Lethal Elections: Gubernatorial Politics and the Timing of Executions

We document the existence of a gubernatorial election cycle in state executions, suggesting that election year political considerations play a role in determining the timing of executions. Our analysis indicates that states are approximately 25 percent more likely to conduct executions in gubernatorial election years than in other years. We also find that elections have a larger effect on the probability that an African American defendant will be executed in a given year than on the probability that a white defendant will be executed, and that the overall effect of elections is largest in the South and Midwest. These findings raise concerns that state executions may fail to meet the constitutional requirements stipulated by the Supreme Court in Gregg v. Georgia for the administration of state death penalty laws.


Introduction
The rapid increase over the past decade in both the number of executions conducted nationally and the number of states that utilize capital punishment has renewed interest in the policy ramifications of death penalty laws and their application. Figure 1 shows the trend in the number of executions by year from 1977 to 2000. In the period from 1976, when the death penalty was again ruled constitutional by the U. S. Supreme Court in Gregg v. Georgia, until the early 1990s there was a gradual increase in the number of executions performed by state governments. However, beginning in the early 1990s, the pace at which states have been executing defendants has accelerated rapidly, from approximately 20 executions per year in the early nineties to a high of roughly 100 in 1999.
There have also been significant increases over this period in the number of states that have reinstated the death penalty and the percentage of death penalty states that have conducted executions. Figure 2 shows the trend in the number of states that have a death penalty over the sample period. At the beginning of the sample, only 28 states had a death penalty, but over the last twenty years 10 more states have added death penalty laws. As the number of states with the death penalty has increased, the percentage of these states that execute a defendant in a given year has also increased. Figure 3 shows the trends in the percentage of states that use the death penalty over time. Over the last 4 years of the sample, almost one-half of states with a death penalty used it in any given year.
In the absence of any consensus on the deterrent effects of capital punishment (see Ehrlich, 1975Ehrlich, , 1977Grogger, 1990;and Ehrlich and Liu, 1999), the focus of recent policy debates has shifted to the possible arbitrary application of the death penalty and the associated implications for defendants' due process rights. 1 This focus is consistent with the conditions set forth by the United State Supreme Court in Gregg v. Georgia, where the Court ruled that states could again impose the death penalty provided that its application was neither arbitrary nor discriminatory. In evaluating whether current state practices meet these criteria, policymakers have for the most part focused on racial and other disparities observed at the sentencing stage of the process, with considerably less attention being paid to possible irregularities existing at the time of execution.
In addition to concerns about possible disparate treatment of defendants based on race, concerns have also been raised regarding the possibility that political and other extra-legal factors may be playing a role in both the sentencing and punishment phases of capital cases (Culver, 1999;Langbein, 1999;Pridemore, 2000). In this paper, we examine one such political factor: Whether gubernatorial elections influence the timing of state executions. State and local elections have previously been shown to exert an independent influence on other public policy decisions. 2 Here, we find that the presence of a gubernatorial election increases the probability of a state execution by approximately 25 percent. We also find that elections have a larger effect on the probability that an African American defendant will be executed in a given year than on the probability that a white defendant will be executed, and that the overall effect of elections is largest in the South and Midwest. Although not definitive, we present some evidence that the cyclical effects we identify lead to reductions in the amount of time that executed defendants spend on death row. 3 These results suggest that concerns about legal due process should not be restricted to the sentencing phase, but should also extend to the manner in which defendants are selected for execution.
The issue of how gubernatorial discretion is exercised in capital cases has taken on increased importance over time as the availability of post-conviction judicial review has been increasingly limited at both the state and federal levels (Langbein, 1999;Pridemore, 2000). More generally, our findings raise questions about the extent to which states are in compliance with the constitutional requirements for executions set forth by the United States Supreme Court in Gregg v. Georgia.
The paper is organized as follows. In Section 2, we briefly review a number of previous studies of state executions. We describe the data in Section 3. In Section 4, we discuss our empirical methodology and present our main findings. Section 5 explores, to the extent possible, alternative mechanisms through which the timing effect we document may be implemented. Concluding remarks are offered in Section 6.

Previous Studies
We are aware of three previous studies that have examined the determinants of state executions in the post-Gregg era. The first, by Culver (1999), provides a useful overview of the way in which conflicts between competing branches of state governments have influenced the frequency with which states perform executions. Culver documents the widespread politicization of the death penalty at the state level and the sometimes intense political pressure that is brought to bear on elected officials who oppose capital punishment. A well-known example is the removal of Rose Bird and two of her colleagues from the California Supreme Court, the first time in the state's history that appellate judges were removed from office. In a similar case, Penny White, a Tennessee Supreme Court justice, was the first appellate judge in Tennessee to lose a retention election, primarily due to her support for a controversial decision that overturned a death sentence in a high-profile murder case.
Culver also discusses the apparent political pressures that capital cases create for governors.
Examples include New Mexico's Toney Anaya (D: 1983-1986), who in his last months in office commuted the death sentences of all five men on New Mexico's death row, and Ohio governor Richard Celeste (D: 1983Celeste (D: -1991, who commuted the death sentences of seven death row prisoners just four days before leaving office. The timing of executive decisions in these examples suggests that political considerations have played a role in the disposition of capital cases. The second study, by Langbein (1999), is the first empirical analysis of capital punishment to focus on executions rather than sentencing. She examines whether the same racial and political factors that appear to play a role in determining which defendants receive the death penalty carry over to the decision to perform an execution. Using data on a panel of death penalty states from 1977 to 1992, she finds that the number of executions performed in a state are significantly related to measures of black political power and the adoption by states of restrictions on the post-conviction legal options of defendants. She also finds some evidence that the race and gender of victims plays a role, as does the severity of the crime.
A final study, by Pridemore (2000), examines the determinants of governors' commutation decisions. Using data on 4800 persons sentenced to death in the United States between 1974 and 1995, he finds that defendants are less likely to have their sentences commuted (are more likely to be executed) if they are older, male, imprisoned in the South, or have the final disposition of their sentence determined in a gubernatorial election year. He finds no evidence that the defendant's race plays a role in the decision to execute.
Although Pridemore's finding of a gubernatorial election cycle in commutation decisions is suggestive of the type of political influence that we seek to quantify in our paper, our analysis differs from his in several important ways. First, his study seeks to explain all of the factors that influence the decision to execute, while our paper focuses more specifically on the role of election year politics.
Second, although his study is based on a relatively long panel of data, he does not control for either national trends in executions or state-specific differences in the propensity to execute. Third, he considers only cases that end in a commutation or an execution, while we allow for additional transitions off of death row and account for the censoring problem that arises when a subset of death row stays have yet to be completed at the time of the analysis. Finally, given his broader objective, he does not examine other (related) outcomes that may be influenced by elections, such as the number of executions performed, differential effects of elections by race, and the impact of elections on the amount of time that prisoners spend on death row. In the present paper, we conduct a more comprehensive analysis of the impact of gubernatorial politics on state executions.

Data
The execution data come from two sources. The first is a panel of U.S. states, with yearly We also have information on all persons sentenced to death since 1972 from the Bureau of Justice Statistics publication, Capital Punishment in the United States: 1972States: -1999. This data set contains information on the demographic characteristics of death row inmates, their criminal backgrounds, and the amount of time that each spent on death row. For each year that a state has at least one execution between 1977 and 1999, we calculate the average time on death row for the defendants executed that year. 5 On average, the wait on death row is slightly less than 10 years.
Data on the timing of gubernatorial elections is taken from the Book of the States. Election cycles vary across states for several reasons. First, some states have gubernatorial elections every two years 4 These restrictions leave a data set that includes 842 state/year cells. 5 There are 199 state/year cells with at least one execution between 1977 and 1999. while most states have elections every four years. Also, most states schedule their elections on even calendar years, but there is a significant minority of states that hold elections in odd years. Finally, among states with a four-year election cycle during even years, some hold elections in presidential election years while others have elections at the midpoint of presidential terms. There is a similar staggering for states with four-year cycles that hold elections in odd years.

Election Cycles in State Executions
To measure the effect of gubernatorial elections on executions, we begin by estimating a probit model of the form: where i indexes states and t indexes time. The coefficient of interest is β , which measures how having a gubernatorial election in a state affects the probability that the state has an execution that year. The year dummies control for national trends in executions that may be correlated with gubernatorial elections. The state fixed effects control for any fixed state-specific omitted variables that may be correlated with the propensity of states to hold executions, and the state trends control for linear changes over time in the propensity of a state to perform executions that might be correlated with elections. Therefore, β is identified by differences in execution behavior in states with and without a gubernatorial election in a given year that are different than their linear trends.
The estimates of this probit model are presented in Table 3, using our sample of executions from 1977-2000. Column (1) displays the results of the estimation of equation (1). The coefficient on the election indicator is positive and statistically different from zero. The estimated marginal probability suggests that a gubernatorial election increases the probability of a state execution by slightly less than 6 percentage points. Evaluated at the mean execution probability observed in our sample, this estimate indicates that states are about 25% more likely to perform an execution in an election year than in other years.
We are concerned that the state linear trends might not be adequately controlling for time-varying omitted variables that are correlated with elections and the probability that a state holds an execution. Therefore, we investigate the robustness of our results to two alternative specifications. In column (2) 7 We have also estimated these relationships using logit and linear probability models and obtain qualitatively similar estimates of the marginal effects using these other models. execution of slightly less than 7 percentage points, which represents more than a 25% increase over the baseline execution probability.
Given our finding that elections increase the probability of an execution in a state, we next examine whether the effect of an election on the likelihood of a state execution varies by the race of the defendant. We re-estimate equation (1) with two separate dependent variables: the first is an indicator for whether a state executes at least one white defendant in a given year and the second is an indicator for whether a state executes at least one African-American defendant in a given year. These results are presented in Table 4. Columns (1) through (3) present the results for the executions of white defendants using: in column (1), our basic probit model with state-specific trends; in column (2), division× year effects; and in column (3), governor fixed effects. In all specifications, the effect of gubernatorial elections is positive but small and not statistically different from zero; a gubernatorial election only increases the probability that a state executes a white defendant by between 7% (column (1)) and 13% (column (2)).
On the other hand, as shown in columns (4) through (6), there is a large effect of elections on the probability that a state executes an African-American in all specifications. The effect of a gubernatorial election is positive, large and statistically different from zero in all specifications, implying that a gubernatorial election increases the chance there is at least one execution of an African-American defendant by between 29% (column (6)) and 37% (column (5)). 9, 10 9 In the literature on sentencing, it is often noted that attempts to determine the pure effect of race on the receipt of the death penalty are confounded by the fact that African-Americans are more likely to be involved in murders with aggravating circumstances (Langbein, 1999, p.634). To investigate this possibility, we estimated the same probit models as were used for black defendants, but changed the dependent variable to the probability that the state executes at least one defendant who was involved in a multi-victim homicide in a given year. The election coefficients from these models were two to three times smaller than the coefficients from the corresponding models for black defendants and were not statistically different from zero. 10 In contrast to many of the studies on racial disparities in sentencing, we did not find evidence of disparate treatment based on the race of the victim. One possible reason for this is the potentially different motivations that arise at the sentencing and punishment stages. At the time of sentencing, there is a substantial focus on the victims of the crime; however, by the time an inmate is scheduled to be executed, news accounts typically focus on the race of the inmate rather than the victim.
There are also differences in the effect of gubernatorial elections on executions by region of the country. We divide the U.S. into the four census regions and estimate a regression model that allows the effect of gubernatorial elections to vary across these regions. 11 The model specification is: ε is an error term, and the other variables are defined as before. 12 The results of this estimation are displayed in Table 5. Column (1) presents the basic estimates, column (2) adds division × year effects to the regression specification and column (3) adds governor effects.
All specifications produce similar patterns in the coefficients of the region/election interactions.
Although the coefficients are imprecisely estimated, Table 5 suggests that the positive effect of gubernatorial elections on the probability of executions is concentrated in states in the South and North Central regions, especially the South. In other regions, the effect is much smaller and is even negative for states in the West.
Using a similar methodology, we also examine whether there are differences in the effect of gubernatorial elections on executions based on the party affiliation of the governor. The model specification is:  Table 7 shows the annual probability of these transitions in our data set. In our multinomial logit model, the probability of outcome j occurring is given by: and 13 We have also estimated this model using a specification that allows for a different effect of elections if the governor is a Republican, Democrat or independent. Again, we find little difference in the effect of elections across these states. 14 We have also attempted to examine whether the election effect varies based on the perceived tightness of the gubernatorial election. One way to classify the tightness of the election would be to examine the actual vote totals; however, we worry that the outcome of the vote might be a function of whether there were executions during the election year. Polling data on the popularity of the incumbent governor (sufficiently far in advance of the election to permit a reaction by the governor) would be a better measure, but consistent polling information across states and over time is not readily available. We have tried two methods for classifying the tightness of the election. The first indicator is whether the governor's party lost seats in the previous legislative election. If the governor's party lost seats, that might be an indication that the governor is vulnerable. The second is an indicator for whether a governor is running for reelection for the first time. Governors again might be more vulnerable when they have the least tenure. Using both methods, we failed to find a differential effect of elections based on the tightness of the race. The results of this estimation are presented in Table 8. We are most interested in two transitions out of death row: executions and overturned sentences. Therefore, the coefficients we present measure the effect of gubernatorial elections on the probability that a defendant is executed instead of remaining on death row and the probability that a defendant has his sentence changed instead of remaining on death row.
Columns (1)  year. To do this, we estimate a count model in which the independent variables are the same as in our probit models, but the dependent variable is a measure of the number of executions that a state holds in a year.
Results for three negative binomial regressions are presented in Table 9. The coefficient on the indicator for a gubernatorial election in column (1)  16 Because the election indicator only varies at the state level, we adjust the standard errors to take into account the within-state correlation of the observations. 17 The count models in Table 9, as well as the probit models in Tables 3 and 4, were also estimated using data for the pre-Furman era from the publication, Executions in the United States, 1608-1991: The ESPY File. For the period

Issues Related to Timing
In this section, we explore the process that leads to additional executions in election years relative to non-election years. Within the limits of our data, we examine two issues. First, we study whether the election cycle in state executions is generated by, on average, moving executions up in time or holding them back. This distinction is of interest because the latter scenario would not appear to compromise the due process rights of death row inmates while the former scenario might. Second, we test a simple model of this process, examining whether there is a one-to-one offset between the number of executions held in election years and the number of executions held in the year after an election. As discussed below, the data do not permit us to directly test whether the additional executions held in election years represent "extra" executions that would not have occurred otherwise, or whether they instead reflect substitutions over time from among a fixed stock of executions. Of course, these two possibilities are not mutually exclusive.
We begin by examining how gubernatorial elections affect the time that defendants who are executed stay on death row. We investigate whether defendants executed during election years stayed on death row for shorter periods of time than defendants executed in other years using information on all persons on death row between 1973 to 1999. For each year that a state has at least one execution, we calculate the average time the people executed in that state were on death row. 18 The regression specification is of the form: 1935-1968, we failed to find any large or significant effect of elections on either the probability of conducting an execution (either in general, or broken down by race), or on the number of executions performed. This result is perhaps not surprising in light of the fact that executions were nearly ubiquitous during this period, implying that the marginal political benefit from holding an additional execution was probably negligible. is the average number of months that defendants who were executed in state i in year t waited on death row, and the other variables are defined as before.
The results of the OLS regressions of equation (5) are presented in Table 10. The coefficient on the election indicator is negative and statistically different from zero in column (1), suggesting that defendants executed during election years spent about 19% (exp(-.1777) -1) less time on death row than people executed in other years. Given an average stay on death row of approximately 10 years, this implies that inmates executed during election years have their stays shortened by slightly less than two years on average. In column (2), division × year effects are added to the specification; the coefficient becomes smaller in absolute value and is not statistically different from zero. By this estimate, a defendant executed during an election year has about a 9% shorter spell on death row (approximately 11 months) than other defendants who are executed. The final column adds the governor effects. The coefficient implies that inmates executed during election years had about a 14% shorter stay on death row (approximately 17 months) than other executed defendants, but the standard error is too large for this effect to be statistically different from zero.
Given that inmates who are executed during election years appear to spend less time on death row than other inmates, a natural question to ask is whether the increased executions during election years represent "extra" executions, or whether they would have occurred at a later date anyway. This is a difficult question because the evidence we have generated does not allow us to discriminate among these two possibilities. To definitively distinguish between the two scenarios, one would need to know the counterfactual of how states would behave in the absence of an election cycle. But because all states have election cycles, and have them all the time, it is impossible to know this counterfactual.
To informally analyze this issue we restrict attention to states with four-year gubernatorial terms, and examine whether the increase in the number of executions during gubernatorial elections is followed by a decrease in executions the following year. If there is a large decrease in executions in the year following an election, then there is some evidence that executions are being moved forward one year.
The negative binomial model specification is: is a dummy that it is the year after a gubernatorial election in state i, and the other variables are defined as above. The results of the estimation of this model are presented in Table 11. Column (1) shows the estimates of equation (6). As before, there is a large and positive election year effect; the coefficient on the election indicator is almost identical to the estimate in Table 9. These results do not appear consistent with the view that states are simply moving some of their executions forward one year. However, more complicated substitutions may be taking place that we are unable to test for directly. Also, given that a relatively small percentage of death row inmates are ultimately executed, it is also possible that election cycles lead to executions that might not have otherwise taken place. Sorting out the exact mechanisms through which election cycles in executions are implemented does not appear to be possible with our data, but remains an important area for future research.

Conclusions
Our analysis indicates that holding other factors constant, states are approximately 25 percent more likely to conduct executions in gubernatorial election years than in other years. Moreover, elections have a larger effect on the probability that an African American defendant will be executed in a given year than on the probability that a white defendant will be executed. We also find evidence that the total number of executions performed is higher in election years and that the relationship between elections and executions is strongest in the South and Midwest. Further, we find some evidence that the existence of politically-timed executions reduces the average time that executed defendants spend on death row, suggesting that the increased executions observed in election years may result from an acceleration of the process by which inmates are selected for execution.
Taken together, our results indicate that election year political considerations influence both the timing and racial composition of executions, a finding which seems in conflict with the Supreme Court's requirement that states administer the death penalty in a consistent and nondiscriminatory manner.
Although not a legal analysis, this work does point to the need for further research on how death sentences are carried out by state governments.             1977 1979 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999 Year Number of Executions  1977 1979 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999 Year Number of States with Death Penalty