Community Revitalization in Post-Katrina New Orleans: A Critical Analysis of Social Capital in an African American Neighborhood

ABSTRACT

This research explored the role of social capital, particularly civic engagement and social trust, in community revitalization efforts in a primarily African American post-Katrina neighborhood (n = 153). Findings reveal high levels of participation in neighborhood and political activities but low levels of social trust. Eighty-four percent of this primarily African American sample reported that they do not trust people of other races as compared to 23 to 32% of African American respondents in the national study. Drawing from critical theoretical perspectives, we offer a critique of the limits of social capital theory as well as a discussion of the importance of building social and racial trust as central components of community development practice. Implications include emphasizing organizational capacity-building activities, community organizing training, and racial reconciliation efforts in post-disaster environments.

KEYWORDS:

• Social capital
• race, community development
• Hurricane Katrina
• disaster
• critical theory

INTRODUCTION

August 29, 2005 was a day marked by a fierce natural disaster in the Gulf Coast, in addition to one of the most destructive technological disasters as seen by the failure of the U.S. Army Corps of Engineers' levee system. The natural disaster occurred first, leaving the marks of a devastating hurricane—destruction of communities and personal property. The technological disaster occurred second and further increased the damage to communities and personal loss. Debates exist on whether the effects of Hurricane Katrina will mimic more of the natural disaster or technological disaster responses. Human influence determines the difference between the two types of disaster. Natural disasters are environmentally occurring conditions such as floods, earthquakes, and tornadoes. Technological disasters are the result of human error, such as collapsing bridges, plane crashes, and toxic poisoning. Human error is a major determinant of the two types, with both producing very different responses in communities (Fischer, 1998). With responses to natural disasters there is no one to blame due to the lack of human influence. However, in technological disasters blame can be assigned to groups and organizations, including governmental institutions; such situations can create (and exacerbate existing) feelings of distrust and anger.

Natural disasters are often followed by a period of community togetherness; residents and volunteers may bond together to rebuild. Technological disasters, with the ability to place blame upon individuals and agencies, often foster distrust and discontent among residents. Much of the damage sustained by New Orleans would not have occurred had the levees not failed, suggesting that the response in New Orleans might mimic more of the responses to technological disasters. The success of disaster recovery appears to be related to community participation (Haines, Hulbert, & Beggs, 1996). However, technological disasters, “have the tendency to weaken relationships in communities” (Rosenfeld, Caye, Ayalon, & Mooli, 2004, p. 165). Distrust of institutions can be a major hindrance toward disaster recovery, in part due to the fracturing of relationships between government agencies and the community. This in turn can hinder progress toward outside recovery efforts. To achieve community redevelopment, successful collaboration between governmental responses and residents of the community is crucial (Cigler, 1999; Farquhar & Dobson, 2004; Mathbor, 2007).

Vulnerable populations often struggle the most in terms of impact and recovery from disasters (Harrell & Zakour, 2000; Morrow & Peacock, 1997; Ozerdem, 2003). In the case of New Orleans, many of the vulnerable populations—communities with lower socioeconomic status and African American populations—received the greatest amount of devastation (Hartman & Squires, 2006). More than 2 years after Katrina, many individuals and business have yet to return to their communities. In addition, research reveals that African Americans continue to be disproportionately affected in health, financial, employment, and housing indicators in post-Katrina New Orleans (Kaiser Foundation, 2007). Hurricane Katrina impacted not just the built environment but also the social environment (Campanella, 2006), and thus the evacuation and dispersal of residents has been a major barrier to neighborhood revitalization post-Katrina. Given that some of the most important work of neighborhood revitalization, including community building and policy advocacy, has been conducted by neighborhood associations, the limited return of citizens presents a problem for areas most affected (Pyles, 2007).

Harrell and Zakour (2000) have highlighted the importance of volunteers and self-help networks in disaster recovery as they can increase the participation of isolated communities. Involvement of communities and the ability of outside responders to understand, collaborate with, and foster this involvement is critical to neighborhood revitalization. Social capital is not only one of the key factors of neighborhood revitalization, but also a major aspect development workers and community organizers need to be cognizant of when assisting with recovery. The recovery efforts in New Orleans take place in a larger context of a city, like many American cities, that has been suffering from blight, neglect, and other social problems, including racism. These social conditions have been coupled with community development efforts to alleviate these problems. Thus, it is appropriate that this research is embedded in the larger community development literature. Jennings (2001) has argued that the current context retrenched public social welfare programs, which has had a negative impact (weakened the fabric) on the capacities of community-based and civic organizations to engage in community and economic development work. Because many local organizations are focused on securing jobs and other basic necessities, the larger community and economic development work has been somewhat ignored recently. There is some evidence that the positive effects of the presence of neighborhood organizations are higher in poor African American neighborhoods than in nonpoor African American neighborhoods (Quane & Rankin, 2006). In all types of neighborhood community development, citizen participation is dependent on mutual trust and cohesion (Ohmer & Beck, 2006). Finally, researchers have recently noted that local community development work is also challenged significantly by global and national political and economic processes; some have argued that such work is better enhanced by partnerships between local and national or global stakeholders (Karger, Iyiani, & Shannon, 2007).

Social capital also appears to play an important role in neighborhood revitalization and community development. Though few studies have inquired into the role of social capital in a disaster context, it is clear that civic engagement and trust would be important in disaster recovery. Therefore, we sought to understand the levels and relationships of social capital in one post-Katrina neighborhood. Thus, the two major research questions of this inquiry were, (a) what are the levels of social capital and well-being in a post-Katrina African American neighborhood, and (b) do relationships exist among these neighborhood revitalization factors? By gaining more insight into these levels, community practitioners are better able to respond to a disaster recovery context through community organizing strategies, community building activities, and organizational capacity building.

After analyzing social capital approaches and introducing critical theory as a guiding framework, we present the methods and results of this study followed by a discussion that includes limitations and implications.

SOCIAL CAPITAL AND CRITICAL THEORY

Social capital has been shown to be associated with positive social outcomes including income, health, happiness, and safety (Putnam, 2000a; Putnam, 2007; Tolbert, Lyson, & Irwin, 1998). It has also been revealed as a factor in a community's capacity to respond to environmental challenges and promote change (Mitchell & LaGory, 2002). Because complex disasters are often the result of unresolved development challenges (Ozerdem, 2003) and low-income neighborhoods are likely to suffer a downward spiral after a disaster (Harrell & Zakour, 2000; Morrow & Peacock, 1997), the levels of social capital are highly relevant in the context of the revitalization of marginalized neighborhoods in a post-disaster context. Research indicates that in the post-Katrina New Orleans milieu, community building and neighborhood organizing have been critical activities to address the needs of citizens in the rebuilding process (Pyles, 2007; Rathke & Laboistrie, 2006). The purpose of this study was to learn about levels of social capital, particularly civic engagement and social trust, in a post-Katrina neighborhood and their relationships to well-being.

Social Capital Theory

The work of Putnam (2000a, 2007) and other social capital theorists has been influential in helping researchers and communities understand the nature of social capital, or the commodity known as community connectedness, and its influence on particular outcomes in a community (Lelieveldt, 2004; Sander, 2002). Putnam (2007) has noted that like tools (physical capital) and training (human capital), community connectedness, or a social network, has a value. He thus conceptualizes social capital as the “social networks and the associated norms of reciprocity and trustworthiness” (p. 137).

Researchers have inquired into a variety of aspects of social capital including political participation, civic participation, religious participation, connections in the workplace, informal social connections, altruism, volunteering and philanthropy, reciprocity, honesty and trust, and social movements (Healy, Hampshire, & Ayres, 2004; Miller, 1997; Putnam, 2000b; Tolbert, Lyson, & Irwin, 1998). These components of social capital tend to be correlated, including civic engagement and trust. High amounts of social capital in a community, Putnam's research has shown, tends to result in better outcomes in education and children's welfare, safe and productive neighborhoods, economic prosperity, health and happiness, participatory democracy, and tolerance.

Putnam (2000a) contends that most types of social capital have been eroding since World War II in the United States, theorizing that this erosion of local capital is the cause of the decline of American democracy. He offers several reasons for the erosion of social capital including geographic shifts (moving to the suburbs), generational changes, and an increase in solitary activities such as watching television. The metaphor that our society is now “bowling alone” rather than participating in group bowling leagues as was done in the past has been a guiding thesis of his work. Putnam's studies of social capital in Italy and the United States focus on the fact that healthy communities have high levels of social capital; however, his research does not explain how to move communities that are unhealthy or have low levels of social capital in the other direction (Miller, 1997).

Putnam (2007) recently has inquired into ethnic diversity and found that diverse neighborhoods in the United States tend to “hunker down” and have less trust (even in one's own race) and lower levels of community cooperation (except in the category of agitation for social reform). Ethnicity and socioeconomic class are two of the strongest predictors of trust in one's neighbors where African Americans, Latinos/Hispanics, individual with low levels of education, and individuals with low incomes are less trusting (Putnam, 2007).

Critiques of social capital have focused on the fact that the types of civic engagement have simply shifted. Some have noted that there is more emphasis on participation in professional associations and less on local citizen-based organizations (Warren, 2001). In addition, engagement in social movement activities has certainly increased since World War II, a fact that Putnam is not really able to explain (McLean, Schultz, & Steger, 2002; Putnam, 2000a). Lelieveldt (2004) studied social capital and neighborhood forms of participation and found that although social capital was shown to be important, the majority of residents' needs focused on the necessity for resources such as money, personnel, and improved social policies. The potential hazard when analyzing social capital data is the belief that increasing social capital is a panacea for community problems. Without confronting power structures and changing policies, the practices of building community and strengthening assets may fall short of remedying inequities. Thus, social capital can be viewed as a heuristic framework or variable to be utilized with caution; it is neither an end in itself nor a sufficient condition for neighborhood revitalization or other positive outcomes. Nonetheless, it can provide meaningful data about an important component of community development.

Critical Theory

In addition to social capital theory, critical theory will be invoked in this article to interpret the findings of the study, as well as to offer some critical insights into the shortcomings of social capital theory. Critical theory is social theory oriented toward critiquing and changing society, in contrast to traditional theory oriented only to understanding or explaining it. Grounded in the Marxist tradition, critical theory attends to sociohistorical context and issues of differential power (Kincheloe & McLaren, 2000). Recent developments in critical theory have drawn from post-discourses emphasizing the social construction of race, gender, and class as a valuable hermeneutic device in social analysis. Thus, the findings of this study will be vetted by considering who wins and who loses in community participation. According to Kincheloe and McLaren (2000), “to seek critical enlightenment is to uncover the winners and losers in particular social arrangements and the processes by which such power plays operate” (p. 281). Additionally, critical theorists contest the notion of the “consent of the powerless” and highlight resistance to oppression in their analysis.

Drawing from the insights and tools of critical theory, we also offer a critical analysis of the discourse of “social capital,” which may silence other discourses related to power and oppression. For example, critics have pointed out Putnam's omission of the realities of the political economy in his concept of social capital. Some have questioned why Putnam does not discuss historic, systemic inequalities as well as policy trends such as neoliberal economic policies when theorizing the decline of democratic participation (McLean et al., 2002). Putnam fails to incorporate this political dimension into his analysis, in particular the exclusion of a political dimension to the concept of trust. Indeed, the recognition that the government may exhibit a corporate agenda over a people-centered agenda quite likely has eroded interest in community participation and social trust. Mohan and Stokke (2000) have pointed out that social capital theories often over-emphasize localism, failing to recognize larger forces that influence local power dynamics. The research study presented here is grounded in the central variables of previous social capital research relevant to the post-Katrina context. These variables include racial trust and community participation. Drawing from similar methodology utilized by other social capital researchers, we present the methods we utilized, including the innovations employed that were appropriate to the context.

RESEARCH METHODS

This research emerged from the work of a group of inter-disciplinary practitioners and researchers (social work, public health, community development, and urban planning) who were focused on developing the capacities of neighborhood associations in post-Katrina New Orleans. This work was convened by and funded by an international relief and development nongovernmental organization (NGO) that appeared on the scene after Hurricane Katrina.

Individuals in a post-disaster neighborhood devastated by levee failures were queried in a face-to-face survey about their current civic engagement activities and their perceptions of community trust during September and October of 2006, approximately 12 to 14 months after Hurricane Katrina. Prior to Katrina, according to the 2000 census, this neighborhood had a population of approximately 14,000 people, primarily African American and more than 50% of the population had incomes under $20,000 per year (http://www.gnocdc.org/orleans). The neighborhood was chosen becomes of its tremendous community redevelopment needs and its strengths in the areas of community activism and organizing. Focus groups were conducted in July 2006 to inform the development of the survey that was used in this study. Two focus groups (n = 11) with neighborhood residents centered on visioning what their revitalized neighborhood would look like as well as the assets and barriers present to achieve the vision. Transcripts from the focus groups were coded for major themes and directly informed the constructs for the survey.

Criteria for participation in the survey included some affiliation with the neighborhood, either living or owning a business there prior to or after the disaster. Sampling strategy was a snowball, convenience sample. The original intention of this research was to conduct a randomized, neighborhood cluster sample. However, after attempting this strategy for several weeks, researchers realized that there were not enough people living in their homes to make it a viable option; thus, a convenience sample was employed. Individuals who self-identified as neighborhood members were instead interviewed at public locations in the neighborhood including community centers, recovery centers, and churches. All but a few individuals who were approached agreed to participate. Participants received a gift card for their participation. The sample size was based on a goal of 150 that was set by the inter-disciplinary team that was determined by several factors including funding limitations and timeline.

The Harvard Social Capital Community Benchmark Survey—Short Form (Putnam, 2000b) was adapted for part of the larger study of which this is a component. Key social capital subdomains and indicators were (a) Political participation, which is indicated by being registered to vote, having voted in the last local election, and having participated in a political meeting or rally; (b) civic associational involvement includes participation in a cultural activity, a parent–teacher group, a neighborhood association, or other recovery group; (c) faith-based engagement includes participation in religious activities other than services; and (d) social and inter-racial trust includes the degree of trust people have in people in their own neighborhoods; the degree of trust in the police; and the amount of trust people have in people of other races. Indicators of individual well-being included single-item questions focused on income, current health status, and levels of happiness.

Descriptive statistics along with correlations—Spearman Rho and point biserial—and chi-squares were conducted to assess the two research questions. Prior to running the analyses, the demographic variable, Living Situation, was dichotomized into FEMA trailer vs. other. Nonparametric correlations (Spearman Rho) were used due to non-normal distributions. Chi-squares were conducted when the purpose of the analysis was used to assess relationships between two dichotomous/nominal level variables.

FINDINGS

A total of 153 individuals participated in the survey. Table 1 shows the frequency and percentages for the demographic variables, where the majority of participants both currently live in the neighborhood (76.5%) and either lived in the neighborhood or owned a business prior to Hurricane Katrina (79.5%). The majority of participants also reported African American or Black racial affiliation (86.9%) and were male (57.6%). The largest proportions of participants lived in a FEMA trailer, were between 35 and 49 years of age, and reported a monthly income from $500 to $999.

TABLE 1. Frequency and percentages of demographic variables

Variable	Category	Freq.	%
Currently live in area (N = 153)	Yes	117	76.5
	No	36	23.5
Currently own a business in area (N = 152)	Yes	26	17.1
	No	126	82.9
Live/own business before Katrina in area (N = 151)	Yes	120	79.5
	No	31	20.5
Living situation (N = 152)	In a home that your family owns	46	30.3
	In a FEMA trailer	61	40.1
	In a rented house or apartment	45	29.6
Gender (N = 151)	Male	87	57.6
	Female	64	42.4
Age (N = 153)	18–34	41	26.8
	35–49	46	30.1
	50–64	43	28.1
	65–74	16	10.5
	75–84	7	4.6
Race/ethnicity (N = 153)	Black	133	86.9
	White	8	5.2
	Mixed	4	2.6
	Other	8	5.2
Monthly income (N = 124)	No Income	11	8.9
	$1–$499	8	6.5
	$500–$999	32	25.8
	$1000–$1499	19	15.3
	$1500–$1999	13	10.5
	$2000–$2499	15	12.1
	$2500 and above	26	21.0

Well-being

Well-being was measured by two items concerning health and happiness. Participants were asked to rate their overall happiness on a scale from 1 (not happy at all) to 4 (very happy). The mean score for happiness was 2.19 (SD = 0.91), suggesting that on average participants were somewhat happy. Participants were asked to rate their health on a scale from 1 (poor) to 5 (excellent). The mean score for health was 2.89 (SD = 1.14), suggesting that on average participants reported good health.

Fifteen nonparametric correlations were conducted to explore the relationships among the well-being variables (Health and Happiness) and demographic variables (Age, Gender, Income, and Living Situation). There was no relationship between demographic variables and the measures of well-being, although there was a correlation between the two measures of well-being, happiness and health.

Trust

Participants were asked three questions regarding the extent to which they trusted people in their neighborhood, the police, and people of other races. Table 2 presents the frequency and percentages of the trust items, where the majority of participants reported very little or no trust at all of people in their neighborhood and of people of other races. Trust items were summed together to obtain a Total Trust score; the minimum Total Trust score (N = 122) was 6 and the maximum was 12 (M = 9.36, SD = 2.08).

TABLE 2. Frequency and percentages of trust items

Item	A lot		Some		Only a little		Not at all
	Freq.	%	Freq.	%	Freq.	%	Freq.	%
People in the neighborhood (N = 146)	18	12.3	25	17.1	50	34.2	53	36.3
Police (N = 136)	34	25.2	30	22.2	49	36.3	22	16.3
Other races (N = 140)	7	5.0	16	11.4	62	44.3	55	39.3

Forty-five nonparametric correlations were conducted to explore the relationships among the trust variables (Police Trust, People Trust, and Racial Trust) and Total Trust with the well-being (Health and Happiness) and demographic variables (Age, Gender, Income, and Living Situation). Results are presented in Table 3, where higher Police Trust ratings and higher Racial Trust ratings are associated with higher Health ratings. Significant relationships were also revealed among all of the trust variables. No other significant relationships were revealed between the trust variables, well-being and demographic variables.

TABLE 3. Nonparametric correlations among trust, well-being, and demographic variables

	Age	Income	Living situation	Happiness	Health	People trust	Police trust	Racial trust	Total trust
Gender	.03	−.10	.13	.10	.01	.06	−.08	−.06	−.01
Age	—	−.01	−.03	− .03	−.02	−.06	−.12	−.06	−.07
Income	—	—	−.09	.05	.09	.09	.09	−.03	.10
Living situation	—	—	—	.03	−.03	−.14	.02	−.10	−.11
Happiness	—	—	—	—	.20 ∗	.03	.02	.12	.05
Health	—	—	—	—	—	.08	.19 ∗	.18 ∗	.18
People trust	—	—	—	—	—	—	.31 ∗∗	.37 ∗∗	.74 ∗∗
Police trust	—	—	—	—	—	—	—	.39 ∗∗	.73 ∗∗
Racial trust	—	—	—	—	—	—	—	—	.79 ∗∗
Note. Significant coefficients are bolded.
∗p < 0.05. ∗∗p < 0.01.

Civic Engagement

Table 4 presents the frequency and percentages of the civic engagement variables, where the majorities of participants were registered to vote and did so in the last local election. Civic engagement items were summed together to obtain a Total Civic Engagement score, the minimum score for Total Civic Engagement (N = 148) was 0 and the maximum was 8 (M = 3.58 SD = 2.08).

TABLE 4. Frequency and percentages of the civic engagement items

Item	Yes		No
	Freq.	%	Freq.	%
Registered to vote (N = 150)	131	87.3	19	12.7
Voted in last election (N = 152)	106	69.7	46	30.3
Political (N = 151)	54	35.8	97	64.2
Religious (N = 151)	65	43.0	86	57.0
Cultural (N = 152)	38	25.0	114	75.0
School/parent (N = 153)	35	22.9	118	77.1
Neighborhood assoc. (N = 153)	69	45.1	84	54.9
Recovery(N = 152)	45	29.6	107	70.4

Eight chi-squares were conducted to assess if significant associations existed between each of the civic engagement items with Gender. Results suggest that none of the Civic Engagement items were related to Gender. Eight chi-squares were conducted to assess if significant associations existed between each of the civic engagement items with Living Situation (FEMA Trailer vs. Other). Results suggest that none of the Civic Engagement items were related to Living Situation.

Sixteen nonparametric correlations were conducted to assess if significant relationships exist among the eight civic engagement Items and Total Civic Engagement with Age, Income, Police Trust, Racial Trust, People Trust, Total Trust, and the well-being variables (Health and Happiness). Results revealed that higher Election participation ratings are associated with higher Age ratings. Higher Cultural Event participation ratings are associated with lower Age ratings. Results also show that higher Income ratings are associated with higher rates of participation in Cultural Event, Neighborhood Association, and a Recovery Organization. Finally, higher Happiness ratings are associated with higher rates of participation in a Cultural Event. No other significant relationships were revealed (see Table 5).

TABLE 5. Nonparametric correlations among civic engagement, trust, well-being, and demographic variables

	Age	Income	People trust	Police trust	Racial trust	Happiness	Health	Total trust
Registered to vote	.15	.16	.03	−.01	−.05	.12	.02	−.04
Voted in last election	.27 ∗∗	.15	−.11	−.11	−.11	.09	−.02	−.13
Political	.03	.15	−.13	.04	−.07	.04	.04	−.07
Religious	.19 ∗	.03	−.13	.00	−.05	.10	−.01	−.09
Cultural	−.29 ∗∗	.21 ∗	−.06	.06	.07	.19 ∗	.10	.05
School/parent	−.07	−.01	−.04	.11	.02	.06	−.02	.01
Neighborhood Assoc.	.25 ∗∗	.22 ∗	−.09	−.12	−.11	.04	.03	−.15
Recovery	.11	.18 ∗	−.11	.16	.05	.06	−.07	.07
Civic Engagement	.13	.20	−.01	−.09	.16	.15	.01	.06
Note. Significant coefficients are bolded.
∗p < 0.05. ∗∗p < 0.01.

Social Capital

The social capital variable was created by the sum of Trust and Civic Engagement. The minimum score for Social Capital was 8 and the maximum was 20; the mean score was 13.03 (N = 119, SD = 2.79). Six nonparametric correlations were conducted to assess the relationships among Social Capital with the well-being and the demographic variables. Results suggest that no significant relationships exist for Social Capital with Happiness (rho= .14), Health (rho = .16); Living Situation (rho = −.09), Gender (rho = .09), Age (rho = .03), and Income (rho = .12).

DISCUSSION

Overall, these findings reveal high levels of civic engagement and low levels of social trust when compared to the Harvard national aggregate survey, in which all comparative data is based on the answers of African American respondents in the Putnam (2000b) study due to the predominance of African American respondents in this sample. Thirty-six percent of the sample reported participating in a political meeting or rally in the last 6 months as compared to the national average of 17% participation rates for African Americans. Forty-five percent of the sample had participated in a neighborhood association as compared to only 25% of a national sample of African Americans. Participation in religious groups, parent–teacher groups, and voter turnout was similar to national findings. The results of this study, including levels of participation in disaster recovery groups and cultural activities, echo recent studies and anecdotal accounts of the post-Katrina environment as one where civic engagement has been highly elevated (Axel-Lute, 2006; Hildebrand, Crow, & Fithian, 2007; Rathke & Laboistrie, 2006). Indeed, these findings mirror the disaster recovery literature that shows that neighborhood associations and other self-help organizations are robust areas of civic engagement in a post-disaster context (Harrell & Zakour, 2000). Kieffer (1984) has argued that a provocative event that offends people's self-interest is necessary for community mobilization; Hurricane Katrina and the subsequent flooding caused by levee failures appears to be such an event.

Some of the high levels of civic engagement activities in this study are likely focused on rebuilding ties in neighborhoods. These ties include the cultural activities that are correlated with income and happiness. Other types of activities are focused on holding officials accountable to implement positive disaster recovery policies, and advocating for access to policy provisions, such as the Road Home program, relevant to rebuilding the neighborhood. In the early days after the storm, citizens came to believe that the city was proposing to turn many neighborhoods into “green space” in the new city plans. Outraged by this prospect, a movement of neighborhood associations emerged demanding a “bottom-up” approach to city planning rather than a “top-down” approach, with the voices of neighborhoods at the forefront (Pyles, 2007). The high participation rates in this study not only reveal the tremendous strength and resilience of people in the neighborhood, but also exemplify a resistance to conditions and policies that have worked against this African American sample.

A large proportion of the individuals in this sample had extremely low incomes. In a study conducted at roughly the same time as this survey, 66% of individuals surveyed in New Orleans reported that their financial situation was worse after Katrina (Kaiser, 2007). To be sure, this neighborhood prior to Katrina had very low income levels, so it is unclear as to what impact Katrina might have had on people's income levels (a question that is beyond the scope of this study). This study did reveal, however, that income levels are related to civic engagement activities. In other words, the financially vulnerable populations are less able to participate in their neighborhood recovery than those with more resources.

Social Trust

Findings about social and inter-racial trust reveal egregiously low levels of trust compared to national aggregate data (Putnam, 2000b). Though African Americans do tend to have lower levels of trust according to Putnam (2007), 70% of the respondents in this study reported that they did not or only slightly trusted people in their own neighborhood as compared to the national results that indicate 17% of African Americans do not trust people in their neighborhood. Fifty-two percent of this sample do not trust (only a little or not at all) the police as compared to 18% of the national sample of African Americans. Eighty-four percent do not trust people of other races (only a little or not at all) as compared to 23 to 32% of African American respondents in the national study. This is a unique finding and is commensurate with what the literature on technological disasters says about low levels of trust (Rosenfeld et al., 2004). What is curious about this data, however, is that according to Putnam's theory, civic engagement and trust tend to be correlated and yet in this instance these variables are clearly not.

The results confirm Putnam's (2000a) findings that trust is positively associated with health. Thus, the lower levels of racial trust, for example, have a negative correlation with people's ratings of their health. One interpretation could be that the experience of racism is hazardous to one's health. We posit several reasons, grounded in critical theory, for the low levels of social and inter-racial trust that were revealed in this study of a primarily African American neighborhood: (a) the pre-Katrina historical racism in the South and particularly in New Orleans as manifested in systemic racism via education, housing, and employment (Mann, 2006); (b) the ethnic trauma experienced by thousands of African American New Orleans residents at the Superdome and Convention Center, as well as through forced evacuation to cities across the United States (Pyles, 2006); and (c) post-disaster community rebuilding policies that have discriminated against low-income African American citizens in the arenas of public housing, health care, and education.

Social capital is a multifaceted notion; trust, however, is the most central facet, according to Putnam. The reciprocity that arises from such trust is indeed what contributes to favorable outcomes over time. Without trust, it is possible that the positive effects of civic engagement activities are merely an empty shell without the necessary substance of trust. Because of a lack of trust within the neighborhood, community healing and the capacity of the neighborhood to recover are seriously impaired. On the other hand, it may be the case that for some populations, particularly marginalized communities such as the one detailed here, low levels of trust actually propel people to become more civically engaged.

Limitations

This research is limited by several methodological factors: use of a nonrandomized sample, a cross-sectional design, a lack of pre-Katrina baseline data, and no comparison to White neighborhoods. The sample is arguably representative of the neighborhood (Louisiana Health and Population Survey, 2006). Nonetheless, the use of a randomized sample would give a clearer picture of responses in this neighborhood. Because the study is cross-sectional, it is impossible to truly evaluate the nature of the relationships between significant variables. Although it is possible to compare social capital variables to national data, there is no existing baseline data for this neighborhood about social capital and thus it is not completely clear whether the frequencies of civic engagement activities, for example, are higher than pre-Katrina levels. It should also be noted that the lack of information in this study about the quality of civic engagement limits our interpretation. That being said, additional analyses of this data including multivariate analyses could reveal more nuanced relationships that are not explored in this article.

An additional limitation exists that is more substantive in nature and it concerns the fact that we do not ask respondents about political trust, i.e., “do you trust the government?” Indeed, the political dimension of trust is silenced in Putnam's work and this omission has been a source of vigorous criticism of his theory (Boggs, 2002). In fact, critics have argued that it is the decline of political trust, not social capital, that has caused the most problems in American life. This deterioration is likely due to factors related to globalized corporatization, particularly the rise of corporate power and its ultimate effect on people's perceptions of trust in the government. Clearly, an understanding of the levels of trust of the government in the wake of a technological disaster caused by failures of public infrastructure would have been of central interest to this study.

CONCLUSION

Transforming racial inequities requires deep commitments to consciousness-raising and other forms of empowerment of the most vulnerable citizens of a community. Teaching community practitioners and citizens about racial reconciliation from an empowerment perspective that focuses on social constructions of racism could be a useful way to address these issues (Gutiérrez, 1995; Witkin & Saleebey, 2007). These activities can be coupled with community development activities focused on culture and the arts, which have shown to leverage a community's assets into community power (Kretzmann & McKnight, 1997).

Despite the fact that we do not ask respondents about their trust levels of the government, it is quite possible that these trust levels are impacted by an overriding distrust of the government. If this is the case, engaging in federal, state, and local policy advocacy that holds policymakers accountable would be of utmost importance. Indeed, it appears that this sample is doing just that. Also, disaster recovery policies can acknowledge and be responsive to preexisting social inequities. Educating the community about the possibilities and problems with policies such as the Stafford Act and the United Nations Convention on Internally Displaced Persons could be a critical capacity-building task. Thus, the importance of organizational capacity-building activities, community organizing training, and racial reconciliation efforts can be emphasized by community practitioners working in post-Katrina New Orleans and other similar contexts. Neighborhood groups and other politically oriented advocacy groups are often in need of training in organizational capacity-building, asset-based community development, and in community organizing and advocacy strategies. If taught such topics in their educational programs, social workers could become vital resources as community organizers and supporters of neighborhood associations and other community development related organizations.

High levels of civic engagement are indeed critical to a well-functioning democracy but they may not be enough to confront long-standing social inequities. Putnam's theory of social capital fixates on local networks and may be silencing the realities of the political economy, inadequate social welfare policies, and sociohistorical context (Mohan & Stokke, 2000). Thus, increasing social capital should not be viewed as a cure-all for neighborhoods and communities in need of revitalization.

The authors would like to acknowledge and thank the community members who participated in this research, Mercy Corps, who provided funding and support, and Jennifer Day and Erica Dudas who provided research assistance.