A Progress Report on The Comparative Study of Electoral Systems October 11, 1994

As part of its call for an internationally collaborative program of cross-national electoral research, in March 1994 the International Committee for Research into Elections and Representative Democracy (ICORE) circulated to directors of election studies in 63 consolidated and emerging democracies a stimulus paper, "The Comparative Study of Electoral Systems." 1 That paper identified several themes around which collaborative data collection might be organized, sketched a study design, and suggested how the planning process might unfold. A Steering Committee (comprised of Jacques Thomassen, Steven J. Rosenstone, Hans-Dieter Klingemann, and John Curtice) invited colleagues to comment on the paper and to participate in an initial planning conference held on August 20-21, 1994 in Berlin at the Wissenschaftszentrum Berlin fur Sozialforschung. Eighty-five social scientists from 44 different polities responded with comments and suggestions that were summarized in a second paper, "Comments on: 'The Comparative Study of Electoral Systems.'" 2 Colleagues from around the world expressed enthusiasm for this endeavor, endorsed the proposal for international collaboration, and indicated a willingness to join in the creation of a module of survey questions that would be asked in their respective national election studies.

Fifty social scientists, representing 31 consolidated and emerging democracies, participated in the Berlin Planning Conference. Discussion focused on seven topics: 3

The purpose of this report is to summarize the deliberations in Berlin as they pertain to the study design, study content, and study planning process.

The Structure of the Comparative Study of Electoral Systems

Deliberations in Berlin produced agreement on a set of principles, processes, and standards that would guide the Comparative Study of Electoral Systems:


The Comparative Study of Electoral Systems will focus on the nature of electoral choice in democratic polities (consolidated democracies, those undergoing transitions to democracy, and those that are recovering from recent democratic breakdowns).

Beginning in 1996, collaborators will include in their national election studies a module of common questionnaire content. The module will contain about 20 questions (about 10 minutes in length) and will be asked in its entirety in a post-election survey.

Beyond the module of common context, there may be a second, optional module containing additional survey questions.

Collaborators will also provide data on a (to be specified) set of demographic characteristics of respondents, coded to be agreed upon set of standards.

Collaborators will aspire to a (to be specified) set of scientific standards concerning sample quality, study administration, and the like.


All collaborators will be able to participate broadly in setting the substantive agenda for the study, in specifying the questionnaire module, and in specifying the demographic and macro- level data to be collected.

A Planning Committee will meet in Ann Arbor, Michigan (USA) in January 1995 and is charged with formulating recommendations on:

Members of the Planning Committee are: Rita Bajarunieni (Lithuania), John Curtice (United Kingdom), Juan Diez Nicolas (Spain), Oscar Hernandez (Costa Rica), Soren Holmberg (Sweden), Hans-Dieter Klingemann (Germany), Marta Lagos (Chile), Filipe B. Miranda (Philippines), Yoshitaka Nishizawa (Japan), Steven J. Rosenstone (United States), Jacques Thomassen (Netherlands), and Gabor Toka (Hungary).

In addition, several colleagues will serve as consultants to the Planning Committee: Gary Cox (University of California, San Diego), Herbert Kitschelt (Duke University and Humboldt University), Ekkehard Mochmann (Zentralarchiv fur empirische Sozialforschung), Richard Rockwell (Inter-university Consortium for Political and Social Research), Herman Schmitt (European Election Study), and W. Phillips Shively (Comparative Political Data Board).

Prior to its January 1995 meeting, the Planning Committee will solicit from all collaborators advice on the priorities that should be assigned to the study themes, questionnaire content, demographic variables, and macro-level data. The Planning Committee will also gather information on the study designs that are employed by election studies around the world.

Following its January 1995 meeting, the Planning Committee will circulate its recommendations to all collaborators to elicit comments and suggestions.

All collaborators will be invited to participate in a second Planning Conference that will be convened in the Spring of 1995 (before the pilot work) to further refine and reach consensus on the Planning Committee's recommendations.

The American National Election Study, with funding from the U.S. National Science Foundation, will provide the initial administrative support for the Comparative Study of Electoral Systems and for the Planning Committee's activities.


The Planning Committee will identify collaborators who are willing to pilot the questionnaire module and who will disseminate the pilot data in a timely fashion.

In the summer of 1996, pilot studies will be conducted in a diverse set of polities to test the module of survey questions.

The Planning Committee is responsible for analyzing the pilot data and for formulating recommendations for revisions of the questionnaire module. This analysis should be conducted during the fall of 1995. The Committee's recommendations should be circulated to all collaborators for comment. Final revisions to the questionnaire module should be completed by January 1996.


In each democracy, indigenous teams of researchers will conduct a national election survey that includes a common module of questions and demographic variables.

Each team of researchers will be responsible for securing funding to finance their national data collection, though the Planning Committee will make efforts to identify sources of support that may subvent the costs of individual or multiple data collections.

Teams of researchers will collect macro-level data. Some of these data may be collected in each country by the researchers collecting the survey data. Some of these data may be collected by separate teams of researchers working cross-nationally.


Each team of researchers will deposit its data in a central archive in a timely fashion in accordance with a (to be specified) set of standards.

Each team of researchers will provide documentation for its data as well as documentation about the sampling process, response rate, and study implementation.

Micro- and macro-level data from all polities will be merged into a single, cross-national data set.

Data will be placed in the public domain in a timely fashion.


The questionnaire module developed for the initial collaboration will be asked in national elections held between 1996 and 1999.

Planning for the next round of collaboration will begin in 1997. The second, and subsequent rounds may focus on a subset of the themes covered in the first collaboration, or it may focus on an entirely different set of themes.


The initial stimulus paper, written responses to that paper, and discussion at the Berlin Planning Conference identified three broad themes around which this collaborative effort might be organized:

The initial stimulus paper developed the first two themes in some detail (see pp. 1-8; 10-11). The third theme surfaced in several of the written responses to the stimulus paper (see pp. 10-11 of "Comments on . . .") and in discussion at Berlin. Few colleagues (either in their written comments or in Berlin) expressed much enthusiasm for collaborating around a fourth theme articulated in the initial stimulus paper -- the role that political parties play in encapsulating political conflict around economic, racial, ethnic, and environmental issues.

Discussion at the Berlin Planning Conference did not further develop the theoretical underpinnings of these lines of inquiry, nor was any effort made to reach consensus on the priority the Planning Committee should assign to each theme. Instead, conference participants elaborated on the kind of survey questions, demographic information, and macro-level data that might be developed in service to the three themes. The ideas to emerge from this discussion are as follows:


Demographic Characteristics of Respondents that Should be Coded in a Comparable Manner:



performance evaluation of:
- government
- the political system
- the way democracy works

legitimacy of :
- the political system
- the former political system
- the state

- state structure v. regime structure
- linkages to the political system
- trust in institutions (parties, church, etc.)
- tolerance for authoritarian government
- support for democracy
- trust in electoral commission
- attachment to the political system

Conference participants emphasized the need for instrumentation that will successfully distinguish among evaluations of the economy, government, regime, democratic structures, and parliament.


Attachment to parties
- party identification
- non-partisanship
- negative party preferences

Perceptions of political parties
- placement of all parties on likes/dislikes scale
- perceived closeness/distance between respondent and party positions (perhaps on the most important cleavage)
- placement of parties and respondent on left-right scale (or on some basic ideological division; or let respondent select the dimension)
- perceptions of party leaders
- electoral viability
- party strength
- groups to which parties are attached
- sources of conflict


Several Conference participants questioned whether we should focus on political parties alone or whether we should also focus on social movements.


- interest in politics (distinguish between local and national politics)
- political knowledge of:
- politics
- electoral system
- parties
- candidates
- political campaigns
- leaders
- meaning of democracy
- perhaps measure from "don't know" responses to other items in module
- sense of political involvement
- political efficacy
- media use
- perceived credibility of the media


- voter turnout
- activism (campaign and other forms of participation, peaceful and non- peaceful)
- mobilization to participate (how much and by whom)
- activities respondents say they would never take part in
- reasons for abstention

Vote Choice
- vote for parliament / president / legislature
- who respondent would never vote for
- meaning of vote choice
- strategic voting
- protest voting
- negative voting
- time of vote decision
- rationale for vote choice: personal characteristics of candidates v. party
- previous vote
- fears of reprisals
- split-ticket voting (and across levels)


Institutional Arrangements
- degree of centralization v. decentralization of the political system
- selection of the chief executive (parliamentary or presidential; type of presidential system)
- fiscal structures of local and central budgets

Electoral Laws
- number of deputies elected from respondent's district
- thresholds
- number of constituencies
- closed or open list; ease with which the voter can revise or create a voter list
- inclusiveness of the franchise
- are leaders running outside of parties?

Electoral Data
- district-level election results for all parties
- electoral results for upper house
- turnout at district level

Political Parties
- party programs
- parties that presented candidates in the respondent's district
- parties that are participating in the government
- parties' organizational strength in the district (e.g. number of party members)
- parties' electoral strength in the district (e.g. current and past vote in district)

General Overview of the Polity
- general perception of the political situation
- nature of macro-level cleavages
- symbolic positions of parties and other leaders on political cleavages and on nature of the political system
- religious, ethnic inequality

Economic Data

- unemployment data
- data should be at the district level, if possible
- fiscal structure of local v. central budgets
Other Data
- policy outputs
- conflict
- civic involvement (e.g. memberships; newspaper readership at the district level)

Issues Concerning the Collection of Macro-Data
- Which of these data are available from published sources?
- Which data should be gathered by the team of researchers conducting the national election survey and which data should be gathered through a more centralized cross-national team of researchers?

The Comparative Study of Electoral Systems
Overview of Activities, 1994-1996
(October 11, 1994)

August 1994:

Berlin Planning Conference Create Planning Committee

October 1994: Circulate Report Summarizing Berlin Meeting

October 1994 - December 1994:

January 26-29, 1995: Meeting of Planning Committees in Ann Arbor, Michigan

February 1995: Planning Committee's Recommendations Circulated for Comment

Spring 1995: Second Conference of all Collaborators to be Convened

Summer 1995: Pilot Studies Fielded

Fall 1995: Pilot Data Distributed, Analyzed, Pilot Study Reports Written and Circulated

January 1996: Set Final Study Specifications