Florida State University
Reevaluation of published research on racial bias in criminal sentencing and of data on execution rates by race from 1930 to 1967 and on death-sentencing rates from 1967 to 1978 indicates that, except in the South, black homicide offenders have been less likely than whites to receive a death sentence or be executed. For the 11% of executions imposed for rape, discrimination against black defendants who had raped white victims was substantial, but only in the South. Evidence for noncapital sentencing also largely contradicts a hypothesis of overt discrimination against black defendants. Although black offender-white victim crimes are generally punished more severely than crimes involving other racial combinations, this appears to be due to legally relevant factors related to such offenses. Crimes with black victims, however, are less likely than those with white victims to result in imposition of the death penalty. The devalued status of black crime victims is one of several hypothetical explanations of the more lenient sentencing of black defendants.
The legitimacy of the legal systems of modern democracies depends heavily on the degree to which the systems operate in a manner consistent with their own stated procedural standards of justice. It has been argued that Western societies, including the United States, are undergoing a “legitimation crisis” and that this is occurring specifically in criminal justice systems at least partly because they fail to live up to their stated commitments to treatment of defendants without regard to ascribed personal characteristics such as race, ethnicity, and gender or partially ascribed characteristics such as status or class position (Quinney, 1974; Chambliss and Seidman, 1971).’
It is widely believed, and frequently stated, that the criminal justice system has been in the past, and remains, racially discriminatory (e.g., Sutherland and Cressey, 1970; Clark, 1970). The most frequently cited category of evidence for this assertion has been research indicating more severe sentencing of black criminal defendants than white defendants, especially in imposition of the death penalty. As there have been at least sixty empirical studies of adult criminal sentencing published which refer to race, it is not surprising that at least one critic of the criminal justice system has asserted that evidence on racial discrimination in sentencing is probably the strongest evidence of racial bias in the criminal justice system (Over by, 1971:575). Because the outcomes of sentencing decisions are among the most visible of legal processing, the legal system’s claim to legitimacy is especially dependent on the public’s perception of the pattern of such outcomes. Therefore, it seems particularly important to take a close look at evidence bearing on this issue.
The first part of this paper attempts a comprehensive assessment of the published scholarly empirical research on racial bias in criminal sentencing in the U.S. in connection with both capital punishment and noncapital sentencing. One of the principal sources of distortion regarding this issue in the past has been selective citation of studies supporting one position or another; therefore, great care has been I As used in this paper the word “race” is socially defined, referring to “a human group that defines itself and/or is defined by other groups as different from other groups by virtue of innate and immutable physical characteristics” (van den Berghe, 1967:9). taken to be as exhaustive as possible in finding relevant studies. The second part of the paper presents new evidence on race and capital punishment, consisting of an analysis of execution rates for blacks and whites over the period 1930 to 1967, for the United States as a whole and for the South, and of death-sentencing rates for the period from 1967 to 1978.
The Varieties of Racial Bias: Some Conceptual Distinctions
At least five different practices can produce racial differentials in criminal sentences which are likely to be viewed as illegitimate or unjust.
(/) Overt racial discrimination against minority defendants. This refers to the imposition of more severe dispositions on members of a subordinate racial group, independent of their legally relevant individual merits, and primarily as a direct result of the conscious or unconscious racial prejudice of the sentencing decision makers.
(2) Disregard for minority crime victims. This would include the failure to sentence offenders (of any race) who victimize minority-group members as severely as those who victimize nonminority-group members.
(3) Class discrimination. This refers to more severe treatment of lower-class defendants as a consequence of class prejudice. It may be due to hostility or indifference of middle-class decision makers toward culturally different defendants, or because lower-class defendants better fit popular stereotypes of serious or dangerous criminals. Because blacks in. the United States are disproportionately members of the lower class, class discrimination would affect them more heavily than whites, independent of any overt racial discrimination. This assumes that among the set of criminal defendants blacks are more likely to be lower class than whites, a debatable assumption considering the overwhelmingly lower-class character of criminal defendants of all races.
(4) Economic discrimination. When a society’s legal system is structured so that significant private economic resources are required in order to effectively obtain full legal protection, this constitutes economic discrimination, even where there is no class discrimination (as defined above). If low-income defendants receive more severe sentences than middle-income defendants because they cannot afford to hire an outside private attorney or cannot make bail, this constitutes economic discrimination and could produce racial differentials in sentencing outcomes.
(5) Institutional Racism. This refers to the application, possibly in a universalistic fashion, of decision making standards which in themselves have considerable consensual support (possibly even among minority members) but which result in less favorable outcomes for minority defendants. As used in the past, the term seems to have referred to, among other things, practices (2) through (4); but institutional racism in sentencing can take other forms as well. For example, if racial-minority defendants are likelier to have prior criminal convictions, then the use of prior record as a criterion for sentence determination will tend to produce less favorable outcomes for minority defendants and would therefore be an instance of institutional racism (for this view, see Burke and Turk, 1975, or Farrell and Swigert, 1978a). The establishment, by legislatures, of higher statutory penalties for crimes committed more frequently by racial minority members than by others (such as violent interpersonal crimes) would also constitute institutional racism, regardless of the behavior of judges, prosecutors, and others who influence sentencing outcomes.
For the sake of verbal and conceptual clarity, it is misleading to label either (3) or (4) ” racial” bias or discrimination. Although clearly unjust and certainly related to race in the United States, these practices are not directly racial in themselves, since they affect whites as well as blacks and could occur in jurisdictions or societies where no racial distinctions of any sort were made. Therefore, although reference will be made to them in examining the evidence regarding overt racial discrimination, these practices are not themselves primary objects of the analysis.
The concept of institutional racism is highly problematical. It is so flexible that any practice producing unfavorable sentencing outcomes for racial-minority members can be characterized as racist, no matter how the outcomes were produced, whether they were intentionally sought, and regardless of what criteria were involved in the decisions producing the outcomes. Any pattern of sentencing involving the crimes commonly dealt with by U.S. criminal courts could be construed as institutional racism, since blacks commit a disproportionate share of such crimes relative to their share of the population (see Hindelang, 1978, for a thorough discussion of the evidence for this statement) and therefore are bound to receive a disproportionate share of the criminal punishment, no matter how fairly the sentencing process is administered. Only an alteration of the social conditions producing differentials in the racial distribution of criminal behavior or a radical redefinition of which crimes the courts focus on could eliminate institutional racism of this sort. For these reasons, neither this study nor any other study of sentencing per se could reject the hypothesis of institutional racism in sentencing. Consequently, attention in this analysis will be primarily focused on overt racial discrimination and secondarily on disregard for minority crime victims.
ASSESSMENT OF PRIOR RESEARCH
This review is intended to be an exhaustive assessment of all scholarly empirical studies of race and criminal sentencing of adults in the United States published up through 1979. It does not cover the few studies relating ethnicity and sentencing, such as Castberg (1971) or Hall and Simkus (1975), nor studies of conviction, as opposed to sentencing, such as Forslund (1969). It does cover studies of the determination of degree of homicide of which defendants were convicted, such as Farrell and Swigert (1978a; 1978b), since such a determination is tantamount to determination of sentence. Also included are studies of commutations of death sentences, as these have commonly been cited as sentencing studies.
When two or more studies of the same data set, using very similar methods, have been published, only one is included, or the studies are treated as a single study (e.g., Wolfgang and Riedel, 1973 and 1975; Farrell and Swigert, 1978a and 1978b, and Swigert and Farrell, 1977; Lotz and Hewitt, 1977 and Hewitt, 1976). Studies of juvenile court dispositions are excluded since, properly speaking, juveniles are not sentenced to criminal penalties and determination of disposition is in any case substantially different from the adult sentencing process.2 Further, the following kinds of studies are excluded: anecdotal or journalistic accounts, case studies, hypothetical-case sentencing studies (e.g., Johnston et al., 1973), purely theoretical studies, and other reviews of the literature. The review also ignores studies such as those of Sellin (1928, 1935) which simply compare sentence lengths of persons sentenced or executed, without any comparison to numbers arrested, convicted, etc. Finally, unpublished studies such as dissertations are excluded because the availability of such material is limited.
Studies which fit the selection criteria were located through an iterative search process. An initial list of relevant studies was compiled through a search of Sociological Abstracts, Crime and Delinquency Abstracts, Legal Abstracts, and Hagan’s (1974) bibliography. (About one-third of the studies reviewed here were also included in Hagan’s review.)
2 Liska and Tausig (1979) reviewed eight studies which examined race and juvenile court dispositions, noting that five of them show a significant race effect in the conventionally expected direction. However, only two of the five simultaneously controlled for offense seriousness and prior record. The crude state of research in this field is also indicated by the almost complete absence of attention to controlling for effects of juveniles’ family circumstances or disposition. Although the philosophy of the juvenile court explicitly defines family stability as a legitimate factor to be considered, Liska and Tausig do not say a word suggesting that racial differences in family intactness could account for differences in juvenile court dispositions. 3 However, for a sampling of recent doctoral dissertation on the subject the reader may consult Dison (1976), Hutner (1977), and San Marco (1979). A reading of their abstracts indicates findings highly congruent with the conclusion of my review
Then the references listed in each of these studies were examined for further relevant studies, and so on, until no further leads were uncovered.
Capital Punishment Sentencing Studies
Table 1 summarizes in compact form the prior scholarly empirical research on racial discrimination and use of the death penalty, first regarding murder and then rape. The last column indicates whether racial differences were statistically significant and in the predicted direction. In a minority of studies my assessment of the evidence presented in a study differs from that of the authors. Two examples of this are noteworthy.
Bedau (1964) concluded that there was racial bias in the final disposition of persons sentenced to death, despite his own acknowledgment that his data showed no significant relationship between race and final disposition, and despite the fact that the observed relationship was in the opposite direction to that indicating discrimination against nonwhites. In New Jersey 66.2% of nonwhites sentenced to die were executed, compared to 68.4% of whites (Bedau, 1964:19). Further, Bedau claimed there was no significant relationship between race and execution when felony and non-felony murders were separated, but reanalysis of his data reveals that in fact nonwhite murderers were significantly less likely to be executed for felony killings than were white felony murderers (X2 = 8.114, 1 degree of freedom, p = .01). There was no race difference for non felony killings.
Bowers (1974:81-107) claimed to have found evidence which pointed “unmistakably to a pattern of racial discrimination in the administration of capital punishment in America” (p. 10). His conclusion was not limited to the South or to use of the death penalty for rape, although he laid particular stress on these areas. Regarding racial bias in use of the death penalty for murder, the evidence that Bowers produced was of two kinds: he argued first that the lower mean age at execution of nonwhites compared to whites was evidence of racial discrimination, and, second, that the lower percentage of cases appealed by nonwhites compared to whites indicated discrimination. While Bowers acknowledged that the age differentials at execution may simply reflect age differences at commission of the offense and arrest, he argued that the difference at arrest was not as large as that at execution. In fact, there was a 2.4 year difference in mean age at execution in Bowers’ data (p. 80), while Wolfgang’s (1958:70) homicide arrest data indicate an almost identical 2.3 year difference in the median age of black and white male homicide arrestees. Regarding the data on appeals, racial differentials in percentage receiving appeals of death sentences for murder since 1940 are largely confined to the South (see especially Bowers’ Table 3-9), consistent with our interpretation of the findings of previous death-penalty discrimination studies. Nevertheless, Bowers concludes that, as indicated by age differentials at execution and differentials in percentage receiving higher appeals, ” racial discrimination in northern and western states began to rival that in the South, at least for the period of decline in capital punishment” (1974:104).
Probably the most serious shortcoming of death-penalty discrimination studies is that they nearly all fail to control for prior criminal record. The one study which introduced such a control, Judson et al. (1969), found no evidence of racial discrimination, suggesting that apparent racial differences in other studies may actually have been due to racial differences in prior criminal activity. This hypothesis is supported by Wolfgang’s (1958:175-6) findings that black homicide offenders were significantly more likely to have previous criminal arrest records than white homicide offenders. It is further supported by Hagan’s (1974:366-8) reanalysis of Nagel (1969), which indicated that where crude controls for prior record (record/no record) were introduced, racial effects shrank. Where more adequate controls for prior record (number of prior convictions) were introduced, the racial difference disappeared altogether (Judson et al., 1969:1366-76), suggesting that the dichotomous measure of prior record may be inadequate for control purposes (see also Green, 1961:11 on this point).
All of the studies purporting to find racial bias in use of the death penalty for murder failed to control for income, class, or occupation of the defendants.4 However, the most methodologically sophisticated study of the subject, which did control for defendant’s occupation, found no racial effect on whether or not a death sentence was imposed for murder by California penalty juries (Judson et al., 1969:1366-76). Further, they found no relationship between the victim/offender racial relationship and sentence imposed, suggesting that the findings of Johnson (1941) and Garfunkel (1949) may have reflected regional and temporal peculiarities characteristic of North Carolina (or more generally, the South) in the 1930s and earlier.
Several points should be noted about the pattern of findings on discrimination in use of the death penalty for murder. First, every single study consistently indicating discrimination towards blacks was based on older data from Southern states, and three of these four studies were based on overlapping data from North Carolina. Second, all of the studies finding discrimination in administration of capital punishment for murder were not in fact studies of sentencing, although most of them have been cited in the research literature as if they were. Mangum (1940), Johnson (1957), Wolfgang et al. (1962), and Bridge and Mosure (1961) all studied commutation of death sentences, not sentencing itself, while Bowers’ (1974) data largely concerned appeals of death sentences. Third, all of these studies failed to control for prior criminal record of the defendant, for the defendant’s class or income, or for the distinction between felony and non felony killings. Since studies which do introduce such controls find that they reduce the sentencing differentials between blacks and whites (Green, 1961; Judson et al., 1969; Nagel, 1969), even in the South the racial differential may have been due to differences in criminal record, income, or type of homicide committed rather than discrimination. The evidence considered as a whole indicates no racial discrimination in use of the death penalty for murder outside the South, and even for the South empirical support for the discrimination hypothesis is weak.
Regarding the use of capital punishment for rape, the evidence strongly suggests overt discrimination against black defendants. Four of the studies of this issue found evidence of discrimination, while the relationship between race and the carrying out of the death sentence (as opposed to commutation of the sentence) was not significant in the Johnson (1957) study, according to Hagan’s (1964:370) reanalysis. The relationships found in the other four studies (Johnson, 1970; Florida Civil Liberties Union, 1964; Partington, 1965; Wolfgang and Reidel, 1973, 1975) are very strong, and the evidence indicates that the death penalty for rape was largely used for punishing blacks who had raped whites. Although all of these studies were methodologically crude, it is doubtful if additional controls could eliminate the huge racial differentials in use of the death penalty. The importance of this conclusion, however, is limited by several facts. First, the death penalty has rarely been used to punish rape. Only 11.8% of the executions from 1930 to the present were for rape. Virtually all the rest were for murder. Second, the use of the death penalty for rapes has always been, at least since national data on executions were first gathered in 1930, strictly a peculiarity of the South. Not a single execution for rape occurred outside the South or the border states during that period (U.S. Federal Bureau of Prisons, 1970). Third, the imposition of death sentences for rape has virtually disappeared. Of the 183 persons who were sentenced to death during 1978, only one was sentenced for rape (U.S. NCJISS, 1979:25). Thus, the rape discrimination conclusion is of historical significance with regard to capital punishment in the South, but has limited relevance to current debates over capital punishment, especially since the United 4Although both Wolfgang et al. (1962) and Johnson (1957) had data on occupation of offenders as well as race, they did not attempt to simultaneously control for race and occupation. No explanation of this conspicuous omission was offered in either study. States Supreme Court declared, in Coker vs. Georgia (1977), that the use of the death penalty for the rape of an adult woman was a disproportionate penalty and therefore unconstitutional (U.S. NCJISS, 1979:2).
Noncapital Punishment Sentencing
Studies summarizes the empirical research on sentencing involving penalties other than the death penalty. The results of each study are simply summarized in the last column of the table (“Discrimination?”), with each study characterized as to whether its findings largely supported a discrimination hypothesis (indicated by “Yes”), were mostly inconsistent with such a hypothesis (“No”), or were only partially consistent with the hypothesis (“Mixed”). Mixed findings most frequently occurred when more than one crime was studied, when sentencing patterns of a number of judges were reviewed, or when more than one measure of sentencing outcome was examined.
Studies were classified, in somewhat arbitrary fashion, according to what proportion of their findings were in favor of the discrimination hypothesis. They were characterized as mixed if from one-third to one-half (inclusive) of the findings favored the discrimination hypothesis and as favorable to the hypothesis if more than one-half of the findings favored it. For example, if a study examined eight different offenses, it would be labeled “Yes” if evidence of bias against black defendants was found for four or more offenses, as “Mixed” if such evidence was found for three of the offenses, and as “No” if two or fewer offenses showed such evidence. Since it could -be argued that evidence of discrimination even for one crime or sentence-outcome measure out of many is evidence worth taking very seriously, readers must judge for themselves the significance of the “mixed” findings.
Under the heading “Sentencing Measure” in Table 2, the dependent variable in each study is noted. The term “disposition” indicates that the dependent variable distinguished between categories like probation, jail sentence, prison sentence, etc., while “sentence severity” denotes a single scale of severity of disposition or sentence constructed by the researcher. The other terms are self-explanatory.
Table 2 also indicates whether the authors of these studies in any way controlled for the type of criminal offense involved. In those studies where only one type of crime was involved, or where several very similar offenses were studied, such a control was obviously unnecessary. However, where several different offense types were lumped together, as in Cargan and Coates (1974) or Pope (1975a), differences in sentence received by black and white defendants could be at least partly attributable to differences in the seriousness of the types of offenses for which they were prosecuted.
Of the 40 studies listed in Table 2, only eight consistently support the racial discrimination hypothesis, while 12 are mixed and the remaining 20 produced evidence consistently contrary to the hypothesis. Since a study’s findings were characterized as mixed even if as few as a third of them favored the discrimination hypothesis, this means that a substantial majority of all of the findings of these 40 studies contradicted the hypothesis.
However, the evidence for the hypothesis is even weaker than these numbers suggest, since of the minority of studies which produced findings apparently in support of the hypothesis, most either failed completely to control for prior criminal record of the defendant, or did so using the crudest possible measure of prior record-a simple dichotomy distinguishing defendants with some record from those without one. This is probably the most important flaw in studies drawing a conclusion of racial discrimination, 5 Use of these standards occasionally resulted in characterizations of findings which differed from those of the original authors (e.g., Bedau, 1964). Since I attempted to accurately reflect the studies’ theoretical or ideological preferences, when the original authors’ conclusions did not seem congruent with their data, their conclusions were discounted. Skeptical readers are encouraged to examine the original studies in order to judge for. Themselves the accuracy of my characterizations. It, NJ since the most methodologically sophisticated sentencing studies have consistently shown various measures of prior record to be either the strongest predictor, or among the strongest predictors, of sentences received (Chiricos and Waldo, 1975; Bernstein et al., 1977; Lotz and Hewitt, 1977; Lizotte, 1978). It appears to be the case that the more adequate the control for prior record, the less likely it is that a study will produce findings supporting a discrimination hypothesis.
Table 3 summarizes the whole body of prior research on race and sentencing, both capital and noncapital. Simply adding up the number of studies favoring or not favoring the discrimination hypothesis could be somewhat misleading, since some studies are clearly better than other s and should therefore be weighted more heavily than others in assessing the body of evidence as a whole. Therefore, although it would be difficult to assign exact weights, some simple quality distinctions can be made, such as distinguishing between studies which control for prior criminal record and studies which do not. Regarding noncapital punishment, Table 3 makes clear the importance of such controls-one-third of the studies without a control for prior criminal record support a discrimination conclusion, while less than a tenth of those with such controls support a discrimination conclusion. Regarding capital punishment, separate tallies of studies with and without such controls are unnecessary, since only one study, that of Judson and his colleagues (1969), controlled for prior record, finding no evidence of racial discrimination either in the sentencing of black defendants in general or in sentencing of those who had victimized whites.
Interracial Relationship of Offender and Victim
It has long been argued that racial bias in sentencing is not to be detected only by looking at the race of the defendant, but by noting the racial relationship of the offender and the victim (e.g., Johnson, 1941). Specifically, it is asserted that crimes involving black offenders and white victims are punished more severely than crimes involving the other three racial combinations, either because crimes involving black victims are taken less seriously or because the crossing of racial lines in the commission of a crime is taken very seriously (Johnson, 1941; Garfunkel, 1949). While black offender-white victim crimes, especially homicides and rapes, are punished more severely than crimes with other racial combinations, it is unclear whether this is due to the racial character of the crime, or to related, confounding factors. Black offender-white victim killings are more likely than other killings to involve an offender and a victim who are strangers to each other, and such killings are much more severely punished regardless of the races involved (Lundsgaarde, 1977:232). Such killings are also more likely to be committed in connection with some other felony, like robbery. Data in Block and Zimring (1973:8) indicate that for Chicago homicides in 1970, 38% of killings with black offenders and white victims were robbery killings, while only 5% of the white offender-white victim killings were robbery killings. Felony killings are punished more severely than other homicides, regardless of races involved (Wolfgang et al., 1962; Bedau, 1964; Wolf, 1964). Finally, black-white killings are less likely than black-black killings to be victim precipitated, and victim-precipitated killings in turn are less likely to be premeditated (Wolfgang, 1958), leading one to expect less severe punishment of black on black killings for this reason, rather than the racial relationship per se. Eleven studies have examined sentencing outcomes by racial combination, and of these, seven (Johnson, 1941; Garfunkel, 1949; Florida Civil Liberties Union, 1964; Howard, 1967; Southern Regional Council, 1969; Wolfgang and Reidel, 1973; Zimring et al., 1976) found more severe punishment for black-white offenses.6 However, none of these studies controlled for the possibly confounding factors we have mentioned. The only four studies which did introduce such controls (Green, 1964; Judson et al., 1969; Farrell and Swigert, 1978b; Myers, 1979) all found no evidence of such sentencing patterns. Thus, consideration of the pattern of findings as a whole strongly suggests that the interracial relationship itself does not affect the sentencing decision, except in connection with the punishment of rape in the South (Florida Civil Liberties Union, 1964; Howard, 1967; and Wolfgang and Reidel, 1973, 1975 support this limited assertion of discrimination).
Table 3. Summary of Prior Research
Yes Mixed No Total
All Studies 7 4 6 17
Murder 3 4 5 12
Rape 4 0 1 5
All Studies 8 12 20 40
Control for prior record 2 8 13 23
No control for prior record 6 4 7 17
NOTE: See the discussion of prior literature in the text for an explanation of the classification of studies by their results.
Examination of prior studies on the question of racial discrimination and use of the death penalty for murder has suggested that many of their conclusions may be seriously time-bound and regionbound. Their findings may not be generalizable to areas outside the South, considering the generally contrary findings of studies of non-Southern jurisdictions using more recent data. Given these considerations, it would seem reasonable to study national sentencing practices, making regional comparisons, using data covering as long a period of time as possible.
EXECUTION RATES AND DEATH SENTENCING BY RACE
It has been claimed that “racial discrimination is strongly suggested by the national execution figures” (NAACP, 1971: 51-2). Clearly, blacks have been executed in numbers far out of proportion to their numbers in the population. Over the period 1930-1976, 53.6% of all legally executed persons in the United States were black, although blacks constituted only about 10-11% of the U.S. population during that period (U.S. Federal Bureau of Prisons, 1971:8; U.S. Bureau of the Census, 1977:25). This disproportion, however, cannot in itself be taken as evidence of racial discrimination, since blacks also commit a large proportion of U.S. homicides, the crime most frequently punished by death. A more meaningful measure of capital punishment sentencing outcome would be an indicator of execution risk, i.e., an execution rate. A true rate compares a number of events (such as executions) with the number of times the event could have occurred. Therefore, the ideal base for the execution rate could be the number of persons convicted of a capital offense, i.e., a crime for which, in a given jurisdiction, the offender could be sentenced to death. However, there are no national data on the number of such crimes committed or on persons arrested for the crimes. Therefore, a surrogate measure is needed.
In this analysis, execution risk by race is measured as the number of executions (for murder) of persons of a given race in a given year, divided by the number of homicide victims of that race who died in the previous year. The number of homicide victims of a given race is used as an approximation of the number of persons of that race who committed a homicide, whether a capital murder or a noncapital murder.7 Since 92-97% of all homicides involve killers and victims of the same race (Garfunkel, 1949:371; Harlan, 1950:745; Wolfgang, 1958:379; Bensing and Schroeder, 1960:51; U.S. Federal Bureau of Investigation, 1977:9), the racial distribution of homicide victims can be used to describe the racial distribution of homicide offenders with very little error (Wolfgang, 1958:223).
6 Although Bullock (1961) has been cited in connection with the issue of sentencing and interracial relationships (e.g., Hindelang, 1969 and Baab and Ferguson, 1968), his study did not actually contain any data on victim-offender racial relationships.
7 Execution rates for rape cannot be computed because there are no comparable data to use for the base of the rate. There were no national data on rape victimizations by race up until 1973 (by which time the judicial moratorium on execution had begun and even the imposition of death sentences for rape had virtually disappeared). In any case, the evidence showing discrimination in capital punishment of rape is fairly conclusive, making the computation of such rates redundant.
Since there is a median lag between arrest and a court trial for criminal homicide of slightly under six months (Wolfgang, 1958:296, 299), and a mean lag of about one year between conviction and execution (Lunden, 1962:1043; McCafferty, 1967:95; U.S. Federal Bureau of Prisons, 1970), the appropriate comparison for our purposes is between the executions in year t and arrests in year t- 1, or possibly year t-2. For the sake of simplicity, the execution rates assume a one-year lag between commission of the homicide and execution of the offender over the time period studied. In any case, the results assuming a two-year lag would be substantively identical.
Table 4 presents the computed execution rates for blacks and whites, covering the entire period for which national execution data is available, 1930-1967.8 In the final column, a ratio greater than one indicates a black execution rate higher than the white execution rate; therefore, for 25 of 38 of the years examined, the black execution rate was lower than the white execution rate. Since they are based on fairly small numbers of executions, race specific execution rates and ratios of execution rates are somewhat unstable for single years, especially for the later years in the time series. Therefore the rates for the entire period were computed. For the period 1930-1967 there were 1,663 executions of whites for murder and 1,638 executions of blacks, while for the period 1929-1966 (lagged one year behind the other period) there were 159,482 white homicide victims and 168,518 black homicide victims (and presumably roughly equal numbers of homicide offenders). Therefore, the white execution rate for the entire period was 10.428 executions per 1,000 homicides and the black rate was 9.720 executions per 1,000 homicides. Thus, over the entire period, blacks were subject to a lower execution risk than whites.
Given the regional pattern of discrimination findings of previous studies of capital punishment sentencing, it may be the case that execution rates are higher for blacks than for whites in the South and that this fact is obscured in national data. It is also possible that the relative execution risks of blacks and whites changes over time and by region. These possibilities are addressed using the data in Table 5.
These data indicate that the execution risk of black homicide offenders (actually nonwhites in this analysis) has indeed been greater than that of white homicide offenders in the South, while the opposite has been true in the rest of the United States. However, the excess of the black execution risk over the white execution risk in the South has declined over time, to the point where execution rates were roughly equal in the period since 1950. The evidence, considered in combination with prior research on capital punishment sentencing outcomes, suggests that use of the death penalty is not inevitably or inherently discriminatory, but rather that racial discrimination in its administration has been highly variable over time and between regions. These data support the racial discrimination hypothesis in connection with death penalty sentencing only for the South. Of particular interest is the somewhat surprising finding that in the recent past, outside of the South, the white execution risk has been substantially higher than the nonwhite risk, a fact which apparently has gone unnoticed in the literature. Possible explanations of this phenomenon will be discussed later in the paper.
8 The number of homicide victims of each race excludes executions and killings committed by policemen in the line of duty. Executions are excluded because it is undesirable to have a common component in the numerator and denominator of the execution rate. Police killings are excluded because they are nearly always considered justifiable homicides and therefore not criminal. These exclusions make the homicide victim figures somewhat better surrogates for figures on criminal homicide offenders.
9 It is debatable whether statistical tests of significance are appropriate where population data are involved, although Blalock (1972:238-9) has argued that they can serve to rule out an alternative explanation of a set of results-that the data could have been generated by chance processes rather than causal ones. A two-sample test of the difference between the proportions of persons executed among blacks and whites indicates the difference is significant at the .05 level (two-tailed test, Z=2.03).
Table 4. Execution Rates by Race, 1930-1967
Black-Black, White-White, Black-White Ratio of Black Executions Homicide Executions Homicide Execution to White
Year for Murder Victims a for Murder Victims a Rate B Rate C Execution Rates
1967 1 1 – 0.168 0.191 0.880
1966 0 5,945 1 5,230 0.000 0.205 0.000
1965 1 5,408 6 4,879 0.203 1.336 0.152
1964 4 4,926 5 4,492 0.893 1.197 0.748
1963 6 4,478 12 4,176 1.375 2.918 0.471
1962 15 4,364 26 4,112 3.583 6.468 0.554
1961 15 4,187 18 4.020 3.568 4.688 0.761
1960 26 4,204 18 3,840 6.394 4.826 1.325
1959 26 4,066 15 3,730 6.619 4.260 1.554
1958 20 3,928 20 3,521 5.040 6.073 0.839
1957 22 3,968 32 3,293 5.479 9.718 0.564
1956 31 4,015 20 3,239 8.105 6.240 1.299
1955 24 3,825 41 3,205 5.954 12.387 0.481
1954 33 4,031 37 3,310 8.317 11.315 0.735
1953 25 3,968 25 3,270 5.840 7.492 0.779
1952 36 4,281 35 3,337 9.217 10.965 0.841
1951 31 3,906 55 3,192 7.463 16.965 0.456
1950 32 4,154 36 3,362 7.402 9.882 0.749
1949 56 4,323 49 3,643 12.216 12.626 0.967
1948 61 4,584 32 3,880 13.475 8.095 1.664
1947 89 4,527 40 3,953 18.924 10.005 1.891
1946 61 4,703 45 3,998 15.877 12.879 1.233
1945 52 3,842 37 3,494 14.790 12.445 1.188
1944 48 3,516 45 2,973 13.829 14.227 0.972
1943 63 3,471 54 3,163 14.593 16.162 0.903
1942 58 4,317 57 3,341 13.075 16.681 0.784
1941 46 4,436 55 3,417 10.426 14.773 0.706
1940 61 4,412 44 3,723 13.610 11.429 1.191
1939 63 4,482 79 3,850 14.338 18.283 0.784
1938 63 4,394 89 4,321 13.011 20.408 0.638
1937 62 4,842 67 4,361 12.086 13.405 0.902
1936 93 5,130 86 4,998 18.383 15.826 1.162
1935 66 5,059 115 5,434 12.028 18.338 0.656
1934 89 5,487 64 6,271 17:056 9.672 1.763
1933 74 5,218 75 6,617 16.122 12.093 1.333
1932 63 4,590 62 6,202 13.322 10.003 1.332
1931 57 4,749 76 6,198 12.800 12.722 1.006
1930 57 4,453 90 5,974 13.106 16.474 0.796
1929 4,349 5,463
SOURCEU:. S., Federal Bureau of Prisons, National Prisoner Statistics, Bulletin No. 46 (1971),p . 8; U.S., National Center for Health Statistics, Vital Statistics of the U.S.: Mortality, (annual issues, 1937-1966); U.S.
Bureau of the Census, Mortality Statistics, (annual issues, 1929-1936).
NOTE: There were no executions in the U.S., 1968-1976; complete execution figures by race for the U.S. before 1930a re not available. Mortality figures for 1929-1932 refer to the death registration area rather than the entire U.S. (95.7% of the U.S. population was covered in 1929; 96.3% was covered in 1932). In 1929-1932 black homicide victim figures were estimated from “colored” homicide figures; the difference is very slight.
a. Excluding executions and killings by policemen in the line of duty for 1950-1967. Figures before 1950 exclude executions but include killings by police.
b Black execution rate is number of black executions in year t per 1,000 black homicide victims in year t-1.
C White execution rate is number of white executions in year t per 1,000 white homicide victims in year t-1.
Possible Biases in Computation of Execution Risk Our estimates of execution risk by race could be biased if the homicides which blacks commit are less likely to be capital murders than those committed by whites. If this were true, using the number of homicide victims of each race as the base of the execution rate would be misleading for comparative purposes, since a smaller proportion of the black offenders could be a In regional and state breakdowns of mortality by cause of death, the data refer only to white/ nonwhite, while the published accumulations for the United States refer to white, black, and other races. considered to be at risk of execution, compared to white offenders. In this case, the black execution would be understated relative to the white rate.
Table 5. Execution Rates for Groups of Years by
Region and Race
Black Rate/ Years White Black and White Rate
1930-1939 14.38 14.24 0.99
1940-1949 12.80 14.07 1.10
1950-1967 5.94 4.57 0.77
1930-1967 10.43 9.72 0.93
Nonwhite Rate/ White Nonwhite a White Rate South
1930-1939 11.01 f4.41 1.31
1940-1949 11i24 14.26 1.27
1950-1967 5.02 5.30 1.06
1930-1967 8.39 10.47 1.25
1930-1939 15.08 12.56 0.83
1940-1949 13.61 13.78 1.01
1950-1967 6.11 3.34 0.55
1930-1967 11.00 9.32 0.85
Three studies report figures on the percentage of criminal homicides designated as first degree (capital) murders, by race. One found the percentage designated first degree murder, both at indictment and at conviction, to be higher for blacks than for whites (Garfunkel, 1949:372), and one found the opposite (Bensing and Schroeder, 1960:43, 45, 88), while the third study found no significant difference (Wolfgang, 1958:302-03). Thus, no consistent relationship was found between race and proportion of criminal homicides designated first degree murder.
However, it has been argued that the designation of degree of homicide could itself be racially biased (Garfunkel, 1949). Would there be a racial difference if the degree of homicide were designated without bias? Given the hypothetical nature of the question, this is not easy to answer directly; however, we can evaluate it indirectly. It is generally agreed that there are a number of factors which can legitimately affect the designation of degree, including whether or not the homicide was committed in connection with another felony (called “felony killings”), whether the killing involved excessive violence or brutality, and of course whether or not the crime seemed to be premeditated. There is no direct evidence on premeditation by race. Wolfgang’s (1958:376) data indicate that a higher proportion of killings committed by whites are committed in connection with robberies than are killings by blacks, suggesting that a higher proportion of white killings might be felony killings, compared to killings by blacks. However, the difference is slight, and this finding has not been corroborated elsewhere. On the other hand, Wolfgang (1958) found no significant relationship between race and the tendency to inflict multiple acts of violence, while black homicide arrestees were significantly more likely to have prior arrest records than white homicide arrestees (Wolfgang, 1958:160, 175-6). Therefore, there is little evidence that would indicate that killings committed by blacks are significantly less likely to be capital murders than those committed by
There is another potential source of bias peculiar to the use of victim data by race as a surrogate for offender data by race. It was assumed that the number of offenders of one race would be roughly equal to the number of victims of that race. This assumption could be substantially incorrect, if, for example, killings involving black victims and white killers were more numerous than killings with white victims and black killers. If such were the case, the number of black victims would overestimate the number of black killers relative to white killers, and therefore underestimate the black execution rate relative to the white execution rate. Data relevant to this question are contained in the 1976 Uniform Crime Reports, which reported offender-victim racial relationships for murders and non negligent manslaughters (U.S. F.B.I., 1977). These data indicate that black offender-white victim killings are more numerous than white offender black victim killings. They further indicate that while only 53.2% of the 10,538 homicide victims were black, 55.8% of the known offenders were black. This finding suggests that use of victim data by race involves a bias whose correction would only strengthen our findings. Similar conclusions on victim-offender racial relationships could be drawn from data reported in smaller scale studies of criminal homicide (Garfunkel, 1949:371; Wolfgang, 1958:379; Bensing and Schroeder, 1960:51; and the seventeen-city study of Curtis, 1974:21).
Death-Sentencing Rates Execution rates, as we have measured them, reflect not only rates at which defendants are sentenced to death, but also the extent to which such sentences are successfully appealed or commuted to a lesser penalty. Therefore, a purer measure of the rate at which defendants are sentenced to death is desirable. Annual data on the number of persons sentenced to die has been compiled, by race of the offender, for the United States since 1967.
While this does not allow computation of death-sentencing rates for a very long period of time, it does update our analysis by providing information on the administration of capital punishment since the de facto moratorium on executions began in 1967. Death-sentencing rates were first computed in a fashion similar to the computation of execution rates: the number of death sentences (actually, persons received by U.S. prisons from the courts, sentence of death) for murder is compared with the number of homicides in the previous year, for each race. Then a second measure of the death-sentencing rate was computed. It could be argued that a better measure of the risk of receiving a death sentence would use persons arrested for, or convicted of, capital crimes as the base for the rate, since it is only such persons who are actually at risk of receiving a death sentence. While there are no national data on convictions for murder, there are national figures on persons arrested for murder or non-negligent manslaughter. Therefore, rates were computed using these figures for the base of the death-sentencing rate, again in a manner similar to the computation of execution rates.
Table 6 shows the computation of death-sentencing rates. The resulting rates, whether based on homicide deaths or homicide arrests, indicate that nonwhites were subject to a lower risk of being sentenced to death than whites over the period from 1967 to 1978. Because of the small numbers of death sentences each year, rates for single years are somewhat unstable, especially for 1972 and 1973. Nevertheless, the findings are on the whole quite consistent with the findings for execution rates.
This aggregate-level analysis does not show that there is never overt racial discrimination in the administration of the death penalty for murder outside of the South. There may be discrimination in particular jurisdictions, in specific individual cases, or at specific, previously unstudied, stages in the legal process leading up to execution, although a close reading of previous studies of various stages in this process, such as arrest, indicate that, at least outside the South, overt racial discrimination may be more apparent than real, just as seems to be the case with sentencing (regarding arrest, see Green, 1970; Black, 1971; Monahan, 1972; Lundman et al., 1978).
What the present analysis does show is that regardless of whatever discrimination there may or may not be at particular stages in the legal process, the outcome is a lower execution rate for blacks than for whites. If there is discrimination against blacks at one or more stages, then, given the observed net result, it seems that there must also be some compensating effects, favoring blacks, at other stages.10 Likewise, if there is discrimination against blacks in one or more regions, jurisdictions, or specific cases, or with particular subtypes of homicides, then there must be some counterbalancing effects elsewhere.
The simple computation of execution and death-sentencing rates obviously does not in any way control for differences in prior criminal record (or other legally relevant variables, for that matter). Considering the stress laid on controlling for prior record earlier in the paper, this omission might seem to undercut confidence in the death penalty findings. However, because introduction of such controls has consistently reduced differences in sentencing outcome attributable to racial discrimination, correcting this omission would only tend to strengthen the conclusion of no overt discriminatory effect of homicide offenders’ racial identity. In connection with capital punishment of rape, controlling for prior criminal record would reduce the apparent discriminatory effect to some degree, but this effect is so large to begin with that it is doubtful if the conclusion of discrimination would have to be altered.
10 This possibility of a pattern of compensating discriminatory effects has been raised by Nagel and Neef (1977:185-8).
Table 6. Death-Sentencing Rates by Race, 1967-1978
Year Nonwhite White Ratio: Nonwhite/White Sentences per 1,000 Homicide Deaths a
1978 74/ 9,230 = 8.02 108/ 10,730 = 10.07 0.80
1977 64/ 9,439= 6.78 68/ 10,115= 6.72 1.01
1976 88/ 10,377 = 9.51 136/ 10,973 = 12.39 0.69
1975 143/ 10,817 = 13.22 121/ 10,648= 11.36 1.16
1974 65/ 10,291 = 6.32 67/ 9,789 = 6.84 0.92
1973 23/ 10,498 = 2.19 11/ 8,840 = 1.24 1.76
1972 40/ 10,226 = 3.91 26/ 8,561 = 3.04 1.29
1971 51/ 9,045 = 5.64 45/ 7,803 = 5.77 0.98
1970 52/ 8,461 = 6.15 64/ 7,016 = 9.12 0.67
1969 36/ 7,880 = 4.57 49/ 6,806 = 7.20 0.63
1968 45/ 7,027= 6.40 51/ 6,009= 8.49 0.75
1967 38/ 6,077 = 6.25 36/ 5,230 = 6.88 0.91
1967-78 719/109,328= 6.58 782/102,529= 7.63 0.86
Sentences per 1,000 Homicide Arrests
1978 74/ 9,256 = 7.99 108/ 7,866 = 13.73 0.58
1977 64/ 7,083 = 9.03 68/ 5,792 = 11.74 0.77
1976 88/ 8,592 = 10.24 136/ 6,581 = 20.67 0.50
1975 143/ 7,567 = 18.90 121/ 4,879 = 24.71 0.76
1974 65/ 7,677 = 8.74 67/ 5,236 = 12.80 0.66
1973 23/ 8,661 = 2.66 11/ 5,145 = 2.14 1.24
1972 40/ 8,586 = 4.66 26/ 4,716 = 5.51 0.85
1971 51/ 7,344 = 6.94 45/ 4,503 = 9.99 0.69
1970 52/ 6,669 = 7.79 64/ 3,743 = 17.10 0.46
1969 36/ 5,922 = 6.08 49/ 3,536 = 13.86 0.44
1968 45/ 5,018 = 8.97 51/ 3,200 = 15.94 0.56
1967 38/ 4,203 = 9.04 36/ 2,911 = 12.37 0.73
1967-78 719/ 86,578 = 8.30 782/ 58,126 = 13.45 0.62
Sources: Persons sentenced to death: U.S., NCJISS, Capital Punishment (1971-72; 1973; 1974; 1975; 1976; 1977; 1978).
Homicide deaths, 1966-76: U.S., NCHS, Vital Statistics of the United States: Mortality (Year) (1968-79).
Homicide deaths, 1977: U.S. NCHS, Monthly Vital Statistics Report: Advance Report: Final Mortality Statistics 1977 (1979).
Arrests, 1966-77: U.S., F.B.I., Crime in the United States (Year) (1967-1978).
A. Death sentences for murder, year t, per 1,000 homicide deaths, year t- 1.
B. Death sentences for murder, year t, per 1,000 homicide arrests, year t- 1.
The conclusions which can be drawn from the available evidence on the racial patterning of sentencing may be briefly summarized as follows:
(1) The death penalty has not generally been imposed for murder in a fashion discriminatory toward blacks, except in the South. Elsewhere, black homicide offenders have been less likely to receive a death sentence or be executed than whites.
(2) For the 11% of executions which have been imposed for rape, discrimination against black defendants who had raped white victims was substantial. Such discrimination was limited to the South and has disappeared because death sentences are no longer imposed for rape.
(3) Regarding noncapital sentencing, the evidence is largely contrary to a hypothesis of general or widespread overt discrimination against black defendants, although there is evidence of discrimination for a minority of specific jurisdictions, judges, crime types, etc.
(4) Although black offender-white victim crimes are generally punished more severely than crimes involving other racial combinations, the evidence indicates that-his is due to legally relevant factors related to such offenses, not the racial combination itself.
(5) There appears to be a general pattern of less severe punishment of crimes with black victims than those with white victims, especially in connection with imposition of the death penalty. In connection with noncapital sentencing, the evidence is too sparse to draw any firm conclusions.
None of these findings are inconsistent with the assertion of institutional racism or income discrimination in sentencing. It is quite possible that low income makes it more difficult to make bail, hire a private attorney genuinely independent of the court, etc., for both blacks and whites, and that these factors in turn result in more severe sentencing outcomes, as Lizotte’s (1978) research indicates-” If black criminal defendants are poorer than white criminal defendants, then income discrimination would produce racial differentials in sentences received. Nor are the data inconsistent with a hypothesis of overt discrimination at earlier stages of the criminal justice process. We might expect violations of stated values such as equal protection and justice for all to occur most commonly in connection with the least visible decisions, such as the decision to arrest, charge, prosecute, or release a defendant on bail (e.g., see Hagan, 1975 on the decision to charge). However, these decisions are less well studied than the sentencing decision, so the evidence for discrimination is necessarily even weaker than that regarding sentencing, quite apart from the actual prevalence of discriminatory practice.
The findings of this study do not suggest a different explanation for a well-known phenomenon. Rather they point to a phenomenon to be explained which differs from that conventionally addressed by American students of the legal reaction to crime and criminals. Students of the criminal justice system, concerned with the contemporary consequences of a historical pattern of racism, have sought to explain patterns of more severe treatment of blacks, while overlooking or downplaying the pattern of more lenient treatment of black defendants.
Blacks in the United States, both in the recent and more remote past, have been less likely than whites to receive a death sentence if they committed a homicide. Furthermore, this pattern is apparently not entirely limited to the sentencing of capital offenders. For a variety of specific crimes, jurisdictions, and judges, various researchers have produced data indicating more lenient treatment of black defendants than whites, although the admittedly scattered findings were usually deemphasized or discounted as merely anomalous results attributable to some flaw in the analysis or research design.12 For example, Bullock (1961) found significantly shorter prison sentences were assigned to blacks convicted of murder; Levin’s (1972) Pittsburgh data indicate that blacks received more lenient dispositions than whites for eight out of nine offense categories; and Bernstein and her colleagues (1977) found that blacks received significantly less severe sentences than whites. Gibson (1978:469) studied sentences given by individual judges and found that seven of eleven judges gave a higher percentage of severe sentences to whites than to blacks.
” It is interesting that Lizotte’s path-analysis findings indicate that defendant’s race affects whether the defendant makes bail (which in turn affects sentence), but that it does not affect the bail amount set. This suggests that there is no overt racial discrimination in bail setting, but that there are income discriminations. Nice Lizotte had no measure of defendant’s income, it is possible that the race effect was found simply because the race variable was serving as a rough surrogate for defendant income.
12 This may be one of the more important subsidiary findings of the literature review. It is a chronic problem in this area, and perhaps in sociology as a whole, that researchers fail to recognize the significance of anomalies, which should alert them to the possible need for alterations in their fundamental assumptions rather than just their methods (See Kuhn, 1962: Ch. VI on this issue).
The specification of phenomena to be explained is in a way a more fundamental scientific task than the development of explanations, since the former obviously must occur before the latter can even be imagined. The pattern of lenient treatment of black defendants in the South was recognized in the 1940s and before by observers such as Dollard (1937) and Myrdal ( 1972), and various explanations were developed to account for it. Today, however, this phenomenon is largely disregarded.
A number of factors which may help account for this pattern can be briefly outlined.
(1) Blacks as devalued crime victims. Perhaps the most plausible explanation of lenient treatment of black offenders who commit predominantly interracial crimes such as homicides, assaults, and rapes is that crimes with black victims are considered by predominantly white social control agents to be less serious offenses, representing less loss or threat to the community than crimes with white victims (Myrdal,  1972:551). Thus, paradoxically, racist sentiments would produce more favorable treatment for members of the subordinate racial group who commit interracial crimes.
(2) White paternalism. Students of criminal justice in the South have suggested a widespread view among whites of blacks as child-like creatures who were not as responsible for their actions as whites were, and who therefore could not be held accountable to the law to the extent that whites are (Dollard, 1937; Myrdal,  1972; Garfunkel, 1949). Therefore their perceived diminished responsibility presumably earned them more lenient sentences.
(3) Sociology-based tolerance. White paternalism may account for Southern sentencing patterns of the past but not patterns in the rest of the country in more recent times. However, it may have been replaced by a new form of white tolerance for black crime, involving the following line of reasoning: “Blacks commit crimes because of poverty, racism, and/or the resulting black poverty-subculture, which accepts or encourages criminal behavior. Their crimes are due to forces beyond their control or at least are to be expected in this light. Therefore blacks should not be held as responsible for their actions as whites.” Of course, this is largely speculative; however, criminal court informants questioned by Bernstein and her colleagues (1977:753) stated that “some judges and prosecutors assume that nonwhites commit crimes because the nonwhite subculture accepts such behavior. These subcultural differences are considered by the judges and prosecutors, thereby making the offenses of nonwhites seem less pernicious.”
(4) Affirmative action in the courts. White guilt over acknowledged past discrimination could motivate liberal criminal-justice decision makers to consciously or unconsciously compensate with more lenient treatment of black defendants.
(5) Compensation for institutional racism. Recognizing the handicaps of low income and greater prior criminal records which black defendants bring into court, some decision makers may attempt to compensate in determining sentence.
(6) Compensation for unconscious prejudice. Johnston et al. (1973) conducted a study involving criminal court judges sentencing hypothetical defendants and found that the hypothetical white “defendants” were sentenced more severely than the black “defendants.” Two of the judges who participated in the study explained that they consciously sought to compensate in their sentencing for any unconscious prejudice on their part against minorities (p. 870).
Various combinations of these explanations can be used to account for less severe sentencing of blacks when and where it has occurred. Factors (1) and (2) may be primarily responsible for lenient sentencing in interracial cases in the South in the 1940s and before, while different combinations of factors (3) through (6) account for leniency when it occurs elsewhere today and in the relatively recent past say, since the late 1960s. Only future research designed to test the hypotheses can determine which of these are more than merely plausible and actually produce the patterns observed.
The findings of this paper should not be interpreted as being incompatible with conflict, critical, or Marxist approaches to law, but rather only with the more simplistic, instrumentalist versions of these perspectives, which heavily stress the failure of the criminal justice system to operate according to its own stated standards of equity and proper procedure (e.g. Quinney, 1974; Chambliss and Seidman, 1971). In contrast, Beirne (1979) has pointed out the partial autonomy from particular social classes which the legal system enjoys, and has stated that “The capitalist class as a whole cannot be well served by frequent and visible abuses of due process” (p. 379). The criminal justice system can routinely operate to further legitimate the existing order through obedience to its own rules and limits to its power, even when overt class or race bias in specific situations would otherwise aid particular segments of the ruling class. Clearly then, a more intellectually mature version of conflict theory would not necessarily predict overt class or race bias in the allocation of penalties among criminal defendants.
However, there are forms of class bias in the legal systems which are not so clearly a threat to the legitimacy of the system. For example, the American legal system openly permits differing economic resources to be used in mounting a criminal defense, and such differences render legal advantages in avoiding conviction or obtaining lenient sentences if convicted, even though the advantages may operate indirectly and may involve no intentional prejudice on the part of any system decision maker. Lizotte (1978) has demonstrated how criminal sentence is affected by whether the defendant made bail and by the type of attorney the defendant had (private attorneys who were not courtroom regulars were more successful in negotiating light sentences than other types). While Lizotte had no measure of income, these are both clearly advantages more available to defendants with greater income. If equal protection of the law is a commodity which must be purchased, then this “equality” cannot be anything more than a legal fiction as long as the resources for such a purchase are distributed in an unequal fashion.
Serious though this economic or income discrimination in court processing may be, there is a far more fundamental bias in criminal sentencing. No studies of court processing of criminal defendants can address the issue of how legislatures criminalize behaviors common to lower-class persons, while either failing to criminalize or assigning slight penalties to equally harmful behaviors common among middle- or upper-class persons, such as poisoning of the air and water, manufacture of food, drugs, and other products harmful to human health, price-fixing, and consumer fraud. Detailed study of the use of wealth and power in controlling the ideological composition of legislatures and enforcement agencies, thereby influencing selection of behaviors to be criminalized, the original setting of penalty ranges, the determination of enforcement priorities, and allocation of enforcement resources, is likely to reveal far more about why blacks and lower-class persons are overrepresented in arrest, court, and prison data than studies of processing within the criminal justice system. The focus on the influence of ascribed characteristics of individual criminal defendants on processing decisions has, at least up to now, failed to yield the empirical support which would justify the attention that continues to be lavished on the subject.
Atkinson, David N. and Dale A. Newman 1970 “Judicial attitudes and defendant attributes: some consequences for municipal court decision making.” Journal of Public Law 19:68-87.
Baab, G. A. and W. R. Ferguson 1968 “Texas sentencing practices: a statistical study.” Texas Law Review 45:471-503.
Bedau, Hugo Adam 1964 “Death sentences in New Jersey.” Rutgers Law Review 19:1-55. 1965 “Capital punishment in Oregon 1903- 1964.” Oregon Law Review 45:1-39. 1967 “Criminal justice: the general aspects.” Pp. 405-14 in Hugo Adam Bedau (ed.). The Death Penalty in America. Garden City, N.Y.: Doubleday.
Beirne, Piers 1979 “Empiricism and the critique of Marxism on law and crime.” Social Problems 26:373-85.
Bensing, Robert C. and Oliver J. Schroeder 1960 Homicide in an Urban Community. Springfield, Ill.: Charles Thomas.
Bernstein, Ilene Nagel, William R. Kelly, and Patricia A. Doyle 1977 “Societal reaction to deviants: the case of criminal defendants.” American Sociological Review 42:743-55.
Black, Donald 1971 “The social organization of arrest.” Stanford Law Review 23:1087-111.
Blalock, Hubert A. 1972 Social Statistics. New York: McGraw-Hill.
Block, Richard and Franklin E. Zimring 973 “Homicide in Chicago, 1965-1970.” Journal of Research in Crime and Delinquency 10:1-12.
Bowers, William J. 1974 Executions in America. Lexington, Mass.: D.C. Heath.
Bridge, Franklin M. and Jeanne Mosurel 1961 Capital Punishment. Staff Research Report No. 26. Columbus, Ohio: Ohio Legislature Service Commission.
Bullock, H. A. 1961 “Significance of the racial factor in the length of prison sentences.” Journal of Criminal Law, Criminology and Police Science 52:411-7.
Burke, Peter and Austin Turk 1975 “Factors affecting post arrest dispositions: a model for analysis.” Social Problems 22:313-32.
Cameron, Mary Owen 1964 The Booster and the Snitch. New York: Free Press.
Cargan, Leonard and Mary A. Coates 1974 “Indeterminate sentence and judicial bias.” Crime and Delinquency 20:144-56.
Castberg, A. Didrick 1971 “The ethnic factor in criminal sentencing.” Western Political Quarterly 24:425-37.
Chambliss, William J. and Robert B. Seidman 1971 Law, Order, and Power. Reading, Mass.: Addison-Wesley.
Chiricos, Theodore G. and Gordon P. Waldo 1975 “Socioeconomic status and criminal sentencing: an empirical assessment of a conflict proposition.” American Sociological Review 40:753-72.
Clark, Ramsey 1970 Crime in America: Observations on Its Nature, Causes, Prevention and Control. New York: Simon and Schuster.
Clark, Stevens H. and Gary G. Koch 1976 “The influence of income and other factors on whether criminal defendants go to prison.” Law and Society Review 11:57-92.
Conklin, John E. 1972 Robbery and the Criminal Justice System. Philadelphia: Lippencott. Curtis, Lynn A. 1974 Criminal Violence. Lexington, Mass.: D.C. Heath.
Dison, Jack Everitt 1976 “An empirical examination of conflict theory: race and sentence length.” Ph.D. dissertation, Department of Sociology, North Texas State University, Denton, Texas.
Dollard, John 1937 Caste and Class in Southern Town. New Haven: Yale University Press.
Farrell, Ronald A. and Victoria Lynn Swigert 1978a “Prior offense as a self-fulfilling prophecy.” Law and Society Review 12:437-53. 1978b “Legal disposition of inter-group and intragroup homicides.” Sociological Quarterly 19:565-76.
Florida Civil Liberties Union 1964 Rape: Selective Electrocution Based on Race. Miami: Florida Civil Liberties Union.
Foley, Linda A. and Christine E. Rasche 1979 “The effect of race on sentence, actual time served and final disposition of female offenders.” Pp. 93-106 in John A. Conley (ed.), Theory and Research in Criminal Justice: Current Perspectives. Cincinnati: Anderson.
Forslund, Morris A. 1969 “Age, occupation and conviction rates of white and negro males: a case study.” Rocky Mountain Social Science Journal 6:141-6.
Garfinkel, Harold 1949 “Research note on inter- and intra-racial homicides.” Social Forces 27:369-81.
Gerard, Jules and T. R. Terry 1970 “Discrimination against negroes in the administration of criminal law in Missouri.” Washington University Law Quarterly 1970:415-37.
Gibson, James L. 1978 “Race as a determinant of criminal sentences: a methodological critique and a case study.” Law and Society Review 12:455-78.
1961 Judicial Attitudes in Sentencing. London: MacMillan.
1964 “Inter- and intra-racial crime relative to sentencing.” Journal of Criminal Law, Criminology and Police Science 55:348-58. 1970 “Race, social status, and criminal arrest.” American Sociological Review 35:476-90.
Greenwood, Peter W., Sorrel Wildhorn, Eugene C. Poggio, Michael J. Strumwasser, and Peter De Leon
1973 Prosecution of Adult Felony Defendants in L.A. County: A Policy Perspective. Law Enforcement Assistance Administration, Washington, D.C.: U.S. Government Printing Office.
1974 “Extra-legal attributes and criminal sentencing: an assessment of a sociological viewpoint.” Law and Society Review 8:357-83.
1975 “Parameters of criminal prosecution: an application of path analysis to a problem of criminal justice.” Journal of Criminal Law and Criminology 65:536-44.
Hall, E. L. and A. A. Simkus 1975 “Inequality in the type of sentences received by native Americans and whites.” Criminology 13:199-222.
Harlan, Howard 1950 “Five hundred homicides.” Journal of Criminal Law and Criminology 40:736-52.
Hewitt, John D. 1976 “Individual resources, societal reaction, and sentencing disparity.” Western Sociological Review 7:31-56.
Hindelang, Michael J. 1969 “Equality under the law.” Journal of Criminal Law, Criminology and Police Science 60:306-13.
1978 “Race and involvement in crimes.” American Sociological Review 43:93-109.
Howard, Joseph C., Jr. 1967 Rape death penalty study reported in the New York Times, September 18, 1967, p. 33.
Hutner, Michael 1977 “Sentencing in Massachusetts Superior Court.” Ph.D. dissertation, Department of Sociology, Syracuse University, Syracuse, N.Y.
Jacob, Herbert 1962 “Politics and criminal prosecutions in New Orleans.” Tulane Studies in Political Science 8:77-98.
Janos, Dean and Robert I. Mendelsohn 1967 “The judicial role and sentencing behavior.” Midwest Journal of Political Science 11:471-88.
Johnson, Guy 1941 “The negro and crime.” The Annals of the American Academy of Political and Social Science 217:93-104.
Johnson, Elmer H. 1957 “Selective forces in capital punishment.” Social Forces 36:165-9.
Johnson, Oakley C.  “Is the punishment of rape equally ad- 1970 ministered to negroes and whites in the state of Louisiana?” pp. 216-28 in William L. Patterson, (ed.), We Charge Genocide. New York: International Publishers.
Johnston, Barbara L., Nicholas P. Miller, Ronald Schoenberg, and Laurence Ross Weatherly 1973 “Discrimination in felony sentencing-a study of influencing factors.” Washington Law Review 48:857-89.
Judson, Charles J., James J. Pandell, Jack B. Owens, James L. McIntosh, and Dale L. Matschullat 1969 “A study of the penalty jury in first degree murder cases.” Stanford Law Review 21:1297-431.
Kelly, Henry E. 1976 “Comparison of defense strategy and race as influences in differential sentencing.” Criminology 14:241-9.
Kuhn, Thomas S. 1962. The Structure of Scientific Revolutions. Chicago: University of Chicago Press.
Kulig, Frank 1975 “Plea bargaining, probation, and other aspects of conviction and sentencing.” Creighton Law Review 8:938-54.
Lemert, Edwin M. and Judy Rosberg 1948 The Administration of Justice to Minority Groups in L.A. County. Berkeley: University of California Press.
Levin, Martin A. 1972 “Urban politics and judicial behavior.”‘ Journal of Legal Studies 1:220-1.
Liska, Allen E. and Mark Tausig 1979 “Theoretical interpretations of social class and racial differentials in legal decision making for juveniles.” The Sociological Quarterly 20:197-207.
Lizotte, Alan J. 1978 “Extra-legal factors in Chicago’s criminal courts: testing the conflict model of criminal justice.” Social Problems 25:564-80.
Lotz, Roy and John D. Hewitt 1977 “The influence of legally irrelevant factors on felony sentencing.” Sociological Inquiry 47:39-48.
Lunden, Walter A. 1962 “Time lapse between sentence and execution: the U.S. and Canada compared. “A BA Journal 38:1043-5.
Lundsgaarde, Henry P. 1977 Murder in Space City: A Cultural Analysis of Houston Homicide Patterns. New York: Oxford University Press.
Lundman, Richard J., Richard E. Sykes, and John P. Clark 1978 “Police control of juveniles: a replication.” Journal of Research in Crime and Delinquency 15:74-91.
Mangum, Charles S., Jr. 1940. The Legal Status of the Negro. Chapel Hill: North Carolina Press.
Martin, Roscoe 1934. The Defendant and Criminal Justice. University of Texas Bulletin No. 3437. Austin, Texas. Bureau of Research in the Social Sciences.
McCafferty, James A. 1967 “The death sentence, 1960.” Pp. 90-103 in Hugo Adam Bedau, (ed,.). The Death Penalty in America. Garden City, N.Y.: Doubleday.
Mileski, Maureen 1971 “Courtroom encounters: an observation of a lower criminal court.” Law and Society Review 5:473-538.
Mohahan, T. P. 1972 “The disposition of juvenile offenders by race and sex in relation to the race and sex of police officers.” The International Review of Modern Sociology 2:91-101.
Myers, Martha A. 1979 “Offended parties and official reactions: victims and the sentencing of criminal defendants.” Sociological Quarterly 20:529-40.
Myrdal, Gunnar . An American Dilemma. New York: Pan- 1972 theon.
National Association for the Advancement of Colored People (NAACP) 1971 “Legal defense fund brief for Aikens 1971.” Cited on p. 71 in William J. Bowers (ed.), Executions in America. Lexington, Mass.: D.C. Heath.