This database updates and expands the coding of electoral systems presented in Gaviria et al.'s (2003) Database of Particularism. Data now cover up to 180 countries from 1978-2005 and distinguish electoral systems by the degree to which electoral institutions create incentives for candidates to cultivate a personal vote - as described theoretically in Carey and Shugart (1995) and Gaviria et al. (2003) - including the amount of vote pooling among co-partisan candidates, the amount of parties' control over ballot access, and whether voters cast their votes for candidates or parties. The database also contains several variables that rank-order electoral systems by tier, distinguish mixed-member and other multi-tier electoral systems, capture district magnitude (in two ways), and record election years. Database created 2007. Database last updated 2010.
Last updated by source: 2012-03-24
| Dataset type: | Time-Series |
| Dataset level: | Country |
(Johnson & Wallack,
2012)
Country-level weighted averages of Party Control over Ballot - SMD (lower/only house) (jw_smdballot) and Party Control over Ballot - MMD (lower/only house) (jw_mmdballot), where the weights are the percentage of members that originate from each tier. This variable thus reflects the value of ballots for the average member sitting in the lower house. The ballot variables focus on the amount of party control over candidates' access to a competitive position on the ballot. The variables equal (in order of increasing personal vote incentives): (0) where parties control access to ballots as well as the order in which individuals will fill the seats that the party wins (closed list multi-member districts, open list multi-member districts with little or no de facto change in list order); (1) where parties control access to the ballot, but not the order in which candidates will receive seats (open lists where intra-party preference votes seem to have a significant influence on which candidates are selected, and single-member districts where parties control access to the list); (2) where there are few or no impediments to individual candidates' ability to appear on the ballot (single-member districts where parties do not control access, e.g. allowing independent candidates and/or use primaries to select candidates).
More about this variableCountry-level weighted averages of Party Control over Ballot - SMD (upper house) (jw_smdballot2) and Party Control over Ballot - MMD (upper house) (jw_mmdballot2), where the weights are the percentage of members that originate from each tier. This variable thus reflects the value of ballots for the average member sitting in the upper house. The ballot variables focus on the amount of party control over candidates' access to a competitive position on the ballot. The variables equal (in order of increasing personal vote incentives): (0) where parties control access to ballots as well as the order in which individuals will fill the seats that the party wins (closed list multi-member districts, open list multi-member districts with little or no de facto change in list order); (1) where parties control access to the ballot, but not the order in which candidates will receive seats (open lists where intra-party preference votes seem to have a significant influence on which candidates are selected, and single-member districts where parties control access to the list); (2) where there are few or no impediments to individual candidates' ability to appear on the ballot (single-member districts where parties do not control access, e.g. allowing independent candidates and/or use primaries to select candidates).
More about this variableCountry-level weighted averages of Sharing of Votes among Candidates - SMD (lower/only house) (jw_smdpool) and Sharing of Votes among Candidates - MMD (lower/only house) (jw_mmdpool), where the weights are the percentage of members that originate from each tier. This variable thus reflects the value of the pooling of votes for the average member sitting in the lower house. The Pool variables measure the extent to which votes among candidates from the same party are shared. The variables equal (in order of increasing personal vote incentives): (0) where pooling of votes occurs across all candidates in a party in a district; (1) where pooling of votes occurs across some, but not all, candidates in a party in a district, or, where there is vote pooling across all candidates in a party in a district, but where the average district accounts for 5% or less of a legislature's membership; (2) where no pooling of votes occurs across candidates in a party (including single-member districts).
More about this variableCountry-level weighted averages of Sharing of Votes among Candidates - SMD (upper house) (jw_smdpool2) and Sharing of Votes among Candidates - MMD (upper house) (jw_mmdpool2), where the weights are the percentage of members that originate from each tier. This variable thus reflects the value of the pooling of votes for the average member sitting in the upper house. The Pool variables measure the extent to which votes among candidates from the same party are shared. The variables equal (in order of increasing personal vote incentives): (0) where pooling of votes occurs across all candidates in a party in a district; (1) where pooling of votes occurs across some, but not all, candidates in a party in a district, or, where there is vote pooling across all candidates in a party in a district, but where the average district accounts for 5% or less of a legislature's membership; (2) where no pooling of votes occurs across candidates in a party (including single-member districts).
More about this variableCountry-level weighted averages of Candidate- or Party-specific Voting - SMD (lower/only house) (jw_smdvote) and Candidate- or Party-specific Voting - MMD (lower/only house) (jw_mmdvote), where the weights are the percentage of members that originate from each tier. This variable thus reflects the value of votes for the average member sitting in the lower house. The Vote variables focus attention on the distinction between casting votes for either parties or individual candidates. The variables equal (in order of increasing personal vote incentives): (0) where voters have only one vote for a party; (1) where voters can vote for a party or a candidate (as in open lists), where voters have multiple votes for multiple candidates (as in runoff or single-transferable vote systems), or where votes for a party or candidate are observationally equivalent (as in single-member districts); (2) where voters have one vote for an individual candidate.
More about this variableCountry-level weighted averages of Candidate- or Party-specific Voting - SMD (upper house) (jw_smdvote2) and Candidate- or Party-specific Voting - MMD (upper house) (jw_mmdvote2), where the weights are the percentage of members that originate from each tier. This variable thus reflects the value of votes for the average member sitting in the upper house. The Vote variables focus attention on the distinction between casting votes for either parties or individual candidates. The variables equal (in order of increasing personal vote incentives): (0) where voters have only one vote for a party; (1) where voters can vote for a party or a candidate (as in open lists), where voters have multiple votes for multiple candidates (as in runoff or single-transferable vote systems), or where votes for a party or candidate are observationally equivalent (as in single-member districts); (2) where voters have one vote for an individual candidate.
More about this variableEquals 1 whenever a country has a bicameral legislature.
More about this variableThis variable ranks countries in increasing order of incentives to cultivate a personal vote according to their most dominant or populous tier (or tier with the greater number of legislators). The variable varies from 1 to 13, corresponding to the thirteen positions in Carey \& Shugart's (1995) ranking. For example, a country with a ranking of 1 would have a tier with the lowest possible rank of personal vote incentives, and that tier would account for the majority of the members in the assembly.
More about this variableDummy variable, 1 if year of election to lower house.
More about this variableDummy variable, 1 if year of election to upper house.
More about this variableEquals 1 wherever independent candidates are legally allowed (even where the legal requirements are strict), and 0 otherwise. This complements the cases where the ballot variables above equal 1 or 2, since they are adjusted to capture de facto practice. jw_indy instead captures the de jure rules. A user could adjust the ballot variables above to be de jure if (s)he replaced values of 2 with values of 1 when jw_indy = 0. Refers to lower house elections. The ballot variables focus on the amount of party control over candidates' access to a competitive position on the ballot. The variables equal (in order of increasing personal vote incentives): (0) where parties control access to ballots as well as the order in which individuals will fill the seats that the party wins (closed list multi-member districts, open list multi-member districts with little or no de facto change in list order); (1) where parties control access to the ballot, but not the order in which candidates will receive seats (open lists where intra-party preference votes seem to have a significant influence on which candidates are selected, and single-member districts where parties control access to the list); (2) where there are few or no impediments to individual candidates' ability to appear on the ballot (single-member districts where parties do not control access, e.g. allowing independent candidates and/or use primaries to select candidates).
More about this variableSame as jw_indy, but for upper house elections. The ballot variables focus on the amount of party control over candidates' access to a competitive position on the ballot. The variables equal (in order of increasing personal vote incentives): (0) where parties control access to ballots as well as the order in which individuals will fill the seats that the party wins (closed list multi-member districts, open list multi-member districts with little or no de facto change in list order); (1) where parties control access to the ballot, but not the order in which candidates will receive seats (open lists where intra-party preference votes seem to have a significant influence on which candidates are selected, and single-member districts where parties control access to the list); (2) where there are few or no impediments to individual candidates' ability to appear on the ballot (single-member districts where parties do not control access, e.g. allowing independent candidates and/or use primaries to select candidates).
More about this variableThe number of legislators coded in the dataset. These may not account for the total number of legislators if there are appointed legislators that have no electoral rules to code.
More about this variableThe number of legislators coded in the dataset. These may not account for the total number of legislators if there are appointed legislators that have no electoral rules to code.
More about this variableIn keeping with the emphasis on the incentives faced by individual legislators, this variable measures the district magnitude considering the viewpoint of the average legislator in the lower house. It is scored as a weighted average of the various district sizes, where weights are computed as the number of legislators running in the district of each magnitude divided by the total number of seats. For example: A country with 300 seats divided among one national district with 200 members and 100 single-member districts has a magnitude for the average legislator of [(200*200) + (100*1)]/300, which yields a figure of 133.67.
More about this variableThis is the district magnitude of the average legislator in the upper house.
More about this variableThis is the standard magnitude of the average district in the lower house. For example: A country with 300 seats divided among one national district with 200 members and 100 single-member districts would have an average district magnitude (jw_mdist) of 2.97 (i.e., 300/101).
More about this variableThis is the average district magnitude in the upper house.
More about this variableBallot (coded as above) for multi-member district tiers in elections to the lower house. The ballot variables focus on the amount of party control over candidates' access to a competitive position on the ballot. The variables equal (in order of increasing personal vote incentives): (0) where parties control access to ballots as well as the order in which individuals will fill the seats that the party wins (closed list multi-member districts, open list multi-member districts with little or no de facto change in list order); (1) where parties control access to the ballot, but not the order in which candidates will receive seats (open lists where intra-party preference votes seem to have a significant influence on which candidates are selected, and single-member districts where parties control access to the list); (2) where there are few or no impediments to individual candidates' ability to appear on the ballot (single-member districts where parties do not control access, e.g. allowing independent candidates and/or use primaries to select candidates).
More about this variableBallot for multi-member district tiers in elections to the upper house. The ballot variables focus on the amount of party control over candidates' access to a competitive position on the ballot. The variables equal (in order of increasing personal vote incentives): (0) where parties control access to ballots as well as the order in which individuals will fill the seats that the party wins (closed list multi-member districts, open list multi-member districts with little or no de facto change in list order); (1) where parties control access to the ballot, but not the order in which candidates will receive seats (open lists where intra-party preference votes seem to have a significant influence on which candidates are selected, and single-member districts where parties control access to the list); (2) where there are few or no impediments to individual candidates' ability to appear on the ballot (single-member districts where parties do not control access, e.g. allowing independent candidates and/or use primaries to select candidates).
More about this variablePool for multi-member district tiers in elections to the lower house. The Pool variables measure the extent to which votes among candidates from the same party are shared. The variables equal (in order of increasing personal vote incentives): (0) where pooling of votes occurs across all candidates in a party in a district; (1) where pooling of votes occurs across some, but not all, candidates in a party in a district, or, where there is vote pooling across all candidates in a party in a district, but where the average district accounts for 5% or less of a legislature's membership; (2) where no pooling of votes occurs across candidates in a party (including single-member districts).
More about this variablePool for multi-member district tiers in elections to the upper house. The Pool variables measure the extent to which votes among candidates from the same party are shared. The variables equal (in order of increasing personal vote incentives): (0) where pooling of votes occurs across all candidates in a party in a district; (1) where pooling of votes occurs across some, but not all, candidates in a party in a district, or, where there is vote pooling across all candidates in a party in a district, but where the average district accounts for 5% or less of a legislature's membership; (2) where no pooling of votes occurs across candidates in a party (including single-member districts).
More about this variableVote for multi-member district tiers in elections to the lower house. The Vote variables focus attention on the distinction between casting votes for either parties or individual candidates. The variables equal (in order of increasing personal vote incentives): (0) where voters have only one vote for a party; (1) where voters can vote for a party or a candidate (as in open lists), where voters have multiple votes for multiple candidates (as in runoff or single-transferable vote systems), or where votes for a party or candidate are observationally equivalent (as in single-member districts); (2) where voters have one vote for an individual candidate.
More about this variableVote for multi-member district tiers in elections to the upper house. The Vote variables focus attention on the distinction between casting votes for either parties or individual candidates. The variables equal (in order of increasing personal vote incentives): (0) where voters have only one vote for a party; (1) where voters can vote for a party or a candidate (as in open lists), where voters have multiple votes for multiple candidates (as in runoff or single-transferable vote systems), or where votes for a party or candidate are observationally equivalent (as in single-member districts); (2) where voters have one vote for an individual candidate.
More about this variableThe variable indicates whether there are run-off elections. These are usually for SMDs with absolute majority requirements. Where jw_multiround is equal to 1, voters have more than a single vote to cast, albeit votes occur on separate election days.
More about this variableIndicates whether there are two or more tiers to the legislature.
More about this variableEquals 1 wherever there are multiple allocation tiers, regardless of whether they are the result of mixed member systems that incorporate different members under different rules, or systems that have upper tiers within a single electoral system to compensate for disproportionality in lower tiers.
More about this variableDummy variable, 1 if single-party system.
More about this variableCoded 1 if multiple tiers are elected in parallel fashion, 0 when they are elected in (at least some-what) compensatory fashion. Is coded only when jw_multitier = 1.
More about this variableThis variable ranks countries in increasing order of incentives to cultivate a personal vote according to their more personalistic tier (or tier with the greater incentives to cultivate a personal vote). The variable varies from 1 to 13, corresponding to the thirteen positions in Carey \& Shugart's (1995) ranking. For example, a country with a ranking of 13 would have a tier with the highest possible rank of incentives to cultivate a personal vote, although that tier may only account for a minority or small fraction of its members.
More about this variableShows the proportion of total legislators (elected and non-elected) that are included in the database (i.e. those that are elected).
More about this variableThis is the proportion of the total number of legislators (elected and non-elected) that are coded.
More about this variableProportion of seats from Multi-Member District (lower/only house).
More about this variableThis is the proportion of coded legislators elected in multi-member districts.
More about this variableThe proportion of legislators that are elected via a national tier.
More about this variableThis is the proportion of coded legislators that are elected via a national tier. This is often (but not always) similar to the proportion elected via multi-member districts (jw_propmmd): some electoral systems have proportional representation based on regional multimember districts as well as national tiers (e.g. Hungary).
More about this variableProportion of seats from Single-Member Districts.
More about this variableThis is the proportion of coded legislators elected in single-member districts (Note: In the original data for Kyrgyzstan propsmd2=60 in 1997-1999 and propsmd2=45 2000-2004. We have replaced these figures with missing values).
More about this variableEquals 1 in two circumstances: where voters may rank order candidates according to preference, or where citizens have multiple preference votes for multiple candidates, even if they may not specifically rank the candidates. Otherwise, jw_rank is equal to zero. Refers to lower house elections.
More about this variableSame as jw_rank, but for upper house elections.
More about this variableBallot for single-member district tiers in elections to the lower house. The ballot variables focus on the amount of party control over candidates' access to a competitive position on the ballot. The variables equal (in order of increasing personal vote incentives): (0) where parties control access to ballots as well as the order in which individuals will fill the seats that the party wins (closed list multi-member districts, open list multi-member districts with little or no de facto change in list order); (1) where parties control access to the ballot, but not the order in which candidates will receive seats (open lists where intra-party preference votes seem to have a significant influence on which candidates are selected, and single-member districts where parties control access to the list); (2) where there are few or no impediments to individual candidates' ability to appear on the ballot (single-member districts where parties do not control access, e.g. allowing independent candidates and/or use primaries to select candidates).
More about this variableBallot for single-member district tiers in elections to the upper house. The ballot variables focus on the amount of party control over candidates' access to a competitive position on the ballot. The variables equal (in order of increasing personal vote incentives): (0) where parties control access to ballots as well as the order in which individuals will fill the seats that the party wins (closed list multi-member districts, open list multi-member districts with little or no de facto change in list order); (1) where parties control access to the ballot, but not the order in which candidates will receive seats (open lists where intra-party preference votes seem to have a significant influence on which candidates are selected, and single-member districts where parties control access to the list); (2) where there are few or no impediments to individual candidates' ability to appear on the ballot (single-member districts where parties do not control access, e.g. allowing independent candidates and/or use primaries to select candidates).
More about this variablePool for single-member district tiers in elections to the lower house. The Pool variables measure the extent to which votes among candidates from the same party are shared. The variables equal (in order of increasing personal vote incentives): (0) where pooling of votes occurs across all candidates in a party in a district; (1) where pooling of votes occurs across some, but not all, candidates in a party in a district, or, where there is vote pooling across all candidates in a party in a district, but where the average district accounts for 5% or less of a legislature's membership; (2) where no pooling of votes occurs across candidates in a party (including single-member districts).
More about this variablePool for single-member district tiers in elections to the upper house. The Pool variables measure the extent to which votes among candidates from the same party are shared. The variables equal (in order of increasing personal vote incentives): (0) where pooling of votes occurs across all candidates in a party in a district; (1) where pooling of votes occurs across some, but not all, candidates in a party in a district, or, where there is vote pooling across all candidates in a party in a district, but where the average district accounts for 5% or less of a legislature's membership; (2) where no pooling of votes occurs across candidates in a party (including single-member districts).
More about this variableVote for single-member district tiers in elections to the lower house. The Vote variables focus attention on the distinction between casting votes for either parties or individual candidates. The variables equal (in order of increasing personal vote incentives): (0) where voters have only one vote for a party; (1) where voters can vote for a party or a candidate (as in open lists), where voters have multiple votes for multiple candidates (as in runoff or single-transferable vote systems), or where votes for a party or candidate are observationally equivalent (as in single-member districts); (2) where voters have one vote for an individual candidate.
More about this variableVote (coded as above) for single-member district tiers in elections to the upper house. The Vote variables focus attention on the distinction between casting votes for either parties or individual candidates. The variables equal (in order of increasing personal vote incentives): (0) where voters have only one vote for a party; (1) where voters can vote for a party or a candidate (as in open lists), where voters have multiple votes for multiple candidates (as in runoff or single-transferable vote systems), or where votes for a party or candidate are observationally equivalent (as in single-member districts); (2) where voters have one vote for an individual candidate.
More about this variableEquals 1 when citizens are given a separate vote for deputies in each legislative tier.
More about this variableEquals 1 when citizens are given a separate vote for deputies in each legislative tier.
More about this variable