- Electoral courts — Electoral districts — Scheduling of elections — Electoral systems — Restrictions on voting — Right to vote
Published under the direction of the Max Planck Foundation for International Peace and the Rule of Law.
General Editors: Rainer Grote, Frauke Lachenmann, Rüdiger Wolfrum.
1 An electoral district (also known as a constituency) is a whole or a part of a state’s territory or a territorial subdivision, in which voters choose their representatives in elections for specified posts (eg president) or in order to act in specified functions (eg to represent a given group—parliamentary or local elections). The character of electoral districts is one of three fundamental elements, apart from the electoral formula and the vote’s structure, of each electoral system, which determine its type as well as its political consequences. The decision regarding the level of establishing electoral districts (one nationwide district or a territorial district) as well as the number of levels (eg apart from territorial districts, establishing one nationwide district, or a territorial district divided into regional and sub-regional districts) together with the relations between these districts determine the way in which representatives will represent their voters.
2 The most significant political consequence related to electoral districts is the level of proportionality of the electoral system, which is determined by the districts. Researchers who pursue studies concerning the electoral system claim that the crucial element of each proportional system is the size of districts (Eggers and Fouirnaies 267). It can be described as follows: the greater the number of seats, the closer to proportionality the system is (Mackenzie 61). It is often emphasized that this parameter has even greater impact than the type of applied electoral formula (Farrell 161).
3 Each electoral district can be described by four parameters: 1) the number of seats to be filled in a district, ie the size of a district; there can be distinguished single-mandate and multi-mandate districts (mandate); 2) the number of residents/voters living in a district (the district’s size); 3) its territorial delimitation (the shape of electoral district); and 4) the district’s levels. All of these parameters are closely related to each other and involve other political effects. The number of mandates filled in a district implies the height of the natural threshold and—as a consequence—determines the proportionality of elections. The number of residents is related to the question of malapportionment. Further, the size of the electoral district is determined with reference to the standard of representation. Boundaries of districts may be manipulated by gerrymandering. Moreover, when it comes to district levels, three basic levels can be stated: territorial, regional, and nationwide.
B. The Number of Seats—Single-Member and Multi-Member Districts and the Natural Threshold
4 An electoral district may be both single-member and multi-member, and, in the case of multi-member electoral districts, they can be of equal number, or a various number of members can be elected. In the case of election to one post, eg a presidential election, the state territory is an electoral district (eg France, Poland, Slovakia, or Romania).
5 An example of a single-member district is the electoral system in Great Britain, where one Member of Parliament (in the House of Commons) is elected in each electoral district. By using a majority system, it is also possible to elect more members in one electoral district; this formula was applied in Poland in elections to the Senate, where two or three senators were elected in one district within the period 1989–2001, and two, three, and four senators were elected within the period 2001–11 (bicameralism).
6 Generally, multi-member electoral districts are related to proportional representation. In such districts, the number of mandates varies from a few to more than one hundred. It is reported that in proportional systems, small electoral districts have a great impact on a party system (Taagepera and Soberg Sugart 113–14; political parties or fractions in legislative body). For obvious reasons, it is not possible to apply a single-member system in such cases. For instance, the Knesset in Israel is elected through at-large elections. This means that during parliamentary elections, the territory of the whole country is one electoral district, and all 120 members of parliament are elected from nationwide lists of candidates (political candidates and campaigns). The lack of electoral districts is one of the most characteristic features of the Israeli electoral system. In the Netherlands, there is a similar system, and 150 members of the Tweede Kamer are elected in one electoral district.
7 It is obvious that, firstly, the size of an electoral district should be generally proportional to the number of voters in the specific district, and, secondly, particular electoral districts should be of a generally similar size. Only meeting these two conditions guarantees all voters an equal impact on the election results. In the case of elections held in a majority system, other aspects of the process become essential. In single-member districts, the question of the number of seats to be filled (which is very difficult and often complicated in the system of proportional representation) becomes irrelevant. Nevertheless, it is still a significant issue that districts should be composed of a similar number of voters or at least a similar number of residents.
8 The most significant parameter that characterizes an electoral district as well as definitively influences the final distribution of seats in a parliament is the number of members elected, which determines the natural threshold, understood as the minimal level of votes that must be obtained by an election committee in order to secure at least one seat. The more seats to be filled in a given district, the lower the natural threshold, which means that less votes are necessary in order to achieve a seat.
C. The Number of Residents (Voters) and Malapportionment
9 The concept of providing equal or at least a similar number of residents or voters for one seat is not a new idea. It was in the eighteenth century when the concept of a representation standard was presented. It was promoted by French politician Emmanuel Sieyés, who believed that one representative should represent 50,000 voters, while Alexander Hamilton, one of the Founding Fathers of the American Constitution, believed that 30,000 should be the representation standard. The latter idea was even proclaimed in the United States (‘US’) Constitution (Constitution of the United States of America: September 17, 1787 (as Amended to May 7, 1992)), which stated in its initial version of Article 1 section 2 that one mandate in the House of Representatives should correspond to 30,000 free men. As the number of people increases, applying such a representation standard involves a constant increase in the number of parliament members. For instance, in Article 49 of the Constitution of Belgium: 7 February 1831 (Belg), as it was originally adopted, the lawmakers stated that one mandate corresponded to 40,000 residents, which necessitated a constant increase in the number of representatives in the House of Representatives. The number increased from 95 in 1839 to 212 in 1949.
10 The abovementioned rule is also related to the significant reform of the electoral system in Great Britain in 1832, which involved the removal of two types of pathological electoral districts, ie pocket boroughs, in which the residents were said to be in the pockets of their patrons who controlled the selection of members of Parliament from these specific districts as well, and rotten boroughs, with very small electorates, in which only a few citizens had the right to vote but mostly lived outside these boroughs. The most characteristic example of such a district was the infamous Old Sarum in the county of Wiltshire, which was once blooming but started depopulating in the seventeenth century despite the right to elect two members of the British parliament. It is worth noting that the aim of the reform was not only to provide equality in an election but also to prevent corruption, which was common in these districts (corruption and bribery).
11 In the US, the Supreme Court in Reynolds v Sims (1964) (US) ruled that the electoral districts of state legislative chambers must be roughly equal in population. Along with Baker v Carr (1962) (US) and Wesberry v Sanders (1964) (US), it was part of a series of Warren Court cases that applied the principle of ‘one person, one vote’ to US legislative bodies.
12 The phenomenon of non-equal representation between electoral districts, called ‘malapportionment’, is generally caused by long-term natural demographic processes (eg a change in population growth, migration of people, etc), which can lead to a change in the number of people in particular electoral districts, resulting in a change in the proportion of the number of residents to the number of representatives, ie the ratio of voters to one seat in a given electoral district varies from others in particular electoral districts. This phenomenon is rarely a result of authorities’ intentional activity, nevertheless, it can be tolerated. For instance, in France, the scale of differences between electoral districts in the 1960s was even 70,000 voters. Districts with less voters were generally in the country, and big ones were generally in the cities, which was in favour of backers of General de Gaulle.
13 Nowadays, determining a similar number of voters (residents) in relation to one seat in parliament has become standard in elections to lower houses of parliament (or unicameral parliaments). There is, however, a significant difference between Europe and South America. In Europe, specific rules of the distribution of mandates among electoral districts are generally set forth in electoral laws, which provide algorithms that aim to create a distribution that would correspond to the material equality of a vote. Only in some Scandinavian states, eg Denmark, is it stated in a constitution that ‘[i]n determining the number of seats to be allotted to each area account shall be taken of the number of inhabitants, the number of electors, and the density of population’ (s 31(3) of the Constitution of Denmark: 5 June 1953 (Den)). In Finland, there is a similar mechanism: ‘[f]or the parliamentary elections, the country shall be divided, on the basis of the number of Finnish citizens, into at least twelve and at most eighteen constituencies. In addition, the Åland Islands shall form their own constituency for the election of one Representative’ (s 25(2) of the Constitution of Finland: 11 June 1999 (Fin)). This issue is regulated more generally in the Swedish Constitution, (Instrument of Government (SFS nr 1974: 152): 1974 (as Amended to December 7, 2010) (Swed), Chapter 3, Sec. 5) and in the Constitution of Luxembourg, (Constitution of the Grand Duchy of Luxembourg: October 17, 1868 (as Amended to March 12, 2009) (Lux), Art. 51(6)).
14 However, the most common situation is to provide regulation within this matter in laws and statutes, sometimes on the basis of an explicit order provided for in a constitution (eg Art. 54(1) of the Constitution of Greece: 9 June 1975 (Greece), which states that the ‘[e]lectoral system and electoral districts shall be determined by a statute’; Art. 16(2) of the Constitution of Ireland: 1 July 1937 (Ir): ‘Dáil Éireann shall be composed of members representing electoral districts established by a statute’; Art. 51(3) of the Constitution of Luxembourg: ‘[a] law passed under the provisions of Art. 114 (232) sets the number of deputies to be elected in each of the constituencies’). In South America, there is a different approach towards this issue, as these rules are often stated in the constitutions of South American countries. There can be also found even more strict mechanisms that involve constitutional provisions for determining the number of seats to be obtained in particular electoral districts. Such a method is, for instance, applied in Article 176 of the Constitution of Columbia (Constitution of the Republic of Colombia: July 5, 1991 (as Amended to November 4, 2011) (Colom)) and in Article 46 of the Constitution of Argentina: 1 May 1853 (Arg). In Argentina, a mechanism has even been introduced for verifying the number of mandates without the need for a constitutional amendment (amendment or revision of constitutions). Such regulation brings far-reaching consequences, as each change in such a determined distribution of mandates among electoral districts requires amending the constitution, which can sometimes cause wide-ranging difficulties of practical character, particularly occurring when a parliamentary majority or significant minority tends to stop the proposed alterations for political reasons related to potential electoral benefits.
15 Division into electoral districts is made with reference to the representation standard, counted by dividing the number of residents or voters by the number of elected deputies, eg to the parliament. An effective representation standard should be considered as a guarantee of material equality in elections when the redistribution of seats (eg in the parliament) is made on the level of territorial electoral districts rather than one nationwide electoral district. In the case of single-member electoral districts, observing the representation standard involves the fact that in each electoral district, there is a similar (preferably the same) number of residents (voters), and the number of electoral districts equals the number of seats. In the case of multi-member electoral districts, the representation standard determines the number of seats to be obtained in each district. This can be explained with the following example: assuming that the representation standard is 50,000 residents (voters), and, in an electoral district, there are 200,000 residents (voters), for this district, based on the representation standard, there should be four mandates (4 × 50,000 = 200,000). Therefore, the idea of a representation standard is to ensure that seats are distributed equally (in accordance with the same rules) regarding the number of residents or voters.
16 It is also worth noting that the representation standard may be counted in two manners: 1) in advance and on a permanent basis by determining the number of voters (residents) for one mandate, eg 30,000 or 50,000 residents (voters) for one mandate; and 2) each time or on a temporary basis but always episodically by specified calculation (proportion) of the general number of residents (voters) to the total number of seats distributed in a given election, eg ten million residents (voters) to the total number of 200 parliamentary mandates gives a representation standard in the amount of 50,000. As a consequence, the two adopted methods for establishing a representation standard provide two options for the standard: 1) a constant representation standard and 2) a variable representation standard. Adoption of a constant (fixed) representation standard necessitates providing a variable amount of generally distributed mandates, which depends on demographic circumstances. An increase in the number of residents (voters) will cause an inevitable increase in the number of seats, and contrarily, a decrease in the number of residents (voters) will determine a decrease in the total number of mandates.
17 Introducing a constant representation standard will be followed by the fact that on the constitutional or statutory level, the number of representatives will not be directly stated, and only the representation standard will be indicated, ie the number of residents (voters) related to one representative. Romania can be indicated as an example, where, in the Chamber of Deputies, one mandate is dedicated to 70,000 residents, whereas, in the Senate, one mandate is dedicated to 160,000 residents. Establishing a variable representation standard is followed by other problems, ie ensuring systematic correction in the distribution of mandates among electoral districts. Therefore, the option that involves determining the proportion of the number of a country’s residents (voters) to the total number of mandates requires corrections to a country’s division into electoral districts, which should be performed more often with more significant and frequent demographic changes (eg related to voters’ mobility). For these reasons, it is advisable to perform corrections to the representation standard before each election.
18 In a comparative approach, it is quite difficult to determine the optimal volume of the representation standard. It is difficult to point out any commonly applied standard as well. A comparative analysis of solutions applied in contemporary states has allowed the indication of a significant fluctuation of its level. For instance, the representation standard is close to 149,000 in Australia, 133,000 in Spain, and 131,000 in Germany, whereas, it is not more than 39,000 in Switzerland, 37,000 in Greece, and only 27,000 in Ireland. Sometimes, this issue is regulated on a constitutional level. In Austria, for instance, it is stated in Article 26(2) of the Constitution of Austria: 1 October 1920 (Austria) that
the federal territory will be divided into self-contained constituencies whose boundaries may not overlap the provinces’ boundaries; these constituencies shall be sub-divided into self-contained regional constituencies. The number of deputies will be divided among the qualified voters of the constituencies (electoral bodies) in proportion to the number of nationals who in accordance with the result of the last census had their principal domicile in a particular constituency plus the number of those who on the day of the census did not have their principal domicile in federal territory, but were entered on the electoral register of a municipality pertaining to that particular constituency; the number of deputies allocated to a constituency will be divided in the same way among the regional constituencies.
19 An important, as well as an increasingly more frequently discussed, theoretical issue is the selection of the manner of establishing the representation standard, which involves answering the question of whether the representation standard should be related to the number of residents or, rather, to the number of voters. When it comes to the essence of representation, it is more adequate to correlate a mandate to the number of residents, which can be rationalized by the principle of sovereignty—the whole sovereign must be represented—whereas, the correlation of the representation standard to the number of voters or even, according to some postulates, to effective voters (ie those who exercise their right to vote) supports the performance of the rule of material equality. In practice, in spite of theoretical disputes, the representation standard is generally calculated on the basis of the number of residents.
D. The Shape of the Electoral District and Gerrymandering
20 In establishing boundaries of electoral districts, it is recommended to apply specified rules, for example: the boundaries of districts should not cross the boundaries of territorial units (which guarantees resistance to electoral manipulations); the district area should be coherent in terms of its geographic (if possible, also cultural and political) dimension(s); in the case of multi-member districts, variance in the number of mandates should be minimized; topological coherence of a district should be preserved; territorial integrity should be retained; all parts of a district should fit together; and there must be no ‘holes’ in an election map (Haman 177; di Cortona et al). The abovementioned conditions are often not implemented by provisions of election law or are even in contrast to the expectations of specified political groups.
21 There are manipulation techniques related to electoral districts regarding altering their shape in order to ensure the most advantageous electoral result, ie ‘gerrymandering’ (originally: ‘gerrymander’). The term originated in the US and consists of two words: the surname of an American politician associated with the Democratic Party—Elbridge Gerry (1744–1814)—and a mythical salamander. In 1812, Elbridge Gerry, as governor of the state of Massachusetts and the party responsible for determining the boundaries of electoral districts, decided, in accordance with the then provisions in force, to determine the boundaries of an electoral district in the county of Essex to the north of Boston so as to consolidate as many supporters of the Federalist Party in this district as possible. Gerry believed that such action would ensure the majority of supporters of his party in other districts. The shape of the newly created district inspired illustrator and cartoonist Elkanah Tisdale (1768–1835), working with the Boston Gazette, to draw a mythical salamander and publish it in an edition dated 26 March 1812. As a result, the notion of the gerrymander was created. The authors of the notion were never identified, nevertheless, it is highly probable that the first to use this term were Nathan Hale and Benjamin and John Russel. Over time, it was popularized across the world.
22 This surprising and simultaneously effective form of manipulation ensured a clear victory for the Democratic Party, although it gained less votes, and Gerry became a part of the electoral law lecture. He lost the post of governor in the subsequent elections, however, the party compensated him for this loss by offering him the post of the fifth vice-president of the United States (during the presidency of James Madison), which he held until 4 March 1814 when he died. Soon, similar forms of manipulation were applied in other states, generating new forms and names created by the combination of the word ‘mander’ and the name or surname of the person who practiced the specified form of manipulation. Hence, the following occurred: ‘henrymandering’ (originating from the name of Gustavus Henry), ‘jerrymandering’ (originating from the name of the then Governor of California, Jerry Brown), and ‘perrymandering’ (originating from the name of the then Governor of Texas, Rock Perry).
23 The basic goals of ‘gerrymandering’ are to maximize the voting power of supporters and to minimize the voting power of opponents. This technique may be applied by those who are in power and control the electoral process. It may be achieved by using several techniques, which are described in detail by American electoral law:
a) the technique of cracking involves decreasing support for a political party or a candidate by dispersing their electorate among few districts;
b) the technique of packing leads to a situation where supporters of a political party or a candidate are placed in one electoral district so the party or candidate they support receives the majority of votes in said district;
c) the technique of hijacking involves the altering of two electoral districts so that two major candidates of a party compete with each other in one district, which leads to the elimination of one of them and enables the candidate of another political party to win in the other district; and
d) the technique of kidnapping leads to the altering of electoral districts’ boundaries such that the areas where a given political party or a given candidate have significant support are moved to another district in which it is much more difficult to acquire the majority of votes on the basis of new electorate. The technique of kidnapping is often applied against politicians who represent numerous urban areas by excluding big cities from a district and making the district more rural.
24 Gerrymandering may be practiced only in electoral systems with multiple electoral districts, ie it cannot be applied in countries like Israel and the Netherlands where there is one electoral district. This technique has been applied many times in the US (it has also been practiced in many other states, including Australia, France, Greece, Hong-Kong, Canada, Lithuania, Germany, Singapore, Venezuela, and Hungary). Taking into consideration threats posed by electoral district manipulation, generally, the boundaries of electoral districts are verified by courts. For instance, in the US, there are many judgments before each election in which gerrymandering is considered to be illegal.
25 As an example, in Germany in the election of 2005, The Left Party (Die Linke) (the successor of the Party of Democratic Socialism (PDS)) gained 8.7 per cent of the votes and, thus, qualified for top-up seats. The number of Bundestag seats of the parties, which traditionally receive over 5 per cent of the votes, cannot be affected very much by gerrymandering because the seats are awarded to these parties on a proportional basis. However, when a party wins so many districts in any one of the 16 federal states that those seats alone count for more than their proportional share of the vote in that same state, the districting does have some influence on larger parties—those extra seats, called Überhangmandate, remain. In the Bundestag election of 2009, Angela Merkel gained 24 such extra seats while no other party gained any; this skewed the result so much that the Federal Constitutional Court of Germany (Bundesverfassungsgericht) issued two rulings declaring the existing election laws invalid and requiring the Bundestag to pass a new law limiting such extra seats to no more than 15. In 2013, Germany’s Supreme Court ruled on the constitutionality of the Überhangmandate, which, from then on, have had to be added in proportion to the second vote of each party, thereby, making it impossible for one party to have more seats than earned by the proportionate votes in an election.
26 In political practice, the technique of cracking is most common and allows for the determination of the electoral districts’ boundaries such that supporters of one candidate or a given political party are dispersed in as many districts as possible so that in none of the districts could they constitute a majority and obtain a mandate. This is often practiced in order to minimize the voting power of racial, national, ethnic, and religious minorities who live in a given area (racial discrimination; protection of religious minorities). This technique was applied concerning Latinos in the Rio Grande Valley in Texas, African Americans in the county of Durham in North Carolina, and Poles in the Vilnius region in Lithuania. As an example, the US Supreme Court issued an opinion on the case in League of United Latin American Citizens v Perry (2006) (US). While the Court stated that states were free to redistrict as often as desired, the justices ruled that Texas’s 23rd congressional district was invalid, as it violated section 2 of the 1965 Voting Rights Act (US) by racial gerrymandering. In 2019, the Supreme Court ruled that federal judges have no power to stop politicians from drawing electoral districts to preserve or expand their party’s power, a landmark ruling that dissenters said will empower an explosion of extreme partisan gerrymandering. The Supreme Court emphasized the role of state legislatures in districting:
[t]he Framers were aware of electoral districting problems and considered what to do about them. They settled on a characteristic approach, assigning the issue to the state legislatures, expressly checked and balanced by the Federal Congress (Rucho v Common Cause (2019) (US)).
According to Chief Justice Tom Roberts, for federal courts to ‘inject [themselves] into the most heated partisan issues’ by deciding partisan gerrymandering claims ‘they must be armed with a standard that can reliably differentiate unconstitutional from “constitutional political gerrymandering”’.
27 Electoral districts are also a means of implementing the principle of material equality of votes, which consequently leads to ensuring proportional representation for particular territories. Districts are generally filled on a territorial basis, however, they can have personal character as well, eg by assignation to a specified group of people. For instance, in New Zealand, there are special electoral districts for the Maori people. Representatives of this group can declare willingness to vote in such a personal electoral district by registering in this district instead of the general territorial district associated with their place of residence. In such districts, only members of the distinct ethnic group can vote. Furthermore, Bolivia has introduced extraordinary electoral districts for Indian people. They can be established only in a rural area in those departments in which Indian people constitute a minority population (Art. 146(7) of the Constitution of Bolivia: 25 January 2009 (Bol)). Similar regulations have been applied in Taiwan and Venezuela as well. The aim of such regulations is to guarantee a political representation for specified minority groups who, in traditional territorial districts, could be dominated by the majority, composed of people who differ from the minority group in terms of a socially significant feature.
E. The Number of Levels
28 Electoral systems are generally equipped with one (territorial) level of seats distribution in electoral districts. However, different (multi-level) structures of districts can also be found. A comparative analysis of solutions applied in contemporary democratic states leads to the conclusion that there are three basic levels of seats distribution: 1) the territorial level (the lowest level—concordant with the basic units of territorial division of the state); 2) the regional level (covering some territorial districts and/or concordant with additional administrative division of the state); and 3) the nationwide level (covering all of the state’s territory). One (selected) level of the structure of electoral districts may be applied in proportional electoral systems (eg Austria) while at least two levels of electoral districts are applied in mixed systems, in which seats are distributed with the application of two different electoral formulas (majority and proportional—eg Germany).
29 The multi-level structure of electoral districts is related to the issue of the apportionment of electoral districts’ functions between particular structures as well. The fact that there are parallel grids of electoral districts crossing with each other makes it easier to transfer or mix functions between them. Nevertheless, this is also possible and applicable in one-level systems (eg the Netherlands). As a result, an expectation can be created that is not possible to fulfil, wrong electoral strategies may be adopted, and there can even be competition not between electoral committees of political parties but between candidates within those committees, which can be treated as a sign of progressing pathology in a political system.
30 The processes of people’s migration cause constitutional law makers to organize electoral districts for voters who live abroad. It is worth mentioning Article 48(3) of the Constitution of Italian Republic: 27 December 1947 (It), which stipulates: ‘[t]he law lays down the requirements and modalities for citizens residing abroad to exercise their right to vote and guarantees that this right is effective. A constituency of Italians abroad shall be established for elections to the Houses of Parliament; the number of seats of such constituency is set forth in a constitutional provision according to criteria established by law’ (provision introduced by Art. 1 of the constitutional law of 17 January 2000, No. 1, and the Constitutional Law No 1 of 17 January 2000, GU No 15 of 20 January 2000 (It)). A similar project of constitutional amendment was introduced in Poland in 2018, however, there was no required support for it (the formal bill of legislative initiative was prepared by the Senate).
31 The presented features of electoral districts clearly confirm their significance for general elections. Concerning studies performed related to this matter, it must be stated that the number of seats filled in an electoral district, the number of residents/voters living in the district, as well as the delimitation of a district area together with its number of levels are some of the most significant elements of an electoral system, which, in fact, decide its impact on the election results.
32 It has been proven by comparative studies, which have shown that, on the one hand, too much concern is given to electoral formula effects, and, on the other hand, too little attention is drawn to other elements of the electoral system that influence election results, including electoral districts (Rae 124). Therefore, the modification of the number of seats filled in a district, the number of residents/voters living in a district, its shape, and the number of district levels may substantially influence election results.
Benoit, K, ‘District Magnitude, Electoral Formula, and the Number of Parties’ (2001) 39 European Journal of Political Research.
Cox, GW, Making Votes Count. Strategic Coordination in the World’s Electoral Systems (CUP 1997).
di Cortona, PG, et al, Evaluation and Optimization of Electoral Systems (SIAM 1999).
Eggers, AC, and Fouirnaies, AB, ‘Representation and District Magnitude in Plurality Systems’ (2014) Electoral Studies.
Farrell, DM, Electoral Systems. A Comparative Introduction (2nd edn Red Globe Press 2011).
Gallagher, M, and Mitchell, P, (eds), The Politics of Electoral Systems (OUP 2008).
Griffith, EC, The Rise and Development of the Gerrymander (1907; Kessinger 2010).
Haman, J, Demokracja, Decyzje, Wybory (Scholar 2003).
Lijphart, A, Patterns of Democracy. Government Forms and Performance in Thirty-Six Countries (Yale University Press 1999).
Mackenzie, WJM, Free Elections: An Elementary Textbook (Rinehart 1958).
Martis, KC, The Original Gerrymander (2008) 27 Political Geography 833.
Michalak, B, ‘Okregi wyborcze: podstawowe narzedzie inzynierii wyborczej’ in Szymanek, J, (ed.), Niedemokratyczne wymiary demokratycznych wyborów (Sejm Publishing House 2016).
Rae, D, The Political Consequences of Electoral Laws (Yale University Press 1971).
Renwick, A, and Pilet, J-B, Faces on the Ballot: The Personalization of Electoral Systems in Europe (OUP 2016).
Sieyès, E, Qu’est-ce que le Tiers-État? (Paris 1789).
Sokala, A, Michalak, B, and Uziębło, P, Leksykon prawa wyborczego i referendalnego oraz systemów wyborczych (Wolters Kluwer Polska 2013).
Taagepera, R, and Soberg Shugart, M, Seats and Votes, The Effects and Determinants of Electoral Systems (Yale University Press 1989).
Taagepera, R, Predicting Party Sizes. The Logic of Simple Electoral Systems (OUP 2007).
Zukowski, A, ‘Gerrymandering—manipulacja granicami okregów wyborczych w systemach demokratycznych’ in Kowalski, M, (ed.), Przestrzen wyborcza Polski (Polish Academy of Sciences 2003).