Our comments are sought on whether it is constitutionally permissible to exclude persons serving a period of imprisonment (hereinafter referred to as prisoners) without the option of a fine from being registered as voters.

1. Clause 4 of the Electoral Laws Amendment Bill, 2003, hereinafter referred to as "the Amendment Bill", provides as follows:

"The following section is hereby substituted for section 8 of the Electoral Act, 1998 (‘the Act’):

1. is detained under the Mental Health Act, 1973 (Act No. 18 of 1973);

2. The purpose hereof is to consider the constitutional validity of the proposed new section 8(2)(f) as contained in clause 4 of the Amendment Bill.

We shall deal with the issue as follows:

1. relevant constitutional provisions will be considered in relation to the proposed new section 8(2)(f).
2. The position of prisoners in foreign countries will be analysed.
3. Thereafter, we shall consider Canadian constitutional law.
4. Finally, we shall comment on the constitutionality of the proposed new section 8(2)(f).

Before dealing with the issues in question, we shall briefly deal with the Act itself.

The Act does not, at present, contain provisions, which restrict the right of prisoners to register as voters and to vote. Unless the Amendment Bill is passed by Parliament, all prisoners will have to be afforded an opportunity of registering as voters and being permitted to exercise their right to vote

This appears to be in accordance with the Constitution, which, unlike the Interim Constitution, does not contain any specific provisions, which disqualify persons from voting.

The 1993 Electoral Act, disqualified prisoners who were convicted and sentenced without the option of a fine for murder, robbery with aggravating circumstances and rape or any attempt to commit such crimes from voting. This Act was passed in accordance with the provisions of section 21(1) of the Interim Constitution.

In the absence of a specific constitutional provision, Parliament, if it wishes to limit the rights of prisoners to vote, "will be obliged to do so in terms of a law of general application which meets the requirements of reasonableness and justifiability as set out in section 36 (of the Constitution)".

3. The question of constitutionality flows directly from the provisions of section 2 of the Constitution, which provides as follows:

In regard to the limitation of a prisoner’s right to vote, the following was stated in August v Electoral Commission, hereinafter referred to as "the August judgment":

4. Central to the consideration of the questions posed above are the interpretation and application of section 36(1) of the Constitution.

5. Section 36(1) of the Constitution provides as follows:

6. On numerous occasions the Constitutional Court has interpreted and applied the provisions of section 36(1) of the Constitution. A precondition for the application of section 36(1) of the Constitution is that a right in Chapter 2 (Bill of Rights) thereof is limited and such limitation is effected by way of "a law of general application".

It is trite that the proposed new section 8(2)(f) would, when enacted, be a law of general application, as it would apply to all prisoners sentenced to and serving a period of imprisonment without the option of a fine. (See in this regard, Larbi-Odam v Member of the Executive Council for Education (North-West Province) 1998 (1) SA 745 (CC) at p. 759, paragraph [27]. In other words, the proposed new section 8(2)(f) meets the precondition.

7. The justification of a limitation of a right provided for in Chapter 2 of the Constitution in the context of section 36(1) thereof involves a two-stage approach. Firstly, it must be established that the law of general application in question limits such a right. Secondly, the limitation must be justifiable according to the criteria prescribed by section 36(1) of the Constitution. The burden of proving that the limitation is justifiable is borne by the person or body that relies on the limitation. (See in this regard, S v Makwanyane 1995 (3) SA 391 (CC) at pp. 435 to 436, paragraph [102] and S v Zuma 1995 (2) SA 642 (CC) p. 654, paragraph [23].)

8. According to Madlanga AJ in S v Steyn 2001 (1) SA 1146 (CC) at p. 1162, paragraph [30], -

9. In De Lange v Smuts NO 1998 (3) SA 785 (CC) at p. 822, paragraphs [86] – [88], Ackermann J stated the following concerning the application of section 36(1) of the Constitution:

(See also Mistry v Interim Medical and Dental Council of South Africa 1998 (4) SA 1127 (CC) at p. 1141, paragraphs [23] et seq; National Coalition for Gay and Lesbian Equality v Minister of Justice 1999 (1) SA 6 (CC) at pp. 30 to 31, paragraphs [33] to[35]; Sonderup v Tondelli 2001 (1) SA 1171 (CC) at p. 1184, paragraphs [29] to [30]; and Prince v President, Cape Law Society 2002 (2) SA 794 (CC) at p. 814, paragraph [[45].)

10. In S v Manamela and Another (Director-General of Justice Intervening) 2000 (3) SA 1 (CC) at p. 20, paragraphs [32] to [33], it is stated that the factors referred to in paragraphs (a) to (e) of section 36(1) of the Constitution are not presented as an exhaustive list, but they are key factors that have to be taken into account in an overall assessment to determine whether the limitation is reasonable and justifiable in an open and democratic society.

11. It is important to note that any effort to demonstrate that the limitation of a right is justifiable in terms of section 36(1) of the Constitution involves a more factual enquiry than the question of the interpretation of that right. In this regard the following is stated in Johan de Waal et al, The Bill of Rights Handbook, 4^th ed, 2001, at p. 147:

As has been mentioned above, the person or body relying on the limitation bears the burden of proving that the limitation is justifiable in terms of section 36(1) of the Constitution and that person or body would therefore have to present appropriate evidence to the court concerned.

12. The proposed new section 8(2)(f) provides that –

13. The first question that must be considered is whether the proposed new section 8(2)(f) limits a right that is provided for in Chapter 2 of the Constitution. Chapter 2 of the Constitution does not provide for a right to be registered as a voter. Nevertheless, registration as a voter is so inextricably connected to a citizen’s right to vote as provided for in section 19(3) of the Constitution that a limitation placed on a citizen’s right to register as a voter would necessarily limit that citizen’s constitutional right to vote. We are of the view that there is no doubt that the new proposed section 8(2)(f) limits the right to vote as provided for in section 19(3) of the Constitution.

14. The second question that must be answered is whether the limitation placed by the proposed new section 8(2)(f) on the right of certain citizens to vote as provided for in section 19(3) of the Constitution can be justified in accordance with the provisions of section 36(1) of the Constitution. In order to answer this question we shall conduct a limitation analysis based upon the criteria provided for in section 36(1) of the Act.

15. The right to vote and the existence of a national common voters roll are two of the values upon which the Republic of South Africa is founded. In this regard section 1(d) of the Constitution provides as follows:

Section 19(3) of the Constitution provides as follows:

"Every adult citizen has the right –

1. to vote in elections for any legislative body established in terms of the Constitution, and to do so in secret; and
2. to stand for public office and, if elected, to hold office."

16. Concerning the constitutional provisions quoted above, Sachs J stated the following in the August judgment at p. 8, paragraph [17]:

It is apparent from what has been quoted above that the right to vote (adult suffrage) is a foundational value upon which the new democratic order in the Republic is built. It is therefore a very important right and its importance is highlighted also by the history of disenfranchisement of most South Africans under apartheid. In applying the principle of proportionality in the weighing-up process one would therefore have to accord the right to vote considerable importance or weight.

17. In order to consider the importance of the purpose of the limitation, it is necessary first to determine the purpose of the limitation effected by the proposed new section 8(2)(f). It is clear that the purpose of the limitation is to exclude certain prisoners during their incarceration from being registered as voters and thereby excluding them from voting. In a document entitled Memorandum on the Objects of the Electoral Amendment Bill, which has been submitted to us, the following is stated in paragraph 4.3 thereof concerning the purpose of the limitation:

18. If one compares the proposed new section 8(2)(f) with the quotation in paragraph … above, the following are apparent:

• First, the wording of the proposed new section 8(2)(f) more or less reflects the purpose enunciated in the quotation above, namely, that prisoners convicted of committing or attempting to commit serious offences and serving a term of imprisonment of more than 12 months without the option of a fine should be disqualified from voting. It may be assumed that persons who have been sentenced to a period of imprisonment without the option of a fine will have been convicted of a serious offence.

• Second, the quotation incorrectly refers to the disqualification of certain prisoners from voting. Strictly speaking, the proposed new section 8(2)(f) disqualifies a prisoner from being registered as a voter while serving a prison sentence of more than 12 months without having been given the option of a fine. However section 24B (2) stipulates that "only prisoners who are awaiting trial may vote". In other words, prisoners who may have registered prior to their having been imprisoned will not be able to use that fact to demand the right to vote. In our view the disqualification is not only a law of general application, but also applies equally to all prisoners who have been incarcerated for a period of imprisonment without the option of a fine. The equality provisions of section 9 of the Constitution appear not to be breached by the amendment in question.

• Third, the use of the disqualification contained in section 47(1)(e) of the Constitution as a guideline appears not to be misguided. The use of that guideline would appear to suggest that its use would be a safe option constitutionally speaking, simply because it is provided for in the Constitution. However, it should be noted that section 36(1) is not applicable to section 47(1)(e) of the Constitution, while it is applicable to the proposed new section 8(2)(f). Moreover, there are material differences between section 47(1)(e) of the Constitution and the proposed new section 8(2)(f). In our view, though, aligning section 8(2)(f) with section 47(1)(e) appears to be reasonable and justifiable in an open and democratic society. The logical (and indeed the only) conclusion to be drawn from sections 1(d) of the Constitution, is that the right to vote encompasses the right to be voted for, as well. This accords with section 19(3). The fact that anyone who is convicted of an offence and sentenced to imprisonment without the option of a fine is not eligible to be a member of the National Assembly suggests that the limitation on prisoners’ right to vote may well be fair and reasonable. It should be noted that a similar prohibition applies to members of provincial legislatures. In both instances, the prohibition applies for a period of five years after the sentence has been completed.

19. Since we are required to comment on the proposed new section 8(2)(f) we are required to interpret the purpose of that provision as it is currently worded. It is apparent from the wording of that provision that its purpose is general in nature and is to prevent a prisoner who is serving a particular type of sentence from being registered as a voter during the course of serving that sentence. The imposition of such a sentence could imply that the prisoner was convicted of a serious offence. However, it is not inconceivable that prisoners convicted of less serious offences could also receive such a sentence. Therefore, in our view the type of offence committed is not necessarily a determining factor in every case. It would appear therefore that the purpose of the proposed new section 8(2)(e) is to disqualify prisoners from being registered as voters while they are serving a sentence of imprisonment without the option of a fine.

20. In order to indicate or determine the importance of the purpose, one would have to determine reasons for the purpose of the limitation that reveal the great value or significance of the purpose. In other words, the question would be: What value or significance can be attached to the reasons for the purpose of the limitation? It would appear that one of the reasons for the limitation would be to exclude certain persons from being registered as voters because they are prisoners serving a particular type of prison sentence and to align the section in question with both sections 47(1)(e) and 106(1)(e). The difficulty lies in the importance that should be attached to such a reason. We have not been furnished with any reason for the limitation or why the limitation is important. We are reasonably satisfied that the two criteria envisaged by the proposed new section 8(2)(f), namely the fact that a person is a prisoner and the fact that such a prisoner is serving a particular type of sentence are sufficient to indicate that the purpose of the limitation is important. However, when read with sections 47(1)(e) and 106(1)(e), the purpose of the limitation assumes a new dimension and importance.

21. It is apparent from what has been stated above that the extent of the limitation is quite considerable. The limitation would affect all prisoners irrespective of the type of offence they were convicted of and who are serving a term of imprisonment without the option of a fine and who had not been registered prior to their incarceration.

22. However, it is clear that the proposed new section 8(2)(f) in effect limits the right of a particular class of prisoner to register and, by necessary implication, the right to vote in an election. In our view the particular class of prisoner to which the proposed new section 8(2)(f) applies is a prisoner who is serving a sentence of imprisonment without the option of a fine and who during that period of time wishes to register as a voter. It does not apply to a prisoner who is sentenced to a period of imprisonment or the payment of a fine and such prisoner has to serve the term of imprisonment because he or she cannot afford to pay the fine. It also does not apply to a person after he or she has completed the said period of imprisonment.

23. As has been mentioned above, we have not been furnished with any reason why a prisoner’s right to vote should be limited and any reason why the purpose of the limitation should be considered important. Therefore, it is difficult if not impossible to consider the relation between the limitation and its purpose, except within the context of section s 47(1)(e) and 106(1)(e).

24. In view of what has been stated above, we are of the view that we are not in a position to consider this aspect of the analysis more carefully. It would be difficult for Parliament or the Government to find less restrictive means to achieve the purpose in the light of the relationship between the proposed amendment and sections 47(1)(e) and 106(1)(e). Indeed, we suggest that there are no less restrictive means.

25. In Charter: Sauvé v Canada (Attorney-General) [1993] 2 SCR 438 the Canadian Supreme Court held that section 51(e) of the Canada Elections Act, RSC 1985, c E-2, which prohibited all prison inmates from voting in federal elections, regardless of the length of their sentences, was unconstitutional as an unjustified denial of the right to vote as guaranteed by section 3 of the Charter. In response to that decision the Canadian Parliament replaced section 51(e) of the Canada Elections Act with a new section 51(e) (SC 1993, c 19, s 23) which denies the right to vote to all inmates serving sentences of two years or more.

26. Section 1 of the Canadian Charter of Rights and Freedoms, hereinafter referred to as "the CCRF", provides as follows:

Section 3 of the CCRF provides as follows:

It is important to note that section 3 of the CCRF is not subject to section 33 thereof.

27. In Sauvé v Canada (Chief Electoral Officer) 2002 SCC 668 a majority of the Justices of the Supreme Court held that the new section 51(e), which denies the right to vote to all inmates serving sentences of two years or more, was unconstitutional.

28. In the majority judgment, which was read by McLachlin CJ, it is stated that the respondents conceded that the voting restriction at issue violated section 3 of the CCRF. Therefore, the restriction would be constitutionally invalid unless the respondents could show that it is demonstrably justified under section 1 of the CCRF. In this regard the following is stated in the majority judgment in paragraph 7:

29. In the majority judgment the government’s argument that denying the right to vote to penitentiary inmates is a matter of social or political philosophy, requiring deference, was rejected. The rejection is based on the following considerations: Rights in the CCRF are not a matter of privilege or merit, but a function of membership of the Canadian polity that cannot lightly be cast aside. The CCRF charges courts with upholding and maintaining an inclusive, participatory democratic framework within which citizens can explore and pursue different conceptions of the good. It is precisely when legislative choices threaten to undermine the foundations of the participatory democracy guaranteed by the CCRF that courts must be vigilant in fulfilling their constitutional duty to protect the integrity of this system. Deference is not appropriate in a decision to limit fundamental rights. The right to vote is the cornerstone of democracy and a stringent justification standard should be applied.

30. The government’s objectives for limiting the right to vote were (1) to enhance civic responsibility and respect for the rule of law and (2) to provide additional punishment. In regard to the objectives’ analysis the following is stated in paragraphs 23 and 24 of the majority judgment:

31. In the majority judgment it was held that the government had not established a rational connection between section 51(e)’s denial of the right to vote and the objectives of enhancing respect for the law and ensuring appropriate punishment.

32. Concerning the test of minimal impairment, the following is stated in paragraph 55 of the majority judgment:

33. In Sauve the government’s argument that denying the right to vote to penitentiary inmates is a matter of social or political philosophy, requiring deference, was rejected. Deference is not appropriate in a decision to limit fundamental rights. In respect of the proposed amendment, the issue is not deference to the prevailing social or political philosophy, but the need to bring the right to vote and be voted for in line with the general exclusions mentioned in sections 47(1)(e) and 106(1)(e) of the Constitution. Clearly no suggestion has been made that the real purpose of the amendment is to enhance civic responsibility and respect for the rule of law and to provide additional punishment. The purpose of the limitation in question is clear, leaving those affected by the limitation in no doubt about the reasons for the limitation of their rights. The objective is accurately and precisely defined and provides a clear framework for evaluating its importance, and to assess the precision with which the means have been crafted to fulfill that objective.

Clearly though, sections 47(1)(e) and 106(1)(e) provide additional punishment for those who were sentenced to more than twelve months’ imprisonment without the option of a fine. That punishment lasts for five years from the date of completion of the sentence and is sanctioned by the Constitution itself.

It should be noted that the logical conclusion to be drawn from the right to vote is that persons enjoying that right are also entitled to make themselves available for membership of the National Assembly or a provincial legislature. Prison populations throughout South Africa are very high. Well-organised prisoners may be able to muster enough votes to ensure a seat in parliament for one or two of their fellow inmates. Would they then be allowed to take their seats, especially in the light of the current debate on prison sizes and conditions? Clearly not, because of the constitutional provisions mentioned herein.

It should be noted that the Canadian Supreme Court shied away from a similar thought, despite its clear and unambiguous stand on the right to vote of prisoners.

The position in Canada in respect of the franchise differs, in our view, fundamentally from the position in South Africa. In Canada every citizen has the right to vote and to be qualified for membership of the House of Commons or of a legislative Assembly. The right is unqualified.

However in South Africa, every citizen has the right to vote and to stand for public office and, if elected, to hold office. The right is qualified by sections 47(1)(e) and 106(1)(e), which prohibits a person from being a member of the National Assembly or a provincial legislature for a period of five years after having completed a sentence of imprisonment without the option of a fine. Because these provisions do not form part of the Bill of Rights, section 36 (Limitations of Rights) does not apply to these two sections.

The difference suggests to us that our Constitutional Court is not likely to follow the approach adopted by the Canadian court, if requested to consider the matter.

It would in any event be cynical in the extreme to state to prisoners serving a sentence of imprisonment without the option of a fine that they have the democratic right to vote, that that right is the cornerstone of our democracy, yet not allow them to participate fully in the democratic processes by making themselves available for election. That would be tantamount to suggesting that prisoners are simply "voting fodder".

34. According to Rottinghaus, dozens of countries (other than the United States of America) place either temporary or permanent voting restrictions on individuals who have been or are currently incarcerated, causing distressing numbers of individuals to be excluded from the civil process.

Legal exceptions to the right to vote are generally regarded as an additional punishment for those who have broken the law.

In this regard and in respect of the position of incarcerated persons internationally, please see the annexure hereto.

33. In the light of what has been stated above, including the instructive remarks made by the majority of Justices in Sauvé v Canada (Chief Electoral Officer) 2002 SCC 668, which are quoted above and which have been considered in relation to the constitutional position in South Africa, we are of the view that the proposed new section 8(2)(f), if enacted, would not necessarily be declared constitutionally invalid. In our view the Constitutional Court could be persuaded to accept the importance of the purpose of the proposed section 8(2)(f).

We do not believe that a constitutional amendment is necessary to effect the proposed amendment. Obviously, that route would remove any doubts about the matter.

Bosnia Norway

Croatia Ukraine

Canada Montenegro

Iran Pakistan

Albania Peru

Czech Republic Poland

Denmark Sweden

France South Africa

Iceland Slovenia

Ireland Switzerland

Israel Puerto Rico

Finland Bangladesh

Greece Kenya

Latvia Macedonia

Lithuania Serbia

NOTE: Countries classified if supporting documentation could be found in constitutions, electoral laws or other corroborating documents. Nations included here expressly document the rights of prisoners to vote (either inelection qualifications or in procedures for special voting needs).

Certain countries have taken a middle ground approach to extending voting rights toprisoners. Table 2 lists the countries that allow prisoners to vote (or not to vote) and the specific circumstances that govern when prisoners are eligible to vote. In general, regardless of country, the more serious the crime involved, the less eligible to vote prisoners become. For instance, in China prisoners on death row and in Lesotho prisoners on death row or serving a life sentence cannot vote. Individuals convicted of crimes considered to be particularly egregious, such as treason or electoral fraud in Germany, those involved in "ideological or anarchistic" activities in Turkey, or those convicted of any felony in Kosovo, are banned from voting as well. In Slovakia and Australia, certain prisoners are allowed to vote in particular elections. In Australia, individuals serving less than five years in prison can vote in federal elections, and in Slovakia prisoners can vote only in federal elections but not local elections. This practice generally follows the form described in the "Theory of Disenfranchisement" section above, where individuals considered to have most egregiously broken the implied or explicit social contract by violating the norms of the community are limited in the rights they can practice while incarcerated.

Australia (twelve months or less can vote; five years and under can vote in federal elections)

Austria (length of sentence)

Belize (no voting for sentences over 1 year)

Benin (sentences of 3 months or more cannot vote)

China (no voting on death row)

Germany (convicted of treason, electoral fraud, espionage or membership in illegal organization

are banned)

Greece (certain felonies)

Kosovo (convicted felons cannot vote)

Italy (certain felons cannot vote)

Jamaica (sentences over 6 months cannot vote)

Japan (certain offenses banned)

Laos (certain offenses banned)

Lesotho (those serving life sentence or on death row are banned)

Macedonia (those prohibited from "practicing their profession" cannot vote)

Mali (sentences over 1 month cannot vote)

Malta (sentences over 1 month cannot vote)

Netherlands (sentences of 1 year or more are banned)

Papua New Guinea (sentences of 9 months or more are banned)

Slovakia (only in presidential elections)

Spain (certain offenses)

Trinidad & Tobago (sentences of more than 1 year are banned)

Turkey (more than one year cannot vote; offenses such as "involvement in ideological or

anarchistic activities.")

Zimbabwe (only those serving a sentence less than six months can vote)

NOTE: Countries classified if supporting documentation could be found in constitutions, electoral laws or other corroborating documents. Those specific offenses which are detailed in election laws or constitutions are included here, but those which are vague are left as such.

Table 3 lists nations that ban prisoners from voting while serving time in prison.

Although no crystal clear pattern emerges, the countries that ban prisoners and ex-prisoners from voting seem to be concentrated in Africa and Latin America. The historical legacy of state-dominated hegemony, the colonial legacy in Africa and the political and social history of strong (and often oppressive) dictatorships in Latin America may contribute significantly to the practice of banning prison voting today and the state power that results from this exercise of this authority. It seems that few traditionally western countries are included on this list, although the United Kingdom is a prominent member of this category.

Azerbaijan Kazakhstan

Angola Kyrgyzstan

Belarus Lithuania

Botswana Uruguay

Bulgaria Malaysia

Comoros Moldova

Equatorial Guinea Mongolia

Estonia Mozambique

Egypt Palestinian Territories

Georgia Panama

Guatemala Poland

Haiti Russia (those awaiting trial can vote)

Honduras Venezuela

Hungary Argentina

Luxembourg Bahamas

Nigeria Barbados

Romania Brazil

Senegal India

Sierra Leone Cape Verde

Vietnam Cyprus

Uganda Equator

United Kingdom (detainees can vote) Latvia

Madagascar Micronesia

Portugal Samon

Sao Tome St. Lucia

St. Vincent Peru

Kenya

NOTE: Countries classified if supporting documentation could be found in constitutions, electoral laws or other corroborating documents.

The countries with the most restrictive practices are catalogued in Table 4 where

prisoners are not eligible to vote while incarcerated and for a period of time after their term is

complete. The countries from Table 3 and Table 4 can be complied together as countries who do not allow prisoners to vote. The United States has the most restrictive practices of this group of countries; some U.S. states permanently ban ex- felons from voting. Nations like Cameroon and Chile ban voting for a ten-year span subsequent to release. Not many countries reach this level of voting restrictions—only eight could be located, but this reflects specific legal and civil society components from each country’s particular history.

Armenia

Cameroon (ten years)

Chile (ten years)

Belgium (those imprisoned for 5 years or mo re are permanently disenfranchised)

Finland (seven years)

New Zealand (seven years)

Philippines (five years)

United States (varies by state – see Case Studies below)

NOTE: Countries classified if supporting documentation could be found in constitutions, electoral laws or other corroborating documents.

Table 5 shows the prison populations in countries that completely restrict voting rights for prisoners and countries that ban ex-prisoners from voting after their term of incarceration is complete. As is clear from the country totals, the vast majority of disenfranchisement occurs in the United States and the Russian Federation where a substantial number of individuals are incarcerated. India and Brazil are the next closest countries in terms of total prison population (at 281,320 and 240,107 respectively) but these are dwarfed by the inmate total in the United States and the Russian Federation.

Almost four and a half million people in the 54 countries on this list do not possess the right to vote as a result of incarceration. Countries that have larger prison populations also seem to be the most restrictive in terms of prisoner access to voting, but there is no proof of causation. It should also be noted that more individuals may be disenfranchised as a result of general oversight during the administration of the election than from technical exclusions (such as those present in the laws of the countries on the list). The impact on civil society, however, is much more difficult to measure. Whether or not these individuals, if released from prison, would positively contribute to the civil society of their countries is unknown.

Angola 49,750

Argentina 38,604

Armenia 4,343

Azerbaijan 17,795

Belarus 55,156

Belgium 8,764

Botswana 6,102

Bahamas 1,280

Barbados 850

Brazil 240,107

Bulgaria 9,283

Cameroon 20,000

Cape Verde 775

Comoros 200

Chile 33,098

Cyprus 369

Equatorial Guinea --

Equator 7,716

Estonia 4,460

Egypt 80,000

Finland 3,617

Georgia 7,688

Guatemala 8,460

Haiti 4,152

Honduras 11,502

Hungary 17,890

India 281,320

Kazakhstan 84,000

Kyrgyzstan 19,500

Latvia 8,437

Lithuania 11,345

Luxembourg 357

Malaysia 28,804

Micronesia 39

Moldova 10,903

Mongolia 7,256

Mozambique 8,812

New Zealand 5,881

Nigeria 39,368

Palestinian Territories --

Panama 10,423

Poland 80,467

Romania 48,053

Russia 905,000

Samoa 176

Senegal 5,360

Sierra Leone --

St. Lucia 243

Venezuela 15,107

Uganda 21,900

United States 2,019,23450

United Kingdom 72,890

Uruguay 5,629

The U.S. figure includes both jail and prison inmates. Almost all of jail inmates are eligible to vote. The actual US prison population is 1,440,655 as of 2002.

NOTE: Data taken from http://www.kcl.ac.uk/depsta/rel/icps/worldbrief/world_brief.html, accessed 6/6/03. The data for Equatorial Guinea, Palestinian Territories and Sierra Leone was not available. The data in this table represent only those who are currently incarcerated; more ex-prisoners in these countries may not be able to vote depending on the country-specific laws.

Blais, et.al. note that "stronger" democracies are slightly less likely to disenfranchise

prison inmates, but "even among ‘strong’ democracies, about one-third disenfranchise all prison inmates, one-third disenfranchise some and another third grant all of them the right to vote." Although their study does not explore the issue of prison voting with the same depth as this paper, the conclusions regarding a discernable pattern of countries and prison voting exclusions similar. There appears to be no strong pattern by geography, age, "strength" of democracy or size, which are typical classificatory categories for comparative politics. Instead, the patters seem to be more country-specific and subtle. Indigenous country history, size of penal population and the public politicization of crime may have significant effects where traditional predictive categories fail. The case studies in the following section will help illuminate these possible causal factors.

This section describes the laws and practices of prison and ex-prisoner voting in several countries. As has been made clear, each nation treats the issue of voting rights for prisoners and ex-prisoners differently, largely as a result of particular historical, social or colonial legacies. These case studies touch on procedures, practices, empirical documentation and impacts in several of the countries listed above. Viewing these countries independent of the larger considerations will illuminate the specific practices and impacts of laws governing prisoner and ex-prisoner voting.

United States. The number of prisoners and ex-prisoners unable to vote in the United States continues to grow rapidly each year as more individuals are incarcerated and those formerly incarcerated are released from prison. The Washington Post of October 1, 2002 notes, "The prison-building boom of the late 1980s and early 1990s, which was fueled by an increase in the number and severity of anti-drug laws, caused the nation's prison population to soar to a current total of more than 2 million."52 The United States retains the second largest rate of incarceration per 100,000 (at 600, Russia is the first at 690) (see Table 5 for a larger comparative analysis). The next closest nation is Belarus at 505. What impact does this have on the electoral system? Forty-eight states and the District restrict inmates from voting while serving time, and at least 32 states bar ex- inmates from voting, either permanently or for several years after their release from prison. Specifically:

Nearly 4 million people, 1 of 50 adults in the country, are denied the right to vote.

Of those, 1.4 million are black men who represent 13 percent of the nation's

black male population.

Table 6 shows the voting-age population in the United States (in total and by state) along with the number of individuals in prison, on probation and on parole and the total number of individuals in each state who are not eligible to vote based upon particular voting laws in each state. The resulting percentage of individuals disenfranchised varies greatly by state, usually hovering at about 1% to 2 % of the total voting age population, with a national average of 1.2 % (or more than two and a half million people). The problem of prison and ex-prisoner disenfranchisement is particularly acute among the population of black males in the United States. Fourteen percent of black men in 1997 were barred permanently from voting based upon specific exclusionary laws in several states. Thirteen states permanently barred voting, which amounts to approximately 510,000 individuals, and more than 950,000 individuals are ineligible to vote because they are currently serving time in prison, on probation or on parole. State specific laws also change the picture: In Alabama and Florida, nearly 1 in every 3 black men is permanently disenfranchised, and in Iowa, Mississippi, New Mexico, Virginia, Washington and Wyoming, the ratio is about 1 in 4.

Alabama 3,401,343 26,741 40,617 5,663 52,713 1.5%

Alaska 439,406 4,546 4,855 522 7,496 1.7%

Arizona 3,960,779 27,710 63,082 3,536 62,787 1.6%

Arkansas 2,031,931 12,159 26,558 10,301 35,739 1.8%

California 25,884,058 159,444 159,444 0.6%

Colorado 3,384,689 17,448 17,448 0.5%

Connecticut 2,537,660 19,196 19,196 0.8%

Delaware 608,981 7,006 19,995 503 17,534 2.9%

DC 459,692 2,750 2,750 .6%

Florida 12,803,101 72,406 294,626 5,891 225,610 1.8%

Georgia 6,357,341 45,937 358,030 20,809 245,761 3.9%

Hawaii 920,403 5,454 5,454 0.6%

Idaho 986,379 6,006 6,006 0.6%

Illinois 9,331,680 44,348 44,348 0.5%

Indiana 4,578,954 20,966 20,966 0.5%

Iowa 2,223,615 7,962 20,797 3,076 21,437 1.0%

Kansas 2,015,152 8,577 8,577 0.4%

Kentucky 3,144,882 15,424 21,993 6,406 32,827 1.0%

Louisiana 3,301,034 35,710 35,710 1.1%

Maine 996,315

Maryland 4,011,170 23,752 80,708 13,415 77,521 1.9%

Massachusetts 4,904,469 10,602 10,602 0.2%

Michigan 7,438,229 48,849 48,849 0.7%

Minnesota 3,745,050 6,606 113,613 3,156 66,569 1.8%

Mississippi 2,127,746 21,460 15,435 1,788 30,966 1.5%

Missouri 4,261,089 28,757 55,767 12,864 69,505 1.6%

Montana 687,113 3,328 3,328 0.5%

Nebraska 1,277,399 3,937 20,847 530 14,891 1.2%

Nevada 1,614,120 10,201 10,454 4,519 19,947 1.2%

New Hampshire 959,733 2,392 2,393 0.2%

New Jersey 6,432,667 28,142 132,846 11,931 106,496 1.7%

New Mexico 1,340,667 5,668 10,335 1,742 12,578 0.9%

New York 14,469,368 67,534 56,719 124,253 0.9%

North Carolina 6,300,761 31,979 110,676 2,954 90,271 1.4%

North Dakota 477,782 1,004 1,004 0.2%

Ohio 8,530,637 45,281 45,281 0.5%

Oklahoma 2,611,041 22,780 30,269 3,406 41,321 1.6%

Oregon 2,656,262 11,455 11,455 0.4%

Pennsylvania 9,367,419 38,062 38,062 0.4%

Rhode Island 799,379 3,241 24,759 375 15,996 2.0%

South Carolina 3,131,422 22,576 42,408 4,100 47,880 1.5%

South Dakota 565,327 2,812 1,532 4,344 0.8%

Tennessee 4,415,247 23,671 41,089 8,074 52,290 1.2%

Texas 15,703,926 162,070 443,684 107,688 491,600 3.1%

Utah 1,586,775 5,343 5,343 0.3%

Vermont 476,162

Virginia 5,534,825 31,603 37,882 4,873 55,417 1.0%

Washington 4,536,596 15,159 159,119 155 94,874 2.1%

West Virginia 1,410,555 4,215 939 5,154 0.4%

Wisconsin 4,076,763 21,533 54,951 9,681 58,690 1.4%

Wyoming 371,023 1,684 4,477 557 4,480 1.2%

The United States is only one of a handful of countries that not only restricts prisoners from voting but also often permanently bars the right to vote in several states.

Table 7 below shows the states that have such laws (as well as those that allow prisoners to vote). The crimes that lead to the exclusion of voting vary widely by state. In some states "major" crimes includedvagrancy, breaking a water pipe, participating in common- law marriage and stealing edible meat. Historical variations in state politics and culture contribute to present day voting standards.

Alabama Maine Washington, DC

Arizona Vermont Hawaii

Delaware Idaho

Florida Illinois

Iowa Indiana

Kentucky Kansas

Maryland Louisiana

Mississippi Massachusetts

Nevada New Hampshire

New Mexico North Dakota

Tennessee Ohio

Virginia Oregon

Wyoming Pennsylvania

South Dakota

Utah

NOTE: Data compiled from http://www.usdoj.gov/crt/restorevote/restorevote.htm, accessed 7/7/03. The states in the first column ban felons from voting in most circumstances, however certain state laws allow for voting to be reinstated after a period of time. The remaining states not listed here ban voting while in prison and for a time after release.

The civil rights movement of the 1960s and 1970s sought to relax the

disenfranchisement of prisoners and ex-prisoners, and a string of court cases in the early 1970s sought to declare such exclusions unconstitutional. The Supreme Court ruled in 1972 that such exclusions were not unconstitutional and that states could decide whom to exclude from the voting rolls. Recently, certain states are also pushing for more stringent voting restrictions, mirroring the civil punishment trend described above. For instance, in Massachusetts in 2001, a ballot initiative passed to deny the right to vote to prisoners, who had previously been able to vote both in state and federal elections. Other states, however, are seeking to expand voting rights for ex- felons. A bill proposed in the Florida State Legislature in 1999 would allow ex- felons to vote one year after the termination of their prison sentence.

Kosovo. As noted in the preceding section, the municipal elections of 2000 allowed for special electoral assistance to "special needs voters," including members of the military and police force not in their home district on election day, those who were hospitalized on election day, hospital staff who were working on election day, those homebound by disability or fear, individuals

with mental impairments housed in medical institutions and those incarcerated in prison. Only prisoners who had not been convicted of a felony, had registered to vote during the "Civil Registration Period" and could produce the requisite documentation or were incarcerated on or before April 19, 2000 (the beginning of the Civil Registration Period) and remained in detention during the full registration period ending July 19, 2000 were permitted to vote.

Prizren 95

Mitrovica 67

Pristina 57

Bondsteel 62

Gnjilane 38

Peja 38

Lipjan 5

Istok-Dubrava 12

NOTE: Source Implementation Plan for Registration and Voting for Individuals with Special Needs,

Joint Registration Task Force, United Nations Mission in Kosovo, 10 August 2000.

Any of the eight prisons listed in Table 8 that allowed their inmates to participate in the program were sent an "Information Package" detailing the necessary requirements for prisoners to be allowed to vote. Figures 1 and 2 below illustrate prisoners taking part in the Special Needs Program days before Kosovo’s National Municipal Election of 2002.64 A Special Needs Team consisting of an international supervisor, a polling station worker, a translator and a driver was dispatched for one day of voting at each prison that allowed its inmates to vote during the week of the scheduled election.