Introduction
Judicial nominations have become the ultimate political battleground. Senate Republicans introduced the filibuster into judicial advice and consent to stall Supreme Court nominee Abe Fortas in 1969[1]. Robert Bork’s rejection by the Democrats in 1987 was due to his strident denial of the implicit “right to privacy” that was cited in the 1973 Roe v. Wade decision to overturn state laws against abortion[2]. Some sixty of Bill Clinton’s judicial nominees were blocked from the bench by Republican partisans[3], and outnumbered Democrats have threatened to filibuster ten of George W. Bush’s appellate nominees[4]. Obviously the politics of the judicial bench are considered worth going to metaphorical war, including what is being called a Republican “nuclear option” (changing Senate rules) and a retaliatory “scorched earth” policy by the minority Democrats (grinding the Senate to a halt).
How significant is political leaning to the decisions made by our judiciary? What other factors influence these decisions, such as race, gender, prior judicial experience or military service? In this paper we will address the methods being used to analyze judicial results and the factors that influence them. We will examine the issue of Fair Housing, ramifications of the Fair Housing Act (Title VIII of the Civil Rights Act of 1968, or “FHA”), and the Fair Housing Amendments Act of 1988 (FHAA), which enlarged the scope of the FHA. We will then analyze the Federal District Court decisions concerning Fair Housing from 3/12/1989 to 12/31/1995, and attempt to shed light on what factors may statistically influence those decisions.
The Politics of Judicial Selection
Federal judges have awesome power. They can order whole states, or even the federal government, to change their policies and operation when they conflict with federal law. They are given lifetime appointments, which should make them immune to outside political pressures. However, these judges are selected according to the political interests of the appointing President and the Senate that confirms them.
Judges themselves almost universally proclaim in their writings their commitment to interpret the law correctly and apply it scrupulously to the case at hand (Rowland & Carp 1996: 149). In practice, since judges are human beings, both the interpretation of law and findings of fact are subject to the judge’s preexisting frame of reference. His or her personal attitudes color both interpretation and decision-making, often in an unconscious way. Political orientation is only one possible factor. Tenure, or experience on the federal bench, may affect the judge’s sympathy for a plaintiff. Judges with prior judicial experience, especially if in the Magistrate courts, may have drastically different visceral leanings than those who have spent their lives in private practice, or as prosecutors or politicos.
The research backs up common sense. “The appointees of Democratic presidents are clearly more liberal in their decision making then are judges chosen by Republican chief executives,” leading researchers are comfortable to declare (Rowland & Carp 1996: 46). The gap seems to be getting wider, possibly accounted for by increased caseload due to expanded protection of minority groups under the 1968 and 1988 Civil Rights Acts. The number of all federal cases ballooned from less than 200 under all the Woodrow Wilson years to upwards of 1500 per year under Kennedy, Johnson, Nixon and Carter (Rowland & Carp 1996: 47).
The concentration of liberal or conservative decisions may be exacerbated by the fact that civil rights “cases are simply not brought in some districts” with low urban populations (Richardson and Vines 1970, quoted in Rowland & Carp 1996: 75). Rowland and Carp suggest that liberal groups supporting minorities are more likely to be found in large cities, whereas in rural areas there is significant pressure to conform to conservative community standards. “Venue shopping” may be a factor skewing the percentages to indicate polarization.
Another factor affecting increased politicization on the bench is a conservative backlash against enforcement of liberal legislation. This is illustrated in the decline of out-party consultation for Senatorial courtesy during selection of candidates. This example shows the contrast between Carter-era collaboration and Reagan-era circumvention of minority party input:
“Before he recommended a candidate… [Democratic Senator Lloyd] Bentsen always consulted informally with his Republican counterpart, Senator John Tower. This consultation, which staffers described as cordial, mirrored exactly the courtesy extended Bentsen by Tower during the Nixon and Ford era. But the courtesy that characterized previous eras disappeared with the election of Republican Phil Gramm…. During this period Senator Bentsen and his staff learned the identity of nominees… only after the formal submission of a nominee to the Senate Judiciary Committee.” (Rowland & Carp 1996: 101, italics in original)
Significant effort is put into political lobbying by various groups for and against judicial nominees, for Supreme Court justices especially, but increasingly for appellate and district court nominees. “The lobbying tactics on federal judicial nominations do not vary much by the type and visibility of the nomination,” conclude Caldeira, Hojnacki and Wright (2000: 62), although their study did not measure breadth and intensity of lobbying activity. Giles, Hettinger and Peppers point out that “the effects of both the policy and partisan agendas on judicial behavior may be conditioned by the political context” (2001: 632). The makeup of the federal judiciary is clearly considered highly important by a wide variety of interest groups on both sides, focusing on ideology as well as concrete interests.
How influential is this maneuvering and selectivity on actual outcomes, and the judicial process? A study of Carter appointees “consistently revealed no substantial differences between white and black jurists.” In fact, “female judges were significantly less libertarian…, more prone to rule in favor of the government…, and less sympathetic to…minorities” (Walker & Barrow 1985: 608).
Existing models of judicial behavior are not unary determiners (Rowland & Carp 1996: 136, passim). Any one or all may be applicable to explain given sets of data, and each perspective can tell us something different. Theories of judicial decision-making are categorized broadly into Behavioral approaches such as Legalistic, Organizational, Attitudinal models, and the more subtle Cognitive model.
The traditional Legalistic model assumes an adversarial relationship on which the judge must take a side. This view assumes there is a “right” side that in fairness will succeed. This model may be correct in the majority of unpublished decisions, which are by definition those that the judges or staffs considered not to be of significant interest as matters of law. In fact, “Republican judges levy significantly higher civil penalties in published cases than unpublished cases” (Ringquist & Emmert 1999: 31). Legalistic theories do not explain the trends of decision-making seen in published cases that showed marked difference in outcomes correlating to political affiliation.
Another problem with the Legalistic model is that contestable cases by definition have plausible legal arguments on both sides. “It follows…that if Supreme Court justices care more about policy than law, it is easy for them to find legal justification for whatever positions they prefer” (Baum 1997: 64). Clearly motivation, perspective and prejudices would impact the decision process even given the straightforwardness of the Legalistic scenario.
The Organizational model includes the entire court structure, attributing decisions to “the shared interests of its major participants,” including the judge, attorneys, and courtroom staff (Eisenstein & Jacob 1974, quoted in Rowland & Carp 1996: 137). This model also deals well with routine cases where issues of law are settled, and includes realistic factors such as caseload, working conditions and deal making. However, it does not explain apparent statistical biases in judicial decisions in contentious areas such as civil rights. The judge is still the final and sole arbiter in these cases.
Attitudinal models evolved to explain statistical evidence that judges’ decisions in non-routine cases are influenced by their personal support level for various constitutional issues. The supposition here is that decisions are directly or indirectly based on political leaning or pandering. It implicitly presupposes that judges are insincere in their explanatory opinions and findings of fact and law. The attitudinal model does serve to explain the statistical facts, however, and sizable bodies of work support it (Rowland & Carp 1996: 141). However, the complexities of strategic behavior add seemingly contradictory results where judges rule against their own presumed interest groups. “For instance, it may be that female and black judges are aware of the expectations associated with their appointments, but are compelled to perform in the opposite manner to counterbalance internal or external criticism…” (Segal 2000: 147).
The central question is whether bias is consciously motivated (Rowland & Carp 1996: 164). The cognitive approach addresses shortcomings in psychologically based attitudinal theories. In this model the judge is not reacting with calculated bias as much as unconscious ideological framing. The critical cognitive process of evaluating facts is itself uniquely individual and affected by perspective. This view allows for sincerity of judges’ legal explanations and motivations. “The honest, indignant reaction of judicial interviewees to the notion that fact-finding might be motivated by their personal policy preferences was an important impetus for our critique of the attitudinal model” (Rowland & Carp 1996: 190, Ch. 6, Note 4).
Most research is based on statistical analysis of case results. The cognitive approach indicates that content analysis of judges’ own reasoning and decision-making processes as recorded in their opinions is an important if often overlooked resource. “One could…define ambiguity operationally in the context of judges’ published opinions – i.e., their own reports of scanty, unreliable or conflicting information…and impose rigorous content analysis on the measurements of those indicators” (Rowland & Carp 1996: 191, Ch. 7, Note 1).
The Politics of Fair Housing
Fair Housing should not even be an issue. The Civil Rights Act of 1866, enacted in the aftermath of the Civil War, prohibited discrimination on the basis of race in access to housing. “By its terms, § 1982 [of] prohibits all racial discrimination in the sale or rental of property. It makes no exceptions” (Walsh 1999: III.A). Despite laws on the books, blacks in particular were subjected to systemic discrimination. Federal and local governments were complicit in, if not actively in support of, this discrimination. This took the form of “restrictive covenants…which forbade the sale of property to blacks” and “widespread violence and intimidation…abetted by refusing local police protection” (Orfield 1974: 784). In addition, the Federal Housing Administration enforced racial segregation in regulations in that, for example, “’incompatible racial elements’ was officially listed as a valid reason for rejecting a mortgage” (Orfield 1974: 786). The result was that minorities were concentrated into urban ghettos, where substandard housing was contradictorily overpriced due to the risks of ownership in bad neighborhoods.
Historically, minorities have suffered from disadvantages in the U.S. legal system, whether because of poverty, culture, or other reasons. “Hispanic defendants receive harsher sentences than either white or black defendants” found Steffensmeier and Demuth (2000: 726) although their study found blacks as well to be convicted and punished more severely than whites. Sentencing can be adjusted for severity of offense, but “measures of severity are higher in poor and minority areas,” which may explain why “there is little difference in average fines when controlling for the seriousness of the offense” concluded Evan Ringquist (1998: 1154) in his study of EPA pollution cases. The difference in severity of offense might be ultimately laid at the feet of housing situations that promulgate poverty, hopelessness and law-breaking behavior.
The prejudicial housing situation and its effects were elevated to crisis level by the attention-getting race riots of the 1960’s. Lyndon Johnson, in his last year in office, signed into law the Civil Rights Act of 1968, including Title VIII, also known as the Fair Housing Act (FHA).
Passage of the FHA was hard-fought, and required compromise from the Senate framers of the bill. This took the form of exceptions written into the law, excluding religious organizations, private clubs, some single-family homes, and “Mrs. Murphy” landlords. Additional exemptions were included in the 1988 FHAA provisions that added familial status and disability (Schwemm 2001).
One example is Section 3603(b)(2), the so-called “Mrs. Murphy” exemption. This clause gives owners the right to discriminate in renting, if the building has four units or fewer, and the owner lives in one of the units. Senator Walter Mondale, a co-sponsor of the Act, understood the political necessity of including this exemption, which put to rest opposition concerns that owners like the archetypal pensioner Mrs. Murphy would sacrifice any selectivity regarding tenants taken by necessity into their own homes. However, some would say “implicit was an understanding the First Amendment right at stake was specifically Mrs. Murphy’s right not to associate with African Americans” (Walsh 1999). The Mrs. Murphy exemption functioned exactly as intended, and took enough steam out of anti-Civil Rights exponents to pass the bill into law.
Mrs. Murphy and other exemptions only cover the actual rental of housing, not hurtful speech that may be associated with it. The 3604(c) provision prohibits discriminatory notices, statements and advertisements. These expressions, which include verbal speech, are explicitly not exempted for almost all FHA situations. With the single exception of familial status in cases of senior housing, “3604(c) is applicable to all housing providers, whether or not they are exempt from the statute’s other mandates” (Schwemm 2001). This creates the odd situation in which otherwise FHA-exempt landlords are allowed to discriminate, as long as they don’t tell their victims why they have been refused.
The exemptions have been abused to circumvent the protections offered by the FHA. The Senior Housing exemption, for example, protects elderly-assistance facilities but also includes certain housing where 80% of the units are merely occupied by “at least one person 55 years of age or older per unit” (Nelson 2003). This has the immediate effect of allowing exclusion of children from developments intended for seniors, to allow them the opportunity for quiet living. A possibly unintended secondary consequence is the exclusion of large groups of minorities, since “Statistically, African-Americans and Hispanics are much more likely to have either a grandchild or a child under 18 living with them than a white family” (Nelson 2003: IV). Thus the Senior Housing exemption has come under fire for its discriminatory effect.
Another exemption with possibly unintended effects is for “reasonable” government occupancy standards. The definition of reasonable is at question, especially in limits on overcrowding according to community standards based on middle-class white family patterns. In addition to an obvious generalized impact on the poor, “there is also clear evidence that many households living closely do so based upon enduring cultural preferences and non-economic interests” (Iglesias 2004).
The passage of the FHAA in 1988 added an important enforcement mechanism for fair housing cases. “The FHAA…is unique among federal statutes in mandating that a complainant’s suit be brought by the government automatically upon finding of reasonable cause and in authorizing governments suits for substantial money damages ‘on behalf of’ private individuals” (Gaetke & Schwemm 1997: C). The bill’s primary sponsor, Senator Edward Kennedy, said the FHAA would “put real teeth into the fair housing laws by giving HUD real enforcement authority” (quoted in Gaetke & Schwemm 1997: B).
The administrative process provides for Magisterial Courts, in which routine Housing and Urban Development (HUD) cases are arbitrated by Administrative Law Judges (ALJs), with plaintiffs retaining the right to appeal to an overseeing federal district court. Government lawyers have quite a good success rate in federal courts, perhaps because of organizational familiarity, but in large part due to expertise (McGuire 1996, in Baum 1997: 53). The system is not perfect, since the government lawyers are not traditional dedicated advocates, but it has given hearing to several thousands of complaints each year, many of which were elevated to become district court cases in this study.
The source for case data was the Legal Research database in LexisNexis Academic. We searched the Federal Case Law section, limiting the search to District Courts, with the keywords “Fair Housing”. The search covered the period from 3/12/89 to 12/31/95, and returned 546 results. These included numerous cases that were not FHA cases, but merely had the words “fair” and “housing” somewhere in the opinion.
I deleted the non-FHA cases, including HUD cases not involving denial or interference with housing. Several of these were filed by builders or investors, and did not concern an actual FHA claim. Also removed from the search results were any Magisterial Court decisions, since the magistrate judges are appointed directly by the district judges and are not subject to the political appointment process[5]. Magisterial cases were included only when the opinions were affirmed or rejected by a federal judge.
The case set may not be comprehensive of FHA district court cases, since my search did not include citations such as “42 U.S.C. §§ 3601, 3604” or “Title VIII of the Civil Rights Act” that may have been used in some opinions. Of course, unpublished opinions were not considered within the scope of this project. However, the final sample group includes 335 total cases, including 207 decision in favor of FHA plaintiffs and 128 against. The aggregate success ratio for all FHA claimants in the study group is 61.8%.
Four fields were used to code results: Pro-FHA, Anti-FHA, Substantive, and Split. Any ruling in favor of the FHA plaintiff was coded as Pro-FHA. The Split decisions category indicated which of these included some Anti-FHA components. So a case in which the FHA plaintiff was granted an injunction, but denied a request for attorney fees, was considered Split. An opinion in which the plaintiff filed eight counts, and all but one were dismissed, was also coded as “Pro-FHA, Split.”
Substantive decisions were tracked to allow us to isolate procedural rulings. I coded 0 (not Substantive) for class certifications, denied motions for summary judgment or fees, and any evidentiary motions. Denial of relief from judgment, amendments granted and reaffirmed decisions were coded as 1 (Substantive). Any decisions awarding damages or fees, and any summary judgments were coded Substantive. Rulings to remand to state courts were Substantive if the FHA claim was thrown out, and Not Substantive if the judge ruled that state law was materially similar and simply relinquished jurisdiction.
Biographical data was obtained from the Federal Judges Biographical Database[6]. The appointing president and date of commission determined the party political orientation, and based on these we created a field for president’s party, with 1 indicating Democrat. I also collected race, sex, prior judicial experience, and military service. These were all coded as 1 or 0, with 1 indicating minority, female, prior judgeship, or history of military service.
We constructed another field for “tenure”, which contained the number of days since commission for each decision. This allowed us to check whether judges became less supportive of civil rights the longer they were on the bench. This might be caused by the scarcity of women and minority judges prior to the Johnson administration, so other factors had to be included to see whether tenure in itself was a significant factor.
Basis for discrimination was divided into seven categories: Race, National Origin, Familial Status, Disability, Religion, Sex and Indeterminate. Another data field was collected to track Interference. This pertained to cases in which the plaintiffs were not denied housing, but claimed their right to “the exercise or enjoyment”[7] of their home was interfered with.
Cases of retaliation against employees were included when the issue was refusal to execute policies in violation of the FHA. The basis of discrimination was determined by the type of housing violation promulgated by the employer.
Finally, cases are divided into five regions based on district court locations. These regions are: Northeast (CT, MA, ME, NH, NJ, NY, PA, PR, RI, VT), Midwest (IA, IL, IN, KS, MI, MN, ND, NE, OH, SD, WI), Border States (DC, DE, KY, OK, MD, MO, WV), South (AL, AR, FL, GA, LA, MS, NC, SC, TN, TX, VA) and West (AK, AZ, CA, CO, HI, ID, MT, NM, NV, OR, UT, WA, WY). Puerto Rico was included in the Northeast region because it is under the First Circuit Court of Appeals, whose jurisdiction with that exception is within that region[8]. The six Border States and the District of Columbia have been recognized in various sources (e.g., Bernstein (2004): B, 2, a, passim.) as a group having Civil Rights characteristics different from Northeast, Midwest, or South, so they are taken separately for analysis.
The case data was imported into the “Intercooled Stata 8.0” statistical program for analysis. For this brief study, only a few summative and speculative trails were followed. Certainly, further analysis would be revealing.
The current study examines demographic factors such as race, sex, region, prior judicial experience, and military background of the judge. My starting assumption was that some of these factors might be as reliable predictors of policy as is political party. Along with basis for discrimination, these demographics are considered as possible explanations for tendencies in Fair Housing decisions.
Analysis of Data
Now we will examine the distribution of opinions in the Fair Housing cases for the period of 3/12/89 to 12/31/95, beginning when the FHAA came into effect. The opinions included procedural and substantive as described above. The difference between Substantive and Procedural outcomes indicates the organizational process of court motions succeeding, at a rate of 80.1%, while the substantive outcomes have a much less rosy 49.2% success rate for the FHA plaintiff (see Figure 1).
Figure 1. Case distribution by type of ruling
|
Ruling |
# |
For FHA |
Anti FHA |
% For |
|
Split |
% Split |
|
Substantive |
199 |
98 |
101 |
49.2% |
|
44 |
22.1% |
|
Procedural |
136 |
109 |
27 |
80.1% |
|
29 |
21.3% |
|
Total |
335 |
207 |
128 |
61.8% |
|
73 |
21.8% |
The effect of the Procedural rulings is to inflate the apparent success rate for the FHA plaintiff. The impact, in general, is that the “% For” in the following tables is about 12.5% lower when you disregard the Procedural decisions. This should be kept in mind throughout discussion of the results. Accordingly, “Substantive” appears as a significant factor in all of the statistical analyses.
Since all Split Decision cases were also counted as “For FHA”, this too may skew the results in favor of FHA cases. However, in these situations the FHA plaintiff did get something, even if only the right to continue the complaint. Therefore the “For FHA” decision totals may be somewhat “soft”, but are indicative of some level of success in court. The percentage of split decisions is about the same for procedural and substantive rulings.
We’ll begin by looking at the breakdown of basis for discrimination in the FHA cases. The most prevalent cause was the 152 Racial claims, which is closely tied to the 37 based on National Origin. There was considerable overlap between these two groups since “national origin” tends to refer to Hispanics, and discrimination is often directed against the two groups together. The suits filed by or on behalf of Hispanics had the highest success rate for all case types, 56.5%, while Racial claims succeeded only 49.4%, close to the overall average (see Figure 2).
Figure 2. Case distribution by basis of discrimination
|
Basis of Discrimination |
|
All |
For FHA |
Anti FHA |
% For |
|
Subst |
For FHA |
Anti FHA |
% For |
Subst Delta |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Race |
|
152 |
97 |
55 |
63.8% |
|
85 |
42 |
43 |
49.4% |
-14.4% |
|
|
National Origin |
|
37 |
25 |
12 |
67.6% |
|
23 |
13 |
10 |
56.5% |
-11.0% |
|
|
Familial Status |
|
42 |
23 |
19 |
54.8% |
|
26 |
14 |
12 |
53.8% |
-0.9% |
|
|
Disability |
|
106 |
68 |
38 |
64.2% |
|
68 |
36 |
32 |
52.9% |
-11.2% |
|
|
Religion |
|
16 |
7 |
9 |
43.8% |
|
10 |
3 |
7 |
30.0% |
-13.8% |
|
|
Sex |
|
20 |
12 |
8 |
60.0% |
|
11 |
3 |
8 |
27.3% |
-32.7% |
|
|
Indeterminate |
|
20 |
11 |
9 |
55.0% |
|
9 |
4 |
5 |
44.4% |
-10.6% |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Interference |
|
20 |
12 |
8 |
60.0% |
|
8 |
1 |
7 |
12.5% |
-47.5% |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
All Decisions |
|
335 |
207 |
128 |
61.8% |
|
199 |
98 |
101 |
49.2% |
-12.5% |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Note: Sum is greater than total of case types because numerous cases claimed multiple bases. |
||||||||||||
Disability had the second highest volume of cases, with 106 claims. These also succeeded at a higher rate than the norm, with 52.9% favorable substantive decisions. Familial Status cases fared well, with 53.8% success in substantive decisions. As an interesting anomaly, the substantive success rate is not substantially different from the procedural success rate, for Familial Status uniquely. Maybe they have better lawyers, or don’t lack them. This may be an effect of HUD lawyer support for familial status cases, which is not uncommon.
The losers in the FHA case history were Religion and Sex discrimination cases, with admittedly small sample groups of 16 and 20 respectively. They succeeded only 30% and 27.3% of the time, indicating lower levels of support across the judiciary, at least in Housing cases.
Cases of interference refer to cases where housing was provided, but the plaintiffs claimed interference with “the exercise and enjoyment” of their property rights. These cases always overlapped with a statutory basis for discrimination. Interference cases succeeded in FHA claims only once out of the small sample group of 8 substantive decisions.
Breaking down the decisions by region we can see that the South and Northeast regions are pretty close, if slightly below the national average (see Figure 3). This seems to indicate that federal scrutiny has succeeded in reducing housing discrimination in the South to a level slightly below the national average. While there are organizational tactics that can discourage cases from being brought, such as awarding attorney fees to the defendant, this does not seem to be the case due to the relatively high plaintiff success rate.
Figure 3. Case distribution by Region
|
Region |
# |
For FHA |
Anti FHA |
% For |
|
Northeast |
131 |
76 |
55 |
58.0% |
|
Border |
15 |
8 |
7 |
53.3% |
|
South |
47 |
28 |
19 |
59.6% |
|
Midwest |
119 |
83 |
36 |
69.7% |
|
West |
23 |
12 |
11 |
52.2% |
|
Total |
335 |
207 |
128 |
61.8% |
Presumably lack of scrutiny allows the Border States to be among the worst places for FHA claims, with only a 53.3% success rate. The West is the worst venue of all based on the case history (52.2%), but some study would have to be done to determine whether this is due to tenuous claims being filed in the courts under the 9th Circuit on the off chance they might buy it. The very low volume indicates either that systemic discrimination does not exist, or that FHA cases are somehow discouraged.
The Midwest had the highest rate of success in FHA decisions. These figures were boosted by numerous protracted cases involving housing discrimination in Chicago (Northern District of Illinois). The courts were dealing with widespread violations, many of which involved blatant discrimination that may have been too clear-cut for political leaning to influence.
One comment about Puerto Rico: the bulk of FHA cases (6 in our sample) are based on disability, especially denial of housing rights to AIDS facilities. These cases tended to be successful, with Reagan appointees accounting for most decisions.
The next breakdown of cases is by presiding Judges’ demographics. The most significant difference, not surprisingly, is the higher support for FHA cases of 69.5% by Democratic judges, compared to 56.2% by Republicans (a gap of 13.3%). This is the largest differential of any of the demographic factors tracked, and its significance is supported by the Stata analysis. Minority judges found for plaintiff 7.5% more often than White judges. Male judges found for FHA causes 6.2% more often than did Female judges, perhaps surprisingly (See Figure 4).
Figure 4. Case distribution by Judges’ demographic
|
Judges' Demographics
|
# |
For FHA |
Anti FHA |
% For |
|
Minority |
38 |
26 |
12 |
68.4% |
|
White |
297 |
181 |
116 |
60.9% |
|
|
|
|
|
|
|
Female |
53 |
30 |
23 |
56.6% |
|
Male |
282 |
177 |
105 |
62.8% |
|
|
|
|
|
|
|
Republican |
194 |
109 |
85 |
56.2% |
|
Democrat |
141 |
98 |
43 |
69.5% |
|
|
|
|
|
|
|
Military experience |
184 |
117 |
67 |
63.6% |
|
Non-military |
151 |
90 |
61 |
59.6% |
|
|
|
|
|
|
|
Prior judicial experience |
149 |
99 |
50 |
66.4% |
|
No prior judicial experience |
186 |
108 |
78 |
58.1% |
|
|
|
|
|
|
|
Total |
335 |
207 |
128 |
61.8% |
Figure 4 also shows the effects of Military experience, which was a common factor amongst many of the older, World War II era veterans. Military veterans on the bench supported FHA causes at a 63.6% rate, only slightly higher than the 61.8% average for all cases. Military experience does not seem to be much of a factor, although it might be interesting to break that down by service to see if ingrained characteristics of any particular branch has an effect. Marines might be more goal-oriented, or Army veterans might support organizational equity to a greater degree.
Prior judicial experience was a factor in favor of FHA claimants. The success rate was 66.4%, with a differential of 8.3% compared to decisions by judges with no prior time on the bench. This should at some future time be isolated to determine whether some other factor is involved. For instance, could non-traditional judges be influencing the results by timidity in supporting plaintiffs in civil rights cases?
The multiple logistic regression analysis shows we have significant determiners for Pro-FHA rulings. The first analysis (Figure 5) includes a host of independent variables for region, judges’ demographics, and basis of discrimination claim. The pseudo R2 of 0.1175 indicates that one or more of the independent variables has a significant predictive effect on the dependent variable, “Pro-FHA Rulings”.
Amongst regions, the Midwest is better for FHA (0.52 coefficient) but is not significant predictor on its own, with a P>|z| of 0.358. This means there is more than a 35% chance that the relationship is accidental. In general, a P>|z| value of less than 0.05 is considered marginally significant, 0.01 significant, and 0.001 highly significant[9].
The most prominent demographic factors for judges, aside from political party, were “Female” and the combined factor “Democrats with Military experience” (military experience alone was not included in the main regression). Both of these factors were slightly negative (both around -0.46 coefficients) but also not significant predictors. The best correlation, for Female judges, still had a 24.7% chance of being coincidence.
Amongst the bases of discrimination, Religion cases got the poorest results, while Disability and National Origin cases got better support than the average. However, none of these factors achieved statistical significance. The only factors that consistently showed P>|z| in the significant range were Substantive and President’s Party. As indicated above, Substantive decisions might be considered a different class, since success in pretrial or procedural motions gives little indication of favorable substantive rulings. This is supported by the large negative determination for Substantive opinions in the field of all FHA rulings.
Figure 5. Multiple logistic regression analysis of case results data
|
Pro-FHA Rulings |
Coef. |
Robust Std. Err. |
z |
P>|z| |
[95% Conf. |
Interval] |
|
Northeast |
0.0605298 |
0.560731 |
0.11 |
0.914 |
-1.038483 |
1.159542 |
|
Border |
-0.2135939 |
0.8243437 |
-0.26 |
0.796 |
-1.829278 |
1.40209 |
|
South |
0.1689064 |
0.6177842 |
0.27 |
0.785 |
-1.041928 |
1.379741 |
|
Midwest |
0.5217353 |
0.5680364 |
0.92 |
0.358 |
-0.5915956 |
1.635066 |
|
Substantive |
-1.521323 |
0.2699927 |
-5.63 |
0.000 |
-2.050499 |
-0.992147 |
|
Judge minority |
-0.0799778 |
0.4446246 |
-0.18 |
0.857 |
-0.9514259 |
0.7914703 |
|
Judge female |
-0.4675247 |
0.4037308 |
-1.16 |
0.247 |
-1.258822 |
0.3237731 |
|
President’s party Dem. |
1.003185 |
0.3937223 |
2.55 |
0.011 |
0.2315037 |
1.774867 |
|
Military + Democrat |
-0.4566731 |
0.4660666 |
-0.98 |
0.327 |
-1.370147 |
0.4568008 |
|
Tenure |
0.000018 |
0.0000546 |
0.33 |
0.742 |
-0.0000891 |
0.000125 |
|
Racial |
0.1174109 |
0.4375006 |
0.27 |
0.788 |
-0.7400745 |
0.9748962 |
|
National Origin |
0.4009189 |
0.4228679 |
0.95 |
0.343 |
-0.4278869 |
1.229725 |
|
Familial Status |
0.0531741 |
0.5546459 |
0.1 |
0.924 |
-1.033912 |
1.14026 |
|
Disability |
0.4204599 |
0.4669903 |
0.9 |
0.368 |
-0.4948243 |
1.335744 |
|
Religion |
-0.7167422 |
0.6262979 |
-1.14 |
0.252 |
-1.944264 |
0.5107792 |
|
Sex |
0.0276864 |
0.4965931 |
0.06 |
0.956 |
-0.9456182 |
1.000991 |
|
_cons |
0.7501576 |
0.6885305 |
1.09 |
0.276 |
-0.5993374 |
2.099653 |
|
Log pseudo-likelihood = -196.61557 Number = 335, Wald chi2(16) = 46.75, Prob > chi2 = 0.0001, Pseudo R2 = 0.1175 |
||||||
President’s party has only a 1% chance of being coincidental in the main regression. Figure 6 isolates Substantive decisions and President’s Party. The overall correlation is less, with a pseudo R2 of 0.0973, so the results are not as completely explained by these two factors alone. Both show significance, but Substantive is such a strong factor that it may be necessary to eliminate Procedural decisions in order to get a true picture of FHA support. Motions are often decided on a strictly legalistic basis, and including non-substantive rulings seems to skew the data. Split decisions in procedural opinions likely have the same effect, whereas split substantive decisions are more likely to indicate real benefit to the plaintiff.
Figure 6. Significant predictors isolated
|
Pro-FHA Rulings |
Coef. |
Robust Std. Err. |
z |
P>|z| |
[95% Conf. |
Interval] |
|
|
|
|
|
|
|
|
|
Substantive |
-1.512359 |
0.2625634 |
-5.76 |
0.000 |
-2.026974 |
-0.9977445 |
|
President’s party Dem |
0.7393636 |
0.2475426 |
2.99 |
0.003 |
0.254189 |
1.224538 |
|
_cons |
1.15139 |
0.2290194 |
5.03 |
0.000 |
0.7025206 |
1.60026 |
|
Log pseudo-likelihood = -21.1284 Number = 335, Wald chi2(2) = 38.2, Prob > chi2 = 0.000, Pseudo R2 = 0.0973 |
||||||
Narrowing the search for FHAA issues (Figure 7), we examine Familial Status and Disability cases in isolation. These cases are presumably more volatile than those with other bases, since these represent a new area of law at the time the data starts. There was no established case history for these decisions, and these early cases present an opportunity to study the establishment of precedent.
Figure 7. FHAA factors isolated
|
Pro-FHA Rulings |
Coef. |
Robust Std. Err. |
z |
P>|z| |
[95% Conf. |
Interval] |
|
|
|
|
|
|
|
|
|
Substantive |
-1.537433 |
0.2656806 |
-5.79 |
0.000 |
-2.058158 |
-1.016709 |
|
President’s party Dem |
0.7481422 |
0.2522167 |
2.97 |
0.003 |
0.2538065 |
1.242478 |
|
Familial Status |
-0.1438993 |
0.4117822 |
-0.35 |
0.727 |
-0.9509776 |
0.663179 |
|
Disability |
0.2848528 |
0.2752821 |
1.03 |
0.301 |
-0.2546901 |
0.8243957 |
|
_cons |
1.093691 |
0.2447356 |
4.47 |
0.000 |
0.6140182 |
1.573364 |
|
|
|
|
|
|
|
|
|
Log pseudo-likelihood = -200.33459 |
||||||
|
Number = 335, Wald chi2(4) = 39.71, Prob > chi2 = 0, Pseudo R2 = 0.1008 |
||||||
The Familial Status cases fared about as well as the norm in this sample group, with a small negative coefficient and a 72.7% chance there is no connection between a given ruling and the basis. Disability cases, including housing for mental health, drug addiction and AIDS sufferers, succeeded at slightly better rate, but still had a better than 30% chance of coincidence, compared to the other isolated factors. This indicates the judiciary responded to the inclusion of these new plaintiffs in about the same way they handled prior protected classes.
We then isolated the statistics for Minority judges in Race cases and Female judges in Familial Status and Sex based cases. Testing for minority judges in cases where both race and national origin were claimed, we found an anomalous 100% positive prediction in favor of the plaintiff. This is rather misleading since there were only five total cases meeting all three criteria. The anomaly may be attributable to the scarcity of minority Republicans on the bench – only 9 out of the 38 minority judges in our group were appointed during Republican administrations, and only a subset, if any, heard FHA cases in which the basis was “racial/national origin.” However, cases based on racial and national origin tend to involve generalized discrimination, often regarding zoning or restrictions, rather than targeting of individuals. The minority ruling anomaly may indicate favor for policy as a matter of principle rather than bias toward personal group.
The results for female judges ruling on familial status cases show no marked tendency that could not also be explained by other factors (Figure 8). Here too, the appointing president’s party is a much more significant factor in favor of FHA causes. The pseudo R2 is 0.1, which indicates there is a significant correlation between the dependent variable, “Pro-FHA Rulings,” and the independent variables. As before, the only significant factors are the very strong negative for substantive decisions and the strong positive for President’s party.
Figure 8. Women judges on familial status cases
|
Pro-FHA Rulings |
Coef. |
Robust Std. Err. |
z |
P>|z| |
[95% Conf. |
Interval] |
|
South + Border |
-0.1270568 |
.2953489 |
-0.43 |
0.667 |
-0.70593 |
0.4518165 |
|
Substantive |
-1.523171 |
.2641152 |
-5.77 |
0.000 |
-2.040827 |
-1.005514 |
|
Female + Familial Status |
0.706075 |
.7266919 |
0.97 |
0.331 |
-0.718215 |
2.130365 |
|
President’s party Dem. |
0.71924 |
.2522943 |
2.85 |
0.004 |
0.2247524 |
1.213728 |
|
Tenure |
0.00000928 |
.0000531 |
0.17 |
0.861 |
-0.0000948 |
0.0001134 |
|
Military Background |
0.1128697 |
.2520103 |
0.45 |
0.654 |
-0.3810614 |
0.6068008 |
|
_cons |
1.071425 |
.308783 |
3.47 |
0.001 |
0.4662211 |
1.676628 |
|
|
|
|
|
|
|
|
|
Log pseudo-likelihood = -200.52555 Number = 335, Wald chi2(6) = 38.3, Prob > chi2 = 0.0000, Pseudo R2 = 0.1000 |
||||||
The last table breaks down the decisions by appointing president of the presiding judge. In Figure 9, the political leaning of the judge is seen to be highly predictable when considering the policies of the respective administrations. The starkest trend is the increased polarization starting with Ford and Carter. The appointees of Presidents Nixon and Johnson both supported FHA causes during the studied period at exactly the same 60% rate. The samples for Eisenhower and Kennedy appointees are too small to be indicative of anything. After Nixon, Democratic appointees consistently ruled for FHA claimants much more often than their Republican counterparts, culminating with the highly supportive 77.8% from Clinton’s contributions to the bench.
Figure 9. Case distribution by Judges’ appointing President
|
Judges' Appointing President |
# |
For FHA |
Anti FHA |
% For |
Yrs. Senate Maj-Min |
|
Eisenhower |
2 |
2 |
0 |
100.0% |
6-2 |
|
Kennedy |
1 |
1 |
0 |
100.0% |
3-0 |
|
Johnson |
20 |
12 |
8 |
60.0% |
5-0 |
|
Nixon |
25 |
15 |
10 |
60.0% |
0-5 |
|
Ford |
23 |
12 |
11 |
52.2% |
0-3 |
|
Carter |
102 |
71 |
31 |
69.6% |
4-0 |
|
Reagan |
114 |
63 |
51 |
55.3% |
6-2 |
|
Bush, GHW |
30 |
17 |
13 |
56.7% |
0-4 |
|
Clinton |
18 |
14 |
4 |
77.8% |
2-0 |
|
Total |
335 |
207 |
128 |
61.8% |
|
Included in Figure 9 is a listing of how many years each President enjoyed a majority in the Senate[10]. These figures are listed only for the period of this research ending in 1995, so there is no data for the next 6 years in which the Clinton administration faced a Republican Senate. President Ford’s very conservative appointees facing a hostile Senate, and President Bush’s continuation of Reagan policies stand out as examples of successful negotiation, or perhaps the last gasp of Senatorial Courtesy.
Conclusion
None of the regional, demographic or basis factors we examined had a significant effect on case outcomes, or these effects were subsumed by and explainable purely by party characteristics. Although minority judges found for minority FHA plaintiffs in 100% of cases studied, the small sample group of five cases makes the significance of this unconvincing. Party predicts policy in our sample, and political affiliation holds sway even over non-traditional judges.
In brief, we found that ideology transcends stereotypical categorization in federal Fair Housing decision-making. This may be attributable to sincere valuation of facts and law through the lens of foundational personal belief systems. If so, our results might actually be less polarized than if inherent checks of propriety and duty were not in place. Our data does not separately test outcomes for appointees who faced out-party approval, another moderating factor. Each President’s appointees could be isolated to evaluate the success of his nominees in promoting his policies on particular issues. However, rigorous content analysis of judicial selection records as well as judicial writings and opinions may be essential to determine whether explicit intent or involuntary inclination is more of a root factor in explaining the apparent political biases in federal judicial decision-making.
Bibliography
Baum, Lawrence. The Puzzle Of Judicial Behavior. Ann Arbor: University of Michigan Press, 1997.
Bernstein, David E. and Ilya Somin. “BOOK REVIEW: Judicial Power and Civil Rights Reconsidered: From Jim Crow to Civil Rights: The Supreme Court and the Struggle for Racial Equality. By Michael J. Klarman.” Yale Law Journal, December, 2004, 114 Yale L.J. 591. LexisNexis Academic. LexisNexis. University Of North Texas, 25 April 2005 <http://www.lexisnexis.com/cis>.
Caldeira, Gregory A., Marie Hojnacki and John R. Wright. “The Lobbying Activities of Organized Interests in Federal Judicial Nominations.” The Journal of Politics Vol. 62, No. 1 (Feb., 2000): 51-69. JSTOR, University Of North Texas, 9 February 2005 <http://www.jstor.org>.
Gaetke, Eugene R. and Robert G. Schwemm. “Government Lawyers and Their Private ‘Clients’ Under the Fair Housing Act.” George Washington Law Review, March, 1997. LexisNexis Academic. LexisNexis. University Of North Texas, 8 February 2005 <http://www.lexisnexis.com/cis>.
Giles, Micheal W., Virginia A. Hettinger and Todd Peppers. “Picking Federal Judges: A Note on Policy and Partisan Selection Agendas.” Political Research Quarterly Vol. 54, No. 3 (Sep., 2001): 623-641. JSTOR, University Of North Texas, 9 February 2005 <http://www.jstor.org>.
Iglesias, Tim. “Thirteenth Annual Symposium On Contemporary Urban Challenges: Urban Equity: Considerations Of Race And The Road Towards Equitable Allocation Of Municipal Services: Clarifying The Federal Fair Housing Act's Exemption For Reasonable Occupancy Restrictions.” Fordham Urban Law Journal, October, 2004. LexisNexis Academic. LexisNexis. University Of North Texas, 8 February 2005 <http://www.lexisnexis.com/cis>.
Nelson, John. “The Perpetuation Of Segregation: The Senior Housing Exemption In The 1988 Amendments To The Fair Housing Act.” Jefferson School of Law San Diego, Thomas Jefferson Law Review, Fall, 2003. LexisNexis Academic. LexisNexis. University Of North Texas, 8 February 2005 <http://www.lexisnexis.com/cis>.
Orfield, Gary. “Federal Policy, Local Power, and Metropolitan Segregation.” Political Science Quarterly Vol. 89, No. 4 (Winter, 1974-1975): 777-802. JSTOR, University Of North Texas, 9 February 2005 <http://www.jstor.org>.
Ringquist, Evan J. “A Question of Justice: Equity in Environmental Litigation, 1974-1991.” The Journal of Politics Vol. 60, No. 4. (Nov., 1998): 1148-1165. JSTOR, University Of North Texas, 9 February 2005 <http://www.jstor.org>.
Ringquist, Evan J. and Craig E. Emmert. “Judicial Policymaking in Published and Unpublished Decisions: The Case of Environmental Civil Litigation.” Political Research Quarterly Vol. 52, No. 1 (Mar., 1999): 7-37. JSTOR, University Of North Texas, 9 February 2005 <http://www.jstor.org>.
Rowland, C. K. and Robert A. Carp. Politics and Judgment In Federal District Courts. Lawrence: University Press of Kansas, 1996.
Schwemm, Robert G. “Discriminatory Housing Statements And 3604(C): A New Look At The Fair Housing Act's Most Intriguing Provision.” Fordham Urban Law Journal, October, 2001. LexisNexis Academic. LexisNexis. University Of North Texas, 8 February 2005 <http://www.lexisnexis.com/cis>.
Segal, Jennifer A. “Representative Decision Making on the Federal Bench: Clinton's District Court Appointees.” Political Research Quarterly Vol. 53, No. 1 (Mar., 2000): 137-150. JSTOR, University Of North Texas, 9 February 2005 <http://www.jstor.org>.
Steffensmeier, Darrell and Stephen Demuth. “Ethnicity and Sentencing Outcomes in U.S. Federal Courts: Who is Punished More Harshly?” American Sociological Review Vol. 65, No. 5 (Oct. 2000), 705-729. JSTOR, University Of North Texas, 9 February 2005 <http://www.jstor.org>.
Walker, Thomas G. and Deborah J. Barrow. “The Diversification of the Federal Bench: Policy and Process Ramifications.” The Journal of Politics Vol. 47, No. 2 (Jun., 1985): 596-617. JSTOR, University Of North Texas, 9 February 2005 <http://www.jstor.org>.
Walsh, James D. “Reaching Mrs. Murphy: A Call for Repeal of the Mrs. Murphy Exemption to the Fair Housing Act.” Harvard Civil Rights-Civil Liberties Law Review, Summer, 1999. LexisNexis Academic. LexisNexis. University Of North Texas, 8 February 2005 <http://www.lexisnexis.com/cis>.
Copyright © Jack Bieler, 2005
[1] “Filibuster Derails Supreme Court Appointment”. U.S. Senate website. Accessed April 25, 2005. http://www.senate.gov/artandhistory/history/minute/Filibuster_Derails_Supreme_Court_Appointment.htm.
[2] “Robert Bork”. Wikipedia, the free encyclopedia. Accessed April 25, 2005. http://en.wikipedia.org/wiki/Robert_Bork.
[3] Seifter, Andrew. “Hannity, Kondracke deceptively labeled Dems' use of filibuster as break with precedent”. Media Matters for America. March 6, 2005. Accessed April 25, 2005. http://mediamatters.org/items/200503160004/.
[4] Coyle, Marcia. “Awaiting fate of the 'filibuster 10'; Circuit nominees may escape logjam. Ideology and control; Trying again”. The National Law Journal. November 15, 2004. The Committee For Justice. Accessed April 25, 2005. http://committeeforjustice.org/cgi-data/news/files/50.shtml.
[5] “Judges, Magistrates, and Other Judicial Workers” http://www.bls.gov/oco/ocos272.htm.
[6] Federal Judicial Center. “Federal Judges Biographical Database.” Accessed April 21, 2005. http://air.fjc.gov/servlet/uSpage.
[7] “US Code Chapter 45 - Fair Housing - Subchapter I, Section 3617”.
[8] “United States Courts of Appeal for the First Circuit.” Wikipedia, the free encyclopedia. Last updated May 1, 2005. Accessed May 5, 2005. http://en.wikipedia.org/wiki/United_States_Court_of_Appeals_for_the_First_Circuit.
[9] “Statistical Significance.” Number Watch. Accessed May 5, 2005. http://www.numberwatch.co.uk/significance.htm.
[10] Senate Historical Office, December 1996. Congress Link. “U.S. Senate Majority and Minority Leaders.” Accessed May 8, 2005. http://www.fountcom.com/folio/sources/leaders.html.