Costs and returns to New England farmers in the farm-to-institution supply chain

Costs and returns to New England farmers in the farm-to-institution supply chain

Summary

As Farm to Institution (FtI) programs in New England expand, new supply chains are being developed to handle the increased flow of regionally-produced goods to regional institutions. While some supply chains deliver product directly to institutions, using no aggregators, distributors, or processors, other supply chains rely upon these additional supply chain actors. These are referred to as “direct” and “intermediated” supply chains, respectively. Each time an additional supply chain actor is added, the costs incurred to perform the supply chain task for which the actor is responsible must be covered. The actor may also charge an additional fee. On the other hand, in the absence of these additional supply chain actors, farmers must absorb the costs associated with performing the supply chain tasks. This research project is developing farmer interviews to identify the transaction characteristics associated with New England farmers’ sales to the institutional market in direct, traditional-intermediated, and coordinated-intermediated supply chain structures. The outcomes of this research will be shared with farmers and FtI practitioners in New England to develop emerging FtI supply chains that explicitly account for transaction characteristics that impact farmers’ profitability. Outcomes will also be used in future research in which a field experiment designed to uncover the transactions costs associated with the various supply chains.  The results from this current research will be important to determining the  distribution of New England farmers’ FtI-specific transaction costs.  The results of the current research and the future experimental work will  assist New England farmers and FtI practitioners in developing FtI supply chains that support farmer profitability

Objectives/Performance Targets

The objectives and methods of this project that have been accomplished to date:

1. Identify farmers’ supply chain transaction costs for direct, intermediated, and coordinated FtI supply chains.

1.1  Share ongoing results and solicit feedback on research methods and strategies with FINE (Farm to Institution New England) partners during FINE Metrics Team conference calls and bi-annual FINE gatherings.

Preliminary results from the interviews have been communicated to the FINE Metrics Advisory Team.  These results have also been used to inform the development of FINE’s 2016 Producer Survey, which will be distributed via a snowball sample through state agencies of agriculture in all six New England states to all producers in New England in early 2016.  Preliminary results have also been discussed with Massachusetts Farm to School staff, who undertook a similar study recently.  In 2015, FINE conducted a Distributor Survey, and results from this research will contribute to questions on future Distributor Surveys, as well.  Preliminary results have been communicated to staff from the John Merck Fund, which funds farm to institution research in New England.  

1.2.  ARMS Data – purchase USDA ARMS (Agricultural Resource Management Survey) 2011 Cost And Returns report data.

The USDA ARMS 2011 Cost and Returns Survey did include a separate question that asked producers to report their sales to institutional purchasers. However, the USDA ARMS 2011 Cost and Returns Survey results will not be useful for our research.  The ARMS Cost and Returns Surveys are designed to elicit detailed information on large commodity growers’ costs and returns.  This has several implications.  The survey sample is weighted towards “commodities” states, or states that produce the nation’s bulk of wheat, corn, cotton, sugar beets, soybeans, etc., and hog, cattle, poultry production.  More farmers in these states are included in the sample.  No New England, or even northeastern, states are large scale producers of the commodities of interest, so these states are included in the sample according to a decreased weight. 

Related to this weighted sampling method, there were so few producers reporting any sales to institutions that ERS is unable to provide anonymous results to us because of the small sample size and reliability of the estimates.  Further, even if we did have access to a sufficiently large set of anonymous results, questions have been raised as to whether the results from a sample weighted towards large scale commodity producing states will be valuable in drawing inferences about our population of producers in New England. ERS will not share the ARMS data when those data could be used for inferences for which the data are not suited.

1.3.  Interviews- Work with FINE to identify farmers to be interviewed.

The team identified potential farmers to interview and the appropriate means of conducting those interviews.  Several unanticipated questions came up during the process of identifying potential farmer interviewees, particularly the issue of how FINE could be included in collecting farmer information in a way that satisfies the University of Massachusetts’ Internal Review Board (IRB) requirements for confidentiality.  The end result of the concerns about confidentiality involve the process of identifying farmers, the different stages at which different people have access to information that has been collected, and how the final list of farmers will be determined.

In coordination with the metrics team, it was  determined that Ms. Fitzsimmons and Mr. Allison would identify key FINE leaders, Ms. Fitzsimmons would speak directly with these individuals to build a list of potential farmer interviewees, and then the metrics team and  Ms. Fitzsimmons, and Dr. Lass would review this list to determine the farmers to prioritize.  This list of potential interviewees would include all farmers that leaders identified, and would include farmers from a range of farm size, products, location, marketing channel, etc.

Ms. Fitzsimmons and Mr. Allison identified key individuals from the FINE leadership team and individual FINE members.   Ms. Fitzsimmons wrote an email request and two attachments (one brief, one detailed) to describe the nature of the request and proposed outcomes, and Mr. Allison sent out the email request to this group.  Ms. Fitzsimmons followed up with individuals. 

The list of potential farmer interviewees will be available to the FINE metrics team, but once the list has been narrowed to those identified to be interviewed, the list and raw interview results will only be accessible to Ms. Fitzsimmons and Dr. Lass. 

The process of identifying farmers agreed to above, and the success in connecting with farmers, asking for a commitment to be interviewed, and then setting an interview date, was more challenging than anticipated.  Because we asked for an introduction, but were not able to share information with FINE about which farmers responded to the introduction and which did not, our ability to follow up was very limited.  The response rate to the introductions was very low, so the inability to ask FINE members to assist with follow up with particular farmers was a challenge.

1.4. – 1.8.  Mail interview questions to farmers one month in advance; Send farmers reminder postcard and email two weeks in advance; Make a reminder call to farmers one week in advance; Interview farmers; Compensate farmers.   These steps all went smoothly.

1.9.  Compile and analyze ARMS data and interview data on supply chain costs, using multinomial, conditional and mixed logistic regression models to determine probabilities associated with market choices. For the reasons described under objective 1.2 above, we will not be doing this.

2. 2. Identify supply chain transaction returns, including those returns that satisfy farmers’ non-financial goals, for direct, intermediated, and coordinated FtI supply chains.

2.1. Share ongoing results and solicit feedback on research methods and strategies with FINE partners during FINE Metrics Team conference calls and bi-annual FINE gatherings.

Preliminary results from the interviews have been communicated to the FINE Metrics Advisory Team.  These results have also been used to inform the development of FINE’s 2016 Producer Survey, which will be distributed via a snowball sample through state agencies of agriculture in all six New England states to all producers in New England in early 2016.  Preliminary results have also been discussed with Massachusetts Farm to School staff, who undertook a similar study recently.  In 2015, FINE conducted a Distributor Survey, and results from this research that will contribute to questions on future Distributor Surveys, as well.  Preliminary results have been communicated to staff from the John Merck Fund, which funds farm to institution research in New England.  

2.2.         ARMS Data – purchase USDA ARMS 2011 Cost And Returns report data – see above.

2.3.         Develop interview questions, including questions eliciting non-sales related returns to farmers, such as increased access to customers, farm tours, CSA sign-ups, etc.

We developed questions, and added an additional objective to test the survey instrument by conducting a pilot interview.  The pilot interview lasted a little over three hours.  The farmer spent a significant amount of time looking at files (both paper and computer) to find accurate answers to our questions.  The farmer was very generous with his time, and among the helpful outcomes of this pilot was a comprehensive list of documents to suggest that interviewees have on hand for the interview.

For further discussion of the instrument design and design process, see Objective 3.3 below.

2.4. – 2.10 Interviews – Work with FINE to identify farmers to be interviewed; Mail interview questions to farmers one month in advance; Send farmers reminder postcard and email two weeks in advance; Make a reminder call to farmers one week in advance; Interview farmers.  For the most part, these steps went smoothly.  See Objective 1.3 above.

2.9. Mail farmer compensation – As per University of Massachusetts guidelines, we compensated farmers directly, in cash, upon the conclusion of the interview.  This method entailed some inconvenience and risk on the part of the interviewer, who was required to travel with large quantities of cash, but was otherwise a very streamlined process.

2.10. Compile and analyze ARMS data and interview data on supply chain returns, using multinomial, conditional.  Removed from project, as per 1-3. Above.

3. 3. Identify other causal factors that relate to supply chain transactions, including farmer preferences, farmer socio-demographic characteristics and farm characteristics.

3.1. – 3.2. and 3.4 – 3.10, please see repeated Objectives above in Objectives 1 and 2.

3.3.         Develop interview questions, including questions eliciting social preferences

We designed the survey to meet many goals.  The survey instrument is intended to be compatible with existing USDA surveys and reports and IRS forms, and additionally include questions that address costs and returns not currently gathered in these existing instruments.  There are multiple purposes for this intention.  Primarily, we will be able to make straightforward comparisons to existing data for those fields that are included in both existing instruments and our instrument.  Farmers will be able to refer to forms that they have already completed in order to answer many of our questions, which will streamline the process and make the interviews go more quickly.  Finally, the questions are designed to complement the existing instruments so that they may be easily inserted into future federal instruments and/ or may serve as a pilot for potential questions, should USDA decide to begin soliciting farm to institution data in future surveys.  

The survey is also designed to be compatible with existing measures of social preferences.  This is more of a challenge, and is the primary reason for requiring a redesign of the first draft of the survey after the completion of the pilot survey.    We originally designed the social preferences section with the intent of using the results to inform a field experiment design.  To achieve this, we reviewed the literature on field experiments, and developed a survey intended to identify a robust Dictator Game to identify farmer social preferences for three social outcomes, as identified in the literature.  In the process of making this design decision, we reviewed and rejected a number of other social preference frameworks.  However, we are no longer intending to use the results of this survey to design an experiment in the near future.  For this reason, we have redesigned this section of the survey to elicit social preferences in a way that is compatible with existing literature.

4. 4. Provide case studies for each of the three supply chains identified – direct, intermediated, and coordinated.

4.1. Share ongoing results and solicit feedback on research methods and strategies with FINE partners during FINE Metrics Team conference calls and bi-annual FINE gatherings. See above Objective 1.1.

4.2. Interviews-Work with FINE to identify farmers to be interviewed and also be willing to serve as a published case study. 

In the course of identifying interviewees and clarifying the purpose of case studies, it became clear that the case studies as proposed did not advance an understanding of farm profitability in farm to institution markets in a way that has not already been demonstrated.  Instead, we have identified three different emerging topics which are important to understanding farmer profitability in these markets, and will be highlighting the experiences of the farms that have been grappling with these topics.  The topics are:  Intermediated sales through non-profit brokers; retail market accessibility and farmer choice to wholesale; the role of farm labor and farm labor regulations.   

Note that we were not able to identify any farms, and therefore did not interview any farmers, who coordinated (in the sense of working together as a co-operative) sales to institutions.  The categorization of “intermediated” sales was not specific enough to identify the relevant aspects of different intermediated supply channels.  However, the idea of establishing a co-operative was mentioned without prompt by three farmers; two suggested establishing a co-operative processing facility, and one suggested a sort of hiring co-operative to address labor issues.

4.3. – 4.7  See Objective 2 above.

4.8. Compile interviews into 3 case studies.  Emerging Topic Studies will be completed for the final report.

4.9. Share case studies with farmers and FINE to verify accuracy.  While we are not writing case studies as originally planned, after each interview, the interview notes were typed and emailed to farmers.  Farmers are given a (flexible) deadline, and asked to review the notes for accuracy, and to edit or redact any information that they consider to be identifying or proprietary. 

4.10. Distribute case studies online.  Emerging Topic Reports will be offered online through FINE and through University of Massachusetts Center for Agriculture and the Environment.

4.11. Submit case studies for publication. Emerging topics will be submitted for publication.

5. 5. Provide results of this research to farmers, distributors, institutions, and FtI advocates in order to inform farmers about this potentially profitable enterprise, to support farm viability, and to enhance FtI markets.

5.1. Share ongoing results and solicit feedback on research methods and strategies with FINE partners during FINE Metrics Team conference calls and bi-annual FINE gatherings.  See Objective 1 above.

5.4. Share tools and materials with farmers and FtI practitioners – Present materials and workshops at conferences for farmers and FtI practitioners.  Results will be shared at the 2016 NH NOFA conference, and additional upcoming conferences.

Accomplishments/Milestones

Accomplishments this year:

1. 1. Submitted survey instrument and contact materials to Institutional Review Board for approval. Received approval.

1. 1. Identified and worked with FINE members to contact 65 farmers across New England.

1. 1. Received 14 agreements to be interviewed; scheduled and completed 11 interviews in 2015.

1. 1. Identified 3 “Emerging Topics” that impact producer profitability in farm to institution markets, which have the potential to affect farmers across the region as farm to institution sales are encouraged to grow.

1. 1. Contributed to the development of the FINE Producer Survey, to be distributed to all farms in New England (not just those that have sold to institutions).

1. 1. Shared preliminary results with the Agriculture of the Middle Working Group, Massachusetts Farm to School, and Northeast Fruit and Vegetable Conference (2015).

1. 1. Used preliminary results to inform a theoretical model for farm viability in the farm to institution supply chain.

1. 1. Secured funding to test the theoretical model in a pilot experimental setting. If successful, we will seek to run experiments in the field.

Lessons learned so far:

1. 1. While it is a little disheartening that the ARMS data is not usable for this study, the process of pursuing that data has confirmed that there are no good resources that identify costs and returns for small to medium sized farms – and none that identify these costs and returns according to marketing channels. As funders, the USDA, and state departments of agriculture continue to promote regionalization and this kind of information will be very important in determining how or whether regional-scale supply chains can support small and medium sized farm operations in New England. 

1. 1. Practitioners are excited about this research. In conversations at FINE events, it is clear that practitioners see a need to have a better idea of how farmers can participate in farm to institution supply chains and remain viable operations.  Practitioner feedback has been very helpful in identifying the anticipated costs and returns that are unique to institutional sales.

1. 1. Feedback from practitioners suggests that most farms that sell to institutions also sell to other marketing channels. We can use this knowledge to ask questions in interviews that provide us with the comparisons between costs and returns in different marketing channels, in order to supplement the information that we originally anticipated would come from the ARMS data.

1. 1. In states with a single large “lead” organization, it has been fairly straightforward to identify farmers to be interviewed. However, in other states, like Maine, identifying which practitioners may have a good idea of potential interviewees has been more difficult. 

1. 1. Farm to Institution practitioners keep very disparate sets of information about farmers, and many large regional organizations that work closely with farmers do not keep information about the marketing channels through which farmers sell their products. In terms of our research, the result of this is that practitioners with relationships and informal knowledge about farmers in their region have been better able to help us identify farmers and provide useful feedback on our research.  

1. 1. The amount of interview time required to complete the interviews was greater than anticipated, and the number of documents that farmers were asked to have on hand is larger than expected.

1. 1. Conducting a pilot test of the survey instrument is critical. While the results of the pilot have set us back on the timeline, the information that we collected through this pilot (both in terms of survey data and the interviewing process) has helped us sidestep some potential and serious challenges.  

1. 1. The IRB approved process of asking FINE members to identify farmers but to maintain confidentiality regarding which farmers agreed to be interviewed and which did not led to some real challenges for follow up. A better process might have been to allow the referring FINE members to have access to the ongoing updates and responses from farmers.  The response rate was about 20%, which was lower than expected.  Timing was likely an additional factor. For other reasons, we were constrained to interview farmers in the autumn, which of course is not ideal as it overlaps with the harvest season.

1. 1. As expected, every farmer keeps records in a unique way. The list of documents we compiled from the pilot interview was of great help, and we also created a “cheat sheet” that linked each type of record with the survey section that it related to.  In general, the survey was well-designed to flow in a natural way with the documents that farmers keep. 

1. 1. The survey was very thorough – many sections were not relevant for certain farmers, but captured something important for one or two others.

1. 1. Originally, we had planned on “clustering” interviews, so that geographically proximate interviewees were scheduled one after the other, to minimize travel costs. This quickly proved to be unrealistic – it was a challenge to schedule interviews with any given farmer, much less to coordinate with many farmers.  This greatly increased the overall travel costs, and time spent traveling, and therefore the amount of time required to complete a number of interviews.

1. 1. Farmers, even those who did not agree to be interviewed, were excited about this research, and indicated that they felt that the question of profitability in the FtI market was important and overdue.

1. 1. Farm businesses are structured in a number of ways. A few farms are considered non-profits and did not file a Schedule F to the IRS, for example.  Other farm businesses maintained a number of corporate entities, so information needed to be pooled in order to fit the survey design.  In our analysis, we use USDA’s classification of farms according to gross income.  Interestingly, only one interviewed farm is classified as a “mid-sized farm,” or a family farm with a gross income between $350,000 and $499,999 a year.

1. 1. Farmers do indicate that they have “social preferences” for farm to institution outcomes, and some farmers do adjust their pricing to offer products at a lower price to institutional customers.

1. 1. Overall, percent of sales to institutions are very small.  Farmers, particularly those with access to a strong retail market, do not necessarily consider institutional wholesale sales as distinct from general wholesale sales, which is an important consideration for farm to institution advocates.  Farmers who consider farm to institution sales as a growing market are those that are further from population centers, or are located in a region that is targeted by preferential purchasing mandates. 

1. 1. Transaction costs that accrue to farmers do appear to be minimized in supply chains where non-profit brokerages are present.

Impacts and Contributions/Outcomes

1. 1. The beneficiaries of this project are New England farmers who participate or are interested in participating in the wholesale institutional market as well as other Farm to Institution practitioners, including those administering state and local programs and institution administrators. This project has had a significant impact on the development of a region-wide survey developed by FINE for New England producers. Additional upcoming outcomes of this project (planning materials, emerging topics reports, and publications) are eagerly anticipated by a growing body of practitioners. 

1. 1. Lessons learned from developing and conducting the interviews, as well as preliminary results, have been shared with the FINE Metrics Team. These lessons have contributed to FINE’s development of a new annual survey of New England producers, which will be distributed online in January 2016 through state departments of agriculture, extension services, and non-profits.  The survey is designed to establish a baseline of producers’ activities in, and opinions about, the farm to institution market in New England (including sales and profitability), and will be conducted regularly to measure progress over time.

1. 1. A workshop, “GroundUp Metrics to Assess Impact” was presented at the 2014 Farm to Cafeteria conference in Austin, TX. Preliminary results were presented informally at the Northeast Fruit and Vegetable Conference (2015) as part of the FINE trade show table.

1. 1. We will offer the results of this project in printed form through a UMass Center for Agriculture website and the FINE website. To date, the methodology has been shared with the FINE Metrics Team, Massachusetts Farm to School, and the Agriculture of the Middle group. The methodology is available to be shared to all other interested parties via FINE’s Metrics Dashboard, which is in development (a prototype was shared at the 2015 NESAWG conference in Saratoga Springs, NY).  Tools for practitioners are in development, including a “Understanding Farmer Profitability in the Farm to Institution Supply Chain”.  Tools for farmers that highlight the Emerging Topics are in development, working with the FINE Communications coordinator to present the findings in a format similar to the Planning for Wholesale Success workbook and training.   These materials will be offered to the FINE metrics Team for comment prior to publication.  “Farmers’ Profitability in the Farm to Institution Market: Lessons from New England” is in process for publication in academic journals including the Journal of Extension.

1. 1. Preliminary results have been shared on an ongoing basis with FINE members. Final results will be made available on the FINE Metrics Dashboard, which is in development.

1. 1. The methodology, outcomes, and tools will also be shared with farmers and Farm to Institution/ School/ Cafeteria and food systems practitioners via regional and national conferences and workshops. Slide presentations will be offered online. Target conferences include NESAWG Regional Conference (2016), NOFA NH Winter Conference (2015), biannual Farm To Cafeteria Conference, (2016), Harvest New England conference, (2016), Northeast Fruit and Vegetable Conference (2017). To date, preliminary results have been shared at the 2014 Farm to Cafeteria Conference, and at the Northeast Fruit and Vegetable Conference (2015).

1. 1. Preliminary results indicate that transaction costs (the costs of gathering information, negotiating terms, and monitoring the implementation of those terms) are important to a farmers’ decision to participate in this market, but are not the only factors. The importance of these costs depends on a number of factors – the size of the farm operation, access to a convenient institutional market, and the role of wholesale in the overall farm operation.   

1. 1. A number of new non-profit local food brokerages have started up in the last few years in New England; a couple have existed for many years. These brokerages often have online platforms through which farmers can list products and prices, and buyers, including wholesalers, can order, receive delivery, and be billed.  Institutional wholesalers make up a small portion of these buyers – from the perspectives of farmers who sell to institutions through these brokerages this is an ideal way to avoid transaction costs and make the institutional wholesale market viable.  There are a number of questions regarding the ability of these sites to become break-even going concerns in the medium run (that is, to not rely on grant funding), and also to preserve a price above wholesale for farmers. 

1. 1. Proximity to an accessible retail market appears to significantly reduce farmers’ interest in selling wholesale, to institutions or otherwise. This is not surprising from an economic standpoint – the price a farmer receives for a locally differentiated product in the direct retail market is likely to be significantly higher than that received in a wholesale market, and for farmers who operate CSAs, farm stands, or sell through farmers’ markets near population centers, the transaction costs for direct sales are sunk.  As some of the larger farms that sell to institutions near cities invest in expanding on-farm retail infrastructure, they are likely to move away from wholesale in the regions where demand is increasing (that is, where there is a higher density of institutions).  At the same time, producers who are very far from accessible retail markets are faced with few options for distribution to those urban areas.  This finding gives weight to the call some are making to regionalize “local food” and invest in regionalized food systems infrastructure.     

1. 1. From the farmers’ perspective, farm labor is a challenge, and a growing market for “off season” wholesale institutional sales is likely to push the limits of that challenge. Three issues related to institutional sales intersect with labor issues:   A number of farmers are responding to institutional food service needs for lightly processed products, like peeled and cubed winter squash;   Food systems advocates are encouraging the regionalization of “local foods”, and; In order to meet the scale requirements, farmers are aggregating product with other farmers on an ad hoc basis to meet delivery volumes.  Each of these three innovations in the farm to institution market, however, could not be undertaken by a farm operation that employs H-2A guest workers, as each of the tasks is in violation of Department of Labor regulations limiting the tasks that may be performed under this visa.  While the wages paid to these workers are similar to the wages paid to non-H-2A guest workers, the challenge is to actually find workers who are available for a farm season that extends to the end of the calendar year at least, and who are willing to choose the work.   Among the farmers who agreed to interviews, as well as farmers with whom we spoke but did not end up interviewing formally, issues of aggregation, processing, and regionalization were repeatedly offered as next steps to building a supply chain for institutional foods. 

Collaborators: