Beyond Organic and Fair Trade? An Analysis of Ecolabel Preferences in the United States

Authors

  • Philip H. Howard,

    1. Department of Community, Agriculture, Recreation and Resource Studies
      Michigan State University
    Search for more papers by this author
    • We are grateful for the participation of the survey respondents, and assistance from Jan Perez. We also express appreciation to three anonymous reviewers for helpful suggestions. Funding for this project was provided by the USDA as part of “A consortium-based program for sustainable agriculture along the Central Coast of California.”

  • Patricia Allen

    1. Center for Agroecology and Sustainable Food Systems
      University of California, Santa Cruz
    Search for more papers by this author

Abstract

The success of alternative food initiatives indicates increasing interest in changing the way food is produced, processed, and sold. Ecolabels such as organic and Fair Trade have entered the mainstream marketplace, and other voluntary identifiers on products are emerging to address criteria not included in these successful initiatives. Little is known about consumer interests in these criteria, however. To anticipate the direction of food-system changes, as well as assist food producers to meet consumer demands, we conducted a national mail survey to assess preferences for criteria that go “beyond” (or could complement) organic and Fair Trade. We utilized a forced-choice paired-comparisons question format to rank five possibilities (humane, local, living wage, small-scale, U.S. grown) that might feasibly be implemented by food producers. Local was the most popular choice, although humane also received a high level of support. Multilevel logistic regression indicated that local was preferred by rural residents, and that humane was preferred by frequent organic consumers and high-income households. Survey respondents also chose product labels more frequently than other potential sources of information about their food. Preferences for local and humane ecolabel criteria should be placed in perspective, as consumers expressed much higher levels of interest in the more individualized concerns of safety and nutrition. The results suggest, however, that consumers are interested in a food system that addresses broader political and ethical values, which has implications for production, marketing, and movement building for sustainable food systems.

Introduction

Consumer interest in alternatives to the conventional food system is growing rapidly, as indicated by increased participation in alternative food institutions and by expanded interest in purchasing foods that meet certain criteria. The number of farmers' markets in the United States, for example, has increased more than 70 percent in the last decade (U.S. Department of Agriculture [USDA] 2008), and community supported agriculture (CSA) projects have expanded from 1 to more than 2,000 in the last 25 years (LocalHarvest 2008). Another way that consumers are making their desire for a better food system known is by purchasing food with ecolabels. Ecolabels are voluntary identifiers on goods and services that represent ecological or ethical criteria. Organic and Fair Trade certified foods are now found in large-scale, mainstream outlets, such as Wal-Mart and McDonald's. These market changes indicate that an increasing number of consumers are making different choices about the food they buy and how they buy it.

While increased participation in alternative food institutions is a major indicator of criticism of and interest in changing the food system, venues such as farmers' markets and CSA are limited in their ability to create significant change and be accessible to a wide range of people (Stagl 2002). The other clear indicator of interest in change, however, that of purchasing food with anticonventional criteria as specified on food labels, has large potential for growth. For example, organic products experienced growth rates approaching 20 percent annually from 1990 to 2008, and now make up more than 3.5 percent of total food and beverage sales in the United States (Martin 2009; Organic Trade Association 2007, 2008). Fair Trade, which embodies criteria such as a fair price for farmers and social/environmental standards, was introduced to the United States later than the organic label, and currently has a market share of less than 1 percent. Fair Trade has achieved even higher growth rates, sometimes exceeding 100 percent per year (Fair Trade Federation 2008). Other food ecolabels, which address criteria other than those already embodied in the organic and Fair Trade labels, are also entering the market. These include domestic versions of Fair Trade, local or regional labels, and labels that certify humane treatment of animals.

There has been insufficient attention to understanding consumer preferences among these newer, emerging ecolabels (McCluskey and Loureiro 2003). Such information is critical because as the food crisis deepens, public support for the current food system is declining (Evans 2007). Recent events and changes, such as food-contamination incidents, increasing oil and food prices, and concerns about obesity, indicate the need for changes in the American food system. The direction of these changes is uncertain, however, and knowing people's preferences can provide guidelines for the future food system. In addition, this information can help producers, marketers, and change agents respond to and engage with the preferences of consumers. For entrepreneurs, this assistance can translate into increased economic viability.

In this article, we review the literature on political and ethical consumerism to highlight the potential for ecolabeling to achieve food-system changes, and the role that social scientists can play in this process. We then discuss the methods we used to determine consumer interests in ecolabeling strategies that go beyond the exemplars of organic and Fair Trade. This included addressing the following questions: (1) which among a narrow set of emerging ecolabel criteria (living wage, local, humane, small-scale, and U.S.-grown) are most preferred? (2) what is the level of support for a labeling strategy compared to other sources of information? and (3) which broader food-system issues generate the most consumer interest? We present the results from a national, random sample mail survey, which indicates interest in integrating a number of political and ethical values more fully into the food system, as well as strong support for the strategy of ecolabeling. The results also suggest that emerging ecolabel criteria focusing on local production and the humane treatment of animals will appeal to the largest numbers of consumers.

Political and Ethical Consumerism

Concomitant with the decline of traditional forms of political participation is the rise of new forms, such as political and ethical consumerism. This activity involves expressing values through individual purchasing/nonpurchasing behaviors to express political or ethical goals, or both. Activism that relies on consumer action is increasing, as indicated by the growth of boycotts (collective action to withhold purchases until a specific demand is met), ethical investing, and ecolabels since the 1960s, as well as evidence from panel survey data (Stolle, Hooghe, and Micheletti 2005; Vogel 2003). These efforts attempt to subsume narrow market logics to wider and more democratic goals (Hinrichs 2000). While economic action for political and ethical ends is not new, recent decades have seen the mobilization of people to achieve such objectives specifically through their identities as consumers (Barnett et al. 2005). Theoretical interest in “reflexive consumption” as a way to overcome a production-consumption divide in agrifood studies has accompanied these trends (DuPuis 2000; Goodman and DuPuis 2002). This in turn has led to further empirical support for the notion that a minority of ordinary consumers consciously attempt to support alternative forms of production, and view their buying decisions within a framework of collective behavior (Moore 2006; Seyfang 2006; Vermeir and Verbeke 2006).

An advantage of such forms of action is that they are well understood and available to almost everyone (Arnould 2007). Although individual income differences result in unequal power to apply these strategies, they can nonetheless be incorporated into people's everyday activities, as compared to lobbying and formal political action. While the strategy of political and ethical consumerism does contain internal contradictions (such as the incompatibility of increasing consumption and ecological sustainability) and emphasizes individual action (Hilton 2007; Johnston 2008), it has, in certain cases, been quite effective in achieving social movement goals. Examples include a Rainforest Action Network campaign that ended Mitsubishi's use of old-growth timber in the 1990s, a grassroots coalition effort to convince retailer Trader Joe's to remove genetically engineered ingredients from its private-label products in 2001, and the Coalition of Immokalee Workers–led movement that pressured fast-food chains to increase piece rates for tomato harvesters in the late 2000s.

Exit, Voice, and Innovation

One source of increasing consumer dissatisfaction with the contemporary food system is the concern that the quality of the food system and of the food available for consumption have declined. Whether it is because food quality has declined in some objective measure or there is a larger gap between desired food quality and available food quality, there is nonetheless a perception of decline. Consumers have essentially two options when faced with declining quality, according to political economist Albert Hirschman's classic framework: exit or voice (1970). Exit involves breaking a relationship with an organization, such as no longer buying their products, while voice involves expressing complaints. Either choice could send the organization a signal of dissatisfaction, but each has its own advantages and disadvantages. Exit provides very limited information for organizations. When demand is highly elastic firms may fail to even recognize such signals until they are already on the verge of failure (Dowding et al. 2000). Voice typically requires greater effort for consumers, and is seldom taken seriously by economic organizations, as those making complaints may be considered unrepresentative. Hirschman suggests that quality is optimized when there is a balance of both exit and voice, but examples are rare. Because eating is essential for survival, yet there is also a limit to what people can eat, overall demand for food is fairly inelastic. Exit may therefore be impossible unless viable alternatives to declining quality exist.

At the intersection of exit and voice is the political and economic tactic of boycott. Boycotts have been successful in some high-profile cases, such as the United Farm Workers–led boycott of grapes beginning in the 1960s and the resulting unionization of farms. In most cases, however, boycotts have had a more limited impact. It is very difficult to organize enough diffuse consumers to give up established purchasing behaviors in order to achieve the desired goals (Friedman 1999).

A limitation of Hirschman's framework is that it deals only with negative responses to declining quality. Consumers may also respond positively to improving quality, therefore entry, in addition to exit, should be included as an option (Neuner 2000). At the intersection of entry and voice is the opposite of a boycott, termed a “BUYcott” (Friedman 1996). It is “positive” in that the strategy asks consumers to do something (buy) as compared to not do something (not buy). In a BUYcott, consumers use voice to express their support of positive changes, while also sending this signal through purchases. Organized efforts to buy union-made products are one example. The growth of ecolabels, however, has most effectively demonstrated the possibilities of BUYcotts in transforming the food system. (See Figure 1 for a graphic presentation of these choices for consumers.)

Figure 1.

Conceptual Framework

BUYcotts appear to be even more effective than boycotts in changing organizational behavior (Friedman 1996), but only if there is an option for consumers to support. What if consumers want products that are not available in the marketplace? One way to determine unmet consumer demands is to bridge information gaps between production and consumption by determining the feasibility and marketability of innovations (Neuner 2000). Social scientists can play a critical role in this process by conducting consumer research and communicating the findings to producers. Such research differs from traditional industry-sponsored market research in that the focus is on broad criteria of interest to consumers, such as sustainability, rather than a narrow focus on increasing profits (Dawson 2003).

Ecolabels

As noted above, ecolabeling is a strategy that allows for the combination of both entry and voice. It is currently the most common type of positive political and ethical consumerism (Vogel 2003), therefore we focused our efforts to determine the feasibility and marketability of food-system innovations on this approach. Ecolabels provide consumers with information about the practices used in goods and services that would otherwise remain hidden. In this sense, they are attempts to defetishize the commodity form (Allen and Kovach 2000). For consumers to have confidence in the claims made by an ecolabel, they must be certain of their veracity and enforcement. Therefore, the most successful ecolabels rely on a third-party certification process—employing an organization without a direct financial interest in the outcome to verify the claims that are made. Ecolabels have been developed for a number of industries, including food, forestry, fisheries, energy, and tourism (Boström and Klintman 2008). Their growth has been encouraged by social-movement organizations, and in some cases involved their direct participation (Busch and Bain 2004).

The market success of ecolabels has helped to achieve some of these movements' political and ethical objectives. Increasing sales of certified organic products, for example, have encouraged alternative production practices on a growing number of acres worldwide, leading to the elimination of synthetic pesticide and fertilizer applications on these lands (Lockie et al. 2006). Farmers who participate in Fair Trade certification have reported better and more stable livelihoods than farmers who participate only in conventional markets (Bacon 2005; Jaffee 2007). The transformative potential of ecolabels should not be overestimated, however, as market pressures tend to eventually erode standards that present obstacles to capital accumulation (Allen and Kovach 2000). Most analysts who study alternative food initiatives have concluded that ecolabels have inherent limitations and should not be relied on as the only strategy to achieve positive food-system changes, but nevertheless remain one effective tactic as part of a more comprehensive approach (Getz and Shreck 2006; Guthman 2004; Lyon 2006; Raynolds and Murray 2007).

While increasing rapidly in popularity, ecolabels are available for only a small number of criteria and a tiny percentage of all foods sold. As a result, few consumers who wish to do so can express their political and ethical goals for all of their food purchases. At the same time, a growing number of producers have expressed interest in developing new ecolabels. Ironically, for example, the success of the organic ecolabel has in some cases marginalized the producers who pioneered it. Some small-scale producers are dropping out of USDA organic certification and developing “beyond organic” or “post-organic” ecolabels to reemphasize values that have been excluded from the current meaning of organic, as well as to recapture lost price premiums (Goodman and Goodman 2007; Howard and Allen 2006; Lockie and Halpin 2005; Moore 2006). One such effort is the Mendocino Organic Network's “Renegade Certification” in California, which uses USDA organic standards as a baseline, but includes other criteria, such as regional production (Howard and Allen 2006). Other initiatives are not necessarily intended to supplant the organic label, but could potentially complement it with criteria not fully embodied in the USDA standards. Examples include the development of domestic counterparts to Fair Trade that would apply to U.S. products, not just those that are imported (Brown and Getz 2008; Jaffee, Kloppenburg, and Monroy 2004); Buy Fresh, Buy Local campaigns (Allen and Hinrichs 2007; Hinrichs and Allen 2008); a U.S. Grown label (U.S. Grown 2009); and various labels that embody the humane treatment of animals1 (e.g. Certified Humane, American Humane Certified, and Animal Welfare Approved). These efforts would benefit from greater consumer input, given that consumer demand is essential to their success.

With the interests of both consumers and producers in mind, our objective was to bridge existing information gaps, in order to increase the opportunities and possibilities of success for market-based collective action. To assist producers to understand consumer preferences for political and ethical practices, our efforts focused on three questions. Our primary question was: Among criteria that are of interest to consumers, as well as feasible to address by producers, which of these have the greatest market potential? To put this strategy in more comprehensive perspective we also asked: Do consumers prefer labels as a means of understanding more about their food? Which broader food-system interests do consumers view as most important? In addition, we explored demographics, behaviors, and attitudes associated with different preferences.

Methods

Survey Design

We developed a series of survey questions to address the questions above. To assess consumer interest in potential ecolabel criteria we developed a brief list of possibilities. We excluded safety and nutrition, although these are known to be important to consumers, because Food and Drug Administration restrictions and the scientific uncertainty surrounding claims made in these areas would make implementation quite difficult. We also excluded environmental concerns, as those that were most commonly identified in our previous focus-group and survey research, such as avoidance of synthetic pesticides and genetically engineered organisms, are already addressed by the organic label (Howard and Allen 2006).2 We chose five single criteria as potentially feasible for operations with interests in new ecolabels and defined them briefly as follows:

  • • Humane: meat, dairy products, or eggs come from animals that have not been treated cruelly.
  • • Living wage: provides above-poverty wages to workers involved in producing food.
  • • Locally grown: grown within 50 miles of the point of purchase.3
  • • Small-scale: supports small farms or businesses.
  • • U.S.-grown: grown in the United States.

We used a forced-choice paired-comparisons format, rather than a Likert scale, in order to develop an ordinal ranking of each respondent's preferences for these criteria and minimize the possibility of scoring them identically. This format was originally developed by psychologists, but is frequently employed in market research. The question asked respondents to imagine a pair of identical food products, each embodying different criteria, and then circle the one they would be most likely to buy, though they were also given a “neither one” option. This approach more closely resembles the types of decisions consumers make when making food purchases (Clarke, Bell, and Peterson 1999; Hanley, Mourato, and Wright 2001). We presented a series of 10 pairs of items to provide all possible combinations.

To determine the level of interest in labeling as a strategy compared to other formats for learning more about food, the survey also listed potential information sources (Table 1)—it instructed respondents to choose up to four (not ranked). To place the ecolabel criteria preferences in perspective with broad food-system interests, we derived eight topics from thematic analysis of focus-group conversations involving consumers.4 We asked respondents to rate their interest in these topics (in Table 1) on a scale from 1 to 10, with 10 meaning “a great amount of interest” and 1 meaning “none at all.”

Table 1. Lists of Information Sources and Food-System Topics
Information sources
Brochure or display at point of sale
Newspapers/magazine articles/books
Product labels
Radio
Talking to seller
Television program/videotape/DVD
Tours of farms and/or processing plants
Web pages/the Internet
Food-system topics
How far your food travels from where it is grown
How nutritious your food is
How safe your food is
The environmental impacts of your food
The influence of large corporations on food production and marketing
The treatment of animals involved in the production of your food
The wages or salaries of workers who grow, make, or sell your food
The working conditions of those who grow, make, or sell your food

Additional survey questions recorded demographics, purchasing behaviors, and attitudes. Prior to analysis, we classified household income into three categories: low (less than $35,000 per year), middle ($35,000 to $75,000 per year), and high (more than $75,000 per year). We used rural-urban commuting area codes to classify respondents into dichotomous residential categories, rural and nonrural, based on their zip code (Rural Health Research Center 2005). We defined frequent purchasing of local food as obtaining food from a household garden, farmers' market, CSA, or roadside stand at least weekly and frequent purchasing of organic as engaging in this behavior at least weekly. We rated agreement with the statement “When I buy products (not just food) I try to consider how my purchase will affect the environment” on a 7-point Likert scale, with 7 representing “strongly agree” and 1 representing “strongly disagree.” We derived this question from the literature on “perceived consumer effectiveness,” which suggests that those who feel they are able to make a difference with purchases are more likely to engage in ecologically sustainable behaviors (Roberts 1996).

This instrument was first implemented as a mail survey in the Central Coast region of California in 2004, with a 48.3 percent response rate. The results indicated that humane and locally grown were the first and second ranked choices, respectively, for ecolabel criteria. Product labels were the most preferred information source. While safety and nutrition were the top food-system concerns, consumers also expressed high levels of interest in ethical issues such as treatment of animals, environmental impacts, and social justice for workers (Howard 2006). Because residents of this region were some of the earliest supporters of organic agriculture in the United States, the applicability of these results to the rest of the country remained an important question.

Survey Implementation

We mailed the survey to 1,000 randomly selected respondents in all 50 states in 2006, with instructions that it be filled out by the primary food purchaser for the household. We employed the Tailored Design Method (Dillman 2000), with a prenotice letter, a $1 bill incentive accompanying the first survey mailing, a postcard reminder, and up to two replacement survey mailings to nonrespondents. The sampling frame comprised names and addresses purchased from a marketing firm, USADATA, a firm that we utilized in the previous survey with a very low rate of undeliverable mailings. The response rate for this survey was 50.7 percent, with 476 of the 1,000 surveys completed, and 62 returned as undeliverable.

Analytic Strategy

We first analyzed the ecolabel criteria preferences by determining the frequency of respondents who selected each option in each pair. Second, we used the ordinal rankings to determine the frequencies for the respondents' top choice (i.e., the criterion they preferred in every possible comparison). Finally, we performed multilevel logistic regression analyses of the top two choices in comparison to the other choices to examine the demographic and behavioral variables consistently associated with preferences for these labels. The “yes” or “no” response to a criterion versus each of the other four criteria composed the first level of the analysis in the model. The second level of the analysis comprised individual-level variables for each respondent. We performed analyses using the software HLM, a program for two- and three-level (hierarchical) linear and nonlinear modeling (Raudenbush et al. 2004). We used multiple imputation as the missing data strategy for independent variables in the analysis. This involves utilizing information from the existing data to generate plausible values for missing data, while also representing uncertainty by generating a small number of data sets, each with different imputed values (typically 3 to 5) (King 2002). Five data sets (m= 5) were generated using the program Amelia II (Honaker, King, and Blackwell 2006). These were combined for analysis with HLM's multiple-imputation function.

We summarized preferences for the eight information sources with frequencies and 95 percent confidence intervals, and interests in the eight food-system topics with mean scores and 95 percent confidence intervals for each question. We then analyzed these food-system topic scores with Ordinary Least Squares (OLS) regression to determine the demographic and behavioral variables associated with these interests.

Results

Table 2 shows the distribution of variables included in the analyses. Comparison of the demographics of the respondents to U.S. Census data indicated that they were more likely to be older, to be of non-Latino white ethnicity, and to have higher educational attainment and household incomes than the general population (see Howard and Allen 2008 for more details). Respondents were also more likely to be women, though this was not surprising as women are more likely to be the primary household food purchaser. The mean age was a little over 50. Household income was divided fairly evenly into low, medium, and high categories. As expected, this variable had the highest amount of missing data. Rural residents accounted for a small minority of the sample. One-third of respondents reported purchasing local foods frequently, while a smaller percentage reported purchasing organic foods frequently. The mean score for considering the environment when making purchases represented slight agreement.

Table 2. Descriptive Statistics for Variables in the Analyses (N=476)
 PercentageMean (S.D)Percentage Missing
Categorical variables   
Gender  5.3
 Men41.0  
 Women53.8  
Income  11.3
 Low income (<35 K)26.9  
 Middle income (35–75 K)34.9  
 High income (≥75 K)26.8  
Local purchasing  4.6
 Frequent33.6  
 Not frequent61.8  
Organic purchasing  4.0
 Frequent17.4  
 Not frequent78.6  
Residence  8.4
 Rural13.4  
 Nonrural78.2  
Continuous variables   
 Consider environment when purchasing 4.2 (1.8)6.7
 Age (years) 51.6 (16.0)7.4

Table 3 shows the results of each paired comparison for the sample as a whole. The ecolabel chosen by the larger percentage of the sample is listed in the “winner” column with the ecolabel chosen by the smaller percentage of the sample in the “loser” column. This table also indicates the percentage choosing neither option. Viewed this way, both local and humane won three of four comparisons, although local won these by the largest margins. Local had an advantage in that one of the comparisons was with a more distant geographic criterion, U.S. grown. More than three times as many respondents chose local in this pair. Humane won three comparisons by very narrow margins, but interestingly, it was selected by the largest number of respondents in a “victory” over local. Pretesting the instrument indicated these types of decisions were very difficult for some respondents, and the percentage selecting “neither” provides an indication of which comparisons posed the most challenging trade-offs. The comparisons between humane, living wage, and small-scale all had “neither” responses of 8.4 percent or above. The comparisons involving the geographic criteria, local or U.S. grown, all had “neither” responses of 5.3 percent or lower.

Table 3. Results of Paired Comparisons of Potential Ecolabel Criteria (N=439)
“Winner” (%)“Loser” (%)Neither (%)
Local (72.9)US (21.9)5.2
Local (65.1)Small (29.6)5.3
Local (58.1)Living (36.9)5.0
U.S. (53.1)Humane (44.0)2.9
U.S. (51.5)Living (44.0)4.5
Small (51.3)U.S. (44.4)4.3
Humane (49.2)Local (46.9)3.9
Humane (47.8)Living (42.6)9.6
Humane (46.7)Small (44.4)8.9
Living (46.7)Small (44.9)8.4

Figure 2 shows the percentage of respondents selecting a criterion as their top-ranked choice. The leading criterion was local, preferred by a little over a quarter, followed by humane, living wage, U.S. grown, and small-scale in last place, with less than 1 in 20 choosing this label over the others in every comparison. These results are different from the California Central Coast survey. The Central Coast respondents ranked humane over local (30.5 percent and 22.0 percent respectively), and also expressed lower support for U.S. grown (5.9 percent) (Howard and Allen 2006).

Figure 2.

Preferences for Post-Organic Ecolabel Criteria (N= 439)

Table 4 shows multilevel logistic regression results for local in comparison to other ecolabels. Respondents were significantly more likely to choose local when paired with U.S. grown, small-scale, and living wage, in comparison to the default pairing with humane. They were three times more likely to select local when it was paired with U.S. grown than when paired with humane, for example. Only rural residence was associated with a preference for local in comparison to the other labels at p < .05. Controlling for the other variables in the analysis, rural residents were nearly twice as likely as nonrural residents to prefer local over the other label options.

Table 4. Multilevel Logistic Regression Model of Choosing Local versus Other Potential Criteria (N=439)
 Odds RatioVariance componentCoefficientS.E.S.D.
  • = p < .05,

  • *** 

    = p < .001

Fixed effects     
Intercept  −0.2980.251 
Gender     
 Men (default)     
 Women1.016 0.0160.152 
Age     
 Years (continuous)1.001 0.0010.005 
Income     
 Low income (default)     
 Middle income1.244 0.2180.175 
 High income1.166 0.1530.195 
Residence     
 Nonrural (default)     
 Rural1.823 0.5980.224* 
Purchasing behaviors     
 Frequent organic1.044 0.0430.200 
 Frequent local1.122 0.1150.161 
 Consider environment (7 pt. scale)0.978 −0.0220.043 
Comparisons     
 Humane (default)     
 Living Wage slope1.585 0.4610.136*** 
 U.S. Grown slope3.093 1.1290.144*** 
 Small-scale (slope)2.111 0.7470.139*** 
Random effects     
Intercept variance 1.249  1.118***

Table 5 presents results of the multilevel logistic regression model for choosing humane over other potential ecolabels. Controlling for the variables in the model, respondents did not differ significantly in their preferences for humane when paired with each of four other criteria. Respondents who reported frequently purchasing organic foods were more than twice as likely to choose humane in comparison to other labels. High-income respondents were almost twice as likely to choose humane.

Table 5. Multilevel Logistic Regression Model of Choosing Humane versus Other Potential Criteria (N=439)
 Odds RatioVariance componentCoefficientS.E.S.D.
  • = p < .05,

  • ** 

    = p < .01,

  • *** 

    = p < .001,

  • = p < .10

Fixed effects     
Intercept  −0.7950.284** 
Gender     
 Men (default)     
 Women1.291 0.2550.175 
Age     
 Years (continuous)0.994 −0.0060.006 
Income     
 Low income (default)     
 Middle income1.314 0.2730.201 
 High income1.642 0.4960.225* 
Residence     
 Nonrural (default)     
 Rural0.890 −0.1190.250 
Purchasing behaviors     
 Frequent organic2.206 0.7900.233** 
 Frequent local0.972 −0.0290.187 
 Consider environment (7 pt. scale)1.067 0.0650.049 
Comparisons     
 Local (default)     
 Living Wage slope0.945 −0.0560.138 
 U.S. Grown slope0.785 −0.2420.138+ 
 Small-scale (slope)0.891 −0.1150.138 
Random effects     
Intercept variance 2.075  1.440***

Figure 3 shows the mean scores of preferences for information sources, with error bars indicating their 95 percent confidence intervals. Product labels were the top choice, selected by over three-fourths of respondents. A brochure or retail display was a close second, also selected by more than three-fourths, and its 95 percent confidence interval overlapped with that of product labels. These results support the use of ecolabels in efforts to encourage political and ethical consumerism, although other means of providing information at the point of purchase may also have strong appeal.

Figure 3.

Preferences for Information Sources (N= 472)

Interest in the remaining options dropped to slightly over half for print sources, such as books and magazines, and less than half for web sources. Despite the growing popularity of television programs that focus on what goes on behind the scenes in food production (e.g., the Food Network's Unwrapped), only a third of respondents selected this option. Tours and talking to the seller were among the least preferred options, both less than 20 percent. These results are also very similar to those from the California Central Coast survey, although a lower proportion of Central Coast residents (26.3 percent) selected video as an information source.

Figure 4 displays interests in specific food-system topics. Safety received the highest mean score, over 9, with Nutrition a clear second at almost 9. Three topics had mean scores between 7 and 7.5 and overlapping 95 percent confidence intervals: treatment of animals, environmental impacts, and working conditions. These were followed by influence of large corporations and wages with scores under 7. Interestingly, wages scored significantly lower than working conditions. How far food travels received the least amount of interest, at slightly over 5. The results are quite similar to those in the previous California Central Coast mail survey, both for the mean scores and rank order of these topics. Although the political and ethical topics rated lower than personal concerns of safety and nutrition, the scores suggest a relatively high degree of interest in issues that typically receive little attention in the conventional food system.

Figure 4.

Level of Interest in Food-System-Related Topics (N≥ 449)

Tables 6 and 7 show the results of OLS regression models for each of the eight food-system topics. Considering the environment when making purchases was the variable most strongly and consistently associated with higher levels of interest in these topics. Women were more interested than men in most of the topics as well. Increasing age was associated with higher interest in the influence of corporations, but lower interest in environmental impacts. High-income respondents were less likely to be interested in working conditions than those with low incomes. Rural residents expressed less interest in wages than nonrural residents. Frequent organic consumers were more interested in both environmental impacts and treatment of animals. This provides further evidence that organic consumers may be more willing to support a humane ecolabel than other consumers.

Table 6. Association of Demographic and Behavioral Variables with Interest in Food-System Topics (N449), Standardized Betas for Ordinary Least Squares Regression of Interest in Food-System Topics, First Four Topics
 SafetyNutritionTreatment of AnimalsEnvironmental Impacts
  • = p < .05,

  • ** 

    = p < .01,

  • *** 

    = p < .001,

  • = p < .10

  • Note: N is greater than or equal to 449, depending upon the model, due to missing data.

Gender    
 Men (default)    
 Women0.088+0.064*0.147*0.117*
Age    
 Years0.055−0.0450.053−0.023*
Income    
 Low income (default)    
 Middle income0.0150.0300.0250.053
 High income−0.0120.081−0.096+−0.089+
Residence    
 Nonrural (default)    
 Rural−0.041−0.090+−0.036−0.022
Purchasing behaviors    
 Frequent organic0.0750.083+0.118*0.100*
 Frequent local0.0040.0320.0370.033
 Consider environment0.091+0.190**0.254***0.377***
Model    
Adjusted R20.020.070.140.21
F2.09*5.56***10.23***15.90***
N463449470449
Table 7. Association of Demographic and Behavioral Variables with Interest in Food-System Topics (N449), Standardized Betas for Ordinary Least Squares Regression of Interest in Food-System Topics, Last Four Topics
 Working ConditionsInfluence of Large corporationsWagesHow Far Food Travels
  • = p < .05,

  • ** 

    = p < .01,

  • *** 

    = p < .001,

  • = p < .10

  • Note: N is greater than or equal to 449, depending upon the model, due to missing data.

Gender    
 Men (default)    
 Women0.128*0.0730.186*0.077
Age    
 Years0.0590.158***0.085+0.047
Income    
 Low income (default)    
 Middle income−0.056−0.004−0.018−0.032
 High income−0.162**−0.011−0.096+−0.023
Residence    
 Nonrural (default)    
 Rural−0.099+−0.048−0.124*−0.021
Purchasing behaviors    
 Frequent organic0.0500.0700.079+0.055
 Frequent local0.0650.079+0.0010.126**
 Consider environment0.225**0.289***0.221**0.270***
Model    
Adjusted R20.130.150.140.12
F9.52***11.05***10.34***8.77***
N467449468449

The variance explained in the models for safety and nutrition were 2 percent and 7 percent, respectively. Because nearly all survey respondents expressed a great amount of interest in these two topics, the independent variables were quite poor predictors of the small amount of variation in this sample. For the remaining topics the models explained 12 percent to 21 percent of the variance. These R2 values are at expected levels, as previous surveys have reported demographic variables tend to be only weakly associated with ecolabel preferences or ecologically conscious behaviors (Gatersleben, Steg, and Vlek 2002; Roberts 1996; Wessels, Johnston, and Donath 1999).

Discussion

Local was the postorganic ecolabel criterion receiving the highest level of support, as it was the top-ranked choice for 26.7 percent of respondents. Due to the nature of the question, which forced respondents to choose from a limited number of options, this ranking should be interpreted with some caution. Those who are less interested in political and ethical consumerism may have selected this criterion for other reasons. Consumer interest in buying local is multifaceted, as indicated by a spokesperson for the northeast supermarket chain Hannaford Brothers, who said, “Our research tells us consumers have about five or six reasons for wanting local: freshness and taste; keeping farmland in the community and having open spaces; a desire to be close to the food source and know where it comes from; support of local farmers and keeping money in the community. Embedded in all of this is concern about food safety. All this creates pretty powerful interest” (Burros 2008).5 The preference for local as an ecolabel criterion may therefore reflect personal concerns about safety, nutrition, or taste for some respondents, even though for others this preference may result from wider political and ethical concerns. This may actually provide a marketing advantage, as research suggests the success of the organic label is partially due to consumer perceptions that these foods are “healthier” than their conventional counterparts (Oberholtzer, Dimitri, and Greene 2005). The depth of this support for local needs further investigation, however.

While local appears to have widespread appeal, the pattern of the results suggests that producers should also give strong consideration to a humane ecolabel, which was the top choice of 22.1 percent of respondents. Humane won the same number of paired comparisons as local (three each), and narrowly won the direct comparison between the two. In addition, humane was more likely to be preferred by frequent organic consumers, who have demonstrated a willingness to pay price premiums, as well as high-income consumers, who are better able to afford price premiums. Preferences for humane are also difficult to disentangle from personal concerns, however; our previous focus-group research indicated that consumers worried about the human health impacts of inhumane practices (Howard 2006).

The lower rankings of living wage, U.S. grown, and small-scale do not mean that consumers are uninterested in these criteria, only that the respondents expressed stronger preferences for local and humane criteria. Surveys have suggested a willingness to pay more for a label that embodies a living wage and safe working conditions (Howard and Allen 2008; McCluskey and Loureiro 2003), and demonstrated overwhelming (≥80 percent) support for country-of-origin labeling requirements (Schupp and Gillespie 2001; Solomon 2007). Willingness to pay for country-of-origin labels has also been found to be high in surveys, but low in more realistic experimental auctions (Loureiro and Umberger 2003; Umberger et al. 2003). Consumer interests in these lower ranked criteria may therefore be less likely to translate into mainstream market success if they also require price premiums (although they may prove to be viable niche markets).

The results of this study may help guide prioritization when multiple criteria are being considered for an ecolabel. The nonprofit Food Alliance label currently includes standards for environmental stewardship, safe and fair working conditions, and humane treatment of animals, for example (Food Alliance 2008). Other multicriteria labels are also in development (Leonardo Academy 2009). One impetus for combining multiple criteria is to simplify certification requirements for producers, while another is to simplify choices for consumers. A potential objection to the emergence of new ecolabels, particularly those with a single criterion, is the concern that consumers will suffer from “label fatigue” by having too many options to choose from (Goodman 2004). Empirical research remains to be conducted to assess the extent to which label fatigue is actually a factor in consumer behavior. The fact that children in first grade can distinguish among 200 corporate logos (McNeal 1999) suggests that such fears may be exaggerated.

This study has several limitations. One is that we did not examine willingness to pay for these criteria. Future research should utilize methods such as experimental auctions and conjoint analysis to obtain these estimates (e.g., Conner and Christy 2004; De Pelsmacker, Driesen, and Rayp 2005). We also did not directly compare these criteria to organic. Although organic is likely to have an advantage due to being more available and well known, such a comparison would provide a better indication of the current market potential for other criteria. Another limitation involves using the term small-scale rather than family farm. Although small-scale could refer to multiple stages of the food system, not just production, invoking the family-farm frame may have more resonance with consumers due to the pervasiveness of agrarian ideologies in the United States (Howard and Allen 2006). An organization in Chicago, for example, is piloting a “family farmed” label (FamilyFarmed.org 2008).

Conclusion

Social scientists have an opportunity to not only apply traditional market research tools in determining the feasibility and marketability of innovations but to develop new methods of bridging information gaps between producers and consumers. While corporations are increasingly applying qualitative methods such as ethnography, more democratic approaches such as participatory-action research are needed. Fostering greater dialogue and deliberation between producers and consumers is another possibility. Methods for doing so might include deliberative polling, citizen's juries, or holding face-to-face meetings using participatory processes such as Open Space Technology (Owen 2008). This study indicates that consumers are interested in changing the direction of the food system to one that places a greater emphasis on political and ethical values, particularly local production and the humane treatment of animals. Producers committed to ecological sustainability and/or social justice can utilize this information to adjust their production or marketing practices to be congruent with consumer preferences found in this study. More substantive changes could be catalyzed through the greater engagement of both producers and consumers in food-system planning and social movements to create a more sustainable food system.

Footnotes

  1. 1 While humane criteria are included in the USDA organic standard, they currently include only vague provisions, such as “access to pasture.” These provisions have been criticized for being insufficiently enforced by the USDA, such as allowing feedlot dairies with up to 10,000 cows in arid regions (Guptill 2009). Nongovernmental third-party certifications have been developed with more detailed requirements than the USDA organic standard.

  2. 2 A small minority of consumers are interested in environmental criteria not included in organic standards, such as the energy consumption and waste involved in food transportation and packaging, but these issues were not raised as frequently as the items tested here.

  3. 3 There is no consensus on the definition of local. We used a 50-mile distance here because it was the most common description provided by participants in previous focus-group research (Howard 2006).

  4. 4 We conducted five focus groups with participants recruited from retail food outlets in the Central Coast of California in 2004. For more details, see Howard 2006.

  5. 5 Our own consumer research supports this analysis, although we identified one additional motivation, the reduction of fossil-fuel consumption for the transport of food.

Ancillary