Skip to main content Accessibility help
×
Home
Hostname: page-component-59b7f5684b-s82fj Total loading time: 0.837 Render date: 2022-10-03T03:57:15.795Z Has data issue: true Feature Flags: { "shouldUseShareProductTool": true, "shouldUseHypothesis": true, "isUnsiloEnabled": true, "useRatesEcommerce": false, "displayNetworkTab": true, "displayNetworkMapGraph": false, "useSa": true } hasContentIssue true

Participation in cost-offset community-supported agriculture by low-income households in the USA is associated with community characteristics and operational practices

Published online by Cambridge University Press:  13 April 2022

Karla L Hanson*
Affiliation:
Department of Public and Ecosystem Health, Shurman Hall, Cornell University, Ithaca, NY14853, USA
Lynn Xu
Affiliation:
Department of Public and Ecosystem Health, Shurman Hall, Cornell University, Ithaca, NY14853, USA
Grace A Marshall
Affiliation:
Department of Public and Ecosystem Health, Shurman Hall, Cornell University, Ithaca, NY14853, USA
Marilyn Sitaker
Affiliation:
The Evergreen State College, Olympia, WA, USA
Stephanie B Jilcott Pitts
Affiliation:
Brody School of Medicine, East Carolina University, Greenville, NC, USA
Jane Kolodinsky
Affiliation:
Center for Rural Studies, University of Vermont, Burlington, VT, USA
April Bennett
Affiliation:
Cornell Cooperative Extension of Jefferson County, Watertown, NY, USA
Salem Carriker
Affiliation:
University of South Carolina, School of Medicine, Columbia, SC, USA
Diane Smith
Affiliation:
Washington State University, Extension of Skagit County, Burlington, WA, USA
Alice S Ammerman
Affiliation:
Center for Health Promotion and Disease Prevention, University of North Carolina at Chapel Hill, Chapel Hill, NC, USA
Rebecca A Seguin-Fowler
Affiliation:
Texas A&M AgriLife Research, College Station, TX, USA
*
*Corresponding author: Email kh289@cornell.edu
Rights & Permissions[Opens in a new window]

Abstract

Objective:

Subsidised or cost-offset community-supported agriculture (CO-CSA) connects farms directly to low-income households and can improve fruit and vegetable intake. This analysis identifies factors associated with participation in CO-CSA.

Design:

Farm Fresh Foods for Healthy Kids (F3HK) provided a half-price, summer CO-CSA plus healthy eating classes to low-income households with children. Community characteristics (population, socio-demographics and health statistics) and CO-CSA operational practices (share sizes, pick up sites, payment options and produce selection) are described and associations with participation levels are examined.

Setting:

Ten communities in New York (NY), North Carolina (NC), Vermont and Washington states in USA.

Participants:

Caregiver–child dyads enrolled in spring 2016 or 2017.

Results:

Residents of micropolitan communities had more education and less poverty than in small towns. The one rural location (NC2) had the fewest college graduates (10 %) and most poverty (23 %) and poor health statistics. Most F3HK participants were white, except in NC where 45·2 % were African American. CO-CSA participation varied significantly across communities from 33 % (NC2) to 89 % (NY1) of weeks picked up. Most CO-CSA farms offered multiple share sizes (69·2 %) and participation was higher than when not offered (76·8 % v. 57·7 % of weeks); whereas 53·8 % offered a community pick up location, and participation in these communities was lower than elsewhere (64·7 % v. 78·2 % of weeks).

Conclusion:

CO-CSA programmes should consider offering a choice of share sizes and innovate to address potential barriers such as rural location and limited education and income among residents. Future research is needed to better understand barriers to participation, particularly among participants utilising community pick up locations.

Type
Research Paper
Creative Commons
Creative Common License - CCCreative Common License - BY
This is an Open Access article, distributed under the terms of the Creative Commons Attribution licence (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted re-use, distribution, and reproduction in any medium, provided the original work is properly cited.
Copyright
© The Author(s), 2022. Published by Cambridge University Press on behalf of The Nutrition Society

Fruit and vegetable (FV) consumption is associated with reduced risk for chronic disease and other positive health outcomes(Reference Boeing, Bechthold and Bub1,Reference Srinath Reddy and Katan2) . However, most US populations do not consume recommended amounts of FV(Reference Lee-Kwan, Moore and Blanck3). One approach to improving FV intake is community-supported agriculture (CSA), which directly connects local farms to consumers by allowing community members to purchase a share of a farm’s anticipated harvest(Reference Wharton, Hughner and MacMillan4). Because a lump sum payment is typically required before the growing season begins, CSA may be more accessible to households with higher incomes(Reference Allen5Reference Hinrichs and Kremer9). However, the economic and health benefits of CSA participation for low-income households have been documented(Reference Galt, Bradley and Christensen10,Reference Miewald, Holben and Hall11) . To broaden access to CSA, some programmes offer a subsidy, or cost-offset (CO-CSA), to low-income participants who are at greater risk of food insecurity and poor nutritional intake(Reference White, Jilcott Pitts and McGuirt12,Reference Guthman, Morris and Allen13) . Understanding the community characteristics and operational practices that support participation in CO-CSA would provide useful implementation information to farms and other CO-CSA programme operators.

To our knowledge, twelve prior studies have examined CO-CSA effectiveness in improving FV access and intake, as well as household food security(Reference Abbott14Reference Hanson, Volpe and Kolodinsky25) and most, but not all, report beneficial changes in outcomes. For example, a pilot study reported that all participants consumed a greater variety of vegetables, learned new methods for cooking and preparing vegetables and liked new vegetables after participating in the CO-CSA(Reference Izumi, Higgins and Baron20). Descriptive studies suggest that CO-CSA participation is associated with improved FV access(Reference White, Jilcott Pitts and McGuirt12,Reference Izumi, Higgins and Baron20) and intake(Reference Hanson, Kolodinsky and Wang18,Reference Hanson, Volpe and Kolodinsky25) , and that participants bought more fresh produce, tried new recipes and cooked with more vegetables after joining a CO-CSA(Reference Andreatta, Rhyne and Dery16). Longitudinal studies of CO-CSA more often reported increases in FV intake(Reference Agboola15,Reference Berkowitz, O’Neill and Sayer17,Reference Hoffman, Agrawal and Wirth19,Reference Johnson, Beaudoin and Smith21,Reference Wilkins, Farrell and Rangarajan23) than no changes(Reference Abbott14,Reference Izumi, Higgins and Baron20,Reference Quandt, Dupuis and Fish22) . Findings from one randomised controlled trial supported the effectiveness of CO-CSA relative to unconditional cash transfer in terms of improved diet quality and reduced food insecurity(Reference Berkowitz, O’Neill and Sayer17). Taken together, these studies suggest that CO-CSA participation can have positive effects on FV access, dietary intake and related behaviours.

These studies examined CO-CSA implemented in different community contexts and enrolled participants who had varying characteristics. Seven studies operated in urban areas(Reference Abbott14,Reference Agboola15,Reference Hoffman, Agrawal and Wirth19Reference Quandt, Dupuis and Fish22,Reference Kato and McKinney24) , and five in predominately rural and micropolitan communities(Reference Andreatta, Rhyne and Dery16Reference Hanson, Kolodinsky and Wang18,Reference Wilkins, Farrell and Rangarajan23,Reference Hanson, Volpe and Kolodinsky25) . Almost half enrolled participants from communities that were predominately African American and/or Hispanic(Reference Abbott14,Reference Agboola15,Reference Hoffman, Agrawal and Wirth19,Reference Quandt, Dupuis and Fish22,Reference Kato and McKinney24) . Most CO-CSA participants were women (71–92 %) of all ages, but sometimes were limited to caregivers of children(Reference Andreatta, Rhyne and Dery16,Reference Hanson, Kolodinsky and Wang18,Reference Hoffman, Agrawal and Wirth19,Reference Hanson, Volpe and Kolodinsky25) , Head Start staff(Reference Hoffman, Agrawal and Wirth19) or focussed on older adults(Reference Johnson, Beaudoin and Smith21,Reference Kato and McKinney24).

The CO-CSA programmes examined in prior studies also varied in operational practices. Five studies examined CO-CSA programmes that were free(Reference Agboola15,Reference Andreatta, Rhyne and Dery16,Reference Johnson, Beaudoin and Smith21,Reference Quandt, Dupuis and Fish22,Reference Kato and McKinney24) , but others required either a contribution based on a fixed percentage of share price – 45 %(Reference Abbott14), 50 %(Reference Hanson, Kolodinsky and Wang18,Reference Wilkins, Farrell and Rangarajan23,Reference Hanson, Volpe and Kolodinsky25) or 67 %(Reference Hoffman, Agrawal and Wirth19) – or a sliding scale (43–75 %) based on ability to pay(Reference Berkowitz, O’Neill and Sayer17,Reference Izumi, Higgins and Baron20) . CO-CSA programmes varied in the options offered to participants, with multiple farm partners resulting in multiple share sizes(Reference Abbott14,Reference Andreatta, Rhyne and Dery16) , multiple pick up locations(Reference Andreatta, Rhyne and Dery16,Reference Wilkins, Farrell and Rangarajan23) and/or delivery(Reference Andreatta, Rhyne and Dery16,Reference Johnson, Beaudoin and Smith21,Reference Kato and McKinney24) , or these options were offered at some farms but not others(Reference Hanson, Kolodinsky and Wang18,Reference Hanson, Volpe and Kolodinsky25) . Four of five CO-CSA programmes that partnered with just one farm offered only one option for share size and pick up location(Reference Agboola15,Reference Hoffman, Agrawal and Wirth19,Reference Izumi, Higgins and Baron20,Reference Quandt, Dupuis and Fish22) . Three studies assessed CO-CSA programmes that operated with market-style selection of produce items(Reference Berkowitz, O’Neill and Sayer17,Reference Izumi, Higgins and Baron20,Reference Kato and McKinney24) , or in which item selection was available at some farms(Reference Hanson, Kolodinsky and Wang18,Reference Hanson, Volpe and Kolodinsky25) . Of the seven studies of CO-CSA programmes that required payment, five accepted Supplemental Nutrition Assistance Programme (SNAP) benefits as payment(Reference Abbott14,Reference Berkowitz, O’Neill and Sayer17,Reference Hoffman, Agrawal and Wirth19,Reference Izumi, Higgins and Baron20,Reference Wilkins, Farrell and Rangarajan23) and two accepted SNAP at some farms(Reference Hanson, Kolodinsky and Wang18,Reference Hanson, Volpe and Kolodinsky25) .

Farm Fresh Foods for Healthy Kids (F3HK), the focus of this manuscript, was a summer growing season CO-CSA programme (mean = 21 weeks, interquartile range 19–23)(Reference Garner, Jilcott Pitts and Hanson26), provided at 50 % cost-offset (half-price)(Reference Seguin, Morgan and Hanson27). The CO-CSA was supported by nine CSA-tailored, in-person nutrition education lessons that featured seasonal produce items via food tastings, demonstrations, hands-on cooking activities, handouts and recipes, and by providing participants with two to four larger cooking tools (e.g. food processor and crockpot)(Reference Seguin, Morgan and Hanson27). The F3HK intervention was implemented in the context of a randomised controlled trial with 1:1 random assignment of child–caregiver dyads to intervention and control groups(Reference Seguin, Morgan and Hanson27). F3HK reported overall improvements relative to control in caregiver nutrition attitudes, self-efficacy, FV intake, skin carotenoids, preparation of FV as snacks for children and household food security(Reference Seguin-Fowler, Hanson and Jilcott Pitts28). However, F3HK was implemented in small and micropolitan communities across four states, each with a unique context for implementation. Although F3HK required farms to receive weekly payments from participants and to accept SNAP as payment(Reference Seguin, Morgan and Hanson27), all other operational practices were decided by each partner farm.

The focus of this manuscript is to explore the context of F3HK implementation with respect to community and participant characteristics and CO-CSA operational practices and to examine associations with participation levels. First, we describe the population characteristics of communities in which the F3HK intervention was implemented. Second, we explore differences in F3HK participant characteristics across communities and states and contrast participant and population characteristics. Third, we describe CO-CSA operational practices at F3HK partner farms and explore differences across communities and states. Fourth, we examine associations between community characteristics, participant characteristics, CO-CSA operational practices and participation levels (i.e. percent of CO-CSA weeks picked up, percent of CO-CSA nutrition lessons attended). These analyses inform recommendations for the development and implementation of future CO-CSA programmes in varied settings.

Methods

Setting and participants

Target locations for F3HK implementation were small and micropolitan (≤ 50 000 population) communities in New York (NY), North Carolina (NC), Vermont (VT) and Washington (WA) in the USA. Each community also had to be served by a partner farm experienced in CSA and a nutrition educator (through cooperative extension in three of four states). F3HK was implemented in a total of twelve communities. In most communities, one farm provided the CO-CSA. However, in VT1 and WA1 participants chose between two different farms, and in all communities in NC, participants were selected from among the same farms. Throughout this article, ‘community’ refers to the location and ‘farm’ refers to the CO-CSA provider.

In spring 2016 and 2017, F3HK enrolled English-speaking adult caregivers of a child 2–12 years of age who had self-reported household income ≤185 % of the Federal Poverty Level and had not participated in CSA for the past 3 years. Participants agreed to use SNAP benefits or money to pay for the CO-CSA share weekly and to attend nine CSA-tailored nutrition education classes(Reference Seguin, Morgan and Hanson27). A total of 685 caregivers were screened for eligibility, 542 (79·1 %) were eligible and 305 of those enrolled (56·3 %), of which 148 were assigned to the intervention group(Reference Seguin-Fowler, Hanson and Jilcott Pitts28). Two communities (one each in NC and VT) were excluded from this analysis due to low F3HK enrolment (≤7 participants, >1 sd below mean sample size of 12), and the remaining 137 intervention participants in 10 communities are the focus of this analysis. Figure 1 depicts the states, communities and participants included in these analyses.

Fig. 1 F3HK implementation states, communities, and participants included in analyses

Measures

Six community characteristics and seven county health statistics were obtained from publicly available data for the approximate time period of F3HK implementation. Community characteristics included population(29), percent of adult population with at least a bachelor’s degree(30), percent of households in poverty, percent of children in poverty, race and ethnicity(31). We subsequently categorised towns as small (population ≤15 000) or micropolitan (population >15 000). Rural-urban continuum (RUCA 2010) codes also were recorded(32). County health statistics included percent of persons food insecure(33), and percent of adult residents with diabetes, high cholesterol, high blood pressure, overweight, obesity(34) and cancer(35).

Seven CO-CSA operational practices were extracted from partnership agreements, farm websites and personal communications and dichotomised for analysis: location (outside or within the participant community), length of summer CSA (≤21 weeks or longer), number of pick up sites offered (one or multiple), number of share sizes offered (one or multiple), payment options (credit or debit cards accepted or not) and whether market-style self-selection of produce items or ‘u-pick’ options were offered or not. U-pick allows consumers to harvest larger quantities of certain produce items themselves. Participant selection of type of pick up site (farm, community location or farmer’s market/farm stand) and type of share size (small (≤8 produce items/week) or large (>8 items)) were recorded as part of a baseline survey.

Participant characteristics were recorded during eligibility screening or online baseline survey. Caregiver characteristics included caregiver sex, general health, marital status, education level, employment status, race and ethnicity. Child’s age, sex and general health also were reported by the caregiver. Household characteristics included number of adults and number of children in household, annual household income and receipt of food assistance benefits through the Special Supplemental Nutrition Program for Women, Infants and Children or SNAP in the past month(Reference Seguin, Morgan and Hanson27). Few participant characteristics were missing (maximum 0·7 %), and missing items are noted on tables.

Participation levels were assessed as: (1) percent of CO-CSA weeks picked up as recorded on logs maintained by partner farms and (2) percent of CSA lessons attended as recorded on lesson sign-in sheets(Reference Garner, Jilcott Pitts and Hanson26).

Analyses

Community characteristics and county health statistics were rounded before reporting to maintain the confidentiality of F3HK communities. Descriptive statistics were used to summarise participant characteristics, CO-CSA options selected by participants and participation levels by community and state. Pearson’s chi-square, Fisher’s exact tests and one-way and Welch’s ANOVA with Bonferroni correction were used to test for differences across communities and states. One-way ANOVA was used to test differences in participation levels across community size, participant characteristics and CO-CSA operational characteristics. Two farms (NY4 and VT2) did not provide reliable CO-CSA pick up data, and the twenty-five participants at these locations were excluded from analyses of weeks picked up. All analyses were performed using SPSS version 26 (IBM Corp.). Results are reported at a 95 % confidence level.

Results

Community and participants characteristics

Community characteristics varied across the 10 F3HK communities included in this analysis (Table 1). Populations ranged from 7000 to 60 000 (just above the upper bound typically used when classified as micropolitan). The largest communities (60 000, 50 000 and 40 000 population) also had the most education (77 %, 44 % and 53 % held a college degree, respectively). The two communities in NC were distinct from one another: NC1 had the highest percentage of adults with at least a bachelor’s degree (77 %), whereas NC2 had the lowest rate of college graduates (10 %) and the highest rate of poverty (23 % of households). The smallest community (NY4) had the lowest rate of poverty (4 % of households).

Table 1 Characteristics of communities where F3HK was implemented

Sources:

* US Census Bureau. Total population, 2011–2015 American community survey (ACS) 5-year estimates: US Census Bureau; 2015 (Available from: https://data.census.gov/cedsci/).

US Census Bureau. Population density, household income, families in poverty, people in poverty, race, ethnicity, 2014–2018 American community survey (ACS) 5-year estimates: US Census Bureau; 2018 (Available from: https://data.census.gov/cedsci/).

US Census Bureau. Educational Attainment, 2015–2019 American Community Survey (ACS) 5-Year Estimates: US Census Bureau; 2019 (Available from: https://data.census.gov/cedsci/).

§ Feeding America 2017 (Available from: http://map.feedingamerica.org/).

Centers for disease control and prevention, 2017 BRFSS survey 2017 (Available from: https://www.cdc.gov/brfss/annual_data/annual_2017.html).

Centers for Disease Control and Prevention and National Cancer Institute. Cancer Incidence, 2013–2017: Centers for Disease Control and Prevention and National Cancer Institute; 2017 (Available from: https://gis.cdc.gov/Cancer/USCS/DataViz.html).

Residents of most communities were predominantly white and non-Hispanic, except in NC2 where 22 % of residents were Black or African American and 43 % were Hispanic and in WA2 where 30 % of residents were Hispanic. NC2 also had the highest rates of diabetes (12 %), high cholesterol (36 %), high blood pressure (38 %), overweight (68 %) and obesity (32 %) but also the lowest rate of cancer (438 per 100 000). All NY locations had high cancer rates (478–541 per 100 000). Across all communities, 11–13 % of households were food insecure.

Some F3HK participant characteristics varied significantly across communities and states (Table 2) and mirrored some of the overall community differences described above. For example, participant race and household income differed significantly by state; NC had the most Black or African American participants (45·2 %). Population data showed that both NC2 and WA2 had high percentages of Hispanic residents (43 % and 30 %, repectively), and both recruited relatively low percentages of Hispanic participants (4 participants or 25 %, and 1 or 9 %, respectively). NC was also the state with the lowest percentage of participants with household income ≥$25 000 (29·0 %). Household income also varied significantly across communities, with NC2 having a notably low percentage of participants with income ≥$25 000 (12·5 %). Three respondents in NC and three in VT (10 % each) had a child in fair or poor health whereas none did in either NY or WA.

Table 2 F3HK participant characteristics by community and state

Differences across communities and states were tested using Pearson’s chi-square or Fisher’s exact tests all with Bonferroni correction.

* Indicates difference at > 95 % confidence.

One observation missing from VT1 for this measure.

CO-CSA operational practices

Four farms (30·8 %) were located within the participant community, two of which were within VT1 (Table 3). Six farms (46·2 %) operated summer CSA shares lasting 22+ weeks, all of which were in NY and NC. Most farms offered multiple share sizes (9 or 69·2 %) or a community pick up location (7 or 53·8 %), and many offered market-style produce selection (6 or 46·2 %), payment by credit/debit card (6 or 46·2 %) and multiple pick up locations (4 or 30·8 %). When participants were offered a choice of share sizes (n 85), the majority (63·5 %) selected a larger share (>8 produce items/week); and, when offered multiple locations for CO-CSA pick up (n 41), most (61·0 %) chose a community pick up site (data not shown).

Table 3 Cost-offset community-supported agriculture (CO-CSA) operational practices by community

* Partner farms shared enrolment of participants in NC1 and NC2.

Farms in VT and WA offered the most flexibility in share size and pick up options, but pick up locations often shifted over time. For example, in WA1 pick up initially occurred at a public housing complex and shifted to a downtown non-profit serving families with low incomes for the second year. In WA2, one farm initially held pick up at a downtown visitors’ information center and the other at the community programmes office of the local hospital, but both farms shifted for the second year (the first to the local hospital and the second away from the hospital and to their own farm stand). Other states offered fewer options. Almost all NY farms offered a choice in share size, but no NY farm initially offered a community pick up site or a choice of pick up locations. In NY2 however, pick up locations also shifted over time to include a community location at which participants could pick up a missed share and later the farm used that location exclusively when the farmers’ market closed for the season. In NC, on the other hand, no farm offered a choice of share sizes, but two of three farms offered a choice of pick up locations and all three offered a community pick up site.

Participation level

Participation varied significantly across communities: participants picked up the CO-CSA share 33–89 % of weeks and attended 15–58 % of the education lessons (Table 4). In most communities, participants picked up their CSA share three-quarters of the weeks or more often. CO-CSA participation was lowest in NC2 (33·4 %) and highest in NY1 (89·2 %). Attendance at nutrition education classes also differed significantly across communities, with the lowest attendance in NC2 (15·3 %) and highest attendance in NC1 (57·8 %). CO-CSA pick up and class attendance behaviours were aligned in NC2 (low participation) and in NC1 and VT1 (high participation). In five communities, however, percentages for class attendance were 40–60 percentage points below CO-CSA pick up participation level (NY1–3 and WA1–2). There were no significant differences in participation level across states.

Table 4 Mean participation level by community and state

CO-CSA, cost-offset community-supported agriculture.

* Indicates difference > 95 % confidence.

Differences in means were tested using one-way and Welch’s ANOVA with Bonferroni correction.

Compared to households in small towns, households in micropolitan communities had significantly higher participation in the CO-CSA (80·6 v. 54·4 % of weeks) and education classes (36·7 v. 24·2 % of lessons; Table 5). CO-CSA participation also was higher among households that included at least two adults (78·5 v. 54·5 % of weeks), included a young child (80·3 v. 62·6 %), and those with incomes at least $25 000/year (84·5 v. 57·1 %). Caregivers with a college education (82·6 v. 57·8 %) and those who were married (84·5 v. 60·6 %) also picked up CO-CSA shares a greater percentage of weeks than their counterparts. Married caregivers also attended significantly more education lessons than unmarried caregivers (38·2 v. 25·3 % of lessons), which was the only variation in lesson attendance across any participant characteristic. CO-CSA participation also was notably lower in the six households that included a child in fair or poor health (38·9 v. 72·3 % of weeks).

Table 5 Mean participation level by community size, participant characteristics, and cost-offset community-supported agriculture (CO-CSA) operational practices

Differences in means were tested using t-tests with Bonferonni correction.

* Indicates difference at > 95 % confidence.

One observation missing for this measure in both analyses.

When considering CO-CSA operational characteristics, CO-CSA participation level was higher when farms offered multiple share sizes (76·8 v. 57·7 % of weeks) and lower when farms offered a community pick up site (64·7 v. 78·2 % of weeks; Table 5). Farm location in the community, longer summer CSA length, multiple pick up sites, payment by credit/debit card, market-style selection of produce and u-pick options all were not associated with participation level. Attendance at education lessons did not differ by any CO-CSA operational practices.

Discussion

This study reported significant differences in CO-CSA and education participation levels across communities but did not vary across the four states in which F3HK was implemented. This provides novel evidence that suggests that the local setting may matter more in supporting participation than does the state context. Further, we identified four inter-related characteristics (micropolitan location, education, income and spouse or other adult in the household) and two distinct operational practices (offering multiple share sizes and community pick up) that were associated with CO-CSA participation level and should be considered when adapting and implementing CO-CSA for rural and micropolitan communities. Five prior studies of CO-CSA took place in predominately rural and micropolitan areas(Reference Andreatta, Rhyne and Dery16Reference Hanson, Kolodinsky and Wang18,Reference Wilkins, Farrell and Rangarajan23,Reference Hanson, Volpe and Kolodinsky25) , but these studies provided few details about their local contexts. A few prior CO-CSA studies documented widely varying participation levels(Reference Abbott14,Reference Berkowitz, O’Neill and Sayer17,Reference Hoffman, Agrawal and Wirth19,Reference Quandt, Dupuis and Fish22) , but only one was in a micropolitan community(Reference Berkowitz, O’Neill and Sayer17), none operated CO-CSA in multiple states and no study examined how community characteristics, participant characteristics or programme operations were associated with participation levels.

Characteristics that may support CO-CSA participation

Participants in micropolitan communities had higher participation levels in both the CO-CSA and the CSA-tailored education classes. To our knowledge, no prior research contrasted CO-CSA participation levels in rural and micropolitan communities. Our data provide novel evidence that residents of micropolitan areas may be better able to participate in CO-CSA than their rural counterparts. We noted some population characteristics that differed between micropolitan and rural communities (particularly education) which may contribute to this difference. However, across all communities, more F3HK participants had finished college than was typical in their communities, even in NC2 where a college degree was rare. Three prior CO-CSA studies in rural and micropolitan areas also noted high levels of education among participants(Reference Berkowitz, O’Neill and Sayer17,Reference Hanson, Kolodinsky and Wang18,Reference Hanson, Volpe and Kolodinsky25) . This suggests that CO-CSA programmes may have more difficulty reaching residents with lesser education either due to recruitment approaches or because programme operations do not meet their needs. Prior research has documented low levels of awareness of CSA among both urban(Reference Cotter, Teixeira and Bontrager36) and rural(Reference Hanson, Garner and Connor37) residents with low incomes. Since awareness of and knowledge about CSA is a necessary precursor to participation(Reference Zepheda and Deal38), a lack of familiarity with CSA may inhibit enrolment. Extension educators could conduct community-wide CSA awareness activities to support reach among those with fewer resources. Future research should explore methods for improving reach of CO-CSA to residents with less formal education.

All F3HK participants had incomes at 185 % Federal Poverty Level or below, but participants who had relatively more education or other resources (higher income, and spouse or other adult living in the household) picked-up CO-CSA shares a greater percentage of weeks than their counterparts with fewer resources. Married participants also attended education sessions more often. This suggests that, even if successfully recruited to a CO-CSA, participants from low-income households may need both a cost-offset and relatively more household resources to fully participate. Conversely, CO-CSA programmes may need to provide a cost-offset and other resources or supports to facilitate full participation. F3HK provided nutrition education and cooking tools as additional supports, but the results presented here suggest that more foundational resources like money and time were needed for some participants to fully participate. Future research is needed to disentangle the effects of these inter-related resources on participation and to explore approaches to meeting these needs.

Our data also illustrate that attendance at education sessions was consistently lower than CO-CSA pick up participation, and in five communities was 40–60 percentage points lower. This suggests that local context may matter differently for pick up and attendance, and programme design may require different adaptations for CO-CSA and for education. Future CO-CSA programmes should aim to identify local factors that could potentially hinder participation levels (rural location, low education, lower income, and no other adults in the household), and address these barriers through local programme adaptation. Research is needed to better understand how to encourage increased CO-CSA participation levels among participants with lower socioeconomic status, so as not to exacerbate existing health disparities.

CO-CSA operational practices that may support participation

This study identified offering multiple share sizes and a community pick up location as operational practices associated with CO-CSA participation levels. Formative research for F3HK had suggested that potential CO-CSA participants desire control over the variety and quality of produce provided through mechanisms such as multiple share sizes and market-style selection(Reference Hanson, Garner and Connor37), as well as convenient pick up location(Reference Hanson, Garner and Connor37,Reference McGuirt, Jilcott Pitts and Hanson39,Reference McGuirt, Sitaker and Pitts40) . Most F3HK partner farms offered flexibility in share size, pick up location or both, which is consistent with the operational characteristics of CO-CSA programmes in prior studies which frequently offered multiple share sizes(Reference Abbott14,Reference Andreatta, Rhyne and Dery16Reference Hanson, Kolodinsky and Wang18,Reference Hanson, Volpe and Kolodinsky25) and multiple pick up locations(Reference Andreatta, Rhyne and Dery16Reference Hanson, Kolodinsky and Wang18,Reference Wilkins, Farrell and Rangarajan23,Reference Hanson, Volpe and Kolodinsky25) . Farms that offered multiple share sizes generally had participants who both selected the larger share, and had higher levels of CO-CSA pick up, suggesting both positive attitudes toward FV and the logistical capacity for consistent pick up among this subset of participants.

When offered a choice of pick up locations, participants often selected a convenient community site. Prior research also suggests that convenience as perceived by participants also may include a familiar socio-demographic environment in which they feel welcome(Reference McGuirt, Sitaker and Pitts40), which many community locations provided (e.g. housing complex, community church). However, participants at community pick up sites also had lower levels of CO-CSA pick up than at other types of locations. For example, in WA1, participants lived in the housing complex that served as the setting for both CO-CSA pick up and classes. Yet despite the convenience of this community location, 9 (43 %) participants attended no classes and 7 (33 %) participants missed pick ups (most of whom eventually dropped out). Together these findings suggest that participants who prefer community pick up locations also may have additional barriers to participation. NC2 operated solely with a community pick up site and the overall poor participation in this location may have influenced these results.

Some prior studies reported on participants’ satisfaction with cost(Reference Abbott14,Reference Andreatta, Rhyne and Dery16Reference Hanson, Kolodinsky and Wang18,Reference Hanson, Volpe and Kolodinsky25) , pick up location(Reference Abbott14,Reference Andreatta, Rhyne and Dery16Reference Hanson, Kolodinsky and Wang18,Reference Hanson, Volpe and Kolodinsky25,Reference McGuirt, Sitaker and Pitts40) and share volume(Reference Abbott14,Reference Andreatta, Rhyne and Dery16Reference Hanson, Kolodinsky and Wang18,Reference Hanson, Volpe and Kolodinsky25) . However, no prior study that reported on participation level(Reference Abbott14,Reference Berkowitz, O’Neill and Sayer17,Reference Hoffman, Agrawal and Wirth19,Reference Quandt, Dupuis and Fish22) also tested associations between programme operations and participation levels. Our study, therefore, provides novel evidence on associations between the choice of share size and a community pick up location with participation levels. In addition, prior studies suggest that CO-CSA farms that offer some flexibility to participants may be more effective than more rigid programmes. Three CO-CSA effectiveness studies offered little flexibility to participants and all reported no change in FV intake(Reference Abbott14,Reference Izumi, Higgins and Baron20,Reference Quandt, Dupuis and Fish22) . CO-CSA studies that included flexibility to participants by offering a few options(Reference Agboola15,Reference Hoffman, Agrawal and Wirth19,Reference Johnson, Beaudoin and Smith21) , many options(Reference Berkowitz, O’Neill and Sayer17) or options that varied because they partnered with multiple farms(Reference Andreatta, Rhyne and Dery16,Reference Hanson, Kolodinsky and Wang18,Reference Wilkins, Farrell and Rangarajan23,Reference Hanson, Volpe and Kolodinsky25) , tended to report positive outcomes.

Assessing and understanding local residents’ needs and desires related to CO-CSA operational practices (especially choice of share size and community pick up location) prior to programme implementation could support the local tailoring of operations to match these needs and, thus, may support participation. Future research should test associations between CO-CSA operational practices and CO-CSA effectiveness using rigorous research design and larger samples.

Limitations

Several limitations of this study deserve note. First, the sample sizes are small which limits statistical power. Although the F3HK intervention enrolled a total of 148 caregiver–child dyads into the intervention group, the contexts for implementation resulted in small local samples (two of which had fewer than eight participants and were excluded from analyses). In particular, NC2 emerged as unique as having the lowest education and income, and was also the only community classified as entirely rural according to RUCA codes(32). Second, selection of communities was primarily guided by the availability of both an interested partner farm and an available local extension educator. This was a difficult pairing to identify and was most difficult in NC where one community slightly exceeded population criteria and no appropriate and willing extension educator could be identified and therefore staff provided the CSA-tailored nutrition education lessons. The inclusion of only one entirely rural town and this selection process together may hinder the generalisability of these results to other locations. Third, although the F3HK intervention trial required partner farms to accept both weekly payments and to include SNAP as an accepted form of payment, other operational characteristics were decided upon and implemented by partner farms. Although acceptance of SNAP benefits was required, only 24 % of participants used SNAP benefits all or most weeks(Reference Seguin-Fowler, Hanson and Jilcott Pitts28). Farms likely select their practices in consideration of the local community context which limits our ability to disentangle associations among contextual characteristics, CO-CSA operational practices and participation levels.

Conclusion

Small towns and micropolitan communities are highly varied in their population characteristics, and CO-CSA does not appear equally well-suited to all implementation contexts. Understanding community characteristics and familiarity with CSA and adapting models to address potential participation barriers such as limited education and financial resources are important to local CO-CSA programme adaptation. Flexibility in CO-CSA operational practices, and particularly offering multiple share sizes, may support recruitment and participation levels. However, although community pick up locations are desired by participants, these enrollees may face challenges to participation. Future research is needed to better understand barriers to participation in CO-CSA, particularly in rural communities and among participants utilising community pick up locations.

Acknowledgements

Acknowledgements: The authors wish to thank Leah Volpe, Jen Regan and Audrey Baker for their support in preparing this manuscript. Financial support: This work was supported by the National Institute of Food and Agriculture, U.S. Department of Agriculture (USDA), under award number 2015–68001-23,230. USDA had no role in the design, analysis or writing of this article. Authorship: K.L.H., M.S., A.S.A., S.B.J.P., J.K. and R.A.S.-F. conceptualised the study and obtained funding. A.B., S.C. and D.S. implemented the intervention. K.L.H. and G.A.M. led the analyses. L.X. reviewed the literature, extracted data from public sources and performed statistical analyses. K.L.H., L.X. and G.A.M. drafted the manuscript. M.S., A.S.A., S.B.J.P., J.K., A.B., S.C., D.S. and R.A.S.-F. revised the manuscript for important content. All authors have read and approved the final manuscript. Ethics of human subject participation: This study was conducted according to the guidelines laid down in the Declaration of Helsinki and all procedures involving research study participants were approved by the Cornell University (protocol ID: 1501005266) and University of Vermont (protocol ID: CHRBSS 16-393) Institutional Review Boards. Written informed consent was obtained from adults for themselves and their children, and assent was obtained from children.

Conflict of interest:

There are no conflicts of interest.

References

Boeing, H, Bechthold, A, Bub, A et al. (2012) Critical review: vegetables and fruit in the prevention of chronic diseases. Eur J Nutr 51, 637663.CrossRefGoogle ScholarPubMed
Srinath Reddy, K & Katan, MB (2004) Diet, nutrition and the prevention of hypertension and cardiovascular diseases. Public Health Nutr 7, 167186.CrossRefGoogle ScholarPubMed
Lee-Kwan, S, Moore, L, Blanck, H et al. (2017) Disparities in state-specific adult fruit and vegetable consumption – United States, 2015. Morb Mortal Wkly Rep 66, 12411247.CrossRefGoogle ScholarPubMed
Wharton, CM, Hughner, RS, MacMillan, L et al. (2015) Community supported agriculture programs: a novel venue for theory-based health behavior change interventions. Ecol Food Nutr 54, 280301.CrossRefGoogle ScholarPubMed
Allen, P (1999) Reweaving the food security safety net: mediating entitlement and entrepreneurship. Agric Hum Values 16, 117129.CrossRefGoogle Scholar
Guthman, J (2008) “If they only knew”: color blindness and universalism in California alternative food institutions. Prof Geogr 60, 387397.CrossRefGoogle Scholar
DeLind, LB & Ferguson, AE (1999) Is this a women’s movement? The relationship of gender to community-supported agriculture in Michigan. Hum Organ 58, 190200.CrossRefGoogle Scholar
Ostrom, MR (1997) Toward a community supported agriculture: a case study of resistance and change in the modern food system. University of Wisconsin.Google Scholar
Hinrichs, C & Kremer, KS (2002) Social inclusion in a Midwest local food system project. J Poverty 6, 6590.CrossRefGoogle Scholar
Galt, RE, Bradley, K, Christensen, L et al. (2016) What difference does income make for community supported agriculture (CSA) members in California? Comparing lower-income and higher-income households. Agric Hum Values 34, 435452.CrossRefGoogle Scholar
Miewald, C, Holben, D & Hall, P (2012) Role of a food box program in fruit and vegetable consumption and food security. Can J Diet Pract Res 73, 5965.CrossRefGoogle ScholarPubMed
White, MJ, Jilcott Pitts, SB, McGuirt, JT et al. (2018) The perceived influence of cost-offset community-supported agriculture on food access among low-income families. Public Health Nutr 21, 28662874.CrossRefGoogle ScholarPubMed
Guthman, J, Morris, AW & Allen, P (2006) Squaring farm security and food security in two types of alternative food institutions. Rural Sociol 71, 662684.CrossRefGoogle Scholar
Abbott, C (2014) Evaluation of the food bank of Delaware community supported agriculture program. Thesis, University of Delaware.Google Scholar
Agboola, F (2016) Implications of community supported agriculture as alternative food networks. PhD Thesis, Loma Linda University.Google Scholar
Andreatta, S, Rhyne, M & Dery, N (2008) Lessons learned from advocating CSAs for low-income and food insecure households. J Rural Soc Sci 23, 116148.Google Scholar
Berkowitz, SA, O’Neill, J, Sayer, E et al. (2019) Health center-based community-supported agriculture: an RCT. Am J Prev Med 57, S55S64.CrossRefGoogle ScholarPubMed
Hanson, KL, Kolodinsky, J, Wang, W et al. (2017) Adults and children in low-income households that participate in cost-offset community supported agriculture have high fruit and vegetable consumption. Nutrients 9, 726.CrossRefGoogle ScholarPubMed
Hoffman, JA, Agrawal, T, Wirth, C et al. (2012) Farm to family: increasing access to affordable fruits and vegetables among urban head start families. J Hunger Environ Nutr 7, 165177.CrossRefGoogle Scholar
Izumi, BT, Higgins, CE, Baron, A et al. (2018) Feasibility of using a community-supported agriculture program to increase access to and intake of vegetables among federally qualified health center patients. J Nutr Educ Behav 50, 289.e1296.e1.CrossRefGoogle ScholarPubMed
Johnson, DB, Beaudoin, S, Smith, LT et al. (2003) Increasing fruit and vegetable intake in homebound elders: the Seattle senior farmers’ market nutrition pilot program. Prev Chronic Dis 1, 19.Google ScholarPubMed
Quandt, SA, Dupuis, J, Fish, C et al. (2013) Feasibility of using a community-supported agriculture program to improve fruit and vegetable inventories and consumption in an underresourced urban community. Prev Chronic Dis 10, E136.CrossRefGoogle Scholar
Wilkins, JL, Farrell, TJ & Rangarajan, A (2015) Linking vegetable preferences, health and local food systems through community-supported agriculture. Public Health Nutr 18, 23922401.Google ScholarPubMed
Kato, Y & McKinney, L (2015) Bringing food desert residents to an alternative food market: a semi-experimental study of impediments to food access. Agric Hum Values 32, 215227.CrossRefGoogle Scholar
Hanson, KL, Volpe, LC, Kolodinsky, J et al. (2019) Knowledge, attitudes, beliefs and behaviors regarding fruits and vegetables among cost-offset community-supported agriculture (CSA) applicants, purchasers, and a comparison sample. Nutrients 11, 1320.CrossRefGoogle Scholar
Garner, JA, Jilcott Pitts, SB, Hanson, KL et al. (2021) Making community-supported agriculture accessible to low-income families: findings from the farm fresh foods for healthy kids process evaluation. Transl Behav Med 11, 754763.CrossRefGoogle ScholarPubMed
Seguin, RA, Morgan, EH, Hanson, KL et al. (2017) Farm fresh foods for healthy kids (F3HK): an innovative community supported agriculture intervention to prevent childhood obesity in low-income families and strengthen local agricultural economies. BMC Public Health 17, 306.CrossRefGoogle ScholarPubMed
Seguin-Fowler, RA, Hanson, KL, Jilcott Pitts, SB et al. (2021) Community supported agriculture plus nutrition education improves skills, self-efficacy, and eating behaviors among low-income caregivers but not their children: a randomized controlled trial. Int J Behav Nutr Phys Act 18, 112.CrossRefGoogle Scholar
U.S. Census Bureau (2015) Total Population 2011–2015 American Community Survey (ACS) 5-Year Estimates. https://data.census.gov/cedsci/ (accessed June 2020).Google Scholar
U.S. Census Bureau (2019) Educational Attainment, 2015–2019 American Community Survey (ACS) 5-Year Estimates: U.S. Census Bureau. https://data.census.gov/cedsci/ (accessed June 2020).Google Scholar
U.S. Census Bureau (2018) Population Density, Household Income, Families in Poverty, People in Poverty, Race, Ethnicity, 2014–2018. American Community Survey (ACS) 5-Year Estimates. https://data.census.gov/cedsci/ (accessed June 2020).Google Scholar
Economic Research Service (2020) 2010 Rural-Urban Commuting Area Codes. https://www.ers.usda.gov/data-products/rural-urban-commuting-area-codes/ (accessed June 2021).Google Scholar
Feeding America (2017) Overall Food Insecurity Rates, 2017 Map the Meal Gap. http://map.feedingamerica.org/ (accessed June 2020).Google Scholar
Centers for Disease Control and Prevention (2017) Behavioral Risk Factor Surveillance System. https://www.cdc.gov/brfss/annual_data/annual_2017.html (accessed June 2020).Google Scholar
Centers for Disease Control and Prevention & National Cancer Institute (2017) Cancer Incidence, 2013–2017. https://gis.cdc.gov/Cancer/USCS/DataViz.html (accessed June 2020).Google Scholar
Cotter, E, Teixeira, C, Bontrager, A et al. (2017) Low-income adults’ perceptions of farmers’ markets and community-supported agriculture programmes. Public Health Nutr 20, 14521460.CrossRefGoogle ScholarPubMed
Hanson, KL, Garner, J, Connor, LM et al. (2019) Fruit and vegetable preferences and practices may hinder participation in community-supported agriculture among low-income rural families. J Nutr Educ Behav 51, 5767.CrossRefGoogle ScholarPubMed
Zepheda, L & Deal, D (2009) Organic and local food consumer behaviour: alphabet theory. Int J Consum Stud 33, 697705.CrossRefGoogle Scholar
McGuirt, JT, Jilcott Pitts, SB, Hanson, KL et al. (2018) A modified choice experiment to examine willingness to participate in a community supported agriculture (CSA) program among low-income parents. Renew Agric Food Syst 35, 140157.CrossRefGoogle Scholar
McGuirt, J, Sitaker, M, Pitts, S et al. (2019) A mixed-methods examination of the geospatial and sociodemographic context of a direct-to-consumer food system innovation. J Agric Food Syst Community Dev 9, 159177.Google Scholar
Figure 0

Fig. 1 F3HK implementation states, communities, and participants included in analyses

Figure 1

Table 1 Characteristics of communities where F3HK was implemented

Figure 2

Table 2 F3HK participant characteristics by community and state

Figure 3

Table 3 Cost-offset community-supported agriculture (CO-CSA) operational practices by community

Figure 4

Table 4 Mean participation level by community and state

Figure 5

Table 5 Mean participation level by community size, participant characteristics, and cost-offset community-supported agriculture (CO-CSA) operational practices

You have Access Open access

Save article to Kindle

To save this article to your Kindle, first ensure coreplatform@cambridge.org is added to your Approved Personal Document E-mail List under your Personal Document Settings on the Manage Your Content and Devices page of your Amazon account. Then enter the ‘name’ part of your Kindle email address below. Find out more about saving to your Kindle.

Note you can select to save to either the @free.kindle.com or @kindle.com variations. ‘@free.kindle.com’ emails are free but can only be saved to your device when it is connected to wi-fi. ‘@kindle.com’ emails can be delivered even when you are not connected to wi-fi, but note that service fees apply.

Find out more about the Kindle Personal Document Service.

Participation in cost-offset community-supported agriculture by low-income households in the USA is associated with community characteristics and operational practices
Available formats
×

Save article to Dropbox

To save this article to your Dropbox account, please select one or more formats and confirm that you agree to abide by our usage policies. If this is the first time you used this feature, you will be asked to authorise Cambridge Core to connect with your Dropbox account. Find out more about saving content to Dropbox.

Participation in cost-offset community-supported agriculture by low-income households in the USA is associated with community characteristics and operational practices
Available formats
×

Save article to Google Drive

To save this article to your Google Drive account, please select one or more formats and confirm that you agree to abide by our usage policies. If this is the first time you used this feature, you will be asked to authorise Cambridge Core to connect with your Google Drive account. Find out more about saving content to Google Drive.

Participation in cost-offset community-supported agriculture by low-income households in the USA is associated with community characteristics and operational practices
Available formats
×
×

Reply to: Submit a response

Please enter your response.

Your details

Please enter a valid email address.

Conflicting interests

Do you have any conflicting interests? *