Content uploaded by Hayley Michelle Finetti
Author content
All content in this area was uploaded by Hayley Michelle Finetti on Sep 24, 2020
Content may be subject to copyright.
1
The Impact of Ex-Felon Disenfranchisement on Voting Behavior
Hayley Finetti and John Ramirez
California State University, Chico
Dr. Diana Dwyre
September 24, 2020
2
Abstract
The 2000 presidential election between Vice President Al Gore and then-Governor George W. Bush was
undeniably close, especially in Florida. Ultimately, Bush was declared the winner of the Florida popular
vote, resulting in his receiving enough electoral ballots to win the presidency. Scholars have debated
whether the outcome might have changed, had Florida ex-felons been able to vote. Further, could ex-felon
voting rights have impacted other elections? In this study, we aimed to determine the impact of ex-felon
re-enfranchisement on the 2000 and 2016 Florida presidential elections. We also investigated the potential
effects on the 2020 Florida presidential election, should ex-felons be able to exercise their right to vote.
Citing other studies on ex-felon voting behavior, we predicted ex-felon voter turnout and vote choice for
all three elections, analyzing the influence of race, class/income, and misinformation on voting behavior.
We found ex-felons are ill-informed of their voting rights after incarceration, reducing their likelihood of
voting. In addition, the ex-felon population is disproportionately black and lower income, factors which
would contribute to greater support for Democratic candidates. As such, we found if Florida ex-felons’
voting rights were restored, their rate of turnout would increase if properly informed of their voting rights,
and they would be more likely to vote for the Democratic candidate. Due to the narrow 2000 Florida
presidential race, we determined that if ex-felons had voted, Gore would have earned enough additional
votes to win the election. In contrast, predicted rates of ex-felon voter turnout in 2016 would likely have
failed to change the election outcome; we predict this same result for the 2020 Florida presidential
election. We conclude that in close Florida presidential races, ex-felon re-enfranchisement can change
election outcomes in favor of the Democratic candidate, especially when ex-felons are properly informed
of their voting rights.
3
Introduction:
Despite the United States Constitution’s several amendments addressing voting rights, no
federal laws address or protect voting rights for incarcerated and formerly incarcerated
persons (Sherman 2018). States decide to what extent felons and ex-felons have the right to
vote, and state partisanship has led to differing state laws (Sherman 2018). Scholars have
found that felon disenfranchisement disproportionately affects non-white, low income
incarcerated individuals (Uggen and Manza 2002; King and Erickson 2016; Klumpp, Mialon,
and Williams 2017; Gelman, Kenworthy, and Su 2010).
How would the restoration of Florida ex-felons’ right to vote impact voter turnout
and number of vote shares for major political parties in Florida presidential elections? In
this study, we examine different approaches to voting eligibility for ex-felons. We predict
restoring voting rights for Florida ex-felons would increase voter turnout and vote shares
for Democratic candidates, because felons are a significant portion of the American
population and are more likely to be non-white and low income, both groups more likely
to vote for Democratic candidates. We will test this hypothesis by analyzing the existence
of a relationship between our independent variable, ex-felon re-enfranchisement, and our
dependent variables, voter turnout and vote shares for Democratic candidates. We will
measure our independent variable by examining how restoring ex-felons’ right to vote in
Florida would have or will influence the outcomes of Florida presidential elections in three
election years (2000, 2016, and 2020) and will construct data sets analyzing the influence
that ex-felon re-enfranchisement in Florida would have had or might have on voter turnout
and vote shares for Democratic candidates in each of these elections.
4
Literature Review
The US prison population disproportionately includes non-white, low income individuals.
Prison Policy Initiative (PPI) used data from the Bureau of Justice Statistics and found that, in
2015, the median income of felons in the US aged 27 to 42 was $19,185 (PPI 2015). On the
other hand, the median personal income of US non-felons among the same age group was
$32,505 (PPI 2015), nearly 41% higher than the median income among felons during the same
year. In 2017, over 60% of incarcerated persons were people of color, and 33.1% were black,
in spite of the fact that blacks make up only 15% of the total US population (The Sentencing
Project 2017, 5). Scholars have illustrated the influence of income and race on voting
behavior, some specifically evaluating the impact of these on incarcerated and formerly
incarcerated persons. In our study, we will evaluate survey, interview, and experimental data
demonstrating the influence of race, income, and misinformation on ex- felon voting behavior.
Race
Researchers have found that because the American prison population is primarily non-white
and that non-whites are more likely to identify with the Democratic Party than whites, felon
disenfranchisement reduces vote shares for Democratic candidates in elections. For example,
Miles analyzed self-reported data from black and white participants of the 1986 to 2000
November Voter Supplements to the U.S. Census Bureau’s Current Population Survey
(CPS) and matched this information to felon racial demographics to estimate that
incarcerated blacks, who make up a majority of the prison population, would be more likely
than whites to vote for Democratic candidates if their voting rights were restored (Miles
2004, 93-99). Similarly, Klumpp, Mialon, and Williams evaluated US states that have
restored voting rights for ex-felons and compared election results prior to and following the
5
reinstatement of their right to vote (Klumpp, Mialon, and Williams 2017, 41). They found
that in re-enfranchised states, black male voter turnout had increased by up to 6.7% in states
fully restoring ex-felons’ right to vote, and these voters were more likely to vote for
Democratic candidates in elections (Ibid., 41-49).
In 2007, 150,000 Florida ex-felons had their voting rights restored, until 2011 when
their rights were reversed (Meredith and Morse 2018). Through analysis of public records and
election data, Meredith and Morse found that, between 2007 and 2011, an estimated 28% of
these re- enfranchised ex-felons, 16% black and 12% nonblack, exercised their right to vote. In
the same survey, of the participants who reported party preference, 87% of black ex-felons
(over 7,500) and 34% of nonblack ex-felons (over 3,500) revealed support for the Democratic
Party (Ibid.).
These findings support our hypothesis that restoring ex-prisoners’ right to vote increases
vote shares for Democratic candidates. However, Klumpp, Mialon, and Williams (2017) and
Miles (2004) expressed that although their studies indicate a positive correlation between
reinstating voting rights for black ex-prisoners and vote shares for Democratic candidates, the
results are insignificant because prisoners often belong to sociodemographic groups that are
less likely to vote in general, due to factors such as race, income, and education level.
Meredith and Morse (2018) provide us a more detailed prediction of how ex-felons in Florida
might vote, if given the opportunity. We will utilize these data to predict voter turnout and
choice among black and nonblack ex-felons, which will also allow us to observe claims by
Klumpp, Mialon, and Williams (2017) and Miles (2004) that ex-felons vote at lower rates than
the general population. Miles (2004) based a portion of his evidence on self-reported data,
which is vulnerable to false reporting. While our data will also rely on survey data, we will
6
minimize the impact of false reporting by incorporating public record data from The
Sentencing Project and Meredith and Morse to supplement survey data (The Sentence Project
2016, 17-18; Meredith and Morse 2018).
Class / Income
In the United States, income and social class have significant influence on the likelihood of
voting and vote choice, because income inequality influences whether citizens have access to
resources necessary to vote and affects the economic policies they support. Gelman,
Kenworthy, and Su examined the differences in voting behavior between lower, middle, and
high income voters between 1940 and 2008 and compiled various data sets revealing higher
income persons were most likely to turn out to vote and to vote Republican; lower income
constituents were least likely to vote and more likely to support Democratic candidates; and
middle income individuals had higher voter turnout and were more likely to vote Republican
than lower income voters, but less likely to do either than higher income persons (Gelman,
Kenworthy, and Su 2010, 1205- 1214). National election exit poll data conducted from 1994
to 2018 suggests these patterns have remained generally consistent, though higher income
voters were only moderately more likely to vote Republican in the 2016 and 2018 national
elections (New York Times 2018).
Scholars examining the effect of felon disenfranchisement on elections have utilized
similar research to quantify the effect of income and class on voter turnout and choice of
incarcerated persons. For example, Uggen and Manza estimated ex-felon voting behavior by
compiling data from the Current Population Survey (CPS) and National Election Studies
(NES) to compare voting behavior of non-felons of various incomes to felons with similar
income levels (Uggen and Manza 2002, 784-787). They found that because a higher
7
percentage of felons are low income than the general population, they would be less likely to
vote in elections and more likely to vote for Democratic candidates (Ibid., 791-799). Likewise,
King and Erickson created data sets comparing felon disenfranchisement laws by state and
broke down sociodemographics of the general and prison population in each state to find that,
in particular, African-American low-income voters are more likely to become incarcerated
and, thereby, disenfranchised (King and Erickson 2016, 807-809).
In May 2019, the Brennan Center for Justice (BCJ) used Florida census data and the
state voter file to estimate the sociodemographics of ex-felons who had registered to vote
since the implementation of Amendment 4, which restored the voting rights of about 1.4
million Florida ex-felons in January 2019 (BCJ 2019, 3). The organization found that newly
eligible ex-felons who had registered to vote in the first quarter of 2019 had income that was
about $15,000 less than the total Florida population average income.
These data support our hypothesis that felon disenfranchisement reduces voter turnout
and Democratic votes in elections. However, Uggen and Manza (2002) incorporated self-
reported data into their research that may complicate the results due to possible false reporting.
We will control for this issue in our study by including data less likely to be skewed due to
false reporting, such as data by The Sentencing Project documenting rates of felon
disenfranchisement in Florida and data collected by Brennan Center calculating the average
income of re- enfranchised registered Floridians as compared to the Florida voting population
(The Sentencing Project 2016; BCJ 2019).
Voting Limitations
Scholars have found that felon disenfranchisement often continues even after voting rights for
ex-felons are restored, because officials do not ensure these previously incarcerated
8
individuals are informed of their voting rights. Meredith and Morse conducted research
analyzing the effects of implementation of Executive Order 42 in Iowa in 2005, which
automatically granted ex-felons their voting rights without the need to file a petition (Meredith
and Morse 2015, 45). Although this law was followed by an increased voter turnout among
ex-felons in 2006, this increase in voter turnout was minimal (Meredith and Morse 2015, 59-
61).
Meredith and Morse found that one factor contributing to low ex-felon voter turnout
was the lack of information provided to ex-felons regarding their automatic right to vote
(Meredith and Morse 2015, 61-71). The scholars asserted that, if provided access to additional
resources and information describing the restoration of their voting rights, ex-felons might be
more likely to exercise those rights (Meredith and Morse 2015, 77-78). The authors observed
that after being notified of their right to vote, voter turnout among ex-felons increased by 8%
in the 2008 election, 3.7% in 2010, and 5.3% in 2012 (Meredith and Morse 2015, 71).
Prior to Meredith and Morse (2015), Drucker and Barreras also analyzed how
misinformation about felon enfranchisement laws impact felons and ex-felons and influence
their likelihood of voting (Drucker and Barreras 2005). Drucker and Barreras conducted a
survey of 334 felons and ex-felons in New York, Connecticut, and Ohio and found that over
half of the participants were unaware or misinformed of their right to vote (Drucker and
Barreras 2005, 4). Manza and Uggen also researched how misinformation about voting rights
restoration affects prisoners and ex-prisoners (Manza and Uggen 2006, 200). The researchers
interviewed 33 Minnesota prisoners and ex-felons on probation or parole and found that most
incarcerated persons interviewed believed their voting rights would never be restored (Manza
and Uggen 2006, 200). Contrary to this belief, Minnesota ex-felons are immediately re-
9
enfranchised after the full completion of their sentence, highlighting the lack of information
provided to ex-felons regarding their rights after incarceration (NCSL 2019).
Meredith and Morse (2015), Drucker and Barreras (2005), and Manza and Uggen
(2006) agree that granting ex-felons the right to vote is not enough to ensure these rights are
being exercised. Many incarcerated and formerly incarcerated persons are unaware of when
and if their voting rights will be restored, which affects their likelihood of voting; as such, the
researchers acknowledge access to information describing re-enfranchisement laws in their
state and how to exercise these rights would be a helpful resource to support ex-felons in
exercising their voting rights (Meredith and Morse 2005, 223-241; Drucker and Barreras
2005, 10-11; and Manza and Uggen 2006, 225).
Their research supports our hypothesis that felon re-enfranchisement would lead to an
increase in voter turnout and, further, Meredith and Morse (2005) and Manza and Uggen
(2006) support our hypothesis that voter restoration for ex-felons would lead to an increase in
vote shares for Democratic candidates. However, their analyses call into question whether
voter turnout would be significantly improved because factors such as lack of access to
information affect the probability of voting. For our analysis, we will estimate ex-felon voter
turnout after notification of voting rights in Florida in the 2000, 2016, and 2020 presidential
elections, using data collected by Meredith and Morse (2015) to support our hypothesis.
Data and Methodology
By predicting the voting behavior of ex-felons in Florida in the 2000, 2016, and 2020
elections, we aim to demonstrate that the restoration of ex-felon voter rights contributes to
increased voter turnout and vote shares for Democratic candidates in Florida presidential
elections. We will test our hypothesis by utilizing qualitative and quantitative data by
10
Meredith and Morse (2018), Burch (2010), Meredith and Morse (2015), and The Sentencing
Project (2016) to estimate voter turnout and choice among black and non-black ex-felons,
with and without notification of their voting rights, in Florida in the 2000, 2016, and 2020
presidential elections.
In 2016, The Sentencing Project conducted a study in which researchers tabulated total
felon disenfranchisement by state, also calculating the number of disenfranchised persons in
each state that are felons (in prison) or ex-felons (on probation, parole, or post-sentence), and
the number of disenfranchised persons in each state who are black. Using the 2016 Sentencing
Project study, we will estimate the number of black and non-black ex-felons during the 2016
and 2020 election years. For total voter turnout in the 2000 and 2016 elections, we extracted
data from the US Election Atlas (US Election Atlas 2019).
For 2020 Florida presidential election projections, we utilized data from the Office of
Economic and Demographic Research to calculate the projected voting age population in
2020 (Office of Economic and Demographic Research 2015, 1). Then, by using 2016
Florida presidential election voter turnout data in Florida, we predict there will be a 75%
total voter turnout in 2020 (Florida Department of State 2020). Due to the onset of the
COVID-19 pandemic affecting the number of voters who arrive at the polls, we recognize
voter turnout may be lower this year. On the other hand, the current extreme partisanship
may motivate more constituents to vote this year in spite of - or even because of - the
pandemic.
Further, we estimated the number of vote shares per candidate in 2020 by first
analyzing data from a statewide poll conducted by St. Pete Polls from April 16, 2020, to April
17, 2020, in which thousands of registered Florida voters reported whether they are planning
11
to vote for President Donald Trump or Vice President Joe Biden this November, or if they are
“Undecided” (St. Pete Polls 2020, 6). Out of the 5,659 total survey respondents, 47.5%
favored Trump, 48.3% favored Biden, and 4.2% were undecided (St. Pete Polls 2020, 6). We
then will estimate the total voting age population in Florida in 2020 and apply the previous
data points to determine statewide total turnout.
Unfortunately, because 4.2% responded that they were undecided, we are unable to
predict how these respondents will vote in the 2020 Florida presidential election. Additionally,
because these are projections based on poll data, we do not know for sure that those who said
they would vote for President Trump or Vice President Joe Biden will do so come Election
Day. Further, we only used one poll to predict 2020 Florida general election results; however,
this was a relatively decently sized sample of 5,659 respondents, and it contained the most
recent data available regarding projected vote choice in the 2020 presidential election. As
such, we are confident we have utilized a reliable source for estimating these data. Then, using
previous data from The Sentencing Project (2016) and Meredith and Morse (2018), we
projected voter turnout and choice among Florida’s newly enfranchised 1.4 million ex-felons
who in January 2019 were granted voting eligibility under Amendment 4. Although there are
over 1.5 million ex-felons in Florida, those with murder or sexual offenses are exempt from
eligibility under Amendment 4 (Florida Department of State 2018, 10-11).
For our study, we will apply rates derived from Meredith and Morse (2018) to estimate
that voter turnout among black and nonblack ex-felons would have been or will be 16% and
12%, respectively, for all three Florida election years. We recognize these voter turnout
estimates may not be entirely accurate, especially because ex-felons might vote at a higher rate
this year than expected due to the passage of Amendment 4. On the other hand, as previous
12
studies have shown, until notification laws are implemented and enforced in Florida, ex-felons
might vote at a lower rate than expected. This is especially probable because Florida Governor
Ron DeSantis has attempted to limit the impact of Amendment 4 since it passed in November
2018 (Mower 2019). However, these Florida voter turnout predictions were the most recent
and reliable we could find, so we decided they would be the best available data to use for our
study.
Further, we will predict 87% of black ex-felons and 34% of white ex-felons would
vote for the Democratic candidate, and that about 3% of black ex-felons and 40% of white ex-
felons would vote for the Republican candidate. With these calculations we added the total
additional ex-felon Democrat and Republican votes to each respective candidate to predict the
difference in election results.
Meredith and Morse (2015) indicate that voter turnout in Iowa increased by an
average of 5.67% after notifying ex-felons of their right to vote. We will apply these
estimates to the 2000, 2016, and 2020 Florida presidential elections to observe the
difference in voter turnout and number of vote shares per candidate with and without
notification of voting eligibility.
We will examine our data by comparing voter turnout rates and percentage of shares
for Democratic and Republican candidates in the 2000, 2016, and 2020 Florida presidential
elections. We will subdivide these data in tables and bar graphs based on the correlation
between turnout and choice with race and income. We will use those data to investigate how
the Florida 2000 and 2016 presidential elections might have been impacted if ex-felons were
able to vote in those elections. Further, we will predict the potential impact of ex-felon re-
enfranchisement on the upcoming 2020 presidential election in Florida.
13
Findings
In November 2018, voters in Florida voted in favor of Amendment 4, which expanded voting
rights for ex-felons by granting them voting eligibility upon the full completion of their
sentence (Florida Department of State 2018, 10-11). Prior to the November 2018
congressional election, Florida experienced the highest rates of felon disenfranchisement in
the United States, with 10.1% of its population ineligible to vote due to various levels of
incarceration (Sentencing Project, 2016). Nationally, Florida’s disenfranchised population
made up 27% of the national disenfranchised population, while Florida’s population
constitutes only about 6.5% (21.48 million (2019)) of the total US population (Sentencing
Project, 2016; United States Census Bureau 2020).
2000 Presidential Election - Without Notification
In the 2000 presidential election, President Bush carried a narrow win over Vice President Al
Gore; although Gore earned over 500,000 more votes than Bush, he lost in the Electoral
College, falling five electors short of taking the election (US Election Atlas 2019). In Florida,
Bush earned at least 2,912,790 vote shares and Gore earned at least 2,912,253 votes, the
former winning the popular vote by a mere 537 votes (US Election Atlas 2019). It is, of
course, important to note that the actual number of votes each candidate earned was never
determined, because the Supreme Court ordered Florida to stop recounting votes before the
state had finished (FactCheck.org 2008). If Democrats had taken Florida in 2000, Gore
instead, would have earned its 25 electors, and ultimately would have won the election (US
Election Atlas 2019). Under Florida’s pre-2018 felon disenfranchisement law, over 600,000
ex- felons were barred from voting in the 2000 election (Meredith and Morse 2018).
In Figure 1, applying estimates of voter turnout of ex-felons in Florida between 2007
14
and 2011 calculated by Meredith and Morse (2018), we show that if Florida’s over 600,000
ex- felons had had the right to vote in 2000 and if voter turnout among ex-felons was 28%,
over 34,560 black ex-felons and more than 46,080 nonblack ex-felons would have voted (see
Figure 1). Further, if we apply reports of party preference from the authors’ survey, more than
30,067 (87%) black ex-felons and over 15,667 (34%) nonblack ex-felons would have voted
for a Democratic candidate, while about 1,037 (~3%) black ex-felons and over 18,432 (40%)
nonblack ex-felons would have voted for a Republican candidate (see Figure 2).
In Figure 3, we have calculated the total number of vote shares each candidate would
have earned in 2000 under these circumstances, and indicated the winning candidate.
Assuming these formerly incarcerated persons would have voted for the major presidential
candidates, Gore would have earned at least an additional 45,734 votes and Bush would have
gained over 19,469 more votes. By adding 45,734 to Gore’s actual vote shares in 2000
(2,912,253), we calculate that, with ex-felon Democratic voters, Gore would have earned a
total of approximately 2,957,987 votes. Using the same calculations for Republican vote
shares, by adding 19,469 votes to Bush’s actual number of vote shares in Florida in 2000
(2,912,790), we estimate that, with ex-felon Republican voters, Bush would have accumulated
a total of about 2,957,987 votes. By these estimates, Gore would have won the state of Florida
by 25,728 votes, thereby earning enough electors to win the election (see Figure 3).
2016 Florida Presidential Election - Without Notification
In the Florida 2016 presidential election, Donald Trump surpassed Hillary Clinton by 112,911
votes, earning 4,617,886 votes to Clinton’s 4,504,975. In Figure 1 we applied estimates of ex-
felon disenfranchisement in Florida in 2016 (The Sentencing Project 2016, 17-18) and used
data by Meredith and Morse (2018) to estimate voter turnout and choice of black and
15
nonblack ex- felons. We show that, if Florida’s 443,811 black ex-felons and 1,135,130
nonblack ex-felons could have voted in 2016, 71,010 black ex-felons and 136,216 nonblack
ex-felons would have voted (see Figure 1).
Illustrated in Figure 2, using estimates of party preference by Meredith and Morse
(2018), we find that 61,779 (87%) black ex-felons and 46,313 (34%) nonblack ex-felons
would have voted for Clinton, a 108,092 increase in vote shares. By contrast, we find that
about 2,130 (~3%) black ex-felons and 54,486 (40%) nonblack ex-felons would have voted
for Trump, an additional 56,616 votes (see Figure 2). By adding these vote shares to the
actual total vote shares for each candidate, we estimate that, if Florida ex-felons had had the
right to vote in 2016, Clinton would have earned 4,613,067 votes and Trump would have
earned 4,674,502 votes. In this instance, Trump still would have won the election by 61,435
votes (see Figure 3).
2020 Florida Presidential Election - Without Notification
We have utilized the same estimated ex-felon voter turnout and vote choice percentages to
develop projections for how the implementation of Amendment 4, which would restore the
right to vote for about 1.4 million previously disenfranchised Florida ex-felons, will impact
the 2020 presidential election in Florida. To do so, we applied the same data that we used to
make ex- felon voting behavior predictions in the 2000 and 2016 presidential election to
estimate how eligible ex-felons might influence the 2020 Florida presidential election.
Because The Sentencing Project documents the most recent data on the population
of disenfranchised persons by state, we have used their 2016 study to estimate the
percentage of black and nonblack ex-felons who will be eligible to vote in the 2020
election, under Amendment 4. In 2016, approximately 28% of ex-felons in Florida were
16
black and 72% were nonblack (The Sentencing Project 2016, 17-18). If these same rates
are true in 2020, of the approximately 1.4 million newly eligible ex-felon voters, about
392,000 are black and about 1,008,000 are nonblack.
Calculating estimated turnout among these ex-felons, we predict that about 62,720
(16%) eligible black ex-felons and 120,960 (12%) eligible nonblack ex-felons will vote in the
2020 presidential election (see Figure 1). Further, we estimate 54,566 (87%) eligible black ex-
felons and 41,126 (34%) eligible nonblack ex-felons will vote for Biden, the Democratic
candidate, and that about 1,882 (~3%) eligible black ex-felons and 48,384 (40%) eligible
nonblack ex-felons will vote for Trump, the Republican candidate. In total, we estimate 95,692
Democrat vote shares and 50,266 Republican vote shares among 2020 eligible ex-felons in
Florida (see Figure 2).
From there, we estimated the total voting age population (VAP) in Florida in 2020 to
be 16,952,432, using 2020 Florida population predictions projected by the University of
Florida, Bureau of Economic and Business Research, and Florida Population Studies at the
2015 Florida Demographic Estimating Conference (Office of Economic and Demographic
Research 2015, 1). We applied the predicted 75% voter turnout to this projected VAP to
estimate that about 12,714,324 Floridians will vote in 2020. Of these 12.7 million voters,
using poll data from St. Pete Polls, we estimate that 6,039,304 Florida constituents will vote for
Trump and that 6,141,018 citizens will vote for Biden. Because 4.2% of respondents were undecided,
we were unable to determine whether these constituents would have voted for Trump or Biden, or if
they would have voted at all.
In Figure 3, we show the number of vote shares each candidate would earn in this
scenario, and which candidate would win the election. Adding the number of estimated
Florida ex-felon Democratic vote shares (95,692) to Biden’s total non-felon projected votes
17
(6,141,018), we predict that, with ex-felon voters, Biden will earn 6,236,710 votes in the
2020 Florida presidential election. Following the same calculations for Trump’s votes, we
estimate that, with ex-felon voters, Trump will gain 6,089,570 votes. Although 2020
projections already predict Biden will win in Florida, by subtracting the number of expected
ex-felon Republican votes (50,266) by the number of expected ex-felon Democrat votes
(95,692), we can estimate that the ex-felon voting population will increase his lead by
45,426 votes (see Figure 3).
Figure 1
Figure 2
18
Figure 3
2000 Florida Presidential Election - With Notification
In Figures 4, 5, and 6, we estimate how the implementation of notification laws in Florida
may have affected voter turnout in each of the elections we previously analyzed. Using data
from Uggen and Manza for correlation notification laws and voter turnout, we estimate a
5.67% voter turnout increase in Florida in the 2000, 2016, and 2020 presidential elections, if
notification laws had been enforced (Uggen and Manza 2002).
In 2000, an increased voter turnout of 5.67% revealed a total of about 85,209 estimated
ex-felons, 4,629 more than without notification, would have voted. Applying previous
estimates by Meredith and Morse (2018), about 36,205 black ex-felons - an additional 1,645 -
and 49,004 nonblack ex-felons - an increase of about 2,924 - would have voted in comparison
to estimated turnout without notification (see Figure 4). We predict about 31,498 (87%) of the
black ex-felons and about 16,661 (34%) of the nonblack ex-felons would have voted for the
Democrat candidate, for a total of about 48,159 additional votes. Further, we estimate that
about 1,086 (3%) black ex- felons and about 19,602 (40%) nonblack ex-felons would have
voted for Bush, a total of 20,688 votes (see Figure 5).
In Figure 6, we summarize the estimated total vote shares for Gore and Bush, including
noting who would have won the election in this scenario. By adding the total number of
19
estimated ex-felon Democratic votes to Vice President Gore’s original 2,912,253 votes, we
estimate Gore would have earned 2,960,412 votes if ex-felons could have voted and if
notification laws were in place; conversely, by adding the total number of predicted ex-felon
Republican vote shares, 20,688 votes, to 2,912,790, his actual number of earned votes in 2000,
we calculate Bush would have earned 2,981,100 votes, if ex-felons could vote in 2000 and if
these ex-felons had been notified of their right to vote. In this scenario, Gore would have won
the Florida election by about 26,934 votes, 1,206 more votes than if ex-felons had not been
notified (see Figure 6).
2016 Florida Presidential Election - With Notification
In 2016, predicting an increase of 5.67% voter turnout among Florida ex-felons if notified of
their right to vote, we estimate that about 74,300 black ex-felons and 144,676 nonblack ex-
felons would have voted in the presidential election, for a total of 218,976 ex-felon voters, an
increase of 11,750 votes without notification (see Figures 4 and 1). Of these, we predict
64,641 (87%) black ex-felons and 49,190 (34%) nonblack ex-felons would have voted for
Clinton, a total of 113,831 votes. Further, we estimate about 2,229 (~3%) black ex-felons and
57,879 (40%) nonblack ex-felons would have voted for Trump, adding up to 60,099
Republican vote shares (see Figure 5).
In Figure 6, we have tabulated the total estimated number of votes for each presidential
candidate, adding the estimated ex-felon votes. By adding the estimated ex-felon Democratic
votes to Secretary Clinton’s actual total votes, we find that she would have earned 4,618,805
votes, equating to 5,738 more votes than if these ex-felons had not been notified.
Adding predicted ex-felon Republican vote shares to his actual total vote shares, we predict
Trump would have accumulated about 4,677,985 votes, 3,483 more votes than when ex-felons
20
were not informed of their right to vote. In this instance, President Trump would have won the
2016 presidential election in Florida by about 59,180 votes, 2,255 less than when ex-felons
were not notified of their voting rights (see Figure 6).
2020 Florida Presidential Election - With Notification
To project election results for the 2020 presidential election, we matched our previous
estimate of increased turnout among ex-felons after notification of their right to vote,
predicting that informing ex-felons of their voting eligibility would have increased voter
turnout by 5.67%. With this increase in ex-felon voter turnout, we estimate that about
65,636 black ex-felons and 128,459 nonblack ex-felons will vote in Florida, for a total of
194,095 ex-felon voters (see Figure 4). Moreover, we predict 57,103 (87%) black ex-felons
and 43,676 (34%) nonblack ex- felons would have voted for Biden, for a total of 100,779
votes. Further, we predict about 1,969 (~3%) black ex-felons and 51,384 (40%) nonblack
ex-felons would have voted Republican, adding up to 53,353 votes in favor of Trump (see
Figure 5).
In Figure 6, we show the total number of estimated vote shares for each candidate,
adding predicted ex-felon vote shares. We added Biden’s projected total vote shares in the
2020 election (6,141,018) to our calculated predictions of the number of ex-felons who will
vote Democratic (100,779) to arrive at an estimated total of 6,241,797 votes for Biden. This
total is 5,087 more ex-felon votes for Biden than when ex-felons were not informed of the
restoration of their voting rights. By adding Trump’s projected total number of votes for the
upcoming election (6,039,304) to our estimates of ex-felon voter turnout after notifying ex-
felons of their 5right to vote (53,353), we predict a total of 6,092,657 votes for Trump in
Florida in the 2020 presidential election. This estimate reveals an increase of 3,087 ex-felon
21
Republican Party votes than in the scenario in which ex-felons had not been informed of their
voting rights prior to voting in the 2020 election (see Figure 6).
Figure 4
Figure 5
Figure 6
Impact of Income on Voting Behavior
Table 1 shows Roper Center (2020) income data from the 2000 national election results,
22
demonstrating that low income voters were more likely to vote for Vice President Gore and
high income voters more often voted for President Bush (see Table 1). Further, Table 2
displays income data from the 2016 national election results (CNN 2020), reporting that low
to low- middle income voters were more likely to vote for Secretary Clinton, while high-
middle to high income voters more often voted for President Trump (see Table 2).
Table 1: Impact of Income on Vote Choice (2000)*
* Table created using income data from Roper Center (2020)
Table 2: Impact of Income on Vote Choice (2016)*
* Table created using income data from CNN (2020)
We examined additional research with the goal of predicting how re-enfranchised Floridian ex-
felons would vote based on their income, but found that discrepancies between the datasets made it
difficult to estimate ex-felon vote choice for the Florida presidential elections in 2000, 2016 and
2020. As can be seen in Table 1 and Table 2, the income brackets differ from each other, and also
23
differ from the income brackets available in national polling (Brookings Institute 2018), and also
different within the available state polling of all Floridians of voting age (U.S. Census
Bureau 2000; U.S. Census Bureau 2016).
Figure 7 depicts detailed breakdowns of ex-felon income in a national study;
however, we were unable to find similar data for Florida only (Brookings Institute 2018). We
also were only able to find detailed breakdowns of ex-felon income in a national study. In
Figure 8, data on Floridian ex-felon income indicates that the average income for ex-felons is
approximately $15,000 less than the total Floridian voting population; however, a detailed
breakdown of this datapoint was not provided (Brennan Center for Justice 2019). Given the
discrepancies between the income brackets and the very limited information about ex-felon
income and vote choice, it would have been difficult to apply the data with confidence. As
such, we chose to not investigate this further. This is an area that would benefit from
additional research.
Figure 7*
*Figure created using income data from Brookings Institute (2018)
24
Figure 8*
*Figure created using income data from Brennan Center for Justice (2018)
Conclusion
In this study we examined the impact of the restoration of Florida ex-felons’ voting rights on
turnout and the number of votes expected for major political parties in Florida presidential
elections. Based on the gathered information, we determined that restoring ex-felon voting
rights would be unlikely to shift the political outcome of most presidential elections. That
said, in a very tight election, such as the 2000 Presidential Election in Florida, a strong
turnout of re-enfranchised ex-felons could change the election outcome. Using predictions by
Meredith and Morse (2018), we predict an estimated 28% voter turnout among Florida ex-
felons, supporting our hypothesis that restoring ex-felons’ right to vote in Florida would
increase voter turnout.
Felon disenfranchisement disproportionately affects minority, low-income populations.
Non-whites comprise a higher percentage of the prison population in the United States. In
addition, research has shown that the views and perspectives of non-whites are more likely to
align with those of the Democratic Party (Klumpp, Mialon, and Williams 2017, 41-49). As
such, the disenfranchisement of ex-felons reduces vote shares for Democratic candidates
(Miles 2004, 93-99). This assertion is supported by Meredith and Morse, who found that when
voting rights were restored to Floridian ex-felons from 2007 to 2011, 87% of black ex-felons
25
and 34% of non- black ex-felons revealed their support for the Democratic Party (Meredith
and Morse 2018).
We found that, of the over 600,000 ex-felons in Florida in 2000, an estimated 36%
were black and 64% were non-black. In 2016, an estimated 28% of ex-felons were black and
72% were nonblack. In 2020, we project similar proportions. Considering blacks constitute
only 15% of the total American population, these findings support our assertion that felon
disenfranchisement and incarceration disproportionately affects blacks. Further, because our
data estimates show black ex-felons in Florida would be much more likely to vote Democratic
(87%) than Republican (~3%), and that black ex-felons would be much less likely to vote
Republican (34%) than their nonblack non-felon counterpart (63.5%), this supports our
hypothesis that restoring voting rights for ex-felons in Florida would increase number of vote
shares for Democratic candidates.
Class and income are also a significant indicator of voting behavior. Gelman,
Kenworthy, and Su found that people with higher socioeconomic status were more likely to be
politically involved, and were more likely to vote for Republican candidates. On the other
hand, communities with lower socioeconomic status were more likely to identify with
Democratic candidates, but they were less likely to participate in the political process. Those
with a middle income were more likely to vote, and were also more likely to support
Republican candidates as compared to lower income individuals (Gelman, Kenworthy, and Su
2010, 1205-1214). Election result data in 2000 and 2016 by Roper Center and CNN support
our hypothesis that high income voters are more likely to vote, and to vote for a Republican
candidate, and low income voters are not as likely to vote, but more likely to vote for a
Democratic candidate when they do.
26
As indicated previously, there is evidence that ex-felon income correlates with their
vote choice; however, the datasets we found utilized different income brackets between
election years, as well as between national and state data. As such, we chose not to make
predictions regarding how re-enfranchised ex-felons in Florida would vote according to their
income because we knew we would not be able to apply the data with confidence. There
would be great benefit in having more detailed data, so we recommend future investigation in
this area.
We also found that misinformation and lack of notification about restored voting rights
influences ex-felon voter turnout. Meredith and Morse estimated an average of 5.67% increase
in voter turnout among Iowa ex-felons after being notified of their restored voting eligibility
(Meredith and Morse 2015, 61). Drucker and Barreras found that, of the 334 felons and ex-
felons surveyed across New York, Connecticut, and Ohio, over half were not notified or were
misinformed of their right to vote (Drucker and Barreras 2005, 4). Further, Manza and Uggen
found that, of the 33 Minnesota felons and ex-felons interviewed, most incarcerated
interviewees never thought their voting eligibility would be restored (Manza and Uggen 2006,
200).
Applying predictions by Meredith and Morse that notification would lead to a
5.67% increase in voter turnout, we found that, in 2000, an additional 4,629 ex-felons
would have voted; in 2016, an additional 11,750 ex-felons would have voted; and, in
2020, we project an additional 10,415 ex-felons will vote (see Figure 4). These findings
show that notification laws have a small impact, but in a particularly close election, such
as the 2000 Florida presidential election, notification could be the difference between a
win and a loss.
27
We estimate that, in the 2000 Florida presidential election, both with and without
notification, Vice President Gore would have earned enough additional votes to win the
presidential election, reversing the outcome in favor of the Democratic Party. In the 2016
presidential election, by our predictions of ex-felon voter turnout and vote shares, Clinton
would not have earned enough additional votes to win the election, with or without
notification. In the 2020 presidential election, our projections estimate President Trump will
not earn enough additional votes to win the election, with or without notification. With these
data, we can predict that, when election results are very close and in favor of the Republican
candidate, ex-felon voters could influence election results enough to flip the election in favor
of the Democratic candidate. Following the 2020 Florida presidential election, we will revisit
the projected outcome and compare our predictions with the actual election results.
Due to lack of access to Florida ex-felon income data and lack of access to current
predicted rates of ex-felon turnout and choice in Florida, our data is limited. This is partially
due to the fact that exit polls and other election data sources do not usually ask if the
respondent has been incarcerated. Also, most predictions of ex-felon voter turnout and choice
are based on national data, and differences in population and sociodemographics across states
make using these national estimates to calculate a state’s voter turnout and choice
problematic. We suggest future research estimate ex-felon voter turnout and choice by state.
We further recommend scholars explore the income levels of ex-felons in each US state and
estimate how income level impacts ex-felon voting behavior in each state. We additionally
propose scholars research the correlation between ex-felon race and income, and voter
turnout and choice, by estimating the influence an ex-felon’s race and income, rather than
just their race or just their income, has on their voting behavior.
28
References
Akee, R. 2019. “Voting and Income.” EconoFact. Retrieved from https://econofact.org/voting-
and-income (accessed on March 3, 2020).
Burch, Traci. 2012. “Did Disfranchisement Laws Help Elect President Bush? New Evidence
on the Turnout Rates and Candidate Preferences of Florida's Ex-Felons.” Political
Behavior 34, 1–26.
Bureau of Economic and Business Research. 2016. Florida Demographic Estimating
Conference, University of Florida. “Florida Population by Age Group.”
http://edr.state.fl.us/Content/population-demographics/data/Pop_Census_Day.pdf
(accessed on April 7, 2020).
CNN. n.d. “Exit Polls 2016.” Retrieved from https://www.cnn.com/election/2016/results/exit-
polls/florida/president (accessed on April 3, 2020).
Dave Leip. 2019. “2000 Presidential General Election Results - Florida.” Dave Leips Atlas of
U.S. Presidential Elections.
https://uselectionatlas.org/RESULTS/state.php?year=2000&fips=12 (accessed on April
7, 2020).
Dave Leip. 2019. “2016 Presidential General Election Results.” Dave Leips Atlas of U.S.
Presidential Elections.
https://uselectionatlas.org/RESULTS/state.php?year=2016&fips=12&off=0&elect=0&
f=0 (accessed on April 7, 2020).
ElectProject. 2020. “Voter Turnout Demographics.” Retrieved
from http://www.electproject.org/home/voter-
29
turnout/demographics (accessed on March 14, 2020).
Florell, Matt. 2020. “Florida Statewide survey conducted for FloridaPolitics.com.”
Floridapolitics.com. Retrieved from
http://stpetepolls.org/files/StPetePolls_2020_State_President_April17_G84JW.p
df (accessed on April 18, 2020).
Florida Department of Corrections, Georgia Department of Corrections, and North Carolina
Department of Corrections. 2020. “Did Disfranchisement Laws Help Elect President
Bush? New Evidence on the Turnout Rates and Candidate Preferences of Florida’s Ex-
Felons.” Retrieved from https://link.springer.com/article/10.1007/s11109-010-9150-
9/figures/3 (accessed on April 7, 2020).
Florida Department of State. 2018. “Voter Turnout.” Retrieved from
https://dos.myflorida.com/elections/data-statistics/elections-data/voter-turnout/
(accessed on April 22, 2020).
Florida Department of State. 2020. “Voter Turnout.” Retrieved from
https://dos.myflorida.com/elections/data-statistics/elections-data/voter-turnout/
(accessed on April 22, 2020).
Florida Department of State. n.d. “Proposed Constitutional Amendments and Revisions
for the 2018 General Election.” Retrieved from
https://dos.myflorida.com/media/699824/constitutional-amendments-2018-
general-election-english.pdf (accessed on April 22, 2020).
Gelman, Andrew, Lane Kenworthy, and Yu Sung Su. 2010. “Income Inequality and Partisan
Voting in the United States.” Special Issue: Inequality and Poverty: American and
30
International Perspectives 91(5): 1203-1219
Jackson, Brooks. 2008. “The Florida Recount of 2000.” Politifact. Retrieved from
https://www.factcheck.org/2008/01/the-florida-recount-of-2000/ (accessed on September
08, 2020).
King, Bridgett and Laura Erickson. 2016. “Disenfranchising the Enfranchised: Exploring the
Relationship Between Felony Disenfranchisement and African American Voter Turnout.”
Journal of Black Studies 47(8): 799–821.
Klumpp, Tilman, Hugo M. Mialon, and Michael A. Williams. 2017. “The Voting Rights of
Ex-Felons and Election Outcomes in the United States.” International Review of Law
and Economics 59: 40-56.
Lai, Rebecca K.K. and Alison McCann. 2018. “Exit Polls: How Voting Blocs Have Shifted
From the “80’s to Now.” New York Times. Retrieved from
https://www.nytimes.com/interactive/2018/11/07/us/elections/house-exit-polls-
analysis.html?mtrref=undefined&gwh=38290A854BBEF2941E5F0879537C9227&gwt=
pay&assetType=REGIWALL (accessed on April 14, 2020).
Leip, David. 2019. “2000 Presidential General Election Results - Florida.” Dave Leip's
Atlas of U.S. Elections, LLC. Retrieved from
https://uselectionatlas.org/RESULTS/state.php?year=2000&fips=12. (accessed
April 20, 2020).
Leip, David. 2019. “2016 Presidential General Election Results 2016 Presidential General
Election Results .” Dave Leip's Atlas of U.S. Elections, LLC. Retrieved from
https://uselectionatlas.org/RESULTS/state.php?year=2016&fips=12&off=0&elect=0&
31
f=0 (accessed on April 20, 2020).
Lewis, Nicole, Aviva Shen, and Anna Flagg. 2020. “What Do We Really Know About the
Politics of People Behind Bars?” The Marshall Project. Retrieved from
https://www.themarshallproject.org/2020/03/11/what-do-we-really-know-about-the-
politics-of-people-behind-bars (accessed on April 3, 2020).
Meredith, Marc and Michael Morse. 2014. “Do Voting Rights Notification Laws Increase
Ex- Felon Turnout?” Quarterly Journal of Political Science 10: 41–100.
Meredith, Marc and Michael Morse. 2015. “The Politics of the Restoration of Ex-Felon
Voting Rights: The Case of Iowa.” Quarterly Journal of Political Science 10: 41–
100.
Meredith, Marc and Michael Morse. 2018. “Why letting ex-felons vote probably won’t swing
Florida.” Vox. https://www.vox.com/the-big-idea/2018/11/2/18049510/felon-voting-
rights-amendment-4-florida (accessed on April 3, 2020).
Miles J. Thomas. 2004. “Felon Disenfranchisement and Voter Turnout” The Journal of
Legal Studies 33(1): 85-129.
Morris, Kevin. 2019. “Thwarting Amendment 4.” Brennan Center for Justice. Retrieved from
https://www.brennancenter.org/sites/default/files/analysis/2019_05_FloridaAmendme
nt_FINAL-3.pdf (accessed on April 22, 2020).
Mower, Lawrence. 2020. “Limits on Florida's Amendment 4 Are Unconstitutional,
Federal Panel Rules.” Tampa Bay Times. Retrieved from
https://www.tampabay.com/florida-politics/buzz/2020/02/19/limits-on-floridas-
amendment-4-are-unconstitutional-federal-panel-rules/ (accessed on April 22, 2020).
32
Office of Economic & Demographic Research. n.d. “Florida Population by Age Group.”
http://edr.state.fl.us/Content/population-demographics/data/Pop_Census_Day.pdf
(accessed on April 17, 2020).
Purtle, Jonathan. 2013. “Felon Disenfranchisement in the United States: A Health
Equity Perspective.” American Journal of Public Health 103(4): 632-637.
Roper Center. 2020. “How Groups Voted in 2000.” https://ropercenter.cornell.edu/how-
groups-voted-2000 (accessed on April 14, 2020).
Ruth, Terrance, Jonathan Matusitz, and Demi Simi. 2017. “Ethics of Disenfranchisement and
Voting Rights in the U.S.: Convicted Felons, the Homeless, and Immigrants.”
American Journal of Criminal Justice, 42(1), 56-68.
Sentencing Project. 2016. “Disenfranchisement News: How Florida’s felony
disenfranchisement laws impact elections.” Retrieved from
https://www.sentencingproject.org/news/disenfranchisement-news-floridas-felony-
disenfranchisement-laws-impact-elections/ (accessed on March 16, 2020).
Sherman, Amy. 2018. “Understanding felon voting rights restoration.” PolitiFact. Retrieved
from https://www.politifact.com/article/2018/apr/25/understanding-felon-voting-
rights-
restoration/ (accessed on April 2, 2020).
Sherman, Amy. 2019. “Have more Florida felons registered as Republicans since
Amendment 4 took effect?” Politifact.
https://www.politifact.com/factchecks/2019/apr/08/jared-
kushner/have-more-florida-felons-registered-republicans (accessed on April 2, 2020).
33
StPetePolls. n.d. “Florida Statewide Survey Conducted for FloridaPolitics.com.” Retrieved
from http://stpetepolls.org/files/StPetePolls_2020_State_President_April17_G84JW.pdf.
(accessed on April 18, 2020).
Turner, Nicholas, and Adam Looney. 2018. “Work and Opportunity Before and After
Incarceration.” The Brookings Institution. Retrieved on
https://www.brookings.edu/wp-
content/uploads/2018/03/es_20180314_looneyincarceration_final.pdf (accessed on
April 2, 2020).
Uggen, Christopher and Jeff Manza. 2002. “Democratic Contraction? Political Consequences
of Felon Disenfranchisement in the United States.” American Sociological Review
67(6): 777-803.
Uggen, Christopher, Sarah Shannon and Jeff Manza. 2012. “State-Level Estimates of Felon
Disenfranchisement in the United States.” The Sentencing Project. Retrieved from
https://www.sentencingproject.org/wp-content/uploads/2016/01/State-Level-
Estimates-of-Felon-Disenfranchisement-in-the-United-States-2010.pdf (accessed on
March 27, 2020).
Uggen, Christopher, Ryan Larson, and Sarah Shannon. 2016. “6 Million Lost Voters:
State-Level Estimates of Felony Disenfranchisement.” The Sentencing Project.
Retrieved from https://www.sentencingproject.org/publications/6-million-lost-
voters-state-level-
estimates-felony-disenfranchisement-2016/#II.%20Disenfranchisement%20in%202016
(accessed on April 2, 2020).
34
United States Census Bureau. 2002. “Voting and Registration in the Election of November
2000.” Retrieved from https://www.census.gov/prod/2002pubs/p20-542.pdf
(accessed on April 11, 2020).
United States Census Bureau. 2018. “Characteristics of Voters in the Presidential
Election of 2016.” Retrieved from
https://www.census.gov/content/dam/Census/library/publications/2018/demo/P
20-582.pdf (accessed on April 11, 2020).
United States Census Bureau. 2020. “QuickFacts: Florida.” Retrieved from
https://www.census.gov/quickfacts/FL (accessed on April 11, 2020).
