In order to determine the total population and total black male population who are disenfranchised due to felony status in each state we utilized the 1996 Department of Justice publication, Civil Disabilities of Convicted Felons as a reference for the states criteria for disenfranchisement. The criteria for disenfranchisement are not always precise in each state and sometimes are defined by offense. In a few states where a number of common felonies were listed as disqualifying offenses, it was assumed that all felons would generally be subject to these restrictions in practice. In Mississippi, for example, the state Constitution defines a number of specific felonies as voting disqualifications. In 1996 two prison inmates who contended that their crimes were not covered by the ban were turned down by the Department of Corrections in their request to vote by absentee ballot. In response to the challenge, a constitutional amendment was introduced in the state legislature to add additional crimes to the disqualifying offenses.83
The data presented in this report, particularly for those states with lifetime disenfranchisement policies, should be considered as estimates. In many cases, necessary data at the state level were only available for selected years or were incomplete. In other instances estimates of felony conviction rates were based on national data from the early 1990s. The degree to which these data reflect practices in different states over a period of time may vary. Also, migration between states may affect the estimates of disenfranchised populations depending on such factors as the degree to which individuals with a felony record move from one state to another.
The estimates of disenfranchisement were developed in two stages of data collection. First, the number of disenfranchised persons was tabulated for the 47 jurisdictions which exclude various categories of felons currently under correctional supervision. For the fifteen states which disenfranchise some or all ex-felons, it was necessary to estimate the total number of people in each state whose prior convictions qualify them as disenfranchised. Most of these states disenfranchise all ex-felons. However, in Tennessee and Washington disenfranchisement depends on the year of conviction, in Arizona and Maryland only the second conviction disqualifies one from voting for life, and in Texas ex-felons are disenfranchised for two years after their release from supervision.
We selected 1996 as the year for analysis because it was the most recent year for which the Bureau of Justice Statistics (BJS) has published state-level data on prison, probation and parole populations. For probationers, the national figure of 55 percent was used to estimate the number of probationers in each state serving a sentence of felony probation.
Racial composition data were available by state only for 1995. In order to estimate the number of disenfranchised black males in 1996 we first estimated the share of each appropriate category (prison, probation, or parole) in 1995 who were black males, and then applied those proportions to the 1996 totals. Racial breakdowns were reported for the majority of states based on counts or estimates. When states reported some offenders of an unknown racial category, the black share was estimated whenever possible. For example, in some states the number of unknowns was known to represent the number of Hispanics. In other cases the "unknown" cases were virtually identical to the number of Hispanics, and so it was assumed that these categories were synonymous. For these states, it was estimated that 5 percent of the Hispanics were black, derived from national figures for the total population. The black male proportion of the correctional population for 1996 was estimated using a base of only those whose race was known or could be reasonably estimated for 1995. The number of black males in each category was obtained by multiplying the estimate for all blacks by the portion of the national population in each category who were male.
Some states (Alabama, Kentucky, Rhode Island, and West Virginia) did not report racial breakdowns for probation and/or parole. The share who were black was estimated by assuming that the ratio of the portion of blacks in those correctional categories to the portion in prison was the same for the state as it was for the nation as a whole. Establishing these ratios allowed us to estimate the portion who were black in each case of non-reporting.
The next stage involved the fifteen states that disenfranchise some or all ex-felons no longer under correctional supervision. Law enforcement agencies in three statesFlorida, Virginia, and Wyoming were able to furnish estimates and race/gender breakdowns of the number of ex-felons in their states. Based on the information available for the remaining states we analyzed data on total arrests, felony filings, and felony convictions for the years 1970 through 1995 to develop estimates on the number of ex-felons in each state. The year 1970 was selected as a beginning point with the knowledge that it would omit felons convicted prior to 1970 but overcount felons convicted after 1970 who had since died. These data were used to estimate the number of persons convicted of a felony in each state by year. In order to avoid double-counting those convicted multiple times it was necessary to estimate the number of offenders who had no prior felony convictions. This percentage, 55.8 percent, was the average of three national semiannual estimates conducted by the Bureau of Justice Statistics. The total number of ex-felons for Tennessee was the sum of felony convictions for the years 1970-1986. Convictions after those years do not involve lifetime forfeiture of voting rights with the exception of the offenses of first degree murder, aggravated rape, treason, and voter fraud. In Washington state, felony convictions for the years 1970-1984 result in disenfranchisement. It is likely that there is a relatively modest number of individuals in Tennessee and Washington who are double-counted. This group is composed of individuals who had a felony conviction in the period 1970-1986 and 1970-1984, respectively, who are currently under correctional supervision. We lacked sufficient data with which to develop estimates of the number of ex-felons in Kentucky and Nevada and therefore, the totals for these states represent low estimates.
The laws concerning disenfranchisement in Arizona and Maryland necessitated a slightly different method of estimating persons with felony convictions. In these states one becomes disenfranchised only after a second felony conviction. In order to avoid counting those with no prior convictions and those who had already received a second felony conviction, it was necessary to estimate for each year the proportion of offenders convicted of a second felony. The national figure derived from the three BJS estimates described above is 17.6 percent. To obtain the total disenfranchised population for these states it was necessary to add the total number of second time felons to the number of first-time felons currently under correctional supervision. The share of first time parolees, probationers, and prisoners was based on national data on sentencing of felony offenders broken down by the number of prior felony convictions.
A special estimate was also developed for Texas, which disenfranchises ex-felons for two years after release from supervision, in order to determine the number of felons who have been released from correctional supervision for two years or less and have not been reconvicted. After recording data on the number of people unconditionally released from prison and successfully completing parole and probation in 1995 and 1996, we used data on arrest and recidivism rates nationally to estimate the number of these ex-felons who were not convicted of a new felony in the two-year period.
After estimating the total number of disenfranchised ex-felons, we computed the share who were black males. We determined that 1986 represented the mid-point year for total felony convictions during the period 1970-1995 in the nine states for which data were available. The proportion of the correctional population for that year that was black represented 89.7 percent of the black proportion for 1995. Therefore, for each state which disenfranchised ex-felons, we calculated the portion who were black males as 89.7 percent of the black male portion of the correctional population of that state in 1995.
Note for Tables: Data in the tables displayed in the text are all rounded to the nearest hundred and so totals may not always exactly match the sum of each row or column.
83 Beverly Pettigrew Kraft, Legislators to address inmates right to vote, Clarion Ledger, January 26, 1997, p. 1A.