Introduction
Direct markets, including farmers' markets, have been increasing in sales volume, number of customers and the number of farmers that depend on this type of outlet throughout the 1990s and 2000sReference Payne1, as well as over a longer time span from the post-World War II eraReference Wann, Cake, Elliot and Burdete2–Reference Shakow4. Ironically, these local markets have grown in the context of globalization, perhaps as consumers seek an identity to otherwise anonymous and geographically distant productsReference Lyson and Green5–Reference Savage, Bagnall and Longhurst7. The marketing for these local products, or products with characteristics of a specific place, differ from commodity goods because consumers who seek out these products are looking for attributes not found in globally produced commodity goods, such as higher quality (freshness, for food), authenticity and a sense of local communityReference Payne1,Reference Warde6.
The development of local agricultural markets has the potential to add value to farm sales as supply chains are shortened, as well as the potential to generate social and environmental benefits for the community, such as agricultural open space, a public good that contributes to the scenic value of communitiesReference Loomis, Rameker and Seidl8,Reference King and Anderson9. Likewise, farm business managers with a keen eye to the external benefits of their local production may be able to capitalize on the added value of marketing with local place identities10–12.
Surveys of vendors and consumers at farmers' markets were developed to test the hypotheses that:
1. in direct market situations, consumers and producers exert influences on each other that may alter the kind of product offered and/or the agricultural practices used to provide a product;
2. farmers' markets are inherently social places with consumer decisions being affected by such motivations as wanting to help farmers and supporting agricultural open space; and
3. consumers can be sorted into meaningful groups to help explain their purchase behaviors.
These surveys were implemented during 2 weeks in August 2004 after being reviewed by staff in the Division of Market Production and Development at the Maine Department of Agriculture.
Farmers in this survey sample engage in different practices, production decisions and marketing approaches than other farmers, likely due in part to proximity to urban populationsReference Shakow4, generally being small and medium-sized farm operationsReference Payne1, or having a desire to engage in entrepreneurial activities. Likewise, farmers' market consumers are also generally different from the typical produce consumerReference Wolf13. However, others have noted the economic significanceReference Guptill and Wilkins14 of these population subsets and their role in bridging urban–rural differencesReference Sharp and Smith15. Future applications of these concepts in marketing should recognize that these different consumer groups have sophisticated attitudes concerning place, community and their personal identities and derive value from these interactions because of their of authenticity and personal relationships.
Literature Review and Background
Direct marketing and local production
Local markets offer producers that cannot compete effectively through economies of scale production an opportunity to compete by diversifying, differentiating their products, entering specialty or niche-markets, value-adding on farm and assuming a greater share of distribution costs internallyReference Roth16. However, institutional barriers within state governments and agricultural land grant universities, as well as farmer perceptions, that emphasize volume production may have to be overcome to shift agricultural production to maximize value per acre rather than yield per acre. Despite the institutional, cultural and infrastructural barriers (e.g., lack of access to rural value-adding businesses like butchers, packagers and creameries)Reference Shorto17, there has still been a trend in the emergence of local and regional food systemsReference Lyson and Green5,Reference Guptill and Wilkins14,Reference Hinrichs18, indicating that there is an incentive for producers to enter local direct markets, the rate of which may be economically suboptimal if there are barriers to entry.
The challenge is for farmers to tap into that local demand by marketing to consumers ‘who think there's nothing like eating a fresh egg, who like knowing where their food came from’Reference Shorto17. Through the face-to-face interactions typical for direct markets, producers receive feedback on new products and develop personal connections with customersReference Roth16,Reference Gilman19. The link between place, product and purchase is salient to the goals of farmland preservationReference Kirby20,Reference Daniels21. For example, in the United Kingdom, The Countryside Agency links consumer food purchases and rural tourism with land preservation through their ‘Eat the View’ promotion11. While food quality remains the most important motivator for consumer purchases, other attributes of the marketplace including food production method, a locally grown product and wanting to help farmers are important to consumersReference Kuches, Toensmeyer, German and Bacon22,Reference Govindasamy, Italia, Zurbriggen and Hoaain23.
Farmers' markets are social places and offer a different type of shopping experience than supermarkets or other retail outletsReference Pyle3,Reference Shakow4. Lyson et al. Reference Lyson, Gillespie and Hilchey24 describe this difference of shopping at farmers' markets:
As bridges between the formal and informal sectors of the economy, they enable individual entrepreneurs and their families to contribute to the economic life of their local communities by providing goods and services that may not be readily available through formal mass markets. Thus, they embody what is unique and special about local communities and help to differentiate one community from anotherReference Lyson, Gillespie and Hilchey24.
Payne echoes these values, that ‘many farmers’ markets also provide their customers with a unique shopping experience that is often lacking in today's fast-paced world’Reference Payne1. Likewise, social interaction at farmers' markets is perceived as being more festive, more friendly and personal than supermarketsReference Hughes and Mattson25.
Social embeddedness of economic interactions
Within farmers' markets, economic interactions coexist with social interactions through the context of community and place, making these exchanges socially and geographically embeddedReference Hinrichs18. GranovetterReference Granovetter26 and BlockReference Block27 develop a theory of social embeddedness to enrich the understanding of neoclassical economics by indicating that social interactions, such as those through trust relationships, help rational actors to mitigate risk. HinrichsReference Hinrichs18 first applied social embeddedness to direct agricultural markets and Lyson et al. Reference Lyson, Gillespie and Hilchey24 have noted specifically of farmers' markets that:
From a neoclassical standpoint, farmers' markets may not make good economic sense. From a community perspective, however, they can nurture local economic development, maintain diversity and quality in products, and provide opportunities for producers and consumers to come together to solidify bonds of local identity and solidarityReference Lyson, Gillespie and Hilchey24.
Thus, the consumer is socially embedded with the interactions of the market vendors and physically embedded through the consumption of products identified with the local physical environmentReference Jackson, Thrift and Miller28.
Warde's theory of communification is that in the context of globalization, consumers face a need for ‘re-embedding’ themselves in a sense of belonging, communicated through place, identity and communityReference Warde6. Mobilizing symbolic values of food, place and identity using such elements as tradition, nostalgia and authenticity has the ability to convey identities of place directly to consumers through their purchasesReference Savage, Bagnall and Longhurst7,Reference Munro, Edgell, Hetherington and Warde29,Reference Warde, Edgell, Hetherington and Warde30. As such, identity of place and community can play a role in farmland preservation and environmental qualityReference Jarosz31. However, the outcomes of these interactions, in terms of land preservation, environmental quality and other tangential benefits, require guidance and involvement from the communityReference Feenstra32.
Previous research
Previous surveys generally focus on identifying consumer characteristics and purchasing habitsReference Wolf13,33–Reference Brooker, Eastwood and Gray39, with few relating consumer and vendor data simultaneouslyReference Hughes and Mattson25. Only one identified survey had the explicit goal of studying the connection between customer perceptions, direct markets and open space, yet did not include data on direct market vendorsReference Manalo, Sciabarrasi, Haddad and McWilliam40. Data from this research identify that some consumers believe that their purchases of local products contribute to agricultural open space (Table 1).
Table 1. Attitudes used for cluster analysis (left-hand side) and significance (right-hand side).
* t-statistic significant at P<0.1.
** t-statistic significant at P<0.05.
*** t-statistic significant at P<0.01.
Of particular interest to this point is that Kezis et al. Reference Kezis, Gwebu, Peavey and Cheng33, another survey conducted in Maine, indicated that quality was a primary reason for shopping at the market and that the next most important reasons were to enjoy the market atmosphere and a desire to support local farmers. Manalo et al. Reference Manalo, Sciabarrasi, Haddad and McWilliam40, in a random telephone survey in New Hampshire, identified that 98% of respondents agreed that ‘buying local produce is an effective way to keep New Hampshire farms viable’. Both studies suggest that consumers have an interest in connecting their local food purchases to farm viability.
Through Kezis et al.'s Orono, Maine consumer surveyReference Kezis, Gwebu, Peavey and Cheng33, Kuches et al.'s random mail survey in DelewareReference Kuches, Toensmeyer, German and Bacon22, Brooker et al.'s two farmers' market surveys in TennesseeReference Eastwood38,Reference Brooker, Eastwood and Gray39, and Hughes and Mattson's Kansas consumer and vendor surveyReference Hughes and Mattson25, social processes were identified as key components of farmers' markets. For example, consumers generally learn of markets most often through word of mouth recommendationsReference Kuches, Toensmeyer, German and Bacon22,Reference Hughes and Mattson25,Reference Kezis, Gwebu, Peavey and Cheng33,Reference Brooker, Eastwood and Gray39. Kezis et al. Reference Kezis, Gwebu, Peavey and Cheng33 noted that shoppers generally bring others with them; only two-fifths shop alone. Also, Hughes and MattsonReference Hughes and Mattson25 indicate that the personal nature of the consumer and producer interaction gives public recognition for the vendor's labor, satisfying both emotional and psychological needs, and contributes to a consumer's sense of belonging.
In order to better describe consumers, responses from surveyed consumers were segmented based upon self-identified values through cluster analysis, a technique used in market researchReference Wedel and Kamakura41. While not applied previously to an agricultural marketing study, others have used cluster analysis to better understand how consumers make trade-offs between car prices, dealer relationship, and accessory packagesReference Odekerken-Schroder, Ouwersloot, Lemmik and Semeijin42; preferences for different types of food using Likert-scale attitudinal questionsReference Honkanen, Olsen and Myrland43; and for developing a new marketing approach for tourists who vacation in Canada's rural regionsReference Carmichael and Smith44. A potential shortcoming of this methodology is its reliance upon consumer attitudes and unobserved behaviors, however Odekerken-Schroder et al. indicated that results from their analysis were in line with observed behaviorsReference Odekerken-Schroder, Ouwersloot, Lemmik and Semeijin42.
Methods
Survey design and implementation
Surveys were designed with input of the Maine Department of Agriculture and a review of other implemented surveysReference Kezis, Gwebu, Peavey and Cheng33,35,Reference Hunt45,Reference Drake46 to gather information regarding the motivations and demographics of both customers and vendors of farmers' markets. Eight communities were selected as representative of both geographic location and farmers' market size. The locations included: the inland towns of Farmington, a small town with a University of Maine branch campus; Orono, home of the main University of Maine campus; Falmouth, a suburban community outside of Portland, and the coastal towns of Portland, Maine's largest city; Blue Hill, a small vacation town, Bar Harbor, located at the entrance to Acadia National Park; Camden, a quintessential regional service center; and Belfast, a small working port. Informal interviews with farmers, consumers, community members, non-governmental organizations, agricultural researchers at the University of Maine and North Carolina State University and government officials provided relevant background information. Vendor and consumer surveys were pre-tested at the Auburn, Maine farmers' market, and then revised before implementation. A second set of revisions was made to the vendor survey after implementation at the Farmington and Orono markets: data from the revised questions were kept separate throughout the analysis.
In order to solicit consumer respondents at the farmers' markets, a small card table was set up with the phrases ‘Farmer market research’, ‘Fill out a 5 minute survey’, and ‘Help this student graduate’ printed on paper to advertise the research. Also, the surveyor wore a T-shirt with ‘Survey Team’ on the front and ‘Help this student graduate. Fill out a 5 minute survey,’ on the back. A total of 216 consumers responded to the survey. Self-selection methods have the potential to introduce bias by attracting only people interested in the survey topic. In this case, this effect is considered negligible since people at the farmers' markets already support the market, as indicated by their presence.
Vendor surveys were distributed to all vendors, including non-farmers, with instructions to fill out the surveys when they had time between customers. A total of 65 farmers and 16 other vendors were surveyed, with one refusal.
Market segmentation
Cluster analysis is seen as an exploratory method of data analysis used to generate hypotheses about market segmentsReference Wedel and Kamakura41 and though it relies on self-identified attitudes and preferences, predicted behaviors will generally, but not exactly, conform to observed consumer decisionsReference Odekerken-Schroder, Ouwersloot, Lemmik and Semeijin42,Reference Honkanen, Olsen and Myrland43. An advantage of using cluster analysis is that grouping consumers by using attitudinal questions helps to explain why consumers shop at farmers' markets rather than who shops as identified through demographic data alone.
Consumers were segmented based on their responses to 15 attitudinal Likert scale questions (see Table 1 for values and exact question wording). Ward's method of hierarchical analysis assumes that individual responses are independent and do not overlap (i.e., one consumer's response is not predicated by another consumer's response)Reference Wedel and Kamakura41. This method compares the similarity of responses between individual consumer attitudes (represented in rows of data) by computing the average distance or linkage between values. Each row begins as its own ‘cluster’, and responses are grouped together until they are all in one cluster: the cut-off point for group selection is arbitraryReference Wedel and Kamakura41. However, following a logical procedure and applying statistics to back-check the difference between each cluster, the validity of these groups can be establishedReference Everitt, Landau and Leese47. The dendrogram in Figure 1 is a visual illustration of the process, conducted by using SAS 9.1.
Figure 1. Cluster dendrogram. Dashed line represents where the three-cluster selection was made. The branch at the left represents the Seasonal Shoppers, the middle branch represents the Lifestylers, and the far right branch represents the Utilitarians. Each individual vertical line at the x-axis represents one individual's preferences, with the horizontal lines representing the groupings of similarities between preferences.
No one method has yet emerged as a standard to determine the statistical relevance of these groups with various researchers using K-means clusteringReference Honkanen, Olsen and Myrland43, MANOVAReference Odekerken-Schroder, Ouwersloot, Lemmik and Semeijin42 and chi-squareReference Carmichael and Smith44. Cut-off groupings of five, four and three were developed. The three cluster grouping had the greatest statistical significance as identified by t-tests on the means for each of the 15 Likert question values, compared 1 to 2, 1 to 3 and 2 to 3. The other consumer responses were then added back into the dataset to allow for a more robust description of each group (Table 2), similar to the way Carmichael and SmithReference Carmichael and Smith44 added their consumers' demographics after clustering on ‘lifestyle’ values.
Table 2. Selected consumer cluster characteristics.
For convenience, each consumer segment was named based on characteristics of the three clusters, comprised of a total of 211 consumers. Cluster 1 is the largest group, representing 72% of the total survey population. These consumers have the highest levels of concern about factors external to food production, show an interest in visiting farms, talk with vendors the most and have the greatest diversity in their purchases, and thus are called the ‘Lifestylers’. Cluster 2 represents 26% of the survey population and is most concerned about food quality and less concerned about external factors. Because consumers in Cluster 2 tend to purchase mostly fruits, vegetables and flowers but few other goods, this group is termed the ‘Seasonal Shoppers’. Despite only having five respondents (2.5% of the survey population), Cluster 3 remains statistically different with marked differences in attitudes concerning external factors to food. With no respondents concerned about synthetic chemicals or desiring contact with farmers, and a focus on high food quality, consumers in Cluster 3 are called ‘Utilitarians’. It should also be noted that of the three groups, this was the cluster least interested in the social aspects of the markets, which may have also played a role in their self-selection decision to complete a survey.
Statistical analysis
Multiple regression analysis was performed on the dependent variables of farmer household income and consumer spending to identify significant influences on these variables. A binary choice probit model identified significant influences on consumers meeting someone they know at the farmers' markets. Correlations were used to identify the relationships between farmers changing their products to meet consumer demand. These analyses were performed using STATA Intercooled 8.2. Throughout the data, distributional tests including t-tests, Pearson's chi square and the Kolmogorov–Smirnov test (notated by D statistic) were applied. SPlus 6.2 was used to perform these distributional tests.
Results
Vendor and consumer profiles
Of the 81 vendors, 65 identified themselves as farmers with the remaining 16 being bakers, artisans, crafters, etc. Where relevant and when comparable data are available, farmer responses are separated as a subset of all vendor responses.
The surveyed farmers differed from the demographics of other farmers by age and education. The surveyed farmers' age distribution was skewed significantly (P=0.0002, D=1) towards the younger age groups and had a mean of 44 compared to a mean of 54 for Maine farmers48. Surveyed farmers also reported higher levels of educational attainment with 53% of market farmers completing 4-year degrees compared to 19% of other farmers in the Northern Crescent region49.
Vendors indicated that their most important motivations to sell at a farmers' market were to have a direct relationship with customers (62%), more profit at farmers' markets (36%), independence and control of their business (27%), and ‘other’ (27%) of which promotion and proximity of the market to their farm were the most frequent responses. No attempt was made to identify whether the desire to have a direct relationship with customers was primarily for social or economic motivations, or both.
Surveyed consumers differed significantly from the Maine population50 by education, income and gender. Surveyed respondents had significantly higher levels of educational attainment (D=0.7, P=0.0123), with three-fifths (60%) of respondents having earned a bachelor's degree or a graduate/professional degree compared to a quarter of Maine residents. Similarly, the respondent income distribution was skewed significantly to the higher end (D=1, P<0.001) compared to the Maine population. While all income ranges were represented, the mean respondent household income was approximately $70,000 compared to approximately $45,000 for the Maine population. A greater proportion of respondents were female, about three-quarters (74%) compared to half (51%) of the Maine population; however, this result is consistent with Kezis et al.'s Orono, Maine surveyReference Kezis, Gwebu, Peavey and Cheng33.
Consumers ranked freshness as the most important reason to shop at a farmers' market out of eight possible choices. Quality, availability of specialty products, helping the farmer and farmer contact were the next most important reasons and were statistically greater for the Lifestyler cluster compared to the Seasonal Shoppers, using a t-test (P>0.05 for each reason). Price was consistently the lowest or next to lowest reason to shop at the markets. Respondents were also asked in an open response format to list three words or phrases that explained why they shop at farmers' markets. More than half of the responses (57%) were factors that related to the intrinsic quality of the food (e.g. freshness, local, quality and organic), while social interaction and market qualities made up about two-fifths (43%) of responses (e.g. atmosphere, fun, friendly and support farmers).
Social interactions
Social factors, such as having fun at the market and interacting with farmers, are important aspects of shopping at farmers' markets. Nearly all respondents (98%) had fun at the farmers' markets. More than half of the survey respondents (59%) make the farmers' market a family event. Also about half of respondents (48%) indicated that they had visited a farm from where they had bought products, while four-fifths (82%) would like to do so. Lifestylers had statistically higher preferences for visiting farms (P<0.01), had visited farms more often (P<0.1) and interacted with vendors more (P<0.001) when compared to Seasonal Shoppers (Table 2). Topics of conversation with farmers differed significantly across clusters, with Lifestylers being the most interactive (Table 3).
Table 3. Meeting familiar people and topics of discussion with vendors.
Distribution of discussion topics for all clusters differ significantly at P<0.01 using the Kolomogorov–Smirnov test: Lifestylers/Seasonal Shoppers D=0.875, P=0.0025; D=1, P=0.0002; Lifestylers/Utilitarians D=1, P=0.0002; and Seasonal Shoppers/Utilitarians D=1, P=0.0002.
* Significantly different between Lifestylers and Seasonal Shoppers: talk with vendors t=2.338, P<0.05; production methods t=1.545, P<0.15; and sampling products t=1.900, P<0.1.
Word of mouth was an important method for people learning about established markets (in general, surveyed markets advertised infrequently, if at all). More than two-fifths of consumers (45%) found out about the markets by word of mouth and a third (32%) learned of the markets from roadside signs or driving by the market. However, for the newest market in the survey, the Falmouth market, word of mouth was responsible for only a small percentage (6%) of how people found out about the market, with driving by (38%), roadside signs (13%) and publications (13%) being more important factors.
Most consumers (94%) talk with vendors and more than four-fifths of consumers (81%) meet people they know at the market, other than vendors (Table 3). A binary choice probit regression model was used to identify factors that influence a consumer meeting someone she knows at the farmers' market (Table 4). Being a permanent state resident (a dummy variable), an increase in monthly shopping frequency, an increase in age and an increase in household size have positive effects on meeting someone a consumer knows at the market. Significant negative influences are an increase in the distance from home to the farmers' market and an increase in the region size which the respondent considered local.
Table 4. Probit model results of influences on consumers meeting others they know.
n=210.
All values significant at P=0.050 except 1 which is significant at P=0.056.
Wald chi-square 25.32, P=0.0003.
Both consumers and vendors identified Maine, or the state level, as the largest region which they consider local. About two-fifths of consumers (39%) and half of vendors (54%) chose the state level (Maine) as the largest region they would consider food to be ‘locally produced’. More than four-fifths of farmers (84%) identify their products as being local and three-fifths (60%) of farmers indicate that this is one of their main selling points (definitions of ‘local’ may differ by each market's bylaws).
Nearly all market farmers (95%) feel more involved with the community since they began selling at the farmers' market. Interaction with customers was the greatest source of this interaction (91%). More than half (58%) found that the interaction improved their overall product visibility. Two-fifths of farmers (41%) changed their products due to consumer demand, demonstrating that there is direct customer feedback for farmers. However, less than half of farmers (38%) felt that the community was supportive of providing the market location, often a public space or privately owned parking lot.
Economic implications
Market farmers had a median annual household income of $42,500 (the survey household income question was patterned after the US Census household income question), which was greater than those reported for Maine by the 2000 US Census (Table 5)50. Census data reported at the state level, the smallest geographic unit available for household income by occupation, were reported only as a median, so a statistical comparison could not be performed51. Compared with national data, market farmers had household incomes at least as high and generally higher than other farmers nationally (Fig. 2)52; however, these differences were not statistically significant, likely due to the variation within the relatively small sample (n=48, 35% non-response rate). Farmer and farm manager income levels both are reported since surveyed farmers were not separated by these occupational distinctions. The data suggest that the sub-set of market farmers have higher incomes than other farmers, perhaps related to their higher levels of education (see Table 5) or as a result of more market farmers being engaged in agriculture as a full-time occupation. The majority of these farmers indicated farming as their full-time occupation (68%) and a third (32%) indicated farming as a part-time occupation (n=63). Three-fifths of market farmers (61%) worked more than 2000 hours a year (or 38.5 hours a week on average) compared to less than a third (31%) of farmers in the Northern Crescent region49. Only one farmer indicated in a personal comment that this was his second career. The age distribution of these farmers was skewed to the younger age ranges, which reduces the possibility for a majority of the market farmers to be drawing upon a pension or retirement benefits from a previous career.
Figure 2. Market farmer income compared with other farmers nationally. Distribution of farmers' market farmers not statistically different from farm managers (D=0.3333, P=1) and farmers (D=0.3333, P=0.6) using the Kolmogorov–Smirnov test.
Table 5. Comparison of market farmer and Maine farm household income.
1 As classified by the US Census Bureau (1999 data).
Influences on producer income were modeled through regression analysis. The error distributions were found to be normal and unskewed, so error was assumed to be homoskedastic. The natural logarithm was applied to the dependent variable, income, which is in units of dollars. Effort, as represented by total hours worked per year, and experience, represented as number of years at the market, were important economic influences on farmer income. There were diminishing returns to total hours worked per year and years worked at the market, so a quadratic expression of those variables was included. [It should be noted that the variable (years at market)Reference Wann, Cake, Elliot and Burdete2 is not significantly different from zero, although its effect is not positive. Again, this variation is likely attributable to the small sample size.] One product type proved to have an important impact on farmer income, selling nursery or floriculture products, though it is imperfectly collinear with years at the market (correlation coefficient of 0.21). This is likely due to sampling frame error associated with variations in the small sample size (n=39); however, the performance of these factors generally conforms to economic theories of diminishing returns. Formally, these factors may be modeled as:
Each year at the market was found to increase farmer income by 7.6% with an effect of diminishing returns reducing income by 0.3% per each additional year squared at the market. Each additional 1000 hours worked significantly increased income by 66.5% but has an associated 13% negative effect because of diminishing returns for each additional unit of 1000 hours squared. The sale of nursery or floriculture products has a significantly positive effect of 44% (Table 6). Overall, the regression had an F-statistic of 3.868, a P-value of 0.00579, and an R-squared of 0.3205.
Table 6. Regression results of influences on market farmer income.
n=39.
Multiple R-squared: 0.3205.
F-statistic: 3.868, P=0.005798.
Nearly all farmers at the farmers' markets (97%) indicated that there were good future prospects at the market (Table 7). When asked about more specific, financial details in an open-ended format, the response rate decreased to 33 respondents. Nearly all (94%) of 33 farmer respondents indicated an increase in sales with an average increase of $21,300 and a median increase at $10,000. Three-quarters of farmer respondents (74%) indicated a positive change in their overall financial situation and a quarter (26%) stated that their financial situation stayed the same (n=43). Few farmers (14%) indicated that their debt had increased since selling at the farmers' market. Approximately half saw no change in their debt level (47%) and a third (33%) did not have debts. During the survey administration, some farmers indicated that their debt levels were not related specifically to the farmers' market but to purchases of farm vehicles and equipment.
Table 7. Prospects for farmers at the market.
1 Likert scale question where 5=very good, 4=good, 3=neutral, 2=bad, and 1=very bad.
2 Likert scale question where 5=improved, 4=improved a little, 3=neutral, 2=a little worse, and 1=worse.
3 Likert scale question where 3=increased, 2=stayed the same, and 1=decreased.
Consumer spending at farmers' markets is affected by both economic and social considerations, as indicated in Table 8. This regression output explains nearly a third (31%) of the variation in consumer spending at the farmers' markets (n=212). Consumer income, shopping frequency at the farmers' market, enjoying the market, talking with vendors about seasonal products (while this was the only significant discussion topic, it is interesting to note that even one discussion topic was significant, suggesting that talking with vendors has a role on influencing consumer spending) and the market being a family event are all positive influences on spending. While the marginal effect of a $10,000 increase in household income is an increase of spending by $1, enjoyment, talking about product seasonality with vendors and the market being a family event each have a greater marginal effect on spending than income alone. However, attitudes towards freshness have a strong negative influence. Some possible explanations could be that fresher products are important (as evidenced earlier) but may not be the ultimate factor in the buying decision, or that social interactions are better predictors of spending than the intrinsic quality of the food. The importance of these social aspects indicate that consumer spending decisions are socially embedded, which is consistent with research on social capital theoryReference Frenzen and Davis53,Reference Flora54.
Table 8. Regression results of influences on consumer spending.
n=212.
Multiple R-squared: 0.3115.
F-statistic: 6.17, the P value is 0.0000.
Environmental indicators
Market farmers often use their production methods to add value to their products. The most frequently used value-added production methods, in decreasing order of use, are heirloom varieties (44%), low spray or low chemical use (36%), certified organic (34%), organic practices without organic certification (34%) and free-range animal production (31%). Practices such as using heirloom varieties have the potential to increase the genetic diversity of agricultural landReference Goland and Bauer55. Also, many of these processes have the potential to reduce environmental impacts from decreased synthetic inputs.
Customer feedback has a role in changing environmental quality by influencing farmer production practices. Applying fewer chemicals and using the environment as a selling point are correlated with farmers changing their products to meet consumer demand (Table 9). A third of farmers (32%) who practiced low chemical applications have also changed their products to meet consumer demand. About a quarter of farmers (27%) who use the environment as a selling point have changed products to meet consumer demands.
Table 9. Correlations of farmers changing products to meet customer demand.
n=65.
* Significant at a 95% confidence interval.
A majority of the farms included in the survey (78%) produce on less than 25 acres of land, two-fifths (39%) of which produce on less than 5 acres. However, these farmers own more land than what they produce on: half of farmers (53%) own farms less than 100 acres in size. A total of 1909 acres were in agricultural use by the surveyed farmers, which represents two-fifths (38%) of the land that they own. Their total land area owned is 4986 acres, likely with most of the non-agricultural land being forested, as 90% of Maine is forest coverReference Irland56,Reference Spear57. If this ratio of land in non-agricultural use is consistent across the 65 farmers' markets in Maine, state-wide market farmers' land in non-agricultural use would consist of approximately 2% of Maine's land in farms—however, further research would be needed to identify the scale of farmer's markets' contribution to non-agricultural open space.
Discussion
Social interactions are an important aspect of farmers' markets and influence consumer spending and promote change in some farmers' production practices. The social interactions facilitate feedback that allows farmers to be responsive to consumer demands. However, by focusing their marketing efforts on freshness, locality and organic production, farmers have not made an overt connection to the consumer's desire to support a working rural landscape. For example, consumers rated their belief that the purchase of local products is an effective way to help preserve Maine's rural landscape only behind personal enjoyment of the market (Table 1). Fresh products, a desire to know how their food is produced and the beliefs that their purchases improve the local economy and support open space were the next highest-scoring attitudes. As providers of a scenic value which adds value to the property tax base and can increase housing demand for a townReference King and Anderson9, farmers could make a case to receive payment for their open space services, of which marketing their open space benefits to interested consumers is one mechanism. Farmers' markets, as one route for farmers to sell products directly to consumers, have already established this connection economically and socially (Fig. 3), and have the potential to do so environmentally (Table 9). Models exist for programs that certify farmer's scenic and environmental services through product labels that incorporate recognizable scenic landscape features, which often result in a price premium for labeled farm products10–12,Reference European58.
Figure 3. Vendor sign at the Portland, Maine farmers' market.
While the environment was not identified as a primary purchasing motivator, aspects of certain production methods and practices including free-range and pasture-based methods have the potential for environmental benefits, though they may be marketed as linked to food quality. Future studies should aim to quantify the level of environmental literacy of farmers' market consumers and the relative importance of customer feedback compared with other influences that affect producer's practice adoption. To establish the potential contribution of social interactions to environmental quality, customer interactions and farm production methods would have to be tracked and evaluated over a period of time.
Conclusion
Compared to agricultural systems based on long supply chains, market farmers have the opportunity to interact directly with consumers and respond to shifts in demand as quickly as the agricultural cycle will allow. In so doing, the land managed by these farmers is under the control of a different set of market forces than land managed to be part of a production system of long supply chains. Economically, farmers in this survey indicated incomes higher than those for other farmers in Maine, and at least as high if not higher incomes than other farmers nationally. In addition, there are social benefits and potential environmental benefits associated with selling in this type of direct market.
Social interactions at farmers' markets, such as the market being a family event and having fun at the market are commonplace, with social interactions contributing to consumer spending. The social atmosphere, friendliness of the markets and the ability for consumers to meet other people that they know indicate that community interactions are part of shopping at a farmers' market. Beyond being places of exchange between producers and consumers, farmers' markets are embedded within the context of communities. This emebeddedness allows consumers and producers to influence each other, most notably through producers responding to consumer demand for products with low chemical application and consumer spending being influenced by farmers talking with consumers about product seasonality.
Aside from social proximity, geographic proximity and locality are important contexts for farmers' markets, allowing consumers to identify visibly with the farmers as being part of the community and its physical landscape. Most of the surveyed farmers (85%) live within 50 miles of their markets. Producers and products are thus linked to place, the value of which is signaled to consumers through product freshness, conversations with the producer, and (some) production methods. The transparency and ease of information-flow in farmers' markets allows for consumer oversight of market farmer behaviors, which has the potential to provide positive landscape benefits instead of agricultural uses being perceived as local environmental hotspotsReference Myers and Mittermeier59.
Consumers respond to vendor signals in different ways and were able to be grouped according to their attitudes towards locally produced foods. By recognizing that consumers have different preferences—here characterized as Lifestylers, Seasonal Shoppers and Utilitarians through cluster analysis—farmers' market vendors have an opportunity to better understand who their consumers are and why they make their purchase decisions.
Consumers in this survey have indicated an interest in connecting their purchases with helping farmers, improving the rural economy, supporting their communities and supporting agricultural open space. Social interactions are a key component of farmers' markets, supporting information sharing between consumers and producers and fostering community interactions. Since the economic interactions are socially embedded, product and community identities are linked through these exchanges. Thus, the potential exists for social interactions to influence the adoption of practices with environmental benefits, promote community development and offer producers a unique, uncommodifiable identity of place with which to market their products.
Acknowledgements
I thank the Maine Department of Agriculture, Division of Market Production and Development for help in constructing the survey instruments; the Edna Bailey Sussman Foundation for a research grant; my advisor Randall Kramer, PhD, Professor of Resource and Environmental Economics, Duke University, for his support of and encouragement of this project; Jon Eisen-Hect, PhD, Duke University, for his technical expertise in analytical methods; Charles Becker, PhD, Duke University, for his advice on econometric modeling; and the farmers' market vendors of Auburn, Bar Harbor, Belfast, Blue Hill, Camden, Cumberland, Damariscotta, Farmington, Orono and Portland for their willingness to participate in this research.
