转基因食品的强制性标签外文文献英文_第1页
转基因食品的强制性标签外文文献英文_第2页
转基因食品的强制性标签外文文献英文_第3页
转基因食品的强制性标签外文文献英文_第4页
转基因食品的强制性标签外文文献英文_第5页
已阅读5页,还剩4页未读 继续免费阅读

下载本文档

版权说明:本文档由用户提供并上传,收益归属内容提供方,若内容存在侵权,请进行举报或认领

文档简介

Contents lists available at ScienceDirect Journal of Behavioral and Experimental Economics journal homepage How will mandatory labeling of genetically modified food nudge consumer decision making Brandon R McFadden a Trey Maloneb aDepartment of Applied Economics and Statisitcs University of Delaware 531 South College Avenue 224 Newark DE 19716 United States bDepartment of Agricultural Food and Resource Economics Michigan State University 306 Morrill Hall of Agriculture East Lansing 48824 MI USA 1 Introduction ThepassageoftheNationalBioengineeredFoodDisclosure Standard NBFDS 1will create a new choice architecture for consumers when purchasing at home food The NBFDS requires a mandatory label for genetically modified GM foods or manufactured foods containing GM ingredients The mandatory labeling of GM material in the United States will occur during a proliferation of voluntary labels that indicate the absence of GM material i e USDA Organic seal and the Non GMO Project verification label Even the most simplistic policies can induce changes in behavior and choice Debnam and Just 2018 For example Debnam 2017 found that some consumers actually increased soda consumption in re sponse to the Berkeley soda tax Thus policies can nudge choices and nudges do not necessarily generate the intended change If taxes have varying influences on consumer behavior it follows that nudges from various labeling schemes will also have systemically different influ ences on consumers Previous research has indicated the existence of asymmetries stemming from consumer valuation of GM and non GM labels Crespi and Marette 2003 examined the effects of labels that read does not contain compared to does contain to determine the welfare maximizing choice under a mandatory labeling system They concluded the two labels had asymmetric effects on welfare and the appropriate choice was dependent on the level of consumer concern Furthermore Costanigro and Lusk 2014 found a higher willingness to pay WTP to avoid a mandatory does contain label compared to WTP to gain a voluntary does not contain label Thus there is no reason to assume a priori that all consumers either seek to avoid GM presence labels or to gain GM absence labels or that the same consumers who seek to avoid GM presence labels are similar to consumers who seek to gain GM absence labels The objective of this paper is to identify heterogeneity in WTP for a consortium of food labels that are specific to the new choice archi tecture established by the NBFDS Determinants of WTP are estimated in a manner that allows the comparison of the characteristics of consumers who sought to avoid a GM presence label compared to the characteristics of consumers who sought to gain a GM absence label This estimation method provides insight into how the choice archi tecture with both GM presence and GM absence labels will nudge choice for various market participants Furthermore this paper adds to the literature by accounting for recent exposure to information about GM foods and concern about GM food safety when examining characteristics associated with WTP More knowledgeable experienced market participants are less likely to ex hibit anomalous consumer behavior List 2003 Malone and Lusk 2018 so it follows that nudges are likely to be primarily effective at inducing change for less knowledgeable participants While previous studies have included subjective importance placed on nutrition and freshness Loureiro and Hine 2002 and opinions about organic and GM products He and Bernard 2011 as controls when determining characteristics associated with WTP for organic and non GM options it is unclear how exposure to recent information and concerns about GM foods affect WTP for non GM options relative to GM options The remainder of this article is organized as follows First the case of GM presence and GM absence labeling in the United States is de scribed Next the experimental design and empirical methods which included a stated preferences survey of 969 respondents are described The third section outlines the results from the empirical estimation and the final section concludes by discussing what the findings mean for the literature and limitations of the study 2 Presence and absence GM labeling The NBFDS requires a mandatory label for GM foods or manu factured foods containing GM ingredients this will create a new choice environment for consumers when purchasing at home food The man datory labeling of GM material in the United States will occur during a proliferation of voluntary labels that indicate the absence of GM ma terial i e USDA Organic seal and the Non GMO Project verification label The current wording of the NDFDS allows companies to indicate the https doi org 10 1016 j socec 2018 09 004 Received 18 April 2018 Received in revised form 3 August 2018 Accepted 12 September 2018 Corresponding author E mail addresses foodecon udel edu B R McFadden tmalone msu edu T Malone 1https www congress gov 114 bills s764 BILLS 114s764eas pdf Journal of Behavioral and Experimental Economics 77 2018 186 194 Available online 13 September 2018 2214 8043 2018 Published by Elsevier Inc T presence of GM material by using a symbol text or QR code This study focuses on the latter two labeling options text and QR code 2QR codes when scanned by a smart device display information about a product The NBFDS required a study to be performed to better understand the potential challenges of allowing the use of QR codes to indicate the presence of GM material The study concluded that most consumers did not state having difficulty accessing the information however due to technological challenges almost all consumers did have diffi cultieswhenactuallyattemptingtoaccesstheinformation Deloitte Consulting 2017 Moreover the study by McFadden and Lusk 2017 determined that consumers were willing to pay more for a QR code that communicated the presence of GM material than text that communicated the exact same thing The Non GMO Project verification label and the USDA Organic seal both indicate the absence of GM material and are commonly found sometimes jointly on products To become Non GMO Project verified a product must have no more than 0 9 of GM material 3By rule of certification a USDA Organic symbol implies a food is non GM Thus the Non GMO Project verified symbol explicitly states the absence of GM material while the USDA Organic symbol just implies absence While organic is implicitly non GM it is unclear that the profiles of consumers who prefer the USDA Organic label and the Non GMO Project verification label are comparable Loureiro and Hine 2002 were the first to examine consumer characteristics associated with WTP for organic and non GM separately They determined that higher levels of education and income estimated as a cross effect and greater sub jective importance placed on freshness and nutrition measured by five point Likert scales were associated with higher WTP for organic Greater subjective importance placed on freshness was associated with higher WTP for non GM Age sex and the presence of a child in the household were not a significant determinant at an alpha level of 0 05 of WTP for either organic or non GM Bernard et al 2006 examined consumer characteristics associated with WTP for non GM when an organic option was present for three products i e potato chips tortilla chips and milk chocolate They determined that age income education sex and number of children in the household were all associated with WTP for a non GM option In terestingly none of these characteristics were associated with WTP for an organic option The authors concluded extra attributes of or ganic foods beyond being non GM such as lack of synthetic pesticides did not significantly increase subject bids Building upon this work He and Bernard 2011 made the distinction between whole and manufactured foods i e potatoes versus potato chips and corn versus tortilla chips when examining consumer characteristics associated with WTP for organic and non GM options They determined WTP was greater for whole foods compared to manufactured foods when the options were labeled either organic or non GM Additionally higher levels of income and being male significantly decreased WTP for or ganic No demographic characteristics were associated with WTP for non GM however a positive opinion about organic food increased WTP for an organic option whereas a positive opinion about GM foods de creased WTP for a non GM option This study is unique because Non GMO Project verification labels and USDA Organic seals are commonly found on products and the premium for these symbols are relative to labels indicating the presence of GM material i e text and QR code The premium associated with organic and non GM labels in the aforementioned studies were relative to conventional options that did not include labels indicating the pre sence of GM material which will now be required by the NBFDS 3 Methods The primary contribution of this study is to identify heterogeneity in the effectiveness of changes in the choice architecture for GM food products The requirement of the mandatory label GM as established by the NBFDS changes the choice architecture for many categories of food products in the U S As such we were primarily concerned with identifying a sample of U S consumers who were representative of the total population to better understand how labels that may exist in the marketplace due to the NBFDS will affect choice To examine how mandatory labeling will nudge decision making our empirical results are based on contingent valuation CV questions collected within a survey of 969U S consumers Applied economists havecommonlyusedtheCVmethodforquitesometime Loomis 2014 While some concerns have been voiced regarding the empirical precision of this approach a key benefit to CV is that it is an effective way to collect data from a large sample of consumers Haab et al 2013 Hausman 2012 It is particularly useful for collecting data for products that are currently unavailable in the marketplace as is the case with the GM labels of interest 3 1 Description of the data Responses were collected in collaboration with the December 2016 issue of the Food Demand Survey at Oklahoma State University Lusk 2017 This method was considered sufficient because opt in online panels have been found to produce similar estimates as do tel ephoneinterviewsandmailresponses Ansolabehereand Schaffner 2014 Respondents were identified via an opt in online panel maintained by Survey Sampling International and associated partners Respondents were randomly assigned to experimental treat ments to identify how mandatory labels that may exist in the market place due to the NBFDS will nudge decision making The survey and experimental treatments are described in more detail in the following paragraphs Three portions of the survey are relevant to our study To determine the difference in WTP associated with the GM absence and GM presence labels respondents were randomly assigned to one of six label treat ments In each label treatment respondents were asked CV questions for two options that varied by label Respondents were prompted with a cheap talk strategy prior to the CV questions in an effort to reduce overstating WTP Loomis 2014 Overstating WTP may occur due to the lack of incentive compatibility associated with hypothetical elici tation however cheap talk makes respondents more sensitive to the possibility of overstating WTP Fifer et al 2014 and can reduce this problem Cummings and Taylor 1999 List 2001 Moreover the WTP examined in this study was the difference between two label options and Lusk and Schroeder 2004 determined that overstating WTP is less likely for differences in WTP compared to WTP for one product While the GM presence labels used in this study included text and a QR code that both read contains genetically engineered ingredients the QR code must be scanned to reveal the words contains genetically engineered ingredients The GM absence labels used were Non GMO Project and USDA Organic The six treatments used to determine dif ferences in WTP were as follows 1 Non GMO Project verification vs QR code 2 USDA Organic vs QR code 3 Non GMO Project ver ification combined with USDA Organic vs QR code 4 Non GMO Project verification vs text indicating contains genetically engineered ingredients text 5 USDA Organic vs contains genetically engineered ingredients text and 6 Non GMO Project verification combined with USDA Organic vs contains genetically engineered ingredients text These treatments allow for the examination of consumer characteristics associated with differences in WTP for GM and non GM options 2For more information about the NBFDS read McFadden 2017 3Information about the standards for Non GMO verification and USDA Organic certification ban can be found at the following links http www nongmoproject org wp content uploads 2016 08 Non GMO Project Standard pdfandhttps www ams usda gov sites default files media Labeling 20Organic 20Products 20Fact 20Sheet pdf B R McFadden T MaloneJournal of Behavioral and Experimental Economics 77 2018 186 194 187 Table 1 Mean difference in WTP for the six label treatments Granola Bars 12 count Apples 1 pound Label treatmentMean difference in WTP Number of observationsMean Difference in WTP Number of observations 1 0 09 0 90 1620 32 0 64 160 2 0 17 0 91 1620 32 0 65 160 3 0 20 1 12 1600 21 0 51 160 4 0 35 1 00 1600 35 0 68 162 5 0 09 0 90 1620 40 0 65 162 continued on next page B R McFadden T MaloneJournal of Behavioral and Experimental Economics 77 2018 186 194 188 Respondents were randomly assigned to one of the six treatments for two food products a box of 12 granola bars and a pound of apples Possible WTP values were bounded from 0 to 10 for granola bars and 0 to 5 for apples all currency is USD It was necessary to elicit WTP for both manufactured and whole foods because preferences to avoid GM foods Lusk et al 2015 and preferences to gain non GM foods He and Bernard 2011 can vary by product form Thus granola bars a manufactured food and apples a whole food were chosen so any re sults could be more generalizable The differences in WTP for the six label treatments are shown in Table 1 In the second relevant portion respondents were asked questions to measure exposure to information and concern about GM foods The exposure question asked how much a respondent had heard or read about GM foods in the past two weeks Answer responses were a five point Likert scale that ranged from Nothing to A great deal The concern question asked how concerned a respondent was that GM foods posed a health hazard in the next two weeks Answer responses were scoredonafive pointLikertscalethatrangedfrom Very Unconcerned to VeryConcerned ThePearsonCorrelation Coefficient between the two measurements was 0 357 suggesting that there is a strong positive relationship between awareness of and con cern for genetically modified foods The survey concluded with the respondents answering demographic questions about age education household size income marriage status political ideology and sex Demographic characteristics were not significantly different across label treatments at an alpha level of 0 05 Table 2 shows the coding and sample means for exposure con cern and demographic characteristics The average respondent had completed some college or two year degree was between 35 and 44 years old and self identified as politically moderate Approximately 62 of our sample identified as female and 38 identified as male 3 2 Empirical approach The empirical analysis used a two step estimation approach similar to Nayga 2000 where nutrition knowledge was examined by demo graphic characteristics and then food label use was examined by de mographic characteristics and nutrition knowledge A similar frame work was used to determine demographic characteristics associated exposure to information or concern about GM foods and then differ ences in WTP were examined by demographic characteristics and ex posure and concern Response categories for exposure to information and concern about GM foods were measured using five point Likert scales so the first es timation used ordered probit models to examine the demographic characteristics of consumers who were recently exposed to information about GM foods or who were concerned about the safety of GM foods An ordered probit model is often used when the dependent variable is a Likertscalebecauseoftheordinalformofthemeasurement Daykin and Moffatt 2002 The dependent variables in the first esti mation were Exposure and Concern represented by the vector Y and modeled as YXef ii1i 1 where X is a vector of demographic characteristics i e Age Education Female Household Size Income Married Politically Conservative IDK Politically Ideology and e captures unobservable determinants of ex posure and concern Exposure and Concern were then included as en dogenous variables in the second estimation which examines determi nantsofdifferencesinWTP alongwiththedemographic characteristics The second estimation used censored regression models to examine the determinants of WTP similar to Batte et al 2007 A contribution of this paper was the estimation of determinants in a manner that al lows the comparison of the characteristics of consumers who sought to avoid a GM presence label and consumers who sought to gain a GM absence label Censored regressions were estimated because difference in WTP within a label treatment could vary from 10 to 10 for granola bars and from 5 to 5 for apples Difference in WTP was regressed on demographic characteristics exposure to information concern about GM foods and indicator variables that account for the randomly assigned label treatment To examine the determinants of WTP to avoid GM labels differences in WTP for label treatments one through three were used to create the dependent variable WTP to Avoid QR Code GM Label and diff

温馨提示

  • 1. 本站所有资源如无特殊说明,都需要本地电脑安装OFFICE2007和PDF阅读器。图纸软件为CAD,CAXA,PROE,UG,SolidWorks等.压缩文件请下载最新的WinRAR软件解压。
  • 2. 本站的文档不包含任何第三方提供的附件图纸等,如果需要附件,请联系上传者。文件的所有权益归上传用户所有。
  • 3. 本站RAR压缩包中若带图纸,网页内容里面会有图纸预览,若没有图纸预览就没有图纸。
  • 4. 未经权益所有人同意不得将文件中的内容挪作商业或盈利用途。
  • 5. 人人文库网仅提供信息存储空间,仅对用户上传内容的表现方式做保护处理,对用户上传分享的文档内容本身不做任何修改或编辑,并不能对任何下载内容负责。
  • 6. 下载文件中如有侵权或不适当内容,请与我们联系,我们立即纠正。
  • 7. 本站不保证下载资源的准确性、安全性和完整性, 同时也不承担用户因使用这些下载资源对自己和他人造成任何形式的伤害或损失。

评论

0/150

提交评论