This article uses newly collected data on communities receiving Neighborhood Legal Services Programs (NLSP) grants between 1965 and 1975 to evaluate the impact ofNLSPs on civil disorders and resulting changes in property values in African American communities. We employ several empirical strategies, all of which confirm the NLSP's effectiveness in combatting civil disorders and indicate a robust, positive relationship between NLSPs and property values. We find that NLSP funding increased property values by 2 percent. These results are consistent with a substantial reduction in riotpropensities due to target government funding, and further support claims by the Kerner Commission report that the NLSP mitigated the damage resulting from the civil disorders.


legal services, War on Poverty, Kerner Commission, property values, riots

Starting in 1964, a series of civil demonstrations escalated into widespread uprisings across the United States. One of the many policy responses was to include the Neighborhood Legal Service Program (NLSP) in the War on Poverty. The NLSP was introduced to equip the poor with alternative paths to remediate grievances, particularly those concerning local police misconduct.1 Under the program, lawyers redressed grievances by bringing civil cases against local police departments and other government agencies. The early success of the NLSP was highlighted by the Kerner Commission report (1968), which called for its expansion to reduce the likelihood of future civil disorders. This qualitative assessment has since [End Page 144] been verified, the NLSP found to reduce the number of riots by 3.6 percent and the severity by as much as 56 percent (Cunningham 2018). However, although the NLSP was deemed successful in reducing uprisings, the impact of the program on individuals' economic well-being has not been assessed.

The primary mechanisms through which the 1960s riots undermined the economic well-being of many in the African American community have been a permanent depression in the value of African American properties and worsened labor market outcomes for African Americans (Collins and Margo 2007; Collins et al. 2004).2 Given the substantial reduction in riot propensities found in Cunningham (2018), it is plausible that the NLSP mitigated the damages from civil disorders and violent protests. To examine the possibility, we exploit the variation in timing and location of the establishment of NLSP projects to estimate the causal impact of the NLSP on racial disparities in wealth by focusing on property values.

Our analysis uses recently collected data on the communities receiving legal services grants between 1965 and 1975. We rely on the differential timing of NLSP's implementation in cities across the United States as well as variation in the location and intensity of treatment to imply a causal relationship. In addition, we use the age of local law schools to isolate the relationship between funding and riot propensities, exploiting the fact that neighborhood law firms were frequently affiliated with nearby law schools.

Our results show that legal services reduced both the number of riots and their severity, the effect being substantially larger on severity. Over the long run, the results further show that the NLSP has a robust positive relationship with property values. Using our most conservative estimates, we find that federally funded legal services increased black-owned property values by 2 percent. These findings are consistent with the narrative that the NLSP created access to social justice by providing additional channels for blacks in urban communities to settle disputes. This is consistent with the Kerner Commission report call for the expansion of legal services. Given renewed attention on police-community relations, this analysis contributes to the literature by deepening our understanding of policy initiatives that deal with unresolved community grievances that at times lead to violent rebellions.3


In March of 1876, the first legal aid society opened its doors in New York City (Hollings-worth 1977; Johnson 2014). The German Legal Aid Society, financed through membership dues, provided legal assistance to German immigrants unable to acquire legal counsel. In 1890, the organization opened its doors to all ethnic groups, changing its name to the Legal Aid Society. To meet the needs of the poor, the organization expanded and opened additional branches, one for seamen and another for women. The women's branch dealt with legal needs, largely pertaining to divorce and domestic relations. In response to underwhelming demand by women, the mission of that branch shifted and the office serviced the legal needs of all, but the majority of the clientele were men.

The need for legal aid by women led to the first legal aid office in Chicago. In 1886, the Protective Agency for Women and Children provided legal assistance to young women in areas related to unpaid wages, debt schemes, and licentious employment practices (Katz 1982). In 1888, the Bureau of Justice joined the Protective Agency for Women and Children providing legal aid for Chicago's poor regardless of ethnicity and gender. At the turn into [End Page 145] the twentieth century, legal societies began springing up across the country from Jersey City to Denver. Harvard opened its first legal clinic for the poor in 1913. This massive expansion in legal aid provision led to the first national convention for legal aid societies in 1914 in Chicago and the formation of the National Alliance of Legal Aid Societies, which would later become the National Legal Aid and Defender Association.

By the 1960s, more than two hundred legal aid societies were servicing the legal needs of the poor (Hollingsworth 1977).4 Although legal assistance was free, many services were not provided due to a lack of funds, excess demand, and individual beliefs on the type of services the poor should receive. However, a reliance on charitable donations to fund legal aid threatened lawyers' ability to take on controversial cases, and lawyers turned away cases that would bring negative attention and influence charitable donations (Johnson 1974). Lawyers also refrained from cases involving bankruptcy and divorce, shunned challenging large corporations, and refused to challenge government agencies (Wright 1967; Stumpf 1975; Katz 1982). Furthermore, the lack of funds restricted lawyers' ability to appeal cases. As a result, the tendency was to focus on simple cases or to just provide advice to clients rather than pursue justice. The limitations were not restricted to the kinds of service provision. The location and hours of operation of legal aid offices often prevented the poor from seeking justice (Levitan 1969). Many of these offices were located in the city center, away from the poor. Regular business hours made it difficult for the typical worker to make the commute to the legal aid office and not miss work.

Legal Services and the War On Poverty

The legal aid movement of the 1950s and the early 1960s coincided with social movements that sought the inclusion of marginalized groups into the greater democracy of America. The Kennedy administration and the Ford Foundation, through demonstration projects, financed antipoverty programs to deal with issues resulting from wartime migration into urban areas (Boustan 2010; Hinton 2016). Demonstration programs in New York City, New Haven, Connecticut, and Washington, D.C., included legal agencies to deal with civil and criminal matters the poor often encountered (Johnson 1974). These demonstration programs provided the blueprint for the community-oriented social service programs introduced in President Lyndon Johnson's War on Poverty.

The incorporation of the "civilian perspective" into the War on Poverty was motivated by Jean and Edgar Cahn in the Yale Law Review.5 The Cahns proposed that university-affiliated, neighborhood law firms be established to serve as intermediaries between the community and antipoverty bureaucracies (Cahn and Cahn 1964). Law firms would provide free legal representation in areas related to divorce, eviction, welfare fraud, coerced confessions, arrest, police brutality, and installment buying. The article details their experience and advocates for the development of a nationwide program. A draft of the manuscript was first circulated among colleagues for comment and eventually landed in the hands of Associate Justice Arthur Goldberg. Justice Goldberg was persuaded to send letters to President Johnson advocating for the inclusion of a nationwide legal services program under the War on Poverty (Johnson 1974). As a result, Sargent Shriver brought the Cahns into the Office of Economic Opportunity (OEO) to spearhead the development and implementation of the program. The support of the American Bar Association (ABA) was vital to the rollout of the legal services program. National support from the ABA helped insulate [End Page 146] the program from the attacks of local bar associations when controversial cases were undertaken against powerful entities. On February 8, 1965, the ABA fully endorsed the NLSP.

Following the Cahns' proposal, the Neighborhood Legal Services Program was launched as part of the War on Poverty in 1965. Neighborhood law firms were financed by grants from the OEO and operated under the Community Action Program (CAP). The OEO was responsible for the antipoverty programs, and CAP projects were one of its largest initiatives. The community-based approach created wide variation in how federally funded grants were not only used but received. The OEO grants avoided local and state roadblocks and went directly to community organizations, allowing federal funds to be spent rapidly with wide variation for intended purposes (Johnson 1974; Gillette 1996).

The first year of the legal services program under the OEO resulted in the issuance of more than 155 grants backed by a total budget of more than $20 million. By 1967, the legal services program doubled in size, issuing more than three hundred grants with an annual budget of more than $40 million. By the end of 1967, the federal legal service program was funding 250 projects and providing legal assistance in forty-eight states (Levitan 1969). The rapid expansion of the program was the result of existing legal aid societies' willingness to adopt the neighborhood approach of the NLSP. This included opening legal services offices in poor neighborhoods with nontraditional hours of operation. NLSP agencies were able to provide services in areas that existing legal aid societies were reluctant to handle. This included divorce, bankruptcy, as well as challenging laws, government agencies, and large corporations. Nearly 40 percent of the initial grantees were existing legal aid societies (Levitan 1969). Law schools were just as vital to the roll-out of the program. As stated earlier, the program was designed to take advantage of cheap labor and expertise provided by a local law school. Law schools provided newly trained lawyers to staff legal services offices, designed new curriculum in poverty law, and opened and operated legal services offices in nearby communities.6

Legal Services and Civil Disorder

The neighborhood law firms established by the NLSP were inundated with new clients. In 1968, legal services offices handled 282,000 cases. The typical NLSP lawyer had fifty to one hundred new cases a month, including ten to twenty related to criminal matters and juvenile delinquency (Levitan 1969). Although NLSP attorneys could not represent clients in felony criminal procedures, many criminal issues were related to perceived illegal police activity, thus allowing legal services to serve as a community advocate for the poor and disenfranchised in these circumstances. According to Legal Services Agency Survey of 1970, NLSP lawyers spent approximately 30 percent of their time in community advocacy and educational activities, playing a role in the enforcement of new laws related to police conduct and judicial procedure (Champagne 1974). In addition, approximately 7 percent of NLSP cases were devoted to law reform. These cases challenged local, state, and federal law in areas primarily related to welfare rights (Levitan 1969). Test cases were often brought against police departments, challenging police procedures and practices that negatively impacted the poor (Cunningham 2016).

Legal services lawyers brought lawsuits on behalf of the black community against police departments in Los Angeles, Cleveland, Washington, D.C., and Camden, New Jersey (Los Angeles Sentinel 1971). Advocates of the program claimed that NLSP lawyers were influential not only in reducing police brutality, but also in reducing the likelihood and severity of riots. Legal services lawyers often showed up at demonstrations to ease frustrations and prevent violence. For example, at a Senate subcommittee hearing, NLSP lawyers were lauded for averting violence after a police shooting of residents in Cleveland (U.S. Congress 1969). NLSPs often served as clearinghouses for local complaints [End Page 147] of police brutality. The purpose was to reduce the impulse to partake in violent demonstrations and to help build cases against illegal police behavior. Even in instances when violent protests occurred, NLSP lawyers served the community through city panels and organizations to help local and police officials redress the community's grievances.

NLSP lawyers often represented individuals at the heart of conflicts between police and the black community, such as the taxi cab driver whose physical confrontation with police sparked the 1967 Newark riot (Finman 1971). This led to a federal lawsuit accusing Newark's police department of violating the constitutional rights of black residents and requested a complete overhaul of the Newark Police Department. The success of the program was highlighted by the Kerner Commission report, which called for the expansion of the program as an antiriot initiative (Kerner Report 1968).

Expected Effects of the Legal Service Program

Neighborhood law firms provided representation, consultation, and referrals for the poor at a cost substantially lower than private law firms would. NLSP lawyers were also involved in community organizing and community advocacy. In particular, NLSP lawyers were willing to articulate grievances against state institutions, which was a dramatic shift from previous behavior. The NLSP may have influenced the decision to riot and subsequent outcomes related to civil disorders through several mechanisms. More precisely, legal services should have both direct effects, increases in the number of police complaints (Cunningham 2016; Pedroza 2017), and indirect effects, changes in riot behavior (Cunningham 2018). The direct effect is a result of the indigent using lawyers to access public services, in this case, adequate and nondis-criminatory policing. The indirect effect, however, could improve the relationship between the police and the community (reduce riot propensities) or escalate tension (increase riot propensities).

An improvement in police-community relations may occur if the police become less likely to use excessive force or if citizens are more likely to use the judicial system to resolve conflicts. A statement from the Office of Economic Opportunity at a Senate Subcommittee Hearing in 1969 provides anecdotal evidence to this effect: "Legal services lawyers have won the confidence of angry young men and women and have channeled their grievances into democratic procedures. This capability and achievement mark a major victory for those concerned with maintaining law and order" (U.S. Congress 1969, 102). Conversely, free legal aid and information may embolden citizens into more militant actions. For instance, a police officer accused NLSP lawyers of organizing the protest that escalated into the Newark Uprising of 1967. He criticized NLSP lawyers for emboldening criminals and interfering with the police ability to control the crowd stating, "You can carry a machete through the streets of Newark and not get locked up" due to the presence of NLSP lawyers (Herbers 1967, 24). Police retaliation against the community for filing lawsuits or citizens exaggerating claims of police use of force to NLSP lawyers are also possible. The indirect effects on riot propensities and severities are ambiguous. These claims have been tested, providing evidence that NLSPs reduce riot propensities and severities, but the indirect mechanisms are unclear (Gillezeau 2015; Cunningham 2018).

Given that rioting caused a dramatic decrease in the property values of African American homes and reduced the labor market opportunities for blacks, we could expect the NLSP to have noticeable impacts on these outcomes (Collins and Margo 2007; Collins et al. 2004). Because most individuals accumulate wealth through homeownership, the 1960s uprisings contributed significantly to generational wealth disparities (Toney 2016). The NLSP, either directly (through legal consultation) or indirectly (by reducing riot propensities and severities), should have a positive impact on wealth accumulation through property appreciation.


To test the impact of the NLSP on wealth accumulation of African Americans, we first must establish a link between NLSPs and riots. After establishing that linkage, we then analyze the impact of the NLSP on property values. We follow [End Page 148] the general approach used by Jamein Cunningham to identify the relationship between NLSPs and riot propensities (2018). However, the dependent variable in our analysis is the cumulative funding of an NLSP, not funding in a given year. In our case, the impact of funding does not depreciate after one year, allowing a NLSP to build on existing efforts.

The empirical strategy used to determine the effectiveness of the NLSP as an antiriot policy is a continuous difference-in-difference analysis. The treatment variable of interest is cumulative NLSP funding in millions of dollars.7 For cities in the control group, this variable is always zero (never receive a grant). For those in the treated group, it is initially zero (before treatment) and increases as a city receives federal grants over time. Our analysis accounts for key cross-sectional differences between funded and unfunded cities by controlling for observable demographic characteristics as well as including fixed effects to capture unobserved heterogeneity that is time invariant.8 The dependent variable of interest is the number of riots in a city in a given year or the severity of those riots. The riot severity measure is the total number of arrests, arsons, injuries, and people killed due to rioting in a city relative to the total share of arrests, arsons, injuries, and people killed due to rioting between 1964 and 1971 (see Collins and Margo 2007).

To identify a causal effect, NLSP funding must be unrelated to existing trends in riot propensities. Urban civil disorders, regarding violence and destruction of property within the black community in response to perceived injustices, were relatively rare before 1964. Riots or mobs resulting in interracial conflict occurred before the 1960s. However, civil disorders similar to those that occurred in the 1960s did occur in Detroit and Harlem in the 1940s as well as in Harlem in 1935. Despite riots being relatively rare events prior to the 1960s (Lieberson and Silverman 1965), the NLSP was promoted as an antiriot program and more than likely implemented in communities more likely to explode. If so, this will attenuate the impact of the program on riot propensities.

As a robustness check, we use the age of the oldest nearby law school as an instrument to deal with endogeneity related to the timing, location, and intensity of treatment.9 Law schools were directly related to the implementation and rollout of the program as well as unrelated to existing trends in riot behavior due to their age and the fact that legal clinics mostly came about through NLSP funding (Cunningham 2018; Johnson 2014). Law schools are an ideal instrument because the program was designed to be affiliated with law schools. The exclusion restriction is satisfied by the fact that urban riots occur well after the implementation of legal education through universities in the United States.

We link the NLSP to wealth accumulation by using property values as a measurement of wealth. For robustness, we use two techniques to examine the relationship between NLSPs and wealth. The first approach is a difference-in-difference regression model. The dependent variable is the log of the median residential property value for black homeowners.10 The [End Page 149] treatment variable is cumulative NLSP funding in a city prior to census year 1960, 1970, or 1980. In addition, an indicator variable identifies cities that experienced a high-severity riot and another indicator variable identifies cities that experienced a medium-severity riot. The zero and low-severity group are the reference group.

The second approach follows the specification outlined by William Collins and Robert Margo (2007). Now the dependent variable is the change in the log median residential property value for all black homeowners. Similar to the previous specification, the treatment variable is cumulative funding of the NLSP. We include the riot severity indicator variables. We also estimate a specification that includes a riot severity measure where a city with zero riots or low-severity riots is coded as 0, medium-severity riots as 1, and high-severity riots as 2. The key difference in this specification is the inclusion of regional fixed effects instead of city fixed effects and conducting a cross-sectional analysis. Also the covariates vary by the specification but include total population, the percentage of the population black, the proportion of workers in manufacturing, the change in property value between 1950 and 1960, the crime rate in 1962, and a measure of residential segregation. These covariates are used to replicate and compare our results to those of Collins and Margo (2007).11


Data on the recipients of federal legal services grants were compiled from National Archives Community Action Program (NACAP) files. NACAP provides information on the city, county, and state for which funds were received. These data also include the date the grant was issued, the amount of the grant, the name and address of the grantee, and a brief description of the intended purpose of the grant. Data on property values and other city characteristics are taken from 1960, 1970, and 1980 census city and county data books. The city-level demographic information was constructed by linearly interpolating between 1960, 1970, and 1980 data. Data on civil disorders has been provided courtesy of Collins and Margo (2007) and originally collected by Greg Carter (1986). The Carter data include the location and duration of race riots between 1964 and 1971 as well as the number of people killed, injured, or arrested, and the number of arsons reported or discovered by police during a race riot. Because the riot data contain civil disorders only between 1964 and 1971, we focus on legal service spending and provision before 1972.

The riot data provided by Collins and Margo differ from other sources of race riots (2007). For comparison purposes, Susan Olzak provides detailed accounts of ethnic protests in standard metropolitan statistical areas (SMSAs) during our sample period, 1964 to 1971 (2015).12 Figure 1 highlights the differences in the frequency of protests by events and data source. Carter race riots are represented by the line without a marker (1986), and Olzak by the line with a circle marker.13 The two main differences from these series are the definition of a riot and the source information. According to Carter, a race riot is defined as a demonstration involving at least thirty participants, some of whom must be black, that results in some property damage or violence (1986). In addition, the event has to occur outside of a school setting or an organized civil rights demonstration. Olzak defines a riot as a demonstration involving at least fifty participants, involving some form or act of violence, and lasting longer than two hours (2015). Additionally, Olzak identifies ethnic conflicts using the New York Times and Carter collects information on riots from the New York Times, the Washington Post, Lemberg [End Page 150] Center for the Study of Violence, Congressional Quarterly Almanac, and congressional reports.14

Figure 1. Frequency of Riots by Data Source Source: Authors' calculation based on tabulations from Collins and Margo 2007 and Olzak 2015. Notes: original source of race riots between 1964 and 1971, which are identified by the line without a marker. Race riots are identified by the line with a circle marker. Black protests (including riots) are identified with a triangle marker. All ethnic conflicts recorded by are identified by the line with a square marker. The sample is restricted to capture events between 1964 and 1971.
Click for larger view
View full resolution
Figure 1.

Frequency of Riots by Data Source

Source: Authors' calculation based on tabulations from Collins and Margo 2007 and Olzak 2015. Notes: Carter (1986) original source of race riots between 1964 and 1971, which are identified by the line without a marker. Race riots are identified by the line with a circle marker. Black protests (including riots) are identified with a triangle marker. All ethnic conflicts recorded by Olzak (2015) are identified by the line with a square marker. The sample is restricted to capture events between 1964 and 1971.

Despite the differences in the data collection efforts, we use the Carter data to take advantage of city level variation in the outcome variables of interest (rioting and property values) and the treatment variable (legal services). A SMSA will include multiple cities of interest such as Dallas-Fort Worth. Our estimating strategy takes advantage of the fact that Dallas received an NLSP grant before Fort Worth. This allows Fort Worth to serve as a comparison for Dallas before Fort Worth is treated. Another example is Raleigh-Durham. Durham received an NLSP grant, but Raleigh was not treated. Considering the similarities of these two cities, we use Raleigh as a comparison for Durham. We also use Carter to compare our results with the literature (1986).

The final sample consists of city-level observations of federal legal service funding, riots, property values, and census demographic information for 185 cities. The selection of cities is based on the availability of median black-owned property values in published census tables for 1960, 1970, and 1980. The final sample contains 122 cities that received NLSP grants, the treatment group, and sixty-three nongrant cities, the comparison group. As shown in table 1, treated cities are typically larger, denser, and more affluent. Within these cities, median property values of black homeowners are less than all homeowners. Black-owned property values are higher in treated cities, but these cities experience more rioting and more severe rioting when compared to nongrant cities.


Our results show that NLSP funding is inversely related to the number of riots. Table 2 reports estimates for the effects of NLSP funding on the number of riots and riot severity. Columns 1 and 4 report estimates from the full sample; columns 2 and 5 restrict the analysis to cities that received NLSP grants. Columns 3 and 6 limit the sample to cities included in the Collins and Margo analysis (2007). Last, columns 4 through 6 report estimates using the age of [End Page 151] the oldest law school as an instrumental variable. Column 4 reports estimates from our preferred specification.

Table 1. Summary Statistics
Click for larger view
View full resolution
Table 1.

Summary Statistics

According to our preferred specification (column 4), a $1,000,000 increase in NLSP funding reduced the number of riots by 3 percent. The typical size of an NLSP grant was $200,000, implying a treatment effect of 0.6 percent decrease in rioting due to NLSP.15 Although we exploit the variation in timing and location, columns 2 and 5 provide suggestive evidence that the results are not driven by cities that never received an NLSP grant. Point estimates are typically larger in the NLSP-only sample relative to the full sample, but the results are not statistically different. Estimates using age of the oldest law school as an instrument in panel A are smaller than ordinary least squares (OLS) results but remain statistically significant in columns 4 and 5.

Although the treatment effects in the first panel are small, most of the direct antiriot efforts of poverty lawyers operating NLSPs involved efforts to reduce the severity of riots. Advocates of the NLSP and the Kerner Commission Report viewed poverty lawyers as vital to resolving grievances that lead to civil disorders. Similar to the first, the second panel shows that NLSP funding is inversely related to riot severity. According to column 4, a $1,000,000 increase in NLSP funding reduced riot severity by 48 percent. This implies that the typical NLSP grant reduced the severity of riots by approximately 10 percent, which is a substantial treatment effect. The results are statistically significant in columns 1, 4, and 5. Moreover, results in columns 4 through 6 are at least 50 percent larger than OLS estimates. These results suggest that public officials were more likely to fund a NLSP in more volatile [End Page 152] communities. If so, OLS underestimates the true effect of the NLSP on riot severity.

Table 2. Estimates of Cumulative NLSP Funding on Riot Propensities
Click for larger view
View full resolution
Table 2.

Estimates of Cumulative NLSP Funding on Riot Propensities

Legal Services and Property Values

Table 3 displays the results for the impact of NLSP on wealth accumulation. Columns 1 through 3 refer to the difference-in-difference approach where the dependent variable is the log of the median residential property value for black homeowners. Columns 4 through 7 refer to the cross-section OLS approach where the dependent variable is the change in the log of the median property value of black homeowners. Column 1 excludes NLSP funding to compare results using a difference-in-difference approach with Collins and Margo cross-section analysis. Column 2 includes NLSP funding and column 3 limits the sample to the cities used in Collins and Margo (2007). For comparability, columns 4 and 5 replicate results from table 3B in Collins and Margo (2007), which highlights the impact of riots on black-owned property value not accounting for the endogeneity of riots. Columns 6 and 7 replicate the results from table 6 in Collins and Margo (2007), which uses rainfall in April of 1968 as an instrument for the severity of riots.

Using the difference-in-difference approach, the results in columns 1 through 3 identify an inverse relationship between property values and riots. The results are statistically significant only for black-owned homes in a community that experienced severe riots. Including NLSP funding reduces the negative impact of high-severity riots on black-owned property values.16 According to column 2, high-severity riots reduce property value by 14 log points (12.9 percent). Restricting the sample to the cities in [End Page 153] Collins and Margo reduces the impact of high-severity riots even further than estimates in columns 1 and 2 (2007). The marginal effect in column 3 implies high-severity riots reduce black-owned property values by 10 log points (9.5 percent). The impact of NLSP funding is statistically significant or marginally statistically significant and positive. According to results in column 2, a $1,000,000 increase in NLSP funding would have increased black-owned property values by 1.8 percent. Limiting the sample implies a 2.3 percent increase in property values.

Table 3. Estimates of NLSP and Riots on Log of Median Black-Owned Property Value, 1960-1980
Click for larger view
View full resolution
Table 3.

Estimates of NLSP and Riots on Log of Median Black-Owned Property Value, 1960-1980

Similarly, columns 4 and 5 highlight the inverse relationship between riots and black-owned property values. Column 4 reproduces estimates from Collins and Margo (2007). Column 5 augments their analysis by including NLSP funding. Using the cross-sectional approach, high-severity riots and medium-severity riots are found to have a negative and statistically significant effect. Also, legal services appear to be an important omitted variable in this analysis, which is evident from table 2 as well as column 5 in table 3. Including NLSP funding increases the estimate of high-severity riots by 39 percent. As shown in column 5, high-severity riots reduce property values by 19 log points (17.6 percent), medium-severity riots reduce them by 9 log points (8.9 percent), and NLSP funding increases them by 2 log points (2.2 percent).

Columns 6 and 7 report estimates using rainfall in April of 1968 as an instrument. Column 6 reproduce estimates from Collins and Margo (2007), and column 7 adds NLSP funding to the analysis. As in columns 4 and 5, including NLSP funding dramatically increases the marginal effects of riot severity on black-owned property values, by 52 percent. Also, using rainfall as an instrument increases the size of the treatment effect. According to column 7, a $1,000,000 increase in NLSP is associated with a 4.4 percent increase in black-owned property values.17

We estimate the overall effect of the NLSP [End Page 154] on black-owned property values by using the estimated effects in column 2 of table 3, to predict the counterfactual log of median property values in 1980 for each city. These predicted values are calculated by subtracting the estimated value added due to NLSP funding in treated cities. Using the number of owner-occupied housing units in each city as weights, the weighted average of property values in 1980 across cities is calculated to construct an average counterfactual value of black-owned homes. The weighted average of property values in the non-NLSP counterfactual in 1980 is $10,486. The weighted average of the actual property in 1980 is $12,136. The difference between the actual and counterfactual property values implies an additional $2,139 increase in property value due to NLSP funding. The average number of black owner-occupied housing units in 1980 across cities is 6,394. Using this number, the NLSP is associated with a $10.5 million increase in property values by 1980.


In closing, we reflect on these results in light of the broader questions framing this volume. These questions reflect on the historical success of the government's legal services program in discouraging civil disorders, the kind of progress we have made today, and the lessons that we should take from this work into the future.

What Worked and What Did Not Work?

Together, the results present a compelling story for the importance of the NLSP. The program may have only had a modest impact in reducing the number of riots that occurred in the United States, but spending on the NLSP substantially reduced the severity of the rioting that occurred. This is a particularly important finding given that riot severity has important impacts on urban development and African American outcomes over the long-run. A one-time investment of $1,000,000 in NLSP may have increased black-owned property values by as much as 2 to 4 percent over the long term. This is a substantial and persistent impact that shows just how cost-effective antiriot programs can be. The finding is particularly salient given that riot prevention was not even the primary goal of the program. The findings of long-run mitigating impacts on the accepted economic costs of riots are a further verification that the NLSP was truly an effective program in lessening riot severity.

More broadly, these results suggest that the War on Poverty and many of its component programs may provide a template for future efforts to discourage or mitigate civil disorders. Despite limited evidence that direct income transfers were successful in discouraging riots, this work is part of a growing literature suggesting that programs designed to empower individuals and address grievances were effective in discouraging civil disorder.

How Far Have We Come?

The NLSP is one of several War on Poverty programs that, in some form, survived the dismantling of the Office of Economic Opportunity by the Nixon administration in the form of the independent Legal Services Corporation (LSC) and its depoliticized mandate. However, even in this new form funding has been rocky throughout the decades. In the 1980s, the LSC budget was cut severely, although Democrats successfully blocked efforts to discontinue the LSC. The following decades marked a period of stability for LSC, although the Trump administration sought to eliminate the LSC in the president's 2017 proposed budget. Despite the current political climate, there appeared to be some degree of consensus that the LSC serves an important role even if funding is limited.

More broadly, however, there is still a long way to go. Unaddressed grievances and unequal treatment helped drive the riots of the 1960s, but it is clear that unfair treatment by state institutions, including the police and the judiciary, are driving factors behind the riots that have occurred since 1970. Although publicly funded legal services programs could potentially help address these injustices, other systematic [End Page 155] changes in policing and the legal treatment of officers involved in shootings of unarmed African Americans may have more substantial impacts.

What Are the Implications for the Twenty-First Century?

The results are of particular relevance to the ongoing Black Lives Matter movement. Because grievances are so explicitly grounded in accusations of police abuses, reasons are strong to believe that legal supports designed to challenge local and state institutions could prove an effective mechanism to calm tensions. Given how often existing legal institutions are perceived to favor police officers engaged in shootings of unarmed African American civilians, the case to be made that a more activist form of legal services as envisioned in the 1960s could have a strong impact in lessening the associated protests is a strong one. [End Page 156]

Jamein P. Cunningham

Jamein P. Cunningham is assistant professor of economics in the Fogelman College of Business and Economics at the University of Memphis.

Rob Gillezeau

Rob Gillezeau is assistant professor of economics at the University of Victoria.

Direct correspondence to: Jamein P. Cunningham at, University of Memphis, Fogelman College of Business and Economics, 3675 Central Ave., Memphis, TN 38152; and Rob Gillezeau at, University of Victoria, Department of Economics, 3800 Finnerty Rd., Victoria, British Columbia V8P 5C2.

This project was generously supported by an NICHD center grant to the Population Studies Center at the University of Michigan (R24 HD041028). The authors would like to thank William Collins and Robert Margo for sharing their data.


Ariel, Barak, William A. Farrar, and Alex Sutherland. 2015. "The Effect of Police Body-Worn Cameras on Use of Force and Citizens' Complaints Against the Police: A Randomized Controlled Trial." Journal of Quant Criminology 31(3): 50935.
Boustan, Leah Platt. 2010. "Was Postwar Suburbanization 'White Flight'? Evidence from the Black Migration." Quarterly Journal of Economics 125(1): 417-43.
Cahn, Edgar, and Jean Cahn. 1964. "The War on Poverty: A Civilian Perspective." Yale Law Journal 73(8): 1317-52.
Carter, Gregg Lee. 1986. "The 1960s Black Riots Revisited: City Level Explanations of Their Severity." Sociological Inquiry 56(2): 210-28.
———. 1987. "Local Police Force Size and the Severity of the 1960s Black Rioting." Journal of Conflict Resolution 31(4): 601-14.
Champagne, Anthony. 1974. "An Evaluation of the Effectiveness of the OEO Legal Service Program." Urban Affairs Review 9(4): 466-89.
Collins, William J., and Robert Margo. 2007. "The Economic Aftermath of the 1960s Riots in American Cities: Evidence from Property Values." Journal of Economic History 67(4): 849-83.
Collins, William J., Robert Margo, Jacob Vigdor, and Daniel Myers. 2004. "The Labor Market Effects of the 1960s Riots." Brookings-Wharton Papers on Urban Affairs 2004: 1-46. Accessed April 29, 2018.
Collins, William J., and Fred H. Smith. 2004. "A Neighborhood-Level View of Riots, Property Values, and Population Loss: Cleveland 19501980." Explorations in Economic History 44(3): 365-85.
Cunningham, Jamein P. 2016. "An Evaluation of the Federal Legal Services Program: Evidence from Crime Rates and Property Values." Journal of Urban Economics 92 (March): 76-90.
———. 2018. "The Language of the Unheard: Legal Services and the 1960s Race Riots." Working Paper. Memphis: University of Memphis.
DiPasquale, Denise, and Edward L. Glaeser. 1998. "The Los Angeles Riot and the Economics of Urban Unrest." Journal of Urban Economics 43(1): 52-78.
Downes, Bryan T. 1968. "Social and Political Characteristics of Riot Cities: A Comparative Study." Social Science Quarterly 49(3): 504-20.
Finman, Ted. 1971. "OEO Legal Service Programs and the Pursuit of Social Change." Wisconsin Law Review 1971(1): 1001-1084.
Gillette, Michael L. 1996. Launching the War on Poverty: An Oral History. New York: Twayne Publishers.
Gillezeau, Rob. 2015. "Did the War on Poverty Stop the 1960s Race Riots?" Working Paper. Ann Arbor: University of Michigan.
Herbers, John. 1967. "Newark Jailer Says Poverty Aides Stirred Riots." New York Times, August 8.
Hinton, Elizabeth. 2016. From the War on Poverty to the War on Crime: The Making of Mass Incarceration in America. Cambridge, Mass.: Harvard University Press.
Hollingsworth, Ellen J. 1977. "Ten Years of Legal Services for the Poor." In A Decade of Federal Anti-poverty Programs, edited by Robert H. Haveman. New York: Academic Press.
Johnson, Earl, Jr. 1974. Justice and Reform. New York: Russell Sage Foundation.
———. 2014. To Establish Justice for All. Santa Barbara, Calif.: Praeger.
Katz, Jack. 1982. Poor People's Lawyers in Transition. New Brunswick, N.J.: Rutgers University Press.
Kerner Commission. 1968. Report of the National Advisory Commission on Civil Disorders. Washington: Government Printing Office.
King, Mary C. 2003. "Race Riots and Black Economic Progress." Review of Black Political Economy 30(4): 51-66.
Levitan, Sar. 1969. The Great Society's Poor Law. Baltimore, Md.: Johns Hopkins University Press.
Lieberson, Stanley, and Arnold Silverman. 1965. "The Precipitants and Underlying Conditions of Race Riots." American Sociological Review 30(6): 887-98.
Los Angeles Sentinel. 1971. "Legal Services: Challenging the Power." May 27, A8.
Myers, Daniel J. 1997. "Racial Rioting in the 1960s: An Event History Analysis of Local Conditions." American Sociological Review 62(1): 94-112.
Olzak, Susan. 2015. Ethnic Collective Action in Contemporary Urban United States—Data on Conflicts and Protests, 1954-1992. ICPSR34 341-v1. Ann Arbor, MI: Inter-university Consortium for Political and Social Research [distributor], 2015-03-04. DOI: 10.3886/ICPSR34341.v1.
Olzak, Susan, Suzanne Shanahan, and Elizabeth H. McEneaney. 1996. "Poverty, Segregation, and Race Riots: 1960 to 1993." American Sociological Review 61(4): 590-613.
Pedroza, Juan M. 2017. "Making Noncitizens' Rights Real: The Safety Net, Legal Aid Services, and Immigration Legal Services Fraud." Working Paper. Stanford, Calif.: Stanford University.
Spilerman, Seymour. 1970. "The Causes of Racial Disturbances: A Comparison of Alternative Explanations." American Sociological Review 35(4): 627-49.
Stumpf, Harry. 1975. Community Politics and Legal Services. Beverly Hills, Calif.: Sage Publications.
Toney, Jermaine. 2016. "Is There Racialized Legacy in Wealth Across Generations? Evidence from Panel Study, 1984-2013." Working Paper. New York: The New School.
U.S. Census Bureau. 1960. 1960 Census of Population and Housing. Washington: Government Printing Office. Accessed April 29, 2018.
U.S. Congress. Senate. Subcommittee on Employment, Manpower and Poverty of the Committee on Labor and Public Welfare. 1969. Legal Services Program of the Office of Economic Opportunity, Hearing, November 14. Washington: Government Printing Office.
Wasow, Omar. 2015. "Nonviolence, Violence, and Voting: Effects of the 1960s Black Protests on White Attitudes and Voting Behavior." Working Paper. Princeton University, Princeton.
Wright, Eric W. 1967. "Competition in Legal Services Under the War on Poverty." Stanford Law Review 19(3): 579-92.


1. This approach is consistent with the broader redirection of War on Poverty organizing resources in the Community Action Program towards hearing and addressing grievances from the African American community. The success of this approach is documented (Gillezeau 2015).

2. Beyond that, evidence suggests that the 1960s civil disorders may have hastened the white flight phenomenon from America's urban core and decreased voter participation (Boustan 2010; Wasow 2015).

3. We do not provide an in-depth review of the civil disorder literature. It is robust, largely originating from sociology, regarding the causes of the racialized civil disorders of the 1960s (Spilerman 1970; Downes 1968; Olzak, Shanahan, and McEneaney 1996; Myers 1997) and corrective policy measures (Carter 1987; Wasow 2015; Ariel, Farrar, and Sutherland 2015) as well as economic outcomes (DiPasquale and Glaeser 1998; King 2003; Collins and Smith 2004; Collins and Margo 2007).

4. Most of the growth in legal aid societies occurred in the 1950s. American lawyers and bar associations were energized by Great Britain's federally funded legal aid society. Acting in opposition to a centralized authority in the law profession, bar associations across the country established charitable legal aid offices to meet the needs of the poor. An estimated forty-nine legal aid societies were operating in 1949. By 1961, the number rose to 236 (Johnson 1974).

5. The Cahns operated one of the three neighborhood law firms financed by the Ford Foundation in the early 1960s.

6. Clinical legal education gained attention and the requisite funding through the expansion of legal aid in the 1960s and 1970s through the legal services program. Before the program, only a select few law schools provided legal aid to the poor through legal clinics (Johnson 2014).

7. This variable captures cumulative legal services funding and is lagged by one year. So in census year 1970, this will include NLSP funding through 1969. And for 1980, this will account for NLSP funding through 1979.

8. Demographic characteristics are from the 1960, 1970, and 1980 censuses as covariates. The covariates are the proportion of the nonwhite population, the percentage of the population with more than twelve years of education, population per square mile, median age, and family median income. We include city fixed effects to capture unobserved heterogeneity that varies across cities but is time invariant. We also include region-by-year fixed effects to deal with riot contagion that varies by region and year. The literature on riots highlights the importance of the geographical region of a city.

9. For this variable, zeroes refer to cities without a law school.

10. Property value serves as a measurement of wealth. We use the median property value of black-owned homes for the city due to the fact that riots will have direct impact on the values of homes in rioting communities as well as indirect impact on nearby areas including other black communities in the city. For instance, white flight hastened by the riots would influence the median property value of all homes in a city (Boustan 2010). Second, Collins and Margo show that riots influence the median property value of all homes and black-owned homes in a city (2007).

11. In addition, we use rainfall in April of 1968. Collins and Margo show that rainfall is an important predictor of riot severity (2007). To compare our estimates with theirs, we instrument for riot severity group using the rainfall variable they provide.

12. Olzak documents ethnic conflicts from 1954 to 1992.

13. Olzak race riots are black initiated riots according to the event type and initiating ethnic group (2015).

14. It is quite possible that Carter overstates the number of riots in the 1960s and Olzak understates them (Carter 1986; Olzak 2015). Both sources identify major severe riots during this period.

15. Both Jamein Cunningham (2018) and Rob Gillezeau (2015) find much larger effects of NLSP on the number of riots. However, both analyses rely on different sample periods and sample selection criteria.

16. Legal Services as an antirioting program should reduce the duration and severity of riots and as a result have a positive effect on property values relative to places that are not funded or receive very little funding. Moreover, NLSP should reduce the likelihood of additional riots, which will also have a positive effect on property values. Because legal services are correlated with rioting, the specification that does not include legal services will capture that positive relationship, making the point estimates smaller. Therefore, accounting for legal services should increase (in absolute terms) the coefficient on riot severity, though this is not the case in each specification.

17. NLSP funding is likely correlated with unobserved factors correlated with property values. In regard to property values, law schools no longer serve as a valid instrument due to possible violation of exclusion restrictions. When using age of law school as an instrument for NLSP, the coefficient on NLSP funding and severity group are dramatically larger in magnitude (four times larger for NLSP funding and two times for severity group) but statistically insignifiant.

Additional Information

Print ISSN
Launched on MUSE
Open Access
Back To Top

This website uses cookies to ensure you get the best experience on our website. Without cookies your experience may not be seamless.