Skip to main content
Log in

Did Disfranchisement Laws Help Elect President Bush? New Evidence on the Turnout Rates and Candidate Preferences of Florida’s Ex-Felons

  • Original Paper
  • Published:
Political Behavior Aims and scope Submit manuscript


This paper re-examines the impact of Florida’s disfranchisement law on the 2000 Presidential election. The analysis simulates outcomes in Florida under scenarios consistent with the turnout rates of Georgia and North Carolina ex-felons in 2000 and Florida ex-felons in 2008. Survey evidence on candidate preferences as well as data on ex-felon party registration in Florida and North Carolina are used to produce estimates of support for Bush and Gore among ex-felons. Based on the simulations, the ex-felon population in Florida would have favored Bush in 2000. Assuming that ex-felons supported Gore at rates similar to GSS respondents with at most a high school diploma, Bush would have defeated Gore by 4,925 and 7,048 votes, assuming turnout of 10 and 15%, respectively.

This is a preview of subscription content, log in via an institution to check access.

Access this article

Subscribe and save

Springer+ Basic
EUR 32.99 /Month
  • Get 10 units per month
  • Download Article/Chapter or Ebook
  • 1 Unit = 1 Article or 1 Chapter
  • Cancel anytime
Subscribe now

Buy Now

Price excludes VAT (USA)
Tax calculation will be finalised during checkout.

Instant access to the full article PDF.

Fig. 1
Fig. 2
Fig. 3
Fig. 4
Fig. 5

Similar content being viewed by others


  1. Although party registration is not partisan identification or vote choice, it is the best information available on the preferences of offenders (Finkel and Scarrow 1985).

  2. Party registration obtained by matching eligible ex-felons to the voter registration files in each state. These data were obtained from the Boards of Elections and Departments of Corrections of each state (see Burch 2007b).

  3. Party registration in 2008 may not accurately reflect what the party registration of Florida’s ex-felons would have been in 2000. For instance, 2008 was a much more Democratic year than 2000 in Florida. Also, the Florida Division of Elections attempted to alert Florida’s eligible ex-felons that they could vote in 2008.

  4. Analyses using offender counts weighted by U. S. life expectancy by race and gender are available upon request.

  5. Information on matching offender records to voter files available upon request.

  6. Please see the appendix for more information on the choice to use the GSS, including the study results calculated using vote choice estimates from other surveys.

  7. Using all southerners, rather than just Floridians, might bias the results in favor of President Bush if Florida voters were more supportive of Gore than those in other southern states, As discussed in the appendix, at least two Florida-specific polls indicate that Florida’s white males supported Gore at rates higher than those in other southern states. Please see the appendix for more details of how Florida-specific survey data differs from survey data of all southerners.

  8. Confining the sample to individuals with low socioeconomic status does not appear to bias the results in favor of Bush. Among southern respondents with a bachelor’s degree in the 2002 GSS, 20.8% of white men and 28.0% of white women report voting for Gore. All but one southern black GSS respondent with a bachelor’s degree voted for Gore in 2000; however, the sample sizes, at five black men and seven black women, are very low. These preferences are similar to those of southern GSS respondents with lower educational attainment shown later in Fig. 4.

  9. Because the original number of Latinos found using the Florida DOC data seemed low, surnames were used to further identify Latinos. These surnames were identified by the Census Bureau as the 639 most common for Hispanics. See Word and Colby Perkins (1996). “Building a Spanish Surname List for the 1990’s—A New Approach to an Old Problem.” U. S. Bureau of the Census Technical Working Paper No. 13. Retrieved January 27, 2010 (

  10. The GSS contains too few Latinos across gender and educational groups for analysis. However, one can get a sense of what Florida’s Latinos would have done from another survey. According to the CBS/NY Times Florida State Poll, conducted in October 2000, 33.2% of Latino males with a high school diploma or less and 28.0% of Latino males with at least some college expressed an intention to vote for Gore in that election. Among Florida Latinas, 25.2% with at most a high school diploma and 39.9% of Hispanic females with at least some college expressed an intention to vote for Gore that November. This poll probably overestimates Gore’s support among voters however, because the CBS/NY Times poll does not account for the decline of Gore support that took place in the month leading up to the election (see Wlezien 2001).

  11. The paper also estimates ex-felon turnout rates in the 2000 general election between 10 and 15% in Michigan and Missouri as well (Burch Forthcoming).

  12. In R, the rtnorm function calls for the standard deviation rather than the variance; as a result, the standard deviation will be referenced throughout the paper even though the function uses the variance to make the random draws.

  13. One could also simulate the vote choice of each ex-felon individually using the binomial distribution. However, simulating 660,000 individual data points for 10,000 hypothetical elections for 8 different scenarios requires vast computational resources.

  14. Simulated outcomes weighted for death rates by race and gender of offenders and outcomes using income as the SES proxy available upon request.

  15. Density plots of these simulated elections can be found in the appendix.


  • Bartels, L. M. (2005). What’s the Matter with What’s the Matter with Kansas? Annual Meeting of the American Political Science Association. Washington, DC.

  • Beck, A. et al. (1993). Survey of State Prison Inmates, 1991. NCJ 136949, United States Department of Justice, Bureau of Justice Statistics.

  • Behrens, A., Uggen, C., et al. (2003). Ballot manipulation and the menace of Negro domination: Racial threat and felon disfranchisement in the United States 1850–2002. American Journal of Sociology, 109(3), 559–605.

    Article  Google Scholar 

  • Biemer, P., et al. (2003). An evaluation of procedures and operations used by the voter news service for the 2000 presidential election. Public Opinion Quarterly, 67, 32–44.

    Google Scholar 

  • Blalock, H. M. (1967). Toward a theory of minority-group relations. New York: Wiley.

    Google Scholar 

  • Brewer, M. D., & Stonecash, J. M. (2001). Class, race issues, and declining white support for the democratic party in the South. Political Behavior, 23(2), 131–155.

    Article  Google Scholar 

  • Brown-Dean, K. L. (2003). Culture, context, and competition: Explaining state-level variation in felon disfranchisement laws. Annual Meeting of the Midwest Political Science Association. Chicago, IL.

  • Bunche, R. J. (1973). The political status of the Negro. Chicago: University of Chicago Press.

    Google Scholar 

  • Burch, T. (2007a). Punishment and participation: How criminal convictions threaten american democracy. Program in Government and Social Policy Ph.D., Harvard University, Cambridge, p. 250.

  • Burch, T. (2007b). A Study of Felon and Misdemeanant Voter Participation in North Carolina. The Sentencing Project. Available online from

  • Burch, T. (Forthcoming). Turnout and party registration among convicted offenders in the 2008 general election. Law and Society Review.

  • Campbell, A. (1960). Surge and decline: A study of electoral change. Public Opinion Quarterly, 24(3), 397–418.

    Article  Google Scholar 

  • Campbell, A., Converse, P., et al. (1960). The American voter. New York: Wiley.

    Google Scholar 

  • Citrin, J., Schickler, E., & Sides, J. (2003). What if everyone voted? Simulating the impact of increased turnout in senate elections. American Journal of Political Science, 47(1), 75–90.

    Article  Google Scholar 

  • Conn, J. B. (2005). Felon disfranchisement laws: Partisan politics in the legislatures. Michigan Journal of Race and Law, 10, 495–539.

    Google Scholar 

  • Dawson, M. C. (1994). Behind the mule. Princeton, NJ: Princeton University Press.

    Google Scholar 

  • DeNardo, James. (1980). Turnout and the vote: The joke’s on the democrats. American Political Science Review, 74(2), 406–420.

    Article  Google Scholar 

  • Diedrich, J. (2006). Signs detailing voting rules for felons removed, returned; wording differs from state message. Milwaukee, Wisconsin: Milwaukee Journal Sentinel.

    Google Scholar 

  • Erikson, R. S. (1995). State turnout and presidential voting. American Politics Quarterly, 23(4), 387–396.

    Article  Google Scholar 

  • Federal Elections 2000: Election Results for the U.S. President, the U.S. Senate and the U.S. House of Representatives (2001) Federal Elections Commission.

  • Finkel, S. E., & Scarrow, H. A. (1985). Party identification and party enrollment: The difference and the consequence. Journal of Politics, 47, 620–647.

    Google Scholar 

  • Fording, R. C. (1997). The conditional effect of violence as a political tactic: Mass insurgency, welfare generosity, and electoral context in American States. American Journal of Political Science, 41(1), 1–29.

    Article  Google Scholar 

  • Fraga, L. R., & Leal, D. L. (2004). Playing the latino card: Race, ethnicity, and national party politics. The DuBois Review, 1(2), 297–317.

    Google Scholar 

  • Frank, T. (2004). What’s the matter with Kansas?. New York: Metropolitan Books.

    Google Scholar 

  • Grofman, B., Owen, G., et al. (1999). Rethinking the partisan effects of higher turnout: So what’s the question? Public Choice, 99, 357–376.

    Article  Google Scholar 

  • Grose, C. R., & Yoshinaka, A. (2002). Electoral Institutions and Voter Participation: The Effect of Felon Disfranchisement Laws on Voter Turnout in the U. S. Southern States, 1984–2000. Annual Meeting of the American Political Science Association. Boston, MA.

  • Harlow, C. W. (2003). Education and correctional populations. Bureau of Justice Statistics.

  • Harvey, A. E. (1994). Ex-Felon disfranchisement and its influence on the black vote: The need for a second look. University of Pennsylvania Law Review, 142(3), 1145–1189.

    Article  Google Scholar 

  • Highton, B., & Wolfinger, R. E. (2001). The political implications of higher turnout. British Journal of Political Science, 31, 179–223.

    Article  Google Scholar 

  • Hillygus, D. S., & Shields, T. G. (2005). Moral issues and voter decision making in the 2004 Presidential Election. PS: 201–209.

  • Holland, P. W. (1986). Statistics and causal inference. Journal of the American Statistical Association, 81(3), 945–960.

    Article  Google Scholar 

  • Hull, E. A. (2006). The disfranchisement of Ex-Felons. Philadelphia: Temple University Press.

    Google Scholar 

  • Key, V. O., Jr. (1949). Southern politics in State and Nation. New York: Knopf.

    Google Scholar 

  • Keyssar, A. (2000). The right to vote. New York: Basic Books.

    Google Scholar 

  • Konner, J., et al. (2001). Television’s Performance on Election Night 2000: A Report for CNN. Available online from Accessed 17 Sep 2010.

  • Knack, S. (1994). Does rain help republicans? Theory and evidence on turnout and the vote. Public Choice, 79, 187–209.

    Article  Google Scholar 

  • Krajick, K. (2004). Why can’t ex-felons vote? Washington, DC: The Washington Post.

    Google Scholar 

  • Kristof, N. D. (2004). Living poor, voting rich. New York: The New York Times.

    Google Scholar 

  • Kurtz, H. (2000). Exit wounds: Polls led networks astray; data, human error tied to Botched calls. The Washington Post, 9 November, A31.

  • Leighley, J. E., & Nagler, J. (2007). Who votes now? And does it matter? Paper prepared for presentation at the 2007 Annual Meeting of the Midwest Political Science Association, Chicago, IL, April 12–15, 2007.

  • Manza, J., & Uggen, C. (2004). Punishment and democracy: Disfranchisement of nonincarcerated felons in the United States. Perspectives on Politics, 2(3), 491–506.

    Article  Google Scholar 

  • Manza, J., & Uggen, C. (2006). Locked out: Felon disfranchisement and American democracy. Cambridge: Oxford.

    Google Scholar 

  • Miles, T. J. (2004). Felon disfranchisement and voter turnout. Journal of Legal Studies, 33(1), 85–129.

    Article  Google Scholar 

  • Nagel, J. H., & McNulty, J. E. (1996). Partisan effects of voter turnout in senatorial and gubernatorial elections. American Political Science Review, 90(4), 780–793.

    Article  Google Scholar 

  • Palast, G. (2001). The wrong way to fix the vote. Washington DC: The Washington Post.

    Google Scholar 

  • Pew Hispanic Center. (2008). Among Hispanics in Florida, 2008 Voter Registration Rolls Swing Democratic. Online; Accessed December 10, 2009, from

  • Pierre, R. E. (2001). Botched name purge denied some the right to vote. Washington, DC: The Washington Post.

    Google Scholar 

  • Pierson, P. (2004). Politics in time. Princeton: Princeton University Press.

    Google Scholar 

  • Piven, F. F., & Cloward, R. A. (2000). Why American still don’t vote. Boston: Beacon Press.

    Google Scholar 

  • Preuhs, R. R. (2001). State Felon disfranchisement policy. Social Science Quarterly, 82(4), 733–748.

    Article  Google Scholar 

  • Rosenstone, S. J., & Wolfinger, R. E. (1978). The effect of registration laws on voter turnout. American Political Science Review, 72(1), 22–45.

    Article  Google Scholar 

  • Shaffer, S. D. (1982). Policy differences between voters and non-voters in American elections. Western Political Quarterly, 35(4), 496–510.

    Article  Google Scholar 

  • Stonecash, J. M. (2000). Class and party in American politics. Boulder: Westview Press.

  • Tucker, H. J., Vedlitz, A., et al. (1986). Does heavy turnout help democrats in presidential elections? American Political Science Review, 80(4), 1291–1304.

    Article  Google Scholar 

  • Wlezien, C. (2001). On forecasting the presidential vote. Political Science and Politics, 34, 25–31.

    Article  Google Scholar 

  • Word, D. L. & Colby Perkins, R. Jr. (1996). Building a Spanish surname list for the 1990’s—A new approach to an old problem. U. S. Bureau of the Census Technical Working Paper No. 13. Retrieved January 27, 2010, from

  • Yoshinaka, A., & Grose, C. R. (2005). Partisan politics and electoral design: The enfranchisement of felons and ex-felons in the United States, 1960–1999. State and Local Government Review, 37(1), 49–60.

    Article  Google Scholar 

Download references


The author would like to thank Jennifer Hochschild, Sidney Verba, Kay Schlozman, Gary King, Dan Galvin, and Alec Ewald for their comments on earlier drafts of this paper.

Author information

Authors and Affiliations


Corresponding author

Correspondence to Traci Burch.

Appendix: Data Selection

Appendix: Data Selection

The analysis in the main body of the paper uses the southern respondents to the GSS to estimate the support for Gore by race, gender, and education. However, many other data sources might have been used to provide these estimates. These include the 2000 American National Election Study, the 2000 CBS/NYTimes Pre-Election Poll of Florida, and the 2000 Voter News Service Florida Exit Poll. Using the GSS best avoids the problems that plague these other surveys: the sample sizes of blacks are large enough for analysis, the reliability of the GSS has not been questioned, and it is a post-election report of actual vote choice.

Each of the Florida-specific samples is deeply flawed. The CBS/NYTimes pre-election poll of Floridians is a pre-election poll. As a result, it contains the opinions of citizens who might have changed their minds or failed to vote. More importantly, however, this survey was fielded too early in the campaign season to account for the drop in Gore support that took place in the month prior to the election (see Wlezien 2001). As a result, this survey probably overestimates the support for Gore in the Florida electorate.

The 2000 Voter News Service Florida Exit Poll also is problematic, particularly because it has been widely discredited. This particular poll is the reason why the networks called Florida for Gore early in 2000. Numerous articles have been produced in both the popular press and in scholarly journals regarding its unreliability (For examples, see Biemer et al. 2003; Konner et al. 2001; Kurtz 2000). In particular, the study is thought to be biased in favor of Gore. Because of this study, the entire methodology of exit polling has changed, and networks remain much more cautious about forecasting victory. Anyone would hesitate to base any analysis on a data source with such a controversial and tainted reputation.

The sample sizes of the southern and Florida ANES respondents are too small when disaggregated by race, gender, and education within many of the cells to calculate reliable vote choice percentages, particularly for blacks. With respect to the 2000 wave of the ANES the sample size for southern black males who completed at most high school was 17; for black females, the sample size was 20. The Hispanic sample size was even smaller: there were only 4 male and 12 female southern Hispanic respondents across the educational distribution for the 2000 ANES. However, 24% of white men, 40% of white women, 94% of black men, and 100% of black women reported voting for the Democratic Presidential candidate in 2000. (These percentages are based on sample sizes of 29 white men, 48 white women, 17 black men, and 20 black women). These estimates are similar to those found using the southern GSS.

To show how the decision to use the southern GSS respondents rather than the other surveys affect the results, Table 5 reports the simulated results of the 2000 election using the estimates for white males and females from these three alternative surveys. (The sample sizes for blacks in these alternative surveys are too small to use; therefore these simulations use the estimates for blacks from the southern GSS). The GSS, ANES, and CBS/NYTimes analyses in both the body of the paper and the appendix rely on the vote for Gore for the combined category of people with a high school diploma or less. The distributions of people with and without a high school diploma in the GSS, ANES, and CBS/NYTimes Poll are skewed heavily in favor of people with a high school diploma but the difference in the proportion voting for Gore between people with and without diplomas in these surveys is small, making the proportion voting for Gore in the combined educational category similar to that found by taking weighted mean of the two categories. However, combining the lowest educational categories in the 2000 VNS Exit Poll in this way produces very different results from what one would get if these educational categories were represented in the Exit Poll sample more equally (as they are among ex-felons in the real world). Because of this issue, the election results for this survey are simulated twice based on two sets of numbers from the 2000 VNS: percent voting for Gore by gender of the combined educational categories and the percent voting for Gore by gender for the separate educational categories weighted for the relative proportion of ex-felons in the population.

Table 5 Simulated results of 2000 election using alternative data sources

The table summarizes for each survey the direction of the bias, the proportion of white males and females voting for Gore, the simulated Bush vote total and the proportion of the simulated elections won by Bush. As the table shows, the results produced using both the 2000 ANES and the 2000 CBS/NYTimes Poll predict a Bush victory, even though the CBS/NYTimes Poll is probably biased in favor of Gore for the reasons noted above. The 2000 VNS Florida Exit Poll, which also is biased in favor of Gore, predicts a Gore victory. The totals based on the combined high school diploma or less category predicts that Gore will win by an average of 2,876 votes; however, Bush still wins 42.6% of the simulated elections even in this best-case-scenario for Gore. The Exit Poll totals based on the weighted mean of the separate educational categories predict that Gore will defeat Bush by 288 votes; Bush wins 48.5% of the simulated elections (Fig. 6).

Fig. 6
figure 6

Density plot of simulated election results assuming 10 and 15% turnout. N = 10,000 simulated elections

Rights and permissions

Reprints and permissions

About this article

Cite this article

Burch, T. Did Disfranchisement Laws Help Elect President Bush? New Evidence on the Turnout Rates and Candidate Preferences of Florida’s Ex-Felons. Polit Behav 34, 1–26 (2012).

Download citation

  • Published:

  • Issue Date:

  • DOI: