Determinants of Voter Participation in Neighborhood Council Elections

Abstract

Neighborhood governance has emerged as an important arena for voluntary civic engagement, but little is known about how election participation varies across neighborhoods. We examined the election participation of 88 Los Angeles neighborhood councils (NCs) and found much of the variation can be explained by neighborhood demographic and socioeconomic characteristics. Consistent with the voting participation literature, NCs with a higher percentage of Hispanic or poor residents had lower participation. Contrary to expectations, NCs with a higher percentage of Asian/Pacific Islander residents or lower educational attainment had higher participation. Interestingly, the presence of nonresidential NC stakeholders, such as business owners, workers, and nonprofit organizations, was not significantly related to participation after controlling for other factors.

Keywords

neighborhood council voter participation governance civic engagement

Neighborhood councils (NCs) emerged in Los Angeles in the late 1990’s to promote collaborative governance and increase voluntary civic engagement in policymaking in the midst of increasing municipal dissatisfaction and fragmentation (Berry, Portney, & Thompson, 1993; Box & Musso, 2004; Chaskin & Abunimah, 1999; Chaskin & Garg, 1997; Kathi & Cooper, 2005; Musso, Weare, Oztas, & Loges, 2006). Neighborhood-based governance could offer opportunities to more effectively engage local decision making (Chaskin, 2003). Although previous research indicates voter participation is generally lower at the municipal level than at the state or national levels (Hajnal & Lewis, 2003) and community participation varies along racial/ethnic, socioeconomic, and social dimensions (Costa & Kahn, 2003; Dekker & van Kempen, 2008; Lien, Collet, Wong, & Ramakrishnan, 2001; Marschall, 2001; O’Brien, 1975), we know little about participation in these relatively new submunicipal institutions, and how voter turnout varies by NC demographic, socioeconomic, and organizational characteristics. Understanding determinants of NC participation is important for evaluating the effectiveness of neighborhood-oriented governance and for making these new institutions more effective in areas with weaker participation.

In this article, we investigate factors that influence participation in board elections for NCs in the City of Los Angeles which were created in 1999 in response to calls for greater neighborhood autonomy and as an alternative to the movement of subareas to secede into separate jurisdictions (Musso, Weare, & Cooper, 2004). Although the city’s NCs have no formal powers, they are official governmental entities with an elected board designed to foster representation of stakeholders in each neighborhood and to improve service delivery by identifying neighborhood priorities and needs. Stakeholders allowed to participate in elections include those who live, work, own property, or attend school or church in the neighborhood regardless of citizenship status. NCs gained increased importance as the primary mechanism for city governance reform after citizens voted against the secession efforts of the San Fernando Valley and Hollywood areas in November 2002. Previous research provides insight into the formation and board composition of NCs in Los Angeles (Jun, 2007; Jun & Musso, 2007). Analysis of the racial and ethnic composition of NC governing boards suggests that their descriptive representation of their neighborhood is low; they tend to be more disproportionately comprised of long-term residents and are disproportionately older, more affluent, well-educated, and White (Guo & Musso, 2007). These patterns draw into question whether these voluntary neighborhood governance organizations adequately represent the diverse interests of their neighborhood, but we know very little about the influence of neighborhood composition on NC board election participation.

The participation rates in neighborhood governance have been disappointedly low in Los Angeles and other cities (Berry et al., 1993; Musso et al., 2004). This pattern in many ways echoes the poor participation observed during an earlier generation of neighborhood initiatives in the 1960’s and 1970’s (Berry, Portney, Bablitch, & Mahoney, 1984; Moynihan, 1969). Low NC participation in Los Angeles may be related to the fact that they play an advisory role in municipal governance and do not have direct control over local land use and city services (Musso et al., 2004). Although assessing the individual stakeholder motivations associated with participation is beyond the scope of our analysis and data availability, we examine the neighborhood-level factors associated with considerable variation in NC participation to inform the city’s ability to address the often divergent needs and priorities of its neighborhoods. More specifically, we examine the relationship between NC demographic, socioeconomic, and organizational characteristics and participation rates in the most recent board election for the 88 NCs certified in the City of Los Angeles as of May 2008 to (a) identify demographic and socioeconomic characteristics associated with potential barriers to engagement in neighborhood governance and (b) to examine whether the presence of nonresidential stakeholders influences NC election participation since, unlike elections at other levels, area stakeholders such as business owners and those who attend school or work in the area can vote in NC elections.

Background

The establishment of NCs has not resulted in widespread participation in neighborhood self-governance in Los Angeles. A 2004 evaluation of the system indicates more than 20,000 stakeholders voted in the first NC elections, representing about 1.3% of residential population, a level slightly lower than the 4% participation rate in the city’s general elections (Musso et al., 2004). This low participation in NCs has prompted some to criticize the system for not providing an effective mechanism for community input and resulted in the issuance by a city commission of a list of 73 recommendations in September 2007 to address discontent within the system and to improve outreach, participation, and racial representation (Hymon, 2007). Although some debate whether neighborhood organizations adequately represent concerns of residents (Swindell, 2000; Verba, Schlozman, Brady, & Nie, 1993), understanding the factors associated with participation in NC board elections could provide substantial insight into factors underlying dissatisfaction with the system.

On the basis of interviews and surveys of policy makers, Berry et al. (1993) compared 5 cities with functioning NC systems with 10 cities with no comparable neighborhood institutions. They found that although cities with NCs did not have higher levels of participation than comparison cities, the type of activities and participation in cities with NCs was qualitatively more meaningful and required a greater commitment and stronger personal relationships. Residents of cities with NCs also had higher levels of tolerance and political efficacy. Furthermore, Berry et al. (1993) found that smaller NCs representing fewer than 10,000 residents enabled greater interaction.

Understanding participation in municipal governance is problematic because involvement is voluntary and individual benefits could vary substantially. Neighborhood governance institutions offer the potential of encouraging enhanced provision of public goods from the city including public expenditures, land use oversight, and city services. Residents cannot be excluded from the benefit of these public goods because of a lack of participation, resulting in a free-rider problem (Olson, 1971). Stakeholders are not forced to participate, and even if they participate they may not benefit. Given smaller groups reduce the cost of coordination, tend to reduce the free-rider problem, and provide many public goods more efficiently (Olson, 1971), more localized governance could connect stakeholders more directly to benefits and prompt greater engagement. These incentives may be limited in Los Angeles NCs, however, as they only have an advisory role on the allocation of public goods and have no formal fiscal and governance powers.

Although the free-rider problem could reduce overall levels of participation, individual and contextual factors could produce variations in the level of engagement across neighborhoods. Individuals may participate for reasons including a sense of civic responsibility embedded in norms and values, nontangible rewards such as prestige, or to promote the work of existing organizations. Variations could also exist because of differences in personal opportunity costs and barriers to civic and political engagement. Voting participation tends to be influenced by socioeconomic and demographic factors, social and political connectedness, and political context (Lien et al., 2001). Individual level factors such as age, gender, household composition, home ownership, and length of residence have been shown to affect neighborhood involvement (Dekker & van Kempen, 2008). There are also very distinctive variations by race, with non-Hispanic Whites and African Americans having much higher levels of voting than Asians and Latinos, largely due to the disproportionately higher numbers of noncitizen immigrants among the latter two groups. Even among those who have become naturalized, there are structural barriers such as voter eligibility requirements and the lack of bilingual voting materials (Lien et al., 2001; Wong, Lien, & Conway, 2005). Although the literature on electoral voting is useful, it is important to note that noncitizen Latinos and Asians are eligible to vote in NC elections.

Previous research suggests that neighborhood engagement varies by nearby racial/ethnic and socioeconomic residential characteristics and that the social context of neighborhoods and the degree to which individuals are connected socially affects involvement in local politics. Marschall (2001) suggests that involvement of African Americans and Latinos is strongly associated with attachment to neighborhood and religious institutions. She suggests that immigration status, English language ability, and education may factor into lower participation in schools and community safety organizations among Latinos (Marschall, 2001). Costa and Kahn (2003) found lower civic engagement in more heterogeneous communities. O’Brien (1975) argues that, despite their sizeable needs, the most disadvantaged areas may not be the most likely to be civically engaged or to support neighborhood organizations.

Previous research on NCs in Los Angeles suggests that participation varies by neighborhood context. Jun (2007) found that more economically heterogeneous communities in Los Angeles tended to establish NCs earlier, whereas more racially heterogeneous communities were less likely to have early formation of NCs. The length of residence and English-language ability were positively related to earlier NC formation. Interestingly, areas with higher home-ownership rates were less likely to have an early NC formation leading the author to speculate that community interests in these areas may have already been represented to some degree by home owner’s associations. Contrary to expectations, areas with older residents were slower to form NCs (Jun, 2007). Jun and Musso (2007) examined the composition of NC boards and found that Hispanic/Latino representation on Los Angeles NC boards was higher in areas with higher residential segregation and higher income, but that this pattern did not hold for African American representation.

Unlike most other forms of municipal governance, a range of stakeholders can participate in NCs to influence local planning priorities and investments. Business owners may have a strong incentive to participate to influence neighborhood vitality or local development and to enhance the business environment and profitability, whereas area students and church congregations may participate to improve safety and quality of life. Landlords may participate to protect the viability of their rental properties. Staff of nonprofit organizations may have an incentive to become involved to promote their interests in prioritizing and addressing community needs. Although the interests of these groups could converge, very little is known about the extent to which the presence of these stakeholders influences NC participation.

Community groups could also represent existing social capital in a neighborhood and prompt greater participation among residents and their clientele. Jun (2007) found that the presence of more community-based organizations was positively related to earlier NC formation (Jun, 2007). Nonprofits organizations in urban areas, particularly organizations that receive government funding, play a role in encouraging their clientele to vote and contact elected officials (LeRoux, 2007). The presence of more organizations, however, could negatively affect participation if local interests are being addressed in community organizations, business groups, and/or home-owner associations. Among persons engaged with preexisting strong informal networks of social capital, many may not necessarily become involved in more formal associations (Hays & Kogl, 2007). Furthermore, despite community consensus on issues and concerns, a high number of neighborhood associations could be related to community-level factionalism if organizations complete (Meyer & Hyde, 2004).

For instance, business organizations such as business improvement districts (BID) could have a complementary and substitution effect on NC engagement. They are local organizations that have been largely successful in promoting the enhancement of local public goods. They are a neighborhood institution in which property owners in a municipal subarea volunteer for additional taxation to provide neighborhood-wide services such as cleaning, security, and maintenance that are not being provided at a satisfactory level by the city. BIDs solve a collective action problem, as once a majority of merchants and commercial property owners vote to establish a BID, they are all legally bound to pay the tax. BIDs have played a major role in urban revitalization and reducing crime and vandalism (Brooks, 2008; Hoyt, 2005; Mitchell, 2001). The existence of BIDs may compete for the attention of business and property stakeholders, given these stakeholders may have scarce time to devote to a new organization.

Data and Method

Information on NC election patterns were obtained from Los Angeles Department of Neighborhood Empowerment (DNE) and included information on NC certification date and status, election dates, and vote counts since formation. About 90% of the City of Los Angeles fell in one of its 88 NCs certified as of May 2008 and are included in the analysis (Figure 1). About one third of these NCs held their first election of board members in 2002, and another 40% held their first election in 2003. NC geographic boundaries were obtained from the Los Angles Geographic Information Systems Division. We assigned blockgroup (BG)-level population, demographic, and socioeconomic data from census 2000 and census 2010 to NCs based on BG centroid location. We examined two participation rates to take into account that nonresidential stakeholders are allowed to participant in NC elections. Participation Rate 1 is the number of votes in the most recent NC election as a percentage of voting-age residents (18 years or older) in 2010. Participation Rate 2 is the number of votes in the most recent NC election as a percentage of the number of voting-age residents in 2010, employees who worked in the NC area in 2006, and firms in the area in 2000. We derived the number of area firms in 2000 based on BG-level data from InfoUSA and the number of area workers in 2006 based on block-level data from the Longitudinal Employer–Household Dynamics (LEHD) program at the U.S. Census Bureau. Rate 2 provides a more complete representation of potential participants even though, due to data limitations, it does not include other stakeholders such as landlords. Ideally, the population, employment, and firm data we used in Rate 2 would represent the same time period but unfortunately such data were not available for this study.

Figure 1.

Voter participation rates in neighborhood councils in the city of Los Angeles

Table 1 describes key variables and their expected influence on participation based on the literature. We include residential NC demographic and socioeconomic characteristics that can influence civic and political participation. We include variables to account for area race/ethnicity, educational attainment, age, and income, as these characteristics are associated with variance in voter registration and voting patterns (File, 2008). The size of a NC as measured by Population Log could affect NC participation, as smaller NCs could allow for more meaningful involvement (Berry et al., 1993). We also include the percentage of housing units occupied by renters because previous research suggests home ownership may be associated with neighborhood involvement (Dekker & van Kempen, 2008).

Table 1.

Description of Key Variables

Variable	Description (source)	Expected net influence on participation
Participation Rate 1	The number of participants in NC elections as a percentage of 2010 voting-age residents (DNE, U.S. 2010 Census)
Participation Rate 2	The number of participants in NC elections as a percentage of 2010 voting-age residents, 2006 employees, and 2000 firms (DNE, U.S. 2010 Census, 2006 LEHD, InfoUSA, 2000)
% API (single race), 2010	Percentage of single-race Asian/Pacific Islander persons (U.S. 2010 Census)	—
% African American (single race), 2010	Percentage of single-race African American persons (U.S. 2010 Census)	—
% Hispanic, 2010	Percentage of Hispanic persons (U.S. 2010 Census)	—
% Poverty, 1999	Percentage of persons in households below the Federal Poverty Level in 1999 (U.S. 2000 Census)	—
% less than high school, 2000	Percentage of persons aged 25 years or older with less than a high school education (U.S. 2000 Census)	—
% older people, 2000	Percentage of persons aged 65 years or older (U.S. 2000 Census)	+
Log of total 2010 population	Log of the number of residents (U.S. 2010 Census)	—
Jobs per resident, 2006	The number of jobs as a percentage of 2010 residents (U.S. 2010 Census, 2006 LEHD)	?
Firms per resident, 2000	The number of firms as a percentage of residents (InfoUSA, 2000)	+/−
Organizations per resident, 2006	The number of community-based organizations and BID per resident (multiplied by 100)	+/−
Years since first vote	The number of years since the NC held its first election (DNE)	+

Note. DNE = Department of Neighborhood Empowerment; LEHD = Longitudinal Employer–Household Dynamics; NC = neighborhood council; API = Asian/Pacific Islanders; BID = business improvement districts. “-” Indicates the expected influence is negative, “+” indicates the expected influence is positive, and “?” indicates the expected influence is unclear.

We also account for the presence of community and business groups within a NC because preexisting social capital and history of civic engagement could complement and heighten involvement in NCs. Conversely, the presence of community and business groups could dampen turnout if existing organizations already provide a conduit for addressing community concerns. We obtained information on nonprofit organizations involved in activities such as the arts, education, health care, and human services in the county of Los Angeles in 2006 from the Urban Institute’s National Center for Charitable Statistics (http://nccs.urban.org/). These data include 501(c)(3) public charities and private foundations which filed an IRS form 900 and had annual gross receipts of US$ 25,000 or more. We geocoded more than 90% of the organizations with a valid street address (i.e., not a post office box) and determined that just more than 2,500 were inside a NC boundary in the City of Los Angeles. We also identified the boundaries of 35 BIDs that were located in NCs in 2006 based on original designation information from the City of Los Angeles and information obtained from BIDs. We combine the number of nonprofit organizations and BIDs in each NC for the analysis. We also account for the number of years since a NC’s first election, given that voting and capacity to organize could increase as NCs become more established and publicized; conversely, voting could drop after the interest and activity surrounding NC formation. As NCs vary in size, we normalize our independent variables for the numbers of firms, jobs, and organizations by the number of residents.

We use multivariate analyses to examine the variation in participation in NC elections. Our conceptual model takes the following form:

P = f (R, O, W, B),

where P is the NC participation rate in the most recent election. R is a vector of the residential NC demographic and socioeconomic characteristics that the literature suggests are associated with voter turnout and neighborhood participation. The remaining variables account for the potential influence of other stakeholders: O (organizations), W (workers), and B (businesses).

We developed three models for Participation Rate 1 and three models for Participation Rate 2. For each rate, we specified one model using the full sample (N = 88) and two models using a restricted sample which excluded NCs which were outliers with participation rates more than 5% (N = 86). This approach allowed us to examine the robustness of results after eliminating outliers. Some variables in the multivariate analyses were highly correlated (Table 2) including variables for NC percentage of Hispanic, percentage of poverty, percentage of less than a high school, and percentage of older people, as were variables for NC jobs, firms, and organizations per resident. Our evaluations of the models, including t tests of estimated coefficients and f tests of adding/removing variables, indicate that although we cannot rule out that collinearity may create problems, the correlated variables add to the explanatory power of the models. Furthermore, leaving out one or more of these variables could lead to a model misspecification. Given the potential problem of multicollinearity, though, we report multivariate models which exclude variables with the highest intercorrelations. We did not combine these variables into a single scale variable because we hypothesize based on the literature reviewed above that each could have an independent and different influence on NC voter participation.

Table 2.

Correlation Matrix for All Variables

	Participation Rate 1	Participation Rate 2	% API, 2010	% African American, 2010	% Hispanic, 2010	% poverty, 1999	% less than high school, 2000	% older people, 2000	Log of total 2010 population	Jobs per resident, 2006	Firms per resident, 2000	Organizations per resident, 2006	Years since first vote
Participation Rate 1	1.00	0.96***	0.42***	−0.06	−0.15	−0.02	0.03	0.28***	−0.43***	0.08	0.03	0.17	0.17
Participation Rate 2	0.96***	1.00	0.35***	−0.05	−0.10	−0.04	0.03	0.20*	−0.49***	−0.05	−0.08	0.03	0.14
% API, 2010	0.42***	0.35***	1.00	−0.34***	−0.25**	−0.11	−0.15	0.14	−0.26**	0.15	0.09	0.26**	−0.22**
% African American, 2010	−0.06	−0.05	−0.34***	1.00	−0.01	0.30***	0.10	0.04	0.03	0.00	−0.01	0.05	0.18*
% Hispanic, 2010	−0.15	−0.10	−0.25**	−0.01	1.00	0.65***	0.91***	−0.74***	0.23**	−0.14	−0.26**	−0.26**	−0.08
% poverty, 1999	−0.02	−0.04	−0.11	0.3***	0.65***	1.00	0.85***	−0.61***	0.22**	0.14	0.06	0.19*	−0.03
% less than high school, 2000	0.03	0.03	−0.15	0.10	0.91***	0.85***	1.00	−0.70***	0.21**	−0.06	−0.17	−0.08	−0.02
% older people, 2000	0.28***	0.20*	0.14	0.04	−0.74***	−0.61***	−0.70***	1.00	−0.22**	0.21*	0.26**	0.3***	0.16
Log of total 2010 population	−0.43***	−0.49***	−0.26**	0.03	0.23**	0.22**	0.21**	−0.22**	1.00	0.03	0.04	0.02	0.02
Jobs per resident, 2006	0.08	−0.05	0.15	0.00	−0.14**	0.14	−0.06	0.21*	0.03	1.00	0.95***	0.90***	0.13
Firms per resident, 2000	0.03	−0.08	0.09	−0.01	−0.26**	0.06	−0.17	0.26**	0.04	0.95***	1.00	0.90***	0.13
Organizations per resident, 2006	0.17	0.03	0.26**	0.05	−0.26**	0.19*	−0.08	0.3***	0.02	0.90***	0.90***	1.00	0.12
Years since first vote	0.17	0.14	−0.22**	0.18*	−0.08	−0.03	−0.02	0.16	0.02	0.13	0.13	0.12	1.00

Note. API = Asian/Pacific Islanders.

p < .10. **p < .05. ***p < .01.

Findings

About 1% of the voting-age residential population of NCs voted in the most recent board election of NCs certified as of May 2008 (Table 3), a level slightly lower than the 4% participation rate in the city’s general elections (Musso et al., 2004). This low-participation rate is consistent with previous findings in Los Angeles (Musso et al., 2004) and other cities (Berry et al., 1993). It also extends previous observations that voter turnout is generally lower at the municipal level than at the state or national level (Hajnal & Lewis, 2003) by suggesting that voter turnout may be even lower at submunicipal levels. Although the overall participation in NC board elections was low, variation across NCs could provide insight into the potential neighborhood-level barriers to participation and the influence of nonresidential stakeholders such as workers and business owners on NC participation rates.

Table 3.

Unweighted Means of Key Variables

		Participation Rate 1^a		Participation Rate 2^a
		(Voters / voting-age residents)		(Voters / [voting-age residents + workers + firms])
	All	Lower participation (<0.66%)	Higher participation (>0.66%)	Lower participation (<0.41%)	Higher participation (>0.41%)
NCs (N)	88	44	44	44	44
Participation Rate 1	1.16	0.33	2.00	0.36	1.97
Participation Rate 2	0.79	0.24	1.35	0.22	1.36
% API, 2010	0.13	0.10	0.15	0.10	0.15
% African American (single race), 2010	0.10	0.10	0.10	0.09	0.11
% Hispanic, 2010	0.46	0.57	0.35	0.54	0.38
% poverty, 1999	0.20	0.23	0.18	0.23	0.17
% less than high school, 2000	0.33	0.40	0.26	0.38	0.27
% older people, 2000	0.10	0.09	0.11	0.09	0.11
Log of total 2010 population	10.45	10.65	10.25	10.64	10.27
Jobs per resident, 2006	0.40	0.29	0.52	0.50	0.30
Firms per resident, 2000	0.06	0.05	0.07	0.07	0.05
Organizations per resident, 2006	0.08	0.06	0.11	0.10	0.07
Years since first vote	3.82	3.66	3.98	3.59	4.05

Note. NC = neighborhood council.

NCs with lower participation had a rate below the median participation rate; NCs with higher participation had a rate above the median participation rate; API = Asian/Pacific Islanders.

Participation rates ranged generally from less than 1% to just more than 2% of residential population (Table 3). To analyze characteristics of NCs by participation level, we define NCs with lower participation as those with a rate below the median-participation rate and NCs with higher participation as those with a rate above the median-participation rate. NCs with relatively lower participation (Participation Rate 1 < 0.66%) and higher participation (Participation Rate 1 > 0.66%) were geographically distributed across the city (Figure 1). The highest participation rates occurred near downtown Los Angeles. The Historic Cultural Neighborhood Council (HCNC) including parts of downtown and Chinatown was an outlier with a Participation Rate 1 of about 11% and a Participation Rate 2 of about 6%, and the Elysian Valley Riverside (EVR) NC to the northeast of HCNC had a Participation Rate 1 of about 10% and a Participation Rate 2 of about 8%.

Consistent with expectations, the means of key variables across these categories suggest participation is related to NC residential demographic and socioeconomic characteristics (Table 3). Consistent with our hypothesized relationship between key variables and participation rates (Table 1), NCs with lower participation tended to have a higher composition of Hispanic residents, but contrary to our expectations, those with higher participation tended to have a higher proportion of Asian/Pacific Islanders (API). NCs with higher participation tended to have operated for more years suggesting they may have a more established base of support or awareness. As expected, NCs with higher participation had lower poverty levels, fewer residents, and a higher percentage of older residents. Contrary to expectations, NCs with higher participation had lower educational attainment. When NC workers/firms were not used in estimating the participation rate (Participation Rate 1), areas with higher participation had more jobs, firms, and organizations per resident, suggesting these stakeholders may influence area-level engagement. When NC workers/firms were used in estimating the participation rate (Participation Rate 2), areas with higher participation had fewer jobs, firms, and organizations per resident.

Multivariate results suggest factors associated with NC election participation after controlling for the influence of other variables (Table 4). Heteroscedascity could occur if variables in the model have different variances. The variances of variables in the model could vary given the size of NCs varies which could result in heteroscedascity. To address this potential problem, the models were weighted by the square root of the number of NC residents; the unweighted results, however, were qualitatively the same as the weighted results. We specified three models for Participation Rate 1 and three models for Participation Rate 2. For each participation rate we generated one model for all NCs (N = 88) and two models excluding the HCNC and EVR NCs which had participation rates more than 5% (N = 86). We included all key variables listed in Table 1 in Models 1, 2, 4, and 5 to examine the influence of each on participation after controlling for other factors. We include Models 3 and 6 to examine patterns after excluding variables with the highest intercorrelations. As indicated in Table 2, variables for NC percentage of Hispanic, percentage of poverty, percentage of less than a high school, and percentage of older people were highly correlated, as were variables for NC jobs, firms, and organizations per resident.

Table 4.

Weighted Least Square Regressions^a

	Participation Rate 1			Participation Rate 2
	(Voters/residents)			(Voters / [residents + employees + firms])
	Model 1	Model 2	Model 3	Model 4	Model 5	Model 6
Independent variables	Coefficient	Coefficient	Coefficient	Coefficient	Coefficient	Coefficient
Intercept	6.867**	6.040**	5.748**	6.428***	5.004***	4.843***
% API, 2010	5.580***	1.927	2.798**	3.067**	1.122	1.292
% African American, 2010	0.271	−0.231	−0.119	0.252	0.047	−0.018
% Hispanic, 2010	−5.050***	−2.858**		−2.800**	−1.731*
% poverty, 1999	−7.062*	−2.801		−4.360*	−2.371
% less than high school, 2000	11.497***	4.251		6.570***	3.033
% older people, 2000	11.874*	−0.065		5.007	−0.318
Log of total 2010 population	−0.797***	−0.482**	−0.540**	−0.665***	−0.413**	−0.437***
Jobs per resident, 2006	0.177	0.093		0.107	0.051
Firms per resident, 2000	0.165	0.572		−0.219	−0.057
Organizations per resident, 2006	−2.592	−4.241		0.228	−1.177
Years since first vote	0.558	−1.007	0.138*	0.594	0.587	0.061
N	88	86	86	88	86	86
Adjusted R²	.42	.11	.13	.35	.08	.09

Note. Dependent variable: neighborhood council voter participation; Weighted by the square root of the number of NC total 2010 population; API = Asian/Pacific Islanders.

Models 2, 4, 5, and 6 exclude two NC areas with participation rates more than 5%.

p < .10. **p < .05. ***p < .01.

Models with the full NC sample performed well (Models 1 and 4) explaining a good portion of the variation in participation rates across NCs (42% and 35%, respectively); models with the reduced sample (Models 2, 3, 5, and 6) explained about 9% to 13% of the variation. Contrary to expectations, we found some evidence that NCs with a higher percentage of API residents or lower NC educational attainment were associated with higher voter participation. Consistent with expectations, having a higher percentage of Hispanic residents, poor residents, and more residents overall was associated with lower participation. Interestingly, variables capturing African American NC composition, area jobs per resident, firms per resident, and organizations per resident were not significant in the models after controlling for other factors. The models which excluded variables with the highest intercorrelations (Models 3 and 6) explain about 8% to 13% of the variation in participation and consistently suggest that NC’s with larger populations have lower participation. Model 3 also provides evidence that NC percentage of API and years since first NC vote were associated with greater participation.

Discussion and Implications

Although the overall participation in NC board elections was low, we found sizeable variation in rates across NCs, and that neighborhood socioeconomic and demographic factors play a significant role in explaining the variation in participation. Racial representation emerged as a major concern in 2007 as the NC system was being evaluated by the city, perhaps in part because NCs were less representative of the city’s Hispanic population than the African American population (Jun & Musso, 2007). The percentage of African American residents was not significant in our final model; this finding is interesting because it differs from the literature on electoral voting, which suggests African American participation is generally lower. The percentage of Hispanic residents was significant and negatively associated with NC election participation after controlling for other factors. Although we do not examine individual-level voting patterns for Hispanics directly, this result may reflect that Hispanic voters are less likely than non-Hispanic voters to participate even when there are fewer legal barriers to participate.

Interestingly, we found some evidence that NCs with a higher percentage of Asian/Pacific Islander residents had higher participation rates. Previous research indicates lower participation rates among Asian Americans in part because of citizenship requirements in most municipal elections (Lien et al., 2001), and that naturalized Asian Americans voted at rates similar to U.S.-born adults. Citizenship is not required for voting in NC elections, and this may promote greater participation among API community residents, workers, and business people. We contacted two NCs to investigate at a cursory level whether areas with a high percentage of NC API residents had significant API NC participation. The HCNC includes parts of downtown and Chinatown and was an outlier in our analysis with a participation rate as high as 11%. Almost half (46%) of HCNC residents were Asian/Pacific Islander according to the 2010 census. Through discussions with HCNC board and staff members and by reviewing NC participant surnames, we found that more than 50% of board members and officers were API and represented business, nonprofit, and residential interests. The Wilshire Center-Koreatown (WCK) NC was about 35% API in 2010 and had a NC board-election rate of about three times as high as the average for all NCs (about 3% for Participation Rate 1 and about 2% for Participation Rate 2); more than 90% of board members and officers were API. Although we cannot conclude from our finding that NCs with higher percentages of Asian/Pacific Islanders means that Asian/Pacific Islander residents were the ones who participated, these patterns based on our cursory-level review of the HCNC and WCK NCs suggest that APIs are actively engaged in NCs in areas with a significant API residential constituency. These patterns are also consistent with some previous analysis of voter registration that suggests geographic concentration of Asian ethnic and immigrant groups can be associated with greater political participation (Cho, Gimpel, & Dyck, 2006). Future research is needed to understand whether Asian Americans are participating at higher levels and, if so, to assess the implications of this pattern for the future of NCs in a region with many immigrant groups.

Interestingly, the presence of nonresidential NC stakeholders such as business owners, workers, and nonprofit organizations was not significantly related to participation after controlling for other factors. Although we do not have sufficient data to assess whether these stakeholders participated in NC elections, this result suggests that business owners who may have an interest in promoting local economic viability, workers who may have an interest in ensuring safety, mobility and convenient amenities near their place of work, and staff and clients of nonprofit organizations who may have an interest in addressing community needs have not in aggregate had a significant influence on increasing or decreasing NC participation rates.

The overall low level of participation in board elections suggests concerted efforts are needed to bolster awareness, interest, and engagement in order for NCs to meaningfully represent local concerns. We have identified several neighborhood factors that influence participation. The city cannot easily alter the demographic and socioeconomic characteristics underlying much of the variance in participation, but it may be possible to work with and through existing neighborhood institutions to mobilize participation. One of the limitations of our analysis is that it does not test for the effects of actions taken by the city. Future research is needed to understand how strategies such as increasing NC funding and outreach, expanding NC authority, and enhancing the relationships between NCs and city-wide governance could affect participation for various stakeholders, and how the success of these strategies differs across communities.

Future research should also investigate the broader implications of NCs for understanding political engagement, given that, unlike traditional elections, noncitizen and nonresident stakeholders can participate in NCs. Our findings suggest that although NC stakeholders faced many of the barriers to participation identified in previous analysis of electoral voting and community engagement, the less restrictive NC participation requirements may foster participation among Asian communities equal to or greater than that of non-Hispanic White and African American communities, which are predominantly U.S.-born.

Finally, future research is needed to assess the implications of the city’s current fiscal crisis on the operation and effectiveness of NCs. In response to budget shortfalls, the Los Angeles City Council recently reduced the staff of the Department of Neighborhood Empowerment which coordinates the NC system by 50% and cut funding to NCs by 10% (Linthicum, 2011; Zahniser, 2010). Such actions could have detrimental impacts on NC election participation if they hinder organizational innovation and improvement and undercut the ability of NCs to conduct outreach, advocacy, and education about local issues and the importance of NCs in municipal governance.

Footnotes

Acknowledgements

The authors thank the Los Angeles Department of Neighborhood Empowerment and the Los Angeles Geographic Information Systems Division for their assistance in obtaining NC election records and boundaries. Megan Emiko Scott’s case studies of two NCs provided valuable insights into the participation of APIs. Leah Brooks of the University of Toronto shared helpful insights on neighborhood patterns and participation, and provided valuable assistance in obtaining information on BIDs. The Haynes Foundation provided support for the initial analysis of community and business group data.

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.

Bios

Douglas Houston is an assistant professor in the Department of Planning, Policy, and Design, University of California, Irvine, United States. His research focuses on neighborhood socioeconomic and environmental disparities, urban air quality, and transportation. His recent publications examine the environmental, land use, and public health implications of near-roadway-vehicle-related air pollution.

Paul M. Ong is a professor in School of Public Affairs and Department of Asian American Studies, University of California, Los Angeles, United States. He has conducted research on immigration, civic and political participation, economic status of minorities, welfare-to-work, health workers, and environmental inequality.

References

Berry

J. M.

Portney

K. E.

Bablitch

M. B.

Mahoney

(1984). Public involvement in administration: The structural determinants of effective citizen participation. Nonprofit and Voluntary Sector Quarterly, 13(2), 7-23.

Berry

J. M.

Portney

K. E.

Thompson

(1993). The rebirth of urban democracy. Washington, DC: Brookings Institution.

Box

R. C.

Musso

J. A.

(2004). Experiments with local federalism: Secession and the neighborhood council movement in Los Angeles. American Review of Public Administration, 34, 259-276.

Brooks

(2008). Volunteering to be taxed: Business improvement districts and the extra-governmental provision of public safety. Journal of Public Economics, 92, 388-406.

Chaskin

R. J.

(2003). Fostering neighborhood democracy: Legitimacy and accountability within loosely coupled systems. Nonprofit and Voluntary Sector Quarterly 2003, 32(2), 161-189.

Chaskin

R. J.

Abunimah

(1999). A view from the city: Local government perspectives on neighborhood-based governance in community-building initiatives. Journal of Urban Affairs, 21(1), 57-78.

Chaskin

R. J.

Garg

(1997). The issue of governance in neighborhood-based initiatives. Urban Affairs Review, 32, 631-661.

Cho

W. K. T.

Gimpel

J. G.

Dyck

J. J.

(2006). Residential concentration, political socialization, and voter turnout. Journal of Politics, 68(1), 156-167.

Costa

D. L.

Kahn

M. E.

(2003). Civic engagement and community heterogeneity: An economist’s perspective. Perspectives on Politics, 1(1), 103-111.

10.

Dekker

van Kempen

(2008). Places and participation: Comparing resident participation in post-WWII neighborhoods in northwest, central, and southern Europe. Journal of Urban Affairs, 30(1), 63-86.

11.

File

(2008). Voting and registration in the election of November 2006. Washington, DC: United States Census Bureau.

12.

Guo

Musso

J. A.

(2007). Representation in nonprofit and voluntary organizations: A conceptual framework. Nonprofit and Voluntary Sector Quarterly, 36, 308-326.

13.

Hajnal

Z. L.

Lewis

P. G.

(2003). Municipal institutions and voter turnout in local elections. Urban Affairs Review, 38, 645-668.

14.

Hays

R. A.

Kogl

A. M.

(2007). Neighborhood attachment, social capital building, and political participation: A case study of low- and moderate-income residents of Waterloo, Iowa. Journal of Urban Affairs, 29(2), 181-205.

15.

Hoyt

L. M.

(2005). Do business improvement district organizations make a difference? Crime in and around commercial areas in Philadelphia. Journal of Planning Education and Research, 25(2), 185-199.

16.

Hymon

(2007, November 27). Chief of L.A. neighborhood agency. Los Angeles Times. Retrieved from http://articles.latimes.com/2007/nov/27/local/me-tharp27

17.

Jun

K.-N.

(2007). Event history analysis of the formation of Los Angeles neighborhood councils. Urban Affairs Review, 43(1), 107-122.

18.

Jun

K.-N.

Musso

(2007). Explaining minority representation in place-based associations: Los Angeles neighborhood councils in context. Journal of Civil Society, 3(1), 39-58.

19.

Kathi

P. C.

Cooper

T. L.

(2005). Democratizing the administrative state: Connecting neighborhood councils and city agencies. Public Administration Review, 65, 555-567.

20.

LeRoux

(2007). Nonprofits as civic intermediaries: The role of community-based organizations in promoting political participation. Urban Affairs Review, 42, 410-422.

21.

Lien

P.-T.

Collet

Wong

Ramakrishnan

S. K.

(2001). Asian Pacific-American public opinion and political participation. Political Science and Politics, 34, 625-630.

22.

Linthicum

(2011, May 19). Los Angeles council approves budget cuts to police, fire and parks in closing a $336-million gap. Los Angeles Times. Retrieved from http://articles.latimes.com/2011/may/19/local/la-me-0519-la-budget-20110519

23.

Marschall

M. J.

(2001). Does the shoe fit? Testing models of participation for African-American and Latino involvement in local politics. Urban Affairs Review, 37, 227-248.

24.

Meyer

Hyde

(2004). Too much of a “good” thing? Insular neighborhood associations, nonreciprocal civility, and the promotion of civic health. Nonprofit and Voluntary Sector Quarterly, 33(3), 77S-96S.

25.

Mitchell

(2001). Business improvement districts and the “new” revitalization of downtown. Economic Development Quarterly, 15(2), 115-123.

26.

Moynihan

D. P.

(1969). Maximum feasible misunderstanding: Community action in the war on poverty. New York, NY: Free Press.

27.

Musso

Weare

Cooper

T. L.

(2004). Neighborhood councils in Los Angeles: A midterm status report. Los Angeles: University of Southern California Urban Initiative.

28.

Musso

J. A.

Weare

Oztas

Loges

W. E.

(2006). Neighborhood governance reform and networks of community power in Los Angeles. American Review of Public Administration, 36(1), 79-97.

29.

O’Brien

D. J.

(1975). Neighborhood organization and interest-group processes. Princeton, NJ: Princeton University Press.

30.

Olson

(1971). The logic of collective action: Goods and the theory of groups. Cambridge, MA: Harvard University Press.

31.

Swindell

(2000). Issue representation in neighborhood organizations: Questing for democracy at the grassroots. Journal of Urban Affairs, 22(2), 123-137.

32.

Verba

Schlozman

Brady

Nie

(1993). Citizen activity: Who participates? What do they say? American Political Science Review, 87, 303-318.

33.

Wong

J. S.

Lien

P.-T.

Conway

M. M.

(2005). Group-based resources and political participation among Asian Americans. American Politics Research, 33, 545-576.

34.

Zahniser

(2010, January 28). Plan to cut L.A. jobs would hurt neighborhood councils and art programs, report shows. Los Angeles Times. Retrieved from http://articles.latimes.com/2010/jan/28/local/la-me-la-budget28-2010jan28