Setting the table for policy intrapreneurship: public administrator perspectives on local food system governance

Abstract In communities across the U.S., there is growing awareness of food system issues which exerts greater pressure on public servants to help build a better food system for their community. Drawing from interviews with local and state-level public administrators and elected officials in two metropolitan regions in Ohio and New Jersey, and supporting interviews from food producers and nonprofit leaders in those areas, this paper examines what roles public administrators believe they could and should take regarding food system development in their community. From this qualitative data analysis, the research identified commonalities in administrators’ positive and negative perceptions of the potential for development of their community’s local food system (LFS) despite their lack of background knowledge regarding these systems. Given food systems’ interdisciplinarity and complexity, LFS development likely requires multi-sectoral alliances via partnership governance. The alliance itself and each of its components is simultaneously a complete entity and a part of a larger, more complex entity; such entities are called holons. These alliances have greater capacity to manage more complicated problems than can be addressed by subordinate holons. In examining the potential role of local public administrators in LFS development, this paper constructs the concept of the policy intrapreneur to clarify our understanding and discussions of how public administrators, decision-makers, and other stakeholders view their roles and responsibilities in the creation and governance of local food systems.


Introduction
Recently in the U.S., major metropolitan areas' local governments are recruiting staff to manage food system development for their urban communities. Examples of position titles include food system administrator, food system policy director, urban agriculture director, and similar terminology. These positions borrow, likely inadvertently, from the call to action that Pothukuchi and Kaufman (1999) posed to the urban planning discipline to include food systems planning in that field. In part, they speculated what role municipal governments could have in planning for local food systems (LFS). Part of this speculation suggested creating a department of food to help develop local food systems. While creating an entire department might be impractical for most municipal governments, with growing popularity of local food systems, there are several roles and responsibilities that public administrators could undertake in developing these local food networks.
From content analysis of interviews and an extended discussion of the concepts of partnership governance, policy entrepreneur, and public entrepreneur, this article offers a new term, policy intrapreneur, to specifically describe government actors leading local food system innovation. This article adapts the concept of a public entrepreneur, advanced by Grossman and Holzer (2015) as part of their larger theory of partnership governance, in discussions around LFS development and governance.
After a brief review of the food systems literature from a U.S. public administration (PA) lens, this article reexamines a content analysis to explore what role local-level administrators believe they could and should play regarding food system development in their communities. The authors draw on data from interviews of local public officials, as well as food producers and nonprofit leaders, in two metropolitan regions in Ohio and New Jersey. These data were originally collected during one author's dissertation research (Jones 2018). Additionally, this research relies on the authors' informal interviews, observations, and experiences from work with public administrators in New Jersey and Ohio, as well as interactions with producers, community members, and nonprofit organizations working on food issues in those communities. From a PA perspective, this study assesses local administrators' perceptions of food system governance and offers a new term to describe governmental actors who champion local food system development.

Literature review
Globally, and especially in the U.S., communities are calling for the development and strengthening of local food systems. A food system can be defined as a network of all the stages, stakeholders, structures, and systems that play a role in feeding a population (B en e et al. 2019;Hospes and Brons 2016). The definition of a local food system (sometimes called a community food system) is often contested by differences in determining conceptual and geographical boundaries around the systems themselves (Hospes and Brons 2016;Martinez et al. 2010). In this work, the features distinguishing an LFS from the global industrialized food system include geographic proximity and/or a shortened food supply chain, a focus on food sovereignty, and an overall objective of sustainability (Bloom and Hinrichs 2011;Hospes and Brons 2016).
With a growing focus on LFS development, there are a number of roles and responsibilities that local public administrators could undertake in the development and management of these local food networks. To understand what role local public administrators could or should play in food system creation and administration in U.S. cities, this section begins with a brief overview of food system governance. This is followed by a consideration of PA perspectives regarding food systems, including a discussion on traditional governmental structures and partnership governance.

Food system governance
Food systems are those people and processes involved in getting food from farm to fork and back. We understand the term governance to mean the dynamic processes of managing and configuring relationships between stakeholders to guide public affairs and decision-making (Hospes and Brons 2016). Thus, we define local food system governance as the ways in which various stakeholders and sectors in a community come together to deliberate and decide issues related to food (Siddiki et al. 2015). An LFS should reflect and evolve to meet a specific community's needs, resources, and constraints and there are various governance structures available to manage these systems.
Food policy councils (FPCs, sometimes called food partnerships, food advocacy coalitions, food policy networks, etc.) are nonpartisan groups formed by a broad range of stakeholders and sector representatives, which serve as a, often voluntary, communitybased outlet for coordinated action and advocacy on local food issues (Halvey et al. 2020). Generally, these FPCs promote collaboration between sectors and stakeholders, assess and advocate for policy, develop and support programs and services to meet local needs, and offer an arena for discussion of food issues (Harper et al. 2009;Scherb et al. 2012;Siddiki et al. 2015). Given FPCs' focus on public policies, these groups are inherently connected to government officials and organizations.
These community-driven food councils involve or interact with local public officials and institutions to varying degrees in their pursuit of food system policies and programs. The stronger links to local government may include formal endorsements, member appointments, enabling legislation, resource support, accountability structures, etc. (Bassarab et al. 2019;Martinez et al. 2010). FPCs frequently face problems of limited participation and conflict over overlapping missions between overtaxed partner organizations, as well as a lack of centralized and formal authority (Cuy Castellanos et al. 2016). FPCs may also have less formal relationships with their associated local governments, and differences in worldview could lead to adversarial relationships. With the most informal of connections, some FPCs are composed of civil society organizations and have no official linkages to local government (Halvey et al. 2020). As FPCs are a relatively new phenomenon, research in this area is nascent but expanding.
A growing body of literature addresses food councils as a form of food governance (Bassarab et al. 2019;Harper et al. 2009;Moragues-Faus and Morgan 2015;Scherb et al. 2012;Siddiki et al. 2015), how local governments are linked to these food system governance structures (Bassarab et al. 2019;Forster and Getz Escudero 2014;Hospes and Brons 2016), and multilevel food governance and the need for integration (Forster and Getz Escudero, 2014;Hospes and Brons 2016;Ruhf et al. 2017). However, most public administration scholars do not directly employ lines of questioning focused on what role U.S. local government administrators could and should play in local food system governance. This type of inquiry guides the research presented here.

Public administration perspectives on local food system development in urban communities
Within the industrialized global food system, it is easy to see why U.S. local governments have had limited focus on food issues until relatively recently. This is especially true in urban settings, as local governments and the public generally have long treated food issues as overtly within the rural agricultural sphere and, therefore, inherently not urban (Moore 2006;Moragues-Faus and Morgan 2015;Pothukuchi and Kaufman 1999). However, "the food system is indeed a significant urban system" (Pothukuchi and Kaufman 1999, p. 213) and "cities are emerging as key transition spaces where new food governance systems are being fashioned" (Moragues-Faus and Morgan 2015, p. 1558). As urban communities are paying greater attention to food issues than previously, especially around urban agriculture and food production, there is a need for PA scholarship to inform and assess these efforts.
While the focus on food issues by local governments has grown since the early 2000s, the field of PA in the U.S. particularly has only recently given attention to food issues in specifically urban settings (Halvey et al. 2020;Moragues-Faus and Morgan 2015;Siddiki et al. 2015). In developing their local food systems, U.S. communities worked with whichever mechanisms were feasible in their specific circumstances. However, the insertion of food issues into the traditional structures, activities, and authorities of urban municipalities can create conflict between stakeholders. The source of this conflict is clear: LFS development is a complex problem that does not fit neatly into traditional municipal government silos, and likely requires socio-political and administrative innovation to overcome structural barriers and resource limitations inherent in this type of wicked problem (Bassarab et al. 2019;Peters and Pierre 2014). Such innovation requires multi-sectoral action, in which government can and should participate.
Due to ongoing conflicts of adequately integrating food into traditional activities of urban local governments (van de Griend, Duncan, and Wiskerke 2019), the inclusion of public administrators in community-wide coordination seems necessary. However, attempts at forging community-wide partnerships have been met with limited and inconsistent success due to limited financial and human resources, little political will and administrative capital from allied organizations, indistinct or conflicting goals, and/or ineffective management (Cuy Castellanos et al. 2016;Scherb et al. 2012;van de Griend, Duncan, and Wiskerke 2019;Winne 2008). But these issues may be best addressed from the frame of multisectoral partnering, also called partnership governance.

Partnership governance and public entrepreneur
Partnership governance is a "process of multisectoral partnering" which creates "expanded public management capacity" with the potential of addressing more complicated problems than can be addressed by any of the partnership's constituent parts (Grossman and Holzer 2015, p. 2). Partnership governance rejects the principal-agent relationships of New Public Management as suggested by Osborne and Gaebler (1992), instead advancing a formalized partnership of organizations that collectively has greater transformational power than would any individual principal-agent relationship. Trust is a central requirement in partnership governance (Grossman and Holzer 2015), and any partnership must actively work to reduce mistrust.
Integral to the concept of partnership governance is the idea of holons (Grossman and Holzer 2015). A holon in this context is simultaneously a complete concept, or a "wholepart"; but also a subpart of a larger, more complex holon. Often visually depicted as growing concentric circles, holons should be understood as a hierarchy of increasing complexity where a higher-order holon transcends its lower-order holons. Partnership governance seeks to build an alliance focused on a larger contextual holon that is more equipped to solve problems that individuals organizations focused on lower-order holons could not (Grossman and Holzer 2015). Using an LFS lens, an example of greater holon could be expanding from the departmental level to encompass the whole government within a given locality, as well as moving from that single locality to encompass multiple governance structures within a given region. This notion of a holon can be defined along administrative or geographic dimensions. Thus, a food policy organization focused on a metropolitan region could be equipped to promote food system development across its component parts and not exclusively in an individual jurisdiction or department.
Borrowing from Schumpeter's conceptualization of entrepreneurship, Grossman and Holzer (2015) note that social entrepreneurship, either public or private, as "a new way of getting things done", is necessary for partnership governance (Schumpeter 1947). This conceptualization of entrepreneurship is deliberately broader than the colloquial understanding of entrepreneurship as the creation of small, private businesses. Despite its inclusive, formal definition, entrepreneurship is generally used to describe private sector actors. Thus, to better distinguish usage of the term social entrepreneur in a governmental context, others have offered the notion of a public entrepreneur.
No uniform definition exists for public entrepreneurs (Grossman and Holzer 2015; Roberts and King 1991), but their activities and traits are studied by researchers in various fields (see Mintrom 2019; Roberts and King 1991). Among potential characteristics of public entrepreneurship, authors have suggested that public entrepreneurs champion public policy goals but, due to their position as an employee of a government or civil society organization, are unable to internalize the same benefits (i.e. profit) of a traditional private sector entrepreneur (Moore 2006).
Distinguishing between the sectoral position of these entrepreneurial actors, other research has referred to the concept of a policy entrepreneur to mean those "public entrepreneurs who, from outside the formal positions of government, introduce, translate, and help implement new ideas into public practice" (Roberts and King 1991, p.152). These policy entrepreneurs are generally seen as, "energetic actors who engage in collaborative efforts in and around government to promote policy innovations" and deliberately seek to shift the status quo (Mintrom 2019, p. 14). Beyond sectoral distinctions, Roberts and King (1991) also categorize policy entrepreneurs based on their leadership level. The common feature of these policy entrepreneurs is their focus on promoting public policy innovation. To better understand the potential for policy entrepreneurship in food system development and public administrators' perspectives on their role in these governance efforts, the authors reanalyzed interview data from two metropolitan regions in New Jersey and Ohio.

Research design
This work explores public administrators': 1. perspectives on how food fits into the local policy agenda; 2. current role in creating and managing their local food system; and 3. understanding of the role that local administrators could and should play in developing and governing local food systems.
In addressing these objectives, this research draws on semi-structured and unstructured interviews conducted by Jones for his dissertation (Jones 2018). Participants included a combination of local administrators and elected officials across two metropolitan regions (greater Dayton, Ohio and greater Newark, New Jersey). The two regions are both examples of post-industrial communities in the U.S., an important component of Jones's dissertation research. Jones's prior professional experience in both regions informed the selection of the two regions over similar post-industrial communities. This site selection method may limit the generalizability of these findings.
A combination of purposive sampling, based on the author's previous work in those regions, and snowballing, based on suggestions from early respondents, methods identified interview respondents. When possible, Jones sought to interview individuals of similar job title or authority within their respective region. For example, Jones interviewed a senior planning official working for the central city government of each metropolitan region. All interviews adhered to the requirements of the New Jersey Institute of Technology's Institutional Review Board (IRB), including the use of signed consent forms.
In total, 24 public sector respondents participated in interviews; five semi-structured and seven unstructured interviews occurred in the greater Newark region, and nine semi-structured and three unstructured interviews occurred in the greater Dayton region. These public sector respondents included both elected officials and bureaucrats. Examples of respondents' job titles include: Mayor, County Commissioner, Economic Development Director, Chief Zoning Administrator, Public Health Inspector, and Planning Director. These individuals are appropriate respondents for an exploratory study of this nature for two reasons. First, as Pothukuchi and Kaufman (1999) observe, there is no Department of Food, or similar governmental body, in local U.S. governments. Thus, drawing respondents from different segments of local government creates the best opportunity for content saturation (Mason 2010). Second, these individuals collectively represent a significant amount of administrative and political decisionmaking power within their respective jurisdictions.
The author also conducted an additional 40 semi-structured interviews of urban food stakeholders and nonprofit leaders working in local food systems across the two urban regions studied. Examples of respondents' job titles include: owner of an urban food-producing business, Executive Director of an urban food-producing nonprofit, Executive Director of an emergency food assistance nonprofit, Farmers' Market Manager, and Director of Neighborhood and Community Development of a neighborhood improvement nonprofit. These interviews similarly adhered to IRB guidelines and used a combination of convenience and snowball sampling based on Jones's interactions with these individuals over the course of the research. Jones attempted to create regional parity in sampling by interviewing individuals with similar job titles in both regions.
In the dissertation, Jones performed a content analysis of transcripts from interviews relying on Hsieh and Shannon's (2005) conventional content analysis methods. Conventional content analysis is appropriate for such a study, as little established theory or scholarly consensus exists to inform this work. Further, the intention to use conventional content analysis informed the heavy use of open-ended questions in the semi-structured interview instruments.
The interviewer transcribed each semi-structured interview word-for-word, omitting stammers and pauses, and created field notes during unstructured interviews. The conventional content analysis included phrases, sentences, and paragraphs. Coding occurred over three rounds, and the first coding generated 10 primary categories based on the dissertation's seven research questions, with additional categories examining major challenges facing the locations, food system entrepreneurs, and the potential for expansion of food distribution. The second round of coding generated 16 secondary categories based on themes discovered across the primary level of codes. A tertiary round of coding generated an additional 96 categories based on themes present in the secondary level of codes. In some cases, tertiary codes were exclusively linked to one secondary code, while other tertiary codes were linked to several secondary codes.
The research of this project reexamines Jones's (2018) data to address the research objectives listed above. Most data examined by this research was generated from interviews with public administrators. In limited instances, data drawn from nonpublic administrators interviewed (like business and nonprofit leaders described above) provided additional context for this reevaluation of the prior research. For example, interview questions for urban farmers focused on the application of zoning regulations help to contextualize interview data from the respective zoning administrator. Three secondary codes from the dissertation specifically inform the focus of this research: Positive Perceptions of Local Food Systems, Negative Perceptions of Local Food Systems, and Lack of Knowledge of Local Food Systems. Tertiary codes subordinate to these secondary codes provide a deeper context. Examples of relevant tertiary codes include: Living Wage Jobs, Conflict over Urban/Rural Divide, Potential for Food-based Adaptive Reuse, and Shifting Economic Development Strategies. In seeking to address the current research objectives, this work reassesses the previously collected data and identifies the following themes.

Findings and discussion
When discussing how food fits into the local policy agenda, administrators and elected officials in the two communities held a mix of positive and negative perceptions of their community's food system. These public officials offered cautious support for the role that local food system development can play in 21st century cities, but also skepticism about the sustainability of the economic impact of such development and the potential role that local government might play in the local food system's development. These findings are discussed below.

How food fits on the local policy agenda
Many public officials interviewed see a place for local food systems as their communities continue their transition into the 21st century. Implicit within that optimism was a recognition that public policy would need to shift, at least on a modest level, to accommodate this largely novel activity, especially urban agriculture and food production. This local food production, a core component of an LFS, is often conceptualized as not part of the urban sphere and, when it is, it is poorly documented (Moragues-Faus and Morgan 2015). At the same time, many officials are skeptical of the efficacy of spending limited public resources on local food systems in favor of other community and economic development goals. The allocation of resources, both human and financial, was identified as a constraint on local government's entry into new policy areas.
A central point throughout the interviews was the lack of foundational knowledge of food systems and entrepreneurship; specifically, what these concepts are, could be, and cannot be. This lack of knowledge, discussed more below, is understandable to some degree, given the emerging nature of LFSs as a transdisciplinary sub-discipline and the relatively recent focus on food systems in urban communities. These themes were consistent across interviews with officials from both regions. No specific regional differences emerged through the content analysis process.
Viewed collectively, these findings align with the idea that individual municipalities or specific government departments, as lower-level holons, lack the requisite knowledge and, thus are unable to effectively incentivize and regulate a regionally and departmentally interconnected food system. This suggests the need for formation of a greater-level holon to address these issues. Importantly, this research specifically reexamined data drawn from LFSs in urban communities and similar studies of rural areas may generate different data.

Lack of prior experience and current perceptions of local food systems
Most officials acknowledged they had little knowledge of both the food production in their communities and about local food systems in general. When pressed, most officials admitted they had little previous experience, either personal or professional, in food system development or agriculture. Respondents commonly believed this lack of knowledge did not affect their subsequent interactions with urban food stakeholders in their community, either positively or negatively. Given the recent emergence of the transdisciplinary sub-field, this answer is not surprising as the vast majority of public officials likely received no formal exposure to food system concepts in their professional academic training (e.g. urban planning, public administration, law school) (Fisher et al. 1996).
Despite a common lack of knowledge or awareness of their communities' food systems, public officials have mixed perceptions of LFS development, with each reporting some variety of positive and negative aspects. None of the respondents was completely in favor of or completely opposed to food system development in their locality. Although rarely acknowledged explicitly, officials frame many of their comments to align with the narrative of urban/rural divide that dominated community development throughout much of the 20th century (Moore 2006;Moragues-Faus and Morgan 2015). This conceptual divide perceives food production as an inherently rural activity and, consequently, anathema to the urban sphere.
The interdisciplinarity of food systems is poorly served by this lack of food knowledge and expertise of local officials. From the frame of partnership governance, the multi-sectoral partnering required of LFSs necessitates cross-disciplinary and trans-sectoral expertise. These roles are often filled by non-governmental leaders in the community, referred to as social entrepreneurs; or, as Roberts and King (1991) would say, policy entrepreneurs. However, when these tasks are undertaken by government officials or within formal government structures, reframing the idea of the social or policy entrepreneur is prudent for clarity of terms. Further, the presence of many micro and small-scale profit-seeking farmers and manufacturers working in food systems can also add to confusion around the term social entrepreneur. It is critical, then, to specify the sectoral position of these local food leaders pushing for policy innovation.

Positive perceptions
Of their positive perceptions, most administrators and elected officials believe that LFS development, urban agriculture, and a local food industry have a role to play in the future of their communities. When speaking of upcoming revisions to the zoning code, a senior planning official for the City of Newark noted, "we can't just focus on traditional manufacturing. The manufacturing base isn't going to come back here. We have to be proactive about writing our code. What do we want to bring? What will diversify our economic development? That could be food. What gives access to food? What gives access to jobs?". Similarly, one former staff member for then-Mayor of Newark Cory Booker contended that Booker administration leadership believed that hydroponic vegetable production was the future of cities, especially for the reuse of former light industrial spaces.
Other respondents also referenced the potential for LFS development as a response to existing post-industrial challenges in urban centers. The county administrator of Montgomery County, Ohio speculated on the potential of hydroponic farming as an ideal reuse for abandoned big box stores (e.g. Walmart and Target), which he perceived as a growing problem for less dense urban regions (e.g. many Midwestern communities). The assistant city manager for Springfield, Ohio echoed this idea as he detailed his city's desire to renovate a former factory site for use by a yet-to-be-identified hydroponic farming company. He noted, "we would love to see [more] hydroponics. We think there is a lot of potential, because it doesn't matter where you're at in terms of growing, it's about the facility you're in and you design the facility around it". The reuse of spaces is a common discussion in post-industrial cities, and local food leaders can create connections between local food systems and potential reuse of formerlyindustrial facilities.
In addition to post-industrial potential, a number of public officials saw connection points between local food producers and retailers. Broadly, these respondents acknowledged the value of LFS development as a type of small-scale entrepreneurship. But when pressed, few could identify what specific barriers food system entrepreneurship faced that were different from challenges facing small business entrepreneurship generally. In this dimension, the officials' lack of background knowledge on LFSs may lead them to overlook the distinctions between local food stakeholders and other small business entrepreneurs. Further, this cautious support from respondents might be a manifestation of and indistinguishable from general support for small business entrepreneurship that is common amongst local-level officials. Future research should seek to identify any nuances between local officials' support of small business entrepreneurs and support of local food producers and distributors specifically.

Negative perceptions
While local officials shared some positive views of LFS development, some negative perceptions were commonly voiced as well. The central theme of these negative perceptions focused on the novelty and limited impact of jobs directly created by urban food entrepreneurship (Jones 2018) . Most respondents who articulated this point emphasized the importance that any jobs created by food producers or manufacturers, either for the proprietors or their employees, offer a living wage. The Mayor of Dayton expressed this point by saying, "we tend to support economic development strategies that are not [focused on] startups … we are looking for gazelles. Food might create two or three jobs. We have rules that if we invest in a company, it has to be a livingwage job, and these generally aren't living-wage jobs". While many industries and sectors suffer from a lack of living-wage jobs, at least one administrator in the Dayton region acknowledged that low-income food system jobs may offer greater job satisfaction than an alternative job. They noted, "if someone is gonna choose between making $12 an hour as a waiter and $12 an hour making donuts in their house and selling them at a farmers' market, more people might choose the latter". Respondents also expressed concern about the raw number of potential jobs that a strong LFS might be capable of creating, even in ideal conditions. Similarly, officials also perceived changing consumer preferences toward locally-produced foods as a fad that will quickly end and, thus, unworthy of policy shifts and interventions.
The final shared negative perception focused on the day-to-day manifestation of LFS activities, as opposed to more conceptual objections. Some respondents expressed concern about urban food producers engaging in activities that would be in conflict with other nearby land uses. For example, one public health official in a suburban community in greater Newark observed that "[urban farmers] can't just put up a tent to sell products [on vacant lots in residential areas]. What is the point of having a business-zoned district? [We need to keep] residential districts with residential use … else you're gonna start creating chaos, because everyone will start putting up tents on vacant lots". Other respondents were similarly troubled by the esthetics of urban food production, specifically urban agriculture. The zoning administrator in the Dayton region noted that "some people feel that perhaps curb appeal could be lost from unwieldy plant growth". While these are legitimate concerns of local officials, it is also these same officials who have the capacity to address such issues through the power of their office; issuing zoning guidance and regulations, offering public support for urban agriculture management, incentivizing local food procurement, etc.
Understood collectively, these findings suggest a cautious optimism among local officials across both regions. However, running through many of the interviews was a lack of knowledge on LFSs' capacities and limitations. Interviews with local food business leaders and nonprofit officials working in food systems support this cautious optimism of public administrators and elected officials, especially given public officials' lack of knowledge on local food issues. Both farmers and manufacturers, along with public servants themselves, want local government officials to champion food system development efforts and improve how current and prospective entrepreneurs are regulated. However, several food producers expressed frustration at the lack of desire by many public officials to educate themselves about the rapidly changing nature of local food systems in urban communities. This frustration with the lack of knowledge of local officials supports the need for a policy champion to specifically address and lead on LFS issues.

Role that local governance could and should play local food system development
While single municipal departments may struggle to overcome traditional governmental silos, food networks of the larger holon may be better equipped to overcome the silo challenge and make changes across various geographical and departmental boundaries. Linked to Ruhf et al. (2017) discussion of complex food systems projects, the complexity of food systems that transcend the scope of a single municipal department aligns with Grossman and Holzer's idea that larger holons are needed to adequately address the challenges to which lower-level holons are unable to respond effectively (2015).
The notion that local government should have some role in LFS development wove throughout many interviews, despite respondents' minimal knowledge on the topic. While no respondent specifically called for expansion of food system governance, either geographically or administratively, neither did they explicitly speak against it. However, there was a lack of consistency in responses as to how or to what degree a government should be involved, as well as which administrator or administrative unit should lead in overcoming traditional departmental silos to develop their community's food system.
One potential response to these data is to adopt the lens of Grossman and Holzer (2015) social entrepreneur, as this complexity points toward the need for greater holons to address more complex problems like the creation and governance of an LFS. But what happens when that social entrepreneur is located in a formal public administrator role? How can we distinguish those official governmental actors from the traditional use of the term social entrepreneur, as well as the public's common use of the term entrepreneur? 4.5.1. Proposing the term policy intrapreneur At first glance, the concept of public entrepreneurship may be similar to the idea of the social entrepreneur. However, public entrepreneurship is different from social entrepreneurship in that public entrepreneurs are leaders of partnerships involving governments, while social entrepreneurs are less interested in matters of government (Grossman and Holzer 2015). This notion of partnerships has been discussed by other local food governance scholars using different terminology like translocal governance and collaborative food governance (Moragues-Faus and Sonnino 2019; Siddiki et al. 2015) and, at a regional level, multi-stakeholder audience outreach (Ruhf et al. 2017). But, using a partnership governance lens to examine the literature on local food systems suggests that many efforts to improve local food systems use holons of insufficient size to effectively address larger systemic issues. Grossman and Holzer (2015) embed the idea of the social entrepreneur into the need for a larger holon, as the social entrepreneur is one of the mechanisms by which a greater holon is created and sustained to address more complex problems. Applying the principles of partnership governance to the development of an LFS is appropriate, especially in an urban context, as it creates a larger holon capable of addressing more systemic problems than its subordinate holon parts are either unwilling or unable to address individually. To advance and coordinate food system efforts across those subparts of that larger holon may require a public sector entrepreneur. Roberts and King (1991) distinguished between four types of public entrepreneurs: those "political entrepreneurs, who hold elected leadership positions in government; executive entrepreneurs, who hold appointed leadership positions in government; bureaucratic entrepreneurs, who hold formal non-leadership governmental positions; and policy entrepreneurs are those "who work from outside the formal governmental system to introduce, translate, and implement innovative ideas into public sector practice" (Roberts and King 1991, p.152). The first three of these types are included in the term proposed here: policy intrapreneur. Innovation in local food policy requires government actors that push for change and, regardless of the type or level of government position they hold, the term policy intrapreneur describes these governmental actors. This conceptualization of policy intrapreneur as a change agent aligns with Mintrom's (2019) explanation of policy entrepreneurs as those that seek, "to transform policy ideas into policy innovations and, hence, disrupt status quo policy arrangements" (p.1). But the challenges described in this research seek a leader to champion these policy innovations from within government in their formal role; hence, intrapreneur.

Conclusion
Interviews with local government officials in metropolitan regions of two U.S. states suggested that many administrators have little knowledge of both a conceptual understanding of their LFS and specific knowledge about extant characteristics of their individual community's food system. Despite this general lack of knowledge, both food producers and the respondents themselves believed that local government should play a role in championing their community's food system policy and governance. However, given the complexity of local food system development within a metropolitan region, the disaggregated nature of LFS, and emerging awareness that government intervention is necessary to improve the LFS, use of the term social entrepreneur may be confusing when describing these experts in different sectors. For greater specificity and identification of these government actors, we propose the term policy intrapreneur to situate the work of a public entrepreneur in the LFS context working toward food system partnership governance. The concept of a policy intrapreneur offers a way to facilitate the creation of partnership governance in food system development, which is especially critical given the interdisciplinary nature of food systems.
This analysis suggests several avenues for future research at the intersection of LFS development and the term policy intrapreneur. First, this research did not address which agency within a given area could or should host a policy intrapreneur. Potential examples of host agencies include county or municipal governments or a subordinate department, as well as planning associations, regional-scale hospital systems, or universities. Cost-benefit analyses of potential host agencies along standardized dimensions would be informative.
Second, future work should seek to assess how municipalities have positioned a policy intrapreneur in relation to their food council or other LFS governance structure. Thirdly, this research did not address what skills sets are necessary for success in an individual serving as a policy intrapreneur. Food system development likely requires a skilled champion who is knowledgeable about food systems and possesses institutional memory of their community and necessary social capital and community organizing skills to lead diverse populations. Deeper investigation would help a host organization determine what manner of individual they should seek when recruiting someone into a newly created food policy intrapreneur position. Finally, while this research assesses data collected from interviews about local food systems in specifically urban communities, there may be lessons for the administration of LFSs in rural and suburban contexts also that should be examined.