Hostname: page-component-745bb68f8f-5r2nc Total loading time: 0 Render date: 2025-01-30T05:34:51.707Z Has data issue: false hasContentIssue false

“Still Doin’ Time”: An Analysis of the Conditional Effects of Legislative Term Limits on State Incarceration Rates

Published online by Cambridge University Press:  27 January 2025

Jason S. Byers
Affiliation:
University of Connecticut
Laine P. Shay*
Affiliation:
Texas A&M University-Corpus Christi
*
Corresponding author: Laine P. Shay; Email: [email protected]
Rights & Permissions [Opens in a new window]

Abstract

A vast body of work investigates the consequences of legislative term limits for public policy. However, considerably less research has delved into their effects in noneconomic policy domains. In this article, we develop the argument that implemented term limits increase the effect that a state government’s ideology has on the state’s incarceration rate. When analyzing incarceration rates among all states between 1979 and 2017, we find evidence to support our theoretical expectation. Specifically, for states with term limits, we find that an increase in state government conservatism is associated with a higher incarceration rate. Conversely, for non-term-limited states, we find that the policy preferences of the state government have little influence on the incarceration rate. These findings deepen our insight into how institutional design can affect public policy.

Type
Article
Copyright
© The Author(s), 2025. Donald Critchlow and Cambridge University Press

“Arkansas has the fifth highest incarceration rate in the country—a per capita rate of 942 per 100,000 people that greatly exceeds that of all independent democracies around the world, including the broader United States,” an Arkansas Times article began, “[b]ut Gov. Sarah Sanders, Attorney General Tim Griffin and legislative leaders aren’t settling for fifth place in the contest over who can spend the most taxpayer dollars locking away the most people for the longest amount of time.”Footnote 1 The article goes on to describe how the Arkansas legislature is drafting a truth-in-sentencing policy that would limit the possibility of parole for inmates who are convicted of certain crimes.Footnote 2 Advocates of the criminal justice policy assert that it will help fight crime by keeping criminals off the streets.Footnote 3 Experts expect that the policy would increase Arkansas’s incarceration rate,Footnote 4 which is already well above the national average.Footnote 5 Arkansas’s tendency to implement punitive criminal justice policies that result in a high incarceration rate stands in stark contrast to those of other states. For instance, in Massachusetts, the incarceration rate is 275 inmates per 100,000 people.Footnote 6 Even more, the Massachusetts legislature still actively strives to lower its incarceration rate, with the state recently scrapping a few “mandatory minimum”Footnote 7 sentence laws for nonviolent crimes.Footnote 8 These two divergent examples raise the following question: What political factors account for a state’s incarceration rate?

Scholars suggest that the government’s policy preferencesFootnote 9, public opinionFootnote 10, and racial demographicsFootnote 11 partially explain the number of inmates imprisoned within a state. One political factor that has not been taken into consideration is state legislative term limits. Currently, 16 states have implemented legislative term limits.Footnote 12 A well-established body of works that focuses on term limits finds that these laws influence the policy-making process and policy outcomes. For example, one study finds that term limits are associated with an increase in state government debt.Footnote 13 However, most prior research focuses primarily on economic policy outcomes. We suggest that if term limits can broadly influence the policy adoption process and they can influence economic policy outcomes, then term limits have the potential to also influence noneconomic policy domains, such as criminal justice.

Building on the literature that outlines how term limits strengthen the influence of chamber leadership, we assert that legislative term limits should strengthen the effect that a state government’s ideology has on a state’s incarceration rate. When analyzing incarceration rates among all states between 1979 and 2017, we find evidence to support our theoretical expectation. Specifically, for states with term limits, we find that an increase in conservatism is associated with an increase in the number of inmates. Conversely, for states without term limits, we find that the policy preferences of the government have little effect on the state’s incarceration rate.

This analysis contributes to the literature by advancing our understanding of the implications that term limits have on public policy in several ways. First, as previously mentioned, scholars have thoroughly analyzed the influence that term limits have on economic policy outcomes. However, considerably less research has delved into the influence that legislative term limits can have on noneconomic domains. Our results suggest that term limits can affect a broad array of policy areas. Second, toward the end of the twentieth century, the number of incarcerated inmates increased dramatically.Footnote 14 This has resulted in a robust discussion within the public about the appropriate criminal justice policies to implementFootnote 15 and influenced American culture.Footnote 16 Our findings suggest that term limits have a polarizing effort on the imprisonment rate. Conservative states with term limits tend to experience an increase in the number of inmates serving. Conversely, liberal states frequently experience declines in their imprisonment rates. Third, this research highlights the issue of disparate justice among the US states, where individuals may receive varying punishments depending on their state of residence. Our findings suggest that institutional design and the policy preferences of lawmakers may enhance this uneven distribution of justice across the United States. Finally, several studies find that electoral considerations can influence criminal justice policy outcomes. One could reasonably argue that term limits potentially dampen the electoral incentive for lawmakers to implement punitive policies that result in a higher incarceration rate. Our results suggest this is not the case for more-conservative states. Taken together, these findings deepen our insight into how institutional design can influence public policy.

State Context and Justice

One pillar of a healthy democracy is the presence of elections.Footnote 17 Consequently, there is a flourishing of literature investigating how elections influence policy responsiveness and the policy-making process.Footnote 18 In this section, we briefly outline two subsets of studies that are relevant to our research question: (1) the influence that elections have on criminal justice policies and (2) the effects of legislative term limits on the policy-making process.

First, a vein of relevant research finds that electoral considerations encourage policy makers to pursue punitive criminal justice policies.Footnote 19 For instance, lawmakers who support more punitive criminal justice policies are sometimes awarded by their constituents in the next elections.Footnote 20 This pattern emerges because punitive criminal justice policies tend to be popular among the public.Footnote 21 Furthermore, research finds that more competitive legislative elections are associated with conservative states experiencing an increase in their incarceration rates.Footnote 22 Conservative states enact these more punitive criminal justice policies, which result in higher incarceration rates, in an effort to please their constituents and boost their reelection efforts.Footnote 23 Afterall, policy makers who cannot get the crime rate under control are potentially punished in their next election.Footnote 24 Thus, there appears to be an electoral connection stemming from criminal justice policies.

A second important body of works analyzes the implications of term limits.Footnote 25 Regarding the influence that term limits have on public policy, most studies focus on the implications that term limits have on fiscal policy outcomes. For instance, the implementation of term limits corresponds with lower state bond ratings.Footnote 26 This pattern emerges because the lack of policy expertise in term-limited chambers results in the lawmakers adopting less sound fiscal policies. Likewise, states with term limits correspond with lower general fund balances.Footnote 27 Term limits have also been shown to incentivize lawmakers to spend more lavishly.Footnote 28 Although a great deal of scholarly attention has been devoted to the influence that term limits have on fiscal policy outcomes, considerably less research focuses on their effects in noneconomic policy areas. The scant research in this area has shown that term limits result in an increase in the likelihood of policy congruence between voter preferences and policy outputs.Footnote 29 This pattern occurs because terms limits cause lawmakers to represent their constituent’s policy preferences more accurately.Footnote 30 In another prominent example, term limits are shown to undermine policy innovation within a state.Footnote 31

Although these two impressive veins of literature deepen our understanding of term limits, public policy, and elections, there is still a great deal to learn. For instance, whereas there appears to be a connection between competitive elections and punitive criminal justice policy outcomes, it is unclear what the policy implications are when the electoral connection is completely severed via term limits. One could reasonably assume that the dampened electoral incentives could result in lawmakers pursuing less-punitive criminal justice policies, resulting in lower incarceration rates. However, other studies suggest that term limits can strengthen interest groups and influential leaders in a state who may wish to adopt more punitive policies.Footnote 32 Furthermore, most research on the policy implications of term limits tends to focus on fiscal policy outcomes. Few studies examine whether legislative term limits influence noneconomic policy outcomes. The purpose of this study is to address some of these gaps in the literature by analyzing the relationship between legislative term limits and a state’s incarceration rate.

Linking Legislative Term Limits to State Incarceration Rates

Polling and other evidence suggest that term limits tend to be broadly popular with the public.Footnote 33 Since the implementation of term limits, scholars have uncovered numerous empirical patterns. For instance, legislative term limits have resulted in legislators devoting less attention to cultivating the “personal vote” with their constituents.Footnote 34 Afterall, as lawmakers in term-limited states will be in office for only a short period, there is little incentive for them to focus on constituency-oriented activities. In another prominent example, legislative term limits have been shown to increase voter participationFootnote 35 and competitive elections.Footnote 36 Students of legislative politics have also shown that term limits influence policy outputs.Footnote 37 In this section, we outline how legislative term limits, conditioned on the government’s policy preferences, may affect the incarceration rate within a state.

Before outlining how term limits influence policy outcomes, it is important to briefly outline the history of term limits. Peverill Squire notes that in colonial America one of the earliest experimentations with legislative term limits was in Pennsylvania.Footnote 38 Moreover, Edward López points out that “de jure term limits” were embedded in the Articles of Confederation and that term limits “almost entirely disappeared until the 1990s.”Footnote 39 Specifically, in the early to mid-1990s, numerous states adopted legislative term limits via citizen initiatives.Footnote 40 It should be noted that there is considerable variation in terms of the provisions within the laws among the states with implemented term limits. Some states, such as Oklahoma, have a lifetime ban, and other states, such as Ohio, have consecutive bans.Footnote 41 Moreover, there is variation among term-limit laws in terms of time. Maine limits lawmakers to eight years, and Louisiana has a limit of 12 years.Footnote 42 A few states, such as Utah, had their term limits laws repealed.Footnote 43 Currently, 16 states have term limits, with North Dakota being the most recent state to adopt term limits in 2022.Footnote 44

The way in which legislative term limits influence incarceration rates might not work the same way for every state. In particular, the state government’s ideology could be a critical factor. Previous research finds that ideology is a powerful factor in terms of political behavior and policy outcomes.Footnote 45 We assert that more-conservative state governments should be associated with higher incarceration rates.Footnote 46 Conservative lawmakers often attempt to reinforce the image to the public that they will maintain “law and order.”Footnote 47 Thus, it is not surprising that more-conservative states are noted for their support of fighting crime by advocating for more punitive criminal justice policies and for having higher incarceration rates.Footnote 48 For instance, more-conservative states are associated with quickly adopting “Three Strikes” laws, which require that repeat offenders receive minimum mandatory sentences.Footnote 49 Furthermore, more-conservative state governments have been shown to fund legal defense programs at lower levels than do more-liberal states.Footnote 50 These and other state policies partially explain why conservative states are associated with higher incarceration rates. Conversely, liberal governments are more likely to implement rehabilitation policies, such as those states that have advocated for the decriminalization of drugs.Footnote 51 Furthermore, research shows that more-liberal governments are more likely to support prison furlough programs.Footnote 52 Thus, more-conservative governments should be associated with higher incarceration rates relative to liberal governments.

However, how exactly might legislative term limits affect a state’s incarceration rate? We assert that the presence of legislative term limits should enhance the influence that the state government’s ideology has on a state’s incarceration rate. Legislative term limits are known to empower the chamber leaders in a state.Footnote 53 One consequence of term limits is that they deprive legislators of their own institutional and policy knowledge.Footnote 54 Because state lawmakers do not have their own expertise to rely on when making policy decisions, they are forced to depend on other political actors, such as their party’s leadership, for decision making.Footnote 55 Previous research has shown that the implementation of term limits has resulted in lawmakers delegating more procedural tools to control the lawmaking process to their leadership relative to non-term-limited states.Footnote 56

It is important to recognize that party leaders themselves are termed out of office; however, as noted by Christopher Mooney, lawmakers delegate procedural power to their leadership to overcome “collective action problems” in the legislature.Footnote 57 Lawmakers without the institutional knowledge and experience, due to term limitsFootnote 58, cannot easily usher in their policy agenda, which is a paramount objective for many lawmakers.Footnote 59 Furthermore, there is evidence that term limits result in party leadership having more influence over the recruitment of candidates and the candidate’s success in fund-raising.Footnote 60 This might result in states adopting more ideologically consistent policies. Therefore, we suggest that lawmakers in term-limited states are more likely to delegate power to their party leaders.

We suggest that the strong chamber leaders that emerge in states with term limits will try to steer the legislature to implement more policies that are consistent with their ideology.Footnote 61 Chamber leaders help to maintain more ideologically consistent policies in the hopes of protecting their party’s brand and to help their colleague’s reelection chances.Footnote 62 In the case of criminal justice policies, we expect that more-conservative state governments with term limits should pursue more-punitive criminal justice policies and, thus, be associated with a higher incarceration rate. On the other hand, more-liberal states with term limits should be less inclined to support policies that result in an increase in the prison population. Conversely, for states without term limits, the legislators have their own experiences and policy expertise that they can rely on. This gives the legislators, in this political context, the luxury of not having to rely on their party leaders in the decision-making process and not being incentivized to adopt more ideologically consistent policies. Thus, we should observe ideology having a weaker effect in non-term-limited states. This discussion leads us to the following theoretical expectation:

Term Limits Hypothesis: The state government’s ideology will have a larger effect on state incarceration rates for states with legislative term limits than on states without term limits.

However, it is also possible that we may find that the effect that term limits have on a state’s incarceration rate does not vary by the state government’s preferences. Several scholars find that criminal justice is a unique policy area with low levels of polarization and that the policy-making process tends to be bipartisan.Footnote 63 Furthermore, combating crime is widely considered an issue in which there is a broad consensus among politicians and the public.Footnote 64 Therefore, lawmakers in term-limited states may have little incentive to pursue policy outcomes (i.e., state incarceration rates) that are consistent with their ideology. If this is the case, then we would find that the effect of legislative term limits does not vary by a state government’s ideology. However, if we do indeed find that the effect of term limits on state incarceration rates does vary by state ideology, this would suggest that term limits can influence a wide array of policy domains.

Data and Methods

We have presented our theoretical rationale for the influence that term limits potentially have on criminal justice policy, and now we turn to an empirical investigation of our rationale. Our data consist of information relating to the incarceration rates of each US state from 1979 to 2017. Specifically, we use panel data, with our unit of analysis consisting of a state in a given year.

Dependent Variable

Our dependent variable, incarceration rate , is measured as the count of prisoners serving a minimum one-year sentence in a state correctional facility per 100,000 population in a state. This data is from the Bureau of Justice Statistics.Footnote 65 The average value is 292.99, and the standard deviation is 144.01. The minimum value is 34, and the maximum value is 870. Because our dependent variable is a nonnegative continuous variable, we estimate our coefficients with a linear model.Footnote 66

Main Independent Variables

Our main independent variables are the government liberalism, legislative term limits, and an interaction between the two. For government liberalism , we use William Berry and his colleagues’ state ideology scores.Footnote 67 The creators of these ideology scores use the state’s congressional delegation’s NOMINATE Common Space scoresFootnote 68 to approximate the ideological preferences of the state government actors. We use this measure at suggested by previous researchers and considering the lengthy time span we are analyzing.Footnote 69 Higher values indicate that the state government is more liberal. The average value is 49.18. The minimum value is 17.51 and the maximum value is 73.62.

Regarding legislative term limits , it is important to use a measure that captures the variation of term limit laws across the United States. We use Travis Baker and David Hedge’s continuous measure of term limits.Footnote 70 This measure incorporates the level of turnover, type of term limit ban, and the ability of legislators to recycle through the chambers. Higher values indicate a more stringent term-limit law. States without term limits are coded as zero.Footnote 71 We code the term-limits variable based on when the term limit law is implemented.Footnote 72 We analyze the implementation of term limits rather than their adoption because the literature suggests that it is the implementation of term limits that influences the policy-making process.Footnote 73 All our explanatory variables are lagged by one year. Finally, we test the interaction between the legislative term limits and government liberalism variables to determine whether the effect that term limits have on the state incarceration rate varies based on the policy preferences of the state government.

Additional Explanatory Variables

We control for other factors that could be associated with a state’s incarceration rate, and several of these variables come from the Correlates of State Policy database.Footnote 74 Previous research has found that the public’s policy preferences can influence the level of incarceration in a state.Footnote 75 Specifically, more-liberal citizens should elect and put pressure on their state government to be less supportive of punitive criminal justice policies. The citizen liberalism variable controls for the ideology of the public. This variable is coded so that higher values indicate a more-liberal public.Footnote 76 We expect the coefficient to be negatively signed and statistically significant.

Michael Olson and Jon Rogowski find that legislative professionalism can enhance the influence of legislative term limits.Footnote 77 Therefore, we use Peverill Squire’s measure of legislative professionalism, which includes a state’s legislator salary, calendar length, and staffers.Footnote 78 Squire provides measures in the following years: 1979, 1986, 1996, 2003Footnote 79, and 2015.Footnote 80 Higher values indicate that the legislature is more professionalized. We also control for the level of interest-group activity in the state legislature. This is measured by taking the total number of registered interest groups that relate to the criminal justice system (i.e., police, fire departments, and correctional workers) and divide it by the total number of groups that are formally registered with the state government.Footnote 81 Higher values indicates that law enforcement agencies might have more influence within the state. This measurement of interest-group activity is similar to the one developed Daniel Lewis and his colleagues.Footnote 82

States with a higher percentage of non-White individuals in the population have been shown to be associated with higher state imprisonment rates.Footnote 83 This variable is measured as the percentage of non-White individuals living in the state.Footnote 84 We expect the coefficient for this variable to be statistically significant and positively signed.Footnote 85

States with higher crime rates might be associated with higher incarceration rates. We account for this factor with two different variables. The violent crime rate variable is the violent crime rate per 100,000. Higher values indicate a higher violent crime rate.Footnote 86 The property crime rate variable is the property crime rate per 100,000. Higher values indicate a higher property crime rate.Footnote 87 We expect both coefficients to be statistically significant and positively signed.

A poor economy can increase the crime rate in a state,Footnote 88 which in turn may affect the number of inmates in a state. Therefore, we include multiple economic indicators in our models. The unemployment rate is the percentage of a state’s labor force that is currently not working.Footnote 89 We also control for the state’s minimum wage .Footnote 90 Higher values are associated with a higher state minimum wage. The poverty rate is the percentage of a state population that lives in poverty.Footnote 91 We control for the income per capita in a state. Higher values indicate that the average state citizen is wealthier. Finally, the economy size variable is a state’s GSP.Footnote 92 Higher values indicate a stronger state economy.Footnote 93

Previous research has found that the current incarceration rate in a state influences future incarceration rates.Footnote 94 Therefore, we include a lagged dependent variable. This variable, incarceration rate t-1, also accounts for autocorrelation in the model and unobserved heterogeneity that is correlated with the units.Footnote 95 We also estimate our model with state and year fixed effects. We present in parenthesis robust standard errors that are clustered on each state.

Findings

The estimates from our linear model are presented in Table 1. First, we will interpret the interaction components. The interaction component for legislative term limits is statistically significant and positively signed. When the state government variable equals zero, this indicates that an increase in values for the legislative term limits variable is associated with an increase in the state incarceration rate. For the state government liberalism variable, it is not statistically significant. For non-term-limited states (i.e., when the legislative term limits variable equals 0), this indicates that an increase in state government liberalism does not affect their incarceration rates.

Table 1. State Incarceration Rates, 1979–2017

Note: Unit of analysis is a state in a given year. Dependent variable is the number of inmates in a state prison (per 100,000). All explanatory variables are lagged by one year. The coefficients are estimated from a linear model. The model includes state and year fixed effects. Presented in parenthesis are robust standard errors clustered on each state.

* p ≤ .05 (all one-tailed tests).

Importantly, our interactive legislative term limits × government liberalism variable is significant and negatively signed. This indicates that legislative term limits have a polarizing effect. Conservative state governments with term limits are associated with an increase in their incarceration rates. Conversely, more-liberal state governments with term limits tend to have a lower incarceration rate.

The effect of our interactive variable on incarceration rates can be seen graphically in Figure 1. Figure 1 demonstrates the predicted incarceration rate across a range of state government ideology scores. We also plot states without term limits (i.e., solid line and light purple confidence band) and states with the most stringent term limit policy (i.e., dashed line and gray confidence band). For states with the most stringent term-limit policy, the most conservative state government is associated with an incarceration rate of 308. When we shift the state government liberalism variable to the most liberal value, it falls to 289. Conversely, for states without term limits, the most conservative state governments correspond to an incarceration rate of 290. The most liberal state governments without term limits tend to have an incarceration rate of 295.

Figure 1. Predicted Incarceration Rates. This figure is derived from the estimates presented in Model 1. The solid line represents states without term limits. The dashed line represents the legislative term limit set at its maximum value. Higher values of government liberalism correspond with greater government liberalism. All control variables are held constant. Shaded areas represent 95% confidence band.

Most of our additional explanatory variables perform as expected. States with more-liberal citizens, lower minimum wage, higher unemployment rateFootnote 96, and fewer violent crimes tend to have a lower incarceration rate. We also find that more-professionalized legislatures are associated with a higher incarceration rate. Conversely, none of the remaining variables appear to have a significant effect on the incarceration rate within a state.

Robustness Check: Alternative Models

It is important that our results remain robust to various model specifications and alternative measures of legislative term limits. In this section, we conduct three additional robustness tests with the results displayed in Table 2. Unless stated otherwise, we use the same variables and methods as described above. It should also be noted that we have conducted an additional robustness check with data provided by Boris Shor and Nolan McCarty.Footnote 97 The results are placed in the Appendix (Table A1).Footnote 98

Table 2. Additional Alternative Model Specifications

Note: Unit of analysis is a state in a given year. Dependent variable is the number of inmates in a state prison (per 100,000). All explanatory variables are lagged by one year. Coefficients estimated from a linear model. Models 2 and 3 include state and year fixed effects. Model 4 only includes year fixed effects. Presented in parenthesis are robust standard errors clustered on each state. Model 2 uses a dichotomous term limits measure. Model 3 includes Sarbaugh-Thompson’s measure of term limits. Model 4 is the placebo test.

* p ≤ 0.05 (all one-tailed tests).

First, we use the traditional measure of legislative term limits: a dichotomous measure of term limits. States with implemented legislative term limits are coded as one and all other observations are coded as zero. The results are displayed in Model 2 of Table 2. The legislative term limits × government liberalism interaction is still significant and negatively signed.

Second, Marjorie Sarbaugh-Thompson has created various measures of legislative term limits.Footnote 99 We use her measurement that includes chamber turnover, ability to recycle within chambers, and type of ban, weighted by each chamber.Footnote 100 Higher values indicate a more-stringent term-limit law. The results are shown in Model 3. Again, our interactive variable is correctly signed and statistically significant with this alternative measurement.

Finally, like Susan Miller and her colleagues, we conduct a test that is similar to a “placebo test.”Footnote 101 This test analyzes whether unobserved state factors that are correlated with the implementation of term limits could be influencing our findings. For this analysis, we limit our data to all states before 1990. We use this period because 1990 was the year the first state adopted term limits. Next, we coded each state that eventually implemented term limits as “one” and all other observations as “zero.” If this new term limit × state ideology interaction is statistically significant, then this would imply that an unmeasured state attribute that is associated with the implementation of term limits might be influencing a state’s incarceration rate rather than term limits; thus, we would have a spurious finding. However, if the new term limit × state ideology interaction is nonsignificant, this would strengthen the claim that term limits are influencing our findings. The results are shown in Model 4. The interaction is not statistically significant. This placebo test casts doubt on the notion that some other state factor that is correlated with the implementation of term limits is driving our findings.

It should be noted that we have also estimated a model in which the dependent variable involves correctional expenditures with data from the US Census. The results are presented in Appendix B (Table B1). We find little evidence that term limits have an effect with this measure. We suggest that this pattern emerges due to the increase in the use of private facilities in the United States.Footnote 102 Private facilities are not reflected in the expenditure data.

Discussion and Conclusion

A Politico article highlighting Oklahoma’s incarceration policy asserts that “Oklahoma had the highest incarceration rate in the United States. If it were a country, it would have led the world. That year there were 1,079 people incarcerated in Oklahoma facilities, including jails and state and federal prisons, per every 100,000 people in the state.”Footnote 103 Although in recent years the incarceration rate in Oklahoma has slightly decreased relative to its earlier peak,Footnote 104 several indicators suggest that the imprisonment rate is starting to increase again in more recent years.Footnote 105 What political factors might have incentivized the state to implement policies that would result in this incarceration rate? Our findings suggest that legislative term limits and the policy preferences of the state government are two components that help to explain the high number of inmates serving in the Oklahoma’s correctional facilities.

In this article, we develop the theoretical argument that term limits should increase the influence that state ideology has on the incarceration rate. Because the lawmakers have less independent policy expertise and knowledge because of term limits, the legislators are forced to rely more heavily on the chamber’s leadership. As chamber leaders wish to improve their party’s reputation with the voters via a unified front, this should result in the state enacting more policies, including criminal justice policies, that are consistent with their ideology. When analyzing state incarceration rates between 1979 and 2017, we find evidence to support our theoretical expectation. Specifically, for term-limited states, we find that an increase in conservatism is associated with a higher incarceration rate. Conversely, we find that the government’s ideology has little effect for states without term limits. Our findings are fairly robust to alternative model and measurement specifications.

Before concluding, it is important to discuss an important limitation to our study. Our results do not indicate which specific state policies influence the size of a state’s prison population. It is certainly possible that term limits cause states to implement certain criminal justice policies more frequently that influence the incarceration rate. We encourage future scholars to analyze which specific criminal justice policies are enacted by term-limited states and their effects on the incarceration rate in greater detail.

These findings deepen our insight into the criminal justice system and political institutions in four ways. First, the United States tends to have a larger prison population relative to that of other countries, with significant variation among the states.Footnote 106 The high incarceration rate in the United States has resulted in an increased interest from scholars, activists, and the general public.Footnote 107 In particular, critics of such policies assert that the criminal justice policies are not cost effective and raise important ethical considerations. Our findings suggest that advocates of criminal justice reform may struggle to advance their cause in conservative states with term limits. Instead, they may find more success in conservative states without term limits.

Second, scholars have noted a connection between punitive criminal justice policies and electoral considerations.Footnote 108 In particular, electoral factors may incentivize lawmakers to pursue more punitive policies that might result in a higher incarceration rate. Our findings suggest that electoral connections that are severely diminished via legislative term limits result in more policies being implemented that result in a higher incarceration rate in conservative states. However, such diminished connections would result in a decrease in the inmate population for more-liberal states.

Third, our research question addresses the potential issue of potential unequal justice across the US states. It is possible that individuals are receiving more punitive rulings, in part, simply because of their state of residence. In other words, an individual who is convicted of a certain offense in Louisiana may face a different penalty than might an individual in Washington with nearly identical circumstances.Footnote 109 In fact, there is some historical research to support this claim.Footnote 110 Our findings suggest, depending on the ideological context of the state government, that term limits may be exacerbating this issue.

Finally, scholars of legislative term limits have devoted considerable attention to the implications of term limits. Regarding the effects on policy, most scholarly attention has been devoted to the consequences of fiscal policy. Considerably less research focuses on the effects for noneconomic policies. Our findings help to fill this void. We find that term limits do influence noneconomic policy domains. Specifically, term limits influence criminal justice policy outcomes. We encourage future scholars to analyze the effects in other noneconomic policy domains and to determine whether term limits have a polarizing effect in those domains as well. For instance, it is unclear whether term limits might influence education policies or cultural issues. We encourage future scholars to analyze these and other noneconomic policy domains in greater detail.

We also believe that these findings pave the way for new areas of research. Scholars of term limits have found considerable variation in turnover among state legislatures.Footnote 111 We believe that the effect of the term-limit policy is not due to only the turnover rate. For instance, we believe that lawmakers anticipate the turnover that will occur due to the term-limit law and that this partially motivates their behavior. We encourage future scholars to more precisely capture the mechanisms that are driving these findings. Taken together, there is still a great deal to learn about the implications of term limits on public policy.

Supplementary material

The supplementary material for this article can be found at http://doi.org/10.1017/S0898030624000058.

Acknowledgments

An earlier version of this manuscript was presented at the 2024 Policy History Conference at Arizona State University. We would like to thank Matthew Hitt, Susanne Schwarz, Logan Strother, Jordan Butcher, Eric Moore, Wendi Pollock, Deborah Sibila, the anonymous reviewers, and the Editors of the Journal of Policy History for their valuable comments and feedback.

References

Notes

1 Lindsey Millar, “Leaning into Mass Incarceration in Arkansas,” Arkansas Times, January 27, 2023, https://arktimes.com/news/2023/01/27/leaning-into-mass-incarceration-in-arkansas.

2 Millar, “Leaning into Mass Incarceration.”

3 See, James Wootton, “Truth in Sentencing-Why States Should Make Violent Criminals Do Their Time,” University of Dayton Law Review, 20 (1994): 779.

4 Hunter Field, “Arkansas Sees Uptick in Incarceration as Crime Drops Slightly,” Arkansas Advocate, January 3, 2024, https://arkansasadvocate.com/2024/01/03/arkansas-sees-uptick-in-incarceration-even-as-crime-drops-slightly/.

5 Millar, “Leaning into Mass Incarceration.”

6 “Massachusetts Profile,” Prison Policy Initiative, 2021, https://www.prisonpolicy.org/profiles/MA.html.

7 Katie Lannan, “Baker Signs Justice Bill Despite ‘Serious Concerns,’” WGBH, April 13, 2018, https://www.wgbh.org/news/politics/2018-04-13/baker-signs-justice-bill-despite-serious-concerns.

8 Will Brownsberger, “Final Criminal Justice Policy Released,” Will Brownsberger-State Senator, March 23, 2018, https://willbrownsberger.com/final-criminal-justice-package-released/#summary.

9 Thomas D. Stucky, Karen Heimer, and Joseph B. Lang, “A Bigger Piece of the Pie? State Corrections Spending and the Politics of Social Order,” Journal of Research in Crime and Delinquency 44, no. 1 (2007): 91–123.

10 Sean Nicholson-Crotty, David A. M. Peterson, and Mark D. Ramirez, “Dynamic Representation (s): Federal Criminal Justice Policy and an Alternative Dimension of Public Mood,” Political Behavior 31 (2009): 629–55.

11 Kevin B. Smith, “The Politics of Punishment: Evaluating Political Explanations of Incarceration Rates,” The Journal of Politics 66, no. 3 (2004): 925–38.

12 “How Many States Have Term Limits on Their Legislatures?” U.S. Term Limits, June 8, 2018, https://www.termlimits.com/state-legislative-term-limits/.

13 Jonathan Day and Keith Boeckelman, “The Impact of Legislative Term Limits on State Debt: Increased Spending, Flat Revenue,” Politics & Policy 40, no. 2 (2012): 320–38.

14 Leah Wang, “Updated Charts Show the Magnitude of Prison and Jail Racial Disparities, Pretrial Populations, Correctional Control, and More,” Prison Policy Initiative, April 1, 2024, https://www.prisonpolicy.org/blog/2024/04/01/updated-charts/.

15 Robert E. Rubin and Nicholas Turner, “The Steep Cost of America’s High Incarceration Rate,” Wall Street Journal, December 25, 2014, https://www.wsj.com/articles/robert-rubin-and-nicholas-turner-the-steep-cost-of-americas-high-incarceration-rate-1419543410.

16 For example, the title of our article alludes to this phenomenon. See, George Jones, “Still Doin’ Time,” track 1 on Still the Same Ole Me, Epic (1981).

17 Robert A. Dahl, Democracy and Its Critics (New Haven, CT: Yale University Press, 2008).

18 Brandice Canes‐Wrone and Kenneth W. Shotts, “The Conditional Nature of Presidential Responsiveness to Public Opinion,” American Journal of Political Science 48, no. 4 (2004): 690–706; Jeffrey R. Lax and Justin H. Phillips, “The Democratic Deficit in the States,” American Journal of Political Science 56, no. 1 (2012): 148–66; Warren E. Miller and Donald E. Stokes, “Constituency Influence in Congress,” American Political Science Review 57, no. 1 (1963): 45–56.

19 Joseph Dillon Davey, The Politics of Prison Expansion: Winning Elections by Waging War on Crime (Boston: Greenwood Publishing Group, 1998).

20 Brandice Canes‐Wrone, William Minozzi, and Jessica Bonney Reveley, “Issue Accountability and the Mass Public,” Legislative Studies Quarterly 36, no. 1 (2011): 5–35.

21 Mark D. Ramirez, “Punitive Sentiment,” Criminology 51, no. 2 (2013): 329–64.

22 Thomas D. Stucky, Karen Heimer, and Joseph B. Lang, “Partisan Politics, Electoral Competition and Imprisonment: An Analysis of States over Time,” Criminology 43, no. 1 (2005): 211–48.

23 Stucky, Heimer, and Lang, “Partisan Politics.”

24 Jeff Cummins, “Issue Voting and Crime in Gubernatorial Elections,” Social Science Quarterly 90, no. 3 (2009): 632–51.

25 Thad Kousser, Term Limits and the Dismantling of State Legislative Professionalism (Cambridge: Cambridge University Press, 2005).

26 Daniel C. Lewis, “Legislative Term Limits and Fiscal Policy Performance,” Legislative Studies Quarterly 37, no. 3 (2012): 305–28.

27 Jeff Cummins, “The Effects of Legislative Term Limits on State Fiscal Conditions,” American Politics Research 41, no. 3 (2013): 417–42.

28 Abbie H. Erler, “Legislative Term Limits and State Spending,” Public Choice 133 (2007): 479–94.

29 Lax and Phillips, “The Democratic Deficit in the States.”

30 Lax and Phillips, “The Democratic Deficit in the States,” 159.

31 Susan M. Miller, Jill Nicholson-Crotty, and Sean Nicholson-Crotty, “The Consequences of Legislative Term Limits for Policy Diffusion,” Political Research Quarterly 71, no. 3 (2018): 573–85.

32 John M. Carey, Richard G. Niemi, Lynda W. Powell, and Gary F. Moncrief, “The Effects of Term Limits on State Legislatures: A New Survey of the 50 States,” Legislative Studies Quarterly 31, no. 1 (2006): 105–34.

33 Ashley Lopez, “Term Limits for Congress Are Wildly Popular. But Most Experts say they’d Be a Bad Idea,” NPR, October 29, 2023, https://www.npr.org/2023/10/29/1207593168/congressional-term-limits-explainer.

34 Carey et al., “The Effects of Term Limits.”

35 Robynn Kuhlmann and Daniel C. Lewis, “Legislative Term Limits and Voter Turnout,” State Politics & Policy Quarterly 17, no. 4 (2017): 372–92.

36 Seth E. Masket and Jeffrey B. Lewis, “A Return to Normalcy? Revisiting the Effects of Term Limits on Competitiveness and Spending in California Assembly Elections,” State Politics & Policy Quarterly 7, no. 1 (2007): 20–38.

37 George A. Krause, David E. Lewis, and James W. Douglas, “Politics Can Limit Policy Opportunism in Fiscal Institutions: Evidence from Official General Fund Revenue Forecasts in the American States,” Journal of Policy Analysis and Management 32, no. 2 (2013): 271–95; Lewis, “Legislative Term Limits and Fiscal Policy Performance”; Miller, Crotty, and Crotty, “Consequences of Legislative Term Limits.”

38 Peverill Squire, “Historical Evolution of Legislatures in the United States,” Annual Review of Political Science 9 (2006): 19–44.

39 Edward J. Lopez, “Term limits: Causes and Consequences,” Public Choice 114, no. 1 (2003): 1–56.

40 Jeffrey A. Karp, “The Influence of Elite Endorsements in Initiative Campaigns,” in Citizens as Legislators, ed. Shaun Bowler, Todd Donovan, and Caroline J. Tolbert (Columbus: Ohio State University Press, 1998), 149–65.

41 Marjorie Sarbaugh-Thompson, “Measuring “Term Limitedness” in US Multi-state Research,” State Politics & Policy Quarterly 10, no. 2 (2010): 199–217.

42 Sarbaugh-Thompson, “Measuring Term Limitedness.”

43 Kousser, Term Limits, 9.

44 James MacPherson, “North Dakota Voters OK Term Limits for Governor, Legislators,” AP News, November 8, 2022, https://apnews.com/article/bismarck-north-dakota-term-limits-government-and-politics-ebba348663d0a8e715f9b90714602354.

45 Kristen L. Swigart, Anuradha Anantharaman, Jason A. Williamson, and Alicia A. Grandey, “Working While Liberal/Conservative: A Review of Political Ideology in Organizations,” Journal of Management 46, no. 6 (2020): 1063–91.

46 Jeff Yates and Richard Fording, “Politics and State Punitiveness in Black and White,” The Journal of Politics 67, no. 4 (2005): 1099–1121.

47 Michael C. Campbell and Heather Schoenfeld, “The Transformation of America’s Penal Order: A Historicized Political Sociology of Punishment,” American Journal of Sociology 118, no. 5 (2013): 1375–1423; David Jacobs and Jason T. Carmichael, “The Politics of Punishment across Time and Space: A Pooled Time-Series Analysis of Imprisonment Rates,” Social Forces 80, no. 1 (2001): 61–89; Vesla M. Weaver, “Frontlash: Race and the Development of Punitive Crime Policy,” Studies in American Political Development 21, no. 2 (2007): 230–65.

48 Smith, “The Politics of Punishment”; Jon Sorensen and Don Stemen, “The Effect of State Sentencing Policies on Incarceration Rates,” Crime & Delinquency 48, no. 3 (2002): 456–75; Stucky, Heimer, and Lang, “Partisan Politics.”

49 Andrew Karch and Matthew Cravens, “Rapid Diffusion and Policy Reform: The Adoption and Modification of Three Strikes Laws,” State Politics & Policy Quarterly 14, no. 4 (2014): 461–91.

50 Andrew Lucas, Blaize Davies, and Alissa Pollitz Worden, “State Politics and the Right to Counsel: A Comparative Analysis,” Law & Society Review 43, no. 1 (2009): 187–220.

51 Daniel J. Mallinson and A. Lee Hannah, “Policy and Political Learning: The Development of Medical Marijuana Policies in the States,” Publius: The Journal of Federalism 50, no. 3 (2020): 344–69.

52 Todd Makse and Craig Volden, “The Role of Policy Attributes in the Diffusion of Innovations,” The Journal of Politics 73, no. 1 (2011): 108–24.

53 Rick Farmer and Thomas H. Little, “Legislative Power in the Buckeye State: The Revenge of Term Limits,” in Legislating without Experience: Case Studies in State Legislative Term Limits, ed. Rick Farmer, Christopher Z. Mooney, Richard J. Powell, and John C. Green (Lanham, MD: Lexington Books, 2007), 43–54; Michael P. Olson and Jon C. Rogowski, “Legislative Term Limits and Polarization,” The Journal of Politics 82, no. 2 (2020): 572–86; Laine P. Shay, “Do Term Limits ‘Limit’ the Speaker? Examining the Effects of Legislative Term Limits on State Speaker Power,” State Politics & Policy Quarterly 21, no. 2 (2021): 139–64.

54 Jennifer Drage Bowser and Gary Moncrief, “Term Limits in State Legislatures,” in Institutional Change in American Politics: The Case of Term Limits, ed. Karl Kurtz, Bruce Cain, and Richard Niemi (Ann Arbor: University of Michigan Press, 2007), 10–21; Moncrief and Thompson, “On the Outside Looking In.”

55 John W. Kingdon, Congressmen’s Voting Decisions (Ann Arbor: University of Michigan Press, 1998).

56 Shay, “Do Term Limits ‘Limit’ the Speaker?”

57 Christopher Z. Mooney, “Explaining Legislative Leadership Influence: Simple Collective Action or Conditional Explanations?” Political Research Quarterly 66, no. 3 (2013): 559–71.

58 Moncrief and Thompson, “On the Outside Looking In.”

59 Richard F. Fenno, Home Style: House Members in Their Districts (New York: Harper Collins, 1978).

60 Seth Masket and Boris Shor, “Polarization without Parties: Term Limits and Legislative Partisanship in Nebraska’s Unicameral Legislature,” State Politics & Policy Quarterly 15, no. 1 (2015): 67–90.

61 Gary W. Cox, Thad Kousser, and Mathew D. McCubbins, “Party Power or Preferences? Quasi-Experimental Evidence from American State Legislatures,” The Journal of Politics 72, no. 3 (2010): 799–811.

62 Gary W. Cox and Mathew D. McCubbins, Setting the Agenda: Responsible Party Government in the U.S. House of Representatives (Cambridge: Cambridge University Press, 2005).

63 Michelle Alexander, The New Jim Crow: Mass Incarceration in the Age of Colorblindness (New York: New Press, 2012); Jacob M. Grumbach, “From Backwaters to Major Policymakers: Policy Polarization in the States, 1970–2014,” Perspectives on Politics 16, no. 2 (2018): 416–35; Weaver, “Frontlash.”

64 Donald E. Stokes, “Spatial Models of Party Competition,” American Political Science Review 57, no. 2 (1963): 368–77.

65 E. Ann Carson, “Imprisonment Rate of Sentenced Prisoners under the Jurisdiction of State or Federal Correctional Authorities per 100,000 U.S. Residents, December 31, 1978–2019,” Bureau of Justice Statistics, October 14, 2020.

66 We have also estimated the coefficients with a robust regression model. The results are similar to those presented here.

67 William D. Berry, Evan J. Ringquist, Richard C. Fording, and Russell L. Hanson, “Measuring Citizen and Government Ideology in the American States, 1960-93,” American Journal of Political Science 42, no. 1 (1998): 327–48.

68 Keith T. Poole, “Recovering a Basic Space from a Set of Issue Scales,” American Journal of Political Science 42, no. 3 (1998): 954–93.

69 William D. Verry, Richard C. Fording, Evan J. Ringquist, Russell L. Hanson, and Carl Klarner, “A New Measure of State Government Ideology, and Evidence That Both the New Measure and an Old Measure are Valid,” State Politics & Policy Quarterly 13, no. 2 (2013): 164–82.

70 Travis J. Baker, Travis J. Hedge, and David M. Hedge, “Term Limits and Legislative‐Executive Conflict in the American States,” Legislative Studies Quarterly 38, no. 2 (2013): 237–58.

71 It should be noted that the Baker and Hedge scores are derived from scores created by Marjorie Sarbaugh-Thompson. We rely more heavily on the Baker and Hedge scores because some of the term-limit scores created by Sarbaugh-Thompson have negative values—suggesting that those term-limited states have lower turnover than non-term-limited states. Theoretically this is possible, but data from the Book of the States suggests that this tends not to be the case.

72 Oregon briefly had term limits implemented, but the legal system later invalidated the law. We have reestimated Model 1, however, with the inclusion of a term limit score for Oregon. The results are similar to those presented here.

73 Miller, Crotty, and Crotty, “Consequences of Legislative Term Limits.”

74 Matthew Grossmann, Michael P. Jordan, and Joshua McCrain, “The Correlates of State Policy and the Structure of State Panel Data,” State Politics & Policy Quarterly 21, no. 4 (2021): 430–50.

75 Peter K. Enns, “The Public’s Increasing Punitiveness and Its Influence on Mass Incarceration in the United States,” American Journal of Political Science 58, no. 4 (2014): 857–72; Mark D. Ramirez, “Punitive Sentiment,” Criminology 51, no. 2 (2013): 329–64.

76 Again, NOMINATE scores are used to construct these ideology scores. For more details, see Berry et al., “Measuring Citizen and Government Ideology.”

77 Olson and Rogowski, “Legislative Term Limits.”

78 Peverill Squire, “Measuring State Legislative Professionalism: The Squire Index Revisited,” State Politics & Policy Quarterly 7, no. 2 (2007): 211–27.

79 Squire, “Measuring State Legislative Professionalism.”

80 Peverill Squire, “A Squire Index Update,” State Politics & Policy Quarterly 17, no. 4 (2017): 361–71.

81 Virginia Gray and David Lowery, “Interest Group Politics and Economic Growth in the U.S. States,” The American Political Science Review 82, no. 1 (1988): 109–31; David Lowery, Virginia Gray, and John Cluverius, “Temporal Change in the Density of State Interest Communities 1980 to 2007,” State Politics & Policy Quarterly 15, no. 2 (2015): 263–86.

82 Daniel C. Lewis, Saundra K. Schneider, and William G. Jacoby, “Institutional Characteristics and State Policy Priorities: The Impact of Legislatures and Governors,” State Politics & Policy Quarterly 15, no. 4 (2015): 447–75.

83 Smith, “The Politics of Punishment.”

84 Nathan J. Kelly and Christopher Witko, “Government Ideology and Unemployment in the U.S. States,” State Politics & Policy Quarterly, 14, no. 4 (2014): 389–413.

85 It is important to mention that we use single variable imputation to address missing data in our data set. We also updated a few of the variables in the data set that we used to run the analysis.

86 Jason Sorens, Fait Muedini, and William P. Ruger, “US State and Local Public Policies in 2006: A New Database,” State Politics & Policy Quarterly 8, no. 3 (2008): 309–26.

87 Sorens, Muedini, and Ruger, “US State and Local Public Policies in 2006.”

88 Augustine J. Kposowa, Kevin D. Breault, and Beatrice M. Harrison, “Reassessing the Structural Covariates of Violent and Property Crimes in the USA: A County Level Analysis,” British Journal of Sociology 46, no. 1 (1995): 79–105.

89 Iowa Community Indicators Program, Urban Percentage of the Population for States, Historical (U.S. Census Bureau) [data set], Iowa State University, 2023, https://www.icip.iastate.edu/tables/population/urban-pct-states.

90 University of Kentucky Center for Poverty Research, UKCPR National Welfare Data, 1980-2017, Lexington, KY, http://ukcpr.org/resources/national-welfare-data.

91 University of Kentucky Center for Poverty Research, UKCPR National Welfare Data.

92 University of Kentucky Center for Poverty Research, UKCPR National Welfare Data.

93 We rescale the income per capita and state economy variable to make them easier to interpret. These transformations have no effect on our findings.

94 Sean Nicholson-Crotty and Kenneth J. Meier, “Crime and Punishment: The Politics of Federal Criminal Justice Sanctions,” Political Research Quarterly 56, no. 2 (2003): 119–26.

95 Nathaniel Beck and Jonathan N. Katz, “Modeling Dynamics in Time-Series–Cross-Section Political Economy Data,” Annual Review of Political Science 14 (2011): 331–52.

96 Some caution should be drawn for the unemployment rate variable. When estimating coefficients for a robust regression model, this variable becomes statistically nonsignificant.

97 Boris Shor and Nolan McCarty, “The Ideological Mapping of American Legislatures,” American Political Science Review 105, no. 3 (2011): 530–51.

98 We have also analyzed the residuals of our model. When analyzing estimates for a robust regression estimator, our results do not substantively change.

99 Sarbaugh-Thompson, “Measuring ‘Term Limitedness.’”

100 It should be noted that Travis Baker and David Hedge’s term-limit scores are constructed with Marjorie Sarbaugh-Thompson’s measure of term limits (see Sarbaugh-Thompson, “Measuring ‘Term Limitedness.’”

101 Miller, Crotty, and Crotty, “Consequences of Legislative Term Limits,” 578.

102 Anna Gunderson, “Why Do States Privatize Their Prisons? The Unintended Consequences of Inmate Litigation,” Perspectives on Politics 20, no. 1 (2022): 187–204.

103 Taylor Miller Thomas and Megan McCrink, “How Oklahoma Popped Its Prison Bubble, in Charts,” Politico, April 23, 2020, https://www.politico.com/interactives/2020/justice-reform-decarceration-in-oklahoma/.

104 Thomas and McCrink, “How Oklahoma Popped Its Prison Bubble.”

105 Taylor Mitchell, “After Years of Decline, Oklahoma Prison Population on the Rise: Advocates Point Out Overcrowding Issues,” KFOR-TV, December 5, 2023, https://kfor.com/news/local/after-years-of-decline-oklahoma-prison-population-on-the-rise-advocates-point-out-overcrowding-issues/.

106 John Gramlich, “America’s Incarceration Rate Falls to Lowest Level since 1995,” Pew Research Center, May 2, 2018, https://www.pewresearch.org/fact-tank/2021/08/16/americas-incarceration-rate-lowest-since-1995/.

107 Charis Kubrin and Carroll Seron, “The Prospects and Perils of Ending Mass Incarceration in the United States,” The ANNALS of the American Academy of Political and Social Science 664, no. 1 (2016): 16–24.

108 Canes‐Wrone, Minozzi, and Reveley, ““Issue Accountability and the Mass Public”; Cummins, “Issue Voting and Crime in Gubernatorial Elections.”

109 Darienne Gutierrez, “Punishment Rate Measures Prison Use Relative to Crime,” The Pew Charitable Trust, March 23, 2016, https://www.pewtrusts.org/en/research-and-analysis/issue-briefs/2016/03/the-punishment-rate.

110 Herbert Koppel, “Sentencing Practices in 13 States,” U.S. Department of Justice: Bureau of Justice Statistics, 1985, https://bjs.ojp.gov/content/pub/pdf/sp13s.pdf.

111 Jordan Butcher, “Be Careful What You Count: Updating Legislative Turnover in the 50 States,” American Politics Research 50, no. 4 (2022): 503–10.

Figure 0

Table 1. State Incarceration Rates, 1979–2017

Figure 1

Figure 1. Predicted Incarceration Rates. This figure is derived from the estimates presented in Model 1. The solid line represents states without term limits. The dashed line represents the legislative term limit set at its maximum value. Higher values of government liberalism correspond with greater government liberalism. All control variables are held constant. Shaded areas represent 95% confidence band.

Figure 2

Table 2. Additional Alternative Model Specifications

Supplementary material: File

Byers and Shay supplementary material

Byers and Shay supplementary material
Download Byers and Shay supplementary material(File)
File 18.3 KB