514 10TH S TREET NW, S UITE 1000
WASHINGTON, DC 20004
TEL: 202.628.0871 • FAX: 202.628.1091
S TAFF@S ENTENCINGPROJECT.ORG
WWW.S ENTENCINGPROJECT.ORG

A ‘Crazy-Quilt’ of Tiny Pieces:
State and Local Administration of
American Criminal Disenfranchisement Law

By Alec Ewald
Department of Political Science, Union College

November, 2005

This report was written by Alec Ewald, Department of Political Science, Union College.
The Sentencing Project is a national non-profit organization engaged in research and advocacy
on criminal justice policy issues. Funding for this project was made possible by support from the
JEHT Foundation, Open Society Institute, and the Tides Foundation.
The Sentencing Project is a partner in the Right to Vote Campaign, a national campaign to
remove barriers to voting faced by people with felony convictions. Further information can be
found at www.righttovote.org.
Copyright © 2005 by The Sentencing Project. Reproduction of this document in full or part in
print or electronic format only by permission of The Sentencing Project.
For further information:
The Sentencing Project
514 10th St. NW
Suite 1000
Washington, D.C. 20004
(202) 628-0871
www.sentencingproject.org

EXECUTIVE SUMMARY
“A Crazy-Quilt of Tiny Pieces” presents results from the first nationwide study to document the
implementation of American felony disenfranchisement law. Data came from two main sources:
a 33-state survey of state elections officials and telephone interviews with almost one hundred
city, county, town, and parish officials drawn from ten selected states.
The report reaches seven major conclusions:
1. Broad variation and misunderstanding in interpretation and enforcement of voting
laws
•
•

More than one-third (37%) of local officials interviewed in ten states either described
their state’s fundamental eligibility law incorrectly, or stated that they did not know a
central aspect of that law.
Local registrars differ in their knowledge of basic eligibility law, often within the same
state. Differences also emerge in how they are notified of criminal convictions, what
process they use to suspend, cancel, or “purge” voters from the rolls, whether particular
documents are required to restore a voter to eligibility, and whether they have
information about the criminal background of new arrivals to the state.

2. Misdemeanants disenfranchised in at least five states
•
•

•

The commonly-used term “felon disenfranchisement” is not entirely accurate, since at
least five states – Colorado, Illinois, Michigan, South Carolina, and Maryland -- also
formally bar some or all people convicted of misdemeanors from voting.
It is likely that misdemeanants in other states who do retain the formal right to vote could
have difficulty exercising that right, given ignorance of their eligibility and the lack of
clear rules and procedures for absentee voting by people in jail who have not been
convicted of a felony.
Maryland excludes persons convicted of many misdemeanors, such as “Unlawful
operation of vending machines,” “Misrepresentation of tobacco leaf weight,” and
“Racing horse under false name.”

3. Significant ambiguities in voting laws
•
•

Disenfranchisement in Tennessee is dependent on which of five different time periods a
felony conviction occurred between 1973 and the present.
In Oregon, disenfranchisement is determined not by conviction or imprisonment for a
felony, but for being placed under Department of Corrections supervision. Since 1997,
some persons convicted of a felony and sentenced to less than 12 months’ custody have
been sent to county jails and hence, are eligible to vote.

i

4. Disenfranchisement results in contradictory policies within states
•

•
•

The “crazy-quilt” pattern of disenfranchisement laws exists even within states. Alabama
and Mississippi have both the most and least restrictive laws in the country, a result
which is brought about by the fact that certain felonies result in the loss of voting rights
for life, while others at least theoretically permit people in prison to vote.
Most felonies in Alabama result in permanent disenfranchisement, but drug and DUI
offenses have been determined to not involve the “moral turpitude” that triggers the loss
of voting rights.
In Mississippi, ten felonies result in disenfranchisement, but do not include such common
offenses as burglary and drug crimes.

5. Confusing policies lead to the exclusion of legal voters and the inclusion of illegal voters
•

The complexity of state disenfranchisement policies results in frequent misidentification
of voter eligibility, largely because officials differ in their knowledge and application of
disqualification and restoration law and procedures.

6. Significant variation and uncertainty in how states respond to persons with a felony
conviction from other states
•

•

No state has a systematic mechanism in place to address the immigration of persons with
a felony conviction, and there is no consensus among indefinite-disenfranchisement
states on whether the disqualification is properly confined to the state of conviction, or
should be considered in the new state of residence.
Interpretation and enforcement of this part of disenfranchisement law varies not only
across state lines, but also from one county to another within states. Local officials have
no way of knowing about convictions in other states, and many are unsure what they
would do if a would-be voter acknowledged an old conviction. Because there is no
prospect of a national voter roll, this situation will continue even after full HAVA
implementation.

7. Disenfranchisement is a time-consuming, expensive practice
•

Enforcement requires elections officials to gather records from different agencies and
bureaucracies, including state and federal courts, Departments of Corrections, Probation
and Parole, the state Board of Elections, the state police, and other counties’ elections
offices.

ii

POLICY IMPLICATIONS
The overwhelming majority of local officials interviewed for this study were generous with their
time and eager to understand the law, implement it fairly, and facilitate voting by everyone
eligible under state law. Despite this, serious problems persist in the interpretation and
application of disenfranchisement law.
1. Policies disenfranchising people living in the community on probation or parole, or who
have completed a sentence are particularly difficult to enforce
•

States which disenfranchise only persons who are currently incarcerated appear able to
enforce their laws more consistently than those barring non-incarcerated citizens from
voting.

2. Given large-scale misunderstanding of disenfranchisement law, many eligible persons
incorrectly believe they cannot vote, or have been misinformed by election officials
•
•

More than one-third of election officials interviewed incorrectly described their state’s
law on voting eligibility.
More than 85% of the officials who misidentified their state’s law either did not know the
eligibility standard or specified that the law was more restrictive than was actually the
case.

3. Occasional violation of disenfranchisement law by non-incarcerated voters not
surprising
•

Given the complexity of state laws and the number of state officials who lack an
understanding of restoration and disqualification procedures, it should come as no
surprise that many voters are ignorant of their voting status, a fact that is likely to have
resulted in hundreds of persons with a felony conviction registering and voting illegally
in recent years.

4. Taken together, these findings undermine the most prominent rationale for
disenfranchisement: that the policy reflects a strong, clear consensus that persons with
a felony conviction are unfit to vote and constitute a threat to the polity
•
•

•

First, when significant numbers of the people who administer elections do not know
important aspects of disenfranchisement law, it is hard to conclude that the restriction is
necessary to protect social order and the “purity” of the ballot box.
Second, because they are all but invisible in the sentencing process, “collateral” sanctions
like disenfranchisement simply cannot accomplish the denunciatory, expressive purposes
their supporters claim. We now know that disenfranchisement is not entirely “visible”
even to the people running American elections.
Third, deep uncertainty regarding the voting rights of people with felony convictions who
move from one state to another indicates that we do not even know what purpose
disenfranchisement is supposed to serve – whether it is meant to be a punishment, or
simply a non-penal regulation of the franchise.

iii

RECOMMENDATIONS
States should take the following steps – utilizing HAVA resources where possible – in order to
move the enforcement of disenfranchisement law closer to fairness and consistency.
1. Clarify Policies Regarding Out-of-State Convictions
•

State officials should clarify their policies and incorporate into training programs the means
by which a felony conviction in another state affects an applicant’s voting eligibility. For
example, sentence-only disenfranchisement states should clarify that newcomers with old
felony convictions from indefinite disenfranchisement states are eligible to vote. And those
states which bar some people from voting even after their sentences are completed must
clarify whether new arrivals with old felony convictions from sentence-only
disenfranchisement states are automatically eligible, and must explain what procedures, if
any, should be followed for restoration.

2. Train Election Officials
•

Clarify disenfranchisement policies and procedures for all state and local election officials
through development of materials and training programs in each state. At a minimum, this
should include distribution of posters, brochures and FAQ sheets to local and state elections
offices.

3. Train Criminal Justice Officials
•

Provide training on disqualification and restoration policies for all correctional and criminal
justice officials, particularly probation and parole staff. Correctional and criminal justice
officials should also be actively engaged in describing these policies to persons under
criminal justice supervision.

4. Review Voting Restrictions on Non-Incarcerated People
•

Given the serious practical difficulty of enforcing laws disqualifying people who are not
incarcerated from voting – problems which clearly include both excluding eligible people
from voting and allowing those who should be ineligible to vote -- state policymakers should
review such policies to determine if they serve a useful public purpose.

iv

“I have so many people that believe you can never vote again.”
- Local elections official, Indiana.
“We get felons calling all the time who don’t know they are eligible to vote.”
- State elections official, Missouri.
“When you get these minutes, you almost have to be a lawyer to understand them.”
- Local Louisiana elections official, explaining the importance of interpreting court
minutes in order to determine whether a particular person is disenfranchised or not.

-

“It can be a full-time job.”
Local Tennessee elections official, describing the work of handling rights-restoration
petitions from formerly-incarcerated people in the months prior to elections.

“I’d hate to have to answer that.”
- Local elections official, Indiana, when asked whether a new state resident with a long-ago
criminal conviction in another state could register under Indiana law, which disqualifies only
incarcerated people.

INTRODUCTION
In 1996, the U.S. Department of Justice described American laws barring people convicted of
crime from voting – disenfranchisement law – as “a national crazy-quilt of disqualifications and
restoration procedures.”1 The D.O.J. had state law in mind. But the pieces of that metaphorical
crazy-quilt are not just states, but the counties, cities, towns and parishes within them – the
governments that actually run our localized suffrage system. Drawing on surveys and interviews
with a nationwide sample of state and local elections officials, this report confirms that because
state disenfranchisement laws and practices are so haphazard, a great deal of responsibility for
interpreting and implementing disenfranchisement law remains with county and town elections
officials – and concludes that these officials, while conscientious, hardworking, and wellintentioned, often lack the information and resources to implement the law fairly and effectively.
Across the country, procedures for implementing disenfranchisement law are fraught with error,
excessive variation, uncertainty and outright ignorance.
Basic voter eligibility law is formally set at the state level, but in practice, the ability of people
convicted of crime to vote often depends on local officials. There are two central reasons for
this. First, most disenfranchised people are not in prison: nationwide, about three-quarters of the
disenfranchised are on probation or parole, or have completed their sentences entirely (thirty-five
percent).2 Second, most states lack accurate statewide voter rolls, most lack clear statewide rules
and procedures for suspending citizens from the rolls,3 and many have confusing, cumbersome
rights-restoration procedures. All this leaves important voter-eligibility decisions in the hands of
local registrars.
Particularly since the contested election of 2000, advocates and scholars have battled over the
wisdom of disenfranchising people with criminal convictions. But this debate has proceeded in a
kind of vacuum, almost entirely ignoring how America’s last surviving formal restriction on the
voting rights of adult citizens is actually implemented.
Critics and defenders of
disenfranchisement have parsed judicial decisions, state and federal statutes, even centuries-old
treatises in political theory – but, with a few important exceptions, have not asked the people
who run American elections what the policy really “looks like” in practice. Indeed, the New
York Times recently editorialized that “there is a stunning lack of information and transparency
surrounding felon disenfranchisement across the country.”4
This report aims to help remedy this problem, and to improve our understanding of how
disenfranchisement and restoration happen – or don’t happen – in twenty-first-century America.
It seeks to enable more fair implementation of these laws by state and local authorities, to
identify problems in enforcement, and to suggest avenues for reform and further research.

1

U.S. Department of Justice, Office of the Pardon Attorney, Civil Disabilities of Convicted Felons: A State-by-State
Survey (1996), p. 1.
2
Jeff Manza and Christopher Uggen, “Punishment and Democracy: Disenfranchisement of Nonincarcerated Felons
in the United States,” 2 Perspectives on Politics 491 (2004), p. 495.
3
This was made quite clear in a recent report. See The American Civil Liberties Union, Demos, and the Right to
Vote Campaign, “Purged! How a Patchwork of Flawed and Inconsistent Voting Systems Could Deprive Millions of
Americans of the Right to Vote,” October 2004.
4
“Felons and the right to vote,” New York Times, July 11, 2004.

1

This is a time of transformation in American election administration: states are racing to meet the
deadlines of the Help America Vote Act (HAVA), which requires robust statewide voter lists,
among other improvements. Indeed, HAVA mandates that states coordinate their new
computerized voter rolls “with State agency records on felony status.” 5 In many states, this
means that enforcement of criminal-disenfranchisement policy is beginning to change, moving
towards centralization. But such change is likely to be gradual – HAVA implementation in other
areas has been slower than expected – and parts of the old systems will survive in many states.
For example, even under the new rules, local registrars will continue to have the authority to
“purge” individuals from the rolls in most states. 6 Provisional voting, another HAVA-mandated
reform, caused problems in many states when adopted for the 2004 elections. And Florida’s
2004 experience with a flawed statewide list of ineligible voters showed that centralization alone
is no panacea.
Unfortunately, disenfranchisement’s disarray is symptomatic of the general state of American
election administration – dogged in many areas by inadequate record-keeping, outdated
technology, and excessive variation. But disenfranchisement is uniquely important because it
concerns the fundamental question of who is legally permitted to vote. As states design new
databases, new procedures, and new training regimes for local registrars, it is crucial that they
understand problems with the current, haphazard way disenfranchisement is practiced. State
governments must focus time and resources on this problem if they are to comply with the letter
and spirit of HAVA in time for the 2006 Congressional elections.

5
6

See 42 USC 15483(a)(2)(A).
See Electionline.org, “Assorted Rolls: Statewide Voter Registration Databases Under HAVA,” June 2005, p.3.

2

STATE PRACTICES AND PROCEDURES
As the controversy over criminal-disenfranchisement law has bloomed, advocates and scholars
have read carefully through state constitutions, voting-rights statutes, and case law in order to
understand the formal terms of each state’s eligibility laws. But we have learned relatively little
about how disenfranchisement law is implemented by courts, corrections, and elections officials
in the forty-eight states now barring some people from voting because of criminal convictions.
In the spring of 2004, a two-page survey consisting of questions regarding disqualification and
restoration procedures was sent to the offices of the statewide elections director in each of the
fifty states. Because state practices vary so widely, it was necessary to use open-ended
questions. Responses were collected through the summer and early fall of 2004; thirty-three
states responded, some by typing or writing answers on the survey form, others by attaching a
typewritten letter or list of answers; a few chose to reply via e-mail or phone.
These surveys yielded six central findings. After explaining these overall results, this section of
the report offers a snapshot of some of the ways states disqualify and restore to the rolls people
with criminal convictions.
No state currently administers and enforces its criminal disqualification and restoration laws in
an efficient, universally-understood, equitable way. Some do not appear to notify local elections
officials of convictions, or do not do so in a clear and timely way; others risk “false positives” in
disqualification, particularly with suspended sentences or offenses not subject to
disenfranchisement; many ask local officials to handle disqualification and restoration with little
or no guidance or supervision from the state; none have clear policies regarding new arrivals
from other states with old convictions.
The central conclusions that emerged from these surveys are the following:
•
•
•
•
•
•

Significant variation and ambiguities in voting laws
Misdemeanants disenfranchised in at least five states
Disenfranchisement results in contradictory policies within states
Confusing policies lead to the exclusion of legal voters and the inclusion of illegal
voters
Variation and uncertainty in how states respond to interstate movement of persons
with a felony conviction
Disenfranchisement is a time-consuming, expensive practice

Significant Variation and Ambiguities in Voting Laws
Disenfranchisement law contains profound ambiguities which contribute to official confusion
and uneven enforcement. The debate over disenfranchisement commonly assumes simply that
people convicted of felonies lose the right to vote, and are restored to the franchise later. But in
many states, the truth is far more complicated. For example, in Oregon and Louisiana, not the
conviction but the sentence brings about disqualification, meaning that some people convicted of

3

felonies remain eligible. Tennessee law has changed so often that officials and voters must
implement five different disenfranchisement policies, and several states’ disqualification and
restoration rules change frequently according to the discretionary decisions of officials such as
attorneys general and governors.
In many states, even the formal disenfranchisement law is complicated in ways that appear
irrational, even baffling.
The problems resulting from Washington state’s contested
gubernatorial election of 2004 centered around the document by which former inmates can prove
their eligibility: state and local officials disagree over who should supply and sign the document,
and the laws governing the form have changed three times in recent years. But in many states,
disenfranchisement law contains deeper complexities – legal oddities which can take on great
significance given our difficulty enforcing these laws. Consider, for example, the laws and
policies of Oregon, Louisiana, Pennsylvania, Colorado, Illinois, Tennessee, Iowa, Alabama, and
Mississippi.
Oregon
Oregon’s disenfranchisement laws challenge the assumption that one is disfranchised for
committing a felony, for being convicted of a felony, or even for being incarcerated for a felony:
Oregon disfranchises only those felons under Department of Corrections (DOC) supervision.
Most felons wind up there, but since 1997, some of those convicted of a felony in Oregon and
sentenced to less than 12 months’ custody have gone into the county jails. Therefore, “they do
not enter DOC custody and will not lose their voting privileges,” as an official at the Oregon
Secretary of State’s office explained in a memo responding to this survey. “Felons may
generally vote except when in DOC custody,” the official wrote. Meanwhile, it is legal for even
DOC inmates to register to vote.7
Louisiana
In Louisiana, two interviews with parish officials revealed a similar wrinkle in that state’s
disenfranchisement laws. Asked how she learned which local residents were disqualified, the
parish official explained the state’s relatively comprehensive notification system: federal and
state courts, the Secretary of State’s office, and the state Department of Public Safety all supply
local elections officials with conviction records, and the local registrars then purge people from
the rolls. But there is one difficulty – “a stickler,” as this parish registrar put it – with the court
records. “We get the court minutes, and you have to read through them carefully. Sometimes,
someone may be convicted of a felony, but the judge says ‘imposition of sentence suspended.’
You would think he’d be disqualified, but he’s not – when you get these minutes, you almost
have to be a lawyer to understand them.” Another parish official volunteered the same point: “It
could be a felony, but the sentence could be ‘suspended or deferred.’ We have to read the
minutes, and look for that phrase.” Indeed, as the Louisiana Constitution plainly states,
Louisiana suspends the voting rights not of anyone with a felony conviction, but only of a person
who “is under an order of imprisonment for conviction of a felony.” 8 State elections statutes
repeat this constitutional language at least twice.9
7

Oregon publishes and distributes relatively comprehensive voter-assistance flyers for released prisoners, and
instructs the DOC in what information to communicate to those leaving DOC custody.
8
Louisiana Constitution, Art. I, §10. Emphasis added.
9
See Louisiana Rev’d Statutes §18:102(1), and 18:171.1 A(1).

4

So, being convicted of a felony actually does not disqualify a person from voting in Louisiana;
being sentenced to an order of imprisonment following such a conviction does. (Some advocacy
groups have interpreted “order of imprisonment” to mean that voting rights in Louisiana are
restored upon release from prison. But state officials explain that here, the “order of
imprisonment” actually covers the full sentence.) None of the other eight parish officials
interviewed mentioned the need to parse court minutes looking for this language, suggesting that
in Louisiana – as in many other states – a person’s ability to vote may depend ultimately on how
a state or local official understands, interprets and enforces the ambiguities of
disenfranchisement law.
Pennsylvania/Colorado
As the Secretary of State’s office explained in their survey reply, Pennsylvania law does not
explicitly bar convicted people from voting. Instead, “it implicitly does so by precluding an
individual who is confined in a penal institution from being deemed a resident of the district
where the institution is located and from being deemed a ‘qualified absentee elector.’” In other
words, as a matter of law, Pennsylvania disfranchises incarcerated felons by barring them from
registering, not from voting. In Pennsylvania, the residency-based restriction does not extend to
incarcerated misdemeanants, who are permitted to vote absentee. In Colorado, however, it does:
“once they’re convicted and confined,” says an elections official, misdemeanants “lose their
residence in the precinct where they previously lived,” and are not eligible to vote.
Illinois
An interview with an Illinois Board of Elections official revealed the profound indeterminacy of
disenfranchisement law in a different way. Responding to straightforward questions about the
eligibility of certain people to vote – those on probation or parole, those confined for a
misdemeanor – the official repeatedly used the phrase “our opinion.” When asked about this
language, the official said dryly, “We can offer our ‘opinion’ only, and it’s not a legal opinion,
the state Attorney General tells us.” Again, these were not matters of great interpretation or
discretion: the questions dealt with formal eligibility. While this official reticence would clearly
make sense if a specific individual were asking about his eligibility to vote, one would think
fundamental citizenship rules should admit to objective, factual answers – at least where the
disenfranchisement rules are relatively simple, as they appear to be in Illinois.
Tennessee
Tennessee, by contrast, may have the country’s most complicated disenfranchisement law:
actually, the state has five different laws, covering different time periods between 1973 and the
present. 10 During one interval, someone convicted of a felony in Tennessee suffered no
disenfranchisement at all; a few years later, the same conviction could bring lifetime
disqualification; and the process for restoring voting rights varies depending on the year of
conviction. Obviously, whether or not a formerly-incarcerated person will be restored to the
rolls under such Kafkaesque rules will depend on how well Tennessee’s judges, Corrections
staff, and state and county elections officials are able to navigate the shoals of the law.

10

To their credit, Tennessee makes the statutes available on-line. See <www.state.tn.us/sos/election/webcon1.htm>.

5

Iowa
The most recent major change to disenfranchisement law came in 2005 in Iowa, where Governor
Tom Vilsack restored the voting rights of all people with felony convictions who have completed
their sentences through an executive order. The New York Times headlined its story on this
development “Iowa Governor Will Give Felons the Right to Vote.” 11 But this headline is a bit
misleading: Governor Vilsack’s executive order amounts to a mass restoration for those who
have finished felony sentences, but it does not change the state’s eligibility rules. A future
governor could easily reverse course, returning to a case-by-case restoration process. And even
before the Governor signed his executive order, Iowa Republicans were exploring legal ways to
block the mass restoration of voting rights. 12 Iowa demonstrates the significance of
gubernatorial discretion in disenfranchisement law, particularly in those states which bar some
people from voting even after their sentences are over. (This is also the case in Florida, where
the governor and three colleagues on a special clemency board decide whether to restore voting
rights to some former felons only after an explicit evaluation of their moral character and
personal qualities.)13

Misdemeanants Disenfranchised in at Least Five States
Both critics and supporters of laws barring people convicted of crime from voting usually refer
to “felon disenfranchisement.” But on the national scale, this term is not accurate. Five states
formally bar some or all people convicted of misdemeanors from voting during their sentences:
Colorado, Illinois, Michigan, and South Carolina prohibit anyone in jail or prison after any
criminal conviction, including misdemeanors, from voting. Maryland, meanwhile, disqualifies
those misdemeanants whose offenses appear on the state’s regularly-revised list of “infamous
crimes.” In each of these states, survey responses, statutes, or press accounts suggested that
misdemeanants might be formally excluded, and follow-up calls were made to state elections and
legal officials inquiring whether confined misdemeanants are eligible to vote in these states.
There was usually some uncertainty among officials in the Secretary of State’s and Attorney
General’s offices about this question, with one or two staffers unable to answer, often suggesting
that the caller try another state agency. But in each of these states, a state official stated
definitively that misdemeanants are formally ineligible to vote while confined. Meanwhile, in
Mississippi and Kentucky, constitutions and statutes suggest that misdemeanants might be
disqualified, but state officials in both states said they should retain the right to vote. And in
Indiana, a legal official in the Secretary of State’s office explained that misdemeanants serving
time in jail are not eligible to vote, even though Indiana case law suggests that misdemeanants
should remain eligible.

11

Kate Zernike, “Iowa Governor Will Give Felons the Right to Vote,” New York Times, June 18, 2005.
“Legislators Looking into Another Lawsuit Against Vilsack,” the Associated Press, June 24, 2005.
13
See Abby Goodnough, “Disenfranchised Florida Felons Struggle to Regain Their Rights,” New York Times,
March 28, 2004. Governor Jeb Bush asked one recent applicant, “How’s the anger situation going?” He told
another, “I’m praying that you’re not going to start drinking again.”
12

6

Disenfranchisement Results in Contradictory Policies Within States
Disenfranchisement policies result not only in broad variation among the states, but in often
contradictory policies within states. This leads to the odd situation whereby the states of
Alabama and Mississippi have both the most and least restrictive disenfranchisement policies in
the nation, based solely on the particular felony of conviction. In both states, significant numbers
of offenders are permanently disenfranchised (unless their rights are restored by the state) but
others never lose the right to vote, including while they are in prison.
Disenfranchisement law is changing in Alabama. There, all people with felony convictions are
barred from voting indefinitely, while a new restoration procedure enacted in 2003 removes the
need for an official pardon for most offenders. But in May of 2005, the state Board of Pardons
and Paroles announced that the state Attorney General had determined that people incarcerated
for drug and DUI offenses remain eligible to vote, because their offenses did not involve the
“moral turpitude” emphasized in the state constitution. Elections officials immediately asked the
state to produce a clear list of which crimes do and don’t bring about disenfranchisement, “in
order to avoid different interpretations in each county.” 14 If this new understanding of state law
holds, Alabama’s policies could be simultaneously the most and least restrictive in the country:
some people with felony convictions would be disenfranchised for life, while others would
remain eligible while incarcerated.
This may already be the case in Mississippi, another state which bars some people from voting
indefinitely. Not all felonies are disqualifying, however: Mississippi’s constitution specifically
lists the offenses which bring about disenfranchisement. But the ten-infraction list – reproduced
on the state voter-registration form – is only the beginning. In their response to our survey, state
elections officials listed twenty disqualifying crimes, and in subsequent interviews, have
mentioned that there are now twenty-one. In Mississippi, the state Attorney General regularly
must decide which offenses are disenfranchisable, responding to court rulings and to questions
from state and local elections officials. For example, the Fifth Circuit Court of Appeals held in
1998 that armed robbery is a disqualifying offense – the Mississippi Constitution disenfranchises
those guilty of “theft,” but there is now no crime called “theft” in Mississippi, so the court had to
decide whether armed robbery is equivalent to theft. 15 They held that it is, and after state
elections officials formally asked for a ruling, the state Attorney General agreed. Burglary,
however, is not a disqualifying offense according to the Attorney General, since it does not
necessarily entail the “wrongful taking of property.” More recently, local and state elections
officials queried the Attorney General about whether the offenses of sexual battery, statutory
rape, and carjacking are sufficiently similar to crimes listed on the Constitution that they should
be disenfranchised. The Attorney General reasoned that all of them should. 16
Mississippi law reveals the ambiguities of disenfranchisement law in different ways. First, while
some offenses bring about lifetime disenfranchisement, other felonies – notably, drug infractions
14

Samira Jafari, “Parole Board: Drug, Alcohol Convicts Eligible To Vote From Prison,” Associated Press, May 18,
2005.
15
See Cotton v. Fordice , 157 F.3d 388 (1998).
16
See “Attorney General’s Opinions,” available through the Mississippi Secretary of State’s web pages, at
<http://www.sos.state.ms.us/elections/2004/Handbook/AGOpinions.pdf>.

7

– are not disenfranchised at all, and such offenders may vote even while incarcerated. Second,
the constitutional list of disqualifying offenses, while appearing on the on-line voter-registration
form, is not in fact the list that voters and elections officials need to know. Currently, that list is
longer, and it is produced by the Attorney General’s office, working in conjunction with the
courts, responding to questions emerging from the changing laws of the state and the changing
experiences of election administrators. In order to implement state law consistently, Mississippi
circuit court clerks and state and local elections officials must know not only who has been
convicted of crimes, but what the current list of disenfranchising offenses includes.

Confusing Policies of Restoration and Disqualification
Any disenfranchisement state with an election as close as Washington’s 2004 gubernatorial
contest would likely experience problems as serious as those encountered by Washington voters
and officials. This is particularly true in those states which bar non-incarcerated people from
voting. As noted above, nationally, about three-quarters of the disenfranchised are not in prison:
they are on probation, parole, or – as with about thirty-five percent of those barred from voting –
have completed their sentences entirely. Combined with the risk of “false positives,” erratic
notification procedures, the lack of good statewide voter databases, the need to register before
voting, and high population mobility, these numbers present serious problems for local elections
officials.
In Washington state, votes cast by people with felony convictions were a key focus of
Republican Dino Rossi’s failed challenge to Democrat Christine Gregoire’s narrow victory.
After a series of interviews with state and local officials – and with many of those accused of
voting illegally – the Seattle Times observed that the system designed to restore voting rights to
those eligible while preventing illegal votes is “so bewildering that almost nobody negotiates it
well.” “You need a degree in government to figure it out,” one official told the paper.17 This
problem is not unique to Washington.
While the Washington case illustrates how ineligible people are sometimes permitted to vote,
interviews with many local officials demonstrate that misunderstanding of basic eligibility laws
is likely to contribute to significant numbers of eligible people being denied the right to vote. As
the findings in our survey reveal (see page 15), more than a third of officials interviewed in ten
states either described their state law incorrectly or stated that they did not know a central aspect
of the law. Of those who misidentified the law, more than 85% either did not know the
eligibility standard or specified that the law was more restrictive than was actually the case.
While states have widely varying procedures for disqualification and notification, one theme
appears common to many states: counties make disqualification records only of previouslyregistered voters. That is, elections officials only strike a disqualified voter from the rolls if the
person was registered to vote at the time of their conviction. Other conviction records are often
simply not retained, creating a potential major gap in enforcement in those states where parolees,
probationers, or other non-incarcerated people are barred from voting. Several state survey

17

“Felon-voting laws confusing, ignored.” Seattle Times, May 22, 2005.

8

responses hinted at this, and subsequent phone interviews confirmed it in many states. New
statewide voter lists required by HAVA should help rectify this problem.

Variation and Uncertainty in Addressing Out-of-State Convictions
No state has a procedure set up by which possible criminal backgrounds of new residents of the
state are investigated, and not a single local official interviewed said they could check – though a
few told stories of random notifications from other states. Data problems aside, our interviews
with local elections officials revealed deep uncertainty as to whether disenfranchisement is a
sanction that remains attached to a person moving into a new state.
Among the ten states where local officials were interviewed, five – Delaware, Maryland,
Nebraska, Tennessee, and Wyoming – bar at least some persons from voting even after all
aspects of their sentences are completed. This may present a special problem in the form of new
residents with felony convictions received in other states: neither voter-registration forms nor
state statutes tell voters and officials whether or not these people are eligible voters. State
constitutions and statutes offer little guidance. Delaware and Maryland, for example, both bar
some people from voting for a waiting period post-sentence, and others for life. On its face, this
suggests some new residents might well be ineligible. But constitutions and statutes make no
mention of where the offense and the sentence occur. Delaware’s constitution refers only to
disenfranchisement of “persons convicted of a crime deemed by law a felony,” and Maryland’s
allows the General Assembly to restrict the voting rights “of a person convicted of infamous or
other serious crime.”18 The same silence appears in the constitutions of sentence-only states like
Georgia and Louisiana, which refer only to the disenfranchisement of a “person convicted of a
felony involving moral turpitude” and someone “under an order of imprisonment for conviction
of a felony,” respectively. 19 Statutes in all four states use very similar language.
County officials interviewed in Delaware, Maryland, Nebraska, Tennessee, and Wyoming were
presented with this scenario: a new resident of the county tells you directly that in his previous
state, he had a felony conviction, served the sentence, was automatically restored to the
franchise, and voted – but may not be eligible under his new state’s laws. Would you register
such a person?
Within each state, county registrars were divided on the correct course of action. In Delaware,
one said yes, one said no, and one did not know and would ask the state. In Maryland, one said
yes, five said no, and four said they did not know – three would ask the state for guidance, and
one would ask the county attorney. In Nebraska, none said yes outright, one said no, five would
say yes only if the applicant could provide evidence of restoration in their previous state, and
four did not know and would ask the state. In Tennessee, five would say yes only if the
applicant could prove their previous restoration, four would register only if the applicant were
18

Delaware Constitution Art. V, §2; Maryland Constitution Art. I, §4. The Delaware constitution, however, does
resolve this question with regard to the status of those convicted of crimes of public fraud or bribery, who are
forever barred from voting. Here, the language refers to those convicted of “any like offense, under the laws of any
state or local jurisdiction.”
19
Georgia Constitution Art. II, §1; Louisiana Constitution Art. I, §10.

9

officially restored by a Tennessee official, and one did not know. In Wyoming, two said yes,
two said no, five would say yes only with proof of previous restoration, and one did not know
and would ask the state.
In two other states – Georgia and Louisiana, both of which disenfranchise during the full
sentence – I presented a similar scenario, except that the hypothetical new resident was still
under some form of supervision, or still owed fines or restitution. Again, however, the person
had been eligible to vote and had voted in their previous home state. In Louisiana, four parish
officials would register the person; one would not; one would do so only with proof of
restoration; and four did not know what they would do. In saying no, one parish official said, “if
they tell us they’re a felon, we have to tell them they can’t register.” The very next official
interviewed, however, saw the legal obligations differently: “If they’re not in our database as a
convicted felon, as long as they sign [the registration form], I have to register them.” In Georgia,
one said yes, seven said no, one would register with proof of restoration elsewhere, and one did
not know.
In Indiana, which disenfranchises only during incarceration, officials were asked about a
hypothetical person moving in from an indefinite-disenfranchisement state, where they had
completed a sentence but still were not permitted to vote. Eight Indiana officials responded that
they would register the voter; two were uncertain. “I’d hate to have to answer that,” said one
staffer, before asking the county attorney, who happened to be in the office. “It’s so bizarre how
the states are so different – it’s kind of scary,” said another.
While some registrars showed some familiarity with neighboring states’ disqualification rules,
others openly acknowledged that they did not know other states’ laws. When asked about a
hypothetical person moving into Delaware who had finished a felony sentence but had been
allowed to vote in their previous state while on probation, one Delaware official said, “You mean
there are states that don’t have waiting periods after the sentence before you can vote?” Asked a
similar question, one Maryland staffer said, “I think disqualification for ‘Infamous Crimes’ is
federal law, actually.” (In fact, there is no federal law.) One Louisiana parish registrar summed
up a common sentiment about new arrivals: “There’s no way for us to know, until there’s a
national database. And waiting for that is not something I’m going to lose any sleep over.”

Disenfranchisement is a Time-Consuming, Expensive Practice
The states responding to our survey have widely divergent ways of notifying the people at the
“front lines” of American voting – county and town clerks – of criminal convictions. Some do
not have any formal reporting procedure at all. In notification states, at monthly, quarterly, or
random intervals, some combination of county, state, and federal courts, the Department of
Corrections, sheriffs (in Indiana), or the State Police (in Virginia), notify state elections officials
of criminal convictions. In those few states with comprehensive, functional voter databases,
convictions are checked against voter rolls, and counties notified; in other states, the statewide
information goes straight to local officials. Sometimes county court clerks also convey
conviction information directly to their own elections boards. Finally, the county boards check
conviction lists against their own voter rolls, and purge, cancel, or suspend registrations. “Since

10

registration records are maintained at the various county election board offices,” an Oklahoma
official explained in a typical summary, “it is the Secretary of each county election board office
who actually processes the cancellation.” California, Minnesota, Montana, North Carolina, and
Virginia, among others, are good examples of this type of process.
There are exceptions, however. Among states disfranchising only incarcerated people, a few
indicated that they had no systematic reporting system at all. Michigan, North Dakota, and
Oregon all did so, with Michigan indicating that it is simply “not necessary,” and others
indicating that local officials could investigate odd-looking absentee-ballot requests as they
wished. Kansas may have been the only state disfranchising former inmates which lacked any
reporting system as late as 2004, and they have begun using one in 2005. Our survey did not ask
detailed questions about the cancellation process, but its results certainly provide ample implicit
support for the conclusion reached by the American Civil Liberties Union in a recent report:
states “conduct their purges with great unevenness,” and do not seem to have codified specific
criteria for officials to use in ensuring that cancellations are accurate.20
How do local registrars determine that voters are eligible? States as diverse as Louisiana,
Wyoming, Massachusetts, and Colorado clearly explained state law in their survey responses,
but noted that they simply did not know what the parish, county, and municipal authorities are
doing to check. Remember, even local officials in states barring only incarcerated felons should
have some way of checking voters’ status, since imprisoned pre-trial detainees always remain
eligible, and most or all misdemeanants usually do. In that category, officials in Massachusetts,
Hawaii, Oregon, New Hampshire, and North Dakota responded that it’s up to local officials
whether and how to investigate voter status when they receive an absentee-ballot request with a
return address that may be a prison. “This is a small state, so people often know who has been in
jail,” wrote a New Hampshire official. “County election officials would exercise due diligence
to verify eligibility,” wrote a Hawaii staffer.
North Dakota is in a class of its own: since the state has no voter registration, physical
appearance at the polls is the only proof of eligibility required; the official I spoke to had never
heard of an absentee-ballot request from a state prison in fourteen years of service. In states
barring people from voting after incarceration – such as Colorado, Connecticut, North Carolina,
and Oklahoma – local officials bear even more responsibility, since not only absentee ballots but
the most routine walk-in voter-registration requests may come from people still under
supervision, and the state criminal-justice bureaucracy may not have notified local officials of
their status. In Colorado, people in prison and on parole are not eligible, but probationers and
former inmates are; proof of eligibility “is decided at the county level by the county clerk.”
How? “I don’t have the answer to that question,” a staffer at the Secretary of State’s office
wrote.
In an August 2004 memo to local registrars, Texas’ Director of Elections issued several cautions
regarding cancellation of voters with felony convictions. The memo, which Texas returned
along with its survey responses, reveals the importance of working carefully – whether at the
state or county level, under an old system or a new HAVA-quality regime. The state’s “official
advice,” the memo says, is “not to immediately cancel a voter whom we have identified as a
20

See Purged!, p. 2, 8.

11

possible felon.” Instead, the counties are instructed to “investigate” the registration by sending a
letter to the person in question warning of suspension. One detail is particularly noteworthy:
“DPS [the Department of Public Safety] has cautioned us that felons are frequently convicted
under false names.”
Almost no states now have a centralized procedure to inform county staff of the end of a
person’s incarceration or sentence – and many lack even a simple way for them to check. This
leaves local officials with a great deal of discretion and responsibility. Minnesota explains that
restoration information is centralized with the “state court administrator,” but offers no
information about how the process occurs. Delaware has some of the country’s most restrictive
laws (lifetime disenfranchisement for some offenders, a five-year waiting period post-sentence
for others), but also has one of the most technologically and bureaucratically coherent policies:
county officials must check every registration applicant against a statewide criminal-justice
database, and follow the appropriate steps if a record appears. Virtually no other state described
such a centralized procedure, but HAVA requires all states to create and use statewide voter
databases.
Many states with different terms of disfranchisement require only an oath from the would-be
voter. Among incarceration-only states, Indianans must swear the same affidavit as everyone
else, saying “I am not currently in prison following conviction.” In Montana, the former inmate
“swears on the voter registration card that he or she is not incarcerated.” Oklahoma, a fullsentence disenfranchisement state, uses only a universal oath: “[t]he applicant need not provide
‘proof’ that he/she is again eligible for registration, but must sign a sworn oath, which states in
part . . . If I have been convicted of a felony, a period of time equal to the sentence has expired,
or I have been pardoned.” 21 Similarly, in Missouri, where the offender is disfranchised during all
aspects of the sentence, one must simply swear to a county official that he is no longer under
sentence.
Eleven of our thirty-three states indicated some documentary requirement for restoration. The
list cuts across conventional categories: it includes Hawaii, which bars only incarcerated felons
from voting; Connecticut, which excludes only those incarcerated and on parole; Arkansas,
which adds probationers to the list; and Virginia and Iowa, which indefinitely disenfranchise
felons. In Connecticut, the would-be voter “must present proof of release [from the DOC] from
confinement and/or discharge of parole.” Some full-sentence states, such as North Carolina and
Idaho, have no documentary requirement. In North Carolina, “the county board of elections will
assume the allegation of citizenship is correct.... We do not cross check the current DOC
database with SEIMS [the statewide voter database].” In Idaho, the norm is for local officials to
“[t]ake the individual’s word when they sign the registration card, under oath, that they have no
‘legal disqualifications.’”
Another variable is whether or not those completing sentences are informed of their restored
rights. Until recently, few states employed any procedures designed to inform people leaving
prison of the restoration process. Policy changes have now established such procedures in some
states, handled either by corrections officials or advocacy groups. About one-third of responding
21

Oklahoma’s written survey response continues, “[v]oter registration applications are not ‘investigated’ to
determine whether the applicant has provided correct information....”

12

states – thirteen of thirty-three – indicated either that a state agency is obligated to inform a
released prisoner of their voting rights, or that non-governmental organizations familiar to state
officials regularly do so. Responses to this question cut across categories: most incarcerationonly states do nothing, while several post-incarceration states are relatively active. In Indiana,
the “state DOC is now required to give prisoner notification” of rights restoration after
incarceration; Oklahoma’s DOC “may provide some information,” but there is no statutory
mandate that it do so; Oregon and Nevada both distribute informational flyers; and Montana’s
Secretary of State “works with advocates for prisoners to release information to the media.”
North Carolina and Nevada also have non-governmental advocacy groups’ materials on file, and
returned them with their completed surveys; Connecticut state officials are well-connected with
local advocacy groups.

13

LOCAL PRACTICES AND PROCEDURES
It is important to know how state officials describe laws governing voter disqualification and
restoration, but America’s highly-localized voting system puts most of the work of running
elections in local hands. Registering voters, maintaining the rolls, financing and planning
election day, choosing polling locations, staffing the polls, helping voters, counting and recounting ballots, and paying for it all – in most states, most of these jobs have long been handled
by county, city, and town officials. How are they administering disenfranchisement law? Do
local officials know state rules and regulations? Do they implement them fairly and
consistently?
In order to address these questions, ten states were chosen for phone interviews: Connecticut,
Delaware, Georgia, Illinois, Indiana, Louisiana, Maryland, Nebraska, Tennessee, and Wyoming.
These states were selected for their differences, including varying disenfranchisement policies:
Illinois and Indiana disenfranchise during incarceration only; Connecticut bars voting by those
on parole, as well as incarcerated people; Georgia, Louisiana, and Tennessee (in most cases)
disqualify for the length of the sentence, including both probation and parole; and Delaware,
Maryland, Nebraska, and Wyoming barred some or all people convicted of felonies from voting
even after they’d completed all aspects of their sentences. (Nebraska changed its law from
indefinite disenfranchisement to a two-year post-sentence waiting period just after the
completion of interviews there; the other states impose post-sentence waiting periods on some
offenders, and may bar some people from voting indefinitely.)
Ten counties, towns, or parishes were selected in nine of the states; Delaware has only three
counties. 22 While such a sample is too small to support strong generalizations, it has proven
large enough to yield some insights and draw some conclusions – particularly those which
confirm concerns other studies have identified.
Key findings include:
•
•
•

Election officials ignorant of basic eligibility law
Election officials interpret and enforce disenfranchisement laws inconsistently
Election officials are uncertain of the status of people not registered to vote at the time of
conviction

22

In each state, on-line lists of counties or town elections officials were used. Typically, seven or eight localities
were chosen randomly, usually from the beginning or the end of the alphabet, and then a few large population
centers were selected. Occasionally, state maps were consulted to avoid having the chosen localities cluster in one
part of the state.
Local registrars around the country were simply too busy preparing for primary and general elections to
answer questions during the summer and fall of 2004, when the surveys of state officials were conducted, so these
interviews were conducted during the winter and spring of 2004-2005. Interviews lasted between five and fifteen
minutes, consisting of about ten open-ended questions regarding disqualification and restoration procedures.
Registrars were informed that the caller was conducting research, and officials were never asked their opinion of
disenfranchisement law’s merits

14

Ignorance of Basic Eligibility Law Among Officials
The first question simply asked the official which people convicted of crime were ineligible to
vote under state law, usually with a single follow-up question about whether they would register
an individual who was on probation or parole after a felony conviction. Constitutions, statutes,
judicial rulings, and some states’ official elections web pages were consulted to determine basic
eligibility rules
On these initial questions, of ninety-three local officials interviewed, thirty-four – about 37% either described their state’s fundamental eligibility law incorrectly, or stated that they did not
know a central aspect of that law. In six of ten states, at least three officials either erred or did
not know the eligibility basics. (These figures do not include those who later erred in describing
disqualification or restoration procedures.) Of those thirty-four, seventeen erred in an
exclusionary direction – incorrectly stating, for example, that probationers may not vote in a state
where only incarcerated people are excluded, or wrongly declaring that only an official pardon
restores eligibility in a state where a routine Corrections procedure will do. Just five erred in an
inclusionary direction, wrongly stating that some people legally barred from voting may
participate. And twelve of those thirty-four officials gave an answer which did not indicate a
clear “direction” – stating, for example, that they did not know whether someone on parole
should be allowed to register. Thus, 85% of the officials who misidentified their state’s law
either did not know the eligibility standard or specified that the law was more restrictive than in
practice.
This finding is consistent with the results of published studies previously conducted in seven
other states, each of which has found significant gaps in local officials’ ability to implement the
law. As the table below indicates, single-state studies have consistently found that at least onefifth of those contacted either did not know state law or described it incorrectly.

15

Local Officials’ Knowledge and Enforcement of Disenfranchisement Law
Published Studies of a Single State*

State
Colorado

Research:
Colorado
Voting
Project

Date
Summer/
Fall 2004

Central Findings
Officials in one-third of
Colorado counties described
eligibility law incorrectly; many
asserted that it had been years
since the Sec’y of State’s office
had notified them of felony
disqualifications.

Florida

Multiple
sources,
cited in
Stuart 2004

Winter
2000-2001

Idaho

The Idaho
Statesman

Summer
2003

Believing it to be inaccurate,
twenty-four of Florida’s fortythree counties chose not to use
the state’s list of ineligible
felons in the 2000 election.
Almost one-third of Idaho’s 44
local elections offices did not
know state law.

Minnesota

Minnesota
Lawyers’
Committee
For Civil
Rights
A.C.L.U. of
New Jersey

Winter
2001-2002

“Confusion among the local
county agencies and the state
agencies” with regard to
restoration information.

Fall 2004Spring
2005

New York

Brennan
Center for
Justice

Fall 2002

Ohio

Prison
Reform
Advocacy
Center

Summer
2004

Nine of the state’s 21 counties
required documentary proof of
eligibility, in violation of state
law.
More than half of state’s 62 local
elections boards required
documentary proof of eligibility,
in violation of state’s automaticrestoration law.
“An Ohio ex-offender’s right to
vote may well depend on where
he or she lives in the state:”
about one-quarter of Ohio
counties wrongly stated that
probationers and parolees may
not vote; another sixth did not
know whether they could.

New
Jersey

* For sources, see Appendix A.

16

Results
The Colorado Voting
Project began a public
information and
registration process; Sec’y
of State met with county
officials to prevent
ineligible felons from
voting in Nov. 2004
election.
Florida’s 2000 troubles
helped spark passage of
HAVA, but the state’s
“purge list” remained
flawed into 2004.
Sec’y of State officially
reminded county clerks of
the law, supplying new
quick-reference sheets.
One county (Hennepin)
examined and clarified its
practices; others unknown.

State Attorney General
announced plans to
improve communication
with counties.
State Board of Elections
issued new policy
instructing local boards of
presumption of eligibility.
Unclear. Ohio Dept. of
Rehabilitation and
Correction may instruct
parole officers to inform
former inmates of their
rights, but state action slow
as of November 2004.

Interpretation And Enforcement of Disenfranchisement Inconsistent
When asked, “Do you assume officials in all counties follow the same procedures in criminaldisenfranchisement law?,” local officials generally believed they did. Overall, 30 of 79 officials
who were asked this question directly23 answered clearly “yes,” while 25 gave some kind of
qualified “yes.” Fourteen answered that they did not know; seven said they expected some
variation existed within their state.
Despite this relatively optimistic assessment, in fact there was significant variation on some
matters in every state. Indeed, local officials in some states were unsure about relatively simple
matters like whether their offices could use a statewide database to ascertain voter eligibility: in
four of ten states, a third of county officials either answered this question incorrectly or stated
that they did not know.
Local registrars varied in many other areas, as well. Interviews showed that local registrars in
states around the country differ in their knowledge of basic eligibility law; in how they are
notified of criminal convictions; in what process they use to suspend, cancel, or “purge” voters
from the rolls; whether particular documents are required to restore a voter to eligibility; and
whether they have any information at all about the criminal background of new arrivals to the
state.

Uncertain Status of People not Registered to Vote at the Time of Conviction
As noted above, state surveys hinted that in practice, many states and localities may not
disenfranchise someone not registered at the time of their conviction. This was confirmed by
phone interviews. Rarely do counties keep records of the criminal convictions of local residents
who are not registered to vote. It is even more rare to conduct criminal background checks on
new registrants.
Of course, this only presents a problem for states which bar people from voting after they leave
prison: where only incarcerated people are disqualified, anyone able to walk up to a clerk’s
counter is eligible to register.

23

Most interviews included this direct question; in some conversations, an answer had already been implied, so it
was not asked.

17

CONCLUSION
Purely from the perspective of administrative effectiveness, policies disenfranchising only
incarcerated people appear much more sound than those disqualifying people who are not in
prison. In incarceration-only states, “purges” are not necessary, though some states use them; no
conviction reporting is needed, though many states do this, as well; no complicated restoration
process is necessary; and there are no worries about the eligibility of new arrivals to the state, nor
about people who were not registered to vote back when they were convicted. It should be easy
to make sure all public officials know the law, and know how to implement it effectively and
fairly: registrars would simply need to have the mailing addresses of the state’s correctional
facilities, and investigate any absentee-ballot requests appearing to come from prisons.
Inevitably, states which bar non-incarcerated people from voting are going to disenfranchise
people who are legally eligible, and fail to disqualify some who are not. Washington state made
this painfully clear in 2004, following the example of Florida four years earlier.24 The
administrative simplifications HAVA legislation offers may well reduce the numbers of people
wrongly excluded and included, but are very unlikely to eliminate them entirely. Meanwhile, it
appears that the administration of new, HAVA-approved statewide voter rolls will involve both
state and local officials in many states – that is, both state and local officials may be involved in
adding and deleting voters from the rolls. In this setting, it is imperative that states clarify rules
for voter disqualifications and restorations and invest in substantial training programs,
particularly where non-incarcerated people are barred from the ballot box.
Haphazard administration of suffrage rules should be of great concern to lawmakers and others
interested in guaranteeing that our basic citizenships laws are fairly enforced. Even if one is not
convinced on philosophical grounds that all non-incarcerated people should have the right to
vote, the practical problems documented here virtually ensure that the policy cannot be enforced
with perfect fairness and accuracy. Indeed, even minimizing the numbers of wrongly-excluded
and wrongly-included voters in post-incarceration disenfranchisement states will require
substantial investments of time, money, and resources.

24

As Guy Stuart wrote in his analysis of the 2000 election in Florida, that election “revealed a problem presented by
felon disenfranchisement laws in those states that disfranchise felons who are not incarcerated.” Stuart refers to the
disenfranchisement of unincarcerated felons as an “eighteenth-century policy,” one which ostensibly would have
worked better in a “close-knit society,” where “everyone knew who the felon was.” By contrast, “[t]oday, we have a
highly mobile society, in which we are not expected to police each other.” Guy Stuart, “Databases, Felons, and
Voting: Bias and Partisanship of the Florida Felons List in the 2000 Elections,” 119 Political Science Quarterly 453,
474, 475.

18

APPENDIX A: SUMMARY OF INTERVIEWS BY S TATE

Connecticut
Connecticut changed its disenfranchisement law in 2001, allowing probationers to vote. Reform
advocacy groups have been particularly active in Connecticut, and the state does a number of
things well: Connecticut posts specific notification and restoration procedures on-line, and state
elections staff appear to have worked hard to disseminate knowledge to local officials, including
regular training sessions. All ten towns reported being informed of felony convictions. Given
this combination of factors, it was surprising that of ten town officials interviewed, seven either
erred in describing eligibility or did not know the law; two others could explain who was eligible
only after retrieving and reading the statute. Among those who erred, two were under-inclusive,
three over-inclusive, and two had no “direction.”
However, these troubling numbers may not accurately capture what appears to happen in
Connecticut when a formerly-ineligible person attempts to register. In Connecticut, knowing the
restoration procedure appears to serve as a proxy for actually knowing the disqualification law.
While most town officials in this sample did not know the law, almost all knew that they must
wait for documentary proof of eligibility from the Departments of Corrections and Probation
before restoring someone. While a few towns indicated that the new policy caused some
confusion at first – citizens were being sent to the clerk of court, instead of the Department of
Probation – towns now agree that no one who’s been on the record as convicted will be restored
without documents. Most recently, the Connecticut legislature passed a reform measure that
would further streamline restoration by eliminating the “Certificate of Discharge” now needed to
restore voting rights.
Delaware
Delaware is unusual in a few ways. Despite enacting a significant constitutional reform in 2000
abolishing lifetime disenfranchisement for most offenders, Delaware now has one of the
country’s more complicated eligibility laws: it is one of only a few states with a post-sentence
waiting period for most felons. In Delaware, one must wait five years after the completion of all
aspects of the sentence before registering. Other offenders, meanwhile, including those
convicted of murder, rape, and other “sexual crimes,” are permanently barred from voting. Such
a mixed system would seem inherently difficult to administer. However, the state has only three
counties, and the “county” officials who run elections are actually state employees. This level of
centralization and professionalization might enable Delaware to administer their
disenfranchisement law more consistently than others – despite that law’s complexity.
Indeed, this appears to be the case. Officials in all three counties described eligibility law and
restoration procedures accurately – though in two counties, the person answering the phone did
not know the law, because one person in the office handles all disqualification and restoration
business. (This phenomenon recurs in large counties in states across the country.) The state’s
statewide voter list, meanwhile, does seem to improve enforcement. Indeed, one county official
noted that HAVA will enable Delaware to “tweak some things” – expanding the DMV’s role in

19

registration, for example – but that generally “we’re in wonderful shape” concerning the required
statewide voter lists required by HAVA.
Still, the state is not without its problems. Someone seeking restoration does not need
documentary proof of eligibility in Delaware – the statewide voter list should take care of that.
Indeed, the list also allows officials to check the criminal background of new registrants, which
very few other states can do. But one local official and one state official interviewed volunteered
that court records and the statewide list contain mistakes – “there have been times a person has
come in, and even though it showed up as a felony, it was plead down to a misdemeanor,”
explained the county staffer. For such voters, documentary proof of eligibility is essential. Like
other states, Delaware has no procedure for notifying elections officials that an individual has
been restored to eligibility. Meanwhile, it is not clear who decides which infractions are “sexual
crimes.”
Finally, Delaware – like other states – is simply not able to implement its law barring all
residents with felony convictions from voting for five years after their sentence – and some
permanently – because the state does not know whether new residents are eligible under state
law.
We have no national data estimating how many formerly-incarcerated people move across state
lines each year. The U.S. Census estimates that overall, about twenty-two million Americans
moved across state lines between 1995 and 2000. 25 About seven percent of Americans have a
felony conviction. If they were proportionally represented among interstate migrants, more than
a million moved across state lines during that period, and some indefinite-disenfranchisement
states probably received many thousands.
So, it is a near-certainty that every year, some people move into Delaware fewer than five years
after completing a felony sentence, but Delaware officials simply have no way of knowing this.
They sometimes receive conviction reports from other states – one county official and one state
official noted this – but such reporting is erratic.
What would they would do if a new resident, standing at their counter, disclosed that he had been
convicted, punished, and restored in another state – but did not meet Delaware’s eligibility rules?
As noted above, Delaware officials were divided on this question. The state official interviewed,
meanwhile, responded that the person should not be registered, but acknowledged some
uncertainty here – and reiterated that unless the person volunteered that information, the state
would be extremely unlikely to know it. Of course, this is true of every state. But in those few
states which disenfranchise post-sentence, it is a major gap in enforcement – and one which is
not going to be rectified any time soon, since a national voter roll has no significant proponents
at this time.

25

See U.S. Census Bureau, “Migration by Race and Hispanic Origin: 1995 to 2000,” Table 1, p. 3.

20

Georgia
Georgia has won praise from many election observers for the speed and skill with which its
election-administration bureaucracy has been improved in the last few years. However, several
significant inconsistencies appeared in interviews with county staffers.
If Delaware, with only three counties, is at one extreme, Georgia is near the other: it has 159
counties running elections, under state supervision. When asked whether they believed officials
in all counties followed the same procedures in this area, only three gave a clear “yes;” one gave
a qualified yes, four said they did not know, and two said they believe there is some variation.
“With 159 counties, I’m sure there’s not as much uniformity as I’d like to believe,” said one.
Only six of ten knew basic eligibility law – Georgia disenfranchises those convicted of crimes of
“moral turpitude” during incarceration, probation, and parole, but not thereafter. Of those who
erred, one was overly exclusionary, believing no one could vote without a pardon; one was overinclusive, believing someone on probation could vote; and two simply did not know whether
someone on probation or parole was eligible. Meanwhile, only one county staffer knew that in
terms of disqualification law, “moral turpitude” is effectively synonymous with “felony.” (When
asked what that phrase means, some officials suggested I contact the courts or the state law
library.) All ten counties noted that they are informed of relevant convictions, though their
accounts of how notification happens diverged: half said the courts inform them directly, while
others gave various different explanations. The Purged! report found that the Secretary of
State’s office is supposed to centralize conviction information, particularly for out-of-state and
federal convictions, before forwarding it to the counties.26 But only three local registrars
mentioned receiving information directly from the state.
Uncertainty emerged as to whether someone seeking restoration must prove their eligibility with
documents. One county staffer, when asked about restoration, said that state officials had
recently informed county registrars that the oath is sufficient, or “self-affirming.” But no other
counties mentioned such a rule. When asked if the voter’s oath was sufficient, two said yes,
while five said no. A state official interviewed explained that while there should be some
documents available, he was not certain whether they were required. Meanwhile, several
Georgia officials indicated that despite the presence of a statewide voter roll, the state’s
disenfranchisement regime has the same major flaw as that of most other states: counties are
very unlikely to learn, or to record, that someone not previously registered to vote has been
convicted of a crime of “moral turpitude.” Should such an offender try to register for the first
time while on probation or parole, he would likely succeed.
Like all other states, Georgia lacks any systematic procedure for checking conviction records in
other states – all ten county officials agreed on this. One county registrar recounted that she’d
once received a new-voter registration notice from Mississippi: a person had moved there from
her Georgia county. That person had a felony conviction, however, and she had passed that
information along to Mississippi. (The official did not know that Mississippi is one of the states
where such an individual would be legally barred from voting permanently, even after all aspects
of the sentence were over.) Such a contact had happened only once in ten years, she said.
26

See Purged!, p. 17.

21

Illinois
Illinois bars only those in prison or jail from voting, but not all officials understood this. Five
knew basic eligibility rules; three described them incorrectly, and two simply answered that they
did not know whether a person on probation or parole retains voting rights. The “no” answers
were all exclusionary, assuming probationers and parolees are disqualified.
These figures do not include the misdemeanor question. Asked whether misdemeanants were
allowed to vote absentee while confined, two registrars said yes, three said no, and five did not
know. (One staffer mentioned that someone in her county jail had recently voted, with the
assistance of county personnel, but she was not sure if the person was a pre-trial detainee or a
misdemeanant.) This confusion is understandable. One document included with Illinois’
response to our survey refers to “convicted felons” 27 in its summary of eligibility rules. But the
relevant state elections statute declares that no person “legally convicted . . . of any crime” may
vote while under confinement. 28 This would seem to include misdemeanants. Later, a state
Board of Elections official explained that indeed, it does: confined misdemeanants are not
eligible to vote in Illinois. Only three local officials knew this.
One county official raised another problem: voting by pre-trial detainees. The U.S. Supreme
Court long ago held that pre-trial detainees retain voting rights. But by Illinois law, according to
this official, if someone is physically in the county on election day, they must vote in person.
Counties, however, do not have the staff to transport them securely, explained the registrar. A
state official interviewed later, however, disagreed, saying that such a person should be eligible
to use an absentee ballot. But the state’s statute governing absentee balloting – as well as the online instructions for absentee-ballot applications – do not name pre-trial detainees among the
groups allowed to vote absentee.29
Indiana
Nine of ten Indiana registrars knew the basic law, which is that only incarcerated people are
barred from voting. One registrar volunteered that many local residents do not know they
automatically regain eligibility after leaving prison: “I have so many people that believe you can
never vote again,” she said – this after saying she lived in “a podunk little town.”
But including a question about misdemeanants made this record much less clear. A legal
official in the Secretary of State’s office explains that Indiana misdemeanants serving time in jail
are not eligible: “If you’re dangerous enough to be confined, you can’t vote.” Four of ten county
officials said so, as well. But three others said they did not know if misdemeanants were

27

See “Convicted Felons Voting and Running for Office,” undated memorandum provided by the state of Illinois,
on file with the author.
28
See 10 ILCS 5/3-5.
29
See 10 ILCS 5/19-1. The phrase “physical incapacity” could be interpreted to cover confined pre-trial detainees,
but this is the only language in the statute which appears to deal with those beset by physical illness. No context
clarifies or expands its meaning.

22

eligible, and three more said they remain eligible. One registrar answered, “I’m pretty sure
they’re allowed – ‘cause one year our sheriff let some of ‘em vote."
Incarceration-only states do not really need to notify local officials of criminal convictions:
anyone who can walk up to the counter is eligible, and the only problematic absentee-ballot
requests would have return addresses in the prisons. It is interesting, then, that Indiana appears
to have multiple notification procedures. No two counties gave the same account; convictionnotification sources included the sheriff, Corrections, the Secretary of State, and the local jail.
One registrar explained a quite labor-intensive process she goes through: “the sheriff’s office is
required to notify us of individuals who are going to court, and then I check to see what happens.
I get a list of people in jail. And then I have to do some research to see if they have been
convicted.”
Finally, more than one Indiana registrar volunteered that the Department of Corrections is
supposed to advise released prisoners of restoration possibility. (The state’s response to our
survey had also indicated this.) State officials interviewed also noted that halfway houses often
help with re-registration.
Louisiana
As explained at some length above, Louisiana’s disenfranchisement law contains an important
ambiguity, one shared with a few other states: a person is disqualified not for a felony
conviction, but for a felony sentence. For parish registrars, this is not a trivial difference: two
explained that they must parse court records lest they wrongly disqualify someone convicted of a
felony, but not sentenced to confinement. As one put it, “you almost have to be a lawyer to
understand these court minutes.” Clear notification, therefore, would seem particularly
important in Louisiana, and the state does appear to have redundant notification systems.
Notification is not without its flaws, however: one clerk noted that the records have contained
mistakes, and another said, “they weren’t very good at it, but they’re doing it now.” The biggest
potential problem, however, remains the possibility that some registrars are getting conviction
information from the courts and the state, rather than sentencing information.
Almost all Louisiana registrars interviewed understood that the state disenfranchises for the full
sentence, but not thereafter. But one state official did not, confidently explaining that felons are
disqualified from voting for fifteen years after the expiration of the sentence. In a subsequent
interview, the official corrected the mistake, explaining that the fifteen-year wait referred to
running for elective office, not voting.
A second terminological complexity is the unique use of the term “pardon letter” by some
Louisiana officials. It emerged that first-time offenders routinely receive a “pardon letter” from
their probation officer which enables them to re-register to vote. When asked about this phrase,
one parish registrar said, “hold on, and I’ll just read one to you: ‘Full rights of citizenship have
been restored, but this does not constitute a full pardon.’” Two other officials, however, seem to
have misunderstood this language. When asked basic eligibility rules, they explained that a felon
“has to be pardoned” in order to vote. This language might easily lead some former offenders to
believe they are not eligible.

23

Maryland
Maryland’s basic eligibility rules include a range of classifications. As a result of reform
legislation passed in 2002 abandoning blanket lifetime disenfranchisement, the state now has
three statuses. The law is quite complex, but in most cases, it means this: first-time non-violent
offenders are automatically restored upon completion of their sentences; non-violent recidivists
may be restored three years after the sentence; and violent recidivists are permanently
disenfranchised.
But there is a catch: Maryland disenfranchises those convicted not just of felonies, but of
“infamous crimes.” The state Attorney General revises the list of “infamous crimes” each year –
an official involved in preparing the list acknowledged that it is a difficult, sometimes subjective
process, and that a number of misdemeanors are on the list. “I try not to put on terribly minor
things,” said the official. “It’s not easy to decide which misdemeanors involve an element of
deceit, fraud, or corruption.” New felonies from the Assembly are added, as are more minor
infractions that involve deception. “Some things that might be horrendous might not be on there,
if it doesn’t involve lying,” said the official. For county registrars, the infamous crimes list adds
difficulty to administering the law. Nine of ten local officials interviewed knew the eligibility
basics, though three did not mention the concept of infamous crimes at all.
The list is not at all a secret – all counties should have it, and the state sent one to the author. It
is a striking document. A few disenfranchisable crimes in Maryland: “Fraud – Alcoholic
Beverages;” “Misrepresentation – Tobacco Leaf Weight;” “Racing Horse Under False Name;”
“Willful Unauthorized Use of ID Badge or Card;” “False Representations – Sale of Bedding and
Upholstered Furniture;” “False Statements – Application for Angler’s License.” The list names a
few common offenses that are not “infamous” and therefore do not bring about disqualification,
such as prostitution, harboring a fugitive, and second-degree assault. Drunk driving is not
“infamous” in Maryland.
Puzzlingly, a few infractions name an individual county. These include “Perjury on Application
for Alcoholic Beverage License, Caroline County;” “False Plat Certificate, Queene Anne’s
County;” and “Willful Failure to Pay Beer Tax – Garrett County.” 30 Apparently, some counties
have passed ordinances, the Attorney General has learned of them, and they have been labeled
“infamous.” This adds a new twist to the localized nature of American disenfranchisement: it
seems that in Maryland, whether a person loses voting rights for an infraction may depend not
just on the vagaries of election-law enforcement, but on what county the offense occurs in.
The state’s process for removing ineligible voters from the rolls appears to be quite disorganized.
The A.C.L.U.’s Purged! report describes a process through which all courts report to the
Secretary of State, which then carefully verifies names, makes sure each infraction is on the
infamous crimes list, and only then sends names to the counties.31 However, neither state nor
county officials interviewed described that process as common. Only three counties followed
30

All offenses are listed in “July 2004 Infamous Crimes List,” a document supplied to the author by the state of
Maryland in .pdf form.
31
See Purged!, p. 22-23.

24

what one state official called a “dream world” scenario: receiving conviction lists from courts,
checking each infraction against the infamous crimes list, verifying the offender’s identity, and
only then sending out a cancellation letter. Five other counties seemed to treat court-generated
conviction lists as final, whether or not it had been effectively screened by either the court or the
county.
Maryland, like most states, does not seem to have anything beyond the oath that would prevent
an ineligible voter not formerly registered from registering after leaving prison. A conviction is
only likely to enter county and state voter records if the person is already registered. And
Maryland, like others disqualifying people from voting after their sentence is over, has no way of
dealing with new arrivals. Asked about a statewide database, one county registrar revealed an
awareness of inter-state movement: “That might be helpful, but we’d need a file for the whole
United States.” Would county registrars enroll an honest newcomer who reported that he’d
finished a sentence fewer than three years ago, but had been restored in the state he lived in prior
to Maryland? One said yes, five said no, and four did not know – three would ask the state for
guidance, and one would ask the county attorney. State officials did not give clear answers,
either. One staffer first said no, then said, laughing, “Actually, I would call our A.G. and say
‘what the heck do I do?’” Another official implied that Maryland might make its own
determination of how effectively the person had been punished in the previous state: “If they
come from some lenient state, we wouldn’t necessarily put them on the rolls.”
Nebraska
Nebraska is an indefinite-disenfranchisement state which has seen advocacy and litigation32 for
reform – that is, it was an indefinite-disenfranchisement state until March 10, 2005, when the
unicameral state legislature overrode the governor’s veto and passed a law restoring voting rights
to felons two years after the completion of their sentences. A special state panel had
recommended reform – previously, only a formal pardon restored eligibility – and the legislature
apparently was convinced. The veto override occurred one week after calls to county and state
officials were completed for this report.
Seven county officials knew Nebraska’s old policy; one did not know restoration was possible at
all, and two were not sure about probationers and parolees. With its two-year post-sentence
waiting period under the new law, Nebraska will still need to deal with newcomers whose
eligibility could be cloudy. Interviewed just before the change, five county registrars said they
would enroll an honest newcomer who disclosed a felony conviction in another state. One
registrar told a simple but poignant story:
“Generally the philosophy is, we enforce the law, but the benefit of the doubt goes to
the citizen. I’m going to tell you a situation that happened to me. A family moved in
– [the father] had been convicted in Louisiana. They moved here. Started a new life.
He said he had a felony. But when somebody’s honest, we thank them for it.”

32

Nebraska’s ex-felons had received a discharge document restoring their “civil rights,” but a 2002 decision held
that this did not restore their right to vote. Ways v. Shively, 646 N.W. 2d 621 (Neb. 2002).

25

Four registrars answered that they would call the Secretary of State; there, they would learn that
the newcomer is eligible to vote.
Sometimes, officials in sparsely-populated towns or counties in various states would plead
ignorance of some aspect of the law, explaining that it simply was not necessary: “everybody
knows everybody here.” This happened several times in Nebraska. One such registrar preferred
to err on the side of gentleness in implementing the law. A man came in to register: “We knew
him, knew the family, and knew he had a felony and so wasn’t eligible. But he was with some
other people, so we didn’t want to embarrass him. So we called him later, and told him. And he
did ask about the procedures for getting restored.”
Another small county developed a novel way to purge the rolls: jury duty. Records used in jury
selection and elections are shared, and anyone who indicates on a jury-pool form that he has
been convicted of a felony is automatically disqualified from voting. No other county, in any
state, mentioned a similar practice.
Tennessee
Tennessee may have the country’s most complex formal disfranchisement law. Tennessee is
sometimes described in advocacy reports simply as a state which used to disfranchise
indefinitely, but since 1986 no longer does so. This account does not capture the policy, under
which, as a state webpage explains, “[t]he manner in which a person may restore a lost voting
right depends on the crime committed and the year in which the conviction occurred.”
Tennessee now has five different laws, covering five different periods between 1973 and the
present. 33 Some people with felony convictions are indefinitely disqualified, some never were
disenfranchised, and some are in between. The Tennessee House recently rejected a bill that
would have allowed most former felons to vote after their sentences ended, so it appears this
complicated system will remain in place.
County registrars knew the law – all ten did, in fact. But this must be qualified. All understood
the fact of variation; most read excerpts from the law itself, rather than even trying to summarize
it; all understood that restoration is run through the state government. But virtually none could
explain the disqualification and restoration particulars of each of the periods. None volunteered
that those convicted during a certain period – 1973 to 1981 – were never disqualified, and so
should not need to follow the restoration process. Instead, most Tennessee officials interviewed
tended to do two things: they do not even attempt to remember the law, instead locating the text
and reading from it; second, they essentially compile a restoration file and forward it to the state
for decision. Indeed, final responsibility for all restorations appears to rest with the state. It may
be that necessity is the mother of invention: the law is so complicated that county officials
simply cannot be relied upon for correct implementation, so the state has centralized restoration.
This might bring a modicum of consistency to a chaotic legal environment.
Nonetheless, serious questions emerged about the consistency of Tennessee’s restoration
procedures. First is the question of just what diverse former felons need to do to be restored to

33

See <www.state.tn.us/sos/election/webcon1.htm>

26

the rolls. Depending on their date of conviction and the nature of their offense, some never lost
the right to vote; others need only a routine certificate of restoration; others need to go to court
and get a judicial order. Local staffers described this variation differently, and were clearly
much more practiced in navigating it than others. One registrar had gone so far as to scan an
important restoration document into her own computer – the form used by those convicted
between 1986 and 1996 – so that she can help applicants fill it out. A Nashville official
described getting the restoration document as no big deal: “usually [the process is] something
minor – not notarized, or certified – it’s actually not difficult.” A few others agreed. “In eight
years, I don’t know of any that were turned down,” said one registrar. This official had
recounted that prior to elections, she needed to work through “sixty or seventy restorations a
week – it can be a full-time job.”
Key decisions regarding rights restoration for some former felons lie with Tennessee’s judges,
and officials are not sure just how those judges make their decisions. Asked whether the
decision is pro forma or a matter of some discretion, one state official said, “that may be on a
judge-to-judge type basis. I don’t know.” One county official thought it was virtually automatic,
but others had very different impressions of the judicial-restoration process. A Shelby County
registrar said, “Well, they’ve got to have an attorney to [pursue restoration], and I’m sure a lot of
them just can’t afford it.” A Montgomery County official seconded this point, noting that “most
people don’t go back – it usually takes an attorney.” Those who need judicial restoration should
be able to go to either their county of current residence or the county where they had been
conviction, but only one registrar explained this.
Other officials raised further questions. For example, it is not clear whether all rights are
restored together – specifically, gun rights alongside voting rights. One can imagine this being a
significant factor to judges, but a definite answer to the question proved elusive. Meanwhile, an
attorney with the state elections office noted that “the D.A.’s in some counties get very involved”
with the restoration process – another potential source of variation.
Asking about new arrivals from other states drew several striking comments from Tennessee
registrars. One volunteered that “we have quite a few of those happen; they get quite distressed”
when they read the eligibility law. Another said sarcastically, “they’ve got a panel of lawyers up
there [in Nashville] to answer those questions.” Another, unprovoked, responded “I wish all
these laws would be changed, tell the truth, but that’s Tennessee law right now. In other states,
it’s automatic. Everybody says, ‘I was in Texas, and I didn’t have to do any of this.’” Another
registrar encountered this situation twice last year, involving individuals on parole for offenses in
states (Ohio and Michigan) where parolees are eligible. The official did quite a bit of legwork
trying to figure it out, ultimately passing the decision to state authorities.
Another registrar distinguishes among felonies in administering the law to new residents. “I felt
so bad!,” she recounted, telling a newcomer that his long-ago rape conviction disqualified him
forever in Tennessee. “I think it upset his wife more than him.”

27

Wyoming
The change did not get much national press, but Wyoming – which had been a lifetimedisqualification state – revised its disenfranchisement law in 2003. As the Board of Parole
explains the new policy, five years after the expiration of their sentences, first-time, non-violent
offenders may apply to the Board for restoration of rights. 34 For first-time, non-violent
offenders, then, it is no longer necessary to receive a full gubernatorial pardon.
Unfortunately, of the ten local officials interviewed, only two described something like the new
policy, stating that there were different ways to get rights restored after the sentence. The
remaining eight stated simply that felons could not vote without a “pardon,” though one said she
had heard of a case where someone got his voting rights back through a judge while still on
probation. As noted in the discussion of Louisiana, the word “pardon” can mean different things.
But in Wyoming, it is not accurate to say that one needs a pardon to be restored to the voting
rolls – as a Board official explained, the Board’s certificate is not a pardon. Asked how many
restoration requests the Board typically gets, the official replied, “Oh, not very many – below 30
in the whole time we’ve been doing this, since 2003.” Local uncertainty about the process could
be one reason why.
Questions about the state’s disqualification procedures revealed the amount of work some local
officials need to do to disenfranchise people with felony convictions. One county registrar
explained that she had tried unsuccessfully last year to work with the courts to get conviction
lists. Then she read an alarming newspaper story which said that in neighboring Colorado, the
voter rolls included 6,000 felons. “So, I went to the county clerk; visited with a judge; went to
the courts; and we ended up down at the county attorney. She could only go back three years.
So, I took the voter list down to her. I eventually pulled two cards off the list. One fellow, when
he heard he was off – goes ‘I really want to vote, I really want to vote....’ I said, ‘Buddy, you
should have thought of that before you did the crime.’”
Such a self-initiated purge process appears unusual: two said they received conviction lists from
courts, three said they got lists from the state, and one said she received none at all. When asked
about a statewide database, six county officials volunteered that the new HAVA list was “on its
way,” and would improve their work. Eight local officials said that only the oath prevents
ineligible registrations; one registrar made vague mention of “the state checking” backgrounds,
though this is not currently done; and one registrar described her county’s idiosyncratic
verification process: “We do research if need be, and have done so. We’ve gotten anonymous
tips before; we’ve had judges at the polls who are familiar with the community, and they’ve
come back with doubts, and we look into it.”
Given such comments, perhaps it is not surprising that when asked if they believed all counties
followed the same procedures, not one Wyoming registrar gave a clear “yes” answer. Five gave
a qualified affirmative answer, such as “well, I would think so, but some of them have a better
court system that works with ‘em.” One answered simply: “Every county’s different.”

34

See “Wyoming Board of Parole Policy and
<http://bop.state.wy.us/pdf/parole_board/04PAGE1_1.pdf>

28

Procedure

Manual,”

p.

52.

Available

at

While divided, Wyoming’s registrars were generally willing to register a new resident who
disclosed an out-of-state conviction. Two would do so outright, five would do so with some
proof of restoration elsewhere, two would not, and one would ask the Board of Parole for advice.
The Parole official interviewed was uncertain, but would like to see some proof of restoration
elsewhere.

29

APPENDIX B: STATE SURVEY RESPONSES

Survey Responses: Disqualification
Following are the survey questions regarding disqualification, with summaries of responses.
Thirty-three states responded, demonstrating both the great variation in how states try to
administer disenfranchisement law, and just how important the work of local officials is in that
process.
Q.: In practice, who determines that a person is ineligible to vote?
Local elections authority: 15.
Courts: 9 (Answers to subsequent questions appear to indicate that in 5 of these 9 states,
practical eligibility decisions are actually made by county officials, following felony
convictions in the courts.)
State elections official: 7.
N.A. or unclear: 2.
Q.: How is that determination made?
Survey responses here ranged too widely to be listed in categories. Some local elections
authorities receive information at regular intervals from courts, others irregularly; others
are informed of disqualifications by state elections officials and others by Departments of
Corrections. Sometimes, that notification consists of a formal instruction to disqualify a
given individual; more often, disqualification lists consist of local residents recently
convicted of felonies – but sometimes the list is statewide, with local officials left to
identify local residents. Sometimes the list merely tallies all criminal convictions, and
the local official must check carefully for felony sentences. A few states answered
simply “felony conviction” when asked about how the disqualification determination is
made – without explaining how elections officials learn of such a conviction.
Typically, local officials will then disqualify the voter, after notifying that person by mail
and allowing them to respond.
Q.: Is the fact of a person’s ineligibility communicated to state and local elections officials?
Yes: 23
No: 10
Notably, seven of the ten non-notifying states disenfranchise only during incarceration –
a fact which diminishes, but does not eliminate, the importance of local awareness of
ineligibility – but three bar some former inmates from voting. Also, one state, Colorado,
responded that state offices, but not local officials, are notified by courts; another,
Kansas, responded that officials “may” be notified, but until 2005, there was no legal
reporting requirement.

30

Q.: How is ineligibility communicated to the appropriate elections officials, in those states which
do report convictions?
(Tallied here are the 23 states which responded that some kind of reporting occurs. Most
states did not explain how frequently such reports are made, and several did not explain
the notification process in any detail.)
Survey response indicates that courts send conviction lists directly to appropriate
counties: 6 (Hawaii; Iowa; Tennessee; Oklahoma; Arkansas; Idaho. In Idaho, the county
clerk of court and the chief local elections official are the same person.)
Survey response specifies that some state office – such as Attorney General, Department
of Corrections, or Board of Elections, supplies appropriate information to counties: 5
(Connecticut, Maryland, Mississippi, Montana, Texas).
Survey response indicates that courts report to a state elections bureaucracy, such as the
Secretary of State or Department of Elections, which then forwards appropriate
information to counties: 4 (Kentucky, Minnesota, Utah, Wyoming).
Survey response indicates that courts report conviction information to a statewide voter
list: 4 (Delaware, North Carolina, Virginia, South Carolina).
Survey response indicates that courts, state elections officials, and federal judicial
officers each supply information directly to county officials: 3 (California, Indiana,
Louisiana).
Unclear or incorrect information supplied: 1 (Missouri, where an official said that
counties are notified by the federal Department of Justice – this, according to the
Department of Justice, is not accurate.)

Summary of Survey Responses: Restoration
Following are the survey questions regarding restoration, with summaries of responses. Thirtythree states responded.
Q.: What person or office formally determines that a disqualified person’s eligibility to vote has
been restored?
Automatic/Voter alone: 13
Local official: 11
State Bureaucratic Official, such as Secretary of State, Board of Parole, or Courts: 5
Elected State Official, such as Governor: 4

31

Q.: In practice, are eligibility determinations made at the state level, or do county, city, or town
elections officials sometimes determine whether a formerly-disqualified voter is now eligible?
Local government: 20
(16 county; 3 city or town; 1 parish)
State agency or un-elected official: 7
Voter, usually with oath: 5
Governor: 1
Q.: What conditions must be met in order for a disqualified voter to become eligible?
Specifically, is documentary proof of eligibility required?
No: 19
(Nine of these 19 noted that in unusual circumstances local official "may"
investigate eligibility, though this is not the norm. Notably, eleven of these 19
states disenfranchise only during incarceration; 8 formally bar some nonincarcerated citizens from voting.)
Yes, unspecified or uncertain: 4
Yes, governor’s letter: 4
Yes, from DOC, courts or other criminal-justice institution: 3
Don’t Know: 3
Q.: How do the relevant state, county, or municipal officials determine whether those conditions
have been met?
(So many different answers were given for this question that not all responses are
summarized)
Varying or unspecified county-level research: 12
N.A. – no such research conducted: 9
(Four of these nine – Idaho, Kansas, North Carolina, and Oklahoma –
disenfranchise during the full sentence; others are incarceration-only).
Voter’s oath only: 3
Don’t Know: 3
Q.: Does any state or local official communicate with a formerly-disqualified person to explain
the restoration process, or to inform them that they have been restored to eligibility?
No: 16
No, but aware of some kind of voluntary action by bureaucrats or NGOs: 5
Yes, through Department of Corrections or criminal-justice system: 8
Don’t Know: 4

32

APPENDIX C: SOURCES FOR TABLE

“ACLU-NJ Announces Results of County Elections Survey.” Press release, American Civil
Liberties Union of New Jersey, May 9, 2005.
“Criminal Disenfranchisement in Minnesota: Denying the Right to Vote to Felons.” A report
issued by the Lawyers’ Committee for Civil Rights, January 2002.
“Felons and the Right to Vote.” New York Times, July 11, 2004 (editorial).
“Voting Rights Victory Secured Throughout New York State.” Press Release, Brennan Center
for Justice at N.Y.U. School of Law, Nov. 3, 2003.
Michael do Yoanna, “Ex-felons Can Vote,” INews, Sept. 9-15 2004.
Jeffrey Gold, “ACLU Says Counties Hinder Ex-Con Voting.” Philadelphia Enquirer, May 10,
2005.
Susan Green and Jeffrey A. Roberts, “6,000 Felons on Voter Lists.” Denver Post, Oct. 10, 2004.
Wayne Hoffman, “Some Counties May Not Let Felons Vote.” The Idaho Statesman, August 25,
2003.
Paul E. Kostyu, “Suit Says Election Boards Gave Bad Advice to Felons.” Copley News Service,
August 17, 2004.
“AG Revising Elections Communications on Ex-Cons.” New Jersey Lawyer, May 10, 2005.
John Nolan, “Advocates: Ohio Backing Away from Vow to Tell Released Felons of Voting
Rights.” Associated Press, September 24, 2004.
Carol Peeples, “Colorado’s Felons and the Vote: An Open Letter.” (Copy on file with the
author.)
Guy Stuart, “Databases, Felons, and Voting: Bias and Partisanship of the Florida Felons List in
the 2000 Elections.” 119 Political Science Quarterly 453.

33