Community Supported Agriculture: Research and Education for Enhanced Viability and Potential in the Northeast

Objective 1: Determine the extent to which Community Supported Agriculture (CSA) constitutes an economically viable production and marketing strategy in the northeast region.

In order to determine the extent to which CSA constitutes an economically viable means of farming, it was necessary to determine the costs and returns associated with CSA operations. To satisfy this objective, cost and return data were collected via mail survey from CSA operations across the Northeast Region for three years: 1995, 1996 and 1997. The survey questionnaire was developed through cooperation of the project personnel during the first months of the project and was pre-tested during March of 1996. Virtually the same questionnaire was used in each of the three survey years. Only minor changes were made prior to the 1996 survey after preliminary analyses of the 1995 data.

The survey questionnaire was very detailed and included questions on numerous categories of farm costs and returns. In addition to data on costs and returns, data on labor use (in hours) and asset levels were collected. Long term viability of CSA operations, as with any farm, depends upon covering the full economic costs of production. Full economic costs include not only those costs that represent day to day expenditures, but also factors that are not paid an explicit wage or rental price. These latter factors typically include land, buildings, capital items and operator labor. In addition, a number of open-ended questions were included that allowed the operator to detail techniques employed to manage soil fertility, insects, weeds and diseases. These open-ended questions allowed us to assess the types of production technologies employed by CSA operations and demonstrate their use of low-input/sustainable technologies.

Lists of CSA operations in the Northeast were obtained from Stephen Gilman, coordinator of the CSA Farm Network project, and from CSA of North America. To reduce the number of confounding factors, we chose to focus on CSA farms that primarily produced vegetable crops for their shareholders. Results from surveys conducted by Stephen Gilman (see Objective 2 below) were used to identify CSA farms that primarily produced vegetables. A sample of 50 CSA farms was drawn from the lists and survey questionnaires were mailed to those operations in the middle of April 1996 asking for 1995 data. CSA operators were offered a payment of $100 to complete the survey as stipulated in the proposal. We anticipated that the $100 inducement and the timing of the survey near the federal tax filing deadline would result in a high response rate. After written and telephone follow-ups, total of 21 CSA farms responded to the 1995 survey. The response rate of 44% was disappointing and was attributed to what, in fact, amounted to bad timing. A very rainy spring forced CSA operators to delay field work until just about the time our survey arrived.

The CSA lists were updated prior to the 1996 and 1997 surveys using information provided by Stephen Gilman. While keeping the survey date close to the federal tax deadline remained an objective, the 1996 survey was conducted earlier, in March of 1997. Anticipating non-response, the number of surveys mailed was increased to 75. Twenty-six CSA farms responded to the 1996 survey, a response rate of 35%. To encourage a higher response rate for the 1997 survey, the questionnaires were mailed in February, 1998, and respondents were offered $150 to complete the survey. With an even greater appreciation for non-response, despite the financial offer, the sample size was increased to 108 CSA farms. Completed questionnaires were received from 34 CSA operations, a response rate of about 31%. These data were merged with the 21 observations for 1995 and the 26 observations for 1996.

One issue that affected response in each of the survey years was the length of the survey and the detail that was requested. While the length of the survey questionnaire clearly inhibited participation, the information was crucial to our findings. For example, it would not have been possible to compute the full economic costs of production with out the detailed information on operator and hired labor hours of work.

Analyses of the 1995, 1996 and 1997 data have been conducted and provide interesting results. On average, share prices charged by CSA operators did cover cash costs as reported on the survey responses. In 1995, the average share price was about $445, while the average reported costs were $408.50. The average 1996 share price was $460 while costs were $365. In the final year of the study, average share price was $350 and costs were $260. This is consistent with the CSA concept; the share price should cover the farm’s budget. However, further analyses suggested that CSA operations did not cover their full economic costs of production. One very important component of full economic cost is the wage paid to the farm operator. A number of operators were not paid a competitive wage for the hours spent working at the CSA. We imputed additional wages of $197 in 1995, $163 in 1996, and $149 in 1997. When these additional costs are added to the CSA budget, the share prices would need to be higher by $161 in 1995, $68 in 1996, and $59 in 1997. CSA share pricing and the inclusion of all costs is a critical factor for long-term viability. Our results and the implications for pricing of shares will be important pieces of information to disseminate to CSA operators.

The ability of CSA operations to cover the full costs of production is important for long term viability. A number of factors may contribute to CSA performance; there are theoretical reasons to consider factors such as farm size, human capital measures, resource endowments and technologies employed. In addition, we considered the organization of the CSA as a factor that may affect success. The two broad categories we considered were core-group and non core-group CSAs. This is reasonable given that core-groups often work with the grower in establishing the share price and share composition. Core-group CSA operations were typically larger and produced more produce for their shares. These differences in amount of product were apparent in the price of a full share. In 1995, core-group farms charged $487 on average while non core-group farms charge $350. However, the price per pound for core-group farms was lower, $0.95 per pound versus $1.03. Eleven of the 23 CSAs in 1995 had a core group. Seven of those CSAs had income that met or exceeded costs. Of the 12 non core-group CSAs, only two had income that met or exceeded costs. In 1996, a similar picture emerged. Nine out of 12 core-group CSAs had income that was greater than or equal to costs, while only 3 of 14 non core-group CSAs had income that met or exceeded costs. These statistics suggest that core-group CSA operations were more likely to be viable.

To assess whether the use of a core-group was a causal factor in farm success required multivariate statistical techniques. Net income was used as a dependent variable in multiple regression analysis. A test was conducted to determine whether the three years of data could be pooled. The test suggested that the data could be pooled, which improves the statistical reliability of the analysis. The results indicated a number of expected results. Larger farms had higher net incomes and human capital measures, experience and education, both resulted in higher net income as well. Importantly, we did find that core-group CSAs had higher net incomes by more than $7,000, even after controlling for farm size and human capital measures for the operator/manager. This result was statistically significant at the 10% level or better. Our conclusion from analyses was that core-group CSAs are more likely to remain viable.

The analyses suggested that core-group farms were more successful based upon monetary measures. The core-group management style was also analyzed in terms of risk management using stochastic efficiency criteria. The results confirmed that most CSA operators would find the risks associated with the core-group farms to be preferable to risks associated with non core-group farms. Mean-variance, stochastic dominance and mean Gini analyses were performed using the data. The analyses showed that the core-group farms dominated non core-group farms when levels of risk associated with each were considered.

Research was also conducted on the value of CSA shares in cooperation with the Department of Nutrition at the University of Massachusetts. A Masters student, Jack Cooley, considered the nutritional aspects of CSA membership. As part of his research, he gathered detailed data on the composition of shares for three CSA farms in the Amherst- Massachusetts area. He also sought the help of Dr. Lass in comparing measures of the retail value of a CSA share to the share price the consumer paid. Retail values computed for the 1995 crop year indicated that retail values were as much as double the CSA share prices. Coupled with the findings on costs of producing a CSA share, these additional results provide at least anecdotal evidence that CSA operators should be able to cover full economic costs of production if the share is priced accordingly. This research was published in the Review of Agricultural Economics in 1998 and in the 1997 CSA Farm Network. In addition, interviews were given to the New York Times, and several national magazines during the summer of 1997.

The final research focus was on attitudes and beliefs regarding low-input/sustainable production. A literature review was conducted focusing on different behavioral models of farm level decision-making. The focus was on models other than the standard neoclassical economic models of profit maximization or cost minimization that are not appropriate for CSA operations. We focused on models that linked manager/operator preferences or beliefs with production decisions and models of cooperation. Little theoretical work has been done on CSA decision-making. We completed the development of one such theoretical model of CSA behavior, a simple bargaining model. In the empirical investigations reported above, it was observed that a number of CSA farms did not cover full economic costs. In the theoretical model, parameter requirements that would result in such a solution were considered. The bargaining model revealed that farmers may accept a less than desirable outcome, in terms of net returns, if they feel significant warm-glow effects from production. These warm-glow effects could be due to a commitment to the CSA concept or from sustainable or environmentally safe production. The model provides a plausible explanation of CSA operator behavior. The conclusion is consistent with community benefits that the CSA farmers detailed in open-ended questions on the survey. For example, 63% of the 1995 CSAs responded that preserving the environment was an important benefit they provided. Equally important in 1995 was the use of CSA as an educational and outreach tool for the community. In addition, 40% or more of the CSAs in both 1995 and 1996 reported that integrating the community was an important benefit that their CSA provided.

Objective 2. To publish an annual CSA Farm Network publication to link CSA projects with area specialty production farms and resource providers and disseminate research based information developed under objective 1.

The CSA Farm Network (CSANET) project has enlisted the seven state NOFA organizations, plus MOGFA in Maine and the Bio Dynamic Association in Pennsylvania to create listings of all the CSAs they could find in the Northeast. These efforts were undertaken during each of the three project years. A total of 177 CSAs were identified in 1995. The CSA operations were sent surveys to elicit information about the scopes of their operations and about practices that they employed; 55% of the CSAs responded to the survey. The database was updated in 1996 and again in 1997. A second survey was conducted during fall 1996 to help update the database; 46% of the CSA operations surveyed provided information. The database was shared with the Bio Dynamic Association, CSA North America and was used to select samples of CSAs that were asked to provide further data on costs and returns (Objective 1). The combined resources represent a comprehensive listing of CSA operations in the US. A Resource Provider survey and database was also created for a Resource Directory.

Winter and spring of 1996 saw the development of the first CSANET publication, Volume I. The coordinator (Stephen Gilman) conducted research, gathered materials and began writing articles for the publication. The coordinator conducted a number of interviews and developed explanatory articles from the interviews. A number of articles were solicited from CSA operators across the Northeast on their practices. Academic researchers conducting investigations of the new (to the U.S.) CSA phenomenon were also contacted for articles. The production editor began formatting the publication, commissioned artwork and gathered graphics. In July 1996, the first issue of CSA Farm Network was published. Five hundred copies were initially distributed to the CSA operations, major participants, and to the participating state organizations for distribution in their areas. Additional copies have been distributed to help offset unanticipated production costs.

Work also began on the 1997 (Volume II) CSA Farm Network publication. Researchers and practitioners were contacted for potential articles and materials for the 1997 publication. A number of CSA farmers were contacted to write up some of their particular practices for the publication, a number of farmers and shareholders were interviewed to develop additional articles. The production editor began to format the 1997 CSA Farm Network publication. Artwork was commissioned and graphics were gathered, the coordinator conducted further research and developed materials for the 1997 publication. The coordinator and production editor finalized the Volume II CSA Farm Network publication layout and established Internet linkage.

The 1997, Volume II CSA Farm Network was published in February 1998, a 96 page book complete with articles, research papers, a resource directory and a Northeast CSA directory. Five hundred copies were mailed to CSA operations in the directory, 150 copies were sent to Northeast SARE for distribution, and 75 copies were sent to reviewers and other organizations listed in the resource database. Copies have also been made available to participating organizations for sale at conferences and workshops. The Coordinator has been interviewed by several national magazines and publications and has given permission to use excerpts of printed CSA Farm Network materials, which stimulated wider interest in project materials. Copies of Volumes I and II are still being sent out to groups and individuals who see listings or ads for the project through nationwide publications and Web Site listings.

In addition to work on the 1997 CSA Farm Network, Stephen Gilman organized and facilitated a CSA forum at the NOFA Summer Conference, held at Hampshire College, August 8. Approximately 55 CSA farm operators, shareholders and other interested individuals attended the forum. He also served as a resource to organizers of the Northeast Community Supported Agriculture Conference and was a presenter on two workshop panels.
Volume III of the CSA Farm Network publication is now in production. The CSA database and connections with the participating organizations have been updated and a number of researchers and practitioners have contributed articles and other materials. Discussions with the production editor have been aimed at creating a larger publication for the Volume III edition with updated and expanded directory sections and a wider scope that includes CSA practices and activities in other parts of the country. In this regard, use of the Internet to disseminate Volume III CSA Farm Network is being investigated. NOFA has been given a large scale Internet Web site by SOHONET of New York City (www.nofa.org) and a portion of the site is being made available to the CSA Farm Network. Due to the highly positive feedback regarding Volumes I and II, plans are to perpetuate the CSA Farm Network project in electronic form beyond the 3 year period of this SARE project. The target electronic publication date is Fall1999. Toward these ends, the Coordinator completed a Web Site Authoring Course to learn the use of PageMill software. The shift to an ongoing CSA Farm Network web site will allow rapid updates and is expected to overcome limitations associated with printing. This will give the project a life of its own beyond the scope of the initial Project. A CSA Farm Network web site will provide a direct and efficient means of disseminating information to larger numbers of CSA farmers, shareholders, researchers and resource providers as well as to consumers looking for a CSA to join in their area through their own personal Internet access, or through Libraries.

Objective 3. Provide direct education and mentoring to CSA farms to help solve specific CSA problems.

A direct stimulus for seeking SARE funding for this objective was based upon findings at the CSA Veteran's Trouble Shooting Forum held in Rowe, MA, in 1993. Among other things, it was determined that many of the regional CSA projects lacked core groups and that this was a significant obstacle to their sustainability. It was also determined that there were many site-specific problems that could very likely be solved in a smaller group atmosphere. Based upon this feedback, the direct education and peer mentoring components of the project were proposed. Community Supported Agriculture of North America (CSANA), under the direction of Robyn Van En and Cathy Roth, was responsible for providing direct consultation and peer mentoring. Robyn Van En died on January 8, 1997, having completed many of the objectives and laid the groundwork for finalization of the project. Elizabeth Keen stepped in after January 8 to fulfill the objectives.

Peer Education and Mentoring for Problem Solving forums were organized by Dr. Cathy Roth and Robyn Van En. Four forums were conducted during the winter of 1995/96. These forums were centrally located amongst the six northern Northeast SARE states to give equal access to as many CSA projects as possible. Two additional forums were organized in New York and Pennsylvania during the fall of 1996. There were separate sessions for the farm production crew and the administrative/membership body. Farmer, apprentice and production-related questions and issues were addressed in the first half of the day; core group, membership and administrative questions and issues were addressed during the second half.

One of the most significant findings from the forums has been that most of the attending CSA now have core groups in place and were prepared to discuss and plan the next levels in their CSA's evolution and sustainability. With core groups established and the individual CSA stabilizing, another goal that Dr. Roth and Ms. Van En had set for the sessions was to stimulate the formation of CSA Associations. These associations or groups of CSA would continue, on their own initiative, to discuss development and pooling of resources in the given state or region.

Part of the process of moving toward the CSA association level was to create a forum for relationship amongst the various CSA operations with a CSA specific agenda. A significant amount of peer education and problem solving was accomplished in the course of each session and Ms. Van En and Ms. Keen were available for further follow up and telephone consultation on the more enduring or site specific issues.

During the general problem solving a prominent area of concern was the difficulty of individual CSA operations to educate their respective community members as to the value of and reasons for CSA. A solution that was broached, and has since been enacted in Maine, Vermont and New York, was to organize a CSA Education and Member Sign up Faire in cooperation with another annual statewide event. Each of these 'faires' had very positive results with the community members, while one of the most important results was the first time formal collaboration between the regional CSA to organize and participate together at the events. Notes taken at each session and the resulting solutions and strategies were to be included in an updated and expanded edition of Community Supported Agriculture for Agriculture Supported Communities.

Robyn Van En provided direct education and peer mentoring to CSA farms through telephone consultations from September, 1995, to January 7, 1997. At least four telephone consultations were logged each week during that period. She primarily addressed specific issues related to the creation and sustainability of CSA. Robyn was active in supplying written materials, names and addresses of contacts for new CSA farmers who required answers to specific questions or had needs for further information. Robyn also was instrumental in putting members of the media in touch with CSA farmers and supplying consumers with information on CSA operations near them. These efforts served to further education the general public about CSA and increase the demand for CSA by the public.

After January 8, Elizabeth Keen continued to conduct telephone consultations, provide referrals and disseminate information approximately three hours per week. In addition, Elizabeth created a comprehensive resource list of publications, organizations, periodicals, and on-line services to assist CSA farmers and members develop to their fullest potential. CSANA completed the directory through a survey to over 600 CSA farms with the assistance of Stephen Gilman (objective 2), CSA West, the Bio-Dynamic Farming and Gardening Association, the state NOFA associations and MOFGA (Maine). The directory provides a sense of belonging among CSA farms to a larger movement and aids to solidify future relationships among CSA farms and provides tangible evidence of CSA growth over the last five years. The directory also responds to the growing interest in CSA among consumers, media, researchers, and the general public. An initial printing was made available at the Northeast CSA Conference and to members of the media, USDA offices, CSA farms in the Directory, and those requesting copies from CSANA.

Northeast CSA Conference planning became the focus for Robyn, Elizabeth and Cathy in November 1996. The planning committee consisted of representatives from UMass Extension, Just Food (NY), Wilson College-Center for Sustainable Living (PA), Cornell Cooperative Extension-Dutchess County (NY), and Ann Reid, member of Caretaker Farm, MA. The planning committee was successful in securing additional Northeast SARE funding for the conference. The conference planning began by sending a brief survey to farmers and members of over 250 Northeast CSA farms asking them to set the agenda for the conference. The clamor for a larger regional CSA meeting was confirmed by survey response, setting the structure as well as creating excitement in the region for the conference. The conference, entitled CSA: Building A Future for Farming in the Northeast, represented a culmination of this portion of the project and was held November 7-8, 1997 in western Massachusetts. The planning committee organized the schedule to maximize discussion that would allow for peer mentoring and trouble shooting. The schedule consisted of six main sessions with six to seven workshops offered in each session, most of the thirty-nine workshops were done in panel discussion format. All workshops were divided into the following categories: Membership, CSA Crop Production, CSA Marketing, CSA Administration, and Other Related Topics. The objective was to encourage in-depth discussion on each topic. Three panelists along with a moderator led each workshop by presenting personal experience on the topic.

In written evaluations, people responded positively to the panel format of the conference, which allowed in-depth discussion and reflection. Attendees loved the sole focus on CSA rather than including other farming issues. For many, the conference was, as one woman put it, “Confirmation of our experience that there is tremendous variation among CSA farms, rather than one true ideal.� Attendees spoke positively, in both written and oral comments, of the incredible opportunity to “network with other farms and farmers� and “share different approaches--both the successes and difficulties.� Participants had the chance to gain insight into CSA from a regional perspective and to develop contacts for the future. Most participants highlighted a strong sense of being part of a larger CSA movement through their attendance at the conference and felt the potential for a supportive network. Even before the conference was over people were talking about the need to have an annual event. The main constructive criticism offered was to allow more chances to participate in the workshops. So many pertinent topics were offered at concurrent workshops that some participants felt they missed out on something in a workshop they could not attend. Hopefully this will be remedied when the audio tapes of all workshops are made available later in November. In addition, the revised CSA Handbook co-written by Robyn Van En and Elizabeth Henderson will contain entries from many of the farms present at the conference. In conclusion, the work completed in the last two years overwhelmingly points to the continued support and enthusiasm for CSA in the Northeast region as a viable farming alternative.