STRATEGIES OF SAFETY AND HEALTH AMONG CITRUS FARMWORKERS IN FLORIDA: A CULTURAL AND SOCIAL NETWORK APPROACH By JOSE ANTONIO TOVAR AGUILAR A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PAR TIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 201 4
Â© 2014 Jose Antonio Tovar Aguilar
To Serena y Paulo
4 ACKNOWLEDGMENTS I would like to thank the man y individuals and organizations that made this work possible. First and foremost, I must thank the University of Florida and all the professors who made it possible: the chair of my supervisory committee, Christ opher McCarty , offered constant support and t ir eless mentoring. Similar encouragement and guidance I received from m y supervisory committee members: Paul Monaghan, Fritz Roka , and Maria Stoikova. Special gratitude is also extended to Russ Bernard , Alan Burns, Adrian Felix, Joan Flocks , Linda McCauley , Kelli McCormack Brown , Sharleen Simpson, James Stansbury, and Neil Sullivan . Financially, this work could not be possible without the continued support of the CDC funded Florida Prevention Research Center (FPRC) at the College of Public Health in the Un iversity of South Florida and her Co Director, the Anthropologist Carol Bryant. She is the heart behind all the exemplary work that the FPRC has done; but I cannot ignore the many other members of the FPRC who have made my work much easier and effective: R icardo Contreras, Mahmooda Khaliq, Jason Lind, James Lindenberger, Juan Luque, Robert McDermott, Tony Panzera, Leah Phillips, and Tali Schneider. No less important is the support and guidance of the Farmworker Association of Florida, the Farmworkers Self H elp Inc., and the Immokalee Lions Club members: Ana Alamillo, Holly Baker, Elvira Carbajal, Juan Concha, Marcos Crisanto, Bertin Cruz, Jeannie Economos, Yolanda Gomez, Claudia Gonzalez, Adan Labra, Maru Martinez, Macarena Moraga, Tirso Moreno, Margarita Romo, Sister Gail, Victor Grimaldo, the William TreviÃ±os, and particularly Ellvia Zamora. Having a network of academics, industry, and community based organizations open the possibility of multiple research
5 locations and entry levels. I am very grateful to all of them. Nevertheless, abov e all past recognitions is my debt with the citrus harvesters who opened their homes and workplaces to me so I could learn first hand of their living conditions. First, I have to honor the memory of Santos de la Cruz and Quirino Velazquez , citrus harveste rs, colleagues, and friends . On May 16 , 2008 in Pasco County, these two migrant workers were gunned down outside Resurrection Park in the Dade City area known as Tommytown. I had trained these men as health care promoters for two years in a row for the Par tnership for Citrus Workers Health (PCWH); they left broken families, and nobody was ever persecuted. You will never be forgotten. All farmworkers were always kind and open with me, but I have to recognize the pickers who went beyond sharing their knowledg e to become my colleagues and ile sharing their lives with me, deserved my public gratitude. I also owe some officials of the citrus companies who collaborated with the PCWH a big thank you. They provided access to large crews for training and services, fund PPE distribution, and offer ed information for this work. The members of the citrus industry who embrace safety as a fundamental principal are from Evans Harvesting: Mike Wide and Horacio Garza; from US Sugar Southern Garden: Andy Esposito, Armando Gonzalez, A.J. Lozano, and Jack Men diburo; from Statewide Harvesting: Adam Patel and Omar Ruiz; and finally from Fellsmere, I had the invaluable support of Cesar Gomez, an honorable, wise and, hard working chivero .
6 My family and friends were the basic support on this journey. But nobody des erves more honor than Rosana Resende, my wife. She has allowed me to get where I am; we both had to work hard for this. She is a caring mother, a natural intellectual, and a modest muse. She is the dynamo of the family, and the perfect companion. I am very lucky to be her husband -not just because of her, but also for her immediate and extended family, especially her parents Maria and Eduardo. To my children who inspire every moment of life, this work is a small token of the endless months I was away from them. The list of UF friends to whom I owe my gratitude is extensive, and I am sorry if for some reason I have forgotten a name or two: Tit a Alvira, Faith Amon, Luis Arreola, Roberto Barrios, James Ciaravino, Felipe Cuevas, Antonio de la Pena, Ana Alice Eleuterio, Amanda Holmes, Pauline Kulsta d, Flavia Leite, Leonardo Martinez, Asmeret Mehari, Maria Morera, Mutsuo Nakamura, Alicia Peon, Ana Cristina Puentes, Leonardo Rada, Kathleen Ragsdale, Debora Rothman, Tracy Van Holt, Rodrigo Vergara, and Richard Wallace. One UF student organization in particular helped me keep my sanity: Mexicans in Ga inesville, I am indebted to all the people in this group. Finally, som e non UF friends in town who I cannot forget to thank are Gary Morrison, Pablo Solano, James Taylor and Perry Trenka, they were important in different parts of this journey that is now ending. In Mex ico, my genuine gratitude and love to my parents, Juan R. Tovar and Herlinda Aguilar M., is endless. Similarly for my admired grandpa Don Max and grandma DoÃ±a Lala . To my siblings, Leonardo and Linda, their significant others Marta
7 and Jaime, their childre n Laura, Mariana, Leo Leo, and Ayrton as well as to my cousins Jorge , Fito , Maxi, Cesar , and Helena. Your support and example was always important. Kruse: I could not have gotten this far without their kindne ss support and wisdom. My large family friends in Mexico could not be forgotten neither for their intellectual c on tri bution on my formation : Enrique Aviles, Guillermo Cebreros, Gonzal o Cebreroz, Agustin Labrada, Mario M armolejo, Gustavo Martinez, and Aureliano Ortega. Finally, my gratitude to Lily Wan, who edited parts of this manuscript ; she is not responsible for any content mistake or omission I have made.
8 TABLE OF CONTENTS page ACKNOWLEDGMENTS ................................ ................................ ................................ .. 4 LIST OF TABLES ................................ ................................ ................................ .......... 11 LIST OF FIGURES ................................ ................................ ................................ ........ 12 LIST OF ABBREVIATIONS ................................ ................................ ........................... 16 ABSTRACT ................................ ................................ ................................ ................... 18 CHAPTER 1 INTRODUCTION: THE AGRICULTURAL CONTEX ................................ ............... 20 Introduction ................................ ................................ ................................ ............. 20 Historical and Phy sical Context of Agricultural in Florida ................................ ........ 23 Speaking the Same Language: Concepts and Terminology ................................ ... 37 2 LITERATURE REVIEW: THE THE ORETICAL SCAFFOLDING FOR MIGRANT AGRICULTURAL WORKERS SAFETY AND HEALTH ................................ .......... 39 Introduction ................................ ................................ ................................ ............. 39 International Migration Theory ................................ ................................ ................ 40 The Political Economy of Agriculture ................................ ................................ ...... 47 The Scientific Evidence: Health, Safety, and Justice ................................ .............. 52 Connecting Dots: Suggestions of Social Network Analysis ................................ ..... 57 3 METHODS: THE PAD FOR COLLABORATION ................................ .................... 68 Introduction ................................ ................................ ................................ ............. 68 Logic Models ................................ ................................ ................................ ........... 69 ................................ . 73 Primary Data: The Settings of the Native Expert ................................ .................... 76 Ethnography: Living the Citrus Worker Experience ................................ ................ 81 Introduction ................................ ................................ ................................ ....... 81 Location ................................ ................................ ................................ ............ 84 Casa y sustento: Food and housing ................................ ................................ . 90 The s ack ................................ ................................ ................................ ........... 93 The f ruit ................................ ................................ ................................ ............ 98 Harvesters ................................ ................................ ................................ ...... 100 Luck ................................ ................................ ................................ ................ 102 The g rove ................................ ................................ ................................ ....... 104 Drinks ................................ ................................ ................................ ............. 105 BaÃ±os (boxes) ................................ ................................ ................................ 107
9 Guest w orkers (H2A) ................................ ................................ ...................... 111 Women in the groves ................................ ................................ ..................... 113 The l adder ................................ ................................ ................................ ...... 114 Changer and g anchear : Other methods of picking ................................ ......... 123 The last f ruit: Palo l impio ................................ ................................ ................ 125 Clippers and Ergonomics ................................ ................................ ............... 126 Outfit and g ear ................................ ................................ ................................ 127 Transportation ................................ ................................ ................................ 128 Multiple r isks ................................ ................................ ................................ ... 129 Health care search and a ccess ................................ ................................ ...... 132 The c linic ................................ ................................ ................................ .. 132 The h ospital ................................ ................................ ............................. 133 The h ealth Fair ................................ ................................ ......................... 134 Remedies ................................ ................................ ................................ . 136 The PCWH and the CH Ws role ................................ ................................ 138 New c hallenge ................................ ................................ ................................ 143 The native a nthropologist ................................ ................................ ............... 146 Mixe d Methods Data Analysis Process ................................ ................................ . 147 4 RESEARCH RESULT: MULTILEVEL ANALYSIS OF MIX DATA ......................... 149 Introduction ................................ ................................ ................................ ........... 149 Descriptive Statistics: Discovering the Invisible Harvester ................................ .... 149 Harvesting s easons 2004 2005 and 2005 2006 ................................ ............. 151 Harvesting s eason 2006 2007 ................................ ................................ ........ 152 Harvesting s eason 2008 2009 ................................ ................................ ........ 153 Harvesting s eason 2010 2011 ................................ ................................ ........ 154 ................................ ................................ ................... 155 Inferential Bivariate Statistics: Exploring Relations for the Use of PPE ................ 158 Harvesting s easons 2004 2006 ................................ ................................ ...... 158 Harvesting s eason 2006 2007 ................................ ................................ ........ 160 Harvesting s eason 2006 2007 ................................ ................................ ........ 161 Harvesting s eason 2008 2009 ................................ ................................ ........ 163 Harvestin g s eason 2010 2011 ................................ ................................ ........ 164 Script Analysis: Making Sense of the Ethnographic Experience ........................... 166 SNA: Knowledge, Support, and Safety ................................ ................................ . 171 C rew A social network a nalysis ................................ ................................ ...... 175 Crew B social n etwork a nalysis ................................ ................................ ...... 179 Crew C social network a nalysis ................................ ................................ ...... 184 Crews c omparison ................................ ................................ .......................... 189 Multivariate Statistics: Explanatory Models for Injury, Safety, and Health ............. 191 Harvesti ng s eason 2006 2007 ................................ ................................ ........ 191 Harvesting s eason 2010 2011 ................................ ................................ ........ 192 5 DISCUSSION: SOCIO CULTURAL ANALYSIS OF RISK ................................ .... 194 Introduction ................................ ................................ ................................ ........... 194
10 Working Conditions ................................ ................................ ............................... 194 The Economic Context: What is the Cost Benefit Ratio of Safety? ....................... 196 Worker Beliefs ................................ ................................ ................................ ...... 203 Social Structure ................................ ................................ ................................ .... 206 Me thodological Contribution Finding ................................ ................................ ..... 207 Future Research and Interventions ................................ ................................ ....... 209 APPENDIX A SURVEY 2004 ................................ ................................ ................................ ...... 214 B SURVEY 2007 ................................ ................................ ................................ ...... 216 C SURVEY 2010 ................................ ................................ ................................ ...... 218 LIST OF REFERENCES ................................ ................................ ............................. 219 BIOGRAPHICAL SKETCH ................................ ................................ .......................... 235
11 LIST OF TABLES Table page 1 1 World citrus production, 1000 x thousand tons ................................ ................... 28 1 2 Mexican production of citrus by state, 1,000 x tons ................................ ............ 35 3 1 Mixed Methods Design ................................ ................................ ....................... 71 3 2 2009 2010 Main citrus producer counties in Florida and data collection ............ 79 4 1 Relation of quantitative data collection by season ................................ ............ 150 4 2 Results for Safety Glasses Trial Seasons 2004 2005 and 2005 2006 ............. 159 4 3 Effect of CHWs on PPE usage in a 2007 quasi experimental test. ................... 160 4 4 Injuries and perceptions of PPE usage seasons 2006 2007 and 2008 2009 ... 161 4 5 Cross tabulation of PPE usage and injuries in Season 08 09 .......................... 163 4 6 Cross tabulation of PPE usage and promotion of PPE usage Season 08 09 ... 163 4 7 Cross tabulation of PPE usage and having a CHW help Season 08 09 ........... 163 4 8 Cross tabulation of PPE usage and having eye debris Season 2010 2011 ...... 165 4 9 Cross tabulation of PPE usage and knowing the C HW Season 2010 2011 ..... 165 4 10 Cross tabulation of PPE usage and experience using PPE Season 2010 11 .. 165 4 11 Crew A normalized cent rality measures ................................ ........................... 178 4 12 Crew B normalized centrality measures ................................ ........................... 18 4 4 13 Crew C normalized centrality measures ................................ ........................... 187 4 14 Logistic regression results for use of PPE Season 2006 2007 ......................... 191 4 15 Logistic regression A results for use of PPE Season 2010 2011 ...................... 192 4 16 Logistic regression B results for use of PPE Season 2010 2011 ...................... 192 4 17 Logistic regression C results for use of PPE Season 2010 201 1 ..................... 193
12 LIST OF FIGURES Figure page 1 1 Citrus growing region ................................ ................................ .......................... 22 1 2 Florida soil orde rs distribution map ................................ ................................ ..... 25 1 3 South Florida water flow diversion ................................ ................................ ...... 26 1 4 Vegetation change in South Florida ................................ ................................ .... 26 1 5 Frequency distribution of the main global citrus producers ................................ . 28 1 6 Mexic an main citrus producer states ................................ ................................ .. 36 3 1 Chronologic research logistic model ................................ ................................ ... 69 3 2 Mixed methods design data collection and analysis logic model ........................ 71 3 3 Schematic collaboration developed of the PCWH program ................................ 72 3 4 Citrus production by county, 2009 2010 ................................ ............................. 78 3 5 Florida commercial citrus production areas ................................ ........................ 80 3 6 Incorporated and unincorporated areas in Indian River C ounty ......................... 85 3 7 Ethnic composition in Fellsmer e, Florida ................................ ............................ 86 3 8 Origin self identification of Hispanics in Fellsmere, Florida ............................... 86 3 9 Indian River Country h ydrography ................................ ................................ ...... 87 3 10 Road map of Indian River County ................................ ................................ ...... 89 3 11 Grab and twist, so the citrus does not peel ................................ ......................... 93 3 12 A full Californian sack stands on the ground ................................ ...................... 94 3 13 A jumbo sack sits in one of the steps of the ladder while worker keep picking ... 95 3 14 Picking fruit from the ground pulling an ayate ................................ ..................... 96 3 15 Twelve tons of tangerines ................................ ................................ ................... 97 3 16 A monum ent in the city of Alvarado, Veracruz honor citrus workers ................... 97 3 17 ................................ ................................ .. 98
13 3 18 A crew leader cle an an empty 15 passengers van ................................ ........... 100 3 19 ............................... 103 3 20 Workers arriving to a m eeting place ................................ ................................ . 105 3 21 Worker changing places ................................ ................................ .................. 106 3 22 A empty baÃ±o ................................ ................................ ................................ .. 107 3 23 A modify diesel truck, call the chiva harvesting ................................ ................ 108 3 24 Harvester next to packing citrus boxes ................................ ............................. 109 3 25 Flatbed trail er tied ............................... 109 3 26 Chiva dumping citrus from a hydraulic hi lift metal basket ................................ 110 3 27 Full se mi truck trailer in the field ................................ ................................ ....... 110 3 28 Full trailers park outside a juice plant ................................ ............................... 111 3 29 Some 18 inch aluminum ladders rest with j umbo sacks on trees ..................... 115 3 30 Wooden ladders rest in an orange tree clos e to Alamo, Veracruz, Mexico ...... 116 3 31 Worker looks at the h orizon carrying an empty ayate ................................ ....... 117 3 32 A ladder rests on branches that can get poke through the rungs ..................... 118 3 33 A ver y instable la dder is a big risk ................................ ................................ .... 118 3 34 Worker secure ladder in the soil, before going up ................................ ............ 119 3 35 Sand accumulation on ladder and gloves ................................ ......................... 119 3 36 Shoes carrying the sand to the ladders ................................ ............................ 120 3 37 A broken branch, a risk hazard ................................ ................................ ......... 121 3 38 A worker barbeando with short sleeves shirt and long sucks in both harms ..... 122 3 39 A worker holds a hook ( gancho ) to pick citrus out of his reach ......................... 124 3 40 R eaching out for the last fruit ................................ ................................ ............ 125 3 41 Citrus worker with clippers getting ready for harvesting ................................ ... 126 3 42 Buses of a large harvesting company rest on S unday ................................ ...... 128
14 3 43 Large anthills are all around the groves ................................ ............................ 129 3 44 A dry tree in the field is a signal of a decaying grove, either by age or illness .. 130 3 45 Chemical dew on fruit and leaves ................................ ................................ ..... 130 3 46 A blooming tree will deliver additional debris and resin ................................ .... 131 3 47 Tree with dry moss, which deliver a green dust ................................ ................ 131 3 4 8 Members of the PCWH participating on a health fair at the regional hospital ... 134 3 49 The Hendry County Health Department holds a Sunday health fair ................. 136 3 50 T he Community Advisory Board select logo for the program ........................... 139 3 51 T he PCWH trains promotores de salud .................... 139 3 52 Constitution of the Lions Club in Immokalee ................................ .................... 140 3 53 The Immokalee Lions Club after holding their first eye health fair in 2003 ....... 140 3 54 Graduation of a promotor de salud ................................ ................................ .. 141 3 55 The author organizing a soccer game of Glasses User vs Non glasses users . 142 3 56 The last PCWH CAB meeting , 2010 ................................ ................................ . 142 3 57 The author conducting a glasses test in the field ................................ .............. 143 3 58 ................................ ........................ 145 3 59 The author spending a collective birthday party with his co workers ................ 147 4 1 Average age and working experience of surveyed worker ............................... 155 4 2 Breakdown place of origin of interviewed workers for season 2010 2011 ........ 157 4 3 Accumulative relation of farmworkers place of origin ................................ ........ 157 4 4 PPE usage and CHWs help for Season 2006 2007 ................................ ......... 161 4 5 Overall percentage of use at baseline and follow up observations ................... 162 4 6 PPE usage and knowing the CHW for Season 2008 2009 ............................... 164 4 7 Chart comparing workers using glasses with previous experience with PPE ... 165 4 8 Crew A single component social network PPE usage, origin, and CHWs ........ 175
15 4 9 Crew A social network degree centrality and PPE usage. ................................ 176 4 10 Crew A social network closeness centrality and previous PPE experience ...... 178 4 11 Crew B social network PPE user, CHW, and place of origin ............................ 181 4 12 Crew B social netwokl degree centrality , PPE usage , and trained CHW ........... 182 4 13 Crew B social network betweenneess and self report PPE usage ................... 183 4 14 Crew C social network PPE us age , CHW, and place of origin ......................... 185 4 15 Crew C social network degree centrality , PPE usage , and past PPE usage .... 186 4 16 Crew C social network betweenness centrality, PPE u sage , and past PPE usage ................................ ................................ ................................ ................ 188 4 17 Crew C closeness centrality, self reported PPE usage , and past PPE usage .. 188 4 18 Centrality mea sures of crews and PPE usage comparison .............................. 190
16 LIST OF ABBREVIATIONS BLS Bureau of Labor Statistics, US Department of Labor CAB Community Advisory Board CBPM Community based Prevention Marketing CBPR Community based Parti cipatory Research CDC Center for Diseases Control and Prevention CHW community health worker, promotor de salud CNC Confederacion Nacional Campesina Mexican National Confederation of Farmworkers CONASUPO CompaÃ±Ãa Nacional de Subsistencias Populares National Company of Popular Subsistence DEO Florida Department of Economic Opportunity DOL United States Department of Labor EPA U nited S tates Environmental Protection Agency FAO Food and Agricultural Organization of the United Nations FDACS Florida Department of Agriculture and Consumer Services FDOC Florida Department of Citrus FPRC Florida Prevention Research Center FWAF Farmworker Association of Florida H2A U . S . Department of to bring nonimmigrant foreign workers to the United States HIV Human Immunodeficiency Virus IFAS University of Florida Institute of Food and Agricultural Sciences N Total of workers in the sample analyzed or eliminated NAFTA North American Free Trade Agreement NASS United S tates Department of Agriculture, National Agricultural Statistics Service
17 NIOSH National Institute for Occupational and Safety Health OUSTR Office of the United States Trade Representative: Executive Office of the President OSHA Occupational Safety and Health Administration PCWH Partnership for Citrus Workers Health PEMEX Petroleos Mexicanos Mexican Oil State Company PPE Personal Protective Equipment PRI Partido Revolucionario Institucional Mexican Political Party UF University of Florida USF Univers ity of South Florida S.A. de C.V. Sociedad AnÃ³nima de Capital Variable. Public Limited Corporation SNA Social Network Analysis SIAP SAGARPA Servicio de InformaciÃ³n Agroalimentaria y Pesquera Secretaria de Ag ricultura , Ganaderia , Desarrollo Rural, Pesc a y AlimentaciÃ³n . Agricultural and Fishery Statistic Services Mexican Department of Agriculture. STATSGO U nited S tates Department of Agriculture, Natural Resources Conservation Service STI Sexually Transmitted Infections SFWMD South Florida Water Mana gement District U.S. United States of America USCB United States Census Bureau USDA United States Department of Agriculture WPS regulation
18 Abstract of Dissertation Presented to the Graduat e School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy STRATEGIES OF SAFETY AND HEALTH AMONG CITRUS FARMWORKERS IN FLORIDA: A CULTURAL AND SOCIAL NETWORK APPROACH By Jose A ntonio Tovar Aguilar August 2014 Chair: Christopher McCart y Major: Anthropology Florida is one of the main global producers of citrus. In 2011 , it generated $3.8 billion of farm gate sales. These sales represented more than 70 % agricultu ral output and nearly 90 % of all farm sales from just crops. Florida alone supplies close to 70% of the total orange juice consumed in the United States each year. Citrus is handpicked, which translates to a labor force of at least 60,000 workers each seas on. Agriculture is among the most dangerous occupations. The U.S. Bureau of Labor Statistics reported that between 2003 and 2011, 5,816 workers died from work related injuries. In addition, the rate of injuries for agricultural workers is over 40 percent h igher than for workers in other industries. Citrus harvesting is physically demanding: Pickers stand on an 18 foot aluminum ladder carrying a bag weigh ing up to 90 pounds. The hazards citrus workers face include ergonomic injuries, falls, heat stress, wild life encounters, but particularly eye injuries cause d by tree debris, dust, mildew and fungus on tree leaves, insects, citrus flowers, sand, and water droplets contaminated by chemicals. Less frequent but more dangerous are traumatic eye injuries from bran ches
19 and fruits , as well as the development of eye illnesses caused by long exposure to UV light . In collaboration with the Partnership for Citrus Workers Health , an evidence based intervention program that for over 10 years has promote d the adoption of protective eye safety equipment among Spanish speaking farmworkers in Florida , this study used mixed methods to explore the strategies followed by citrus pickers to avoid the potential hazards encountered in the fields and the processes they followed when injured. Workers relied on experienced co workers to learn strategies for safe and effective methods for picking, and when looking for health care. Mexican rural culture played a role in the beliefs, strategies, behaviors, and social relations conducive t o Working conditions, however, are still the main risk factor. Social network analysis usage has the potential to increase the dissemination of safety practices by selecting health promotores with spec ific network structures.
20 CHAPTER 1 INTRODUCTION : THE AGRICULTURAL CONTEX Introduction Agriculture has been the engine of major changes in human history: it made sedentary life possible, encouraged technological advances that spilled into all other industri es, empowered land conquest and human expansion, and made possible the establishment of complex civilizations (Manning 2004). Today, agriculture and food are still behind many industrial and technological shifts. While the discourse of globalization is cen tered on newer and faster industries, some argue that despite the bound to make some of the most critical contributions to the reconfi world (Bonann o et al. 1994:3) . Food cultivation, processing, storage, and distribution have always borne in mind foods (good to eat) are foods that have a more favorable balance o f practical benefits who stand to gain from the process. Those whose very lab or is the engine of these processes (farmworkers, peasants, sharecroppers) , however , do not often have their needs considered. My growing concern for farmworkers came from a long lasting interaction with this population dating back to 2006, when I started working at the Farmworker Association of Florida. I was well aware of the important role of agriculture in Florida, a state to which I moved from Mexico in 1999, but had never had the opportunity to work
21 directly with the people who made farming possible. The fascination encountered in the day to day interaction with persons coming from my own country and other Central American and Caribbean countries where I worked and lived as a journalist created an intense commitment to work with this population. In g eneral, my most striking observation, the one from which I was also ignorant from seven previous years in Florida, was the shameful invisibility of farmworkers in modern society. Those who harvesting the food we eat every day are , for the most part, distan ced from public consciousness. Some authors have expressed how farm labor conditions have remained hidden from the public eye (Bletzer, 2004:532) even though systematic consti tution of inequality and suffering , according to Benson (2008:591). T he unhealthy , despicable situation of farmworkers is widespread, and documented cases of structural violence against them can be found through out history and all around the world. In the United States this phenomenon is not different; in fact, it is illustrative of many national policies and practices around the world. Agriculture is an important factor of any country independenc e and grow th . The United States has relied on it to mainta in steady economic development for decade s. The combine d effects of land clearance, scientific and technological development, as well as cheap labor have raised the country to a position of prominence in worldwide agricultural production. Nevertheless, the social and environmental effects of such development have been incommensurable and for the most part ignored by the mass media and the general public . Not even the recent success of the Coalition of Immokalee Workers (CIW) and
22 their Fair Food Program has conditions in the fields. Citrus is the largest fruit crop in the world with about 60,000,000 MT (1 MT = 1 000kg) grown, slightly exceeding grape production (FAO, 1999). Citrus is grown in two belts on both sides of the equator from about 20 to 40 degrees of latitude. Although Figure 1 1 ignored the important Mediterranean production region, it is a good indicator of the geographic distribution of the crop . All citrus is thought to have originated in the Himalayan region of southwestern China and northern India from where it expanded to the old continent, south Asia, and later into Oceania and the Americas. Figure 1 1. Citrus growing region (Source: Bates et al. 2001)
23 Historical and Physical Contex t of Agricultural in Florida Historically f rom indigenous displacement and genocide, to slavery and civil war , and to the large importation of farm labor from developing countries the working conditions of the agricultural work force in the United States has d epended on large numbers of underpaid, unprotected, and easily controlled populations. It is not surprising that farmworker s were excluded from any kind of protection under U.S. labor laws passed in the 1920 30s the National Labor Relations Act and the Fai r Labor Standard that protect ed almost all other classes of workers. This exclusion was done explicitly at the behest of the agricultural industry, which wields tremendous power in the U.S. Agricultural workers as well as domestic employees are excluded f rom the eight hour day, forty hour workweeks, and overtime payments. These two sectors were also excluded from child labor protection and until 1966 were unable to organize and form unions. In similar fashion, safety and labor conditions were not recognize d until 1983, when the Migrant and Seasonal Agricultural Worker Protection Act was passed. The law also established the need of farm labor contractors to register with the U.S. Department of Labor and to assure necessary protection to farmworkers, agricult ural associations, and agricultural employers. Although the first legislation regulating the use of chemicals in the U.S. was enacted in 1910, it was not until the seventies that the U.S. Environmental Protection Agency had the authority to regulate heal th and environmental exposure of pesticides. However, older pesticide risks were not addressed until legislation was enacted in 1988, requiring reregistration and 1996, requiring that pesticide food standards are
24 I n Florida, agricultural development was initially limited to the north part of the state and some of its offshore islands to the east, where climate and reliable soil conditions allowed for early plantations. In the south, however, this development was pre are fav orable for farming for long periods of time, soil conditions and overflowing did not allow for farming in most parts of the peninsula. After the end of the Seminole Wars, agricultural expansion mainly faced environmental limitations, which had to be overco me by scientific development dealing with soil conditions and water drainage. The recent soil formation in the peninsula and the topographic flatness of the area contributed to the lack of nutrients and drainage; basic conditions for the development of agr icultural production were absent. As seen in Figure 1 by the orange color on the map. Also known as podzols, most of these soils, like in the case of Florida, are not appropriat e for agriculture due to their sandy characteristics. Sands have low levels of moisture and nutrients, they either have excessive drainage or if in shallow rooting zones, as in the case of south Florida, have poor drainage. In addition, their low pH conten t, phosphate deficiencies, and aluminum toxicity compounds the need for loamy drainage and large amount of lime and fertilizers (FAO 2006:91).
25 Figure 1 2. Florida soil ord ers distribution map (Source: Grunwald, 2002) To solve the problem associated wi th these poor conditions, especially after the fires and floods of the 1940s ( Ingebritsen et al. 2013) , extensive drainage systems were developed (Figure 1 2). In the winter of 1947, a massive rainfall flooded millions of agricultural acres, opening the cr eation of the U.S. Army Corps of Engineers Central and Southern Florida (C&SF) Flood Control Project (Harwell, et al. 1996); which transformed 50 percent of the Everglades to permanent agricultural and urban areas (Figure 1 3) This also greatly benefited f rom the development of commercial air conditioning in the 50s (Storper et al. 2009)
26 Figure 1 3. South Florida water flow diversion (Source: Ingebritsen et al. 2013 ) Figure 1 4. Vegetation change in South Florida (Source: Davis et al.1994 )
27 Moreover, s ince weather and soil conditions fungus, bacteria, and pests are prevalent in Florida soil; food production can be difficult to accomplish. To improve nutrient deficiency and control pests and infections, massive amounts of agrochemicals are used on a regu lar basis. hundreds of diff erent fertilizers, herbicides, pesticides, and fungicides enabled the (FDACS 2011b). The Economi c and Human Impact of Agricultur e In terms of revenue generated, Florida's top five agricultural products are greenhouse and nursery products, oranges, tomatoes, sugar cane, and cattle and calves; citrus production, though, surpasses that of most countries in the world . T able 1 1 shows the historic citrus production of the current four larger producers of citrus in the world, more than 50 percent of the global production in reach by them . China recently surpassed Brazil as the lead grower, while Florida mai ntains the comfortable third place (Figure 1 4).
28 Table 1 1. World citrus production, 1000x thousand tons 1980 1981 & 1988 1989 * 1990 19 91 1998 1999 * 2004 2005 2005 2006 2006 2007 2006 2007 2008 2009 2009 2010 2010 2011 Forecast Brazil 11.6 16 .9 18.9 20.4 20.8 19 .0 19.1 17.5 22.7 China 1.7 7.3 14.9 14.5 15.9 18.9 21.4 23.9 22.9 USA 11.2 14.4 10.3 10.6 9.5 11.6 10.7 10 .0 10.4 Mexico 2.5 4.4 6.4 6.7 6.7 7.4 7 .0 6.8 6.7 Others 30.7 38 .0 46.9 50.4 52.8 52.7 60.3 51.2 52.7 Total 57.7 81 .0 97.4 102 .6 105 .7 109.4 118.5 109.4 115.4 *Average of both seasons. Source: FAO 2012. Figure 1 5. Freq uency distribution of the main global citrus producers for the season 2010 2011 (Source: FAO 2012) According to the Florida Department of Agri culture and C onsumer Services , an executive agency charged with marketing, research and regulation for the industry, Florida Brazil 20% China 20% USA 9% Mexico 6% Other countries 45% GLOBAL CITRUS PRODUCTION 2010 2011
29 California totaled 32 percent, and Texas and Arizona combined prod uced the remaining ( F DACS 2011b:12 ) . Introduce d to 1493, oranges were the first citrus to grow roots in the Americas, and later in the mid 1500s at the shores of Florida . The s t ate's unique sandy soil and subtropical climate proved to be ideal for growing the seeds that the early settlers planted and have flourished ever since. Commercial groves have existed since the 1800s, and since then grapefruit has been introduced to the st ate . As the citrus industry grew, so did Florida economy and its need for farmworkers , since more than 90 percent of the fruit is handpick ed (Roka, 2012). Per the e stimates of the Florida Department of Agriculture and Consumer Services, Florida Agricult ural Statistics Service in the 2009 2010 citrus seasons, Florida farmworkers were responsible for harvesting 554,037 acres of citrus trees or 165,850,000 boxes of all kinds of citrus ( DEO, 2011; USDA, 2012). Around the s tate, the groves cover small and lar ge patches from the light slopes of Central Florida to the sandy former swamps of South Florida. From coast to coast, picking from tree to tree, thousands of harvesters have put Florida as second only to Brazil in global orange juice production and the sta te remains the world's leading producer of grapefruit (Figure 1 5) . Monetarily, t he production of citrus during 2011 generated $3.8 billion of farm gate sales. These sales represented more than 70 percent output and nearly 90 percent of all farm sales from crops alone (Roka, 2013) . Today it is a $9 billion industry, emp loying nearly 76,000 people (FD O C 2010 ), of which at least 60,000 workers are citrus harvesters ( DEO 2011).
30 As with any agricultural work, harvesting citrus i s a risky enterprise: the job is conducted i n extreme environmental (high temperatures and humidity), ecological (exposure to chemicals), ergonomic (heavy loads, irregular terrains, and constant use of ladders) , and social (low payment, and invisibility) c onditions. Crews of 10 to 40 pickers dispersed around C entral and S outh Florida work o n isolated land with little supervision. Citrus harvesting is physically demanding: Pickers stand on 18 foot aluminum ladder s that lean on the branches deep inside the ca nopy of orange tree s . Starting at the top of the ladder, with a large canvas bag across one shoulder, they descend, snapping off oranges using both hands and filling the bag. A full bag can weigh 90 pounds. Once on the ground, the picker drags the full b ag over to a large bin, lifts it to the edge, and empties it. Each bin holds up to 600 pounds of fruit (Monaghan 2008:36) and, when full, can be worth $7.00 to $15.00 each to the migrant , depending on the set price that day. Each bin is supposed to hold t en 90 pounds bags, but that measurement has been disputed (Moreno 2007). Some experienced workers can do it, but only if the fruit is a good size. An experienced worker can fill 8 18 bins (more than two tons of oranges or grapefruit) in a day and sometimes more if the fruit is plentiful and easy to harvest . Although that may seem to indicate that workers could earn upwards of $250 a day, in reality that is not the case. Bigger fruits, like grapefruit, means more bins in a day (nearer to 18 for the best work ers) but also pay less than smaller fruits like tangerine. So, in reality, a good worker can make around $120 a day, bearing in mind that this is only on optimal days with good weather and healthy grove. The hazards citrus workers face include ergonomic in juries, falls, heat stress, wildlife encounters, and particularly eye injuries cause d by tree debris, dust, mildew and
31 fungus on tree leaves, insects, citrus flowers, water droplets contaminated by chemicals, and especially sand coming from the particulate heavy soil s conditions of the state. Environmental conditions are extreme: the groves are typically wet and humid in the mornings and hot and dusty in the afternoons, with temperatures rising to 90 Â° F and above increasing eye injuries . When the fields ar e wet, the shoes, boots, and long pants of workers get filled with sand and when they scale the ladders the sand stays on the steps and eventually ends up falling in their eyes as they move the ladder from tree to tree. In dry conditions, the dust caused b y moving vehicles leaves the trees with additional debris. Less frequent , but more dangerous are traumatic eye injuries cause by branches and fruits. Constant exposure to UV light also increases the risk of developing pterygia, pinguecula, cataracts, and r etinal damage (Quandt et al. 2008, Lacey et al. 2007, Taylor et al. 2006, Villarejo et al. 1999). Moreover, because workers are paid a by piece rate there is a tendency to work fast without further risk considerations , especially since supervision is limit ed, safety equipment is inadequate, and, in most cases, not readily available. When injuries occur agricultural workers often are deterred by complicated reporting procedures, expensive treatment, reliance on crew leaders and subcontractors for access to h ealth services, and a lack of health insurance (Monaghan et al. 2008) . In some instances w orkers receive d safety training s , but in many cases training is limited to protection standard (WPS) 30 minute video and workers are not given approp riate time and instruction to sufficiently learn the materials.
32 Citrus harvesting operations are conducted in two significantly different settings impacting the possibilities of safety trainings. One kind of production is carried out by large growers/harv esting companies, many of which use the agricultural guest program (H2A visa workers). The remaining production is conducted by a large number of small and medium companies who work as independent contractors and even intentionally keep small crews to avoi d regulatory oversight. There is not data available in terms of which operation harvests more, but the tendency for agricultural enterprises in the United States is to have large companies consolidate at the cost of smaller farmers and harvesters. Large c ompanies complain of the cost associated with injuries and engag ing i n safety programs . They are well aware of regulations and avoid infractions that can also carry public backlash against their name. For these companies, the conduct of WPS and food safety trainings are regularly practiced and occasionally the companies will host or support safety trainings to avoid falls, high voltage lines, lightning, wildlife encounters, and run overs. One of these cases is the training for farm labor supervisors and cre w leader conducted by the University of Florida (UF), though IFAS extension (Institute of Food and Agricultural Science) at the Southwest Research and Education Center in Immokalee (Morera et al. 2014). Eye safety is a particular concern for these companie s, as it is the most common kind of injury among harvesters. The companies are well aware of the problem, but share difficulties promoting the use safety equipment (PPE) among workers. Even when the company provided safety glasses or offered incentives for crews without accidents, several eye injuries were reported each season. Research conducted by the
33 Partnership for Citrus Workers Health (PCWH) proved the importance of proper equipment (Tovar et al. 2014), marketing strategies (Monaghan et al. 2008, 2011 b), and strategies for safety practices, however, have not been deeply explored (Monaghan 2008). That is one of the purposes of this study. A more significant limitation o companies with a significant increase of H2A workers. In the case of smaller and medium companies, their work force is mainly comprised of undocumented migrant workers. Their safety trainings are sporadic and concentrated on the regulatory WPS. Because these crew leaders usually have a bad reputation for wage theft and mistreatment of their workers, the turnover rate of workers is high. Safety equipment (PPE), including eye safety goggles, is never provide d, and workers usually live in unhealthy housing conditions (crowded, with limited services, and excessive alcohol consumption). 1 realities in each setting; large and small harvesting com panies. ranks among the most hazardous industries. Farmers and farmworkers are at very L 2013a) reported that 245 crop production workers die d of work related injuries and over an additional 20 thousand more suffered non fatal injuries on the job. BSL, part of Department of Labor had observed a decline in the number of fatal accidents in thi s despite the declines over the last two years, agriculture recorded the 1 Neither if which contribute to safety conditions.
34 highest fatal injury rate of any industry sector at 21.2 fatal injuries per 100,000 FTE workers in 2012 . Moreover, the number of injuries may be even larger if we consider that operations with less than 11 workers do not have to report injuries. Information from the 2007 Bureau of Labor Statistics (BSL 2007) indicates that the annual rate of eye injuries among farmworkers is more than twice the incidence rate among workers in private industry (3.8/10,000 versus 8.7/10,000). Furthermore, the rate is higher among farmworkers working on fruit trees (11.5/10,000) ; the difference may be even larger if we consider that many of the agricultural workers are less likel y to report injuries as a result of cultural stigma, vulnerability in the work place (especially if they are undocumented), and lack of familiarity with laws and regulations in agriculture (Kandula 2004, Monaghan et al. 2008) . More importantly, Quandt et a l. (2001) estimate that 90% of agriculture related eye injuries could be prevented with the proper use of protective eyewear. Florida Citrus Workers the Cultural Context Citrus pickers come to Florida from Mexico, Central America, and Haiti; although by a nd large, as a mirror of the n ational ethnic composition on agriculture, pickers from all over Mexico are the backbone of labor force. Of particular importance are the workers coming from the main citrus production area o n the gulf c oast of Mexico , Veracruz ; although pickers from traditional migrant states (Guanajuato, Michoacan, Guerrero, and Hidalgo) as well as the growing migrant indigenous population of southern state s of Chiapas and Oaxaca are also found in the groves. Mexico has been a traditional supplier of labor for the United States agricultural industry. It is not only its condition of being an under developed country with an army of
35 cheap labor nearby, but also the fact that it has been a rural country. Even when demography is changing, Mexico still has important agricultural populations, as its local production can show. Table 1 2 highlights the six main citrus producing states in Mexico; together they represented 77.6% of the total production, although Veracruz alone cont ributes 43.8%. The main citrus varieties grown are oranges (Valencia, Navel, and Navelina); grapefruits (Marsh, Ruby Red, Star Ruby); Tangerines (mandarins, tangerines, and tangelos); and limes (Mexico is the second world producer after India). Together, t he mentioned varieties occupied 94% of the dedicated land for citrus production ( SIAP SAGARPA 2014) . For simplicity, other citrus like lemons and source oranges were added in the lime category. Table 1 2. Mexican production of citrus by state, 1,000 x tons State Orange Limes Tangerines Grapefruit Percentage Veracruz 1,983 520 380.1 261.5 43.8 Tamaulipas 545 57.2 48.9 35.5 9.6 Michoacan 3.4 463.9 0 40.5 7.1 Colima 5.6 494.1 0 0.2 7 San Luis Potosi 374.5 8 13.9 0.1 5.5 Puebla 258.7 32.5 44.9 4.2 4.6 Ot her States 910.2 573 46.5 73.5 22.4 Total 4,080 2,149 534 416 7,179 Percentages correspond to the total production contribution. Source: SIAP SAGARPA 2014 Mexico is one of the main producers of citrus in the world (Figure 1 5), 29 out of the 32 states i n the country register some kind of citrus production. Figure 1 6 illustrates the states who concentrate over 92% of the total production on citrus. The economic and cultural impact of this crop is significant; it is estimated that over 90,000 families
36 dep end of this production which is localized in over one million acres. Figure 1 6. Mexican main citrus producer states (Source: Info Rural with data from SIAP SAGARPA 2012) The migrant conditions of workers coming from small rural areas open the door to q uestion about the cultural implications of their behavior in relation to the risky working conditions they face. Do they have specific strategies to avoid injuries while working? What do they do when the injury happens? Do they think ahead to prevent accid ents? Agriculture, a key development in the rise of sedentary human civilization, develops thinking, materials and instruments of human culture. Crops around the world define farming, eating, religions, and other social practices of human groups. Citrus workers are no different: to their national cultural background, we may include their farmworker condition, and more specifically, their socialization as naranjeros (citrus harvesters). The current study explores the cultural strategies followed by citrus pickers
37 in Florida to avoid the potential hazards encountered in the fields and the processes they follow when injured. Speaking the Same Language: Concepts and Terminology Some concepts and terms used during this work need some clarification. The concep Bustamante (1977), Portes (1978), and Chavez (1988) or the extensive work of De spatialized socio political conditio (De Genova 2004:161) of farmworkers in the United States. Most of the workers, as mentioned before, are from Mexico, and although as noted by Passel and Taylor (2009) the concept of Hispanic/Latino as an ethnic group corresponds more to an adopted labe l of the U.S. Census than a self identify concept, for the purpose of this work, I use the term indistinctly, recognizing its limitations. For instance, the term far from properly describes the characteristics of some of the workers, since a growing number of indigenous people are working in agriculture. The argot used in the groves of Florida is extensive and significant; most of it is in Spanish and in a pseudo translated English. For instance, the crew leaders are call chiveros ( goat keeper) because the y drive la chiva (the goat is the truck in which they collect the citrus before dumping it in to large trailers) and because they take care of chivos (goats are in this case the workers who move in the field as a group). When introduced in this text, argot s will be explained as the tools ( baÃ±os, ayates, ganchos ), strategies ( escalerear , changear, barbear, puntear ), and working conditions ( huerta baraÃ±uda , de pechito , matudo) are presented. Spanglish is a debatable concept, but words such as la troca (truck) , la traila (trailer) , dumpear (to dump), cuitear (to quit) , and lunchear (to lunch) , are just examples
38 of the pseudo translated English that occurs and are sporadically mentioned in this work. Proper English translation is added in parenthesis, as they ap pear above. Finally, there are several Spanish terms used because they originated in the language. This is the case of promotores de salud, known by many different names in the English academic literature, it is also used sporadically in this dissertation since I prefer the terms community health workers (CHW) ; as per 2009 Management and Budget formally recognized CHW as an occupational classification and it is now listed in the 2010 2010: 816 ). Other terms mentioned are Maquila a term coin by the Mexican government in 1965 as a program to offset effects of the end of the Bracero program in the border; and Bracero , a term rooted on the word brazos (harms) and that literally means the act of gri nding wheat brought to the miller by farmers to be converted to flour. Additional Spanish words use in the text are conventional foods ( pupusas, tacos, chicharrÃ³n), clothing ( paliacate, guayavera), organizacions ( La Via Campesina , CONASUPO ), and places ( p upuseria, tienda) , which will be properly contextualize or translated.
39 CHAPTER 2 LITERATURE REVIEW : THE THEORETICAL SCAFFOLDING FOR MIGRANT AGRICULTURAL WORKERS SAFETY AND HEALTH Introduction On the eve of an international conference for Community Hea lth Workers (CHW) in San Diego, California, I drove to Miami to meet with the only promotor in the Partnership for Citrus Workers Health (PCWH) program able and willing to risk deportation in order to attend the meeting. Four years earlier, he had spent th ree days without water or food in the Arizona desert in order to leave decrepit farmland disputed by government troops, organized crime, and an indigenous peasant revolt not far from the Guatemalan border. His family members mother, sister, brother, and c ousin had only been able to make it to Tijuana where they are now working at a maquila . In Miami, we were able to get his Mexican passport in order to catch our flight the first one he had ever taken in his 24 years of life . W hile waiting for his passport we ate at a Honduran pupuseria and after picking up the document we ended up on the beach where we spent the evening socializing with a Russian tourist group, Cuban families, and some Canadian snow birds. During our red eye flight we shared seats with a f riendly Philippine who calmed my companion with his stories of long flights until both fell asleep. At the conference, we were among the 5 percent of male participants; the women of mainly Mexican origin were impressed with his courage and humility. On thi s trip, he opened an e mail account, learned to swim by observing American children in our hotel pool, and was taught how to dance by a Puerto Rican nurse, while I provided English instruction and taught him the secret to knotting a tie. Before leaving Cal ifornia, my friend and I attempted to meet his family through a border fence in a beach park in Tijuana. However, the US side of the park was closed
40 down and only I could enter Tijuana. In an effort to provide my friend and his family with at least glimpse s of each other, I packed a digital camera with recent pictures of my friend and crossed the border to share them with his family. I had also packed the $200 in remittances that my friend had intended to pass through the fence. Later that day I returned to San Diego toting a bag of homemade Mexican food and a digital camera with twice as many Photo s. After the conference, my friend and I bid each other farewell at our Chicago stop he was on his way to pick apples and I was on my way back to Miami. Internati onal Migration Theory International migration theory has shifted from bipolar models to transnational models because the realities of the past have been transformed by the mobility of the present. It is not that the pull and push dynamic has ceased; as lon g as poor countries cannot offer their people opportunities for survival, their people will look for a way out by migrating either to a different region, city, or country. The same is true for economies in search of unskilled and cheap labor. What has chan ged is that connectivity among the host and home nation has closed gaps between places and created multiple identities. Even while many of the farmworkers I encounter cannot go to Mexico nor can their families visit them, they spend several hours every wee kend on the phone, texting, or emailing pictures. And some actually do venture back and forth between the U.S. and Mexico. The proximity of the two countries is an advantage and similarities among them are making their transition less traumatic, though not less difficult . Perhaps, as Juan necessarily stretches the idea of national belonging by disengaging it from its presumed
41 territorial and linguistic imperative, de centering it in relation to any putative core values (Smith 2008:3). In traditional bipolar models that emphasize linear movement and settlement from home to host nation we can find the employment of different c oncepts, assumptions, and frames of references. One of these is the neoclassical economics school of thought which focuses its attention on the differences in wages and employment conditions between countries, and on migration cost. It was originally conce ived to explain labor migration in the context of Economic Development (Lewis 1954). Countries with a large endowment of labor relative to capital have a low equilibrium market wage, while countries with a limited endowment of labor relative to capital are characterized by a high market wage. The resulting difference in wages causes workers from low wage countries to move to high wage countries. And v ice versa, the flow of investment capital will flow from capital rich countries to capital poor countries be cause the rate of return will be higher by international standards. The simple and compelling explanation of international migration offered by neoclassical macroeconomics has strongly shaped public thinking and has provided the (Massey 1993:433). Based on this theory, the U.S. labor market is attractive to developing countries because it offers better salaries; the same way that U.S. investment would prefer to invest in other countries because they will have b etter returns. The problem with this position is that it does not incorporate rapid changes in the job market, demographics, and the pressure of the financial markets.
42 orresponding with this perspective is the microeconomic model based on individual choice (Sjaastad 1962 , Torado 1969, Todaro and Maruszko 1987 ). This Neoclassical explanation holds that the cost benefit calculation of people, expecting a positive return, encourages their migration. International migration is conceptualized as a form of investment in human capital . People choose to move to where they can be most productive, given their skills (Massey 1993:435). The problem with such reasoning in the context of migration to the U.S. is that it does not account for the active recruitment of cheap labor for the benefit of capital growth, nor does it explain migration in search of freedom, democracy, or simple family reunification rather than exclusive economic gain. om 1985), in contrast with the former positions, considers conditions in a variety of markets, not just labor markets. It views migration as household decisions taken to minimize risks to family income or to overcome capital constraints on family producti on activities (Massey 1993:432). Because there are differences between developed and developing economies, this position analyzes the strategies with which poor countries deal with u nstable conditions, such as restrictions in accessing credit, technology, crop insurance, unemployment insurance, or assessing market uncertainties. Dual labor market theory links immigration to the structural requirements of modern industrial economies. According to Piore (1979), immigration is not caused by push factors in s ending countries, like low wages or high unemployment, but by pull factors in receiving countries always in need of foreign workers. For Massey (1993) this demand for immigrant labor stems from four characteristics of industrial societies: 1)
43 Structural in flation linked to occupational hierarchy, such that the importation of migrant workers is needed because they will accept low wages; 2) Motivational problems observed at the bottom of the job market for which immigrants will simply expect monetary compensa tion; 3) Economic dualism in reference to the inherent bifurcation between labor and capital for which migrants will be willing to take secondary sector jobs (agriculture, construction, etc . ) compared to the natives dominating the primary market jobs (fina nce, academics, and so on); and 4) D emography of labor supply, shape d by the increase of U.S. participation in the primary labor market and U.S. teenagers obtaining higher education levels, such that the secondary markets can only count on migration worker s. Dual labor market theory neither posits nor denies that actors make rational, self interested decisions, as predicted by microeconomic models, but recognizes the importance that the job market plays in driving this migration. The problem with this persp ective in the face of U.S. migration is that once again economics is the only motivation driving people in, while there may be other factors that account for that migration. In addition, not all migration is responding to calls from secondary sector jobs. In the face of a growing demand for high tech jobs and in the service industry the labor market is reaching for workers to fulfill demands in primary sectors. World systems theory sees immigration as a natural consequence of economic globalization and mark et penetration across national boundaries. Building on the work of Immanuel Wallerstein (1974), a variety of sociological theorists likened the origins of international migration not to the dichotomy between particular national economies, but to the struct ure of the world market that has developed and expanded since the
44 sixteenth century. In this scheme, the penetration of capitalist economic relations into peripheral, neo capitalist societies creates a mobile population that is prone to migrate abroad. It is from this school of thought that we can trace what has been called the transnational perspective. World systems theories argue that international migration follows the political and economic organization of an expanding global market, a view that , for M assey (1983) , yields six distinct hypotheses: 1) The penetration of the global economy into peripheral regions is the catalyst for international movement ; 2) Capitalist investment foments changes that create an uprooted, mobile population in peripheral cou ntries while simultaneously forging strong material and cultural links with core countries ; 3) International migration is especially likely between past colonial powers and their formal colonies ; 4) Political and military interventions to protect investmen ts, when they fail, produce another form of international migration ; and 5) International migration ultimately has little to do with wage rates or employment differences between countries and more to do with the dynamics of market creation and the structur e of the global economy. In the case of the U.S. many of these hypotheses apply, but others are not accounted for in world systems theory. For instance, many of the U.S. direct and indirect military interventions in Latin America cannot be categorized as f ailures and nevertheless they had an impact on the number of migrants coming to this country. The Mexican American war was won by the U.S. and officially we cannot call Mexico a U.S. former colony even though more than half of current im migrants living in the U.S. come from its southern neighbor. Furthermore, wage rates and employment differences have opened new kinds of mobility: the influx of Asian high tech professionals is not the
45 creation of this market but rather the creation of human resources . Forei gn professionals educated in the United States tend to stay here instead of looking for jobs in their home countries. What happens in complex societies, like the United States, is that more than one theory may not fit into the equation. For instance, goin g back to my earlier narrative: my because the household unit decided for him that his migration would better provide for the family. Yet in the case of our Philippine passenger who moved to the U.S. because he understood that migration was the best way to maximize his technical knowledge, the n eoclas s ical perspective represents a better explanation. Transnationalism may not be able to account for all the different levels of migr ation because it is a concept that is still evolving and debates about its meaning are still in progress or because some other theories are still in effect. However, for Portes (2003) , some of the empirical and conceptual points on which the until recentl y contentious literature on transnationalism has reached a measure of consensus are: 1) transnationalism represents a novel perspective, not a novel phenomenon ; 2) transnationalism is a grassroots phenomeno n; 3) not all immigrants are transnationals ; 4) im migrant transnationalism has macro social consequences ; and 5) the extent and forms of transnational activism vary with contests of exit and reception. Transnationalism also recognizes that the field is multidisciplinary. For instance a newer perspective i s coming from geographers (Jackson 2004) and they suggest that transnationalism is a social space that can be occupied by a wide range of actors, not all of whom ar e directly connected to global communities themselves . In other words ,
46 they see societies as transnational entities in which an actor can choose to participate in the exchange. dominant struc ture of the U.S. labor market and the cheap and disposable nature of the immigrant work force. Using the Bracero program as a historical reference and comparing the U.S. Mexican relation to the colonial relations of France with Algeria and England with Ind ia, Gonzalez offers an alternative explanation to push pull migration theory , which centers its attention on the supply and demand of the labor market. Civil War period a widespread and vocal groundswell of businesspeople argued for the forced military occupation and annexation of Mexico, but it was overpowered by then minister to Mexico, William Rosecrans, who insisted that the absorption of Mexico ensured more problems than benefits. He argued that there existed an alternate method for realizing 8: 38 ). Regardless of the validity of this position, other academicians have traced Latino migration to the impacts of U.S. policies in Latin America; large numbers of Latinos in the U.S. have come from Puerto Rico, Cuba, Dominican Republic, and Honduras al l of which at some point in their history have been occupied by the U.S. military (Gonzalez 2001). Prior to the U.S. annexation of the Mexican northeast provinces, Mexican settlers owned and operated all rural enterprises. The conquest not only disenfranch ised these landowners through state directed
47 violence, but as a result many ended up as sharecroppers, working the land they once owned (Luna 1997 1998). Some historic similarities can be observed in the migration patterns of Florida in relation to the eco nomic, political, and military influence of the United States with non Spanish speaking Caribbean nations like Haiti, Granada, and the Virgin Islands. Florida was first dependent on slavery for agricultural production, but with the end of the Civil War and the expansion of the railroad the demand for Caribbean migrants to support an expanding industry grew stronger. Not surprisingly, sugarcane was the main crop and Caribbean workers with experience in this crop were ideal for work in its harvesting . A simil ar situation can be observed with citrus harvesting: once important sections of the Everglades swamps were drained and the mechanization of sugar harvesting stopped the flow of Caribbean workers, farmers interest ed in cheap labor focused their attention on experienced citrus pickers. Mexico has grown citrus for a long time and recruiters from large citrus companies have been present in citrus production areas for at least 20 years; many workers come from Veracruz, the largest citrus producing state in Mexic o . It is important to remember , as shown in Figure 4 1, that Florida alone is the third largest producer of citrus in the world, trailing Brazil and China, and far ahead of Mexico (FAO 20 12 ). The Political Economy of Agriculture The imbalanced U . S . Mexic o relations are of particular importance in the face of their economic ties and impact on agricultural production. For instance, the impact that the North American Free Trade Agreement (NAFTA) has had in rural Mexico over the past 17 years has been devasta ting. The agreement lowered, and sometimes
48 maize, beans, rice, milk, and eggs normally produced by Mexican peasants and small ranchers. The result was the destruction producers could not compete with heavily subsidized agriculture. By 2000, Gonzalez (2007) 3.5 million corn farmers cannot survive the 45 percent drop in maize prices ltural products to Mexico totaled 18.9 billion, the corn consumption (Turrent FernÃ¡ndez et al. 2012) . While the end result for the farmworkers in Mexico was to look for jobs outside rural areas, for U.S. agriculture it meant greater product demand thus access to cheap and experienced farm labor. Mexico went to a subsistence production and small scale marketing production covering the food demand of the nation to a network of migrants sending money back to Mexico to pay for the staples produced in the country they migrated from. The destruction of Mexican agriculture is related to the political cooptation of a corrupt and inefficient one party system in Mexico and the agricultu ral welfare system in the U.S. After the end of the Mexican Revolution , the caudillos formed a single party in order to keep disputes from getting out of control the National Revolutionary Party (currently known as the Institu tional Revolutionary Party or PRI) . The single party system ruled Mexico with different names for 70 years and was organized in three sectors: industrial, agricultural, and governmental. Each sector ensured power sharing among unions, farm producers, and a large governmental class. The National Confederation of Farmers (CNC) ensured that the PRI maintained steady control of the rural electorate in Mexico through incentives, corruption, or coercion. Now, due in large part to neoliberal policies rlos Salinas de Gortari, the CNC is the
49 weakest sector in the party . Structural changes depressed labor markets while offering huge benefits to close circles, like the political godfather of President Salina. s son is married to the dau ghter of the main benefactor of the maize open market, Roberto Gonzalez Barrera (Martinez 2012) . Gonzalez Barrera, the owner of Maseca, used the open market to access cheap maize coming from the U.S. while his competitors fresh maize producers were strippe d of the price guarantee regulated by the Mexican government. Then the Government stacked the market so much in his favor that it left millions of Mexicans with little choice but to buy (DePalma 1996) . In addit ion, Maseca offered incentives to tortilla makers if they converted their production from fresh maize to the use of maize flour Mr. Gonzalez's tortilla conquest represents perhaps the most blatant example of the cronyism and unfair (thou gh not necessarily illegal) practices that prevailed at the time Mr. Salinas was touting the The Mexican bureaucracy, with its paternalistic views, has only invested in rural areas when they have served its own interests. Direct government intervention in half of 1930 until the beginning of the 1990s when it was eliminated under NAFTA (Yunez Naude 2003). Since its cre ation in the mid 1960s, the National Company of Popular Subsistence (CONASUPO) played a key role in Mexican agricultural policies, shaping food production, consumption, and rural incomes, but it was mainly utilized as a political tool to bribe local commun ities and leaders, and in many cases enrich its directors. Raul Salinas de Gortari, who was the brother of Carlos and died in prison,
50 was accused among other crimes of diverting part of CONASUPO money to his own benefit like buying radioactive milk from Ru ssia to be re sold in Mexico (Zamora 1997). Beyond electoral times, when corruption is not central and government is disinterested in agriculture, peasants many of whom come from indigenous areas live in ignominy. No less damaging for the Mexican agricultu ral sector is the institutionalized American farming practice of exceptionalism. Agricultural exceptionalism holds that the farming industry is different from most economic sectors in modern societies as a result of unstable weather and market conditions (Daugbjerg and Swinbank 2006: 63 3). Furthermore, it has often been argued that contributes to broader national interests and goals (Skogstad 1998:468). The U . S . Congress declared that the oducts by farmers and ranchers is of C ode, Title 7, Chapter 56 [ 1998 ] ). In reality both concepts unpredictable agricultural markets and nationalism have been used to increase farm production and income at the cost of cheap and disposable labor. For instance, in 1936 Congress passed a law to improve the working conditions of the labor market, but the agricultur al sector did not have to comply with these requirements: for those who w ork in agriculture it is illegal to form unions, there is no overtime pay, restrictions governing child labor are more lax, retirement plans or health insurances are unheard of, and safety conditions are for the most part unregulated or badly enforced. As collective activities of agricultural enterprises while disallowing collective action from
51 farmworker s. Since the New Deal, the National Labor Relations Act has protected agricultural , (US C ode, Title 29 [ 1998 ]) leaving farmworkers without the possibility to unioniz e . In the immigration arena, these tools have been used just as often to recruit workers as to send them back to their countries when the economy sours (Gonzalez 2007). In particular, the Immigration Reform Control Act of 1986 (IRCA) increased the usage of farm labor contractor s who recru it for agricultural enterprises and determine farmworker salaries (Luna 1998). In Florida, such contractors provide most workers for small and medium growers. Statewide, the extended use of contractors to maximize income by m inimizing the cost of workers basically transfers the risks of agricultural employment to workers (Luna 1998), leaving them at the mercy of injuries and in some cases even slavery (Bowe 2007). Safety is also impacted by legislation and beliefs: The Occupat ion Safety and Health Act covers a wide range of safety issues in all work places, but has some farm related exemptions ( Runyan 2001 ), while American farmers see risk as inherent to their way of life (Sorensen 2008). Yet significant differences exist betw een the workers hired by contractors and the workers hired by the large producers themselves more a symptom than a basic cause of the difficulty, the reason being the lack of regulation and inspection in the sector. Perhaps agricultural exceptionalism was once valid, particularly when most farming was conducted by small family operations. However, this is no longer the case. The number of farms in the e astern U . S . has declined from 5.3 million in 1950 to 2.1
52 million in 2002 while the average age of principal operator continues to increase and the number of family members living and working on farms declines . Thus agriculture has become more commercial in nature , yet the laws regulating agricultural l abor remain based on the antiquated family farm model (Arcury and Quandt 2009:19). Technological advantages (machinery and chemicals) have reduced the need for labor force in the fields, but this is not the case for all crops. Fruits and vegetables continu e to be crops with high demand for workers. The Scientific Evidence: Health, Safety, and Justice Green Revolution and an increased demand for standardized produce translate to exte nded use of chemicals in the fields. A growing body of scientific literature has uncovered the heavy health toll paid by farmworker s, but little has changed in terms of regulations or enforcement. For example, suspicious cancer clusters have been detected in the farmers of California since the 1970s; in the rural town of Fowler between 1981 and 1984 children were diagnosed with leukemia at an incidence thirty five times the anticipated rate for a town this size (Luna 1998:502). In Florida, for instance, the State Department of Agriculture has a mere 1 5 field inspectors to regulate, inspect, and enforce environmental law (Estabrook 2011) in 47,500 commercial farms utilizing 9.25 million acres of land (FDA S 201 1a ). Socially and economically, farm work is cle arly at a disadvantage. Most farmworkers have incomes that place them near or below the poverty level. Farmworkers wages seldom exceed minimum wages, and at times fall below minimum wage. Most farmworkers also lack an understanding of English and of the s carce safety regulations that do exist .
53 when it comes to protecting its farmworkers, employing only about fifty inspectors we know hurt appro Estabrook 2011: 41) that allowed for poor enforcement of regulatory practices, including safety training for workers. Because, the Florida Department of Agriculture and Consumer Services (FDACS) is also the entity in charge of promoting the State agri Shelly Davis and Rebecca Schleifer of the Farmworker Justice Fund have abuse investigation and eyewitnesses outside of the presence of supervisory personnel and with adequate translation, failed to obtain relevant medical records, routinely a ccepted uncorroborated employers claims of compliance, used checklists as a substitute for thorough on site inspection, and ignored evidence of employer retaliation Agriculture , by nature, is a dangerous industry, with rates of occu pational fatality and injury that are seven times the national average ( BLS 2007). Some strategies to improve the working conditions of workers are based on research, education, and legislation. Unfortunately, most actions have limited impact, compared to the political lobbying and economic leverage the industry has when regulations are at stake. The actions of community academic alliance have had little transformational swing to result in any real change on the fields while the American public is largely u naware of the conditions under which its food is produced. Social justice, human rights, and public health should be arguments to engage the American consumer and their
54 representatives on the inequalities and risks associated with farm work while tools lik e Facebook and Twitter should be seriously considered as platforms for carrying out these arguments. Social justice is a process that seeks fairness or equity in the distribution of social burdens and resources across all social groups, and provides all pe ople the opportunity to realize their full potential, which includes working and living in environments where health and safety hazards are addressed, earning a living wage, living in communities free of discrimination, and having access to appropriate hea lth, education, and social services (Arcury and Quandt 2009). Social justice is based on basic h uman r ights principles and involves progressive taxation, income redistribution, economic egalitarianism, and equal opportunities ( McLauglin and Baker 2007 ). Th ese principles are allegedly embedded in the US constitution and in the International Labor Organization (of which the US is member since 1934), but their general application has , Wiggins, an d Quandt, (2009) is composed of three arms : f irst, efforts must be made to change the perspective of American consumers about the human cost of producing the food they eat ; s econd, research needs to systematically document the characteristics of farmworker s, the conditions of farm work, and the injuries and illnesses experienced by farmworkers to develop targeted policy and programs to improve safety and justice ; t hird, specific changes in policy and regulation will be needed, including improving farm labor laws, increasing enforcement of the laws, and increasing support and protection for workers who organize.
55 In th e 2011 annual meeting of the American Academy of Health Behavior , Dr. John Lowe a renowned health educator in the tobacco prevention field, sug gested that in examining all that has been done to curb the epidemic, it was clear to him that more so than education, regulation has had the greatest impact in reducing the use of tobacco products . He did not ignore the importance of education in its suc cess with prevention, but he concluded that from the epidemiological trends and the timing of restricted smoking laws in the United States , the correlation was so clear that he now emphasizes legislation and enforcement along with education in his effort t o prevent skin cancer and obesity as well. Likewise, if any change is to take place on the farm fields, appropriate legislation and strict enforcement of regulations protecting farmworkers are dire necessities . Debates about the cost of illegal immigration in the U.S. were rampant even before the 2008 financial crisis (Simon 1999; Tanton 1996; Vernez 1996). However, until the general public realizes the public health impact that not just the region , but the country as a whole, has had o n the structure of ag r i business, little can be done in order to rever se the trend. Poverty is growing even as free trade increases, perhaps because, as Bowe (2007) concludes, the super rich want an egalitarian society in which the world population works cheap, consumes cheap, and dies young. If it is at all possible to change this apocalyptic vision, society must undergo a complete transformation to ensure that each person values food production over consumerism; someone who is more concerned with environmental conditions than with cheap and beautiful produce; individuals who stand for human rights and not racist prejudice based on alleged nationalism.
56 From a less utopian stance, academicians and farmworker organizations should keep working to provide the general public with s ound data that prove workers suffer harsh and discriminatory practices in order to pressure legislators and regulators. Data that show farmworker s are not a cost to the American economy, but instead a pillar of the country should be made available to the g eneral public to apply that same pressure . Finally, in the international arena, countries like Mexico should encourage and be supportive of their local food production and less invested in wars that are in the interest of mainly U.S. corporations and not t heir farmworkers . In the meantime, important steps must be taken to recognize farmworkers and improve their working conditions everywhere. This progress should happen from multiple perspectives: the academic, the industrial, and the advocacy arenas. Let us take the specific case of a community based prevention market program conducted by the Florida Prevention Research Center at the College of Public Health in the University of South Florida to promote safety glasses for citrus pickers. The program worked w ith community based organizations (CBOs), orange growers, and local health departments. During this process, when one of the growers realized that one of the CBOs had complained of pesticide violation ( as a result of pesticides being applied in fields next to where workers were present ) the grower simply left the project. The program was only able to work with large companies, since contractors were unable and unwilling to participate. Besides the pesticide violation, researchers and community members witne ssed other violations and did not report them in order to keep the project alive. It was also clear that safety regulations were not enforced and that piece rate payment made workers disregard most safety standards.
57 Piece rate means that workers are paid by the amount they pick. Orange harvesters can pick up to 2 tons of citrus a day, using canvas bags capable of holding up to 9 0 pounds. Few companies provide safety equipment or even harvesting tools, like the canvas bags or gloves, much less safety glasse s. Moreover, safety glasses tend to fog in hot and humid conditions, slowing down picking rates. Technology is available to prevent glasses from fogging, yet it is not adopted because its cost reduces profits. Some years ago, Margarita Romo , director of th e advocacy group Farm Workers Self Help Inc., mentioned that local legislation was going to force citrus companies to provide canvas bags for harvesters; but orange pickers in small crews told us they are still being charged for the canvas bags. Safety con ditions of workers depend on the laws that directly or indirectly encourage safety strategies. Workers without proper residency status in the United States are regularly discouraged by farm labor supervisors to report any kind of violation with the argumen t that any investigation will bring light to their legal status. Then, they would be deported and economic consequences would be greater than the risks they are exposed to in their daily life. And so, they remain silenced. Ending agricultural exceptionalis m is one realistic approach, but it needs to be followed by comprehensive migration reform and enforcement to ensure compliance with all laws affected by both changes in policies (exceptionalism and migration). The cost of not acting will not only permanen tly impoverish the entire population, but also will continue to erode the principles this nation stands for. Connecting Dots: Suggestions of Social Network Analysis Social networks in the citrus industry are everywhere: from the big conglomerates that mana ge most of the orange juice distributing in the country to the
58 growers who own the trees. The spectrum is wide and also can include the government departments and agencies that regulate production and distribution, the crew contractors that provide most of the harvesters collecting the fruit, and the farmworker s who hand pick the fruit. The universe of analysis can range from big companies like Tropicana Pepsi Co., Minute Maid operations producing citrus in on e county. We can research how growers interact locally by looking at their regional association or we can research how growers interact at the state level by looking at the structure of their state organization, the Florida Citrus Mutual. The citrus indust ry includes complex linkages at the international level that can also be analyzed; for example Coca Cola and Pepsi use citrus juice concentrate from Brazil and Mexico in their products. The citrus industry in Florida is so large that according to Florida Citrus Mutual the industry generates roughly 76,000 full time and part time jobs, providing 30 percent of ions a year (FD O C 2010 ). The possibilities for analyzing any big industry a re enormous. However, we also have to understand the restrictions that reality can impose on the collection of network data. For instance, the imprecision of the number of workers reported by Citrus Mutual (roughly 76,000) reflects one of the main obstacle s faced by any inquiry into the social extent, composition, and structure of the labor force employed in the industry: its farm harvesters. In migration studies, estimating the exact number of people moving in and out of the host country is nearly impossib le. In the particular case of farm work migrants in the United States, the situation is exacerbated by the constant entrance of disenfranchised
59 farmworkers coming through the Mexican border. For the most part migrants from Mexico, who are the bulk of most agricultural operations in the US, lack legal status. Thomas Arcury and Sara Quandt (2009) estimates that at least half of these workers in the East coast of the United States do not have proper working permits. The legal situation of workers decreases the trust afforded to researchers while the industry is also mistrusted in inquiries regarding its employees. For some scholars, the migration process has been a planned strategy of the capitalist system to access cheap and disposable labor (De Genova 2003; Gonzalez 2007), while others see it as the natural development of economic growth and the American dream myth (Tanton 1996). Regardless of the most appropriate explanation for the condition of the significant number of undocumented workers in the United St ates this condition extremely limits the scope of scientific studies. The limitation to collecting social network data increases by the fact that the industry has a high level of worker turnover due not just to their undocumented status, but to the nature of the agricultural citrus industry itself: m obility of workers is demanded by production. For example, some workers may start picking grapefruit or limes in September October and may end up picking Valencia oranges in June July. Changing crops does not o nly imply different locations, but also diverse companies and even operations. In Florida, we can encounter two complementary harvesting systems: large companies who manage their own crews, have their own machinery, and hire workers directly, even with the use of guest workers brought in from Mexico; meanwhile there are smaller growers who rely on independent contractors. Some mechanization is also in place, but while they harvest less than 5 percent of the total production, smaller
60 growers still require pi cking crews that collect the remaining fruits hanging in the trees thus workers are embedded in the same scheme whether hired by a company or contractor. In social network analysis we encounter two kinds of approaches: whole/complete network or sociocentri c network analysis and personal or egocentric network analysis. In general both perspectives can be complementary viewpoints on the same data (Marsde n 2005), but for the most part , particularities arising through sampling and data selection mean that it is rarely possible to move with any ease from studies are concerned with the structural properties of networks at the global level, whereas egocentric studies focus on t he network as it appears from the standpoint of those situated at particular locations within it (Carrington 2005:3). Additionally, socio centric network analysis, like egocentric network analysis, can have a 1 mode or 2 mode structure (Wasserman 1994). Th e first case refers to the connection between persons or organizations, while the second makes reference to a combined link between persons or organizations and events or facts. Each of the four quadrants of these approaches can be utilized to understand t he use of safety glasses among migrants involved in the citrus industry, and naturally each will have particular advantages and disadvantages in doing so. The usage of social networks is extensive , p articularly so in the last 20 years . T hey have been used to explain a variety of social phenomena including the diffusion of innovation, the assessment of social capital, and the usage of social support or the influence of peers. Commonly, network data are collected and attention is focused on
61 the structure or c omposition either using the sociocentric or the egocentric perspective. Pointing out particularities in each perspective, Marsde n (200 2 ) notes that whole network studies examine sets of interrelated objects or actors that are regarded for analytical purpos es as bounded social collectives, while egocentric studies focus on the focal actor or object and the relationship in its locality. Egocentric and whole network designs are usually distinguished sharply from one another, but they are interrelated. A whole network contains an egocentric network for each object within it (Marsden 2002). Conversely, if egos are sampled densely, whole networks may be constructed using egocentric network data (Marsden 2005). Social network analysis, when applied towards understa nding social influence, examines the social foundations for influence; it is strategic since it links the structure of social relations to the attitudes and behaviors of the actors who compose a network (Marsde n 1993:127) by appealing to specific features of interconnections among the elements (Laumann 1979). Network analysts have developed a distinctive approach to social influence that entails a structural conceptualization of social proximity. The general hypothesis is that the proximity of two actors in social networks is associated with the occurrence of interpersonal influence between actors. During field work in the groves of Florida I witnessed this precise phenomenon: peer health educators were hired and trained in order to increase the use of safet y glasses, and when observed in the fields using their glasses, usually their peers were also using safety glasses. One way to see it is in the context of diffusion of innovation; Burt (1987:1288) describes the social structural circumstances that make ego (the person) and alter (his/her network
62 adoption. In the same field of diffusion of innovation, Valente (1999:38) suggested that individuals who are connected to one a nother in a group are more likely to share information with each other and thus reach common understanding and perceptions when faced with a product or idea. Consequently, individuals in the same group can be expected to adopt roughly at the same time. Whi le specific network data was not collected in the project with citrus harvesters, it was nevertheless clear through our participant observation that many of the peer workers adopting the use of safety glasses were also living in the same migrant camp, or e ven in the same room. Marsden (1993:128) notes that Cartwright (1965:3), following March (1955) and substantive processes that underlie influence are diverse and for Marsden (1993:128) include relations of authority, identification, expertise, and competition. Lippitt, Polansky and Rosen (1952:37) called this behavioral contagion or spo ntaneous and it is an imitation by other actors of a behavior initiated by one member of the group . T he initiator did not display any intention of getting the other to do what he did , but affect ed their perception, making that behavior acceptable. I n the c ase of citrus peer educators getting in fields with the expressed condition of wearing safety glasses at all times. The use of peer influence and social networks analysis in scientific journals has mainly focused its attention on risky behavior (tobacco a nd alcohol use, substance abuse, delinquency, unprotected sex, bulimism) concerning adolescents. In a recent
63 search of journals published between 1990 and 2011 (Science Direct) 16,831 articles were identified using the words peer influence and social ne twork analysis, most of which were concerned with adolescent behavior (Brandy 2009; Guryan 2008; Levy Paluck 2011 just to mention a few recent ones dealing with positive peer influence), but some of which include network modeling (Moody 2001), academician influences (Rawling 2011), and adult sexual behavior (Yang 2010). Perhaps one reason behind the inclination towards studying peer influence in adolescents lies in the fact that susceptibility to peer influence in adolescence follows a bell curve, increasi ng during early adolescence, peaking around age 14, and declining thereafter (Steinberg, 2007). Moreover, it is not just the fact that influence can be more effective, but that some behaviors adopted at this age can have long term consequences. It is pos sible peer influence is more effective during adolescence than later stages in life. As Rawling (2010) and Yang (2011) demonstrate, peers act with the perception of what their equals are doing. Now, for the most part peer influence studies are based on an ego centered network approach, but some should be considered whole networks when a school or a classroom is clearly defined as the universe. During the implementation of the Partnership for Citrus Workers Health (Monaghan 2008), annual surveys consistently estimated the mean age of workers at around 27 years of age with high levels of homophily, which may explain how, without being adolescents, these workers were affected by a peer to peer program. Homophily is the tendency of people from an associated bond with individuals similar to them; similar background, educations, even taste. These effects can be measured and accounted for using social
64 network analysis with different levels of success according to the use of the four quadrants outlined above. Selecti ng the kind of network analysis we want to use is determined by the kind of question we want to answer. A one mode whole networks analysis can help us understand, for instance, how the structure of the citrus industry functions, or how communication works among Citrus Mutual members. A two mode whole network analysis in which the use of safety glasses is linked to the various levels of social exchange in the workplace is an ideal approach that, unfortunately, would not be feasible considering the realities of the agricultural industry and the labor market. In general, this kind of social network analysis could be a powerful method in which the use of personal protective equipment and the interaction of workers, supervisors, and employers are explained. Reg ardless of the limitations described, socio centric networks can be collected from these workers if the boundaries are clearly established; for instance if the units of analysis used are the crews or the companies (as mentioned earlier), or even the towns or counties with the highest levels of orange production. Crews and companies may be the easiest mechanisms for data collection in these settings and could even be used in a more sophisticated way to evaluate over time the impact of the community health wo rkers in the Partnership for Citrus Workers Health to identify the concentration of power, enforcement, or other resources affecting the use of safety equipment. A s noted earlier, setting the boundaries for a whole network analysis tends to be artificial a nd unclear ; in this case crews as will see are well established boundaries.
65 T he whole network approach can imply limitations to the extent to which the workers link form and interactions outside the established limits we set for the data collection. Perso nal networks or ego networks , on the other hand , expand the focus on the interactions that workers may have with other peers outside their crew, company, or even region. U sing personal networks we can distinguish three types of approaches: relation based, position based, and resource based; all with particularities appropriate to the question we want to answer. For our particular case, relation based could have be en the mo st appropriate since most workers may not differentiate themselves from a position or a resource point of view. Particularities of all ego networks include homophily, size, average strength of ties, heterogeneity, density, composition, and range. For McPherson et al. (2001) , ications for the information they receive, the attitudes they form, and the interactions they experience; opportunity to increase specific strategies for safety under a peer influence model. Two mode egocentric social network analyses over time can suggest proper strategies for peer educators to place them in effective positions. Health education has focused its interventions on the transition of knowledge and practices, but has not fully taken advantage of the ways in which this knowledge is shared. Social marketing has overcome some of the early barriers that public health and epidemiologic perspectives had; now, it is time that other aspects of marketing are enhanced i n order to improve health and safety conditions for the most disadvantaged populations.
66 Finally, in the four quadrants of social network analysis (one and two mode egocentric and socio centric approaches) tasks to be addressed when using these analyses to look at the influence effect should include a) elucidating the substantive process that underlies claims that there should be structural effects on the attitudes and behaviors of actors, b) defining interpersonal proximity in a network in an appropriate ma nner given these processes, and c) assessing the predictive success of the approach using available mathematical and statistical models of social influence processes (Marsden, 1993:128). Including social network analysis in community health worker s (CHW) programs, such as the Partnership for Citrus Workers Health, can be highly beneficial for different reasons: First, it can help identify the more potentially effective promotor based on the position held in his crew, either by bridging ties, centralizing information, or even having the most amount of ties. During the selection of current promotores , we observed early adopters people using safety glasses and invited them to the program; we also asked crew leaders to select natural leaders amongst the crews. The results were an uneven group of CHWs with different levels of impact in their crews; with a more accurate model to select these workers we may be able to increase our impact. Second, longitudinal social network analysis can offer a better evaluation o f how CHWs progress in two areas: at the interior of their crew, by looking at how they are gaining influence among peers, and at the personal level, by looking at how they are concentrating knowledge and redistributing it to their crews. Third and finally , ego centered social network analysis can facilitate the development of effective ways of placing CHW in communities to improve their actions; if networks can help us understand the best ways
67 to communicate our messages we should use them to our benefit. H omophily , is somewhat like agreement on cognitive anthropology; in the end, it is a manifestation of culture. Understanding culture as a set of shared beliefs and values behaviors, and attitudes, homophily is a by product of wanting to bond with people wh o share in this culture and culture marks actions as much as the context of our reality. Workers share origin, vocation, crew, transportation, and labor camp, but deciding who they form relations with is up to who they sympathize with or who can bring them (Scott 200:82), these measurements will help understand how individual actors are positione
68 CHAPTER 3 METHODS: THE PAD FOR COLLABORATION Introduction Science creates theories, methods and, sometimes, solution s. Complex realities could, however, disprove theories and deem methods limited in scope. For a long time now, multiple disciplines have collaborated in the search for better solutions. In this process, the use of multiple methods is also common practice. are collected, the better picture we may have of the phenomenon we are trying to describe. The present study uses data from two distinct sources. The f irst one comes from the long standing collaboration of the Farmworker Association of Florida with the University of South Florida (14 years). The second one, a whole season (9 months), comes from my independent research using social network data and ethnog raphy in the citrus fields of Florida. Because the data collected under the umbrella of the Partnership for Citrus Workers Health (PCWH) was designed, collected, coded, and analyzed for members of this collaboration including this author I will refer to it as secondary data. Primary data will refer to data collected on the citrus season of 2010 2011 since th ese data w ere specifically designed, collect ed , cod ed , and analyzed by this author. The secondary data complements the primary. Access to the data alre ady collected by the PCWH and storage at the University of South Florida (USF) was granted by Carol Bryant Ph . D., Director of the Florida Prevention Research Center (FPRC). Because my study was concurrent with my work at the PCWH, my primary data collectio n was approved by the Institutional Review
69 Board of the University of South Florida , where the FPRC is located. Secondary data used, either collected or not by this author, was reviewed, cleaned, coded or recoded, and analyzed or re analyzed when needed. T o clarify the process in which the data was collected and how the different actors contributed and expanded the PCWH program we developed three logic models. Logic Models The logic model adopted by the PCWH can be seen in three distinctive and complementar y dimensions: as a sequential process (Figure 3 1) as a mixed method data collection and analysis mode (Figure 3 2), and as a collaboration scheme (Figure 3 5). Given the complexity of the process, the three models are explained individually, even though i n practice the work was carried out in an integrated fashion. Figure 3 1 . Sequential research lo gic model Ethnography Surveys Social Network Analysis Script Analysis Safety and Health Strategies Pilot test Quasi experimental test Greco Latin square glasses test Surveys Evaluation Participant observation Focus Groups Glasses test Surveys Community Health Workers trainning Evaluatiion Formative Research Community Advisory Board Capacuty buiilding Literature review Formative research and intervention design planning Partnership for Citrus Worekrs Health Social Marketing Community based Participatory Research Disease Prevention and Health Promotion Comuty based Prevention Marketing
70 Initially, the idea to develop a community based partnership to improve public health practices of citrus workers came from Florida Prevention Res earch Center director Carol Bryant, participating in the Together for Agricultural Safety project which sought to reduce the adverse health effects of pesticide exposure among fernery and nursery workers (Flocks et al. 2001). Given the relevance of the cit rus industry in the state and with the purpose of testing the Community based Prevention Marketing model fully on disadvantaged rural communities, the FPRC joined the FWAF proposing a research intervention program to improve the conditions of harvesters. F ollowing this model, the process started with the formation of a community advisory board (CAB) composed by farmworkers; citrus industry managers; academic, advocacy, and service group representatives. The first task of the PCWH was to build the capacity o f its members on social marketing techniques that included the selection of the intervention and data collection. 2) the formative research of the project followed an initial exploratory mixed method design for the taxonomy of the social marketing intervention, while the evaluation used an experimental embedded approach. The final methodological design used for this project to identify the reason behind the adoption of safety p ractices employed a concurrent mixed method strategy (Figure 3 4). The end product represents more a hybrid mixed method design as defined by Harrison and Reilley 2011, being that it involves a combination of all the mixed methods used.
71 Table 3 1 . Mixed Methods Design (Source Harrison and Reilley 2011) Design Type Variants Timing Mixing Concurrent Convergence Concurrent Merging the data during the analysis Embedded Experimental Correlational Concurrent or sequential Embed one type of data in the other Explanatory Follow up explanations Sequential: quantitative followed by qualitative Connect the data between the two phases Exploratory Instrument dev. Taxonomy Sequential: qualitative followed by quantitative Connect the data between the two phases Figure 3 2 . Hybrid m ixed method design logic model PCHW is the centerpiece of this work; for that reason, it important to outline the schematic collaboration development of the program. Figure 3 5 illustrates this logic model and outlines the process o f data collection, corporate and institutional involvement, and the outcomes of the intervention . The dynamic created by the intervention of multiple institutions created special conditions to successfully implement strategies for health and safety. The di rection of the arrows signals the influence of each component to its proxy and the years of implementation are included in parentheses. If there is no year listed, this means the process is still ongoing. Formative Research Participant observation Focus Groups Surveys Evaluation Pilot test Quasi experimental test Greco Latin square glasses test Safety and Health strategies Ethnography Surveys Social Network Analysis Script Analysis Data Analysis Descreptive Inferential bivariate Script analysis Social network analysis Logistic regresion
72 Figure 3 3 . Schematic collaboration developed of the PCWH program
73 An important distinction is that the ethnographic study was only conducted in the small crew, whereas no such study was conducted for the large scale harvesting crews. For over 10 years the PCWH has been an evidence based intervention program that promotes the adoption of protective eye safety equipment (PPE) among Spanish speaking farmworkers of Florida. At the root of this program is the systematic use of community based preven tive marketing (CBPM) and the training of community health workers (CHW) among citrus harvester using popular education. CBPM is a model that combines the organizational system of community based participatory research (CBPR) and the strategies of social m arketing. This particular program relied on formative research data using a mixed methods approach and a multilevel stakeholder analysis that allowed for rapid dissemination, effective increase of PPE usage, and a subsequent impact on adoptive workers and companies. The Florida Prevention Research Center (FPRC) in the College of Public Health at USF is one of 37 centers that form the network of Prevention Research Centers funded by the Centers for Disease Control and Prevention (CDC). The centers were crea ted to conduct prevention research and are leaders in translating research results into policy and public health practice ( CDC 2013). Aware of potential impacts to citrus , the FPRC committed itself to the implementation of a program to prev ent injuries and also to test the feasibility of a CBPM in a rural area, with low literacy workers, and traditionally opposed groups like agricultural management and farmworkers advocacy groups. CBPM is a data driven model that combines community organizat ion principles
74 placement, and promotional strategies to design, implement, and evaluate health interventions (Bryant et al . 2000). The use of mixed methods in marketing researc h is well documented, though it is rarely recognized or cited as such (Harrison and Reilly 2011). Similarly, the field of social marketing relies on an array of quantitative and qualitative methods, but the systematic use of mixed methodologies is not usua lly recognized or cited as one of its strengths. The CBPM in the PCWH followed nine steps: 1) mobilize the community ; 2) develop a community profile ; 3) prioritize and select target behavior ; 4) build community capacity ; 5) conduct formative research ; 6) d evelop marketing plan ; 7) develop program materials ; 8) implement program ; and 9) monitor and evaluate (Monaghan et al. 2008) . Each step, with the exception of points 1 and 4, imply some degree of data collection and analysis. Because the marketing plan de pends on the formative research and the effectiveness of the program is based in its monitoring and evaluation, the use of mixed methods appear natural. Prior to my arrival at the PCWH in 2006 , the project conducted quantitative methods in the form of int ercept interviews ( n = 58), post intervention surveys ( n = 185), attendance records, and non random site visits ( n =108). Qualitative data was collected as key informant interviews ( n = 10), focus groups interviews with workers ( n = 50 in eight groups) and CHW ( n = 5), attendance minutes, and participant observations. Once I was part of the project we evaluated the program with a dditional mixed method approach es like annual surveys ( n =330 in the largest sample of any season in 2006), a Greco Latin square d esign test , a quasi experimental test, field observations , and CHW
75 reports on the quantitative side ; in addition to face to face interviews with CHW; and participant observation on the qualitative field . Except for some key informant interviews, all data were collected and recorded in Spanish. FPRC personnel analyzed the data and prepared summaries for the FWAF and PCWH boards to review, provide feedback, and offer interpretation. Focus group transcripts, individual interview notes, and field notes were tr anscribed, translated, and conceptual framework (product benefits and costs, placement, and promotional wear. Observational data were tabulated by hand (Monaghan e t al. 2012) . All closed ended information of survey instruments, observations, and the Greco Latin square and quasi experimental tests followed standardized statistical analysis. Partial results of the data collected and analyzed by the PCWH have been published multiple times (Tovar Aguilar et al. In press; Monaghan et al. 2012, 2011, 2008; Monaghan 2011, 2008; Bryant et al. 2009; Mashburn et al. 2009; Luque et al. 2007). The main focus of these wo rks is to explain the implementation process of the program (Tovar Aguilar et al. In press; Monaghan et al. 2012, and 2008; Monaghan 2011; and Bryant et al. 2009); however, some reported the effectiveness of the program (Monaghan et al. 2011, Mashburn et a l. 2009), and other highlight the safety perceptions of farmworkers (Monaghan 2008, Luque et al. 2007). The current study does not pretend to be an integral analysis of the PCWH work or a historical recount of its implementation, but will tap into data bot h published and unpublished.
76 Following the mix ed method s classification of Creswell and Clark (2007) the PCWH m e rge d , connected, and embedded procedures over its course . Methodologically, this means that a) on occasion , results are a combination of quanti tative and qualitative data; b) at other times , data coming from one method is use d to construct the other and; c) sometimes one method quantitative or qualitative is embedded in the other, as is the case when open ended questions complement surveys. The a rray of multifaceted mix ed method s used here, however, is closer to explanatory, concurrent, and embedded designs. Th e basic premise for the use of mixed bination of quantitative and qualitative approaches provide a better understanding of research usage of this methodology, I was able to appreciate how quantitative data can show some trends on the general demographic of the workers, their injuries, and strategies to deal with them , while qualitative data deepens the understanding of motivations, beliefs, and practices of the workers when dealing with risky work conditions . For the same reason, my primary data also depends on an explanatory concurrent mixed method design. Primary Data: The Settings of the Native Expert The orange harvesting season is a nine month long process that starts with Navel oranges, Orlando tangelo s, and Red and White grapefruits in late October or early November and ends with Valencia oranges and Pink grapefruits in late June or early July. From October 2010 to June 2011 additional field observations, surveys, ego centered s ocial n etwork s ( n = 3) , and participant observation was conducted in central
77 Florida. Quantitative and qualitative data were collected concurrently to explain the strategies follow by workers to avoid injuries and reach for health care when injuries occurred. Field observations, surveys, and ego centered s ocial n etwork s data were collected with the support of one large harvesting company, who managed ten harvesting crews and occupied around 300 guest workers (H2A) from Mexico. The company requested support from the PCWH in 2009 to overcome a sharp increase in the management and they agreed to pilot the program in three of their crews located in two counties (Polk and Okeechobee). Figure 3 4 and T able 3 2 , respectively, shows the geography and extent of the present work. Most of the principal citrus producing counties serve in one way or another as fieldwork locations; although there are several citrus towns that were central locations of this wor k. It is also important to notice that even when workers live in a particular county, many times they are transported to groves far away to harvest ripe citrus. Noted on Table 3 2 are the types of data collection used in each county and if the data was co llected by the PCWH (secondary data) or if it was collected by the author (primary data). The different kind of data collected is listed in the last column and a detailed recount of each method used is given in each chapter; however, it is important to dis tinguish here the difference between field observation and participant observation. While the first one refers to visits in the groves in which the crews were harvesting and the sole purpose was to either see workers with PPE or test the PPE itself, the se cond one refer to moments in which the researcher trained, interviewed, or mingled with workers, supervisors, or either other experts such as extension or healthcare officials.
78 Figure 3 4 . Citrus production by county, 2009 2010 (Source: USDA NASS 2011)
79 Table 3 2 . 2009 2010 Main c itrus p roducer c ounties in Florida and data collection Counties by Highest Production Citrus Production Prim. or Sec. Data Research Activity 1. Polk 27,875 Primary Field Observations, Surveys, SNA 2. Highlands 21,592 Prim. & Sec . F ield Observations 3. De Soto 17,956 Secondary Field Obs. 4. Hend r y 16,330 Secondary Field Observations , Surveys, Participant Obs ervations 5. Hardee 13,027 Secondary Key Informants Interviews 6. St. Lucie 10,219 Primary Participant Observation 7. Indian River 9, 603 Primary Participant Observation , Ethnography 8. Collier 7,416 Prim. & Sec. Field Observations 9. Manatee 6,111 Secondary Field Observations 10. Lake 4,438 Prim. & Sec. Participant Observation 11. Martin 3,132 Primary Participant Observation 12. Osceola 3,11 5 Primary Participant Observation 13. Hillsborough 3,023 Prim. & Sec. Data Depository 14. Pasco 2,902 Secondary Field Observations , Surveys, Participant Obs ervations 15. Charlotte 2,680 Secondary Field Observations 16. Lee 2,493 Primary Field Observations, Sur veys 17. Glades 2,132 Secondary Field Obs. 18. Okeechobee 1,678 Prim. & Sec. Field Obs., Surveys, SNA, Participant Obs. Adapted from : USDA NASS 2011. All ten crews at the company who collaborated with the PCWH received basic eye safety training and the PPE recommen ded by the program at their arrival for the citrus season 2010 2011. In the three company selected crews, this researcher
80 recruited and conducted additional trainings for CHWs on their houses. The number of farmworkers trained as promotores was similar in two crews (A and B = 3), but in crew C one third of its members requested the training ( n = 10). The number of participants differ, I argue, as a direct relation to the composition of the crew and their location their background, social networks, and socia l context. The logic model allowed the researcher to cover all Florida commercial citrus production areas (Figure 3 5 ) and setting s . Figure 3 5 . Florida commercial citrus production areas (Source: USDA NASS 2011 )
81 Unannounced field observation was conduct ed once with the crews that did not receive additional training; pilot crews, on the other hand, were observed on three occasions. During these observations of the pilot crews, all workers were interviewed on site, and their use of safety glasses was recor ded. Social network interviews were guest workers program (H2A visas) the company had to provide housing and transportation for them; this fact facilitated the intervention and research logistics. While quantitative data followed institutional channels, qualitative data came from my incorporation into a small harvesting crew composed mainly of undocumented farmworkers scattered throughout a small town. Participant observation in volves lived and worked alongside citrus harvesters during the majority of the picking season. Audio field notes were recorded each evening while in the field and were transcribed when I returned to my permanent residence in Gainesville , Florida . Ethnography: Liv ing the Citrus Worker Experience Introduction The initial process of becoming a citrus picker presented singular challenges; the first of which was my condition as a native anthropologist. While native anthropologists have distinct advantages, such as bein sensitive with cultural norms, and to have easier access to and more profound corollary challenges as well. Zentella describe s one such challenge
82 What I discovered as a native anthropologist in the fiel d is that the term itself is relative even among nationals. After several years of working with citrus pickers, I realized my distance from the workers in a series of layers. First, in most cases, I was introduced to the crews as a safety specialist by ma nagers or crew leaders; I was not only more educated, but I was given a superiority or a supervisory place among them. Second, d espite being Mexican, as were most of the citrus harvesters , being originally from Mexico City did not necessarily facilitate in tegration , most of the workers came from rural areas and we re distrustful of city people who have a reputation for being lazy and corrupt. Third, my mobility never created a sense of belonging among them, not just in terms of self transportation, but in t erms of legal status. For my integration in the small crew, h owever, I avoided at least the first condition. The crew leader only introduced me as a new and novice worker; I was not the expert. They were the experts to whom I asked all my questions. Mainly I wanted to know what a new citrus farmworker should know in order to make money and not hurt himself in the ordeal. I did have some advantages given my native condition and previous experience . First, I have some close family members dedicated to agricul ture in a small town in Mexico and I even farmed a couple of years in Yautepec, Morelos. Second, for professional and personal reasons, I have visited 30 out of the 32 States in Mexico and all the countries in Central America. This familiarity allowed me t o better relate to the places the workers were coming. Finally, my previous experience with the
83 PCWH meant knowing the citrus harvesting process and its argot , which also facilitated connection with the workers and inhabitants of town. Acceptance into the crew I worked was evident by the changes my nickname underwent over time. My regular urban , desk job work was evident to my new co workers the first week of work. They called me referencing how my face bore no evidence of t he hard life lived outdoors, exposed to slang demonym for a person born in the suburbs or surrounding areas of Mexico City . In the Antonio el de los lentes , a simple descriptive term to differentiate me from the other Antonio in the grove . The first ch allenge was to explain my true intentions to the gatekeepers of citrus production. El chivero as the orange crew leaders are known did not understand how or why I would have to spend some months collecting oranges to complete a study. The fie ld supervisor the boss of my boss and his son did a long interview with me before I started working to understand why I wanted to be there picking fruit since I was obviously well educated (the inference being that I was above that kind of work). Of course , he as well as many other people in the industry was concerned about work place violation complaints . Being aware of this, I made sure to declare: just hap py to know that I was not expecting to be paid by hour , but by the piece rate for whatever I could pick ; like any other member of the crew . The supervisors were not the only ones surprised by my desire to participate alongside the workers. Even the workers themselves did not see why I would subject
84 myself to such heavy work even for one day when I had advantages they did not. My knowledge (university degrees), and even m y very old car (mobility) represented my flexibility to do any other kind of job or move to a place with better opportunities. In fact, many months after my fieldwork ended, I ran into the owner of the grove s where I worked at an industry event and he also asked many of the same questions I had already answered to his foreman when I was hired. My relationship with the workers, as with the evolution of my nickname, grew as the season passed and I was able to pick as much citrus as some of the workers in the crew. But these were not my first friendships among harvesters; as with many workers, months picking oranges, created bonds to the extent to visit their families in Mexi co or have my family staying in the workers homes. It is true that my nativism was an advantage in developing my ethnography, but also helped me recognize the differences that pre exist at the interior of each culture. If this level of self identification with my participants rose between deep understanding of harvester culture and the liability of its interpretation, triangulation of data and awareness of the situation are the strategic balance I followed. Location The geographical coordinates of the locat ion selected for the qualitative data collection are 27Â° 46.00 N, 80Â° 35.55; one hour east of Orlando and three hours north of Miami. Fellsmere is an inland town in the eastern coastal Indian River County. One of many locations where tourism, retirees, and wealthy families keep pushing urban development, while the rest of the economy depends on agricultural production.
85 Although the incorporated city of Fellsmere has two large tacks linked by a short stretch of highway only the west side is populated (Figure 3 6 ). Figure 3 6 . I ncorporated and unincorporated areas in India n River County , Florida , Fellsmere highlighted in red (Source: Wikipedia Commons 2007) Fellsmere 3 7 and 3 8 ). For instance, according to the last census of 2010, the place has not only a significant percentage of inhabitants who identify themselves as Hispano or Latino (81.1%), but also has one of the largest proportions of Mexican descendants in Florida (77.1%), followed by inhabitants of Central America (1.4% ) , Puerto Rico ( 0.7% ), Cuba ( 0.5% ), and from other places of South America and the Caribbean (0.3%) descent (USCB 2014 ). Nevertheless, s ome caution is needed in interpreting these numbers
86 since Census data may undercount undocumented populations and, more critical to this study, not accurately capture the annual fluctuation of workers during the citrus season. Figure 3 7 . Ethnic composition in Fellsmere, Florida (Source : USCB 2014) Figure 3 8 . Origin self identification of Hispanics in Fellsmere, Florida (Source: USCB 2014) Hispanic 81% White Non Hispanic 13% Black 6% Hispanic Non Hispanic Mexico 96% Central America 2% Puerto Rico 1% Cuba 1% Hispanic Composition
87 S ince its founding in 1910 , has been agriculture . The arrival of Henry F to Indian River County in 1893 , and the creation of the Drainage Districts in 1905 open the possibility of rentable farming. P art icularly the state sponsored Land Reclamation Program opened great extensions of land . T he Drainage Districts, as their name indicates , would drain millions of acres of wetlands turning these uninhabitable areas into highly productive real estate (Johnsto n 2000), as we saw in the first chapter (Figure 1 scars on th eir soil; agriculture and tourism industries have created a net of channels extended throughout the state; Indian River country is not different; when blue lines are forming rectangles on Figure 3 9 , agricultural irrigation section are clear. Figure 3 9 . Indian River Country hydrography (Source: Florida Center for Instructional Technology 2008)
88 At the forefront of the industries born in that time was Indian River Citrus , which rapidly became the benchmark by which all other citrus products were judged (Jo hnston, 2000). Indian River County, where Fellsmere is located, is among the main producer of citrus in Florida, and its proximity with the other citrus farming areas make it an important supplier of farm labor ( Figure 3 4 and Table 3 2 ). The other work av ailable in these area coastal communities are lawn services, construction, hotels and home services. The town is situated 10 miles away from the coast, between Melbourne Beach ( to the North) and Vero Beach ( to the South); most of the streets are unpaved an d only the main road, Country Road 512, has street lighting. Interstate 95 is part of the city limit, although is underdeveloped (Figure 3 10 ) . A lthough it is hard to notice, the town is divided by zones. There is an area of prominent families, the origina l migrants who manage the crews that harvest citrus. And then there is the area of the low income apartments, a project funded by the U.S. Department of Agriculture and the U.S. Department of Housing and Urban Development specifically for people who earn a t least $4,000 a year in the agricultural industry . Two to three blocks away from the main road is another area of single family houses, which precedes the trailers zone, the other housing available in town. Although some workers reported that small camps can be seen where some migrant workers sleep each night, I was unable to see any. There is also a small area where the African American and Caribbean community are concentrated in the outskirts of town. In 2010, there were 1,467 housing units reported by t he census (USCB 2014). T here is a day care and an elementary school in town, but m iddle and high school students are bused to the nearby coastal community of Sebastian Beach. The clinic is not far en route to Interstate 95. Fellsmere is an
89 incorporated tow n, so there is a mayor and 5 people city council, most of the representatives are Hispanic. The old school, a brick building from the time the city was founded serves as the home of city government, while the new school is a combination of concrete buildin gs and portables. Figure 3 10 . Road map of Indian River County (Source: Florida Center for Educational Technology 2009) As expected , and not even counting the people refusing to participate in the Census, from a town with these characteristics, the 2010 US Census (2014) shows that a higher percentage of its inhabitants are male (56. 1 %), young (with average of 2 6.4, it ranks as the 25 th youngest town in Florida) , low educational level (with 6 1 . 9 % of people over 25 years of age with less than a high school education) and low income (with an estimated of $27,190.00 per household and $11,188 per capita) . The dimension of the
90 cultural influence on town can be measured by the amount of homes wherein Spanish is the primary language (66.3%) , and for the infrastru cture of the city. The main church is the Virgin of Guadalupe temple, there are at least two soccer fields, a dancing/concert space. There are three Mexican stores and on weekends a mobile taqueria. Also, as with many rural communities the Dollar Store is in the main street. There are two restaurants in town: one pizzeria and one fancier establishment. The workers never go to either of these places, though the pizzeria is popular with younger generations and the managers of the farming operations. Casa y su stento: Food and housing Although my crew leader own ed one trailer and one house that he rented out to his worker s , he did not offer the se space s because he reserved them for his more experienced workers . He did, however, direct me to one popular and acces sible place i n town: the L aundromat. Essentially, I rented a bed in a three bed efficiency in a tenement housed above that business . I shared the space with other two workers : a talkative Guatemalan and a middle aged Mexican from Veracruz. As mentioned, al though most of the town is a Mexican enclave , many Central American workers join the citrus crews and other low pay ing jobs in the area (construction, gardening, golf course maintenance, packing houses, road maintenance ). My roommates reflected this realit y ; the Mexican picker had harvested citrus from childhood and my Guatemalan neighbor jumped from unskilled job to job depending on the conditions of the work. They did not like each other, the Mexican thought that the Guatemalan did not want to work at all , and the Guatemalan thought that the Mexican was working too much for nothing. The three of us cooked, but only the Mexicans cleaned. In the beginning, I was anxious about the amount of cockroaches in our room they were of all sizes and moved freely
91 acros s the walls, floor, and ceiling. At night, I could feel them crawling on my covers and my hair. Among the first products brought with my insignificant citrus worker check I got cleaning products and insect traps. In one week, the cockroach encounters decre ased, and I could sleep without having the little scratchy legs of the insects on my face. Each morning the noise of the dishes woke me up much too early. Outside our curtain less windows there was no sign of the rising sun . Despite this, the lights c ame o n in our room just like the lights of many other houses across the town, full of activity betraying the hour. In my little living space , it wa s Don Lauriano who was up cooking Mexican chicharrÃ³n in green salsa with potatoes; he was coughing from the strong smell of peppers , making him cough and sneeze . It was 5 a.m. and I knew it was time for me to get up and prepare for the long journey. On the other side of the room, Salomon cover ed his head with a blanket -the only piece of bedding he own ed . I grab bed the working garments I had left out the night before -jeans, a long sleeved shirt, long socks, and boots -and started my day with a three block walk to the store where the orders of tacos, gorditas, and enchiladas were made quickly for workers who did not cook either for breakfast, lunch, or dinner. The restaurant opened at 5 am, by 3 pm it was closed . Income, food, and clothing are topics of relevance for workers . T he price paid per box of harvest is discussed amongst the workers every morning and nigh t. Many workers decide who they will and or will not work for based on the price paid per box and on how honest the chivero s (crew leaders) are meaning how regularly and fairly they pay. Because most citrus workers are males with families away, they tend to cook in teams, sharing recipes and locations where to obtain cheap, fresh, and/or ethnic
92 products. In my crew, three workers share d a small traila (mobile home) and they t ook turns cook ing . One day they joke d that anyone coming to eat at their house wo uld know very easily who cooked that day because if someone found a white hair in the food it meant that it was the oldest worker who cooked, while if there was a black hair, it was the youngest one who was in the kitchen that day. If there was no hair in the food, that meant that the third (and middle worker) cooked not because he was cautious, but because he was bald. Despite hardly being a fashion conscious crowd, much conversation was also dedicated to outfits and gear. D iscussions would revolve aro und their working clothes and tools, and where to find free or inexpensive long sleeved shirts and jeans ; the best deals in town were circulated by word of mouth. Covering the body as much possible despite the heat is important because clothing s erves as a protective layer against the elements and pesticides. All ethnic Mexican stores in the region carry these clothes as well as gloves, long socks, hats, boots, bandanas, belts, water containers, coolers, and citrus bags. The church also collected clothes to give it away to workers and families who came to service and collected food on a regular basis to be distributed Association of Florida, also with an office on town, maintained an open porch where clothing, car seats for children, and shoes were regularly given away to whomever stopped by. My own preparation for the field came from my field experience with PCWH, but little did I know how much I had to learn from my co workers to harvest fruit with less pain or having a serious injuries . How to pick a fruit was knowledge that came from one
93 careful with the skin of the fruit; an orange that d oes not have the stem on top of the fruit will be thrown away at the plant by the selectors. You need to hold the fruit, turn and 11 ). It is a simple operation that can be overlooked by the pace in which workers can do it. Some pickers can even pick two or three citrus with each hand at once. Figure 3 11 . Grab and twist, so the citrus does not peel (Photo: JA Tovar 2009) The Sack When I arrived in town I was carrying an almost n ew sack I used when in the process of collecting data in the g roves for the PCWH . At the time, because there were
94 two hours between observations, I spent part of that time picking with the workers, a way to open the door for informal conversations. This time was different, I was going to be picking for days at a time and my first sack was inadequate. In market s there are 12 ), the ayate, and the regular sack I was carrying, which came in three different sizes (45, 75, and 90 lbs.). My sack was the 90lbs model, the so called jumbo, and i t was not just uncomfortable ; it was dangerous. The California sack had one shoulder strap and two clips on the sides to dump the fruit, and, as the name implies, comes from the West coast. Th e most common sack in Florida also has one shoulder strap, but it has only one clip in the front so that it can be attached to either a long or a short harness; mine was a jumbo sack a n up to 90 pound bag (Figure 3 1 3 ) . Figure 3 12 . A full Californian s ack stands on the ground waiting for the crew leaders to stop by and pick all the full boxes of the worker (Photo JA Tovar 2008)
95 Figure 3 1 3 . A jumbo sack sits in one of the steps of the ladder while work er keep filling it up (Picture JA Tovar 2008)
96 Laur iano saw my blackened shoulder one week o n to the job and gave me used 1 4 is a traditional nylon bag used i n Veracruz. The bag looks like a huge shopping bag with two shoulder straps connected to it and does not have a single clip o n either end. Filling and dumping the fruit is an easy task that can be done o n the side more convenient ly each time. If it were no t for this bag and advice, I would never have lasted for a whole season or been able to equal my co pickers in Mexico use huge baskets called colotes, which they carry to the trucks on their heads (Figure 3 1 5 ). The image is so iconic that a monument honoring the workers has been erected in a well known citrus producing town in Mexico (Figure 3 1 6 ). The big company I was assisting in the CHW program also prohibited the use of the ayate, as they insi sted its roughness damaged the fruit. Figure 3 1 4 . Picking fruit from the ground pulling an a yate ( Photo JA Tovar 2011 )
97 Figure 3 1 5 . Twelve tons of tangerines (Source: Ranchocarballomercado 2009) Figure 3 1 6 . A monument in the city of Alvarado, Vera cruz honor citrus workers ( Photo JA Tovar 2012)
98 The Fruit Figure 3 1 7 = oran ges; yellow = grapefruits; red = tangerines; dark green = limes; and light green = specialties (Source: Tucker et al. 1998 ). The variety of citrus in the area is high, there are at least groves of Navel and Valencia oranges, Ruby and Minneola grapefruits, tangelos and tangerines, even limes.
9 9 Figure 3 1 7 shows one of the most comprehensive lists of growing ci trus in Florida I have encountered so far; however, I cannot say with certainty how many of the fruits listed are harvested in the region I was picking. I am sure of the six varieties I listed above, but I also assume we picked some other varieties I was u nable to recognize. The Indian River area, is particularly strong in grapefruit production. M ost of th ese fruits are picked for packing purposes and have many different characteristics distinctive from the fruits dedicated for juice processing. Fruit dedic ated for juice can be picked at any time, in rain or when it is freezing, the appearance of the fruit is unimportant; it could be of any shape as long as the sugar content is correct and the costumers would never know how the fruit looked before it became juice. Fresh fruit to the contrary is marketed by its looks; the main work of the packing houses is selecting the fruit appealing enough to go to the market; the rest will have the same destiny as the rest of the fruit: juice. Fresh fruit cannot be wet whe n picked because the humidity in the box will create spots on the skin. Because most of the fruit we picked was to be sold to the fresh markets, we never started our picking too early in the day or worked in the rain. Sometimes, we were already on our way to work and were sent back because it was raining in the grove. Nevertheless, the fact that citrus for packing/markets were affected by weather conditions did change the schedule of the crews. The morning dew usually disappeared around 9 to 10 in the morn ing, but most of the workers we re on their way to the groves by 5 to 7 a . m . , depending how far the ir grove wa s ; in some cases I witnessed trips of up to four hours one way from where the workers lived to the grove; but on average it is at least one hour tr ip .
100 Harvesters Two of the Mexican stores in town we re the center of job seekers and employers. The 15 passenger vans c a me and went with their human cargo (Figure 3 1 8 ) . Some pass ed almost empty, others refuse d more workers inside. For the most part e verybo dy k new each other, although new faces like mine appear ed from time to time. In the same way workers will share how honest a crew leader is, the crew leaders also classif ied the uno del montÃ³n in terms of harvest rate ). Workers also talked of the faces they recognized or asked about the whereabouts of the ones who they had not seen in the current season. No , al pato lo (The duck was stop ped last year by i m migration Mira ya llego la llorona, Â¿ para quien estara trabajando ? here, I wondering : who he is working for ? ). As the season progressed more and more workers filled the vans, and some small compani es even had more than one van. My crew was part of a three van company, but not all workers used them. Figure 3 1 8 . A c rew leader clean an empty 15 passengers van (Photo: JA Tovar 2007)
101 The first day I went to the grove, I was given some paperwork to com plete. I was many people come and go between crews without ever observ ing another worker fill out papers. Perhaps this was because I was one of the few an eight workers i n my crew each day since many of them were still up north harvesting other products . E ven then we were a large crew compared to others, which was a good sign for all of us , but most of my co workers already knew that since they were worki ng for the same chivero before. Each day, my journey started and ended in the same store where I my crew grabbing ice, coffee, sodas, lunch and lottery tickets in the morning and . The visi t to the store before and after work was always a fascinating event in my day. As the time passed more and more workers and crews were showing up; looking at the calendar in Figure 3 1 5 , October does not have the amount of citrus varieties November to Marc h have. Workers follow that pattern, because they also have to be aware of other crops they follow in New Jersey, North Carolina, or Michigan. This case does not apply to H2A workers which are regularly hired for a nine month period, from October to June, matching the harvesting period of Navel and Valencia oranges, the most profitable ones. There are even cases in which workers are sent back to Mexico because all the Navel oranges have been harvested and the Valencia oranges are not ripe enough to be send to the processing plants. Making the calculations, it is cheaper to pay for the transportation back and forth than to keep paying them their minimal wage per their contract, which is higher than the agricultural minimum wage, usually higher than the federa l minimum wage. For example, the federal minimum wage was $7.25,
102 Florida agricultural wage was $8.56 an hour, and the H2A minimum wage was $10.26 (DOL 2014a, 2014b). Interestingly enough, before companies hire foreign workers they have to advertise the pos ition(s) locally and offer the same working conditions in terms of salary and housing; several companies commented that not only the application of legal workers was only 10% to 20% of their labor force need, but that within one week only one or two worker according to the production manager of one of the companies who works closely with the PCWH. Luck Lottery tickets are relevant for economic and symbolic reasons. They are such a popular part of each harvesting crew that in 2012 the State of Florida launched a game based on the Mexican loteria (Figure 3 1 9 ), and dropped ticket prices from $3.00 to $2.00, which is a more affordable, thus a more popular price for workers, as are the was a reflection of many realities: first it denoted the firm belief in luck. Things happen not b ecause you act on a strategy of buying five different tickets each day, but because you have luck. In a similar sense, some injuries happened because they were unpreventable regardless of how much care you took to avoid them. Discussions in the van started with how much workers had lost or won, and some workers even had a stack of used tickets to show how committed they were to play. The tickets also functioned as presents or as expressions of gratitude; nevertheless if you won the expectation was that you would return the favor with another gifted ticket. That is the other meaning of the game: Nobody was a better harvester at that time, and everybody
103 was equal, because everybody had the same possibility to win. Finally, the game worked as a distractor and g ave the workers something to daydream about. They dream ed about what could be accomplish ed with $500,000 pri z e on it . Figure 3 1 9 using the translated Mexican loterÃa figures : El Gallo, El Paraguas, La Estrella, El Vena do, El Arbol, El Cantaro, El Loro, La Chalupa, La Corona, El Loro, La Mano, La Bota, La Sandia, El Tabor, El Sapo, y Las Jaras (Source: Florida Lottery 2014) .
104 The Grove The true lottery game, however, does not end in a small store or in van trips to the fi elds. When workers arrive to a new grove, the first step to harvesting is to assign the grove lines to each workers, which insures that no favoritism is taken by the chivero or that any worker takes advantage of the others . Because of the multiple illnesse s in the trees, the amount of fruit in each line is unequal; also the distance between the first line and the last line is vast, with the first lines being in a more advantageous position. Walking with your equipment from one side to the other can cost you half an hour and potentially one less baÃ±o (the main containers that hold the fruit after dumping them from the harvesting bags). Small pieces of paper are numbered from one to the number of workers in the crew and then are identically folded and put in t o a hat. Then, each worker picks one and unfolds it. Odd numbers are assigned to the side where the crew arrived and where the ladders are located, making the number one the more desired number since is also the first line in the block. Even numbers are as signed to the opposite side from where the ladders are placed and the last even numbers is always the least desired. Once the lines are assigned workers start picking on each side of that line. Between each line there is usually a channel or an uneven lin e where the trucks do not go. Once the workers se topan (encounter) each other and finish all the fruit in their line, they move to the following free line on the opposite side (Figure 3 20 ). The number of lines they will cover depends on the distance of t he lines, the amount of trees and fruit in each of them, how fast they pick, and how many trailers they are assigned each day. Generally the same grove will be cleaned out to the end, but in some cases, two different groves with two different kinds of citr us will be harvested on the same day,
105 because the market had requested an order of each kind. Workers did not like to move groves a lot because of the time it took to move the equipment (ladders, boxes, and chiva) from one grove to the next. Sometimes, the equipment will stay in the same groves, many times with the ladders standing in the same trees. Although, most crew leaders will request the workers bring back the ladders to the transportation trailer in order to prevent lightning on the trees or to be r eady to a change of groves the next day. Figure 3 20 . Workers arriving to a meeting place, one on the bottom one on the ladder, next to the baÃ±o a nd a yoga ( green and white water container ) in the far right (Source: Chang W. Lee/ New York Times 2009) Drin ks With your ladder on one shoulder, and your bag, water container, and/or a cooler with your food and drinks in the other, your feet are making surcos in the sand while you are advancing in the fields (Figure 3 21 ). Each worker will carry extra drinks ins ide their water containers (called yoga) and/or the coolers: some sodas (generally Coca Cola),
106 an energy drink (Red Bull or, in most of the cases some other cheaper, extra caffeinated drink), a sports drink (especially Gatorade), and perhaps even a beer (p referably a Bud Light). After a week in the fields, I also realized that solely water was not enough; you really need the extra calories to last, so a sport drink and a soda were in my yoga all the time. Some workers commented, that few crew leaders the o nes with the worst reputations regularly carried beers as a way to bribe the workers to stay with them. Figure 3 21 . Worker changing place s ( Photo : JA Tovar 2011)
107 BaÃ±os (boxes) The baÃ±os or boxes (the industry and the state measures the crop harvest ed crop by box count) are strategically delivered by a chiva (a special modified diesel truck that uses a mechanical hand to pick up the boxes filled by the workers) by the chivero. The crew leader estimates how much fruit is in each line and knows how muc h each picker is able to harvest and drop the boxes accordingly. The main difference between juice citrus and fresh fruit citrus is indicated by the kind of container. The regular baÃ±o is a black round bucket the size of a small Jacuzzi (Figure 3 22 ) , once that bucket is full the fruit is dumped in to a gondola in the back of the chiva , and the same bucket can be used again (Figure 3 2 3 ). Good pickers usually carry between four and five buckets at the time; pickers like me would be happy with two. Fresh fru it on the contrary is dumped into white square toll boxes which are carried on flat bed chivas to semi tractors to be moved to a packing house (Figure 3 2 4 and 3 2 5 ). C hivas holding juicing citrus also dump the fruit into semi tractors to be transported to juice plants (Figure 3 2 6 to 3 2 8 ). The size of packing boxes varies depending on the kind of fruit and/or the packing house requirements; there are boxes of 6 (specific for specialties and tangerine), 9, 10 (the main size), and 11 jumbo bags. Figure 3 22 . A empty baÃ±o (Photo JA Tovar 2007)
108 Figure 3 2 3 . A modify diesel truck, call the chiva uses a mechanic arm to delivery empty baÃ±os and collect them after they are full ( Photo JA Tovar 2010)
109 Figure 3 2 4 . Harvester next to packi ng citrus boxes (Sourc e: Therese Kapaun 2013) Figure 3 2 5 . Flatbed trailer t ie d down with 10 bags worth citrus boxes (Source: Macro P lastics Inc. 2014)
110 Figure 3 2 6 . Chiva dumping c itrus from a hyd raulic hi lift metal basket ( Photo : J A Tovar 2011) Figure 3 2 7 . Full semi t ruck trailer in the field (Photo JA Tovar 2009)
111 Figure 3 2 8 . Full trailers park outside a juic e plant (Photo: JA Tovar 2008) Guest Workers (H2A) This system has made Florida into the most productive citrus industry in the world; Brazil and Mexico have bo th dedicated more land to the production of citrus, but neither one of them has achieved the production level of crop and harvest as Florida has. Even with the growing impact of citrus greening in the fields, the harvesting system used in Florida is highly efficient, and many of the involved Mexican guest workers and managers of this increasing labor force production we have in here [Florida], over there [Mexico] they fill one small truck (Figure 3 1 5 ) on half a day and here they may fill four, five and even six semi trucks in a 10 hour day; because we know this, we usually start with one or two weeks of half day Large compani es not only follow strategies of acclimation, but also monitor their H2A contracts that have strict levels of production that workers must fulfill in order to
112 stay on the job. The system is a very controversial one; many advocacy groups are very critical o f the system. However, the H2A workers we interacted with were, for the most part were satisfied with the contract. The main differences between what companies call local workers and the guest workers is that the latter come to Florida only for nine months , the company has to pay for their visa processing, transportation, and housing; they also insured them at least a minimum wage for the entire season. H2A workers do not get any migration rights, cannot switch companies, and are exempt from paying any kind of taxes or receiving any governmental assistance. Many workers with high productivity were encouraged by companies to return to Mexico and then were brought back as H2A workers; although there were cases in which the US government rejected , most of them were able to come back under the new system. What was clear in the H2A crews I was working with is that most of the workers were part of the same community, since most of this recruitment happened among friends and family members. Of course, the main problem in Mexico is that the recruitment places are becoming the new coyotes (smugglers) in the country, asking for high payment in order to secure a place as a H2A workers. Another official of a company who also hired H2A workers once revealed at the annual Citrus Expo in Fort Myers that the best way to ensure the fairness and honesty of the process is for the companies to go directly to Mexico to hire the workers, since the local agencies and authorities are not trustworthy. The H2A workers I interacted with were, for the most part, recruited directly by the companies they currently worked with. The effect on local (Florida based) workers is that fewer jobs are available for them, although some received free housing when hired for the shortage of H2A workers. The
113 other problem is that only large companies can afford to implement such programs, thus further contributing to the trend of large agroindustry dominating small producers. For these large, consolidated companies, the system has worked in their favor and they not only harvest their own groves, but also have extended their services to harvest the groves of smaller growers. Women in the grove s In my small company back at the grove, the equality achieved with the lottery ends when the grove sections are raffled off after the almuerzo (the second heavy breakfast eaten in the groves). If you do not have your ladder already marked, you had better pick one quickly and then get to your assigned line to begin filling your boxes. Each day, each wor ker competed against his co workers and himself. My first day on the job was humiliating. I hardly finished three boxes, while the rest of my coworkers completed over ten a couple even made over twenty boxes. Since the boxes were counted every day , my co workers continuously offered encouragement, telling me they Ã©chale ganas , each one of them would come to share his wisdom with me, teaching me how to be more efficient and keep safe in the gr ove. A few workers in these small companies pick in teams. When that happens, the teams are usually composed of couples and/or family members. Naturally, these are were usually the people who fill more boxes. H2A crews such as those found in larger compani es do not have this practice of shared picking. Importantly, I never noticed a single female worker in any of the large companies employing guest workers that I visited. On the other hand, this was a regular occurrence in many of the smaller crews, such as the one I worked with, where there were always some women coming to harvest with their husbands, at least on weekends. In one such small
114 company that I worked at briefly, there was even a female crew leader 1 who also picked alongside her significant other . The Ladder Regardless of whether workers pick in teams or solo, appropriate equipment choice and usage is critical. Besides the appropriate bag, knowing how to work the ladder is the most important tool to increase picking rate, and thus make more money . Workers mark their ladders with paliacates ( bandanas ) and plastic bags after selecting the most stable, light, and height appropriate for the size of the trees. As I saw other workers do, I stepped on my ladder o n the floor each morning to check the stab ility of the rungs. But selecting the ladder is not as important for them as knowing how to move and position it. The easiest way to move the ladder once in your picking spot is to keep it straight; keeping the balance of the ladder, especially on windy da ys, can be done by force or experience. Lacking the latter, I relied on force to balance it and ended up with sore muscles for something that is typically considered an easy task by seasoned workers. Making a correct escalerazo (positioning of the ladder) translate s into picking the most fruit with minimum movement of the ladder. The oldest picker in my crew came to talk with me as soon as I did my first escalerazo and explained: if you leave your ladder on the branches , you are not going to reach the fruit in the middle and you risk a broken branch or a slip of the ladder on the side. El Chapatin then rearranged it, took it on the side and slid it in the middle of the tree, shaking it to check stability before giving 1 prominent crew leader both she and her father have a terrible reputation for not paying workers fairly.
115 me the go ahead. Now it was firm and cen tered , so I could reach all around the tree and stay secure. T he trick is to clear the tree of fruit with the least amount of escalerazos . There are mainly two kinds of ladders in the fields: the aluminum ladder (Figure 3 2 9 ) and the less common, but stil l used, wood ladder. This last one also adds the risk of splinters and producers blame the wood ladders for carrying tree illnesses. Workers, including myself, however, were happier with how the ladders fit better inside the tree branches, were easier to t ake off the trees, and carried less sand. In Veracruz, Mexico, the opposite is true, most ladders used are made of wood (Figure 3 30 ). Figure 3 2 9 . Som e 18 inch aluminum ladder s rest with jumbo sacks on trees at the end of the day ( Photo : JA Tovar 2009)
116 Figure 3 30 . Wooden ladders rest in a n orange tree next to a banana plantation close to Alamo, Veracruz, Mexico (Source: R anchocarballomercado 2009) For the most part, fruit is collected from top to bottom, so the sack is empty at the start (Figure 3 31 ) . Arriving to the top of the tree, you will see an ocean of green canopies interrupted by spines of ladders. The ladder rest on the branches (Figure 3 3 2) supporting the workers moving up and down the steps as fast as they can; but before they climb the la dder, they have to be sure it is secure on the ground (Figure 3 3 3 and 3 3 4 ). Carrying the aluminum ladders from tree to tree, around the fields, and securing them on the ground and around the branches makes such loud clanking sounds that they can be heard from long distances; on more than one occasion this sound guided me to find a crew in dense groves. The steps on the ladder are also a tool, but can be a risk factor as well; with the use of jumbo bags, the ladders are very
117 useful to keep the sacks restin g while the harvesting take place (Figure 3 1 3 ); the risk lies in the accumulation of sand in each step (Figure 3 3 5 boots or shoes (Figure 3 3 6 adjusted. Figure 3 3 1 . Worker look s at the horizon carrying an empty ayate , a good way to see where the ot her workers or the crew leader are ( Photo : J A Tovar 2010 )
118 Figure 3 32 . A ladder rest s on branches that can get poke through the rungs at any given time (Photo : JA Tovar 2010) Figure 3 3 3 . A very instable ladder is a big risk, there is a worker in top of this ladder (P hoto: JA Tovar 2008)
119 Figure 3 3 4 . Worker secure ladder in the soil, before going up ( Photo : JA Tovar 2011 ) Figure 3 3 5 . Sand accumulation on ladder a nd gloves (Photo : JA Tovar 2008)
120 Figure 3 3 6 . Shoes carrying the sand to the ladders (Photo JA Tovar 2008 ) find in town. One even replied, when asked why he did not harvest othe r crops, that the citrus was heavy work, but that the trees provide the shade and the rest other crops do eat or rest and the truck can hold your back; there is noth Nevertheless, the same peaceful trees are also the source of insecurity on the job: a broken branch (Figure 3 3 7 ), a loose branch between the ladder (Figure 3 3 2), or a poorly placed ladder on the ground (Figure 3 3 3 ) could m ean dire consequences. I fell from the ladder twice on both occasions the ladder was poorly positioned. Luckily enough, one fall was from the third step and I fell on top of a semi empty ayate; and the other time I fell on my jumbo sack and I was held by t he main branch of the tree that just left bruises on my side and on my spirit.
121 Figure 3 3 7 . A broken branch , a risk hazard (Photo JA Tovar 2009) Windy days add to the instability risk of ladders; to avoid movement of the ladder on the ground, many work ers jump on the first step so the base of the ladder legs gets shoved deep into the sand, while others make a back and forth movement so the ladder gets stuck into the branches. This method is effective, but each tree has its own anatomy, with older groves tending to have larger, more difficult to manage branches while young groves have branches that are too flexible. To improve the positioning of the ladder, workers start with the ladder standing up on its side and then find an opening in the branches and push the ladder hard to enter into the tree. Once inside, they turn the ladder around and jump onto the first step. All the operations have to also be prioritized by where a picker can collect the most fruit; a good picker moves the ladder the least. Ladde rs come in sizes increasing by 2 inches going from 10 to 22 inches.
122 In Florida most of the ladders are from 18 to 22 inches. As mentioned before, l adder usage is also relevant to the age of the grove. Young groves will have shorter trees, so the ladder is only used once the low hanging fruit ha ve been picked ( de pechito ), while older trees should be picked from top ( puntear or copetear , Figure 3 30 ) to bottom ( barbear Figure 3 3 8 ) . De Pechito is an interesting expression because it makes reference to a socc er play in which the player receives the ball with the chest either to score or have control of the game. Workers pick a good sized fruit yield so quickly in these small trees that they keep their jumbo bags open in front of them and just let the citrus hi t their chest and fall into the bag. The expression De Pechito is also used as a term referring to a task that is easy to complete. The last grove I picked with my crew was one such grove and I worked so hard to achieve picking as many boxes as my co worke rs (15 boxes in 9 hours), having their recognition when the notebook, where the daily picking was accounted for, was passed around the truck was my Figure 3 3 8 . A w orker barbeando with short sleeves shirt, but we aring long sucks in both harms (Photo : Jason Lind 2009)
123 Changer or Ganchear : Other methods of picking An older picker explained to me that larger fruits are closer to the trunk, while smaller fruits will hang in the exterior. Larger citrus are usually in the center of the tree and, in many cases, out of reach from the ladder, depending on the type of citrus (packing versus juicing), crew leaders often make the decision to either clean the trunk ( palo limpio) o r leave the fruit behind; sometimes it is the worker who insists on picking the trunk hugging fruit, for which they may need to c hangear . The literary translation of changear activity of p icking fruit on the tru n k is risky and is usually left to the end , unless there is a chal Ã¡ n d by hanging off the trees , just like a monkey . Contrary to other crews I had observed before, the gancho (a hook use d to cut the unreachable fruit or, when used behind a mechanized harvester, is used to finish cleaning the trees, Figure 3 3 9 ) was prohibited in the large company crews since it can damage the fruit , rendering the fruit usele ss for the packing house . In my small crew, only one worker made his own hook, using a large nail and a piece of metal pipe. Many times this gancho was shared, but all the equipment that personal bags, water containers, and marked ladders (unless that worker did not show up to work, which meant the ladder was up for grabs) . The gancho is not well recieved by the industry, especially for use on the packing fruit since it easily damages the stem or peel. Regarding the juicing citrus, though , the gancho was even encouraged to be used by a crew that was trailing the mechanic harvesting; it was provided by the company, contrary to the standard workers who were allowed to pick fruit from the grou nd as
124 exhibited in Figure 3 11 . The gancho 40 ). Figure 3 3 9 . A w orker hold s a hook ( gancho ) to pick citrus out of his reach (Photo : Jason Lind 2008)
125 Fi gure 3 40 . Reaching out for the la st fruit (Photo: Lynne Sladky/AP 2007 ) The Last Fruit: Palo Limpio the patron (the owner of the grove and company) no r the other workers want to deal morning when we arrived to a new grove. The day before, one of my first days on the field, I left some trees with fruit still nestled in the tops of them, as I thought I would go back to collect them. Palo limpio , means no fruit should be left on any tree. The last fruits are the more difficult ones because they are generally more difficult to reach depending on where the ladder can be positioned or where branches may hit workers. Some workers even use other oranges on the ground to throw at unreachable oranges. Citrus picking is a very specialized job: It takes years to learn and you put your body at risk on the job each day.
126 Clippers and Ergonomics The invisible hand of the market is present when picking fruit . Sometimes the crew goes for a specific size fruit, what they call espatear (selecting only the larger fruit). Workers could recognize the groves they have been to by the kind of fruit and trees they have. Each morning the price question comes first: What are we picking (grapefruit, oranges, or tangerines)? The price is directly related to the fruit: grapefruit pay the least, but filling boxes with them is easier than with tangeri nes, which pay the most, but also could be forcefully cut with clippers and take more than double the amount of time it takes to fill a box. The ergonomic impact of this work adds to the heavy load, constant up and down scaling of ladders, and bending, all compounded by the constant repetitive motion of using the clippers, which are occasionally already in bad shape. In large companies, clippers are provided, while in smaller crews, each worker is responsible for providing his own; the usage of clippers als o decreases the usage of gloves, since gloves do not facilitate the cutting as much (Figure 3 41 ). Figure 3 41 . Citrus w orker with clippers g etting ready for harvesting (Photo : Steve Miller 2010 )
127 Outfit and Gear Gloves (Figure 3 11 , 12 , 3 5 , 3 8 , 3 9 , 40 ) , long pa nts (Figure 3 1 3 , 1 5 , 20 , 2 1, 2 2 , 31 , 3 4 , 3 9 , 5 5 ) , long sleeves sh irts (Figure 3 11 , 1 4 , 1 5 , 20 , 31 , 3 9 , 40 , 5 5 ), and closed toe d shoes (Figure 3 1 4 , 20 , 2 1, 3 4 , 3 6 , 3 9 , 5 5 ) are required in large companies sometimes even provided but are also regula rly used on smaller, less regulated crews, like mine. The usage of hats (Figures 3 1 3 , 1 5 , 20 , 2 1, 2 4 , 2 6 , 31 , 40 , and 5 6 ) , large socks (Figure 3 2 1 and 3 8 ), masking taped sleeved shirts (F igure 3 3 9 ), and paliacates (light cotto n Mexican bandanas, Figures 3 2 4 ) is not regulated , but are still very common on small and large crews. Canvas sleeves (Figure 3 2 1) and safety glasses (Figure 3 2 9 and 3 6 ) were less common, but they can also be seen in use, especially among large company crews where they are not en forced, but sometimes provided. Once, while grabbing breakfast at the store, another member of my crew saw my arms full of scratches and asked if I wore long sleeves. I told him I did and he then I use socks. A fter buying some socks later t hat morning he helped me cut holes in them to slide my fingers through. Then onward , I only picked with these in your arms, enough pain we have carrying this entir e weight he said when he saw me with my new tool. Long socks, like gloves, have a limited lifespan, but workers always have more than one pair so they can wash them regularly or have extras in case they tear. The other option is to buy gloves from the st ore, where they are always available. One day in the fields, I realized that I left my sack at home and I did not have anything to pick with; I asked the chivero if he had an extra sack. He did not have any extras, but said that at least one other worker a lways has more than one. Another
128 worker leant me his extra and explained that he always carried two, because if one breaks, there is not store around to get another. For that day he let me use his older sack, while he unpacked the new one he was carrying i n his own car (mueble is what they call their own transportation). Transportation In the height of the season, the company van was always full and many workers relied on their own mueble. I drove to the groves in either my own vehicle or in one of the oth we had to pay for our own gas and the roads were sandy and difficult to maneuver (on more than one occasion the cars had to be pushed out of the sand). The main advantage was th at at the end of the day, we did not have to wait for the crew leader to finish picking through the rest of the boxes of fruit before we could leave. Also, because the van was full, there was usually no space to put the yogas in the back, as it can be seen in Figure 3 1 4 . The operation is quite different when large buses from larger operations are used for transportation, and this way, workers can travel with their equ ipment next to them (Figure 3 42 ). Figure 3 42 . B uses of a large harvesting company rest on Sunday , park next to the workers camp (Photo : J A Tovar 2008 )
129 Multiple Risks Many other risks are constantly present in the field. I was stung by several wasps and bitten by many more ants (Figure 3 4 3 ). A growing complaint is the presence of baraÃ±ero g roves (groves where many trees have dry trunks with a lot of debris and little fruit). When groves were in such decrepit conditions (Figure 3 4 4 ), workers usually asked for a pay increase, which oftentimes resulted in an increase of 10 to 25 cents per bag. On the other hand, healthy old groves were also dangerous because it is difficult to get in the center of the tree and protect your face from the branches (Figure 3 40 ). Finally, there is the issue of chemical contact. Though these citrus grove chemicals are not as toxic as the ones used on tomatoes or strawberries, they still cause skin and eye irritation when the morning dew (Figure 3 4 5 ) gets on the workers, in addition to other debris like small flowers (Figure 3 4 6 ) and dry moss (Figure 3 4 7 ). Figur e 3 4 3 . Large anthill s are all around the groves (Photo : JA Tovar 2013)
130 Figure 3 4 4 . A dry tree in the field is a signal of a decay ing grove, either by age or illness (Photo : JA Tovar 2009) Figure 3 4 5 . Chemical dew on fruit and leaves (Photo JA Tovar 2007)
131 Figure 3 4 6 . A b looming tree will deliver additional debris and resin (Photo : JA Tovar 2010 ) Figure 3 4 7 . Tree with dry moss , which deliver a green dust (Photo : JA Tovar 2011)
132 Health Care Search and Access The Clinic Concerned with the access to health care, I visited the local clinic on one of the days we did not go to the field, because the rain prevented us from going . The clinic is two miles west of the town. It is accessible by public transportation that has hourly trips between Fellsmere and the Sebastian Beach Walmart store. The clinic is part of the migrant clinic that receives federal funding. While there is no requirement of legal status to be seen, an appointment is required to reserve services. The cost is income based, so proof of inco me and address is required. I know of the clinic, because it is on the way into town from Interstate 95. I was also notified of its schedule and services by one of workers also furth er assisted me by informing me of the income/address documents needed. I did call the clinic beforehand, but I did not make an appointment because I did not know when I was free to visit. On the phone, I could tell the woman from the clinic was fluent in Spanish, even though she answered in English. At the clinic, there were some cars parked in its large parking lot. The clinic was located in a large and narrow building shared with other unrelated business. In a white and air conditioned waiting room, some female patients waited to be seen. There were posters about immunization, birth control, and healthy eating on the walls; the sliding office window opened and the receptionist asked me, in Spanish, if I had an appointment. I did not, explained why, and as ked if I could be seen without one. She apologized, offered to book one for me and gave me the card of the clinic to make an appointment by phone if I was unsure when I day, o
133 The Hospital about the clinic, but mainly only the women have gone, with the exception of a documented older male pic ker, who visited the clinic annually. As with my case, workers are limited by the schedule of clinics. Some studies have shown that males are more resistant to seeking out doctors on a regular basis, and male pickers are no different ( Springer and Mouzon 2 011 ). When there is an urgent need to visit a doctor, either the worker ended up in an emergency room or, more generally, the crew leader would bring the worker(s) to the doctor designated by the company. Fellsmere does not have a hospital and the closest ER is in Vero Beach (20 Miles away) 2 or Malabar (24 miles away) 3 . When harvesting, the nearest hospital may be Vero Beach or Fort Pierce (33 miles away from Fellsmere) 4 . When one of our co workers fell from the ladder and broke his leg, he was brought by t he crew leader to Vero Beach which was less than 20 miles away from where we were usually picking; however, this was not the case for many other crews or workers, some of which were even sent back to Mexico to receive treatment. This was the case of a work er in the community who lost an eye. 5 2 Indian River Emergency Medical Services, 1900 27th St, Vero Beach, FL 32960 3 Holmes Regional Medical Center, 1350 S Hickory St, Melbourne, FL 32901 4 Lawnwood Regional Medical Center, 1860 N. Lawnwood Cir. Fort Pierce, FL 34940 5 The legal ramifications of injuries i s also different according to workers legal status, knowledge of their rights, and their injuries. For instance, the first worker received workers compensation, because even though he was undocumented he knew about its rights, the second worker even though he could have been assisted to demand his rights was quickly removed from the country by his crew leader; perhaps, the pain and lost of worker contributed to his decision. In some rural areas of Florida, like Homestead, law firms have marketing campaign t o encourage workers to demand their rights and corporations complain about it, but with cases like the one described before workers have mentioned some improvement, nevertheless most of the workers avoid confrontation with their employees for fear of depor tation or exclusion of jobs, if branded as trouble makers.
134 The Health Fair The other likely opportunity for workers to see a health professional is at health fairs. Health fairs are commonly used by workers because they are usually held on weekends, when people are less likel y to be working in the groves; also, there is a sense of anonymity and services although limited are free. On a warm winter Sunday of 2008, the PCWH was participating in a health fair in Clewiston (Figure 3 4 8 ); the workers of the company I was working wi manager offered free transportation to all workers who wanted to attend. Over 50 workers showed up at the fair. At the fair, some health screenings were offered: blood pressure, glucose level, heart examination, body m ass index, self examination of breast and testicular cancer after educational training, testing for HIV and other STIs, and eye and oral screenings. Each of the workers had most of the screenings. Figure 3 4 8 . Members of the PCWH participating on a heal th fair at the regional hospital in Clewiston (Photo: JA Tovar 2008)
135 Health fairs should not be underestimated. At this health fair above, one of the workers was found to have a heart murmur by a medical student from Florida Atlantic University, and was se nt inside the hospital that was hosting the health fair. That same day the worker was airlifted to the JFK Medical Center in Palm Beach for emergency surgery that lasted over 10 hours. The worker had a genetic condition which affected the arterial valves i n the heart and he had already lost his father and an older brother from heart failure. Doctors stated he would not have survived the week had they not cousins and ano ther friend from the same town. It was a long night; I helped in translating the dire outcomes of the procedure to the two men with me: the worker remained in an induced coma for a week because it was very difficult to attach the artificial valve. Both of my companions had to donate blood, but they needed even more. The patient stayed in the hospital for several weeks until he was released with a very strict regimen 6 . This was the most extreme case of intervention I have ever seen come of a health fair, bu t many other workers also received some kind of assistance. The benefits of the health fair were so great that the county health department was able to organize health fairs in several camps (Figure 3 4 9 ) until the financial crisis of 2009 cut the funds t hat were keeping the program running. For the time being, workers still visit the 6 These two men and the company managers visited him each weekend until he came back to the camp. The company arranged for the recovering worker to be with his cousin in the camp so he could be loo ked after until the end of the season, when he went back to Mexico. He never returned and contrary to the only male at home and supported his mother and t wo younger sisters. Some years later his cousin told me, he had died doing what he liked best, playing soccer.
136 Dade City, four hours away. Two of the citrus workers who requested assistance to their C HW were able to have their cataracts removed at the Benita Spring Lions Club clinic at the end of the citrus season. Though these services continue on an annual basis, the PCWH has ended due to lack of funding. Nevertheless, three of the largest companies in Florida maintain at different levels eye safety programs among hundreds of their workers. Figure 3 4 9 . The Hendry County Health Department holds a Sunday health fair on the workers camp (Photo JA Tovar 2008) Remedies In the field, accessing health ca re is a multi step process ; the first person to ask when an injury or illness affect s the workers is a nearby friend, particularly if he/she is older and more experienced . Stores carry a great variety of supplies, from natural remedies to over the counter Mexican drugs. Cellophane bags of medicinal plants hang next to multi sided packs of dry peppers, sp i ces, and candies; some even have printed
137 the specific illness the plant will treat, but usually whomever was buying it would already know from home or from other workers how to prepare the remedy and more information on its uses . Stores also dispense over the counter products such as painkillers, muscle creams, eye drops, and more health products made in Mexico; it was these processed medicines, and not the traditional treatments, that were more likely to be used . colirio drops to clear red eyes, broucolin , and vickvaporrou for a cold, and m anteca de cacao for chapped lips. If the in jury or illness is more severe, workers will look for an inexpensive and accessible place of treatment . For the most part , male workers, embedded with a macho culture and lacking U . S . immigration documentation and health insurance , refuse medical treatment until there is no other option. On the other hand, f emale workers, concerned with their reproductive health , tend to see physician s more regularly and also assist their male coworkers and companions in search of professional help. Aware of the risks in t he grove, workers actively mitigate the hardship of the journey with the low level tools available to them, such as socks, paliacates , and tape. When these risks cannot be avoided, mitigation of the pain incurred can include regular consumption of alcohol, painkillers, and topical over the counter medications. As such, these workers are not simply passive recipients of the oppressive and risky conditions they encounter, but agents in constructing a new culture that bridges their heritage and their reality. Citrus workers in Florida share cultural features linked to their ethnic background and work experience. Although in the field there is some competition to see who will pick the most amounts of fruit, community emerges when workers generously
138 share advice and best practices with co workers. Citrus pickers are a close community and sharing information on prices, employers, pick ing skills, and resources are a daily practice. Fights among workers are rare, and when they do happen, many times they are related t o alcohol consumption. Longtime residents mentioned racial tension when the Hispanic population grew in the sixties and seventies, but currently most of the Anglo population who have some economic interest in Fellsmere moved to the co a stal cities of Sebast ian and Vero Beach. 7 The PCWH and the CHWs role from word of mouth and self treatment, or visiting a more professional care clinic, I have not seen a more comprehensive p rogram than the PCWH project. With their use of CBPR and inclusion of current farmworkers, the workers themselves have become the owners of the actions that will improve their work conditions (Figure 3 50 ). The use of CHW extends this horizontal model and creates the capacity for workers to organize 51 ). Institutional members of the CAB also have the opportunity to reach out to workers and fo rm new partnerships (Figure 3 4 9 ) or even new organ izations, such as the I mmokalee Lions Club (Figure 3 52 and 3 5 3 ). 7 Considerations to other groups in Fellsmere are important, such is the case of Haitian families coming from Fort Pierce, the previous capital of the region, before the creation of Indian River County and where there is a large and historic Caribbean population presence. They arrived when the region was once the located the princ ipal production to Hawaii (Knight, et al. 1984).
139 Figure 3 50 . Paul Monaghan, PCWH director in 2003 , help s the Community Advisory Board select logo for the program (Photo Luis Arreola 2003 ) Figure 3 51 . Ricardo Contreras, member of the PCWH at USF train s promotores de salud outside a workers trailer (Photo : Jason Lind 2008)
140 Figure 3 52 . Constitution of the Lions Club in Immokalee (Photo Luis Arreola 2005). Figure 3 5 3 . The Immokalee Lions Club after holding their first eye health f air in 2003. William TreviÃ±o, the current president of the club, stands second from ri ght to left in the front row (Photo : Luis Arreola 2003 ) As trainers of CHW, we were call ed their co workers, the CHW members b 51 ). For the workers who came to appreciate the use of PPE, their appreciation stemmed from education by their co workers, so it did not seem like just another imposition to their
141 work (Figure 3 5 5 and 3 5 6 ). One of the main problems in agricultur al labor is that the investment in production is concentrated in the development of more technology geared to increase production with less labor; meanwhile, the workers are left with cheap solutions for their safety. For instan ce, fogginess on safety glasses is the principal problem with their use. I stopped using them and reverted back to using my regular glasses, which did not provide as much protection, but still helped a little. The technology of fog free glasses exists, but it is very expensive and mainly reserved for hunting and military use. After many, many tests with PPE (Figure 3 5 6 ), few models proved to perform significantly better to warrant companies consideration of investment. Other problems with the PCWH program include limitations on tax money funding, locations, and kind of company. Figure 3 5 4 . Paul Monaghan, PCWH Director, Francisco Tovar, a H2A citrus worker , and the author delivering a promotor de salud certificate of completion (Photo : Jason Lind 2007)
142 Figure 3 5 5 . The auth or organizing a soccer game of G lasses User vs N on glasses users (Photo Paul Monaghan 2007) Figure 3 5 6 . Citrus harvesters FWAF Coordinator, FPRC Director, CDC representatives, and Citrus Industry managers met at the RCMA in Immokale e to plan the PCWH strategies in their last CAB meeting (Photo JA Tovar 2010)
143 Fi gure 3 5 7 . The author conducting a glasses test in the field (Photo Jason Lind 2008) New Challenge Finally, any support program for workers also currently faces the same chal lenges the whole industry is facing regarding the advance of the huanglongbing infection : fewer resources and higher cost s (Figure 3 5 8 ). For citrus harvesters this is also bad news since it mean s less emplo yment due to less farming area, less income with the decrease i n production, and also an increase of risks linked to the increase of groves (Figure 3 4 4 ) and the use of more toxic chemical cocktails to treat the trees (Figure 3 4 5 ). In the two y ears following the spring of 2006, when the McKinnon Corporation groves were
144 diagnosed with the disease, Maury Boyd , president of the Winter Garden citrus company, devised a three pronged strategy to combat HLB that has become regular among other growers . To eliminate the Asian citrus psyllid that spread the infection he sprayed his trees aggressively with chemical insecticides. Pre HLB, Boyd, like many of his colleagues, had only sprayed his trees with petroleum distillates that smothered mites and other r elatively innocuous insects. To support the trees' natural immune system, potassium salicylate is applied and, to avoid the deficit of nutrient s , micronutrients like manganese and boron, are directly sprayed onto the leaves (Satran 2013) . Although the dise ase is not eliminated and this increases the yield production cost dramatically the chemical treatment has worked on the trees, but the human costs may have been increased. A company collaborating with PCWH reported that most eye related injuries of 2012 1 3 were chemical related. The combination of all these chemicals has not be evaluated for the short term and long term effects to worker health, and the newness of its usage means there are no rules in terms of a waiting period for harvesting after spraying to safeguard workers. Over the course of PCWH, for instance, workers have often complained of skin irritation. Advocacy groups have blamed chemicals but companies cite allergens such as polyester. The PCWH staff worked on promoting the use of cotton work clothes, but also on following the rules of the WPS for workers to receive proper training and to ensure that the application of chemicals follows federal rules, but that was before the increase of chemical use. Now, that the EPA 8 is trying to modify the 2 0 years old law, action is needed to review the . 8 EPA proposal has positive improvements, but also negative changes like eliminating central postings.
145 Figure 3 5 8 . The i mpact of greening on : Huffington Post 2006)
146 At the end of the season I had been harvesting in, many workers menti oned that it was their last season because it was not worth it to stay in the United States unable to send money to their families. Many, I discovered the following season, moved to their native home voluntarily. Others were forcedly returned in the wave o f deportations led by the Obama Administration, and a few joined the service and construction industries that are slowly coming back to Florida. With the increase of H2A workers also present in er crew leaders, who had made small fortunes, preferred to retire to keep working for less. Still, Fellsmere is, and may be for a while, an agricultural Mexican town. The Native Anthropologist m that functions not only in relation to cultural and ethnic background. Socio economics is a barrier that is rarely accounted for by researchers studying their own nationals and it is part of the Latino diversity. My education, knowledge of English, and m igration status create d a bubble that could only be permeated when my harvesting skill partially match ed that of my co workers. Is it impossible to maintain undisturbed fieldwork when researchers become source s of information? The trade off of my role as a researcher was to offer legal, technological, and occidental medicin al advice as long as workers saw me in the field as another picker competing to collect the most amount of fruit without suffering hechandole (making the effort) . Nevertheless, many workers have remained my friends who I have accompanied to court, visited in the hospital, visited their families back home, stayed in their houses, celebrated milestones from birthday to baptism, and travele d with them across the country (Figure 3 5 9 ).
147 Figure 3 5 9 . The author spending a collective birthday party with his co workers and one visitor in Fellsmere (Photo courtesy of Rosana Resende 2012) Mixed Methods Data Analysis Process As in previous researc h of the PCWH project; most data was collected in Spanish, with the exception of a few conversations in English with specialized and key informants during the participant observation process or as a results of my work at the PCWH. Instruments were entered, coded, and analyzed using IBM SPSS Statistics 21 (IBM 2012) or UCINET 6 (Borgatti et al. 2013) software depending on the kind of data. Qualitative data represented by the field notes was entered, coded, and reviewed with the assistance of Atlas.it 7 softw are (Scientific Software Development 2012). Additional software assistance included Excel (Microsoft 2012) and EndNotes (Thomson Reuters 2014). Four sets of analyse s were applied to the data: 1) descriptive and inferential bivariate statistics, 2) script analysis, 3) social network analysis and, 4) multivariate characteristics, and work related experiences with injuries. Script analysis was used to analyze the wor kers safety routines. Social network analysis was used to understand the process by which workers may influence each another in relation to safety practices. Lastly, multivariate statistics were used to test potential models explaining safety
148 strategies. T ogether, the sets of analys e s try to explain the safety strategies of workers , logical model follows: 1. Descriptive and Inferential Bivariate Statistics: Frequency Distri butions and Two Tailed t Test for Difference of Means summarized the workers demographics and basic discrepancies of workers in relation to the use of PPE, the effects of CHW, and their safety strategies. 2. A script analysis was made of various safety pract ices depicted through the harvesting season. Schank and Abelson (1977) define individuals' routine patterns of activities as "scripts." Scripts represent unconscious detailed knowledge that allows individuals to do less m ental processing during frequently experienced events. In this case, scripts were noted from the ethnography to explore the culture behind adopting safety practices and communicating them. in this case for safety for intelligently applying past knowledge to new experience ofte n seem to require common sense and practical rules of 1977:3). 3. The social position of workers in three crews was determined through their socio networks. I elicited data from almos t of workers in each crew (96 percent of workers among the tree crews) in order to determine how workers were connected with each other and the strength and type of ties they share. Four questions pertained to different levels of relationships; first, I as ked about how often they talked to each other during the week; secondly, I asked them how would they classify their relationship (acquaintance, friend, or family); I also asked if they live in the same house; and finally, I asked them who they will ask for a favor like borrow money or give a message to a family member in Mexico or advice like how to pick more fruit or cure an injury from every single member of their corresponding crew. The responses from the questionnaire were used to generate four networks of social relations among workers, each network reflecting the nature of each question. The process was done by generating four adjacency matrices where the first row and first column lists each worker in the crew and the intersecting cells represents the no binary fashion. 4. Each multiple regression corresponds to a linear equation, where y = a + bx 1 + bx 2 + bx 3 n + , and where e is an error term. This model, already applied before (Monaghan et al. 2011b, Mashburn e l al. 2009) is replicated again to explore which variables safety strategies with the addition of social network structures.
149 CHAPTER 4 RESEARCH RESULT: MULTILEVEL ANALYSIS OF MIX DA TA Introduction This chapter present s the results of primary and secondary data analysis . It explores trends, correlated factors, cultural behaviors, and social structures aimed at strategies for safety and health among citrus farmworkers in Florida. The p rocess follow a logic model from simple to complex; having the possibility of selecting an educated set of variable for the final analysis. Descriptive Statistics : Discovering the Invisible Harvester Descriptive and inferential bivariate statistics results presented in here provide a partial picture of the workers who harvest citrus throughout Florida. More specifically, descriptive statistics are used to summarize workers demographics and their working a nd injury experiences in relation to the use of PPE, the effects of CHW, and injuries. With the exception of season 2002 2003, which was a short exploratory survey ( n = 58) on workers awaiting immigration services, all structured interviews were conducted with workers embedded on large harvesting operations ; the advantage of doing so is the growing tendency of this operation and the easy access to a steady work force. Table 4 1 summarize quantitative data collected on six discontinuous season periods, since there were not surveys collected on seasons 2003 20 04, 2007 2008, and 2009 2010. Discrepancies observed between surveys and workers reached are due to not interviewing all the cr ews visited , either for the lack of resources or because the company partner decline interviewing some crews (only field observa tions) ; or because the additional workers were reached by other means like focus groups , glasses tests . All w orkers counted on the surveys were interviewed at least once during the season
150 that lasted 9 months and spanned from October to July. On season 200 6 20 07 , due to the quasi experimental evaluation design implemented , workers were interviewed twice, a baseline interview before CHW were trained, and a post CHW training interview some weeks after they were promoting the use of safety glasses . During this particular season, a crew leader, who was fired by the company, departed with all the workers who had been baseline interviewed, but not observed in the field or interviewed were conducted with them . Finally, other crew was observed once by mistake, safet y glasses were delivery to workers, but they were not interviewed. Table 4 1. Relation of quantitative data collection by season Season Surveys Workers r eached Crews reached H2A workers Crews with CHW Company Partner ( s ) 2010 2011 88 ~250 10 100% 3 1 2008 2009 291 291 12 ~60% 12 2 2006 2007 a 301 311 13 ~50% 9 2 200 6 200 7 b 3 30 398 1 5 ~ 4 0% 9 2 2005 2006 76 297 5 0% 7 2 2004 2005 109 130 6 0% 6 4 2002 2003 58 >58 0 0% 0 n/a Source: Luque et al. 2 007 , Mashburn et al. 2009 , Mona ghan et al. 2011a, 2011b . Note: 2006 2007a post survey, 2006 2007b baseline survey. The main purpose of the observations in the field was to measure the on site use of PPE. The number of crews observed and the size of each crew varied each season, but we experience a tendenc y for a growing implementation of the U.S. Department of Labor H2A visas or the so called guest worker program, which is also noted on Table 4 1. Ideally, each crew observed should have had a CHW, but in the 2006 2007 season,
151 the evaluation design required assigning some crews as control groups without CHW; while in season 2010 2011, our new company partner requested to train CHW as a pilot program in 3 out of its 10 crews, although I was able to participate in their WPS training, where I was able to talk a bout eye safety and also visit all of the crews in the field for that particular season, but only interview crews with CHW. The limitations of the surveys include inconsistency in questionnaire items from one season to the next, except for seasons 2004 to 2006, and post test 2007 to 2009 which were similar; this means that we have five slightly different surveys. Another issue when analyzing aggregated data of all the surveys is an unknown number of workers who were surveyed across multiple seasons; so, al though I present some general results of the fully comprehensive data, the outcomes must be understood with these issues in mind. A key missing piece of information in many of the surveys was the level of formal education of workers; agriculture is a very specialized enterprise, but the complexity of risk associated to farming require learning processes that require some structural process of learning. Finally, one major restriction of these surveys is the exclusive setting of large harvesting companies; al though the tendency is the consolidation of these large operations, there is still large number of workers harvesting in smaller companies, and even independent crews, similar to the ones I worked for one season. Harvesting S easons 2004 2005 and 2005 2006 In seasons 2004 to 2006 a convenience sample of workers were interviewed, all of them were male. For season 2004 2005, the questionnaires ( n = 109) showed that most works were coming from Mexico ( 98 .5 % ) with a small proportion coming from Central American ( 1.5 %) . Their average age was 30 .5 and their average working
152 experience was 4 harvest ing season s , but about one third were first year citrus workers. Season 2005 2006 surveys ( n = 76) showed a very similar pattern: most of them coming from Mexico (97.4%), few from Central America (2.6%), with an average harvesting experience of 3.6 seasons . One particular and important contribution of these surveys was the inclusion of the level of formal education completed by each worker; in season 2003 2004 workers reac hed sixth grade on average, while in season 2004 2005 the average dropped by over one point to a level less than 5 th grade completion. Unfortunately this information was not collected in subsequent surveys. Harvesting S eason 2006 2007 The citrus harvesters whose baseline interview s were during the 2006 2007 season ( n = 3 30 ) were all male (mean age of 30.9) distributed amongst 14 crews (averaging 23.6 workers per crew). Most of the workers were harvesting citrus in Florida for the first time (32.5%), while a nother significant number were picking for a second (24.3%), third (14.6%), and forth (7.6%) season; the reaming 21% were picking citrus in Florida for over five seasons (average 3.8 seasons), getting to the extreme of a 64 years old picker with 40 seasons experience. There was no demographic information of the workers place of origin, but as I mentioned before, season 2006 2007 also had a post intervention survey ( n = 301) and that data was included. Because most of these workers also are part of the basel ine, survey data is very similar in some cases, but it will also allow for a more complex inferential analysis from this particular season. A ll participants of the post intervention interview were male (mean age of 30. 7 ) working amongst 13 crews (23.2 wor kers per crew on average). Seventy seven percent of the workers were from Mexico (26.5 % from Veracruz, 11.3% from Chiapas, 11.2% from Guanajuato, 6.3% from Oaxaca, 4.3% from Guerrero, and 17.1 from other Mexican
153 states). Three percent were from Central Am erica (Nicaragua and Guatemala) and 20% did not provide information regarding their nationalities. We did not interview workers about their legal status , but participant companies informed us that as many as 50% were H2A guest workers. In terms of experien ce harvesting citrus in Florida, 34% reported less than a year of experience while 17% reported more than a year but less than two years of experience. Twenty eight percent reported between two and five years of experience and 21% reported over five years of experience (average 4.4 seasons) . Harvesting S eason 2008 2009 In season 2008 2009, once the quasi experimental design study was completed all participating crews ( n = 12) had a CHW and all harvesters in these crews were interviewed ( n = 291). The avera ge number of workers per crew was 24.3. Besides one female worker who was harvesting with her husband, all workers were male. For the first time in the survey process, one Haitian worker was interviewed; this was not strange, as many crews in Florida are c omposed of only Haitian workers, but the project had not worked with any of these crews, which are mainly restricted to smaller operations. Besides this exception and two Guatemalans (which sum 1% of the sample), all other participants were of Mexican orig in (23% from Veracruz, 13.7% from Guanajuato, 6.2% from Chiapas, 5.5% from Michoacan, 3.4% from Guerrero, 2.4% from Oaxaca, 2.1% from Estado de Mexico, 2.1% from Puebla, 1.7% from Hidalgo, 1.4% from Queretaro, 1.3% from Tamaulipas, and another 1.5% from Co lima, Distrito Federal, San Luis Potosi, Tabasco, and Zacatecas in equal proportions). It is important to note that 34.4% of respondents did not specify location or origin and only said they were from the country of Mexico.
154 In terms of experience harvestin g citrus in Florida, 36.5% reported the current season as their first harvesting, 20.6% were in their second season, 12.2% were in their third season, 7% were in their fourth season, 3.8% were in their fifth season, 2.4% were in their sixth season, 2.1% re ported having picked in Florida for 7 and 9 seasons, and 3.1% picked for 8 and 10 seasons. The remaining 7% reported over 10 seasons with three older workers reporting up to 20, 22, and 37 seasons in the fields of Florida. The average number of seasons wor ked was 3.3. Harvesting S eason 2010 2011 During the season of 2010 2011, all participants ( n = 88) were male (mean age of 31.2) in crews of 24 to 35 workers (the average number was 29.3). One clear difference between these workers and previous ones was the ir legal status and their place of origin. Because all of these workers came to Florida with H2A visas, their place of origin reflects where they were recruited. 100% of the workers were from Mexico (73.9% from Veracruz, 12.5 % from Hidalgo, 3.4% from Puebl a, 2.3% from Chiapas , 2.3% from San Luis Potosi , and the remaining 5.5% from Guanajuato, Guerrero, Morelos, and Nuevo Leon in equal proportions). In terms of experience for the 2010 as their first time harvesting citrus in Florida;14.8% were in their second season, 21.6% were in their third season, and 11.4% were in their fourth seasons, 0.2% and 0.3 % reported having picked in Florida for 5 seasons, 2.3% for 10 seasons, 2.3% for 12 seasons, and the remaining 3.3% workers reported 6, 7, or 8 seasons of experience (they average was 3.6 seasons). When grouped in similar categories of experience, workers from season 2006 2007 are not different from workers in season 2010 2011.
155 Figure 4 1, shows t he average age and working experience of citrus harvesters in Florida. Workers are quite similar in the fact that the age and experience of pickers does not significantly increase over time and there are constantly new workers incorporated into the workfor ce. It is important to notice that the experience of workers collected in these surveys was limited to picking citrus in Florida; but many of them have previous extended experience as farmworkers, even as citrus pickers in their hometowns. It is very commo n for farmworkers to specialize in specific crops, either because they were born where the crop is mainly produced or because they find themselves with the skills to make more than average income harvesting that particular produce. Figure 4 1. Average ag e and working experience of surveyed work er from season 2004 2005 to season 2010 201 1 31.2 30.6 30.7 30.9 32.4 30.5 2.7 3.3 4.4 3.8 3.6 4 0 5 10 15 20 25 30 35 2010-2011 2008-2009 2006-2007a 2006-2007b 2005-2006 2004-2005 Average Age and Experience Working in Florida Experience Age
156 For instance, one of the more senior harvesters I interviewed was a 74 year old man who had been picking citrus in Florida for 32 seasons as of 2006, but we did not ask if he had previous experience harvesting citrus in any other place. However, in 2012, I specifically asked all the workers about their life experience picking citrus and one person had spent 35 years of his 53 year old life harvesting the crop. Perhaps the older workers did not pick citrus in Mexico; that is difficult to know, but an analysis of the last survey (season 2010 2011) revealed that only 30% did not pick citrus, while the workers who had harvesting experience reached up to 19 seasons, perhaps eve n more because the workers whose data was missing ( n = 6) either did not give their age or As mentioned before, place of origin is closely related with the t ime farmworkers have picked specific crops, like citrus. Of the 27 workers with no previous experience in season 2010 2011, over 75% of them came from states other than Veracruz, which for that current season represented 77% (Figure 5 2) of the workers pro viding information about their place of origin ( n = 78). This survey season had the highest number of workers coming from the Mexican state with higher citrus production in the country. Figure 5 3 shows the place of origin of all the workers interviewed fr om seasons 2004 2005 to 2010 2011, excluding the baseline interview of season 2006 2007, which did
157 Figure 4 2. Breakdown place of origin of interviewed workers for season 2010 2011 Figure 4 3. Accumulative relation of f armworkers place of origin 2% 1% 1% 1% 1% 3% 14% 77% Place of Origin Season 2010 2011 Chiapas Guanajuato Guerrero Morelos Nuevo Leon San Luis Potosi Hidalgo Veracruz 0 50 100 150 200 250 300 Overal Place of Origin Season 2010-2011 Season 2008-2009 Season 2006-2007a Season 2005-2006 Season 2004-2005
158 Twenty four states out of the 32 that form the United States of Mexico (the official name of the country) are represented in the five seasons sampled by at least one worker (Aguascalientes, Campeche, Coahuila, Colima, Nuevo Leon, Sonora, and Zacatecas). However, most of the workers by far came from Veracruz, followed by Chiapas, Guanajuato, Oaxaca, and Guerrero. I decided to add the missing data to point out a significant number ( n = 174); as attenuate, for season 2008 2009 ( n = 1 00) workers actually mentioned that they were born in Mexico, but interviewers were not instructed to prompt about the specific state in the country; these were recoded as missing data. Workers from Central America came from Guatemala ( n = 11), Nicaragua ( n = 3), El Salvador ( n = 1), and Honduras ( n = 1). In more than one occasion we found workers from Haiti ( n = 1) and the United States ( n = 2) during visits to the field, but only three of these interviews were conducted. The latter was quite sporadic and crew leaders always joked that these workers usually lasted for less than a week. In the first case, on more than one occasion I met entire crews of Haitian workers in the field. Unfortunately, even if I wished to interview these farmworkers, the limitatio n of communication with these workers was a barrier I do not speak Creole and they had limited Spanish and English knowledge. Inferential Bivariate Statistics: Exploring Relations for the Use of PPE Harvesting Seasons 2004 2006 For season 20 04 2005 and 20 05 2006 safety glasses usage rates were based on 1 8 field observations of 277 workers (Luque, et al . 2009). Ever since then, it was noticed that usage rates may have been markedly lower when observed in the early morning hours because of higher humidity, w hich affects visibility, a fact consistent with future studies (Monaghan et al 2009, Tovar et al. 2014). The results of the observations
159 show that safety glasses use increased over time from a negligible percentage at baseline (<1%), to between 34% and 37% use at the end of the pilot study (Table 4 2 ) . Survey r esults demonstrated that the CHW s had increased influence on crews by helping their fe llow workers with eye washes and other injury related activities (e.g., splinter extractions, minor sprains , etc. ). Pearson chi square tests of association were used for the categorical data to gauge changes in perception of the glasses (Table 4 2) . Complaints about safety glasses declined from 2004 to 2005, with almost three times as compared to complaints in 2004 (43.3%, p < .01) . Other findings include significant increases in 2005 in the percentage of citrus workers who : did not experience dirt in the eyes (38.7%, p < .05); who percei ved that they could harvest more citrus with glasses (27.3%, p <.05); and who perceived that they could harvest citrus faster with glasses (31.7%, p < .05) (Luque, et al . 2009:364). Table 4 2. Results for Safety Glasses Trial Seasons 2004 2005 and 2005 2 006 2004 2005 ( n = 74) % 2005 2006 ( n = 76) % CHW s helped with eye drops 62 64 Concern about eye injuries 80 88 With Glasses, Harvest More* 39.2 53.9 With Glasses, Harvest Faster* 40.6 60.0 With Glasses, Harvest Without Worry 81.1 92.1 No Complaints About the Glasses** 12.2 30.3 Dirt in Eyes Last Week* 40.5 23.7 Wearing safety glasses ( n = 277 ) **** >1 to 34 >1 to 37 Note: * p < .05; ** p < .01; *** p < .001 **** There is a strong significance between start and end of each season usage of PPE (p < .00 1), but there is not statistical significant difference between seasons . Source: Luque, et al. 2009
160 A limitation of the study highlighted above was that c omparison groups were not used to evaluate the pilot project because of time, budgetary, and access re strictions ; and that there was not a correlation build in to the use of glasses and the presence of CHW on the crew. Harvesting Season 2006 2007 An examination of members of the intervention crews, use of safety glasses , and level of contact with the CHW (CHW helped, CHW did no t help, or CHW was un known by the harvester ) during season 2006 2007 revealed a statistically significant relationship ( Â² =39.00, P < .001). Nearly half of the workers who received help from the CHW (48.9%) were observed wearing safety glasses. Workers on crews who knew their CHW , but did not receive help were somewhat less likely to wear glasses (31.8%), and just 24.0% of workers who did not know the identity of their CHW wore them (Table 4 3 and Figure 4 4) . T able 4 3 . E ffect of CHWs on PPE usage in a 2007 quasi experimental test . CHW Helped CHW Did Not Help CHW Not Know Was worker wearing glasses at the time of the i nterview? No. (%) Ye 44 (48.9) 21 (31.8) 6 (24.0) No 46 (51.1) 45 (68.2) 19 (76.0) Total 90 66 25 . Source: Monaghan et al. 2011b
161 Figure 4 4. PPE usage and CHWs help for Season 2006 2007 The fact that most crews had at least one CHW with the sole purpose of increasing PPE usage does not solely explain their decision to use or not use the equipment or, for that matter, increasing prophylactic measures like eyewashes, but it has been shown to be an influencing factor. Harvesting Season 2006 2007 Similar to the longitudinal analysis of season 2004 2005 and 2005 2006, th e analysis for season 2006 2007 uses the baseline and follow up surveys. This analysis excludes a couple of crews that completed the baseline but did not participate in the follow up interview. Table 4 4. Injuries and perceptions of PPE usage seasons 2006 2007 and 2008 2009 2006 2007b ( n = 310 ) % 2006 2007a ( n = 301 ) % Dirt in Eyes Last Week * 29 38 S omeone promote the use of PPE 15 10 Reported any kind of eye injuries in current season 44 49 Overall PPE usage of workers with CHW * * 13.7 33.8 Note: Si gnificant * p< 0.05 ** p < 0.01 0 5 10 15 20 25 30 35 40 45 50 CHW Helped CHW Did Not Help CHW Not Know CHWs Help and PPE Usage PPE usage Without PPE
162 As in the previous season, the difference in the reported proportion of workers with dirt in the eyes last week was significant, but the direction was negative, with an increase in the number of injuries reported. Although w e could not be sure, we considered that the increase in reports may have resulted from the encouragement of CHWs to report injuries. The other two different proportions analyzed, although not significant, also point to this phenomenon. As with previous sea sons, the increase of glasses use was highly significant from the baseline to the end of the season. Figure 4 5 shows results from the baseline observation and the post training of CHW observations on the treatment crews ( n = 10). With the exception of one crew who observed a 3 point decreased of usage, all crews significantly increased. Crews with small increase were attributed to short period of time from training to observations and crews who have not previously exposed to the PCWH program. Each overall observation is composed by a set of four repetitive measures on three independent observations. In other words percentages are the results of observing each crew twelve times. Figure 4 5. Overall percentage of use at baseline and follow up observations 0% 10% 20% 30% 40% 50% 60% 70% 1 2 3 4 5 6 7 8 9 10 0% 2% 0% 26% 24% 14% 37% 20% 0% 14% 8% 12% 32% 53% 38% 34% 35% 66% 20% 33% Crews Overall
163 Harvesting Season 2008 2009 Finally, Seasons 2008 2009 and 2010 2011 only had one observation and survey each, so there was no data to compare PPE usage change over time. However, some bivariate data analysis was completed exploring some of the significant relations we have seen. From Season 2008 2009 cross tabulations (Table 4 5 to 4 7) results shown that there was not significant correlation with the use of glasses and having an injury in the current season, or having a person promote the use of safety gl asses. Nevertheless, similar to the previous season, knowing the CHW was statistically significant with wearing the safety glasses at the time of the observation (Table 4 7and Figure 4 6). Table 4 5. Cross tabulation of PPE usage and injuries in Season 08 09 Eye injuries in the current season Total Yes (%) No (%) Using the glasses when approach for the interview Yes 16 (7.1) 14 (6.2) 30 No 72 (32.2) 121 (54.2) 193 Total 88 135 223 Table 4 6. Cross tabulation of PPE usage and promotion of PPE usage Season 08 09 Did someone promote the use of PPE Total Yes (%) No (%) Using the glasses when approach for the interview Yes 27 (12.1) 3 (1.3) 30 No 172 (77.4) 20 (9) 192 Total 199 23 222 Table 4 7. Cross tabulation of PPE usage and having a CHW h elp Season 08 09 Did the CHW help you? Total Yes (%) No (%) Using the glasses when approach for the interview * Yes 27 (12.1) 3 (1.3) 30 No 144 (64.9) 48 (21.6) 192 Total 171 51 222 Note: * p = 0.05
164 Figure 4 6 . PPE usage and knowing the CHW for S eason 2008 2009 It may be that the findings of the Season 2008 2009 are influenced by the long period in which the PCWH has worked with some of these crews. Although PPE usage decreased from previous years, each crew maintains regular users and eyewash is a regular practice that is reflected by the low number of injuries reported by the company. This was the last year in which the project was active with this company, but it decided to keep CHWs as an internal position in their crews. Harvesting Season 2 010 2011 Table 4 8 and 4 9 shown results from bivariate analysis of Season 2010 2011. There was not significant evidence that either having debris in their eye or knowing the CHWs had an impact on the use of glasses. These crews were from a company new to the PCWH and were composed exclusively of guest workers (H2A), which may have affected the findings as there was no long term exposure to the program (Table 4 10 and Figure 4 7). A more detailed exploration of this Season is done by social network analysi s and multivariate analysis. 0 20 40 60 80 100 120 140 160 PPE usage Without PPE PPE Usage and Knowing the CHW CHW help CHW did not help
165 Table 4 8. Cross tabulation of PPE usage and having eye debris Season 2010 2011 Have debri s this season Total Yes (%) No (%) Using glasses at the time of interview Yes 14 (16.3) 10 (11.6) 24 No 30 (34.9) 32 (37.2) 62 T otal 44 42 86 Table 4 9 . Cross tabulation of PPE usage and knowing the CHW Season 2010 2011 Do you know the CHW Total Yes (%) No (%) Using glasses at the time of interview Yes 16 (21.9) 5 (6.8) 21 No 37 (50.7) 15 (20.5) 52 Total 53 20 73 Table 4 10. Cross tabulation of PPE usage and exp erience using PPE Season 2010 11 Use glasses in the pas t Total Yes (%) No (%) Using glasses at the time of interview * Yes 17 (20) 7 (8.2) 24 No 24 (28.2) 37 (43.5) 61 Total 41 44 85 Note: * p > 0.05 Fi gure 4 7 . Chart comparing workers using glasses with previous experience with PPE 0 5 10 15 20 25 30 35 40 With PPE Without PPE PPE Users and PPE Experience Experiece No Experience
166 Script Analysis: Making Sense of the Ethnographic Experience Coaching between workers is a daily routine embedded in verbal and non verbal communication , analyzed in here as scripts ( Schank and Albelson 1977) . Examination of field notes highlight safety and health trends link to the use of ladders, clippers, and containers; rehydration and food intake; utilization of proper working outfit; prevention and treatment of injuries and illness; and living conditions. Examination, marking, and positioning of ladders ; baÃ±os (boxes) usages; liquid and food sharing are examples of non verbal peer teaching experiences. Pre and post harvesting conversations regarding or (like careless driving, unpaid work, abandonment on the fields, dismissal of injuries); hazardous working condition warnings (for example ant locations of cheap or free safety clothing (long sleeve shirts, socks, gloves, and hats) ; discussion of injuries or ill treatments (like over the counter drugs, traditional healing, or home remedies); and sharing information on the schedule and locations of health care services are examples of verbal str ategies for safety and health. These routine activities and verbal communications represent casual conversations, scripts transmitted as strategies to other workers. main concerns for safety in the groves revolved around the use of ladders and man y routine activities , i.e. scripts, showing strategies taken to avoid accidents. When arriving to a new grove to pick, the early morning routin e was to select ladders and checking their structure . For that purpose, they laid the ladders on the ground and w alked over them looking for cracks, missing rungs, or bent corners. O nce its safety was test ed, picker s marked it with any object on hand ( bandana, a plastic bag, a sweater). The script was
167 non verbal, but it became verbal if the ladders had a structural p roblem . Sometimes mainly in large operations they were mark ed with red tape , and definitely ended at the bottom of the pile . Of course, occasionally even defective ladders h ad to be use d, because there was not enough ladders for all the crew workers, but specific instructions about temporary fixes (count missing rungs or tie a bandana for example) and use limitations were given (like gentle escalerazo or limited heights). Learning to escalerear is like learning to dance, you look for the good dancer and i mitate their steps. An eighteen foot ladder weighs 27 pounds, and moving that . Moreover, carrying the ladder in the proper position facilitates quick harvesting and safety of location amongst the tree canopies. The goal is to move the ladder the least to collect the most fruit, but the workers must also ensure that operations are conducted in a safe manner. The learning experience came many times from painful experiences: when I fell fr om the ladder the first time and spoke with co workers on our way home, all of them confessed having fallen from a ladder at least once. They went on to describe their falls and engaged in conversation about managing a fall. Some advice from these scripts sideways to move it to the center of the tree, where larger branches will hold it; jump on the first steps to secure the base in the ground; move it side to side to look for b roken branches; if you use a jumbo sack, hold it on the steps so it will help to stabilize the ladder; and if picking with ayate, avoid sudden turns so the ladder will not twist around you (as was the case with my fall). All these acts are taken by workers in an automatize script learned out of habit from years of experience.
168 Wooden ladders, while scarcely used are still are used in smaller crews from time to time, and require different inspection and management to ensure they do not have cracks. And while they might not have bent pieces like aluminum ladders, wooden ladders could have splinters, water damage or termite damage. Wooden ladders are heavier than aluminum ladders, but casually are easier to position and remove from the branches because they are easier to turn around. Because most workers came from Veracruz, where ayates and wooden ladders are common, they complain of modern jumbo sacks and aluminum ladders for reasons dealing more with productivity than safety. Clipper and ganchos, while rarely u sed, came on casual conversation (scripts) among workers. Conversations about clippers usually focused on blade sharpness, springs resistance, and handle additions and bolts conditions (inspection). For hook usage and positioning of tool, an end attachment to avoid traumas was also mentioned. BaÃ±os (boxes) are another basic tool for harvesting and also the topic of verbal and non verbal scripts related to safety usage. Non verbal examples were centered around management of boxes in the field: location (wher e to locate baÃ±os for safe and easy access), transportation (how to move baÃ±os to avoid injuries), and harvesting (dropping of oranges and use of yogas for faster filling, as well as standing on them to collect last citrus on surrounding trees). Verbal scr ipts concentrated on constant coordination with crew leaders about the number of baÃ±os and their location in relation to trees and the road. Other risks were cover ed using scripts, but were not equally use d among crews setting. For example, scripts related to PPE usage were only present i n crews where
169 CHW were present, which also may cognitive construction of ) . It was clear from the analysis that workers in companies wher e proactive safety policies had higher exposure, discourse and verbalized issues related to eye safety were more prevalent. In smaller crews with limited participation of crew leaders in the safety management, such discourse was not as prevalent. Script p revalence indicated safety priority and if PPE was not mentioned amongst a crew, discourse about eye injuries was a combination of scripts about how to it is our daily bread) , meaning such injuries are part of the work that puts their daily food on the the counter eye drops, company provided eye, or chamomile eye drops. Beverage and food shari ng among workers is also part of their natural daily behavior and serves as an example of non verbal script. Many times we exchanged with positive or negative comments (scripts) of the intake: beer and spicy food in the morning to help with the hangover; c affeinated drinks only in the afternoon after a lot of water and lunch; heavy breakfast and light lunch to avoid cramps; and finally, extremely high protein in each meal to provide the energy workers need. Pre and post harvesting conversations regarding c behavior were very common and are in many ways related to health and safety. For example, crew leaders were identified and avoided by workers for their poor or careless driving, salary theft, abandonment on the fields, and dismiss al of injuries record. Workers with drinking or drug problems, violent backgrounds, regular use of sex
170 workers, and unpaid debts were also branded as bad influences who were risks to other workers safety, health, or even detection, detention, and other con sequences of workers by law enforcement officers. Hazardous working condition warnings (e.g., ant hills, wasp nests, chemical residue, risky wildlife sightings, or power lines) were immediately reported at the time of their encounter. Such hazards were shou ted out by workers. Given the spread of workers, many repeated the risk and its location; at the end of the day, more details of the event were brought up in conversation, especially if the crew was scheduled to return to the same grove. Proper clothing a nd gear for the daily journey is generally assumed. All harvesters know, non verbally, that long sleeve shirts, socks, and gloves protect against tears and branch scratches; hats prevent sand and branches eye injuries in addition to shading the harsh sun. Closed toed shoes and long pants protect from injuries in the feet and legs. Procurement of the proper outfit was typically low in cost and local deals were shared amongst workers. For example, when a church, organization, or store had items that could be of interest for other members of the crews, workers notify each other with a phone call. Discussion of injuries and illness treatments occurred when one member of the crew, a worker in the camp, or someone close to a worker was in poor health. Over the cou nter drugs, traditional healing, huesero (traditional T
171 the workers and how the company was managing the case. Co workers followed the outcome of such workplace injury cases, so they would know what to expect if they we re ever injured. Finally, workers shared information regarding the hours and locations of health care services such as clinics, health fairs, and emergency rooms and their cost, requirements, and benefits. Information on pregnancy and child services was al so shared, but only when a health issue occurred that necessitated such information . SNA: Knowledge, Support, and Safety Socio centered networks focus on bounded social groups. In this particular case, the crews are the unit of analysis. The goal is to und erst and how the structural position of actors within the group explain s both individual and group level phenomena (Wasserman and Faust 1994). Social network analysis was used to exp lore the links between workers i n three crews. First we explored the overal l links within each crew . One common practice is to elicit alters (members of the crew with who m they may have a connection ; however, rather than that, I used the list of crew members to revel four spheres of the relationship: a) if they were living in the same house , b) how often they talk to each other, c) what kind of relation they have with each other (family, friend, acquaintance), and d) whom they feel comfortable ask ing for favor s such as asking for money or giving a message to a family member in Me xico or advice (like how to pick more fruit or cure an injury). The last question was used to generate an adjacency matrix where the first row and the column list each of the workers in the crew and the intersection of each of the cells represent s the tie between workers (in a yes or no binary fashion).
172 Within each crew, a ll workers were able to identify each other, but not all workers were present at the time of the social network interview . Those absent are not shown their corresponding matrix ( n = 5, t wo in two crews and one in the third one). The social network interviews in their homes lasted 45 minutes on average and workers were compensated by selecting donations from the FWAF (PPEs, hygiene products, clothing, and /or food). The workers did not ove rlap between crews; there was no overlap because they came from differen t locations in Mexico, traveled in separate buses, and lived i n distant towns of Florida. The only time when workers from different crews were in the same location was when they arrive d from Mexico at the central location of the company to finalize their paperwork and receive their safety training . Even then they stay ed within their own crews. To insure the boundaries of crews , I probe d during each visit about how much workers knew of t he other contratados (H2A workers like them hire d in Mexico) either in their hometown or the other workers camps, and if they kept communication with them after the time when they were in the central location . After several visits , it was clear that they o nly ever hear d about the other crews if their chivero (crew leader) mentioned where the other crews were working . They did, however, ask me about how the other crews were doing in terms of work, payment, and location , further indicating that they did not c ome by this knowledge on their own. Somehow, the crew leader and I were the only link between crews. In this study I focus on three centralit y measures of worker s in between crews as a way to see how the use of PPE could have been influence d by the worker s social network. These mea sures are (1) degree centrality, (2) betweenness centrality, and (3)
173 closeness centrality . Degree centrality is a measure of network activity (McCarty 2002); it reveals how many direct ties an individual has to others in a netwo rk. Degree is an indicator of access to sources of information. Actors who have more ties have greater opportunities because they have more choices. This autonomy makes them less dependent on any one specific actor, and hence more powerful (Hanneman and Ri ddle 2005) and influential among other workers. Closeness centrality is a measure of independence from the control of others (McCarty 2002) . Closeness is measured by the number of ties an actor has to go through in order to reach another actor. These dista nces, called path lengths, can go from one, when an actor has direct tie with another, to several, depending on the size and characteristics of the network. An individual who has many one path length contacts is in a favorable to position to spread and rec eive information (Hanneman and Riddle 2005). Betweenness centrality is a measure of information control (McCarty 2002) and indicates the extent to which an actor can be a broker or a bridge in a network, a position which might indicate a potential for cont rol over others (Scott 2000) or control over the flow of information. Betweenness, in other words, measures how often a person lies along the shortest path between two other persons. Actors that lie between other pairs of actors have the capacity to isolat e people (Hanemann and Riddle 2005) as well as bringing together separate groups. Social networks were analyzed with UCINET software which provided visual and statistical analysis of networks. Results were then correlated against measures relevant for this study: (1) PPE usage; (2) and CHW influence . In addition to generating the SNA, participants were re interview ed with the network results to clarify the links and nodes
174 formed. The combination of field observations and follow up interviews with the SNA pr ovide a survey The visual impact of SNA is perhaps one of its more convincing explanatory tool s . I combine visual representations, previous analysis, and centrality measurements to explain PPE usage and CHW influence . In visual representations, each actor is presented with a dot, ( node ) ; the reported relations among actors are presented with lines conne cting the dots. The lines have arrows that note the direction of the relation. When an actor is mentioned by another actor the arrow point s in the direction of the person mentioned. The bigger or denser the network is, the more complex and difficult it is to draw conclusions or even observe the arrows , as illustrated i n each f igure (Figures 4 8 to 4 17) . The networks are presented in alphabetical order as crew A (Figure 4 8 to 4 10 ), crew B (Figure 4 11 to 4 13 ), and crew C (Figure 4 14 to 4 17 ). The visua l representation s of crews A and B are similar ; all actors are connected, although crew B has a denser network, or more connections between nodes. Visual representation of crew C , in contrast, shows a completely different structure, including one dyad and two isolate s. Isolates are actors who neither mentioned nor were mentioned by anyone who m they would ask for help or advice . Crew B also has one isolate . Colors represent usage of eye safety glasses at the time of the observation: in magenta are the worker s using safety glasses, blue are workers without glasses, and black are workers who were resting when I visited them in the field (therefore, safety glasses usage was
175 Mexico. Crew A is the most homogeneous, with workers coming from only three different states. As expected, the network with the most components (Crew C) is the most heterogeneous in terms of place of origin. Finally, the size of the nodes reflects whether or not w orkers received CHW training, with the larger nodes representing those who participated in the training for four weeks, approximately 10 hours each. Smaller nodes correspond to workers untrained through CHW. Crews A and B have three CHWs each, and crew C h as 12 . The higher proportion of workers trained in crew C was possibly due to its isolation and layout of th e camp with little else to do, training became a desirable activity. The social network analysis section concludes with the incorporation of the centrality measurements of each crew in bivariate and logistic models. Crew A Social Network Analysis Figure 4 8 . Crew A single component social network. Color distingui sh PPE user (magenta) and not users (blue) , bigger nodes are CHW, shape of the nodes indicate place of origin (n =3) , direction of arrows shows who they will look for help or advice
176 Crew A ( n = 24), represented by Figure 4 8, was a very experienced harvest ing crew (15.4 seasons on average); most of their members came from the well known citrus town of Alamo, Veracruz (circle shape). As expected, at least two of the CHW were central in the network and appear to have influenced other workers in the usage of P PEs, reaching 33% usage at the time of the visit. That means two thirds of the crew either was not wearing glasses or could not be observed (resting.) Workers coming from Hidalgo ( n = 2) were also wearing safety glasses and one of them was also a CHW, whic h also reinforces the idea of social influence among their peers. Curiously, the other CHW who was trained was not wearing PPEs, but seems to have been influencing other workers who have ties with him in terms of asking for favors or advice; interestingly enough one of the CHW who wear the glasses requested help from the CHW who does not wear his glasses, but the association was not reciprocated. If we examine degree centrality we can better explain each of such influences (Figure 4 9). Figure 4 9 . Crew A degree centrality represented by size, and PPE usage .
177 There are four actors in Crew A with the highest degree centrality (OMC, JLSJ, JSHT, and ESP); three of them are wearing glasses and may be influencing other workers to do so. The other one, ESP, who i s not wearing PPE may be influencing ESJ, the CHW who is not practicing eye safety practices in fact, ESP is the father of ESJ so this influence is quite likely. In addition, the workers who are closer to ESP may be also influenced by him as a senior perso n. Additionally, ethnographic data helps explain the cluster of non users around ESP. Although data on religious practices were not observation that this crew was divided into t wo groups: Evangelical Christians and Catholics. ESP was a middle aged, well spoken man and the Evangelical leader in the crew. Company management even confided that, although he barely made the harvesting quota, he has a positive moral influence among the other workers that was as important to the company as his harvesting skills. During the general safety meeting, ESP requested more control over the visits of sex workers and alcohol consumption in Within this network, alters with the highest degree centrality also scored the highest levels of betweenness and closeness. Figure 4 10 represent the degree centrality of actors combined with the previous experience of workers with the usage of safety glasses. Many workers had used PPEs in previous jobs, but most such experiences were in heavy industries (oil industry, foundries), construction, or gardening. Contrary to what we expected, previous use of safety glasses did not necessa rily translate to adopting the practice in the fields. About half of those with previous experience used them now, while the other half stopped using them.
178 Figure 4 10 . Crew A closeness centrality represented by size and previous experience of safety gl asses usage Although the graphs provide a clearer picture of the interactions happening on power and influence. Besides the possibility of using these measurements in mult iple regression analysis, Tables 4 11 to 4 13 present outcomes of the main actors for each crew, including CHW, in addition to descriptive statistics for each measurement. Table 4 11 . Crew A normalized centrality m easures ID Degree Closeness Betweenness O CM * 95.652 95.83 23.68 JLSJ * 86.96 88.46 15.64 JSHT 73.91 79.31 11.02 ESP 69.57 76.67 10.40 ESJ* 30.43 58.97 0.17 Mean 34.42 61.972 2.981 St. Dev. 23.4 5 11.268 5.962 *Trained as CHW
179 OCM and JLSJ were young workers (28 and 25 years old respectively), but with plenty of harvesting experience. They both are from the citrus region of Alamo, Veracruz. In Mexico, OCM started picking oranges at 15, when he had to support his family and JLSJ started picking with his father starting at age 12. OCM and JLSJ kn ew each other back home, and remained close in their land. OCM was in his third season in Florida, while JLSJ was coming for the first time. They also had experience using PPE in other industries; they both worked in an electrical cable factory back home, and OCM was also required to use safety goggles in a packing house at the Mexico USA border. JSHT at 23 years of age was one of the youngest members of the crew and one with the least experience back in Mexico. He came from a region in Hidalgo where there is no citrus production; however, he and the other workers from the same region were among the first to arrive as guest workers (H2A) to this company, making him among the most experienced in Florida. This made him a sought after source of information for other workers, and although he declined to be trained as CHW, he clearly saw the Conversely, ESJ (as mentioned before) was trained as CHW but graphs and measurements o utcome indicate he had only a marginal influence among the other workers. Crew B Social Network Analysis Crew B ( n = 29) was the network with the highest density (1.67) compared to the other two crews (crew A = 0.23, and crew C = 0.05); it was also the cre w with the lowest usage of safety glasses (7%) and with the highest number of eye injuries among the participant crews that season (crew A = 1, crew B = 4, crew C = 0). Paradoxically, it was located near the headquarters of the company, and shared the larg e camp with two
180 injuries in the other 7 crews of this company, but the total number of eye injuries among all the workers in this company all that season was 23 (down 5 0% from the previous year, and the main reason was because the company looked for the assistance of the PCWH). We worked with this company the following season, but there was no data collection, or follow up training for CHW. All workers, as in the first s eason, were trained in the use of glasses and eye wash when they were arriving from Mexico and receiving their WPS training and finalizing their paper work. Injuries, however, went up again this year ( n = 35), due to one crews having 7 eye injuries in one day for chemical residue in the leaves; which was not the only case. Currently, due to financial restrictions, the company does not give PPE to all workers, but only to workers who have had an injury; although eye wash training and practices have stayed in place. The workers in crew B came from five different states in Mexico, but the vast majority (80%) of the crew was composed of harvesters coming from Atotonga, a small city in the hills of Veracruz, near the coastal citrus growing region of the state. Th e crew of 29 workers has a mean working experience of harvesting citrus in Mexico of 10.07 seasons, and of 2.67 seasons picking citrus in Florida. In Figure 4 11, usage of safety glasses at the time of the observation is denoted by color: magenta for users ( n = 2), blue for non users ( n = 27); trained CHWs are represented by larger nodes (ACM, JDFS, and LTC); and shape of the node indicates place of origin. Surprisingly, although most of the workers came from Veracruz none of the workers who volunteered to be trained as CHW was from that state. This may be another reason why their influence among other crew members was limited. Also contrary to the other two crews most of
181 the workers who completed their CHW training ( n = 3) were not using safety glasses at t he time of the observation. In Figure 4 12, degree centrality attributes represented by the larger size of the nodes were used in combination with the previous experience of workers using eye safety equipment in another job (magenta color), and trained as CHW (shaped as circles). A few of them had such experience ( n = 7), but, unlike the case with Crew A where about half of previous users utilized glasses while picking, in Crew B none of them do including the only trained CHW who admitted using safety glas ses before. Of particular importance is also the fact that only this same worker was among the most central members of the crew, due to his condition of the most experienced worker picking citrus in Florida. Figure 4 11 . Crew B social network. Color dis tinguish PPE user (magenta) and not users (blue) , bigger nodes are CHW, shape of nodes indicate place of origin, direction of arrows determine who they will look for help or advice
182 Figure 4 12. Crew B degree centrality reflected by the size of the nodes , previous experience using PPE (magenta color), and trained CHW (shaped as circles) One more analysis was added to explain the crew behavior. Figure 4 13 represents the betweenness centrality and the self reported use of PPE in the season. According to wo rker accounts, most of them used the glasses for at least one day, including the users at the time of the observation and the CHW (as expected). Among workers with the highest centrality measurements was a CHW with experience using PPE when he worked in a nursery in Mexico. But his position did not translate to a bigger impact on worker usage of safety glasses. One possible, negative effect of this low usage observation is the fact that five workers at the time of the visit reported having damaged safety gl asses; which were replaced but not counted as used. Safety glasses, on a daily base usage do not last more than two weeks due to scratches on the exterior of the glasses and deterioration of anti fog film on the interior, which in both cases reduces visibi lity of workers.
183 Figure 4 13 . Crew B help or advise social network, size reflect level of betweenneess , color self report PPE usage Table 4 12 gives us the specific centrality measurements for crew B. As with Crew A, only the main actors are included: th e workers with the highest centrality and CHWs. Of the five more central alters, one of them was a CHW: JDFS. While he was a trained CHW, not only was he not wearing safety glasses in the field but also suffered an ant bite in his eye which got infected an d kept him from work for one week. This likely made him lose credibility as a CHW within the crew, promoting eye safety while not attending to his own. In exactly the same central position, there were two other workers who were not using glasses, increasin g my belief that workers were not motivated to use glasses when looking at their peers with the most influence among them. The other two trained CHW, one of which was using PPE at the time of the visit ich may explain the limited
184 Table 4 12 . Crew B normalized centrality m easures ID Degree Closeness Betweenness EOR 96.43 50 12.94 EZA 96.43 50 12.94 JDFS* 96.43 50 12.94 FZA 89.29 48.2 8 9.53 PHG 85.71 47.4 6 8.17 ACM* 25 36.84 0.28 LTC * 21.43 36.36 0.11 Mean 37.69 39.32 2.05 St. Dev. 26.49 4.69 4.31 Note: * Trained as CHW Crew C Social Network Analysis Crew C was unique in several ways. It was the furthest away from the company headquarters, it was the most heterogen eous in terms of place of origin in Mexico, and it was housed in a semicircle of portable trailers, many of which were abandoned. From the first visit, harvesters were eager to engage with the researchers and change the routine. Most of them wanted to be t rained as CHWs although some missed one or two of the trainings and therefore did not become CHWs per se. Beyond this common willingness to be trained and spend time together a few Sundays while in these classes, this was not a unified group. Distance amon g them was palpable. In visiting this crew, I had to be careful to spend similar amounts in each trailer to avoid appearing as if I favored one group over another. While there was no tension felt or expressed, it was still clear that there were three separ ate groups in the camp as Figure 4 14 demonstrates. In this diagram, the colors represent the workers who were wearing glasses at the time of the observation (users in magenta, non users in blue, and one worker not present at the time in blank); node size shows workers who were trained as
185 CHW; and the shape denotes places of origin of each worker (there are a total of nine locations, two of them in Veracruz). Crew C (n=35) network has five components (two isolates, one dyad, and two larger networks). It is very dispersed (low degree of interconnectedness) (Figure 4 14), which also challenges the notion of influence in adoption of eye safety practices. Half of the workers who were trained as CHWs were wearing glasses ( n = 6), plus another seven non CHW worker s totaling 37% PPE usage at the time of the unannounced observation. The composition of the groups is quite interesting I had expected the two larger components to be grouped around state of origin but this was not quite so. While it is true that in each l arger group one place of origin is predominant (Veracruz/circles and Hidalgo/triangles) there were also people from different places in each one of them. In other words, the origin of each group did not determine the structure of the network. Figure 4 14 . Crew C network components. Color distinguish PPE user and not users, bigger nodes are CHW, shape of nodes indicate place of origin, direction of arrow is who they will look for help or advice
186 Figure 4 15 combines the degree centrality of the network (re presented by the size of the nodes), the usage of PPE at the time of the observation (magenta user, blue non users), with previous experience using safety glasses (circles=past users, squares= without previous experience, and triangles are missing informat ion). A large number of workers had previous experience using safety glasses (73%) and apparently it was more likely that workers with previous experience were likely to be wearing glasses harvesting citrus. Figure 4 15 . Crew C social network degree cen trality (size of nodes), PPE use (magenta user and blue non users), and past PPE usage (on circles) Table 4 13 shows the actual centrality measurements of only the first six more central actors. Because of the large number of CHWs ( n = 12), not all are inc luded.
187 Centrality measures are much smaller for Crew C than for the other crews, given the high dispersion of the 35 alters and the five different components that emerged. Table 4 13 . Crew C normalized centrality m easures ID Degree Closeness Betweenness D MS 29.41 6.42 24.19 HEMB 20.59 6.3 3.91 MGE * 17.65 3.82 1.16 RMV * 17.65 3.82 1.16 BMH 14.71 6.28 2.16 TOM * 14.71 3.82 3.20 Mean 8.07 5.23 2.70 Std. Dev. 6.18 1.21 4.93 Note: * CHWs As expected, the largest component of the network contains the pe rson with the highest centrality measurements for the entire network, DMS. Because he had family members in the crew, and also was very social, he serves as the link between four other components or clusters. In other words, if DMS was not present in the c rew, the larger cluster would become four unconnected components. Let us look more closely at him. This worker began going to the field back in Veracruz at age eight. As is common for children, his job then was changeando (picking the fruit in the middle o f the tree by climbing the tree trunks). His lack of previous experience using safety glasses may be due to his young age, 21 (Figure 4 16 ) . DMS was convinced that it was impossible to wear glasses while picking and he never wore them (Figure 4 17) despit e having been provided with them.
188 Figure 4 16 . Crew C reflecting b etweenness centrality measure by the size of the nodes , PPE usage (magenta user, blue non users) , and past PPE usage on circles (squares have not previous experience using glasses) . I solate s and dyad, although not present in here are accounted for the drawing. Figu re 4 17 . Crew C closeness centrality measure represented by size of nodes, self reported PPE usage (magenta users, blue non users) , and self report of PPE usage on circles (squares represent workers reporting no wearing safety glasses in the current season )
189 Importantly, all of the workers in Crew C with previous experience wearing PPE were using eye safety protection at the time of the observation. None of the workers who had not used PPE in the past were wearing them (Figure 4 1 6 ); however, half of these w orkers self reported using the safety glasses at some point in this current season while harvesting (Figure 4 1 7 ) when asked about wearing the PPE . Curiously, one of the workers, RVM, had self reported not wearing the glasses during the seaso n but was observed with them. This may reflect a contamination effect by the presence of the observer or by misunderstanding the question of past use. Crews Comparison Figure 4 18 compares the means of the centrality measurements of all three crews (A, B, C) and the PPE usage at the time of the observations. As described above, the crew (B) with the highest density (1.67) and mean degree centrality (37.69) also showed the lowest eye safety glasses usage (7%) of all the crews, while the crew (C) with the low est density (0.05), degree centrality (8.07), and closeness centrality (5.23) had the highest PPE usage (37%). Although more analysis is needed, this initial finding contradicts our expected prediction of central CHW influencing peer workers to adopt the u sage of safety glasses. Perhaps the differences in PPE usage in these crews (B and C) are not completely related to the overall structure of their social network, but also to the number of workers trained, and the specific location and commitment of the w orkers. Crew A is the closest to what we were expecting regarding the effects of CHW on the network, with the highest closeness (61.97) and betweenness (2.98) centrality measures. The centralized CHW may have had a real diffusion of innovation effect among their peers;
190 the same way the central workers who did not use glasses also may have affected other workers in their resistance to wear them. When centrality measures are incorporated in a logistic regression model, the analysis is conducted at the crew l evel ( n = 3) since each has its particular structure. Results of a bivariate model using each centrality measure to predict the adoption of eye safety equipment were not significant. Although in one crew it is significant in a multivariable analysis, and m oreover, betweeness centrality was a statistically independent variable among the two crews with greatest eye safety equipment usage: crews A (p =0.193) and C (p=0.124). Figure 4 18 . Centrality measures of crews and PPE usage comparison 0.227 34.42 61.972 2.981 33% 1.672 37.69 39.32 2.05 7% 0.05 8.07 5.23 2.7 37% DENSTY DEGREE CLOSENESS BETWEENNESS PPE USAGE A B C
191 Multivariate Sta tistics : Explanatory Models for Injury, Safety, and Health Lastly are presented the results of general logistic regression models for harvesting seasons 2006 2007 and 2010 2011 . Harvesting Season 2006 2007 Logistic regression was used to analyze the data from the 2006 2007 season and to explore the use of PPE, controlling for years of harvesting experience and age group. Examination of the parameter estimates for the individual categories showed that workers in intervention crews with 1 2 years of experien ce (OR=2.89, 95% CI 1.11, 7.55) and who knew and received help from the CHW (OR=3.73, 95% CI 1.21, 11.57) were significantly more likely to use glasses than other intervention crew members. Table 4 14 . Logistic regression results for use of PPE Season 2006 2007 Parameter B(SE) OR (05% CI) Experience, y 1 2 1.062 (0.49) * 2.89 (1.11, 7.55) >2 35 (0.41) 0.70 (0.31, 1.57) Younger (<29 y) 1.24 (38) * 0.29 (0.14, 0.61) Know CHW and received help 1.32 (0.58) * 3.73 (1.21, 11.57) Know CHW but received no he lp 0.52 (0.58) 1.68 (054, 5.25) Note: OR = Odds Ratio CI = Confidence Interval , * p > . 05 Source : Monaghan et al. 2011b. Regarding other variables, workers in intervention crews who were less than 29 years of age (the mean age of the workers interviewed ) were significantly less likely to wear the glasses than were older workers (OR=0.29, 95% CI 0.14, 0.61). In this model, knowing the CHW was insufficient by itself to improve the use of glasses (OR=1.68, 95% CI 0.54, 5.25) as was having more than two year s of experience (OR=0.70, 95% CI 0.31, 1.57).
192 Ha rvesting Season 2010 2011 For season 2010 2011, I analyzed the data using three multivariate logistic regression models (A, B, and C) : Model A explores similar variables relevant for Season 2006 2007(Table 4 15) . C ontrary to the previous model, associations in this model were not significant. Table 4 15 . Logistic regression A results for use of PPE Season 2010 2011 Parameter B(SE) Or (05% CI) Experience, y < 2 0.409 (0.69 ) 1.5 (0.38, 5.87 ) Younger (<29 y) 0.230 (0.66 ) 1.25 (0.34, 4.63 ) Know CHW and received help 0.654 (0.75 ) 1.92 (0. 41 , 8 . 38 ) Know CHW but received no help 0.432 (0.78 ) 0.64 (0.14, 2.99 ) Note s : OR = Odds Ratio CI = Confidence Interval M odel B incorporate s the previous experience of worker s ( n = 81) using safety glasses (significant in Table 4 10), the time they had picked citrus in Florida, and experiencing any kind of eye injury in the current season (Table 4 16) . Table 4 16 . Logistic regression B results for use of PPE Season 2010 2011 Parameter B(SE) O R (05% CI) Wear PPE in the past 1.3 94 (0.54 3 ) * 4.032 (1.3 9 , 1 1 . 62 ) Time working in the USA 0 .083 (0.108) 1.09 (0.88, 1.34 ) Injury in the season 0. 402 (0 .913 ) 0.67 (0.11, 4.0 1 ) Note s : OR = Odds Ratio CI = Confidence Interval * p = .01 As in the bivariate logistic analysis, examination of the parameter estimates for the individual categories showed that workers who used safety grasses in the past (OR=3.783, 95% CI 1.3, 10.98) were more likely to be using glasses at the time of the inter view. The relationship between current and previous experience using PPE, was more evident by the addition of the time the workers have harvested citrus in
193 Florida and their experience with injuries. One possible explanation is that workers with previous e xperience using safety glasses are coming from other industries and have less time harvesting citrus in Florida. Model C tests the closeness centrality measure of Crew C ( n = 31) in relation to the past experience of workers using PPE and their amount of t ime harvesting in Florida. Table 4 17 represents the only model in which a centrality measurement is significantly associated with the usage of safety glasses. All other attempts to link the social network structure to the usage of glasses were not signifi cant. One explanation for these outcomes is the reduced number of cases resulting from the estimation of a logistic regression model for each individual crew. Table 4 17 . Logistic regression C results for use of PPE Season 2010 2011 Parameter B(SE) OR (05% CI) Wear PPE in the past 1. 183 (1.01 3) 3.27 (0.45 , 23 . 77 ) Time working in the USA 0 .242 (0. 179 ) 1.27 (0.90, 1.08 ) Closeness 0. 615 (0 .322 ) * 0. 54 (0. 29, 1 .0 2 ) Notes: OR = Odds Ratio CI = Confidence Interval * p = .05 Contrary to the previous model, th e examination of the parameter estimates for the individual categories showed that past experience of workers using safety glasses do not influence the use of safety glasses as much as the negative effect of closeness centrality measurement (OR = 0.54, 95% CI 0.29, 1.02). Crew C, as noted, has multiple components that enforce closeness centrality measures in a negative way. Figure 4 17, illustrates the effect of this relation: participants high in closeness centrality are also less likely to wear safety gla sses at the time of the interview
194 CHAPTER 5 DISCUSSION : SOCIO CULTURAL ANALYSIS OF RISK Introduction The results of this study point to four major areas of findings: working conditions, worker beliefs, social structure, and methodological contributions. I discuss the effects of each condition in worker strategies for safety and health. Additionally, I explain the benefits of employing a mixed method approach. Working Conditions The first and most important finding is the difference between large industrial harvesting operations and the small and medium harvesting crews. The increased participation of H2A workers in the former is changing the dynamics of labor relations in general, and even of the participation of growers in the safety of harvesters. A more pro safety strategies. When that safety net is not available, workers rely on unstructured verbal and non verbal communication about safety practices that may or not be adopted by other workers. In other words, company involvement helps formalize communication about safety practices through specific channels. The increased involvement of large companies in safety strategies has a dual purpose: on the one hand, it is a cos t benefit strategy that can save companies financial resources through risk reduction and prevention. On the other hand, it helps with public relations in dealing with increasing pressure by advocacy groups about the living conditions of farmworkers. Recal ling that this is primarily the case where there is a high proportion of H2A workers, this increased visibility of company policies also helps deflect local criticism of conservative groups regarding the legal status of farmworkers.
195 Smaller companies are a lso under pressure, but do not implement similar policies there is little enforcement on the part of the state and they are unlikely to face penalties for lack of compliance. Likewise, since both advocac y and conservative groups tend to target the larger, more visible companies, public relations is not a major concern for smaller companies. A second finding regarding working conditions is the same for both small and large operations the pay rate formula. The pay rate formula is tied to production so that workers will often take unnecessary risks in order to pick more fruit, more quickly. H2A workers in large companies take risks to meet the quota and avoid being sent back, and workers in small operations i ncrease their pace which puts them at higher risk, in order to maximize the low payment they receive per bag. It was an often cited t have to be concerned with how many baÃ±os Moreover, although legal status of workers was not in any way the focus of this study, it bears repeating that undocumented citrus workers exist under vulnerable conditions. They are unable t and avoid seeking legal and health care services when needed. The oversight and public pressure on large companies that helps ensure H2A worker access to services should also be extended to sma ller companies. This would help ensure that undocumented workers are guaranteed the same safeguards as guest workers. This is not an unrealistic goal: the value of some undocumented workers does not go unnoticed and it helps them become a part of the forma lized system. Much like athletic scouts and
196 college recruits, t hose who are noticed by large companies as productive harvesters are encouraged by these companies to go back to Mexico so they can be brought back as H2A workers. Why not just recognize the va lue of farmworkers and keep them here as the experts for the industry? Finally, the combination of the financial crisis, the decrease in orange juice consumption, and the spread of greening infections have caused speculation about the collapse of the citru s industry in Florida (Figure 3 58) . For hand picking harvesters, this means increased pressure from the companies (of any size) which, in turn, also increases their risks in the field. For larger operations that can afford it, the rush to maximize product meaning more chemical exposure for farmworkers. Smaller operations, on the other hand, cannot typically afford the investment that these cocktails require. In their case, the increased ris k comes from harvesting in groves that are in decay, exposing workers to more debris in their eyes and increased risk of falls and injuries associated with broken branches. The Economic Context: What is the Cost Benefit Ratio of Safety? A recurrent questio n on the implementation of the PCWH program is the cost benefit associated with it. To consider this for this study, there are at least four distinctive perspectives that apply. They correspond to the political position of each of the entities that form th e PCWH: worker , industry , advocacy , and the academic perspective. From the point of view of the workers, the cost benefit is centered on the immediate effect the use of safety glasses has in their income, and the perspective they may have for the future . For companies, aside from similar concerns over productivity, safety policies are adopted either for corporate responsibility or for financial
197 considerations. For advocacy groups, the human cost is the only principle able to guide a fair and just agricul tural system. Finally, the academic perspective has the decrease premature death, disease, and disability and increase the quality of life for yant et al. 2009:330) in an ethical and sustainable manner. capital. If financial balances are jeopardized by either public relations or economic impact; or the benefits surpass th e cost of injuries more generous and aggressive safety policies could be implemented. The PCWH conducted a productivity test, but the self report of workers and the focus on the perception of workers cannot be compared to the actual production of workers o r the actual use of safety glasses in the field (Luque et al, distinction as the most productive worker among 250 fellow harvesters. 1 Although there is some suggestion that the use o f glasses should not affect productivity (Luque et al. 2007) , there is no study that looks at the specific cost benefit of the worker and the company in terms of productivity. In the same way there is no detailed analysis of the financial cost benefit of p articipant companies considering the savings over injury costs. For this project, a uniquely trained group of CHWs was employed to improve citrus worker health and to reduce health care costs by treating eye injuries in citrus groves or work camps before t hey became more severe or chronic 1 This worker was encouraged by the citrus company to go back to Mexico, so they could bring him back as a H2A workers. He did return, but the State Department denied his visa, because he was cut once crossing the border illegally. The US immigration policy for these skill workers block them any possibility of legally coming here to work independently; a way. to perpetuate social control and cheap labor.
198 problems. In addition, CHWs attempted to improve access to health care by making referrals to community health centers in the event of non work related injuries or conditions. Although there was a particular interest of th e social marketing exploratory research to analyze the economic cost benefit of injuries in the industry; the data needed to provide correct estimates for that has been scarce. Attempts to evaluate the cost benefit of the program combining the CHW reports of activities (glasses delivered and eye washes practiced) with the amount and severity of injuries reported by the company where they work was done in Season 2007 2008, but (un)fortunately, the official report of injuries was so small that any quantitativ e analysis was compromised. Other difficult to quantify factors for a cost benefit analysis of preventing eye injuries are the indirect cost of injuries and the investment return evaluation of implementing safety policies and programs. Companies know that injuries impact labor costs through loss of production and the additional cost of the employers caring for injured workers, but there is no accounting of this cost. Moreover, the economic impact of specific policies and programs is difficult to evaluate. For instance, one regular policy when I joined the program was the practice of compensating crews who did not report injuries in the season: workers received additional economic benefits or pre paid telephone cards and crew leaders received up to 1,000.00 season. Perhaps it is the investment of less than $2,000 a month per crew that decreased the number of reported injuries rather than the systematic implementation of safety practices. It is easy to understand the motivation of cr ew leaders and workers to hide the injuries. I recorded anecdotes of workers who were brought by the crew leader
199 to private doctors either because they found it more beneficial to pay out of pocket for the injury than to lose his monthly bonus, maybe becau se the companies prefer to treat injuries out of sight of adjusters that could increase their insurance cost. Almost all workers surveyed had experienced eye injuries and many confessed to hiding the injury to avoid losing income, receiving the phone card, The current tendency of the large companies who used to promote this policy have stopped the practice in favor of implementing safety programs (e.g. the PCWH and the Farm Labor Supervisor Program or the co production of citrus specific safety videos at IFAS UF 2 ) to reduce the cost of injuries. For instance, three of the PCWH harvesting members maintained different levels of the intervention after our partnership ended. In o ne case, the company fully maintained the CHW system and provided eye safety glasses on demand to help prevent eye injuries, while the other two promoted additional eye safety trainings and have a regular supply of eye wash liquid and eye safety glasses on demand and when workers experience an injury. Unfortunately, the program only reached a small proportion of the industry and the safety policies applied in the industry is in direct relation with the size of the company. The large companies we worked wit h are also engaged in the implementation of the H2A program. The adoption of this model implies more oversights and thus more care for safety practices. Fragmented, spread out, and with little supervision, smaller 2 The University of Florida Institute for Food and Agricultural Sciences (UF/IFAS), initiated the Farm Labor Supervisor (FLS) Training program in 2010 as a voluntary educational program targeting anyone with supervisory or management responsibilities over migrant and seasonal farm workers. The goal of this program is to effectively teach supervisors the rules and regulations regarding: farm workers including the prevention of injuries (Roka et al. 2014)
200 crews are at the mercy of the crew leader or small operator (grower or harvester) willingness to follow workers comp law. My opinion is that older crew leaders in this setting have the tendency to reject preventive safety practices because it has never been part of their operations, while younger crew leaders tend to search for a professionalization of their work. I would expect that if smaller companies adopt safety programs workers would adopt these practices. The problem is that most operators see the education of the workers as adding to the co st of improving worker conditions. For small or large operations the cost of implementing any safety program has to be weighed against the benefits and savings, both in terms of the prevented injuries as well as the savings obtained by reduced liability in surance costs. Some insurance companies are not only aware of the PCWH program but even suggest its implementation to clients who do not count with a safety program (offering up to 10% reduction on deductible, according to a representative of an insurance company). In fact, involving insurance companies may offer a good o pportunity to quantify more accurately the cost benefit of eye injuries , as they likely have algorithms in place to do just that . For instance, how can one properly measure the effect of r egular eye washes in the field in terms of time spent and injuries prevented? Or the number of times safety glasses prevent sand from getting into an eye, or a branch prick? The mathematical models developed for the insurance industry could account for sa ving s in injuries as well as for the proper investment needed . For example, t he usual lifespan of safety glasses when used regularly has not been recorded but participant observation and comments from users suggest that it is less than a month. This would represent a frequent cost, a fact that has moved companies to offer workers glasses
201 only on demand , but that has not been properly quantified in terms of the cost benefit of prevented injuries . Citrus c ompanies are not required by law to enforce safety gla sses usage, and doing so would significantly increase their equipment costs , but perhaps may also save them the costs of injuries . Finally, another deterrent for these companies is that the adoption of safety glasses is a process that takes time ; any mathe matical model has to take the cost of the implementation and full adoption timeline into account . The best formula in terms of investment and return could greatly b enefit workers, yes, but also companies looking to reduces production costs in difficult con ditions. The production cost of citrus has increased not only by the severe illness of groves but for the real shortage of labor markets. Companies who adopt the H2A program invest in different ways to make it solvent. For instance, they avoid the risk of corruption generated by third party involvement ; recruit directly in Mexican citrus productive areas ; and identify local productive workers who can go back to Mexico and come back as guest workers. For some experts (Roka 2014), the economic model of manual harvesting is not viable for the industry . The physical cost of citrus harvesting and the poor socio economic advantages available for workers call into question the current sustainability of the model . In general , the consumer price of food will rise wit h associated labor cost s. The actual costs of inputs for production as well as increased worker demands are likely to affect the end product of this industry in ever growing ways. H ow much is the public willing to pay for a box of orange juice? 3 A lternativ e 3 It i s not only the labor cost of farmwork, which is bad pay to keep cheap fresh and vegetables in the product section of supermarkets, but also the same depress salaries for the rest of the workers in the food chain system what has to be question.
202 s olutions for the production of fresh food , like mechanization 4 or socialized harvesting , described below, must be discussed . While essential work, h arvesting fruits and vegetables is a form of labor that is ignored and even despised in industrialized s ocieties. U nderstanding and appreciating this hard work is tied to proximity to food production. Residents of urbanized metropoles have little direct relation to the land and food production, which tends to happen away from urban poles. For many decades, a nd in many places still, urban jobs are considered preferable to rural ones just as city living is considered more advanced, with more opportunities, and better infrastructure. For Latin Americans in general, hailing the draw toward the city is very stro ng (Hillman ) . Many farmworkers confessed to only be ing in the field while waiting for other opportunities to come along such as construction which, while also being physically demanding pays much be tter than agricultural work, and in the service industry in various capacities. Those who chose, or were force d to stay and make a ca r e er out of agriculture pay a very high price to do so with their bodies through repeated stress and contamination (Holmes 2013) . Part of the issue is the concentration of the burden of agricultural labor on one small segment of the population, the ethnic niche of Mesoamerican males. Possible solution would be to socialize the harvesting labor, approximating consumers to thei r products through contact, and helping divide the burden more equitably. This is not so 4 The idea o f mechanize the citrus harvesting process is not new , it goes back till 1950. However, it was and it reach its pick in Florida in Season 2009 2010 when it surpassed 7 % of the total p roduction of citrus . Nevertheless, the grow n of citrus illnesses and a rebound in the supply of labor has limited the use of this method.
203 far fetched. Prior to highly industrialized agriculture, farming in the U.S. was primarily a family or community enterprise, as is the case with the Pennsylvania Dutch to this day. I n many former communist countries , youth had to work in the fields during the summers , providing the state with cheap and replenishable labor. In the case of the citrus industry, perhaps we should consider ways in which the current model ca n be adapted so as to distribute the labor more broadly so as to become more sustainable. It is clear that the current rotation of young and experienced workers is restricted by the demographic change of Mexico and the limited incentive s offer ed by the job . Is it possible to think of a collective model that will bring awareness to the difficulties of farm work and alleviate the labor shortage? A controversial proposal, no doubt, but a discussion that may be needed in the face of the limited supply of labor, the life expectancy of new generations, and the growing demand for a more equal and healthy food chain system . Workers recognized that the use of safety glasses could provide protection from foreign objects but few wore safety glasses because there preva iled a negative attitude towards the costs associated with the practice. They are afraid of losing their already limited income or, worse, jeopardizing the production requirement to keep working. The recurrent solution suggested by the workers is the imple mentation of an hourly pay rate to overcome the loss of production by the slowing down of harvesting; but that is a marginal productivity. Worker Belie fs When it com es to worker beliefs, company size again becomes a relevant factor. For the most part, small and medium harvesting companies have little involvement in
204 workers strategies for safety and health, leaving workers to rely exclusively on their own strategies of cultural beliefs. In larger companies, safety policies and the presence of CHWs likely standardize the strategies for safety and health, subordinating worker beliefs in terms of relevance for crew practices. It may be that larger companies without safety policies which were not observed will more closely resemble the smaller crews with the wisdom of older and or more experience workers becoming of primary importance. Safety strategies in these cases are linked to the set of cultural beliefs of workers mor e than to labor regulations. As exposed by Campbell et al. (2006) rural masculinities are subordinated to the location in which they are developed. The citrus worker culture is a reality associated with the social origin of the farmworkers. On the one hand , some workers reject the use of safety glasses because being concerned with their physical well being can be seen as a sign of weakness, contradicting their social codes that shape them as macho Mexican men (see Gutmann 2006 for a discussion on Mexican ma sculinity). One endures pain, does not complain, boys do not cry, and so on. For instance, when we were showing different types of goggles, several workers were concerned that certain was also employed as a justification for glasses use, because to wear them implies discomfort. drawbacks and one even composed a glasses themed corrido , a type of Mexican ballad as sociated with macho rural culture. Strongly tied to this concern with fulfilling their masculine roles is the concern with being providers. When workers quite a few confessed to me that they were not
205 planning to keep on working, their main reason was becau se they were no longer able to send money to Mexico under the current situation. A common concern was that ated with less income, there is less incentive to wear them. Another finding is that injuries or lack thereof are dismissed as the result of luck, fate, or the faith of the workers. For many workers there is the sense that an accident is going to happen a nd that the person who suffers such an accident is predestined to be injured. An important material concern is the immediate need to make more money through increased production, a powerful deterrent to full implementation of prevention practices which are more oriented toward the future. As glasses are associated with lower production rates, it may be that the choice to neglect the safety eyewear is justified through a discourse that displaces focuses on this belief in bad luck or the will of the gods that absolves workers of personal responsibility in managing their own risk 5 Finally, another finding was that workers hold a lot of respect for experience. It was clear that older workers, or eve n younger workers with long term experience harvesting citrus in Florida, have a strong reputation among their peers. This can work in a positive or negative ways for eye safety. If the experienced worker dismisses the idea of safety, others will follow an d the same holds for those who prize it. On the other 5 Interestingly, in a study of sex workers in Cancun (Otis and Tovar 1999) that examined safety in sexual practices and self esteem, I found that sel f esteem ratings did not predict the avoidance or engagement in risk behaviors. Rather, the amount of money a client offered was the main factor in sex workers decisions to engage in risk behaviors. These women were often single mothers, so, like the citru s workers, the concern with providing for the family in the present overrode concerns for long term health.
206 hand, institutional hierarchy does not work well when imposing safety practices. Workers distrust managers because they believe that these hold company interests over the well being of workers. Moreover , they feel that safety practices may be another layer of impositions they suffer in their daily routine. The key here is that time and experience, not rank, become important. For this reason, when CHW have good reputation and are convinced of the benefits of safety practices, they are able to convert members of their crew in to PPE users. Social Structure crews plays a small part in the adoption of safety goggles. Originally , I hypothesized that the denser the network, the easier the diffusion of innovations was going to happen. However, the particular results of the social network analysis implemented suggested that the more fragmented the network, the more likely the worker was going to be using PPE. One possible explanation of this phenomenon is that workers in fragmented crews competed even in the adoption of safety glasses; another explanation is that the level of exposure to the use of safety practices by the large numbe r of CHWs trained in that particular crew change the perspectives for adoption. A second finding is related to the way in which SNA discovers cultural influences, such as religion. For instance, there was one central worker who was not a particularly prod uctive harvester; he was brought by the company each year because he had more influence on his co workers in terms fostering good behavior. SNA results show that he kept a central role in the crew network structure, in part for having been at the heart of the Evangelical Christians in the crew. This influence extended, to our dismay, to
207 eyeglasses wear. His age and beliefs prevented him from adopting safety practices, which influenced other workers away from their use as well. Methodological Contribution Fi nding A fourth finding is not directly related to the results of the study but with the unique and effective way in which it was conducted: the systematic and hybrid (Harrison and Reilley 2011) use of mixed methods. It is clear to me that, without the use of this planned model and the point of entry at which they were conceived, some important information would not have emerged. For instance, small crews composed of mainly undocumented workers would not be accessible using survey instruments. In large comp anies, participant observation helped us gain entry in the field and face to face interviews and focus groups helped establish rapport with the workers, important factors to gain worker trust and help ensure greater reliability in the survey phase. Survey construction itself benefitted from the mixed methods approach, without which we would not have learned the specific worker concerns to incorporate into the survey in their own jargon (e.g., have you suffered a ramazo , how many baÃ±os have you fill today, h ow much a baraÃ±udo grove increases your risk of injuries), which was extracted from the observations (Figure 3 1 and 3 2) . Finally, the use of glasses benefited from mixed testing methods by revealing complex realities not accessible by a single approach. Neither the quantitative nor the qualitative data would have sufficed in providing as comprehensive a picture of worker strategies for safety and health as the combined approach yields. The ethnography yielded important information about how harvesters lea rn about and come to adopt safety and health strategies in Florida. Further, ethnographic data helps complement existing knowledge when it comes to comparing large and small
208 operations, as well as citrus juice processing versus fresh fruit production. Whil e the PCWH intervention was mostly focused on citrus operations destined for juice, the ethnographic work was centered on crews picking for the fresh fruit market. While the overall risks associated with harvesting citrus may be quite similar in each setti ng, the specific working conditions of each system influences the prevention of injuries and its consequences. This hybrid methodology has applicable benefits beyond the groves. Advocacy groups and growing numbers of consumers have denounced the poor work ing conditions for farmworkers. The use of mixed methods here directly demonstrates the benefits of corporate involvement for workers, but also help corroborate the prevalent injustices suffered by these workers due to broader structural conditions (e.g., prosecutorial immigration laws), poor regulation enforcement (especially in small crews), and the almost universal piece rate system (that encourages workers to engage in riskier behaviors). rception of risks and serious concern for the economic impact of PPE usage (Luque et al. 2007 ; Table 4 2 ) and significantly increase the number of users (Monaghan 2009 ; Figure 4 5 ), while decreasing the number of eye infections and increasing the access to health care services (Tovar et al. 2014). Mixed methods were essential to the planning, execution, and evaluation of these processes: from the mathematical model used for testing safety glasses performance in the fields using a Greco Latin square design e valuation (Tovar et al. 2014) to the ethnographic experience, the combined information complements our understanding of the industry, its workers, managers, and regulators.
209 The choice to test the effects of social network structures on the adoption of eye safety glasses was logical after regular observations of workers while conducting the quasi experimental evaluation of the CHW program (Monaghan et al. 2009), where it th Another indication toward Social Network Analysis emerged from training these promotores de salud attrib utes and qualities these workers should have, which led me to the idea of selecting workers based on their social network attributes. The implementation of this study disseminati on and adoption of eye safety glasses. The results although not conclusive, show there is potential benefit in using SNA for the selection of CHWs (Table 4 17) . Further testing is needed to establish a clear relation of SNA structure and the use of CHW in the increased use of PPE equipment after and before the selection and training of the promotor . To reach this conclusion the renewed use of mixed methods is recommended. Future Research and Interventions Improving the working conditions of farmworkers i s not a simple and straightforward task. By virtue of the competing interests and multiple stakeholders, seemingly rigid structural challenges, amid shifting politics and policies, it is a very complex and challenging process in constant need of evaluation and strategic intervention. Take, for instance, the increased participation of H2A workers in the citrus industry. How many H2A visas are available depends on intricate political interests of the Department of State and whichever administration is in offi ce. Lawmakers approach
210 immigration in general by balancing the appeals of advocates, industry lobbyists, and their constituents. How these visas then get distributed is another issue. In any case, higher numbers of H2A visa holders in citrus opens up the p ossibility of recruitment criteria and interventions that may improve the implementation of safety practices, like the adoption of eye safety glasses. Our last study (Season 2010 2011) shows a strong asses and the current adoption of PPE equipment (Table 4 10). Changing the political will, however, is a long and winding process that may not yield changes quickly enough if at all. More immediately, working conditions may also be ameliorated by the stra tegic use of the promotor model to help disseminate knowledge and skills for better health and safety. This study suggests that SNA may be used in the process to help select a good promotor that will influence the group. Moreover, SNA also can be utilized to the social structures that could facilitate the adoption of eye safety glasses and their usage maintenance. Who is the more appropriate person to become a good CHW? To insure the success of any CHW program the selection, training, and evaluation of pro motores is key. network structure, work experience, communication skills, and productivity. If possible, all should be taken into consideration when selecting intervention m odels based on CHWs. If the structure of the CHW network is as relevant as I suspect, good promotores should hold a critical position in their network. Central positions have often been equated with opinion leadership or popularity, both of which have been shown to be associated with adoption behaviors (Becker 1970; Rogers 2003; Valente 1995; Valente
211 & Davis 1999 ; Valente and Pumpuang 2007; Valente 2010 ). This can be tested out more directly with the promotor model in other studies to add to this body of li terature. Any discussion of interventions must take into account finances. There are current efforts of the FPRC and one of the harvesting members of the PCWH to refine the cost benefit analysis of implementing the CHWs program. According to , Mark Wade (20 13), safety manager of Evans Properties Inc. in one year alone, crews utilizing a promotor showed reductions in both frequency and financial cost of eye injuries. There was a 54% reduction in the frequency of incidents and a 76% in the costs of injuries re lative to those crews without a CHW . Similarly, Omar Ruiz (2012), safety manager of Statewide Harvesting and Hauling Inc. also reported over 50% reduction in the inciden ce of injuries in the first year of implementation, although it went up again the follo wing year (in this case, several members of the same crew were affected in the eye by a particular chemical in a grove). In light of these findings, Southern Gardens Citrus, the first company that joined the PCWH reported only a single eye injury in Season 2008 2009 and it eventually decided to institutionalize the program and run it on its own, without the support, or logistic of the PCWH. This can made good financial sense, if as the rough estimate of Mr. Ruiz, confirm that the cost of the entire season s afety glasses supplies for a company of 250 harvesters can offset the cost of a single eye injury. This kind of study has implications for employers seeking reduced harvesting costs and a healthier environment for workers. The study and evaluation of poli cies that affect citrus workers is another attractive field to conduct more research and intervention. Does the enforcement of safety policies generate a rejection or the improvement of worker conditions? My initial
212 reaction was to dismiss the practice of enforcement, because I thought that it is not through imposition of rules that workers may change their safety practices but by means of education, economic compensation, and regulatory oversight. Now, I propose that more evaluation is needed to suggest in ternal policies that could benefit workers. How can we use the fact that the workers who were forced to use safety glasses in other jo bs were more likely to wear them again? The total amount of information collected by the PCWH has not been completely ana lyzed. A more comprehensive longitudinal analysis of all the data collected over ten years should provide some tendencies of harvesters and companies. Even the current social network data can enhance our knowledge of workers by using other technics like c reating a two mode network in which workers are tied by the use or safety glasses; or by adding the kind and intensity of alters relationships. Several focus groups of workers, audiotapes of the PCWH community advisory board meetings, and data from a surve y in 2002 should complement the results highlighted in this study, enhance the study of the PCWH structure, and plan for the future. Finally, the study of the citrus industry could not be completed without reaching other components of the production syste m. One stakeholder largely left out of this study is the consumer. What is the role that consumers could play in the improvement of workers conditions and even survival of the industry? How can the market compensate workers for rough labor conditions? What is the future of the industry with the current immigration impasse created by a divided and dysfunctional political system? And lastly, citrus production can be considered through a more global perspective. This could include comparative studies of nation al productions and their
213 groves has a direct relationship to the demand and access of specialized labor. This affects not only Florida but Veracruz, and perhaps even oth er Mexican states that may wish to compete for this market of workers or become niche producers of other types of produce harvesters. What can we learn from the perspectives of workers and the permanent flows of workers? Is there is another way to adopt an d enforce international standards for safety? Can increasing global awareness of poor labor conditions 6 also create new international policies? As is the fact with most commodities nowadays, the particularities of one indust ry in one state cannot be extracted from the broader context of the global flows of capital, products, and workers. There is a lot to think about in one glass of orange juice. 6 For instance, as evidenced by support for garment workers in Bangladesh and sherpas in Nepal .
214 APPENDIX A PILOT TEST SURVEY 2004 Pilot test of Eye Inju ry Survey for Citrus Employees Date ______ Crew ______ Interviewer______ Age of picker _______ 1. In the past week, did you have a day when your eyes were red and irritated? Can you provide details? Probe: a. What caused it? b. Did you report it to the chivero? c. Did you treat it yourself? d. Did anyone give you help? 2. In the past week, did you have basura , dust or chemicals fall in your eyes? Can you provide details? Probe : a. How were you picking? b. Did you report it to the chivero ? c. Did you treat it your self? How? 3. Did you suffer from a ramalazo or trauma in the past week? Can you provide details? Probe: a. How were you picking? b. Did you report it to the chivero? c. Did you treat it yourself? How? 6, How many years have you picked citrus 7. Did you pick for this crew leader or this company last year? (Circle each) 8. Last year, did you wear safety glasses?
215 9. Last year, was there someone on your crew that encouraged you to wear safety glasses? 10. Do you wear safety glasses now? Why or why no t? Demographics 11. What is your age? 12. How many seasons have you picked in Florida? 13. Are your married? 14. How many years did you went to school? 15. Where are you from (which Mexican state)?
216 APPENDIX B FOLLO W UP INTERVIEW FOR SEASON 2006 2008 (SPANISH VERSION) Fe cha ________ Entrevistador ___ __ Using glasses ( Do not ask ) Yes ____ No ____ Cuadrilla ________ Edad ____ Origen __________ 1. Â¿Ha tenido el ojo rojo o irritado algÃºn dÃa de la semana pasada? No ____ (Pase a pregunta 2 ) Si ___ (ContinuÃ© con la siguiente pregunta) a. Â¿QuÃ© lo causo? b. Â¿Lo reporto al chivero? No _____ Si ______ c. Â¿Lo trato por su cuenta? No _____ Si ______ Â¿CÃ³mo? d. Â¿Alguna persona le ayudo? No _____ Si ______ Â¿QuiÃ©n? 2. Â¿Ha tenido basura, polvo, o spray en el ojo algÃºn dÃa de la semana pasada? No ____ (Pase a pregunta 3 ) Si ___ (ContinuÃ© con la siguiente pregunta) a. PodrÃa darnos mÃ¡s detall es, por ejemplo Â¿CÃ³mo estaba piz cando? b. Â¿Lo reporto al chivero? No _____ Si ______ c. Â¿Lo trato por su cuenta? No _____ S i ______ Â¿CÃ³mo? 3. Â¿Ha sufrido un ramalazo o un golpe algÃºn dÃa de la semana pasada? No ____ (Pase a pregunta 4 ) Si ___ (ContinuÃ© con la siguiente pregunta) d. PodrÃa darnos mÃ¡s detall es, por ejemplo Â¿CÃ³mo estaba piz cando? e. Â¿Lo reporto al chivero? No ____ _ Si ______ f. Â¿Lo trato por su cuenta? No _____ Si ______ Â¿CÃ³mo? 4. Â¿CuÃ¡nta s temporadas lleva pizcando naranjas? ___________________
217 5. Â¿En que fecha empezÃ³ a pizcar en esta temporada? __________ 6. Â¿Pi z co con la misma cuadrilla el aÃ±o pasado? Si_______ No ______ Â¿Para quiÃ©n trabajo? ________________ 7. Â¿Uso lentes de seguridad el aÃ±o pasado? No ___ Si _____ 8. Â¿Hubo alguien en la cuadrilla que promoviera el uso de lentes la pasada temporada? 9. Â¿RecibiÃ³ lentes en esta temporada? No ___ Si _____ Â¿Qu iÃ© n se los dio? 10. Â¿Ha utiliz si o porque no? 11. Â¿Conoce al promotor? No ___ Si _____ Â¿ CÃ³ mo se llama? ______________ 12. Â¿El promotor le ha ayudado? No ___ Si _____ Â¿ CÃ³ mo ?
218 APPENDIX C SOCIAL NETWORK QUESTIONS 2010 1. How often do you talk with you co worker __________ (list each member of the crew in here)? a. Every day b. Six days a week c. Five days a week d. Four days a week e. Three days a week f. Two days a week g. One day a week h. 2. Is__________ (l ist each member of the crew in here) a family member, a friend, or acquaintance? 3. Would you ask __________ (list each member of the crew in here) for a favor (like borrowing money or give a message to a person in Mexico) or advice (like how to harvest or h ow to cure an injury)? a. Yes b. No 4. Who do you share house with?
219 LIST OF REFERENCES Arcury, Thomas A. and Sara A Quandt 2009 Latino Farmworkers in the Eastern United States: Health, Safety, and Justice. New York: Springer Verlag. Arcury, Thomas A., Mel inda F. Wiggins, and Sara Quandt 200 9 Conclusions: An Agenda for Farmworker Social Justice in the Eastern United States. In Latino Farmworkers in the Eastern United States: Health, Safety, and Justice. Tomas A. Arcury and Sara A. Quandt, eds. Pp. 221 234. New York: Springer Verlag. Arreola, Luis 2003 2005 Photos. All imagines appeared in here were taken during field work for the Partnership for Citrus Workers Health and were approved for publication by Carol Bryan, Director of the Florida Prevention Rese arch Center. Bates R.P., Morris J.R. and Crandall, P.G. 2001 Principles and practices of small and medium scale fruit juice processing. FAO agricultural service bulletin 146. Rome: FAO publications. Becker, M. H. 1970 Sociometric location and innovati veness: Reformulation and extension of the diffusion model . American Sociological Review 35: 267 282 . Benson, Peter 2008 EL CAMPO: Faciality and Structural Violence in Farm Labor Camps. Cultural Anthropology 23(4): 589 629. Bernard, Russ 2002 Research M ethods in Anthropology: Qualitative and Quantitative Approaches. Walnut Creek: Altamira Press. Bletzer, Keith V. 2004 Open Towns and Manipulated Indebtedness among Agricultural Workers in the New South, American Ethnologist 31(4): 589 629. BLS U.S. De partment of Labor, Bureau of Labor Statistics 2013a Census of Fatal Occupational Injuries, April 25, 2013. http://www.bls.gov/iif/oshwc/cfoi/cftb0259.pdf accessed January 21, 2014 . BLS U.S . Department of Labor, Bureau of Labor Statistics 2013b Economics News Release: Census of Fatal Occupational Injuries Summary. http://www.bls.gov/news.release/cfoi.nr0.htm 2012 .
220 BLS U.S. Depa rtment of Labor, Bureau of Labor Statistics 2007 Incidence rates for nonfatal occupational injuries and illnesses involving days away from work per 10,000 fulltime workers by industry and selected parts of the body [Table R6]. http://www.bls.gov/iif/oshwc/osh/case/ostb1662.pdf , a ccessed October 26, 2013. Bonan n o, Alessandro, Lawrence Bush, William H. Friedland, Lourdes Gouveia, and Enzo Mingione 1994 From Columbus to ConAgra: the Globaliz ation of Agriculture and Food. Lawrence: University Press of Kansas. Borgatti, S teve P., M.G. Everett, and L.C. Freeman 2002. Ucinet 6 for Windows: Software for Social Network Analysis. Harvard: Analytic Technologies. Bowe , John 2007 Nobodies: Moder n American Slave Labor and the Dark Side of the New Global Economy. New York: Random House. Brandy, Sonya S., Margaret M Dolcini, Gary W. Harper, and Lance M. Pollack 2009 Supportive Friendships Moderate the Association between Stressful Life Events and Sexual Risk Taking among African American Adolescents. Health Psychology 28(2): 238 248. Bryant, Carol A., MS Forthofer, Kelli McCormack Brown, D. Landis, Robert McDermott 2000 Community Based Prevention Marketing: The next steps in disseminating behavi or change. American Journal of Health Behavior 24(1), 61 68. Bryant Carol A., Kelly R. McCormack Brown, Robert J. McDermott, Rita D. Debate, Moya L. Alfonso, Julie A. Baldwin, Paul Monaghan, Leah M. Phillips 2009 Community Based Prevention Marketing: A New Framework for Health Promotion Interventions. In Emerging Theories in Health Promotion Practice and Research. Ralph DiClemente, Richard A Crosby, and Michelle C. Kekler, eds. Pp. 331 356. San Francisco: John Wiley & Sons, Inc. Bureau of Labor Statisti cs. 2007 Incidence rates for nonfatal occupational injuries and illnesses involving days away from work per 10,000 fulltime workers by industry and selected parts of the body [Table R6]. h ttp://www.bls.gov/iif/oshwc/osh/case/ostb1662.pdf , accessed October 26, 2013. Burt, R. S. 1987 Social Contagion and Innovation: Cohesion versus Structural Equivalence. American Journal of Sociology 92: 1287 1335.
221 Bustamante, Jorge A. 1977 Undocumente d Immigration from Mexico: Research Report. International Migration Review 11(2):149 177 Calavita, Kitty 2010 Inside the State: The Bracero Program, Immigration and the I.N.S. New Orleans: Quid Pro Books. Campbell Hugh, Michael Mayerfeld Bell, and Marga ret Finney 2006 Masculinity and Rural Life: An Introduction. In Country Boys: Masculinity and Rural Life, Campbell H, M Mayerfeld Bell, and M Finney (edit). University Park: The Pennsylvania State University. Carrington , Peter J., John Scott, Stanley Was serman 2005 Models and Methods in Social Network Analysis. New York: Cambridge University Press. Cartwright, Dorwin 1965 Influence, Leadership, Control. In Handbook of Organizations. James G. March, eds. Pp. 1 47. Chicago: Rand McNally. Center for Dis ease Control and Prevention 2013 About the Prevention Research Center. http://www.cdc.gov/prc/about prc program/index.htm , accessed October 26, 2013 Center for Diseases Control and Preven tion 2014 Agricultural Safety. CDC NIOSH Workplace Safety and Health Topics. http://www.cdc.gov/niosh/topics/aginjury/ accessed January 21, 2014. Chavez, Leo R. 1988 Shadowed Lives: Undocument ed Immigrants in American Society. Stamford: Cengage Learning. Continental Trading S .A. de C.V. 2011 Naranjeros tienen buen aÃ±o; 2012 serÃ¡ malo [Citrus producers have good year; 2012 is going to be bad]. http://www.continentaltraiding.com.mx/productos/naranjeros tiene buen ano 2012 sera malo , accessed February 5, 2014. Creswell, John W. and Vicki L. Plano Clark 2007 Designing and Conducting Mixed Me thods Research. Thousand Oaks: Sage Publications. Davis , S. M. And J. C. Ogden 1994 The Everglades: The ecosystem and its restoration. Delray Beach, St. Lucie Press.
222 Daugbjerg, Carsten and Alan Swinbank 200 8 Curbing Agricultural Exceptionalism: The EU The World Economy 31(5): 631 652. De Genova Nicholas Anthropology 31: 419 447. De Genova, Nicholas and Ana Yolanda Ramos Zayas 2003 Latino Crossings: Mexicans, Puerto Ricans, and the Politics of Race and Citizenship . New York: Routledg e. De Genova, Nicholas 185. De Genova, Nicholas Durham: Duke University Press Books. De la PeÃ±a, Antonio Politics of Participation and Organization in a Rural Ecuadorian Community. Ph.D. dissertation, Department of Anthropology, University of Florida. DEO Florida Department of Economic Opportunity 2011, Agricultural Services Outreach Plan. http://www.floridajobs.org/PDG/Memos/Ag_Services_Plan_2011_Draft.pdf , accessed February 3, 2 014. DePalma , Anthony 1996 How a Tortilla Empire Was Built in Favoritism. New York Times, February 15: A1, A12. Diehm, Jan /Huffington Post 2013 Why Oranges Have Gotten so Expensive: A Huffington Post Graphic in the note of Joe Satran Citrus Greening Fo rces Florida Growers to Trust a Controversial Savior posted on the Huffington Post digital version. http://www.huffingtonpost.com/2013/08/30/citrus greening_n_3780984.ht ml , accessed March 21, 2014. Estabrook, Barry 2011 Tomatoland: How Modern Industrial Agriculture Destroyed Our Most Alluring Fruit. Kansas City: Andrew McMeel Publishing, LCC.
223 FAO 2006 World Reference Base for Soil Resources 2006. Rome: FAO publicatio ns. FAO 2012 Citrus Fruit Fresh and Processed: Annual Report. http://www.fao.org/fileadmin/templates/est/COMM_MARKETS_MONITORING/ Citr us/Documents/CITRUS_BULLETIN_2012.pdf , accessed February 3, 2014 Fairlie. Margaret Carrick 1935 History of Florida. Kingsport: Privately printed. Florida Center for Institutional Technology, University of South Florida. 2008 Florida waterways: I ndian River County. http://fcit.usf.edu/florida/maps/pages/11200/f11243/f11243.htm , accessed March 19, 2014. Florida Center for Instructional Technology , University of South Fl orida. 2008 Florida County Maps. Florida Waterways: Indian River County. http://fcit.usf.edu/florida/maps/pages/11200/f11243/f11243.htm , accessed April 5, 2014. Florida Cente r for Instructional Technology , University of South Florida. 2009 Florida County Maps: Indian River County road network. http://fcit.usf.edu/florida/maps/pages/12200/f12277/f122 77.htm , accessed March 19, 2014 . Florida Department of Citrus (FDOC) 2010 Florida Grapefruit: About FDOC. http://www.gofloridagrapefruit.com/about fdoc/ , accessed February 3, 2014. Florida Department of Agriculture and Consumer Services 2011a Florida Agriculture Overview and Statistics. http://www.freshfromflorida.com/Divisions Offices/Marketing and Development/Education/For Researchers/Florida Agriculture Overview and Statistics , accessed February 3, 2014. Florida Department of Agriculture and Consumer Services 2011b Florida Ci trus Statistics 2009 2010. http://www.nass.usda.gov/Statistics_by_State/Florida/Publications/Citrus/fcs/2009 10/fcs0910.pdf , accessed February 3, 2014. Florida Lottery 2014 Scratch Offs, $2 Games, 1158 LoteriaÂ®. http://www.flalottery.com/scratch offs.do ; accessed February 21, 2014.
224 Goldman, Lynn R. 2007 Managing Pesticide Chronic Heal th Risks: U.S. policies. Journal of Agromedicine 12(1): 67 75. Gonzalez, Gilbert 2007 Guest Workers or Colonized Labor?: Mexican Labor Migration to the United States. Boulder: Paradigm Publishers. Grunwal d , Sabine 2002 Inventory of Physico Chemical Soi l Properties and Spatial Pattern Analyses in the Everglades Ecosystem. http://soils.ifas.ufl.edu/faculty/grunwald/research/projects/NRC_2001/STATSGO.gif , acce ssed February 3, 2014. Guryan, Jonathan, Brian Jacob, Eric Klopfer, and Jennifer Groff. 2008 Using Technology to Explore Social Networks and Mechanics Underlying Peer Effects in Classrooms. Developmental Psychology 47(2): 350 358. Gutmann, Matthew C. 2 006 The Meaning of Macho: Being a Man in Mexico City (Men and Masculinity). Berkley: University of California Press Hanneman , Robert A. and Mark Riddle 2005 Introduction to Social Network methods. Riverside: University of California Harris, Marvin 198 5 Good to Eat: Riddles of Food and Culture, New York: Simon and Schuster. Harrison, Robert L., and Timothy M. Reilly 2011 Mix Method Designs in Marketing Research. Qualitative Market Research 14(1): 7 26. Harwell, Mark A., John F. Long, Ann M. Batursk a, John H. Gentile, Christine C. Harwell, Victoria Myers, and John Ogden 1996 Ecosystem Management to Achieve Ecological Sustainability: The Case of South Florida. Environmental 20(4): 497 521. Hillman , Richard S. and Thomas J. 2011 Understa nding Contemporary Latin America. Bolder: Lynne Rienner Publishers. Holmes , Seth 2013 Fresh Fruit, Broken Bodies. Berkley: University of California Press. IBM Corp. 2012 IBM SPSS Statistics for Windows, Version 21.0. Armonk: IBM Corp.
225 Info Rural 201 2 Principales Estados Productores de Citicos en Citricos, produccion nacional. http://www.inforural.com.mx/spip.php?article100389 , accessed March 29, 2014. Ingebritsen, S.E., Christopher McVoy, B. Glaz, Winifred Park 2000 Florida Everglades: Subsidence threatens agriculture and complicate ecosystem restoration. The U.S. Geological Survey Circular 1182. http://pubs.usgs.gov/circ/circ1182/ , accessed March 11, 2014. Jackson, Peter, with Philip Crang and Claire Dwyer 2004 Transnational Spaces. New York: Routledge . Johnsto n, Sidney J. 2000 A History of Indian River County. Vero Beach: The Indian River Historical Society. Kandula N amratha R . , M Kersey, N Lurie 2004 Assuring the Health of Immigrants: What the Leading Health Indicators Tell Us. Annual Review of Public Health. 25:357 76. Kapaun, Therese 2013 LREC Fukumoto navel harvest. http://ucanr.edu/blogs/LindcoveNews/ ; accessed February 20, 2014. Knight, Robert J. Jr., Mary Lambert, Janet S. Bunch 1984 World and local importance of some tropical fruit crops grown in Florida. Proceedings of the Florida State Horticulture Society 97:351 354. Lacey SE, LS Forst, RE Petrea, LM Conroy 2007 Eye injury in migrant farm workers and suggested hazard controls. Journal of Agricultural Safety and Health 13(3):259 74. Lee , Chang W. /New York Times 2010 Deal sets back Everglades restoratio n. http://www.nytimes.com/slideshow/2010/03/07/us/20100307 everglades slideshow_index 7.html , accessed March 19, 2014. Levy Paluck , Elizabeth 2011 P eer pressure against prejudice: A high school field experiment examining social network change. Journal of Experimental Social Psychology 47: 350 358. Lewis, W. Arthur 1954 Economic Development with Unlimited Supplies of Labor. Manchester School of Econo mic and Social Studies 22:139 191. Lippitt, Ronald, Norman Polansky , and Sidney Rosen 1952 The Dynamics of Power. Human Relations 5:37 64
226 Luna, Guadalupe T. 1997 1998 An Infinite Distance?: Agricultural Exceptionalism and Agricultural Labor. Universit y of Pennsylvania Journal of Labor and Employment Law 1: 487 510. Laumann, Edward O. and Peter V. Marsden 1979 The Analysis of Oppositional Structures in Political Elites: Identifying Collective Actors. American Sociological Review 44: 713 732. Lind, Ja s on 2008 2011 Photos. All im agines appeared in here were taken during field work for the Partnership for Citrus Workers Health and were approved for publication by Carol Bryan, Director of the Florida Prevention Research Center. Luque John S., Paul F. Mo naghan, Ricardo B. Contreras, Euna August, JA Baldwin, Carol A. Bryant, Robert J. McDermott 2007 Implementation Evaluation of a Culturally Competent Eye Injury Prevention Program for Citrus Workers in a Florida Migrant Community. Progress in Community He alth Partnership Research and Education Action 1(4):359 369. Manning, Richard 2004 Against the Grain: How Agriculture has Hijacked Civilization. New York: North Point Press. Macro Plastics Inc. 2014 Safe Handing and Practices: Bin Loading on Trailers. http://www.macroplastics.com/wp content/uploads/2013/04/Florida Tie Down 1.jpg ; accessed February 27, 2014. March, James G. 1955 An Introduction to the Theory and Measurement of Influence. American Political Science Review 49:431 451 Marsden, Peter V. 2005 Recent Developments in Network Measurement. In Models and Methods in Social Network Analysis. Peter Carrington, John Scott, and Stanley Wasserman, eds. Pp. 8 30. New York: Cambridge University Press Marsden, Peter V. 2002 Egocentric and Sociocentric Measures of Network Centrality. Social Networks 24(4): 407 422. Marsden, Peter V. and Noah E. Friedkin 1993 Network Studies of Social Influence. Sociologica l Methods and Research 22(1):127 151
227 MartÃnez, JosÃ© 2012 Las EnseÃ±anzas del Profesor: Grupo Atlacomulco. De Hank Gonzalez a PeÃ±a Nieto [The Professor Lessons: Atlacomulco Group. From Hank Gonzalez to PeÃ±a Nieto]. Mexico City: Oceano. Mashburn, Diane, Pa ul Monaghan, Amy Harder, Glenn Israel, Tracy Irani 2009 Lesson Learned f rom a Camp Health Aide Safety Program for Farm Workers. Paper presented at the 25 th Annual Meeting of the Association for International Agricultural and Extension Education, San Juan de Puerto Rico, May 24 27. Massey, Douglas S., Joaquin Arango, Graeme Hugo, Ali Kouaouci, Adela Pellegrino, and J. Edward Taylor 1993 Theories of International Migration: A Review and Appraisal. Population and Development Review 19(3):431 466 McCarty, C hirstopher 2002 Structure in Personal Networks. Journal of Social Structure 3(1). Electronic Journal. http://www.cmu.edu/joss/content/articles/volume3/McCarty.html , accessed Apr il 15, 2014. McLaughlin, Eithne and John Baker 2007 Equality, Social Justice and Social Welfare: A Road Map to the New Egalitarianisms. Social Policy and Society 6(1): 53 68. Microsoft 2013 Microsoft Excel. Redmond: Microsoft Miller, Steve 2010 Port rait of a picker: in his own words, a local farmworker tell a story of his life and work. Santa Maria Sun web edition July 6, 2010 11(17). http://www.santamariasun.com/cover/4862/ portrait of a picker/ ; accessed February 27, 2014. Monaghan, Paul 2007 Photo . All imagines appeared in here were taken during field work for the Partnership for Citrus Workers Health and were approved for publication by Carol Bryan, Director of the Fl orida Prevention Research Center. Monaghan, Paul 2008 Using Cultural Consensus Analysis to Improve a Safety Campaign for Farmworkers Practical Anthropology 30(1): Monaghan, Paul 2011 Lessons Learned from a Community Coalition with Diverse Stake Holders : The Partnership for Citrus Workers Health.
228 Monaghan Paul F., Carol A. Bryant, JA Baldwin, Y Zhu, B Ibrahimou, Jason Lind, Ricardo B. Contreras, J. Antonio Tovar Aguilar, Tirso Moreno, Robert J. McDermott 2009 Using community based prevention marketin g to improve farm worker safety. Social Marketing Quarterly 14:71 87. Monaghan Paul F, Bryant CA, McDermott RJ, Forst LS, Luque JS, & Contreras RB. 2011a Adoption of Safety Eyewear among Citrus Harvesters in Rural Florida. J Immigrant Minority Health. 1 4:460 466. Monaghan Paul F, Linda S Forst, J. Antonio Tovar Aguilar, Carol A. Bryant, Glenn D Israel, Sebastian Galindo Gonzalez, Zac Thompson, Y Zhu, Robert J. McDermott 2011b Preventing Eye Injury among Citrus Harvesters: The Community Health Worker M odel. American Journal of Public Health. 101(12):2269 2274 Moody, James 2001 Peer Influence Groups: Identifying Dense Clusters in Large Networks. Social Networks 26(1): 261 283 Morera, Maria, Paul Monaghan, J. Antonio Tovar Aguilar, Sebastian Galindo Go nzalez, Fritz Roka, and Cesar Asuaje 2014 Improving Health and Safety Conditions in Agriculture through Professional Training of Florida Farm Labor Supervisors . Journal of Agromedicine 19(2) . Otis, Martha and Antonio Tovar 1999 , consumismo y violencia. Revista Mexicana del Caribe 4(7):153 171 OUSTR Office of the United States Trade Representative: Executive Office of the President 2012 Mexico. http://www.us tr.gov/countries regions/americas/mexico , accessed February 5, 2014 Piore, Michael J. 1979 Birds of Passage: Migrant Labor in Industrial Societies. Cambridge: Cambridge University Press. Passel, Jeffrey and Paul Taylor : Pew Hispanic Center. Pletcher, David 1958 Rails, Mines and Progress: Seven American Promoters in Mexico, 1867 1911. Ithaca: Cornell University Press. Portes , Alejandro 1978 Introduction: Toward a Structural Analysis of Illegal (undocumented) immigrati on. International Migration Review 12(4): 469 484.
229 Portes, Alejandro 2003 Conclusion: Theoretical Convergencies and Empiral Evidence in the Study of Immigrant Transnationalism. Internatio nal Migration Review 37(3):874 892 Prell, Christina 2012 Social N etw ork Analysis: History, Theory, and Methodology. Thousand Oaks: Sage Publication. Quandt S ara A, R Elmore , T Arcury, D Norton 2001 Eye symptoms and use of eye protection among seasonal and migrant farmworkers. South Med ical J ournal 94:603 607. Quandt Sara A., SR Feldman, QM Vallejos, MR Schulz, A Verma, AB Fleischer, TA Arcury 2008 Vision Problems, Eye Care History, and Ocular Protection among Migrant Farmworkers. Arch Environ Occupational Health 63:13 16. Ranchocarballomercado 2009 Veracruz, Mex ico, Panoramio Google Maps. http://www.panoramio.com/photo/30408124 , accessed February 24, 2014. Rawling, Craig M. and Daniel A. McFarland 2011 Influence Flows in the Academy: Using Affiliation Netw orks to Assess Peer Effects among Researchers. Social Science Research 40(3): 1001 1007. Roka, Fritz 2013 Outlook on Agricultural Labor in Florida, 2012 13. UF/IFAS, Southwest Research & Education Center, Immokalee, FL. http://www.fred.ifas.ufl.edu/pdf/webinar/labor_roka.pdf , accessed October 26, 2013. Roka, Fritz, Cesar Asuaje, Paul Monaghan, Maria Morera, Sebastian Galindo and J. Antonio Tovar In print. Developing a Mode l for Ext ension Based Training o f Farm Labor Supervisors . Journal of Extension. Rogers, Everett M. 2003 Diffusion of Innovation. New York: Free Press. Ruiz, Omar 2012 Personal communication. Runyan, Jack 2001 Federal Laws and Regulations Affecting Farm Safety . http://nasdonline.org/document/1842/d001777/federal laws and regulations affecting farm safety.html , accessed February 5, 2014.
230 Satran , J oe 2013 Citrus Greening Forces Florida Growers to Trust a Controversial Savior posted on the Huffington Post digital version. http://www.huffingtonpost.com/2013/08/30/c itrus greening_n_3780984.html , accessed March 21, 2014. Schank, Roger and Robert Abelson 1977 Scripts Plans Goals and Understanding. Hillsdale: Lawrence Erlbaum Associates. Scientific Software Development 2012 Atlas.ti. Germany: Scientific Software Development. Scott, John 2000 Social Networks Analysis. Thousand Oaks: Sage Publications. Skogstad, Grace 1998 Ideas, Paradigms and Institutions: Agricultural Exceptionalism in the European Union and the United States. Governance: 11(4): 463 490. Slad ky , Lynne / A ssociated P ress 2007 Photo from the Note: Farmworkers Say Contractors Cheat Them: Legal Guest Workers Abused, Vulnerable to Middlemen, Advocates Say. Print by NBC 6/1/2007. http://msnbcmedia.msn.com/i/msnbc/Components/ArtAndPhoto Fronts/COVER/080518/g 080518 cover immig2 12p.jpg , accessed February 4, 2014. Short, Nicholas M. Ecosystem. https://www.fas.org/irp/imint/docs/rst/Sect3/Sect3_8.html accessed February 3, 2014. SIAP SAGARPA 2012 Anuario Estadistico de la Produccion Agricola por Estado. Most c urrent data available was 2012. http://www.siap.gob.mx/cierre de la produccion agricola por estado/ , accessed March 28, 2014. Simon, Georg 1957 Models of Man. New York: Wiley. Simon, Julian L. 1999 The Economic Consequences of Immigration. Ann Arbor: The University of Michigan Press. Sjaastad, Larry A. 1962 The Costs and Returns of Human Migration. Journal of Development Economics. 70S:80 93.
231 Smith, Michael P. and Matt Bakker 2008 Citizenship across Borders: the Political Transnationalism of el Migrante . Ithaca: Cornell University Print. Sorensen, Julie A., John J. May, Kris Paap, Mark A. Purschwitz, and Maria Emmelin 2008 Encouraging Farmers to Retrofit Tractor s: a Qualitative Analysis of Risk Perceptions among a Group of High risk Farmers in New York. Journal of Agricu ltural Safety and Health 14(1): 105 117. South Florida Water Management District 1995 Land Usage and Coverage. http://fcelter.fiu.edu/about_us/everglades/restoration/ , accessed January 16, 20014. Springer, Kristin W. and Dawne M. Mouzon Soc ial Class. Journal of Health and Social Behavior 52(2):212 227. Stark, Oded and David E. Bloom 1985 The New Economics of Labor Migration. American Economic Review 75:173 178 Steinberg, Laurence and Kathryn C. Monahan 2006 Age Differences in Resistance to Peer Influence. Developmental Psychology 43(3): 1531 1543. Storper, Michael and Allen J. Scott 2009 Rethinking human capital, creativity, and urban growth . Journal of Economic Geography 9(2):147 167. Strathmore Ladders 2014 Fruit Picking Supplies. http://strathmoreladder.com/fruitpickingsupplies.html ; accessed February 20, 2014. Tanton, John, Denis McCormack, and Joseph Wayne Smith 1996 Immigration and the Social Contract of West ern Societies. Burlington: Avebury. Taylor SL, ML Coates, Q Vallejo, SR Feldman, MR Schulz, S ara A Quandt, AB Fleischer, T homas A Arcury 2006 Pterygium among Latino migrant farmworkers in North Carolina. Archives of Environ Occupational Health 61:27 32. Torado, Michael P. 1969 A Model of Labor Migration and the Urban Unemployment in less developed Countries. The American Economic Review 59:138 148
232 Todaro , Michael P. and Lydia Maruszko 1987 Illegal Immigration and US Migration Reform: A Conceptual Framework. Population and Development Review 13:101 135. Tovar, Jose Antonio 2007 2013 Photos. All imagines appeared as Figures were taken during field work or while working by the Partnership for Citrus Workers Health and were approved for publication b y Carol Bryant, Florida Prevention Research Center Director . Tovar Aguilar, J. Antonio, Paul Monaghan, Carol Bryant, Mark Wade, Andrew Esposito, Omar Ruiz, and Robert McDermott 2014 Improving Eye Safety in Citrus Harvest Crews through the Use of Persona l Protective Equipment, Community based Participatory Research, Social Marketing, and Community Health Workers . Journal of Agromedicine 19(2) . Tucker, D.P.H., S.H. Futch, F.G. Gmitter, and M.C. Kesinger. 1998 Florida Citrus Varieties. SP 102 University o f Florida Institute of Food and Agricultural Science http://citrusagents.ifas.ufl.edu/agents/futch/PDF/Fla%20Citrus%20Varieties.pdf ; accessed March 27, 2014. Turrent FernÃ¡ndez, Antonio, Timothy A. Wise, and Elise Garvey http://www.ase.tufts.edu/gdae/Pubs/wp/12 03TurrentMexMaize.pdf , accessed February 5, 2014. USCB 2014 State and County Quick Facts: Fellsmere (city), Florida. http://quickfacts.census.gov/qfd/states/12/1222100.html , accessed February 4, 2014. US Departme nt of Agriculture 2012 2010 2011 Citrus Summary Production, Price and Value Production by County and Per Tree. http://www.nass.usda.gov/Statistics_b y_State/Florida/Publications/Citrus/cspre/cit9 2211.pdf , accessed February 3, 2014. US Department of Agriculture National Agricultural Statistics 2011 Florida Citrus Statistics 2009 2010: Florida Agriculture Statistics Services. Maitland, USDA NASS Fl orida Field Office Valente, Thomas W. 2010 Social Networks and Health: Models, Methods, and Applications. New York: Oxford University Press
233 Valente, Thomas W. 1999 Network Models of the Diffusion of Innovation . Cresskill: Hampton Press. Valente, Thom as W. 1996 The Diffusion Network Game. Connections 19(2):30 37 Valente, Thomas W. Kathryn Coronges, Cynthia Lakon, and Elizabeth Costenbader 2013 How Correlated Are Network Centrality Measures? Connect 28(1):16 26. Valente Thomas W. and Patchareeeya Pu mpunang 2007 Identifying Opinion Leaders to Promote Health Behavior Change. Health Education and Behavior 34(6):881 896. Valente, Thomas W. and Rebecca L. Davis 1999 Accelerating the Diffusion of Innovation Using Opinion Leaders. The Annals of the Ameri can Academy of Political and Social Sciences 566:55 67. Vernez, Georges and Kevin F. McCarthy 1996 The Cost of Immigration to Taxpayers. Washington: RAND Corporation. Villarejo D. and SL Baron 1999 The occupational health status of hired farmworkers. Oc cupational Medicine State of the Art Reviews 14:613 635. Wade, Mark 2013 Email communication Wallerstein, Immanuel European World Economy in the Sixteen Century. New York: Aca demic Press. Wasserman , Stanley and Katherine Faust 1994 Social network analysis: Methods and Applications. New York: Cambridge University Press. Wikipedia Commons 2007, Indian River County Incorporated and Unincorporated areas, Fellsmere Highlighted u sing USCB data . http://en.wikipedia.org/wiki/File:Indian_River_County_Florida_Incorporated_a nd_Unincorporated_areas _Fellsmere_Highlighted.svg , accessed March 19, 2014. Yang, Cui, Carl Latkin, Rongsheng Luan, and Kenrad Nelson. 2009 Per Norms and Consistent Condom Use with Female Sex Workers among Male Clientes in the Schuan Province, China. Social Science in Medicin e 71(4): 832 839.
234 Yunez Naude , Antonio 2003 The Dismantling of CONASUPO, a Mexican State Trader in Agriculture. The World Economy 26(1): 97 122. Zamora, Guillermo 1997 Caso CONASUPO : La Leche Radioactiva : El Crimen MÃ¡s Atroz contra el Pueblo Mexicano [ The CONASUPO Case: Radioactive Milk: The Worse Crime Again the Mexican People]. Mexico City: Planeta. Zentella, A.C. 1997 Growing up Bilingual: Puerto Rican Children in New York. Blackwell Publishers. Zohar , Dov and Gil Luria 2004 Climate as a Social Cognitive Construction of Supervisory Safety Practices: Scripts as Proxi of Behavior Patterns . Journal of Applied Psychology 89(2): 322 333.
235 BIOGRAPHICAL SKETCH Jose Antonio Tovar Aguilar has an MA in anthropology from UF. He holds a BA in Philosophy from the Universidad de Guanajuato , where worked as a journalist. For 10 years, he covered political and cultural events in Mexico, Central America, and the Caribbean, working for radio, press, and cable. Since enrolling at UF, he has worked on several researc h projects concentrating on health among Latino communities in the United States . In Florida, he has worked for College of Nursing, Bureau of Economic and Business Research College of Medicine and College of Public Health . In California, Illinois, New York, and Texas he has worked for the healthcare branch of Harris Interactive evaluating services and wellness plans. Over the last 7 years he has worked at the FWAF conducting research on the impact of agricultural working condit collaboration with the Nell Hodgson Woodruff School of Nursing at Emory University; on the socioeconomic effects of non timber forest production on rural communities in association with the Department of G eosciences at Florida Atlantic University; on the economic impact of natural disasters on Florida farmworkers with participation of Samuel Proctor Oral History Program; and, currently, on the global organizational strategies of citrus harvesters with La Via Campesina in Brazil, Mexico, and the United assessing the nutritional impact of displace children in Tegucigalpa, Honduras post hurricane Mitch.