Bureaucratic Effectiveness and Civil Rights Enforcement

Abstract

This article compares federal, state, and local civil rights agencies’ effectiveness in enforcing the Fair Housing Act. Two factors primarily define effective enforcement: whether agencies’ conciliation efforts are more likely to lead to agreements between the parties involved in complaints and whether agencies are more likely to provide remedies to complainants in cases in which there is cause to believe discrimination occurred. The analysis shows that state and local agencies are generally more effective than the U.S. Department of Housing and Urban Development (HUD) both at conciliating complaints and at providing remedies. HUD does appear to be more effective than state and local agencies in terms of the dollar amount of monetary relief awarded when successful conciliations occur, but HUD’s remedial effectiveness disappears after controlling for the likelihood of successful conciliations.

Keywords

federalism bureaucratic effectiveness fair housing HUD

Discrimination and segregation have long characterized the American housing market. In recent years, although the most flagrant forms of discrimination have decreased, substantial levels have continued in housing rentals and sales (Turner et al. 2002; Turner et al. 2013). In addition, residential segregation persists despite declines since the 1980s (e.g., Lichter, Parisi, and Taquino 2015; Logan and Stults 2011; Massey and Tannen 2015). Because these problems continue, scholars have recommended improvements in fair housing enforcement (Lamb 2005; Massey and Denton 1993; Oliver 2010). One important improvement would be to increase the effectiveness with which civil rights agencies at all levels of government resolve housing discrimination complaints and provide remedies for victims of discrimination.

Scholars have explored numerous aspects of governmental effectiveness (e.g., Carpenter 2001; Scholz 1991; Scholz and Wei 1986; Wilson 2000), even though defining and measuring the concept has posed problems (Cameron and Whetten 1983; Lee and Whitford 2009). Conceptions of effectiveness often emphasize the success or competence of government actors or institutions in achieving their most fundamental objectives (Gilmour and Lewis 2006; MacDonald and Franko 2007; Wolf 1993), recognizing that bureaucratic effectiveness varies across government agencies and levels of government (Carpenter 2001; Ingraham, Joyce, and Donahue 2003; MacDonald and Franko 2007; Wolf 1993). Taking these ideas into account, this article focuses on bureaucratic effectiveness in enforcing the Fair Housing Act of 1968, commonly known as Title VIII. The approach is unique in providing a longitudinal comparison of federal, state, and local agencies’ success in enforcing the same policy in contrast with research that typically compares federal versus state success in policy implementation. This article examines a basic question in federalism: when a national program can be enforced by federal, state, and local governments, does the level of government at which the remedy is pursued result in different outcomes? This research here compares civil rights agencies at each level of government to determine which operates most effectively. This leads to the fundamental question of how to define effectiveness within this intergovernmental context.

According to the U.S. Department of Housing and Urban Development (HUD 1998), “The critical measure of the effectiveness of a [fair housing enforcement] program is how much it actually helps the people it is designed to assist” based on “the number of cases that it conciliates and efforts to get compensation for housing discrimination victims” (p. B-24; see also HUD 2009). This article focuses on two key variables in order to conceptualize and measure bureaucratic effectiveness in Title VIII enforcement: whether civil rights agencies’ conciliation efforts lead to an agreement or settlement between the parties involved in a complaint and whether a remedy is provided. A rich comprehensive Title VIII data set obtained from HUD through a Freedom of Information Act request (HUD 2005) permits an analysis of this cooperative federalism program (e.g., Agranoff 2001; Elazar 1962). These data allow conceptualization of the notion of effectiveness as defined by HUD and measure it in terms of how well federal, state, and local civil rights agencies carry out their conciliation and remediation responsibilities. For example, because effectiveness may be evaluated by determining which level of government is most successful at conciliating Title VIII complaints, first-attempt success rates provide a basic measure of how effective agencies are at conciliating complaints. Agencies that conciliate in their first attempt are more effective than agencies needing multiple attempts.

The article proceeds as follows. The first section describes fair housing enforcement in the American federal system and the concept of substantial equivalency followed by theory, hypothesis, and data for measuring effectiveness in Title VIII enforcement at the federal, state, and local levels. Bureaucratic effectiveness will be evaluated in terms of conciliation success and remedies provided in Title VIII enforcement. The final section states conclusions and suggests implications.

Intergovernmental Enforcement

Various states and localities passed fair housing laws before and during the 1960s (Collins 2006). Yet the most pathbreaking changes came at the federal level in the Fair Housing Act of 1968 and, two decades later, the Fair Housing Amendments Act of 1988. The 1968 Fair Housing Act forbids all of the following on grounds of race, color, religion, sex, and national origin: (1) discrimination by refusing to rent or sell housing after a bona fide offer is made; (2) discrimination in the terms, conditions, or privileges of rentals or sales in housing; (3) advertising for the rental or sale of housing that indicates a preference, limitation, or discrimination; (4) misrepresenting the availability of housing for rent or sale; (5) discrimination in loans for purchasing, constructing, improving, or repairing housing; and (6) discrimination in access to or membership in multiple listing services or organizations for real estate brokers. The Fair Housing Amendments Act of 1988 expanded fair housing protections to include persons with disabilities and families with children. It also strengthened enforcement by creating administrative law judges (ALJs) at HUD who decide, when conciliation fails, whether there is reasonable cause to believe discrimination occurred (see Seabrook, Wilk, and Lamb 2013).

HUD has been the lead federal agency in the fair housing enforcement effort ever since Congress passed the Fair Housing Act. HUD implements federal fair housing policy by investigating, conciliating, and closing Title VIII complaints. State and local civil rights agencies that assist in Title VIII enforcement under HUD’s Fair Housing Assistance Program (FHAP) use a similar process (Bullock, Lamb, and Wilk 2017; Bullock, Wilk, and Lamb 2015). In fact, the 1968 legislation created a cooperative federalism arrangement: state and local governments that pass housing discrimination laws substantially equivalent to Title VIII in terms of rights, procedures, remedies, and the availability of judicial review have the first opportunity to enforce Title VIII, ahead of the federal government (Lamb and Wilk 2010; HUD 2014). FHAP, initiated in 1979, provides federal funds to state and local governments that have a substantially equivalent law. At the close of fiscal year 2016, HUD recognized thirty-five states and fifty localities as having substantially equivalent laws (HUD 2017), but the number of FHAP agencies has fluctuated through the years because HUD may certify or decertify them for various reasons.

The first step in this intergovernmental enforcement process requires that the alleged victim of discrimination file a Title VIII complaint. The agency that receives the complaint then determines whether it has jurisdiction. If the agency lacks jurisdiction, the complaint is referred to the appropriate federal, state, or local agency, and the respondent is then notified that the complaint has been filed. At that point, the agency investigates the complaint by interviewing the complainant, the respondent, and other relevant witnesses. The enforcement agency then attempts conciliation by helping the parties reach an agreement in the case (see HUD 2008).

Conciliation

The conciliation process is crucial to effective enforcement, and civil rights agencies at each level of government have an obligation to try to conciliate fair housing complaints that fall within their jurisdiction. Whether a particular complaint is processed by HUD or by a state or local agency is not random; if there is both a state and local FHAP agency where the alleged discrimination occurred, the complainant may initially file a complaint with either HUD or the local or state agency. If the complainant files with HUD under these circumstances, HUD sends the complaint to the local agency for processing. Complaints filed with the local agency get processed there. The state agency processes the complaints it receives. If there is no substantially equivalent agency in the geographical area where the discrimination allegedly occurred, then HUD has jurisdiction. Conciliation is not always achievable and HUD in particular usually needs multiple attempts to conciliate.

Conciliation efforts continue until the parties reach an agreement and the case is closed or until a determination has been made concerning whether reasonable cause exists to believe that the respondent violated Title VIII. The case is closed if the enforcement agency finds no reasonable cause to believe that discrimination occurred, although the complainant retains the right under Title VIII to file suit in federal district court. However, if the enforcement agency concludes that reasonable cause does exist, legal action is taken before an ALJ or in a federal district court (e.g., HUD 2014, 2017).

Remediation

In addition to successful conciliation, a second critical indicator of effectiveness involves compensating victims of housing discrimination. Conciliation agreements provide a remedy to the complainant or an ALJ can order a remedy based on a finding of reasonable cause (HUD 2014, 2017). Remedies fall into two broad categories: monetary relief and other forms of compensation, specifically the provision of the housing unit that was originally denied to the complainant (or a comparable housing unit), or some form of affirmative action on the part of the respondent so others will not experience discriminatory treatment in the future. These two general categories of compensation furnish this research with three measures of remedial effectiveness: whether monetary relief was awarded, the amount of monetary relief, and overall relief if both money and a housing unit or affirmative action were part of the remedy.

Why FHAP Agencies May Be More Effective than HUD

A general theory of bureaucratic effectiveness in fair housing enforcement should consist of at least three basic principles (see, e.g., Carpenter 2001; Scholz 1991; Scholz and Wei 1986). First, a law’s objectives are more likely to be accomplished if effectively enforced. Second, bureaucratic enforcement agencies responsible for receiving and handling complaints are effective at carrying out those duties if they are more likely to process and close complaints successfully and to provide remedies where appropriate. Third, law enforcement agencies in charge of prosecuting violators are effective at carrying out those responsibilities if they conduct investigations and file lawsuits when they have reasonable cause to believe the law was violated.

Focusing on the second principle here, one might think that the Fair Housing Act’s substantial equivalency requirement would result in little or no difference in how effectively federal, state, and local agencies enforce Title VIII because essentially the same legal rules and enforcement procedures apply to each level of government. As the lead federal agency, HUD’s rules and procedures are spelled out by the Fair Housing Act and administrative regulations, and the same rules and procedures apply to FHAP agencies under Title VIII’s substantive equivalency requirement. To be certified by HUD as a FHAP agency, a state or local government must pass and enforce a fair housing law that provides the same rights, procedures, remedies, and availability of judicial review as required in federal enforcement. However, for reasons set forth below, it is reasonable to hypothesize that several differences between HUD and FHAP enforcement will result in state and local agencies being more effective than HUD.

First, FHAP agencies receive US$2,400 for each Title VIII complaint investigated and closed (Wilk and Lamb 2011). Moreover, HUD makes additional funds available if FHAP agencies provide satisfactory final investigative reports or perform well in conciliating complaints. An extra US$800 has been available for each complaint closed by a FHAP agency under these two circumstances (U.S. Commission on Civil Rights 1994). Because state and local civil rights agencies are typically underfunded, HUD’s financial assistance may constitute a meaningful portion of a FHAP agency’s budget. By contrast, Congress provides no financial incentive for HUD to investigate and close complaints. Not surprisingly, then, HUD has long had a large case backlog (e.g., U.S. Commission on Civil Rights 1994).

Second, Title VIII places an important time constraint on both HUD and FHAP agencies: they have only 100 days to complete the investigation of a complaint after it is filed unless, in the words of the Fair Housing Act, “it is impracticable to do so.” HUD strictly adheres to this 100-day time limit (U.S. Government Accountability Office 2005), but the executive branch and Congress have not held HUD to the same standard. The U.S. Commission on Civil Rights (1994) found that between 1989 and 1993, HUD took an average of 118, 184, 206, 140, and 151 days, respectively, to process complaints. Similarly, based on 2004 data, the Government Accountability Office (2005) estimated that HUD took longer than 100 days to complete investigations in 98 percent of all complaints it closed unless there was a finding of reasonable cause. This 100-day requirement should encourage FHAP agencies to carry out conciliation and remedial responsibilities more swiftly than HUD.

Third, HUD has the authority to decertify FHAP agencies if they fall short of their substantive equivalency requirements (U.S. Commission on Civil Rights 1992, 1994). This possibility may place additional pressure on FHAP agencies to maintain high performance standards, including those for conciliations and remedies. Conversely, little pressure appears to be placed on HUD to enforce Title VIII more effectively (National Commission on Fair Housing and Equal Opportunity 2008; U.S. Commission on Civil Rights 1994). Executive oversight primarily comes from HUD’s Office of Inspector General, whereas legislative oversight is largely vested in the Senate Committee on Banking, Housing, and Urban Affairs. However, these principals do not regularly focus on HUD’s Title VIII enforcement, much less conciliations and remedies.

Other considerations also suggest that state and local enforcement may be more effective than HUD’s. Requiring that state and local governments demonstrate that they are prepared and able to enforce national standards probably increases the likelihood they will faithfully carry out federal civil rights laws. FHAP agencies that make a conscientious effort to establish and sustain substantial equivalency status over time would be expected to function effectively in conciliating complaints and providing remedies. In addition, substantially equivalent agencies may often be more knowledgeable than HUD officials about housing conditions and patterns of discrimination in their states and localities, and this knowledge, in turn, could affect conciliations and remedies (Schwemm 2017; U.S. Commission on Civil Rights 1992, 1994). Finally, generally speaking, state and local agencies may be more effective at enforcing national legal requirements because they are smaller and more flexible (Scholz and Wei 1986).

The Title VIII data set (HUD 2005) permits an examination of conciliations and remedies. It contains information for over 120,000 complaints filed with HUD and FHAP agencies between 1989 and 2004. Each individual complaint serves as the unit of analysis. Over 55,000 of these complaints entered the conciliation process. For each complaint, a number of variables are available, such as whether the complaint was conciliated successfully, the number of days it took to reach conciliation, the number of conciliation attempts made for each complaint, the agency responsible for processing the complaint, the type of discrimination alleged by the complainant, and the issue involved. These data allow us to assess bureaucratic effectiveness in enforcing Title VIII’s version of cooperative federalism. Our methods for doing so are explained in the following sections.

Conciliation Effectiveness

The theory of cooperative federalism is anchored in a tradition of cooperation among different levels of government that share power to resolve similar problems (e.g., Elazar 1962; Grodzins 1966). Which level of government deals most effectively with housing discrimination complaints? The first measure of effectiveness, conciliation success, looks at whether a federal, state, or local civil rights agency achieves the most successful conciliation rate, regardless of which party might be considered the winner. Failure to achieve conciliation represents ineffectiveness. The dependent variable is coded 1 for effective and 0 for ineffective outcomes. Table 1 presents the descriptive statistics for all effectiveness indicators.

Table 1.

Descriptive Statistics for Effectiveness Measures, 1989–2004.

	HUD	State	Local
Conciliation process
Number of complaints entering the conciliation process (N)	(31,274)	(9,445)	(5,351)
1. Conciliation success
Percentage of conciliations with a resolution	34	36	44
2. First attempt success
Percentage of conciliations resolved on first attempt	17	81	64
Conciliation remedies
3. Monetary relief
Percentage successful conciliations resulting in monetary relief	53	54	47
4. Monetary amount
Median monetary relief provided in successful conciliation	US$950	US$760	US$500
5. Overall relief
Percentage successful conciliations providing relief (money, housing, other)	71	67	60
Conciliation remedies controlling for failure
6. Monetary relief
Percentage conciliation attempts resulting in monetary relief	22	26	25
7. Monetary amount
Median monetary relief provided in cases where conciliation was attempted	US$875	US$850	US$500
8. Overall relief
Percentage conciliation attempts providing relief (money, housing, other)	32	33	33
Cause remedies
9. Monetary relief
Percentage cause determinations resulting in monetary relief	1	21	15
10. Monetary amount
Median monetary relief provided in cause determinations	US$6,750	US$4,000	US$3,150
11. Overall relief
Percentage cause determinations providing relief (money, housing, other)	2	27	18

Note: HUD = U.S. Department of Housing and Urban Development.

Of 55,583 complaints that entered the conciliation process, 36.4 percent were successfully resolved. On its face, this seems rather low, but agreement is often difficult when one party is alleged to have discriminated. Comparing conciliation success across different levels of government, the first row of Table 1 reveals that local civil rights agencies have the highest conciliation success rate, 44 percent. In comparison, HUD has the lowest success rate (34 percent), which is only slightly worse than state civil rights agencies’ 36 percent success rate.

Since it may take an agency several attempts to reach a conciliation agreement, achieving an agreement on the first attempt demonstrates greater bureaucratic effectiveness. The second measure of effective conciliation focuses on complaints successfully resolved on the first attempt. This variable is coded as 1 if an agreement was reached on the first attempt and as 0 if more attempts were required.

The first attempt success variable examines only complaints that were successfully conciliated, not the total number of complaints in which conciliation was attempted. The second row of Table 1 shows a large discrepancy across the levels of government. HUD succeeds at conciliation on its initial attempt in 17 percent of conciliated cases. In 83 percent of the complaints resolved through conciliation, HUD had to make multiple attempts. HUD’s performance dramatically lags behind state civil rights agencies, which successfully conciliate on the first attempt 81 percent of the time. Local agencies also outperform HUD with 64 percent of the conciliations achieved in the initial negotiation.

The first assessment of effectiveness estimates the likelihood of successful resolution by conceptualizing effectiveness using two independent variables. One model has successful resolution as the dependent variable; in the second model, resolution on the first try serves as the dependent variable. The main independent variables of interest are the processing agency dummies, which indicate whether a state or local agency handled the complaint; HUD-processed complaints serve as the base category. Another dummy variable, returned, involves complaints returned from FHAP agencies to HUD for any number of reasons. For example, Title VIII gives state and local agencies 100 days to resolve complaints; because returned complaints take more than 100 days for FHAP agencies to close, some of them may be more complex than the typical Title VIII complaint. Failure to control for returned complaints could, therefore, result in the appearance that HUD’s complaint processing is less effective than it is in actuality.

Control variables that should theoretically affect outcomes are taken into account. These include the issue(s) cited in a fair housing complaint and the type(s) of discrimination claimed. The types of discrimination include race, disability, family status, gender, color, national origin, and religion. The dummy variable trap does not apply to either issues or discrimination type because more than one issue or type of discrimination can be alleged in each Title VIII complaint; the categories are not mutually exclusive. The dummy variable trap refers to situations where dummy variables are created for categories that are mutually exclusive. This prevents the reader from making a direct comparison to the base category whose effect is captured by the intercept. Reflecting the provisions of Title VIII described earlier, the possible issues could include terms and conditions, advertising, financing, refusal to rent, refusal to sell, coercion, and false representation, among others.

The degree of segregation in the geographical area where a complaint was filed is also included in the models. Farley and Frey (1992) develop a continuous measure that ranges from 0 (no segregation) to 100 (complete segregation). The variable included in the models here is dichotomous, coded as 1 in counties with a score of .60 or higher and as 0 in all other areas. This variable is expected to negatively impact effectiveness. A final control variable takes those complaints that allege a pattern or practice of discrimination into account. These cases are considerably more complex than most Title VIII complaints and generally require much more time to resolve (Schwemm 2017).

A fixed effects approach is employed with respect to time and location. Dummy variables capture the year the complaint was filed, and additional variables indicate the complaint’s state of origin. To avoid the dummy variable trap, 1989 and Alaska are the excluded categories. However, for ease and focus in interpreting the variables of interest, the coefficients for years and states are not presented.

Since the dependent variables are dichotomous, logistic regression is the appropriate technique. Table 2 presents the results with the conciliation success model in column (a) and the first-attempt model in column (b). The logit coefficients are presented along with robust standard errors. The pr Δ coefficients indicate the change in predicted probabilities of a favorable outcome by increasing the given independent variable from its minimum value to its maximum (Long and Freese 2006). The excluded category in column (a) is complaints retained by HUD in which conciliation was attempted and in column (b) the complaints retained by HUD and successfully conciliated.

Table 2 shows that HUD significantly lags behind FHAP agencies on both conciliation indicators. According to column (a), both state and local FHAP agencies are more effective than HUD at successfully achieving conciliations, and local agencies are most effective. HUD’s ineffectiveness relative to FHAP agencies is even more pronounced in terms of the ability to achieve conciliation on the first attempt. The logit and probability change coefficients for state and local agencies are both very large and significant in column (b). An interpretation of the change in probability coefficients reveals that a state agency is 88 percent more likely than HUD to conciliate a complaint successfully on the first attempt and local agencies are 73 percent more likely to succeed. FHAP agencies seem much more likely than HUD to conciliate successfully on the first attempt.

Table 2.

Coalition Effectiveness: Influences on Effectiveness in the Conciliation of Title VIII Complaints, 1989–2004.

Dependent Variable	Column (a)		Column (b)
	Conciliation Success		First Attempt Success
	Coef		Coef
	(SE)	pr Δ	(SE)	pr Δ
Responsibility^a
State agency	0.285***	.06	4.10***	.88
	(.030)		(.079)
Local Agency	0.541***	.13	3.585***	.73
	(.036)		(.088)
Returned	−0.298***	−.06	−0.020
	(.031)		(.061)
Discrimination type
Race	−0.349***	−.08	0.059
	(.026)		(.057)
Disability	0.052		−0.092
	(.029)		(.063)
Family status	0.311***	.07	−0.320***	−.08
	(.027)		(.058)
Sex	−0.128***	−.03	0.122
	(.031)		(.064)
Color	0.209***	.05	0.671***	.10
	(.046)		(.092)
National origin	−0.189***	−.04	−0.014
	(.034)		(.073)
Religion	−0.356***	−.07	−0.167
	(.072)		(.159)
Issue
Terms and conditions	−0.028		0.096
	(.024)		(.054)
Advertising	0.432***	.10	−0.062
	(.046)		(.103)
Finance	−0.041		−0.154
	(.048)		(.111)
Refusal to rent	0.016		0.056
	(.025)		(.055)
Refusal to sell	−0.554***	−.11	0.319*	.05
	(.063)		(.142)
Coercion	−0.380***	−.08	−0.075
	(.031)		(.068)
False representation	0.011		−0.016
	(.057)		(.120)
Other issues	0.209***	.05	0.180*	.06
	(.032)		(.074)
Segregated area	0.011		0.067
	(.021)		(.044)
Pattern or practice	−1.478***	−.23
	(.365)
Constant	0.239		0.046
	(.194)		(.318)
N	55,583		20,180

Note: Yearly coefficients omitted from presentation. State coefficients omitted from presentation. Robust standard errors are given in parentheses.

^aU.S. Department of Housing and Urban Development serves as the reference category.

*p < .05.

**p < .01.

***p < .001.

These results are not surprising, given the descriptive statistics presented in the first two rows of Table 1. However, the conciliation process represents only one component of effectiveness. We next explore the compensation component.

Remedial Effectiveness

Federal, state, and local civil rights agencies’ ability to award remedies is the second facet of bureaucratic effectiveness. Three measures are used to assess remedial effectiveness: relief in successfully conciliated complaints, relief in all complaints entering the conciliation process, and relief in complaints resulting in cause determinations. Individual complaints serve as the units of analysis. The monetary relief variable is coded as 1 if money was awarded to the complainant and 0 if it was not. The descriptive statistics for these variables, presented in rows 3, 4, and 5 of Table 1, show that HUD made monetary remedies available to complainants in 53 percent of successfully conciliated complaints between 1989 and 2004. The monetary remedy rate was 54 percent for state agencies and 47 percent for local agencies.

The monetary amount variable is the dollar amount provided in relief. Monetary compensation awarded by HUD substantially outpaces amounts from state and local FHAP agencies. Row 4 of Table 1 reports that the median monetary relief HUD, state, and local agencies provided is US$950, US$760, and US$500, respectively. Because the actual dollar amounts are severely skewed, the dependent variable is transformed using the natural log of the actual values for the multivariate analysis that follows. The overall relief variable is coded as 1 if any type of remedy was granted (i.e., monetary, housing, affirmative action, or other) and 0 if not. Row 5 of Table 1 shows that HUD secures some type of remedy in 71 percent of successful conciliations, compared with 67 percent and 60 percent for state and local agencies, respectively.

Logistic models examine bureaucratic influences on monetary relief and overall compensation. An ordinary least squares (OLS) regression model explores correlates of the amount of monetary relief awarded. These multivariate models appear in the three columns of Table 3. Column (a) reveals that HUD performs relatively well in providing compensation to successful complainants. Local agencies lag behind both HUD and state agencies; the difference between HUD and the states is not statistically significant. Cases returned to HUD from state or local agencies award compensation to the complainant less often than complaints initially retained by HUD.

Table 3.

Remedy Effectiveness Influences on Effectiveness in Remedies Provided in Successfully Conciliated Title VIII Complaints, 1989–2004.

Dependent Variable	Column (a)		Column (b)	Column (c)
	Monetary Relief		ln (Monetary $)	Overall Relief
	Coef		Coef	Coef
	(SE)	pr Δ	(SE)	(SE)	pr Δ
Responsibility^a
State agency	0.056		−0.250***	−0.178***	−.04
	(0.047)		(0.045)	(0.050)
Local agency	0.095		−0.697***	−0.091
	(0.046)		(0.055)	(0.059)
Returned	−0.103*	−.03	−0.058	−0.194***	−.04
	(0.050)		(0.057)	(0.051)
Discrimination type
Race	−0.016		0.073	0.096*	.02
	(0.044)		(0.051)	(0.047)
Disability	−0.708***	−.18	−0.038	−0.412***	−.09
	(0.049)		(0.059)	(0.051)
Family status	0.060		0.193***	0.073
	(0.045)		(0.049)	(0.049)
Sex	0.175***	.04	−0.154**	0.225***	.05
	(0.052)		(0.059)	(0.058)
Color	0.009		0.021	0.100
	(0.074)		(0.094)	(0.082)
National origin	−0.189***	−.05	−0.010	0.044
	(0.056)		(0.066)	(0.061)
Religion	−0.271*	−.07	−0.168	−0.254*	−.06
	(0.123)		(0.142)	(0.125)
Issue
Terms and conditions	0.047		−0.084*	0.164***	.04
	(0.038)		(0.040)	(0.042)
Advertising	0.383***	.09	0.168***	0.128
	(0.069)		(0.061)	(0.073)
Finance	1.058***	.24	1.069***	0.697***	.13
	(0.088)		(0.090)	(0.093)
Refusal to rent	0.537***	.13	0.079	0.749***	.15
	(0.040)		(0.042)	(0.044)
Refusal to sell	0.511***	.12	0.531***	0.469***	.09
	(0.117)		(0.131)	(0.128)
Coercion	0.002		0.227***	0.028
	(0.052)		(0.061)	(0.054)
False representation	0.073		0.311***	0.266*	.05
	(0.089)		(0.087)	(0.106)
Other issues	−0.018		0.589***	−0.076
	(0.051)		(0.058)	(0.052)
Segregated area	0.054		−0.048	0.009
	(0.033)		(0.039)	(0.035)
Pattern or practice	1.364		0.890	0.986
	(1.211)		(0.490)	(1.150)
Constant	−0.591		4.283***	−0.153
	(0.312)		(0.470)	(0.323)
n	20,069		10,168	20,190

Note: Yearly coefficients omitted from presentation. State coefficients omitted from presentation. Robust standard errors are given in parentheses.

^aU.S. Department of Housing and Urban Development serves as the reference category.

*p < .05.

**p < .01.

***p < .001.

The model in column (b) reinforces the disparities shown on row 4 of Table 1. HUD awards significantly exceed amounts from FHAP agencies, with the difference especially great for local agencies. The final column in Table 3 demonstrates that HUD is more effective at providing not only monetary remedies for complainants but housing units and affirmative action as well. The differences between each level of agency are not large, yet they are significant. Achieving some beneficial output is especially great when HUD is compared with local agencies. Complaints returned to HUD, which may have proven particularly difficult, are significantly less likely than cases retained by HUD to provide some form of relief.

Thus far the findings appear mixed: HUD resolves fewer complaints than FHAP agencies, but HUD is most likely to secure relief for complainants in successful conciliations and to provide the largest amount of monetary relief. In light of the conflicting findings, it is appropriate to consider both stages of the conciliation process—the initial success or failure of conciliation and remedy awarded if success is achieved. Simultaneously considering both stages helps determine by level of agency the likelihood that compensation will be available. Relief results for all conciliation attempts appear in rows 6, 7, and 8 of Table 1. Figures in rows 6 through 8 differ from those in rows 3 through 5 since the former consider all conciliation attempts, while the latter focuses on successful conciliations. On two dimensions, HUD’s advantage over state and local agencies in providing remedies disappears once failed conciliation efforts are included. Yet the amount of monetary relief presents another story. HUD outcomes remain the most generous (US$875), but the difference between HUD and state results (US$850) has largely disappeared, although they remain significantly greater than awards from local FHAP agencies (US$500).

The multivariate analysis presented in Table 4 includes both successful and unsuccessful conciliation attempts. This provides a sense for which type of agency is most likely to furnish an effective outcome for a complainant by considering both the probability of a successful conciliation and the provision of relief. When all complaints are considered, HUD provides less compensation to successful complainants than do state agencies and about the same as local agencies. Most surprising is the evidence that with controls, local FHAPs provide the same level of payouts as HUD, a finding not foreshadowed by the descriptive data. The discrepancy seems to result from the transformation of the compensation variable. A regression on the untransformed dependent variable yields differences similar to those shown in Table 1, in which there is very little difference in the amount HUD and state agencies awarded, with local agencies lagging behind. This suggests that the descriptive findings reflect a few large outliers of monetary awards provided by HUD. In all, the findings do not support a conclusion that FHAP agencies outperform HUD in monetary amount when only considering successful conciliations. However, on the whole, Table 4 indicates that HUD’s awards lag behind FHAP agencies when complaints enter the conciliation process—regardless of whether there is ultimately a successful conciliation. Complaints returned to HUD produce significantly less relief than those it retained.

Table 4.

Influences on Effectiveness in Remedies Provided in All Title VIII Complaints Entering the Conciliation Process, 1989–2004.

Effectiveness: Conciliation Remedy
Dependent Variable	Column (a)		Column (b)	Column (c)
	Monetary Relief		ln (Monetary $)	Overall Relief
	Coef		Coef	Coef
	(SE)	pr Δ	(SE)	(SE)	pr Δ
Responsibility^a
State agency	0.303***	.05	0.537***	0.342***	.07
	(.029)		(.050)	(.031)
Local agency	0.421***	.08	0.108	0.432***	.09
	(.035)		(.061)	(.039)
Returned	−0.325***	−.05	−0.291***	−0.402***	−.08
	(.033)		(.053)	(.032)
Discrimination type
Race	−0.264***	−.04	−0.238***	−0.244***	−.05
	(.029)		(.041)	(.027)
Disability	−0.385***	−.06	−0.442***	−0.085**	.02
	(.033)		(.044)	(.030)
Family status	0.304***	.05	0.395***	0.254***	.05
	(.030)		(.043)	(.028)
Sex	0.050		0.012	0.009
	(.033)		(.048)	(.031)
Color	0.038		0.024	0.105**	.02
	(.038)		(.072)	(.046)
National origin	−0.252***	−.04	−0.321***	−0.118***	−.02
	(.038)		(.051)	(.034)
Religion	−0.594***	−.08	−0.553***	−0.564***	−.10
	(.089)		(.094)	(.078)
Issue
Terms and conditions	0.010		−0.061	0.077**
	(.026)		(.038)	(.024)
Advertising	0.582***	.11	0.562***	0.379***	.08
	(.046)		(.082)	(.047)
Finance	0.606***	.12	1.043***	0.301***	.07
	(.049)		(.085)	(.048)
Refusal to rent	0.273***	.05	0.332***	0.302***	.06
	(.027)		(.039)	(.025)
Refusal to sell	−0.283***	−.04	−0.073	−0.322***	−.06
	(.069)		(.095)	(.064)
Coercion	−0.214***	−.03	−0.215***	−0.300***	−.06
	(.035)		(.044)	(.031)
False representation	0.056		0.129	0.119*	.03
	(.061)		(.089)	(.056)
Other issues	0.072		0.275***	−0.021
	(.037)		(.058)	(.034)
Segregated area	−0.045*	−.01	0.020	−0.029
	(.021)		(.032)	(.021)
Pattern or practice	−0.262		−0.615	−0.883*	−.15
	(.334)		(.509)	(.350)
Constant	−0.882***		1.094***	−1.072***
	(.061)		(.226)	(.240)
n	55,420		38,696	55,583

Note: Yearly coefficients omitted from presentation. State coefficients omitted from presentation. Robust standard errors are given in parentheses.

^aU.S. Department of Housing and Urban Development serves as the reference category.

*p < .05.

**p < .01.

***p < .001.

The final approach for evaluating intergovernmental bureaucratic effectiveness in fair housing enforcement analyzes remedies when there is reason to believe discrimination has occurred. This measure conforms closely to HUD’s definition of remedial effectiveness discussed earlier (HUD 1998, B-24).

Effectiveness and Cause Determination Remedies

In addition to managing the conciliation process, Title VIII requires HUD or an appropriate FHAP agency to investigate each complaint. When an agency concludes that discrimination has occurred or is about to occur, the case goes to an ALJ or a federal district court. Just over 8,000 of the total complaints resulted in a cause determination.

The analysis here does not treat a cause or no-cause determination as a measure of effectiveness because there is no way of knowing whether discrimination actually occurred. In theory, a cause finding could only be regarded as effective if the complainant did in fact experience discrimination; for an agency to find cause in the absence of discrimination would be to operate ineffectively. Effectiveness is therefore measured in terms of an agency’s ability to award remedies following a cause determination. Three remedies are possible: monetary relief, monetary amount, and overall relief.

Rows 9 and 11 of Table 1 indicate that HUD rarely (1 percent) provided relief following a cause determination. Local FHAP agencies, which have almost exactly the same number of findings of discrimination, are far more likely to provide some form of relief (15 percent). Following a finding of discrimination, state FHAP agencies provide relief at the highest rate (21 percent). Even though HUD rarely finds discrimination, when it does so, row 10 indicates that its monetary awards are the most generous (US$6,750), being more than twice as large as those from local FHAPs (US$3,150) and two-thirds larger than state agency awards (US$4,000).

Table 5 includes three models with each of these dependent variables and examines only cases that resulted in cause determinations. Coding is the same as in Tables 3 and 4. One minor change is the way in which state effects are taken into account. The lower number of observations prevents a fixed effects approach because too many of the state variables drop out due to perfectly predicting the dependent variable. Instead, robust standard errors are clustered on the state of origin. Like the fixed effects approach, this method makes it more difficult for variables to achieve statistical significance. Columns (a) and (c) in Table 5 reinforce the descriptive figures in Table 1: HUD is significantly less likely to provide relief than state or local agencies. In cause determinations, local agencies are 33 percent more likely than HUD to award monetary and other forms of relief. State agencies are 14 percent (monetary relief) and 15 percent (other forms of relief) more likely than HUD to do so. Further support for the descriptive data appears in column (b), which shows that when HUD provides relief, it is far more generous than from FHAP agencies. This point is highlighted by the positive effects shown in column (a) as compared to the negative coefficients for the state and local variables in column (b). Complaints returned to HUD yield significantly fewer awards of relief than those HUD retains.

Table 5.

Influences on Effectiveness in Remedies Provided in Title VIII Complaints Resulting in Cause Determinations, 1989–2004.

Dependent Variable	Column (a)		Column (b)	Column (c)
	Monetary Relief		ln (Monetary $)	Overall Relief
	Coef		Coef	Coef
	(SE)	pr Δ	(SE)	(SE)	pr Δ
Responsibility^a
State agency	3.239***	.14	−1.421**	3.238***	.15
	(0.406)		(0.325)	(0.406)
Local agency	3.291***	.33	−1.432***	3.293***	.33
	(0.410)		(0.363)	(0.410)
Returned	−0.359**	−.01	0.090	−0.362***	−.01
	(0.116)		(0.145)	(0.116)
Discrimination Type
Race	−0.206		0.253	−0.212
	(0.166)		(0.183)	(0.166)
Disability	−0.082		0.112	−0.081
	(0.203)		(0.226)	(0.203)
Family status	−0.098		−0.127	−0.094
	(0.152)		(0.173)	(0.151)
Sex	−0.069		0.379*	−0.064
	(0.160)		(0.183)	(0.160)
Color	0.178		0.136	0.186
	(0.315)		(0.231)	(0.315)
National origin	−0.055		0.479*	−0.053
	(0.209)		(0.212)	(0.209)
Religion	−0.757*	.01	−0.669	−0.753*	−.01
	(0.349)		(0.433)	(0.349)
Issue
Terms and conditions	−0.001		−0.013	0.003
	(0.109)		(0.135)	(0.108)
Advertising	0.488**	.01	−0.045	0.492***	.01
	(0.190)		(0.189)	(0.190)
Finance	−0.474		−0.403	−0.467
	(0.708)		(0.370)	(0.707)
Refusal to rent	0.270**	.01	−0.142	0.271**	.01
	(0.092)		(0.136)	(0.092)
Refusal to sell	0.221		0.728*	0.230
	(0.217)		(0.288)	(0.216)
Coercion	−0.189		0.118	−0.189
	(0.153)		(0.186)	(0.153)
False representation	0.116		0.105	0.120
	(0.258)		(0.217)	(0.257)
Other issues	−0.215		0.117	−0.216
	(0.157)		(0.224)	(0.156)
Segregated area	−0.165		−0.160	−0.163
	(0.130)		(0.128)	(0.130)
Pattern or practice	0.039		0.290	0.043
	(0.449)		(1.094)	(0.449)
Constant	−5.594***		11.843***	−5.602
	(0.914)		(1.216)	(0.917)
n	8,032		713	8,044

Note: Yearly coefficients omitted from presentation. Robust standard errors clustered on states are given in parentheses.

^aU.S. Department of Housing and Urban Development serves as the reference category.

*p < .05.

**p < .01.

***p < .001.

Conclusion

This article compares bureaucratic effectiveness of Title VIII enforcement by HUD and state and local FHAP agencies between 1989 and 2004. Multiple reasons support the hypothesis that state and local agencies enforce the Fair Housing Act more effectively than HUD. The reasons include monetary incentives provided to FHAP agencies, HUD’s adherence to the 100-day time limitation on FHAP investigations, and the possibility of FHAP decertification by HUD. Relying on a large data set (HUD 2005) and HUD’s own characterization of enforcement effectiveness (HUD 1998, B-24), measurement of effectiveness is based on civil rights agencies’ ability to conciliate Title VIII complaints and provide remedies.

The analysis shows that state and local FHAP agencies are generally more effective than HUD both at conciliating complaints and at providing remedies. HUD does appear to be more effective than FHAP agencies in monetary relief awarded when successful conciliations occur; however, HUD’s remedial effectiveness disappears after controlling for the likelihood of successful conciliations. Taking these findings into account, FHAP agencies more effectively enforced Title VIII than HUD, meaning that the level of government handling a housing discrimination claim makes a difference. The requirement that states and localities must pass and enforce discrimination laws substantially equivalent to Title VIII in terms of rights, procedures, remedies, and the availability of judicial review in order to participate in FHAP helps explain the findings. State and local governments willing to endure the process of obtaining and retaining substantial equivalency may be more enthusiastic than HUD about carrying out their functions and better equipped to process complaints given their geographical proximity to alleged cases of discrimination.

The findings have several implications. First, differences in outcomes in a cooperative federalism system may become apparent when the responsibility for enforcing a national program is placed in the hands of various levels of government. Second, at least with regard to fair housing, local agencies demonstrate the highest levels of effectiveness in enforcing the national program’s major objectives. Third, Congress might consider expanding the number and role of FHAP agencies in enforcing fair housing policy since only fifty local FHAP agencies existed at the end of fiscal year 2016 (HUD 2017). Numerous other metropolitan areas and counties could pass housing discrimination laws—often their first—and develop effective FHAPs (U.S. Commission on Civil Rights 1992). Fourth, the success of state and local fair housing program implementation suggests the desirability of delegating enforcement of other federal programs. If enforcement of this sometimes contentious civil rights program does not suffer, it is unlikely that other, less controversial policy goals would come up short if given to states or municipalities. Delegation would be especially attractive if it resulted in cost savings, as the U.S. Commission on Civil Rights (1992) has indicated is true of FHAP in fair housing enforcement.

Footnotes

Declaration of Conflicting Interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The authors received no financial support for the research, authorship, and/or publication of this article.

References

Agranoff

Robert

. 2001. Managing within the matrix: Do collaborative intergovernmental relations exist? Publius 31:31–56.

Bullock

Charles S.

III Lamb

Charles M.

Wilk

Eric M.

. 2017. Cooperative federalism and fair housing enforcement. Social Science Quarterly. doi: 10.1111/ssqu.12451.

Bullock

Charles S.

III Wilk

Eric M.

Lamb

Charles M.

. 2015. Fair housing enforcement in the south and non-south. Social Science Quarterly 96:941–54.

Cameron

Kim S.

Whetten

David A.

. 1983. Organizational effectiveness: One model or several? In Organizational Effectiveness: A Comparison of Multiple Models, edited by Cameron

Kim S.

Whetten

David A.

, 1–22. New York: Academic Press.

Carpenter

Daniel P.

2001. The Forging of Bureaucratic Autonomy: Reputations, Networks, and Policy Innovation in Executive Agencies, 1862–1928. Princeton, NJ: Princeton University Press.

Collins

William J.

2006. The political economy of state fair housing laws before 1968. Social Science History 30:15–49.

Elazar

Daniel J.

1962. The American partnership: Intergovernmental co-operation in the nineteenth-century United States. Chicago, IL: University of Chicago Press.

Farley

Reynolds

Frey

William H.

1992. Changes in the segregation of whites from blacks during the 1980s: Small steps toward a more integrated society. Population Studies Center No. 92-257, University of Michigan, Ann Arbor.

Gilmour

John B.

Lewis

David E.

. 2006. Political appointees and the competence of federal program management. American Politics Research 34:22–50.

10.

Grodzins

Morton

. 1966. The American System: A New View of Government in the United States. Edited by Elazar

Daniel J.

Chicago, IL: Rand McNally.

11.

HUD (U.S. Department of Housing and Urban Development). 1998. A New HUD: Opportunity for All: 1997 Consolidated Report Appendices. Washington, DC: U.S. Department of Housing and Urban Development.

12.

HUD (U.S. Department of Housing and Urban Development). 2005. Office of Fair Housing and Equal Opportunity Data Set. Washington, DC: U.S. Department of Housing and Urban Development. (Obtained by the authors under the Freedom of Information Act.)

13.

HUD (U.S. Department of Housing and Urban Development). 2008. Title VIII Complaint Intake, Investigation, and Conciliation Handbook. Washington, DC: U.S. Department of Housing and Urban Development.

14.

HUD (U.S. Department of Housing and Urban Development). 2009. Fair Housing Assistance Program-HUD/U.S. www.hud.gov/offices/cfo/reports/2009/cjs/fheo1.pdf (accessed July 29, 2017).

15.

HUD (U.S. Department of Housing and Urban Development). 2014. Annual Report on Fair Housing FY 2012–2013. Washington, DC: U.S. Department of Housing and Urban Development.

16.

HUD (U.S. Department of Housing and Urban Development). 2017. Annual Report to Congress, Office of Fair Housing and Equal Opportunity, FY 2016. Washington, DC: U.S. Department of Housing and Urban Development.

17.

Ingraham

Patricia W.

Joyce

Philip G.

Donahue

Amy Kneedler

. 2003. Government Performance: Why Management Matters. Baltimore, MD: Johns Hopkins University Press.

18.

Lamb

Charles M.

2005. Housing Segregation in Suburban America since 1960: Presidential and Judicial Politics. New York: Cambridge University Press.

19.

Lamb

Charles M.

Wilk

Eric M.

. 2010. Civil rights, federalism, and the administrative process: Favorable outcomes by federal, state, and local agencies in housing discrimination complaints. Public Administration Review 70:412–21.

20.

Lee

Soo-Young

Whitford

Andrew B.

. 2009. Government effectiveness in comparative perspective. Journal of Comparative Policy Analysis 11:249–81.

21.

Lichter

Daniel T.

Parisi

Domenico

Taquino

Michael C.

. 2015. Toward a new macro-segregation? Decomposing segregation within and between cities and suburbs. American Sociological Review 80:843–73.

22.

Logan

John R.

Stults

Brian J.

. 2011. The persistence of segregation in the metropolis: New findings from the 2010 census. http://www.s4.brown.edu/us2010 (accessed July 29, 2017).

23.

Long

J. Scott

Freese

Jeremy

. 2006. Regression Models for Categorical Dependent Variables Using Stata, 2nd ed. College Station, TX: Stata Press.

24.

MacDonald

Jason A.

Franko

William W.

Jr . 2007. Bureaucratic capacity and bureaucratic discretion: Does congress tie policy authority to performance? American Politics Research 35:790–807.

25.

Massey

Douglas S.

Denton

Nancy A.

. 1993. American Apartheid: Segregation and the Making of the Underclass. Cambridge, MA: Harvard University Press.

26.

Massey

Douglas S.

Tannen

Jonathan

. 2015. A research note on trends in black hypersegregation. Demography 52:1025–34.

27.

National Commission on Fair Housing and Equal Opportunity. 2008. The Future of Fair Housing: Report of the National Commission on Fair Housing and Equal Opportunity. www.nationalfairhousing.org/Portals/33/reports/future_of_fair_Housing.pdf (accessed January 2, 2017).

28.

Oliver

J. Eric

. 2010. The Paradoxes of Integration: Race, Neighborhood, and Civic Life in Multiethnic America. Chicago, IL: University of Chicago Press.

29.

Scholz

John T.

1991. Cooperative regulatory enforcement and the politics of administrative effectiveness. American Political Science Review 85:115–36.

30.

Scholz

John T.

Wei

Feng Heng

. 1986. Regulatory enforcement in a federalist system. American Political Science Review 80:1249–70.

31.

Schwemm

Robert G.

2017. Housing Discrimination: Law and Litigation. St. Paul, MN: Thomson West.

32.

Seabrook

Nicholas R.

Wilk

Eric M.

Lamb

Charles M.

. 2013. Administrative law judges in fair housing enforcement: Attitudes, case facts, and political control. Social Science Quarterly 94:362–78.

33.

Turner

Margery Austin

Ross

Stephen L.

Galster

George C.

Yinger

John

. 2002. Discrimination in Metropolitan Housing Markets: National Results from Phase 1 HDS 2000. Washington, DC: U.S. Department of Housing and Urban Development.

34.

Turner

Margery Austin

Santos

Ron

Levy

Diane K.

Wissoker

Doug

Aranda

Claudia

Pitingolo

Rob

. 2013. Housing Discrimination against Racial and Ethnic Minorities 2012. Washington, DC: U.S. Department of Housing and Urban Development.

35.

U.S. Commission on Civil Rights. 1992. Prospects and Impact of Losing State and Local Agencies from the Federal Fair Housing System. Washington, DC: U.S. Commission on Civil Rights.

36.

U.S. Commission on Civil Rights. 1994. The Fair Housing Amendments Act of 1988: The Enforcement Report. Washington, DC: U.S. Commission on Civil Rights.

37.

U.S. Government Accountability Office. 2005. Fair Housing: HUD Needs Better Assurance that Intake and Investigation Processes Are Consistently Thorough. Washington, DC: U.S. Government Accountability Office.

38.

Wilk

Eric M.

Lamb

Charles M.

. 2011. Federalism, efficiency, and civil rights enforcement. Political Research Quarterly 64:392–404.

39.

Wilson

James Q.

2000. Bureaucracy: What Government Agencies Do and Why They Do It, 2nd ed. New York: Basic Books.

40.

Wolf

Patrick J.

1993. A case survey of bureaucratic effectiveness in U.S. cabinet agencies: Preliminary results. Journal of Public Administration Research and Theory 3:161–81.