supermarkets have over florists. By differentiating between arrangements

and non-arrangements we can appreciate that most of the florists decline in market share

over time can be explained by the great loss in the non-arrangement sector of the market.

Here again, the marketing and product mix seems to have a greater effect in supermarkets

than in florists.

Market Shares across Variables

Figure 3-14 shows cut-flowers market share distribution by expenditures and

transactions across the variables considered in the present study. The graph shows that

the distribution of expenditures and transactions follows the same pattern in all the

variables except purpose. In purpose, gift buying has a greater market share in

expenditures which consequently is the highest level of market concentration when

compared to the rest of the variables. Gender also denotes a marked difference in the

distribution with female buying having more market shares than males. The first three

age groups denote an increasing market share distribution while the last one shows a

slight decrease. Consequently the first age group has the least percentage of market















1.00


0.80


0.60


0.40


0.20


0.00

1.00


0.80


0.60


0.40


1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003

Years


Figure 3-12 Percent of yearly market shares in arrangements for florists and supermarkets

by expenditures and transactions. Source: AFE and Ipsos-NPD group.


Market Shares

Expenditures
Flonsts
-Supermarkets





F ------ T -------------I-------








Transactions
Florists
T -Supermarkets

k ,__ _____ i a_ _


- _____


1993 1994 1995 1996 1997 1998 1999 2000 2001 2002


Years


Figure 3-13 Percent of yearly market shares in non-arrangements for florists and

supermarkets by expenditures and transactions. Source: AFE and Ipsos-NPD

group.


Market Shares
Exenditures
Flonsts
---I --Supermarkets

-I











Transactons
Flonsts
- -Supermarkets






~ ~~ ~ -__ -_ _- -_-_- -_ _-_-_-_-_ _- _-_-_-_- -_ _-_- _ 4 _ _
-

-


0.80


0.60


0.40









shares of all the divisions within the variables. The fact that market share concentration

decreases by nearly 10 percent in the fourth group tells us that people tend to buy cut-

flowers up to a certain age and then on consumption drops to a level similar to people

that fall in the second category. Unlike age, cut-flower consumption does not increase as

income rises. The four income groups follow a more erratic distribution with the first and

third groups having approximately the same market share level at 20 percent. The same is

true for the second and fourth groups with approximately 30 percent of the market each.

Surprisingly, the third income group does not seem to be buying to a level proportional to

their purchasing power. In general, the graph shows that major combinations of variables

peak demand in cut-flowers such as gift, female, of approximately 41 to 54 years of age

and with an income of either the second or fourth income group. The same is true for the

combinations that show a decrease in market share percent relative to the average

consumer such as a male buying for self who is under 25 years and with an income in

either the first or third group. Note that these are simple percentages without any

constraints on the other variables when calculating the distribution.

It was expected that the distribution of the variables differ when comparing cut-

flowers and flowering/green house plants. As presented by the variables for

flowering/green house plants in Figure 3-15, this difference is more noticeable when

comparing purpose and gender. When it comes to the purpose, the distribution is equally

distributed among the two divisions in expenditures. However, in terms of transactions

self buying market shares, at 64 percent, nearly doubles the gift percentage of the market.

In gender, females have approximately 80 percent of the market in both expenditures and










Market Share


Transactions
MAge OIncome
OPurpose OGender
0.36


0.68 i 0.66


0.29 r ( 2r -032 0.~4
0.22 0.20


8-^

/'.<<^ oU~


r t


Demographics

Figure 3-14 Percent market shares of cut-flowers expenditures and transactions by
demographics. Source: AFE and Ipsos-NPD group.


Market Share


Transactions
*Age OIncome
OPurpose OGender

-F 0.34 033
0 7'7 __


S0150 1.50

0.28 (031
0.22 0.19 I 22




0.82
0.64

0.8 f .31 0.36
0.18 0.23 0.18
F ]. .. 1- 0 1 8


Demographics


Figure 3-15 Percent market shares of flowering/green house plants expenditures and
transactions by demographics. Source: AFE and Ipsos-NPD group.


1.00
0.80
0.60
0.40
0.20
0.00



-v


~' ~h~'D


I d










transactions. The rest of the variables follow the same distribution showed in the

previous graph.

Figure 3-16 shows that in terms of cut-flowers florists and supermarkets dominate

the specialty and mass merchandising sections of the market. Florists have approximately

60 percent of the market in terms of expenditures followed by supermarkets with 21

percent. To the contrary, supermarkets have the highest market share concentration with

approximately 45 percent followed by florists with 32 percent. In both cases, the

combined market share of both outlets exceeds by more than 80 percent the total market

share of the cut-flower section of the market.


Outlets

Florists 0.60 0.32


Supermarkets 0.21 0.45,


Warehouses/Price Club 0.02 0.02


Other Specialty 0.0 Exenires 000 Trsac
OCut Flowers Cut Flowers


Other Mass Merchandising -0.03 0.06


Retail Internet -0.02 0.01


Other -0.03 0.04

0.00 0.20 0.40 0.60 0.80 1.00 0.00 0.20 0.40 0.60 0.80 1.00
Market Shares

Figure 3-16 Percent of cut-flowers market shares expenditures and transactions by
outlets. Source: AFE and Ipsos-NPD group.

When considering flowering/green house plants the dominating outlets also fall

within the specialty category and the difference to the mass merchandising shares is of











only 3 percent. Figure 3-17 shows that even though specialty and mass merchandising

stores have the greatest market shares in the industry; neither florists or supermarkets

capture the majority of the percentage in their respective groups. The graph shows that

the other specialty and mass merchandising stores grouped together have more than

florists and supermarkets market shares. In the flowering/green house plants the tendency

to buy through the internet is low as described by the low market share percentage both in

expenditures and transactions. Both Figure 3-16 and 3-17 show the importance of florists

and supermarkets when selecting fresh flowers, particularly in the cut-flower section of

the market.

Outlets

Florists 0.22 0.0O


Supermarkets 0.19 0.2!


Warehouses/Price Club 0.01 0.01

Exhen.tures Transactons
Other Specialty 0.2 IHwg/Green 0.21 lHwg/Green
House Plants House Plants

Other Mass Merchandising 0.24 0.36


Retail Internet -0.01 0.00


Other 0.0 1 0.06

0.00 0.20 0.40 0.60 0.80 1.00 0.00 0.20 0.40 0.60 0.80 1.00
Market Shares

Figure 3-17 Percent of flowering/green house plants market shares expenditures and
transactions by outlets. Source: AFE and Ipsos-NPD group.















CHAPTER 4
THEORETICAL MODEL AND MODEL SPECIFICATION

This chapter presents a general consumer demand concept to illustrate the buying

decision by outlet among U.S. households Later, Heckman's censored model is

developed along with the model specification for the present study. Socioeconomic and

demographic variables incorporated into outlet market share models used in this study to

are presented and discussed. Finally, outlet market share estimates are presented in the

form of econometric parameters and their statistical properties.

Consumer Demand for Flowers

Consumer demand theory analyses the behavior of consumers as they purchase a

set of goods to satisfy personal needs given a specific utility function. Generally,

consumers have a budget constraint that limits the choices that they make in their

purchasing behavior. The theory assumes that the consumer acts as a rational economic

agent, maximizing his utility function given his own budget constraint. In the general

form, maximization of the utility function as described by Girapunthong and Ward

(2003) is given by:

Max u= u(qi,...,qn) (4-1)
n
Spjq = m
J-1
where p, and q, are the price and the quantity of thejth good, respectively, and m is the

total expenditures or income on all n goods.

In this study, flowers are divided in outlet market shares for cut-flowers,

flowering/green house plants and dry/artificial. As shown in a previous chapter cut-









flowers and flowering/green comprise most of the flowers industry in the U.S. The

analysis further divides cut-flowers in arrangements and non-arrangements due to the

importance as a source of business for the two major outlets of this study, florists and

supermarkets.

Details on the Data

Outlet market shares in the cut flower industry can be modeled by changes in

consumer preferences. A set of socioeconomic and demographic data were defined such

as age, income, gender, and purpose for buying to determine their influence on consumer

outlet selection. In models that use information from panel data the question on the

sample representative from the population is always present. Ahl (1970) states that "if

as occasionally happens, the test panel is not perfectly balanced to the test market along

one or more key demographic characteristics, it may be necessary to take these

imbalances to account in the prediction. If that is the case, then the predictions for each

demographic group would have to be weighted according to the percent of the population

it represents. That however is not the case for the data used in this study as it is collected

by Ipsos, a private organization that specializes in retrieving balanced consumer

information from targeted population. That is, the sample is demographically

representative of the population.

Censored Data Model

The data of the study follow a function form with a substantial number of

observations having zero/near-zero and one hundred percent values. This tendency is

often seen in data that models consumer demand of certain commodities where zero is a

possibility. When the data follow this functional form, the problem of censoring of the

dependent variable arises. The researcher may be tempted to erase such occurrences from









the dataset and work with the rest of the sample. However, by doing so, the integrity of

the data, as well as the validity of the findings and policy suggestions are seriously

compromised. The significant portion of observations on cut-flowers and flowering/green

plants taking a zero or one hundred percent values (insert figures) provides justification

for considering censored regression models as an appropriate framework for conducting

the present investigation.

Tobin proposed an estimation method for data with truncation problems later called

the Tobit model (Tobin, 1958). According to Tobin, probability and multiple regression

models fail to present thoroughly information about the dependent variable because of the

probability of limit and non-limit responses.

The general formulation of the Tobit double-censored model is

y*= Xip + e,, ,~ N(O, a2) (4-2)

0 if Y* < 0,
y, = if 0 < y' <1.0,
1 if Y* >1.0

where y, is defined as the latent variable and y, is the dependent variable (Greene,

2003). However, Tobin assumes that the decision to consume is the same as the decision

of how much to consume of the good and this is not always the case (Haines et al., 1988).

In cases in which the decision to consume and the amount of the good consumed differ,

the Tobit model understates the actual magnitude of the dependent variable. Therefore, it

is necessary to redefine the concept of the Tobit model to account for what Cragg called a

double hurdle model (Cragg 1971). According to Cragg, there is a clear distinction

between the propensity to consume and what is actually purchased. In his double hurdle

model, he utilizes a probit model to calculate the probability of buying the good (first









hurdle) and a standard regression model (second hurdle) for the amount of the good

purchased. In this research the analogy is to have selected outlet, you first had to be a

buyer. Hence, the first hurdle is being a buyer. Then the Tobit is part of the second

equation.

The nature of the data makes it necessary to censor the data in the present study and

agreeing with Long (1997) that if such procedure occurs Ordinary Least Squares (OLS) is

inconsistent, an alternative to the Tobit model was considered for the present analysis.

This papers uses the two-step decision proposed by Heckman instead of the double

hurdle presented by Cragg and discussed earlier mainly due to the fact that the portion of

the residual that arises from the use of an estimated value of ki, in place of the actual

value of Xi is not orthogonal to the xi data vector (Heckman, 1979). Heckman proposed a

model similar to Cragg by acknowledging the two-decision approach in purchasing

behavior. However, Heckman used an inverse Mills ratio to link both processes.

As described by Heckman and further explained by Long (1997) the sample

selected in the probit model is given by

y*= Xif + e,, assuming that 8, -N(O, a2) (4-3)
and

y 0 if Y <0,
1 if y >0.

where Xi is a 1 xK, explanatory vector of socioeconomic or demographics variables.

Also, fj is a K, 1 vector of parameters with j=1,2,...,n. and e is a residual that captures

unobserved influences in the dependent variable. The magnitude of the f parameters

reflects the impact of changes in the x vector on the probability. The subscript land 2

denote the probit and Tobit model specification respectively.








To simplify the formulas that follow, let

Pi =Xifl (4-4)
and so the decision to become a fresh flower buyer follows a standard normal probability

function given by


f(Y ul a ) 2-l exp Y 1 2 1 i (4-5)
,1 fT 2[ e, 2 a C, C

with the cumulative distribution function of y* given by

F(yi 0 / i, l)= ) f (zLA,o4)dz=Pr(y*=0) (4.6)

and by default

Pr(yl 1) = 1 F(yi* /pi, o) (4-7)

which can rewritten as:

Pr(yl < O) = Y- = F(yi 0 /1i, al) (4-8)


Pr(yi 0)= P K = 1 F(yi* = i,o u)

since we are dealing with a symmetric standard normal distribution.

In order to calculate the probability distribution function of the censored part of the

distribution the original distribution is divided by the region to the right of zero (positive

purchases). The function is given by

f(yb yl > 0, P1, ol)= f(y 1, -) (4-9)
Pr(y1 = 1)

and since the data has been censored to the left of zero, E(y y > 0) > E(yi *) = i.

Then, the expectation of becoming a buyer is given by










E(yb yl* >0) = +1 -+a (4-10)


where

(1 -0

= O CT (4-11)



is a monotone decreasing function of the probability that an observation is selected into

the sample known as the Inverse Mills ratio. The value of A is saved and used as a

regressor in the second stage estimation.

Heckman's two-stage method generally utilizes a probit for participation

(probability of purchasing) and an ordinary least squares (OLS) procedure for the actual

consumer purchases. Both decisions are independent and thus no corner solution is

observed. However, in this study corner solutions can be seen in the second stage and

would imply that a consumer did not buy in either two outlets. Instead of an OLS

procedure a Tobit was used to model intensity of buying among florists and

supermarkets.

The formulation for the Tobit model is set as

0 if y < 0,
y, = y if 0 1 if y2 >1.0

The second stage Tobit is a function of both the variables considered in the probit

and the Inverse Mills Ratio estimated by the model. The notation for the Tobit is as

follows:


Z =f(X, A1)


(4-13)











and to simplify the formulas that follow let


p2 = Za (4-14)

and since we are dealing with a standard normal distribution the probability that an

observation is censored from below is given by


prob(y = 0 I Zi) = i 2
C 2 2


(4-15)


and censored from above by


prob(y = 1.0 Zi) = 1 ( -Za
C2

with the uncensored portion given by


prob(0 < y <1.0 I Zi) = y Za
C 2


(4-16)


(4-17)


and for facilitation purposes let




0= O-Za 1 -Za
S0 12
I r 2 J I z2


(4-18)


so notation for the expected market share for the two outlets is given by


E(y)= P(O)+ P(l)+ (Di -() Za + oc2K -A I


(4-19)


(1D + Za((l (Do) + 02 (0 )

which is presented in the Tobit model. Next, the model specification is presented for the

probit and Tobit models in terms of the variables of this study.


C2 C2










Tobit Model for Outlet Selection

Several socioeconomic and demographic variables are included in Heckman's

model: the respondents' age, income level, gender, and purpose for buying, flower form,

and months to account for seasonality. Furthermore, the dependent variable can be

expressed in terms of transactions or expenditures levels. First, dummy variables are

defined for the right-hand-side were the sum of the parameters for each set of dummy

variables are restricted to zero. And so the function for the first step of the Heckman

estimation process using the probit model for positive expenditures and transactions is

given by

4 2 2
Xpf = o+ f-/,DAge, + ~/ =4DGnd, + p=6DPur, + (4-20)
1=1 1=1 =1-

4 12 3
DInc, + 12+ DMt, + ,#24+,DForm,
=1 =1 1=1

Since each dummy class is exhaustive and mutually exclusive, inclusion of all

discrete values within a class immediately creates the well known dummy variable trap.

One convenient solution to this problem is to restrict the sum of the coefficients to zero

4
where, for example, Y,8, = 0 then /l4 f1 f2 3. Substituting for f4 (for the other
1=1

appropriate coefficients) then gives equation (4-22) where the dummy variable trap no

longer exists. The rest of the coefficients are described by

f4 =-A P f P32 (4-21)

P6 =-P5
8 f/5

P8& -fP7

Pl2 9- Plo- PfI

f24 =- P13 P14 P15 16 P l- 18 19 P20 -21 22 P2 23










27 P25 P26

and defining

XAge, DAge DAge4 (4-22)

XGnd, = DGnd DGnd2

XPur, = DPur, DPur2

Xlncj = DInc DInc4

XMt = DMt DMt12

XForm, = DForm DForm3

The probit equation variables now corrected for the dummy trap are

3 3
Xp = fo + XAge + fsXGnd + flXPurY + fiXInc, + (4-23)
l=1 11

11 2
S,22+XMt, + Zf24+,XForm,
1=1 1i

defined as the dummy variables in the probit model. The impacts of the variables are then

compared relative to the average household measured with /io.

The second stage is depicted by a Tobit model with the same variables of the probit

model including the Inverse Mills Ratio as a regressor in the equation. The function is

given by

3 3
Za= ao+ a,XAge, + aXGnd, + a XPur + as ,XInc, + (4-24)
1-1 1-1

11 2
a1 XIt, + a24+,XForm, + a 28Mills
1=1 1i

The explanatory variables used in the model are described in Table 4.1. Tables 4.2

and 4.3 present the estimated Probit and Tobit coefficients and t-values. The coefficients

reflect the sign of the relationship to the dependant variable, in this case market shares,










while the t-values reflect the significance of the relationship. Next, the Probit and Tobit

coefficients are discussed outlining the most statistically significant variables in the

model.

First-Step Results: Decision to Become a Buyer

The decision to become a buyer was estimated by a probit model using

expenditures and transactions. The /f parameters in Tables 4.2 and 4.3 denote the effect

on the probability of buying flowers from an initial sample of 27,072 household

observations with that 8,040 households not becoming buyers of fresh flowers. An

econometric programming code using TSP are included in Appendix C. The results

indicated that there was little distinction between the estimated coefficients and t-values

in terms of expenditures and transactions as should be the case since one cannot have

expenditures without transactions. Thus in the second stage Tobit we could have equally

included the Inverse Mills Ratio from each model when establishing both the second

stage expenditures and transaction models.


IOZero Purchases MPost ve Purchases
1.00


0.80


0.60


0.40


0.20


0.00
1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004

Figure 4-1 Distribution of values in the first stage probit model.









Table 4-1 Description for the Variables in the Heckman model
Variables Definition
Po Intercept
Agei 1 if = under 25 years of age, 0 otherwise.
Age2 1 if = 25 to 39 years of age, 0 otherwise.
Age3 1 if = 40 to 54 years of age, 0 otherwise.
Age4 1 if = 55 and more years of age, 0 otherwise.
Gndl 1 if = female, 0 otherwise.
Gnd2 1 if = male, 0 otherwise.
Purl 1 if = gift, 0 otherwise.
Pur2 1 if = self, 0 otherwise.
Incl 1 if = under $25,000 dollars, 0 otherwise.
Inc2 1 if = $25,000 to $49,999 dollars, 0 otherwise.
Inc3 1 if = $50,000 to $74,999 dollars, 0 otherwise.
Inc4 1 if = $75,000 dollars and more, 0 otherwise.
Mti 1 if = January, 0 otherwise.
Mt2 1 if = February, 0 otherwise.
Mt3 1 if = March, 0 otherwise.
Mt4 1 if = April, 0 otherwise.
Mt5 1 if = May, 0 otherwise.
Mt6 1 if = June, otherwise.
Mt7 1 if = July, 0 otherwise.
Mts 1 if = August, 0 otherwise.
Mt9 1 if= September, 0 otherwise.
Mtio 1 if= October, 0 otherwise.
Mtll 1 if = November, 0 otherwise.
Mtl2 1 if = December, 0 otherwise.
Form, 1 if = Arrangements, 0 otherwise.
Form2 1 if = Non-arrangements, otherwise.
Form3 1 if = Flowering, 0 otherwise.
Mills Monotone decreasing function of the probability that an observati
selected into the sample
Sigma Function that describes the probability of using each outlet among
buyers.
Source: AFE and Ipsos-NPD group












Table 4-2 Estimated Probit and Tobit Coefficients for Expenditures
Buyers based on Expenditures Expenditures Shares Models
Florists Share Supermarkets Share
Variables Probit Coefficients T-values Tobit Coefficients T-values Coefficients T-values


XAgel
XAge2
XAge3
XGndl
XPurl
XIncl
XInc2
XInc3
XMtl
XMt2
XMt3
XMt4
XMt5
XMt6
XMt7
XMt8
XMt9
XMtlo
XMt11
XForml
XForm2
Mills1


Po
01
32
P3
P5
P7

PlO
Pl9
P13
014
P15
P16
P17
P18
P019
P20
P21
P22
P23
025
P26


0.88693
-1.16447
0.21657
0.51074
0.51251
0.77493
0.11155
0.32632
-0.16054
-0.19032
0.16311
0.08701
0.26365
0.39670
-0.07695
-0.20605
-0.17209
-0.20298
-0.03584
-0.02847
-0.87114
0.40065


74.69340
-63.36020
11.98300
26.56230
46.35180
65.63670
6.23783
17.49510
-9.12933
-5.79866
4.70352
2.55200
7.15948
10.52040
-2.21238
-6.29466
-5.22000
-6.21451
-1.07392
-0.84993
-57.61850
26.15020


Sigma2
Number of observations = 27072
Number of positive observations = 19042


ao
a(
U2
(3
a5
(a7

a10


a13
a14
(15
(16
(17
Q18
a19
(20
a21
a22
(a23
(a25
(26
s28
(9?o


0.11087
-0.13624
0.01543
0.08026
0.05111
0.32441
-0.00742
-0.00173
0.01908
0.00485
-0.00526
-0.00669
0.01653
-0.00998
-0.03266
0.00705
0.01311
0.00200
0.01908
-0.03811
0.34996
-0.04366
0.14068
0.50350


Number of observations
R2 = 0.462364


8.30978
-9.90005
2.07795
9.99890
9.24886
44.13356
-1.04646
-0.24044
2.63217
0.34961
-0.40329
-0.50571
1.22499
-0.74081
-2.29710
0.50564
0.94799
0.14352
1.41980
-2.83341
38.25374
-6.70899
4.95642
122.35026
19042


0.17627
-0.05390
-0.01065
0.01673
0.05429
-0.04534
0.00330
0.04686
-0.03805
0.03256
0.05487
0.00736
0.00871
-0.04652
-0.02143
-0.02349
-0.03978
-0.02392
0.00729
0.03903
-0.21243
0.25860
0.00215
0.44075


15.78032
-4.60476
-1.71780
2.47201
11.53462
-7.65187
0.56182
7.79006
-6.26596
2.83263
5.08086
0.67069
0.77439
-4.11431
-1.82304
-2.02520
-3.44277
-2.06175
0.65248
3.50810
-26.50132
46.64365
0.08898
139.67058












Table 4-3 Estimated Probit and Tobit Coefficients for Transactions
Buyers based on Transactions Transaction Shar
Florists Share
Variables Probit Coefficients T-values Tobit Coefficients T-values


XAge
XAge2
XAge3
XGndl
XPurl
XIncl
XInc2
XInc3
XMtl
XMt2
XMt3
XMt4
XMt5
XMt6
XMt7
XMt8
XMt9
XMtlo
XMt11
XForml
XForm2
Mills1


Po
p1


Sigma2
Number of observations


0.88724
-1.16425
0.21645
0.51070
0.51278
0.77521
0.11139
0.32619
-0.16076
-0.19054
0.16294
0.08682
0.26350
0.39657
-0.07716
-0.20629
-0.17233
-0.20322
-0.03606
-0.02868
-0.87143
0.40047


74.70430
-63.34360
11.97540
26.55770
46.36870
65.64910
6.22818
17.48720
-9.14078
-5.80531
4.69842
2.54620
7.15493
10.51640
-2.21815
-6.30172
-5.22679
-6.22130
-1.08034
-0.85625
-57.63000
26.13730


ao
(a
U2
U3
(a5
(a7
(9
a(o
all
a13
a14
a15
(16
a17
a18
a19
a20
(a21
(22
a(23
a25
0(26
(s28


27072


0.13627
-0.04337
-0.00168
0.03642
0.01399
0.23679
-0.00684
-0.01935
0.02446
0.00713
-0.02147
-0.01278
0.00287
-0.02293
-0.01844
0.01329
0.02113
0.00993
0.01822
-0.02865
0.39821
-0.07376
0.02866
0.44281


11.61261
-3.58532
-0.25718
5.15586
2.87573
36.56254
-1.09601
-3.05088
3.83456
0.58409
-1.86785
-1.09623
0.24169
-1.93285
-1.47370
1.08297
1.73545
0.80951
1.54068
-2.41960
49.61143
-12.88202
1.15040
122.35598


Number of observations


es Models


Supermarkets Share
Coefficients T-values
0.20188 18.22941
-0.10275 -8.84949
0.00033 0.05444
0.03830 5.70522
0.07338 15.70986
-0.01069 -1.81563
-0.00620 -1.06278
0.05371 8.99824
-0.03832 -6.36662
0.03923 3.44324
0.06568 6.12739
0.01680 1.54508
0.00937 0.83857
-0.04283 -3.81800
-0.02892 -2.48029
-0.02450 -2.13158
-0.04379 -3.82450
-0.02705 -2.35192
0.00745 0.67294
0.02818 2.55180
-0.23908 -30.13249
0.27534 50.09437
0.03889 1.62000
0.43836 138.51323
19042


Number of positive observations = 19043
R2 .462495









This study assumes a significance level of ninety five percent also denoted by an

absolute t-value of 1.96 or greater. Figure 4-1 presents the distribution of the values for

the first stage of the model considering either positive or negative responses to becoming

a buyer of flowers. The graph shows that approximately on average 70 percent of the

sample used chose to become a buyer of any type of flowers in any outlet. Although

small, the percent of positive responses has increased by 3 percent over the 10 years. The

graph also shows the importance of using a probit model for the first stage as a way to

focus on actual flower consumers. The magnitude of the t-values indicated that all the

variables were significant in the models. The variables that had the greatest influence on

the propensity to become a buyer were purpose and form with t-values of 65.64 and -

57.63 respectively. In purpose, gift buying had the greatest positive significant impact on

the decision to purchase fresh flowers. Clearly, the probability of becoming a buyer

increases when the reason for buying is not for self consumption. For forms,

arrangements had a negative and significant impact on the probability to become a buyer.

Thus suggesting that if a consumer decides to buy fresh flowers he will least likely

choose arrangements preferring choosing instead non-arrangements and to a lesser degree

flowering and green plants.

The ages of the household have a significant impact on the household's decision to

become a buyer. However, the under 25 years of age group had a negative /f coefficients

indicating a negative relationship to the average buyer. In other words, people who fall in

this category are less inclined to become a buyer than the average age. The coefficients

indicated that the probability of becoming a buyer increases as the consumers age

increases peaking in the 40 to 54 years age group and then a slight decline in the last









group. Recall that the last group equals the negative sum of the parameters for each

dummy class.

As for the gender variable, females had the greatest effect on the decision to

become a buyer. This result seems to be consistent with previous studies that argue that

the majority of purchases of products available at supermarkets are done by females since

they do most of the grocery shopping within the household.

Compared to the rest of the demographic variables, income is less significant to the

decision of becoming a buyer. The first two groups, under $25,000 and $25 to $49,999

showed a positive relationship to the decision to become a buyer while the third level

$50,000 to $74,999 showed a negative one. Clearly the probability of becoming a buyer

does not consistently increase as income rise.

Apparently seasonality influenced the decision to become a buyer to a lesser degree

than the rest of the variables. It is important to notice that the /f coefficients showed

positive signs from January through May and negative the rest of the year. This may not

be surprising since the first period captures some of the most important calendar

occasions like as Valentines and Mother's Day. Throughout the year, only the months of

October and November were insignificant to the decision to become a buyer and that is

consistent with the well known demand problems in the fall months

Second-Step Results: Intensity of Buying

The a parameters in equation 4-19 reflect the decision of much to consume in either

florists or supermarkets once the decision of buying has been taken. Because this study

only considers florists and supermarkets purchases to model their expected market shares,

the a parameters could yield zero responses also known as "corner solutions". The corner

solutions in this particular case could mean that a consumer is a buyer but did not buy









through either two mentioned outlets. Unlike the probit estimates the coefficients and t-

values estimated for the second-stage Tobit differ considerably in terms of expenditures

and transactions as would be expected. It is important to recognize that while the probit

model dealt with either positive or negative answers, the Tobit model focuses on

consumers that chose not to become buyers, buy some, or buy all the time in either

florists or supermarkets. Figure 4-2 shows the distribution of values in the second stage

Tobit model with the three possible outcomes in the distribution in for florists. The Tobit

estimates focus on the part of the responses that fall within the middle category in the

graph, which represent 61.7 percent. In Figure 4-3 the same portion for accounted for

49.3 percent of the distribution. In both graphs, the percent of values of the distribution

that either reported no purchases or 100 percent purchases is considerable and had to be

accounted for in the by using a censored model.

Zero
29.7%







One
8.6%


Middle
61.7%


Figure 4-2 Distribution of values in the second stage tobit model for florists.









Zero
39.7%









One
11.0%


Middle
49.3%
Figure 4-3 Distribution of values in the second stage tobit model for supermarkets.

The probit results indicated that purpose and form were the variables that had the

greatest impact in becoming a buyer. In accordance, the Tobit estimates showed that the

same variables had the greatest impact in selecting outlets. In terms of expenditures, the a

coefficients for gift buying were positive for florists and negative for supermarkets and

statistically significant based on their t-values. This shows that when it comes to buying

fresh flowers as a gift the consumer prefers florists than supermarkets. In terms of

transactions, the same relationship was seen in florists with supermarkets (t-value of-

1.81) having no significance from the average.

Form was statically significant for both arrangements and non-arrangements for the

two measurements. The relationship was clear with florists having positive and negative

coefficients in terms of arrangements and non-arrangements respectively. The exact

opposite was the true for supermarkets coefficients. In all, the magnitude of the

significance was slightly greater in terms of transactions. This shows that arrangements









are bought more in florists than in supermarkets. The opposite is true for non-

arrangements and flowering/green plants. Therefore, product type tends to dictate the

type of outlet from where to purchase.

The previous results indicated that females were more inclined than males to

become buyers. The Tobit estimates showed that females were also more predisposed to

choose supermarkets than florists for their respective purchases. This intensity of buying

is considerably greater in terms of transactions.

Income groups describe different intensities of buying between the two outlets.

People under 25 years had negative and insignificant impact on outlet choice. The 24 to

40 years of age group showed negative and significant impact for florists and positive for

supermarkets. The group clearly prefers to buy in supermarkets once a decision to buy

was made. The third group, 40 to 54 years of age presents exact opposites parameters for

florists and supermarkets preferring the former outlet. From this relationship, people with

a less amount of income primarily buy in supermarkets until their income increases to a

point in which they switch to florists.

Seasonality impacts on outlet selection were more heterogeneous than was initially

expected. The months of March, April, June (only in florists), and October showed

insignificant coefficients for the two outlets. Thus, showing that in those months there is

no clear distinction of preference among buyers in the two outlets. The coefficients for

florists were insignificant except in the months of June and November when they have

negative values. To the contrary, supermarkets show significant coefficients with positive

values in January, February and November and negative in the rest.









As stated before, the Inverse Mills Ratio is a decreasing function of the probability

selected into the sample. The Inverse Mills Ratio coefficients were insignificant except in

terms of florists' expenditures. This means that except for florists' expenditures a regular

Tobit model could have been estimated using just these households that were buyers. The

positive and significant coefficient in florists' expenditures shows a more complicated

decision making process. A Tobit estimation in this case could have erroneously

oversimplified the decision process faced by consumers. The sigma values represent the

decision to become buyers in either florists or supermarkets. In this case the significant

impact of the sigma coefficients, which represents the Inverse Mills Ratio for outlet

selection, means that by running the model only among consumers of a particular outlet

will create a sample selection bias. Clearly, the coefficients show that something

influenced consumers' outlet selection once a decision to buy has been taken.














CHAPTER 5
MODEL SIMULATIONS

This chapter uses the estimates from the previous chapter to show simulated

changes in the model variables. Simulations were conducted over each of the economic

and socioeconomic variables in the model. Two simulations in particular combine the

variables that had the greatest impact on outlet selection. Then, a ranking of the variables

that showed the largest effect of the coefficients and range is discussed. Finally, time

recursive methods are estimated to analyze the variation of the coefficient for the average

consumer.

For any give combination of variables included in the models from Chapter 4, one

can calculate the estimated market share either for those shares when the share lies

between zero and one or for estimated shares across the full range of values including the

limits. Since, each variable influences both the continuous and limits in the shares as well

as the probability of being a buyer (i.e., as captured with the Inverse Mills ratio in the

model), the following simulations will be based on the expected shares including the full

range of share possibilities. For example, in Figure 5-1 the distribution of the shares

clearly reflects the need for the Tobit estimates as presented in Chapter 4. Give the

estimated model then the expected share can be for the non-limited share or for the full

range as suggested with the three arrows combined in the upper expectation in Figure 5-

1. Since most marketing policies intended to influence the outlet shares are likely over

the full range of the independent variables suggest with Figure 5-1, the more useful

information would be to have the expected shares over the full range.











-CI ^








SE(Sr0/hE(Share0share < 1.0)
SF ,,O- e. **.



S* *

*. ".
/ *
/ .*' S











Independent variable

Figure 5-1 Distribution of shares

Hence, in all of the following simulations or sensitivity analyses, the expectations include

the upper and lower limits alone with the continuous portion of the estimates. The exact

procedures are demonstrated in the Appendix C.

Expected Share for the Average Conditions

Recall from Chapter 4 that each dummy variable was estimated by imposing the

restriction that the sum of the coefficients for each dummy was set to zero. What this

means is that each estimated coefficient is a deviation from the average household and

the intercept in the model represents the average. Hence, as a reference point of the

subsequent expectations across a number of variables, the expected shares using just the

intercept for the outlet (florist and supermarket) and measurement (expenditures and


transactions) equations provide those average shares. Later simulations are measured by








58



adjusting the variables relative to these means. Note also that these means are over the


full range as discussed with Figure 5-1.


Figure5-2 shows the expected market shares for florists and supermarkets based on


both expenditures and transactions. As discussed earlier, it is possible that someone may


be a buyer but chose not to use either a florist (or supermarket) in a given period. That is


the zero share is feasible with the right set of circumstances. Similarly, the circumstances


could be that one type outlet is always used, thus giving the share of 100 percent. In this


figure, on average florists and supermarkets account for approximately 58 percent of the


total fresh flower expenditures and 55 percent of the transactions for the average


household. Between the two outlets the distributions are quite similar especially for the


expenditures. Note that supermarkets, on average, capture a slightly larger share of the


transactions as would be expected given the volume of traffic through supermarkets


versus florists.

Outlet probability by expenditures
0.60
0.55 -
0.50 -
0.45 -
0.40 -
0.30 ---0.28 -027





Outlet probability by transactions
0.6025
0.55
0.15


0.50 -
0.05

0.045
Outlet probability by transactions
0.60 -

0.55 --------------------------

0.30 ----------------
0.25 -
0.40
0.35

0.10
0.05
0.00



Figure 5-2 Average outlet shares the fresh flower market for florists and supermarkets.
Figure 5-2 Average outlet shares the fresh flower market for florists and supermarkets.









Expected Outlet Shares by Demographics

Outlet Shares by Household Age Groups

Figure5-3 represents the outlet expected shares of the market for fresh flowers with

the probabilities (shares) based on the four age groups defined in Chapter 4 and shown on

the bottom axis of Figure 5-3. Each bar in Figure 5-3 shows the deviation from the

average (see Figure 5-2) with bars below the average reflecting negative impacts and

values above the mean indicating positive impacts. In addition to the direction, the

magnitude of the change is directly comparable since all units are now in terms of market

shares. Hence, one can draw meaning conclusions about the direction and degree of

impact across the four age groups. For the most part in the graph, expenditures and

transactions levels displayed similar distributions with each type outlet. Between outlets

the impact of age is considerably different both in magnitude and direction. Supermarket

selection probabilities showed the expected upward slope as age increases while florist's

probabilities show a more erratic increase up to the 40 to 54 age group but then decreased

among those 55 and older. The first and second age groups consistently showed

probabilities either below or close to the means for both outlets. The third and fourth

groups present opposite outlet selection inclinations with the 40 to 54 years old group

more inclined to purchase in florists and the 55 years and more age group showing a

higher probability of selecting supermarkets. Recall from the previous chapter that the 25

to 40 years age group showed the least impact on the decision to become a buyer in either

outlet. In this case a less insignificant impact on the decision to become buyers could

mean that people from that category are indifferent as to which outlet to choose for their

purchases, thus showing probabilities closer to that of the average consumer. Note also

that the range of impact of age somewhat greater for supermarkets than for florists. This









is particularly true for the transactions. Overall, the impact of aging on the use of outlets

is far more consistent both in the direction and magnitude for supermarkets.

Outlet Share over Gender

Gender was one of the variables showing considerable impact on outlet selection as

initially suggested with the Tobit responses from Table 4-2. Again using expected share

relative to the average shares, Figure5-4 shows the outlet probabilities between male and

female buyers. As expected, the simulations show that females have a higher probability

than males of selecting the two outlets for fresh flower purchases. This difference is

greater in terms of supermarkets than in florists and could possibly be related to grocery

shopping periods where females are more likely buying the groceries. Going back to the

estimates in the previous chapter the first Inverse Mills Ratio showed that the propensity

to become a buyer was significant only when considering florists. This indicated that

something influenced the consumers' decision to buy in florists. In this case, a more

complicated outlet selection decision is reflected with a probability that is closer to the

model means of selecting that particular outlet. From a marketing perspective, the clear

implication from Figure 5-4 is that florist targeting gender would likely have far less

impact than expected with supermarket programs. At least the potential gains for florists

are substantially less than for supermarkets as evident from the relative sizes of the bars

in Figure 5-4.

Buying Purpose Impact on Outlet Shares

The model estimates showed that purpose had the most significant impact among

the variables expected to influence outlet selection. The simulations in Figure5-5 show

major variations between outlet selections with gift buying depicting a higher probability

for florists than for supermarkets. In addition, supermarkets showed a higher probability







61



of selection when considering buying for personal use (self). Also, the range of shares for


supermarkets according to purpose was considerably less that the range for florists. As


expected, the difference in outlet probabilities when considering purpose suggests that


consumers tend to choose florists for gifts and to lesser degree supermarkets for self use.


The range difference in florists could probably be attributed to customary behavior,


where flowers are generally purchased for gifts on either calendar or non-calendar


occasions.


Florist probability by expenditures










Supermarket probability by expenditures
__ L _
_ _ 1 _ L _
__ __ __ _1_ __I _
_ _ k J _
_ 1 _ _
I_ _ _ 1 _ L _
Suemre poailtbyepniu s
__ __ 1__ ___ __
_ _ _ 1 _ L _
__ __ __ __ __ __ 1_ __ _


Florist probability by transactions


Supermarket probability by transactions






-- p-o
___ _
S_4
1 _ _ J _ L _

I L J
I I I


Figure 5-3 Florist and supermarket probabilities over age.

Outlet Shares Across Incomes

Outlet selection probabilities in the simulations in Figure5-6 show that the two


lower income groups are more inclined to shop in supermarkets while the two upper


income groups show a higher probability of buying in florists. This behavior suggests a


~p~


isp
L,
,o
b












shift from supermarkets to florists as income rises and likely reflects with increased


purchasing power buyers move to the more valued added goods found more in florists





Florist probability by expenditures Florist probability by transactions


Figure 5-4 Florist and supermarket probabilities over gender.


Recall from Chapter 1 that florists usually charge a premium for the creative value added


for their products while supermarkets offer lower priced products. The graphs also show


that the first and last income groups have probabilities closer to the mean of the average


consumer. In the case of the under $25,000 income groups, a probability slightly above


the mean is seen only when considering supermarket expenditures and below the mean


for the rest of the measurements. The second income group, $25,000 to $49,999, showed


probabilities above average for supermarkets and below average for florists suggesting


that supermarkets are generally preferred. The third group, $50,000 to $74,999 shows the


opposite behavior, preferring florists for their fresh flower purchases. Finally, the last


group, $75,000 and above, shows small increases above the mean for florists and below


I - - -_ -- _

_ _- -__ _ _- - .
I I




Supermarket probability by expenditures


I
Supermarket --probability by transactions-
-- ----------------I------ -

-------4---------------I--------














-------------- -------










63




for supermarket. The distribution of probabilities in the four income groups suggests that



in the first and fourth group the demand for fresh flowers is more inelastic with respect to



income than the third and fourth groups.


Florist probability by expenditures














Supermarket probability by expenditures


_ - - -_ _ T _ _

-------------- -------



-------
_ ^ _ _ _^ _ _


Florist probability by transactions















Supermarket probability by transactions
II









Suprmaketproabiityby ranacton


Figure 5-5 Florist and supermarket probabilities over purpose.


Florist probability by expenditures














Supermarket probability by expenditures

- - - - - - - -
i


Florist probability by transactions


Figure 5-6 Florist and supermarket probabilities over income.


- - F - I



Supermarket probability by transactions
I I
-----------1------- ------- ----
I




-- - 1- - -p T-7- -


I


I









Outlet Shares over Flower Forms

One of the hypotheses in Chapter 1 was that outlet selection decision was closely

related to the form of the product. This was validated by the significance of the estimates

in Chapter 4. In addition, Chapter 3 showed that florists main source of business

comprise the arrangement sector of the market while supermarkets focused on the non-

arrangement sector. Figure5-7 presents the probabilities of selecting each outlet based on

the product form with the forms being arrangements, non-arrangement, and flowering

plants. As expected, when it comes to arrangements florists showed a higher probability

(more than fifty percent) of being selected with supermarkets showing approximately the

same probability for non-arrangements purchases. The probability distribution suggests a

clear outlet selection decision for arrangements and non-arrangements. However,

flowering/green plants probabilities are below the mean of the average consumer

suggesting that the two outlets are less probable choices when it comes to this product

form. Probably more than any other figure, the response to forms capture the

fundamentally differences between florists and supermarkets in terms of the product

offerings.

Combined Effect of Purpose and Form

Going back to the estimates in Chapter 4, purpose and form were the two variables

that had the greatest impact on both the decision to become a buyer and on the outlet

selection. Showing the combined effects of these two important variables give additional

insight into the expected outlet shares. In Figure5-8 the horizontal dotted lines represent

the probability of outlet selection for the average consumer as initially estimated in

Figure 5-2. The graph illustrates changes in the probabilities of outlet selection as

different combinations of form and purpose are considered. The graph shows that the











combinations for gift buying have a higher probability of selection than the combinations


for self. Also, gift combinations were the only ones that showed above average

probabilities of selecting florists, thus proving that gift buying weights considerably on

the decision to choose florists. Furthermore, the arrangements sector was the only

category that showed probabilities above the mean for the three flower forms.


Florist probability by expenditures Florist probability by transactions

0.50
0.40 r
0.30- -
0.20 ---
0.10-
0.00
0.60 Supermarket probability by expenditures Supermarket probability by transactions

0.50 ------ -
0.40 -
0.30 -
0.20 -
0.10
0.00






Figure 5-7 Florist and supermarket probabilities over form.

Flowering/green houseplants showed probabilities well below the mean suggesting that


such fresh flower form is less probable of being bought at florists when compared to the

other two. Note that the patterns are almost identical when based on expenditures and

transactions.


Figure5-9 shows the probabilities of supermarket selection for the combination of


purpose and form. Unlike florists, the distributions of the probabilities for supermarket

selection show are more evenly distributed when considering purpose. Recall from











Chapter 3 that non-arrangements were the biggest component of supermarket fresh


flower demand. As expected, the consumers' probability of selection was larger for non-


arrangements than for the other two product forms. Unlike florists, some combinations


did not show a large increase in the probability of selection in terms of purpose. In other


words, self and gift displayed similar probabilities when compared against arrangements,


non-arrangements, and flowering/green plants. As expected, arrangements had the lowest


probability of selection among the three product forms with supermarkets.





Florist probability by expenditures
1.00 -
0.80
0.60 1
0.60 - I t -- I -- -L- -- L- --- L ---- - -
0.40 -


0.00
Florist probability by transactions
1.00
0.80
0.60 -----
0.40
0.20



/0 V - --' --


Figure 5-8 Florist probabilities over purpose and form.

Simulations by Seasons

Seasonality was expected to influence outlet selection in fresh flowers consumption


to the extent that calendar occasions could influence where one buyers things for those


special occasions. However, the model estimates showed that seasonality was only


significant when considering supermarkets. This is consistent with the distribution of















Supermarket probability by expenditures
1.00

0.80

0.60 -----

0.40 --

0.20

0.00 -1
Supermarket probability by transactions
1.00 -

0.80



0.40

0.20

0.00 --J-


g% g%/ g ~ tc t/ 0/ 01/



Figure 5-9 Supermarket probabilities over purpose and form.

probabilities in Figure 5-10 where supermarkets show a higher variance relative to the

means of the model. The graph also shows that supermarkets have a higher probability of

outlet selection in the first four months of the year while florists experience the same in

the fall months. The graphs show that supermarkets are more probable to be chosen for

Valentine's Day relative to the average consumer. Recall from Chapter 3 that demand for

fresh flowers peaked in calendar occasions particularly in Valentines (February) and

Mother's Day (May). The remaining calendar occasions do not affect the probability of

outlet selection for florists and supermarkets. In fact, one could argue that the above-

average probabilities seen in florists during Fall could be attributed to non-calendar

occasions.
















Florist probability by expenditures



- 1 L -I I -I L -I I -I -
I I I I I I I I I I
II --- I-L L 11 1
II I I I I I I I I I
CLOT? uu ....,,




Supermarket probability by expenditures




I - - I
T r 1 I T -r T- r 1
_ 1 1 1 1 _ __ | I I
-^ ^ ^ ^ ^ ^ __I_ _I


Florist probability by transactions


Figure 5-10 Florist and supermarket probabilities over months.


Under $25,000 dollars
38


34
32


8 -





liiii
$50,000 to $74,999 dollars


- min
lIIi


$25,000 to $49,999 dollars


L I P I L T A I L




$75,000 dollars and more
L__II IL_ L
I I I I I L_ _L
L__II IL_ L
I I I I I I_ L I I
I I II I I I I I
L ~III I I I 1


Figure 5-11 Florist probabilities over income and months by expenditures.


The last simulation combines income and months to show the monthly variation in


the probabilities of selection as income rises. The variation in transactions and


Supermarket probability by transactions



O W - --I
- ^ _-1-- ----- ---- ------ ----
I I^ I^- ^ - -I I I^ ^ -


I__ _4 B ^ H "^ i _
T ^ _ _ _ _







69


expenditures levels is minimal and so they provide essentially the same information. In

this case expenditures levels were selected to discuss changes between the two variables.

From a previous simulation it was clear that outlet selection shifts from supermarkets to

florists as income rises. Figure5-11 shows the same behavior among income groups

throughout the year. The first two income groups showed probabilities either below or

near the mean. For these groups the lowest probabilities of selection coincided with the

February, May and June months showing that these groups are less inclined to purchase

in florists in calendar occasions. The probability of selecting florists increases in the third

income group especially in non-calendar months and decreases in the fourth group. In

fact, the fourth group behaves more closely to the average consumer depicting a relative

homogeneous probability of selecting florists.


Under $25,000 dollars $25,000 to $49,999 dollars
0.38 ..
0.36L +-- +
0.34 --- ~ n
0.32

0.28 l
0.26 .. .
0.24
0.22
$50,000 to $74,999 dollars $75,000 dollars and more
0.38 -------------
0.36
0.34
0.32
0.30
0.28
0.26 I I
0.24 1 7 IJ _1 _L__L____L
0.22



Figure 5-12 Supermarket probabilities over income and months by expenditures.

Figure5-12 shows the combined probabilities of income and months to account for

the seasonality effect of different income groups. Unlike the previous graph the variation









in expenditures and transactions is clear. The first income group shows a probability of

distribution above the mean from the January to April period then below the mean from

May to September and then peaks back above the mean in the last three months. This

distribution clear shows a higher probability in February coinciding with Valentine's Day

and the lowest in May. The second income group shows above the mean probabilities

except in May. The graph shows that summer months show the least probability of

selecting supermarkets but still above to the probability for the average consumer. The

third income group shows a different behavior showing above the mean probabilities for

February and below the mean the rest of the year. The last group shows above

probabilities on January, February and November and below the mean in the rest of the

year. The graphs groups show that in the summer months the probability of selecting

supermarkets decreases. It was expected that calendar occasions influence the probability

of selecting supermarkets. However, only Valentines appears to influence the probability

of selecting supermarkets.

Rankings Factors Impacting the Outlet Shares

In the previous figures a range of probabilities were shown and since the

probabilities are comparable, they can be ranked in terms for the magnitude and direction

of change. Figure5-13 shows the ranking of both the range of the variables as well as the

absolute high and low expected market shares for florists. Recall from the model

estimates that form and purpose were the variables that had the greatest impact on outlet

selection decisions. In the florists' case, the same variables that showed the biggest range

within each variable categories. The range difference is influenced by the magnitude of

the coefficients and by the difference between the variable that showed the least and the

variable that showed the greatest probability of outlet selection within each demographic









and socioeconomic variable category. Form showed 43 percent outlet selection

probability difference in expenditures between arrangements (highest probability) and

flowering/green houseplants (lowest probability). In addition, purpose showed a 32

percent outlet selection probability in expenditures between gift (highest) and self

(lowest). The ranking in terms of transactions was similar to that of expenditures. The

rest of the variables showed probabilities differences of less than 10 percent denoting a

more homogenous distribution between each of the divisions within those variables.

What is most apparent in Figure 5-13 is the relative low level of importance of

demographics relative to form and purpose when considering what truly impacts the

lower probabilities of using florists, i.e., product form and purpose of buying.

Figure 5-14 shows the variable ranking for supermarkets following the same

criteria for florists except that the magnitude of change is somewhat less for

supermarkets. Purpose was also the variable that showed the biggest range in outlet

selection probabilities. Unlike florists, gender and age were the variables that followed

purpose in the variable ranking with approximately 8 percent difference in each of them.

The graph showed that purpose had the least probability range in the case of

supermarkets. For supermarkets, form is the dominate variable impacting the likelihood

for using supermarkets.

Dynamics in the Outlet Share Coefficients

Figure 5-15 shows the variation in the coefficient for the average consumer from

1993 to 2004. Recall from the previous chapter that fo is the coefficient for the average

consumer after the dummy trap was corrected in the model. The coefficient was

calculated recursively from 1993 to find out if there were any other variables that would

influence the model above and beyond the socioeconomic and demographic ones












Expenditures


Forms -

Purpose

Age

Month-

Income-

Gender-


Forms

Purpose

Month -

Age-

Income -

Gender -


Figure 5-13 Variable rankings for florist.


Expenditures
-------- ------------------


--------- ------- ----




----T sa s-----

Transactions



------ --- ------ --
------ --- ----. -

:------- -- -- ------ -^ ----- -- ----------
I,;,7 7 ,7 7~ r -^ ;--r-^-,


0.10 0.20 0.30


g 0.29
0.07
0.06
0.06
0.06
0.05


0.32
00.09
00.09
00.07
10.06
S0.03
H0


0.50 0.(


Figure 5-14 Variable rankings for supermarket.


considered in the study. The graph shows that when choosing an outlet for flower


purchases, florists as an outlet choice is becoming less important for consumers. To the


- - 0.42
II I I / 0.32

------ ------------- -------- 0.06
----- ---- --- ------- 0.04

-------- ------ ------- ------- ------- 0.03

------ ------------- -------------- 0.03

Transactions
- - 0.44
-I------ ------ ------ 0.27

------ ----- 0.04

S- 0.03
-- 0.03

[------ ~-- 1- I-- 0.01

00 0.10 0.20 0.30 0.40 0.50 0.60


Forms
Gender
Month

Age
Purpose
Income


Forms
Age
Gender
Month
Income
Purpose








73



contrary, supermarkets are becoming more important to the average consumer as an


outlet selection choice. With the exception of the age coefficients, the rest of the


coefficients showed no significant variation over the period considered and further


discussion was not deemed necessary. Furthermore, any changes seen in the coefficients


seemed to affect florists and not supermarkets which validate the information presented


in the previous graph.


Florist Transactions Coefficient


1993-1993-1993-1993-1993-1993-1993-1993-1993-
1996 1997 1998 1999 2000 2001 2002 2003 2004



Supermarket Expenditure Coefficient
0.30
MIntrcept


-0.0 ------------------- ---
0.20
0.10
0.00
-0.10
-0.20
1993-1993-1993-1993-1993-1993-1993-1993-1993-
1996 1997 1998 1999 2000 2001 2002 2003 2004


1993-1993-1993-1993-1993-1993-1993-1993-1993-
1996 1997 1998 1999 2000 20012002 2003 2004



Supermarket Transactions Coefficient
0.30
Mlntercept
0.20
0.10
0.00
-0.10
-0.20
1993-1993-1993-1993-1993-1993-1993-1993-1993-
1996 1997 1998 1999 2000 20012002 2003 2004


Figure 5-15 Time Varying Coefficients for the Average Consumer














CHAPTER 6
CONCLUSION

Introduction

This chapter presents the summary, findings and recommendations of the study. A

brief summary of the previous chapters is presented outlining the major findings. Then,

the conclusions of the chapter are discussed with a particular emphasis placed on whether

the hypotheses from Chapter 1 were either validated or refuted. Finally, the implications

and limitations of the study and the recommendations for further research are presented.

Overview of Outlet Analyses

The main objective of this study was to analyze outlet market share changes given

a change in demographic and socioeconomic variables associated with fresh flower

consumers. By focusing in the fresh flower section of the market, the study covers

approximately 90 percent of all indoor flowers. A two-step estimation model was used to

describe the outlet selection process faced by buyers in the fresh flower industry. In the

first stage decision, the model used a probit model to differentiate between buyers and

non-buyers of flowers. In the second stage, the analyses used a Tobit model to estimate

the variables that had the greatest impact on the decision to choose either florists or

supermarkets once they decided to buy fresh flowers. Estimates from the Tobit model

were used to simulate the probability of selecting these two outlets. Finally, the

parameters were recursively estimated to test if there were changes in the parameters over

time. Parameter changes could suggest structural and/or preference changes not initially

captured in the original Tobit estimates. Also, often there are not specific variables to









measure these changes and the use of time-varying parameters is an indirect way to test

for such changes.

The flower industry in the United States was grouped into three categories: cut-

flowers, flowering and green house plants, and artificial. This study focuses on the fresh

flower portion of the market since it comprises nearly 90 per cent of the flower industry.

Based on the major outlets used by consumers, the outlets were divided into 4 categories:

mass merchandising, specialty, internet retail, and others. The data showed that

supermarkets comprised the majority of mass merchandising purchases with florists

capturing most of the purchases in the specialty category. Data for the retail internet

purchases were not available until the year 2000 and, hence, were not included in the

outlet share models. Florists and supermarkets were chosen in the study for the relative

importance in their respective categories but also because in the last decade major

structural changes have occurred in the industry as described by market share changes

particularly in these two outlets.

Chapter 1 introduced the problematic situation and the major hypotheses of this

study focused on the changes in market shares for both outlets. One of the strengths of

the study is that the sample used was drawn from an extensive database which covered

many aspects of the flower demand in the US. The database used is maintained by Ipsos,

a private company who along with the National Panel Diary (NPD) collected the

information from approximately 9,000 demographically balanced households every two

weeks through the use of consumer diaries. The diaries included comprehensive

information on actual purchases recording flower type, outlet selection, number of

transactions, and occasions.









Chapter 2 presented a literature review that was divided in three parts covering the

articles on choice theory and consumer preferences, market share theories, and

econometric models for censored data. Chapter 3 presented an overview of the fresh

flower industry for the years from 1992 through 2004. The chapter showed the relative

change in market shares among outlets using expenditures and transaction market share

levels based on cut-flowers and flowering/green house plants. The chapter covered

seasonal and yearly trends over florists and supermarkets, as well, as changes in both

arrangements and non-arrangements. The chapter showed the importance of florists and

supermarkets in the flower industry and the relative changes in market shares over the

study period.

Chapter 4 explained the theoretical framework and the model specification to

model consumer behavior in the fresh flower industry. The model specification started

with the neo-classical utility maximization theory and then explained the nature and

problems of censored data with its alternatives for estimation. Unlike the original

Heckman two-stage decision process which used a probit and a ordinary least squares

(OLS) procedure, this study used a tobit model to account for the "corner solutions"

previously discussed. In the first stage probit model, which assumed a significance level

of 95 percent, the estimates showed that purpose and form were the two variables that

had the greatest impact on the decision to become a buyer. The same variables had the

greatest impact on the second stage decision to choose either florists or supermarkets for

their flower purchases.

Chapter 5 presented simulations drawn from the model estimates to simulate

expected outlet shares over a range of variable values. Purpose and form were combined









to show the effect of both variables on the outlet use. A ranking of the variables from

the largest to smallest impacts on the probability of outlet selection was presented

Finally, a time recursive model was used to determine if there were underlying structural

changes taking place within the outlets.

Major Outlet Selection Conclusions

Estimates from the two-stage model showed that purpose and form were the

variables having the greatest impact on both the decision to become a buyer and on the

decision to choose either florists or supermarkets for their purchases. Assuming a

significance level of 95 percent, almost all variables included in the probit and tobit

models were statistically significant with the significance test being measured against the

average consumer. The probit estimates showed purpose, gender, and form were the most

importance factors impacting the likelihood of becoming a buyer. The combination of

females buying non-arrangements for gifts had the largest impact on market penetration

or attracting buyers. Among the variables that showed the least probability to become

buyers were people of less than 25 years of age having an income of $50,000 to $74,999

and buying flower arrangements. In general, the majority of the variables were

statistically significant given the mentioned confidence level. Seasonality showed the

typical decline in the likelihood of buying flowers in the later half of each year, thus

again highlighting the seasonal problem found throughout the flower industry.

The second stage tobit model estimates showed that purpose and form were the

variables that had the greatest impact on the decision to buy in either florists or

supermarkets. The estimates showed a marked difference between the two outlets in

terms of the combination of the variables that increased their probability of selection.

Recall from Chapter 3 that florists and supermarkets specialized in different sectors of the









industry as described by their market share levels. The estimated coefficients for the tobit

model described the same level of market segmentation between the two outlets. For

example, in the florists case the variables that had the greatest impact on outlet decision

such as the second age group from 40 to 54 years old, females, gift, and arrangements are

the same variables that has the least impact on the decision to choose supermarkets

except for females who are important on both outlets. In the supermarkets case, the third

age group from 40 to 54 years of age, non-arrangements and buying for self use had the

greatest positive impact on consumer outlet selection decision. The result from the

estimates reflected the market share conditions explained in Chapter 3 but also introduced

some of the variables that have influenced fresh flower demand in the last few years. The

demographic and socioeconomic variables used in the models therefore explained part of

the difference in market shares between the two outlets.

Starting from Chapter 3 it is apparent that the fresh flower industry was

experiencing major restructuring at the retail level. Furthermore, when analyzing florists

and supermarkets alone the models show that over the past decade florists have lost

market shares while supermarkets have improved their market position. The results

indicated that florists' main source of business falls in the arrangement sector of the

market while supermarkets are in the non-arrangement sector. The results also show that

1998 was a turning point for both outlets either widening the gap among them and even a

switching of market share position relative to other outlets. The statistically significance

of the estimates show that the variables considered for the study were accurate indicators

of what the flower industry has been experiencing in recent years.









The perceived characteristics of the product play a major role in fresh flower

demand as described by the purpose and form variables. The fact that both outlets

specialize in different segments of the market with statistically significant coefficients

associated with them implies that a degree of concentration in a sector could be achieved

by product differentiation. Recall from Chapter 1 that florists charge a premium for the

value added to their products in the form of arrangements and supermarkets focus on

large quantities of non-arrangements. Since florists share of the market remain essentially

the same in the arrangement section, the systematic market share loss in the non-

arrangement section could be the reason why florists' market share levels have declined

over time. In addition, the increasing market share in the non-arrangement section of the

market could be the reason why supermarkets have experienced a steady increasing in

total market participation. However, it is important to clarify that the results of this study

do not imply that florists share loss has been totally captured by supermarkets. Recall that

while the study focused on two outlets in particular all the outlets, some of the declines in

market shares could be reflected in the smaller outlets not included in the modeling

efforts.

Important to the overall analysis was the hypotheses that rising incomes and other

demographics could impact the types of outlets used. Drawing from the outlet share

model estimates, simulations where used to explicitly show the range of probabilities of

selecting florists or supermarkets as each variable was considered. In Chapter 5, each

simulation was completed by adjusting the variables relative to the mean of all the other

variables with the expected market shares being shown for florists and supermarkets

based on both expenditures and transactions. The probability means for florists were 28









percent in expenditures and 25 percent for transactions while the probability means for

supermarkets were 27 percent in expenditures and 28 percent in transactions. As a

general rule variations in income had a greater impact on using supermarkets than for

using florists and generally the florists share gains while the supermarket shares decline

as income increases.

When considering age, the simulations showed that the probability of selecting

florists increased until the third age group 40 to 54 years of age and then dropped for

people of 55 years of age or older. In direct contrast, supermarket shares consistently

increased over the age groups.

Purpose of buying showed marked differences for the two outlets. The fact that

florists main source of business comes from gift buying and supermarkets from self

buying were apparent from the distributions in Chapter 3. However, in the simulation

over purpose and interesting fact arises in that supermarkets probability of selection is

approximately the same as of the average consumer. When it comes to gender, the

simulations show that females have a higher probability of choosing both outlets than

males. Yet, the gender effect on supermarkets is more than twice as great for

supermarkets compared with florists.

The variable rankings in Chapter 5 showed that purpose and form were the two

factors showing the biggest variation within their divisions for florists with

approximately 40 and 30 percent, respectively. The ranking shows the variables that had

the biggest range from the least and most probable selection. Interestingly, the variables

form with 30 percent and gender to a lesser degree with approximately 10 percent were

the ones that showed the biggest variation when considering supermarkets.









In general, the statistically significance of the estimates provided important

information to modeling the fresh flower outlet demand given a set of socioeconomic and

demographic variables. However, as it is sometimes the case in econometric analysis

some variables are not considered in the study because of data limitations among others.

To account for this, a time recursive models were estimated to see if the variables

considered in the study became more or less important in modeling fresh flower outlet

demand. The results indicated that the coefficients for the majority of the variables

considered did not vary appreciably over time suggesting that the importance of the

variables included in the analysis was fairly stable. In fact, Chapter 5 only presents the

time varying coefficients for the average consumer for both florists and supermarkets. It

is safe to conclude that when considering the two outlets only florists appear to become

less important to the outlet selection decision faced by consumers.

Limitations

While many studies focus on pricing considerations and volume of sales as an

approach to estimate market share changes in an industry, this study assumes a

relationship between the characteristics of the buyer and the demand for fresh flowers.

Unfortunately, there is not a wide array of studies to compare the results of this study.

However, the methodology and the results are similar to several studies in agricultural

commodities such as away-from-home food consumption, cigarette, and fresh vegetable

consumption. Because a two-stage process was used to model first the decision to

become a buyer and then the propensity to buy in either florists or supermarkets, the

results from this study could be of interest from an outlet category or an industry point of

view. While any member of the flower industry could use this information to better its

position within the industry, efforts were made to present the information in the most









impartial manner possible. Therefore, this study does not provide any particular

suggestions for improvements to any particular outlet category and only analyses and

tries to explain the current trend in market share levels within the flower industry. A

limitation of the study was that the simulations chapter only dealt with the combinations

of variables that had the greatest impact on outlet selection yielding higher market share

levels. One interesting aspect for further research could be to combine the variables that

would represent the least market share level and analyze its implications to the industry.

The negative impacts highlight targets for potential promotions or other marketing efforts

to offset negative influences.

Recommendations

Information of the impact of the relative impact of various socioeconomic factors

on the consumption of fresh flowers discussed in this study can benefit producers and

consumers and may facilitate the decision making of policy makers. It is important to

investigate the effect of socioeconomic and demographic variables on the decision to

consume as a proxy to future market shares changes. The lost in market shares by the

florists sector continues to be a troubling factor for that portion of the industry and

anytime that helps them to counter the loss in shares would be beneficial. The negative

factors impacting florists provide areas for marketing and promotion efforts that need to

be explored in more detail. Also, it is important to remember that this study deals only

with household data and that commercial transactions were not incorporated into the

model. It is likely that florists capture a larger share of the commercial market but our

analysis does not specifically show that. Therefore, incorporating commercial data into

the analysis would be beneficial. Realistically, however, getting the commercial data is

very difficult and often impossible. One of the strengths of the study was its ability to






83


pinpoint the divisions within each variable, for example the low probability of selection

in the first age group, in which promotions or advertisement are needed to stimulate

demand. Another is that by understanding what products comprise the main source of

business of both outlets we can understand the effect that generic or brand promotions

would have in the flower industry. An important extension would be to specifically

design the marketing strategies using the targets suggested with this research. That has

not been done since it was beyond the scope of the study.
























APPENDIX A

INDUSTRY OVERVIEW


Market shares


0.60

0.40


0.00 -

1.00
Transactions

0.80 -- -- he


0.60 ---


0.40 --


0.20 -


1993 1994 1995 1996 1997 1998 1999 2000 2001 2002

Years

Figure A-i Percent of yearly market shares for specialty based in cut flowers by

expenditures and transactions. Source: AFE and Ipsos-NPD group


1.00

0.80


Exenditure-
Florists
-1I- -
I- I Ir,


2003












































0.60


0.40


0.20


Market shares

Enxendimres
S1 1 1 1 Flonsts
















Transactions
-- O ter



____


I I I
-


1993 1994 1995 1996 1997 1998 1999 2000 2001 2002


2003


Years



Figure A-2 Percent of yearly market shares for specialty based in flowering/green house

plants by expenditures and transactions. Source: AFE and Ipsos-NPD group


Market shares

Expenditures
-Supermarkets
Warehouses/Pnce Club
Other




__ I_ II I _









Transactions
-Supermarkets

Other
F - r -


1993 1994 1995 1996 1997 1998 1999 2000 2001 2002


2003


Years

Figure A-3 Percent of yearly market shares for mass merchandising based in cut flowers

by expenditures and transactions. Source: AFE and Ipsos-NPD group.



























0.40


0.20


0.00

1.00


0.80


0.60


0.40


0.20


Market shares


1993 1994 1995 1996 1997 1998 1999 2000 2001 2002


2003


Years

Figure A-4 Percent of yearly market shares for mass merchandising based in

flowering/green house plants by expenditures and transactions. Source: AFE

and Ipsos-NPD group.


Market shares


0.40 -


0.20


Transactions
-SpalEty
__ -Mass MerchanLdsng
-Rtal Internet
OI I I I er
--IL- +it














Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov De

Months


Figure A-5 Percent of monthly market shares based in cut flowers

transactions. Source: AFE and Ipsos-NPD group.


by expenditures and


Expenditures
-Sup markets
- - r - - Warehouses/Pne Club
Other

-I

--- ---I I I---- I


Transactions
-Supermarkets
- Warehouses/Pnce Club
Other




---1 ----


1.00


0.80


0.60


Exmenitumes
-Spealty
-.. -r I Mass Merchandsng

OI I I I I Other
-


I


0.80


0.60