Digging in: Toward a more just urban garden land policy

,


Introduction
In a park on the edge of downtown Minneapolis, college students and neighbors tend tomatoes and marigolds in raised beds built over an old horseshoe court (Figure 1). Across town to the north, sunflowers tower over rows of onions, squash vines, and collard greens on a lot left vacant after a devastating tornado in 2011. In South Minneapolis, nearly two dozen garden plots are squeezed into a narrow strip next to a fire station. While each site is unique, reflecting their caretakers' varied aspirations and labors, all three are among the City's hundred-plus gardens on public land made possible by recent policies.
In response to surging interest in urban agriculture, many cities across the country have developed policies that provide access to publicly owned land. Such policies are often touted as a way to address environmental injustice and racial equity, and indeed they can offer exciting possibilities for doing so in meaningful ways. However, supporters of urban agriculture come with a wide range of political commitments and goals-what geographer Nathan McClintock calls the "radical, reformist, and garden-variety neoliberal" (2014, p. 147) contradictions of urban agriculture-and, if not carefully crafted, land policies can end up exacerbating existing inequalities. With a vibrant and diverse urban agriculture movement (Homegrown Minneapolis, 2019a) and a reputation as a progressive bastion (Thompson, 2015), but with some of the most egregious racial inequalities in the country (Minnesota Compass, 2021c;Nickrand, 2015), Minneapolis is a particularly fitting place to dig into the messy details and muddy debates over justice in garden land policy.

Figure 1. Grape Tomato Vines Reach Skyward in a Raised Bed at a Community Garden in Loring Park
Journal of Agriculture, Food Systems, and Community Development ISSN: 2152-0801 online https://foodsystemsjournal.org We present two case studies of the urban garden policies adopted by the City of Minneapolis and the Minneapolis Park and Recreation Board, a semiautonomous institution (hereafter referred to as the City and the Park Board, respectively). This study has two aims. First, we illustrate the ways in which seemingly minor, common-sense details systematically shaped who benefited from the garden land policies. Second, we highlight the key role of public participation-particularly during the stage where the details were determined-in crafting policies that reduce barriers to racial equity. Beyond the particularities of urban agriculture policies and of Minneapolis, the study offers concrete insights for scholars and practitioners working to develop more just policies and urban futures.

Literature Review
Many scholars propose justice as a normative goal in urban policy and planning, though there remains debate on its meaning in practice (Agyeman, 2013;Davidoff, 2016;Fainstein, 2010;Fischer, 2009;Harvey, 1992;Krumholz, 1982;Marcuse et al., 2009;Steil, 2018). Drawing mainly on the work of planning scholars Agyeman (2013) and Fainstein (2010), and environmental justice scholar Schlosberg (2007), we take recognition, participation, and equity as distinct but interrelated elements that form a conceptual core in justice theory.
Recognition entails the respect for and meaningful consideration of social groups differentiated on multiple grounds of identity, such as class, race, ethnicity, national origin, gender, age, and ability (Young, 1997). Rather than seeking to homogenize or reduce intergroup differences, proponents of recognition envision a pluralistic urban space with policies that afford different treatment to social groups according to their needs or interests (Young, 1990). Adequate recognition provides standing for the participation and consideration of groups' distinct interests (Schlosberg, 2004). Burgeoning academic literature has foregrounded the particular salience of race in the production of spatial inequality-or the uneven distribution of resources, opportunities, and hazards across spacein the U.S., and its importance as an analytic approach in struggles for greater justice (Anderson, 2015;Lee et al., 2021;Lipsitz, 2007;Pulido, 2017;Song, 2015;Williams, 2020).
Participation or procedural justice involves opportunities for laypeople to substantively influence policymaking (Fainstein, 2010). The essential questions for participation are "who is involved, how, and on whose terms" (S. White, 1996, p. 14). There is considerable variation in the stage of policy development where participation occurs (e.g., agenda setting, design, implementation), the required time commitment (e.g., a single event versus sustained engagement), barriers for laypeople to participate (e.g., meeting times, location, compensation, caretaking responsibilities), and the degree of influence laypeople have (e.g., token participation versus decision-making power). These factors are determined both by the state's willingness to share power and the ability of laypeople to mobilize to exert pressure (Cornwall, 2008). While participation is often conceptualized as a normative spectrum from manipulation to community self-determination (Arnstein, 2019;Pretty, 1995), ambiguity, contestation, and power asymmetries are inherent in any participatory process (Cornwall, 2008;S. White, 1996).
Equity is concerned with the fairness of policy outcomes-the distribution of goods, opportunities, burdens, and risks (Agyeman, 2013;Schlosberg, 2004). Under strict equality, benefits and burdens are divided equally among individuals. In contrast, equity recognizes that in practice individuals and groups begin from unequal positions, and that policies based on strict equality can uphold and even exacerbate existing injustice. Instead of distributing benefits equally without regard to present status, an equitable policy prioritizes the needs and desires of the people with the fewest resources and the least power (Fainstein, 2010;Krumholz, 1982).
Clearly, recognition, participation, and equity are enmeshed. Schlosberg (2007) explains: "it is not just that political and cultural institutions create conditions that hamper equity and recognition, but that both distributive inequity and misrecognition hamper real participation in political and cultural institutions" (p. 28). For Fainstein (2010) and Fraser (1995aFraser ( , 1997, distributional justice ought to Journal of Agriculture, Food Systems, and Community Development ISSN: 2152-0801 online https://foodsystemsjournal.org be given the most weight, but Schlosberg (2007) argues that certain elements will assume greater salience in particular contexts or moments in history.
Urban gardening is often touted as a concrete manifestation of urban and food justice (Heynen et al., 2012), where equity, participation, and recognition can be integrated. Beyond its contribution to equity concerns, such as household food security (Meenar & Hoover, 2012), many proponents view urban gardens as a way to model alternative land governance arrangements that allow for and require greater participation, such as holding land in common (Aptekar, 2015;Eizenberg, 2012;Morrow & Martin, 2019;Schmelzkopf, 2002;Staeheli et al., 2002). Furthermore, scholars have documented urban garden projects that fostered greater participation and recognition by enabling grassroots political power, self-determination, and community empowerment (Irazábal & Punja, 2009;Saldivar-Tanaka & Krasny, 2004;Sbicca, 2019;Schmelzkopf, 1995;M. White, 2010M. White, , 2011. In terms of recognition, urban gardens serve as important sites of cultural reproduction, particularly for Indigenous, migrant, immigrant, and other communities for whom the gardens may have unique culinary, medicinal, and spiritual importance (Airriess & Clawson, 1994;Saldivar-Tanaka & Krasny, 2004). However, urban gardening does not inherently contribute to justice. Urban gardeners may be motivated solely by interest in fresher food or recreation rather than broader structural change (Horst et al., 2017;Lovell et al., 2014). Furthermore, even urban agriculture projects that espouse justice goals may reinforce racial inequalities in practice (Guthman, 2008;Kato, 2013;Passidomo, 2014;Reynolds, 2014;Safransky, 2017). Additionally, urban agriculture has long been entangled with the projects of settler colonialism and racial capitalism (McClintock, 2014(McClintock, , 2018. Despite these seeming contradictions, urban gardening and land access have repeatedly emerged as a central concern of justice movements precisely because land dispossession has been central to the oppression of low-income people and people of color (Gilbert & Williams, 2020;McClintock, 2018). In sum, while urban gardening on its own is limited in addressing the root causes of injustice (Agyeman, 2013;Alkon & Mares, 2012), nevertheless, it can make important material and symbolic contributions to broader efforts of low-income people and people of color to exercise greater power in the food system and in urban space (Block et al., 2012;Horst et al., 2017).
One key to fostering more just urban agriculture is stable, long-term land tenure, a common and persistent challenge for gardeners (Diaz et al., 2018;Lavallée-Picard, 2018;Vitiello & Wolf-Powers, 2014). It is difficult for urban gardens to compete with profit-or tax-generating land uses under urban governance regimes that prioritize the principle of highest-and-best-use (Vitiello, 2022). Prioritization of profit-and tax-generating potential pushes gardens onto economically marginal land; and, even then, access is often temporary in the face of changing property markets. Uncertain tenure has substantial socio-ecological consequences. It makes it less likely for gardeners to invest in practices with significant upfront costs but longer-term benefits such as soil health, perennial crops, and water infrastructure, to develop relationships with neighbors, and to foster broader social or political change. When a community garden is successful at providing benefits to a neighborhood but lacks stable tenure, it may even contribute to increases in surrounding property values, risking the displacement of the garden or the gardeners themselves (Glennie, 2020;Sbicca, 2019).
Many urban agriculturists have sought access to publicly owned land to avoid development pressures and secure long-term tenure. Given systemic racial disparities in wealth and land ownership, providing equitable land access is one way municipalities can foster a more just urban agriculture system (Desjardins et al., 2011;Horst et al., 2017;Thibert, 2012). However, municipalities must design such policies carefully to avoid reinforcing existing injustices (Cohen & Reynolds, 2014;Jerme & Wakefield, 2013). The emerging literature on urban agriculture policy underscores the need for a ground-level understanding of how policy actors grapple with this challenge. Through case studies of the City and the Park Board, we examine whether and how policy actors sought to integrate equity, participation, and recognition throughout the policymaking process and in the final policy language itself. In both case studies, policy actors used "racial equity" as the dominant frame to articulate goals and formulate policy alternatives. The term foregrounds racialized groups and distributive justice as critical sites of analysis. However, it was often used more expansively to include questions of participation and recognition of multiple identity categories in addition to race, including physical ability, primary language, housing status, and immigration status. While we situate the case studies in the broader academic scholarship on justice, we primarily use the frame "racial equity" to acknowledge the intellectual contributions of the research participants.

Methods
We examine garden land access policies and policymaking by the City and the Park Board using a contextualized case study approach to examine similarities and foreground differences (Locke & Thelen, 1995;Simmons & Smith, 2017). This approach is apt for probing the ways the two institutions are shaped by similar social and political pressures that are mediated by distinct missions and cultures, bureaucratic structures, participation processes, and the particularities of the individuals involved (Locke & Thelen, 1995). Data collection included participant observation at Park Board and City council meetings, public planning meetings, and community group meetings from 2015 to 2019. In addition, in the fall of 2019 semi-structured interviews were conducted with 35 elected officials, staff, and community organizers and gardeners who participated in the policymaking processes, and we collected plans, reports, meeting minutes, and draft and final policies from public websites and at planning meetings. In analyzing the data, we employed a semi-open iterative coding strategy (M. Williams & Moser, 2019) to allow themes to emerge while keeping a keen eye on how seemingly minor details of policies and policymaking processes can influence racial equity.

Minneapolis Context
Straddling the Mississippi River in the upper U.S. Midwest, Minneapolis has a temperate climate with a growing season from about May through September (Minnesota Department of Natural Resources, 2021). The nationally recognized Park Board manages 6,800 acres (2,750 hectares) of parks, ornamental gardens, golf courses, and trails, accounting for 15% of the City land area (Minneapolis Park and Recreation Board, 2020; Trust for Public Land, 2020). In 2015, when data on vacant lots was last publicly available, the City was the single largest owner of vacant land, with more than 700 parcels primarily intended for future multifamily or business development (Shoquist, 2015).
Minneapolis has roughly 430,000 residents, of whom 37% are people of color (U.S. Census Bureau, 2021). Though the metropolitan region boasts high overall levels of educational attainment, median income, and homeownership, these aggregate measures mask deeply entrenched racial inequalities (Table 1; Minnesota Compass, 2021c, 2021dNickrand, 2015;Thompson, 2015). Public policies have inscribed these racial inequalities on the landscape. Racial covenants, redlining, interstate highway construction (often deliberately routed through Black and Brown communities, with racialized use of eminent domain), discriminatory siting of public housing, and other policies created distinct areas of poverty and affluence that closely correspond with racialized groups (Figure 2, Mapping Prejudice, 2021;Metropolitan Council, 2015;Nelson et al., 2020;Shelton, 2018). In the run-up to the 2007− 2008 financial crisis, banks disproportionately targeted Black homeowners for subprime mortgages, deepening the racial homeownership gap and further concentrating vacant parcels in a handful of neighborhoods (Metropolitan Council, 2015). These broad contours of racial inequalities shape the landscape of possibilities for urban gardening.
Urban gardening has a long history in Minneapolis, punctuated by a few notable upswings in discursive and material support. For example, in the 1910s the Minneapolis Garden Club led a massive beautification campaign that eventually grew to encompass 400 acres. The social reformers who founded the garden club sought to reshape the physical and social landscape, and prominent real estate developers provided enthusiastic supportsometimes in combination with racial covenantsto boost property values (Walker et al., 2023). During the Great Depression, the Family Welfare   Association managed a massive unemployment relief garden tended by hundreds of families (Salsberry, 1931). Local leaders also promoted patriotic gardening campaigns during both World War I and II (Pack, 1919;"Victory Garden Goals Doubled" 1943). Over the following decades, urban gardens attracted interest from apartment dwellers, counterculture hippies, and immigrant communities which included the growing number of Hmong refugees (Wascoe, 1981). By the 1990s, Minneapolis ranked second nationally in community gardens per capita (Lawson, 2005). By 2016, there were 295 community gardens and urban farms on public or private land in Minneapolis (Homegrown Minneapolis, 2017). Around 2010, a loose coalition of gardeners and community organizers began advocating for better access to public land to grow food. Some gardeners had been able to access City or Park Board land for some years on an ad hoc basis, but access often hinged on well-resourced and wellconnected individuals to champion the projects. Advocates hoped that formal policies would increase land availability and make the process more transparent and equitable. The Homegrown Minneapolis Food Council-an advisory group established in 2012 and made up of community members, City staff, and elected officials-served as the key forum for discussing land access policies (City of Minneapolis, 2011;Homegrown Minneapolis, 2019a), and the Park Board eventually created its own urban agriculture advisory committee a few years later.

Access to Public Land for Gardens in Minneapolis
We begin with a brief overview of each case and then examine the details of the respective policies and their implications for justice. While Table 2 summarizes key elements of the City and Park Board polices, we limit discussion to three illustrative examples: land availability, costs to gardeners, and application selection criteria. Throughout, we highlight the crucial dynamics of participation in each case and their impact on the final policy language.
The City's Garden Lease Program allows community and market gardeners to lease some vacant City-owned parcels (we focus in this study on community gardeners). It was launched in response to the 2009 Homegrown Minneapolis report, drawing from feedback from over 100 community groups, which identified land access as a top priority for strengthening the local food system (City of Minneapolis, 2009Minneapolis, , 2010. The City acted quickly, launching a pilot program the following year with 21 parcels. In 2015, the City directed staff to draft a formal policy and the Land Access Committee of the food council provided comments and recommendations. By the end of the year, the City council had approved new parcel selection criteria, application requirements, and lease terms (City of Minneapolis, 2015aMinneapolis, , 2015b. With these changes, the number of available lots shot up to over 100 but has fluctuated since (Homegrown Minneapolis, 2017Minneapolis, , 2019b. The policy does not fit comfortably within a single department, requiring coordination between the Department of Community Planning and Economic Development (which owns most of the parcels and executes all leases), Public Works (which owns some of the largest parcels), the Sustainability Department (whose staff handle public outreach), and Homegrown Minneapolis (which makes policy recommendations).
The Park Board's community garden policy applies to gardens within parks, as well as a few scattered tax-forfeited parcels (we focus on the former). The Park Board policy development process began later and lasted substantially longer than that of the City. From initial community engagement in 2012, it was nearly eight years before the pilot program was launched, compared to a single year for the City (see Appendix A for a more detailed timeline). Staff presented a draft Urban Agriculture Activity Plan (UAAP) to the board of commissioners in early 2014. Public testimony convinced Commissioners to amend the final plan to include explicit racial equity goals and metrics (MPRB, 2014b(MPRB, , 2014c. 1 By 2015, the UAAP Implementation Teammade up of staff and community members-convened to develop a garden policy. This allowed gardeners and community organizers to engage in long-term, intensive participation, not only debating overarching policy goals but also directly shaping granular policy details. Furthermore, several organizers brought their experiences with the City's policy to the Park Board and insisted on defining some implementation procedures. Previously, implementation procedures were set internally by staff, so allowing public input at this stage in the policy process was a major departure for the Park Board, an important point to which we will return. The 2019 season was a pilot program with four gardens tended by staff and volunteers. By 2021, the policy was in full swing, with eight community gardens and plans to add nine more over time. Similar to the City policy, the Park Board requires substantial collaboration between departments, including Planning, Environmental Stewardship, Asset Management, and Recreation, which are responsible for policy development and funding, staffing, maintenance, and on-site oversight, respectively.
The two institutions differ substantially in how they determine what land is available for gardening, with important implications for justice. The City primarily bases its decision on a parcel's lack of development potential while the Park Board primarily decides based on neighborhood interest, site conditions, and funding availability. This difference is in part a result of the distinct missions, cultures, and political leanings of the institutions, but perhaps more importantly because of the underlying legal structures that constrain what the City and Park Board may do with land they own.
When the City department of Community Planning and Economic Development (CPED) acquires a parcel, typically it "seeks to quickly return the property to a tax-generating use" (Berkholz, 2009, p. 10) and has often acquired the parcel from the state or county on the condition that it is returned to private ownership. Thus, CPED offers only lots that are undevelopable, undersized, or that "pose marketing challenges" through the Garden Lease Program (City of Minneapolis, 2015a, p. 1). Using lack of development potential as the guiding criteria led to several challenges. First, the selection process does not consider suitability for gardening. A 2016 volunteer assessment found that nearly a quarter of available lots had deep shade, no water access, or other challenges. Unsurprisingly, nearly all of these challenging parcels remained unleased. Second, in the context of racialized spatial inequality, the use of market-based criteria led to a clustering of available lots in neighborhoods with some of the highest proportions of people of color and people in poverty (Figures 2 and 3; Minnesota Compass, 2021aCompass, , 2021b. This distribution could potentially expand gardening opportunities for residents of these neighborhoods, but proximity is not the same as genuine access. Other aspects of the policy-such as costs, complex requirements, and others discussed below-can make it disproportionately difficult for low-income people and people of color to participate in the program even when there are eligible lots nearby. Third, there was a spatial mismatch between available lots and demand for garden lots. Several parcels in South Minneapolis received multiple applications, while some suitable parcels in other parts of the City received none. Fourth, availability of lots is tied closely to shifting market conditions. From 2016 to 2020, the number of lots available ranged from 80 to 100. City staff predicted that 25% of garden lots could be developed from 2020 to 2025. Furthermore, land tenure is largely short-term and tenuous. The City offers leases that last one, three, or five years, but can terminate leases at the end of any growing season, which leaves gardens on CPED-owned parcels especially vulnerable to shifts in property markets. (The few eligible parcels owned by Public Works are typically held permanently, but gardens may be displaced by other departmental mandates such as infrastructure maintenance.) Indeed, even when the total number of lots has remained stable, there has been substantial turnover in the particular lots available through the Garden Lease Program. When selling a parcel before the lease term, the City sometimes offers displaced gardeners an alternative site. However, gardeners asserted that this does not adequately account for the loss. They cannot transfer their investments in soil health or relationships with neighbors to a new site. Garden sites are not fungible.
The Park Board's mission and enabling legislation provide it with much wider leeway to acquire and hold land for public use in perpetuity. For gardens on existing parkland, rather than competing with the exchange values of land, gardening is largely weighed against other recreational uses. (The Park Board does hold some tax-forfeited parcels for community gardens, but this is relatively rare.) Potential garden sites are identified through the park master planning process, based primarily on interest from neighborhood residents and site characteristics. These factors avoid several concerns about spatial distribution, site quality, and tenure involved in the City's policy. However, the Park Board policy poses other challenges related to the amount of land available, engagement of under-resourced groups, and tenure for individual gardeners. First, in terms of land availability, demand for garden plots far outstripped supply. In 2020, there were 130 plots across eight garden sites, but the Park Board received twice that number of applications. Nine additional park sites are planned, but it will take several years to build them. As to where land is available, the Park Board's garden sites are distributed more evenly than the City's ( Figure 3). Second, gauging community interest through the master planning process is undoubtedly a positive step toward matching demand for garden space with access to land. However, the Park Board faces common challenges in engaging underrepresented groups, such as non-native English speakers and low-income people. Third, under the Park Board's policy, garden site tenure is assured, but advocates were split on whether an individual gardener's plot tenure should be guaranteed year-to-year. Some saw long-term tenure for individuals as the foundation of responsible soil management practices and lasting social bonds. Others were concerned that guaranteeing the same plot year-to-year could lock out newcomers and over time disproportionately favor white, well-off residents. This second faction proposed giving no preference for returning gardeners; eventually, the implementation team and Park Board staff settled on this approach, which means that there is no guarantee an individual gardener will have a plot from season to season.
Costs for gardeners to access land through these 2 Initially, the City also recommended a refundable $250 damage deposit. Public comments and a formal recommendation from Homegrown forcefully argued that the additional cost represented a substantial financial barrier. Instead, the City agreed to charge gardeners if parcels required grass reseeding when the lease expired. 3 The Homegrown Council and community members objected to the insurance requirement because of the high cost burden, but a powerful council member insisted that it was necessary to protect the City risk. policies can create barriers that exacerbate raceand class-based disparities. As such, they reflect an institution's (un)willingness to commit resources to reduce systemic barriers. Overall, the City's policy involves higher financial, in-kind, and labor costs for gardeners than the Park Board's policy.
Under the City's final policy, leasing a typical garden site involves roughly $600 in direct costs before the season begins. This includes $51 in administrative and annual lease fees. 2 The City also requires gardeners to carry a US$1 million liability insurance policy, which costs roughly US$400 per year (A. Diamond, personal communication, 2021). 3 Once a parcel is leased, gardeners incur Journal of Agriculture, Food Systems, and Community Development ISSN: 2152-0801 online https://foodsystemsjournal.org additional costs for access to water from a public fire hydrant (a US$150 upfront cost, plus metering) or they may reach a private agreement with neighbors to access water. Gardeners also assume costs for plants, soil amendments, tools, and temporary infrastructure such as raised beds, fencing, and pathways. However, Homegrown staff are often able to coordinate deliveries of free or low-cost compost and recruit outside partners to offer free seeds.
Regarding labor, the City lease agreement requires gardeners to mow all lawn areas in the summer and clear snowy sidewalks in the winter. Many community members objected to this requirement because it involves costly equipment and additional labor, some occurring outside the growing season, that is not directly related to gardening. However, by shifting maintenance to gardeners the City saved an estimated $3,600 per leased parcel, precisely what made the policy attractive to some elected officials. One council member stated, "In terms of all the other good stuff with regards to healthy foods and all that, that's great and all. For me, the important part of this is save [the City] money" (Minneapolis City Council, 2015). This provision prioritizes cuts in municipal spending over reducing financial barriers to land access.
The Park Board policy involves no fees or liability insurance, reducing upfront financial costs to gardeners by hundreds of dollars compared to the City policy. This was primarily a result of insistence by organizers on the implementation team. Many gardeners lauded this as making the policy more equitable, though some staff suggested a modest fee (e.g., US$10) to ensure greater follow-through among participants, a tactic that had increased attendance for other space-limited programming.
Another major point of contention arose around funding for garden construction. Initially, the Park Board followed the City in making gardeners responsible for the labor and costs of creating the gardens, while many community organizers and some staff advocated for more funding, especially for expensive infrastructure such as accessible paths and water access. The Park Board was reluctant to commit additional funds until an internal vetting process revealed that the Park Board's existing liability insurance would only cover employees or contractors engaged in construction, but not garden volunteers. In the end, the Park Board agreed to fully fund garden construction, allocating US$75,000 per year. By covering construction and liability insurance, the Park Board dramatically decreased direct costs to gardeners compared to the City policy. However, it also limits how quickly the program can grow under current budget allocations. For reference, a single water line costs roughly US$30,000, 40% of the annual budget allocation for gardens. Furthermore, it gives the Park Board more power in determining the design and layout for gardens, whereas gardeners have considerably more freedom under the City's policy.
Regarding other costs, similar to City policy gardeners provide most plants, tools, and other materials. However, the Park Board is often able to offer a selection of transplants from their own greenhouses or suppliers. Unlike the City, the Park Board policy does not require gardeners to invest in any maintenance beyond their garden plots.
The application process determines who can access the benefits of the policy, which is fundamental to questions of equity and recognition. Under the City's policy, applications typically represent a group of gardeners-which may be organized formally through a neighborhood organization or church, or informally as a few neighbors interested in gardening together-whereas under the Park Board's policy applicants are typically individual gardeners. While the policies are not precisely analogous, comparing the application processes is still helpful for interpreting the intended beneficiaries and potential barriers.
The application for the City's program is a multi-step process, includes dense legal language, and requires a fiscal sponsor as well as proof of insurance. Organizers report that the process is especially challenging for gardeners with limited English proficiency, or limited knowledge of, trust in, or time to navigate City bureaucracy. Staff eventually developed a helpful 26-page handbook for prospective gardeners; while the handbook is clear and well-designed, the fact that it is necessary underscores the complexity of the process.
At first, the City did not clearly define how staff would select prospective lessees when multiple garden groups applied for the same lot. Despite well-intentioned staff, the lack of clarity allowed room for doubt to grow among gardeners, leading to distrust and accusations that the process was inequitable. Now the Garden Lease Program materials explicitly list four criteria staff use to select applicants. First, community gardens receive priority over market gardens. Proximity is also considered: garden organizers who live closest to the desired lot receive priority. Garden groups with existing leases receive priority over new applicants. Fourth, if multiple applicants for the same parcel meet all the above criteria, applicants who are willing to share the lot receive priority. If no party wishes to share it, the lease is awarded by lottery.
For the Park Board, staff initially considered applicant selection as a procedure to be determined internally without public consultation, similar to City policy. However, community organizers on the implementation team continually pressed to expand the scope of their work to include application selection procedures, and the Park Board eventually assented. The implementation team quickly rejected a first-come, first-served approach over concerns that it would favor people most likely to hear about the opportunity through the Park Board's website and email lists, which according to the Park Board's own analysis skewed whiter and wealthier. The implementation team also considered a lottery approach where any on-time application would have an equal chance of being selected-the same process used to grant use of many other park amenities (e.g., berths in canoe storage racks). Community organizers viewed this approach as formally equal but not equitable; that is, it did nothing to affirmatively address inequalities.
Instead, advocates proposed using criteria to rank applications. The first three were uncontroversial, giving preference to applicants who (1) are Minneapolis residents, (2) plan to grow food rather than ornamentals, and (3) have a connection to the site (e.g., proximity to home, school, or work). 4 Defining a criterion to address racial equity was a much thornier question. Advocates initially discussed criteria such as income, race/ethnicity, homeownership, and citizenship status, but these posed a number of challenges. First, it was impossible to agree on their relative importance. In addition, the Park Board's legal counsel warned that considering race in the provision of benefits would risk violating the constitution. Furthermore, many advocates worried that such detailed questions could feel invasive, stigmatizing, and potentially threatening, particularly for undocumented immigrants. Instead, the implementation team eventually settled on a criterion that prioritized applicants who do not otherwise have access to garden space. This decision passed legal muster and many advocates felt it to be a reasonable-if impreciseproxy for disproportionately low rates of land ownership among low-income people and people of color. If necessary, based on the number of applicants and plots available, a committee of staff and community members then considers open-ended questions, such as why the applicant wishes to garden. Compared to the City's application, the Park Board's application is quite simple, consisting of a handful of questions that fit on a single page.
In summary, regarding recognition and equity, advocates at the City and the Park Board continually sought to center the needs and wishes of gardeners who are low-income or people of color in determining the details of the policies. This approach included arguing for more land, secure and long-term tenure, lower costs, equity-informed applicant selection, and other terms intended to reduce barriers to land access. Many staff and elected officials at both institutions also voiced support for these goals. However, at the City, although community members helped get the policy on the agenda, they were primarily limited to an advisory role; decisions about specific policy details and procedures remained internal to City staff and council members. Furthermore, even if elected officials had evinced a greater willingness to protect land permanently for urban agricultural use, it would have required establishing new legal structures to do so. Ultimately, the City took relatively small steps, such as reducing some fees and lengthening some lease terms, to address community member concerns.
In contrast, at the Park Board's implementation team, community members participated directly in setting the policy terms, though the process was more protracted and openly contentious at times. The result was a Park Board policy that goes much further: it attempts to match land availability with community interest, funds garden construction, has a short and simple application, and includes selection criteria that aim to address inequality in land access. Indeed, a Park Board planner proudly described the final policy as "a collective vision" that "reflects community voice." Community participants were less effusive but still positive; one stated, "This is a huge thing for the Park Board to do, but it's not happening out of the goodness of their heart. It took a lot of pushing, a lot of trainingIt is not perfect but it's way better than a lot of the outcomes we've seen previously." The policy cases illustrate both the surprisingly large influence that seemingly minor details can have on racial equity and the ways public participation, particularly at the stage where policy details and procedures are set, can contribute to more equitable policy.

Discussion and Conclusion
Drawing on theories from urban planning and environmental justice, this paper explores the radical possibilities, thorny tradeoffs, and contentious disputes involved in providing access to public land for gardening. Urban agriculture is often imbued with a taken-for-granted goodness in policy discussions, which can serve to gloss over questions of power and justice. A close examination of the policy development process and the policy details helps to re-politicize the policy, revealing the underlying agendas that are ultimately served (Hammelman, 2019;Swyngedouw, 2015). In these cases, the details of the City's policy reveal an emphasis on cutting costs, preserving the ability to sell parcels for development, and a reluctance to commit substantial resources to reducing barriers to garden land access. What Schmelzkopf (2002) describes as the "hegemonic project of the government to maximize exchange values" (p. 323) is visible both in the political attitudes of key City council members and the deeper structures that constrain City policymaking. In contrast, the Park Board's statutory authority includes permanently holding land for public use, loosening the grip of the otherwise dominant market logic. Ultimately, the City policy suggests a view of gardening as marginal to its core functions, while the Park Board policy views gardening as squarely within its scope.
A justice lens highlights how concrete policy details distribute benefits and burdens, the degree of public participation, and the social groups that may be (mis)recognized. Practitioners-elected officials, staff, gardeners, and community organizers-involved in garden land policies can incorporate equity concerns by carefully considering access and barriers to long-term tenure including not only proximity, but also the complexity of the application process, fees, and other requirements. Practitioners must also take into account the distribution patterns of soil and air pollution, as well as access to other resources necessary for growing food, such as water, compost, soil testing, and raised beds. As the present cases illustrate, substantive participation by gardeners is essential. There should be opportunities to meaningfully shape land access policies at all stages, from goal setting through evaluation, as well as to determine garden sites (at all stages, from design through dayto-day maintenance). Based on my findings, we contend that participation processes for urban garden policies should also be sensitive to potential variation in availability of participants with respect to the local growing season. We argue that addressing recognition requires leeway for a wide range of gardening practices, aesthetics, cultural meanings, and ways of relating to other-than-human species. Furthermore, recognition requires accounting for historical and ongoing traumas experienced by many in terms of land, agriculture, and food, including land dispossession, enslavement, forced migration, and labor exploitation. Policies should include provisions for redress and repair. Finally, the present study illustrates the value of attending to the ways in which equity, recognition, and participation are linked. Rather than arguing over the relative importance of recognition versus distributive concerns (Young, 1997;Fainstein & Fainstein, 2013;Fraser, 1995bFraser, , 1997 or of participation and the discursive turn (Fischer, 2009), these elements were tightly bound in both policy cases.
While much of the scholarly literature emphasizes the role of public engagement in defining broad goals and shaping the policy agenda, the present case comparison illustrates the importance of public participation throughout the policy development process, including during the implementation and evaluation stages (Bryson et al., 2012;Cooper et al., 2006). At the City, Homegrown Minneapolis members facilitated a community evaluation process to formulate recommendations presented to the City council, some of which were eventually adopted. At the Park Board, advocates worked to expand their scope of influence to include setting some implementation and evaluation procedures. The role of participation in the City and Park Board cases reflects the insights of Majone and Wildavsky (1995) about the contingent nature of policy and the power that lies in implementation: If problems are best understood through solutions, then implementation includes not only finding answers, but also framing questions. Reformulating problems means changing solutions. Policy ideas in the abstract … are subject to an infinite variety of contingencies, and they contain worlds of possible practical applications. (p. 149) In spelling out Park Board procedures, advocates sought to protect their vision of the policy's intent by laying out terms and procedures in much greater detail than is typically done, thus narrowing the space for staff to impose their own interpretations of what the policy should be. In conducting an independent policy evaluation, advocates sought to assert their own benchmarks of success and underlying visions for the policy. We do not wish to over-romanticize participation or to argue that more participation is always better. These cases highlight many of the complexities, nuances, and difficult tradeoffs involved in participatory policymaking that other scholars have identified (Bryson et al., 2012;Slotterback & Lauria, 2019). Through the act of gardening, community members already participate directly in the practice and implementation of any land access policy, and gardeners may not have an interest in other stages of the policy process or be able to afford an investment of unpaid time. In the present study, most advocates who could sustain engagement with the policy process over the long term were community organizers with paid positions at nonprofits. On one hand, this reduced the burden of unpaid labor for community members. On the other hand, the organizers were largely white and well-educated, raising questions about the representativeness of participants, which is a challenge for any uncompensated public engagement process.
The study has several limitations. First, these cases represent only one particular period in time. My fieldwork ended in the fall of 2019, the pilot year of the Park Board policy, do the case study does not cover policy implementation. Second, data collection ended before the COVID-19 pandemic and the attendant economic instability, as well as before the uprisings following the police murder of George Floyd. Nevertheless, the present study provides a window onto one strand of racial justice activism and policy work preceding these multiple overlapping crises. Future research should explore the influence of these dynamics on urban agriculture land access policies.
Urban garden projects can advance social justice (Irazábal & Punja, 2009;Saldivar-Tanaka & Krasny, 2004;M. White, 2011) but can also exacerbate existing injustice (Guthman, 2008;Kato, 2013;Reynolds, 2014;Safransky, 2017). If protecting public land for gardening is to contribute to a more just urban agriculture system, we must attend to how land is protected and who has the power to make such decisions. This study contributes a rich ground-level examination of policy language and development processes to the growing scholarship interrogating the role of urban planning and policy in supporting urban agriculture (Cohen & Reynolds, 2014;Halvey et al., 2021;Hammelman, 2019;Horst et al., 2017;Jerme & Wakefield, 2013;Meenar et al., 2017;Pothukuchi, 2015;Thibert, 2012). The goal of increasing public land for gardening attracted support from policy actors with quite different underlying values and purposes in mind. During the policy development process, contentious debates over policy details laid bare tensions among these underlying values and whose agendas would ultimately win out. This case study comparison illustrates how what would appear to be minor details can uphold or undermine racial inequality. Policies must be deliberately designed to reduce structural barriers and ensure that benefits flow to the most marginalized communities. The best chance of doing so comes from the meaningful participation of gardeners from such communities throughout the policy process.