|UFDC Home||myUFDC Home | Help|
This item has the following downloads:
1 FOOD SAFETY STANDARD S IN TOMATOES AND TH EIR EFFECTS ON RISK MITIGATION FOR GROWE RS By GABRIELLE ALEXANDRA FERRO A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQU IREMENTS FOR THE DEGREE OF MASTER OF SCIENCE UNIVERSITY OF FLORIDA 2010
2 2010 Gabrielle Alexandra Ferro
3 T o my mom, dad, grandparents and sister, Dominique. Without their support I would not have accomplished all that I have already accompl ished in life. I would also like to dedicate this to Brian, who for the last two years has stuck beside me through thick and thin. His unwavering support has helped me to accomplish this and so much more.
4 ACKNOWLEDGEMENTS I would like to acknowledge my committee both Dr. John VanS ickle and Dr. Mark Brown for the help they have provided me and the flexibility they have both given me in writing this. Without the support and funding of Dr. VanS ickle this would have been a much more difficult task for me to accomplish. I would like to acknowledge, Reggie Brown from the Florida Tomato Committee for assisting in distributing my survey to the tomato growers so that I could begin my data collection. I would also like to acknowledge the UF/IFAS Food and Resourc e Economics Department, and especially Jessica Herman, for both their guidance and support throughout my c ourse of study.
5 TABLE OF CONTENTS page ACKNOWLEDGEMENTS ................................................................................................... 4 LIST OF TABLES ................................................................................................................ 6 LIST OF FIGURES .............................................................................................................. 7 ABSTRACT .......................................................................................................................... 8 CHAPTER 1 INTRODUCTION .......................................................................................................... 9 2 LITERATURE REVIEW .............................................................................................. 12 3 METHODOLOGY ....................................................................................................... 22 Purpose ....................................................................................................................... 22 Regression Analysis ................................................................................................... 22 VOI Analysis ............................................................................................................... 25 4 RESULTS .................................................................................................................... 28 Flexibility Analysis ....................................................................................................... 30 VOI Analysis ............................................................................................................... 32 5 SUMMARY AND CONCLUSIONS ............................................................................. 38 APPENDIX A INFORMED CONSENT .............................................................................................. 40 B SURVEY ..................................................................................................................... 42 C SAS OUTPUT ............................................................................................................. 47 D SAS INPUT ................................................................................................................. 53 E FLEXIBILITY ............................................................................................................... 54 F PROBABILITY DATA ................................................................................................. 55 LIST OF REFERENCES ................................................................................................... 57 BIOGRAPHICAL SKETCH ................................................................................................ 60
6 LIST OF TABLES Table page 4 -1 OLS Regression Estimates of the Model (3-1) ...................................................... 37 4 -2 Estimates of Model (31) Corrected for First Order Autocorrelated ..................... 37
7 LIST OF FIGURES Figure page 3 -1 VOI Chart ................................................................................................................ 27
8 Abstract of Thesis Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Master of Science FOOD SAFETY STANDARDS IN TOMATOES AND THEIR EFFECTS ON RISK MITIGATION FOR GROWERS By Gabrielle Alexandra Ferro May 2010 Chair: John J. Vansickle Cochair: Mark Brown Major: Food and Res ource Economics There is limited research done on the impacts of risk mitigation practices that are being implemented with food safety standards in specific produce industries. Extensive research has been done on cattle and some leafy green produce, but very few topics touch on the issues with tomato production. There has been a significant decrease in the probability of contracting a food borne illness from Florida tomatoes, mainly due to the enforcement of food safety standards. Yet it remains unknown what the cost of a food safety incident i n the market is. This research address es both of these concerns utilizing regression analysis to quantify the impact of an incidence on Florida tomatoes and a v alue of information analysis to determine the impact of food safety measures to mitigate these impacts There is evidence that an incidence o f s almonella results in a significant price decrease. In addition, the findings of the present research indicate that there has been a significant dec rease in the probability of contracting a food borne illness from Florida tomatoes. There are numerous opportunities for further research on this topic that could be done with larger data set s
9 CHAPTER 1 INTRODUCTION Beginning in the e arly 1900 s food safety becam e a n issue that the government attempted to tackle. With the advent of modern transportation and the means to send goods quickly across borders there is an even stronger need for regulations. In 2007 the state of Florida implemented their own set of food safety guidelines in respo nse to the multi -state salmonella outbreak i n July of that year. Many of the growers in Florida utilize various audit companies to ensure that their product will be accepted by the packing houses and suppliers. This is in respo nse to a push by the supplier to maintain strict tracking and credibility standards in the produce market to limit their risk of food safety incidents. M aintain ing standards that Florida wants for growers and the additional costs of private audits increas es the cost to produce tomatoes. This cost is ultimately pass ed on to the consumer, and assuming that the consumer can get cheaper produce from another source, there is no incentive to maintain these standards. Starting in 1998 food safety became a larg e concern for the tomato industry because 14 out of the 78 outbreaks were linked to fresh tomatoes from 1998 through 2008. It was during 1998 the federal government issued the GAPs guide, which provides general guidelines to producers of fresh produce. I n 2004 the Food and Drug Administration (FDA) issued a Produce Safety Action Plan (Action Plan) that had four general goals associated with it. They intended to (1) prevent contamination in fresh produce, (2) limit the impact of contamination to the publi c, (3) facilitate better communication and (4) improve relevant research on food contamination. Another facet of this Action Plan was to create a multi year tomato safety initiative which began in
10 2007. This program was in coordination between the states of Virginia and Florida, the FDA and several educational institutes. It is through this research and funding that the commodity specific guidelines for tomatoes have originated (Guidance 2009). G rowers are concerned that there is no risk mitigation w ith these food safety standards, and that there is not any improvement before and after implementation of the standards In 2007 mandatory food safety audits were enforced in the Florida tomato industry. It remains to be seen if these standards have caus ed an improvement in food safety. The standards ultimately provide an indication that the product is safe, the guidelines are followed and the food safety outbreak is not or iginating at their farm. Yet recent studies done by Arnade (2009) and Lloyd (2001) suggest that even if a grower is not responsible it will affect them regardless. In the case of the spinach E.C oli outbreak, one year after the incidence there was still evidence of a decrease in demand that can be attributed to the food safety outbreak This research looks at how food safety standards have impacted producers in Florida By utilizing Value of Information (VOI) analysis it can be determined if there is a difference between growers that utilize private audit firms and growers that only utilize Florida Department of Agriculture and Consumer Services (FDACS) audits. This data was not available during this study and therefore the VOI provides an analysis of what will occur when a food safety incident occurs. Additional research was done t o determine what the price impact on tomatoes due to food borne illnesses was. Regression analysis was employed to examin e the impact that s almonella and h epatitis A have on the price of tomatoes. A survey was created and distributed to determine the imp act that food safety regulations and guidelines have had on individual growers in
11 terms of profit, price and their beliefs. These methods have not previously been employed to study Florida tomatoes and food safety. This resear ch is intended to help producers understand the impact of food safety regulations are on their bottom line. Furthermore it will allow some understanding of what regulations provide to consumers and producers alike.
12 CHAPTER 2 LITERATURE REVIEW The first studies on United States food saf ety date back to the early 1900s. In 1902 according to the Food and Drug Administration (FDA), funds were first allocated to a study to determine the safety of food preservation. It was at this point in time, after studies were done that it was d etermined the onus of food safety should be on the producer of food, and that it should be food labels that provide consumers with all the in formation that they need The only thing that was conquered during this time was the actual ingredients that went into the product would be included in labeling and not the manner in which the product was handled. It was in 1906 that the first law was signed by Theodore Roosevelt and prohibited the introduction of misbranded and adulterated foods, drinks, and drugs in interstate commerce; prohibited the addition of color additives to conceal inferiority; and prohibited the use of "poisonous" colors in confectionary. This was called the Pure Food and Drugs Act and was enforced by the Department of Agriculture and th e Bureau of Chemistry (Janssen 2009) It was also at this time that a n important piece of literature was published, The Jungle by Upton Sinclair (1906) This highlighted the conditions in which meat was produced and catapulted the Meat Inspection Act at the same time the Pure Act was passed. Then in 1913 another law was passed which was referred to as the Gould Amendment which required listing ingredients on the packaging. It was in 1930 that the Agency controlling these standards, the Bureau of Chemis try, became the Food and Drug Administration that we know today. The first law to pass under this new organization was the result of individuals dying due to a chemical in the food product. Finally in 1939 the first act passed relating to tomato products and it dealt specifically
13 with canned tomatoes, paste and puree, and aided in determining actual ingredients. Different laws and regulations were enacted in the following years, some focusing on specific products others on problems with production or lab eling (Janssen 2009). It was in 1997 that President Clinton presented a proposal for a food safety initiative titled Food Safety f rom Farm to Table. This proposal was intended to further reduce the incidence of food borne illness using strategies outl ined by the six agencies in charge of regulating food safety. It was the intention of the government at this point in time (1994) that private and public organizations work together to combat food safety issues. The report stated that the health system w a s not able to adequately monitor and control food pathogens to an extent necessary to prev ent deaths and illness. There we re six fundamental agencies that deal t with food safety two under the Department of Human and Health Services, three under the Department of Agriculture and one under the Environmental Protection Agency. It was t hese agencies who pushed for implementation of the Hazard Analysis and Critical Control Point (HACCP) system in order to identify points of contamination and clearly outline who is responsible for each part of the regulatory system for food safe ty. This type of program existed on a F ederal level for meat, poultry and seafood, at the time of this report. The 1967 Federal Meat Inspection Act requires that poultry and meat be i nspected at the state level. The requirements ensure that the level of inspection at the state level is either equal to or beyond the requirements at the Federal level (Federal Food Safety Laws 2010) Unfortunately there is no federal regulation at a state level for these procedures. It has become the responsibility of the states to create their own standards and enforce them (United States 1997).
14 Prior to t he s almonella outbreak of 2007 that was at first attributed to tomatoes the Florida tomato industry developed BMPs and GAPs that everyone is mandated by law to follow (Ashby 2009). BMPs refer to the best management practices while GAPs refer to good agricultural practices, both are frequently used in agricultural audits. The most recen t outbreak occurred because of s almonella Saintpaul and was attributed to tomato product ion in June of 2007 According to the Center for Disease Control and Prevention (CDC) as of 1990 there have been 12 multi -state produce outbreaks in addition to various smaller outbreaks that can be attributed to the production of tomatoes (CDC Salmonella 2010) In July, 2002 an outbreak occurred in Orlando Florida resulting from th e strain s almonella Javiana that was traced back to roma tomatoes (Toth 2002). Furthermore in Au gust, 2002 another outbreak occurred resulting from the strain s almonella Newport while there has been no report by the CDC linking this outbreak back to tomatoes; the media has implicated a packing house in the Northe ast region of the United States (Shin 2008). In 2004 CDC reported an 18 state outbreak of s almonella resulting from infected roma tomatoes, i n 3 separate outbreaks All of the CDC studies were based on a case control study using an equal number of well and sick individuals to determine the source of the o utbreak. Once they obtained the s e data a multivariate analysis was performed to determine that the illness resulted from the consumption of fresh sliced roma tomatoes, at least in the 2004 case (Corby 2005). Then from 2005 to 2006 there were 519 confirmed cases of s almonella resulting from tomatoes grown in Ohio, Virginia and Florida fields. Again a case control study was conducted and three separate outbreaks of salmonella were traced back to tomato production. The mo st recent outbreak i n April
15 2007 was ultimately traced back to Jalapeno peppers, but not before irreparable damage was done to the tomato industry. T he tomato industry is still experiencing the aftershock of being linked to the salmonella scare in the form of a decrease in r evenue of ten percent (Ashby 2009) There appears to be greater implications than just a loss of revenue though as the California growers for leafy greens have shown food safety costs to have more than doubled since their link to E. Coli. Like the state of Florida, California has adopted standards for growers that enable buyers to trace back and ensure that the produce is properly grown. California leafy greens are subject to the California Leafy Green Products Hand ler Marketing Agreement (LGMA), which i s similar to the standards enforced by the Florida Tomato Committee in the form of audits (Growers 2009). Fortunately for the growers of Florida Tomatoes long before standards were in place independent audits existed. Companies like Primus performed audits on the growers in order to ensure packing houses were getting clean produce. This was not standardized across the industry and as a result the state stepped in with government audits. The problem now is that some growers still have to perform the i ndependent audits in addition to the government mandated audits which is resulting in increased costs for the growers. These audits are in some cases mandatory before the shipping house will package and ship the tomatoes to the supplier. No data yet exis ts on what the risk is for not performing these audits on the tomato growers themselves, but there is some information on the impact that these food scares have had on other markets. Lloyd et al. (2001) focus on food safety scares and the impact they have on beef and other markets, specifically in relation to Bovine Spongiform Encephalopathy (BSE).
16 In addition they are focusing on what has occurred because of the link between BSE and Creutzfeld-Jakob disease (CJD). They find that a link between a particul ar product and a disease or problem results in a market collapse for the product. Furthermore they focus on the supply chain and the effect that the food scare has on each of three stages in the marketing chain: retail, wholesale and producer. Their findings indicate that the producer is the hardest hit financially, with more than double the fall out than what can be expected for the retailer. This is not a constant effect but rather a generalization of what has occurred to the beef sector as a result o f the linkage between beef and CJD as a result of BSE. An increase of poultry consumption is also reported as beef consumption falls because of the link to CJD. This can be expected as poultry is a substitute for beef in the event of a price change or in this case a food scare. This has direct implications to the tomato industry because there are substitutes for tomatoes F urthermore the link between a food safety incident and tomatoes will cause a decrease in the production of tomatoes according to pri ce and demand (Lloyd 2001). Calvin states that in 1998 the FDA submitted voluntary guidelines to limit microbial contamination using GAPs. It is important to note that there are no mandatory federal guidelines that growers and shippers must use, and regul ations vary state to state with Florida being the first to influence tomato standards in the form of an audit Calvin makes a point to not e that Federal guidelines leave it to the discretion of each grower to determine if the risk outweighs the cost to im plement certain procedures which would help to eliminate microbial diseases in fresh produce. Furthermore growers can opt to have a private audit done, which some packing houses and purchasers require, the cost for this in some areas is substantial O ut breaks are actually increasing as measured
17 through the 1990s which would indicate that not enough is being done to curtail the risk associated with growing fresh produce (Sivapalasingam 2004) This does not distinguish between domestic and international produce that is sold in the United States, the study we will be conducting looks at the impact that any food borne outbreak on tomatoes has on the overall Florida tomato industry. It remains to be explained why the actual number of outbreaks has increased and whether th is is a result of better oversight by the FDA and other regulatory bodies or whether there actually are more cases (Calvin 2003 ). Carter and Smith (2006) estimate d the market effect of a food scare using genetically modified StarLink corn. This study ended up b eing a natural study because StarLink corn was leaked by accident into the non modified corn and was consumed by humans. A natural study means that this study was not controlled but occurred in the public arena. The public found out about the introduction of the genetically modified corn into their food and the effects are studied in this paper. The authors highlight three methods previously used to determine the impact that StarLink had on the corn market, yet end ed up using a new mod el that created a relative price for a substitute (RPS). This model was adequate because it takes into account other products that can affect demand and uses time series data to adjust for the food shock. In essence they isolate d the shock from any a nd all other changes such as income and technology, allowing them to estimate the effect that StarLink had on the market for corn. The point is m ade that the exact time that StarLink hit the market is often unknown which suggests that using a model esti mated with time series data may be the best way to optimize results. This study ultimately found that there was a significant decrease of 7
18 percent in the price of corn which la sted for approximately one year. O verall individuals were willing to accept t he risk of eating genetically modified products at a price reduction (Carter 2006). The question of whether food safety information impacts U.S. meat demand is answered at least partially by Piggott and Marsh (2004) They introduce d the point that food safety concerns are easier to detect and remain a shock to the system longer than concerns over health. Such concerns are usually in the form of outbreaks where information to the public is readily available and tend to shock demand due to the influx of i nformation. Their theoretical model infers that such shocks are a result of belief about quality due to the safety scare and thus they relate quality and food safety as inversely related. This research concludes by finding that the economic impact of food safety issues is small but statistically important for the product and its substitutes (Piggott 2004). There is an argument that the overall benefit of tracking far outweighs the costs associated with the process. Cox et al. (2005) focus ed on this i n determining a probability model that allows costs to be estimated for each decision made. Dealing with BSE they determine that discovery of BSE in the United States regardless of where it came from can severely damage ex port trade. In order to mitigate this risk they determined probabilities using a probability tree with not tracking and tracking and not testing and testing. They were able to develop costs associated with each of these models and determined that the cost associated with not tracking wa s the equivalence of $90 million per year, while tracking costs were only $ 10 million. This study showed that tracking resulted in lower risks and overall less cost to the market It was
19 determined that the ideal outcome was for tracking and testing catt le from Canada in order to mitigate risk associated with BSE (Cox 2005 ). Arnade, Calvin and Kuchler (2009) studied the impact of a food safety shock on the market in relation to the 2006 outbreak of E. coli in bagged spinach products. This is one instanc e where the FDA made an announcement at the time of the outbreak and not after the incident had already occurred. This study focused on determining the overall impact of substitutes in the market and the market for spinach itself. Using the Almost Ideal Demand System (AIDS) in a two -staged analysis, the upper and lower stage, they obtained two equations to estimate impacts. They concluded that following the FDA announcement in 2006 of the problem with bagged spinach, the demand for this product dropped s ignificantly, and remained low even at the end of 2007. This drop in demand also affected the purchase of other bagged leafy greens, but did not appear to have as large of an effect on bulk leafy greens including spinach. Overall there was a decrease in the total expenditures on leafy greens by 1%. This demonstrates that there is a measurable impact of a food safety scare that affects both substitutes and the products and extends far beyond the actual outbreak (Arnade 2009). The early 1900s forced the cr eation of several different agencies and programs to regulate problems resulting from food production. This was really the advent of food safety hazards and the research that followed. Just in the last ten years (2000 -2010) various administrative leaders have made it a point to propose legislation related to food safety, from Clinton with the Food Safety from Farm to Table report and just this past year with Obama developing the Food Safety Administration (United States 1997) While there have been numer ous outbreaks of s almonella and other f ood borne
20 illnesses related to t omatoes, the research does not exist that demonstrates the risk a ssociated with food safety that f armer is taking to produce tomato es and furthermore what has been done to mitigate that risk. In 2008 another outbreak of s almonella Saintpaul occurred and while it was n ot traced back to tomatoes, tomatoes were implicated early on (Ashby 2009) This had a direct impact on the sale of tomatoes across the nation and affected the farmers in Florida specifically (Ashby 2009) The Florida Department of Agriculture and Consumer Services (FDACS) had begun enforcement of Tomato Good Agricultural Practices (T -GAP) and Tomato Best Management Practices (T -BMP) using audits enforced by the state of Florida. These standar ds we re widely recognized by the industry, but Florida was the first state to enforce them on a state wide basis. This enabled the state to say that tomatoes from Flor ida were not the culprit in the 2008 outbreak. The impact s of f o od safety shocks and scares have been investigated by numerous authors but primarily with respect to beef and its substitutes. A natural study by Carter and Smith (2006) provides some basis in which we can look at the market effects of a food safety shock using corn. Furthermore Piggott and Marsh dissect what happens with demand when a food safety shock occurs. We will primarily be focusing on what type of risk exists for producing fresh tomatoes, and whether the risk has in some part been mitigated by audits required by the State of Florida. At this time (2010) there is draft guidance for Federal legislation related to tomato production and food safety issues. This legislation is the nearly the same combination of measures and cont rol that the State o f Florida uses to prevent food borne illness. This legislation does not call for mandatory federal inspections. The need for private
21 inspections may still exist with this legislation as buyers may still require growers to use these services, thus increas ing costs. This legislation is pending and utilizes the same checklists and ideologies as the mandatory Florida inspections; the data presented in this paper is likely to reflect that risk associated with passage of this legislation.
22 CHAPTER 3 METHODOLOGY Purpose The purpose of this study is t o determine the impact s of food safety incidences on tomato growers and whether audits decrease the risk s that they assume. It is hypothesized that salmonella will have a large negative eff ect on the price of tomatoes when there is an inciden t Furthermore it is believed that in addition to public audits, private audits have the best ability to reduce risk s for tomato growers. Regression Analysis Surveys wer e sent out to tomato farmers in Florida with several reminders sent to encourage a response. Unfortunately only two surveys were retur ned with useable data. This altered the manner of data collection but not the purpose of the study as a whole. As a resu lt this study used data already reported in the scientific community. Every year the CDC releases a report on the number of food borne disease outbreaks by etiology. These data are used in a regression analysis. This information is readily available on the CDC website (Outbreak Surveillance Data). Based on these data, only incidents w h ere tomatoes were listed as the vehicle of transmission were included in the study Starting in 2006 these data included the numb er of hospitalizations as well as the deaths due to each individual outbreak. These data provided the location of outbreak, the month of outbreak, the number of ill and the confirmed etiology. Not all cases were confirmed with some cases only stat ing the suspected etiology or vehicle of transportation. These data are included in the overall data even though they are not
23 confirmed and only suspected. In many instances there is more than one vehicle of transportation ; in these instances the d ata are included. Regression analysis was performed to determine the overall impact on the price of tomatoes as a result of food safety incidence s in the tomato market. Using a simple regression analysis with six independent variables and one dependent variabl e, the overall impact of a food scare on price was estimated The price of tomatoes was specified as a function of the number of cartons produced (quantity) in Florida and Mexico, the price of a substitute, consumer income and finally whether or not the month had a food safety scare. This model uses the price of Florida tomatoes as the dependent variable. This price is used in order to analyze the effect that food safety scares associated with tomatoes as a whole have on the price of Florida tomatoes. The q uantity of Florida and Mexican tomatoes is measured in cartons (25 pounds) of t omatoes. The price of Florida cucumbers was included in the model to account for substitution effects The price of Flori da green peppers was tried as a n alternative to Florida cucumbers as a substitute but proved to be an insignificant factor. Consumer income is measured by the national consumer income reported by the Bureau of Economic Analysis. It is a report of earnings as received by employee per month. In addition an interaction term between consumer income and the price of Florida cucumbers was added to the model,. Finally there are two dummy variables, each measuring a different food etiology that has occurred; s almonella and hepatitis A. Two other dummy variables were tried in the model, n orovirus and unknown, but both were found to be insignificant and removed
24 SAS computer analysis program software w as used to estimate the model discussed above. Formally, this model can be written as (3 -1) The variables in equation (3 1 ) are defined as follows : : t he price of Florida t omatoes, as reported by the Florida Agriculture Statistical D irectory : t he quantity of Florida t omatoes and Mexican t omatoes as reported by the Florida Tomato Committee Report : c onsumer income as reported by the Bureau of Economic A nalysis, measured as comp ensation received by empl oyees (not deflated, seasonally adjusted at an annual rate ). cuc : p rice of Florida cucumbers as reported by the Florida Agriculture Statistical Directory : a dummy variable with respect to whether there was an outbreak of s almonella during the month in question, as reported by the CDC. : a dummy variable with respect to wh ether there was an outbreak of h epatitis A during the month in question, as reported by the CDC cic : an interaction term between consumer income and cucumber price. Equation (3 -1) in level level form was estimated using the OLS method. A log-log model was attempted but the results did not fit well T he model was corrected for autocorrelation after determining the model was subject to a fir st order autocorrelation. Equation (3 -1 ) is a price dependent demand equation and the impacts of the explanatory variables on price in terms of percentage changes, which are referred to as flexibilities are often considered The flexibilitie s of the variables can be derived from The data analysis for this paper was generated using SAS software, Versi on 9.2 of the SAS System for University of Florida. Copyright 2002 2008 SAS Institute Inc. SAS and all other SAS Institute Inc. product or service names are registered trademarks or trademarks of SAS Institute Inc., Cary, NC, USA.
25 the results as follows: (3 -2 ) In equation (3-2) x stands for CI, QQS and CUC The partial derivative 1 + 6 CU when x = CI; 2 5 + 6 CI when X=CU. Sample m ean values are used for p and the oth er variables present in the regre ssion to estimate the flexibilit ies In this manner the impact that changes to the food safety environment have on price of Florida t omatoes can be estimated. VOI Analysis Additional data from the Florida Tomato Committee annual report s will be used to calculate the incidence rate for food borne illness. A simple computation was used to estimate the incidence rate for each food borne illness by dividing the number of ill by the number of tomatoes produced. Originally data from various private audit firms were to be used to determine the probability of a food borne illness from audited farms as compared to nonaudited farms. Various private audit firms were approached in order to obtain this information. Unfortunately t his information wa s not made available and therefore the only analysis that could be completed to determine the probability of suffering from food borne illness for all tomatoes produced regardless of whether they were for privately audited farms or not The incidence rate is estimated as follows: (3 3 )
26 This equation will be used to determine the probabilities necess ary to estimate the impacts of implementing food safety s tandards in Florida in June 2007 and how they impacted the incidence of food safety outbreaks. This will help to determine wh ether the cost associated with food safety programs actually mitigates an y food safety ris k for the producers of Florida t omatoes. By determining average prices, when there is no salmonella and price per month with incidences of 200 7 using Equation (31) a revenue stream can be determined. The difference in the two revenue s treams represents the overall impact of incidences of salmonella. In addition it can be found what percentage decrease in revenues will occur due to food safety incidences. Figure (3 1 ) demonstrates the decisions that a grower has to make prior to shi pping his product to suppliers. The first decision is inherent within the decision to produce tomatoes; the State of Florida requires that all producers undergo an FDACS audit. The second decision is one that is either pushed upon the producer or the producer makes the decision to participate in. Some suppliers require that producers have private audits done to ensure the quality of the product. The funds for these audits come out of the growers budget and are not in any way subsidized or funded by a third party. The audits are meant to be comparable to the FDACS audit but in some cases are more thorough. Had the data been obtainable from private audit firms, the probability of a farm passing on a food borne illness with and without private audits coul d have been determined. Without this data the value of the private audits cannot be determined
27 Figure 31. VOI Chart. D1: The original decision point Ill: represents an incidence of food borne illness that is reported. No ill: represents no incidenc e of a food borne illness These two data sets will be combined to determine the overall effec t of food safety scares in the tomato market Utilizing both probability data and the regression the actual impact that these events have on the t omato market can be determined.
28 CHAPTER 4 RESULTS The OLS method was initially used to estimate model (31). The Durbin Watson statistic, however, indicated an autocorrelation problem, and the model was reestimated under a first order autocorrelation structure for t he error term. The OLS estimates are shown in table (41), while model estimates corrected for autocorrelation are shown in table (42). The results in table (41) and table (42) both indicate that an incidence of s almonella as related to tomato product s results in a lower price for tomatoes. The results suggest that lower price s result from a decline in demand for the product as a result of food safety concerns. The coefficient on the sa lmonella dummy variable is significant at the 1 0% level of signif icance, with a negative sign, implying a decreased price with incidence. T hus, t his variable demonstrates that food safety incidents can significantly affect the price of tomato es In addition, the estimated coefficient for the h epatitis A dummy variable was negative, indicating that this variable also negatively impacts the price. However, t he pvalue for the h epatitis A coefficient is rather large (0. 2 365) in comparison to those for the rest of the variables in the model, so there is no solid evidenc e that a n incidence of h epatitis A will result in a decrease in price. Despite, the mixed level of statistical significance on these two dummy variables, the hypothesis that both of these variables measuring incidences of a food safety outbreaks, a ffect the price of a tomato tomatoes would seem to be an interesting area for future research. T he data used in this analysis only spanned four years from 2004 to 2007, and variation in the data was limited, possibly resulting in reduced significance of the model estimates. Nevertheless, some of t he variables util ized in this regression had impact s
29 which were expected by the researcher The coefficient on the quantity of Florida and Mexican tomatoes that were produced each m onth was significantly different than zero at the 10% level and had the correct sign, negative; that is when quantity falls price tends to rise and vice versa C onsumer income and the price of cumbers however had neutral effects (not significantly different than zero), taking into account both the direct and interaction coefficients (the income effect was measured as the direct coefficient for income plus the interaction coefficient times the mean value of the cumber price, while the cumber price effect was measured as the direct coeff icient for this variable plus the interaction coefficient times the mean value of income). Another i mportant note to make about the estimated model is that its R -squared value at 0.2664 is r elative ly low. Again, t he length of data and limited variation o f the variables may be underlying this result Had there been more data available such as a ten year data set a larger R -squared may have been found as well as higher t statistics for the estimated coefficients. This speculation suggests future research might consider repeating this analysis based on a larger data set. As discussed earlier in the literature review there is a concern about the lasting effect of an incidence of a food borne illness. Although it is possible to pinpoint the instance that the food borne illness begins to occur, it is impossible to determine when exactly it hit the market. The regression model takes into account the month in which the incidence is registered with the CDC, and when there is an actual illness associated with it. A time d ecaying variable was added to the model but it was found to be insignificant. This does not necessarily imply that the effect only lasts the month that the outbreak occurs but rather that it is hard to quantify the impact that the outbreak
30 m ay have on the market for a longer length of time. A longer time series data set w ould be ideal in this situation; it would need to include more than t he four years of data that were available at the time of this study. Further analysis of this topic may find, for example, an impact like that estimated for the leafy green market ; in this case a severe outbreak resulted in an overall decreas e in consumption and revenue for a much longer period than the incidence lasted. Seasonality is another factor that was not considered in this study due to the limited data. With an expanded data set this factor might also be explored. As indicated above, a test was ru n on the OLS regression results to determine if there was an issue with autocorrelation. For t he original OLS regression the DurbinWatson was 0.952. This value fa lls below the lower limit with six variables and an intercept with 44 observations. The autocorrelation correction was run using a lag of 1 and a new Durbin-Watson was found of 1.5479. This is in the indeterminate range and is considered here to be acceptable Flexibility Analysis Price flexibilities are utilize d in this model with the dependent variable being price. The flexibilities simply indicate percentage changes in price for one percent changes in the dependent variables; i.e., they are elasticities in context of a price dependent demand s pecification. Generally, price flexibilities represent the response of price to changes in costs and demands when both are changing in a com petitive market (Moore 1955). T his concept allows the research to create comparisons between price and the various variables in the model through percentage changes. Referring back to E quation ( 3 -2 ) we can find the flexibilities for each variable. Th e flexibilities in this study are calculated at the sample mean values of the variables
31 involved. The price flexibilit y f or consumer income is 1.8344 which implies that a 1% increase in consumer income will result in a price de crease of 1.8344% It is i mportant to recognize that the flexibility for consumer income on the tomato price is a linear combination of two coefficients (the coefficient on consumer income and the coefficient on the interaction term). Based on the F -test, this flexibility is not s ta ti stically different from zero at the 10% level of significanc e. Therefore it does not suggest that tomatoes are an inferior good but rather that tomatoes are neutral with respect to income. It should be noted that, i ncome roughly follows a time trend variable and may be picking up other factors beyond the effect of income on the price of tomatoes. The price flexibilit y of quan tity is -0.59 5 indicating that a 1% increase in quantity produced results in a 0.595 % decrease in price For cucumbers the price flexibility is 0.0604 which suggests that a 1% increase in the price of cucumbers results in a 0.0604 % increase in the pr ice of tomatoes. As is the case with consumer income, the impact of the cucumber price on the tomato price is a result of a lin ear combination of the cucumber price coefficient and the coefficient on the interaction term. This flexibility is also not statistically significant based on the F -test at the 10% significance level. The last two variables for salmonella and hepatitis A are binary; and elasticities for these variables were not calculated. The regression coefficient for salmonella is -2.61, meaning an incidence of salmonella will decrease price by $2.61. To examine what percentage of price this is consider the sample mean price of $11.39 If price were to decrease by $2.61 from $ 11.39 the percentage change would be 22.9%. This value provides a measure of the percentage decrease in price when there is an incidence of salmonella in the tomato market Using the same method, the impact of hepatitis A on
32 price can be examined. It should be noted that hepatitis A was found to be statistically insignificant. However, the sign of its coefficient was consistent with expectations; hence, the following does not have a stron g statistical underpinning. The coefficient associ ated with hepatitis A is 4.10 which i mplies a decrease in price of $4.10 with an incidence of hepatitis A. This is a change in price of 36.0 % from the base price of $ 11.39 Both of these percentage cha nge values show the large extent that an incidence o f a food safety outbreak may have on the price for tomatoes. VOI Analysis Using E quation ( 3 3 ) it is possible t o find the probability of food borne ill ness for tomato es produced for the United Stat es market It was not possible to find out the locations of all the outbreaks that were reported to the CDC, so the percentages are based on national output. From 2005 to 2006 the percentage opportunity of contracting a food b orne illness from a tomato w as 6.86 106%. This implies that the chanc e not becoming ill is 99.9999314 %. In 2007 these percentages changed slightly and the oppo rtunity of becoming ill was 5.45 106% or 99.999 9 95 % of not becoming ill. The CDC does not report data to the public as to where they believe the outbreak originated unless there is conclusive evidence. While it can be argued that the chance of getting a consumer ill has d ecreased substantially from 2005 to 2007, the question is if it was a result of increased food safety standar ds or something else entirely. F armers on average pay about fifteen hundred dollars to have a private audit done on their property, in addition to the cost to have the government perform th eir audit. This cost comes with no conclusive evidence that enforcing these standards results in fewer outbreaks of food borne illnesses. These percentages do not take into account where the tomatoes were produced. If the same data were to be run using only Florida
33 tomatoes, the difference would be substantial. M ost of the tomato production comes from California and Florida and the ability to pinpoint the source of the outbreak would do a lot to further the understanding of percentage risk of contracting a food borne illness. It is imp ortant to note that in Fi gure ( 3 1 ) the first decision made is not actually a decision but more of a requirement. All tomato producers in the State of Florida ar e required to participate in an FDACS audit if they deliver their product to the general public. The second choice is on e that the farmer can make on their own, or as required by their supplier. Private audits are supplied by many different firms and essentially do the same audits that FDACS does with some more in-depth analysis. Further research could be done in this area to deter mine if private audits provide enough additional benefit over public audits. Private audits are in most cases required by the packing house or the buyer of the tomato products and are paid by th e grower. I f these audits are essentially doing t he same job that the FDAC S audits are providing, then it really questions the need for both and the costs for gr owers to pay for private audits. It is hypothesiz ed that the optimal decision rule is to participate in both the public and private audits. T here are a number of costs associated with this decision rule; the public audit cost, the private audit cost and the cost of an incident of a food safety outbreak To determine the overall cost to the Florida tomato industry data from the Economic Researc h Service USDA division was utilized (U.S. Tomato Statistics 2010) U.S. fresh market field grown tomato production for the state of Florida in pounds was used to calculate revenue (Lucier 2008) From the regression equation( 3 1 ) an price is calculate d for every month and used to determine price received. Using the
34 regression data and calculating an average price when there is no incident of food safety an average price of $11.15 was found. The cost for an audit is $75.00 an hour and the average audit takes about three and a half hours. This places the cost of a public audit at $262.50. A priv ate audit on average costs between $1500.00 and $2000.00. A simulation of the tomato markets from the regression model w hen there is no incidence o f s alm onella in the tomato market estimates the state of Florida tomato industry earns on average $452,571,587 in revenue. When the model is simulated with the salmonella incidents that occurred the revenue is found to be $434,559,368. From these revenue estim ates an expected value can be calculated which will help to determine at what cost audits are valued. The difference between these two revenues is $18,012,218. This is the value that can be placed on food safety audits to eliminate food borne illness ris ks to growers. This is the cost that growers as an industry can expect to pay to have audits still be worthwhile in terms of revenue streams. This amount is 3.98% of the actual expected revenue assuming there are no outbreaks of food borne illness. This value calculation is made with the assumption that the audits that are performed result in no incidences of food borne illness associated with salmonella or hepatitis A It is important to note that while the expected value does not differ greatly from t he value without salmonella the difference between revenue with salmonella and revenue without salmonella is substantial. Regardless of whether the incident occurs on Florida farms or outside the state of Florida the impact remains the This information was furnished by Charles Beasley who is the Bureau Chief of the Division of Fruits and Vegetables for the Florida Department of Agriculture and Consumer Services. He provided this information on a personal phone call on 3/29/2010 at 1 pm. He informed me that these were average estimates of prices for audits.
35 same, which is large in terms of the revenue lost. Furthermore, in relation to the revenue for three months in which the outbreaks occur ( June July and October ) the impact represents 28.32% of the revenue stream. By decreasing the incidence of salmonella in the market it i s feasible that a large revenue increase can occur. The implications are such that even though the probabilities appear low at first look, the substantial income that is lost due to an outbreak appears to justify the added cost of $262.50 for a public au dit. It remains to be seen whe ther or not the additional expense for the private audit decreases the probability of selling infected produce. The probabilities imply that there needs to be better monitoring in accordance with the audits. There is a que stion of the audits and whether they cover enough of the issues with tomatoes or the institution that is responsible for doing the audits. While there has been a significant decrease in the percentage chance of contacting a food borne illness from tomatoe s, it remains to be seen where the critical point is, i.e., at what point is there less pressure placed on the grower to submit to public and private audit s at regular periods in time. Evidence exists that infer s food safety standards are working and affe cting the price of Florida tomatoes. As reported by the National Agricultural Statistical Service (NASS) there was a large decline in t he price of Florida tomatoes i n 2007 when the food incidents were at their highest, in terms of consumer awareness. Pri ces fell to $31.90 per 100 lb in 2007. This was with quantity produced more or less the same as in 2006 when price was at $40.90 per 100 lb. With the implementation of food safety standards and a decrease in quantity due to the low returns to growers, pr ice rose in 2008 to $59.50 per 100 lb. While some of the price increase can be attributed to the quantity
36 decrease, this provides anecdotal evidence to support the idea that food safety standards in Florida did cause some of this price increase (U.S. Tomato Statistics 2010). Further research on this topic c ould impart knowledge on how the media affects the demand for tomatoes during and after a food safety incident. A study was done on apples that determined the impact the Alar crisis had on the public (Herrmann 1997). In addition to the research that has been done so far on tomato risk mitigation, looking at the media impact would certainly add to and enhance what is already known about the price impact on tomatoes in relation to an outbreak. In conc lusion there is evidence that food safety standards work to reduce the risk that a consumer has of getting ill from consuming a Florida tomato. Furthermore there is also evidence that the producer is mitigating their risk of producing a product that has a food borne illness by participating in the audits. It would be interesting to determine if the risk is further mitigated by participating in private audits as well as the mandatory public audits. This data was not available but hopefully in the future i t can be determined if there is an additional benefit for a producer to pay for a private audit. This study is c onclusive that an incidence of s almonella significantly decreases the price of Florida tomato es ; but the data do not show a decaying factor. With a larger data set it is hypothesized that there c ould be a significant decaying factor present. Further research on this topic will aid in public policy creation and food safety standards.
37 Table 4 1. OLS Regression Estimates of the M odel (3 1) VARIABLE PARAMETER Pr > |t| INTERCEPT 97.24650 0.1871 CI 0.03814 0.1184 QQS 0.00048862 0.0232 SALMONELLA 3.55695 0.0966 HA -5.59330 0.2263 CUC 9.78701 0.0721 CIC 0.00322 0.0735 Table 4 2. Estimates of Model (3-1) Corrected for First Order A utocorrelated VARIABLE PARAMETER Pr > |t| INTERCEPT 91.8992 0.1987 CI 0.0363 0.1233 QQS 0.000576 0.0142 SALMONELLA 2.6138 0.1068 HA -4.1024 0.2365 CUC 9.0225 0.0784 CIC 0.002946 0.0802
38 CHAPTER 5 SUMMARY AND CONCLUSIONS T hree methods of data analysis were applied to fresh market tomato markets to reach a conclusion regarding policies changes for food safety standards. The regression analysis provides an important prospective into the cost associated with a food safety outbreak, in this case s almonella and h epatitis A. Both can have significant impacts on growers not associated with the food borne illness incident. The flexibility analysis provides insight into the percentage effects of each of the variables on price. Finally the Value of Inform ation Analysis provides an estimate associated with the cost of a food safety outbreak and the value of an audit that mitigates the risk The probability of contracting a food borne illness from tomatoes as calculated in this study is extremely low but the impact on the grower of a food safety outbreak is high. Overall it is estimated that the Florida tomato i ndustry suffered a n $ 1 8 million dollar loss on revenue from the food safety incidences of 2007 This is the impact of a food safety outbreak that did not necessarily occu r on Florida farms but just the presence of an outbreak in the market that impacted all farmers This $ 1 8 million dollar loss is what the industry can justify in spending on food safety audits, to ensure that there are no outbreaks of food borne illnesses. The policy of enforcing food safety audits has appeared to work in terms of decreasing the opportunity of selling an infected tomato. While the food safety audits are currently being enforced in Florida but not in other states, this enforcement appears to have affected the opportunity of contracting a food safety illness from tomatoes. The added cost of the audits is offset by the lowered probability of a food safety incident and the savings associated with those lower probabilities It remains to be seen whether
39 private audits lower the probability of a food safety incident enough to affect the expected net revenues of the tomato grower to warrant the additional audits. F urther research could help to provide this know ledge. Further research on this topic would further determine the need for audits, whether they are private or public. Public audits are required by the s t ate of Florida and the private audits at this time do not substitute for the se public audit s It ma y be shown that private audits provide an added benefit to Florida grower s there is reason to believe t hat food safety outbreaks can further limited by better enforcement of food safety standard s U.S. government policy could be implemented to force all t omato growers to have public audits done in order to reduce outbreaks of food safety illnesses. In the meantime Florida tomato growers can account for their improved food safety practices with the use of audits, in many cases both private and public. Thi s might become a marketing tool that could mitigate some of the costs associated with food borne illnesses that result from tomatoes grown outside the state of Florida.
40 APPENDIX A INFORMED CONSENT Protocol Title: The Potential Impacts Throughout from Food Safety Practices Implemented by Florida Tomato Growers and the Impact of Recent Food Safety Scares Please read this consent document before you decide to participate in this survey. Purpose of the research study: The purpose of this study is to determine the impact of implemented food safety standards in the state of Florida on the Florida tomato industry. What you will be asked to do in the study: Answer survey questions related to your production, harvesting and packing of tomatoes in the state of Flor ida as well as 4 demographic questions. Time required: 10 minutes Risks and Benefits: There are no Potential Risks and the Benefits extend to the research community. Compensation: There is no compensation for this study it is entirely voluntary. Confide ntiality: Your identity will be kept confidential to the extent provided by law. Your name will not be used in any report. Right to withdraw from the study: You have the right to withdraw from the study at anytime without consequence. Whom to contact if you have questions about the study: Gabrielle Ferro, Food and Resources Economics, UF/IFAS 561-3734822 John Vansickle, International Agricultural Trade & Policy Center, UF/IFAS 352-3921881 ext 221 Whom to contact about your rights as a research parti cipant in the study:
41 IRB02 Office, Box 112250, University of Florida, Gainesville, FL 32611 -2250; phone 392-0433. Agreement: I have read the information above. I voluntarily agree to participate in the survey. Furthermore by returning the survey, I indi cate that I agree to participate in the study.
42 APPENDIX B SURVEY Food Safety is becoming an ever important issue in todays society with recent outbreaks of food borne illness. There is limited data about the cost that farmers and growers will incur as a result of increased food safety standards. This survey will help us determine the overall impact on you as a grower due to increasing demands for food safety standards in tomato production. We hope you enjoy taking this survey and we thank you in advance for your help and time. For each of the questions please write in the space provided. 1 The following three questions relate to T -GAP and T -BMP as enforced by FDACS in 2007 a How much do these programs add to your growing costs per acre? _______________ ________________________ b How much do these programs add to your harvesting and hauling costs per carton? _______________________________________ c How much do these programs add to your packing costs per carton? _______________________________________ 2 Do yo ur buyers require private audits? YES _______ NO________ 3 What year did your operation begin private audits? _____________________________________ 4 On average how many private audits does your operation perform per year? _____________________________________ 5 What is the average cost of a private audit for your operation? _____________________________________ 6 Did your operation have a food safety incident prior to beginning use of private audits? YES _______ NO________ a If so, What was the fiscal cost of this incident? ___________________________________
43 7 Did your operation have a food safety incident after beginning use of private audits? YES _______ NO________ a What was the fiscal cost of this incident? ___________________________________ 8 What was the fiscal cost to your operation due to the 2008 Saint Paul Salmonella tomato food scare? ______________________________________ a Was this in part du e to produce that you were unable to sell? YES _______ NO________ 9 The year following implementation of mandatory T -GAP and T -BMP what was the percentage increase in cost for production of your crop? ___________________ _____________________ 10. Do you believe private audit implementation added value to your products? YES _______ NO________ a If Yes, How much value did it add ($/carton)? ___________________________ The following question is based on your thoughts about food safety standards implementation in the State of Florida 11. Do you feel that the T -GAP and T -BMP have increased consumer confidence in Florida Tomatoes? Yes No a Why do you feel this way? _________________________ _____________________________________________ ______________________________________________________________________ ____ b What other steps that have not currently been implemented, do you feel could be taken to increase confidence and demand for Florida Tomatoes? ______________________________________________________________________ ______________________________________________________________________
44 The following section focuses on information about yourself. 1 Age ________ 2 What is your gender? Male Female 3 What do you consider yourself? Full time Farmer Part time Farmer Hobby Farmer 4 How many years have you farmed tomatoes?_____________________
45 Survey The purpose of t his study i s to determine the reduction in ris k by implementing food safety standards, and how this affects the bottom line of the Florida tomato producer. A paper survey was administered to allow producers to provide accurate input into developing these estimates The survey is described below. T he survey began by asking responders to identify the costs associated with T GAP and T -BMP, as a means of measuring overall mandatory costs for implement ation. The following question wa s asked to determine whether audit from private companies were require by buyers given that mandatory inspections by the State of Florida were already required. The following three questions related to the overall cost associated with th ese private audits including t he year they were started and the overall cost of the audits over that period of time. The succeeding two questions attempt ed to determine the incidence of outbreaks prior to and after implementing audits, specifically private audits. Furthermore, the questions look ed to uncover the fiscal cost of such an inci dent to the specific producer. By asking for information on sp ecific incidents such as the 2008 salmonella food safety scare, the survey attempted to collect data that could provide a specific cost estimate associated with a food safety scare. The 2008 sc are was later determined not to be caused by tomatoes but there was a lasting mark on the industry and there is an attempt to uncover this cost in this question. There is a belief that implementing food safety standards has increased cost of production b ecause there are additional steps that producers have to take in order to be compliant. Yet there is also a belief that because most producers were already using private audits, there has been little to no increase in cost of production. The survey
46 attem pted to determine if producers feel that audits increase the actual value of their product. This does not directly relate to the price that they are getting for the product but more about what they feel they receive for the product in relation to the addi tional cost of audits. The final question attempts to collect the producer s opinion of T -GAP and T -BMP. By determining if producers feel that consumer confidence is affected by the Florida audits that have been implemented there is a foundation that inc reased cost of production may be supported by adding value to the product. The end of the survey asks basic demographic questions to be utilized in determining the overall impact of producer demographics on their beliefs. The survey as a whole is an att empt to determine the overall risk that is being mitigated by standardizing audits. There is inherently a cost related to an audit. In order to have someone perform the audit, they must be paid. It is yet to be determined if that actual increase in prod uction cost as a result of these audits is actually mitigating the food safety risk that the producer assumes in growing the product and returns to the producer
47 APPENDIX C SAS OUTPUT
53 APPENDIX D SAS INPUT Florida Tomato Committee and Florida Statistical Directory
54 APPENDIX E FLEXIBILITY
55 APPENDIX F PROBABILITY DATA
56 Centers for Disease Control
57 LIST OF REFERENCES Arnade, Carlos, Linda Calvin, and Fr ed Kuchler. "Consumer Response to a Food Safety Shock: The 2006 FoodBorne Illness Outbreak of E. coli O157: H7 Linked to Spinach." Review of Agricultural Economics 31.4 (2009): 734750. Print. Ashby, Elizabeth. "Time Will Heal." Citrus and Vegetable 15 A ug. 2009: 12-14. Print. Calvin, Linda. "Produce, Food Safety, and International Trade." International Trade and Food Safety: Economic Theory and Case Studies. Ed. Jean C Buzby. N.p.: n.p., 2003. 74-81. Print. Carter, Colin, and Aaron Smith. Estimating the Market Effect of a Food Scare: The Case of Genetically Modified StarLink Corn. N.p.: n.p., 2006. Print. CDC Salmonella. Center for Disease Control and Prevention, 23 Jan. 2010. Web. 15 Mar. 2010. http://www.cdc.gov/salmonella/. Corby, R, V Lanni, and V Kistler. "Outbreaks of Salmonella Infections Associated with Eating Roma Tomatoes ---United States and Canada, 2004." MMWR. CDC, 7 Apr. 2005. Web. 25 Feb. 2010. http://www.cdc.gov/mmwr/previ ew/mmwrhtml/mm5413a1.htm Cox, Louis Anthony, Jr, Douglas A. Popken, John J. VanSickle, and Ranajit Sahu. Optimal Tracking and Testing of U.S. and Canadian Herds for BSE: A Value of Information (VOI) approach. N.p.: n.p., n.d. N. pag. Print. Cox, Louis A nthony, Jr, Douglas A. Popken, John J. VanSickle, and Ranajit Sahu. "Tracking and Testing of US and Canadian Cattle Herds for BSE: A Risk Management Dilemma." CHOICES Winter 2004: 5154. Print. Crop Profile for South Florida Tomatoes. Glades Crop Care, INC, 3 Dec. 1999. Web. 26 Mar. 2010. http://www.gla descropcare.com/CP_tomatoes.pdf "Federal Food Safety Laws." National Conference of State Legislatures. NCSL, n.d. Web. 19 Mar. 2010. http://www.ncsl.org/?tabid=19077. Grainger, 2007. Annual Report 2007, Maitland:Florida Tomato Committee "Growers' Food Safety Costs Doubled." Growing Produce. N.p., 22 Sept. 2009. Web. 25 Feb. 2010. http://www.growingproduce.com/news/gp/?storyid=2596. "Guidanc e for Industry: Guide to Minimize Microbial Food Safety Hazards of Tomatoes; Draft Guidance." FDA. Food and Drug Administration, 3 1 July 2009. Web. 30 Mar. 2010. http://www.fda.gov/Food/GuidanceComplianceRegulatoryInformation/Guidan ceDocuments / ProduceandPlanProducts/ucm173902.htm
58 Herrmann, Robert, Rex Warland, and Arthur Sterngold. "Who Reacts to Food Safety Scares?: Examinig the Alar Crisis." Agribusiness 13.5 (1997): 511-520. Print. Janssen, Wallace F. "The Story of the Laws Behind the Labels." U.S. Food and Drug Administration.FDA, 18 June 2009. Web. 25 Feb. 2010. http://www.fda.gov/AboutFDA/WhatWeDo/History/Overviews/ucm056044.htm Klugh, 2008. Florida Agriculture Statist ical Directory, Florida:Florida Department of Agriculture and Consumer Services Lloyd, T, et al. The Impact of Food Scares on Beef and Inter -Related Meat Markets. N.p.: n.p., 2 001. N. pag. Print. Lucier, Gary. "Background Statistics: Freshmarket Tomatoes ." Newsroom. Economic Research Service, 10 June 2008. Web. 13 Apr. 2010. http://www.ers.usda.gov/News/tomatocoverage.htm. McClure, 2006. Annual Report 2006, Maitland:Florida Tomato Committee Murrah, 2005. Annual Report 2005, Maitland:Florida Tomato Commit tee Murrah, 2004. Annual Report 2004, Maitland:Florida Tomato Committee Moore, John R., and Lester S. Levy. "Price Flexibility and Industrial Concentration." Southern Economic Journal 21.4 (1955): 435440. Print. Outbreak Surveillance Data. Centers for Dis ease Control and Prevention, 20 Jan. 2010. Web. 12 Mar. 2010.
59 "U.S. Tomato Statistics." Economic, Statistics, and Market Information System. Economic Research Service, n.d. Web. 13 Apr. 2010. http://usda.mannlib.cornell.edu/MannUsda/view DocumentInfo.do?documentID=12 10. U nited States. Food and Drug Administration. Food Safety from Farm to Table : A National Food Safety Initiative. Center for Disease Control and Preventation. N.p., May 1997. Web. 10 Mar. 2010. http://www.cdc.gov/ncidod/foodsafe/report.htm
60 BIOGRAPHICAL SKETCH Gabrielle Alexandra Ferro was born in B ethesda, Maryland She was raised in Lake Worth, Florida and attended Suncoast Community High School, graduating in 2004 from the International Baccalaureate Program. She completed the Master of Science in food and resource e conomics from the University of Florida College of Agricultural and Life Sciences in May 2010 Gabrielle completed her bachelors degree in e conomics from the University of Florida, Warrington College of Business in 2008. Sh e plans to eventually pursue a Ph.D. in Economics with a concentration in Public Policy. After graduating she intends to find a job that focuses on data analysis and contracting.