Scott, H. R., Extension Agent, WVU Extension Service
Boone, H. N., Professor, West Virginia University
Boone, D. A., Professor, West Virginia University
Brown, C, Associate Professor, West Virginia University

ABSTRACT

The purpose of the study was to determine the role management structure had on the economic success of farmers markets in West Virginia. A descriptive correlational research design compared the management styles of the markets and effects on economic profitability. Findings showed the variables of market management structure, volunteer/paid status of the manager, size of the market, and age of the market did not demonstrate a significant difference in market economic success.  Discriminant analysis determined that “years of operation” was the only factor which impacted the “successful” market status. Additional research should be conducted to determine the management structure which is working in successful markets.

Introduction

Background/Theoretical Framework

Farmers markets were a part of the culture in Europe and early settlers to the United States had a working knowledge passed onto them by their ancestors. Farmers markets have been a part of the food industry in the United States for almost as long as history (Webber, 2010).  Pyle (1971), in “Farmers Markets in the United States: Functional Anachronisms?” reported the number of farmers markets were recorded in 1880, 1918, 1946, and continued through 1969. In the early 1900s, nearly 40 percent of Americans lived on farms, compared with 1 percent in 2000, and much of the food bought and consumed in the United States was grown locally (Pirog, 2009).  Bachmann (2008) states farmers markets are an ancient method used across the world for selling produce directly to customers.  There have been ups and downs in farmers markets with a decline after World War II followed by some reestablishment in the 1970s and 1980s (Webber, 2010).  The most recent resurgence began in the 1990s and is still going today (Stephenson, Lev, & Brewer, 2006).  Speculation as to the driving forces behind this expansion was the use of farmers markets to build communities and this became a social life for the farmers and customers.

Several markets have established a record of economic success.  Markets such as the Pike Place Market in Seattle, Washington and the Soulard Market in St. Louis, Missouri are well established markets that have survived over the years (Webber, 2010).  Many markets get started, but the ones which survive have had economic success as they matured.  Markets that did not become self-sustaining were often supported by cities, local government agencies, and nonprofit organizations or grants.

Oberholtzer and Grow (2003) in a survey of the Mid-Atlantic Region found that the key ingredient was the market manager.  Jolly (2005) states the success of any market, and the financial success of the vendors, depends a great deal on the manager of the market.  The life span of new markets is that half will close within a year (Stephenson et al., 2006).  Being able to have enough administrative revenue to provide sustainable management (manager) is a struggle for markets.  This is a challenge to markets of all sizes, but seems to have a larger impact on the small scale markets. 

The Farmer’s Market Model (Stephenson et al., 2006) provided the theoretical framework for the research.  Stephenson et al. (2006) examined successful farmers markets in the Northwest and synthesized a model that illustrated how farmers market organizers successfully adapted to barriers and challenges in their environment.  One aspect identified was the area of managing to maximize atmosphere, products, and community.  The synthesized farmers market model and description were:

Market managers identified atmosphere, product, and community as key elements of good farmers markets. These elements may be seen as a target or goals for market organizers. Markets operate under wide-ranging external influences. Skilled management supports successful markets as they adapt to these influences, flourish, and reach market goals (Stephenson et al., 2006, p. 7).

The synthesized model in Stephenson et al. (2006) places farmers markets in an environment made up of natural and political influences ranging from the dependence of crop production on local agro-ecozone conditions to the impacts of state and federal regulations.  Markets adapt to these conditions through management and their adaptations are visible: they create an atmosphere conducive to socializing and sales, they procure a variety of high quality products, and they build community support via a loyal customer base and integration into local social and economic systems.

Much of the ability to excel in the key traits of successful farmers markets is based on the use of management tools. Individual farmers markets have access to varying quantities of resources in terms of people, time and revenue. The availability of these resources impacts the ability of market organizers to manage and therefore, impacts the level and quality of management markets receive (Stephenson et al., 2006, p. 7).

 

Purpose(s)/Objective(s)

With farmers markets increasing in popularity and numbers, a need exists to determine why some markets thrive while others fail. This researcher defined a market as economically successful when the market’s income is sufficient to pay for all costs associated with the operation of the market. A review of literature yielded some regional information however no information specific to West Virginia was found. The problem addressed was whether the market management structure of farmers markets was a contributing factor in the markets economic success.

This research examined farmers markets in West Virginia to determine how management structure impacted the economic success of farmers markets as compared to previous research. Descriptive correlational research techniques were used that involved examining associations between the variables of management structure, volunteer paid status of the manager, size of the farmers market, and age of the farmers market to economic success.  Theoretical implication from previous research was that management appeared to be the common factor that impacted successful markets. The information gained as a result of this study will provide recommendations regarding farmers market organization and planning that may enhance the success and longevity of individual farmers markets. In addition this information will be utilized in educational materials to benefit farmers market managers, board of directors, and others who assist with current management and strategic planning for farmers markets. The following hypotheses were used in determining management impact on economic success: 

Hₒ ꞊ Farmers market management structure has no effect on market economic success.

H_A ꞊ Farmers market management structure has an effect on market economic success.

Hₒ = Paid manager for farmers market has no effect on market economic success.

H_A ꞊ Paid manager for farmers market has an effect on market economic success.

Hₒ ꞊ Size of the farmers market has no effect on whether the manager is volunteer or paid.

H_A ꞊ Size of the farmers market has an effect on whether the manager is volunteer or paid.

Hₒ ꞊ Age of the farmers market has no effect on whether the manager is volunteer or paid.

H_A ꞊ Age of the farmers market has an effect on whether the manager is volunteer or paid.

 

Methods and Procedures

Research Design

The research design used in examining the farmers markets as to their economic profitability according to their management style was descriptive correlational research. This research involved comparison of the management styles in the market and the effect it had on economic profitability. Ary, Jacobs and Sorensen (2010) stated “correlational research assesses the relationships among two or more variables in a single group” (p. 349). “Correlational research is useful in a wide variety of studies. The most useful applications of correlation are: (1) assessing relationships, (2) assessing consistency, and (3) prediction” (Ary, Jacobs & Sorensen, 2010, p. 351).

 

Population

The target population for this research study was the managers of farmers markets operated in the state of West Virginia during 2012. Due to the small number of known farmers markets operating in the state (80-100), the survey was sent to all of the markets making the target population the accessible population. Market information was obtained via the West Virginia Department of Agriculture, National Resource Conservation Service, West Virginia Famers Market Association and West Virginia University Small Farms Center. All farmers markets that were operated in communities and were not a commercially owned and operated business were included in the population.

By including all the farmers markets in the survey, sampling error was not a concern. In assembling the lists from the various sources and doing physical comparisons of the names to avoid duplications addressed frame and selection errors. Measurement error was addressed by having the validity and reliability of the survey instrument established.

 

Data Collection

Data collection procedures followed recommendations of Dillman, Smyth, & Christian (2009). A total of 90 survey questionnaire packets were mailed on the initial mailing based on the data base that had been compiled from various agencies. It was determined by returns and email responses that seven of the markets had closed or ceased to operate in 2012. In addition, two new markets that were not on any database were identified. This gave a total population of 85 farmers markets with a response from 56 markets for 65.9 percent rate of return.

 

Non-Response Error

To address non-response error, a comparison of early respondents to late respondents was made (Dillman, Smyth, and Christians, 2009). ). Linder and Wingenbach (2002) recommend using a comparison of early to late respondents as one method of addressing non-response error. Late respondents are similar to non-respondents so the responses from the early mailings were compared to the respondents from the late mailings by running a comparison on a selection of variables. If no significant differences were found between the early and late respondents, one can assume the respondents are an unbiased sample of the recipients and one can generalize to the total group. Out of the five questions tested the written job description was statistically significant (α ≤ .05) (see Tables 1 and 2). Because early and late respondents were different, generalizations were limited to the 52 respondents.

 

Examined Areas	Late	N	M	SD	t	df	Sig
Total Sales Producer/vendor	No	33	$54,762.94	$113,843.93	0.704	40	0.486
	Yes	9	$27,566.00	$31,071.21
Number of Producers	No	41	31	74.591	0.736	51	0.465
	Yes	12	15	9.733

*α ≤ .05

Table 1. Comparison of Early and Late Respondents on Selected Variables

 

Examined Areas	Code	Late	N	Chi Value	Sig.
Time of Operation	One day	No	25	1.659	0.198
		Yes	11
	Two day	No	17
		Yes	3
Market Status	Group 1	No	16	1.222	0.269
		Yes	3
	Group 2	No	24
		Yes	10
Written job description	Yes	No	7	6.386	0.012*
		Yes	7
	No	No	32
		Yes	6

*α ≤ .05

Table 2. Comparison of Early and Late Respondents on selected Variables – Part 2

 

Data Analysis

Information from the questionnaire was collected and entered into an excel spreadsheet and then imported into SPSS for Windows. The alpha level of significance was set a priori at α ≤ .05 for all statistical tests.  Descriptive analyses appropriate for the respective scales of measurement were performed on the data including measures of central tendency (mean, median, or mode) and variability (frequencies or standard deviation).  The results were represented as frequencies and percentages as well as mean, median and mode in both table and narrative form.

 

Results and Discussion

Hypothesis #1

For the purpose of the following analysis economic success was determined using respondents’ indicators of their market economic status.  The status categories of beginning, struggling, and getting started were recoded into one category (Struggling) and sustaining and successful were recoded into a second category (Successful).  A chi-square test of independence was used to determine if there was a significant association between the variables state government agency, city-county or municipal government agency, producer-vendor board of directors, community association-non-profit organization, members of the market association, market manager and market status.  The following hypotheses were tested:

Hₒ ꞊ Farmers market management structure is independent of market economic status.

H_A ꞊ There is an association between farmers market management structure and market economic status.

There was no significant difference between the variables state government agency, city-county or municipal government agency, producer-vendor board of directors, community association-non-profit organization, members of the market association, market manager and market status. In each case the researcher failed to reject the null hypothesis.  All the variables were independent (see Table 3). 

 

		Market Status		Value	df	Sig.
		Struggling	Successful
State government agency	Yes	3	9	0.794	1	0.373
	No	16	25
City-county or Municipal government agency	Yes	1	9	3.581	1	0.058
	No	18	25
Producer-vendor operated Board of Directors	Yes	4	12	1.173	1	0.279
	No	15	22
Community association-non-profit organization	Yes	2	2	0.377	1	0.539
	No	17	32
Members of the market association	Yes	7	19	1.768	1	0.184
	No	12	15
Market Manager	Yes	7	11	0.11	1	0.741
	No	12	23

*α ≤ .05

Table 3. Chi-Square Analysis – Farmers’ Market Management Structure by Market Economic Status

 

Hypothesis #2

            A chi-square test of independence was used to determine if there was a significant association between the variable paid market manager and market status.  The following hypotheses were tested:

Hₒ ꞊ The variables paid manager for farmers market and market economic success are independent.

H_A ꞊ There is an association between paid manager for farmers market and market economic success.

The chi-square analysis (χ = .406, df = 1) determined there was not a significant difference between market manager and market status (see Table 4).  The researcher failed to reject the null hypothesis. The variables paid market manager and market status were independent.

 

		Market Status		Value	df	Sig
		Struggling	Successful
Paid Market Manager	Unpaid	13	26	0.406	1	0.524
	Paid	6	8

*α ≤ .05

Table 4. Chi-Square Analysis – Paid Market Manager and Market Status

 

Hypothesis #3

The population for the study consisted of 53 markets reporting the salary status of their manager with 38 markets not paying their market manager and 15 paying their market manager.   A t-test statistical procedure was used to determine if a statistical difference existed between the mean number of vendors in the two groups.  The hypotheses tested were:

Hₒ ꞊ The mean number of vendors is equal between managers who volunteer and those who are paid.

H_A ꞊ The mean number of vendors is not equal between managers who volunteer and those who are paid.

An independent t-test statistical analysis procedure was used to compare the mean number of vendors for the unpaid and paid market managers.  The statistical analysis results (t = -1.595, df = 14.086) were not significant.  Therefore the researcher failed to reject the null hypothesis (see Table 5). There was no difference between the number of vendors for paid managers and volunteer managers.

 

	N	Mean	SD	df	t
Unpaid	38	13.55	10.454	14.086	-1.595
Paid	15	62.4	118.402

*α ≤ .05

Table 5. Comparison of the Mean Number of Vendors for Paid and Unpaid Market Managers

 

Hypothesis #4:

The population for the study consisted of 52 markets who reported the age of their market with 36 markets not paying their market manager and 16 paying their market manager.  A t-test statistical procedure was used to determine if a statistical difference existed between in the mean age of the farmers market when compared by the “paid” status of the market manager.   The hypotheses tested were:

Hₒ ꞊ The mean age of the market is equal between managers who volunteer and those who are paid.

H_A ꞊ The mean age of the market is not equal between managers who volunteer and those who are paid.

An independent t-test statistical analysis procedure was used to compare the mean age of the unpaid and paid market managers.  The statistical analysis results (t = 1.342, df = 50) were not significant.  The researcher failed to reject the null hypothesis (see Table 6). The average age of the market was equal for paid managers and volunteer managers.

 

	N	Mean	SD	df	t
Unpaid	36	10.5	9.416	50	1.342
Paid	16	7.13	5.149

*α ≤ .05

Table 6. Comparison of the Mean Age of Farmers Market for Paid and Unpaid Managers

 

Discriminant Analysis

A stepwise discriminant analysis was conducted on the data to determine the best discriminators among “years of operation, total producer-vendor sales, number of producers, advertising budget, full time year round, part time seasonal, part time year round, and volunteers” as an influence on market status.  The eight discriminators were used as potential discriminating variables in the statistical procedure.  The null hypothesis tested was there would be no impact by the discriminators between the group centroids on the discriminant scores.  At an alpha level of ≤ .05, the null hypothesis was rejected on the discriminator “years of operation” and the research hypothesis was accepted that the discriminator did have an impact on “market status.”  None of the other discriminators loaded into the equation.

One factor, years of operation, loaded on the discriminant function when analyzed by their structure coefficients.  The group centroids for not checked and checked were -.575 and .372 (see Table 7).  The canonical discriminant function coefficients for each attribute were 1.000.

To determine the similarity between a single variable and a discriminant function, the structure coefficient was examined.  The structure coefficient was 1.000 signifying that the function was carrying nearly the same information as the variable (Klecka, 1980).

The Wilks’ Lambda is a multivariate measure of the group difference over the discriminating variables (Klecka, 1980).  Values of the lambda which approach zero indicate high discrimination.  The analysis resulted in a Wilks’ Lambda of .813 indicated that 81.3 percent of the variance was unexplained.  The eigenvalue of .230 indicated that the discriminant function can explain only .230 times as much as not being explained.

The canonical correlation coefficient is used to examine the relationship between the sets of variables.  A large coefficient indicates a strong relationship between the groups and the discriminant function (Klecka, 1980).  The canonical correlation coefficient was .433.

 

Statistic				Value
Centroids
	Beginning, struggling, started not a factor			-0.575
	Sustaining, successful a factor			0.372
Standardized canonical discriminant function coefficient				1.000
Structure coefficient				1.000
Canonical correlation coefficient (Rc)				0.433
Eigenvalue				0.230
Wilks' Lambda				0.813

*α ≤ .05

Table 7. Summary Data: Discriminant Analysis of Discriminating Variables

 

The classification analysis results found that 66.0 percent of the original group cases were correctly classified (see Table 8).  Based on the sustaining-successful factor, the researcher can predict with 66.0 percent accuracy the market status.

In looking at the predicted groups, the numbers were nearly the same between the beginning-started groups, but the difference occurs when examining sustaining-successful factor.   Here the 32 cases are evenly split between beginning-started and sustaining-successful groups.  Years of operation is the classification which influenced the 16 cases that consider themselves as sustaining-successful in market status.  The cases where years of operation classification did influence the market status would indicate that these markets were probably at the average years of operation or older. 

 

Group		No. of Cases	Predicted Group
			Beginning	Sustaining
Beginning-started status not a Factor
Number		18	17	1
%			94.4	5.6
Sustaining-successful status a Factor
Number		32	16	16
%			50.0	50.0

Percent of Cases Correctly Classified: 66.0%

Table 8. Classification of Cases Based on Discriminant Analysis and Years of Operation

 

Conclusions

The results of the four hypotheses examined in this study were as follows:

• Market management structure was independent for the variables of government agency, city-county or municipal government agency, producer-vendor board of directors, community associations-non-profit organization, members of the market association, market manager and market, when comparing market status.

• No significant association between variables paid market manager and market status.

• When comparing the mean number of vendors between unpaid managers and paid market managers, there was no significant difference.

• Examining if there was a difference between the mean age of farmers markets and the “paid” status of the market manager, we found the average age of the market was equal for paid managers and unpaid managers.

Results from a stepwise discriminant analysis of “years of operation, total producer-vendor sales, number of producers, advertising budget, full time year round, part time seasonal, part time year round, and volunteers” resulted in “years of operation” being the only discriminator that had an impact on “market status.”

 

Implications

As Extension Agents work within their communities they need to look at several factors to assist in developing a farmers market into a successful operation.  The Stephenson Farmers Market Model illustrates there are multi-interactions that occur involving such areas as atmosphere, product and community. Producers are successful in growing product, but Extension Agents need to help them understand what consumer expectations are for market products, the social interactions occurring, regulations, quality and presentation of products, and organizational structures that can be used to make farmers markets function economically. In order to do this, Extension can provide training in the areas of business structure, site selection, how to determine the potential mass of consumers in the area, management, marketing, collaboration within the community, and others issues specific to the location. The results of a successful farmers market will not only be to producers, but to the community or region due to the economic multiplier effect. 

In this study “years of operation” was a significant factor because as farmers markets continue in operation they develop the ability to manage their operation economically. Other factors considered as being significant in the operation of farmers market are: location of market, how market is financed, role of manager in the market, governance structure used, collaborations developed within community, and economic status of consumers. Extension can be a partner in helping local farmers markets by providing training and education in these areas which will improve the ability of farmers market achieving success. 

 

Literature Cited

Ary, D., Jacobs, L. C., & Sorensen, C. (2010). Introduction to Research in Education. Belmont, CA, United States: Wadesworth.

Bachmann, J. (2008). Farmers' markets: marketing and business guide. Retrieved September 21, 2010, from ATTRA - National Agriculture Information Service: www.attra.ncat.org/attra-pub/PDF/farmmarket.pdf

Dillman, D., Smyth, J., & Christian, L. (2009). Internet, mail and mixed-mode surveys: The tailored design method (3rd ed.). New York: Wiley.

Govindasamy, R., Zurbriggen, M., Adelaga, A., Italia, J., Nitzsche, P., & VanVranken, R. (1988). Farmers markets: managers characteristic and factors affecting market organization. Rutgers, The State University of New Jersey. New Brunswick: New Jersey Agricultural Experiment Station P-02137-8-98.

Govindasamy, R., Italia, J., & Adelaja, A. (2002). Farmers’ markets: Consumer trend, preferences, and characteristics.  Journal of Extension [On-Line], 40(1). Article 1RIB6. Available at: http://www.joe.org/joe/2002february/rb6.php  

Gwin, L., & Lev, L. (2011). Meat and poultry buying at farmers markets: a survey of shoppers at three markets in Oregon. Journal of Extension [On-line], 49(1), Article 1RIB4. Available at:  http://www.joe.org/joe/2011february/rb4.php

Jolly, D. (2005). Farmers market management skills (Vol. 2). Davis, CA: UC Small Farm Center.

Klecka, W. (1980). In Discriminant analysis: Quantitative applications in the social sciences. London: Sage Publications.

Linder, James R., & Wingenbach, Gary J. (2002).  Communicating the handling of nonresponse error in Journal of Extension Research in brief articles. Journal of Extension [On-line], 40(6), Article 6RIB1. Available at: http://www.joe.org/joe/2002december/rb1.php

Mainville, D. (2010, April 27). Foundations for a successful farmers' market. Retrieved September 21, 2010, from Virginia Tech Extension Service: www.pubs.ext.vt.edu/448/448-502/448-502.html

Oberholtzer, L., & Grow, S. (2003, October). Producer-only farmers' markets in the mid-atlantic region. Retrieved September 21, 2010, from Henry A. Wallace Center for Agricultural and Environmental Policy: http://www.winrock.org/wallace

Pirog, R. (2009). Local Foods: Farm Fresh and Environmentally Friendly. http://www.leopold.iastate.edu/research/marketing_files/WorldBook.pdf.

Pyle, J. (1971). Farmers' Markets in the United Sates: Functional Anachronisms? Geographical Review, 61(2), pp. pp. 167-197.

Ragland, E., & Tropp, D. (2006, May). National farmers market manager survey 2006. Retrieved September 21, 2010, from Agricultural Marketing Service: www.ams.usda.gov/AMSv1.0/getfile?dDocName=STELPRDC5077203

Stephenson, G., Lev, L., & Brewer, L. (2007, December). Understanding the link between farmer's market size and management organization. Retrieved September 21, 2010, from Oregon State Extension Service: http://extension.oregonstate.edu/catalog/pdf/sr/sr1082.pdf

Stephenson, G., Lev, L., & Brewer, L. (2006). Enhancing the success of northwest farmers' markets. An executive summary. Retrieved September 21, 2010, from Oregon State University Extension Service: www.smallfarms.orgeonstate.edu/sites/default/files/TechReport22.pdf

Webber, D. (2010). How to organize and run a successful farmers' market. Retrieved September 21, 2010, from Mass.Gov.org: www.mass.gov/agr/markets/farmersmarkets/start_a_market.htm

---