1 INTRODUCTION

While the recent polls suggest that Americans place great importance on democratic values and principles (Pew Research Center, 2018), it is unclear whether their perceptions of public programs are influenced by the extent of public participation. After peaking during the implementation of 1960s war on poverty programs (Greenstone & Peterson, 1976), the efforts to engage the public in governance today are fraught with challenges and often carried out—on a limited scale, particularly in the field of social welfare (Amirkhanyan & Lambright, 2017). The worldwide New Public Management reforms emphasized, among other things, accountability for results and greater managerial autonomy, with attention paid to the process, including public participation. While it is a given that the public demands effective performance, individuals might in fact also expect democratic participation in the provision of public services. The first objective of this article is to explore if individual perceptions of public programs and their performance are influenced by the observed level of public participation.

Public voice and engagement are not exclusive to the programs delivered by government agencies. Public programs, implemented on the ground by nonprofit organizations and for-profit businesses, also frequently engage service recipients and others (Amirkhanyan & Lambright, 2017). Some studies suggest that individuals hold biased views against government organizations giving them less credit for their work compared to private entities (e.g., Baekgaard & Serritzlew, 2016, 2020; Charbonneau & Van Ryzin, 2015; Hvidman, 2019; Hvidman & Andersen, 2016; James & Van Ryzin, 2017; Marvel, 2015, 2016; but see Meier et al., 2019; Meier et al., 2022). With most empirical studies situated in the fields of health care and postal service, the policy context might be an important factor in how people view public–private differences and think about performance. The study's second objective is to contribute to the growing anti-government bias literature by examining public views of government, nonprofit, and for-profit organizations in the context of food assistance programs for low-income individuals. By moving beyond the more sector-neutral fee-for-service contexts used in the research conducted to date, the study adds the public assistance context to this literature.

Finally, we investigate how performance information shapes public views on organizations. Prior research has highlighted the role of performance data in shaping public perceptions of organizations (James et al., 2020). Building upon this literature, we ask if the public gives fair credit to organizations with different levels of performance information coming from a credible source. Furthermore, we also investigate whether the effect of democratic participation on public perceptions of government programs is altered by clear and unambiguous (positive or negative) performance information about these organizations.

We use a randomized survey experiment among 1000 US residents exposed to a vignette about a hypothetical local organization working on addressing the problem of food insecurity in a community. A global concern (United Nations, 2019), food security presents an insightful case for empirical examination in the United States: an estimated 11.1% of households suffered from food insecurity in 2018 (USDA, 2019). These households often rely on a combination of federal and charitable food assistance, and the impact of food insecurity is compounded by many factors, such as serious health concerns (Amirkhanyan, Jacknowitz, et al., 2019). The legal ownership of hospitals, nursing homes, or postal service providers that served as contexts for the previous sector-bias studies is somewhat less likely to be the focal point of individual choices and attitudes toward these services, at least in the US context. By contrast, food assistance, encompassing the federal Supplemental Nutrition Assistance Program (also known as food stamps) and charitable food assistance programs, being part of the safety net system, may be viewed as more inherently governmental.

This study makes important theoretical and practical contributions to the literature on public participation, anti-government bias, and performance assessment. First, this study examines whether public participation in service planning and implementation influences individual evaluations of organizations, a novel contribution to the literature on public participation, especially in the context of food assistance. Second, this study investigates whether the public is negatively biased against government organizations in the context of “safety net” programs delivered by local organizations (either local governments, nonprofits, or for-profit organizations). By focusing on the context in which public assistance is provided to help address one of society's “wicked problems,” this study advances the knowledge of anti-government views beyond the more sector-neutral, fee-for-service contexts that have been used in the literature to date. Third, we take performance information into account when studying individual perceptions of organizations and public participation. This setting allows us to consider the role of procedures (i.e., participatory arrangements) and outcomes (i.e., performance data) simultaneously.

2 FROM ATTITUDES TO ACTIONS: PUBLIC PARTICIPATION IN GOVERNANCE

Democracies provide fundamental rights for individuals to participate in governance through voting, assembly, petitioning, and other mechanisms of engagement (Pérez, 2009). Deliberative democracy is “nonnegotiable”: it is a key mechanism to avoid domination in a society and to facilitate the distribution of income and wealth (Shapiro, 2003). The many forms of public participation can help identify and understand public values and ensure that they are reflected in public value—what government creates on behalf of the public (Nabatchi, 2012; Schiff et al., 2022). Yet, in the United States (the locus of this study), many individuals are disconnected from each other, being increasingly less likely to vote, sign petitions, and hold memberships in nonprofit associations (Berry, 2005; Putnam, 2000). Bok (2001) describes this “troubling paradox”: while wanting to control their government, Americans are devoting less time to trying to influence it. Meanwhile, greater public participation, especially among those of lower socioeconomic status, is an important strategy to guide policy decisions and help improve government performance (Bok, 2001). Unsatisfactory patterns of civic engagement may exacerbate the government-citizen divide, particularly because those more supportive of government and its programs have lower rates of engagement than those more hostile toward government, and hence, the “loudest voice” belongs to those who, while benefiting, give less credit to government (Mettler, 2018).

Democracy depends on individual participation as much as it depends on its formal institutions. In a democracy, citizens and other individuals can influence public policies and their implementation by staying informed about major public policy issues, commenting on them, serving on advisory boards, taking part in protests, debating during public forums, filling out citizen or client surveys, and many other mechanisms of active political and civic life (Amirkhanyan, Cheon, et al., 2019; Amirkhanyan, Jacknowitz, et al., 2019; Amirkhanyan & Lambright, 2017). Public participationⁱ can potentially enhance public trust, empower individuals, build public support for government programs, and augment the performance of public services (Amirkhanyan, Cheon, et al., 2019; Amirkhanyan, Jacknowitz, et al., 2019). An active public can be particularly important in an era of contracting and network governance (Bryson et al., 2014; Denhardt & Denhardt, 2015).

Public participation takes different forms, ranging from public meetings and advisory boards to various data collection tools such as surveys and interviews (Amirkhanyan & Lambright, 2017). Participation also varies in terms of intensity—“the extent to which participation efforts provide citizens with meaningful opportunities to engage in decision-making or service delivery in ways that influence policy or programs” (Amirkhanyan & Lambright, 2017, p. 43). Among the numerous taxonomies, Arnstein (1969) classified participation modes into eight levels based on the extent of power sharing between public administrators and the general public: manipulation, therapy, informing, consultation, placation, partnership, delegated power, and citizen control. In a similar vein, Fung (2006) proposed the “democratic cube” framework to classify citizen participation efforts based on participants' inclusiveness, communication mechanisms, and empowerment. Generally speaking, participation mechanisms of higher intensity usually involve two-way communication between citizens and public managers, and citizens tend to share more power and are able to exercise more impact on policy processes and outcomes (Arnstein, 1969).

Empirical research on public participation explores many subjects, including the scope, forms, and intensity of participation, managerial motives and challenges in engaging people, and various impacts of public participation. Much of this research is conducted from the perspective of public managers who are responsible for creating and implementing opportunities for participation (Bryson et al., 2014; Denhardt & Denhardt, 2015; Meijer, 2011; Nabatchi & Mergel, 2010; Nalbandian, 2008). Among the key questions is the discussion of the many perceptions and conceptualizations that public administration practitioners and scholars hold about the general public: people as subjects, voters, clients, interest group advocates, customers/clients, volunteers/co-producers, partners/owners, co-learners, or “active citizens” (Frederickson, 1980; Roberts, 2004; Vigoda, 2002). What is unclear is whether ordinary individuals value democratic participation in public service delivery and whether participation factors into their perceptions of how well government does its job.

This study pursues recent calls for more multidimensional assessments of public programs (McDonald III et al., 2022) by examining how participation opportunities affect public views on several dimensions of government performance: efficiency, effectiveness, equity, red tape, as well as respondents' overall comfort and overall approval with the way services were implemented. Public participation—comments, complaints, advising, co-production, and other strategies—is likely to help better align the priorities, parameters, and benefits of public policies and programs with public preferences (Amirkhanyan & Lambright, 2017; Heikkila & Isett, 2007; Mingus & Zhu, 2018). Participation can be instrumental in avoiding situations when the government and the private sector fail to provide goods and services that reflect public values—what Bozeman (2007) characterizes as public values failure. Some prior research suggests that public participation is associated with positive views on public service delivery. Ruder and Woods (2020), for example, find that information about participatory and transparent procedures is associated with increased perceptions of fairness among respondents (see also Pedersen et al., 2017). Based on this, we expect that participation has a positive effect on public views about both the quality and the equity dimensions of performance. Additionally, when a program involves public input, individuals may feel more comfortable with the way service is delivered and thus are more likely to approve the program. We hypothesize that individuals rate the performance of participatory programs higher, as these programs would be perceived as having better value alignment, better service quality, and greater accountability, transparency, and oversight.

While seeing the benefits of public participation, people might also acknowledge that taking the time to engage the public, solicit their input, and come to a consensus may undermine public program efficiency. Engaging the public requires resources for organizing and implementing participation opportunities.ⁱⁱ The process of designing, planning, and implementing collaborative processes can slow down policymaking and administration (Lawrence & Deagen, 2001). Public values pluralism—the existence of diverse fundamental preferences and commitments in a society—is linked to the need to reconcile these values (Nabatchi, 2012). The reconciliation process can include both cooperative and adversarial elements. In both cases, efforts to reach an agreement among stakeholders with heterogeneous perspectives and preferences can lead to inefficiency in the policy process, conflicts among stakeholders, and suboptimal policy decisions (Irvin & Stansbury, 2004). Thus, we hypothesize that respondents acknowledge the costs of participation and perceive participation as undermining public programs' efficiency. Hence, we propose the following hypotheses:

3 ANTI-GOVERNMENT BIAS IN SAFETY NET PROGRAMS

Although there is literature about anti-government bias in several countries (e.g., Mexico [Sandoval-Ballesteros, 2013], the Netherlands [van den Bekerom et al., 2021], Italy [Belle et al., 2021], Denmark [Baekgaard & Serritzlew, 2016], Spain [Casado-Aranda et al., 2022], and South Asia [Baniamin & Jamil, 2023]), the United States appears to be an outlier in terms of the salience of the issue. The US Constitution protects public rights to criticize government agencies and public officials, and Americans have amply used this “prized American privilege to speak one's mind” (Aron, 2012, p. 57). In contrast to many nations where public service has a more dignified reputation, the practice of “bureaucrat bashing” has long historical roots in the United States (Goodsell, 2018). The neutral and technical term “bureaucrat” has acquired negative connotations suggesting laziness, incompetence, rigidity, wastefulness, and thirst for power. Criticizing government bureaucrats became a part of the US political culture and a focal element of presidential campaigns beginning at least with President Carter (Goodsell, 2018).

The ordinary public views of government institutions have mirrored these developments. Since the 1960s, the percentage of Americans considering “big government” the biggest threat facing the country has always surpassed the percentage of those considering “big business” or “big labor” as the biggest threat (Gallup, 2017). Among other factors,ⁱⁱⁱ limited social protections for those in need, ineffective regulations, and poorly designed public programs may have influenced the deterioration of public attitudes toward the government (Bok, 2001). In 2019, when the current survey experiment was conducted, the public trust in government was near historic lows: 17% (with only 9% among African Americans) (Pew Research Center, 2018). An important caveat in these trends is that public attitudes toward government in general are somewhat distinct from their views on specific public services (Christensen & Lægreid, 2005). For instance, in the context of federal education policy, Americans may dislike government in the abstract, but they in fact approve the majority of specific actions it takes, such as federal assistance to schools (Davies, 2008). Hence, public perceptions of government are complex and not uniform.

In contrast to this negative treatment of the public sector, the field of public administration research finds evidence to support a more optimistic view. While documenting the many bureaucratic pathologies, ample empirical evidence finds excellent leadership, management, innovation, and performance in the public sector (Rainey & Steinbauer, 1999). Relatedly, reviews of privatization literature fail to demonstrate clear-cut evidence of the positive effects of transferring public ownership to for-profit and nonprofit institutions in terms of costs, quality, or social equity (Hodge, 2000). Although increasingly disconnected from their government, Americans are, paradoxically, depending more on government programs such as food stamps, social security, Medicare, home mortgage interest deductions, and others—with 96% of adults receiving benefits from at least one of them (Mettler, 2018). Nonetheless, while there is evidence that public organizations fare well compared to private organizations and play an important role in most people's lives, the discrepancy between the “a priori” views of government and the empirical knowledge persists, and the government is still largely viewed more as a liability than an asset (Rainey & Bozeman, 2000; Rainey & Steinbauer, 1999).

The negative views of the public sector are not inconsequential. The hostile atmosphere can lower recruitment, retention, and morale in government organizations (Marvel, 2015). Negative views can also discourage individuals from cooperating with government policies and programs, paying taxes, or following government regulations (Christensen & Lægreid, 2005; Hetherington, 2005; Marvel, 2015). Surveys affirm these expectations: 64% of Americans believe that low trust in government makes it harder to address the nation's persistent problems (Rainie & Perrin, 2019).

A growing body of behavioral public administration research has emerged on public anti-government bias—a term used to describe individual prejudice against government institutions, a propensity to discount positive information about them, and related choices and assessments. Marvel (2015) found that Americans ascribed lower performance to the US Postal Service compared to its private counterparts, and these perceptions were not changed with positive information. Hvidman and Andersen (2016) found that public hospitals in Denmark were perceived more negatively than private hospitals in terms of efficiency, red tape, and effectiveness, while benevolence was in fact perceived to be higher in the public sector. Additionally, Hvidman (2019) examined in-home elderly care and found that, conditional on respondents' beliefs about government, public services were perceived to be less effective. In contrast, Meier et al. (2019) found no evidence of bias against government hospitals for any measure of performance. Similarly, Meier et al. (2022) found a modest anti-for-profit sector bias in US nursing home care (nonprofits were in fact perceived most positively); however, sector biases disappeared when objective performance information was presented.^iv

Given these mixed findings in studies conducted across service fields, the context in which public biases may be more or less pronounced seems important. This is especially the case because governance in action relies on the delivery of public services via nonprofits and for-profit organizations as well as traditional government bureaucracies (McDonald III et al., 2022). Past studies on sector bias, in addition, have focused on services provided to the general population: almost everyone at some point uses one or more forms of postal services, hospitals, and long-term care (Hvidman & Andersen, 2016; Marvel, 2015; Meier et al., 2022). Experiments conducted in the United States, in particular, are focused on services where quality, timeliness, and customer service might outweigh sector-related considerations. In these policy fields, individuals may be particularly unaware of or genuinely uninterested in the provider's sector, as has been shown in prior research (Van Slyke & Roch, 2004). The current study seeks to address this gap: we ask if individuals are biased against government organizations in the context of the safety net programs targeting people with lower socioeconomic status and delivered by public, nonprofit, or for-profit providers; and whether they give credit to organizations irrespective of their sector.

One key question concerning the distribution of public services revolves around the potential trade-offs between equity and other aspects of performance, such as effectiveness and efficiency (Keeley, 1978; Le Grand, 1982). Previous research shows that the general public effectively distinguishes between program effectiveness and equity (Meier et al., 2023). Thus, our study investigates whether the organizational sector influences public perceptions of multiple dimensions of performance—efficiency, effectiveness, equity, red tape, and their overall comfort and approval of programs. Consistent with past studies, we first expect respondents to perceive government organizations as less effective and less efficient compared to private nonprofit and for-profit organizations due to a lack of competitive market pressures (see Hvidman & Andersen, 2016; Marvel, 2016). Meanwhile, public and nonprofit organizations' mission orientation and the lack of profit-maximization motive will likely result in higher ratings on fairness and equity. Prior research also shows that public organizations are often perceived as less effective but more equitable compared to private organizations (see Hvidman, 2019). This supports the idea that the performance ratings of public organizations may be influenced individual views on the potential trade-offs between effectiveness and equity.

In addition to perceptions of equity, individuals may also find it more “logical” or “conventional” for the food assistance programs to be delivered by public and nonprofit organizations. Food assistance is likely to be perceived by the public as an essential “safety net” program designed to aid vulnerable populations. Respondents may, therefore, consider public and nonprofit organizations as more suitable for stepping in and addressing market failure. Prior research in elder care also indicates that people tend to feel more comfortable when using nursing home services provided by nonprofit rather than for-profit facilities (Meier et al., 2022).

Additionally, we hypothesize that respondents are likely to attribute higher levels of red tape to government programs. Government organizations are historically more constrained than private entities by mandates, rules, regulations and administrative burdens affecting both the citizens and the employees (Herd & Moynihan, 2018; Rainey & Bozeman, 2000). We anticipate that people hold prior perceptions of this distinction and will, therefore, associate government ownership with burdensome administrative rules and procedures. The perception of extensive red tape is also consistent with the perception of low productivity and inefficiency in the public sector, as discussed in the literature on anti-public sector bias (Hvidman, 2019).

A related question is whether public participation affects people's views of organizations of all sectors equally or whether its impact varies across public, nonprofit, and for-profit organizations. This is a novel topic in the empirical literature and, therefore, our hypothesis is exploratory and not comprehensively informed by prior research. On the one hand, participation is an essential part of the democratic process, suggesting it should be linked to the implementation of government agencies, somewhat less for nonprofit organizations, and very little for private for-profit organizations (Amirkhanyan & Lambright, 2017). If so, the effects of participation would be greater in the public and nonprofit sectors. On the other hand, information from the external environment (customers, clients, and the general public) and ongoing relationships with these groups could well be helpful in designing programs to meet the needs of clientele, which would suggest that participation would matter for all organizations regardless of sector (Merlo et al., 2014). In the case of food assistance, the latter would be particularly compelling because clients tend to have urgent needs and expect instant help; when public participation helps successful service delivery, its positive effect on public perceptions should outweigh any sector bias (Amirkhanyan & Lambright, 2017). Furthermore, in the vignette we use, food assistance is a publicly financed service implemented across three sectors, and people would likely extend their concept of publicness and its attendant role of democratic engagement to public and private service providers. Furthermore, private organizations will likely be viewed as accountable to the public funding agencies for their responsiveness to the community's needs and would be expected to engage citizens. Therefore, we hypothesize that participation matters for the organizations regardless of their ownership status.

4 THE ROLE OF OBJECTIVE PERFORMANCE INFORMATION

For performance systems to be effective, they need to be understood by users and provide clear comparative information that helps differentiate between high and low levels of performance. This study examines how clearly described performance information combines with participation to influence public evaluation of service performance. Prior research focusing on the effect of objective performance measures on individual perceptions has shown mixed findings: some studies show that positive performance information improves public perceptual evaluations (see Meier et al., 2022), while others suggest that it does not matter much (Marvel, 2015). The latter findings might be partly because provided performance information is often ambiguous or context-specific (Baekgaard & Serritzlew, 2016; Marvel, 2015) or because individuals see performance information as less reliable (Mizrahi & Minchuk, 2020). Thus, we focus on less ambiguous performance information and explore how comparable data shapes public perceptions of performance in the context of food assistance received by clients with low socioeconomic status.

We first argue that positive unambiguous performance information improves public perceptions of effectiveness, efficiency, equity, comfort, and approval and lowers perceived red rape. Recent experimental studies show that a clear performance index, such as five-star quality rating, significantly improves public perceived effectiveness, efficiency, equity, and comfort, in the context of elder care (see Meier et al., 2022). The inclusion of equity is of special importance given its increased role in public administration (McDonald III et al., 2022) and recent experimental work that shows the public responds directly to equity concerns and is unwilling to tradeoff equity for increases in efficiency or effectiveness (Meier et al., 2023).

Additionally, we expect that public participation can counteract information on poor performance. Respondents' exposure to public participation—a deliberative decision-making process involving numerous entities—might imply that the organization is not the only entity responsible for substandard performance. If other involved parties may share the collective responsibility for the outcomes, public evaluations of underperforming organizations might be more lenient. Participation may also be perceived as good in and of itself or as a signal of collective investment in the program that could lead to better outcomes. An extensive participatory process might also be interpreted as an effective management process. By contrast, in the absence of public participation, “clarity of responsibility” for policymaking and implementation is high: that is, people are more likely to blame service providers for poor performance (Marvel & Girth, 2016) and evaluate them harshly. Our hypotheses are as follows:

5 DATA AND METHODS

6 FINDINGS

Table 2 shows that respondents exposed to the participation cue rated performance significantly higher in terms of efficiency, effectiveness, and equity (about one-third of a standard deviation) and slightly lower in terms of red tape. These respondents also were more comfortable with the way the program was implemented (one-half standard deviation) and more likely to approve of the program (about four-tenths of a standard deviation) than those who did not receive the participation cue. Although one might expect that a participation process might affect some dimensions of performance more than others, the relative size of all the coefficients was in fact within a narrow range. These results support Hypothesis 1 on effectiveness, equity, comfort, approval, and red tape (in the latter case, the relationship was in the opposite direction than hypothesized) and reject Hypothesis 2—on efficiency.

Given that several past studies suggest that American respondents are biased against government organizations (see Marvel, 2015, 2016; but see Meier et al., 2019), we examined whether sector biases exist in the context of the means-tested program, how large they are, and whether they change the effect of participation on perceived performance. In Table 3, we regress the sector variable on the six dependent variables measuring performance assessments. For-profit organizations are the excluded category coded as zero for the sector variables so that the coefficients for government and nonprofit organizations indicate how they are different from for-profit organizations. For all six measures—efficiency, effectiveness, equity, red tape, comfort, and approval—the coefficients are statistically insignificant. In effect, this result suggests that people do not evaluate public, nonprofit, and for-profit organizations differently and are essentially neutral toward organizational ownership (factor scores have a mean of zero, and, hence, the best estimate of the impact for all three types of organizations contains zero well within the range of error). Similarly, there are no differences in perceived comfort and approval based on the sector of the organization providing the services. Therefore, Hypotheses 3 and 4 are rejected: we find no evidence of sector bias in the examined policy context.

Given that participation has a positive effect on the evaluation of organizations, our next question is whether organizational ownership matters for the perceived benefits of public participation. We hypothesized that participation could be helpful in designing and implementing programs to meet the needs of clientele irrespective of ownership; therefore it would positively contribute to public perceptions of all organizations regardless of sector. In this case, the interaction between participation and the organizational sector would be insignificant. Table 4 shows the results that are consistent with our expectations. With the exception of red tape, participation improves respondents' perceptions of all aspects of performance, but none of the sector interactions rise to the level of statistical significance. In short, participation appears to improve public views of services irrespective of the legal ownership of the organization implementing these services (Hypothesis 5 supported).

Next, we turn our attention to the role of performance information in people's assessments. For performance systems to be effective, they need to be understood by users and to provide clear comparative information that helps differentiate among levels of performance. A wide variety of organizations use five-star evaluation systems (with five being the highest), including the official US government rating systems for hospitals and nursing homes, with similar systems in place for many public schools. This study employs five-star rating as objective performance information. Table 5 shows that assessments of fictitious food security organizations are highly sensitive to ratings, with an increase of one star moving respondents' evaluations up by about one-third (equity and red tape) to about two-thirds (the other four indicators) of a standard deviation.

Next, in our analysis, we explore if an organization's star rating moderates the effect of participation on public assessments of the food security program. Table 6 shows that, with the exception of red tape, virtually all coefficients in the table are statistically significant; however, the negative interaction between participation and the star rating makes the interpretation a bit more involved. To illustrate, in the efficiency regression, to calculate the rate for star rating and participation organizations, one needs to consider the coefficients for participation (1.042), the coefficient for star rating (0.760), the coefficient for participation interacting with star rating (−0.243), and the intercept (−2.427). For an organization with a four-star rating but one that does not engage public participation, the estimate is 0.613; meanwhile, for a four-star agency that does engage public, the estimate is 0.683. On the other hand, for a two-star agency that does not engage the public the estimate is −0.907; meanwhile, the estimate goes up to −0.351 for a two-stars agency that does involve participation. For all cases, the worst performance ratings occur in the case of two stars and no participation, and the best ratings—in the case of four stars and participation. Participation's impact appears to improve the perceptions of any agency whether it is poor performing or doing well; it cannot turn a poor performance into a positive, but it can mitigate the negatives.

7 DISCUSSION AND CONCLUSION

Public attitudes and choices related to government are important for a well-functioning democracy where active citizens are expected and encouraged to cooperate with, coproduce, and hold government officials accountable for public policies and programs. The public administration literature suggests that not all public services are alike in terms of how their outcomes are produced, experienced, measured, and observed (Resh & Cho, 2019; Wilson, 1989). Accordingly, public perceptions of and biases toward government and government programs may not be uniform. In this study, we explore public views related to public participation—a fundamental element of democratic governance. Building upon the prior research suggesting that public participation is associated with improved service outcomes (Amirkhanyan, Cheon, et al., 2019; Amirkhanyan, Jacknowitz, et al., 2019), this article explores if public assessments of organizations are influenced by the presence of public participation in program implementation.

Our survey experiment shows that people have more confidence in the effectiveness, efficiency, equity, and other aspects of organizational performance when they are told that the public has an active role in designing, implementing, and governing the program. The scenario we presented includes numerous forms of public participation: advisory boards, deliberative meetings, public hearings, and incorporation of public feedback. Public participation appears to matter regardless of organizational ownership. One might expect higher levels of participation expectations when it comes to government entities, but our respondents prefer that all publicly financed services—irrespective of the providers' sector—involved public feedback. Clearly, US respondents value public participation as a meaningful contribution to decision-making on how their tax dollars are spent by providers across sector lines.

This study also seeks to contribute to our knowledge of anti-government attitudes by focusing on the field of food assistance provided at the local level and its interactive effects with participation. Largely subsidized by the federal government but delivered by local government agencies, charitable organizations and for-profit organizations, food assistance has been designed and is likely perceived by the public as an inherently governmental “safety-net” program aiming at clients with markedly low socioeconomic status. As opposed to health care or mail services, frequently used by the general public and discussed in terms of their quality, SNAP is rarely publicly debated in terms of its quality. Instead, the discussion is more likely to focus on the political choices and the extent of subsidizing the program. Anti-government bias in this context may be tied to people's perceptions of the deservedness of various marginalized social groups who use public safety-net programs (a direction we explore through our robustness checks by examining the role of political views).

Our analysis suggests no sector bias among US respondents: people are not prejudiced more against government providers of food assistance than their private counterparts. Similar to other organizations, public organizations are assessed positively when performance data are high and are evaluated poorly when the performance data are low. The absence of bias and attention to performance information are essential in a democracy because the public is expected to make informed decisions, and such knowledge makes “consent of the governed” meaningful. Despite bureaucrat bashing in the media and lack of trust toward government in the abstract, there is no evidence of respondents preferring private organizations to public ones in delivery programs to combat food insecurity. This suggests that anti-government bias might be linked to the level of governance and its impact: the public might hold strikingly different views toward an abstract national government located and operating away from their communities, as opposed to local agencies and programs aiming at solving their local problems.

Whether the organization delivering services is government, nonprofit, or for-profit has no impact on support for public participation. Public participation has a positive effect regardless of the organizational sector. In many policy fields, public, nonprofit, and for-profit organizations deliver services side-by-side, and public and nonprofit actors behave much like their for-profit counterparts. While consuming these services, individuals may be unaware of organizational ownership. Furthermore, the nature of the bias in services frequently consumed by the general population may differ from those in a traditionally publicly subsidized field that serves marginalized social groups. The latter context may also lack robust multi-sector markets: a community is likely to have only a few homeless shelters or a limited number of charitable food pantries; therefore, people may have no choice of providers as they would in the context of heath or postal services. Individuals in this context may care less about the organizational sector and pay more attention to procedural arrangements and service outcomes.

Finally, attention to the role of performance data is critical: the public does not discount success in public organizations and gives them as much credit for it as private entities. Efforts of public organizations to measure and report performance, therefore, are likely to be useful for people to draw conclusions and make political choices. While performance information appears to matter more for individuals than participatory procedure, participation can help mitigate the effect of negative performance information.

While this study contributes to the literature on public participation, sector bias, and performance assessment, it also has limitations. First, our design involved fictitious organizations operating in a fictitious local area (county). Thus, our respondents' knowledge of the organizations would be limited to what they are told in the experiment's vignette. With an identical description of the organizations and randomized ownership, we can be sure to attribute differences in perceptions to the sector because our respondents cannot draw on other knowledge or prior experience with this organization. There is also likely to be greater trust in local governments than in the more distant national government, and this might have contributed to the null results (experiments that focus on different levels of government would be valuable). Second, while public participation has various forms (e.g., direct vs. indirect) and intensity (high vs. low), we did not manipulate the different types of participation. Future research could incorporate various types of participation and elaborate further on the effect of participation on public views.

Additionally, this experiment was focused on a single policy area in the US context and may have limited generalizability to other contexts and nations. At the same time, the characteristics of the experiment are likely to be relatively common in some service areas and in certain other countries. Food insecurity is a social service administered at the local level and such programs operate in most if not all countries around the world; the choice of public, nonprofit, or for-profit organizations, however, is likely to be constrained in many countries either by a lack of government capacity or by the lack of robust nonprofit or for-profit sectors. In developing countries, many food programs are in fact delivered by international nonprofit organizations (Uvin, 1995) because existing governments lack capacity or are plagued with corruption (BouChabke & Haddad, 2021). Countries also vary in their nonprofit capacity and density (Kim & Kim, 2018) and the size of the for-profit sector as the result of government size and constraints. Individuals are most likely to have strong views about service implementation processes that are familiar to them. Additional work in other country contexts is needed to establish the generalizability of the current findings. Furthermore, whether participation and its role and impact might vary across countries and programs especially needs further research. We framed our hypotheses on participation within the literature on democracy and tested them in an established Western democracy. To the extent that countries are less democratic, rely on private providers in the delivery of safety-net programs or have greater levels of participation than the United States could also affect the generalizability of our results.

A single experiment can only be suggestive, but the implications for democratic accountability are clear and should be pursued in further research. In our experiment, individuals responded to both process (participation) and outcomes (performance); in fact, they were willing to forgive some, but not all, of poor performance if the process was participatory. Left unexplored is how individuals might perceive performance given a flawed process that discouraged participation (rather than providing no information) or that lacked transparency. While the respondents cared about the quality of services and the process, they did not care about who delivered the services. The experiment showed no sector biases in the evaluation of either outcomes or processes similar to US studies in health care and elder care, thus, raising the question of whether postal services are an anomaly in this regard. Further studies are clearly merited in the context of other public services. Finally, the existence or lack of a sector bias, while important, is not the same thing as whether individuals might prefer services to be delivered by public, nonprofit, or for-profit organizations. That question which also has implications for democratic accountability remains to be explored.

CONFLICT OF INTEREST STATEMENT

The authors declare no conflict of interest.