<%BANNER%>

Having Our Cake and Eating It Too: Americans' Ambivalence about Social Welfare


PAGE 1

nrr rnrrrnr rnnrrrrnr rrnnnr rnrnrrnrrnrr n rnn

PAGE 2

!"#$%& () *(#* +)

PAGE 3

%#), -+./*&#),/,#-(&/,& .0(.#12

PAGE 4

#3 rr &%(*4.!("/*&)5$"(*,!("/*&)56#0/5,#))/"&(&# *..#&&//5(*,"(*4(*, )) -#(&/)0 "&%/#"!(&#/*-/(*,)+!! "&2

PAGE 5

3 rr !($/ rr222222222222222222222222222222222222 222222222222222222222222222222222222222222222222222 2222222222#3 r2222222222222222222222222222222222222 222222222222222222222222222222222222222222222222222 222222222222222222223## nr222222222222222222222222222222222222 222222222222222222222222222222222222222222222222222 22222222222222222222#7 n2222222222222222222222222222222222222222222 222222222222222222222222222222222222222222222222222 222222222222222222222222227 8nnrrnr222222222 2222222222222222222222222222228 rrr95 rnrrnr5nrrr5nr5rrr5rrr222222222222222222 22222222222222222222222222222222222: .'#3(1/*-/()&()//* *-/!&+(1#;/,(*,/()+ "/,22222222222222222222222222222222222228 %/"/3(1/*-/ 0.'#3(1/*& 1#&#-(1&&#&+,/)222 222222222222222222222222222222222222222222222222228 < &/*&#(1 +"-/) 0.'#3(1/*-/2222222222222222222 222222222222222222222222222222222222222222222222222 222222228= %/ *)/>+/*-/) 0.'#3(1/*& 1#&#-(1&&#&+,/)2 22222222222222222222222222222222222222222222222? .'#3(1/*-/(*,&&#&+,/)' +& -#(1/10("/22222 222222222222222222222222222222222222222222222222@ +&&#*$#&11 $/&%/"6#&% ./! &%/)/)2222222 222222222222222222222222222222222222222222222222222 : *-1+)# *22222222222222222222222222222222222222222 222222222222222222222222222222222222222222222222222 2222222222222222222222A? Arnrrrrn9 rnr22222222222222222222222222222 222222222222222222222222222222222222222222222222222 2222A #"/-&(*,*,#"/-&/()+"/) 0.'#3(1/*-/22222222 222222222222222222222222222222222222222222222222222 A@ +"-/) 0.'#3(1/*-/#*/*/"(1(*,' +& -#(1 /10("/222222222222222222222222222222222?@ $*#/ *01#-&222222222222222222222222222222222 222222222222222222222222222222222222222222222222222 22222222222?@ $*#/B00/-/ *01#-&22222222222222222222222 222222222222222222222222222222222222222222222222222 222228 00/-/B00/-/ *01#-&22222222222222222222222 222222222222222222222222222222222222222222222222222 22222? /!/*,/*&("#('1/)0 *)/>+/*-/) 0 -#(1/10 ("/.'#3(1/*-/222222222222222222@ &%/"*,/!/*,/*&(*, *&" 1("#('1/)222222222222 222222222222222222222222222222222222222222222222222 222< 1#-"/0/"/*-/)222222222222222222222222222222222 222222222222222222222222222222222222222222222222222 22222222222< ,/ 1 $5("&,/*�#-(&# *5n(-/5/*,/"5(*, *./222222222222222222222222222222222: 1#&#-(1* 61/,$/22222222222222222222222222222222 222222222222222222222222222222222222222222222222222 222222222@8 +..("22222222222222222222222222222222222222222222 222222222222222222222222222222222222222222222222222 2222222222222222222222@A

PAGE 6

3# ?rrnrrrnr rrrnr2222222222222222222 2222222222222222222222222222222222@@ "#&#>+/ 0"/3# +)!!" (-%/)& /()+"#*$.'#3( 1/*-/22222222222222222222222222222222222@: #"/-&/()+"/)222222222222222222222222222222222222 222222222222222222222222222222222222222222222222222 222222222222@: *,#"/-&/()+"/)2222222222222222222222222222222222 222222222222222222222222222222222222222222222222222 22222222222<8 *(1)#)2222222222222222222222222222222222222222222 222222222222222222222222222222222222222222222222222 222222222222222222222222+/*-/)2222222222 222222222222222222222222222222222222222222222222222 288 ,/1(*,/)&!/-#0#-(&# *22222222222222222222222 222222222222222222222222222222222222222222222222222 22222222288A n/)+1&)22222222222222222222222222222222222222222222 222222222222222222222222222222222222222222222222222 22222222222222222222222288? +..("22222222222222222222222222222222222222222222 222222222222222222222222222222222222222222222222222 2222222222222222222288= +/*-/) 0 -#(1/10("/.'#3(1/*-/22222 222222222222222222222222222222222222222222228? *-1+,#*$ ../*&)22222222222222222222222222222222 222222222222222222222222222222222222222222222222222 22222222228 nrrnrr222222222222222222222222222222222 222222222222222222222222222222222222222222222222222 22222222222228A8 nr22222222222222222222222222222222 222222222222222222222222222222222222222222222222222 222222228?

PAGE 7

3## r ('1/ !($/ AB8 #)&"#'+&# 0n/)! *)/) *6 #"/-&/()+"/) 0 .'#3(1/*-/222222222222222222222A: AB ,/1 0&&#&+,/)(' +& -#(1/10("/222222222222 222222222222222222222222222222222222222222222222222 22?8 ABA $#0&%/1$ "#&%.22222222222222222222222222222 222222222222222222222222222222222222222222222222222 2222222222222?? AB? "/>+/*-(*,*&/*)#& 0.'#3(1/*-/(' +& -#(1 /10("/222222222222222222222222222222? AB #)&"#'+&# 0n/)! *)/) *(1+/.! "&(*-/*,#(& ")2222222222222222222222222222222222222 AB@ #)&"#'+&# 0n/)! *)/) 3/"*./*&r3(1+(&# *) *,#-(& ")2222222222222222222222222= AB< #)&"#'+&# 0,/ 1 $5("&,/*�#-(&# *5n(/5/*,/"5(*,*./22222222222222@ ?B8 ""/1(&# *)(-" ))#"/-&(*,*,#"/-&/()+"/) 0 .'#3(1/*-/2222222222222222222222222< ?B ""/1(&# *)'/&6//* +"-/)(*, +".'#3(1/*-/ /()+"/)2222222222222222222222222222<@ ?BA *(1)#) 0n/)#,+(1)0" ,/1 0&&#&+,/)(' +& -#(1/10("/22222222222222222222<= ?B? ""/1(&# *)'/&6//* +"-/) 0.'#3(1/*-/(' +& -#(1/10("/222222222222222222222= ?B .!("#*$#"/-&.'#3(1/*-//()+"/./*&!!" (-%/) 222222222222222222222222222222222222=8 ?B@ .!("#*$*,#"/-&.'#3(1/*-//()+"/./*&!!" (-% /)222222222222222222222222222222222=A ?B< +1("#(&/ ,/1) 0 -#(1/10("/.'#3(1/*-/22 22222222222222222222222222222222222222222222= B8 "/3(1/*-/ 0 -#(1/10("/.'#3(1/*-/(-" )),/ 1 $22222222222222222222222222222222222:< B "/3(1/*-/ 0 -#(1/10("/.'#3(1/*-/(-" ))/* ,/"(*,n(-/222222222222222222222:: BA #00/"/*-/)(-" )),/ 1 $ +"-/) 0.'#3(1/*-/ (' +& -#(1/10("/C22288 B? #00/"/*-/)(-" ))/*,/"(*,n(-/ +"-/) 0.'# 3(1/*-/(' +& -#(1 /10("/C2222222222222222222222222222222222222222222 222222222222222222222222222222222222222222222222222 222222222222222228 B +"-/ 0#00/"/*-/)#* -#(1/10("/.'#3(1/*-/ (-" )),/ 1 $2222222222222228?

PAGE 8

3### B@ +"-/ 0#00/"/*-/)#* -#(1/10("/.'#3(1/*-/ (-" ))/*,/"2222222222222222228@ B< +"-/ 0#00/"/*-/)#* -#(1/10("/.'#3(1/*-/ (-" ))n(-/22222222222222222222228< @B8 ""/1(&# *(&"#7 0 -#(1/10("/ 1#-"/0/"/ *-/)(*,r3(1+(&# *) 0 +)%5 *$"/))5(*,&%/+!"/./ +"&22222222222222 222222222222222222222222222222222222222222222222228 8 @B %//,#(&#*$r00/-& 0 -#(1/10("/.'#3(1/*-/ *r3(1+(&# *)2222222222222222288@ @BA %//,#(&#*$r00/-& 0 -#(1/10("/.'#3(1/*-/ (-" )),/ 1 $222222222222222288< @B? %//,#(&#*$r00/-& 0 -#(1/10("/.'#3(1/*-/ (-" ))/*,/"22222222222222222288= @B %//,#(&#*$r00/-& 0 -#(1/10("/.'#3(1/*-/ (-" ))n(-/222222222222222222222288:

PAGE 9

#7 nr #$+"/ !($/ AB8 #)&"#'+&# 0 -#(1/10("/.'#3(1/*-/222222222 22222222222222222222222222222222222222222222222222? @ AB #)&"#'+&# 0 $*#/ *01#-&22222222222222222 222222222222222222222222222222222222222222222222222 2222222 ABA #)&"#'+&# 0 $*#/B00/-/ *01#-&2222222 222222222222222222222222222222222222222222222222222 2 AB? #)&"#'+&# 000/-/B00/-/ *01#-&2222222 222222222222222222222222222222222222222222222222222 22@ AB -#(1/10("/ 1#-"/0/"/*-/)222222222222222222 222222222222222222222222222222222222222222222222222 222222: AB@ 1#&#-(1* 61/,$/#)&"#'+&# *2222222222222222222 222222222222222222222222222222222222222222222222222 2222222@? ?B8-(&&/"!1 &) 03/"1(!!#*$ +"-/)22222222222 222222222222222222222222222222222222222222222222222 2222222222<:

PAGE 10

7 ')&"(-& 0#))/"&(&# *"/)/*&/,& &%/"(,+(&/ -% 1 0&%/*#3/")#& 01 "#,(#*("&#(1+10#11./*& 0&%/ n/>+#"/./*&)0 "&%//$"// 0 -& 0%#1 ) !% nrr rnrrrnr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

PAGE 11

8 rn8 nnrrnr #*-/&%/0 +*,#*$ 0&%/*#&/,&(&/)5&%/! 1#&# -) 0/(-%/"(%()'//*,/0#*/,' ) ./#))+/2*&%//("1/("),/'(&/6()-/*&/"/, *&%/)&"+$$1/0 "&%/"(�#-(&# 0 &%/ *)&#&+&# *'/&6//*&%//,/"(1#)&)(*,&%/(* &#B/,/"(1#)&)2%#))&"+$$1//3 13/, #*& &%/'(&&1/'/&6//** "&%/"*../"-#(1#*&/"/) &)(*,) +&%/"*($"#-+1&+"(1#*&/"/)&) /3/*&+(11"/)+1&#*$#*&%/#3#1("2 11 6#*$&% /#3#1("5#))+/)"/1(&/,& #*,+)&"#(1#;(&# *6/"/&%/!"#.("0 -+)2#*-/&%/ /6/(15#))+/)&%(&%(3/#*3 13/, &%/$ 3/"*./*&)/* .#-" 1/#*!" 3#,#*$0 "&%/ 6/10("/ 0#&)-#&#;/*)%(3/'//*&%/ !"#.(","#3#*$0 "-/'/%#*,! 1#&#-)2 3/")/3 /*,/-(,/)5&%/)/) -#(16/10("/! 1##))+/)%(3/-"/(&/,(,#3#,/#*&%/*#&/,&(&/)& %(&%()'//*(,/0#*#*$-%("(-&/"#)�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

PAGE 12

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n/3/",.#*#)&"(&# *(*,&%/$"#-+1&+"(1,D+)& ./*&,.#*#)&"(&# *E/**/&&(*, /**/&&8::G2)#*$)+"3/,(&(0" .&%/./"#-(* (&# *(1r1/-&# *&+,#/)Er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

PAGE 13

A &%/1()&A/(")2%/&%/",#3#,/,/3/*1 "1/))) 5&%/!+'1#-%()'//*)!1#&+!(1 *$ &%/)/1#*/)0 ") ./&#./2 %/!" '1/.6#&%(",/)(*,1,/*,#-4)EAG#*0/ "/*-/)(*, &%/")&%(&%(3/ +)/,r "#)&%(&&%//.!1 /,)+"3/#*,#-( & ")&%(&0(#1& (-+*&0 "&%/ ))#'#1#&&%(&.(*("/ n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r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r3#,/*-/)+$$/)&)& %(&/1/3(&/,.#,,1/"/)! *)/ 0"/>+/*-#/)("/ 0&/*("/)+1& 0* *(&&#&+,/)56%/ "/!/ !1/%(3/* "/(1! )#&# *&%/ #))+/E)%/"?G5(*,&%/-% )/&%#)"/)! *)/' /-(+)/#&#)* *B..#&&(12&(1)

PAGE 14

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n/)/("-%%() 00/"/,.#7/,"/)+1&)6#&%"/$(",)& #&)!"/3(1/*-/5'+&&%/./()+"/) 0 (.'#3(1/*-/ *' &%)#,/) 0&%/("$+./*&%(3/%(, !" '1/.)&%(&6#11'/,/)-"#'/,#* ,/&(#11(&/"2%+)5&%/>+/)&# *#)* &"/) 13/,2 */&%/1/))5) ./)+$$/)&&%(& (.'#3(1/*-/(' +&) -#(16/10("/#)0(#"1!"/3(1/* &E/1,.(*(*,9(11/"8::F(*&"#1 (*,(*&"#18:::F ,) *5(# 5(*,r))/)8G56% #1/ &%/")("$+/&%(&&%/!+'1#-#) H* &B) I(.'#3(1/*&E&//*'/"$/*(*,"/6/"F ('5G2%/ *-/!&+(1#;(&# *(*,./()+"/./*& 0(.'#3(1/*-/%( 3/* &'//*)&(*,(",(-" )) 1#&#-(1)-#/*-/ &%/",#)-#!1#*/)2%+)5/3#,/ *-/ 0&%/!"/3(1/*-/5) +"-/)(*, *)/>+/*-/) 0)+-%(.'#3(1/*-/#)(1) #**)#)&/*&2%/!"/)/ *&)&+,)+$$/)&)&%(& '/-(+)/&%/*-1+)# *)&%(&%(3/'//*,"(6* *' & %)#,/) 0&%/)/("$+./*&)6/"/

PAGE 15

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

PAGE 16

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rnrn G5 *01#-&#*$0//1#*$)E rn G5 "'/1#/0)#**01#-&6#&%0//1#*$)E rn rn GE&//*'/"$/*(*,"/6/"G2%#)-%(!&/"("$+/ )50#")&5&%(&(11&%"// &!/) 0*01#-&)&#.+1(&/(.'#3(1/*-/(' +&(11 0&%/) -#(16/10("/#))+/)./*&# */, (' 3/F(*,)/*,5&%(&6/-(*())/))&%/*)&"+-& 3(1#,#& 0/(-% 0&%/,#00/"/*& ./()+"/) 0(.'#3(1/*-/'.!("#*$% 66/11&%/& %/ "/&#-(1) +"-/) 0(.'#3(1/*-/ !"/,#-&/(-%"/)!/-/12%/0#*,#*$)#*,#-(&/&% (&(*(!!" (-%(,(!&/,0" /7!/"#./*&(16 "4') -#(1!)-% 1 $#)&)0 "+)/# *1("$/B)(.!1/)+"3/)E"(#$5(*/5

PAGE 17

< (*,("&#*/;F"(#$ G&%(&0 "-/)"/)! *,/*&)& "(&/&%/#"! )#&# 3/(*, */$(/0//1#*$))/!("(&/1#)&%/. )&3(1#,2 0&/")/&&1#*$ *&%/'/)&./()+"/5%(!&/""/3#) #&)!"/3# +)"/)/("-%&%(& (,,"/))/)&%/>+/)&# 06% #). )&1#4/1& '/ (.'#3(1/*&2%/0#*,#*$)!"/)/*&/, %/"/-%(11/*$/!"/3# +)())/"&# *)&%(&1#'/"(1)(" /. "/1#4/1& '/(.'#3(1/*&(' +& ) -#(16/10("/E/1,.(*(*,9(11/"8::F)//(1) ('G2*(,,#&# *5&%/"/)+1&) )+$$/)&&%(&* & *1)/10B#,/*�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

PAGE 18

= % 6&%/)/*-1+)# *))!/(4& 6(", +"+*,/")&(*,#* $ 0&%/ *-/!&+(1#;(&# *5 ./()+"/./*&5!"/3(1/*-/5) +"-/)(*,*)/>+/*-/) 0(.'#3(1/*-/(' +&) -#(1 6/10("/2#*(115(,#)-+))# *#) 00/"/,(' +&% 6 &%#))&+,(,,)& +"+*,/")&(*,#*$ 0 ./"#-(*)0//1#*$)(' +&) -#(16/10("/(*,6%(&# &./(*)0 "! 1#&#-)#*&%/*#&/, &(&/)& ,(2

PAGE 19

: rn rrr95rnr r nr5nrrr5nr5rrr5 r rr /%(3# "(1"/)/("-%/")%(3/'/$+*& /.'"(-/&%/#, /(&%(&!/ !1/, & */-/))("#1%(3/()#*$1/H&"+/I(&&#&+,/ *#))+/) 5'+&"(&%/"()& "/ 0.+1&#!1/(*, ) ./&#./)*01#-&#*$(&&#&+,/)&%(&&%/.#$%&,"( 6+! *(&(*$#3/*&#./E -%)-%#1, 8:=8F/1,.(*(*,9(11/"8::F9(11/"8::Fr($1( *,%(#4/*8::AF13("/;(*, "/%.8::F% .!) *59(**(5(*,"#00#*8::F"#/ )&/"(*,/&&8::@F".#&($/(*, ** "F/00/"&5+$/5(*, ,$/F#1) *5 #*,)/5(*,-% 1/"F (3#*/8F/6'B1("45-"/$ "5(*,9(**(F -"(65()/-4/5(*, *$/" AF)//"(#$(*,("&#*/;(5'0 "("/3 #/6G2%/#,/(&%(&#*,#3#,+(1)("/ 0&/* n 5 "&%(&&%/)#.+1&(*/ +)1! ))/))! )#/(*, */$(//3(1+(&# *) 0 ()#*$1/(&&#&+,/ 'D/-&E13("/;(*,"/%.8::F9 (11/"8::F9(11/"(*,/1,.(*8::F r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

PAGE 20

8 "/)!/-/0 -+)#*& >+/)&# *2)("/)+1& 0&%/) /!" '1/.)5#&#)+*-1/("D+)&% 6 ,#3#,/,&%/./"#-(*!+'1#-#) *&%/#))+/)&%(&) %(!//1/-&# *)(*,! 1#&#-)#*$/*/"(12 %#)-%(!&/""/3#/6)&%/1#&/"(&+"/6#&%"/$(",)& &%/ rnrn 5 n 5 n 5 r (*, rnn 0(.'#3(1/*-/2 11 6#*$&%#) '" (,"/3#/65!"/3# +)"/)/("-%0 -+)/,)!/-#0#-(11 *) -#(16/10("/(.'#3(1/*-/#) ,#)-+))/,2%#)#)0 11 6/,'()+..("&%(&0 ".( 11())/"&)&%/%! &%/)/)&%(&("/ &/)&/,#*&%/!"/)/*&)&+,2 nrnnn %/*-/!& 0(.'#3(1/*-/#)* &*/6E(!1(*8:
PAGE 21

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

PAGE 22

8 (.'#3(1/*-//#&%/"2 "/7(.!1/5#0) ./ */6())+! "/ 0)!/*,#*$0 "&%/! "'+& &)+!! "/ 00 ,)&(.!)5%/ ")%/#)* &*//))("#1(.'#3(1/*&(' +&) -#(1 6/10("/2%#).('/(1(-4 0*)&"(#*&E *3/")/ 8:@?G5* *B(&&#&+,/5 "&%/"/)+1& 0( % )& 0 &%/""/() *)2$(#*5(.'#3(1/*-/"/0/")& rnn )#/(*,*/$(/ /3(1+(&# *)"/$(",#*$( nr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r,(&(5/1,.(*(*, 9(11/"E8::G./()+"/,(.'#3(1/*-/'+*&#*$&%/ *+.'/" 0*01#-&#*$ *)#,/"(&# *)5)! *&(*/ +))&(&/./*&) 0(.'#3(1/* -/5(*,&6 B)#,/,"/.("4)E#2/25 H1&% +$%&%#*4N5*/3/"&%/1/))0(3 "IG50#*, #*$)&" *$)+!! "&0 "&%/!"/)/*-/ 0 (.'#3(1/*-/#*.(* 0&%/"/)! *,/*&)21&% +$%&% /)+')&(*/("$+./*&.('/ .!/11#*$5 &%/")-% 1(")(1) &/&%(&&%#)(!!" (-%#)* &3(1#,E"(#$5(*/5(*, ("&#*/;F13("/;(*,"/%.8::G213("/;(*, "/%.E8::G! #*& +&&%(&(* #*,#3#,+(1K)('#1#&& /7!"/))"/() *)0 "!/".#&& #*$(*,!" %#'#&#*$(' "&# *, /)* &

PAGE 23

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
PAGE 24

8? 13("/;(*,"/%.K)0#*,#*$)$#3/+)"/() *& '/1# /3/&%(&&%/!" '('1("/& ) ./ ,/$"//2+&&%#)) , /)* &(11 6+)& #*0/"# 0(* */#*,#3#,+(1#)(.'#3(1/*&21) 5 /"" "3("#(*-/)("/(*(--+.+1(&# 0/"" ")#*&% /'#*("-% #-/. ,/1(*,.('/( 0+*-&# 0* *(&&#&+,/)5+*-/"&(#*&5/>+#3 -(&# 5 "(% )& 0 &%/"0(-& ")()6/112 %#1/13("/;(*,"/%.$/&#&"#$%&'&#$%&/*#*$ +!&%/*-/!&+(1,/0#*#&# *5&%/#" ./()+"/0(#1)& ,#)&#*$+#)%(.'#3(1/*-/0" .&%/3 /"&%#*$)&%/)+$$/)&("/ !" '1/.(&#-6#&%&%//1,.(*(*,9(11/")&+,2)( "/)+1&5&%/3(1#,#& 0&%/#"./()+"/ #)>+/)&# *('1/2 ,#)&#*-&# *-(*'/.(,/'/&6//*&!/) 0./()+"/) 2())#1#E8::@G)+$$/)&/,&%(& (&&#&+,/)&"/*$&%%()'//*./()+"/,+)#*$ n (*, r #*,/7/) 0 )&"/*$&%2/0#*/,5H./&(B(&&#&+,#*(1#*,/7/)("/'( )/, *"/)! *,/*&)K#.!"/))# *) 0 &%/#" 6*(&&#&+,/)56%/"/() !/"(/#*,/7/)("/ 1#*4/,& &%/D+,$./*&!" -/))/) "/)! *)#'1/0 "(&&#&+,/"/)! *)/)IE())#1#8::@5 !2@A=G2%/)(./!"#*-#!1/#)(!!1#/,& ./()+"#*$&%/(&&"#'+&/ 0(.'#3(1/*-/2 )#.!1#0 &%/&/".#* 1 $0 "&%/"/.(#*,/" 0 &%/!"/)/*&)&+,5 n (*, r ./()+"/)("/"/0/""/,& () (*, n ./()+"/)5"/)!/-/1213("/;(*,"/%.)./()+" / 0(.'#3(1/*-/#) *)#,/"/,(* n ./()+"/ 0(.'#3(1/*-/'/-(+)/#&(&&/.!&)& ()-/ "&(#*&%/,/$"// & 6%#-%(.'#3(1/*-/#)!"/)/*&6#&% +&.(4#*$&%/ "/)! *,/*&(6("/2%+)5#&#) H#*,#"/-&2I&%/")%(3/+)/, ./()+"/) 0(.'#3(1/*-/6%#-%)#.!1()4 "/)! *,/*&)5+)#*$3("# +)6 ",#*$)&"(&/$#/)5& /! "&&%//7&/*&& 6%#-%&%/("/& "*2 "#*)&(*-/5"#/)&/"(*,/&&E8::@G%(,"/)! *, /*&)"/! "& *(*88B! #*&)-(1/&%/ ,/$"//& 6%#-%&%/0/1&*01#-&5#*,/-#)# *5 .#7/,6#&%"/$(",)& (% )& 0(&&#&+,/ 'D/-&)2r.!1 #*$()#.#1("(!!" (-%5% .!) *59( **(5(*,"#00#*E8::G+)/,&/*

PAGE 25

8 >+/)&# *),/3/1 !/,'(.#/) *E8:==G& !" '/6%/ &%/"!/ !1/6/"/H*0+)/,I H& "*2I(-# !! (*,%#)11/($+/)()4/,!/ !1/& "/! "&&%//7&/*&& 6%#-%&%/#" "/(-&# *)& (* 'D/-&6/"/.+,,1/,5,#3#,/,5&/*)/ 5*&"(,#-& "5D+.'1/,5*01#-&/,5 *)#)&/*&5+*#0 ".5(*,%(". *# +)E(-# !! 5(", */"5(*,/"*&) *8::
PAGE 26

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
PAGE 27

8< & #*)&(*-/) 0)&" *$#*&/"*(1#;/,*01#-&6%/"/ #*,#3#,+(1)-1/("1! ))/))! )#/(*, */$(//3(1+(&# *) 0()#*$1/(&&#&+,/ 'D/-&2 %/./()+"//.!1 /,'"(#$(*,%#) 11/($+/)#)&%/. )&*)#)&/*&6#&%&%#)*-/!& +(1#;(&# *2%/!"/)/*&)&+,+)/)(* (,(!&/,3/")# 0&%#)./()+"/2 6&%(&&%/)% "& .#*$) 0!"/3# +)6 "4!/"&(#*#*$ & &%/*-/!&+(1#;(&# *(*,./()+"/./*& 0(.'#3(1 /*-/%(3/'//*,/)-"#'/,56/-(* 3/ *& ,#)-+))#*$&%/!"/3(1/*-/ 0)+-%(.'#3( 1/*-/2 nn %/)&+, 0(.'#3(1/*-/#)$" 6#*$(-" )),#)-#!1# */)2*(*/7&/*)#3/"/3#/6 0 (&&#&+,/"/)/("-%0" .8::@B8:::5;D/*E8G,/3 &/,(*/*&#"/)/-&# *& (.'#3(1/*-/ #*,#-(&#*$#&)$" 6#*$!/"3()#3/*/))2.'#3(1/*-/" /)/("-%%()"/(-%/,#*& .(* )+')&(*/("/()#*-1+,#*$%/(1&%)&+,#/)5.("4/&# *$5(*,.(* &%/"("/()E".#&($/ (*, **/"F **/" 2F/6/11AG2&+,#/)%(3//3/*$ */()0(" ()& +)/'"(#*B#.($#*$&/-%* 1 $& #,/*�&%/*/+" 1 $#-(1(-#&()) -#(&/,6#&% (.'#3(1/*-/E+**#*$%(. 2AG2 6/3/"5&%//.!#"#-(1/3#,/*-/ 0(.'#3(1 /*-/ -/*&/"/, *! 1#&#-(1(&&#&+,/)(*,&%/#.!1#-(&# )&%/"/#*#))o#.#&/,5'+&$" 6#*$2& #)0(#"& )(&%(&&%/*-/!& 0(.'#3(1/*-/#)) &(*,(",#*&%/!)-% 1 $1#&/"(&+"/F(*,5 6%#1/&%/#,/(&%(&(.'#3(1/*-/#)(.! */*& 0! 1#&#-(1(&&#&+,/)#)* &,/'(&/,5#&) !"/3(1/*-/(*,*)/>+/*-/)("/(* &%/">+/)&# *2 %/"/)+1&)%(3/'//*.#7/,6%/*#& ./)& &%/)/>+/)&# *)E)//('F&//*'/" $/*(*,"/6/"G5'+&+!1/, 6#&%&%/./()+"/./*&#))+/)(-" ))&%/1#&/"(&+"/5 &%/"/#)D+)&* &/* +$%/3#,/*-/& ,"(6(*,/0#*#/*-1+)# *)2 n/)/("-%*-/"*#*$&%/!"/3(1/*-/ 0(.'#3(1/*-/% ())+$$/)&/,&%(&.(* ./"#-(*)("/(.'#3(1/*&(' +&(% )& 0,#00/"/*& &%#*$)#*-1+,#*$ rrn E13("/; (*,"/%.8::F"(#$5(*/5(*,("&#*/;G5 E"(#$ G5 r

PAGE 28

8= E/1,.(*(*,9(11/"8::F(*&"#1(*,(*&"#18::: F(#* +)(*,("&#*/; G5 nrnn E% *$(*, 1#*)4B(%.#()G5(*, nnrn )+-%() *$"/))5&%/!"/)#,/*&5(*,&%/+!"/./ +"&E-"(6(*,("&/1)G2 &%/"))+$$/)&&%(&(.'#3(1/*-/#)* &6#,/)!"/(,E &//*'/"$/*(*,"/6/"F ('5G5(&1/()&6%/*#&./)& )/3/" (1! 1#-, .(#*)2A%/"/-/"&(#*1 )//.)& '/. "/"/)/("-%)+$$/)&#*$&%/0 "./"2? nn #&%"/$(",& &%/) +"-/) ")&#.+1(*&) 0(.'#3(1 /*-/5&%/ "%() 00/"/,)/3/"(1 !"#.(")+)!/-&)#*-1+,#*$&%/*01#-& 0"/3(1 +/)E"(#$ F("&#*/; F/1,.(*(*,9(11/"8::F13("/;(*,"/%.8: :Fr($1(*,%(#4/*8::AF (&;(*,())8:==G&%/*01#-& 00//1#*$)E(3#* /(*,&//*'/"$/*F&//*'/"$/* (*,"/6/"G5(*,*01#-&'/&6//*&%/0 "./"( *,1(&&/"E(3#*/ 8::=F &//*'/"$/*(*,"/6/"G2 (1+/*01#-&#)&%/. )&0"/>+/*&1-#&/,) +"-/ 0(.'#3(1/*-/E/1,.(*(*, 9(11/"8::F13("/;(*,"/%.8::Fr($1(*,%(# 4/*8::AF9(11/"8::F(&;(*, ())8:==G2 ("/ 3/"("-%#*$* ".(/!"#*-#!1/)(*,'/1#/0() )+.!&# *)(' +& $ 3/"*./*&5-#&#;/*)%#!5(*,) -#/&E-(**8::
PAGE 29

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rn rn 2 1&% +$%#&#)$/*/"(11'/1#/3/,&%(&(.'#3(1/*-/ --+")6%/*&%/"/#)(*01#-& #*3 13#*$(!/") *)"/3(1+/)E13("/;(*,"/%. 8::Fr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

PAGE 30

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

PAGE 31

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

PAGE 32

(&&#&+,/)(' +&$("#$%&)2&&%/1/()&5&%/)/,#) -"/!(*-#/))+$$/)&. "/#*>+#"#) */-/))("*-/"*#*$3(1+/*01#-&()($*#/ ) +"-/ 0(.'#3(1/*-/2 &%()(1) '//*)+$$/)&/,&%(&(00/-/ "#/*&(&# *).(-1()%& !" ,+-/ (.'#3(1/*-/)+"" +*,#*$(!("&#-+1("(&&#&+,/ 'D/&E(3#*/(*,&//*'/"$/*F &//*'/"$/*(*,"/6/"G2*#*,#3#,+(1)0//1# *$)6#&%"/$(",& ) ./(&&#&+,/ 'D/-&.('/.#7/,20&%/)/(00/-/ "#/*&(&# ))&"+-&+"/(&&#&+,/)(' +&(* &%/" 'D/-&5.#7/,0//1#*$).("/)+1&#*(.'#3(1/*-/(' +&&%(& 'D/-&2%#)!" -/))-(*'/ &% +$%& 0() rn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rn E#*,/"(*,#*&/"8F /1) *8:::F13("/;(*,"/%.8::
PAGE 33

A 8::@F*(%(8:=@G(*, r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�# -(&# *5(*,!"/)#,/*&#(13 &/-% #-/ '+&($(#*5,+/& &%/1#.#&(&# *) 0&%/,(&(ErG 5&%/("/0 "-/,& #*0/"(.'#3(1/*-/ 0" .&%//"" "3("#(*-/#*/(-%"/)!/-/. ,/12 0$*#/ "#/*&(&# *)(*,(00/-/ "#/*&(&# ))&"+-&+"/(&&#&+,/)5#&#)(1) "/() *('1/& /7!/-&&%(&*01#-&'/&6//*&%/&6 .()&#.+1(&/(.'#3(1/*-/E(3#*/ 8::=F&//*'/"$/*(*,"/6/"G2n/)/("-%%()/ .'"(-/,&%/#,/(&%(& #*,#3#,+(1)(&&#&+,/)("/1#4/1)%(!/,'' &%0/ /1#*$)(*,$*#&# *)E"/-41/"(*, #$$#*)8:=:Fr))/)5(,, -45(*,9(**(8::AF#11( "(*,/))/"8:=:G2 ./*&/*, &%(&/. &# *)+*,/".#*/(*#*,#3#,+(1)('#1#&& .(4/"/() */,,/-#)# *)E/(")F @/1) *E8:::G, /))+$$/)&&%(&$*#//1/./*&) 0 +&$" +!(&&#&+,/), .#*(&/ (00/-&#*&%/#"#*01+/*-/ *! 1#!#*# *'+&&%# ))&+,, /)* &0 -+) *&%/!"#.(0 *//00/-& 3/"(* &%/"2

PAGE 34

? /(") 8:=G56%#1/ &%/")%(3/)+$$/)&/,&%(&&%/(00/-& #3/(*,$*#/'()/) 0 (&&#&+,/)("/$/*/"(11*)#)&/*&E"/-41/"8:=?F /)&#*$/"8:+/)&# *('1/2&//*'/"$/*(*,"/6/"EG'+# 1&(. ,/1&%(&#*"! "(&/)(11 0 &%/)/0(-& ")(*,&%/0 +*,3/"1#&&1/(.'#3(1/*/F% 6/3/"5&%/3(1#,#& 0&%/#" ./()+"/ 0(.'#3(1/*-/#),+'# +)2%/(+&% ")* &/ &/*&#(16/(4*/))/)6#&%"/$(",& &%/#"./()+"/51/(3#*$&%/, !/*0 "0+"&%/""/ )/("-%2 nnn r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

PAGE 35

G2(3#*/(*,&//*'/"$/*EG!"/)/*&/3#,/* -/ 0&%#)')% 6#*$&%(&*01#-& &/,#*0//1#*$)& 6(",1#'/"(1(*,*)/"3(/ ) -#(1$" +!)-(*#*01+/*-/-#&#;/*) 3 &#*$,/-#)# *)2 r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n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

PAGE 36

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nn %/"/()) ./())/"&&%(&.(*./"#-(*)("/(.'#3(1 /*&(' +&) -#(16/10("/ E/1,.(*(*,9(11/"8::F(*&"#1(*,(*&"#18:::F ,) *5(# 5(*,r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

PAGE 37

< *+.'/" 0*01#-&#*$*)#,/"(&# *)E/2$25(.#7 01#'/"(1(*,*)/"3(/../*&)G5 )! *&(*/ +))&(&/./*&) 0(.'#3(1/*-/E/2$25H)// ./"#&#*' &%)#,/)IG5(*,&6 B)#,/, "/.("4)E/2$25H/ !1/)% +1,&"& $/&(%/(, *& %/#" 6*5'+&$ 3/"*./*&)% +1,%/1! 6%/**/-/))("IG&%(&6/"/ 00/"/,#*"/)! *)/& !/*B/*,/,!" '/)(' +&) -#(1 6/10("/(&&#&+,/)#*&%/8:=+/)&# */,&%/6 "4 0(*&"#1( *,(*&"#15'+&0 ",#00/"/*& "/() *)2/("$+/,&%(&&%/)//.#*$1*&"(,#-& E "#**)#)&/*&G(&&#&+,/)#,/*�#/, #*&%(&)&+,("/* &(!" ,+-& 0(.'#3(1/*-/F"(& %/"5&%/"/!"/)/*&,#)&#*-&# *).(,/' -#&#;/*)'/&6//*,#00/"/*&&!/) 0$ 3/"*./*&)!/* ,#*$2%#1/('#)!" '('1

PAGE 38

= ""/-&#*())/"&#*$&%(&(.'#3(1/*-/(' +& */) #(16/10("/#))+/, /)* &*/-/))("#1 )#$*#0(.'#3(1/*-/(' +&(* &%/"5%/(1) *&/*,) &%(&H(.'#3(1/*-/#))#$*(1/,. )& -1/("1'&%//7#)&/*-/ 0)//.#*$1*&"(,#-& !(&&/"*) 0 !#*# *IE5!28AG2 #)0#*,#*$),#00/"0" .&% )/ 0(*&"#1(*,(*&" #15'+&%#)./()+"/)+00/")0" .&%/ )(./!" '1/.F nrnn )#&# *)("/* &&%/)(./()%(3#*$ rnn )#/(*, */$(//3(1+(&# *)"/$(",#*$( nr E)//(#* +)(*,("&#*/;G2 &//*'/"$/*(*,"/6/")EG/7(.#*(&# 0&%/ 8::r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rn rn E3(1+/3)23(1+/G5 E$" +!(00/-&3)2$" +!(00/-&G5(*, rn E3(1+/3)2$" +!(00/-&G*01#-&#*&%/#"(*(1)# )5&%/)/(+&% ")

PAGE 39

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r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n !" /-(+)/&%/"/%(3/'//**-/!&+(1(*, !/"(&# *(1 !" '1/.))+"" +*,#*$&%/ )&+, 0(.'#3(1/*-/#*$/*/"(1(*,)!/-#0#-& ) #(16/10("/5#&#)"/() *('1/& "/3#)#& ) ./ 0&%/>+/)&# *)(,,"/))/,'!"# ""/)/("-%5 '+&!/"%(!)6#&%('/&&/"(!!" (-%&

PAGE 40

A ./()+"#*$(.'#3(1/*-/2%/"/("/()/"#/) 0%! &% /)/)&%(&6#11'/&/)&/,&%" +$% +& &%/-%(!&/")#*&%#))&+,2%/&%/ "+*,/"1#*$& %/)/%! &%/)/)#)'"#/01,/)-"#'/, %/"/(*,/1(' "(&/, *&%" +$% +&&%/)&+, # $*#/5$*#/B(00/-/5(*,(00/-/*0 1#-&("/! )#/1"/1(&/,& (.'#3(1/*-/(' +&) -#(16/10("/2(1+/*01#-&#)&%/. )&-#&/,! &/*&#(1) +"-/ 0(.'#3(1/*-/E(&;(*,()) 8:==F/1,.(*(*,9(11/"8::Fr($1(*,%(#4/*8: :AF13("/;(*,"/%.8::F "(#$ F("&#*/; G2)3/"/,/("1#/"#*&%#)-%(!&/"5&%/1#& /"(&+"/ #*,#-(&/,&%(&/$(1#&("#(*(*,#*,#3#,+(1#)&3(1+/) )&"+-&+"/(&&#&+,/)(' +&) -#(16/10("/ E/1,.(*8:==F/1,.(*(*,9(11/"8::F-(**8::< F/1,.(*(*,&//*'/"$/* 8F "/*8F(1) )//#1/*)8::G5(*,&%/"/0 "/5&%/)/("/&%/3(1+/)0 -+)/, %/"/2(1+/)%(3/'//**-/!&+(1#;/,()($*# /*)&"+-&E)//&//*'/"$/*(*, "/6/"G2&%#*4#*$ 0&%#)'" (,1()( rn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r3/*#0$*#/*01#-&#*&%/0 ". 03(1+/* 01#-&#)() +"-/ 0(.'#3(1/*-/5#& #)! ))#'1/&%(& En 4/(-%8:
PAGE 41

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

PAGE 42

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nr (!!/("& '/" +$%1/>+#3(1/*&(. *$6 ./*(*,'1 (-4)5()/1,.(*(*,9(11/" .#$%&1/(,+)& /7!/-&E#*,/"(*,(*,/")8::@F /(&%/"?F(1) )//('G2 /-(+)/6 ./*(*,'1(-4)("/$/*/"(11&% +$%& 0( )) -#(16/10("/1#'/"(1)5(*, &%//3#,/*-/)+$$/)&)&%(&&%/#"3(1+/)("/* &,#) &"#'+&/,&%/6(#*6%#-%/1,.(*(*, 9(11/"())/"&56/)% +1,* &/7!/-&1#'/"(1)& '/ "/(.'#3(1/*&&%(**)/"3(/)2*

PAGE 43

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n/)/("-%%()/7&/*)#3/1/7!1 "/,&%/ rnn 0(&&"#'+&/) 0(&&#&+,/))+-% () rn E" )*#-4(*,'/1) *8::F" )*#-48:==(58:=='F *#*$/" 8::G5 E'/1) *8::F" )*#-4(*,'/1) *8::F" )*#-4 8::AG5 n E+,,8:=@F" )*#-4(*,-%+.(*8:==G5(*, E(;# 8:=@G& *(./(0/65'+&&%/"/%()* &'//**/("1().+-%" /)/("-%-/*&/"/, *&%/! &/*&#(1 1#&#-(1*)/>+/*-/) 0(.'#3(1/*-/20(.'#3(1/* -/#)(*(&&"#'+&/ 0(&&#&+,/)5(*, "/)/("-%#*,#-(&/)&%(&/3/" &%/"(&&"#'+&/ 0(& &#&+,/)#)*)/>+/*&#(15&%/"/#)/3/" "/() *& '/1#/3/&%(&(.'#3(1/*-/#)& 2n/)/("-% %())+$$/)&/,&%(&&%/"/1(&# *)%#! '/&6//*! 1#-!"/0/"/*-/)(*,/3(1+(&# *) 0! 1#& #-(1-(*,#,(&/)(*,#*)&#&+&# *)#) 6/(4/"0 ") ./ */6% #)(.'#3(1/*&(' +&&%/! 1# -E)//"(#$ 2GF%/*-/5 &%#)%! &%/)#)#)&/)&/,%/"/2 +%/./,#(&#*$/00/-& 0(.'#3(1/*-/(' +&) -#(16 /10("/#))&" *$/"0 *)/"3(/)5./*5(*,6%#&/)2%#)#)"/(11(*/7!1 "(& "%! &%/)#)2n/)/("-% %())+$$/)&/,&%(&(.'#3(1/*-/ %()(./,#(&#*$/00/-&5(*,&%(&#&)!"/3(1/*-/3(" #/)(-" ))$" +!)5) #&)//.) "/() *('1/& )//#0&%/./,#(&#*$/00/-&3("#/)(" ))$" +!)20*)/"3(/)5./*5(*,

PAGE 44

A? 6%#&/)("/. "/(.'#3(1/*&(' +&) -#(16/10("/&%( *&%/#""/)!/-/+*&/"!("&)5&%/ #*&+#//7!/-&(&# **-/"*#*$&%/./,#(&#*$/00/ -&(-" ))&%/)/$" +!)#)& /7!/-&&%/ /00/-&& '/)&" *$/"(. *$&% )/6% ("/. "/(.'# 3(1/*&2$(#*5&%#)#)(*/7!1 "(& %! &%/)#)2#&%* "/(11#&/"(&+"/& ,"(60" .5& %#)D+)&)//.)& '/&%/. )&"/() *('1/ /7!/-&(&# *2%(&). "/5&%/>+/)&# *#) */6 "&% (*)6/"#*$2)&%/./,#(&#*$/00/-& 0 ) -#(16/10("/(.'#3(1/*-/)&" *$/"0 ") ./J nn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

PAGE 45

A rnA rnrrrrn9 rnr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rr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

PAGE 46

A@ %/,/&(#1) 0(11&%/)/)+"3/./()+"/)("//7!1#(&/,%/"/2%/#*-1+,/>+/)&# 6 ",#*$(*,"/)! *)/,#)&"#'+&# *)(-" ))(11#*,/! /*,/*&(*,,/!/*,/*&3("#('1/)2#")&5 &%/,#"/-&(*,#*,#"/-&./()+"/) 0(.'#3(1/*-/(' +&) -#(16/10("/! 1#-("/,/)-"#'/,2 %/*&%/%! &%/)#;/,) +"-/) 0(.'#3(1/*-/E$*# /*01#-&5$*#/B(00/-/ *01#-&5(*,(00/-/*01#-&G*-/"*#*$(&&#&+ ,/)(' +&) -#(16/10("/! 1#-("/ ,/&(#1/,2#*(115(11 &%/"3("#('1/)("/,/)-"#'/ ,#*-1+,#*$,/. $"(!%#-)(*,*&" 1)2 %/,#)&"#'+&# *),/)-"#'/,%/"/)+$$/)&&%(&.(* ("/(.'#3(1/*&(' +&) -#(16/10("/2 1) 5.(*! ))/))*01#-&#*$3(1+/)E$*#/G5 *01#-&#*$0//1#*$)(' +&&%/ !/"-/#3/,'/*/0#-#("#/) 0) -#(16/10("/E(00/- /G5(*,*01#-&'/&6//*&%/&6 E$*#/B(00/-/G2 ,nrnn ),#)-+))/,#*&%/!"/-/,#*$-%(!&/"5"/)/("-%/") %(3//.!1 /,&6 '" (, )&"(&/$#/)0 "./()+"#*$(.'#3(1/*-/2*/(!!" (-% #*3 13/)+)#*$ ./()+"/) 0 (.'#3(1/*-/6%/"/"/)! *,/*&)("/)#.!1()4/,& ) &(&/#0&%/("/& "* ".#7/, *-/"*#*$&%/#))+/(&%(*,2%/)/("/-1())#0#/, (),#"/-&./()+"/)'/-(+)/&%/("/ '()/, *&%/)/10B())/))./*& 0&%/"/)! *,/*&2 %/ &%/"(!!" (-%#) n (*,#*3 13/)()-/"&(#*#*$&%/,/$"//& 6%#-% "/)! *,/*&)("/(.'#3(1/*&6#&% +&.(4#*$&%/.(6(" /&%(&(./()+"/./*& 0 (.'#3(1/*-/#)'/#*$*,+-&/,2'3# +)1&%/)/(!! (-%/)("/-1())#0#/,()#*,#"/-& '/-(+)/&%/#"#*&/*&#)& "/. 3/&%/)+'D/-#& 0&%/"/)! *,/*&2 "/!1#-(&/,(&()/&)'()/, *&%/,(&(56%/"/&%/.#) )#*$,(&(#*/(-%"/!1#-(&# *("/ )+')&#&+&/,6#&%,"(6)0" .&%/! )&/"# ",#)&"#'+& # 0&%/.#))#*$3(1+/*,#&# *(1 ')/"3/,3(1+/)E#&&1/(*,n+'#*8:=
PAGE 47

A< %/&6 ,#"/-&./()+"/) 0 rn /.!1 /,#*&%#))&+,("/ *1+)/,#*%(!&/"?2%/)/./()+"/)("/,/"#3/, 0" .()!1#&B)(.!1//7!/"#./*&(1 >+/)&# **&(#*/,#*&%/#*)&"+./*&2(10 0&%/) (.!1/6()"(*, .1())#$*/,&%/ 0 11 6#*$>+/)&# ./!/ !1/&%#*4&%/$ 3/"*./*&)% +1,!" 3#,/0/6 /")/"3#-/)5/3/*#*("/() )+-%()%/(1&%(*,/,+-(&# *5#* ",/"& "/,+-/)! /*,#*$2&%/")0//1#&#) #.! "&(*&0 "&%/$ 3/"*./*&& !" 3#,/. "/)/"3#/)& -#&#;/*)/3/*#0#&./(*) (*#*-"/()/#*)!/*,#*$2%#-% 0&%/)/! )#&# *)# )-1 )/)&& +" 6*3#/6)5 ("/ + rn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n/)! *,/*&)6/"/. "/1#4/1& -% )/&%/H& "*'/ &6//*)#,/)I"/)! *)/E?2TG&%(*

PAGE 48

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n 0(.'#3(1/*-/+)/)()#.#1("./&% ,& &%(& 01 3("/; (*,"/%.E8::58::<58::=5(*,G(*,('E G2$(#*5&%#)./()+"/./*& )&"(&/$#*3 13/)(*(1;#*$&%/"/)#,+(1) 0(. ,/ 1 0&%/(&&#&+,/ 'D/-&+*,/" /7(.#*(&# *2n/)#,+(1)("/&%/,#00/"/*-/'/&6//*& %/!"/,#-&/,3(1+/0" .&%//>+(&# (*,&%/(-&+(1 ')/"3/,"/)! *)/ 0&%(&#*,#3#,+(1 2%/#,/(%/"/#)&%(&&%/0#& 0&%/ />+(&# *& &%/,(&()% +1,'/6 ")/(. *$#*,#3#,+ (1)6% ("/(.'#3(1/*&'/-(+)/&%/ "(*$/ 0(--/!&('1/"/)! *)/)0 "&%/.#)$"/(&/"2 /-(+)/&%/%(3/' &%! )#/(*, */$(/0//1#*$)5&%/#""/)! *)/).(' +*-/0" )#,/& )#,/2%#)#*-"/()/)&%/ 3("#(*-/ 0&%/#""/)! *)/)& &%/#*,#-(& 0&%/ ,/!/*,/*&3("#('1/5(*,&%/"/0 "/5&%/ ,/16#11!/"0 ".! "1(&!"/,#-&#*$&%/#""/)! *)/)2 %/!"/)/*&)&+,/7&"(-&)&%/"/)#,+(1)0" .(. /1 0(&&#&+,/)(' +&) -#(1 6/10("/+)#*$ ",/"/,1 $#&() !! )/,& !" '#&E 13("/;(*,"/%.8::58::<58::=5

PAGE 49

A: G ",#*("1/()&)>+("/)E('G'/-(+ )/ rrn #) ./()+"/,+)#*$&%/ ",#*(1)-(1/,/)-"#'/,(' 3/0 "&%/,#"/-&./()+"/) 0(.'#3(1/*-/2 &%$" +!)E&% )/6% -% )/&%/& "*"/)! *)/(*, &% )/6% 00/"/,(.#7/,"/)! *)/G ("/.'#*/,0 "!+"! )/) 0&%/(*(1)#)%/"/2 ('1/AB8#)&"#'+&# 0n/)! *)/) *6 #"/-& /()+"/) 0.'#3(1/*-/ &(1/"-/*&($/ 00/"/,H "*In/)! *)//6/"/"3#-/)B&" *$18<28T/6/"/"3#-/)B &&" *$1A<2@T "//"3#-/)B &&" *$1?=82=T "//"3#-/)B&" *$18=A2T!rn"n# +/)&# *0 ".& &%/.+1("#(&/(*(1)#)!"/)/*&/, '/1 6& *&" 10 "(*>+/)&# *0 ".(&/00/-&2% //00#-#/*&0 "&%/0 ".,+..6() &"#3#(12 "&%/)(4/ 0!(")#. *5&%#)3("#('1/# ) .#&&/,0" .&%/. ,/1()"/! "&/,#* ('1/AB2A%/)/3("#('1/)6/"/)/1/-&/,'/-(+)/&%/("/&%/ !"#.(" */)#*&%/1#&/"(&+"/+)/,& /7!1(#*(&&#&+,/)(' +&) -#(16/10("/E)// "/* 8F/1,.(*(*,9(11/"8::F (+0.(**(*,/&" -#48:::F#1/*)8:==58::F (*,1+/$/18::AF(&/8::?F #*,/"(*,#*&/"8G2

PAGE 50

? /)&#.(&/)("//7&"(-&/,& -"/(&/(*/63("#('1/2 %/. ,/1, /)* &!/"0 ".6/11E)// ('1/ABG2*1A 0&%/)/3/*#*,/!/*,/*&3("#('1 /)"/(-%&"(,#&# *(11/3/1) 0)&(&#)&#-(1 )#$*#0#-(*-/E!V2G2% )/6% ("/. "/#*,#3#, +(1#)&("/1/))1#4/1& 0//1! )#/1 (' +&) -#(16/10("/2 *3/")/15/$(1#&("#(*#).#) )#/1()) -#(&/,6#&%)+!! "&0 ) -#(16/10("/2n/!+'1#-(*)("/1/))1#4/1& 0//1 )#/&%(*("/#*,/!/*,/*&) /. -"(&)2/#&%/""(-/5$/*,/"* "#*./#))#$*# 0#-(*&2%/!)/+, n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nB>+("/#)(. ,#0#-(&# 0&%/ 7 (*,*/11/00#-#/*&& ())+"/&%(&#& -(*3("0" .& 82%(&#)5($/14/"4/K)n,#3#,/) 7(*,*/11K)n'#&).(7#.+. #* ",/"& (-%#/3/(./()+"/&%(&"(*$/)0" .& 82%/"/0 "/($/14/"4/K)n6#11 ".(11'/%#$%/"&%(*&%/ 7(*,*/11./()+"/' +&6#11&/*,& "+*1 6/"&%(*&%/ ""/)! *,#*$n2 "(*/7(.!1/ 0&%/)(././&% ,+)#*$ ",#*("1 /()&)>+("/)"/$"/))# *)//("4) E?G2

PAGE 51

?8 ('1/AB ,/1 0&&#&+,/)(' +& -#(1/10("/ /00#-#/*& &(*,(",r"" :T *0#,/*-/*&/"3(1 r* .#-*,#3#,+(1#).B2882AB28=B2r$(1#&("#(*#).28<2A2882?//1#*$)O/*/0#-#("#/)B282B282:/. -"(&28:2B282:n/!+'1#-(*B2?:2B2=:B2:1(-42=2:B2=2=?/.(1/28=28B282?=*./2A2=B28A28=B1 $1#4/1#% ,8+/)&# *)&%(&6()#*&" ,+-/,() 0 11 6) K.* 6$ #*$& "/(, +()/"#/) 0 n (' +&&%/4#*,) 0&%#*$)) ./ !/ !1/&%#*4&%/$ 3/"*./*&)% +1,'/, #*$& (,," /))-/"&(#*!" '1/.)&%(&("/ 0(-#*$&%/+*&"20&/"/(-%5K,1#4/ +& "(& /&%/)&(&/./*& *(?B! #*&)-(1/ & #*,#-(&/% 6 r +0//1& 6(",#&20 +, &%(3/(*! )# / 0//1#*$)5$#3/#&&%/1 6/)&"(&#*$ 08F#0 +%( 3/) ./! )#/0//1#*$)5"(&/#&( F#0 +%(3/$/*/"(11! )#/0//1#*$)5"(&/#& (AF(*,#0 +%(3//7&"/./1 )#/0//1#*$)5"(&/#&(?21/()/"(&//(-%)& (&/./*&'()/,) 1/1 *% 6 )#/1 +0//1(' +&#&5 nrnrnrrnn nnr 2%/0#")&)&(&/./*&#)2222

PAGE 52

? %/)&(&/./*&)6/"/&%/*"/(,(*,"/)! *,/*&)6/"/ ()4/,& "(&//(-% */ )/!("(&/120(!/") *)//./,+*)+"/ "*0+)/,( &(*! #*&5#*&/"3#/6/")6/"/& 1,& "/!/(&&%/#*)&"+-&# *)().(*&#./)()*/-/))("2 %/)!/-#0#-()!/-&) 0) -#(16/10("/! 1#-E'()/ *>+/)&# *)0" .&%/r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�#-(&# *5! 1#&#-(1&"+)&5(*,3(1+/) #*,#-(& ")G2%/*&%/#*&" ,+-&# *6() "/!/(&/,/7-/!&6#&%&%/6 ",)H! )#/I(*,H! )# /1I"/!1(-/,'H*/$(/I(*, H*/$(/12I 11 6#*$&%#)#*&" ,+-&# *5"/)! *,/ *&)6/"/"/(,&%/)(./'(&&/" 0 #&/.)()'/0 "/2 /7&5#*,#-(& ") 0 rn 6/"/-(1-+1(&/,0 "/(-%#&/.+)#*$ (*(1$ "#&%.,/3/1 !/,'% .!) *(*,%/"11/($+ /)E8::F(1) )//(!1(*8:
PAGE 53

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

PAGE 54

?? ,#./*)# *(1E)//('8::?G5'+&()*0#".(& 0(-& "(*(1)#)#*,#-(&/)%/"/5 (.'#3(1/*-/(' +&)+-%! 1#-#/)(!!/(")& '/ * /,#./*)# *2* &%/"6 ",)5!/ !1/ &%(&("/(.'#3(1/*&(' +& */()!/-& 0) -#(16/10 ("/! 1#-("/ 0&/*(.'#3(1/*&(' +& &%/")2('1/ABA $#0&%/1$ "#&%. )#/n/(-&# *)"/ /$(/n/(-&# *)"/ */6 %"// +" /$(///1#*$)8222B2 .//$(///1#*$)228282/*/"(11/$(///1#*$)282A22r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

PAGE 55

? ('1/AB?"/>+/*-(*,*&/*)#& 0.'#3(1/*-/(' +& -#(1/10("/ &(*,(", *,#&# /(*"/ /3#(&# /"-/*&.'#3(1/*& /,#-(1*)+"(*-/2?882AA2?T ./1/))2:82<@82?Tr,+-(&# *2A882?<2AT&(*,(", 0#3#*$2:<82@@82?T '+("(*&//2=:82A@@2T%#1,("/2<@82:2@Tn/&#"/./*&/*/0#&)2A@82?@=2T.'#3(1/*-/*,/7?2@@2:?=2AT +.'/" 0()/)P@< (&(("/0" .( rr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

PAGE 56

?@ /n" /(*P?2@< &,2/32P@2:??@ P@< &/(&(("/0" .(1 "#,( &/")+"3/ 0"/$#) &/"/,3 &/")*,+-&/,#* (?2 #$+"/AB8#)&"#'+&# 0 -#(1/10("/.'#3(1/* -/ nn0nn nn #*-/#&6()())/"&/,#* !nr &%(&-#&#;/*)%(3/($*#/.(! 0 1#&#-)E(.!'/11 8:@G(*,&%(&!)-% 1 $#-(10 "-/))%(!/! 1#&#-( 1'/%(3# "5 "/)/("-%0 -+)#*$ *! 1#&#-(1$*#&# *'+"$/ */, E)// !4#*8::8F*#,/".(*5" ,5 (*,/&1 -48::8F)///*$("(*,-+#"/8::A(*, (+(*,/(")8:=@0 "("/3#/6G2 %/'()#-#,/(#))#.!1/2/ !1/.(4/,/-#)# *)(*, 0 ".(&&#&+,/)'+)#*$$*#/ )% "&-+&))+-%()!("&#,/*�#-(&# *5./,#(-+/)5 (*,3(1+/)5(. *$.(* &%/")2%/)/ )% "&-+&)!/".#&#*,#3#,+(1)& .(4/"/() *('1/,/#)# *)6#&%.#*#.(1/00 "&E !4#*

PAGE 57

?< 8::8F#)4/(*,(1 "8::8G2)0 "3(1+/)'/#*$( $*#/)% "&-+&5&%/#,/(#)&%(& #*,#3#,+(1), &*//,() !%#)&#-(&/,#,/ 1 $& ,/&/".#*/&%/#"! 1#&#-(1!"/0/"/*-/)2 %/#"!"/0/"/*-/) "! 1#&#-(1/3(1+(&# *).('/' ()/, *% 6*)#)&/*& "#**)#)&/*& 1#-#/) "! 1#&#-(1(-&# *)("/6#&%-/"&(#*'/1# /0)&%/! ))/))E/1,.(*8:==G2 1#-#/) (*,(-&# *).('/D+,$/,()"#$%& "6" *$'()/, *&%/#"*$"+/*-/6#&%,//!1%/1, 3(1+/)E)//n 4/(-%8:+/)&# *2 5&%//00/-& 03(1+/*01#-& #),/!/*,/*& *&%/!/") *(1#.! "&(*-/ !/ !1/!1(-/ *&%/#"3(1+/)2 %/1#&/"(&+"/)+$$/)&)&%(&3(1+/))+-%() n E/1,.(*8:==F/1,.(* (*,9(11/"8::F-(**8::
PAGE 58

?= ) -#(16/10("/2+(1#&(*, /* .#-#*,#3#,+(1#).()('/1#/0#*&%/0"//, .& (--+.+1(&/6/(1&%21&% +$% ,#00/"/*&)&+,#/).(*-/!&+(1#;/5 !/"(&# *(1#;/ 5(*,1('/1&%/3(1+/)#*,#00/"/*& 6()5&%/"/#)'" (,($"//./*&&%(&$"/(&/"#*,#3#, +(1#).#)()) -#(&/,6#&%1/)))+!! "& (*,$"/(&/"/$(1#&("#(*#).6#&%%#$%/")+!! "&0 ) -#(16/10("/!" $"(.)(*,)!/*,#*$2 %/ ")+$$/)&)&%(&$*#/*01#-&'/&6//*#*, #3#,+(1#).(*,/$(1#&("#(*#).#) rn () +"-/ 0(.'#3(1/*-/(' +&) -#(16/10("/! 1#2 '+#1,(./()+"/ 0*01#-&'/&6//*#*,#3#,+(1#) &(*,/$(1#&("#(*3(1+/)5 )/!("(&/#*,#-(& ") 0/(-%3(1+/6/"/0#")&*)&" +-&/,2n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rnr #*,#3#,+(1#).5 "('/1#/0#* &%/0"//, .& (--+.+1(&/6/(1&%2-% 1(")6#&%(, #00/"/*&)+')&(*/0 -+).#$%& !"/0/"& ./()+"/#*,#3#,+(1#).,#00/"/*&150 "/7 (.!1/5*-/!&+(1#;#*$#&#*&/".) 0( '/1#/0#*0"//, 0/7!"/))# *2

PAGE 59

?: *(11-()/)5"/)! *)/)6/"/,/,0" .8E)&" *$) +!! "&0 "&%/0#")&)&(&/./*&#* &%/!(#"G& E)&" *$)+!! "&0 "&%/)/*,)&(&/ ./*&GF0 "&%/)/*,/$(1#&("#(*#).!(#"5 &%#))"#*$6()"/3/")/,& !" 3#,/*)#)&/*-#* ,#"/-&# 06 ",#*$2%/&6 )/&) 0 #&/.)6/"/&%/*.'#*/,#*& #*,#-/)6#&%)"/)" (*$#*$0" .& 8E%#$%3(1+/) "/01/-&#*$)&" *$/")+!! "&0 "#*,#3#,+(1#)& "/$ (1#&("#(*3(1+/)G2%/)/#&/.),#,* & )-(1/&%(&6/11& $/&%/"E&(+'P28@5!U20 "&%/#*,#3#,+(1#)&#&/.)(*,& (+'P25! U20 "&%//$(1#&("#(*#&/.)G2%/)/#&/.)6/"/ '()/, *./()+"/)0" .r(*,&%/ )-(1/6/11& $/&%/"#*&% )/,(&('+&&%/(!!" (-% %/"/3("#/))1#$%&12%/.!(*# )&(&/./*&)+)/,%/"/#*/(-%#*,#-(& "("/,/"#3/, 0" .&6 )/!("(&/#*,#-(& ")0" r2%#)(11 6/,0 "&%/+)/ 0.+1&#!1/./()+"/) 6#&%0/6/">+/)&# *) *&%/)+"3/2 /"%(!)11(!)#*$&6 #&/.)#*& /(-%#*,#-(& ",/ -"/()/,&%/3(1#,#&5() !! )/,& %(3#*$0 +")/!("(&/>+/)&# *)0 "/(-%2 ./()+"/ 0 rnrn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

PAGE 60

. )& 0&% )/#*&%/)(.!1/,/. *)&"(&/(&1/()&) ./*01#-&2 n /n" /(*P2? &,2/32P2?=? P@< &/(&(("/0" .(1 "#,( &/")+"3/ 0"/$#) &/"/,3 &/")*,+-&/,#* (?2 #$+"/AB#)&"#'+&# 0 $*#/ *01#-& %/!"/)/*&)&+,!" )/)&%(&&%//00/-&) 03(1+ /*01#-& *(.'#3(1/*-/("/ ,/!/*,/*& *&%/!/") *(1#.! "&(*-/!/ !1/!1(-/ *&%/#"3(1+/)"/)!/-/12 rn #)'()/, *"/)! *)/)& &6 )/!("(&/#&/.)5#*&" +-/,()0 11 6)H) + 4* 65* &/3/" */($"//) *&%/,#00/"/*&$ (1) 3(1+/)&%(& +"*(&# +$%&& !+")+/2K.$ #*$& 1#)&&%"//:,#00/"/*&$ (1)(*,%(3/ +&/11./% 6#.! "&(*& /(-% 0 :%/)+"3/#*-1+,/,(./()+"/ 0&%/#.! "&(*-/ 0 &"(,#&# *(1. "(13(1+/)56%#-%6() ()4/,#*'/&6//*&%/&6 #*,#-(& ")+)/,%/"/2

PAGE 61

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n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n1n ),/)-"#'/,#*%(!&/"5"/)/("-%%()/.'"(-/,&% /#,/(&%(&#*,#3#,+(1) (&&#&+,/)("/1#4/1)%(!/,'#*,#3#,+(1)0//1#*$ )()6/11()$*#&# *)E"/-41/"(*,

PAGE 62

('1/AB#)&"#'+&# 0n/)! *)/) *(1+/.! "& (*-/*,#-(& ") &(1/"-/*&($/ *,#3#,+(1#)..! "&(*-/ &.! "&(*&(&118?2AT *1 ./6%(&.! "&(*&8<<:2T.! "&(*&8:A28Tr7&"/./1.! "&(*&8A@2?T @<82T r$(1#&("#(*#)..! "&(*-/ &.! "&(*&(&11A=@2AT *1 ./6%(&.! "&(*&8A2T.! "&(*&A<28Tr7&"/./1.! "&(*&:A?2?T @<82T &/(&(("/0" .( rr )+"3/ 0"/$#)&/"/,3 &/")*,+-&/,#*(?2 #))#*$3(1+/)6/"/"/!1(-/,+)#*$.+1&#!1/#.!+&(& # *2 #$$#*)8:=:Fr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n rn ("/./()+"/,6#&%&6 (,,#/#*,#-/)5 */0 "! )#/ 0//1#*$)(*,(* &%/"0 "*/$(/0//1#*$)2%/)/# *,#-/)("/'()/, *(*)6/")& &6

PAGE 63

A >+/)&# *)&(!!#*$"/)! *,/*&)K(00/-&& 6(",H! !/ !1/I(*,H'1(-4)2In/)! *,/*&) 6/"/"/(,&%/0 11 6#*$#*&" ,+-&# /7&5K,1#4/& &%/)(./&%#*$/7-/!&6#&%(1 #)& 0,#00/"/*& rnn nrn (*, r &%(&("/(-/#*! 1#&#-)2*-/($(#*0 +, &%(3/ nr 0//1#*$)& 6(",&%/#*)&#&+&# "$" +!5$#3/#& &%/1 6/)&"(&#*$ 08F #0 +%(3/ rr 0//1#*$)5"(&/#&(F#0 +%(3/ nr 0//1#*$)5"(&/#&(AF(*,#0 +%(3/ r 0//1#*$)5"(&/#&(?21/()/ "(&//(-%#*)&#&+&# "$" +!'()/,) 1/1 *% 6 )#/1 +0//1(' +&#&5 nrnrnrrnnn nr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
PAGE 64

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o 6/"0 "$*#/B (00/-/E2A@G() !! )/,& $*#/*01#-&E 2?G211#*(115#&(!!/(")&%(&!/ !1/ ("/. "/1#4/1& /7!/"#/*-/$*#/*01#-&2 */&%/1/))5&%#), /)* &./(*&%(& $*#/B(00/-/*01#-&#)* &(! &/*&#(1) + "-/ 0) -#(16/10("/(.'#3(1/*-/2 1n ),/)-"#'/,#*%(!&/"5(00/-/ "#/*&(&# *). (-1()%#*,/!/*,/*&1 0 $*#&# *& !" ,+-/(.'#3(1/*-/)+"" +*,#*$(!("& #-+1("(&&#&+,/ 'D/-&E(3#*/(*, &//*'/"$/*F&//*'/"$/*(*,"/6/"G2#. !15#0(*#*,#3#,+(1%().#7/, 0//1#*$)(' +&) ./&%#*$5(*,&%/)/(00/-/ "#/* &(&# *))&"+-&+"/(&&#&+,/)(' +& (* &%/" 'D/-&5(.'#3(1/*-/.('/&%/"/)+1&2%#) !" -/))-(*'/&% +$%& 0()

PAGE 65

/n" /(*P2A@?A &,2/32P2A@? P@< &/(&(("/0" .(1 "#,( &/")+"3/ 0"/$#) &/"/,3 &/")*,+-&/,#* (?2 #$+"/ABA#)&"#'+&# 0 $*#/B00/-/ *0 1#-& rn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

PAGE 66

@ *01#-&#)1/))0"/>+/*&&%(*$*#/*01#-&( *,$*#/B(00/-/*01#-&2%/ (3/"($/)"/#)1 6/"E2A8G&%(*' &%$*#/*01#-&E2?G(*,$*#/B(00/-/ *01#-&E2A@G21/("1&%#)&!/ 0*01#-&#)* &6#,/)!"/(,5'+&($(#*5&%#), /)* & ./(*#&#)* &() +"-/ 0) -#(16/10("/(.'#3(1/* -/6%/*!"/)/*&2 n /n" /(*P2A=A &,2/32P2
PAGE 67

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rnnnnn2 "n "/3# +)"/)/("-%%())+$$/)&/,&%(&&%/!" '('#1#& 0(*#*,#3#,+(10//1#*$ (.'#3(1/*&(' +&(! 1#-#))+/.('/"/1(&/,& & %/!/") *K)! )#&# *&%(&#))+/ E"(#$ F"(#$ F(#* +)(*,("&#*/;G2" B1#0/3 &/")5 0 /7(.!1/5&/*,& '/. "/(.'#3(1/*&(' +&6%/&%/"( "&# *)% +1,'/1/$(1+*,/" H&"(+.(&#-I-#"-+.)&(*-/)56%#1/!" B-% #-/3 &/") ("/. "/(.'#3(1/*&(' +&&%/ 1/$(1#& 0H/1/-/I(' "&# *)E"(#$ G21) 5!/ !1/6#&%. "/! )#/3#/6)

PAGE 68

= (' +&% )/7+(1#&#*$/*/"(1("/1/))(.'#3(1/*& (' +&$("#$%&) *#))+/)&%(&, nr ,#"/-&1#*3 13/-%#1,"/* ".(""#($/E"(#$ G2#*(115(#* +)(*,("&#*/; EG)% 6&%(&)+!! "&/") 0) -#(16/10("/! 1#("/1/))1#4/1& '/(.'#3(1/*&2 ('1/AB@#)&"#'+&# 0n/)! *)/) 3/"*./*&r 3(1+(&# *)*,#-(& ") &(1/"-/*&($/ r3(1+(&# 0/ "$/2+)%#)(!!" 3/)&" *$1=A<2@T#)(!!" 3/* &)&" *$1<@82T#7/,O*B'/&6//*8:A28T!!" 3/* &)&" *$1=:2@T!!" 3/)&" *$1@A<2T @<82T r3(1+(&# 022 *$"/))#)(!!" 3/)&" *$18A?28T#)(!!" 3/* &)&" *$18?@?28T#7/,O*B'/&6//*@:882?T!!" 3/* &)&" *$18<2:T!!" 3/)&" *$1888@2@T @<82T r3(1+(&# 0&%/+!"/./ +"&#)(!!" 3/)&" *$18A8<2T#)(!!" 3/* &)&" *$18@2=T#7/,O*B'/&6//*=A8A2
PAGE 69

: )!/*,#*$56#&%&% )/6% ("/& "* ".#7/,#*B'/&6/ /*2%/!/"-/*&($/ 0"/)! *)/)' -(&/$ "#),#)!1(/,#*#$+"/AB2*&/"/)&#*$1 &%/"/)! *)/)/1/-&/,. )& 0&/*6()0 "/)/"3#-/)/3/*#0#&./(*)(*#*-"/()/#*$ 3/" *./*&)!/*,#*$E*/("#*$?TG2 n +$%1/>+(1!/"-/*&($/)ED+)&'/1 6TG)/1/-&/, 0/6/")/"3#-/))&" *$15.#7/, & "*5(*,. "/)/"3#-/)* &)&" *$156%#1/ *1(' +&8T-% )/0/6/")/"3#-/)* & )&" *$12 "/)/"3#-/) )&" *$1 "/)/"3#-/) &)&" *$1 .#7/, "& "* 0/6/")/"3#-/) &)&" *$1 0/6/")/"3#-/) )&" *$1 n &/(&(("/0" .(1 "#,( &/")+"3/ 0"/$#) &/"/,3 &/")*,+-&/,#* (?2 #$+"/AB -#(1/10("/ 1#-"/0/"/*-/)r "4"rnn4540n4nr n )/"#/) 0*&" 13("#('1/)("/(1) /.!1 /,#* &%#))&+,2%/)/3("#('1/)("/ +)/,& /7!1 "/,#00/"/*-/)(-" ))$" +!)5!("&#-+1 ("15,#00/"/*-/)#*&%/!"/3(1/*-/ 0

PAGE 70

@ (.'#3(1/*-/(-" ))#,/ 1 $5$/*,/"5(*,"(-/2)* &/,/("1#/"5/1,.(*(*,9(11/" E8::G*-1+,/,&%(&) -#(16/10("/1#'/"(1)56% &/*,& !1(-/(%#$%3(1+/ *' &% /$(1#&("#(*#).(*,#*,#3#,+(1#).5("/. "/!" */& (.'#3(1/*-/&%(**)/"3(/)2%#) !" )#&# *#)"/3#)#&/,6#&%&%/*/6./()+"/ 0(. '#3(1/*-/!"/3# +)1,/)-"#'/,2 ./()+"/#,/ 1 $5(&"(,#&# *(1)/3/*B! #*&#*,#-(& "#)+)/,%/"/2n/)! *,/*&)6/"/"/(, &%/0 11 6#*$)&(&/./*& /%/("(1 & 0&(14&%/)/,()(' +&1#'/"(1)(*, *)/"3(/)2*()-(1/ 0 */ &%" +$%)/3/*56%/"/H8I#) (*,H+/)&# *2%#) -"/(&/,(&"(,#&# *(1
PAGE 71

@8 #))+/)#*,/0#*#*$' &%&%/$/*,/"$(!E#1/*)8:== F(+0.(*(*,/&" -#48:::F "/* 8G(*,&%/"(-#(1-1/(3($/E (*,1+/$/18:: AF(&/8::?F#1/*)8::F#*,/" (*,#*&/"8G#*./"#-(*! 1#&#-))+$$/)&)&%(& $/*,/"(*,"(-/("/#.! "&(*& !"/,#-& ") 0(&&#&+,/)(' +&) -#(15(*,&%/"/0 "/ 5&%/"/.('/,#00/"/*-/)#* (.'#3(1/*-/(-" ))&%/)/$" +!)28/-(+)/#*./#)#.! "&(*&#*)&"+-&+"#*$(&&#&+, /) (' +&) -#(16/10("/E "/*8G#&#)(1) #*-1+, /,()!"/,#-& 0) -#(16/10("/ (.'#3(1/*-/2%/,#)&"#'+&# *(-" ))(11 0&%/)/3 ("#('1/)#)*&(#*/,#*('1/AB<2 6n! r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n#-(*)5'+&(1) /7#-(*)F)//,/1((";( 8::G(*,&%/"/#)(1("$/+'(* !+1(&# *#*1 "#,(2)("/)+1&5#&#)* &)+"!" #)#*$& 1/("*&%(&&%/(&#* )#*&%/ rr )+"3/, &5 *(3/"($/5,#00/")#$*#0#-(*&10" .6%#&/)#*&/".) 0&%/ 3("#('1/)&%(&("/. )&-"#&#-(1& +"(*(1)#)2

PAGE 72

@ ('1/AB<#)&"#'+&# 0,/ 1 $5("&,/*�#(&# *5n(-/5/*,/"5(*,*./ &(1/"-/*&($/ ,/ 1 $#'/"(18<2:T ,/"(&/8=8<2=T *)/"3(/A<@82AT @<82T ("&,/*�#-(&# */. -"(&AAA=2?Tn/!+'1#-(*?A:2:T*,/!/*,/*&8?2?T&%/" =82AT @<82T n(-/ %#&/ 8=?2=T 1(-4 8=2?T #)!(*#-AA2=T&%/" 8=A2T @<82T /*,/"/.(1/A@:2AT(1/ ?+/)&# *)#*/(-%)&+, & ./()+"/#*0 ".(&# *1/3/1)2 %/!"/)/*&)&+,(1) /.!1 )(./()+"/ 0 r'nr 2n/)! *,/*&) 6/"/"/(,&%/0 11 6#*$#*&" ,+-&#

PAGE 73

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

PAGE 74

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

PAGE 75

@ *&%/*/7&-%(!&/"5&%/"/1(/*)&"+-&3(1#,#& 0/(-% 0&%/./()+"/) 0 (.'#3(1/*-/#).!("/,(*,&%/"/)+1&)6#11)+$$/) &&%(&&%/./()+"/(,(!&/,0" .&%/ #*,#"/-&(!!" (-%+)/,'"(#$(*,%#)11/($+/) E5G#))+!/"# "2%/0#*,#*$) (1) )+$$/)&&%(&(11 0&%/(' 3/B,/)-"#'/,! &/*& #(1) +"-/) 0) -#(16/10("/ (.'#3(1/*-/("/5#*0(-&5"/1#('1/!"/,#-& ") 0)+ -%2

PAGE 76

@@ rn? rrnrrrnr rr rnr %/#,/(&%(&!/ !1/)(&&#&+,/)(' +&#))+/)("/( .#7 0*)#,/"(&# *)&%(&.( "/)+1&#*(.'#3(1/*-/#)'/.#*$(--/!&/,(-" )), #)-#!1#*/)2%/!" '1/.& ,(#)&%(& )+"3/"/)/("-%#)) &"/(&#*$(&&#&+,/)()#0& %/("/+*#B! 1("2#.!15)&(*,(",)+"3/ #*,#-(& "), & 00/"(*))&/.(&#-6( 0)/!(" (&#*$&% )/6% %(3/' &%! )#/ (*,*/$(//3(1+(&# *) 0($#3/* 'D/-&0" .&% )/6% %(3/(-1/("! )#&# *2%#) -%(!&/"/3(1+(&/)&%/ (*, n ./()+"/) 0) -#(16/10("/(.'#3(1/*-/&%(& 6/"/,/)-"#'/,#*&%/!"/3# +)-%(!&/"2%/0#*,#*$ )#*,#-(&/&%(&&%/#*,#"/-&./()+"/ 0 ) -#(16/10("/(.'#3(1/*-/&%(&0 "-/)"/)! *,/*&) & "(&/&%/#"! )#/(*,*/$(/ 0//1#*$))/!("(&/1!/"0 ".)'/&&/"&%(*(* 0&%/ &%/"(!!" (-%/)-+""/*&1 00/"/,#* &%/1#&/"(&+"/20+11)!/-#0#/,. ,/1 0) -#(16 /10("/(.'#3(1/*-/#)!"/)/*&/, 0 11 6#*$&%#)(*(1)#)2 %#1/)+"3/"/)/("-%/")%(3/'/$+*& (--/!&&%(& n #)()&(*,(", (&&"#'+&/ 0(&&#&+,/)E/1,.(*(*,9(11/"8::F9( 11/"8::F13("/;(*,"/%.8::F "(#$5(*/5(*,("&#*/;G5&%/%(3/* &!/"0 /-&/,(6(& -(!&+"/&%#) !%/* ./* *2+"./()+"/)%(3/* &-(+$%&+!6#&%&% / "2*(,,#&# *5&%/"/#) *)#,/"('1/,#)($"//./*&*-/"*#*$' &%% 66/*-/!&+(1#;/(.'#3(1/*-/(*,% 6 6/./()+"/#&20(.'#3(1/*-/#)()&(*,(",(&&"#'+ &/ 0(&&#&+,/))+-%() rn E" )*#-4(*,'/1) *8::F" )*#-48:==(58:=='F *#*$/"5" )*#-45 /"/*&8::G5 nn E" )*#-4(*,'/1) *8::F" )*#-4 28::AG5 (*,

PAGE 77

@< n E13("/;(*,"/%.8::G5(*,&%/)/(&&"#'+&/)("/ *)/>+/*&#(1& +" +*,/")&(*,#*$ 0(&&#&+,/)#*$/*/"(1E)//())#1# 8::@0 "("/3#/6G5#&)//.)1 $#-(1&%(& 6/)% +1,)/&&1/ *(*(!!" (-%& ./()+"/&%/(&&" #'+&/ 0(.'#3(1/*-/2/%(3/ )&(*,(",(!!" (-%/)#*)+"3/"/)/("-%& -(!&+"/& %/)/ &%/"!%/* ./*(5) #&)//.) #.!/"(/& +"+*,/")&(*,#*$& ,/3/1 !)+-%(. /()+"/0 "(.'#3(1/*-/2 0)&(*,(",#*,#-(& ")+)/,#*&%/.(D ")+"3/)+& #1#;/,#*&%/,#)-#!1#*/E./"#-(* (&# *(1r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r n &%(&)#.!1()4"/)! *,/*&)#*) ./6(& )&(&/6 %/&%/"&%/("/& "* '/&6//*)#,/) *&%/#))+/(&%(*,(*, nrn &%(&/.!1 ) ././(*)& ()-/"&(#*&%/,/$"//& 6%#-%#*,#3#, +(1)("/(.'#3(1/*&6#&% +&.(4#*$ "/)! *,/*&)(6("/&%(&&%#)!%/* ./* *#)'/#*$./( )+"/,2"# ""/)/("-%#*,#-(&/)&%(& &%/""/1(&# *)'/&6//*,#"/-&(*,#*,#"/-&./()+" /) 0(.'#3(1/*-/("/. ,/)&#*

PAGE 78

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rnrn 5 rn 5(*, rn rn !"/,#-&/(-%./()+"/ 0(.'#3(1/*-/2%#)&/)&%() &6 '3# +)())+.!&# *) 0#")&5&%(&&%/ ")+$$/)&#*$&%(&&%/)/("/) +"-/ ) 0(.'#3(1/*-/#)(--+"(&/5(*,)/*,5 &%(&&%/./()+"/) 0&%/)/*-/!&)("/3(1#,2)0 "&%/0#")&())+.!&# *5&%/ "#*&%/)/ "/$(",)#)0(#"1,/3/1 !/,5#&#)D+)&&%(&./()+" /)&%(&%(3/* &-(+$%&+!2 *-/"*#*$ &%/3(1#,#& 0&%/./()+"/)5&%/#*,#-(& ")+)/,6 /"/(,(!&/,0" .r(*,5(*, "/)/("-%%()"/!/(&/,1,/. *)&"(&/,&%/#"/00#-(6%/*#&./)& !"/,#-&#*$ ,/!/*,/*&3("#('1/)&%(&&%/ ")+$$/)&)#&)% +1,2 06/(--/!&&%#)()(&/)& 0&%/#" 3(1#,#&5#&)//.)("/() *('1/&/)&& .!("/&%/ 3(1#,#& 0&%/./()+"/) 0 (.'#3(1/*-//.!1 /,%/"/2 %/0#*,#*$)#*,#-(&/&%(&&%/&%/ "/&#-(1) +"-/) 0(.'#3(1/*-/!/"0 ".'/)&() !"/,#-& ") 0&%/#*,#"/-&./()+"/&%(&%()"/)! *, /*&)"(&/! )#/(*,*/$(/0//1#*$)

PAGE 79

@: )/!("(&/12%#1/&%#)#*,#"/-&./()+"/.('/&%/ )&3(1#,5&%/!"/)/*&)&+,*&/*,) &%(&#&#)* &&%/. )&!"(-&#-(1'/-(+)/ 00#*(*#(1)&21/("1. "/6 "4*//,)& '/ */5'+&&%//3#,/*-/!"/)/*&/,%/"/)+$$/)&)&%(& &%#)#)(*/*,/(3 "&%(&*//,) +*,/"&(4#*$2/0 "/. 3#*$ *& &%/(*(1)#)5,#"/ -&(*,#*,#"/-&(!!" (-%/)("/ -"#&#>+/,#*. "/,/&(#1&%(*!"/3# +)-%(!&/")2% #)-%(!&/"*-1+,/)6#&%&%/ !"/)/*&(&# *(*,,#)-+))# 0(0+11)!/-#0#/,. ,/1 0) -#(16/10("/(.'#3(1/*-/ +)#*$&%/. )&3(1#,#*,#"/-&./()+"/2n n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

PAGE 80

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
PAGE 81

<8 */.!"#)/, 0 !/"(/!)-% 1 $#-(1!" -/))/) (*,&%/ &%/".!"#)/, 0 #.!"/))# *) 0&%/)/!" -/))/)2 5,#"/-&./()+"/) .('/!#-4#*$+!!/ !1/)"/(-&# & 0//1#*$(.'#3(1/*&"(&%/"&%(*&%/(-&+(1!%/* /*(20"/(-&# *)3("5&%/*&%/,#"/-& ./()+"/#)-(!&+"#*$) ./&%#*$ &%/"&%(*&%/!%/* ./*(#&)/102n(&%/"5#&.('/&%/ '!" ,+-& 0(.'#3(1/*-/2 rn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

PAGE 82

< ./()+"/./*&)&"(&/$#*3 13/)(*(1;#*$&%/"/)#,+( 1) 0(. ,/1 0&%/(&&#&+,/ 'D/-& +*,/"/7(.#*(&# *2n/)#,+(1)("/&%/,#00/"/*-/'/& 6//*&%/3(1+/0 "/(-%"/)! *,/*& !"/,#-&/,0" .&%/!" '#&/>+(&# *(*,&%/(-&+(1 ')/"3/,"/)! *)/ 0&%(&#*,#3#,+(12 %/*&/*,&%(&&%/3("#(*-/ 0&%/)/"/)#,+(1)) % +1,'/%#$%/"(. *$&% )/6% ("/ (.'#3(1/*&2r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nr 5% 6! )#/1(*,*/$(/1&%/0//1& 6(",&% / 'D/-&(&%(*,2 %#)#)6%/"/&%/(!!" (-%+)/,'"(#$(*,%#)11/($+/)#)&%/)&" *$/)&2%/

PAGE 83

+/)&# *).('/(.'#$+ +).(4#*$ &%/.#*)+00#-#/*&()#*,#-(& ") 0!"/0/"/*-/E)// 13("/;(*,"/%.8::58::<58::=G2 %/"/0 "/5&%/! ))/))# 0 !! )#*$*)#,/"(&# )#)* &*/-/))("#1(!" ,+-& 0 "/!"/)/*&(/ 0(.'#3(1/*-/2n/./.'/"&%(&(.'#3( 1/*-/,/0#*/,()(*(&&"#'+&/ 0 (&&#&+,/)"/0/")& *01#-&(' +&()#*$1/(&&#&+, / 'D/-&2r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r"" "3("#(*-/+1,'/(

PAGE 84

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n":"n n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

PAGE 85

< E"#/)&/"(*,/&&8F+11#$(*(*,-"(6F /6'B1("45-"/$ "5(*, 9(**(G5'+&(1) ,/. *)&"(&/)&%(&&%/"/#)* &.+-% 0("/1(&# *)%#!'/&6//*&%/ &6 #*,#"/-&./()+"/)2&(!!/(")&%(&/(-%#)-(!& +"#*$(,#)&#*-&!%/* ./* *2 ('1/?B8 ""/1(&# *)(-" ))#"/-&(*,*,#"/-& /()+"/) 0.'#3(1/*-/ #"/-& H "*I #"/-&H 1+*&//"/,I *,#"/-&n/)#,+(1) *,#"/-& )O/$ #"/-&H "*I82BBBBBB#"/-&H 1+*&//"/,IBB82BBBB*,#"/-&n/)#,+(1)2A8Y2AY82BB*,#"/-& )O/$2@B2=28AY82 &/(&(("/0" .( rr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

PAGE 86

<@ -()/50(-/3(1#,#&1/(,)& *)&"+-&3(1#,#&2n/ )/("-%/")%(3/+)/,&%#) *-/!&+(1#;(&# *& '+#1,&%/ "#/)(' +&&%/! &/* &#(1) +"-/) 0(.'#3(1/*-/2 !/"%(!)5&%#)./()+"/#)'/&&/")+#&/,& &/)&&%/ /7#)&#*$&%/ "E*01#-& 0$*#/(*, (00/-//1/./*&)G20(.'#3(1/*-/#)!" !/"1,/0 #*/,())#.+1&(*/ +)1! ))/))#*$ )#/(*,*/$(/0//1#*$)(' +&) -#(16/10("/ 5&%/*&%/&%/ "/&#-(1) +"-/) 0)+-% (.'#3(1/*-/)% +1,!"/,#-&&%/./()+"/2('1/?B ""/1(&# *)'/&6//* +"-/)(*, +".' #3(1/*-//()+"/) #"/-& H "*I #"/-&H 1+*&//"/,I *,#"/-&n/)#,+(1) *,#"/-& )O/$ $*#/ *01#-&2@28
PAGE 87

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

PAGE 88

<= ('1/?BA*(1)#) 0n/)#,+(1)0" ,/1 0&&#& +,/)(' +& -#(1/10("/ ("#(*-/ 0n/)#,+(1) /3/*/&(&#)&#" '('#1#& (rrn) 3/B(3/"($/ $*#/ *01#-&82: <228 /1 6B(3/"($/ $*#/ *01#-&2: (rrn 3/B(3/"($/ $*#/B00/-/ *01#-&82:: 22:: /1 6B(3/"($/ $*#/B00/-/ *01#-&2 (rrn* 3/B(3/"($/00/-/ *01#-&82: 82A:2? /1 6B(3/"($/00/-/ *01#-&288 -. (&(("/0" .( rr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

PAGE 89

<: n n1n n n1n &/(&(("/0" .(1 "#,( &/")+"3/ 0"/$#) &/"/,3 &/")*,+-&/,#*(?2 #$+"/?B8-(&&/"!1 &) 03/"1(!!#*$ +"-/)

PAGE 90

= ('1/?B? ""/1(&# *)'/&6//* +"-/) 0.'#3(1/* -/(' +& -#(1/10("/ $*#/ *01#-& $*#/B00/-/ *01#-& 00/-/ *01#-& $*#/ *01#-&82BBBB $*#/B00/-/ *01#-&2?82BB00/-/ *01#-&2+(&# *)&%(& ,/1/(-% 0&%/0 +"./()+"/) 0(.'#3(1/*-/()( 0+*-&# 0&%/! &/*&#(1) +"-/) 0 (.'#3(1/*-/6#&%()/& 0*&" 1)2%/,#"/-&. ,/ 1)6%/"/"/)! *,/*&)H3 1+*&//"/,I .#7/,"/)! *)/)!/"0 ".'/&&/"&%(*&% )/6%/"/&%/ 6/"/$#3/*&%/H& "*'/&6//*

PAGE 91

=8 ('1/?B .!("#*$#"/-&.'#3(1/*-//()+"/./*& !!" (-%/) "* 1+*&//"/, ,/18 ,/1 ,/18 ,/1 $*#/ *01#-&2<<2
PAGE 92

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n 6&%(&&%/. )&3(1#,./()+"/ 0(.'#3(1/*-/%() '//*)/1/-&/,5&%#)#*,#"/-& ./()+"/.!"#)/, 0! )#/(*,*/$(//3(1+(&# *)#)+)/,0 "&%/"/.(#*,/" 0&%/ )&+,2%#))/-&# *!"/)/*&)0+11)!/-#0#/,. ,/1) 0(.'#3(1/*-/+)#*$&%#)./()+"/2 %/. ,/1)!"/)/*&/,(' 3/,#,* &#*-1+,/H$/*/"(1 1#-!"/0/"/*-/)I()(* /7!1(*(& "3("#('1/'/-(+)/&%(&#*,#-(& "6()+)/ ,& *)&"+-&&%/&6 ,#"/-&

PAGE 93

=A ('1/?B@ .!("#*$*,#"/-&.'#3(1/*-//()+"/./* &!!" (-%/) *,#"/-&Bn/)#,+(1) *,#"/-&B )O/$ ,/18 ,/1 ,/18 ,/1 $*#/ *01#-&2?<2882A8Y82YE2AGE2AGE2:GE2:G $*#/B00/-/ *01#-&2@BBA2?8YBBE2AAGBBE2A?GBB00/-/ *01#-&BB2ABBA2:YBBE2=GBBE2AGr$(1#&("#(*#)..! "&(*-/B28B28B2A
PAGE 94

=? $r *,#"/-&E )O/$G -#(1/10("/.'#3(1/*-/P(Q Z8 $*#/ *01#-&QZr$(1#&("#(*#)..! "&(*-/QZA*,#3#,+(1#)..! "&(*-/QZ?n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n /$(1#&("#(*#).(1) /7%#'#&/,%#$%/"1/3/1) 0 (.'#3(1/*-/5 2%(&#)%(",1)+"!"#)#*$5 0+")/5#*1#$%& 0 !"/3# +) "/)/("-% *(.'#3(1/*-/#* &%/"! 1#-("/()E#*-1 +,#*$(' "&# *5$("#$%&)5(*,"(-/G2 /(1) )//0" .('1/?B<&%(&3(1+/#.! "&(*-/(+*&)0 "(! "&# 0&%/ 3("#(&# *#*) -#(16/10("/(.'#3(1/*-/n/)! *,/*& )6% "/$(",/,/#&%/"/$(1#&("#(*#). "#*,#3#,+(1#).()#.! "&(*&6/"/1/))1#4/1& /(.'#3(1/*&56%#-%)+$$/)&)&%/ #.! "&(*-/ */(&&(-%/)& "/3(1+/)-(*) ./&#./ )'1 -4 +&*01#-&#*$0//1#*$) (' +&) -#(16/10("/2%#)#)/)!/-#(11) (. *$& % )/6% )"/%#$%#*"/1(/ #.! "&(*-/E"(&#*$ */ 0&%/3(1+/)(' 3/&%/ &%/ "GF&%(&#)5(1("$/" n '/&6//* &%/#.! "&(*-/(&&(-%/,& #*,#3#,+(1#).(*,/$(1#& ("#(*#).#)*/$(/1(*, )#$*#0#-(*&1()) -#(&/,6#&%(.'#3(1/*-/5/3/*6%/ **&" 11#*$0 "&%/1/3/1) 0 #.! "&(*-/(-",/,& &%/3(1+/)&%/.)/13/)2%+)5 6%#1/3(1+/*01#-&* ".(11&/*,)

PAGE 95

= ('1/?B<+1("#(&/ ,/1) 0 -#(1/10("/.' #3(1/*-/ /00#-#/*& &(*,(",r"" :T *0#,/*-/*&/"3(1) # $*#/ *01#-&82A82:2
PAGE 96

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

PAGE 97

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rnrn G5*01#-&#*$ 0//1#*$)E rn G5 "'/1#/0)#**01#-&6#&%0//1#*$)E rn rn G2*(,,#&# *3(1+/#.! "&(*-/5! 1#-!"/0/"/*-/) 5(*,"(-/(11)%(!/) -#(1 6/10("/(.'#3(1/*-/2 0&/")/&&1#*$ *&%/'/)&./()+"/5&%/*/7&-%(!&/ ""/3#)#&)!"/3# +)"/)/("-%&%(& (,,"/))/)&%/>+/)&# 06% #). )&1#4/1& '/ (.'#3(1/*&(. *$1#'/"(1)(*, *)/"3(/)5(. *$./*(*,6 ./*5(*,(-" ))"(-/ 2%/0#*,#*$)!"/)/*&/,&%/"/ -%(11/*$/!"/3# +)())/"&# *)&%(&1#'/"(1)("/. /1#4/1& '/(.'#3(1/*&(' +&) -#(1 6/10("/E/1,.(*(*,9(11/"8::F)//(1) (' G2%/"/)+1&)#*,#-(&/&%(&)/10B

PAGE 98

== #,/*�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

PAGE 99

=: rn nrrrrnrrrnr 5 rrn5nr %/*6/&(14(' +&1#'/"(1)(*,*)/"3(/)#*&% /*#&/,&(&/)5 r r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rn 5 "6%(&&%#))&+,"/0/")& () rnrn 2%/*&/*,&%(& *)/"3(/)!"# "#&#;/#*,#3#,+(1#)& 3/"/$(1#&( "#(*3(1+/)6%#1/1#'/"(1)!1(-/" +$%1 />+(1#.! "&(*-/ */(-%2)("/)+1&5&%/3(1+/) 0(1#'/"(1("/. "/1#4/1& ./#*&

PAGE 100

: *01#-&5)&#.+1(&#*$(.'#3(1/*-/(' +&) -#(16/10 ("/2%/)/("$+./*&)("/*&"("& 0#*,#*$)!"/)/*&/,%/"/2%/("$+./*&#)/7&/*,/,% /"/& #*-1+,/,#00/"/*-/)*-/"*#*$ 0//1#*$)(' +&&%/!/"-/#3/,'/*/0#-#("#/)E rn (*, rn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rnrn G5(00/-/#&/.))+-%()! )#/(*, */$(/0//1#*$)(' +&&%/!/"-/#3/,'/*/0#-#("#/) E rn G5(*,*01#-&

PAGE 101

:8 '/&6//*3(1+/)(*,0//1#*$)E rnrn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

PAGE 102

: r "40n4n5.n n ,n nn ("#*$ En 4/(-%8:+(1#.! "&(*-/ *3(1+/) 0 />+(1#&(*,1#'/"&&%(*("/ *)/"3(/)2#)0#*,#*$)(1) #*,#-(&/&%(&)/10 B#,/*�#/,1#'/"(1)$/*/"(11!"# "#&#;/ /$(1#&("#(*#). 3/"1#'/"&E(4#*& #*,#3#,+(1#).G 6%#1/0 "*)/"3(/)#&#)&%/ "/3/")/2 +-%&%/)(./-(*'/)(#,*-/"*#*$$/*,/"2/(&% /"E?G0 +*,&%(&6 ./* )"/,%#$%/" *+*#3/")(1#).(*,'/*/3 1/*-/&%(* ,#,./*2r$(1#&("#(*#).0#&)+*,/" &%/+.'"/11( 0+*#3/")(1#).2 *3/")/15./*("/ $/*/"(111/))/$(1#&("#(*(*,. "/ #*,#3#,+(1#)&2('EG*0#".)&%/)/0#*,#* $)E)//(1) (&1(*,8::=F ""#)

PAGE 103

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

PAGE 104

:? (' +&) -#(16/10("/! 1#-5(*,&%/"/("/* "/(1 ,#00/"/*-/)#**01#-&(-" ))#,/ 1 $5 $/*,/"5(*,"(-/5&%#)#)(*+*1#4/1) +"-/ 0,#0 0/"/*-/)2 n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n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

PAGE 105

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n1n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

PAGE 106

:@ )#,/ 0&%/.#7/,0//1#*$)*-/"*#*$"(-/2$(#*5 6/)% +1,* &/7!/-&,#00/"/*-/) (-" ))/(-%$" +! 0&%/)(./.($*#&+,/()(00/- /*01#-&,+/& &%/*)#)&/*0 3(1+/!"# "#&#/)(-" ))&%/)/$" +!)2%#)*)#)&/ *-'(1(*-/)&%/! &/*&#(1*01#-& +& 0 ". )&2 $*#/B(00/-/*01#-&#)-/"&(#*1 (!1(+)#'1/&%/ "/&#-(1/7!1(*(&# 0 (.'#3(1/*-/E(3#*/ 8::=G'+&#&#)!" '('1* &()$ ,(*/7!1(*(&# *() (00/-/*01#-&6%/*#&./)& 6%1#'/"(1)6 +1,'/1/))(.'#3(1/*&&%(* *)/"3(/)2 5 ,nnnn r "40n4n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

PAGE 107

:< ('1/B8"/3(1/*-/ 0 -#(1/10("/.'#3(1/*-/ (-" )),/ 1 $ $nn+nn #'/"(1) *)/"3(/) B(1+/ #'/"(1) *)/"3(/) &(*,(", 0#3#*$2@<822?=2=T@?2T ./1/))2<:2:28<2T@2=T '+("(*&//2<:2:2+(13("#(*-/) ())+./,G2H/"-/*&.'#3(1/*&I/*&"#/)#*&%/&('1 /#*,#-(&/&%/!" "&# *6% %(3/ (.'#3(1/*-/)"/)$"/(&/"&%(*;/" 2/"-/*&($/)0 "&%/
PAGE 108

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
PAGE 109

:: ('1/B"/3(1/*-/ 0 -#(1/10("/.'#3(1/*-/ (-" ))/*,/"(*,n(-/ nnnn /.(1/) (1/) B(1+/ /.(1/) (1/) /,#-(1*)+"(*-/2:2:28A2=T?28T ./1/))2:88282A8=2@T@2@Tr,+-(&# *22?@2?2+(13("#(*-/) ())+./,G2H/"-/*&.'#3(1/*&I/*&"#/)#*&%/&('1 /#*,#-(&/&%/!" "&# *6% %(3/ (.'#3(1/*-/)"/)$"/(&/"&%(*;/" 2/"-/*&($/)0 "&%/
PAGE 110

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

PAGE 111

88 ('1/BA#00/"/*-/)(-" )),/ 1 $ +"-/) 0 .'#3(1/*-/(' +& -#(1/10("/ #'/"(1) *)/"3(/) B(1+/ (rn& r$(1#&("#(*#).EB8G2<:2@@2*,#3#,+(1#).EB8G2A<2?@2r$(1#&("#(*#)..! "&(*-/E8B?GA2A82=2*,#3#,+(1#)..! "&(*-/E8B?G2:=A2:2n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

PAGE 112

8 ('1/B?#00/"/*-/)(-" ))/*,/"(*,n(-/ +"/) 0.'#3(1/*-/(' +& -#(1 /10("/ /.(1/(1/ B3(1+/ 1(-4 %#&/ B3(1+/ $*#/&/.) r$(1#&("#(*#).EB8G2<2@@282<<2@:2?*,#3#,+(1#).EB8G2?82?@2<2?2??2A8r$(1#&("#(*#)..! "&(*-/E8B?GA2@2:82A2: 2:28 *,#3#,+(1#)..! "&(*-/E8B?GA2A2<2A2<= A2?2 n/1(/#00/"/*-/EBAG2:82:2:?2:=2: 2@ $*#/ *01#-&EB8G2?82??28A2?2?A 2== 00/-/&/.) )#/" +!00/-&EB8G2A82AA2<22 A?2 /$(/" +!00/-&EB8G2A2?2282 ?2 00/-/ *01#-&EB8G2A2A82<822A 28 $*#/B00/-/ *01#-&EB8G2A@2A< 2?2A2A<28A +.'/" 0()/)A@?<88 -. (&(("/0" .(1 "#,( &/")+"3/ 0"/$#)&/"/, 3 &/")*,+-&/,#*(?2B 3(1+/"/!"/)/*&)&%/!" '('#1#&&%(&6/-(** &"/D /-&&%/*+11%! &%/)#)&%(&&%/"/#)* ,#00/"/*-/(. *$1#'/"(1)(*,*)/"3(/)(-" )) &%/)/#&/.)EB&/)&)B/>+(13("#(*-/) ())+./,G2 2<="/)!/-/1G2%//3#,/*-/(1) #*,#-(&/)&%( &'1(-4)3(1+/)&"+-&+"/)%(3/($"/(&/" "(*$/2%/,#00/"/*-/'/&6//*'1(-4)/$(1#&("#(* (*,#*,#3#,+(1#)&3(1+/)#).+-% $"/(&/"&%(*#&#)0 "6%#&/)2 *-/"*#*$&%/(00//#&/.)5)+"!"#)#*$15'1(-4)%(3/ 0/6/"! )#/0//1#*$)(' +&'1(-4)(*,! "!/ !1 /&%(*6%#&/)E2(*,2A? "/)!/-/1G2*&%/ &%/"%(*,5&%/, %(3/0/6/ "*/$(/0//1#*$)&%(*6%#&/)E28 (*,2?"/)!/-/1G2)"/)+1&5'1(-4)/7%#'#&1 /))(00/-/*01#-&&%(*6%#&/)E2 (*,2A"/)!/-/1G2%/(1) /7%#'#&)1#$%&1 1/))$*#/B(00/-/*01#-&E!P 28AG2$(#*5&%/)/,#00/"/*-/), &,/0#*#/1 &/11+)6%&%/"/("/ ')/"3('1/ ,#00/"/*-/)#*(.'#3(1/*-/(-" ))"(-/5'+&&%/, !" 3#,/%#*&)()& 6%(&6/)% +1, /7!/-&#*(.+1("#(&/. ,/1 0&%/,#00/"/*-/)# *(.'#3(1/*-/2

PAGE 113

8A ,nr "40 n4n5 %/"/)+1&) 0&%/&/)& 0&%/1#4/1) +"-/ 0&%/ ,#00/"/*-/#*(.'#3(1/*-/(-" )) #,/ 1 $("/#*('1/B2 %/1/*,)+!! "&& &%/*&/*&# *&%(&(00/-/ *01#-&5 6%/**&" 11#*$0 "$*#/*01#-&(*,3(1+/# .! "&(*-/5#)&%/'/)&/7!1(*(&# 0 6%*)/"3(/)("/. "/(.'#3(1/*&&%(*1#'/"(1) (' +&) -#(16/10("/5(*,&%(& $*#/B(00/-/*01#-&/7!1(#*)) ./ 0&%/3 ("#(*-/2%/&/)&#*3 13/).!("#*$ "/)+1&)(-" ))()/"#/) 0 ",/"/,1 $#&. ,/1) 0 ) -#(16/10("/(.'#3(1/*-/20$*#/5 (00/-/5 "$*#/B(00/-/*01#-&6/"/&%/ /7!1(*(&# 0&%/,#00/"/*-/#* (.'#3(1/*-/(-" ))#,/ 1 $56/6 +1,/7!/-&#,/ 1 $& '/./(*+*"/1#('1/ !"/,#-& 0(.'#3(1/*-/(' +&) -#(16/10("/6%/* &%//7!1(*(&# 0&%/3("#(*-/ (-" ))#,/ 1 $#)#*&" ,+-/,& &%/. ,/12%/0#" )&. ,/1E ,/1G#*-1+,/) *1 #,/ 1 $()(*/7!1(*(& "3("#('1//7-1+,#*$*)/ "3(/)()&%/"/0/"/*-/-(&/$ "2 ,/1(,,)&%/*&" 13("#('1/)$/*/"(1! 1#!"/0/"/*-/)5#*./5(*,! 1#&#-(1 4* 61/,$/2%#)#)0 11 6/,'#*&" ,+-#*$&%/$* #/*01#-&#&/.E ,/1G(*, &%/3(1+/#.! "&(*-/#*,#-(& ")5&%/$*#/B(00/ -/*01#-&#&/.)E ,/1G5(*, &%/*&%/(00/-/*01#-&#&/.E ,/1G2%/(00 /-/(*,$*#/B(00/-/*01#-& ./()+"/)("/($(#*#*)/!("(&/. ,/1)& (3 #,.+1& #11#*/("#&2%#)!" -/))(11 6)+) & ,/&/".#*/#0&%/)/) +"-/) 0*01#-&5(*,6%#% */)5/7!1(#*&%/3("#(*-/(-" )) #,/ 1 $5$/*,/"5(*,"(-/2 #")&5#* ,/1&%//)&#.(&/ 0&%/"/1(&# *)%#! '/&6//*'/#*$1#'/"(1(*,) -#(1 6/10("/(.'#3(1/*-/#))#$*#0#-(*&(*,*/$(/2/ -(+)/&%/,+..3("#('1/0 *)/"3(/)#)/7-1+,/,()&%/"/0/"/*-/-(&/$ &%#))+$$/)&)&%(&.!("/,& *)/"3(/)1#'/"(1)("/1/))(.'#3(1/*&2%/"/1 #('#1#& 0#,/ 1 $#)*)#)&/*&6%/* &%/*&" 1)("/#*&" ,+-/,& &%/. ,/1)E ,/1 G5) &%/, &(-+*&0 "&%/

PAGE 114

8? ('1/B +"-/ 0#00/"/*-/)#* -#(1/10("/ .'#3(1/*-/(-" )),/ 1 $ ,/1 ,/1 ,/1 ,/1 ,/1 #'/"(1B2?YB2?=YB2A@B2
PAGE 115

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n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

PAGE 116

8@ ('1/B@ +"-/ 0#00/"/*-/)#* -#(1/10("/ .'#3(1/*-/(-" ))/*,/" ,/1 ,/1 ,/1 ,/1 ,/1 /.(1/B2AYB2
PAGE 117

8< ('1/B< +"-/ 0#00/"/*-/)#* -#(1/10("/ .'#3(1/*-/(-" ))n(-/ ,/1 ,/1 ,/1 ,/1 ,/1 0"#-(*./"#-(*B82@YB828=YB8288YB82YB2:?Y E2@GE2@GE2+/)&# *)(' +&&%/())/"&# *&%(&1#'/"(1)("/. /(.'#3(1/*&(' +&) -#(16/10("/&%(* *)/"3(/)2

PAGE 118

8= %/0#*,#*$)!"/)/*&/,#*&%#)-%(!&/" 00/")/3/" (1*-1+)# *)2#")&5)/10B #,/*�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�#/,#,/ 1 $(*,$/*,/"2 (, 0 -+)/) *&%/! 1#&#-(1*)/>+/*-/) 0) -#(16/ 10("/(.'#3(1/*-/2 !/-#0#-(115&%/"/1(&# *)%#!'/&6//*(.'#3(1/*-/ (' +&) -#(16/10("/5! 1#!"/0/"/*-/"/$(",#*$) -#(16/10("/5(*,/3(1+(&# ) 0/ "$/2+)%5&%/22 *$"/))5(*,&%/+!"/./ +"&#)/7!1 "/,2%/0# *,#*$)#*,#-(&/&%(&(.'#3(1/*-/ 6/(4/*)&%/"/1(&# *)%#!'/&6//*! 1#-!"/0/"/*-/) (*,/3(1+(&# *) 0+)%'+&, & ) 6%/*#&./)& /3(1+(&#*$ *$"/))(*,&%/ +!"/./ +"&2

PAGE 119

8: rn@ rrrrrrrrr nr rr %#1/"/)/("-%%()/7&/*)#3/1/7!1 "/,&%/ rnn 0(&&"#'+&/) 0(&&#&+,/) )+-%() rn E" )*#-4(*,'/1) *8::F" )*#-48:==(58:=='F *#*$/" 8::G5 E'/1) *8::F" )*#-4(*,'/1) *8::F" )*#-4 8::AG5 n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rnn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

PAGE 120

88 (.'#3(1/*-/ *&%/"/1(&# *)%#!'/&6//*! 1#-!"/0 /"/*-/)(*,! 1#&#-(13(1+(&# *)("/ "/)&(&/,2 +"&%5()/"#/) 0. ,/1)("//)&#.(&/, & &/)&&%/)//7!/-&(&# *)2%/)/ 0#*,#*$)("/,#)-+))/,'/0 "/. 3#*$ *& &%/*1+,#*$-%(!&/" 0&%#))&+,2 .nn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rn "/0/")& &%/,/$"//& 6%#-%!/ !1/-("/,//!1( +&(*#))+/ (*,#&))#$*#0#-(*-/& &%/#",(#11#3/)E" )*#-4 (*,'/1) *8::F" )*#-48:==(5 8:=='F *#*$/" 8::G2 ** &/),#)&(*-/0" .&%/.#,! #*& &"(,#&# *(1(&&#&+,/)-(1/)E'/1) *8::F" )*#-4 (*,'/1) *8::F" )*#-4 8::AG2&&#&+,/ n "/0/")& &%/,/$"//& 6%#-%(*#*,#3#,+(1#)* 0#,/*&&%(&%#) "%/"(&&#&+,/& 6(",(* 'D/-&#)""/-&E+,,8 :=@F" )*#-4(*,-%+.(*8:==G2 #)&%/)&"/*$&% 0&%/ 'D/-&/3(1+(&# *1#*4#*. /. "E(;# 8:=@F(;# (*,#11#(.)8::G2

PAGE 121

888 n/)/("-%%())+$$/)&/,&%(&&%/)/(&&"#'+&/).(%( 3/)/3/"(1*)/>+/*-/)2 #*)&(*-/5&%/*)#)&/*0(&&#&+,/"/! "&) 3/" &#./#)! )#/1"/1(&/,& #.! "&(*-/ E" )*#-48:=='F-%+.(*(*,"/))/"8:=8G(*,-/"& (#*&E/1%(.8::8G2n/)#)&(*-/& -%(*$/#)! )#/1"/1(&/,& /7&"/.#&Er6#*$8: ?F)$ ,5+-#(*,(**/*'+(. 8:+/*-/) 0(.'#3(1/*-/2%#)1#& /"(&+"/#)* &3/"/,#*#&)/*&#"/& %/"/E)//%(!&/"0 "(. "/&% +$%"/3#/6G2n( &%/"5 *1&%/1#&/"(&+"/&%(&0 -+)/) **)/>+/*-/) 0(.'#3(1/*-/&%(&-(*'//7!1 "/, +)#*$&%/!"/)/*&,(&(("/"/3#)#&/,2 %#1/)&+,#/)%(3/0 -+)/, *(.'#3(1/*-/(' +&) #(16/10("/E/1,.(*(*,9(11/" 8::F&//*'/"$/*(*,"/6/"F('G5&% /%(3/* &1 4/,(&#&) *)/>+/*-/)2"/3# +))&+,#/))+$$/)&&%(&(.'#3(1 /*&(&&#&+,/)/7/"&1/))#.!(-& /3(1+(&# *) 0! 1#&#-(11/(,/")&%(*, *B(.'#3( 1/*&(&&#&+,/)E-"(65()/-4/5(*, *$/"AG2n/)/("-%%()(1) )+$$/)&/,&%(&&%/ /00/-&(!/") *K)! 1#-!"/0/"/*-/)

PAGE 122

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

PAGE 123

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nn ,/&/".#*/6%/&%/") -#(16/10("/(.'#3(1/*-/(00 /-&)&%/"/1(&# *)%#!'/&6//*( !/") *K)! 1#-!"/0/"/*-/)(*,%#) "%/"/3(1+(&# *) 0! 1#&#-(11/(,/")(*,#*)&#&+&# *)5 )/3/"(1. ,/1)("/'+#1&&%(&/)&#.(&/&%//00/-&) 0) -#(16/10("/! 1#-!"/0/"/*-/)5 ) -#(16/10("/(.'#3(1/*-/5(*,&%/#*&/"(-&# *'/& 6//*&%/. */3(1+(&# *) 0/ "$/ 2+)%5&%/22 *$"/))5(*,&%/+!"/./ +"&2 11 0&%/3("#('1/)("/,/,#*&%/

PAGE 124

88? #*&+#/,#"/-&# *2#$%/"3(1+/)"/!"/)/*&'/&&/" /3(1+(&# *)5. "/(.'#3(1/*-/5(*, )+!! "&0 "#*-"/()/,) -#(16/10("/)!/*,#*$2*( 11 0&%/)/. ,/1)5()#$*#0#-(*& )#/#*&/"(-&# *&/".E! 1#-!"/0/"/*-/Y(.'# 3(1/*-/G#*,#-(&/)&%(&(.'#3(1/*-/ 'n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
PAGE 125

88 +!"/./ +"&2$(#*5#0! 1#-!"/0/"/*-/)("/* & "/1(&/,& &%/)//3(1+(&# *)6/ )% +1,* &/7!/-&(.'#3(1/*-/& '/#.! "&(*&2('1/@B8 ""/1(&# *(&"#7 0 -#(1/10("/ 1#-"/0/"/*-/)(*,r3(1+(&# *) 0 +)%5 *$"/))5(*,&%/+!"/./ +"& 1#"/0/"/*-/ +)%r3(1+(&# *$"/))r3(1+(&# +!"/./ +"&r3(1+(&# 1#-"/0/"/*-/82BBBBBB+)%r3(1+(&# *B2Y82BBBB *$"/))r3(1+(&# *B2@2A=Y82BB+!"/./ +"&r3(1+(&# *228=2AY82-. (&(("/0" .? rr )+"3/2('1//*&"#/)("//*,(11)(+'""/1(&# */00#-#/*&)2B&(#1/,&/)&Y!V22 #))#*$3(1+/)6/"/"/!1(-/,+)#*$ .+1&#!1/#.!+&(&# *2 n/)+1&)0 "&%/0#")&)/& 0. ,/1)("/#*('1/@B 5(*,&%/*0#".&%#) /7!/-&(&# *2%/"/#)* &("/1(&# *)%#!'/&6//*(. '#3(1/*-/5! 1#-!"/0/"/*-/)5(*, /3(1+(&# *) 0 *$"/)) "/3(1+(&# *) 0&%/+!"/ ./ +"&5'+&&%/"/#)()#$*#0#-(*& ./,#(&#*$"/1(&# *)%#!#*&%/. ,/1!"/,#-&#*$/3(1 +(&# *) 0+)%2%/!+;;1#*$&%#*$#) &%(&#&#)* &#*&%//7!/-&/,,#"/-&# *2%/*/$(& #3//00#-#/*&#*,#-(&/)&%(&&%/ "/1(&# *)%#!'/&6//*! 1#-!"/0/"/*-/(*,&%/)//3 (1+(&# *)#))&" *$/"0 "&% )/6% ("/ (.'#3(1/*&2%#)0#*,#*$#)*&"("& !"/3# +)0# *,#*$)0 -+)/, *&%/./,#(&#*$/00/-& 0(.'#3(1/*-/E)//"(#$ 2G2%/ "/&#-(115&%/"/#)* "/() *0 "&%#) 0#*,#*$2 /"%(!)5#&#)(*(* .(15 "&%/"/.('/) ./&%#* $!("&#-+1("(*,+*#>+/(' +&&%/ "/1(&# *)%#!'/&6//*) -#(16/10("/(.'#3(1/*-/(*, /3(1+(&# *) 0+)%2 *(>+/)&& +*,/")&(*,&%#)"/)+1&53("# +)*& 13("#('1/)6/"/#*&" ,+-/,& )//% 6&%/(1&/"/,&%/"/1(&# *)%#!2%/"/(11# *&/"/)&#*$&%#*$%/"/#)6%#&)//.)& 3("0 "/3(1+(&# *) 0+)%2&)//.)&%(&!("&# ,/*�#-(&# *#))+-%()&" *$!"/,#-& 0/3(1+(&# *) 0+)%&%(&#&) (4)+!) .+-% 0 &%/3("#(*-/#*&%/!"/,#-/. ,/1) &%(&&%/).(11./,#(&#*$/00/-&(.'#3(1/*-/(!!/(") & %(3/ *&%/#*01+/*-/ 0! 1#!"/0/"/*-/)3(*#)%/)2%#)0#*,#*$)+$$/)&)&%(& !nr E(.!'/11 2

PAGE 126

88@ ('1/@B%//,#(&#*$r00/-& 0 -#(1/10("/. '#3(1/*-/ *r3(1+(&# *) / "$/2+)% 22 *$"/)) +!"/./ +"& 1#-"/0/"/*-/B2A8YB2=2? E2GE2GE2G -#(1/10("/.'#3(1/*-/282828 E28GE28GE28G 1#-"/0/"/*-/Y -#(1/10("/.'#3(1/*-/ B28YE2G B2E2G B2E2G B1 $1#4/1#% ,::2A@88@:2:88:@28 ($/14/"4/)/+, n2:282 @<@<@< -. (&(("/0" .? rr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�#/, *)/"3(/)2)(.'#3(1/*-/ #*-"/()/)5&%/1#*4'/&6//*! 1#-!"/0/"/*-/)(*, /3(1+(&# *) 0+)%#)6/(4/"0 *)/"3(/)() !! )/,& 1#'/"(1)2 8:@G$ &#&"#$%&#*&%/#"0(./, nnr 6%/*&%/)+$$/)&/,&%(&! 1#!"/0/"/*-/))&/.0" .!("&#,/*�#-(&# *(*,&%(& &%/#"/00/-& *3 &/-% #-/#),"#3/*' !("&#,/*�#-(&# *2 &/-% #-/(*,-(*,#,(&//3( 1+(&# *)("/-/"&(#*1"/1(&/,5(*,#& (!!/(")&%(&!("&#,/*�#-(&# *#)(1) ,"#3#*$& %/"/1(&# *)%#!'/&6//*! 1#!"/0/"/*-/)5(.'#3(1/*-/(' +&) -#(16/10("/5(*, /3(1+(&# *) 0/ "$/2+)%2

PAGE 127

88< ('1/@BA%//,#(&#*$r00/-& 0 -#(1/10("/. '#3(1/*-/(-" )),/ 1 $ /00#-#/*& 2r2 :T *0#,/*-/*&/"3(1 1#-"/0/"/*-/B2A2B2?B2 -#(1/10("/.'#3(1/*-/B228B22+..0 *)/"3(/82A?22:@82<+..0 ,/"(&/2A828B282< 1#-"/0/"/*-/Y -#(1/10("/.'#3(1/*-/ B282B2B28 1#-"/0/"/*-/Y -#(1/10("/.'#3(1/*-/Y *)/"3(/ 282822 B1 $1#4/1#% ,8A=2?@ ($/14/"4/)/+, n2? @< -. (&(("/0" .? rr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

PAGE 128

88= ('1/@B?%//,#(&#*$r00/-& 0 -#(1/10("/. '#3(1/*-/(-" ))/*,/" /00#-#/*& 2r2 :T *0#,/*-/*&/"3(1 1#-"/0/"/*-/B2A82B2?8B2 -#(1/10("/.'#3(1/*-/228B282+..0 "(1/28?28AB282?8 1#-"/0/"/*-/Y -#(1/10("/.'#3(1/*-/ B282B28 2 1#-"/0/"/*-/Y -#(1/10("/.'#3(1/*-/Y(1/ B22B282 B1 $1#4/1#% ,88A2?8 ($/14/"4/)/+, n28 @< -. (&(("/0" .? rr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

PAGE 129

88: ('1/@B%//,#(&#*$r00/-& 0 -#(1/10("/. '#3(1/*-/(-" ))n(-/ /00#-#/*& 2r2 :T *0#,/*-/*&/"3(1 1#-"/0/"/*-/B2:2B2A:B2 -#(1/10("/.'#3(1/*-/228B228+..0 "%#&/2@@2A28828 1#-"/0/"/*-/Y -#(1/10("/.'#3(1/*-/ 2828B282A 1#-"/0/"/*-/Y -#(1/10("/.'#3(1/*-/Y%#&/ B228B2?2 B1 $1#4/1#% ,8@A2? ($/14/"4/)/+, n288 @< -. (&(("/0" .? rr )+"3/2('1//*&"#/)("/ ",/"/,B1 $#&/)&#.(&/)2 #))#*$3(1+/)6/"/"/!1(-/,+)#*$.+1&#!1/#.!+&(& # *2 E/3(1+(&# *) 0+)%GF#&#))&" *$/)&0 "*)/"3( /)(*,6%#&/)() !! )/,& 1#'/"(1) (*,'1(-4)2 %/*/7&(*,0#*(1-%(!&/"#*&%#))&+,)+..("#;/) (11 0&%/0#*,#*$)*&(#*/,#* &%#))&+,2%/)/0#*,#*$)("/,#)-+))/,#*1#$%& 0% 6&%/#*0 ". +"+*,/")&(*,#*$ 0 (.'#3(1/*-/5"/(16 "1,!"(-&#&# */")5(*, +"+*,/ ")&(*,#*$ 0! 1#&#-)#*&%/*#&/, &(&/)& ,(2

PAGE 130

8 rn< rnrrrr5rnr r5 rrn %#))&+,%()'//* "$(*#;/,(" +*,&%/ rnrn 5 n 5 n 5 r (*, rnn 0(.'#3(1/*-/(' +&) -#(16/10("/! 1#-2 -",#*$15&%#)*-1+,#*$-%(!&/"#)(1) ) ("" (*$/,2%/>+/)&# *)(,,"/))/,%/"/ 6/"/()0 11 6)8G. *$&%/!"/3# +)(!!" (-%/)& ./()+"#*$(.'#3(1/*-/+)#*$ )+"3/#*,#-(& ")56%#-%%()&%/%#$%/)&3(1#,#&J G 6!"/3(1/*&#)(.'#3(1/*-/ )+"" +*,#*$) -#(16/10("/! 1#-(*,, /)&%/!"/3 (1/*-/3("'$" +!JAG%(&("/&%/ &/*&#(1) +"-/) 0)+-%(.'#3(1/*-/J?G% #). "/1#4/1& '/(.'#3(1/*&(*,6%J G%(&#)&%/"/1(&# *)%#!'/&6//*(.'#3(1/*-/(' +&) -#(16/10("/5! 1#-!"/0/"/*-/)5 (*,/3(1+(&# *) 0! 1#&#-(1-(*,#,(&/)(*,#*)&#&+ &# *)J%/-%(!&/")+..("#;/)&%/ 0#*,#*$)"/$(",#*$&%/()) -#(&/,%! &%/)/)! )#&/ ,#*%(!&/"20&/",#)-+))#*$&%/ 0#*,#*$) 0/(-%"/)!/-/15! &/*&#(1D+.!#*$ 0 0! #*&)(*,0+&+"/>+/)&# *)("/ 00/"/,2%/*5&%#)-%(!&/"(*,)&+,*-1+,/)6#& %) ./0#*(1../*&)"/$(",#*$% 6 &%/)/0#*,#*$))!/(4& +"+*,/")&(*,#*$ 0(.'#3( 1/*-/5& "/(16 "1,!"(-&#&# */")5(*, & +"+*,/")&(*,#*$ 0! 1#&#-)#*&%/*#&/,&(& /)& ,(2 nnnn n )/*-/!&+(1#;(&# *) 0(.'#3(1/*-/#*! 1#&#-( 1)-#/*-/"/)/("-%%(3/1/,& 1/))&%(* !&#.(1./()+"/./*&)&"(&/$#/)E%(!&/") 855AG2-",#*$15&%#))&+, 00/"/,(0"/)%(!!" (-%& ./()+"#*$(.'#3(1/*-/( +&) -#(16/10("/! 1#-&%(&#) (,(!&/,0" ./("1#/"6 "4E"(#$5(*/5(*,("&#*/ ;F"(#$ G2

PAGE 131

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nn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

PAGE 132

8 (.'#3(1/*&(' +&) -#(16/10("/! 1#-&%(*.(1/)5 (*, (0"#-(*./"#-(*)("/1/)) (.'#3(1/*&(' +&) -#(16/10("/! 1#-&%(*6%#&/)2 %/!"/3(1/*-/ 0(.'#3(1/*-/ $/*/"(11(*,(-" ))$" +!)! )/)) .(*>+/)&# *) &%(&&%/-(** &(11'/1#)&/,%/"/2 +&%/"/("/(0/6/7(.!1/))&%/!"/3(1/*-/ 0( .'#3(1/*-/*)#)&/*& 3/"&#./J (3/&%/$" +!,#00/"/*-/)'//**)#)&/*& 3/"&#. /J 6, /)!+'1#-,#)+")/(00/-& &%/!"/3(1/*-/ 0(.'#3(1/*-/J*,5, /),#)+")/ %(3/+*#0 "./00/-&) (.'#3(1/*-/(-" ))$" +!)J n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

PAGE 133

8A (-+*&0 "&%/3("#(&# *#*(.'#3(1/*-/6%#1/#&, /)0 ")/10B#,/*�#/,#,/ 1 $(*, $/*,/"2%+)5!("&#(1)+!! "&#)!" 3#,/,& ) %/3("#(&# *#*(.'#3(1/*-/(-" )) $" +!)#)("/)+1& 03("#(&# *#*&%/) +"-/) 0) +-%(.'#3(1/*-/(-" ))$" +!)2 %#))&+,(1) 00/"/,(*/6&%/ "/&#-(1*&"#'+& # *#*,/. *)&"(&#*$&%(&()&%/ !/") *(1#.! "&(*-/!/ !1/!1(-/ *&%/#"3(1+/)#* -"/()/)"/1(/& &%/"3(1+/) (.'#3(1/*-/,/-"/()/)E%(!&/"?G2%#)%()* &'// *!"/3# +)1/.!#"#-(11 ,/. *)&"(&/,6#&%"/$(",)& ) -#(16/10("/ "(* &%/"(&&#&+,/ 'D/-&2%/0#*,#*$)(1) #*,#-(&/&%(&&%/!/") *(1#.! "&(*-/#*,#3#,+(1)! 1(-/ */$(1#&("#(*#).(*, #*,#3#,+(1#).3("#/)(-" ))#,/ 1 $5$/*,/"5(*," (-/'+&&%#), /)* &(-+*&0 "&%/ ,#00/"/*-/)#*(.'#3(1/*-/(-" ))"(-/2*&%/ &%/ "%(*,5&%//3#,/*-/#*%(!&/")A(*, ?, /))+$$/)&&%(&&%#)#)(*#.! "&(*&!"/,#-& 0(.'#3(1/*-/!" 3#,#*$)+!! "&& $ )#*,#3#,+(1)!1(-/. "/#.! "&(*-/ */3(1+/( ) !! )/,& (* &%/"&%/("/1/)) 1#4/1& '/(.'#3(1/*&(' +&) -#(16/10("/2 %/1(-4 0"/)+1&)6#&%"/$(",)& 6%0"#-(*./ "#-(*))//.& '/) H+*B (.'#3(1/*&I#)(3/"#*&/"/)&#*$* *B0#*,#*$2%/ '/)&(*)6/"56#&% +&&%/,(&(5& &%#) >+/)&# *#)&%(&&%/6/10("/-+1&+"/,#00/")(-" )) "(-/2%#1/.(*. "/6%#&/)("/ 6/10("/&%(*'1(-4)5&%/!" "&# 0'1(-4) *6/ 10("/#)!" '('1%#$%/"2%/"/0 "/5#& #). "/1#4/1&%(&'1(-4)%(3//#&%/"!/") *(11' /*/0#&/, "4* 6) ./ */&%(&%() '/*/0#&/,0" .) -#(16/10("/!" $"(.)5(*,()("/ )+1&5.(%(3/. "/! )#/0//1#*$) (' +&)+-%!" $"(.)2/"%(!)5*&" 11#*$0 "6%/&% /" "* & */%()"/-/#3/,!/") *(1 '/*/0#&)0" .) -#(16/10("/ "4* 6) ./ */&%(&% ()6 +1,'/'/*/0#-#(12*B,/!&% #*&/"3#/6)6#&%'1(-4)6 +1,(1) '/+)/0+1()(*# *0 ".(&# *$(&%/"#*$)&"(&/$!"# "& '+#1,#*$)+"3/#*,#-(& ")2

PAGE 134

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nnn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

PAGE 135

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nn n %#))&+,)&("&/,')+$$/)&#*$&%(&!+'1#!#*# *(' +&) -#(16/10("/! 1#-%() '//*,#3#,/,)#*-/&%/(,3/*& 01("$/)-(1/) -#(1 !" $"(.)#*&%//6/(12& *-1+,/)'(,,"/))#*$&%#)>+/)&# *2"/./"#-(* )()/3/*1,#3#,/,(' +&) -#(1 6/10("/()! 11)0" .&%/1()&/(")%(3/)+$$/) &/,J0&%/1 "#,()(.!1/+)/,%/"/#) "/!"/)/*&(/ 0&%/+*&"()(6% 1/5&%/(*)6/ "& &%#)>+/)&# *#)("/) +*,#*$* 2 n(&%/"5.(*./"#-(*)("/5#*0(-&5 n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

PAGE 136

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

PAGE 137

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
PAGE 138

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n/(16 "1,!"(-&#&# */")("/"/1#*$ *.("$#*(1 ,#)&"#'+&# *)&%(&.('/.#)1/(,#*$2 n/$(",1/)) 0&%/! &/*&#(1+)/0+1*/))0 "-(*,#,(& /) 0! 11)&%(&-(!&+"/ (.'#3(1/*-/5&%//3#,/*-/!"/)/*&/,#*&%#))&+,1/("1)+$$/)&)&%(&&"(,#&# *(1 #*,#-(& ")("/.#))#*$(*#.! "&(*&(&&"#'+&/ 0(& &#&+,/)2%/#,/(&%(&!/ !1/6/"/ 6#11#*$& (*)6/"(&/1/!% */(*,&/11) ./' ,&%( &&%/%(, )#/0//1#*$) (*, */$(/0//1#*$)(' +&&%/)(./&%#*$#)D+)&(.( ;#*$2+"&%/"5&%/6/"/

PAGE 139

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r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

PAGE 140

8A &%(&&%/)/("/&%/!/ !1/6% %/1!& ,/-#,/ *"/! "/)/*&(&# **-/"*#*$) -#(16/10("/ 1#-0 "&%/!+'1#-()(6% 1/5(*,0 "/3/"&%#* $/1)/()6/112/ !1/&%(&6(*&&%/#" -(4/(*,& /(&#&& 2

PAGE 141

8A8 nrrnrr '/1) *5n225#*,/"52n25/&/")5225[#) 4/522E8:=G200/-/(*, )/.(*&#-.! */*&)#*! 1#&#-(1!/"-/!&# *2 .rnr+rnn#r +rr 5?5@8:B@A2 D;/*52E8G2(&+"/(*, !/"(&# 0(&&#&+,/) 2 nn/r+rr 55+#3 -(&# (' +&&%/*&/"*(1n/3/*+//"3#-/2 n.rnr+r#n 5?5 ?8=B?2 13("/;5n225["/%.52EG2 1(r0-n20&nrrn0 nn+3nrn 2"#*-/& *5"#*-/& **#3/")#&"/))2 *,/") *5225[ %*) *522EAG2 %/6% (*,6%/* 0H$/*,/"B'1#*,I(&&#&+,/) "/,#-& ") 0$/*,/"B" 1//$(1#&("#(*#).#*&6 ,#0 0/"/*&, .(#*)2 #/r 5?:5
PAGE 142

8A *) 1('/%/"/525*,/"5225"25[&/6("&52 52E8G2%//00/-&) 0!("&(*, !"/0/"/*-/) **$"/))# *(1" 11B-(113 &#*$2 4#5 5@5 AAB<2 ".#&($/5225[ **/"52EG2&&#&+,#*(1(. '#3(1/*-/&/)& 0&%"//4/ %! &%/)/)2 +rnn#r+rr"n 5@58?8B8?A2 ".#&($/5225[ **/"52EG2&&#&+,#*(1( .'#3(1/*-/(*,! 1#&#-(1 !#*# "/3#/65(*,(3/*+/)0 "0+"&%/""/)/("-%2*22 "(#$[22("&#*/;Er,)2G5 n0+r0n++r E!!28?B8@@G2/6 "4(1$"(3/ (-.#11(*2 )%/"52E?G2 +rnn+26-(n#r7 nr 2()%#*$& *5 "/))2 (#1/525["(,522E8::=G2/&/" $/*/#&(* ,"/!"/)/*&(&# *%/)/*(&/(*, 0"//&"(,/2 n.rnr+r#n 5?5?B??2 (",/)5225[1,/*,#-45n22EAG2 +3nrn2$nn $n 2/1. *&5% .) *(,)6 "&%2 ("$%5225%(#4/*525 3/*,/"5n25["(&& 5 2E8::G2%/$/*/"(1#& 0&%/ (+& .(&#-(&&#&+,/(-(&# */00/-&2 .rnr+rnn#r+rr 5 @5=:AB:82 (""/&&5r225[ 4522E8::8G2 *$"/))# *(1 (&&#&+,/)(*,3 &#*$'/%(3# "* /7(.#*(&# 0)+!! "&0 ") -#(16/10("/2 4#5 58@5A+#)#&/& %/(&/,#*&"(!)-%#*01#-&2n/)! *)/&#././()+"/./*&#*)+"3/"/)/( "-% .! )#+.*,+-&/,(& &%/ )889nn$nr&nnrn#rr +r+rr 5 *&"/(15(*(,(2 ())#1#5225[1/&-%/"522E8::8G2n/)! *)/B &#././()+"/./*&#*)+"3/"/)/("-% ./&% ,0 "(*,(*/61 4(&* *(&&#&+,/)2 +3nrn5 55 AA8BA?@2 /**/&&52225[/**/&&2r2E8::G2 4n4n2n(rnr !":rnn 2(6"/*-/5*#3/")#& 0(*)()"/))2

PAGE 143

8AA /**/&&5225[(""/11522E8:
PAGE 144

8A? (.!'/11525 *3/")/52r25#11/"52r25[& 4/)52r2E8:@G2 !n r 2/6 "4 %*#1/[ *)2 (*&"#15225[(*&"#1522E8:::G2 /n$#n2nnn +3nrnr:rnn 2()%#*$& *5 ," 6#1) */*&/" "/))2 ("" 11522E8:==G2 ./*)(+& .(*,&%/$/* ,/"$(!8:=(*,8:=2*2 +/11/"Er,2G5 !+rr:n: E!!2A@B
PAGE 145

8A "(#$5225("&#*/;5225[(*/522EG 2.'#3(1/*-/(*,"/)! *)/ #*)&('#1#&!(*/1)&+,2*22"(#$[22 ("&#*/;Er,)2G5 n n#r+r3nrn E!!2B+/*-/)5(*,""/1(&/) 0(&&#&+,/(--/))#'#1 #&2*n2r2/&&[22 )*#-4Er,)2G5 #n2nnn(rnn E!!2?
PAGE 146

8A@ (;# 5n225 6/115225[/""522E8:=AG2 6(",(!" -/)). ,/1 0&%/ (&&#&+,/B'/%(3# ""/1(&# *--/))#*$ */)(&&#&+, /+! *./"/ ')/"3(&# 0&%/ (&&#&+,/ 'D/-&2 .rnr+rn n#r+rr 5 ??5+ (1#&#/)"/1(&#*$& &%/)&"/*$&% 0&%/ (&&#&+,/S'/%(3# ""/1(&# *)%#!2 .rnr-n#r+rr 5 8?5 A:=S?=2 /(&%/"522E?G2(1+/""/1(&/) 0(.'#3(1/ *&(&&#&+,/)& 6(",$/*,/""/1(&# *)2 +rnn#r+rr"n 5A5AB82 /,/"#522EG2r,+-(&# *5#,/ 1 $5(*,"( -#(1(.'#3(1/*-/ *01#-& (.!1#0#-(&# "*01#-&"/) 1+&# *J*22"(# $[22("&#*/;Er,)2G5 n0+r0n++r E!!2=AB8G2/6 "4(1$"(3/ (-.#11(*2 /1,.(*52E8:==G2&"+-&+"/(*,*)#)&/*-#*!+ '1#!#*# *%/" 1/ 0"/'/1#/0) (*,3(1+/)2 n.rnr+r#n 5A5?8@B??2 /1,.(*525[&//*'/"$/*52n2E8G2%/%+.(* #&("#(*0 +*,(&# *0 "!+'1#)+!! "&0 ") -#(16/10("/2 n.rnr+r#n 5 ?5@=B@<<2 /1,.(*525[9(11/"52E8::G2%/! 1#&#-(1-+1 &+"/ 0(.'#3(1/*-/,/ 1 $#-(1 "/)! *)/)& &%/6/10("/)&(&/2 n.rnr+r#n 5A@5@=BA<2 /)&#*$/"52E8:
PAGE 147

8A< (#* +)525[("&#*/;522EG2%(&%(!!/*) 6%/*6/)#.+1&(*/ +)16(*& !! )#&/&%#*$)J.'#3(1/*-/(' +&) -#(16/10("/2 *22"(#$[22 ("&#*/;Er,)2G5 n0+r0n++r E!!2?+#)&525[-%#-41/"5r2EG2 +n1n$n2+r +n#r&nrr 2/6(3/*(1/*#3/")#&"/))2 "/*,)&(,525[+*,'(-452EAG2 -# B,/. $"( !%#-/00/-&) *-+1&+"(1'#()/) ",#-)&+, 0$"#,B$" +!&%/ "2 #rrr 5?@5=:BA@2 (,, -452EAG2(4#*$(!("&1/(,/"1/)) 0( !("&./.'/"%/#.!(-& 0 (.'#3(1/*-/ *())#.#1(&# *(*,*&"()&/00/-&)#* 1#&#-(1!("&(&&#&+,/)2 +rr+rr 5?5<@:B<=2 /&%/"#*$& *522EG2 6!$2 ;n"r<. 2 (.'"#,$/5("3(",*#3/")#&"/))2 -%)-%#1,522E8::G2 n?rn<2/0(0n #r r=rn 2"#*-/& *"#*-/& **#3/")#&"/))2 ,) *525(# 52n25[r))/)522E8G2%/ 1/ 0(&&#&+,#*(1(.'#3(1/*-/#* )+)-/!&#'#1#&& *)/*)+)#*0 ".(&# *2 "n#r+rr 5A5 8:
PAGE 148

8A= 1'" 45225[" )*#-4522EG2/&(B!) -% 1 $#-(13/")+) !/"(/ ./()+"/) 0(.'#3(1/*-/#00/"/*&#(&#*$&%/*)/> +/*-/) 0!/"-/#3/,#*&"(B !)-%#-*01#-&(*,"/(1#*&"(B!)-%#-*01#-&2 *22"(#$[22("&#*/; Er,)2G5 nn#r+r3nrn E!!2+/2 +rr"n 5<<5A@8BA<2 (&;525[())5n22E8:==G2n(-#(1(.'#3(1/*-/ (*,(./"#-(*3(1+/*01#-& ""/1(&# *(1(*,!"#.#*$)&+,#/) 0,+(1$*#/ )&"+-&+"/)2 .rnr +rnn#r+rr 55=:AB:2 (&;525(-4/*%+&525[())5n22E8:=@G2n(-# (1(.'#3(1/*-/53(1+/,+(1#&5(*, '/%(3# "2*22 3#,# [22(/"&*/"Er,)2G5 +0
PAGE 149

8A: (+0.(**5225[/&" -#452n2E8:::G2%/-%(* $#*$! 1#&#-) 0./"#-(*./* *,/")&(*,#*$&%/) +"-/) 0&%/$/*,/"$(!2 n.rnr+r #n 5?A5=@?B==<2 #*,/"52n25[(*,/")522E8::@G2 <(rr2/+rn +/*-/) 0(. '#3(1/*-/& 6(",!"/)#,/*&#(1 -(*,#,(&/)2 n.rnr+r#n 5?5:8B::2 (3#*/525[&//*'/"$/*52EG2" +!(.'#3(1 /*-/(*,/1/-& "(1,/-#)# .(4#*$2*22"(#$[22("&#*/;Er,)2G5 n0+r0n+ +r E!!28B@ @ /6 "4(1$"(3/(-.#11(*2

PAGE 150

8? (3#*/525% .)/*52259(**(5225[ "$#,( 5r2E8::=G2*&%/!"#.(0(00/-& #*&%/,/&/".#*(&# 0(&&#&+,/)(*,'/%(3# "%/ ,/"(&#*$#*01+/*-/ 0 (00/-/B$*#/(.'#3(1/*-/2 .rnr-n#r+rr 5 A?5 A:=B?82 (.(*5225[(")/522EG2("&! 1("# ;(&# *(*,H*01#-&/7&/*)# *I#*&%/ ./"#-(*/1/-& "(&/2 n.rnr+r#n 5?@5<=@B=2 #&&1/5n2225[n+'#*522E8:=+/*-/) 0/7! )+"/(*,,/$"// 0(,3/ "&#)#*$-1(#.)/7($$/"(&# * *)+./")'/1#/0))&"/*$&%5'/1#/0*0#,/*-/5(*, (&&#&+,/)2 .rnr$'n / 55@@S=82 ("&#*/;5225"(#$5225[(*/522EG 2" )(*,*)(.'#3(1/*-/(*, !+'1#!#*# *2*22"(#$[22("&#*/;Er, )2G5 nn #r+r3nrn E!!28B8?G2/6 "4(1$"(3/(-.#11(*2 ("&#*/;522522"(#$5(*/5225[(#* + )52EG2.'#3(1/*-/(*,3(1+/ *01#-&&/)& 0&6 #))+/)2*22"(#$[2 2("&#*/;Er,)2G5 n0 +r0n++r E!!2@AB=G2/6 "4(1$"(3/(-.#11(*2 (&1(*,5n2r2E8::=G2 ./*)"/!"/)/*&(&# *#**( &# *(11/$#)1(&+"/)/3/1 !/,(*, ,/3/1 !#*$,/. -"(-#/)2 4#5 5A58:B82 -(**522E8::+/*-/)2*22" (#$[22("&#*/;Er,)2G5 nn#r+r3nrn E!!28B8@G2/6 "4 (1$"(3/(-.#11(*2

PAGE 151

8?8 -"(65225()/-4/5r25[ *$/"52EAG2. '#3(1/*-/5+*-/"&(#*&5(*, !" -/))/) 0-(*,#,(&//3(1+(&# *2 +r+rr 5?5?8B??=2 #11("5225 [/))/"52E8:=@G2r00/-&) 0(00/-/(*,$*# /0 -+) *&%/ (&&#&+,/B'/%(3# ""/1(&# *2 .rnr+rnn#r+rr0 85+/*-/) 0.+1&#,#./*)# *(1! 1#&#-(1(&&# &+,/)2*2r2("#)[22 *#,/".(*Er,)2G5 !&r"2-n&nrnr =rn n(n E!!2@B8G2.)&/",(.%/.)&/",(.-% 1 0 ..+*#-(&# *n/)/("-%5*#3/")#&/#&3(*.)&/",(.5 %//&%/"1(*,)2 +11#$(*525[-"(6522EG2(1+/*01#&%//00/-&) 0.!/&#*$ !"#*-#!1/) *! 1#&#-(1D+,$./*&2(!/"!"/)/*&/,( &&%/ AAnn$nr n+r#nrrn 5 )& *52 /1) *52r2E8:::G2" +!(00/-&(*,(&&"#'+&# #*) -#(1! 1#!#*# *2 .rnr +r 5@85AA8BA@2 /1) *52r2[#*,/"52n2E8::@G2))+/0"(./) (*,$" +!B-/*&"#).#*./"#-(* !+'1#!#*# *2 .rnr+r 5=58B8<=2 /6'B1("452n25-"/$ "525[9(**(522E G2%#*4#*$(*,-("#*$(' +& $*#/#**)#)&/*-%/*(*,0 "6% ., /)(&& #&+,#*(1(.'#3(1/*-/0//1 +*.0 "&('1/J .rnr+rnn#r+rr 5=58+/*-/ )%/-/"&(#*&(*,#.! "&(*-/ 0 )/10B4* 61/,$/2 .rnr+rnn#r+rr 5 @58=BA2

PAGE 152

8? /&/")/*525[+&& *52r2E8:
PAGE 153

8?A #*-1(#"522E8:<=G2%/! 1#-*)/>+/*-/) 0 !("&"/(1#$*./*& -#(16/10("/ 1/$#)1(&# *#*&%/ +)/ 0n/!"/)/*&(/)58:AAB8 :?2 n.rnr +r#n 55=AB82 *#,/".(*5225" ,5n225[/&1 -452r2E8 ::8G2 /rnnn(r2 -rrnn+r+rr /6 "4(.'"#,$/*#3/")#&"/))2 *#,/".(*5225[#(;;(52E8::AG2 !#r/ 2(.'"#,$/5("3(", *#3/")#&"/))2 &//*'/"$/*52n25["/6/"52n2EG2%/* & B) B(.'#3(1/*&!+'1#1#(&&#&+,/)#*&%/! 1#&#-(1-+1&+"/ 0(.'#3(1/*-/2 *2r2("#)[22*#,/".(* Er,)2G5 !&r"2-n&nrnr =rnn (n E!!288B8?G2.)&/",(.%/.)&/",(.-% 1 0 ..+*#-(&# n/)/("-%5*#3/")#&/#&3(*.)&/",(.5%//&%/"1(* ,)2 &#.) *5225(-4+/*5225[r"#4) *5n22E8 ::G2*(.#-"/!"/)/*&(&# *2 n+r#n/ 5=:5?AB@2 & */-()%522EG2 (n+nn+r 2 +1,/"5/)&3#/6 "/))2 &+,1("5225[-11#)&/"52EG /)(-"#& #-(1.())/7#)&J.!("(/(*(1)#) 06 ./*K)1/$#)1(/"/!"/)/*&(&# *)#*-/8:2 -rn.rnr+r / 5 ?85AASA2 6(**5225"25[r15n22E8:=?G2'(&&1/ 0 6#11)/10B3/"#0#-(&# *3/")+) '/%(3# "(1*0#".(&# *2 .rnr+rnn#r+rr 5?@58=
PAGE 154

8?? %1(*/"5225"(5225[("-#(522EG 2,/ 1 $5#))+/)5(*,!("&#)(*)%#! (. *$(&#* )2(!/"!"/)/*&/,(&&%/ AAAnn$nr6n +r#nrrn 5(* )/52 /1-%525[#$/1.(*52E8::AG2%/! 1#&#-) 0 #)!(*#-./"#-(*)*)#$%&)0" (&# *(1+"3/)58:=B8:==2 #r#n5 5+(1#&2*22n#../".(*522(1 ,5[2#17Er,)2G5 +rr:/ E!!2?:B?AG%#-($ *#3/")#& 0%#-($ "/))2 #1) *5225#*,)/525[-% 1/"522E G2. ,/1 0,+(1(&&#&+,/)2 +rr/ 5 8<588B8@2 "#$%&5225"25r"#4) *5n225[-3/"52 2E8:=+/)&# *)3/")+)"/3/(1#*$!"/0/"/*-/)2 n.rnr+r#n 5A@5 <:B@8@2

PAGE 155

8? nr *(./#)() *(#* +)26()' "*#*(4/1(*,5 1 "#,(5#*8:<86%/"/ (&&/*,/,1/3/1(*, +"&r1/./*&("5 +&%6/)&+*# "#$%5(*,(4/1(*,#$% -% 120&/"$"(,+(&#*$56/*&& "/3(", ..+*# & 11/$/5"(*$/ ()& 11/$/5 /)( 11/$/5(*"(*-#)&(&/*#3/")#&5(*, *,#(*n#3/" ..+*#& 11/$/ '/0 "/"/-/#3#*$(22#*! 1#&#-(1)-#/*-/0" 1 "#,(&1(*&#-*#3/")#&2%/*6/*& *& $/&.22#*! 1#&#-(1)-#/*-/0" .&%/*# 3/")#& 01 "#,(2%(3/'//*.(""#/, & %/""(#* +)0 "=/(")(*,6/D+)&%(, +"0 #")&,(+$%&/"5/11("((#* +)2


Permanent Link: http://ufdc.ufl.edu/UFE0013118/00001

Material Information

Title: Having Our Cake and Eating It Too: Americans' Ambivalence about Social Welfare
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: UFE0013118:00001

Permanent Link: http://ufdc.ufl.edu/UFE0013118/00001

Material Information

Title: Having Our Cake and Eating It Too: Americans' Ambivalence about Social Welfare
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: UFE0013118:00001


This item has the following downloads:


Full Text












HAVING OUR CAKE AND EATING IT TOO:
AMERICANS' AMBIVALENCE ABOUT SOCIAL WELFARE















By

JASON GAINOUS


A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY

UNIVERSITY OF FLORIDA


2005






























Copyright 2005

by

Jason Gainous
































This document is dedicated to my family.















ACKNOWLEDGMENTS

I thank my parents, grandparents, wife, dissertation committee, and Frank and

Associates for their patience and support.
















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 ................... ............................ .. ix

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

1 INTRODUCTION AND BRIEF OVERVIEW OF STUDY .............. ................ 1

2 ATTITUDINAL AMBIVALENCE: CONCEPTUALIZATION,
MEASUREMENT APPROACHES, PREVALENCE, SOURCES,
CONSEQUENCES, AND SOME HYPOTHESES .............. ....................... .......9

Ambivalence as It Has Been Conceptualized and Measured ....................................10
The Prevalence of Ambivalent Political Attitudes .................................. .............. 17
Potential Sources of A m bivalence................................................... ............... ...... 18
The Consequences of Ambivalent Political Attitudes..............................................24
Ambivalence and Attitudes About Social Welfare................. ............................26
Putting it All Together with Some Hypotheses .................................. ............... 29
C onclu sion ............................................................... ..... ..... .. ..... 34

3 SOCIAL WELFARE AMBIVALENCE: OPERATIONALIZATION AND
D ESCRIPTIVE STA TISTICS......................................................... ............... 35

Direct and Indirect Measures of Ambivalence ......................................................36
Sources of Ambivalence in General and About Social Welfare.............................46
C cognitive C conflict ............................................... .. .. ........ .... ...... ...... 46
Cognitive-A effective Conflict........................................ ........................... 51
A ffective-A effective C conflict ............................... .............. ............... .... 54
Dependent Variables for Consequences of Social Welfare Ambivalence..................56
Other Independent and Control Variables ....................................... ............... 57
Policy Preferences .............. ....................... ..... ........... ......................... 57
Ideology, Party Identification, Race, Gender, and Income ..............................59
P political K now ledge ............................................ .. .. ........ .... ............6 1
S u m m ary ....................... ................ ...... .......................................... 6 3









4 A TEST OF MEASUREMENT VALIDITY: THE SOURCES OF
AMBIVALENCE ABOUT SOCIAL WELFARE ..............................................66

Critique of Previous Approaches to Measuring Ambivalence ..............................69
D direct M measures ................ .................... ........................ .. ............ 69
In direct M easu res ............................................................ ......... ................7 1
A n aly sis ....................................... ......... ..... .................................................. 7 4
Empirically Comparing Measurement Approaches ........................................74
Fully Specified Models of Social Welfare Ambivalence.............. .....................82
Sum m ary ...................................... ................. ................. .......... 87

5 PREVALENCE: SOCIAL WELFARE AMBIVALENCE ACROSS IDEOLOGY,
G E N D E R A N D R A C E ...................................................................... ..................89

Ideology, Gender, and Race: Potential Sources of Social Welfare Ambivalence
D differences ........................................... ........................... 92
C cognitive C conflict .............................. ....... ................. ..... ...... 92
A effective C onflict........... ............................................................. ..... .... .. 94
Cognitive-A effective Conflict........................................ ........................... 95
Results .......... .............. ............ ........... . ........ .............. 96
Differences in the Prevalence of Ambivalence across Ideology, Gender, and
Race.................... ....... .................. ....................... 96
Multivariate Test of Difference across Ideology, Gender, and Race ..............103
Su m m ary ...................................... ....................................................10 8

6 THE CONSEQUENCES: THE MEDIATING EFFECT OF SOCIAL WELFARE
A M B IV A L E N C E ........................................................................ .......................109

Attributes of Attitudes: The Consequences ............................................................. 110
M odel and Test Specification ........................................................ ............. ..113
R e su lts ..........................................................................................1 1 4
S u m m a ry .......................................................................................1 1 8

7 SOCIAL WELFARE AMBIVALENCE: WHAT DO WE KNOW, WHERE DO
WE GO, AND WHAT DOES ALL THIS MEAN FOR POLITICS ....................... 120

Conceptualization and Measurement of Social Welfare Ambivalence ....................120
Prevalence of Social W welfare Am bivalence ............................................................121
The Sources of Social W welfare Ambivalence ............. ................. ....................122
The Consequences of Social Welfare Ambivalence.....................................124
C including C om m ents ........................................... ....................................... 125

LIST OF REFEREN CES ........................................................... .. ............... 131

BIOGRAPHICAL SKETCH ....................................................................... 145
















LIST OF TABLES


Table p

3-1 Distribution of Responses on Two Direct Measures of Ambivalence .................39

3-2 Model of Attitudes about Social Welfare.... .............................. ..............41

3-3 L ogic of the A lgorithm ......... ..................... .................................... ............... 44

3-4 Frequency and Intensity of Ambivalence about Social Welfare............................45

3-5 Distribution of Responses on Value Importance Indicators................................52

3-6 Distribution of Responses on Government Evaluations Indicators ....................58

3-7 Distribution of Ideology, Party Identification, Race, Gender, and Income..............62

4-1 Correlations across Direct and Indirect Measures of Ambivalence......................75

4-2 Correlations between Sources and Four Ambivalence Measures ..........................76

4-3 Analysis of Residuals from Model of Attitudes about Social Welfare....................78

4-4 Correlations between Sources of Ambivalence about Social Welfare.................80

4-5 Comparing Direct Ambivalence Measurement Approaches .................................. 81

4-6 Comparing Indirect Ambivalence Measurement Approaches .................................83

4-7 Multivariate Models of Social Welfare Ambivalence.......................................85

5-1 Prevalence of Social Welfare Ambivalence across Ideology...............................97

5-2 Prevalence of Social Welfare Ambivalence across Gender and Race ...................99

5-3 Differences across Ideology: Sources of Ambivalence about Social Welfare......101

5-4 Differences across Gender and Race: Sources of Ambivalence about Social
W welfare ................................................................................................... 102

5-5 Source of Differences in Social Welfare Ambivalence across Ideology .............104









5-6 Source of Differences in Social Welfare Ambivalence across Gender................ 106

5-7 Source of Differences in Social Welfare Ambivalence across Race ............... 107

6-1 Correlation Matrix of Social Welfare Policy Preferences and Evaluations of
Bush, Congress, and the Supreme Court................. ............. ... ............ 115

6-2 The Mediating Effect of Social Welfare Ambivalence on Evaluations ...............116

6-3 The Mediating Effect of Social Welfare Ambivalence across Ideology ..............117

6-4 The Mediating Effect of Social Welfare Ambivalence across Gender ................18

6-5 The Mediating Effect of Social Welfare Ambivalence across Race ......................119
















LIST OF FIGURES

Figure page

3-1 Distribution of Social Welfare Ambivalence....... ... ........................................46

3-2 D distribution of Cognitive Conflict ........................................ ........ ............... 50

3-3 Distribution of Cognitive-Affective Conflict ............. ..........................................55

3-4 Distribution of Affective-Affective Conflict................................ ................. 56

3-5 Social W welfare Policy Preferences ........................................ ....................... 59

3-6 Political K now ledge D istribution....................................... .......................... 64

4-1 Scatterplots of Overlapping Sources ................... ............. ..... ............... 79















Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy

HAVING OUR CAKE AND EATING IT TOO:
AMERICANS' AMBIVALENCE ABOUT SOCIAL WELFARE

By

Jason Gainous

December 2005

Chair: Stephen C. Craig
Cochair: Michael D. Martinez
Major Department: Political Science

Behavioral research has begun to embrace the idea that people's attitudes are made

up of a range of considerations and are not necessarily bipolar. People can simultaneously

feel positive and negative about a single object. These people are ambivalent. This study

explores Americans' ambivalence when it comes to social welfare policy. Using a

statewide telephone survey, a series of hypotheses are tested. The findings indicate that

the measure of ambivalence developed here is more valid than those previously designed

to measure social welfare ambivalence, ambivalence is fairly common, it is rooted in the

conflict of cognitive and affective elements, the prevalence of ambivalence varies across

ideology, gender, and race, and finally, social welfare ambivalence has only minimal

consequences on voting behavior.














CHAPTER 1
INTRODUCTION AND BRIEF OVERVIEW OF STUDY

Since the founding of the United States, the politics of each era has been defined by

some issue. In the early years debate was centered on the struggle for the ratification of

the Constitution between the Federalists and the anti-Federalists. This struggle evolved

into the battle between northern commercial interests and southern agricultural interests

eventually resulting in the Civil War. Following the Civil War, issues related to

industrialization were the primary focus. Since the New Deal, issues that have involved

the government's economic role in providing for the welfare of its citizens have been the

primary driving force behind politics. For over seven decades, these social welfare policy

issues have created a divide in the United States that has been a defining characteristic of

party politics at both the elite (Sinclair 1978; Barrett and Cook 1991; Ansolabehere et al.

2001) and mass levels (Berelson et al. 1954; Campbell et al. 1960; more recently, see

Green, Palmquist, and Schickler 2002; Stonecash 2000; Layman and Carsey 2002).

This study focuses on social welfare policy. Specifically, the concentration is on

the nature of public opinion surrounding this controversial set of issues. Most people

probably think of social welfare policy as government checks or food stamps, but it really

includes many issues. While checks and food stamps are certainly part of it, social

welfare policy in general can be thought of more broadly. It has been defined as policy

that provides for government programs in an effort to ameliorate social and economic

disadvantages (Sniderman and Piazza 1993). By this definition, assistance for the poor is

not the only type of social welfare policy. Areas such as social security, public education,









government provided health insurance, and government provided jobs could also be

thought of as social welfare. While various programs may fall under the broad umbrella

of "social welfare policy," public opinion on these various issues is certainly not

consistent across dimensions. Simply, many may support one type of welfare program

(e.g., increased spending on public education), but not support another (e.g., food

stamps).

The earliest polls that focused on the public's position on the social welfare policies

associated with the New Deal suggested that there was great support for programs such as

Social Security and wages-and-hours legislation, but not much support for the National

Recovery Administration and the Agricultural Adjustment Administration (Bennett and

Bennett 1990). Using survey data from the American National Election Studies (NES)

and the General Social Survey (GSS), Bardes and Oldendick (2003) explored the

distribution of public opinion across several domains of social welfare policy for the last

50 years. Their results suggested that the public is fairly evenly split on the general

question of whether or not the government should provide more or fewer services.

While the public is evenly split on the general question, as described above, support

does indeed vary significantly according to the nature of the program. For example,

public support is high for Social Security, spending for schools, and many also think that

the government should help people get doctors and hospital care. But on the other hand,

the public is fairly evenly divided on the question of whether this country should have a

government health insurance plan. Programs that seek to ensure that everyone has a good

standard of living (e.g., food stamps) are generally less well supported. Their results

suggest also that the distribution of attitudes surrounding these issues remained stable for









the last 30 years. Whether divided evenly or less so, the public has been split up along

these lines for some time.

The problem with Bardes and Oldendick's (2003) inferences and others that have

used NES or GSS is that they employed survey indicators that fail to account for the

possibility that many are ambivalent, that is, they simultaneously possess positive and

negative evaluations (see Klopfer and Madden 1980; Katz, Wackenhut, and Hass 1986;

Cacioppo and Berntson 1994; Thompson, Zanna, and Griffin 1995; Priester and Petty

1996; Alvarez and Brehm 1995, 1997, 1998; Craig, Kane, and Martinez 2002; McGraw,

Hasecke, and Conger 2003) about social welfare policy. If it is the case that many are

torn internally between the contending sides of governmental activism versus

governmental restraint with regard to such programs, then polls that are standard in the

discipline are not the most accurate representations of public opinion regarding these

issues. Most standard survey indicators used across academia (NES and GSS) force

respondents to choose one side or the other, or choose the moderate or neutral response.

With no response category that captures those who are torn on the issue, many

respondents may simply choose whatever response comes the closest to their attitude.

They may choose a positive or negative response based on what they have most recently

thought about it (see Zaller 1992). So, where do the ambivalent go? The "middle

response" seems to be the safe response for someone for whom the other responses do

not fit. Perhaps those who are ambivalent are choosing the middle response category for

attitudes about social welfare. Evidence suggests that elevated middle response

frequencies are often a result of nonattitudes, where people have no real position on the

issue (Asher 2004), and they choose this response because it is non-committal. It also









seems that the most logical response for one with both positive and negative feelings is

the middle response. If we overestimate true moderate positions on policy, then our

understanding of social welfare issues is faulty.

The present study designed a measure to capture such ambivalence. Among the

many details that will be discussed throughout, the findings show, amazingly enough,

that a significant number of those surveyed said they had extremely positive feelings

about some social welfare issue, and then just a few minutes later, when asked about their

negative feelings toward the same issue, they said they had extremely negative feelings

toward the same thing. Again, standard survey measures used in the discipline are forcing

respondents to choose one side or the other, and based on the magnitude of those who

expressed mixed feelings about various social welfare issues when given the opportunity,

it would appear that standard indicators are missing this perhaps consequential detail.

The idea that many Americans may be ambivalent about social welfare is not novel.

Research has offered mixed results with regards to its prevalence, but the measures of

ambivalence on both sides of the argument have had problems that will be described in

detail later. Thus, the question is not resolved. Nonetheless, some suggest that

ambivalence about social welfare is fairly prevalent (Feldman and Zaller 1992; Cantril

and Cantril 1999; Hodson, Maio, and Esses 2001), while others argue that the public is

"not-so" ambivalent (Steenbergen and Brewer 2000; Jacoby 2002, 2005). The

conceptualization and measurement of ambivalence have not been standard across

political science or other disciplines. Thus, evidence of the prevalence, sources and

consequences of such ambivalence is also inconsistent. The present study suggests that

because the conclusions that have been drawn on both sides of these arguments were









based on data that do not offer adequate measures of ambivalence, further examination is

needed.

The contention here is also that loose conceptualizations of ambivalence in political

science research have contributed to less than optimal measurement strategies. As

previously mentioned, this study offers a fresh approach to measuring ambivalence about

social welfare policy. This approach is adapted from earlier work (Craig, Kane, and

Martinez 2002; Craig et al. 2005). A comparison of construct validity between this

approach and those used by other researchers suggests that it is higher in construct

validity. With this in mind, this study revisits some of the theoretical puzzles of earlier

research. Specifically, the questions addressed here are as follows: 1) Among the

previous approaches to measuring ambivalence using survey indicators, which has the

highest validity? 2) How prevalent is ambivalence surrounding social welfare policy? 3)

What are the potential sources of such ambivalence? 4) Who is more likely to be

ambivalent and why? 5) What is the relationship between ambivalence about social

welfare, social welfare policy preferences, and evaluations of political candidates and

institutions? In answering these questions, the present study is organized as described

below.

Chapter 2 offers an extensive overview of the research that addresses the

phenomenon of ambivalence. This overview lays out the theoretical foundation for this

study. First, problems in the literature associated with the conceptualization of

ambivalence and related measurement issues are discussed. Second, evidence of the

prevalence and potential sources of ambivalence is covered. Third, the consequences of

ambivalence across a range of attitude objects are explored. Fourth, the literature









specifically addressing social welfare ambivalence is discussed in light of these

measurement and specification issues. Finally, a series of hypotheses are offered based on

the preceding review of the literature.

Chapter 3 explains all of the specifics concerning the measurement of the

dependent and independent variables employed here. This includes four different

approaches to measuring ambivalence about social welfare policy. This specification

includes descriptive statistics across all indicators. These descriptive statistics lay the

empirical foundation for the rest of the study addressing the prevalence of ambivalence.

The findings suggest that many are ambivalent across a range of social welfare issues,

including government spending on medical insurance, assistance for the homeless,

education, programs to improve the standard of living for poor Americans, providing jobs

for citizens, child care programs, and finally, retirement benefits.

Chapter 4 has two goals. The first is to test the validity of the four approaches to

measuring ambivalence described in the previous chapter. In this test, the sources of such

ambivalence are also explored. Previous research has suggested that ambivalence may

result when individuals have conflicting thoughts or beliefs (cognitive conflict),

conflicting feelings (affective conflict), or beliefs in conflict with feelings (cognitive-

affective conflict) (Steenbergen and Brewer 2000). This chapter argues, first, that all three

types of conflict stimulate ambivalence about all of the social welfare issues mentioned

above; and second, that we can assess the construct validity of each of the different

measures of ambivalence by comparing how well the theoretical sources of ambivalence

predict each respectively. The findings indicate that an approach adapted from

experimental work by social psychologists for use in large-sample surveys (Craig, Kane,









and Martinez 2002; Craig et al. 2005) that forces respondents to rate their positive and

negative feelings separately is the most valid.

After settling on the best measure, Chapter 5 revisits previous research that

addresses the question of who is most likely to be ambivalent. The findings presented

here challenge previous assertions that liberals are more likely to be ambivalent about

social welfare (Feldman and Zaller 1992; see also Jacoby 2002). In addition, the results

suggest that not only self-identified liberals, but also females and African Americans are

less ambivalent about social welfare policy than males and whites respectively. This

chapter builds on the findings of the previous chapter to sort out the underlying cause of

these differences across ideology, gender, and race. It argues that we can get a better

understanding of why members across these groups would be more or less ambivalent

about social welfare by teasing out the differences across these groups when it comes to

the potential sources of such ambivalence.

Chapter 6 focuses on the political consequences of ambivalence. Specifically, it

explores the relationships among social welfare ambivalence, social welfare policy

preference, and evaluations of George W. Bush, the U.S. Congress, and the Supreme

Court. The findings indicate that in some cases ambivalence weakens the relationship

between social welfare policy preferences and these evaluations. This mediating effect of

ambivalence is strong for evaluations of Bush, but not for Congress and the Supreme

Court. Further, where this mediating effect does exist (evaluations of Bush), it is

strongest for conservatives and whites as opposed to liberals and blacks.

Chapter 7 summarizes the findings presented here. In doing so, the results of the

tests of the hypotheses formalized in Chapter 3 are discussed. This discussion is based on






8


how these conclusions speak toward our understanding of the conceptualization,

measurement, prevalence, sources and consequences of ambivalence about social

welfare. Finally, a discussion is offered about how this study adds to our understanding of

Americans' feelings about social welfare and what it means for politics in the United

States today.














CHAPTER 2
ATTITUDINAL AMBIVALENCE: CONCEPTUALIZATION, MEASUREMENT
APPROACHES, PREVALENCE, SOURCES, CONSEQUENCES, AND SOME
HYPOTHESES

Behavioral researchers have begun to embrace the idea that people do not

necessarily have a single "true" attitude on issues, but rather a store of multiple and

sometimes conflicting attitudes that they might draw upon at any given time (Hochschild

1981; Feldman and Zaller 1992; Zaller 1992; Eagly and Chaiken 1993; Alvarez and

Brehm 1995; Thompson, Zanna, and Griffin 1995; Priester and Petty 1996; Armitage and

Connor 2000; Meffert, Guge, and Lodge 2000; Wilson, Lindsey, and Schooler 2000;

Lavine 2001; Newby-Clark, McGregor, and Zanna 2002; McGraw, Hasecke, and Conger

2003; see Craig and Martinez 2005a, 2005b for a review). The idea that individuals are

often ambivalent, or that they simultaneously possess positive and negative evaluations of

a single attitude object (Alvarez and Brehm 1995; Zaller 1992; Zaller and Feldman 1992;

Eagly and Chaiken 1993), has been extensively researched across disciplines. However,

previous large-sample work attempting to measure and explain such ambivalence has two

recurring problems.

The first, and perhaps largest, problem is that most survey research focused on

ambivalence uses survey indicators of questionable validity. This, in large part, is due to

the inappropriate conceptualization of ambivalence. Second, models of ambivalence are

often underspecified because they fail to incorporate important potential sources of

ambivalence. Most studies focus on one particular source rather than a range of potential

sources. Failure to control for alternative explanations brings the findings about the









respective focus into question. As a result of these problems, it is unclear just how

divided the American public is on the issues that shape elections and politics in general.

This chapter reviews the literature with regards to the conceptualization,

measurement, prevalence, sources and consequences of ambivalence. Following this

broad review, previous research focused specifically on social welfare ambivalence is

discussed. This is followed by a summary that formally asserts the hypotheses that are

tested in the present study.

Ambivalence as It Has Been Conceptualized and Measured

The concept of ambivalence is not new (Kaplan 1972; Scott 1969). It was initially

observed empirically by social psychologists using experimental data (Katz and Hass

1988; Katz, Wackenhut, and Hass 1986; Klopfer and Madden 1980; Priester and Petty

1996; Thompson, Zanna, and Griffin 1995). While ambivalence is often considered in

attitudinal research generally (see Ajzen 2001), the empirical evidence of ambivalence

surrounding political issues has only blossomed as of late (see Craig and Martinez 2005a,

2005b). Again, the problem is that there has been little agreement in the literature as to

what constitutes ambivalence.

Traditionally, researchers have measured people's attitudes on a bipolar continuum

that ranges from positive to negative with a neutral point in between (Eagly and Chaiken

1993; Thurstone 1928; Thurstone and Chave 1929). At first glance, this theory of

attitudes seems appropriate. Simply, we classify our assessments as good, bad, or so-so,

hence, positive, negative, or neutral. The problem is with the assumption that attitudes are

uni-dimensional. Individuals sometimes feel good and bad about an object

simultaneously. Just as we would suspect that some people possess good and bad feelings

about both political parties, it makes sense to describe something as both good and bad,









or even an ideology as both liberal and conservative. Politicians are seen as being liberal

on some issues and conservative on others (Abelson et al. 1982).

As noted above, behavioral researchers have begun to take note of such

phenomena, but conceptualizations and operational definitions have varied across studies

and disciplines. Generally the psychology literature conceptualizes ambivalence in more

narrow terms than does political science (at least early political science studies). The

present study concurs with the psychology literature and some of the more contemporary

political science research that argues that ambivalence is strong internalized conflict

occurring wherever individuals possess positive and negative evaluations of a single

attitude object (see Cacioppo and Berntson 1994; Bassili 1998; Craig, Kane, and

Martinez 2002; Albertson, Brehm, and Alvarez 2005; Gainous and Martinez 2005). The

contention here is that political science has typically not measured ambivalence in a way

that is consistent with the definition given above. Therefore, these measures are not valid.

They are, simply, capturing something other than the simultaneous possession of positive

and negative evaluations of a single attitude object.

For instance, some political science studies have treated value conflict and

ambivalence synonymously (Hochschild 1981; Feldman and Zaller 1992). This is

problematic because the conflict of separate values is not the same thing as having

positive and negative evaluations of a single object. While this may be a source of

conflicting evaluations (to be discussed below), it is certainly not an appropriate

operational definition if we look to psychology's conception of ambivalence. Others

suggest that the presence of inconsistent positions across separate attitude objects is

reflective of ambivalence (Cantril and Cantril 1999). This does not necessarily signify









ambivalence either. For example, if someone was supportive of spending for the poor but

not supportive of food stamps, he or she is not necessarily ambivalent about social

welfare. This may be a lack of constraint (Converse 1964), non-attitude, or the result of a

host of other reasons. Again, ambivalence refers to conflicting positive and negative

evaluations regarding a single attitude object (see Gainous and Martinez 2005).

Perhaps the loose conceptualizations offered by political science are a result of the

early work that addressed this phenomenon. Hochschild (1981) said ambivalence exists

when an individual cannot resolve conflict between values. Based on these findings

combined with their own results, Zaller and Feldman (1992) concluded that mostot

people possess opposing considerations on most issues, that is, considerations that might

lead them to decide the issue either way" (p. 585). By this definition, the conflict of

values or inconsistent positions across separate attitude objects could constitute

ambivalence. This misleading conceptualization contributed to measurement strategies

that were either not focused on single attitude objects or were centered on the potential

sources of ambivalence, like value conflict, rather than the actual phenomenon itself.

Using responses to open ended questions contained in NES data, Feldman and

Zaller (1992) measured ambivalence by counting the number of conflicting

considerations, spontaneous statements of ambivalence, and two-sided remarks (i.e.,

"Although I think X, I nevertheless favor Y"), finding strong support for the presence of

ambivalence in many of the respondents. Although the substantive argument may be

compelling, other scholars also note that this approach is not valid (Craig, Kane, and

Martinez 2002; Alvarez and Brehm 1995). Alvarez and Brehm (1995) point out that an

individual's ability to express reasons for permitting and prohibiting abortion does not









necessarily signify the presence of an underlying conflict, whether between core values

(Katz and Hass 1988; Schnell 1993) or other idea elements.

"Opposing considerations" may in fact indicate the presence of factors such as

equivocation (when someone is trying not to make a bad impression on the interviewer),

uncertainty (when he/she is unsure of which side of the issue to choose), informedness

(where the respondent has sufficient information to cite both sides evenly while clearly

favoring one or neither), nonattitude (where the respondent has no real position on the

issue), or the questions may be ambiguous making them insufficient as indicators of

preference (see Alvarez and Brehm 1995; Alvarez and Brehm 1997; Alvarez and Brehm

1998). To resolve the validity problem in Feldman and Zaller's work, Alvarez and

Brehm, as does this study, define ambivalence as strong internalized conflict about a

single attitude object.

The problem for Alvarez and Brehm (1995) is not that a loose conceptualization of

ambivalence led to an invalid measure of ambivalence. Instead, they defined it

appropriately but used a measure that does not accurately represent its conceptualization.

They inferred ambivalence from patterns of error variance in heteroskedastic probit

models of binary choice by suggesting that error variance across individuals indicates the

presence of ambivalence within individuals. However, such inferences about an

individual-level concept (ambivalence) based upon aggregate-level data (error variance in

binary choices) are problematic. Error variance is high, by definition, when a larger

proportion of people are not predicted accurately by the binary choice model, whereas

ambivalence exists when an individual person holds both positive and negative feelings

about an issue. High error variance may indeed be, in part, a result of ambivalence, and









Alvarez and Brehm's findings give us reason to believe that they probably are to some

degree. But this still does not allow us to infer if any one individual is ambivalent. Also,

error variances are an accumulation of errors in the binary choice model and may be a

function of nonattitudes, uncertainty, equivocation, or a host of other factors as well.

While Alvarez and Brehm get it right by tightening up the conceptual definition, their

measure fails to distinguish ambivalence from the very things they suggest are

problematic with the Feldman and Zaller study. As a result, the validity of their measure

is questionable.

A distinction can be made between types of measures. Bassili (1996) suggested that

attitude strength has been measured using meta-attitudinal and operative indexes of

strength. Defined, "meta-attitudinal indexes are based on respondents' impressions of

their own attitudes, whereas operative indexes are linked to the judgment processes

responsible for attitude responses" (Bassili 1996, p. 638). The same principle is applied to

measuring the attribute of ambivalence. To simplify the terminology for the remainder of

the present study, meta-attitudinal and operative measures are referred to as direct and

indirect measures, respectively. Alvarez and Brehm's measure of ambivalence is

considered an indirect measure of ambivalence because it attempts to ascertain the degree

to which ambivalence is present without making the respondent aware. Thus, it is

"indirect." Others have used direct measures of ambivalence which simply ask

respondents, using various wording strategies, to report the extent to which they are torn.

For instance, Priester and Petty (1996) had respondents report on an 11-point scale the

degree to which they felt conflict, indecision, or mixed with regards to a host of attitude

objects. Employing a similar approach, Thompson, Zanna, and Griffin (1995) used ten









questions developed by Jamieson (1988) to probe whether people were "confused" or

"torn." Cacioppo and his colleagues asked people to report the extent to which their

reactions to an object were muddled, divided, tense, contradictory, jumbled, conflicted,

consistent, uniform, and harmonious (Cacioppo, Gardner, and Berntson 1997).

Many studies, including the Priester and Petty (1996) and Thompson, Zanna, and

Griffin (1995), have explored the relationship between indirect and direct measures of

ambivalence, finding only a modest relationship. Holbrook and Krosnick (2005) suggest

this is because they are separate constructs.1 While it is not their focus, they also suggest

that direct measures may be indirectly tapping the underlying conflict that results in

polarized evaluations of a single attitude object. They go on to define each of these

constructs as different types of ambivalence. This approach, again, reflects that lack of

consensus as to what actually constitutes ambivalence. The contention that these are

separate constructs makes sense, but allowing measures to define the concept is

problematic. Having a clear conceptualization of the concept makes measurement design

easier.

Craig and his colleagues created an indirect measure of ambivalence using large-

sample surveys that focused on single attitude objects while making fewer assumptions

than the inferential measure employed by Alvarez and Brehm or the direct measures

discussed. They asked respondents to rate positive and negative evaluations separately;

ambivalence was then calculated using an algorithm that captures the intensity of these





1 They refer to direct measures as meta-psychological (MP) measures and indirect
measures as operative (OP) measures.









feelings and the similarity between the two. Using this approach, any inconsistencies

observed are based on a single attitude object. Also, with this measure, it is not necessary

to make assumptions based on the variability of responses. Of course this approach does

not completely isolate ambivalence because respondents may bounce around on the

separate positive and negative components as a result of nonattitude or uncertainty, but it

definitely requires fewer assumptions than the latent measure employed by Alvarez and

Brehm.

Clearly, explorations into ambivalence and political attitudes have had some

conceptual problems but the theoretical insights they offer are still useful, assuming that

we can get around the problems with validity. For instance, Hochschild's (1981), Zaller's

(1992), and Zaller and Feldman's (1992) misconceptualization of ambivalence as value

conflict led to studies that suggested it was a source of ambivalence rather than the

phenomenon itself (see Martinez et al. 2005 for a review). Zaller and Feldman also

contend that ambivalence can help to account for the response instability so often

associated with surveys of ordinary citizens (see Converse 1964). The ideas that value

conflict is a source of ambivalence and that ambivalence may be related to response

instability are reasonable theoretical expectations. The executions of all these studies just

lack appropriate measures.

The present study concurs with the conceptualization of ambivalence that is

predominant in psychology and has been co-opted by more contemporary political

science (see Cacioppo and Berntson 1994; Bassili 1998; Craig, Kane, and Martinez 2002;

Albertson, Brehm, and Alvarez 2005). Simply, the term ambivalence should be restricted

2 Thompson, Zanna, and Griffin (1995) created this algorithm (Ambivalence = [(P +
N)/2] |P N|) to measure ambivalence by modifying an earlier version (Kaplan 1972).









to instances of strong internalized conflict where individuals clearly possess positive and

negative evaluations of a single attitude object. The measure employed by Craig and his

colleagues is the most consistent with this conceptualization. The present study uses an

adapted version of this measure. Now that the shortcomings of previous work pertaining

to the conceptualization and measurement of ambivalence have been described, we can

move on to discussing the prevalence of such ambivalence.

The Prevalence of Ambivalent Political Attitudes

The study of ambivalence is growing across disciplines. In an extensive review of

attitude research from 1996-1999, Azjen (2001) devoted an entire section to ambivalence

indicating its growing pervasiveness. Ambivalence research has reached into many

substantive areas including health studies, marketing, and many other areas (Armitage

and Conner 2000; Conner et al. 2002; Jewell 2003). Studies have even gone as far as to

use brain-imaging technology to identify the neurological activity associated with

ambivalence (Cunningham et al. 2003). However, the empirical evidence of ambivalence

centered on political attitudes and the implications therein is still limited, but growing. It

is fair to say that the concept of ambivalence is standard in the psychology literature; and,

while the idea that ambivalence is a component of political attitudes is not debated, its

prevalence and consequences are another question. The results have been mixed when it

comes to these questions (see Jacoby 2005; Steenbergen and Brewer 2000), but coupled

with the measurement issues across the literature, there is just not enough evidence to

draw any definitive conclusions.

Research concerning the prevalence of ambivalence has suggested that many

Americans are ambivalent about a host of different things including abortion (Alvarez

and Brehm 1995; Craig, Kane, and Martinez 2002), gay rights (Craig et al. 2005), social









welfare (Feldman and Zaller 1992; Cantril and Cantril 1999; Gainous and Martinez

2005), the environment (Chong and Wolinsky-Nahmias 2005), and American institutions

such as Congress, the president, and the Supreme Court (McGraw and Bartels 2005).

Others suggest that ambivalence is not widespread (Steenbergen and Brewer 2000;

Jacoby 2002, 2005), at least when it comes to several policy domains.3 There certainly

seems to be more research suggesting the former.4

Potential Sources of Ambivalence

With regard to the sources or stimulants of ambivalence, theory has offered several

primary suspects including the conflict of core values (Craig et al. 2005; Martinez et al.

2005; Feldman and Zaller 1992; Alvarez and Brehm 1995; Eagly and Chaiken 1993;

Katz and Hass 1988) the conflict of feelings (Lavine and Steenbergen 2005; Steenbergen

and Brewer 2000), and conflict between the former and latter (Lavine et al. 1998;

Steenbergen and Brewer 2000).

Value conflict is the most frequently cited source of ambivalence (Feldman and

Zaller 1992; Alvarez and Brehm 1995; Eagly and Chaiken 1993; Zaller 1992; Katz and

Hass 1988). Values are overarching normative principles and belief assumptions about

government, citizenship, and society (McCann 1997; see also Jacoby 2002; Feldman and

Zaller 1992; Feldman 1988). Egalitarianism, individualism, humanitarianism, and moral

traditionalism are all examples of values people have and rely on to form opinions.


3 Steenbergen and Brewer (2000) explore ambivalence about social welfare, gay rights,
affirmative action, and abortion. Jacoby (2002, 2005) focuses only on attitudes about
government spending.

4 For a complete discussion of the prevalence of ambivalence across policy domains and
various objects see various chapters contained in a two volume set of books focused on
ambivalence and politics edited by Craig and Martinez (2005a, 2005b).









Ambivalence is believed to result when an issue crosscuts an individual's values. For

instance, research has indicated that egalitarian and individualist values structure attitudes

about social welfare (Feldman 1988; Feldman and Zaller 1992; McCann 1997; Feldman

and Steenbergen 2001; Goren 2001; also see Gilens 1995). If an individual places equal

importance on each, social welfare pits them against each other, ambivalence may result.

Following the work of Steenbergen and Brewer (2000), the present study contends that

such value conflict is a cognitive process. Because values are overarching normative

principles and belief assumptions about government, citizenship, and society, and beliefs

are cognitive orientations, the conflict between values can be conceptualized as cognitive

conflict.

Although it is generally believed that ambivalence occurs when there is a conflict

involving a person's core values (Alvarez and Brehm 1995; Eagly and Chaiken 1993;

Katz and Hass 1988), the evidence showing this to be the case is limited. In their study of

political tolerance, Peffley and his colleagues (2001) assumed that value conflict and

ambivalence are interchangeable terms yet failed to demonstrate an actual link between

the two using either indirect or direct measures of ambivalence. As previously discussed,

Alvarez and Brehm (1995) inferred the presence of ambivalence in citizens' attitudes

about abortion from patterns of error variance in heteroskedastic probit models of binary

choice- an approach to measuring ambivalence that is problematic. Although Alvarez and

Brehm can compare the level of ambivalence in the general public across issues, the

relationship between ambivalence on any given issue and value conflict is assumed rather

than tested directly at the individual level.









Jacoby (2002) also explored the effects of value conflict on ambivalence. He refers

to "value conflict" as "value ambivalence" which further muddies what actually

constitutes ambivalence. In an ordinary least squares equation, he used his measure of

value ambivalence to predict attitudes about government spending. His argument was that

the equation's fit to the data should be significantly worse among those with value

ambivalence, as opposed to those who are not conflicted, if value conflict is indeed a

source of ambivalence about government spending. His findings provided only minimal

support for the hypothesis. Again, the lack of significant findings may be a product of his

measure of ambivalence. As previously discussed, variance between predicted and

observed values may result from a host of other factors such as nonattitude or uncertainty.

There is no reason to believe that value conflict is associated with either, so this may

explain why the relationship between response variance and value conflict is small.

Armitage and Conner (2005) assert that empirical evidence suggesting that value

conflict is a source of ambivalence is mixed simply because, first, the values researchers

explore are, in fact, not diametrically opposed (see Albertson, Brehm, and Alvarez 2005).

Second, these values may not be not as stable as theory suggests because most people do

not feel intensely about them; hence, it is possible that value structures are malleable.

They assert that findings supporting the contention that value conflict is a source of

ambivalence have demonstrated only moderate support for the hypothesis at best (see

Craig et al. 2005; Gainous and Martinez 2005). They note that despite the fact that value

conflict is often mentioned as a source of ambivalence, very few studies have actually

explored this possibility. This makes their conclusions a bit premature. Clearly more

work needs to be done before this question is resolved.









There is also, potentially, another interpretation of the previous research Armitage

and Conner use to draw their conclusions. They suggest that this research often fails to

consider the personal importance individuals place on values and as result felt conflict

may be overestimated. They note that Gainous and Martinez (2005) contend that value

importance is negatively related to ambivalence but assert that, because value importance

is a stronger predictor of ambivalence than value conflict, such a relationship provides

only weak support for the value conflict hypothesis. Another interpretation of the results

is that the findings of positive effects are more convincing than previous studies because

Gainous and Martinez controlled for value importance while most models overlook this

factor. So in this instance, cognitive conflict between values may indeed have an effect

on ambivalence that is independent of the personal importance people place on their

values.

Armitage and Conner's (2005) suggestion that values are not always diametrically

opposed can be extended to suggest a conclusion different from theirs. It is possible that

the degree to which values are brought into opposition is dependent on the particular

attitude object. Craig and his colleagues found virtually no zero-order relationship

between value conflict and ambivalence about gay rights, and multivariate analyses

revealed statistically significant associations but small effect sizes. Gainous and Martinez

did find a significant zero order relationship between value conflict and social welfare

ambivalence. The Craig study looks at conflict between egalitarianism, individualism,

and moral traditionalism, and the Gainous/Martinez study at egalitarianism and economic

individualism. The relationship between these latter values and attitudes about social

welfare is stronger than the relationship between egalitarianism, individualism, and









attitudes about gay rights. At the least, these discrepancies suggest more inquiry is

necessary concerning value conflict as a cognitive source of ambivalence.

It has also been suggested that affective orientations may clash to produce

ambivalence surrounding a particular attitude object (Lavine and Steenbergen 2005;

Steenbergen and Brewer 2000). An individual's feelings with regard to some attitude

object may be mixed. If these affective orientations structure attitudes about another

object, mixed feelings may result in ambivalence about that object. This process can be

thought of as affective conflict. It is important to note that affect and cognition are distinct

from each other.5 This is not meant to imply that there is no relationship between the two,

but rather to suggest that they may have independent effects on the stimulation of

ambivalence.

If affect can structure attitudes independently of cognition (Breckler 1984), it is

reasonable to expect conflicting affective orientations to produce ambivalence (see

Lavine and Steenbergen 2005). Studies across a range of disciplines have demonstrated

that these orientations do indeed structure attitudes centered on a host of various objects

(see Marcus 2003 for a review). For instance, group-affect, or how individuals feel about

the members of particular groups (i.e., sexual orientation, race, etc.) is a structural

component of attitudes about gay rights (Wilcox and Norrander 2002; Wilcox and

Wolpert 2000; Nelson and Kinder 1996), affirmative action (Kinder and Winter 2001;

Nelson 1999; Alvarez and Brehm 1997; Kinder and Sanders 1996; Nelson and Kinder


5 In current psychological conceptions, cognition is a surface phenomenon of the
neocortex and affect is thought to be located in the older regions of the brain (MacLean
1990). While the present study is neither concerned with the neurological differences
between cognition and affect nor with the primacy of one over another, it is important to
recognize them as two separate constructs.









1996; McConahay 1986) and social welfare (Kinder and Winter 2001; Nelson 1999;

Sniderman, Brody, and Tetlock 1991; Cook and Barrett 1992; Bobo and Kluegel 1993;

Gilens 1995; see also Jacoby 2005).6 If attitudes about these policies are structured by

how individuals feel about the perceived beneficiaries, and these individuals have

conflicting feelings, then it makes sense that this affective conflict could stimulate

ambivalence.

Lavine and Steenbergen (2005) demonstrate that conflict between feelings about

conservative and liberal groups has consistent effects on a variety of behaviors and

attitudes. Most significantly, people who are most conflicted about group evaluations

tend to decide later whom to support for president, are more likely to split their ticket on

election day, and are less likely to utilize their issue preferences when deciding how to

vote in presidential elections. Their findings suggest that affective conflict is a source of

ambivalence equally across ideology, party identification, and presidential vote choice

but again, due to the limitations of the data (NES), they are forced to infer ambivalence

from the error variance in each respective model.

If cognitive orientations and affective orientations structure attitudes, it is also

reasonable to expect that conflict between the two may stimulate ambivalence (Lavine et.

al 1998; Steenbergen and Brewer 2000). Research has embraced the idea that

individuals' attitudes are likely shaped by both feelings and cognitions (Breckler and

Wiggins 1989; Esses, Haddock, and Zanna 1993; Millar and Tesser 1989). Some contend

that emotions undermine an individual's ability to make reasoned decisions (Sears 2000;


6 Nelson (1999) does suggest that cognitive elements of outgroup attitudes dominate
affect in their influence on policy opinion but this study does not focus on the primacy of
one effect over another.









Sears et al. 1980), while others have suggested that the affective and cognitive bases of

attitudes are generally consistent (Breckler 1984; Festinger 1957; Rosenberg and

Hovland 1960). Lavine and his colleagues (1998) argue, first, that cognition and affect

are not always consistent; second, that affect has a stronger influence on attitudes among

those who have conflicting cognitions, while there is no primacy effect among those with

similarly valenced affect and cognition. While the present study does not address the

primacy of affect vs. cognition, these findings do indicate that cognitive-affective conflict

occurs (see also Steenbergen and Brewer 2000).

Thus, the evidence suggests that there are various potential sources of ambivalence

including cognitive conflict, affective conflict, and cognitive-affective conflict. The

problem is that researchers fail to build models that incorporate all of these theoretical

explanations. Most models are underspecified and, as a result, the findings they generate

are questionable. Steenbergen and Brewer (2000) built a model that incorporates all of

these factors and they found very little ambivalence; however, the validity of their

measure of ambivalence is dubious. The authors note potential weaknesses with regard to

their measure, leaving the door open for further research.

The Consequences of Ambivalent Political Attitudes

Early research suggested that ambivalence plays a role in shaping citizens' feelings

about their national leaders. Abelson and his colleagues (1982) found that positive and

negative affective reactions toward candidates clustered on separate factors; or, in other

words, their reactions formed two factors that were not significantly correlated with one

another. Thus, having good feelings toward a candidate did not necessarily imply an

absence of bad feelings toward him. Subsequent research suggests that ambivalence also

affects the manner in which attitudes are translated into behavior (Armitage and Connor









2000). Lavine and Steenbergen (2005) present evidence of this by showing that conflict

rooted in feelings toward liberal and conservative social groups can influence citizens'

voting decisions.

Evidence also suggests that ambivalence mediates political evaluations (McGraw et

al. 2003), and the relationship between a person's policy preferences and his or her

evaluations of political leaders and institutions (Craig et al. 2005). For example, in an

experiment, Haddock (2003) tests the degree to which candidate and party evaluations

are linked among those that are ambivalent about British politics in general. He used a

direct measure of ambivalence that simply asked subjects if they were mixed or

ambivalent. The results indicated that the correlation between evaluations of the Prime

Minister and the party were highest when participants evaluated them together rather than

separately, and that this correlation was strongest for those who were ambivalent. Other

studies have demonstrated that more negative evaluations of candidates are associated

with ambivalence (Craig et al. 2005; McGraw, Hasecke, and Conger 2003).

Research has suggested that consequences of ambivalence extend beyond

individual-level evaluative effects. Ambivalence can be thought of as an attribute of an

attitude. It has been hypothesized that ambivalence, like other attributes such as

importance, extremity, and certainty, will moderate the stability of attitudes over time

(Armitage and Conner 2000; Bassili 1996; Craig, Martinez, and Kane 2005). The

evidence, however, is decidedly mixed. Some have concluded that ambivalence does

contribute to attitude instability or susceptibility (Zaller and Feldman 1992; Bargh et al.

1992; Hill and Kriesi 2001; Craig, Martinez, and Kane 2005; Foumier 2005), while other









research has failed to uncover such a link (Bassili 1996; Armitage and Conner 2000,

2005).

Craig and Martinez's (2005a, 2005b) two edited volumes offer, perhaps, the most

complete examination of the prevalence of ambivalence and its political consequences to

date. The fact that these are the only books other than Cantril and Cantril (1999) that

focus entirely on ambivalence and its consequences for politics suggests that there is

clearly room for more inquiry. Further, mixed findings concerning prevalence and

consequences indicate that the matter is not settled. Much the same can be said when it

comes to research that focuses on social welfare ambivalence.

Ambivalence and Attitudes About Social Welfare

Whereas some assert that many Americans are ambivalent about social welfare

(Feldman and Zaller 1992; Cantril and Cantril 1999; Hodson, Maio, and Esses 2001;

Gainous and Martinez 2005), others counter that conflicting attitudes are relatively

uncommon among the general public (Steenbergen and Brewer 2000; Jacoby 2002,

2005). The contention here is that conclusions drawn on both sides of the argument suffer

from the same limitations as other studies of ambivalence: invalid measures and

underspecified models. The details of the findings of each of these respective studies will

be discussed in more detail in later chapters. For now, let us address the validity of the

measures used and the specification of the models.

First, concerning the validity of the measures, Feldman and Zaller (1992), Cantril

and Cantril (1999), Steenbergen and Brewer (2000), and Jacoby (2002, 2005) all use

measures that could be capturing phenomena other than ambivalence. Feldman and Zaller

(1992) found that social welfare ambivalence is fairly common, but this finding may be a

function of the indicators they employed. They measured ambivalence by counting the









number of conflicting considerations (e.g., a mix of liberal and conservative comments),

spontaneous statements of ambivalence (e.g., "I see merit in both sides"), and two-sided

remarks (e.g., "People should try to get ahead on their own, but government should help

when necessary") that were offered in response to open-ended probes about social

welfare attitudes in the 1987 NES Pilot Study. The problem is that any or all of these

comments might be offered without ambivalence necessarily being present. For example,

research suggests that many African Americans would probably agree with the statement,

"Although I think people are responsible for their financial conditions, I nevertheless

support social welfare" (see Kinder and Sanders 1996). Under the coding scheme

employed by Feldman and Zaller, black respondents who make such a statement would

be categorized as ambivalent when in fact they may not be. Instead, they may not connect

the two ideas.

Cantril and Cantril (1999; also see Free and Cantril 1967) simply assumed that

ambivalence is present whenever an individual expresses inconsistent opinions about the

size of government on the one hand, and spending for social programs on the other, that

is, when she/he expresses support for both smaller government and higher spending, or

vice versa. As a result of this loose conceptualization, defining ambivalence as

inconsistent positions about government size and government spending, their measure is

not valid.

Jacoby (2005) also questioned the work of Cantril and Cantril, but for different

reasons. He argued that the seemingly contradictory (or inconsistent) attitudes identified

in that study are not a product of ambivalence; rather, they represent distinctions made by

citizens between different types of government spending. While Jacoby is probably









correct in asserting that ambivalence about one social welfare issue does not necessarily

signify ambivalence about another, he also contends that "ambivalence is signaled most

clearly by the existence of seemingly contradictory patterns of opinion" (2005, p. 153).

His findings differ from those of Cantril and Cantril, but his measure suffers from the

same problem; inconsistent positions are not the same as having conflicting positive and

negative evaluations regarding a single attitude object (see Gainous and Martinez 2005).

Steenbergen and Brewer's (2000) examination of the 1992 NES led them to

conclude that the American public is "not-so" ambivalent about social welfare but their

measure of ambivalence is also problematic. Like Alvarez and Brehm (1996), they

inferred ambivalence from patterns of error variance in a heteroskedastic probit model.

As previously discussed, this is problematic because error variance may signify other

things in addition to ambivalence.

Measurement has clearly been a problem, but it is not the only one. The models

developed by Feldman and Zaller (1992), Cantril and Cantril (1999), Gainous and

Martinez (2005), and Jacoby (2002, 2005) all omit important potential sources of

ambivalence. Steenbergen and Brewer (2000) offer the most complete model of the

sources of ambivalence, but, the findings again are questionable because of their measure

of ambivalence. According to Steenbergen and Brewer, studies that assume ambivalence

and value conflict to be synonymous are defining ambivalence too narrowly and,

consequently, overlooking other types of conflict such as that which may involve a clash

between cognition and affect. By looking under every rock when including cognitive-

cognitive (value vs. value), affective-affective (group affect vs. group affect), and

cognitive-affective (value vs. group affect) conflict in their analysis, these authors









attempted to give ambivalence the benefit of the doubt; in the end, they found little

evidence of conflict and assumed low levels of ambivalence. While their model was more

complete than most, the results may, in fact, be a product of the inappropriate measure.

Gainous and Martinez (2005) attempted to remedy this problem by using the previously

discussed indirect approach to measuring ambivalence developed by Craig and his

colleagues (2002, 2005) but, like the models developed by Feldman and Zaller (1992)

and Jacoby (2002), their model only includes a measure of cognitive conflict.

In summary, the existing literature does not provide a definitive answer either way

regarding the prevalence, sources, and consequences of social welfare ambivalence. Each

study makes its own contribution, either theoretically or methodologically, but no one

study integrates each of these contributions. Craig and his colleagues (2002, 2005; see

also Gainous and Martinez 2005) offer a measure that is consistent with the more

narrowly defined conception of ambivalence. Most studies only look to value conflict as

a potential source of ambivalence about social welfare (Feldman and Zaller 1992;

Steenbergen and Brewer 2000; Jacoby 2002; Gainous and Martinez 2005). Steenbergen

and Brewer suggest that affective and cognitive-affective may also be a source of

ambivalence surrounding this issue but no one study integrates all of these potential

sources. The present study does so in an effort to build a more integrative model of

ambivalence based on the idea that ambivalence involves strong internalized conflict

regarding a related set of objects.

Putting it All Together with Some Hypotheses

Because there have been conceptual and operational problems surrounding the

study of ambivalence in general and specific to social welfare, it is reasonable to revisit

some of the questions addressed by prior research, but perhaps with a better approach to









measuring ambivalence. There are a series of hypotheses that will be tested throughout

the chapters in this study. The theory underlying these hypotheses is briefly described

here and elaborated on throughout the study:

H1 Cognitive, cognitive-affective, and affective conflict are positively related to
ambivalence about social welfare.

Value conflict is the most cited potential source of ambivalence (Katz and Hass

1988; Feldman and Zaller 1992; Eagly and Chaiken 1993; Alvarez and Brehm 1995;

Craig et al. 2005; Martinez et al. 2005). As covered earlier in this chapter, the literature

indicated that egalitarian and individualist values structure attitudes about social welfare

(Feldman 1988; Feldman and Zaller 1992; McCann 1997; Feldman and Steenbergen

2001; Goren 2001; also see Gilens 1995), and therefore, these are the values focused on

here. Values have been conceptualized as a cognitive construct (see Steenbergen and

Brewer 2000). By thinking of this broadly as a cognitive process rather than simply

conflict between values, the door is opened to think of potential affective sources of

ambivalence. Theory has suggested, first, that the conflict of feelings (Lavine and

Steenbergen 2005; Steenbergen and Brewer 2000) and conflict between cognitive and

affective elements (Lavine et al. 1998; Steenbergen and Brewer 2000) may stimulate

ambivalence, and second, that attitudes about social welfare programs are structured, in

part, by people's feelings about the perceived beneficiaries of such programs. Therefore,

we should expect cognitive-affective and affective conflict to create ambivalence.

H2 As individuals place more importance on one value as opposed to another they
are less likely to be ambivalent about social welfare.

Even if cognitive conflict in the form of value conflict is a source of ambivalence, it

is possible that value hierarchies (Rokeach 1973; Schwartz 1992; Jacoby 2002) exist and

are structured in ways that sometimes serve to reduce the likelihood of ambivalence









occurring. If someone places more importance on individualism than egalitarianism, and

an issue arises that happens to pit these values against each other, the conflict will not

necessarily matter for the preferred value will prevail and determine the person's response

to the issue in question. Jacoby (2002) presents evidence suggesting that most citizens

can, in fact, place their values in some rank order of importance.

H3 Between the indirect and direct approaches to measuring ambivalence, the
relationship between the above sources of conflict and ambivalence about social
welfare is strongest for indirect measures that gauge positive and negative
evaluations separately.

As described in this chapter, the conceptualization and operational definitions of

ambivalence have varied. As a result, evidence of the sources of ambivalence has been

mixed. If the theory is right that all of these types of conflict are sources of ambivalence,

then the mixed findings are most likely a result of inappropriate measures. The strictest

definition of ambivalence refers to the simultaneous possession of positive and negative

evaluations of a single attitude object as ambivalence (see Klopfer and Madden 1980;

Katz, Wackenhut, and Hass 1986; Cacioppo and Berntson 1994; Thompson, Zanna, and

Griffin 1995; Priester and Petty 1996; Alvarez and Brehm 1995, 1997, 1998; Craig,

Kane, and Martinez 2002; McGraw, Hasecke, and Conger 2003). Indirect measures that

gauge positive and negative evaluations separately are most consistent with this strict

conceptualization of ambivalence. For this reason, the theoretical predictors of

ambivalence about social welfare should perform best when using a measure that

captures the essence of concept of ambivalence from which the theory of sources was

originally derived (of course this assumes that measures of conflict are good). So,

whichever measure of ambivalence is best predicted by the sources of ambivalence









(cognitive, cognitive-affective, and affective conflict) has the highest construct validity,

assuming that the theory of the sources of ambivalence is accurate.

H4 Liberals are less ambivalent about social welfare policy than conservatives.

Hs Females are less ambivalent about social welfare policy than males.

H6 African Americans are less ambivalent about social welfare policy than whites.

H7 The variation in ambivalence across groups is a result of variation in the sources
of such ambivalence across groups.

Feldman and Zaller (1992) concluded that social welfare liberals, who tend to place

a high value on both egalitarianism and individualism, are more prone to ambivalence

than conservatives. According to them, liberals are more likely to experience value

conflict because while equality and individualism are of roughly equal importance for

liberals, conservatives tend to place more emphasis on individualism. Thus,

individualism usually trumps egalitarianism for conservatives, while it is fair fight

between the two values among liberals a fight that supposedly leads to higher levels of

social welfare ambivalence. However, evidence from a number of studies raises questions

about key aspects of the Feldman-Zaller argument. Women and blacks, for example,

generally support social welfare programs in disproportionate numbers (Kaufmann and

Petrocik 1999; Goren 2001; Gilens 1988, 1995; Bobo and Kluegel 1993; Tate 1994;

Kinder and Winter 2001), but the relative importance of equality and individualism do

not appear to be roughly equivalent among women and blacks, as Feldman and Zaller

might lead us to expect (Kinder and Sanders 1996; Feather 2004; also see Jacoby 2002).

Because women and blacks are generally thought of as social welfare liberals, and

the evidence suggests that their values are not distributed the way in which Feldman and

Zaller assert, we should not expect liberals to be more ambivalent than conservatives. In









fact, the distribution of values across these groups evidenced in the literature should lead

us to expect the opposite of Feldman and Zaller. Their logic holds up but from the other

direction. It is conservatives that should be more likely to be in conflict because liberals

prioritize values of equality over individualism, and therefore, their values are not as

likely to come into conflict. While conservatives are more likely to be individualist as

opposed to egalitarian (Feldman and Zaller 1992), the predominance of one value over

the other is not as great for conservatives.

Hs Ambivalence about social welfare policy weakens the relationship between
social welfare policy preferences and evaluations of candidates.

Research has extensively explored the consequences of attributes of attitudes such

as importance (Krosnick and Abelson 1992; Krosnick 1988a, 1988b; Boninger et al.

1995), extremity (Abelson 1995; Krosnick and Abelson 1992; Krosnick et al. 1993),

certainty (Budd 1986; Krosnick and Schuman 1988), and accessibility (Fazio 1986) to

name a few, but there has not been nearly as much research centered on the potential

political consequences of ambivalence. If ambivalence is an attribute of attitudes, and

research indicates that every other attribute of attitudes is consequential, there is every

reason to believe that ambivalence is too. Research has suggested that the relationship

between policy preferences and evaluations of political candidates and institutions is

weaker for someone who is ambivalent about the policy (see Craig et al. 2005); hence,

this hypothesis is tested here.

H9 The mediating effect of ambivalence about social welfare is stronger for
conservatives, men, and whites.

This is really an exploratory hypothesis. Research has suggested that ambivalence

has a mediating effect, and that its prevalence varies across groups, so it seems

reasonable to see if the mediating effect varies across groups. If conservatives, men, and









whites are more ambivalent about social welfare than their respective counterparts, the

intuitive expectation concerning the mediating effect across these groups is to expect the

effect to be stronger among those who are more ambivalent. Again, this is an exploratory

hypothesis. With no real literature to draw from, this just seems to be the most reasonable

expectation. What's more, the question is one worth answering. Is the mediating effect of

social welfare ambivalence stronger for some?

Conclusion

In the next chapter, all of the variables used in this study are described theoretically

and operationally. The adapted survey measure of ambivalence will be discussed in

further detail, as will other measures of ambivalence. Up to this point, the discussion of

the sources of ambivalence has been broad and not specific to social welfare. The next

chapter will cover the details of the cognitive and affective components that shape

attitudes about social welfare. The discussion focuses on how values shape attitudes

generally as well as specifically about social welfare, and on how group affect can serve

to shape these attitudes.














CHAPTER 3
SOCIAL WELFARE AMBIVALENCE: OPERATIONALIZATION AND
DESCRIPTIVE STATISTICS

This chapter describes all of the measures used to explore the prevalence, sources

and consequences of ambivalence. Four different measures of social welfare

ambivalence, indicators of the potential sources of ambivalence, and a host of controls

are detailed. These measures are used, first, to determine the most valid approach to

measuring ambivalence (Chapter 4). Next, they are used to explore the distribution of

social welfare ambivalence across ideology, gender, and race (Chapter 5). Finally, they

are used to examine the consequences of such ambivalence regarding candidate and

government evaluations (Chapter 6).

The data come from a telephone poll conducted between May 10-22, 2004 by the

Florida Voter survey organization. Six hundred and seven respondents were chosen

randomly from a list of all registered voters in the state of Florida. Only those whose

names were drawn from the list were actually interviewed. Up to four callbacks were

attempted on all working numbers and initial refusals. The margin of error is plus or

minus four percentage points. The survey itself included indicators that can be used to

build direct and indirect measures of ambivalence, values, value importance, feelings

about the perceived beneficiaries, attitudes about social welfare in general, and a variety

of control variables.1


1 Additional information can be obtained from the author, or from Florida Voter directly
(954-584-0204). I address the pervasive problem of missing data in analyses of survey
research by using a multiple imputation process. Using this procedure, I created five









The details of all these survey measures are explicated here. They include question

wording and response distributions across all independent and dependent variables. First,

the direct and indirect measures of ambivalence about social welfare policy are described.

Then the hypothesized sources of ambivalence (cognitive conflict, cognitive-affective

conflict, and affective conflict) concerning attitudes about social welfare policy are

detailed. Finally, all other variables are described including demographics and controls.

The distributions described here suggest that many are ambivalent about social welfare.

Also, many possess conflicting values (cognitive), conflicting feelings about the

perceived beneficiaries of social welfare (affective), and conflict between the two

(cognitive-affective).

Direct and Indirect Measures of Ambivalence

As discussed in the preceding chapter, researchers have employed two broad

strategies for measuring ambivalence. One approach involves using direct measures of

ambivalence where respondents are simply asked to state if they are torn or mixed

concerning the issue at hand. These are classified as direct measures because they are

based on the self-assessment of the respondent.

The other approach is indirect and involves ascertaining the degree to which

respondents are ambivalent without making them aware that a measurement of

ambivalence is being conducted. Obviously these approaches are classified as indirect

because their intent is to remove the subjectivity of the respondent.



replicate datasets based on the data, where the missing data in each replication are
substituted with draws from the posterior distribution of the missing value conditional on
observed values (Little and Rubin 1987; see also Horton and Lipsitz 2001). Analyses
which follow are based on the mean results of the five replicate imputed datasets.









The two direct measures of social welfare ambivalence employed in this study are

only used in Chapter 4. These measures are derived from a split-sample experimental

question contained in the instrument. Half of the sample was randomly assigned the

following question:

Some people think the government should provide fewer services, even in areas
such as health and education, in order to reduce spending. Others feel it is
important for the government to provide more services to citizens even if it means
an increase in spending. Which of these positions is closest to your own views, or
are you torn between the two?

The other half of the sample was asked the following question:

Some people think the government should provide fewer services, even in areas
such as health and education, in order to reduce spending. Others feel it is
important for the government to provide more services to citizens even if it means
an increase in spending. Which of these positions is closest to your own views?

After each question, respondents were asked how strongly they felt about their

position. On the first question, a 5-point scale was created ranging from fewer

services/strongly to more services strongly with torn between the two in the middle. In

the second question, interviewers were instructed to record any volunteered response that

indicated the respondent was mixed or torn as mixed/in-between. This also made a 5-

point scale after respondents rated how strongly they felt about their position. Dummy

variables were constructed from each of these indicators. If respondents chose the "torn

between two sides" response in the former indicator they were coded as a 1 and all other

responses were coded as a 0. The same was done for the latter whenever respondents

volunteered a mixed response. Therefore, there are two direct measures of social welfare

ambivalence.

The response distributions on these indicators are as expected (see Table 3-1).

Respondents were more likely to choose the "torn between sides" response (24.0 %) than









to volunteer a mixed response (12.5 %) (t-test p < 0.001). It makes perfect sense that

people would be more likely to choose a response that is offered as opposed to

volunteering one (see Bishop, Oldendick, and Tuchfarber 1980). Again these measures

depend on the direct assessment of the respondents. This is problematic if we strictly treat

ambivalence as the simultaneous possession of positive and negative evaluations because

people may choose or volunteer a mixed response for other reasons. Allowing

respondents to assess the degree to which they possess these evaluations may distort the

measure. For example, research has suggested that those without any real attitude on a

given issue may select the "safe" response (see Asher 2004 for a review). Often the safe

response is the middle or neutral response in a standard survey indicator. A response of

"both good and bad" equates to a middle/neutral or "safe" response.

The first indirect measure of ambivalence uses a similar method to that of Alvarez

and Brehm (1995, 1997, 1998, and 2002) and Jacoby (2002). Again, this measurement

strategy involves analyzing the residuals of a model of the attitude object under

examination. Residuals are the difference between the predicted value from the equation

and the actual observed response of that individual. The idea here is that the fit of the

equation to the data should be worse among individuals who are ambivalent because the

range of acceptable responses for them is greater. Because they have both positive and

negative feelings, their responses may bounce from side to side. This increases the

variance of their responses to the indicator of the dependent variable, and therefore, the

model will perform poorly at predicting their responses.

The present study extracts the residuals from a model of attitudes about social

welfare using ordered logit as opposed to probit (Alvarez and Brehm 1995, 1997, 1998,









2002) or ordinary least squares (Jacoby 2002) because social welfare policy preference is

measured using the ordinal scale described above for the direct measures of ambivalence.

Both groups (those who chose the torn response and those who offered a mixed response)

are combined for purposes of the analysis here.2

Table 3-1 Distribution of Responses on Two Direct Measures of Ambivalence
Total Percentage
Offered "Torn" Response
Fewer Services-Strongly 52 17.1%
Fewer Services-Not Strongly 23 7.6 %
More Services-Not Strongly 48 15.8 %
More Services- Strongly 108 35.5 %
Torn Between Sides 73 24.0 %
Total 304 100.0 %

Volunteered "Mixed" Response
Fewer Services-Strongly 51 16.8
Fewer Services-Not Strongly 37 12.2
More Services-Not Strongly 59 19.5
More Services- Strongly 118 38.9
Mixed/In Between 38 12.5
Total 303 100.0


Note: Data are from a Florida Voter survey of registered voters conducted in May 2004.
Missing values were replaced using multiple imputation.

Social welfare policy preference is modeled as a function of individualism,

egalitarianism, feelings about the beneficiaries, party identification, race, gender, and

income (these variables are described below).3 Then, the residuals from the ordered logit


2 A dummy variable is added for the question form to the multivariate analysis presented
below to control for any question format effect. The coefficient for the form dummy was
trivial. For the sake of parsimony, this variable is omitted from the model as reported in
Table 3-2.

3 These variables were selected because they are the primary ones in the literature used to
explain attitudes about social welfare (see Goren 2001; Feldman and Zaller 1992;
Kaufmann and Petrocik 1999; Gilens 1988, 1995; Bobo and Kluegel 1993; Tate 1994;
Kinder and Winter 2001).









estimates are extracted to create a new variable. The model does not perform well (see

Table 3-2). Only 3 of the seven independent variables reach traditional levels of statistical

significance (p < 0.05). Those who are more individualist are less likely to feel positively

about social welfare. Conversely, egalitarianism is positively associated with support for

social welfare. Republicans are less likely to feel positive than are independents or

Democrats. Neither race, gender nor income is significant. The pseudo R2 is 0.13.4

In the analysis of residuals, the strategy employed here parts from that of Alvarez

and Brehm, and Jacoby. They each analyze the variance of the residuals across different

groups that theory suggests should be more or less ambivalent. The problem is that their

analysis is aggregate in nature because it does nothing to explain why the model would

be a bad fit for any one individual. To remedy this problem, this study will simply use the

absolute value of the individual residuals and model them as a function of a set of

individual-level predictors.5 This allows us to infer why the predicted value is close or

not close to the observed value for each individual.

While exploring residuals at the individual-level may alleviate the problem of

previous studies that did aggregate analyses, this approach presents a problem of its own.

This measure is not consistent with the psychological conception of ambivalence as the

simultaneous possession of positive and negative evaluations of a single attitude object. It

may be the case that model fit will be worse for those that are conflicted, but poor model

4 Nagelkerke's R-Square is a modification of the Cox and Snell coefficient to assure that it
can vary from 0 to 1. That is, Nagelkerke's R2 divides Cox and Snell's R2 by its maximum
in order to achieve a measure that ranges from 0 to 1. Therefore Nagelkerke's R2 will
normally be higher than the Cox and Snell measure but will tend to run lower than the
corresponding OLS R2.

5 For an example of the same method using ordinary least squares regression see Marks et
al. (2004).









Table 3-2 Model of Attitudes about Social Welfare

Coefficient Standard Error 95 % Confidence Interval

Economic Individualism -0.11 0.03 -0.18 -0.05
Egalitarianism 0.17 0.03 0.11 0.24
Feelings/Beneficiaries -0.01 0.05 -0.10 0.09
Democrat 0.19 0.20 -0.21 0.59
Republican -0.49 0.20 -0.89 -0.09
Black 0.28 0.29 -0.28 0.84
Female 0.18 0.15 -0.12 0.48
Income 0.03 0.08 -0.13 0.18

-2 log likelihood 1738.56
Nagelkerke Pseudo R2 0.13
N 607


Note: Data are from a Florida Voter survey of registered voters conducted in May 2004.
Table entries are ordered logit coefficients, associated standard errors, and 95% confidence
intervals (threshold levels are not shown). Missing values were replaced using multiple
imputation.

fit could also result from other factors. For instance, if someone has non-attitude or no

real position on the issue they may guess which could certainly result in a poor

prediction.

This next indirect measure was modeled after that used by Craig and his colleagues

(2002, 2005). Respondents were asked to indicate how positively they viewed several

aspects of social welfare policy, using a battery of questions that was introduced as

follows:

I'm now going to read you a series of statements about the kinds of things some
people think the government should be doing to address certain problems that are
facing the country. After each, I'd like you to rate the statement on a 4-point scale
to indicate how positively you feel toward it. If you do not have any positive
feelings, give it the lowest rating of 1; if you have some positive feelings, rate it a
2; if you have generally positive feelings, rate it a 3; and if you have extremely
positive feelings, rate it a 4. Please rate each statement based solely on how
positively you feel about it, while ignoring or setting aside for the moment any
negative feelings you may have. The first statement is. ...









The statements were then read and respondents were asked to rate each one

separately. If a person seemed unsure or confused at any point, interviewers were told to

repeat the instructions as many times as necessary.

The specific aspects of social welfare policy (based on questions from the NES as

well as recent news stories) that respondents were asked to evaluate are as follows: The

government should....

* ensure that every citizen has adequate medical insurance;

* provide programs to help homeless people find a place to live;

* ensure that every child has access to a good education;

* provide programs that improve the standard of living of poor Americans;

* see to it that everyone who wants a job has one;

* provide childcare programs to assist working parents;

* ensure that the retirement benefits that citizens have built up over the years are
protected.

After this battery was given to respondents, they were given a number of filler questions

(party identification, political trust, and values indicators). Then the introduction was

repeated except with the words "positive" and "positively" replaced by "negative" and

"negatively." Following this introduction, respondents were read the same battery of

items as before.

Next, indicators of social welfare ambivalence were calculated for each item using

an algorithm developed by Thompson and her colleagues (1995; also see Kaplan 1972).6

This algorithm is:


6 This model is derived from a version of the semantic differential (Osgood, Suci, and
Tannenbaum 1957), as modified by Kaplan (1972) in an effort to show that people's
overall attitudes are made up of both positive and negative elements. Thompson and her









Ambivalence = [(P+ N)/2] IP N|

where P is the positive reaction score and N is the negative reaction score. The range of

scores for each of the seven items described above is -0.5 through 4.0, with intervals of

0.5.

We can see that this algorithm is sensitive to the intensity of positive and negative

sentiments (see Table 3-3). For instance, consider a person who reports "some" positive

feelings (score of 2), as well as no negative feelings (score of 1), with regard to one of the

social welfare items; that individual would be characterized by the model as experiencing

a modest amount of ambivalence (overall score of 0.5). If, in contrast, the same

respondent were to express "generally" positive feelings (score of 3) and no negative

feelings (score of 1), his or her score would fall to zero, i.e., no ambivalence at all.

Looking at the extremes, we would expect someone who reported "extremely" positive

and negative feelings (score of 4 and 4 respectively) to be the most ambivalent. The

algorithm produces the highest score possible (4) for these respondents. Conversely,

those with "extremely" positive feelings and "no" negative feelings or vice versa get the

lowest possible score (-0.5).

Next, an additive index of the seven separate items was created. Principal

components factor analysis confirmed that all seven load on a single factor, and the

reliability of an additive index constructed from them is very good (. = .860). It is

important to note that people's attitudes about social welfare policies may be multi-



colleagues (1995) adjusted the model to better account for the presence of polarized
beliefs. See Craig, Kane, and Martinez (2002) for a more complete discussion of this
measure as employed in a large-sample survey.









dimensional (see Jacoby 1994), but as confirmatory factor analysis indicates here,

ambivalence about such policies appears to be on one dimension. In other words, people

that are ambivalent about one aspect of social welfare policy are often ambivalent about

others.

Table 3-3 Logic of the Algorithm
Positive Reactions Score
Negative Reactions Score One Two Three Four

No Negative Feelings 1.0 0.5 0.0 -0.5

Some Negative Feelings 0.5 2.0 1.5 1.0

Generally Negative Feelings 0.0 1.5 3.0 2.5

Extremely Negative Feelings -0.5 1.0 2.5 4.0


Note: Table is in Craig, Kane, and Martinez (2002). Table entries are overall ambivalence
scores assigned to individuals with the indicated mix of positive and negative reactions.

The results shown in Table 3-4 suggest that there is some degree of variability in

the levels of ambivalence observed across the seven program areas that form the basis for

the social welfare ambivalence index, and that these levels are far from trivial. For the

sample as a whole, mean ambivalence scores are higher on policies that would assist the

homeless, improve the standard of living of poor Americans, ensure full employment,

and provide childcare programs to assist working parents; on each of these issues, more

than half of those in the sample have ambivalence scores greater than zero. Ambivalence

is less common with regard to universal medical insurance, programs to ensure that all

children receive a good education, and protecting retirement benefits. Overall, despite the

variation that is evident here, a single seven-item social welfare ambivalence index scales

well (see above) and will be used for the bulk of this study.









Table 3-4 Frequency and Intensity of Ambivalence about Social Welfare

Standard
Condition Mean Score Deviation Percen

Medical Insurance 0.41 1.32
Homeless 0.95 1.27
Education 0.31 1.40
Standard of Living 0.97 1.26
Job Guarantee 0.89 1.36
Child Care 0.76 1.29
Retirement Benefits 0.36 1.46
Ambivalence Index 4.65 6.90


t Ambivalent

35.4%
61.4%
27.3%
61.4%
56.5%
52.6%
28.0%
48.3%


Number of Cases = 607


Note: Data are from a Florida Voter survey of registered voters conducted in May 2004.
Table entries indicate the (a) mean ambivalence score for each item (scores ranging from
-0.5 to +4.0), and for the combined scale (scores from -3.5 to +28.0); (b) associated
standard deviation; and (c) percentage of respondents with scores greater than zero for a
given item (this threshold is arbitrary but it gives a sense of how many people offered
responses that are at least somewhat conflicted (see Table 3-3).

The distribution on this index is contained in Figure 3-1. The mean response is well

over 0 at 4.65. While the distribution is skewed left there are many scores to the right of

the mean indicating that many are ambivalent about social welfare (based on this

measure). Of course, the threshold that constitutes whether or not someone is ambivalent

is arbitrary (Craig et al. 2005). Nonetheless, these results indicate that many

simultaneously offered positive and negative responses. Before the inference can be

comfortably made that ambivalence is common, the relative construct validity of each of

these direct and indirect measures of ambivalence must be compared. This involves

seeing how well the theoretical sources of ambivalence predict each.




































0.00 10.00 20.00
Note: Data are from a Florida Voter survey of registered voters conducted in
May 2004.

Figure 3-1 Distribution of Social Welfare Ambivalence

Sources of Ambivalence in General and About Social Welfare

Cognitive Conflict

Since it was asserted in The American Voter that citizens have a cognitive map of

politics (Campbell et al. 1960) and that psychological forces shape political behavior,

research focusing on political cognition burgeoned (see Popkin 1991; Sniderman, Brody,

and Tetlock 1991; see Iyengar and McGuire 1993 and Lau and Sears 1986 for a review).

The basic idea is simple. People make decisions and form attitudes by using cognitive

shortcuts such as party identification, media cues, and values, among many others. These

shortcuts permit individuals to make reasonable decisions with minimal effort (Popkin









1991; Fiske and Taylor 1991). As for values being a cognitive shortcut, the idea is that

individuals do not need a sophisticated ideology to determine their political preferences.

Their preferences or political evaluations may be based on how consistent or inconsistent

policies or political actions are with certain beliefs they possess (Feldman 1988). Policies

and actions may be judged as right or wrong based on their congruence with deeply held

values (see Rokeach 1973).

What happens when an issue pits opposing values against each other? Cognitive

conflict such as this may result in ambivalence. In fact, value conflict is the most often

mentioned source of ambivalence in the literature (Katz, Wackenhut, and Hass 1986;

Katz and Hass 1988; Eagly and Chaiken 1993; Alvarez and Brehm 1995; Martinez et al.

2005; Newby-Clark, McGregor, and Zanna 2005). The contention here is the possibility

that varying value hierarchies (Rokeach 1973; Schwartz 1992; Jacoby 2002) could be

structured in such a way as to reduce the likelihood of ambivalence surrounding social

welfare (Gainous and Martinez 2005). If an individual places more importance on one

value as opposed to another, and social welfare issues pit these values against each other

(Feldman and Zaller 1992; Feldman 1988; Goren 2001), the conflict will not necessarily

matter. Simply, the preferred value will prevail and determine the person's response to the

issue in question. So, the effect of value conflict is dependent on the personal importance

people place on their values.

The literature suggests that values such as egalitarianism (Feldman 1988; Feldman

and Zaller 1992; McCann 1997; Feldman and Steenbergen 2001; Goren 2001; also see

Gilens 1995) and economic individualism (Feldman 1988; Feldman and Zaller 1992;

McCann 1997; Goren 2001; also see Gilens 1995) are related to citizens' attitudes about









social welfare.7 Egalitarianism can be thought of as a general belief in equality and

economic individualism as a belief in the freedom to accumulate wealth. Although

different studies may conceptualize, operationalize, and label the values in different

ways, there is broad agreement that greater individualism is associated with less support

and greater egalitarianism with higher support for social welfare programs and spending.

Theory suggests that cognitive conflict between individualism and egalitarianism is

potentially a source of ambivalence about social welfare policy.

To build a measure of conflict between individualist and egalitarian values,

separate indicators of each value were first constructed. Respondents were read a series

of companion statements and asked to say which came closer to their own opinion. For

individualism,8 the item pairs were:

* The government should see to it that every person has a job and a good standard of
living; or, the government should just let each person get ahead on their own.

* We need a strong government to handle today's complex economic problems; or,
the free market can handle these problems without government being involved.

For egalitarianism, the item pairs were:

* We have gone too far in pushing equal rights in this country; or, we should do more
to make sure that everyone is treated equally.

* If people were treated more equally in this country, we would have many fewer
problems; or, this country would be better off if we worried less about how equal
people are.



7 Feldman and Steenbergen (2001) contend that "humanitarianism" also is important as a
predictor of citizens' attitudes about social welfare.

8 These questions were designed to tap support for economic individualism, or a belief in
the freedom to accumulate wealth. Scholars with a different substantive focus might
prefer to measure individualism differently, for example, conceptualizing it in terms of a
belief in freedom of expression.









In all cases, responses were coded from 1 (strong support for the first statement in

the pair) to 5 (strong support for the second statement); for the second egalitarianism pair,

this scoring was reversed to provide consistency in direction of wording. The two sets of

items were then combined into indices with scores ranging from 2 to 10 (high values

reflecting stronger support for individualist or egalitarian values). These items did not

scale that well together (taub = 0.16, p < 0.05 for the individualist items and taub = 0.22, p

< 0.05 for the egalitarian items). These items were based on measures from NES and they

scale well together in those data but the approach here varies slightly. The companion

statements used here in each indicator are derived from two separate indicators from

NES. This allowed for the use of multiple measures with fewer questions on the survey.

Perhaps collapsing two items into each indicator decreased the validity, as opposed to

having four separate questions for each.

A measure of cognitive conflict, which captures the magnitude of the difference

between them, was calculated using the same algorithm as the one described earlier for

measuring social welfare ambivalence; that is,

Cognitive Conflict = [individualism + egalitarianism]/2 individualism -
egalitarianism|


with higher values representing more conflict. This item was rescaled to have values

between 0 and 1. The distribution of this measure is displayed in Figure 3-2. Apparent

from the normal curve superimposed on the histogram, cognitive conflict is not normally

distributed. In fact, nearly 60 respondents indicate no conflict at all. On the other hand,









most of those in the sample demonstrate at least some conflict.


Note: Data are from a Florida Voter survey of registered voters conducted in
May 2004.


Figure 3-2 Distribution of Cognitive Conflict

The present study proposes that the effects of value conflict on ambivalence are

dependent on the personal importance people place on their values respectively. Value

importance is based on responses to two separate items, introduced as follows: "As you

know, not everyone agrees on the different goals or values that our nation ought to

pursue. I'm going to list three9 different goals and have you tell me how important each of



9 The survey included a measure of the importance of traditional moral values, which was
asked in between the two indicators used here.









them is to you personally." The importance of egalitarianism and individualism was then

determined based on answers to a pair of questions:

* The first goal is equality, by which we mean a narrowing of the gap in wealth and
power between rich and poor. How important is equality to you extremely
important, important, only somewhat important, or not important at all? ...

* And the third goal is a free marketplace, by which we mean all citizens having a
chance to get ahead on their own without the government getting involved. How
important is a free marketplace to you extremely important, important, only
somewhat important, or not important at all?

Responses were recorded so that higher values represent greater importance. In

addition, the relative importance of one value as opposed to the other was calculated as

the absolute value of individualism importance subtracted from egalitarianism

importance; higher numbers indicate that one of these values has priority over the other

for the individual. For instance, someone who said one value was extremely important

and the other was not important at all would get a score of 3 while someone who said

both were extremely important would get a score of 0. So, the further the distance of

importance between values, the higher the score.

The distribution of responses displayed in Table 3-5 is skewed toward finding both

individualism and egalitarianism important. Near 70% of the respondents selected

important or extremely important for both values. Concerning differences, it appears that

more people think egalitarianism is not important at all than think individualism is not

important at all. The values are certainly in opposition to each other substantively, so it is

unlikely that the respondents are distributed in the same way for each separate value.

Cognitive-Affective Conflict

As described in Chapter 2, research has embraced the idea that individuals'

attitudes are likely shaped by individuals' feelings as well as cognitions (Breckler and









Table 3-5 Distribution of Responses on Value Importance Indicators
Total Percentage
Individualism Importance
Not Important at All 14 2.3 %
Only Somewhat Important 177 29.2 %
Important 195 32.1 %
Extremely Important 221 36.4 %
N 607 100.0 %

Egalitarianism Importance
Not Important at All 38 6.3 %
Only Somewhat Important 135 22.2 %
Important 225 37.1 %
Extremely Important 209 34.4 %
N 607 100.0 %


Note: Data are from a Florida Voter survey of registered voters conducted in May 2004.
Missing values were replaced using multiple imputation.

Wiggins 1989; Esses, Haddock, and Zanna 1993; Millar and Tesser 1989). If we know

that values are a cognitive base of attitudes about social welfare (Feldman and Zaller

1992; Feldman 1988; Goren 2001) and feelings about the beneficiaries of social welfare

are an affective base (Kinder and Winter 2001; Nelson 1999; Sniderman, Brody, and

Tetlock 1991; Cook and Barrett 1992; Bobo and Kluegel 1993; Gilens 1995; see also

Jacoby 2005), then it is reasonable to expect that conflict between these two components

would stimulate ambivalence about this issue. Apart from the obvious (poor people),

many citizens think of African Americans as being among the principal beneficiaries of

governmental welfare policies (Sniderman, Brody, and Tetlock 1991; Cook and Barrett

1992; Bobo and Kluegel 1993; Gilens 1995; see also Jacoby 2005). Accordingly, feelings

about welfare beneficiaries are measured with two additive indices, one for positive

feelings and another for negative feelings. These indices are based on answers to two









questions tapping respondents' affect toward "poor people" and "blacks." Respondents

were read the following introduction:

Next, I'd like to do the same thing except with a list of different government
institutions and groups that are active in politics. Once again: If you do not have
any positive feelings toward the institution or group, give it the lowest rating of 1;
if you have some positive feelings, rate it a 2; if you have generally positive
feelings, rate it a 3; and if you have extremely positive feelings, rate it a 4. Please
rate each institution or group based solely on how positively you feel about it, while
ignoring or setting aside for the moment any negative feelings you may have. The
first group is...

The names of the groups and institutions were then read (including poor people and

blacks) and respondents were asked to rate each one separately. Then, as with social

welfare items, the introduction was repeated except with the words "positive" and

"positively" replaced by "negative" and "negatively" following a number of filler

questions. As before, scores range from 1 (no positive/negative feelings) to 4 (extremely

positive/negative feelings) and from 2-8 after summing each respectively. The positive

items and negative items scaled well together indicating that people shared similar feeling

across these two groups (positive feelings a = .772; negative feelings a = .868).

These indices were used in combination with the values indicators described in the

previous section to construct a measure of cognitive-affective conflict. The idea here is

that we should expect conflict between individualist values and positive feelings about

the perceived beneficiaries and between egalitarian values and negative feelings about the

perceived beneficiaries to stimulate ambivalence. It is not logical to suggest that

individualist values will come into conflict with negative feelings about the beneficiaries

or for egalitarian values to conflict with positive feelings because each has the same

directional effect on attitude about social welfare (see Table 3-2). Rather than doing

separate measures for each, the scale for positive feelings about the beneficiary and









individualist values is inverted and each is added to negative feelings about the

beneficiary and egalitarian values respectively. Flipping them gives them the same

directional effect on attitudes about social welfare. Because there is no reason to expect

that conflict is more likely to stimulate ambivalence for individualist/positive feelings or

egalitarian/negative feelings conflict, inverting the scales permits one measure of

cognitive-affective conflict. After summing the inverted scales with the non-inverted

scales, each is rescaled so that all values fall between 0 and 1. The summed values scale

and the affective scale are then run through the same algorithm used to create scales of

cognitive conflict and ambivalence about welfare. The scale that it creates is also rescaled

to have values between 0 and 1.

The distribution of this scale of cognitive-affective conflict is displayed in Figure

3-3. It is not normally distributed. The distribution is skewed to the left indicating that

most people do not experience a large amount of this type of conflict. On the other hand,

very few people demonstrate no conflict in contrast to the distribution of cognitive

conflict (see Figure 3-2) but the average level of conflict is still lower for cognitive-

affective (0.36) as opposed to cognitive conflict (0.42). All in all, it appears that people

are more likely to experience cognitive conflict. Nonetheless, this does not mean that

cognitive-affective conflict is not a potential source of social welfare ambivalence.

Affective-Affective Conflict

As described in Chapter 2, affective orientations may clash independently of

cognition to produce ambivalence surrounding a particular attitude object (Lavine and

Steenbergen 2005; Steenbergen and Brewer 2000). Simply, if an individual has mixed

feelings about something, and these affective orientations structure attitudes about

another object, ambivalence may be the result. This process can be thought of as affective




































0.00 0.20 0.40 0.60 0.80 1.00
Note: Data are from a Florida Voter survey of registered voters conducted in
May 2004.
Figure 3-3 Distribution of Cognitive-Affective Conflict

conflict. A measure of affective conflict is created by summing the positive responses to

poor people and blacks, summing the negative responses to poor people and blacks, then

running this through the same algorithm used above juxtaposing positive responses

against negative responses. Next, this scale is recorded to have values that range between

0 and 1.

It is apparent in Figure 3-4 that affective conflict is also not normally distributed.

Like cognitive conflict, a significant proportion of respondents demonstrate no conflict at

all. In fact, more than twice the number of respondents demonstrates no affective conflict

as when compared to those who demonstrate no cognitive conflict. So, this type of









conflict is less frequent than cognitive conflict and cognitive-affective conflict. The

average score is lower (0.31) than both cognitive conflict (0.42) and cognitive-affective

conflict (0.36). Clearly this type of conflict is not widespread, but again, this does not

mean it is not a source of social welfare ambivalence when present.


150-



120-



S90-
=



60-



30- 0 00



0-
0.00 0.20 0.40
Note: Data are from a Florida
May 2004.


Mean = 0.3083
Std. Dev. = 0.27433
N =607


0.60 0.80 1.00
Voter survey of registered voters conducted in


Figure 3-4 Distribution of Affective-Affective Conflict

Dependent Variables for Consequences of Social Welfare Ambivalence

Policy preferences are often related to evaluations of political candidates (Aldrich,

Sullivan, and Borgida 1989; Miller and Shanks 1996). Evidence has suggested that the

link between preferences and evaluations of candidates and institutions may be mediated

by ambivalence (Craig et al. 2005). We should expect that the relationship between


~R~7h;









policy preferences and evaluations would be weaker among those who are ambivalent.

Chapter 6 examines this possibility in relation to attitudes about social welfare.

Accordingly, evaluations of George W. Bush, the U.S. Congress, and the Supreme Court

are measured. To measure these evaluations respondents were asked the following

questions:

* Do you approve or disapprove of the way George W. Bush is handling his job as
president?

* Do you approve or disapprove of the way Congress is handling its job?

* And do you approve or disapprove of the way the U. S. Supreme Court is handling
its job?

After each question they were asked if they approved/disapproved strongly or not so

strongly. A five-point scale was created for each ranging from disapprove strongly to

approve strongly with those who volunteered a mixed response coded in the middle.

The distribution of responses on all three of these evaluations is included in Table

3-6. The George W. Bush evaluation indicator is bi-modal with the largest percentages

falling in disapprove strongly (37.6 %) and approve strongly (37.2 %). The distributions

of the U.S. Congress and Supreme Court indicators are more even across responses.

Other Independent and Control Variables

Policy Preferences

Previous research has suggested that the probability of an individual feeling

ambivalent about a policy issue may be related to the person's position on that issue

(Craig et al. 2002; Craig et al. 2005; Gainous and Martinez 2005). Pro-life voters, for

example, tend to be more ambivalent about whether abortion should be legal under

"traumatic" circumstances, while pro-choice voters are more ambivalent about the

legality of "elective" abortions (Craig et al. 2002). Also, people with more positive views









about homosexuality in general are less ambivalent about gay rights on issues that do not

directly involve children or marriage (Craig et al. 2005). Finally, Gainous and Martinez

(2005) show that supporters of social welfare policy are less likely to be ambivalent.

Table 3-6 Distribution of Responses on Government Evaluations Indicators
Total Percentage
Evaluation of George W. Bush
Disapprove strongly 228 37.6 %
Disapprove not strongly 76 12.5 %
Mixed/In-between 19 3.1 %
Approve not strongly 58 9.6 %
Approve strongly 226 37.2 %
N 607 100.0 %

Evaluation of U.S. Congress
Disapprove strongly 134 22.1%
Disapprove not strongly 146 24.1%
Mixed/In-between 69 11.4 %
Approve not strongly 157 25.9 %
Approve strongly 101 16.6 %
N 607 100.0 %

Evaluation of the Supreme Court
Disapprove strongly 103 17.0 %
Disapprove not strongly 126 20.8 %
Mixed/In-between 83 13.7 %
Approve not strongly 151 24.9 %
Approve strongly 144 23.7 %
N 607 100.0 %

Note: Data are from a Florida Voter survey of registered voters conducted in May 2004.
Missing values were replaced using multiple imputation.

Accordingly social welfare policy preference is measured using the same indicator

that contains the direct measure of ambivalence. The split-samples are pooled where

those that chose "torn between sides" or volunteered a "mixed" response are in the same

response category. This creates a 5-point scale that ranges from extremely strong feelings

that fewer services should be offered if it means increasing spending to extremely strong

feelings that the government should provide more services even if it means increasing










spending, with those who are torn or mixed in-between. The percentage of responses by

category is displayed in Figure 3-5. Interestingly the response selected most often was for

more services even if it means an increase in government spending (nearing 40%).

Roughly equal percentages (just below 20%) selected fewer services strongly, mixed or

torn, and more services not strongly, while only about 10% chose fewer services not

strongly.



40-





30-





1 20-





10-


o0 I I *I I I
fewer services fewer services mixed or torn more services more services
strongly not strongly not strongly strongly
Note: Data are from a Florida Voter survey of registered voters conducted in
May 2004.


Figure 3-5 Social Welfare Policy Preferences

Ideology, Party Identification, Race, Gender, and Income

A series of control variables are also employed in this study. These variables are

used to explore differences across groups, particularly, differences in the prevalence of









ambivalence across ideology, gender, and race. As noted earlier, Feldman and Zaller

(1992) concluded that social welfare liberals, who tend to place a high value on both

egalitarianism and individualism, are more prone to ambivalence than conservatives. This

proposition is revisited with the new measure of ambivalence previously described. To

measure ideology, a traditional seven-point indicator is used here. Respondents were read

the following statement:

We hear a lot of talk these days about liberals and conservatives. On a scale of one
through seven, where "1" is very liberal and "7" is very conservative, where would
you place yourself on this scale or haven't you thought much about this?

For the purpose of this analysis this measure was collapsed where those who reported a

score of 1 through 3 were categorized as liberals, 4 as moderates, and 5 through 7 as

conservatives.

Party identification is also used. It was measured a multipart indicator. First,

respondents were asked:

Generally speaking, do you usually think of yourself as a Republican, a Democrat,
an Independent, or what?

Then the strength of their partisanship was gauged with a follow-up question. This

created a traditional 7-point party identification scale ranging from strong Democrat to

Strong Republican. Dummy variables were created for Democrat and Republican

identification (0 = not Democrat and 1 = Democrat, 0 = not Republican and 1 =

Republican, 0 = all Independents and Independent leaners).

Variables are also created for race (0 = nonblack, 1 = black), gender (0 = male, 1 =

female), self reported income (collapsed into 5 categories- Between $0 and $10,000,

between $10,000 and $30,000, between $30,000 and $50,000, between $50,000 and

$70,000, $70,000 or more) are included. The particular importance of social welfare









issues in defining both the gender gap (Gilens 1988; Kaufman and Petrocik 1999; Goren

2001) and the racial cleavage (Bobo and Kluegel 1993; Tate 1994; Gilens 1995; Kinder

and Winter 2001) in American politics suggests that gender and race are important

predictors of attitudes about social, and therefore, there may be differences in

ambivalence across these groups.10 Because income is important in structuring attitudes

about social welfare (Goren 2001) it is also included as predictor of social welfare

ambivalence. The distribution across all of these variables is contained in Table 3-7.

Political Knowledge

Evidence has suggested that there is a relationship between ambivalence and levels

of political sophistication. Zaller (1992) proposes that individuals can reliably resist the

arguments to which they are exposed only to the extent that they possess information

about these arguments. He asserts that, in combination with the idea that most Americans

are not very politically aware, citizens will be unlikely to exhibit high levels of resistance

due to low levels of information. Thus, in an environment wherein communications on

both sides of the issues is evenly distributed, it is likely that individual attitudinal

ambivalence will arise. Alvarez and Brehm (1995, 1997, and 1998) also argue that those




10 Since Latinos in the aggregate are more liberal, at least on certain issues, than whites
(Welch and Sigelman 1993; DeSipio 1996; Uhlaner, Gray, and Garcia 2000; Alvarez and
Bedolla 2003; also see de la Garza, Falcon, and Garcia 1996), we might normally expect
their level of social welfare ambivalence to be similar to that found among blacks and
women. Unfortunately, this proposition cannot be tested because the race indicator used
here does not make distinctions among different groups of Latino citizens. In particular,
we know that Cubans tend to be more conservative than other Latinos (especially Puerto
Ricans, but also Mexicans; see de la Garza et al. 1992) and there is a large Cuban
population in Florida. As a result, it is not surprising to learn that the Latinos in the
Florida Voter survey do not, on average, differ significantly from whites in terms of the
variables that are most critical to our analysis.









Table 3-7 Distribution of Ideology, Party Identification, Race, Gender, and Income
Total Percentage


Ideology
Liberal
Moderate
Conservative
N

Party Identification
Democrat
Republican
Independent
Other


Race
White
Black
Hispanic
Other
N

Gender
Female
Male


Income
Between $10,000 and $30,000
Between $30,000 and $50,000
Between $50,000 and $70,000
$70,000 or more
N


Note: Data are from a Florida Voter survey of registered voters conducted in May 2004.
Missing values were replaced using multiple imputation.

who are least informed are more likely to be ambivalent across several policy domains.

They ask political knowledge questions in each study to measure information levels.

The present study also employs a measure of political knowledge. Respondents


were read the following introduction:


20.9 %
17.8 %
61.3 %
100.0 %


38.4 %
39.9 %
20.4 %
1.3 %
100.0 %


84.8 %
8.4 %
3.8%
3.0%
100.0 %


59.3 %
40.7 %
100.0 %


35.1 %
35.7 %
18.6 %
10.5 %
100.0 %









Here are a few questions about the government in Washington. Many people don't
know the answers to these questions, but even if you're not sure I'd like you to tell
me your best guess.

Then they were asked the following questions:


* First, do you happen to know what job or political office is currently held by John
Ashcroft?

* Who has the final responsibility to decide if a law is constitutional or not is it the
president, Congress, or the Supreme Court?

* Would you say that one of the parties is more conservative than the other at the
national level? If yes: Which party is more conservative?1

Dummy variables were created for each correct response (0 = incorrect, 1 = correct). Those

who responded with "don't know" were also counted as incorrect. Then an additive index

was constructed by adding the three together (c = 0.36). Higher values represent more

political knowledge.

The distribution of correctly answered questions is contained in Figure 3-6. Most of

sample is moderately informed. Only around 8 percent of the sample was unable to answer

any of the questions. 26.4 percent got one correct, 35.3 percent got two correct, and 30.3

percent got all three correct. This is a considerable amount of variance making this a good

measure of political knowledge.

Summary

The preceding chapter focused on the measures used in this study. All of the

specifics concerning the measurement of the dependent and independent variables used in

this study have been explained. This included four different approaches to measuring

ambivalence about social welfare policy. The distribution across the primary indirect


11 "No" responses to the first part of the question were coded as incorrect.

















E None Correct
0 One Correct
o Two Correct
O Three Correct


30.3 %









35.3 %





Note: Data are from a Florida Voter survey of registered voters conducted in
May 2004.

Figure 3-6 Political Knowledge Distribution

measure used here suggests that many are ambivalent across a range of social welfare

issues, including government spending on medical insurance, assistance for the homeless,

education, programs to improve the standard of living for poor Americans, providing jobs

for citizens, child care programs, and finally, retirement benefits. The potential sources of

social welfare ambivalence have also been defined theoretically and operationally. The

distributions across all of these measures indicate that many are conflicted, whether it is

between cognitive or affective elements.






65


In the next chapter, the relative construct validity of each of the measures of

ambivalence is compared and the results will suggest that the measure adapted from the

indirect approach used by Craig and his colleagues (2002, 2005) is superior. The findings

also suggest that all of the above-described potential sources of social welfare

ambivalence are, in fact, reliable predictors of such.














CHAPTER 4
A TEST OF MEASUREMENT VALIDITY: THE SOURCES OF AMBIVALENCE
ABOUT SOCIAL WELFARE

The idea that people's attitudes about issues are a mix of considerations that may

result in ambivalence is becoming accepted across disciplines. The problem today is that

survey research is still treating attitudes as if they are uni-polar. Simply, standard survey

indicators do not offer any systematic way of separating those who have both positive

and negative evaluations of a given object from those who have a clear position. This

chapter evaluates the direct and indirect measures of social welfare ambivalence that

were described in the previous chapter. The findings indicate that the indirect measure of

social welfare ambivalence that forces respondents to rate their positive and negative

feelings separately performs better than any of the other approaches currently offered in

the literature. A fully specified model of social welfare ambivalence is presented

following this analysis.

While survey researchers have begun to accept that ambivalence is a standard

attribute of attitudes (Feldman and Zaller 1992; Zaller 1992; Alvarez and Brehm 1995;

Craig, Kane, and Martinez 2002), they have not perfected a way to capture this

phenomenon. Our measures have not caught up with theory. In addition, there is

considerable disagreement concerning both how we conceptualize ambivalence and how

we measure it. If ambivalence is a standard attribute of attitudes such as attitude

importance (Krosnick and Abelson 1992; Krosnick 1988a, 1988b; Boninger, Krosnick,

Berent 1995), intensity (Krosnick and Abelson 1992; Krosnick et al. 1993), extremity and









certainty (Alvarez and Brehm 1995), and these attributes are consequential to our

understanding of attitudes in general (see Bassili 1996 for a review), it seems logical that

we should settle on an approach to measure the attribute of ambivalence. We have

standard approaches in survey research to capture these other phenomena, so it seems

imperative to our understanding to develop such a measure for ambivalence.

If standard indicators used in the major surveys utilized in the discipline (American

National Election Studies, General Social Survey, etc.) offer no systematic way of

distinguishing those who are ambivalent from those who are not, it is likely that the

results of many of the studies that used these data are questionable, especially when it

comes to those that focus on attitudes about issues where ambivalence may be prevalent.

For example, let's assume that social welfare ambivalence is widespread and that social

welfare policy preferences are related to candidate evaluations. If a model of these

evaluations accounts for policy preferences but not for ambivalence about the policy, the

estimate for the effect of policy preferences will be biased. The relationship may appear

to be stronger or weaker than it actually is. The exclusion of ambivalence does not mean

that previous research results are completely without merit, but it certainly suggests that

some studies may need fine-tuning.

As described in Chapters 1 and 2, researchers have employed direct measures of

ambivalence that simply ask respondents in some way to state whether they are torn

between sides on the issue at hand and indirect measures of ambivalence that employ

some means to ascertain the degree to which individuals are ambivalent without making

respondents aware that this phenomenon is being measured. Prior research indicates that

the correlations between direct and indirect measures of ambivalence are modest in









magnitude (Newby-Clark, McGregor, and Zanna 2002; Priester and Petty 2001). This

implies that they are capturing distinct phenomena. The direct measures may be capturing

a generally "unsure" feeling, which in common vernacular is often described as

ambivalence. If ambivalence is a standard attribute of attitudes such as direction or

intensity, then it is just the way thoughts are organized generally rather than an

individuals subjective experience.

This chapter discusses the strengths and weaknesses of these two approaches. It

offers an empirical test of the validity of each using the four previously described direct

and indirect measures of ambivalence (Chapter 3). If a measure is valid, or at least has

construct validity, the theoretical correlates of the measure should serve as reliable

predictors. Accordingly, this chapter compares how well the theoretical sources of

ambivalence including cognitive conflict, affective conflict, and cognitive-affective

conflict predict each measure of ambivalence. This test has two obvious assumptions:

first, that theory suggesting that these are sources of ambivalence is accurate, and second,

that the measures of these concepts are valid. As for the first assumption, theory in these

regards is fairly developed, it is just that measures that have not caught up. Concerning

the validity of the measures, the indicators used were adapted from NES and GSS, and

research has repeatedly demonstrated their efficacy when it comes to predicting

dependent variables that theory suggests it should. If we accept this as a test of their

validity, it seems a reasonable test to compare the validity of the measures of

ambivalence employed here.

The findings indicate that the theoretical sources of ambivalence perform best as

predictors of the indirect measure that has respondents rate positive and negative feelings









separately. While this indirect measure may be the most valid, the present study contends

that it is not the most practical because of financial cost. Clearly more work needs to be

done, but the evidence presented here suggests that this is an endeavor that needs

undertaking. Before moving on to the analysis, direct and indirect approaches are

critiqued in more detail than previous chapters. This chapter concludes with the

presentation and discussion of a fully specified model of social welfare ambivalence

using the most valid indirect measure.

Critique of Previous Approaches to Measuring Ambivalence

Direct Measures

Direct measures of ambivalence require respondents to state their direct feelings.

For example, respondents may be asked whether their attitudes are one-sided or mixed or

whether they agree with statements like "I have positive and negative feelings about....."

For example, Priester and Petty (1996, 2001) asked respondents1 to complete a series of

10-point scales designed to assess the extent to which their reactions were conflicted,

mixed, and indecisive with respect to the attitude objects under observation. Using a

typical large-sample survey, Mulligan and McGraw (2002) also employed a direct

measure of ambivalence using the following indicator:

Some people feel that there are only good things or bad things about this issue (a.
government wiretapping, b. social welfare spending). Their feelings are consistent.
Other people feel that there are both good things and bad things about this issue.
Their feelings are inconsistent. Thinking about your own views, would you say that
your feelings about this issue are extremely consistent, very consistent, somewhat
consistent, somewhat inconsistent, very inconsistent, or extremely inconsistent?


1 Their study was based on an experiment but the booklets participants were asked to
complete are comparable to surveys. Of course the sample is different than that in a
typical survey (completely random), but as far as question wording goes, it was set up
like a survey.









While other researchers have used slightly different wording, these two examples

exemplify the direct approach, at least broadly.

The strengths of the direct approach are, first, that it is practical. It requires one

simple question that allows respondents to state if they are conflicted, mixed, etc. In

practice, we are often faced with trade-offs in survey research because of financial and

research limitations. We can only have so many questions on any given instrument. With

the direct approach, there is no need to have multiple questions gauging positive and

negative responses separately, thus, it is the least costly method.

On the other hand, the direct approach treats ambivalence as a subjective

experience rather than as an attribute of attitudes in general. If ambivalence is an attribute

of attitudes in general, then perhaps respondents are not in the best position to make a

diagnosis. A medical doctor asks patients what their symptoms are and then makes a

diagnosis. The direct approach essentially asks respondents to diagnose themselves. This

is probably not the best way to capture the phenomenon of an individual simultaneously

possessing positive and negative evaluations of a single attitude object (Newby-Clark,

McGregor, and Zanna 2002; Armitage and Connor 2000; Cacioppo, Gardner, and

Berntson 1997; Priester and Petty 1996; Craig, Kane, and Martinez 2002; McGraw,

Hasecke, and Conger 2003; Albertson, Brehm, and Alvarez 2005; Gainous and Martinez

2005). Allowing respondents to assess the degree to which they possess these evaluations

may allow extraneous factors to distort the measure. Bassili (1996) compares the validity

of indirect and direct measures of attitude strength2, arguing the indirect measures have

more predictive validity. He suggests that two realms of psychological functioning exist:


2 Again, the author refers to these as operative and meta-attitudinal measures.









one comprised of operative psychological processes and the other comprised of

impressions of these processes. So, direct measures may be picking up people's reaction

to feeling ambivalent rather than the actual phenomena. If reactions vary, then the direct

measure is capturing something other than the phenomena itself. Rather, it may be the

byproduct of ambivalence.

Indirect Measures

Indirect measures of ambivalence involve some method of gauging the extent to

which ambivalence exists without depending on an individual's personal assessment.

While there have been several strategies employed, they all share this common thread.

The big difference from early attempts in political science as compared to later attempts

is that they involved the subjectivity of the researchers. For example, Feldman and Zaller

(1992) asked survey respondents to state whatever thoughts came to mind as they

answered two traditional closed-ended policy questions. Then they measured

ambivalence by counting the number of conflicting considerations, spontaneous

statements of ambivalence, and two-sided remarks (i.e., "Although I think X, I

nevertheless favor Y"), finding strong support for the presence of ambivalence in many

of the respondents. This can be thought of as an indirect measure of ambivalence because

respondents were not asked to assess their own degree of ambivalence, but this approach

certainly required the direct assessment of the researchers. Of course there are always

direct decisions made in research, but perhaps some are more consequential than others.

As described in Chapters 2 and 3, Alvarez and Brehm (1995, 1997, and 1998) and

Jacoby (2002) employ another variation of the indirect approach. They inferred the

presence of ambivalence in respondents' attitudes from patterns of error variance in

heteroskedastic probit and ordinary least squares regression models respectively. This









measurement strategy involves analyzing the residuals of a model of the attitude object

under examination. Residuals are the difference between the value for each respondent

predicted from the probit equation and the actual observed response of that individual.

They contend that the variance of these residuals should be higher among those who are

ambivalent. Essentially, Jacoby and Alvarez and Brehm each test to see if the residual

variance is higher among those whom theory suggests should be ambivalent (e.g., those

with values in conflict).

Also previously discussed, others have used another indirect approach to measuring

ambivalence where respondents are asked to rate how positively they feel toward an

attitude object and then asked separately to rate how negatively they feel toward the

object (Craig, Kane, and Martinez 2002; Craig, Martinez, and Kane 2005; Gainous and

Martinez 2005; Martinez et al. 2005). The responses are then combined via a

mathematical algorithm yielding an ordinal measure of ambivalence.

As for the strength of all three of these indirect approaches to measuring

ambivalence, they all treat ambivalence as an attribute of attitudes in general as opposed

to a subjective experience. In this case, attitudes are a mix of multiple considerations

(Zaller 1992), and the Feldman and Zaller approach as well as the strategy used by Craig

and his colleagues give respondents the opportunity to explicitly express these multiple

considerations. Alvarez and Brehm clearly separate the respondent's direct assessments

of how mixed they are from the measurement. But they do not give them the opportunity

to rate, simultaneously, how positively and negatively they feel toward the object at hand.

This is where the approach used by Craig and his colleagues is the strongest. The









operational definition they use most closely fits the conceptual definition offered in the

psychology literature.

Concerning the weaknesses of these indirect approaches, we can address each

individually. Concerning Feldman and Zaller's approach, while attitudes are a mix of

considerations, the possession of "opposing considerations" may in fact indicate the

presence of factors such as equivocation (when someone is trying not to make a bad

impression on the interviewer), uncertainty (when they are unsure of which side of the

issue to choose), informedness (where the respondent has sufficient information to cite

both sides evenly while clearly favoring one or neither), or nonattitude (where the

respondent has no real position on the issue), or the questions may be ambiguous making

them insufficient as indicators of preference (see Alvarez and Brehm 1995, 1997, 1998).

Therefore, the possession of opposing considerations is not necessarily a product of or

representative of ambivalence. Remember that ambivalence defined as an attribute of

attitudes refers to conflict about a single attitude object. Expressing conflicting

viewpoints that are not explicitly about the same object is not ambivalence.

Alvarez and Brehm (1995) and Jacoby (2002) define ambivalence appropriately but

use a measure that does not accurately represent their conceptualization. Inferring

ambivalence from patterns of residual variance is problematic on two levels. First,

ambivalence is an individual-level concept (ambivalence) and their inferences are

essentially based upon aggregate-level data (error variance in a predictive model across

groups). Second, these measures fail to distinguish ambivalence from the very things

Alvarez and Brehm suggest are problematic with the Feldman and Zaller study

(equivocation, uncertainty, informedness, and nonattitude). Error variance could be a









result of these factors. They control for levels of political knowledge to account for such,

but this approach still requires more assumptions than the other indirect approach.

There are a couple of obvious weaknesses with the measure used by Craig and his

colleagues as well as Gainous and Martinez. The first problem is that the question is

rather complicated which may confuse respondents. The next problem is a practical one.

For each attitude object, they ask two questions. Surveys are expensive and if it is

necessary to ask two questions for every one attitude object the cost will potentially

double. This may limit the number of objects or issues that can be included. Also,

because it makes the survey longer, response rates would likely significantly drop off.

Aside from this practical problem, this approach seems to be the closest representation of

ambivalence, properly defined.

Analysis

Empirically Comparing Measurement Approaches

Before moving on to the models that compare the performance of each of the direct

and indirect measures of ambivalence, let us take a look at the relationship between each.

Table 4-1 contains the zero-order correlations between each of the measures of social

welfare ambivalence (detailed in Chapter 3). There is not a strong relationship between

any of these measures. There is a very weak positive relationship between the two

indirect measures (0.13). There is a significant correlation between the direct measures

and the residuals, but this relationship is suspect because the residuals and the direct

measures are based on the same indicator. There was no other measure in these data.

Nonetheless, it appears that these direct and indirect measures certainly are not all

capturing ambivalence. These results mirror the findings of previous research that

suggested there was not a strong relationship between direct and indirect ambivalence









(Priester and Petty 2001; Mulligan and McGraw 2002; Newby-Clark, McGregor, and

Zanna 2002), but also demonstrates that there is not much of a relationship between the

two indirect measures. It appears that each is capturing a distinct phenomenon.

Table 4-1 Correlations across Direct and Indirect Measures of Ambivalence
Direct Direct Indirect Indirect
"Torn" "Volunteered" Residuals Pos/Neg

Direct "Torn" 1.00
Direct "Volunteered" -- 1.00
Indirect Residuals 0.31* 0.23* 1.00
Indirect Pos/Neg 0.06 -0.08 0.13* 1.00


Note: Data are from a Florida Voter survey of registered voters conducted in May 2004.
The two direct variables cannot be correlated because they are each part of the same split-
sample indicator. Table entries are Kendall's Tau b correlation coefficients. 2-tailed test *p
< 0.05.

Table 4-2 contains the zero-order correlations between the potential sources of

ambivalence (detailed in Chapter 3) and each measure of social welfare ambivalence

evaluated here. First, the indirect measure that had respondents rate their positive and

negative feelings separately is most strongly correlated to each of source of conflict.

There is a significant positive relationship between each source of ambivalence and this

indirect measure. There is also a positive relationship between cognitive conflict and the

indirect residuals measure as well as the offered "torn" response direct measure. Oddly,

there is a significant negative relationship between cognitive-affective conflict and the

offered "torn" response direct measure. There are no significant relationships between

sources and the "volunteered" direct measure. These preliminary findings suggest that the

indirect measure has the highest construct validity. This makes sense, again, because it is

operationalized in a way that is consistent with the psychological conceptualization of

ambivalence (simultaneous possession of positive and negative evaluations). So, in this









case, face validity leads to construct validity. Researchers have used this

conceptualization to build theories about the potential sources of ambivalence. So

perhaps, this measure is better suited to test the existing theory (conflict of cognitive and

affective elements). If ambivalence is properly defined as simultaneously possessing

positive and negative feelings about social welfare, then the theoretical sources of such

ambivalence should predict the measure.

Table 4-2 Correlations between Sources and Four Ambivalence Measures
Direct Direct Indirect Indirect
"Tor" "Volunteered" Residuals Pos/Neg

Cognitive Conflict 0.06 0.17* 0.05 0.20*
Cognitive-Affective Conflict -0.03 -0.13* 0.08* 0.46*
Affective Conflict -0.00 -0.09 0.06* 0.55*


Note: Data are from a Florida Voter survey of registered voters conducted in May 2004.
Table entries are Kendall's Tau b correlation coefficients. 2-tailed test *p < 0.05.

Before moving to a multivariate analysis of each of the measures, the approach to

analyzing the indirect residuals measure employed by Alvarez and Brehm (1995, 1997,

1998, and 2002) is replicated. This involves comparing the variance across levels of

cognitive, cognitive-affective, and affective conflict. The idea is that the variance should

be higher among those with more conflict. Again, the problem with this approach is two-

fold. First, the measure of ambivalence is error variance. This variance could be high for

reasons other than ambivalence (non-attitude, uncertainty). Second, the analysis only

gives us aggregate inferences and ambivalence is an individual-level phenomenon. The

latter problem is resolved later by modeling the individual-level residuals. Nonetheless,

this approach has been used; so replicating it may verify that this approach is

problematic.









For this analysis a dummy variable is created for each source of ambivalence that

represents above-average conflict (0 = below mean, 1 = above mean). Next, the between-

group variance is analyzed. A one-way analysis of variance (ANOVA) is done testing for

the homogeneity of variance across groups (below mean, above mean). This involves

getting a Levene statistic (the test statistic for a homogeneity of variances test) and the

associated probability. This probability tells us the likelihood with which we can

confidently reject the null and accept the hypothesis that there is higher error variance

when predicting attitudes about social welfare for those with higher levels of conflict as

opposed to those with lower levels.

The results of the analysis in Table 4-3 indicate that, other than cognitive conflict,

the sources of ambivalence are not strong predictors of the error variance from the model

of attitudes about social welfare in Chapter 3 (Table 3-2). For that matter, the effect of

cognitive conflict is in the opposite of the expected direction. These findings suggest that

those who are less conflicted when it comes to values are actually more ambivalent.

Further, the variance is higher for those with below average cognitive-affective and

affective conflict but this difference is not significant. If anything, this analysis tells us

that the residuals are not a good measure of ambivalence (assuming that these measures

of conflict are sources of ambivalence). This just confirms the results presented in Table

4-2.

It is important that we look at the relationship between the sources of ambivalence

before moving on to the multivariate analysis of the residuals and other measures of

ambivalence. Because each of these measures uses overlapping items to create scales,

multicollinearity is potentially an issue. If these scales correlate highly, they contribute









Table 4-3 Analysis of Residuals from Model of Attitudes about Social Welfare
Variance Levene
of Residuals Statistic Probability

Comparison 1
Above-average Cognitive Conflict 1.90
7.05 0.01
Below-average Cognitive Conflict 2.29

Comparison 2
Above-average Cognitive-Affective Conflict 1.99
0.00 0.99
Below-average Cognitive-Affective Conflict 2.02

Comparison 3
Above-average Affective Conflict 1.90
1.39 0.24
Below-average Affective Conflict 2.11


Note: Data are from a Florida Voter survey of registered voters conducted in May 2004.
The first column gives the within-group variances of the Ordered-Logit equation in Table
3-2. The second column gives the Levene statistic (the test statistic for a homogeneity of
variances test). The observed probability value from this test is in the third column.

redundant information and can cause other variables to appear to be less important than

they really are. The scatterplots contained in Figure 4-1 of cognitive-affective conflict

with cognitive conflict and then cognitive-affective conflict with affective conflict

suggest there is a relationship between the cognitive-affective conflict scale and the

affective scale but not with cognitive conflict scale. The results contained in Table 4-4

confirm the significance of these relationships. As the scatterplots suggested, the only

concern is the relationship between cognitive-affective and affective conflict (Tau b

0.62, p < 0.05).

To address the problem of multicollinearity in all of the models constructed in the

remainder of this study, two separate models will be estimated for each dependent

variable- one with cognitive-affective conflict, and another with affective conflict.





















-=


S060-




S040-


0

020-




000-










100-




080-
-*r

0
U
S060-


<

040-


0

020-




000-


0
0
0
0 0
0
0
0
0


O O
O
oO
O 0
0 o
O
0 0


0
0
0
0
0
0
0
O


0
0
0
0
oo


0
0



0
0



O
0


0
0


o
o o o





o0g 0 0 0
o0 0 0 0
0o o0
O0 0





0 0 0 0
0 00 0

0 0 0
0 O
O o
Ooo O

o o o

0 0
0 0


O O O
O O
0 0
0 0
0


0


0
0
0
O


I I I I I I
000 020 040 060 080 100
Cognitive Conflict






0

o oo
0 0

S o


0 00 I 8 0o 0

0 0 0 o 8
o 8 o o



0 0 0 0
80
8 8 o o


0



8 0


I I I I I I
000 020 040 060 080 1 00
Affective Conflict


Note: Data are from a Florida Voter survey of registered voters conducted in May 2004.


Figure 4-1 Scatterplots of Overlapping Sources









Table 4-4 Correlations between Sources of Ambivalence about Social Welfare
Cognitive Cognitive-Affective Affective Conflict
Conflict Conflict

Cognitive Conflict 1.00
Cognitive-Affective Conflict 0.04 1.00
Affective Conflict 0.07* 0.62* 1.00


Note: Data are from a Florida Voter survey of registered voters conducted in May 2004.
Table entries are Kendall's Tau b correlation coefficients. 2-tailed test *p < 0.05.

Concerning the present models that are used to compare the relative construct validity of

the four ambivalence measures, they are as follows:

Model 1

Social Welfare Ambivalence = a + P1 Cognitive Conflict + p2 Egalitarianism
Importance + 03 Individualism Importance + 04 Relative Importance of Values + s5
Cognitive-Affective Conflict + 37 Female + s8 Black + Income + 39 Political
Knowledge + e

Model 2

Social Welfare Ambivalence = a + P1 Cognitive Conflict + 02 Egalitarianism
Importance + 13 Individualism Importance + 04 Relative Importance of Values + s5
Affective Conflict + 37 Female + s8 Black + Income + 39 Political Knowledge + e

While this prevents the simultaneous estimation of the relationship between cognitive-

affective conflict, affective conflict, and social welfare ambivalence, it does resolve the

problem with multicollinearity.

By looking at how well these models perform across all four measures of

ambivalence, we get a better test of construct validity than the bivariate relationships

presented above (Table 3-2). Tables 3.5 and 3.6 contain the results of the equations that

model each of the four measures of ambivalence as a function of the potential sources of

ambivalence with a set of controls. The direct models where respondents "volunteered"

mixed responses perform better than those where they were given the "torn between









Table 4-5 Comparing Direct Ambivalence Measurement Approaches

Tom Volunteered
Model 1 Model 2 Model 1 Model 2

Cognitive Conflict 0.77 0.73 2.13* 2.16*
(0.56) (0.56) (0.78) (0.79)
Cognitive-Affective Conflict 0.62 -- 3.41 --
(0.33) -- (0.34)
Affective Conflict -- 0.35 -- 3.92*
-- (0.28) -- (0.30)
Egalitarianism Importance -0.10 -0.10 -0.37* -0.25*
(0.09) (0.09) (0.09) (0.09)
Individualism Importance 0.12 0.17 -0.38* -0.28*
(0.09) (0.09) (0.09) (0.09)
Relative Difference 0.04 0.04 -0.05 -0.01
(0.18) (0.18) (0.22) (0.22)
Black -1.12 -1.12 0.86 0.84
(0.65) (0.65) (0.60) (0.60)
Female 0.83* 0.83* -0.26 -0.28
(0.30) (0.30) (0.38) (0.38)
Income 0.02 0.02 0.33 0.33
(0.15) (0.15) (0.20) (0.19)
Political Knowledge -0.21 -0.20 -0.16 -0.16
(0.15) (0.15) (0.20) (0.20)
Constant 1.91 2.02* 0.93 0.93
(1.00) (1.01) (1.37) (1.37)

-2 log likelihood 317.76 318.15 200.40 202.66
Nagelkerke Pseudo R2 0.08 0.08 0.17 0.16
N 304 304 303 303


Note: Data are from 2004 Florida Voter survey. Table entries are logit estimates. 2-tailed
test *p < 0.05. Standard errors are in parentheses. Missing values were replaced using
multiple imputation

sides" option (Table 4-5). Cognitive conflict and egalitarianism importance are

significant and in the expected direction in the volunteered model, but cognitive-affective

conflict is significant and not in the expected direction. The findings here suggest that

ambivalence goes down as this type of conflict goes up, ceterisparibus.









If cognitive, cognitive-affective, and affective conflict are indeed sources of social

welfare ambivalence, this direct measure must be capturing something other than

ambivalence (assuming the measures of conflict are valid). Basilli (1996) might suggest

that the direct measures are capturing people's direct reaction to possessing positive and

negative evaluations about social welfare. Obviously these reactions are not correlated

with the sources of ambivalence, or the model would pick that up. In the "torn between

sides" response model nothing is significant. In sum, the lack of findings across the direct

measures suggests that they are not good measures of ambivalence.

There are no significant estimated effects (Table 4-6) in the model of the

residuals. These findings suggest that this is a poor measure of ambivalence. The final

model that uses the combined positive and negative evaluations performs the best.

Cognitive conflict, value importance and the relative difference, cognitive-affective

conflict, affective conflict, race, and gender are all significant and in the expected

direction (Table 4-6). This makes sense if we think of ambivalence as an attribute of

attitudes. If ambivalence is the simultaneous possession of positive and negative

evaluations of a single attitude object, then this measure makes the most intuitive sense.

Fully Specified Models of Social Welfare Ambivalence

Now that the most valid measure of ambivalence has been selected, this indirect

measure comprised of positive and negative evaluations is used for the remainder of the

study. This section presents fully specified models of ambivalence using this measure.

The models presented above did not include "general policy preferences" as an

explanatory variable because that indicator was used to construct the two direct









Table 4-6 Comparing Indirect Ambivalence Measurement Approaches


Indirect-Residuals


Cognitive Conflict

Cognitive-Affective Conflict

Affective Conflict

Egalitarianism Importance

Individualism Importance

Relative Difference

Black

Female

Income

Political Knowledge


-2 log likelihood
Nagelkerke Pseudo R2
N


Model 1

0.47
(0.30)
0.62
(0.33)


-0.10
(0.09)
0.12
(0.09)
-0.12
(0.09)
-0.19
(0.28)
-0.16
(0.15)
-0.08
(0.08)
-0.12
(0.08)

1759.00
0.03
607


Model 2

0.51
(0.30)


0.35
(0.28)
-0.10
(0.09)
0.17
(0.09)
-0.13
(0.09)
-0.18
(0.28)
-0.17
(0.15)
-0.08
(0.08)
-0.12
(0.08)

1762.48
0.03
607


Indirect-Pos/Neg
Model 1 Model 2


1.31*
(0.29)
3.41*
(0.34)


-0.37*
(0.09)
-0.38*
(0.09)
-0.17
(0.09)
-1.03*
(0.26)
-0.20
(0.15)
0.02
(0.08)
0.08
(0.08)

4239.78
0.29
607


1.52*
(0.29)


3.92*
(0.30)
-0.25*
(0.09)
-0.28*
(0.09)
-0.17*
(0.09)
-0.98*
(0.27)
-0.26
(0.15)
0.06
(0.08)
0.03
(0.08)

4164.25
0.37
607


Note: Data are from 2004 Florida Voter survey. Table entries are ordered-logit estimates
(threshold levels are not shown). 2-tailed test *p < 0.05. Standard errors are in
parentheses. Missing values were replaced using multiple imputation.

ambivalence measures. Therefore, it could not be included and still draw conclusions by

comparing results across models. The models are as follows:

Model 1

Indirect (Pos/Neg) Social Welfare Ambivalence = a + 01 Cognitive Conflict + 02
Egalitarianism Importance + 13 Individualism Importance + 04 Relative Importance
of Values + s5 Cognitive-Affective Conflict + 06 Social Welfare Policy Preference
+ 37 Female + s8 Black + 39 Political Knowledge + e









Model 2

Indirect (Pos/Neg) Social Welfare Ambivalence = a + 01 Cognitive Conflict + 32
Egalitarianism Importance + 03 Individualism Importance + 04 Relative Importance
of Values + s5 Affective Conflict + 06 Social Welfare Policy Preference + 07
Female + s8 Black + 09 Political Knowledge + e

The results contained in Table 4-7 are not much different from those in Table 4-6,

but the new variable (social welfare policy preference) is added. As the ambivalence

index is best seen as an ordinal variable, an ordered logit procedure is employed to

estimate the multivariate model. Table 4-7 shows that there are multiple sources of

ambivalence about social welfare policy among respondents in the sample. Much of the

existing literature focuses on the relevance of value conflict as a precursor to

ambivalence, and that is borne out in these findings. The positive and significant

coefficient on the cognitive conflict variable indicates that people who expressed higher

levels of both individualism and egalitarianism also exhibited higher levels of

ambivalence, ceterisparibus. That is hardly surprising, of course, in light of previous

research on ambivalence in other policy areas (including abortion, gay rights, and race).

We also see from Table 4-7 that value importance accounts for a portion of the

variation in social welfare ambivalence: Respondents who regarded either egalitarianism

or individualism as important were less likely to be ambivalent, which suggests the

importance one attaches to core values can sometimes block out conflicting feelings

about social welfare. This is especially so among those who score high in relative

importance (rating one of the values above the other); that is, a larger difference between

the importance attached to individualism and egalitarianism is negatively and

significantly associated with ambivalence, even when controlling for the levels of

importance accorded to the values themselves. Thus, while value conflict normally tends









Table 4-7 Multivariate Models of Social Welfare Ambivalence

Coefficient Standard Error 95 % Confidence Intervals

Model 1
Cognitive Conflict 1.31 0.29 0.73 1.88
Cognitive-Affective Conflict 3.39 0.34 2.73 4.05
Egalitarianism Importance -0.34 0.09 -0.51 -0.17
Individualism Importance -0.42 0.09 -0.59 -0.25
Relative Difference -0.17 0.09 -0.34 -0.00
Policy Preference -0.13 0.05 -0.23 -0.03
Black -0.98 0.26 -1.50 -0.46
Female -0.18 0.15 -0.47 0.11
Income 0.09 0.08 -0.13 0.18
Political Knowledge 0.07 0.08 -0.09 0.22

-2 log likelihood 4245.39
Nagelkerke Pseudo R2 0.30

Model 2
Cognitive Conflict 1.52 0.29 0.95 2.09
Affective Conflict 3.93 0.31 3.33 4.53
Egalitarianism Importance -0.23 0.09 -0.40 -0.06
Individualism Importance -0.33 0.09 -0.50 -0.16
Relative Difference -0.18 0.09 -0.35 -0.01
Policy Preference -0.15 0.05 -0.25 -0.05
Black -0.92 0.27 -1.44 -0.40
Female -0.24 0.15 -0.52 0.05
Income 0.06 0.08 -0.09 0.21
Political Knowledge 0.01 0.08 -0.15 0.17

-2 log likelihood 4180.05
Nagelkerke Pseudo R2 0.38
N 607 607 607 607


Note: Data are from a Florida Voter survey of registered voters conducted in May 2004.
Table entries are ordered logit coefficients, associated standard errors, and 95%
confidence intervals; threshold levels are not shown. Missing values were replaced using
multiple imputation.

to heighten ambivalence, ambivalence becomes less likely to occur when one value is


held more dearly than the other.









The estimates for cognitive-affective and affective conflict are both significant,

demonstrating a positive relationship between these sources and social welfare

ambivalence (Table 4-7, models 1 and 2 respectively). In fact, the magnitude of these

effects is larger than any other across both models. The estimate for affective conflict

(3.93) is larger than that for cognitive-affective conflict suggesting that affective elements

are a predominant source of attitudes about social welfare (see Table 3-2) and

ambivalence about social welfare. This is consistent with literature that suggests feelings

about the perceived beneficiaries is one the best predictors of attitudes about social

welfare (Kinder and Winter 2001; Nelson 1999; Sniderman, Brody, and Tetlock 1991;

Sniderman and Piazza 1993; Cook and Barrett 1992; Bobo and Kluegel 1993; Gilens

1995; see also Jacoby 2005).

Ambivalence also is asymmetric with respect to policy preferences. The significant

and negative coefficients in Table 4-7 indicate that respondents who are more supportive

of higher levels of government services and spending tend to be less ambivalent about

social welfare policy as a whole. Since ambivalence is related to attitudinal pliability

(e.g., Craig, Martinez, and Kane 2005), this finding has the important political

implication that conservatives may be more likely than liberals to be "talked out of" their

general opposition to social welfare in specific circumstances and conditions. Whereas

Feldman and Zaller (1992) maintained that social welfare liberals are more conflicted

than conservatives, hence more ambivalent, these results suggest that (controlling for

value importance and value conflict), liberals, measured as those with liberal social

welfare preferences are actually less torn between the pros and cons of social welfare

policy (this is explored further in Chapter 5).









Finally, gender, income, and political knowledge are not significant. The

insignificance of these variables does not mean that there are not patterns of ambivalence

across them, but rather, it may mean that other variables in the model do a better job of

explaining the variance explained by these variables. This proposition is explored in the

next chapter. Concerning race, the coefficient indicates that black respondents tend to be

significantly less ambivalent than non-blacks, ceterisparibus.

Summary

Chapter 4 has accomplished two goals. First, it assessed the construct validity of

each of the different measures of ambivalence by comparing how well the theoretical

sources of ambivalence predict each respectively. The findings indicate that the approach

adapted from earlier work (Craig, Kane, and Martinez 2002; Craig et al. 2005) that forces

respondents to rate their positive and negative feelings separately is the most valid. In this

test, the sources of such ambivalence were also explored. The evidence presented in this

chapter also provides support to previous research that suggested ambivalence may result

when individuals have conflicting thoughts or beliefs (cognitive conflict), conflicting

feelings (affective conflict), or beliefs in conflict with feelings (cognitive-affective

conflict). In addition value importance, policy preferences, and race all shape social

welfare ambivalence.

After settling on the best measure, the next chapter revisits previous research that

addresses the question of who is most likely to be ambivalent among liberals and

conservatives, among men and women, and across race. The findings presented there

challenge previous assertions that liberals are more likely to be ambivalent about social

welfare (Feldman and Zaller 1992; see also Jacoby 2002). The results indicate that self-









identified liberals, females, and African Americans are less ambivalent about social

welfare policy than their respective counterparts. Chapter 5 builds on the findings of the

present chapter to sort out the underlying cause of these differences. It argues that we

can get a better understanding of why members across these groups would be more or

less likely to be ambivalent about social welfare by teasing out the differences across

these groups when it comes to the potential sources of ambivalence. The findings indicate

that the variance across ideology and gender can be accounted for by the sources of

ambivalence discussed above, but they do not explain why African Americans are less

ambivalent than their counterparts. Some conjecture is offered as to why this may the

case.














CHAPTER 5
PREVALENCE: SOCIAL WELFARE AMBIVALENCE ACROSS IDEOLOGY,
GENDER, AND RACE

When we talk about liberals and conservatives in the United States, social welfare

policy is often at the center of the discussion. The popular conception and empirical

evidence suggests that liberals generally support welfare programs while conservatives

generally do not (see Cook and Barrett 1992; Jacoby 1991). This broad generalization

may be accurate, but research has also suggested that social welfare ambivalence varies

across ideology (Feldman and Zaller 1992; Steenbergen and Brewer 2000; Jacoby 2002).

If this is the case, and females (Gilens 1988; Kaufmann and Petrocik 1999; Goren 2001)

and African Americans (Kinder and Winter 2001; Gilens 1995; Tate 1994; Bobo and

Kluegal 1993) tend to be social welfare liberals, we should expect ambivalence to vary

for them as well. The specific questions addressed in this chapter are, first, who is more

torn when it comes to this controversial issue- liberals or conservatives, males or females,

blacks or whites? And second, why?

Previous literature has offered varying descriptions of the distribution of social

welfare ambivalence. Feldman and Zaller (1992) pioneered exploration into the varying

prevalence of social welfare ambivalence across ideology. They argue that conservatives

are less ambivalent about social welfare because liberals are more likely to experience

value conflict, or what this study refers to as cognitive conflict. They contend that

conservatives prioritize individualist over egalitarian values while liberals place roughly

equal importance on each. As a result, the values of a liberal are more likely to come into









conflict, stimulating ambivalence about social welfare. These arguments are contrary to

findings presented here. The argument is extended here to include differences concerning

feelings about the perceived beneficiaries (cognitive-affective and affective conflict)

suggesting that Feldman and Zaller did not get it right in their explanation of the

difference in the sources of ambivalence across ideology. Their measure of ambivalence

is suspect (see Chapter 2 and 4), they missed important sources of ambivalence about

social welfare, and finally, their characterization of liberals' and conservatives' values

was inaccurate. As a result of these reasons, their conclusions are incorrect when it comes

to who is most ambivalent.

Jacoby (2002) also finds support for the contention that conservatives are less

ambivalent than liberals. On the other hand, he does not find any significant differences

regarding the individual ranking of values across ideology. Contrary to the findings here

and of Feldman and Zaller (1992), he suggests that social welfare ambivalence is not

common (see also Jacoby 2005). In addition, he fails to look for differences concerning

feelings about the perceived beneficiaries (cognitive-affective and affective conflict). His

measurement approach is also less than adequate (see Chapter 2 and 4).

Steenbergen and Brewer (2000) do include all of these sources of ambivalence and,

contrary to Feldman and Zaller as well as Jacoby, argue that liberals are less ambivalent

than conservatives. They also suggest that such ambivalence is not widespread. On the

other hand, they contend that liberals are less likely to experience conflict between

cognitive items such as values (cognitive conflict), affective items such as positive and

negative feelings about the perceived beneficiaries (affective conflict), and conflict