CHAPTER 5

Partnering with the Public

ESTUARIES AND WETLANDS are an often overlooked instance of a common. Though they do not produce much in the way of immediately useful resources for humans, they are critical to the overall stability and health of the watershed. They buffer the impacts of storms and flooding, recharge the groundwater, store excess carbon, and act as a refuge and breeding ground for many species of wildlife, thereby helping to sustain biodiversity. For this reason, many environmental groups hold the preservation of these unique and valuable areas as a critical part of their mission. Tom Langen, a biologist at Clarkson University who specializes in the study of wetlands, worked with one such group that sought to convince private landowners to donate any wetlands in their holdings for conservation. The environmentalists brought to the landowners what they believed to be an enticing opportunity to contribute to the protection of lands vital to the watershed they called home. This logic appealed to their sense of a common good and their desire to promote it. Unfortunately, it swayed few people.

What did convince people to donate their land to the program, as Langen recounts, was concern for their legacy. Whereas most landowners were not especially motivated by the opportunity to preserve the watershed, they were enticed by the idea that the land would be donated in their name, in which case their contribution to the watershed would be attributed to them in perpetuity. The environmental group quickly reoriented its outreach strategy, concentrating on the opportunity for individuals to leave the land in their name to the conservation project rather than having it pass into other hands through sale or inheritance, at which point their legacy would end.

Just like a 311 system, the wetland preservation story entails a collaboration between an institution and the public in the maintenance of a common. Particularly instructive about this case are the distinct motivations the two parties had. The environmental group was concerned with protecting the wetlands for the broader public good, but the private landowners were attracted by the creation of a legacy. Until the environmental groups understood this, their outreach was ineffective, because they were selling their own motivation rather than speaking to the interests of their audience. A similar thought process is visible surrounding 311, which is one of the newest manifestations of coproduction, or government programs that directly involve constituents in the design and delivery of services. Proponents of “civic tech” often treat 311 and related innovations as expanding the channels for civic engagement or political participation and in turn classify custodianship with behaviors like voting, volunteering, and donating to civic groups. But is this just an assumption based on the fact that it is a government program intended to involve the public in the delivery of services? As the story of wetland preservation demonstrated, targeting the correct motivations is critical to the success of a program; otherwise it will likely fail to garner much participation.

Here we see an opportunity to convert the insights from Part II into practical value—in short, to transition from what we can learn from urban informatics to what communities can directly gain from such work. Building on our newfound understanding of territoriality as a major motivation for custodianship, Part III of this book will address more fully the motivations that drive use of 311 and the ways in which these insights can support additional refinements to such programs. In doing so, it also evaluates the promise held by civic tech for communities and how such innovations are best designed. This chapter further probes the question of why people use 311. There is no reason to believe that territoriality is the only motivation relevant to custodianship, and thus we consider whether it also arises from a desire to be engaged civically or politically, or what I call a civic disposition.

Testing the relative importance of territoriality and a civic disposition for the usage of 311 is more than just a question about a single program. It is also an opportunity to examine coproduction’s more general assumption that a civic disposition is the primary motivator for participation in any government program. We will see here that there is a need to move beyond this narrow perspective to one that acknowledges that a given coproduction program might engage any of the numerous facets of human psychology, depending on the nature of participation it requires. This lesson is especially valuable as civic tech expands the variety and nature of such programs. This work builds on a project I undertook with Dietmar Offenhuber of Northeastern University, Jesse Baldwin-Philippi of Fordham University, Melissa Sands of the University of California Merced, and Eric Gordon of Emerson College. ¹ Chapter 6 then builds on these insights to assess a series of public interventions and experiments that seek to increase participation in 311. Importantly, they are designed and evaluated in light of the motivations that they seek to engage. Overall, these two chapters illustrate how civic tech can enable effective coproduction if paired with a thoughtful examination of when and why individuals would be interested in participating in a given program.

Collaborations between the Government and the Public

As I have noted before, the proliferation of 311 systems and allied programs has been part of the broader trend of civic tech. Just as computational social science has promised to leverage data to transform our understanding of human behavior, this movement has seen digital technology as a catalyst for diversifying and strengthening the ways that people can contribute to their local communities and to society in general. One subset of this work has focused specifically on the lines of interaction between municipal governments and their constituencies. Previously, such interaction centered primarily on town hall meetings and other public forums, which tend to be dominated by a small, relatively vocal subset of the community. The limitations of this system have often left both government officials and the constituents they serve wondering what exactly was accomplished. ² The hope has been that “Gov 2.0” will be able to leverage the two-way communication capacities of modern web tools to solve these sorts of challenges. ³ Stephen Goldsmith and Susan Crawford have also lauded such efforts as establishing a new form of public administration that is responsive to the needs of the public in near real time. ⁴ And others have argued it will transform public deliberation and the crafting of public policy. ⁵

The promise of civic tech lies in more than just expanding communication between government and the public. The hope is that it can be the basis for what Benjamin Barber calls “strong” democracy, in which constituents are consistently involved in various aspects of designing, managing, and delivering government services. ⁶ This idea may not be as novel as it appears at first blush, however. Rather, it puts a technological spin on a long-standing concept in political science and public administration referred to as coproduction, or programs that directly engage constituents in the planning and implementation of government services. ⁷ Whereas classical examples of coproduction have included parent-teacher associations and community policing, civic tech has facilitated other forms of participation in government, including, of course, public deliberation, but also participatory budgeting ⁸ and planning new community developments. ⁹

As coproduction programs increase in number and popularity, they raise a question that has long faced such efforts: What motivates individuals to participate in a government program? This knowledge is critical because the success of coproduction programs stands and falls with constituent participation. Relative to the extensive literature on coproduction writ large, this question has received only a small amount of study. ¹⁰ More importantly, it has not been updated to account for the variety of coproduction programs that have emerged in recent years, including the proliferation of civic tech. Traditional theory has treated participation in a coproduction program as an intrinsically civic (or political) act and thus driven by a broader tendency to take part in activities that contribute to society (e.g., volunteering, voting). I refer to this unidimensional characterization of participation as the public-as-citizen model. Civicness is only one of many human motivations, though, and I argue that a given coproduction program might capitalize on any of them. I refer to this more expansive view as the public-as-partner model, in that it treats each member of the public as a multifaceted partner whose various capacities enable him or her to contribute in multiple ways.

Coproduction: A Brief History

In the 1970s, the Workshop in Political Theory and Policy Analysis at Indiana University introduced the concept of coproduction, offering it as an alternative to the prevailing practice at the time of administering government programs in a wholesale fashion through large, centralized bureaucracies. ¹¹ They argued that this should be tempered by a more localized approach that incorporated the public into service delivery. ¹² Philosophically, coproduction embodied the democratic ideals of citizen access to and participation in government. In a practical sense, such programs would improve services by allowing members of the public to tailor implementation to community needs. If this latter concept sounds similar to Elinor Ostrom and her colleagues’ emphasis on localized institutions as the most efficient solution to the problem of the commons, it should. Ostrom herself was one of the primary actors in the workshop and wrote a handful of prominent pieces on the subject.

One of the most eloquent descriptions of coproduction comes from Ostrom’s 1996 essay on the subject, defining it as “one way that synergy between what a government does and what citizens do can occur.” ¹³ From this, she reasoned that coproduction could make services more efficient and effective, but only if the efforts of these two entities were complementary and nonsubstitutable. Taking 311 as an example, the maintenance of the urban commons consists of two different activities that might be divided between government and the public. Urbanites observe and can report instances of deterioration or denigration in public spaces during their daily movements, and city agencies provide the professional expertise and equipment for fixing them. While this illustrates the potential advantages of coproduction, it also highlights the dependence of coproduction on the participation of the public and, in turn, the importance of understanding when and why members of the public would choose to do so.

Coproduction saw a decline in popularity during the 1990s, owing to the emergence of “the new public management,” an effort to make bureaucracy operate more like private industry. ¹⁴ This perspective cast members of the public as “customers,” rather than potential collaborators in the delivery of government services. Interest in coproduction has seen a resurgence, however, in the last 10–15 years, as the new public management has been replaced with the more consonant paradigm of “new public governance,” which views service delivery as the collaborative coordination of various partners. ¹⁵ Such collaboration might involve multiple government agencies working in concert to address a particular need but has also lent itself well to Ostrom’s stance that coproduction improves delivery systems by leveraging the complementary capacities of both government and citizens. ¹⁶ Enthusiasm for coproduction has been further fueled by the advent of modern web technologies. Proponents of civic tech have capitalized on these novel resources to create new mechanisms for communication and collaboration between the government and the public, enhancing the potential for true partnership.

In the years since Ostrom and her colleagues introduced coproduction, thinkers have expanded it to encompass a variety of different arrangements. The level of activity and responsibility attributed to the public can vary considerably across examples. In the most basic case, Stephen Osborne and his colleagues have argued that all service delivery entails coproduction because the recipient of the service must necessarily participate in its consumption. ¹⁷ At the other extreme, there are cases in which the public is the primary or sole deliverer of the services, such as when expert patients assist in the provisioning of health care. ¹⁸ Additionally, Tony Bovaird has illustrated ways in which coproduction can go beyond service delivery to include the planning and design of policies and programs. ¹⁹ Others have also extended the model to account for collaborative arrangements between government and community organizations or nonprofits. ²⁰ For our purposes here, 311 is specifically a case in which individual members of the public participate in service delivery, but it might also offer insights on individual contributions to the planning and design of services as well.

Public as Citizen or Public as Partner?

By definition, coproduction programs depend on the participation of the public and are unlikely to be effective without it. For this reason, a critical question is why members of the public do or do not choose to participate in a given program. Most work to date, however, has focused on governments and how they can better construct channels or distribute resources in ways that will make programs more accessible. ²¹ This emphasis on whether the public is able to coproduce has left little understanding, however, of why someone would want to do so. ²² Central to this latter question is how one conceptualizes constituents and their role. As recounted by John Clayton Thomas, as public administration has evolved in the last 40 years, so has its perspective on the public. ²³ From the 1970s into the 1980s, a period that also saw the original rise of interest in coproduction, members of the public were “citizens” with a major stake and interest in shaping policy and its implementation. In the 1990s, the philosophy of new public management ²⁴ treated the public as “customers” whose needs must be fulfilled. More recently, new public governance has emphasized collaboration across agencies and sectors, treating members of the public as an additional “partner.” The first and last of the three perspectives on the public are the most important for our purposes here because each presents a particular way of thinking about the active role that constituents might take in governance.

Existing research on participation in coproduction programs has largely focused on the public-as-citizen model. This is embodied by a popular metaphor that refers to coproduction programs as a “bridge to citizenship” by which participation will entrain and encourage involvement in civic life. ²⁵ This perspective treats participation in coproduction as an overtly civic or political action, reflecting a generalized civic disposition that manifests in a broader pattern of political participation, including behaviors such as voting and volunteering. There are clear weaknesses to this approach, the most apparent being that this proposed by-product of coproduction programs has been little tested, ²⁶ and the few studies that have tested it have found little evidence that other civic and political behaviors actually predict participation in coproduction. ²⁷

The lack of support for the public-as-citizen model calls for an alternative conception of the motivations that might drive participation in coproduction programs. As the new public governance and civic tech diversify the ways in which government seeks to collaborate with the public in service design and delivery, it raises the question of what it means to think of members of the public as partners. To inform this, let us think about how agencies and organizations become involved in the process of governance. In such cases, the potential of each organization to contribute to a program is not exclusively dependent on its will to provide a public good (i.e., the equivalent of civic disposition at the organizational level); in fact, there are cases where this does not appear to be the case at all. For example, in Ghana, the Public Road Transportation Union is a private association that collects taxes from buses and taxis on behalf of the government in exchange for extensive control over the transportation sector, ²⁸ an activity that speaks little to a desire to provide a public good. Rather, the manner and extent of a contribution by an agency or organization is determined by their specific capacities in that domain.

We might then define a partner as an entity with characteristic facets and capacities that can contribute to the collaborative process of governance. When one applies this definition to members of the public, it is apparent that the public-as-citizen model takes a narrow view of why constituents would be motivated to participate in a coproduction program. Indeed, humans are endowed with a diverse array of motivations that extend far beyond a capacity for civicness, and in theory a coproduction program might appeal to any one of these. To make effective use of this knowledge, we might turn to the psychological concept of modularity, which states that a given motivation is oriented toward certain goals or tasks. As such, it is responsive to relevant cues and contexts and will manifest itself in behaviors surrounding those goals. ²⁹ In this view, a motivation will be relevant to coproduction if a program’s activities evoke the cues or call on the behaviors associated with that motivation. This perspective enables us not only to expand our attention to motivations beyond a civic disposition but also reason what types of motivations would be relevant for any given program.

The one systematic attempt to identify multiple motivations for coproduction was made by John Alford. ³⁰ He proposed three types of nonmaterial rewards that might contribute to one’s willingness to coproduce: intrinsic rewards, such as increased self-esteem from efficacious action; solidary incentives, arising from a desire to contribute to the group; and expressive values, or normative beliefs about societal issues. Alford’s nonmaterial rewards offered an important advance, but they did not go far enough in that they were still generalized incentives, applicable to any coproduction program. Instead, an important value of the public-as-partner model as I have described it is that it permits the motivations of interest to vary across coproduction programs, depending on the nature of participation. In a practical sense, this broader view becomes increasingly necessary as new public governance in general and civic tech in particular further diversify the ways that constituents might participate in the design and implementation of policy. Indeed, part of the strength and appeal of coproduction programs is that they can leverage actions that have not previously been part of the governance process.

To summarize, the public-as-partner model provides a distinct perspective on participation in coproduction by considering any of the diverse array of human motivations as being potentially relevant to a given coproduction program. It expands on the public-as-citizen model by opening up the possibility that participation in coproduction could entail more than just a civic disposition, and it poses the previously unexplored corollary that different coproduction programs rely on distinct sets of motivations. If this were found to be true, it would call for a program-by-program approach when implementing, evaluating, and promoting coproduction. It also raises a second, complementary question: In actuality, how important is a civic disposition to coproduction?

Applying the Public-as-Citizen and Public-as-Partner Models to 311

The 311 system’s collaborative model for the maintenance of the urban commons makes for an effective comparison of the public-as-citizen and public-as-partner models because the two make divergent predictions about the motivations that would lead a constituent to identify and report an issue in the public domain. The majority of research on 311 has taken the former approach, classifying 311 reports with other forms of political participation or civic engagement, such as voting. ³¹ This would stand to reason, as issues in the commons are everyone’s problem but no one’s formal responsibility, meaning efforts to address them might be consciously civic or “for the greater good.” This logic, however, would appear to be a reiteration of the traditional model of the commons, which characterized contributions as a manifestation of “cooperation.” Part II of this book has already noted that a major weakness of this approach is that it uses the consequences of a behavior to infer the motivations, rather than considering the action itself on its own terms. As importantly, each of these studies has assumed this interpretation of 311 reporting to be true without directly validating it.

In contrast, when thinking of the public as a multifaceted partner, we might ask what other motivations are engaged during acts of custodianship. I have already argued at length that such behaviors are a manifestation of a fundamental human capacity for territoriality, and I have presented multiple lines of evidence to this effect. Nonetheless, it certainly does not preclude the possibility that a civic disposition plays its own independent role in motivating participation in 311 or even that it might account for the previously observed association between territoriality and custodial reports; for instance, it is possible that those with a higher civic disposition also express greater concern for their neighborhoods and that a higher civic disposition in turn is responsible for both that concern and custodianship.

The goal here is to test the distinct hypotheses arising from the two models. The public-as-citizen model predicts that usage of 311 to report public issues would be a function solely of a civic disposition and thereby greater in those who are also more active in other civic and political activities (e.g., voting). The public-as-partner model would predict that 311 reporting would be greater in those with higher territorial motives. Though we have already seen evidence of this relationship in Chapter 3 , here we retest it while simultaneously considering any effect of a civic disposition, which I did not do before. It is worth noting, however, that the hypothesis is not that territoriality would be active across coproduction programs or uniquely responsible for usage of 311, but simply that this motivation is particularly relevant to the form of participation required by 311 systems.

Testing the Public-as-Citizen and Public-as-Partner Models

To test the public-as-partner and public-as-citizen models, we return to the user survey presented in Chapter 3 . The current analyses supplement the two survey scales measuring territoriality—benefiting the local community and enforcing social norms—with two additional pieces of information about individual users. First, the survey also included a series of items about participation in civic activities. Second, my colleagues and I linked survey responses and 311 behavior to public voting records. These two additions provide the necessary measures for civic disposition. Additionally, rather than a simplistic analysis of “participation,” these analyses use the methodologies from previous chapters to describe the custodial habits of 311 users in multiple ways, including whether someone was a custodian or not, the volume and geographical range of such activity, and whether they acted as a custodian in various spaces across the urban landscape. This permits a more nuanced assessment of how civic disposition and territoriality might explain different aspects of 311 reporting. Importantly, the focus on measuring and analyzing individual-level behaviors here is unique among studies that have compared 311 usage to political participation, offering a number of advantages: ³² (1) it avoids the ecological fallacy, permitting a “true” individual-level interpretation; (2) it links objective behavioral measures from multiple databases with self-reported behaviors and motivations; (3) the multilevel models permit us to control for differences between neighborhoods in the number of public issues that might need attention, thereby factoring out similarities inherent in living in the same environment; and (4) this is one of the only studies of this sort that has isolated reports of a particular type of interest rather than just counting 311 calls of all types in an indiscriminate fashion.

Before proceeding with this analysis, it is important to recall a few details from the original presentation of the survey sample. The sample was notably more white, educated, and middle-aged than the overall population of Boston (see Table 3.1 in Chapter 3), though it is unclear how much it differed from the demographic composition of 311 users. The sample was also above average in its use of 311. It had an overrepresentation of custodians, and those custodians made more reports than would be expected by chance. These biases should be taken into account when considering the results reported here, though, as before, we have to consider whether the underrepresented groups would in fact have different relationships between territoriality, civic disposition, and custodianship. It is not clear whether or why this would be the case. The higher level of custodianship in the sample does, however, give us greater power to compare between groups and to understand the motivations of highly active custodians. Additionally, because the survey included 311 users of all types, we are able to compare custodians to noncustodians while accounting for confounding factors that may lead them to know of and use the system in the first place.

Measuring Civic Disposition

The survey asked respondents to indicate whether they had participated in each of nine civic or political activities in the previous six months (e.g., attended a meeting for a local community group or government agency; see Table 5.1 for a complete list of items and their prevalences). Contacting a government official or agency was the most widespread behavior (70 percent), though this might be inflated if some respondents included 311 as being a government official. For most other civic activities, including volunteering, signing a petition, donating to a civic group, and attending a community meeting, about half of the sample reported participating. The average individual participated in three activities (mean = 2.98, median = 3), and 109 respondents (15 percent) had not participated in any. Henceforth we analyze the activities as a sum rather than as individual activities. ³³

TABLE 5.1 Comparison of participation in civic activities, voting, and territorial motives between custodians and noncustodians
		Custodians		Noncustodians		Total
Civic activities and political participation
Signed a petition about a social or political issue		208 (48%)		119 (49%)		327 (49%)
Volunteered with a local or national civic group		194 (45%)		105 (43%)		299 (44%)
Donated to a local or national civic group		244 (57%)		126 (52%)		375 (55%)
Used social media to engage with a local or national civic group		150 (35%)		90 (37%)		240 (36%)
Contacted a government official or agency		316 (73%)		163 (67%)		479 (71%)
Participated in an online discussion or blogged about a political issue		125 (29%)		72 (30%)		197 (29%)
Participated in a march, protest, or demonstration for a political cause		28 (6%)		20 (8%)		48 (7%)
Sent a letter to the editor of a local or national newspaper		36 (8%)		31 (13%)		67 (10%)
Attended a meeting for a local community group or government agency		218 (51%)		112 (46%)		330 (49%)
Voted in 2011 municipal election		181 (42%)		89 (37%)		270 (40%)
Territoriality
Benefit community		4.39 (0.78)		4.20 (0.90)		4.32 (0.83)
Enforce norms		3.42 (1.21)		3.36 (1.28)		3.40 (1.23)
Material interests
Maintain property values		2.56 (1.57)		2.47 (1.49)		2.53 (1.54)
Total		431 (64%)		243 (36%)		N = 674

We identified 562 survey respondents in public voter records by linking on names and addresses. We assumed that those who could not be matched in this way were not registered and therefore did not vote. ³⁴ Voting provides a valuable measure of civic disposition, as it is the most common and basic manifestation of citizenship, and here we focus particularly on whether each individual had voted in the most recent municipal election, which was in 2011. This is for two reasons. First, conceptually, voting in a municipal election is the most appropriate parallel for studying participation in a municipal program. ³⁵ Second, from a measurement perspective, voter turnout in local elections tends to be exceedingly low, especially during odd-numbered years, as there are no concurrent federal elections. This makes it a more salient indicator of civicness. ³⁶ As with reports of civic activities, voting in the 2011 municipal election was elevated in our sample. We had a record of voting for 40 percent of the survey respondents compared to a 24 percent turnout rate for all registered voters that year.

Civic Disposition, Territoriality, and Custodianship

An initial descriptive analysis found little relationship between a civic disposition and custodianship. Custodians were no more likely than noncustodians to participate in any of the nine civic activities or to vote (see Table 5.1 ; χ ² _df=1 = 0.00 − 2.86, all p -values = ns ), nor did custodians and noncustodians differ in total number of civic activities (means: custodians = 2.99, noncustodians = 2.95; t -value = 0.28, p = ns ). In contrast, custodians reported a greater motivation to benefit the community (means: custodians = 4.40, noncustodians = 4.21; t -value = 2.78, p < .01), though the same was not true for the desire to enforce norms (means: custodians = 3.44, noncustodians = 3.39; t -value = 0.55, p = ns ).

Multilevel Analysis

To formalize the analysis of custodianship, civic disposition, and territoriality, I again utilize multilevel models. As in Chapter 3 , these models allow us to control for demographic characteristics and for confounds arising from the opportunities for reporting issues created by one’s neighborhood context. Though a major reason for using multilevel models was the potential for similarities between neighbors owed to a shared residential environment, I place all details testing for the presence of such clustering at the neighborhood level (i.e., τ and its significance) in footnotes as it is a bit tangential to the main theme. One difference between the sample analyzed here and that in Chapter 3 arises from the decision to leave the measure of material incentives out of the analysis given its general impertinence in previous analyses. This allows us to include 12 additional respondents who had omitted that item but had completed all of the other measures of interest and who were not part of the analysis in Chapter 3 . ³⁷ Major results from the analysis are represented visually in Figures 5.1 and 5.2 ; complete results are reported in Appendix C .

FIGURE 5.1 Dot plots illustrating (a) the correlation between the size of an individual’s home range and a combined territoriality score and (b) the lack of the same correlation with total civic and political activities.

The findings from these models were consistent with the descriptive comparison of custodians and noncustodians above. Users of 311 who participated in more civic activities or had voted in the 2011 municipal election were no more likely to act as custodians (civic activities: O.R. = 0.96, p = ns; voting: O.R. = 1.21, p = ns ). Those who expressed a greater desire to benefit their community were more likely to act as custodians (O.R. = 1.33, p < .05), confirming earlier analyses. ³⁸

Custodianship across the Urban Landscape

As in Chapter 3 , we can go further than a simple comparison of custodians to noncustodians to take account of the varied ways that individuals interact with the geography of a city. People have a home neighborhood but also visit other neighborhoods for work, travel, and recreation. We can attend to this in two ways. First, we can flexibly define a person’s “home neighborhood” through the geographic clustering of an individual’s reports. From this, we derive two measures: the number of reports made within one’s home neighborhood and the geographic range of custodianship in the home neighborhood. ³⁹ Second, we utilize items from the survey that asked individuals whether they used 311 in various locales, including their neighborhood of employment, where they visit friends and family, or on their commute. These analyses are necessarily limited to custodians. ⁴⁰

Individuals who expressed a greater desire to benefit their community or to enforce social norms had home clusters that were both larger (benefit community: O.R. = 2.36, p < .01; enforce norms: O.R. = 1.32, p < .05) and had more reports (benefit community: O.R. = 1.26, p < .01; enforce norms: O.R. = 1.22, p < .001). Notably, those participating in more civic activities had home clusters with fewer such reports (O.R. = 0.86, p < .001), but the range of reporting around the home was neither larger nor smaller (O.R. = 1.08, p = ns ). Having voted was unrelated to either measure (reports: O.R. = 0.96, p = ns; size: O.R. = 0.92, p = ns ). Figure 5.1 visualizes these relationships, illustrating how the size of one’s home range correlates with territoriality but does not correlate with civic and political activities. ⁴¹

FIGURE 5.2 Schematic depicting the dichotomous relationship between territoriality, civic disposition, and the geography of custodianship.

In order to better understand reporting outside the home neighborhood, we turn to the survey items about where people report issues. In the survey, 28 percent of custodians reported that they used 311 to report issues while on their commute, 24 percent in the neighborhood where they work, and 13 percent in the neighborhoods of friends and family. Those who expressed a greater desire to benefit the community and participated in more civic activities were more likely to state that they reported while on their commute (benefit community: O.R. = 2.23, p < .01; civic activities: O.R. = 1.57, p < .001), but only civic activities were associated with a greater tendency to report in the neighborhood where one works (O.R. = 1.34, p < .05). Those who voted were more likely to report from the neighborhood of one’s family or friends (O.R. = 1.38, p < .05). ⁴²

Summary: Observing the Public’s Role as Partner

The findings here strongly support the public-as-partner model of coproduction. Consistent with the analyses in Chapter 3 , the two components of territoriality—seeking to benefit the community and enforcing social norms—were associated with custodianship. Those with a greater desire to benefit the community were more likely to act as custodians, and both components of territoriality predicted more reports and a broader geographical range of reporting in the home neighborhood. Additionally, those with a greater desire to benefit the community were more likely to report while on their commute. These relationships are crucial to explaining global usage of 311 because, as we have seen, most reporting occurs in the reporter’s home neighborhood. In contrast, indicators of a civic disposition were predictive of reporting exclusively in contexts outside the home neighborhood, including at work, while on one’s commute, and in the neighborhoods of friends and family. These are much less frequent contexts for the use of 311, diminishing the relative importance of a civic disposition to the functioning of a system as a whole.

Beyond the relative importance of territoriality and civic disposition to 311, the geographic patterns of the results revealed something of a dichotomy in how these motivations interact with the urban landscape, as illustrated in Figure 5.2 . Territoriality explained 311 reporting in the home neighborhood, exactly where such motives should, by definition, be stronger. Meanwhile, 311 reporting outside the neighborhood was more strongly explained by a civic disposition, though, notably, these relationships were only moderately consistent, as three of six possible relationships between indicators of civic disposition and locales beyond the home neighborhood were significant. This seems fitting, as the decision to report throughout the city is similar in spirit to a desire to participate in activities that contribute to society more broadly. In fact, this distinction mirrors the assumptions of the public-as-citizen model and the public-as-partner critique. The former ignores the basic motivation to maintain public spaces over which one has a sense of ownership, which is foundational to the collaborative arrangement of 311. It is only when individuals break through this invisible boundary of personal interest that a civic disposition becomes a relevant factor.

The results confirm that 311 data should not be used to assay political participation, which many authors to date have done. ⁴³ This is especially true when the data are aggregated to assess an individual or neighborhood’s volume of activity. We do find, however, that reports outside a reporter’s home neighborhood (i.e., those not included in the home cluster) could arguably be used as an indicator of a civic disposition. Leveraging this methodology to interpret 311 participation in this manner would require the analysis of reports nested within individual accounts, as done here. This is not to say that the data cannot be used to test hypotheses about how relationships between constituents and government can influence participation, which is the focus of some existing studies, but rather that it would be inaccurate to treat 311 reporting as a proxy for activities such as voting or volunteering or as the primary motivation for such behaviors.

A New Perspective on Coproduction and Civic Tech

Traditional perspectives on coproduction treat a tendency toward civic and political behavior, or what I have referred to here as a civic disposition, as the fundamental basis for participating in and contributing to government programs. A similar set of assumptions is embedded in the stated mission (and very name) of civic tech, which seeks to enable a civic disposition. As an alternative to this public-as-citizen model, I argue that those studying and implementing coproduction programs should think of members of the public as partners—individuals who have an array of facets and capacities, any of which might be incorporated into the collaborative process of governance. The public-as-partner model encompasses and extends the public-as-citizen model, acknowledging a civic disposition as one of many motivations that might be active in coproduction. Consequently, it permits a more nuanced assessment of participation. Whereas the public-as-citizen model assumes that a single generalized motivation is uniquely applicable across programs, the public-as-partner model’s more expansive approach anticipates that programs might differ in the specific motivations that they engage, depending on the particular nature of participation each requires. This latter consideration becomes even more valuable as civic tech rapidly diversifies the number of ways that constituents might contribute to the design and delivery of services.

Importantly, the public-as-partner model does not dismiss a civic disposition or other generalized motivations for participation in govern ment programs as irrelevant but simply distinguishes between them and motivations that are more specific to the participation required by a given program. The analysis of 311 presented here clearly illustrates the complementary roles that these two types of motivations might play, suggesting a series of considerations that we might explore for each.

Program-Specific Motivations

The division that emerged between territoriality and civic disposition was geographic in nature, situating the former in the home neighborhood and the latter in other spaces around the city. This would raise the question, though, of how specific a motivation is to any given program. Attending to concerns in one’s home neighborhood certainly is not unique to 311 systems, in which case “program-specific” refers more to the particular form of a program and others that cue similar motivations. For example, community policing, one of the original inspirations for attention to coproduction, depends on the enforcement of social norms and expectations by local residents and, in turn, their territoriality. ⁴⁴ It would stand to reason that the same motivation might be central to participation in any coproduction program that would require individuals to take action surrounding the upkeep, beautification, or defense of their neighborhood.

Coproduction programs are diverse, however, and many are not centered on neighborhood spaces. In rural France, families participating in the Villa Family program act as hosts to elderly tenants as a localized substitute for nursing homes. ⁴⁵ In the United Kingdom, the Sure Start program provides new mothers with in-home consultations with existing mothers. ⁴⁶ In many developing countries, community members are trained to deliver health services in rural areas in order to supplement an otherwise limited formal health sector. ⁴⁷ Meanwhile, civic tech has only expanded the ways that constituents might engage in governance, as illustrated by efforts to involve residents in community development decisions through virtual environments ⁴⁸ and to crowdsource solutions to Boston’s snow-removal problem during the record-breaking winter of 2014–2015. ⁴⁹ Each of these programs relies on its own characteristic set of motivations.

I have listed a disparate set of programs to illustrate how broadly participation in, and thus motivation for, coproduction might vary. That said, it is likely that there are other themes that link the goals and activities across multiple programs, leading them to call on one or more of the same motivations. For example, Ostrom describes how two Nigerian communities had differing success rates with a public school system owing to the value that local parents placed on formal education. ⁵⁰ In parallel, Melissa Marschall found that parents in Detroit, Michigan, with higher education were more engaged in discussions regarding school issues. ⁵¹ We see then how two programs in different cultures can tap similar motivations thanks to a shared focus on childhood education. Short of proposing a full taxonomy of coproduction programs and the motivations that they engage, this simply demonstrates how the public-as-partner model might inform further theory and practice. Program-specific analysis will reveal both distinctions and overlaps in motivations across programs in a manner that will resemble the organization of human behavior, with programs clustering around particular motivations that are most closely related to the major societal challenges that coproduction programs seek to address. In turn, policymakers and practitioners can leverage these consistencies to give greater context to participation in any given program.

Generalized Motivations

Even if there are motivations that are generalized across coproduction programs, it is clear that it is no longer sufficient to assume that they are universally responsible for all forms and levels of participation. Consequently, a new question of interest is how a generalized motivation is relevant to a given program. Here we saw a civic disposition responsible for custodianship that fell geographically beyond the natural bounds of territoriality. This insight might have implications for other place-based programs but would be less helpful for understanding participation in, say, parent-teacher associations or community health care delivery. While a civic disposition may play a role in these latter two as well, it will manifest in a way specific to the program. Similarly, we might think of Alford’s nonmaterial rewards, including solidarity incentives and normative beliefs, as classes of motivations that might take multiple forms, depending on the program in question. ⁵² For example, territoriality is strongly associated with social relationships within the neighborhood. ⁵³ In contrast, Marschall found no evidence that neighborhood social relationships influenced greater involvement around school issues, likely because that particular social context has little bearing on the behavior in question. ⁵⁴ Individuals and communities also differ in their normative beliefs and therefore might vary in how they respond to the values promoted or embodied by a particular program.

Another way to think about generalized motivations, particularly civic disposition, is along the spectrum of constituent participation, from the receipt of services to the planning and design of programs. The 311 system sits midway along this spectrum in that constituents direct the allocation of services but are not involved in the design of the system. To its one side are programs in which the public’s only role is to receive services. It has been argued that these too are a form of coproduction because individuals are not merely passive vessels but must accept and utilize the services if the program is to have any impact. ⁵⁵ Program-specific motivations would seem largely if not uniquely responsible for participation in such cases, as an individual’s desire for services would be rooted less in a civic disposition than in the anticipated benefits, which will clearly vary depending on the nature of the service. Take the example of flu shots, which are important not only for individuals but also for preventing the spread of disease throughout a population. Though this might support an argument that one has a “civic duty” to receive a flu shot, we have already seen multiple cases, from estuaries to streetlight outages, in which such rhetoric is ineffective. Instead, people are more likely to decide whether to receive a flu shot based on whether they believe that it will keep them or their family members healthy while not exposing them to unwanted side effects. ⁵⁶

For programs on the other end of the spectrum, which involve the public in some mixture of planning, design, and implementation, ⁵⁷ a civic disposition is increasingly relevant. Especially when participation requires one to join an organization that meets regularly, sometimes with government officials, this would seem to appeal to the same motivations that were measured as civic activities in the survey presented here—donating, volunteering, advocating, and voting. Nonetheless, these programs necessarily target particular topics, creating situations, such as the one observed in this study, in which a civic disposition will combine with other motivations to drive participation. In the case of 311, territoriality was an initial requirement for being a custodian in the first place, but custodians with a greater civic disposition were more likely to report across the city. This reflects one particular way that motivations synergize in influencing participation—that the program-specific motivation is a prerequisite, enabling a civic disposition to have an impact. As similar research is conducted on participation in other programs, it will become clear whether this is typical or whether there are other ways in which these motivations interact.

Next Steps: Theoretically Informed Innovation

The public-as-partner model transforms our understanding of how coproduction programs operate. This is timely given the rapid proliferation of civic tech, which is currently experimenting with the number of ways in which the public can be incorporated into governance. Practically, the model offers a framework that can inform the organization, implementation, and evaluation of these programs. We might understand this by revising the metaphor that describes coproduction. I noted earlier that traditional rhetoric casts coproduction programs as a “bridge to citizenship” that calls on and bolsters a civic disposition, ⁵⁸ a perspective that is similarly foundational to the philosophical inspiration of civic tech. The public-as-partner model suggests an alternative metaphor, that coproduction acts as a lever translating motivations into civic impacts. Many of these motivations, such as territoriality in the current case, are not inherently civic, yet the program channels them into positive outcomes for the community. As public administrators continue to use coproduction programs, including those considered civic tech, to engage constituents in new and creative ways, they will likely call on more and more behaviors that were not previously part of the governance process. In many cases, modern data collection will also provide a detailed window into the patterns of participation, as they did here. This in turn will both call for and facilitate a case-by-case investigation of these various levers in order to identify the motivations that each is harnessing for the collective good.

Metaphors, of course, have their limits, but thinking about coproduction programs as levers can assist those managing or implementing them to better understand their work. In this way, the current chapter has set the theoretical basis for a practical reorientation of the management of coproduction programs. If we know which motivations a program is leveraging, one might ask whether it is appropriately designed to do so. For example, returning to the lesson of wetlands preservation at the beginning of this chapter, do outreach efforts speak to the constituent’s interests and concerns, or are they falling on deaf ears? Similarly, program evaluation should consider the motivations whose presence or absence is responsible for the observable level of participation. Especially when a program is seeing disparities across demographic or socioeconomic groups, this information might help administrators avoid the perpetuation of existing inequities. ⁵⁹ Chapter 6 builds on these opportunities, assessing the effectiveness of three innovations that sought to expand the reach of Boston’s 311 system. Distinctively, these projects—from conception, to design, to evaluation—were informed by a theoretical understanding of territoriality, civic disposition, and any other motivations that might drive custodianship.