Abstract
A significant amount of literature has inspected the relationship between public–private partnerships and state and local government. This literature has focused primarily on how these agreements shape financing, economic development, and public policy measures. There is little research, however, on how improvement districts may affect political participation. There are many reasons to believe that these districts may raise levels of political participation, as they deeply affect state and local politics and shape the socioeconomic development of local communities. This article fills this gap in the literature by exploring the relationship between the establishment of local improvement districts and voter participation rates. An original data set is constructed from 18 state assembly districts and 22 local improvement districts in New York City across nine elections between 2002 and 2018, resulting in 198 unique observations across time. Empirical results reflect how the development of improvement districts can serve as signals for rising political participation in surrounding areas, marked by increasing rates of voter turnout across midterm and presidential-year election cycles. These findings are compelling, providing insight into how local organizations designed and sustained through issue ownership and community collaboration have the ability to raise political participation through electoral activity.
Keywords
Establishing consistency in high levels of civic engagement and democratic engagement has not only been shown to be integral to political life, but it is a normatively appealing benchmark as well. Democratic citizenship is predicated on the belief that creating the space for a socially and politically active populace strengthens well-being and representation for communities. Previous research has shown the issues posed to local government when a weaker civic culture exists, as measured by political participation, associational life, and interpersonal trust (Andrews, 2007).
Civic engagement encompasses the ways in which individuals connect with others to create regular social interactions motivated by shared membership in a given group or organization (Manturuk et al., 2012). Research in civic engagement has historically assessed the basis for which social groups form to interact with matters of public life, and how developments of social and political change affect democratic government. It has been argued that there are potentially negative consequences presented by higher levels of civic engagement (Fiorina, 1999); however, the sociopolitical aspects of robust standards of democratic citizenship and social involvement are widely determined to improve government responsiveness and political representation.
Equally integral to the maintenance of stable, functioning communities is the administration of government services. An increasingly prevalent method through which local governments are administering community services are through public–private partnerships (PPPs). PPPs are organizational systems in which public actors and private actors work together to achieve a common goal, or set of goals, through shared resources (Koppenjan, 2005). The ultimate aim of these partnerships is to provide beneficial public services by sharing risk with private actors and reducing overall costs (Bloomfield, 2006). The term PPP is a broad term, however, and encompasses a wide, diverse array of partnerships with unique focuses (Wettenhall, 2003).
One of the most prominent forms of PPPs are “improvement districts,” which are commonly referred to as community improvement districts (CIDs) or business improvement districts (BIDs). Improvement districts are community organizations developed with the authorization of state and local government authorities to oversee the maintenance and improvement of social, physical, and economic standards in the area. Since 1970, the number of these districts has risen to approximately 800 in the United States (Morçöl et al., 2017), with similar improvement district systems found in cities in other countries including Canada, Germany, England, Sweden, Australia, and South Africa (Peyroux et al., 2012). These organizations engage in a range of activities from organizing community events, social gatherings, and service delivery to local neighborhoods.
Research surrounding improvement districts, and PPPs broadly, has focused on the effectiveness and equitability of service delivery provided by these partnerships. No research to this point, however, has explicitly sought to assess the relationship between improvement districts and civic engagement in communities in which these districts are established. As a result, scholars and practitioners are aware of laws, policies, and institutional mechanisms that affect and respond to improvement districts, yet have little to no insight into how these districts affect and respond to community residents. This article seeks to help fill this significant gap in scholarly and practical knowledge.
There are multiple reasons why we would theoretically expect that the establishment of a local improvement district could raise levels of democratic participation and civic engagement among citizens. First, the creation and maintenance of these districts requires collaboration between state and local government and community groups and residents. These collaborative activities provide more opportunities for residents to see their effect on public affairs. Second, these improvement districts allow government officials to become engaged and involved with community residents and other organizations to a greater extent. This connection may raise resident awareness of elected officials and community interests. Finally, these districts increase funds targeted at improving physical, social, and economic standards for community residents and property owners, which would be expected to raise the likelihood that residents would participate in civic life and elections.
The hypothesis that improvement districts will have a relationship with democratic participation is tested empirically in this article by constructing both multivariate ordinary least squares (OLS) regression models and fixed effects difference-in-difference models based around an original data set, which is composed of 18 state assembly districts. These assembly districts are observed across nine elections from 2002 to 2018, and encompass approximately 1.3 million active, registered voters. Findings from these models reflect significant effects on voter participation in these state assembly districts after an improvement district was created and maintained in the area. This allows for a total of 198 unique observations across time in the study.
These findings give new insight into the role of local improvement districts in relation to democratic participation and provide a path for future research to study the role of PPPs in understanding civic engagement and democratic citizenship. Specifically, they show how local organizations designed and sustained through issue ownership and community collaboration have the ability to raise political participation through electoral activity. This reflects the role of local organizations in fostering participation in civic and political life.
Political Participation and Civic Engagement
The study of civic engagement in political life encompasses a wide array of theoretical and methodological approaches to understand social interaction with public institutions and organizations (Nabatchi and Amsler, 2014). An exact definition of civic engagement has been historically difficult to pin down; however, we typically think of civic engagement as the ways in which citizens of a given society interact with public institutions and social organizations (Adler and Goggin, 2005). Democratic theory typically conceives that greater levels of civic engagement are related to government that is more stable, representative, and efficient (Andrews, 2007). The significance of the citizens’ relationship with their political community has certainly been at the center of academic inspection since humans began to coalesce into self-governed polities; however, much of the current research concerning civic engagement research began during the closing decade of the twentieth century.
In 1993, Robert Putnam’s “Making Democracy Work: Civic Traditions in Modern Italy” set off a subsequent line of inquiry into how citizens connect and collaborate in the state of contemporary public life (Putnam, 1993). Putnam’s (1995) work set a theoretical foundation for an issue which others clearly recognized to be meaningful and understudied, and was followed up by an abbreviated, but equally influential essay “Bowling Alone: America’s Declining Social Capital.” Putnam lamented the decline of Americans’ involvement in social organizations, trust in government institutions, and electoral participation in the face of a changing demographic and economic structure in the latter half of the twentieth century. As a result of these changes in civic life, Putnam contended that there was a serious need for further inspection into social connectedness and civil affairs.
Articles soon followed, which provided further insight into the historical role of civic culture and engagement in American political affairs (Rice and Sumberg, 1997; Skocpol, 1997). During this time, a number of shorter articles also began to inspect more focused topics within the area of civic engagement, including the role of corporations in social networks (Heying, 1997), media and interpersonal trust (Shah, 1998), and the relationship between economic cooperation and social capital (Kenworthy, 1997).
The concept of “social capital” was a significant focus of early civic engagement scholarship and has continued to be at the center of research in the area. Social capital refers to “connections among individuals—social networks and the norms of reciprocity and trustworthiness that arise from them” (Putnam, 2000). While this is, in some ways, an amorphous definition, it allows scholars to assess civic engagement through analysis of public participation among citizens and government officials. Previous research has explored aspects of civic life and social capital through studies of local government performance in relation to citizen perception (Rice, 2001), trust in government and national government performance (Keele, 2007), and technological effects on civic engagement (Mandarano et al., 2010).
More recent studies on civic engagement have taken a variety of angles in analyzing civic engagement and social capital in relation to government and politics. Studies on the role of incarceration and democratic participation have been important in shining new light on the ways in which the criminal justice system can have influential effects on citizens’ engagement in civic affairs (Lee et al., 2014; Weaver and Lerman, 2010). Others have looked at how social capital plays a role in shaping economic development of communities (Doh, 2014) and how religion impacts social networks and civic engagement patterns (Lewis et al., 2013). Some studies have also sought to engage with civic engagement among youth (Wilkenfeld et al., 2010), and how social media platforms shape the way citizens engage with local government (Haro-de-Rosario et al., 2018). These studies reflect the breadth and diversity present in the pathways through which citizens engage in public affairs and political processes.
PPPs and Local Improvement Districts
The study of PPPs has historically focused on how state and local government can share costs with private organizations to provide mutual benefit for both public and private actors, as well as citizens. PPPs encompass a wide array of collaborative services between government agencies and private groups, which can develop financing and operation agreements of community services such as transportation, utility, and park needs. There can be a great deal of variation in the structure of partnerships that take place between governments and private organizations, marked by a mix of interests and power sharing that may occur (Wettenhall, 2003). Effective expertise, contract management, and structures of state and local government are particularly important in the overall success and long-term health of many PPPs (Bloomfield, 2006).
Improvement districts are a form of PPP, as they draw on a combination of public and private funding streams, as well as typically being governed by a board of government officials alongside local residents and property owners. These districts require a unique form of hybrid management of local spaces, by tasking private organizations to engage with and deliver services to communities in tandem with government officials (Grossman, 2010). These districts allow for a notable amount of flexibility among different communities, as distinct area needs and varying assessment funding levels can create innovative management and service delivery (Mitchell, 2009). Historical development of different communities in local spaces can also shape ways in which improvement districts address citizen needs and provide distinct services (Ziebarth, 2019).
Improvement districts are influential in many cities across the United States, and increasingly across localities globally. These districts are appealing to socioeconomically diverse communities, as they provide local residents and property owners a significant amount of creative leadership in reshaping public spaces. The demand for citizen empowerment in the future development of communal spaces at the local level can allow for innovative approaches to take root and develop towns and cities. This reflects a distinct form of public management that can be amenable to social and physical redevelopment (Grossman, 2008; Hambleton, 2000).
This also makes improvement districts a unique and exciting phenomenon to study, particularly in relation to democratic participation.
Contemporary studies have focused on the effects of improvement districts on local government and spaces in relation to economic development (Nijaki and Worrel, 2012), public safety (Cook and MacDonald, 2011; MacDonald et al., 2013), and service distribution (MacLeod, 2011). Despite the interesting questions of democratic participation and civic engagement posed by the creation and implementation of these local improvement districts, research has lagged behind in seeking to address these issues. Studies have examined the role of democratic accountability for improvement districts in the community, as they serve as a mechanism of community organization that is beholden to both public and private actors (Meek and Hubler, 2006; Morçöl and Patrick, 2006). Yet these studies are focused primarily on organizational and legal structure, and do not assess active citizen engagement.
Other research has studied PPPs and civic engagement in relation to education reform (Fusarelli et al., 2011; Kowalski, 2010). These studies, however, are confined to theoretical and empirical analysis of the education system. Since education is a single aspect of civic engagement and community participation, there remains a number of other aspects of this area that cannot be explained theoretically or empirically by research on school boards and structures of educational organization. This reflects important places in the literature that remain largely unexplored.
A Civic Engagement Participation Theory
The theoretical expectation underlying this study is that local improvement districts will increase democratic participation in the communities they serve. This is based around the premise that increasing the civic engagement of individuals with government and community organizations will create a greater likelihood that eligible voters will seek to shape the electoral outcomes of who represents their community in government. This follows closely to the resource model of political participation, which conceives of organizational approaches and communication as integral aspects of political participation (Brady et al., 1995).
The resource model counts individual access to time, money, and civic skills as essential in explaining patterns of political participation. The theoretical basis for the research question and empirical investigation of this piece is similar to the resource model, but differs in that it places primary focus upon the way community services and resources affect democratic participation, as opposed to the way in which individual resources affect individual participation. This serves theoretically as a departure by placing emphasis on democratic participation as a socially influenced activity shaped largely by the role of community and efficacy of local civic engagement.
This perspective on voter participation differs from previous theoretical accounts of participation as strictly decisions of rational choice. Rational choice theories, which conceive of political activity as a set of utility maximization choices, have historically had trouble explaining patterns of democratic participation. Early rational choice models of political participation developed a calculus of voting, whereby individuals will be more likely to vote if there is more regard given to how important their vote will be in changing an election (Downs, 1957; Riker and Ordeshook, 1968). These models were later rejected and revised because it was argued that not all individuals would employ this calculus in electoral participation decisions (Ferejohn and Fiorina, 1974). This expanded the rational choice account of predicting when and why citizens will cast a vote. Further updating to rational choice perspectives of political participation took place to contend that voting was a uniformly low-cost, low-benefit decision for individuals, and strategic political activity was heavily influential in moving citizens to vote (Aldrich, 1993).
The civic engagement participation theory conceives of electoral participation as an action influenced by more than a function of individual interaction with campaigns and political media, and individual access to resources. This is not to say that these aspects are not influential as well, as campaigns, media, and individual resources are important in shaping political participation. However, the stability and reach of community civic outreach is considered to be an essential aspect of democratic participation among community residents. Citizens are expected to be more likely and willing to vote in elections when they have stronger ties to activities and organizations in their community that are not specifically political, particularly in state and local elections. While on the surface, improvement districts are not politically powered organizations, they require an inextricable link between state and local government officials and agencies with community leaders, residents, and property owners.
Improvement districts bring together property owners, both residential and commercial, typically in an urban area. For an improvement district to form, a group or individual must come forward to contact the designated local government agency overseeing the registry process. This group or individual must present a plan for the improvement district and receive approval from at least half of the residents and property owners within the geographic boundaries of the district. The improvement districts are then governed by a Board of Directors made up of residents, property owners, and local officials. Importantly, this process displays mechanisms for broad community support, with local officials, community board members, and residents having to both support the establishment and continue the maintenance of the improvement district. Through this distinct governing structure, improvement districts engage in a variety of activities that engage with local residents and must respond to the needs of the given community.
Figure 1 displays the basic theoretical model for civic engagement participation as reflected through the implementation of local improvement districts. The formation of an improvement district requires two simultaneous conditions, whereby issue leadership and community collaboration occur. Issue leadership in the model can be explained in the formation process by the individual or group who takes initial command of the development process, through reaching out to the required local government office, develop and submits the initial proposal for the improvement district, and takes charge in communicating with local residents and property owners about the district.
Political Participation Theory Model.
Community collaboration is the piece tied to broad public support and localized efforts to engage with the improvement district process from establishment to maintenance. This occurs through the requirement that over half of the residents and property owners in a district support the establishment of the district, as well as the Board being governed by various community stakeholders who live in, work in, and help govern the community. Since the establishment of the improvement district not only requires this majoritarian support among community members, but also brings together members of the community to discuss this localized issue, the development of the improvement district itself is central in initially raising public engagement.
Following the formation of the improvement district, not only do leaders of the district continue to engage with the community, but community members also engage with local residents, businesses, and community organizations. This is the mechanism through which it is theorized that community member will be more likely to actively participate in local elections. Improvement districts not only provide many public management services such as sanitation, beautification, and structural updating, but lead community activities. As part of the creative and innovative aspects of reaching out to communities, improvement districts sponsor programs for local residents and families, that include, but are not limited to, outdoor movies, public art, and concerts. Through these continued localized interactions, the theoretical expectation is that community members will be more likely to participate in local elections through a greater interest and connection with their locality.
Data and Methodological Approach
To assess the relationship between local improvement districts and democratic participation, an original data set was constructed. Data were collected from the New York State Board of Elections, the New York State Task Force on Demographic Research and Reapportionment, and the New York City Small Business Services Department from 18 state assembly districts from 2002 to 2018. The New York State Board of Elections data were collected on enrollment levels of active, registered voters in each observed assembly district during each observed year, and vote count levels in each observed election. In total, the study observes 22 improvement districts across 9 elections, giving a sample size, N, of 198.
The New York State Task Force on Demographic Research and Reapportionment data compiled demographic characteristics, per-capita income levels, and education levels for each assembly district, adjusted for each available census over the observed time span. The New York City Small Business Services Department data compiled included BID name, borough, assembly district location, and assessment rate. These state assembly districts include a total of approximately 1.3 million active, registered voters during each election. All state assembly districts observed are located within New York City. New York City is a useful area to assess in a study on improvement districts, as the city itself has the largest number of these districts in the United States and is socioeconomically diverse across its many neighborhoods (Unger, 2016).
As with the majority of improvement districts, in New York City improvement districts are governed by a Board of Directors, which has property owners, residents, and local elected officials on the Board. The Board of Directors is then in charge of overseeing decisions pertaining to staffing and outreach associated with the district. Using New York City as a case study is beneficial for a number of reasons. When selecting a case, or cases, to study one seeks to achieve a representative sample and useful variation on the dimensions of theoretical interest (Seawright and Gerring, 2008). New York City is home to a large number of improvement districts, with the districts being found across a diversity of communities, representative of the city itself. In addition, since New York City is home to a large number of improvement districts, there is greater internal variation within the selected case in comparison to a city with few districts.
It is important to note, however, that this single case study may have some consequences as well. Since New York City is a very large city, there is a possibility the effect of improvement district formation on political participation could differ in smaller cities or towns. In addition, since New York City is demographically and economically diverse, there is the possibility that results could differ in more demographically or economically homogeneous cities. Finally, as with any study, the results presented should be recognized for both their strengths and limitations. While a single case study of a large city puts a certain level of constraint on generalizability, the availability of data on a number of districts and communities across time allows for a more thorough analysis of the subject in question and can provide findings against which future research can compare.
Table 1 provides a list of each of the improvement districts observed for the purposes of this study, as well as the borough in which they were established, the annual assessment rate for the districts, the state assembly district in which each improvement district is located, the number of registered voters in each assembly district, and the assessment fee rate per registered voter in the district. The twenty-two improvement districts included in this study are found in all five boroughs of the city, and range in annual assessment budgets from $165,000 to $2,500,000. Assessment fees can vary not only as a result of individual improvement district regulations, but as a result of the number of properties within the district resulting from differences in population and building density. Since these improvement districts cover diverse communities throughout the city, this property structure variation can affect the overall assessment fee collection when observing a number of improvement districts. We can see, as well, in Table 1 that the number of registered voters in the assembly districts in which the improvement districts are located ranges from 49,764 to 94,990, while the per-voter assessment fee rate has a range from $2.45 per registered voter up to $26.35 per registered voter.
Observed Improvement Districts.
To inspect whether a significant relationship exists between the implementation of an improvement district in a given community and increased democratic participation among community residents, a multivariate model is constructed. The dependent variable for the model is the voter participation rate in the assembly districts during a given election cycle. The independent variable of interest is the binary presence or absence of the improvement district in the assembly district at the time of each election. The variable is coded 0 in years where the improvement district is not present in the district during an election cycle, and coded as 1 in the years where the improvement district is present during an election cycle. The model accounts for the effects of improvement district presence on democratic participation across time by beginning observation of voter participation at a point in time in which no improvement districts are present and staggering the creation of improvement districts in the assembly districts to assess the effects of improvement district presence on changes in voter participation.
Controls are included in the final model for differences among the observed districts concerning disparities in income levels, education levels, racial demographics, and size of the improvement district, as measured by annual assessment fees. Per-capita income is used to account for district income level. Education level is measured by the proportion of district residents who have earned an associate’s degrees, or have earned additional education or degrees above the associate’s level. The income and education variables use log transformations to reduce skewness of the measures. Differences in demographic makeup of the district account for the proportion of a district’s residents who self-identify as either non-Hispanic Asian, non-Hispanic Black, Hispanic, and non-Hispanic White. Size of the improvement district is measured by annual assessment fee. The size of the district is measured by annual fee, as opposed to geographic size, because differences in community residential density may distort comparisons between districts based on such a measure. Instead, annual assessment fee is a better measure of the reach presented by a given improvement district. Controls are included to avoid issues of omitted variable bias in testing the statistical significance of BID presence as a mechanism for spurring democratic engagement. This allows for the model to include adjustments for potentially confounding variables that may affect differences in voter participation outside of the improvement district development.
In addition to the multivariate OLS regression models, two fixed effects panel models are developed to assess the causal effects of improvement district establishment and political participation. This allows for the models, which observe elections across two distinct time periods, to account for district-specific endogeneity by design (Burden and Wichowsky, 2014). In addition, using fixed effects models for the panel data accounts for possible time-invariant confounders at the assembly district level, and controls for omitted variable biases which can occur as a result of unique district characteristics (Fowler and Hall, 2018). By employing this design, the possibility of a causal link between improvement district formation and political participation can be empirically supported to a greater extent. The first model assesses district political participation in the 2002 and 2006 midterm election years. The second difference-in-difference model follows from the first, but instead observes political participation rates from 2004 and 2008 presidential-year elections.
Results and Findings
To first assess whether a difference is present in voter participation rates within the state assembly districts in elections following the creation of a BID within the assembly district, the mean voter turnout rates in assembly districts before the presence of a BID and following the presence of a BID are observed along with the interquartile range and possible outliers. Figure 2 displays a box plot visualization of these data. The box plot on the left, representing voter participation rates in the observed assembly districts prior to improvement district presence, has a mean turnout of approximately 35%among active, registered voters. On the right, the plot showing voter participation rates in the assembly districts following improvement district creation displays a mean turnout close to 42%. Voter turnout in assembly districts following improvement district presence has a wider range of turnout; however, there was no presence of any outlier observations in either set.
Political Participation Rates Pre- and Post-Improvement District.
While there appears to be a meaningful effect on voter participation rates in state assembly districts following the creation of an improvement district, further steps are taken to assess whether there is a significant relationship of improvement district creation on voter participation and to control for possible confounding aspects of districts which may affect voter participation. Table 2 displays the effect of the independent variable of interest, BID presence, along with independent control variables modeled income, education, and demographic variation, on the dependent variable, voter participation in assembly districts.
Effects on Assembly District Voter Participation.
p < 0.001; ** p < 0.01; * p < 0.05.
Notes: All models use ordinary least squares. The dependent variable for all models is the presence of a local improvement district. The Income variable is measured by logged per-capita income for the district. The Education variable is measured by logged proportion of population with an Associates-degree level of education or higher. Demographic variables are measured by the proportion of a given district’s population which identify with a given racial group. The Annual Assessment variable is measured by logged annual assessment for the improvement districts.”
In the table, we can observe that after the addition of a set of variables controlling for the presence of an improvement district remains statistically significant at the 0.01 level, and is expected to affect a rise in voter turnout of approximately 5%. Not unexpectedly, income is significantly associated with voter participation, but appears to have a minimal effect. Furthermore, there does not appear to be a significant relationship between education levels and change in voter participation rates, and only minor, non-significant variations in turnout rates based on assembly district racial demographics. Finally, there is no relationship between the size of an improvement district and the rate of voter participation.
The multivariate OLS regression results presented reflect a meaningful, significant relationship between creation and continuation of a local improvement district and democratic participation in subnational elections. Over the observed time period of nearly two decades, the creation and continuance of an improvement district was associated with higher mean levels of voter participation rates across 18 assembly districts with over 1.3 million active, registered voters. To control for possible random variation, a multivariate model with multiple control factors was developed, which also took into account variation over time through staggered dummy variable coding to match rates prior to and following improvement district presence across all districts.
Across time, we recognize a notable pattern that not only in these assembly districts were there occurrences of voter turnout rates higher than previously observed across multiple elections prior to improvement district presence, but these higher rates of voter turnout were relatively consistent. In Figure 2, we detect visible clustering of voter turnout rates between 45% and 55% in assembly districts following the creation of improvement districts. In the same figure, we saw noticeably fewer turnout rates above 40%prior to improvement district presence, and assembly district turnout rates above 40%were less consistent. The confidence interval for the model shown in this figure is relatively narrow, showing further evidence that it demonstrates an accurate model with minimal noise for understanding improvement district effect on democratic participation.
The regression statistics for the model displayed in Table 1 further support the visible differences from the box plot and visualized regression figures. Even after controlling for a number of possible confounding changes in districts that may affect voter participation, such as income levels, education levels, and demographic differences, the presence of an improvement district remained significantly associated across time with increased voter participation rates. Over time, improvement districts were expected to raise participation rates among registered voters by 5%on average. This additionally reflects the significant role in which improvement districts played across assembly districts and among citizens.
The only control variable that remained significant in predicting voter turnout rates in the final model was income levels. This is not an unexpected finding, as higher levels of income are typically associated with higher rates of electoral participation. However, income levels ultimately displayed no meaningful change in voter participation rates in terms of magnitude. One reason for this may be that the same districts were observed across time, meaning that major shifts in assembly district income levels did not occur. Nonetheless, it is encouraging that differences in education, income, and racial makeup did not have significant, substantive effects on democratic participation, and equally noteworthy that the inclusion of a local improvement district across state assembly districts reflected an expected 5% increase in voter participation rates that was significant at the 0.01 level across nearly two decades.
Table 3 shows the fixed effects panel models for the midterm elections in 2012 and 2016 and the presidential year elections in 2014 and 2018, with both standard errors and robust standard errors, respectively. We see that across both midterm and presidential-year elections, improvement district establishment has a substantive, positive effect on within-district political participation. The midterm effect on political participation represents a 2.6% increase in turnout rates, while the presidential-year effect is even greater at 8.4%. When observing robust standard errors, both the midterm and presidential-year effects are statistically significant, with the presidential-year effect reflecting an especially high level of significance.
Fixed Effects Assembly District Turnout.
p < 0.001; ** p < 0.01; * p < 0.05.
Notes: All models used fixed effects at the improvement district-level. The dependent variable for each model is present of an improvement district. The SE Midterm Turnout and SE Presidential Year Turnout models present the fixed effects regressions with standard errors. The Robust Midterm Turnout and Robust Presidential Year Turnout models present the fixed effects regressions with robust standard errors.
The results from Table 3 further support findings from the previous multivariate OLS models that improvement district implementation is associated with positive rates of change in political participation in state assembly district elections. This provides greater evidence for the previously discussed civic engagement theory of political participation, displaying how across time, the development of these community development organizations can have a meaningful effect on voter turnout. The fixed effects panel models also provide causal support for the theory by observing within-district changes to reduce issues of confounding variables and omitted variable bias that can result from community differences. Accounting for these factors, the results develop an even more empirically extensive picture of improvement district implementation effect on political participation in subnational elections.
Conclusions
Can local improvement districts affect democratic participation in the communities in which they are created? The findings from this study provide evidence that this can, in fact, be the case. After modeling data collected across time from a number of socioeconomically diverse state assembly districts with over 1.3 million active, registered voters prior to and following the implementation of improvement districts, the results reflected a clear relationship between improvement districts and higher levels of political participation. Even after controlling for variation in income, education, and demographic characteristics in a number of multivariate OLS regression models, the presence of an improvement district was significantly and substantially associated with an increase of approximately 5% in voter turnout rates.
In addition, further causal support was provided by two fixed effects panel models. These models reached similar results, reflecting substantive, statistically significant rises in political participation within districts which had implemented improvement districts. It should be noted that although the multivariate OLS models help point to an association between improvement district development and increased political participation rates, and the fixed effects models testing the same relationship integrate a quasi-experimental design to bolster causal claims, future work testing similar hypotheses with either natural experiment or regression discontinuity design (RDD) could provide stronger causal evidence for the theoretical and empirical evidence provided here.
Since improvement districts observed were located in a variety of economically and racially diverse areas, the model was constructed with cognizance that economic and racial disparities occur in voter participation. Economic realities, as well as issues of time and transportation constraints, are placed on voters, particularly those in communities that have been socioeconomically marginalized, making it important to be inclusive of these communities when assessing democratic participation. This is particularly essential when observing local politics in urban areas where spatial segregation and economic inequality can be especially pronounced, and must be addressed. Since local improvement districts are located throughout these areas, they are increasingly important in state, local, and urban politics.
The theoretical framework behind the empirical modeling for this study is based around a civic engagement participation theory of political participation. The development of this was based around previous research surrounding civic engagement and local improvement districts. Studies of civic engagement have reflected the significance of social capital among citizens for efficacy in public affairs (Kenworthy, 1997; Putnam, 2000), as well as the importance of civic engagement in good governance practices (Keele, 2007; Rice, 2001). Research has also shown how perception and interaction among citizens with systems in their communities can affect their participation in democratic citizenship (Weaver and Lerman, 2010). Previous work concerning PPPs has been substantive in addressing issues of economic development (Nijaki and Worrel, 2012), service delivery (MacLeod, 2011), and education (Fusarelli et al., 2011; Kowalski, 2010). There remains a significant gap, however, in studies addressing improvement districts as community groups tied into the fabric of civic affairs in local spaces.
The framework utilized followed most closely with the resource model of political participation, which views political participation as an action shaped largely by individual access to resources, including time, money, and civic skills (Brady et al., 1995). The civic engagement participation theory differed slightly, however, as it considers community resources and activities as a significant aspect of shaping citizens’ political participation. While not rejecting the idea that individual resources shape patterns of political participation, it theoretically contends that citizens’ social environment and community engagement hold meaningful influence on political participation that must be taken into consideration when seeking to explain voter turnout. This theoretical expectation was supported by the empirical results presented, as voter participation rates were consistently and considerably higher across multiple elections and assembly districts following the implementation and management of local improvement districts.
One disheartening finding drawn from the model is that even after the creation of improvement districts, many state assembly districts still had voter turnout rates below 30%. This is particularly an issue, as even a 5% increase in voter turnout rates within an assembly district that had 20%voter turnout rates previously would still leave that district with participation rates at or below 25%. Future studies should put particular focus on the ways in which communities with especially low levels of political participation can address these issues, and how PPPs may be able to connect citizens with political processes. Subsequent research would also be well served to look at other forms of political participation and civic engagement in relation to similar community development groups, leveraging increasingly sophisticated methodological techniques.
This study looks strictly at voter participation, but future studies may look at aspects of civic engagement and democratic citizenship such as running for elected office, social organization membership, and community activism. Studies could also look into differences between urban and rural areas. Since assembly districts observed in this study were all located within New York City, the findings apply more specifically to urban politics and in more sparsely populated rural areas we may see different strengths in effect. Subsequent research should assess this possibility. The substantive and significant relationship between the establishment of an improvement district and democratic participation across socioeconomically diverse areas show encouraging evidence that collaborative community groups and processes between local government and citizen organizations can positively shape democratic participation.
Footnotes
Declaration of conflicting interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.
