Part II Impact

Section 2. Sentencing Neutrality

GOAL: Sentencing is neutral with respect to race, gender, and social and economic status. F.S. 921.001(4)(a)1.

FACT: This study found that an offender's race has NO meaningful effect on the sentencing decisions made by Florida courts under the 1994 and 1995 sentencing guidelines structure.

FINDINGS:

• Before examining any other factors, black offenders were more likely than white offenders to be sentenced to prison (Table 1). Within many offense groups, black offenders received longer prison sentences than white offenders (Table 2).

• However, black offenders had higher rates of characteristics generally considered appropriate for higher rates of imprisonment and longer prison sentences. Black offenders had 14.3% higher total sentence points, were 27% more likely to have a prior record, and were 40.4% more likely to have a prior prison sentence (Table 3).

• When taking into account relevant sentencing factors, race was not found to be an influential factor in determining either the decision to sentence an offender to prison or the length of prison sentences. Of the 32 factors studied which affect the in/out prison decision, 28 were more influential in determining the sentence than the race of the offender (Table 4). Of the 31 factors used to study the length of prison sentence, race was the least influential factor in determining the sentence (Table 5).

* The database used for this section of analysis includes scoresheets with sentence dates outside the FY 1996-97 that is analyzed in the rest of the report.

Does An Offender's Race Affect Sentencing in Florida?

Introduction:

The question addressed in this section is: Does an offender's race affect the sentencing decisions made by Florida courts when punishing felony offenders under the 1994 and 1995 sentencing guidelines structure. FS 921.001(a)(4) states that sentencing guidelines embody the principles that sentencing is to be "neutral with respect to race..". This study will determine whether this racial equity principle is followed when offenders are punished under Florida's sentencing guidelines mechanism.

Two major changes in Florida's sentencing policy have occurred in the past two decades. In 1983, the indeterminate sentencing policy, also known as parole, was eliminated and replaced with sentencing guidelines. In 1994, the 1983 sentencing guidelines structure was replaced with the 1994 sentencing guidelines. Modifications to the 1994 guidelines were made in 1995 for crimes committed on or after October 1, 1995. This study does not examine whether race affected sentencing decisions under the pre-1983 indeterminate sentencing system or during the 1983 guidelines period. It focuses on whether racial disparity exists within the 1994 and 1995 sentencing guidelines system.

Prior Research:

Some evidence suggests that racial disparity in sentencing did exist in Florida prior to the 1983 sentencing guidelines. The Sentencing Study Committee, which was responsible for recommending that Florida implement sentencing guidelines in 1983, conducted an in-depth study of 1,000 felony cases in 1979.¹ The committee examined the decision whether to sentence an offender to jail or prison and, if a prison sentence was imposed, the length of incarceration. The committee found that, after holding legally relevant factors constant, non-white offenders were significantly more likely to receive a jail or prison sentence than white offenders.

Methodology:

For this analysis, all felony offenders sentenced in Florida courts from July 1, 1994 to December 31, 1996 who were sentenced under the 1994 or 1995 sentencing guidelines were examined for racial disparity.² Statistical models were constructed based on variables contained in the sentencing guidelines database maintained by the Florida Department of Corrections. This database contains information on 221,351 offenders sentenced under the guidelines over the period specified.

These statistical models examine the effect of all variables simultaneously, to measure the unique effect of each on the sentencing outcome while holding all other variables constant. This method enables one to determine whether an offender's race influences judicial sentencing when other characteristics about the offender are held constant (seriousness of the current offense, prior criminal record, offenses other than the primary offense,³ and victim injury).

Sentencing is analyzed as a two part decision: first, whether or not the offender is sentenced to prison, and second, if a prison sentence is administered, the length of the prison sentence. Prior sentencing research has conclusively demonstrated that the judiciary utilizes different factors, or the same factors to different degrees, when making the in/out prison decision versus the length of prison sentence decision.

The following factors were included in the models. Details on how these factors were measured are located at the end of this section.

1. Race (black/white)⁴
2. Gender (male/female)
3. Age at Time of Sentencing
4. Most Serious Offense ("Primary Offense")
5. Statutory Felony Class of the Primary Offense
6. Type of Primary Offense (Murder, Sexual/Lewd Assault, Robbery, Other Violent, Burglary, Property, Drugs, Weapons, and Other)
7. Seriousness of Additional Offenses
8. Types of Additional Offenses
9. Seriousness of Offender's Prior Criminal Record
10. Types of Prior Crimes in Offender's Prior Criminal Record
11. Prior Florida Prison Sentences
12. Victim Injury
13. Prior Community Supervision Violations
14. Law Enforcement Enhancement Guidelines Points
15. Case Disposition (Plea/Trial)

Many of these factors are measured in multiple ways to model the sentencing decisions. For example, the types of prior criminal convictions are measured by the number of prior robbery convictions, the number of prior drug convictions, etc. This approach was taken to develop models which explain as much of the sentencing decisions as possible. A total of 32 factors were used to predict the sentencing decisions.⁵

Details on the methodology of the current study can be obtained from the author.⁶

Findings:

This study failed to find evidence that an offender's race has any meaningful effect on decisions made by Florida courts under the 1994 and 1995 sentencing guidelines structure. This leads to the conclusion that the goal of racial equity explicit in the sentencing guidelines law has been met when examining the 1994 and 1995 sentencing guidelines structure.

Overall, the empirical evidence presented in detail in this section documents three primary conclusions:

1. Before examining any other factors, black offenders were more likely than white offenders to be sentenced to prison (Table 1). Within many offense groups, black offenders received longer prison sentences than white offenders (Table 2).
2. However, black offenders had higher rates of characteristics generally considered appropriate for higher rates of imprisonment and longer prison sentences (e.g., more serious crimes and more serious prior criminal records) (Table 3).
3. After taking into account relevant sentencing factors, race was not found to be an influential factor in determining either the decision to sentence an offender to prison (Table 4) or the length of prison sentences for those receiving a prison term (Table 5).

It should be noted that this report addresses only disparity at the final stage in the judicial process of sentencing for felony offenders. The question of whether racial disparity exists at earlier stages in the criminal justice process, such as arrest, prosecution, plea bargaining, or conviction is not within the scope of this report.

Is there racial equity is specific sentencing guidelines factors? Below are the answers to this question.

Question: Are black offenders more likely than white offenders to be sentenced to state prison following a felony conviction when one does not examine any of the factors which are intended to affect sentencing (e.g., seriousness of the current crime, prior criminal record and victim injury)?

Answer: Yes. Table 1 shows that black offenders were more likely to be sentenced to prison than white offenders: black 20.8% versus white 14.1 %. When examining the likelihood of black and white offenders receiving a prison sentence for general crime types (violent, property, drug, and other), black offenders were more likely to be sentenced to prison than white offenders when the primary offenses was one of the following: Violent (+11.4%), Property (+3.4%), Drug (+9.4%), and Other (+3.7%). Within the nine more specific offense types (murder/manslaughter, sexual/lewd assault, etc.) black offenders were more likely to be sentenced to prison than white offenders. Examining 49 specific offense types in Table 1 reveals that black offenders were more likely to be sentenced to prison than white offenders within 39 offense types, while white offenders were more likely in 10 offense types.

Question: If the judge decided a prison sentence is appropriate for a

convicted felon, were black and white offenders given different lengths of prison sentences under the 1994 and 1995 sentencing guidelines when no legally relevant factors are considered?

Answer: Yes, but only for some offense types. Table 2 shows that the average sentence lengths for black and white offenders were identical (4.6 years) when all 38,031 offenders sentenced to prison are considered. However, while the numerical differences are not significant, when examining differences in average sentence lengths within the four broad offense types, black offenders received longer sentences in three of the four groups (Violent: black = +0.3 years, or + 4.7% longer prison sentences; Property: black +0.3 years, + 8.1%; Other: black +0.4 years, +13.6%). White offenders received an average of 0.6 years, or 17.2% longer prison sentences than black offenders for drug convictions.

When examining racial differences within nine specific offense types, black offenders received longer sentences in seven of the groups. Black offenders sentences averaged 20.4% (2.7 years) greater than white offenders for those convicted of murder or manslaughter. Black offenders convicted of robbery received 18.8% (1.2 years) longer sentences, on average, than white offenders convicted of robbery. Overall, white drug offenders received 17.2% (0.6 years) longer sentences than black offenders. However, when examining those convicted of selling drugs, black offenders received 25.5% (0.6 years) longer sentences than white offenders. White drug offenders received longer sentences for possession and trafficking of drugs compared to black offenders convicted of the same crimes.

Question: Are black offenders and white offenders different relative to sentencing factors considered relevant to the in/out prison decision and the length of prison sentence?

Answer: Yes. Table 3 shows that, for most factors, black offenders consistently exhibited higher rates of characteristics generally associated with judicial decisions towards more punitive sanctions. The figures in Table 3 reveal the following differences:

• Black offenders had higher overall sentencing guidelines points, which are a composite of the seriousness of the current primary crime, additional current crimes, prior criminal record, victim injury, supervision violation points, drug trafficking enhancements, and law enforcement protection enhancements. The average for total points was 30.7 for black offenders compared to 26.3 for white offenders -- a difference of 4.4 points, or 14.3%.

• Black offenders had higher average primary offense points than white offenders. The average for primary offense points was 26.2 for black offenders compared to 24.9 for white offenders -- a difference of 1.3 points or 5.0%. For drug crimes, black offenders had 15.5% higher average primary offense points than white offenders.

• Black offenders had more serious prior criminal records than white offenders using four different measures. First, black offenders were 27.0% more likely to have a prior criminal record than white offenders (black offenders = 66.2%, white offenders = 48.3%). Second, the average prior record guidelines points for black offenders was 8.2 compared to 4.8 for white offenders, a difference of 3.4 points or 41.5%. Third, black offenders had an average of 5.8 prior criminal convictions compared to 4.3 for white offenders. Fourth, black offenders were much more likely to have prior Florida prison commitments. Black offenders were 40.4% more likely to have a prior prison sentence than white offenders (black offenders: 9.2%; white offenders: 5.5%). These differences continued within the four broad offense groups. For example, among drug offenders, black offenders were 31.9% more likely to have prior criminal record, had 51.6% higher average prior record points, had 18.3% higher average number of prior record convictions, and were 58.5% more likely to have prior prison sentences than white offenders.

• Black offenders were slightly more likely (0.3%) to be convicted of multiple crimes (i.e., "additional offenses" beyond the primary crime) than white offenders. For all offense types, white offenders who had additional crimes had higher average guidelines points for these offenses than black offenders (5.4 versus 5.1). Examining additional point differences across the racial groups revealed that black offenders had higher point levels than white offenders for violent offenses (+1.7, 16.5%), drug offenses (+0.8, 25.0% ), and other offenses (+0.6, 17.6%). Only within property offenses did white offenders have higher additional points than black offenders (+1.9, 42.2%).

• Black offenders and white offenders did not differ appreciably in terms of the likelihood that their crimes involved victim injury. White offenders had committed crimes involving some level of victim injury 10.0% of the time compared to 9.6% for black offenders.

• Black offenders were more likely to have previously been violated for failure to abide by the conditions of community supervision (22.1%) than white offenders (19.1%). For violent offenses, black offenders were 17.8% more likely than white offenders to have violated conditions of supervision.

Question: Was race a factor used by judges to any meaningful degree when deciding whether to sentence an offender to prison when all other measured factors are held constant?

Answer: No. Table 4 shows that when considering the 32 factors measured in the in/out prison decision model, race was of no meaningful importance. Out of the 32 factors in the model, 28 affected the prison decision to a greater degree than whether the offender was white or black. Only three factors out of 32 were less influential in determining the sentence imposed than the race of the offenders. This leads to the conclusion that the race of the convicted felon had no meaningful impact on the judge's decision whether a prison sentence is warranted. Instead, the number of times the offenders had been sentenced to prison in the past, the seriousness of the current crime, the extent and severity of prior criminal record, the number of prior prison sentences, and the injury inflicted upon the victim are the factors that primarily determined the imprisonment decision.

The statistical models utilized in this study are able to explain just over half (52.2%) of the judicial decision of whether to sentence a criminal to prison. There are obviously a significant number of factors taken into consideration by judges to decide whether to administer a prison sentence that are not accounted for in the models constructed for this study. Without the ability to measure all the factors considered in the prison decision, the true effect of race cannot be quantified.

Question: Once a judge decided to sentence an offender to prison, was race a meaningful factor in deciding the length of the prison sentence?

Answer: No. Table 5 clearly demonstrates that, under the sentencing guidelines, race has no meaningful effect on the length of prison sentences. Of the 31 factors in the model, 28 affected the length of the prison sentence to a greater degree than race. There were only 2 other factors less influential than race is determining the length of the prison sentence. The seriousness of the current offense, extent of victim injury, and the severity of the offender's prior criminal record, are factors which judges apparently consider most when determining the length of the prison sentence.

Conclusion:

This study of 221,351 felons sentenced under Florida's sentencing guidelines policy from July 1994 to December 1996 clearly demonstrates that the goal of ensuring equity in sentences across racial groups has been realized. There is no meaningful empirical evidence to suggest that black offenders and white offenders are treated unequally by the judicial system under these sentencing guidelines. The race of the offender does not have any meaningful bearing on the decision by Florida judges to sentence a felon to prison or how long imprisoned offenders will be incarcerated. What were influential in determining these punishment decisions were factors such as the severity of the crime(s) for which the offender is being sentenced, the extent and seriousness of the offender's prior criminal record, the number of prior prison sentences, and the amount of injury inflicted upon the victim.

Future Research:

Although sentencing guidelines have directed how over one million felons have been punished in Florida since 1983, this study is the first attempt to address the important question of whether there is racial equity in criminal sentencing under sentencing guidelines. There is much more research that will be done to further study the sentencing racial equity issue. Specifically, the following types of analyses will be conducted and reported in the future.

1. Analysis within specific offense types will be conducted to determine if there is any evidence of meaningful racial disparity. The analysis reported here utilized statistical controls to account for differences in the types of crimes. While this method is generally considered valid, studying this issue within specific offense types may educate us further about this issue.
2. Research will be conducted to determine if any meaningful levels of racial disparity occurred during different time periods since the 1994 and 1995 sentencing guidelines have been in place. It is particularly relevant to examine the racial disparity issue examining the 1994 compared to the 1995 sentencing guidelines.
3. The sentencing guidelines structure and resulting data enable periodic reviews of the racial disparity in sentencing issue to be conducted. Such reviews can be completed annually or on a specialized basis when policy changes require them.
4. While judicial circuit was considered in the statistical models developed in this study, further analysis of the racial disparity issue conducted for individual judicial circuits would tell us more about this issue.
5. If possible, comparative analysis with other states with and without sentencing guidelines structures will be made to identify how Florida compares to other states in terms of the issue of sentencing equity.
6. Further enhancements to the sentencing models will be made to increase the extent to which sentencing decisions are predicted with available data. This will involve including additional data when available and more refinements to the data already accessible.
7. Examinations will be made of whether disparity across geographical areas of the state exists. In addition, the issue of gender and socio-economic equity will be addressed since these factors are also included in the guidelines' equity goal.

Details on Measurement of Sentencing Factors:

Age at Time of Sentencing: Measured in years. The age of the offender when the offense occurred was also used in the preliminary analyses. The influence of age at sentencing and age at offense produced identical results in the statistical models.

Most Serious Offense ("Primary Offense"): This was measured using the guideline point value associated with the primary offense.

Statutory Felony Class of the Primary Offense: There are five felony class levels defined by Florida law which are sentenced under the guidelines: life, first degree punishable by life, first degree, second degree, and third degree. Capital crimes are not sentenced under the sentencing guidelines and are therefore not considered here. Four separate factors were created for the models indicating whether the primary offense was or was not each of the felony class levels. These variables were included in the model to determine their unique effect on the sentencing decisions. One could create one continuous factor from the felony class level. However, that would assume that the seriousness of the crime increases to the same degree with each increase in the felony class. For example, it would assume that a second degree felony is twice as serious as a third degree felony. There is no basis for making this assumption.

Type of Primary Offense: The specific primary offenses were categorized into nine groups (murder/manslaughter, sexual/lewd assault, robbery, other violent, burglary, property, drugs, weapons/escape, and other). Nine dichotomous variables (no=0, yes=1) were created for each of the offense groups. Eight of these variables were part of the model (for statistical models you exclude one category to form a comparison point). These variables were treated as control variables and are not reported in the in/out prison decision or length of prison decision tables.

Seriousness of Additional Offenses: Measured as the number of guidelines points assessed for all additional crimes for which the offender was sentenced. This was used as an overall seriousness measure. A measure of the number of additional offenses was created, however, the guidelines points were found to have more explanatory power than the number of crimes. These two measures were highly correlated. Only the point total was used in the model.

Types of Additional Offenses: A measure of the nature of additional offenses was created by developing indicators of the number of additional crimes for each of the nine offense groups detailed above. These nine variables were then used in the statistical models.

Seriousness of Prior Offenses: Measured as the number of guidelines points assessed for all prior crime convictions. This was used as an overall seriousness measure of the offender's prior criminal record. A measures of the number of prior felony convictions was created, however, the guidelines points were found to have more explanatory power than the number of crimes. These two measures were highly correlated. Only the point total was used in the model.

Types of Prior Offenses: The same types of measure explained above for additional offenses was developed for prior record crimes.

Prior Florida Prison Sentences: The number of times an offender has been sentenced to Florida's prison system in the past. This variable only includes new sentences to prison and does not include admissions to prison which resulted from a technical violation of supervision. In these latter cases, the offender is returned to prison to complete a prior commitment. This information is not part of the guidelines scoresheet. It was obtained from the Department of Correction's data system.

Victim Injury: This is measured by the total number of victim injury points assessed on the guidelines scoresheet. Several other measures, such as number of victims involved and number of various types of victim injury, were developed. However, the total victim injury points explained more of the sentence decisions and was used in the final models.

Prior Community Supervision Violations: This was measured by the number of release program violation points assessed on the guidelines scoresheet.

Law Enforcement Enhancement Guidelines Points: If the primary offense on the sentencing guidelines scoresheet is a violation of the Law Enforcement Protection Act, the subtotal sentence points are multiplied by either 1.5, 2.0, or 2.5, depending upon which provision of the law was violated. The measure used for this analysis is the number of additional guidelines points assessed if the multiplier was used.

NOTE: Due to time constraints, this study has not been replicated or expanded. It has been reprinted from the FY1996-97 annual report.

Endnotes

¹ "A Report on the Analysis of Sentencing Procedures in Florida's Circuit Courts," Sentencing Study Committee, February 29, 1979.

² Capital felony cases are not sentenced under the sentencing guidelines and are therefore not a part of this study. The 1994 sentencing guidelines took effect for any crimes committed on or after January 1, 1994. Therefore, offenders sentenced between July 1, 1994 and December 31, 1996 who committed crimes prior to January 1, 1994 are not a part of this study. Offenders sentenced between January 1, 1994 and June 30, 1994 under the 1994 sentencing guidelines were excluded from this analysis. The quality of the guidelines data for this time period is questionable. Data quality improvements as a result of additional training, feedback to judicial circuits, and the implementation of data quality auditing procedures resulted in more accurate and complete guidelines scoresheets after the first six months of implementation.

³ "Primary Offense" is the most serious crime for which the offender is sentenced under the sentencing guidelines. The determination of which offense is primary, if multiple offenses are involved, is based on which crime results in the highest number of total sentencing guidelines points. For almost all sentencing scoresheets, the primary offense will be the one which falls in the highest guidelines level.

⁴ The sentencing guidelines scoresheet allows for the entry of three race categories: white, black, and other. There were 1,889 (0.8%) cases with the "other" race category in the database studied for this report. The relatively low number of cases in "other" and the inability to identify the specific racial group led to the decision not to include these cases in this analysis.

⁵ The judicial circuit which sentenced the offenders was used as a control variable in models not presented in this report. These models produced virtually identical results to those reported in this report in terms of racial effects and the relative importance of the factors in the models. To examine this issue further, analysis will be conducted in the future within specific judicial circuits to further study the racial disparity issue.

⁶ William D. Bales, Ph.D., Florida Department of Corrections, Bureau of Research and Data Analysis, Tallahassee, Florida 32399-2500. Phone (904) 488-1801.