RESEARCH BRIEF

Agenda Setting within a Community-Based Food Security Planning Process: The Influence of Power
C H R I S T I N E M C C U L L U M , P H D , R D ; ' DAVID PELLETIHR, P H D ; DoNALD B A R R , PHD;-' JENNIFER WILKINS, P H D , R D 'Center for Health Promotion and Prevention Research, University of Texas-Houston, Homton,Texas; -Division of Nutritional Sciences, Cornell University, Ithaca, NewYork; 'Department of Policy Analysis and Management, Cornell University, Ithaca, New York groups are substantial, additional theoretical frameworks and public policy-making models are needed. KEYWORDS: critical perspective, community food security, agenda setting, power (J Nutr Educ Behav. 2003;35:189-199.) Design: A critical perspective, case study design, and multiple qualitative methods were used. Pafticipants/Setting: Forty-four participants were purposefully recruited to participate in a community-based planning process called a search conference (SC). Seven additional disenfranchised stakeholders who did not attend the SC were also recruited to participate. Phenomenon of Interest: To assess how power influenced agenda setting and to determine the extent to which disenfranchised stakeholders' most salient interests were incorporated into the fmal SC action agendas. Analysis: The constant comparison method, content analysis, and consensus were used to produce the final analysis. Results: Power influenced agenda setting by managing knowledge, problem framing, trust, and consent. Two of seven of disenfranchised stakeholders' most salient interests, including need for adequate food preparation skills and increased availability of locally produced foods, were incorporated into the final SC action agendas. Conclusions and Implications: Citizen politics can be used to build a community' food security agenda on issues that are not at odds with stakeholders in positions of power. To bring about change on issues in which power differences between
Fiiiidiiig \v.is provided throiigl) grants from the Centers for Di-scasc Control and Prevention and the US Department of Agriculture (USDA) Cooperative Kese.irch. Education, and Exteiisioi] Service (CREES). Additional support was provided through N.itional Institutes of Health Grant #2R25CA57712-l)(i, Behavioral Science Education Cancer Prevention and Control, National Cancer lnstittito. Address for correspondence: Christine McCulIuni. PhD. RD, Center for Mealth Promotion and Prevention Research, University of Texas-Houston, Health Science Center, 7000 Fannin St.. Suite 2568. Houston. TX 77030;Tel: (713) .=>00-976.?: Fax: (713) 500-9750; I2-niail: chriscine.incullum@uth.tnlc.edu. 2003 SOCIETY FOR NUTRITION EDUCATION

ABSTRACT Objective: To determine the effectiveness of using citizen politics as a framework for empowering citizens to build a community food security agenda.

INTRODUCTION Community food security is a relatively new term that has no universally agreed upon definition. It can be viewed as a numerical aggregation of household food security, which is concerned with the ability to acquire food at the household level. Alternatively, it can be viewed as relating to broader concerns underlying social, economic, and institutional factors within a community that affect the quantity and quality of food and its affordability. Communities are unlikely to be entirely "food secure" or "food insecure." Rather, they can be placed on a continuum, with the end goal being a foodsecure comiTiiniity, which has been defined as "all persons in a community having access to a culturally acceptable, nutritionally adequate diet through non-emergency or conventional food sources at all times."' Public dialogue is an important tool that can be used as part of a larger process toframeand prioritize key community food security issues and to generate social action around identified issues and concerns.^'^ As noted by Thomson and coUeagues,"* opportunities to engage citizens in discussions on community food issues can empower*' local people in several ways: 1) citizens become aware of others who share common interests about food system issues; 2) people learn specifics about their community food system (such as the extent of hunger and farmland loss) and how they can put their knowledge to work; 3) citizens become part of a solution that addresses community concerns; 4) people can be prepared to contribute in meaningful ways to public decisions on how to deal with community concerns (such as food safety, sprawl, keeping local farms in
189

190

McCuIlum etal/AGENDA SETTING WITHIN A COMMUNITY-BASED FOOD SECURITY PLANNING PROCESS business, or supermarkets closing in a neighborhood); and 5) citizen dialogue forces other people to consider the ideas and opinions of others as they create a vision for the future of their community food system.'* food system goals and commit to a series of action plans.'^ However, the action plans that emerged from a series of community food security planning processes reflected the values and interests of certain stakeholders more than others, even in the absence of overt conflict."*"'" How power influenced agenda setting within these conmiunity-based food security planning processes has not been investigated. Therefore, this study used a critical perspective, case study design, and quahtative methods (1) to assess how power influenced agenda setting within a community-based food security planning process called a search conference (SC) and (2) to determine the extent to which disenfrancliised stakeholders' most salient interests related to community food security were incorporated into the fmal action agendas that emerged from the SC.The main tenet of critical theory is that the construction of meaning and knowledge is shaped by social rules and larger social structures within which their action is guided and constrained. A critical perspective recognizes the importance of human understanding through multiple, socially constructed meanings and realities of everyday life; however, it also seeks to uncover silences that may reside in communication systems.'-'""''

The research-based conceptual framework of "citizen politics" has been proposed as a way to empower citizens to take positive action on public problems such as poverty, hunger, and food insecurity in their communities.^'' The purpose of citizen politics is to develop and strengthen societal capacity to do public work, influence public policy decisions, and enable disenfranchised stakeholders to participate as active citizens in the policy-making process. Citizen politics is based on 3 core concepts: public, diverse interests, and power.''' Public refers to a "public world" in which citizens use their collective power to act on diverse self-interests to shape their surrounding environments and solve common problems.'' The public work of problem solving and governance is carried out tbrough mediating institutions, for example, schools, churches, and the workplace, which connect people to larger public arenas and issues. Diverse interests refer to the different Hfe experiences, perspectives, knowledge, and talents that people bring to public life. Power, within this framework, is conceptualized as dynamic and interactive and is seen as being created through building relationships among diverse interests and taking action on shared problems and goals."'' Although citizen politics has been proposed as a way to empower citizens to take action on public problems such as hunger and food insecurity, to our knowledge, little research has assessed how power is used to influence agenda setting within this type of an approach. Using Q methodology, PeUetier et al identified 3 diverse groups (social justice advocates, pragniatists, visionaries) who placed varying levels of importance on 4 different aspects of community food security: social justice, healthfulness of the food supply, economic viability of local agriculture, and environmental sustainability.* Social justice advocates were primarily concerned about social justice and the social welfare of low-income communities (eg, their ability to access highquality, healthful food), especially in hght of welfare reform. Pragmatists were concerned about the dechne of local agriculture but were not concerned about the environmental consequences of the current food system. They were also unsympathetic toward social justice and social welfare issues because they viewed such issues as a matter of personal responsibility.Visionaries were primarily concerned about the decline of local agriculture and enviromnental consequences of the current food system and were mildly concerned about the social welfare of low-income communities.These authors concluded that the diverse values and interests associated with the concept of "community food security" were saUent to community members in 6 upstate New York counties but that not aD persons embraced all community food security components simultaneously." Participants representing these 3 different groups (social justice advocates, pragmatists, and visionaries) were able to fuid common ground on various

STUDY PROCEDURES Project Design Tliis research was embedded within a larger action research project entitled "The North Country Food and Economic Food Security Project." Action research is a process that incorporates professional knowledge, local knowledge, process skills, research skills, and democratic values as the basis for cocreated knowledge and social change." A SC was used for the initiation of community-baSed food security action planning processes in 6 upstate NewYork counties.''A SC is a strategic planning process that places citizens side by side with formal decision makers to defuie problems and goals, develop alternatives, and initiate action plans based on a common vision that emerges. Its theoretical premises of democratic decision making, open participation, and respect for difference and diversity'** are closely aligned with the theoretical premises of citizen politics, which include public, diverse interests, and power. ^''' The planning and implementation for the SC were carried out through a series of sequential steps: (1) communitybased organizations committed to conducting the SC; (2) logistical planning occurred; (3) a community-referencing system based on informal relationships within the community was used to determine who was eligible to participate in the SC; (4) a "search question" was determined; (5) statistics and data on the pi-oblenis or concerns of interest were collected by a community advisory committee (which were compiled into a profile and distributed to participants at the SC); (6) 30 to 50 persons participated in a 2- to a 2.5-day SC; and (7) post-SC meetings were convened to further refine

Journal of Nutrition Education and Behavior Volume 35 Number 4

July August 2003

191

and implement action plans, goals, and objectives identified at the SC. A more detailed description has been provided elsewhere.'-* The design elements for the SC where the current research was conducted are detailed in Table 1. Sample Forty-four participants from one county in upstate NewYork were purposefully recruited to participate in a 2.5-day SC on conmiunity food security.The Peer Reference System (PRS), a specific type of snowball samphng technique, was used to determine which individuals in the community would be eligible to participate in the SC (R. Rich et al, unpublished data, 1996). In tliis study, community was defined by a specific geographic location (ie, county). Members of a countywide comniimity advisory committee were responsible for conducting the PRS. Advisory committee members (n = 11) were chosen by representatives from 2 community-based agencies and included 2 dairy farmers, a small crop farmer, a local business owner, 2 consumers (including 1 who was disenfranchised), an agricultural educator, a director of a social services agency, a food stamp administrator, a director of a community-based human services organization, and a health inspector from the county health department. The first step in conducting the PRS was to identify key informants who met 2 eligibility criteria: demonstrated interest in community activity and involvement in 1 sector of the food system. Using these criteria, advisory committee members identified 30 key informants representing 8 different sectors of the food system (farmland protection/natural resources conservation, agricultural production, food processing, food distribution/transportation, wholesale marketing, retail marketing, consumer, and "other"). Each key informant was contacted and asked to identify up to 3 potential SC participants who met the 2 eligibihty criteria. After 3 iterations of the PRS, 204 individuals who met the 2 eligibihty criteria and who also agreed to participate in the SC were identified. From this list of 204 names, members of the advisory conunittee selected the final 50 SC participants. Of these, 44 agreed to attend the SC. Table 2 illustrates selected demographic characteristics and the food system sector representation for the final SC participants (n = 44). Disenfi-anchised stakeholders who attended the SC (n = 7) were recruited to participate in this research through the PRS, where they \vere represented in the "consumer" category Seven additional disenfranchised stakeholders who did not attend the SC were also recruited to participate in this research through use of a similar snowball sampling technique. In this study, disenfranchised was defined as "the denying of a privilege, freedom, or power to a person or group resulting in a limited ability to participate in everyday Hfe." Disenfranchised stakeholders are often marginalized or discriminated against at the local level,'''-" leading to personal hardship and/or the need to turn to the federal government for assistance. To qualify as disenfranchised, each participant had to meet 1 of 3 criteria: participation in 1 or more food

and nutrition assistance programs, receipt of? unemployment benefits, and receipt of disability benefits. iThirteen of 14 (93%) disenfranchised stakeholders in this study were women, had at least a high school degree, and reported participating in 1 or more food and nutrition assistance programs.Ten of 14 (71%) participants reported receiving unemployment benefits, and 6 of 1 4 (43%) reported receiving disability benefits. Data Collection Participant observations were conducted at all advisory committee meetings held prior to the SC to assess how power influenced agenda setting within die larger community food security SC process. In this study, power was defined as the capacity to produce intended, foreseen, or unforeseen effects on others based on the ability to control access to valued resources.-' According to Bachrach and Baratz, agenda setting is shaped by "a set of predominant values, beliefs, rituals, and institutional procedures that operate systematically and consistently to benefit some persons and groups at the expense of others."^- All documents produced prior to and during the SC were also reviewed to assess how power influenced agenda setting and the extent to which disenfranchised stakeholders' interests were incorporated into the final SC action agendas. Document review is considered useful in portraying the values, beliefs, and viewpoints on a specific issue expressed within a particular setting and can provide valuable empirical insights when used in combination with other methods.'-* An open-ended questionnaire was administered at 2 different points during the SC to all attendees (n = 44) to further assess how power influenced agenda setting. Questions developed for use in the open-ended questionnaire were based on theoretical considerations of group behavior and power dynamics-'-^''"^'^ and were pilot tested with community members. A subsample of participants (n = 25), including all disenfranchised (n = 7) and powerful stakeholders (n =18) who attended the SC, were interviewed after the SC to gain further insight into themes that emerged from earlier data collection methods and to exaixdne the extent to which their most sahent interests were incorporated into the final SC action agendas. In this study, a participant's categorization as powerful was determined by his/her organizational decisionmaking authority and access to valued resources.^-* and not by individual-level characteristics. All interviews were tape recorded and transcribed verbatim. Focus groups were conducted widi disenfranchised stakeholders (n = 14) prior to and after the SC to explore their most salient community food security h-iterests within a safe environment, that is, an environment in which authoritative interests and values are not present, and to prevent the presence of ideological false consciousness, that is, a process by which people develop a systematic misunderstanding of themselves through participation in a world in which public ideas and images distort reality in the interests of various dominant groups.'' Conducting focus groups at 2 points in

192

McCullum et al/AGENDA SETTING WITHIN A COMMUNITY-BASED FOOD SECURITY PLANNING PROCESS

Table 1. Design Elements for One County's Search Conference on CofnmLinity Food Security Session Search Conference Design Eiements Day1 Local sponsors welcome, describe baci<ground on search conference Search managers describe philosophy of searching and the design principies of a search conference Selection of participants is described and participants introduce themselves Search conference schedule is reviewed Questions and answers Objective is to piece together the relevant history of the community's food security system. What brought things to the current situation (both internal and external factors)? Longest tenured participant is asked for appropriate starting point in time/history Participants pair up (3 persons maximum) to identify most significant shaping forces, decisions, and events Group draws the history on pieces of paper that are on the wall Participants walk through the cocreated shared history as a group Day 2 Participants meet in small groups divided by peer groups (where each peer group is represented by 1 of 8 different food system sectors) to bring out their special interests Each participant states his/her major expectation or desire for search conference outcome Peer groups report to other groups in large-group format Search conference managers review agenda and comment on match of searching with expectations identified Mixed small groups" with representation from all peer groups (organized beforehand) Groups asked to answer search question, "What should our local community food system look like by the year

Welcome and overview

Shared history

Expectations

Desirable future

2005?"
Report 5-6 major points done by each small group's facilitator Search managers integrate information and drive process to agreement Probable future In large-group format, answer question, "If no substantial changes occur, what will our community food system look like by the year 2005?" Top 5-6 items identified Casual integration (reviewing similarities and differences) Participants complete midpoint search conference evaluation Search managers report resuits to participants during lunch Small groups meet as remixed from peer groups Answer question: "Given what we would like in 2005, what should we keep doing, what should we stop doing, and what new things do we need to be doing?" Group all results of brainstorming Full group organizes by affinity process (piece all materials together) Break Search managers lead process of labeling identified clusters Participants self-select into cluster most motivated to work on Group discusses those clusters not attracting sufficient interest and either incorporate or let go Groups meet in respective action planning groups Groups organize subelements of change Group develops overall change statement; "Our group is trying to..." and subeiement statements Day 3 Small groups use force field anaiysis Groups analyze their results and identify major restraining forces Small groups report in large group; identify where common constraints exist and decide how to approach the common areas Individually, participants identify necessary activities to bring about selected change Group organizes actions into cluster into sequential order Group produces narrative sections of template, goals and objectives, and rationale Break Groups identify immediate action steps (who will convene next meeting, when it will take place, who will be the contact person, who else needs to be invited) Discussion of follow-up and send-off Search conference evaluation

Mid-search conference check-in Keep, drop, create

Areas for change; selection

Areas for change; definition

Action planning; force field analysist

Action planning; work flow diagramming Action planning; template Personal commitments Close

"Each of the small groups had a self-selected group facilitator and group recorder. Both of these group responsibilities were rotated throughout the search conference, and each group member (search conference participant) was required to serve as a facilitator and group recorder during the 2.5-day search conference. fForce field analysis is defined here as an examination of the forces working in support of the search conference participants' vision of "what should our local community food system look like by the year 2005?," the forces working against the identified community food system vision (restraining forces), leverage points (or opportunities for change), and recommendations for action.

Journal of Nutrition Education and Behavior Volume 35 Number 4

Table 2. Selected Demographics and Food System Representation of Search Conference Participants (N = 44) N Demographics* Gender Male Female Age 19-25 26-34 35-43 44-52 53-61 62+ Education < High school High school Associate degree/some coliege Bachelor degree Professionai degree Advanced degree Food System Representation Food system sectors Farmland protection/natural resources conservation Agricultural production Food processing Food distribution/transportation Whoiesaie marketing Retaii marketing Consumer* Other* %

July August 2003

193

informed consent and confidentiality were applied throughout the study. A university committee on human subjects approved all aspects of this research. Data Analysis and Quality

19 25 0 3 8 17 10 6 1 8 20 7 2 6

43 57 0 7 18 38 23 14 2 18 45 16 5 14

4 7 3 4 1 5 10 10

9 16 7 9 2 11 23 23

Data analysis occurred throughout the entire data collection process to identify bhnd spots in the data and to allow the researcher to pursue emergent themes in subsequent data collection.-' Data collected prior to and during the SC (including notes of observation, documents, pre-SC interview and focus group transcripts, responses from open-ended questionnaires, and field notes) were analyzed using constant comparative method and thematic content analysis."^ A second round of data analysis using verbatim post-SC interview and focus group transcripts was conducted to test working hypotheses that emerged from the first round of data collection.''' Identified themes common to all sources were included as emergent categories and were entered into a matrix to organize the data.The matrix included how many times each emergent category was mentioned and the breadth of each category.-**' Inconsistencies in the categories were explored and revisions were made, and negative cases, that is, cases in which data did not fit into any of the emergent themes, were sought." Focus group transcripts were coded to determine disenfranchised stakeholders' most salient interests reported within a safe environment at 2 different points in time (before and after the SC). A salience score for disenfranchised stakeholders' interests was calculated by multiplying the frequency of each response (number of times it was mentioned) by the intensity (low, moderate, high) with which it was expressed. For those disenfranchised stakeholders who attended the SC (n = 7),documents produced within the first small-group session at the SC (expectations) and responses from open-ended questionnaires were coded to determine their reported interests at 3 different points in time during the SC.-'' Two researchers independently coded themes that emerged from all data sources. Initial differences in agreement over themes that emerged were discussed until there was 100% agreement in assigned codes.^' Data quality was ensured by several criteria and corresponding methods: credibility, dependability, and confirmability Credibility was established through the use of prolonged engagement, member checks, peer debriefers, reflexivity, method triangulation, and negative case analysis.'''-* Dependability and confirmability audits were conducted to ensure that shifts and changes in the data were both tracked and trackable (dependability) and through an inspection of the raw products of the research inquiry (confirmability).'FINDINGS Cotnmunity Food Security Action Agendas Six final action agendas and corresponding goals emerged from the SC on community food security, including the need

*Race is not reported here as a demographic variable because all search conference participants (N = 44; 100%) were white. 'Seven disenfranchised stakeholders who attended the search conference were included in the consumer category. *This category included a nutrition teaching assistant with the Expanded Food and Nutrition Education Program (EFNEP), a Community Action Agency director, a food stamp administrator, an associate director for a food and nutrition assistance program, a culinary arts instructor, a professor, a legislator, a food service supervisor, and 2 educators.

time allowed for critical reflection, a deeper probing of issues and concerns, and discussions on ideological false consciousness.'*'' Specific focus group questions were developed in consultation with theories on power-''"' and community outreach specialists who worked with disenfranchised stakeholders. Questions were pretested with members of the target audience for readability and understandability. Focus groups were moderated by a trained community informant as a way to reduce power differences that existed between the researcher and participants, enable participants to be as comfortable as possible when discussing their interests and concerns, and allow the researcher to take field notes on the group interactions and dynamics.-'' Focus group sessions were tape recorded and transcribed verbatim. Procedures for

194

McCuUum et al/AGENDA SETTING WITHIN A COMMUNITY-BASED FOOD SECURITY PLANNING PROCESS

for (1) distribution of surplus food, to strengthen a distribution network for surplus and leftover food; (2) education, to raise awareness of consumers as to the food and fiber system and related life skills through education; (3) family and community values, to encourage and nurture a commitment to personal responsibility and wider relationships through building community: (4) food processing and marketing, to give the North Country producers outlets for their products; (5) legislative initiatives and actions, to develop a plan to bring change to issues brought up at the SC; and (6) new agriculture, to explore agricultural economic opportunities in the North Counti'y. How Power Influenced Agenda Setting Power influenced agenda setting through 4 separate mechanisms, including managing (ie, controlling) knowledge, problem framing, trust, and consent (Table 3). The results have been recontextuaJized in the data and are described below. Managing knowledge. Knowledge was managed by use of "decisionless" decisions on certain issues and by avoiding discussion on sensitive issues. A decisionless decision was used during the pre-SC advisory committee meetings when the concern that some elderly persons in the county may be eating pet food owing to a lack of income was btought to the attention of the advisory committee members by a health inspector from the county health department: "I've had repeated reports from public health nurses that the elderly are eating dog food. But when I called the state to report what I had heard, they told me, 'back off, there's too much money involved.'" Advisory committee members acknowledged the need to make a decision as to whether the SC participants should be made aware of this problem; however, no decision was made, and this knowledge was not shared with, participants. During the SC, knowledge was managed by avoiding open discussion on sensitive issues relevant to the topic of community food security as was reflected in a statement by a powerful stakeholder who attended the SC: "I avoided dis-

cussion on the topic [agricultural pesticide use] because I knew it would be perceived as controversial." Managing problem framing. Problem framing was managed by the pursuit of narrow interests, choice ofthe terminology used, and decisions being made before all identified costs, benefits, and risks were considered. Prior to the SC, problem framing was managed by the pursuit of narrow interests and by using the terminology of "community food systems" instead of "community food security." For example, although it was originally agreed that the SC would focus on community food security, it was later reframed to focus on community food systems by a member of the advisory committee whose primary interest related to agriculture as opposed to hixnger or food insecurity. Problem framing was further managed by the pursuit of narrow interests during the SC. A city planner observed, "Certain individuals were participating because they had their own agendas. They might have needed to justify how their participation was going to benefit their particular organization, and that gets frustrating because we were talking about communities and benefiting everyone." Problem framing was managed during the SC by the choice of terminology used when some SC participants referred to government welfare programs as "easy welfare." Not all SC participants agreed with the use of this terminology, including an associate director o\ a food and nutrition assistance program who commented, "Easy welfareI don't like this terminology; it's used by people with niiddle-class values." This participant, however, reported not feeling comfortable sharing this point of view openly with other SC participants. Finally, problem fiaming was managed by making decisions before all identified costs, bene:fits, and risks were considered.With regard to the issue of welfare reform, a disenfranchised stakeholder who attended the SC commented, "They kept talking about moving us [welfare recipients] off of pubHc assistance and taking personal responsibility. But what I want to know is where I can get a decent paying job so that I can support me and my two kids? Right now the only place I can get a job is at [a fast food restaurant], and

Table 3.

Mechanisms of Power That Influenced Agenda Setting during the Community's Food Security Planning Process Before the Search Conference "Decisionless" decisions Pursuit of narrow interests Choice of terminology used During the Search Conference Avoid discussion on sensitive issues that are reievant to the topic Pursuit of narrow interests Choice of terminoiogy used Aii identified costs, benefits, and risi<s are not considered before decisions are made Perceived credibiiity of the information source Ruies of "anticipated reactions" Argue that politicai issues are really technical issues best left to experts

Mechanisms of Power iVIanaging i^nowledge Managing problem framing

Managing trust Managing consent

Trusted individuals are seiected to serve on advisory committee Screened process is used for final selection of participants Argue that poiitical issues are really technical issues best left to experts

Journal of Nutrition Education and Behavior Volume 35 Number 4

July August 2003

195

Tabie 4. Rank Order of Disenfranchised Stakeholders' Most Salient interests (n = 14) as Reported in Focus Groups Conducted prior to and after the Search Conference on Community Food Security Pre-Search Conference Focus Groups 1. iHigh food prices in iow-income neighborhoods 2. Food safety 3. Lack of access to high-quality foods 4. Lack of adequate food preparation skills 5. Store fraud/false advertising 6. Negative impact of food advertising on diet and health 7. Lack of availability of localiy produced foods Post-Search Conference Focus Groups 1. Food safety 2. iHigh food prices in iow-income neighborhoods 3. Lack of access to high-quality foods 3. Lack of adequate food preparation skills 3. Store fraud/false advertising 4. Negative impact of food advertising on diet and health 5. Lack of avaiiabiiity of locally produced foods

Saiience Score* 54 50 50 40 38 36 22

Salience Score 66 54 50 50 50 43 34

*The salience score was caicuiated by the frequency with which the issue was mentioned multipiied by the intensity.

there's no way I can survive on those wages. They didn't address this bigger issue." Managing trust. Trust was managed prior to the SC through a selection of "trusted individuals" who were chosen to serve on the 1 l-member advisory committee and by the perceived credibility of an information source.This latter mechanism, that is, perceived credibility of an information source, was illustrated during discussions on agricultural pesticide use at the SC. An agricultural educator noted, "A few persons were talking about pesticides, that they were unsafe, but their source [the media] couldn't be trusted." In contrast, a member of the county's soil conservation board stated that "Some participants felt media coverage on this issue [the negative effects of agricultural pesticide use] wasn't trustworthy, but I feel it's critical," although this latter participant reported not feeling comfortable sharing her viewpoint openly during the SC. Managing consent. Consent was managed through using a screened process to select final SC participants, by arguing that political issues were technical issues best left to experts, and by using the rules of "anticipated reactions."With regard

to use of a screened .selection process to decide on the final list of SC participants, one advisory committee member noted that those individuals "who we get along with well" were more likely to be selected. Consent was also managed by arguing that political issues were technical issues best left to experts. For example, a powerful stakeholder who attended the SC commented, "This [recombinant bovine somatotrophin (rBST) use] is an issue that the government and scientists might be asked to talk about or make a decision on, but not your everyday person.... I don't think it was the time or the place to discuss (rBST use]. It's an issue best left to experts." Rules of anticipated reactions were used when a fear of repercussions of what others might say or do at a later point in tinie prevented some participants fi-om openly commenting on issues during the SC. A farmer who attended the SC revealed,"I didn't bring it up [rBST use] because I knew some persons wouldn't agree with what I had to say, and that I might regret it later." Comparison of Disenfranchised Stakeholders' Interests to Final SC Action Agendas
Table 4 illustrates the most sahent community food security interests reported by disenfranchised stakeholders (n = 1 4 ) during a safe environment (ie, focus groups) before and after

Table 5. Rank Order of Disenfranchised Stakeholders' (n = 7) Most Salient Interests on Community Food Security Reported at Three Different Points in Time during the Search Conference Beginning of Search Conference (Expectations)* (n = 7) Food safety concerns (n = 2) Store fraud/false advertising (n = 2) High food prices in low-income neighborhoods (n = 1) Lack of adequate food preparation skills (n = 1) Lack of availability of locally produced foods (n = 1) IVIiddie of Search Conference (After Probable Future) (n = 7) Education (n = 2) Family and community values (n = 2) High food prices in iow-income neighborhoods (n = 1) Hunger owing to lack of income (n = 1) increased avaiiabiiity of localiy produced foods (n = 1) End of Search Conference (Action Planning) (n = 7) Family and community vaiues (n = 3) Education (n = 2) Hunger (n = 1) increased avaiiabiiity of locally produced foods (n = 1)

Expectations refers to the first smaii-group session that was held at the search conference according to peer groups, where each participant was asked his/her main reason for participation in the search conference on community food security.

196

McCullum et al/AGENDA SETTING WITHIN A COMMUNITY-BASED FOOD SECURITY PLANNING PROCESS

the SC. Table 5 highlights the reported interests of disenfranchised stakeholders wlio attended the SC (n = 7) at 3 different points in time during the SC.Two of 7 ofthe most salient interests identified by disenfranchised stakeholders, including the need for adequate food preparation skills and increased availability of locally produced foods, were incorporated into the final action agendas that emerged from the SC (eg, education and food processing and marketing). Five of 7 (71%) of the disenfranchised stakeholders who attended the SC reported that their most salient interests were either downplayed or ignored and that their most saHent interests were not incorporated into any ofthe final action agendas that emerged from the SC. One disenfiranchised stakeholder reflected, "I brought it up at the beginning. You know, that many of the stores in low-income neighborhoods have higher food prices, but they really didn't address it. I think it was an important issue that should have been addressed, but it wasn't." Another disenfranchised stakeholder commented that her most salient concern related to food safety was not incorporated into any ofthe final SC action ageiidas:"My biggest concern was about food safety. Whether or not the fish being caught locally are safe to eat This issue was not addressed by any ofthe groups that came about as a result of the search conference."

Negative Case
There was one negative case to report in this study. The director of a community-based human services organization openly challenged a participant duritig the SC who stated that "'welfare recipients are lazy" by proclaiming, "I told him, 'yo" a^^^ just not in a position to judge.' I mean I feel poverty does that to people. You become chronically depressed as a result of undergoing stress On a daily basis. I don't know what poor mother or father doesn't go through that. It's the problem that needs to be dealt with, not the person."This individual was an exception in that she openly expressed her differing viewpoint to other individuals at the SC and perceived that her viewpoint was respected and taken seriously by other SC participants.

DISCUSSION
The purpose of this study was to assess how power influenced agenda setting within a specific type of community-based food security planning process and to determine the extent to which disenfranchised stakeholders' most salient concerns were incorporated into the final action agendas that emerged from the community food security planning process. Power influenced agenda setting through 4 mechanisms: (1) managing knowledge, (2) managing problem framing, (3) managing trust, and (4) managing consent.The results fi'om this study are similar to previous fmdings reported by Forester, who documented how these same mechanisms of power influenced agenda setting within land use planning.-'' The current fmdings also provide support for mechanisms of power described previously by Bachrach and Baratz-- and Gaventa.--^ These include (a) rules of anticipated reactions or situations in which

B, confronted by A, who has greater power resources than B, decides not to make a demand on A for fear that the latter will invoke sanctions against him, and (b) decisionless decisions, which can be observed through institutional inaction on an identified issue perceived to be politically sensitive. In this research, participants' decision to not address certain issues (eg, agricultural pesticide use and rBST use^''t) may be explained, in part, by theories on agenda building in which it is argued that for "big science"^'* issues, agenda setting and policy advocacy do not use a social problem-solving model.-^'* Rather, they take place within closely linked alliances that are composed of a small group of scientific researchers, federallevel officials and their representative government agencies, interested companies/firms, and chairpersons of congressional committees who have jurisdiction over these matjgjg 3> > Others have noted that mobilizing support for a pub., -. v lic issue may also be difBcult when viewpoints are too diverse or at variance with those in leadership positions.-*' In .such cases, a stronger analysis of power relations, institutions, and politics is needed.-^'* The results from this research demonstrated that disenfranchised stakeholders had only 2 of 7 of their most salient interests incorporated into the fmal action agendas that emerged from the SC. One explanation for this finding is that disenfiranchised stakeholders who attended the SC may have been coerced into adopting other SC participants' interests as their own, given that their most salient interests as reported at the beginning of the SC were strikingly different compared with their interests as reported by the end ofthe SC (see Table 5). These results also support previous findings that revealed the concerns of a group of socially disadvantaged women were often overshadowed hy the concerns ofthe majority and ignored by those in power.^'-' In this previous research, the process of community organizing enabled these women to initiate price changes in, supermarkets in low-income neighborhoods. However, they remained powerless to overcome nutritional inequities that were constructed through the media and political system.''" These women also reported constraints to engaging in community organizing: lack of child care, underdeveloped writing skills, resource shortages, fear of repercussions (ie, fear of having welfare benefits withdrawn prevented any action that challenged the government), and lack of sustained reahzation that their actions lead to change."To increase the likelihood that more of disenfiranchised stakeholders' most sahent interests are incorporated into future SC action agendas, increased use of peer groups could be adopted to reduce actual or perceived status differences among participants."" Individuals working in community-based organizations can also (1) engage more disenfranchised stakeholders during the earhest stages ofthe SC planning (eg, advisory committee formation); (2) oversample disenfranchised stakeholders in relation to more powerful stakeholders; (3) include more time for critical reflection during the SC; (4). use an anonymous voting procedure for reporting priority issues; and (5) work with disenfranchised stakeholders to identify

Journal ofNutrition Education and Behavior Volume 35 Number 4

July August 2003

197

both facilitating factors and barriers to their participation in a SC."*' Tarasuk has reported that, to date, the main emphasis of community food security initiatives in Canada has focused on enhancing food skills and alternative means of food acquisition as opposed to challenging the structural origins of individual and household food insecurity (eg, povert)').''-' Siniilarly, Chavis has noted that community coalitions can he used to maintain the status quo and contain the empowerment of grassroots leadership and marginalized groups.*'' Power differentials and other resource limitations iced by disenfranchised stakeholders may limit their ability to bring about needed changes on their own.'"' As stated by Travers, "there is danger in placing too much responsibility on the shoulders of those with the fewest resources and the least political power to initiate change. For the shift in power relations crucial to social change, advocacy is essential."^' In conclusion, citizen politics can be used to build a community food security agenda on issues that do not conflict with the interests of more powerful stakeholders. However, for issues that generate conflict with stakeholders in positions of power, additional theoretical frameworks (eg, theories on power,^--'*-^'' agenda building,'''"'-^ and agenda setting'''''^') and policy-making models (eg, advocacy and institutional politics) are needed. There were several limitations to this study. First, the sample was very homogenous. AU participants were white and resided in one county in upstate New York. Therefore, these findings cannot be generalized to other groups.The fmal selection of SC participants was based on a convenience sample and not a representative sample, further limiting the generalizabOity of the results. Because focus groups were not conducted with powerful stakeholders (n = 18), a similar comparison between disenfranchised stakeholders' (n = 14) and powerful stakeholders' (n = 18) interests could not be made. Furthermore, because focus groups were not conducted with all food system stakeholder groups, it was not possible to assess the extent to which other groups did not have their most salient interests incorporated into the final SC action agendas. Finally, all but one of the disenfranchised stakeholders recruited to participate in this study were women. Because men are more likely than women to influence macrolevel food security and food policy decisions,''^ gender differencesas opposed to disenfranchisement status as defined in this studymay have explained why the majority of disenfranchised stakeholders' most salient interests were not addressed within the final action agendas that emerged from the SC.

broadly use community organization**** and sociopolitical theories-including theories on power,--'^*-^ agenda buildjj.^g j(..45 gj^jj agenda setting,'''""'to assess how power influences agenda setting within other types of community-based food security planning and decision-making processes. Second, researchers working with disenfranchised stakeholders and other marginalized groups can adopt a critical perspective'-'"'^ and a participatory action research approach-^""''"''''^"^' to assess if community-based approaches designed to build community' food security are truly empowering these groups or if they serve to maintain power differentials and other resource inequities.''''''' Nutrition educators working to build a community food security agenda in their own comnuniities would benefit from training in process-oriented skills, including facilitation, negotiation, conflict resolution, and constituency building, tO transform conflict into greater capacit)', equity, andjustice.-'-*'' When community food security issues and concerns are connected to broader structural issues (eg, poor conditions for low-waged workers) and institutional policies, educators would benefit from training in additional policy-making models, including advocacy and institutional politics and their corresponding strategies (eg, convene meetings with decision makers, hold press conferences, and use mass media portrayals'* to influence public opinion and to help allocate resources needed to bring about desired policy changes).To strengthen efforts that involve an advocacy or social action component, nutrition educators Can expand their use of partnerships from traditional institutional boundaries (eg, government institutions, professional association) to include nongovernmental organizations.''^

ACKNOWLEDGMENTS Funding was provided through grants from the Centers for Disease Control and Prevention and the US Department of Agriculture (USDA) Cooperative Research, Education, and Extension Service (CREES). Additional support was provided through National Institutes of Health Grant #2R25CA5771206, Behavioral Science Education Cancer Prevention and Control, National Cancer Institute.The authors would like to acknowledge the contributions of the 51 participants who participated in this research. An earlier version of this article was presented at the 32nd Annual Meeting of the Society for Nutrition Education in Baltimore, Maryland.

IMPLICATIONS FOR RESEARCH AND PRACTICE These findings have important implications for research and practice. First, future research needs to be conducted to determine how community coahtions can be used to increase the power of grassroots and other citizen-led organizations.'*''To achieve this goal, nutrition education researchers can more

ENDNOTES '*The term "empower" is uSed here to refer to the process of empowerment as proposed by Rappaport: "the process by which people, organizations, and communities gain mastery (i.e., control) over their lives."' 'lyBST, or reconibinant bovine somatotropin, is a hormone manufactured using leconibinant deoxyribonucleic acid

198

McCullum et al/AGENDA SETTING WITHIN A GOMMUNITY-BASED FOOD SECURITY PLANNING PROCESS

18. Eriiery F, Purser R. The Search Cotifercnce: A Powerful Method for Planning Organizational Change and Commtmity Action. San Francisco, Calif: Jossey-Bass Publisliers; 1996. 19. Chambers R. Whose Reality Counts? Putting the First Last. London: Interniediate Technology Publications; 1997. 20. Allen P. Reweaving the food security safety net: mediating entitlement and entrepreneurship. /ifdV HMMIJH I^/HM. 1999; 16:117-129. 21. Barr D, Transforming Pou'ey:A 13-Week Program for Democratic Change in Vbi/f Community, [thaca, NY: Cornell Empowerment Group; 2000. 22. Bachrach P, Baratz M. Power and Poverty. NewYork, NY: Oxford University Press; 1970, 23. Marshall C, Rossman GB. Designing Qualitative Research. 3rd ed.Thousand Oaks, Calif: Sage Publications; 1999. 24. Lukes S. Pon/(?r.vl Radical View. London: Macmillan; 1974.

(DNA) technology that is injected into dairy cows to increase milk production by 10% to 15%. The US Food and Drug Administration (FDA) approved rBST for large-scale commercial use in November 1993. However, its use has elicited controversy as a result of concerns related tp its effects on cows.hiiman health, and the economic viability of small dairy farms.''' *The term "big science" refers to the scale of Science policy initiatives and extends beyond size, cost, or extent of resources used. For a more detailed explanation of this concept, refer to Galison and

REFERENCES
1. Cohen B. Community Food ScairilyAssessnWirHHilkit.'Wiihmgton.DC: Economic Research Service, US Dept. of Agriculture (USDA), EFAN-02-013;July 2002. 2. Bridger J. Community imagery and the built environment. Soc Q. 1996;37:353-374. 3. Smith J, Maretzki AN. Citizen Ditilogiui:A Giiidehook. Universitj' Park, Pa: The Pennsylvania Scace University; 2000. 4. Thomson JS, Abel JK, Maretzki AN; Edible contiections; a model for citizen dialogue used to discuss local food, farm, and commimity issues. JAppI Commwi. 2001;85(1):25-42. 5. Rappaport J. In praise of paradox: a social policy of empowerment over prevention. .i4mj Community Psydiol. 1981:9:1-25. 6. Wagner P. Citizen politics: a conceptual framework for shaping public po^cy. J Niilr Educ. l996;28:80-82.

25. Gaventa J. Power and Powerlessness: Quiescence and Rebellion in an Appalachian Hj/fcy. Urbana, III: University of Illinois Press; 1980. 26. Krueger R, King ]. Involuing Communily Members in Focus Groups. Thousand Oaks, Calif: Sage Publications; 1998. 27. Miles MB, Huberman AM. Qualitative Data Analysis: An Expanded Sotirceboob. 2nd ed.Thousand Oaks, Calif: Sage Publications; 1994. 28. SttAuss hL. Corbin}, Basics of Qfialiintive Reseiirch:Teckniijues and Proceduresfor Developing, GrcundedTheory. 2nd ed.Thousand Oaks, Calif; Sage Publications; 1998. 29. McCullum C. Mechanisms and Effects of Power within a Community-Based Food Security Planning Process [dissertation]. Ithaca, NY; Cornell University; 2000. 30. Falk LW, Bisogni CA, Sobal J. Food choice processes of older adults; a qualitative: investigation. J Nutr Educ. 1996;28;257-265. 31. LincplnYS, Cuba EG. Nammlistic Inquiry. Beverly Hills, Calif: Sage Publicatibris; 1985.

7., Hogg KS, ed. Reini'eriling Citizenship: The Practice of Citizen Politics. 32. SchvvandtTA, Halpern ES. LinkingAuditing and Metaevaluation: EnhancMinneapolis, Minn: Project Public Life; 1994. ing Quality in .Applied Research. Newbury Park, Calif: Sage Publications; 8. Pelletier D, KraakV, McCullum C, Uusitalo U. Values, public policy, 1988. and community food security^ Agric Human Values. 2000;i7:7S-93. 33. Fore.iter J. PLmning in the Face of Power. Berkeley, Calif: University of y. Pelletier D, Kraak V, McCullum C, Uusitalo U, Rich R. Community California Press; 19S9. food security: salience and participation at coilimunity level. Agric Human Values:. 1999;16:401-419. U). Pelletier D, Kraak V, McCullum C, Uusitalo U, Wch R.The shaping of collective values through deliberative democracy: an empirical study from NewYork's North County. /^o/iVy Sdeiues. 1999;32:103-131. 11. McCullum C, Pelletier D, Barr D, Wilkins J. Use of a participatory planning process as a way to build community food security.J/lm Diet Assoc. 2002; 102:962-967. 12. Pelletier D, McCullum C, Kraak V, Asher K. Participation, power, and beliefs shape local food and nutrition policy./Nfc. 2003;133:301S-304S. 13. Travers K. Nutritibn education for social change: critical perspective. JKutrEduc. 1997;29:57-62. 14. Fay B. How people change themselves: the relationship between critical theory and its audietrce. In: BallT, e<i,Pqlitieai Theory and Praxis. Minneapolis, Minn: University of Minnesota Press; 1977:200-233. 34. US Congress, Office ofTechnoIogyAssessment. A Newlechnological Era for American Agriculture. Washington, DC: US Government Printing Office; 1992. OTA-F-475. 35. Galison P, HeVly B, eds. Big Science:Tlie Growth of Large-Scale Research. Stanford, Calif; Stanford University Press; 1992. 36. Madison JJ. Agenda-building and big science. Policy Sciettces. 2000;33:3]-53. 37. Bosso C. Pesticides and Politics:The Life Cycle of a Public Issue. Pittsburgh, Pa: University of Pittsburgh Press; 1987. 38. Moser C, Norton A. To Claim Our Rights: Livelihood Security, Human Rights and Sustainable Developtitent. London: Overseas Development Institute; 2001. 39. Trayers K. Reducing itiequitics through participatdry research and community empowerment. Health Educ Behav. 1997;24:344-356. 40. Travers K.The social organization of nutritional inequities. Soc Sci Med. 1996;43:543-553.

15. Lather P. Research as praxis. Hrtiwfrf frfHmfionfl/few'etf. 1986;56:257-277. 16. Farley J. Justifying conclusions in naturalistic evaluation: a critical per41. Levin M, Martin AW, eds. Power Difference in Discourse. Ithaca, NY: Corspective. Evaluation ttnd Program Planning. 1987;10:343-350. nell University' and The Norwegian Institute of Technology; 1994. 17. Greenwood DL, Levin M. Introduction lo Action Research: Social Research 42. Baker EA, Teaser-Polk C. Measuring community capacity: where do for Social CAflMi'fi. Thousand Oaks, Calif: Sage Publications; 1998. we go from here? Health Educ Behaf. 1998;25:279-283.

Journal of Nutrition Education and Behavior Volume 35 NunibtM' 4

43. TarasukV. A critical examination of communitj'-based responses to Household food iiisectn-icy in Canada. Health Educ Behav. 20Ol;28: 487-499. 44. Chavis DM.The paradoxes and pi-omise of community coalitions, ^m J Community Psyc/io/. 20()l;29:309-3iy. 45. Rochefort DA, Cobbs RW. The Politics of Problem Definition: Shaping the Policy Agenda. Lawrence, Kan: University of Kansas Press; 1994. 46. Kindgon JW. Agendas, .Alimuitii'cs, and Public Policies. New York, NY: Harper Collins; 1995. 47. Van Esterik P. Right ro food; right to feed; right to be fed.The intersection of women's rights and the right to food. Agric Human Values. 1999:16:225-232.

July August 2003

199

48. Minkler M, ed. Community Organising and Community Building for Health. New Brimswick. NJ: Rutgers University Press; 1997. 49. Sarri R, S;irri C. Participatory action research in two communities in Bolivia and the United States. Itmnmionat .Social PVork 19'J2;35:267-280. .TO. Wang C, Burris MA. Photovoice: Concept, methodology, and use for participatory needs nsscssmein. Health Educ Behav. l997;24:369-387. 51. Park P, Brydon-MiUer M, Hall B, Jackson T, eds. Voices of Change: Participatory Research in the United States and Canada. Westport, Conn: Bergin & Garvey; 1993. 52. Greene J. Qualitative program evaluation: practice and promise. In: Denzin NK, Lincohi YS, eds. Collecting and Infei'prctin^ Qualitative Aliiicr/uAt. Thousand Oaks, Calif: Sage Publications; l9'>K:372-399.

VISION, MISSION, AND GUIDING PRINCIPLES OF THE SOCIETY FOR NUTRITION EDUCATION
Vision Healthy people in healthy communities. Mission To enhance nutrition educators' ability to promote healthful sustainable food choices and nutrition behaviors. Guiding Principles Fiscal responsibility Respect for diversity of opinions and perspectives Trust and wilhngness to communicate openly and respectfully Knowledge-based decisions Excellence and lifelong learning Pi"ofessioruilisni and integrity Inclusiveness in membership Equality among members Rewarding and enjoyable experiences for volunteers and supporters