You are here: Home / People / Benjamin Highton / home / Copenhagen2012 / datasets / cses2--codebook--part1--introduction


Plain Text icon cses2_codebook_part1_introduction.txt — Plain Text, 197 kB (202,618 bytes)

File contents



                          CODEBOOK: INTRODUCTION

                       FULL RELEASE - June 27, 2007

                             CSES Secretariat



The Comparative Study of Electoral Systems ( CSES MODULE 2
FULL RELEASE [dataset]. June 27, 2007 version.

These materials are based on work supported by the National Science
Foundation ( under grant numbers SES-0112029 and SES-0451598,
the University of Michigan, and the many organizations that fund election
studies by CSES collaborators.

Any opinions, findings and conclusions or recommendations expressed
in these materials are those of the author(s) and do not necessarily
reflect the views of the funding organizations.










    >>> WEIGHTS







The Comparative Study of Electoral Systems (CSES) is a collaborative
program of research among election study teams from around the world.
Participating countries include a common module of survey questions in
their post-election studies. The resulting data are deposited along with
voting, demographic, district and macro variables. The studies are then
merged into a single, free, public dataset for use in comparative study and
cross-level analysis.

The research agenda, questionnaires, and study design are developed by an
international committee of leading scholars of electoral politics and
political science. The design is implemented in each country by their
foremost social scientists.

By collaborating in this way, the CSES community hopes to forward
scientific inquiry into the relationship between electoral institutions and
political behavior.


CSES Module 2 was in use from 2001 through 2006. It focused on three key
theoretical questions. First, it examined the contrasting views of the
logic of elections - to what extent are elections a mechanism to hold
government accountable, as opposed to a means to ensure that citizens'
views are properly represented in the democratic process? Second, the
module added a new set of items on citizen engagement and cognition across
democratic polities. Third, the module expanded the analyses of the first
module to examine how voters' choices are affected by the institutional
context within which those choices are made.


The CSES Module 2 Planning Committee was responsible for the design of
CSES Module 2, and shared responsibility for its implementation. The
following persons were members of the CSES Module 2 Planning Committee:

W. Phillips Shively, chair
University of Minnesota
United States

Yun-han Chu
National Taiwan University

Gary Cox
University of California - San Diego
United States

John Curtice
University of Strathclyde
United Kingdom

Amaury de Souza

Juan Diez-Nicolas
ASEP, Complutense University

Yilmaz Esmer
Bolgazici University

Ashley Grosse
Director of Studies (-2001), CSES Secretariat
University of Michigan
United States

Soren Holmberg
Goteborg University

David Howell
Director of Studies (2001-), CSES Secretariat
University of Michigan
United States

Hans-Dieter Klingemann

Marta Lagos
Opinion Publica Latinoamericana

Radoslaw Markowski
Polish Academy of Sciences

Ekkehard Mochmann
Universitat zu Koln

Yoshitaka Nishizawa
Doshisha University

Virginia Sapiro
University of Wisconsin-Madison
United States

Hermann Schmitt
European Election Study and MZES

Jacques Thomassen
University of Twente

Gabor Toka
Central European University

Bernhard Wessels


The CSES Module 3 Planning Committee shared responsibility with the CSES
Module 2 Planning Committee for the implementation of CSES Module 2. The
following persons were members of the CSES Module 3 Planning Committee:

Professor Ian McAllister, chair
Director, Research School of Social Sciences
Australian National University

Professor Bernt Aardal
The Norwegian Election Studies

Dr. Kees Aarts
School of Business, Public Administration and Technology,
University of Twente
The Netherlands

Professor John Aldrich
Department of Political Science
Duke University
United States

Professor Ulises Beltr�n
Associate Professor, Divisi�n de Estudios Pol�ticos,
CIDE (Centro de Investigacion y Docencia Economica)

Professor Andr� Blais
D�partement de Science Politique, Universit� de Montr�al

Professor Yun-Han Chu
Institute of Political Science, Academia Sinica

Professor Russell Dalton
Department of Political Science
University of California, Irvine,
United States

Professor Juan D�ez-Nicol�s
ASEP (Analisis Sociol�gicos Econ�micos y Pol�ticos)

David A. Howell, ex officio
Director of Studies
CSES Secretariat, Center for Political Studies, University of Michigan
United States

Professor Ken'ichi Ikeda
Department of Social Psychology, The University of Tokyo

Christiaan Keulder
Institute for Public Policy Research (IPPR)

Professor Marta Lagos
Latinobar�metro, Opini�n P�blica Latinoamericana

Professor Radoslaw Markowski
Institute of Political Studies, Polish Academy of Sciences

Dr. h.c. Ekkehard Mochmann
Zentralarchiv f�r Empirische Sozialforschung, Universit�t zu K�ln

PD Dr. Hermann Schmitt
Mannheimer Zentrum f�r Europ�ische Sozialforschung (MZES)
Universit�t Mannheim

Professor Michal Shamir
Department of Political Science, Tel Aviv University

Professor Sandeep Shastri
Dean of Research and Social Science
International Academy for Creative Teaching,

Dr. G�bor T�ka
Political Science Department, Central European University

Professor Jack Vowles
Department of Political Studies, University of Auckland
New Zealand

Prof. Dr. Bernhard We�els
Wissenschaftszentrum Berlin f�r Sozialforschung (WZB)


The CSES project is extremely grateful to our Module 2 collaborators, who
raised their own funds to include CSES Module 2 in a nationally
representative post-election study in their country or province.

Listed collaborators are our contact persons for election studies that
appear in the CSES Module 2 full release dataset - they are not necessarily
the parties who collected the data.

Within each election study, collaborators are presented in alphabetical
order. The affiliations listed are current as of the the date when the
election study first appeared in the CSES dataset.

- Albania (2005)

Dr. Altin Ilirjani
Albanian Political Science Association /
University of North Carolina at Chapel Hill
United States

- Australia (2004)

Prof. Clive Bean
School of Humanities and Human Services
Queensland University of Technology

Dr. Rachel K. Gibson
ACSPRI Centre for Social Research (ACSR)
Research School of Social Science
The Australian National University

Prof. Ian McAllister
Research School of Social Sciences
The Australian National University

- Belgium (2003)

Prof. Jaak Billiet
University of Leuven

Prof. Lieven De Winter
University of Louvain La Neuve

Prof. Andre-Paul Frognier
University of Louvain La Neuve

Prof. Marc Swyngedouw
University of Leuven

- Brazil (2002)

Alberto Carlos Melo de Almeida
UFF-Universidade Federal Fluminense and
FGV-Funda��o Get�lio Vargas

Rachel Meneguello
Unicamp - Cesop
Cidade Universit�ria "Zeferino Vaz"

- Bulgaria (2001)

Radosveta Popova
Gallup International

- Canada (2004)

Andr� Blais
D�partement de science politique
Universit� de Montr�al

Joanna Everitt
University of New Brunswick - Saint John
Department of History and Politics

Patrick Fournier
D�partement de Science Politique
Universit� de Montr�al

Elisabeth Gidengil
Department of Political Science
McGill University

Neil Nevitte
Department of Political Science
University of Toronto

- Chile (2005)

Professor Marta Lagos
Latinobar�metro, Opini�n P�blica Latinoamericana

- Czech Republic (2002)

Lukas Linek
Institute of Sociology
Czech Academy of Sciences
Czech Republic

Zdenka Mansfeldova
Institute of Sociology
Czech Academy of Sciences
Czech Republic

Adela Seidlova
Czech Republic

- Denmark (2001)

J�rgen Goul Andersen
Aalborg Universitet

Jakob Rathlev
Aalborg Universitet

- Finland (2003)

Heikki Paloheimo
Department of Political Science
University of Tampere

Juhani Pehkonen
TNS Gallup

- France (2002)

Thomas Gschwend
Mannheimer Zentrum f�r Europ�ische Sozialforschung (MZES)
Universit�t Mannheim

Hermann Schmitt
Mannheimer Zentrum f�r Europ�ische Sozialforschung (MZES)
Universit�t Mannheim

- Germany (2002 Telephone)

Hermann Schmitt
Mannheimer Zentrum f�r Europ�ische Sozialforschung (MZES)
Universit�t Mannheim

Bernhard We�els
Wissenschaftszentrum Berlin f�r Sozialforschung (WZB)

- Germany (2002 Mail-Back)

Hermann Schmitt
Mannheimer Zentrum f�r Europ�ische Sozialforschung (MZES)
Universit�t Mannheim

Bernhard We�els
Wissenschaftszentrum Berlin f�r Sozialforschung (WZB)

- Great Britain (2005)

John Curtice
Department of Government
University of Strathclyde

Steve Fisher
Trinity College and Department of Sociology, Oxford
United Kingdom

- Hong Kong (2004)

LI Pang-kwong
Public Governance Programme and Department of Politics and Sociology
Lingnan University
Hong Kong

- Hungary (2002)

G�bor T�ka
Political Science Department
Central European University

- Iceland (2003)

�lafur Th. Hardarson
University of Iceland

- Ireland (2002)

Michael Marsh
Department of Political Science
Trinity College Dublin

- Israel (2003)

Asher Arian
Political Science, University of Haifa

Michal Shamir
Political Science, Tel-Aviv University

- Italy (2006)

Hans Schadee
Dipartimento id Psicologia Ateneo
Universit� delgi Study di Milano - Bicocca

Paolo Segatti
Dipartimento di Studi Politici e Sociali
Universit� di Milano

- Japan (2004)

Hiroshi Hirano
Gakushuin University

Ken'ichi Ikeda
The University of Tokyo

Yoshiaki Kobayashi
Keio University

- Korea (2004)

Hyung Joon Kim
Korean Social Science Data Center

Wook Kim
Department of Political Science
Paichai University

Nam Young Lee
Department of Political Science
Sookmyung Women's University

- Kyrgyzstan (2005)

Kusein Isaev
Center for Sociological, Politological, Social-Psychological Research
Bishkek Humanities University

- Mexico (2003)

Ulises Beltr�n Ugarte

Benito Nacif

Rolando Ocampo Alcantar

Olivia P�rez

- The Netherlands (2002)

Galen A. Irwin
Leiden University
Department of Political Science
The Netherlands

Joop J.M. van Holsteyn
Leiden University
Department of Political Science
The Netherlands

- New Zealand (2002)

Jack Vowles
The Department of Political Studies
University of Auckland
New Zealand

- Norway (2001)

Bernt Aardal
Institute for Social Research

Henry Valen
Institute for Social Research

- Peru (2006)

Catalina Romero
Social Sciences Department
Pontificia Universidad Cat�lica del Per�

David Sulmont
Social Sciences Department
Pontificia Universidad Cat�lica del Per�

- Philippines (2004)

Linda Luz B. Guerrero
Social Weather Stations

Vladymir Joseph Licudine
Social Weather Stations

Gerardo Sandoval
Social Weather Stations

- Poland (2001)

Krzysztof Jasiewicz
Department of Sociology, Washington and Lee University and
Institute of Political Studies, Polish Academy of Sciences
United States/Poland

Radoslaw Markowski
Institute of Political Studies
Polish Academy of Sciences

- Portugal (2002)

Antonio Barreto
ICS-UL, Instituto de Ci�ncias Sociais
Universidade de Lisboa

Andre Freire
ISCTE, Higher Institute for Labour and Business Studies
and ICS-UL, Social Sciences Research Institute
University of Lisbon

- Portugal (2005)

Ant�nio Barreto
Instituto de Ci�ncias Sociais

Andr� Freire
Instituto Superior de Ci�ncias do Trabalho e de Empresa/
Instituto de Ci�ncias Sociais
Universidade de Lisboa

Marina Costa Lobo
Instituto de Ci�ncias Sociais

Peter Magalh�es
Instituto de Ci�ncias Sociais/
Universidade Cat�lica Portuguesa

- Romania (2004)

Gabriel Badescu
The Center for the Study of Democracy
Babes-Bolyai University of Cluj

Andrei Gheorghita
The Center for the Study of Democracy
Babes-Bolyai University of Cluj

Paul Sum
Department of Political Science and Public Administration
University of North Dakota
United States

- Russia (2004)

Timothy Colton
Davis Center for Russian and Eurasian Studies
Harvard University
United States

Henry Hale
Department of Political Science
George Washington University
United States

Polina Kozyreva
JSC "Demoscope"

Michael McFaul
Hoover Institution
Stanford University
United States

- Slovenia (2004)

Janez Stebe
Arhiv druzboslovnih podatkov (ADP)
Faculty of Social Science - University of Ljubljana

Niko Tos
Faculty of Social Science - University of Ljubljana

- Spain (2004)

Juan D�ez Nicol�s
ASEP / Complutense University

- Sweden (2002)

S�ren Holmberg
Statsvetenskapliga Institutionen
Department of Political Science
G�teborg University

Henrik Oscarsson
Statsvetenskapliga Institutionen
Department of Political Science
G�teborg University

- Switzerland (2003)

Peter Selb
Institut f�r Politikwissenschaft
Universit�t Z�rich

- Taiwan (2001)

Department of Political Science
National Chung-Cheng University

- Taiwan (2004)

Department of Political Science
National Chung-Cheng University

- United States (2004)

American National Election Studies (ANES)
Center for Political Studies
Institute for Social Research
University of Michigan
United States

Nancy Burns
Center for Political Studies
Institute for Social Research
University of Michigan
United States

Russell Dalton
Department of Political Science
University of California, Irvine
United States

Donald R. Kinder
Center for Political Studies
Institute for Social Research
University of Michigan
United States

W. Phillips Shively
Department of Political Science
University of Minnesota
United States


While the CSES project and its governance are international in nature, the
CSES Secretariat for CSES Module 2 was housed in the Center for Political
Studies, Institute for Social Research, University of Michigan, United

Primary support for the CSES Secretariat was provided by the American
National Science Foundation (NSF) through two grants, with supplemental
funding from the University of Michigan:

1. Grant SES-0112029, "Political Science Research Infrastructure:
Comparative Study of Electoral Systems" with Principal Investigators
W. Phillips Shively (University of Michigan) and Nancy Burns (University
of Michigan) supported the CSES Secretariat for the period 2002 through

2. Grant SES-0451598, "The Comparative Study of Electoral Systems" with
Principal Investigators Nancy Burns (University of Michigan), Donald
Kinder (University of Michigan), and Ian McAllister (Australian National
University) provided support for the CSES Secretariat for the period
2005 through the release of this dataset, and partial support thereafter.

Professor W. Phillips Shively of the University of Minnesota, and
chair of the CSES Module 2 Planning Committee, oversaw operations of
the CSES Secretariat through 2003. Professor Ian McAllister of the
Australian National University, and chair of the CSES Module 3
Planning Committee, oversaw operations through the remainder of CSES
Module 2.

Various persons staffed the CSES Secretariat throughout the Module 2
period. Ashley Grosse was the Director of Studies during the design phase.
David Howell served as the Director of Studies beginning in December 2001.
Bojan Todosijevic, Angela Pok, Matias Bargsted, Karen Long Jusko, and
Michio Umeda provided research support and data management services.
Laurie Winslow, Carol Kent, and Tom Ivacko served as project administrator.
Laurie Pierson acted as webmaster, Patricia Luevano provided technical
support, and Kelly Ogden-Schuette acted as a research assistant. Sarah
Cookingham, Jennifer Dolfus, Daniel Grossman, and Justin Wilson were
student employees who provided research, technical, and clerical support.

And, while he is not a member of the CSES Secretariat, we would like to
thank James Wagner of Survey Research Operations (SRO) at the University of
Michigan's Institute for Social Research, for volunteering some of his
time to help us think through issues related to the weights.


The CSES module is intended to be administered as a single, uninterrupted
block of questions at the beginning or the end of a national election
survey. There are several features of this component of which analysts
should be aware:

A. The question text is included in the variable documentation of this
codebook. The questions are reported in the order in which they appear in
the CSES questionnaire.

B. Where there are known differences in the way a particular question was
administered in an election study, this is noted in the "Election Study
Notes" following the documentation of the corresponding variable.

C. There are several sets of party and leader evaluation items included
in the module. These correspond to parties labeled A-F, in descending
order of vote share, of the six most popular parties in the lower house
elections (or presidential elections if legislative elections were not
held). Where respondents were asked to evaluate other parties, these
evaluations have been included where possible and are labeled parties
G-I, regardless of their vote shares. The parties and leaders to which
these evaluations apply are identified in Appendix I.

D. There are several questions (including the vote-choice and party
identification items) which ask the respondents to specify a political
party. The codes for these items are also reported in Appendix I.


The district-level variables report the returns of the lower house (first
segment) election for each respondent's district. Wherever possible,
these data have been collected from official electoral commissions (see
Bibliography for details). In other cases, CSES has been grateful for the
compilations of these data provided by projects like "Political
Transformation and the Electoral Process in Post-Communist Europe" project
election database ( The following
district-level data are included in the CSES data file (Xs beneath the
first four columns indicate inclusion; if no district data are included,
the last column is marked with an X):

                         Seats  Candidates/   Vote    Turnout  No District
                                  Lists      Returns              Data

 Albania (2005)                                                     X
 Australia (2004)          X         X          X        X
 Belgium (2003)                                                     X
 Brazil (2002)             X         X          X        X
 Bulgaria (2001)                                                    X
 Canada (2004)             X         X          X        X
 Chile (2005)              X         X          X
 Czech Republic (2002)                                              X
 Denmark (2001)                                                     X
 Finland (2003)            X         X          X        X
 France (2002)                   [NA- Presidential Election]
 Germany (2002)            X         X          X        X
 Great Britain (2005)      X         X          X        X
 Hong Kong (2004)          X         X          X        X
 Hungary (2002)                                                     X
 Iceland (2003)            X         X          X        X
 Ireland (2002)            X         X          X        X
 Israel (2003)             X         X          X        X
 Italy (2006)              X         X          X        X
 Japan (2004)              X         X          X
 Korea (2004)                                                       X
 Kyrgyzstan (2005)                                                  X
 Mexico (2003)             X                    X        X
 Netherlands (2002)        X         X          X        X
 New Zealand (2002)        X         X          X        X
 Norway (2001)             X                    X        X
 Peru (2006)               X         X          X        X
 Philippines (2004)        X                             X
 Poland (2001)             X         X          X        X
 Portugal (2002)           X         X          X        X
 Portugal (2005)           X         X          X        X
 Romania (2004)            X         X          X        X
 Russia (2004)                                                      X
 Slovenia (2004)           X         X          X        X
 Spain (2004)              X         X          X        X
 Sweden (2002)                                  X
 Switzerland (2003)        X         X          X        X
 Taiwan (2001)                                                      X
 Taiwan (2004)                                  X        X
 United States (2004)      X         X          X


To supplement the micro (survey) data, the teams of researchers responsible
for the collection of the public opinion data also compiled and deposited
the following types of data: electoral legislation, political party
platforms, and official electoral returns.

To facilitate this process, a detailed questionnaire was constructed to
serve as a framework for the macro component of the project. The Macro Data
Reports, completed by the CSES collaborators, can be found on the CSES
website in the Module 2 section under "Download Data".

Additional measures thought pertinent to the micro-district-macro design
are also compiled and available in the CSES data files.

A bibliography of the sources consulted during the compilation of macro
data follows the main body of these introductory materials.



There are several components to the CSES documentation. Analysts will want
to become familiar with all of them.

For most election studies, collaborators have provided documentation
to accompany their election studies. These documents, where available,
can be found on the CSES Module 2 download page under "Data Center" on
the CSES website (

The CSES Module 2 questionnaire is also available from the website, or
by referencing the corresponding variables in this codebook.

The codebook consists of three files:

"CSES2_codebook_introduction.txt"  is the Codebook Introduction, the
                                   file you are reading now; it includes
                                   an overview of the study and data,
                                   information about use of the files,
                                   general election study descriptions
                                   and notes, and a bibliography

"CSES2_codebook_variables.txt"     is the Variable Descriptions file,
                                   including questions, code frames,
                                   general notes, and notes specific
                                   to an election study (by variable)

"CSES2_codebook_appendices.txt"    contains Appendix I (Parties and
                                   Leaders) and Appendix II (Primary
                                   Electoral Districts)

The last set of documentation you will want to become familiar with is
the CSES Module 2 errata page. It is accessible from the CSES Module 2
download page under "Data Center" on the CSES website. Information,
updates, and error notifications and corrections are posted there, in
real time, as they become available. Please check regularly for errata
notifications to keep up to date.


In the CSES Module 2 dataset, all variables begin with the letter "B".
This convention helps reduce the possibility of overwriting data when
merging with other CSES data sets.

Variables are presented in five groupings:

B1001-B1999 Identification, weight, and election study variables
B2001-B2999 Demographic variables
B3001-B3999 Micro-level (survey) data (the CSES Module 2 questionnaire)
B4001-B4999 District-level data
B5001-B5999 Macro-level data

In the Variable Descriptions portion of the codebook, the headers for
individual variables are surrounded by two lines of dashes. Variable names
do not exceed eight characters in length.

Most sections of the codebook can be navigated in the electronic
files by searching for the characters ">>>" or ")))" as appropriate.


The CSES Module 2 data files are distributed in a single ZIP file named
"". We recommend that users download this file to their
local computer in a directory named:


The last portion of the directory name ("20070627") indicates the version
of the dataset (2007, June 27). This arrangement allows the distributed
files to run without modification, and also sets your files up in a way
that is well-organized for future distributions of CSES data files.

Within "" are seven (7) files:

"cses2_codebook_part1_introduction.txt" is the codebook introduction file
"cses2_codebook_part2_variables.txt"    is the variable descriptions file
"cses2_codebook_part3_appendices.txt"   is the codebook appendices file
"cses2_rawdata.txt"                     is the raw data file
""      contains statements for reading the data into SAS
""     contains statements for reading the data into SPSS
""    contains statements for reading the data into STATA

You only need to unzip files for the statistical package you wish to
use (SAS, SPSS, or STATA). If you wish to use none of the three, the
"cses2_rawdata.txt" file is stored with a header statement and delimiters
(commas) so that you may read it into a software package of your choice.


In the "" file are six (6) files:

""     contains code labels
""        contains column locations
""        contains format statements
""    contains missing data statements
"" contains variable labels
""            is the file you execute to generate a SAS
                           system file

If you plan to use SAS, unzip these six (6) files into subdirectory"/sas/"
under the default directory "c:/cses/module2/20070627/".

You need only read into SAS the syntax file "". If you
downloaded the files to the recommended subdirectory, you should be
able to execute the file as is and a SAS system file will be
created in the same subdirectory.

The file "" refers to the other files above in the body
of the file - if you wish to exclude any of the other files from running
on your data set (for instance, perhaps you do not wish to have code
labels appear) you may comment out or delete the reference from the
"" file.

If you are storing any of the files in an alternate location (including
the data file "cses_rawdata.txt"), or you want to store the resulting
SAS system file under a different name or location, you will want to
modify the SAS files accordingly.

Additional instructions are found in header of the "" file.


In the "" file are seven (7) files:

"cses2_codelabels.sps"     contains code labels
"cses2_columns.sps"        contains column locations and variable labels
"cses2_formats.sps"        contains format statements
"cses2_missingdata.sps"    contains missing data statements
"cses2_variablelabels.sps" contains variable labels
"cses2_run.sps"            is the file you execute to generate an SPSS
                           system file
"cses2_dataset.por"        is an SPSS portable dataset

If you plan to use SPSS, unzip these seven (7) files into subdirectory
"/spss/" under the default directory "c:/cses/module2/20070627/".

To read in the dataset using the SPSS portable file "cses2_dataset.por":

If you prefer, you can import the "cses2_dataset.por" file into SPSS
directly by launching SPSS, selecting "File" then "Open" and then
"Data". In the file window that appears, change "Files of type" to
"SPSS Portable (.por)", find the file "cses2_dataset.por" and select
it, and then press the "Open" button. If SPSS successfully imports the
dataset, you do not need to make use of the remaining files in this

If you wish to read in an SPSS dataset using the provided syntax files,
or if the SPSS portable file does not work for you:

You only need to read into SPSS the syntax file "cses2_run.sps". If
you downloaded the files to the recommended subdirectory, you should be
able to execute the file as is and a SPSS system file will be created
in the same subdirectory.

The file "cses2_run.sps" refers to the other files above in the body
of the file - if you wish to exclude any of the other files from running
on your data set (for instance, perhaps you do not wish to have code
labels appear) you may comment out or delete the reference from the
"cses2_run.sps" file.

If you are storing any of the files in an alternate location (including
the data file "cses2_rawdata.txt"), or you want to store the resulting
SPSS system file under a different name or location, you will want to
modify the SPSS files accordingly.

Additional instructions are found in header of the "cses2_run.sps" file.

Running the SPSS run file (cses2_run.sps) in SPSS 14.0 successfully will
nonetheless produce warning message number #534 due to a small glitch in
this version of SPSS. This is nothing to worry about.


In the "" file are six (6) files:

""     contains code labels
"cses2_columns.dct"       contains column locations
""        contains format statements
""    contains missing data statements
"" contains variable labels
""            is the file you execute to generate a STATA
                          system file

If you plan to use STATA, unzip these six (6) files into subdirectory
"/stata/" under the default directory "c:/cses/module2/20070627/".

You only need to read into STATA the syntax file "".
If you downloaded the files to the recommended subdirectory, you should
be able to execute the file as is and a STATA system file will be
created in the same subdirectory.

The file "" refers to the other files above in the body
of the file - if you wish to exclude any of the other files from running
on your data set (for instance, perhaps you do not wish to have code
labels appear) you may comment out or delete the reference from the
"" file.

If you are storing any of the files in an alternate location (including
the data file "cses2_rawdata.txt"), or you want to store the resulting
STATA system file under a different name or location, you will want to
modify the STATA files accordingly.

Additional instructions are found in the header of the "" file.


This section provides information about the composition of identification
variables, the standard rules and exceptions for coding missing data, and
the composition of weights. In addition to the documentation included
in this section, please reference specific variable notes.


Each record in CSES Module 2 contains two key identification variables.

The first is ELECTION STUDY, which uniquely identifies an election study
across time. It appears in two variations:


The second is RESPONDENT, which uniquely identifies a respondent across


Notes on their creation and use are available in the variable descriptions
portion of the codebook.


(1) Researchers should be aware that sometimes there are multiple response
categories included in the code "missing" (the code "missing" is usually
indicated with a last digit of 9). For some election studies in which we
could not distinguish among various answers, the code "missing"
may include cases where respondents refused to answer the question,
"don't know" responses, and cases where there a particular question went
unanswered for other reasons.

(2) While CSES guidelines request that the response categories "Refused"
and "Don't Know" be volunteered responses, this was not always consistently
applied. For instance, sometimes the options were offered explicitly to
respondents in mail-back surveys, which do not have the benefit of an
interviewer being present. To identify whether the response options were
volunteered (or not) in a particular election study, please refer to the
original questionnaire provided by the collaborator. The original
questionnaires, where available, are accessible from the CSES Module 2
download page under "Data Center" on the CSES website.


Because of the variance in the study designs and methods used in the
election studies included in this project, the weights provided by the
collaborators also vary greatly. They are described in detail in variables
B1010-B1014. Analysts are advised to read the weight documentation
carefully to ensure that their analyses are weighted appropriately.

The CSES Module 2 Full Release includes original weight data, as
provided by the collaborators, where available. The original weight
variables include the following:


The remainder of the weight variables provided are derivatives constructed
from the original weights.

    B1012_1  >>> POLITY WEIGHT: SAMPLE


The majority of studies that comprise CSES are collected in countries that
have free or partly free elections. However, sometimes a collaborator will
include the CSES module in a study of a country that is a developing
democracy or that is considered not free. If the data collection is judged
to be of sufficiently high quality, the study is included in CSES datasets
even if the country is considered to be not free. The decision is left to
individual users as to whether such countries should be included in their
analyses of CSES datasets.

The CSES Module 2 dataset does not include freedom ratings. Freedom ratings
can be obtained from external sources, such as Freedom House, whose
website as of this dataset release is:

Freedom House is not affiliated with the CSES project.



The Full Release of CSES Module 2 contains data from the following
forty-one (41) election studies (in alphabetic order):

        Election Study                  Cases

        Albania (2005)                  1,116
        Australia (2004)                1,769
        Belgium (2003)                  2,225
        Brazil (2002)                   2,514
        Bulgaria (2001)                 1,482
        Canada (2004)                   1,674
        Chile (2005)                    1,200
        Czech Republic (2002)             948
        Denmark (2001)                  2,026
        Finland (2003)                  1,196
        France (2002)                   1,000
        Germany (2002 Mail-Back)        1,023
        Germany (2002 Telephone)        2,000
        Great Britain (2005)              860
        Hong Kong (2004)                  582
        Hungary (2002)                  1,200
        Iceland (2003)                  1,446
        Ireland (2002)                  2,367
        Israel (2003)                   1,212
        Italy (2006)                    1,439
        Japan (2004)                    1,977
        Korea (2004)                    1,500
        Kyrgyzstan (2005)               2,000
        Mexico (2003)                   1,991
        Netherlands (2002)              1,574
        New Zealand (2002)              1,741
        Norway (2001)                   2,052
        Peru (2006)                     2,032
        Philippines (2004)              1,200
        Poland (2001)                   1,794
        Portugal (2002)                 1,303
        Portugal (2005)                 2,801
        Romania (2004)                  1,913
        Russia (2004)                   1,496
        Slovenia (2004)                 1,002
        Spain (2004)                    1,212
        Sweden (2002)                   1,060
        Switzerland (2003)              1,418
        Taiwan (2001)                   2,022
        Taiwan (2004)                   1,823
        United States (2004)            1,066

        TOTAL                          64,256

For multi-wave panel studies, only those respondents who participated in
the wave of the study that included CSES Module 2 are included in the
CSES Module 2 dataset.


The following studies collected CSES data during the CSES Module 2 field
period, but were unable to be included in the CSES Module 2 Full Release
for the listed reasons. We are highlighting the studies here to promote
their efforts, and as a service to CSES users who might wish to contact
the collaborators directly to inquire about obtaining the data for use in
their own analyses.

If you are interested in obtaining these data, please contact the listed
collaborators directly. CSES is unable to distribute or support the files.

- Belarus (2006)

CSES Module 2 was included in a study after the 2006 election in Belarus,
under the direction of Principal Investigator Stephen White (University
of Glasgow). Fieldwork was carried out by the Center for Sociological and
Political Research at Belarusian State University, and produced 1,000
interviews. While a number of CSES Module 2 questions were present in the
study, not enough CSES Module 2 questions were present to allow for its
inclusion in the CSES Module 2 Full Release.

- Great Britain (2005 Internet Study)

The British Election Study (BES), under the direction of Principal
Investigators David Sanders (University of Essex), Paul Whiteley
(University of Essex), Harold Clarke (University of Texas at Dallas), and
Marianne Stewart (University of Texas at Dallas), included CSES Module 2 in
a Internet survey run by YouGov after the 2005 election in Great Britain.
A total of 3,326 interviews were collected. As of publication time, the
dataset was publicly available from the BES website at address:

The study has also been archived at the UK data archive at the University
of Essex ( as Study Number 5495,
"British Election Study, 2005: Comparative Study of Electoral Systems".

The BES survey that included CSES Module 2 drew respondents from a panel
built using a non-probability sample, not a probability sample, and thus
the study was unable to be included in the CSES Module 2 Full Release. It
did appear previously, however, in the fourth advance release of CSES
Module 2 (April 10, 2006 version).

It is worth noting that John Curtice and Steve Fisher arranged for CSES
Module 2 to be included as a self-completion supplement to the 2005
British Social Attitudes (BSA) survey. The BSA data appear in the CSES
Module 2 Full Release, and having being collected after the same election,
may be interesting to methodologists for mode comparisons with the BES
Internet effort.

- Thailand (2001)

CSES Module 1 was repeated in Thailand after their 2001 election, as
part of a face-to-face data collection that resulted in 2,000 interviews
under the direction of Principal Investigators Thawilwadee Bureekul (The
King Prajadhipok's Institute) and Robert Albritton (The University of
Mississippi). While some questions are in common, not enough CSES Module
2 questions were present in Module 1 to allow for inclusion of the study
in the CSES Module 2 Full Release.

- Ukraine (2006)

CSES Module 2 was included in a study after the 2006 election in Ukraine,
under the direction of Principal Investigator Stephen White (University of
Glasgow). Fieldwork was carried out by the Ukrainian partners of Russian
Research, and produced 1,600 interviews. While a number of CSES Module 2
questions were present in the study, not enough CSES Module 2 questions
were present to allow for its inclusion in the CSES Module 2 Full Release.


The following section provides:

(1) General information about the election after which the CSES Module 2
was administered. In several cases, the election followed the establishment
of new electoral rules. In several others, the election marks a dramatic
change in government. This information is provided with the intention of
alerting the analyst to interesting features of the election. For more
details, please refer to the Macro Reports prepared by the collaborators,
available on the CSES website. Unless specified, the information in this
section has been provided by the collaborator, the Macro Report, or the
Parline database (with permission).

(2) General information about the research and sample designs of the
component election studies. For example, in some cases, the CSES module was
administered in a later wave of a multi-wave study. Additionally, in
several countries, portions of the population were excluded from the sample
frame, usually because of geographic isolation.

(3) Additional information on survey weights, where available, is provided
for some election studies.

Please note that the number of "Election Summaries" will not match the
number of election studies, because sometimes multiple election studies are
conducted concerning a single election. Where multiple election studies are
conducted concerning a single election, they are grouped under a single


(1) The elections were considered a crucial step towards Albania's
accession to the North Atlantic Treaty Organization (NATO) and the European
Union (EU). The EU insisted that free and fair elections would be a
condition for the conclusion of a Stabilization and Association Agreement,
paving the way to EU membership. Both parties that have dominated Albanian
politics since the collapse of Communism in 1991, the PSS, led by Prime
Minister Fatos Nano, and the PDA of former President Sali Berisha,
supported joining the two organizations. They also called for strong ties
with the United States and supported Albanian troop deployments in Iraq,
Afghanistan, and Bosnia. But the parties proposed different policies over
tax reform. The PDA insisted on cutting taxes in half in order to promote
investment, which the PSS criticized as being irrational. Continuing
problems of corruption in the country were also an important issue in the
elections, and the PSS was criticized for not handling the problem
efficiently. The elections were monitored by 3,500 local observers together
with a total of 400 international observers. Less election-related violence
was reported than in previous elections. However, an election official was
shot dead in the capital Tirana on voting day, while another man was shot
dead outside the PDA's office. The man allegedly involved in this killing
was later killed in a reprisal shooting. The OSCE said the conduct of the
poll showed only limited progress since previous elections, criticizing
incorrect procedures related to the use of ink to prevent multiple voting,
the secrecy of the vote and verification of voters' identities. A
delegation from the European Parliament stated that voter lists were
"intentionally inaccurate", which thus disillusioned voters. Following
complaints filed with the Central Elections Commission (CEC), re-runs were
held in three single-member constituencies on 21 August. Following the
allocation of the proportional representation seats, the PDA and its allies
secured an absolute majority of 73 seats, while the PSS and its allies
totaled 64 seats.
(Source: PARLINE database:

(2) The 2005 Albanian post-election survey was administered through
face-to-face interviews to a representative sample of 1,500 households
and was conducted between July 5th and 20th, 2005. A stratified sample
selection model was employed whereby the country was divided into two
parts: i) the six main districts counting for 48 percent of the total
population of Albania, and ii) the rest of the country. In the six main
districts, seven to ten households were interviewed in each Primary Sample
Unit (PSU) selected for every 9,200 households sorted geographically from
North to South (7 or 10 interviews x 83 PSUs for every 9,200 households).
In the rest of the country, seven to ten households were interviewed in
each PSU (90) selected for every 7,520 households. The second household
member over 18 years old to appear to the interviewer was selected as a
respondent. The response rate was 76%.


(1) The election campaign was a tight contest between the Australian
Labor Party (led by Mark Latham, who had been elected its leader in
December 2003) and the conservative Coalition of the Liberal Party
and the National Party (led by Prime Minister John Howard). During
the election campaign, Howard concentrated on the two main issues of
his bid for re-election - economic prosperity (including the lowest
interest rates in a generation) and a steady hand in matters of
national security. Labor's campaign was centered on issues such as
creating a more equitable education system and making health care more
affordable. The bombing of the Australian Embassy in Jakarta the month
prior to the elections made national security a key issue in the campaign.
The two candidates had very different positions on the country's role in
the war on terror. Prime Minister John Howard had sent the third largest
combat force in the United States-led intervention in Iraq, while Latham
said the deployment had made the country less safe and promised that as
Prime Minister he would withdraw most of Australia's troops from Iraq by
the end of the year. The elections saw the Coalition increase its
majority in the House of Representatives with the Liberal Party obtaining
75 seats and its coalition partner, the National Party, obtaining 12
seats. The Australian Labor Party won 60 seats. Additionally, three
independents were elected.

(2) The 2004 Australian Election Study is the seventh in a series of
surveys, beginning in 1987, that have been timed to coincide with
Australian Federal elections. Data collection, using self-completion
(mail out/mail back) questionnaires, was administered between October 8,
2005 and February 15 2005. The survey is based on a stratified systematic
random sample. The sample of electors for all of Australia was drawn from
the Commonwealth Electoral Roll by the Australian Electoral Commission
following the close of rolls for the 2004 election. The Commission
supplied name and address information only, to be used solely for this
study. The sample was selected to be proportional to the population on
a state by state basis. Of a total mailing of 4,250, there were 1,769
completed returns, giving a raw response rate of 42%. An adjusted
response rate of 45% was calculated by removing the out-of-scope sample
(e.g., deceased, incapable, return to sender, n=275).


(1) In the May 2003 elections, Prime Minister Guy Verhofstadt's
centre-left coalition of six parties was primarily opposed by the
Christian Social Parties and far-right parties (Vlaams Blok and
National Front). Ten days before polling day, the French-speaking
Green Party quit the coalition over a conflict about overnight flights
over Brussels. The electoral campaign mainly focused on economic issues
such as tax cuts. With economic growth sluggish and unemployment high,
Verhofstadt vowed to keep cutting taxes and to reform state bureaucracy
and an overburdened judiciary. During the four years in government, the
coalition had passed some of the world's most progressive social
legislation, including legalizing gay marriage and euthanasia. The
Flemish far-right Vlaams Blok, desiring Flemish independence,
campaigned for zero tolerance on crime and for ending immigration.
It expected to improve its score in Antwerp, the country's second city,
where immigrants clashed with police in November 2002. The Liberal
Party and the Socialist Party registered gains in both parts of the
linguistically divided country. The Socialists made the biggest gains,
both in terms of seats and votes. The Christian-Socialists recorded a
slight decline, losing 3 seats. The Vlaams Blok won the largest vote in
its 25-year history winning 3 extra seats, taking its tally to 18 seats.
In the French-speaking southern region of Wallonia, the extreme-right
National Front scored 5.2 percent despite an almost invisible campaign
boycotted by public broadcasters. Support for the Greens, who had been
members of the outgoing coalition, collapsed, as they lost all nine
seats in Flanders and seven out of 11 in Wallonia.

(2) Face-to-face interviewing started on December 15, 2003, seven months
following the parliamentary elections. Interviewing concluded April 30,
2004. The survey was conducted in both Dutch and French. The sampling
frame excluded the German speaking region (population size approximately
72,000). A two stage sampling procedure was used. Primary units were
stratified by province, municipalities, selected according to their
population size. Secondary (or elementary) units were selected at random
within the primary units. Individual respondents were selected at random
from the national registry. Interviewers made four attempts to contact
each one of the respondents. The response rate was 63.4%.

(3) The CSES data file includes two weight variables. The first weight
variable is designed to match the known demographic distribution of
the population by region, age, gender, and education (the original
variable �AGE_Belg�). The calculation method used includes information
on population subclasses and is based on the joint distribution of age
(A), gender (G), and education (E) in the official statistics (NIS data,
2003). The seven age categories are 18-24; 25-34; 35-44; 45-54; 55-64;
65-74 and plus 74. The five education categories are �No Education or
Primary Education�; �Lower Secondary Education�; �Higher Secondary
Education�; �Higher Education� and �University�. The weighting variable
for Belgium was computed to ensure correct distribution between the
regions (Flanders, Wallonia and Brussels). This variable is included as
a demographic weight. The political weight contains weighting coefficients
for joint distribution of region, age, gender, education, and vote in the
2003 General Elections for the Belgian sample. Since the sample was drawn
in such a way that every person had the same chance of being elected,
there is no need to compensate for differential probability of selection;
hence, there is no sample weight.


(1) On October 6, 2002, Brazilians went to the polls to elect a new
President, along with all 513 members of the House of Deputies,
two-thirds of the 81 members of the Senate, and the state governors.
This was the first time in the country that the some 115 million
registered voters across the country could use a computerized voting
system in a presidential election, allowing at least 70 percent of the
vote to be counted four hours after the ballots close and the final
results to be released the following day. The elections were held
against a background of great uncertainty in the economic field and,
above all, in financial markets. Another issue raised during the
electoral campaign was violence, a topic which took on renewed importance
after authorities in Rio de Janeiro warned that criminal gangs had
threatened to disrupt the voting. Federal troops were sent to the city
in an effort to ensure safe voting after the electoral commission had
authorized the Defense Ministry to send reinforcements. The presidential
candidate leading the opinion polls was Luis Inacio da Silva, also known
as "Lula", the leader of the leftist Workers Party. The former metal-
worker had made unemployment his top campaign issue and promised to revive
the economy. About a quarter of Brazil's population live below the poverty
line and the country has the world's fourth worst income distribution. In
the presidential elections, none of the candidates won an outright
majority. Lula da Silva, who topped the poll, and Serra therefore faced a
second round on October 27, 2002. Lula da Silva won the second round with
60 percent of the votes. The left and centre-left candidates also
performed well in the elections for Congress and state governorships.

"The October 2002 elections were a milestone in Brazilian electoral
history. For the first time, a left party, the Workers Party (Partido
dos Trabalhadores, PT), won both the presidential election and a
plurality of seats in the Chamber of Deputies."
(Source: Nicolau, 2004, p. 338).

(2) The CSES survey was conducted between October 31 - December 28, 2002,
after the multiple elections held on October 6 and 27. The primary focus
of the CSES study was the parliamentary election for the Lower House.
The survey used face-to-face interviews, on a random sample of
2514 respondents, with a 72% response rate. To reduce costs of travel,
all municipios with less than 20,000 inhabitants were excluded. Also
excluded was 70 percent of rural households. Thus, 11.4 percent of
the municipios and 3 percent of the population were excluded from the
sample. The total percentage of the eligible population excluded from
the sample frame was about 4%. In the first stage of the sampling, 102
primary sampling units (PSU), or municipios, were selected with
probability proportional to size (PPS). Each PSU consists of a single
municipio as defined by IBGE (institute that conducts census in Brazil).
For the 1996 Contagem, IBGE divided Brazil into 5 census regions with
27 states and 5507 municipios. Twenty seven of these municipios were
selected with certainty as self-representing PSUs and 75 as non self-
representing PSUs. The self-representing PSUs are state capitals. In
the second stage, 280 secondary units (census tracts) were selected PPS
within each of the PSUs. In the third stage, households were selected
PPS within census tracts with one adult respondent being selected at
random. The person who had the date of anniversary closest to the date
of reference (October 27) was then selected to answer the questionnaire.
The survey used two versions of the questionnaire with small differences
(see the corresponding file in the download section of the CSES web site).

(3) One original weight variable was submitted with the Brazilian data
file. It compensates for both disproportional sampling in the state of
S�o Paulo as well as the demographic profile of the population of the


(1) The elections were held on June 17, 2001 according to a proportional
system with ballot lists of parties, coalitions, and independent
candidates registered in 31 multi-mandate constituencies. Thirty-six
parties and coalitions (out of a total of 64), as well as 11 independent
candidates, took part in the elections. Four political forces exceeded
the 4% barrier. This was the first time since the end of the communist
era that Parliament and Government completed their full four-year term.

(2) This is a post-election, face-to-face survey, conducted between July
5-15, 2001. The sample was designed to represent the national population
of voting age by region, urban/rural residence, gender, age and ethnicity.
Double-clustered random sampling was employed, with universe stratification
by region (28 regions in sum). As a basis for the sample, the database of
ESGRAON was used, which contains the full name and address of each
inhabitant. With this information, various other aggregates were later
superimposed, such as election precincts, settlements, municipalities, and
regions. For the purposes of the sampling design procedure, the following
steps were made: 1) universe was stratified by region (28 in total),
2) lists of election precincts in each region were prepared containing
the number of persons aged 18+ in each precinct, and 3) given the
targeted size of the final sample (1620), the number of individuals to
be interviewed in each region was determined directly proportional to the
relative share of the respective region within the universe. The final
stage relates to the random selection of sampling points used in the

The transition from "sampling point" to "persons with addresses" was
achieved through systematic stepped selection from the initial list
of individuals in each election precinct. The final product of this
procedure was a list of 1980 persons with their full three names and
precise addresses, that were grouped per sampling point, settlement
and region. There were 1,864 realized attempts at interviewing during
the fieldwork (94.14%). Effective interviews after quality control
procedures resulted in a sample size of 1,482. Three contacts were
made with the household before declaring it non-sample or non-interview.
Data weights are based on data provided by the National Statistical
Institute and adjust for gender, age, ethnicity, and type of residence
(capital city vs. city vs. town vs. village).


(1) On May 23, 2004, on the advice of the Prime Minister Paul Martin,
the Governor General dissolved Parliament, paving the way for general
elections on June 28, 2004. Martin was appointed to the office of
Prime Minister in December 2003 following his election as leader of
the Liberal Party of Canada and the retirement of Jean Chretien, who
had served as Prime Minister for ten years.

The Prime Minister was trying to take his Liberal Party to its fourth
straight general election victory. This party was challenged by the
Conservative Party, led by Stephen Harper, the left-wing New Democratic
Party (NDP) of Jack Layton, and the Bloc Qu�b�cois, a separatist party
that is strong in Quebec. For the first time since 1993, the country's
two strongest conservative parties, the Canadian Alliance and the
Progressive Conservatives, were united to contest the general election,
having merged into the new Conservative Party in October 2003.

The main issue of the 2004 campaign was the "sponsorship scandal" that
involved the disbursement by the government of $250 million in advertising
contracts over four years to Quebec advertising firms friendly
to the Liberal Party with a significant portion, of the amount, about
$100 million, paid as fees and commissions. Both the Conservative
Party and the Bloc Quebecois used this scandal against the Liberal
Party. Other major issues during the electoral campaign included
improvements in the government-administered healthcare system, taxes,
gun registration, homelessness, and the issue of same sex marriages.

Voter turnout was 60.9 percent, the lowest since 1898. The Liberals won
36.7% of the popular vote. That translated to 135 seats in the House of
Commons, a sharp decrease from the number the Liberals had won in three
landslide victories since 1993, but far more than the 99 seats won by the
Conservative Party. The big winner of the election was the Bloc Qu�b�cois,
whose win of 54 of Quebec's 75 seats was a considerable improvement over
its victory in 38 districts in 2000. The New Democratic Party obtained 19
seats while the last seat went to an independent candidate.

(2) The CSES survey was conducted as a part of the three-component Canadian
Election study: the Campaign-Period Survey (CPS), the Post-Election Survey
(PES), and the Mail-back Survey (MBS)). The CSES survey questions, apart
from the current and previous vote choice items, were part of the MBS.
Demographic variables are recorded in the CPS. The post-election survey
commenced one week after election on July 5 and ran to September 19, 2004.
The MBS, started about one week after the start of the PES, and
questionnaires returned to Institute for Social Research at York University
before the end of November were added to the data file. A rolling cross-
section sample release was employed for the CPS. Modified random digit
dialing (RDD) procedures were utilized to select households. Within
households, the birthday selection method was used to select respondents.
The sample for the CPS was designed to represent the adult population of
Canada: Canadian citizens 18 years of age or older who speak one of
Canada's official languages, English or French, and reside in private homes
in the ten Canadian provinces (thus excluding the territories). Because
the survey was conducted by telephone, the small proportion of households
in Canada without telephones were excluded from the sample population.
Residents of old age homes, group homes, and educational and penal
institutions were excluded from the sample. Statistics Canada estimates
that 3.7% of the private households in Canada do not have a residential
telephone number. To select individual survey respondents for the CPS, a
two-stage probability selection process was utilized. The first stage
involved the selection of households by randomly selecting telephone
numbers. The second stage of the sample selection process was the random
selection of a respondent from the selected household. To be eligible for
the interview the household member had to be an adult (18 years of age or
older) and a Canadian citizen. If there was more than one eligible person
in the household, the person with the next birthday was selected as the
survey respondent. The design of the CES included a longitudinal component
as the CPS respondents were asked to complete the PES and respondents to
the PES were asked to complete the MBS. There was no substitution of
respondents for those who declined to participate in the second or third
wave of the study. The number of completed interviews, the sample size,
for the three studies was 4,323, 3,138 and 1,674 respectively. The
response rate to the CPS was 55 percent and the reinterview rate for the
PES was 73 percent and 53 percent of the PES respondents (representing 39
percent of the CPS respondents) completed the MBS.

(3) In order to produce national estimates it is advisable to correct for
both the unequal probabilities of selection at the household stage and the
unequal probabilities of selection based on province of residence. The
included sample weight (B1010_1) represents the original National Weight,
constructed as the product of the household weight and the provincial
weight and should be used with the National Sample when national estimates
are required. The household weight compensates for the higher probability
of households with smaller number of adults to be included in the sample.
The Provincial weight compensates for over representation of the eight
smaller provinces and a corresponding under-representation reduction in
Ontario and Quebec. The weights are prepared for the entire Canadian
Election Study sample, i.e., not separately for the post-election and
mail-back data sets. The re-interview rates are reasonably high and sample
attrition between the surveys was not associated with household size or


(1) In the wake of the constitutional reforms of August 2005 which
abolished the seats reserved for non-elected senators, parliamentary
elections were held in parallel with the presidential election on 11
December 2005. The main issues at the 2005 elections were how to deal with
crime, unemployment, health, education, and poverty. One of the leading
presidential candidates, former PS defense minister, Ms. Michelle Bachelet,
vowed to implement greater regional trade integration and a better social
security and pension system. Her main rival, billionaire and former
senator, Mr. Sebasti�n Pi�era of the RN, pledged to modernize the national
police force and introduce tough policies on crime. He distanced himself
from his partner in the APC coalition, Mr. Joaquin Lav�n, leader of the
Independent Democratic Union (UDI). The latter was also a presidential
candidate. The UDI promised to create one million new jobs by 2011 and
combat poverty and inflation. The CPD fared well in both polls. It obtained
the majority in both chambers of parliament for the first time since the
return of democracy in 1990, winning 11 of the 20 renewed seats in the
Senate. It thereby increased its number of seats to 20 out of 38. In the
Chamber of Deputies, it secured 65 of the 120 seats, including two
independents allied to it. The number of seats going to the Alliance for
Chile (APC) and allied independents was reduced to 54 in the Chamber of
Deputies and 17 in the Senate. In the presidential run-offs held on 15
January 2006, the CPD candidate, Ms. Michelle Bachelet, secured 53.5 per
cent of the poll, becoming Chile's first woman president.
(Source: Parline)

(2) The Chile 2005 Election Study was conducted between January 1st and
8th, 2006 as a post-election study using face-to-face interviews. The
sample frame included citizens eligible to vote living in urban areas only
and excluded two low density regions of the country (Regions XI and XII).
Around 20% of the population were excluded from the sample frame. The
sample selection was applied in three stages: a) Random selection of
subdistritos (geographical units with approximately two thousand
inhabitants); b) Random selection of houses in the subdistritos; and c)
After the house was selected, the person was randomly selected through the
last birthday method. Three contacts were made with the household before
declaring it non-interview.

(3) The study has one multiplicative weight variable designed to match
the distribution of sex and education in the population and the results
of the first round of the presidential election.


(1) The two main parties involved in the 2002 election were the governing
Czech Social Democratic Party (CSSD) and the main opposition, the centre-
right Civic Democrat Party (ODS) of the former Prime Minister Vaclav Klaus.
During the electoral campaign, both parties focused their programs on
accession to the EU and on promises of strong economic performance. The
main obstacle politicians had to overcome before the 2002 elections was
voter apathy. According to analysts, one of the main reasons for this
apathy was the very few new faces contesting the polls. In fact, turnout
was only 58 percent, much lower than the 74 percent registered in the 1998
elections. In the election to the Chamber of Deputies, the leftist Czech
Social Democratic Party (CSSD) was returned with the largest number of
seats for the second consecutive time. The party failed to secure an
overall majority, however, winning only 70 of the 200 seats in the Chamber,
four less than in the previous 1998 elections. The opposition, the Civic
Democratic Party (ODS) won 58 seats, 5 less than the 63 it secured in 1998.
The Communist Party of Bohemia and Moravia (KSCM) was the only party to
increase its representation, taking 41 seats (17 more than in 1998). The
centre-right Coalition, formed by the Christian Democratic Union-Czech
People's Party (KDU-CSL) and Freedom Union-Democratic Union (US-DEU),
obtained 31 seats, eight fewer than in 1998.

(2) The Czech post-election study was conducted between June 24 and
July 1, 2002 with face-to-face interviews. The sample was designed as
a regular continuous monthly survey of CVVM (Centrum pro vyzkum
verejneho mineni - Centre for public opinion polling). Also included
are respondents between 15-17 years of age. Yet respondents less than
18 years of age are not included in the CSES sample, as they are not
allowed to vote. The study used the quota sampling design. Respondents
were interviewed in their homes. The only requirement for inclusion in
the sample frame was that a person lives in a house or flat. Homeless
people thus were not intentionally excluded. The Quota Sample was
designed, using census data about the structure of the population in
terms of education, sex, region and age. Quotas were designed for each
region (total of 8 regional units). Individual respondents were
identified by interviewers, who selected respondents in the county
area based on fulfilling the quota. The response rate was 52.8%. Note
that respondents who refused to take part in an interview were not
asked again or otherwise persuaded; other people were chosen instead.
The number of refusals reflects the number of people who refused to
be interviewed and were replaced by other respondents.


(1) The Danish 2001 elections brought an end to the Danish Social
Democratic - Radical Liberal coalition government that was in power
since 1993. Voters� swing to the right was such that, for the first
time since 1929, the right obtained a clear majority, so the Liberals
and Conservatives were enabled to form a government without support
from any of the small centre parties. What is interesting is that
the defeat of the governing parties took place "against the background
of an economic miracle that had reduced unemployment by more than 60%,
a significant increase in private and public consumption, and resolution
of the state deficit problem" (Anderson 2003, p. 186). The most decisive
issue was not the war against terrorism, or even welfare, but rather,
the immigration issue. All parties on the right, not only the two
populist parties but also the Liberals and the Conservatives, demanded
tighter rules, especially concerning the family reunion and restricting
the access to social security. Welfare issues were also seen as salient,
but the important change in 2001 was that the social democrats were
perceived as less competent than the bourgeois parties.

(2) The Danish Election Study used face-to-face interviews
conducted between December 7, 2001 and March 14, 2003. The sample is
based on a random systematic selection with clusters of households,
where the primary sampling units were households. Percentage of the
eligible population excluded from the sample frame was 2.6%.
Respondents within a household were selected according to the
birthday criteria. The response rate was 58.7%.

(3) The study does not have weight variable(s). "The data [were]
representative, so there was no need for weights. (This was proved
by actually making weights, which turned out to make no difference)."


(1) This was the first election held under the new constitution of 2001.
The 2003 electoral campaign focused on unemployment, tax cuts, the
country's possible entry into NATO, as well as the criticism of the
government's stance on Iraq by the opposition Centre Party. The leader of
this party, Anneli Jaatteenmaki, accused the four-party ruling coalition of
doing too little to push for a peaceful resolution of the Iraq crisis,
while Prime Minister Paavo Lipponen declared that Finland wanted a
UN-brokered solution. The Prime Minister's Social Democrats narrowly lost
the elections to the Centre Party, as the latter gained just two more seats
than Lipponen's but not an absolute majority. The Centre Party enjoyed the
greatest gain in voter support, obtaining 24.7 per cent of the vote (55 of
200 seats), while the Social Democratic Party won 24.5 per cent with 53
seats. Following government coalition talks between the Centre Party, the
Social Democratic Party, and the Swedish People's Party, a new Centre
Party-led Cabinet was sworn in on 15 April 2003, with Anneli J��tteenm�ki
as the new Prime Minister, making her the first woman Prime Minister in the
country's history.

(2) Face-to-face interviewing started on March 17, 2003, the day following
the parliamentary elections, and ended on April 30, 2003. The survey was
conducted in both Finnish and Swedish. The primary sampling units were
based on two dimensions. The North-South dimension grouped the
municipalities into three groups: South, Central, and Northern Finland.
The urban-rural dimension was used to form two groups: 1) urban
municipalities and 2) semi-urban and rural municipalities. By linking these
two dimensions, six primary level strata were obtained. The Capital Region
(Helsinki, Espoo, Vantaa, Kauniainen) formed an individual stratum. The
second stage stratification grouped area zip codes within the first stage
strata, so that each second stage stratum contained an approximately
similar number of people. Inside each stratum, clusters (zip code areas)
were picked using PPS-sampling. Inside each picked cluster the same sample
size (6 interviews) was used. Inside each sampled cluster (zip code area),
a starting point (address) was chosen randomly. Six subsequent interviews
were made using random walking from each randomly sampled starting address.
From every target household, a person was selected who met the requirements
for sex and age criteria. The original sample was N=1270 which included an
over-sample (134 cases) of the Swedish speaking Finns. In the final dataset
the number of cases is 1196, of which 5% (60 cases) are Swedish speaking
Finns. The over-sample was corrected by dropping out randomly 74 Swedish
speaking Finns (using SPSS SAMPLE command).


(1) Results of the first round of the French 2002 election came as a shock
to many observers. The general expectation was to see Jacques Chirac (The
Rally for the Republic) and Lionel Jospin (The Socialist Party) in the
second round. Due to Jospin's poor showing and the widespread splintering
of the left-wing vote in the first round of the election, Jean-Marie Le Pen
(Front National) faced Chirac in the second ballot. As a consequence,
Jospin, who was also the outgoing Prime Minister, resigned as head of the
Socialist Party and did not stand in the subsequent legislative elections.

For many voters the choice between Chirac, who was at the time under
investigation for actions carried out while he was mayor of Paris and
who was benefiting from Presidential immunity and Le Pen, an extreme
nationalist, was rather difficult. Although Chirac defeated Le Pen by a
landslide, the victory was probably based more on the fear of a victory for
an extreme right leader than enthusiastic support for the incumbent. The
presidential election of 2002 signified the rise of the right wing in
French politics.

(2) The CSES survey was conducted on May 23-24, 2002, after the
presidential election (April 21 and May 5, 2002) and before the
parliamentary elections (June 9 and 16, 2002).

This was a telephone survey (CATI system) that included unlisted telephone
numbers in the population sampled. It is estimated that approximately 4% of
households were without a telephone and, hence, they were excluded from the
sample frame. The response rate was 43%. Primary sampling units were
randomly selected from national telephone numbers, with selection in
further stages according to a quota method, using gender, age, occupation,
city size, and region. Five contacts were made with the household before
declaring it non-interview. There was a maximum of two re-contacts used to
persuade respondents to be interviewed.


(1) The electoral campaign focused on the disastrous floods that had hit
the eastern and northern parts of Germany. Analysts believed that
Chancellor Gerhard Schroeder's confident handling of the crisis helped
the Social Democratic Party (SPD) to recover the Germans' confidence as
expressed in the opinion polls. Other major issues in the agenda during
the electoral campaign were unemployment and the state of the country's
economy. Edmund Stoiber, leader of the Christian Social Union (CSU),
presented himself as the man who could reproduce Bavaria's economic success
at the national level. The elections were the first in German history to
feature televised debates between the two main candidates for the post of
Chancellor. Foreign policy also dominated the speeches of the two
candidates and, in particular, Germany's relations with the United States.
Gerhard Schroeder picked up votes by speaking out against U.S. plans for
military action against Iraq. However, he was criticized by conservatives
for wrecking the U.S.-German relationship. The final results gave the SPD
251 seats, 16 less than in the outgoing legislature. The CDU increased its
representation by ten seats, reaching 190, while its partner, the Christian
Social Union of Bavaria (CSU) won 58 seats, 15 more than in 1998. The Green
Party took 55 seats, 12 more than in the previous elections. The Free
Democratic Party (FDP) obtained 47 seats, while the Party of Democratic
Socialism (PDS) obtained only 2 seats, a net loss of 30. A few weeks after
the elections, the two parties in the ruling Red-Green coalition (the SPD
and the Green Party) reached an agreement on the new coalition.

(2a) Telephone survey
This is a post-election survey, focused on the Bundestag elections of
September 22, 2002. The interviews were conducted by telephone, between
October 21 and November 12, 2002. Some 4% of the eligible population was
excluded from the sample frame. Furthermore, the new Laender (East Germany)
was over-sampled. The survey includes 1007 respondents from East Germany
and 993 from West Germany. The sample can be analyzed separately for East
and West, or, if appropriately weighted, for Germany, as a whole.

The sample is a single-stage random household sample. An initial sample
was drawn from the Infratest Telephone Household Master Sample (ITMS)
which comprises a multi-stratified, largely unclustered sample that is
distributed in proportion to the number of private households in micro-
cells, thereby compensating for regional or local differences in the
telephone density of households. The sample frame covers all listed and
unlisted numbers in Germany. A second sample for the German post-election
study was drawn from a pool of respondents interviewed before the election
who agreed to participate in further studies. Respondents were interviewed
from mid-August, 2002 up to the election on September 22, 2002. 85.6% of
the second sample was willing to provide another interview. The response
rate of the first wave of the panel study was 66.3%.

(2b) Mail-Back survey
This study is part if a Pre-Post Election Panel Study. The survey was
applied as a self-completion supplement drop-off to the face-to-face
pre-election survey. The interviews were completed between September 23 and
October 31, 2002. The primary sampling units were randomly selected, and
the individual respondents were identified using the Birthday rule. Total
panel attrition was 49 percent.

(3) Before the federal election of 22 September 2002, the standard number
of seats in the German federal parliament, the Bundestag, was reduced from
656 to 598. To achieve this, the boundaries of the 167 constituencies were
redrawn. As a result, there were 299 constituencies in 2002, 29 fewer than
in 1998.


(1) On April 4, 2005, Prime Minister Tony Blair called a general election
for 5 May 2005 after asking Queen Elizabeth II to dissolve Parliament. At
this election, the statutory number of members was decreased to 646 from
659, due to changes in constituency boundaries in Scotland where there are
now 59 constituencies instead of 72. During the campaign, Blair asked
voters' support for eight years of economic growth and low unemployment
rate. The Conservative leader, Michael Howard, accused the Labour
government of broken promises and insisted that his party would focus on
the issues that matter to hard-working Britons, promising a �4 billion tax
cut. He also criticized Labour's immigration and health care policies.
The Conservatives also emphasized the crime issue, with the idea to link
immigration, terrorism, and crime together in the public mind. The Liberal
Democrat leader promised to address people's hopes, not playing on people's
fears. He said his party would be the real alternative, promising to
replace the council tax with a local income tax and to provide free
personal care for the elderly. Blair was criticized by both opposition
parties, and also within his own party, over Britain�s role in Iraq. The
final results gave Blair a historic third term in office, though the party
lost 56 seats compared to the previous election. It also won the lowest
share of the vote for a ruling party since the Great Reform Act of 1832:
35.2 percent. The Conservative Party increased its number of seats by 33
to a total of 197 by securing 32.3 percent of the vote, merely 2.9
percentage points less than Labour. The Liberal Democrats won 62 seats,
11 more than in 2000.
(Source: PARLINE database:

(2) The CSES module was implemented between June 20 and November 24, 2005,
as a self-completion post-electoral supplement to the annual British Social
Attitudes (BSA) survey. This self-completion supplement was left behind to
a sub-sample of BSA respondents by the interviewer at the end of a face-to-
face interview, and picked up later. The sample frame is intended to cover
Great Britain (Northern Ireland was excluded and as well as Scotland north
of the Caledonian Canal). The exact percentage of the eligible population
excluded from the sample frame is not available, given that the sample
frame - the Post Address File (PAF, a list of addresses maintained by the
Post Office)- includes the people that needed to be excluded for CSES
purposes. Nonetheless we know that the proportion of the population
excluded is more than 3.1%. Estimates based on the 1991 Census Validation
Study suggest that PAF has a 96.4% coverage of households and a 96.9%
coverage of individuals. The primary sampling units - postcode sectors-
were randomly selected and stratified by region, population density and
percent owner-occupiers. Within each selected postcode sector, addresses
were drawn using regular intervals and a random start. At selected
addresses, the interviewer selected one person to be interviewed. The
weight factor corrects for respondents at multi-household addresses and/or
in large households which had less chance of selection than respondents
at single household addresses and/or in small households. The response
rate was 40%.

(3) There is one weight variable, which adjusts for a) unequal selection
probability of respondents, b) non-response rates and c) demographic
features (gender, age and region). The BSA weights were created using the
2004 mid-year population estimates from ONS/GROS. The weight factor was
estimated considering the total sample of 4,268 individuals, from which
CSES is a random sub-sample (with 860 respondents).


(1) The Hong Kong 2004 Legislative Council Election was held on September
12, 2004, which returned 30 members from the five geographical
constituencies and another 30 members from the 28 functional
constituencies. This was the first general election held after the March of
July 1, 2003, in which about half a million Hong Kong people marched to
demand more democracy and protest against the resumption of the second
reading of the National Security Bill. The 2004 election was fought between
the pan-democrats and the pro-government (the leftists and the
conservatives) camps. After its victory in the 2003 District Council
Election held on November 23, 2003, the pan-democrats camp had planned but
failed to get hold of a majority of seats in the Legislative Council.
However, the number of seats that the pan-democrat camp commanded has
increased from 21 in 2000 to 25 in 2004.

(2) The Hong Kong 2004 Election Study was conducted between October 5 and
December 19, 2004, as a post-election study based on face-to-face
interviews. The sample frame included citizens eligible to vote (registered
electors). A pool of 4,000 household addresses was provided by the Census
and Statistics Department (CSD), Hong Kong Special Administrative Region
Government. According to the CSD, household addresses were selected by
systematic replicated sampling from its Register of Quarters. 2,500
household addresses were then randomly selected from the pool by the
investigator. Interviewers then visited the selected households to see if
there was a registered elector(s). Within the selected household, one
respondent was selected randomly among the registered electors by using
the Kish Table. The response rate was 39.2%.


(1) Outgoing Prime Minister Viktor Orb�n, who led the center-right Fidesz-
MPP party, focused his electoral campaign on the country's strong economic
growth as well as on his nationalist agenda, which gives perks to ethnic
Hungarians living abroad. One of the most important components of this
agenda is "the reunification of the nation across the borders". This
nationalism was highly criticized by the opposition Socialists, who
contested that it harms Hungary's image abroad. They accused the Prime
Minister of being populist.

In the first round, more than 71% of the country's 8.1 million eligible
voters cast ballots, an absolute record. The results for this round showed
that the Socialists won 42.1% of the votes, against 41.1% for the governing
coalition of Fidesz-MPP and MDF.

In the closely fought contest, voter turnout was even higher in the second
round elections than in the first round. Moreover, the results of the
second round of parliamentary elections produced a shift in power. The
coalition of opposition Socialists and Liberals managed to narrowly defeat
the ruling conservatives of Prime Minister Viktor Orb�n. Fidesz-MPP and
MDF, two of the three government parties before the election, ran a joint
list and joint candidates in every single-member district. While these two
parties together won more seats than the socialist MSZP, they formed
separate factions in the new parliament. Thus, MSZP ended up as the single
biggest party in the new parliament; hence, there was no serious dispute
about which party's prime ministerial candidate should be asked by the head
of state to make the first attempt at forming a new government after the

(2) The survey was conducted between April 11 and 19, 2002, between
the two rounds of the 2002 Parliamentary elections. The Central Statistical
Office�s census list served as the starting point for sample construction.
Localities were stratified by county and population size. Within each
stratum, primary sampling points were selected at random with a probability
proportional to their population size. A random route procedure was
followed to select households. A Kish Table was used to select respondents
within households. The response rate was 60.5%.


(1) Before the election, the country was governed by a centre-right
coalition government formed by the Independence Party and the Progressive
Party. The Independence Party leader and Prime Minister, Davi Oddsson, who
had been in power for twelve years, was strongly opposed to EU membership.
The Progressive Party, on the other hand, was in favor of EU membership. In
his campaign, Oddsson focused on his economic record and offers of tax
cuts. He had the advantage of having provided several years of steady
economic growth. Inflation and unemployment were also low, compared to
elsewhere in the European Union and the Nordic countries. The opposition
Social Alliance, headed by former Reykjavik Mayor Ingibjorg Sorlun
Gisladottir, had pushed a campaign based largely on voter fatigue and a
growing awareness that Iceland must do more to prevent a growing gap
between rich and poor. The Alliance also capitalized on criticism that the
Independence Party had moved too slowly to reform a complex system set up
in the 1980s, of distributing fishing quota licenses among Iceland's
fisheries. However, the Social Alliance got less mileage out of its
campaign to hold a referendum on whether Iceland should join the European
Union. Oddsson's conservative Independence Party won 22 of the 63 seats at
stake, just two more than the rival Social Alliance. The Independence Party
remained the biggest force in Parliament despite losing four seats. The
junior partner in the outgoing coalition, the Progressive Party, kept all
of its 12 seats.

(2) This is post-election, telephone election study, conducted between May
15 and June 29, 2003. The sample was randomly selected from The National
Register (on which the electoral register is based). People under 18
years of age on election day and people over 80 years of age were excluded.
At least 10-20 attempts at contacting respondents were made over maximum
45 days before declaring a household a non-interview. The response rate
was 64.3%.


(1) The outgoing legislature was the only one to have served its full term
of office since 1943. The three-week electoral campaign focused on social
services, as the government had been criticized over the state of health
and education. Another issue was law and order, since the death of two
police officers at the hands of joyriders in April 2002. The government
was also credited with popularity as it was a signatory to the Good Friday
Agreement of 10 April 1998, aimed at bringing peace between Protestants
and Roman Catholics in Northern Ireland.

The final results showed that the Prime Minister's Fianna F�il party won
80 of the 166 seats, while the main opposition, the Fine Gael party,
obtained 31 seats, 23 less than in the outgoing legislature. The third
force in the House, the Labour Party, won 21 seats (an increase of four
seats). The Progressive Democrats also performed well, doubling their
representation to eight seats. The Green Party increased its presence in
the newly elected House of Representatives, obtaining six seats. The Sinn
Fein, the Irish Republican Army-linked party, won five seats, a large
increase from the previous one seat obtained in the outgoing legislature.

(2) This is a post-election, face-to face study, focused on the
parliamentary elections of May 17, 2002. Persons living in institutions,
approximately 2.7% of the population, were omitted from the sample frame.
A three-stage clustered sampling approach was used. In the first stage,
a random sample of PSUs was selected. In the second stage, a random sample
of households  was selected. In the third stage, a random person within a
household was selected. The sampling frame used for this study was the
most up-to-date national electoral register. Electors are recorded in the
electoral list in so-called Polling Books. For sample selection purposes,
these polling books were reconstituted into area units known as District
Electoral Divisions. There is a total of 3,400 District Electoral
Divisions (DEDs) in Ireland. These DEDs are the most spatially
disaggregated area units in Ireland for which census data are available.
Furthermore, they are the standard PSU building block for random sample
selection. Once the Electoral Register was re-structured into the
District Electoral Division structure, a random sample of 220 PSUs was
selected. Each PSU was made up of the District Electoral Division or
aggregate thereof using a minimum population threshold criteria. A sample
of 25 addresses was selected from within each of the 220 PSUs. This was
the second stage of sample selection. Households were randomly selected
from within the PSU. Individuals were then selected from within
households using the 'next-birthday' rule. This was the third stage of
sample selection. The response rate was 60%.


(1) Elections were held for all the seats in Parliament following its
premature dissolution in November 2002. General elections had previously
been held in May 1999. In November 2002, Prime Minister Sharon dissolved
parliament and called early elections, nine months before the scheduled
date, after the Labour Party, his largest coalition partner, withdrew
from government in a dispute over the 2003 budget. The general elections
were preceded by party elections to choose a leader for the Labour Party
in November 2002 and a leader for the rightist Likud Party one month later.

The electoral campaign focused on security issues, as the elections were
held against the background of the conflict with the Palestinians, the
potential U.S. attack on Iraq, and the global "War on Terrorism".
Voter turnout was close to 68.5%, an all-time low. The previous record
for low voter turnout was set in 1949, when 75% of all eligible voters
cast ballots.

Official results showed that Ariel Sharon was the first incumbent Prime
Minister to win re-election since the 1980s. His Likud Party won 37 seats
while the Labour Party had its worst showing ever, losing seven of the 26
seats it had picked up in the last Knesset elections in 1999, when it
became the largest single faction in Parliament. The Shinui Party became
the Knesset's third largest party, more than doubling its strength from
six seats in the outgoing Parliament to as many as 15 in the incoming one.

(2) The Israeli CSES module was applied immediately after the Knesset
elections of January 28, 2003. It was a telephone survey that included
only listed phone numbers. It is estimated that approximately 8% of
households do not have telephones; as a result, they were excluded from
the sample frame. The response rate was 13% (52% refused, in 27% of cases
contact was not established, and 8% were non-sample lines). Three
contacts were made with the household before declaring it non-interview.


(1) The outgoing President of the Republic, Mr. Carlo Azeglio Ciampi,
dissolved parliament on 11 February 2006 and called parliamentary elections
for 9 and 10 April 2006. Because of changes to the electoral law, the
revised proportional representation system favored the formation of
coalitions, and the provision relating to constituencies for Italians
abroad (electing six senators and 12 deputies) came into effect for the
first time. Two coalitions dominated the 2006 elections: the House of
Freedoms, led by Italy's longest-serving Prime Minister, Mr. Silvio
Berlusconi, and the Union, which incorporated the Olive Tree coalition.
Mr. Romano Prodi, former prime minister and former president of the
European Commission, was chosen in October 2005 to lead the Union
coalition. The economy and the presence of Italian troops in Iraq were
the main issues on the agenda during the election campaign. Prime
Minister Berlusconi had pledged a stronger economy under his leadership.
However, the GDP growth rate stagnated at 0.8 per cent per year and youth
unemployment remained high. During the election campaign, he promised to
abolish property tax, while his rival Mr. Prodi promised to boost the
economy by reducing labor costs and the country's deficit. Both leaders
pledged to withdraw Italy's troops from Iraq by the end of 2006.
Allegations of corruption and bribery related to the Prime Minister's
private business interests formed a backdrop to the campaign. Although
the share of the vote was very close, the electoral system provides a
bonus for the party or coalition with the highest score. Therefore,
after including the results from the constituencies abroad, the Union
won 348 seats in the Chamber of Deputies, while the House of Freedoms
held 281 seats. In the Senate, the Union narrowly won with 158 seats of
the 315 elective seats, only two more than the House of Freedoms.
(Sources: PARLINE database:

(2) The Italy 2006 Election Study was conducted between May 10th and 16th,
2006, as a post-election study using face-to-face interviews. The
sample frame included citizens eligible to vote. The sample frame
excluded residents of the Trentino Alto Adige region. Also excluded were
people who work in advertising or marketing, and those who were surveyed
during the last six months. Consequently, an estimated 5% of the eligible
population was excluded from the sample frame. Only one contact was made
with each household before declaring it non-sample or non-interview.

(3) The study includes a demographical weight variable based on population
distribution which adjusts the sample distribution of gender, level of
education and geopolitical Cattaneo areas. The employed reference to design
this weight was the Italian census of 2005.


(1) The July 11, 2004 elections were held to renew half of the House of
Councilors, the upper house of the Parliament. Opinion polls had shown
that Koizumi's Liberal Democratic Party (LDP) could lose several seats in
the upper house as his coalition was under criticism for its changes to
the pension systems and for deploying Japanese troops to Iraq.
Nevertheless, the election would not affect control of Parliament because
the ruling coalition had a majority in the lower house. During the
electoral campaign, Prime Minister Koizumi talked about his conservative
administration's economic policies and said that, under the leadership of
his Liberal Democratic Party, the world's second largest economy had
recovered to a degree better than had been forecast. On the other hand,
opposition Democratic Party (DPJ) president, Katsuya Okada, appealed to
voters for a change in leadership. The opposition was looking for gains
in the House of Councilors, appealing to voters angry about higher
premiums and benefits cuts in the pension system and public opposition to
Japan's military deployment in Iraq. Final results showed that the ruling
Liberal Democratic Party had won 49 seats, one fewer than its 50 up for
re-election, while New Komeito gained one seat to take it to 11. This
gave the ruling coalition 139 of the 242 seats in the upper house. The
DPJ increased its representation, winning 50 seats, up from the 38 it
previously held.

(2) This survey was a part of a large panel survey (Japan Election Study 3:
2001-2005 (9 waves)). The CSES Module 2 was part of the post-election wave
that was conducted between July 15 - 26, 2004. The survey is based on
face-to-face interviews. Sampling was based on the voting registry. The
voting registry contains all the eligible voters over 20 years old.
Included in the sample frame was name, gender, and date of birth. Once the
samples were selected, their contact addresses along with names, gender,
and age were copied in the sample master file. Primary sampling units were
Electoral districts "chiten". The chiten (precincts) were selected by PPS
selection mechanism, where the probabilities of the particular chiten�s
selection are proportional to the relative size of the chiten. From each
chiten, about 16 samples were selected with systematic selection, as all
the eligible sample are contained in the list, systematic selection of
sample resulted in EPSEM sample from each PSU. Response Rate: 56.6%

(3) The original sample was selected with EPSEM design, therefore no sample
weights were created, however due to the disproportionate attrition by PSU,
the entry of new sample in latter waves caused non-EPSEM sample, therefore,
this difference in selection probability were reflected as a sampling
weight. It was not attempted to match sample demographic with known
population estimates of the universe for the following reasons: 1> Often,
during the years between each census (e.g., 5 years) the estimates from
population registry are used as the estimates of the universe (this survey
uses both population registry and the voting registry as sample frame.).
Therefore the sample frame used in this study is comparable to a population
frame; 2> Only a few studies examining coverage error of the voting
registry and population registry were conducted in the 1970s at which point,
the coverage error was minor. No other study was conducted ever since.
Response weights were made by means of propensity weighing  adjustment,
using the information gathered in previous waves of the survey. As there
are multiple waves of surveys in JESIII, the dataset were first
re-constructed into one record per wave format (like that of survival data)
and logistic model was fit, using response to each wave as a dependent
variable and using various measurements from previous waves as predictors.
As a result, the weighting did not try to correct/match distribution of the
sample to known universe (i.e., census) but they referred to the
distribution of variables in previous waves. The variables included in the
propensity mode were, gender, age group (in category), college education as
a dummy variable, income as categorical variable, home ownership as a dummy
variable, employment status as a dummy variable, and summary statistics of
"don't know" responses.


(1) The electoral campaign got under way after months of political
confrontation that culminated on March 12, 2004 with the impeachment
of President Roh Moo Hyun. Roh's powers were suspended until a decision
by the Constitutional Court was taken. Three opposition parties, the
conservative Grand National Party (GNP), the former ruling Millennium
Democratic Party and the right wing United Liberal Democrats, pushed the
impeachment vote in the National Assembly to oust the reformist President,
on charges of election law violations, incompetence and corruption. The
President's impeachment, the first one since the founding of South Korea
in 1948, polarized the country's politics and sent tens of thousands of
demonstrators into the streets condemning the vote by the National
Assembly as a "parliamentary coup". On 14 May 2004, Roh was reinstated
after the Constitutional Court rejected two of the impeachment counts
against him and ruled that the third one (violation of electoral
neutrality) was insufficient grounds for his removal from office. Two
months later, on May 20, the President joined the Uri Party.

Ahead of the general elections, the National Election Commission (NEC)
issued a series of new regulations in order to ensure a free and fair
exercise. Final results showed that the pro-government Uri Party obtained
152 seats of the 299 in the National Assembly. The Uri Party's emergence
came at the expense of the main opposition Grand National Party, which
will no longer command a simple majority as it obtained only 121 seats.
The Progressive Democratic Labour Party made its debut with 10 seats,
in sharp contrast with the Millennium Democratic Party, which was reduced
to a minor party with a mere 9 seats.

(2) This is a post-election, face-to-face study, conducted between April
15 and 22, 2004; that is, immediately after the Parliamentary election
that took place on April 15, 2004. Sampling procedure was based on
randomly selected electoral districts. For each electoral district,
a quota of respondents was then determined, reflecting the whole
population distribution in terms of age and sex (quota sampling). The
sample frame excluded Cheju island and other small islands since people
in these regions, who comprise a very small percentage of the total
population, are difficult and costly to contact.

(3) "The election was conducted under a new electoral system, a mixed-
member majoritarian system that combines 243 single-seat districts (SSDs)
with 56 proportional representation (PR) seats, elected from a single
nation-wide district. Each voter casts two votes, one for an individual
in the SSD tier, and one for a closed party list in the PR tier."
(Cho, 2005, p. 526).


(1) Parliamentary elections were held in 27 February 2005. The Kyrgyz
opposition attempted to unite ahead of the poll, forming electoral blocs
and pledging to work together to ensure free and fair elections. The
opposition parties criticized the first round election results and called
for the cancellation of the parliamentary elections and the holding of
early presidential election. They held demonstrations in central Bishkek
and in the south of the country. The runoff elections were again marred
by accusations of electoral fraud. Opposition parties criticized the
disqualification of a number of opposition candidates. Opposition leaders,
supported by thousands of voters, continued to call for the entire
parliamentary elections to be declared null and void. They also established
a coordinating committee to control the situation in the country, and
demanded that President Akayev resign and early presidential and
parliamentary elections be held. Protest rallies had spread from the
capital Bishkek to many other provinces by 24 March 2005. The government
headquarters in the capital as well as state administrations in a number
of provinces were occupied by protesting voters. On 24 March 2005,
President Akayev fled Bishkek to Moscow, and Prime Minister Nikolai Tanayev
submitted his resignation to the Parliament. Parliament held an emergency
session on 24 March 2005, and appointed the head of the opposition
coordinating committee, Mr. Kurmanbek Bakiyev, as acting prime minister. He
then formed an interim government. The new parliament held its first
session on 27 March 2005, and elected Mr. Omurbek Tekebayev as speaker. Mr.
Akayev resigned on 4 April 2005. Early presidential elections were held on
10 July 2005 and Mr. Kurmanbek Bakiyev, who had been the acting head of
State since 24 March 2005, was elected President.
(Source: PARLINE database:

(2) The survey was conducted in February and March 2006, using face-to-face
interviews in the Kyrgyz and Russian languages. This CSES study
is primarily focused on the early Presidential election of July 10, 2005.
Earlier in 2005, there were also Parliamentary elections (a first round
conducted on February 27, and a second round on March 12, 2005). Sampling
was based on combination of random route and quota sampling. Sample
stratification was based on 7 geographic regions (oblast) and the capital
of Kyrgyzstan. The number of interviews was defined for each region in
proportion to the population size. In each sampling units quotas were
established based on gender, age, level of education, employment and
distribution of economics branches. Households were selected according to
the random route method. In each household, only a single person was

(3) Note that in items that code parties from A to F (or I) (e.g., B3037,
B3038, B5011, etc.), Kyrgyz parties are not ordered according to the
election returns, but according to their popularity as expressed in the
CSES party identification item B3029_1. The Kyrgyzstan CSES election study
is focused on the early presidential election of July 10, 2005, but the
candidates basically did not represent parties that respondents were asked
about in the aforementioned items. As a result, the list of the parties
is not applicable to the presidential election results items (B5005).
These items list election results of the candidates that participated in
the presidential election (see the corresponding election study note).


(1) Elections were held for all the seats in the Chamber of Deputies when
the members' terms of office expired.

Opinion polls showed that while President Fox's approval ratings remained
high, many voters across the country were disillusioned by his failure to
fulfill the promise to create millions of jobs -- a promise he had made
when he was elected in July 2000. Voters were also reported to be
disappointed by his government's inability to push his main reform
proposals through a divided Congress, where the opposition Institutional
Revolutionary Party (PRI) was the largest party in both Houses, and had
stymied the President's proposals for changing Mexico at almost every turn.
Mr. Fox's party, the second largest in the Chamber since July 2000, with
206 seats, was unable to pass bills such as opening the electricity
industry to more private investment and levying a tax on food and medicine
to increase spending on social services. President Fox and his aides
acknowledged the lack of new jobs but argued that they had done all they

The PRI had held uninterrupted power in Mexico for 71 years, until Mr. Fox
won the presidency in 2000.

In the run-up to the election, Fox traveled all over the country announcing
new joint private-public investment projects to spur growth, while his top
aides met with U.S. officials to announce a number of programs to
encourage U.S. investment in Mexico's most job-poor areas. Other major
campaign issues revolved around election financing scandals involving both
major parties.

For the 2003 elections, only 42% of the 64 million eligible voters cast
votes -- a record abstention rate. This was seen as a clear sign of
voters' disenchantment with the biggest political parties.

(2) This is a face-to-face, post-election study focused on the Mexican
legislative elections of 2003. The primary sampling units were electoral
precincts clusters. The clusters were defined as groups of all of the
precincts with similar electoral results and belonging to the same county
(municipality). The second stage was the precincts within the clusters.
The third  stage was households within the precinct and the last stage was
the respondent in the selected household. Respondents in the household
were not selected randomly. Interviewers followed quotas of gender and age,
taken from 2000 census data. Non-sample replacement method followed the
same systematic selection used for the original sample line. Three contacts
were made before declaring a sample line non-sample and/or non-interview.
One recontact was used to persuade respondents to be interviewed. The
response rate was calculated to be 52%. Sampling units were selected with
unequal probabilities.


(1) The 2002 election campaign revolved about the single person - Pim
Fortuyn - a publicist and columnist, a newcomer on the Dutch political
scene. His exceptional success at the municipal elections before the
parliamentary elections demonstrated the electoral impact he could have
in parliamentary elections. The two major coalition parties, PvdA and
VVD, had intended to base their campaign on the strong economic
performance of the purple coalition. However, Fortuyn managed to shift
the campaign agenda and to turn the election into a referendum on the
coalition�s policies on immigration, health, crime, and education. The
strongest impact on the outcome of May 2002 elections had Pim Fortuyn�s
assassination just 9 days before the election. The campaign was
immediately suspended until after the funeral. Thousands paid their
respects, and the ceremony was broadcast live on television. Thereafter,
campaigning was kept very low key, especially by PvdA and the Green Left,
which had been accused of 'demonizing' Fortuyn, so contributing to a
political climate that brought about the murder. The List Pim Fortuyn,
with 26 seats, made the best showing ever of a new party, ending up as
the second largest party. The two parties that had long competed to
become the largest party, VVD and PvdA, ended up in a close contest in
third and fourth place.
(Source: Adapted from Irwin and Van Holsteyn, 2004)

(2) In the Dutch 2002 survey, the CSES module was included in a post-
election panel, conducted immediately after the election, between May 16
and June 27, 2002. It is a face-to-face survey, but in the post-election
wave, 374 persons were given the opportunity to answer the questions on a
paper questionnaire, a computer diskette, or via Internet. 65 percent of
these 374 respondents chose one of these options; 35 percent preferred a
face-to-face interview. The primary sampling units were telephone
connection (fixed lines). Each connection was checked to see if it belonged
to a household. The household then became a sampling unit. Each household
was sent a letter emphasizing the importance of participation in the survey
and invited to participate in the survey. The response rate for the first
wave was 28%, while the panel attrition for the second wave was 17.5%.

(3) The study has two multiplicative weight variables, constructed by
using the computer program Bascula. The demographic weight includes the
characteristics sex, age, marital status, size of household, region of
residence, and degree of urbanization. The political weight includes
voting behavior as an additional population characteristic.


(1) Elections were held for all the seats in the House of Representatives
following the dissolution of Parliament. The Prime Minister, Helen Clark,
called an early election at a time when support for the Labour Party, in
opinion polls, exceeded 50% and the Alliance, the junior coalition
Government party, was disintegrating. The election campaign was less in
the mold of the traditional contest between the two main parties, Labour
and National, than a preoccupation with which of the smaller parties would
support a main party in forming a government. The same number of parties
were returned to the House but in rather different configurations. The
senior government party, the center-left Labour Party, increased its
representation in the House by 3 seats, to 52 seats total. Yet, it was
insufficient for an outright majority in order to govern alone. The main
opposition party, the center-right National Party, obtained 27 seats, 12
fewer than in the 1999  election and the lowest result since the party's
formation in 1938. The populist New Zealand First Party more than doubled
its seats, winning 13, 8 more than in the 1999 elections. The center-right
United Future New Zealand Party, a merger of United and Future New Zealand
(the latter formerly the Christian Coalition Party), made an unpredicted
strong showing, gaining 8 seats. The Labour Party entered into a
coalition government agreement with Jim Anderton's Progressive Coalition
Party on August 8, 2002.

(2) The 2002 NZES deposit to the CSES was based on a postal self-
completion questionnaire. The sampling frame was the electoral rolls
(containing an estimated 94% of New Zealanders of voting age). Samples
were drawn randomly from each of 69 electorates, the numbers sampled
being proportionate to the numbers of names listed, except in the case
of the seven Maori electorates, which were over-sampled.

(3) New Zealand used a self-administered mail questionnaire, with "Don't
Know" explicitly listed for many questions.


(1) The turnout for the Storting 2001 elections was the lowest since 1927,
with about 75% of the eligible voters casting votes.

Results showed that the Labour Party suffered its worst election in 77
years, losing 22 seats, while the Socialist Left got the best results ever,
gaining 14 new seats. The far-right Progress Party made a gain of 6,
up from the 20 it had in the outgoing legislature.

The electoral campaign focused on two issues: taxation and the state of
public services. While the outgoing government promised continuity and
focused on the fact that the United Nations had proclaimed Norway the
country with the best standard of living in the world, the opposition
parties declared that government should use the country's oil wealth to
cut taxes and improve health and education.

On 17 October 2001, the Labour Prime Minister, Jens Stoltenber, tendered
his government's resignation. This paved the way for Kjell Magne Bondevika,
the leader of the Christian People's Party, to head a right-wing coalition
composed of the Conservatives, the Christian People's Party and the Liberal
Left. Between the three they held a total of 62 seats. The backing from
the far right would yield a total of 88 seats, five more than the required

(2) The study was designed as a pre-election and post-election panel study.
Most of the demographic variables were asked in the pre-election survey.
The respondents were randomly selected from population registers in
selected primary sampling units. Three contacts were made with the
household before declaring it a non-interview. Three recontacts were used
to persuade respondent to be interviewed. Total panel attrition was 4%.


(1) "On 9 April 2006, parliamentary elections were held in parallel with
presidential elections. Poverty alleviation was a key issue in the 2006
elections, in a country where over 50 per cent of the population lives
below the poverty line, despite economic growth averaging 4.5 percent
in recent years. The final parliamentary election results showed a major
defeat for the ruling Per� Posible Party, which won only two seats, a
sharp decrease from 45. The biggest winner was the UPP, which took 45
seats, up from six previously. The APRA and the National Unity Party also
made gains, winning 36 and 17 seats respectively. The Alliance for the
Future took 13. Ms. Keiko Fujimori was elected with the highest individual
score nationwide. In the presidential race, Mr. Garc�a of the APRA
defeated Mr. Humala in run-off elections on 4 June 2006. After the
elections, the Centre Front, the Per� Posible Party and the National
Restoration formed a new party in Congress and allied to the APRA."
(Source: Parline)

(2) This study was conducted as a post-election face-to-face survey over
the period May 18th to May 21st, 2006. The primary sampling units
were housing blocks in urban areas, and the village in rural areas, all
of which were randomly selected. In each block a predefined number of
houses were selected using random systematic jumps. Individual respondents
were selected by the proportional equivalent system using official
demographic information. The response rate was 38.7%, All interviews were
conducted in Spanish.

(3) Demographic weights are calculated according to the proportion of the
population in the regions where the sample was taken. They match the
spatial distribution of the population among the regions. Political
weights are calculated according to the official results of the first
round of the presidential election.


(1) On May 10, 2004, Philippine voters went to polls to elect the
president, vice-president, twelve senators, and all the members of the
House of Representatives, together with 17,000 local government posts.
Incumbent President, Gloria Arroyo, was looking to win her first popular
mandate, as she had inherited the presidency in 2001, when Joseph Estrada
was deposed by street protests. Under the Constitution, an elected
President cannot run again, but Arroyo was vice-president under President
Estrada and completed his term. Her main rival in the race was Fernando
Poe Junior, the country's best-known film star, who was allegedly backed by
Joseph Estrada. There were three other runners: Raul Roco, a former
education secretary, Panfilo Lacson, the former Police Chief, and Eduardo
Villanueva, an Evangelical preacher. The election battle was fought mainly
on personality and ignored the important current issues in the country such
as poverty, an underperforming economy, a looming debt crisis, Islamic
militant terrorism, Muslim and Communist insurgencies, and birth control.
Arroyo presented herself as a "safe pair of hands" for the economy and was
backed by the business sector, while Poe ran on his film star reputation as
"Mr. Clean", adored by the poorest classes because of his screen roles
playing an underdog superhero battling for the oppressed. The elections
were marred by numerous complaints that many people had been left off the
voters' lists or could not find their voting precincts. There were also
many reports of violence during the campaign. The released figures showed
that 141 people were killed and 192 wounded in election-related violence
since December 2003. The removal of former President Joseph Estrada in a
military-backed uprising was the main source of bitterness that had divided
the country. About 230,000 police and troops were deployed at polling
stations throughout the country in an effort to contain the violence.
According to reports in the local press, some voters were handed envelopes
of cash or free lunches if they voted for certain candidates. Other voters
complained that the indelible ink aimed at preventing double voting washed
off with suspicious ease. Poe's supporters protested in the streets
disputing the fairness of the election. Opposition candidates also claimed
that the extremely slow manual counting of votes, which was completed about
a month after the election, was also another avenue for cheating. Reacting
to criticism by international observers, President Arroyo conceded there
were flaws in the elections, but declared the problems were isolated and
warned the opposition against stirring up trouble. Official results showed
that of the 12 Senate seats at stake, the ruling K4 coalition (Coalition
for Truth and Experience for the Future, led by the President's Lakas-
Christian Muslim Democrats) obtained 7. It also won 184 of the 212 seats of
the House of Representatives (Lakas-CMD 93, the Nationalist People's
Coalition 54, the Liberal Party 34, Kampi 3) while the opposition Laban ng
Demokratikong Pilipino (LDP) obtained 11 seats.

(2) This post-election face-to-face survey was conducted between August 5
and 22, 2004, in five languages: Tagalog, Cebuano, Ilocano, Ilonggo, and
Bicolano. For sampling purposes, the Philippines was divided into four
study areas: National Capital Region (NCR), Balance Luzon, Visayas, and
Mindanao. The sample size for each of the four study areas was 300
voting-age adults. Multi-stage probability sampling was used in the
selection of sample spots (60 spots per area). The three study areas, apart
from NCR), were further divided into regions, and then provinces were
selected from the regions according to the probability proportional to
population size (PPS) of the region. Once the sample provinces were
selected, 60 spots for each of the major areas were allocated among the
sample provinces. Using the quota set for each spot in each region, the
spots were distributed in such a way that each province was assigned a
number of spots roughly proportional to its population size. For the third
stage, within each sample spot, five households were established by
systematic sampling (random route procedure). The individual respondents
are identified using a probability selection table or the Kish grid.

(3) The weights are used to match the major study areas (National Capital
Region, Balance of Luzon, Visayas, Mindanao areas) with the distribution
of the adult population given recent official (census) projections.


(1) One of the main subjects of the 2001 electoral campaign was the bad
economic situation. Unemployment had jumped from 10 to 16 percent in the
year prior to the elections and the government budget deficit had
ballooned, forcing leaders to search for deep cuts and depressing

The economic development that the country had witnessed in the immediate
post-communist period had benefited mainly the major cities. Outside the
city center, small farmers have had to deal with rising costs and foreign
competition, while steelworkers, factory workers, and coal miners face big
job cuts.

(2) This is a post-election survey, focused on the elections of September
23, 2001, when the new Sejm and Senat (Lower and Upper house) were elected.
The sample used in this survey was a random sample, representative of the
adult population of Poland. The sample was selected by applying a multi-
level sampling scheme with stratifying on the first level. First level
sampling units were statistical regions created by the Central Statistical
Office (GUS) for census purposes and representative studies. Second level
sampling units included households in the previously sampled regions. The
sampling process was carried out by means of specially prepared computer
software, adequate to the adopted scheme, using generators of random
numbers of monotonic distribution. First and second level units were
sampled from the data sets about regions and districts, belonging to GUS.
Third level sampling units were adults who permanently inhabit selected
households. The interviewer selected one person as a respondent by means
of the "Kish method" among the inhabitants of a given household.

The study was conducted from September 29th to October 11th, 2001.
Announcement cards were sent by CBOS prior to interviewing to selected
addresses. Initially, 3240 addresses were selected. 1794 interviews were
carried out and sent for analysis, which accounts for about 55.4% of the


(1) The March 2002 elections were held for all seats in the Assembly
following the premature dissolution of this body in December 2001. General
elections had previously been held in October 1999.

On March 5, 2002, the official campaign for the general election on March
17, 2002 opened, with economic issues at the fore. The election, the second
in the country in two and a half years, followed the sudden resignation of
the Socialist Prime Minister, Antonio Guterres, after his party's heavy
defeat in local elections in December 2001.

During the campaign, both parties, the incumbent Socialist and the main
opposition party, the center-right Social Democrats (PSD), focused on
stabilizing the State budget, introducing corporate tax reductions, raising
productivity and reforming the education system. Additionally, the Social
Democrats and the Popular Party agreed to form a coalition government.

(2) CSES Modules 1  and 2 were part of the same election study for
Portugal 2002. This is a post-election study focused on the parliamentary
elections in 2002. The survey was conducted between March 23 and April 8,
2002, immediately after the elections held on March 17, 2002. The sample
frame was restricted to mainland Portugal (Azores and Madeira Islands were
excluded from the sample frame), and included approximately 95.4% of the
eligible population.

The sample was stratified by NUTS II (5 regions in the mainland: North,
Centre, Lisbon and Tagus Valley, Alentejo and Algarve) and HABITAT
(eleven categories of localities by number of inhabitants). For each cell
within the NUTS II and HABITAT frames, and according to the proportion of
residents in each cell, the number of interviews was defined. Then, the
number of localities inside each cell of the same frame was randomly
selected, trying to ensure that no more than 10 interviews were done in
the same locality. In each locality, the method of random route was used.
In the household, respondents were selected using the following criteria:
next person living in the household to have his/her birthday. If, on the
first attempt to find someone in the housing unit, no one was there,
no callback was done to this unit - rather, it was immediately substituted
by another one. Three contacts were made with the household before
declaring it a non-interview. The response rate was 81.4%.


(1) After dissolving the unicameral 230-member Parliament (Assembleia da
Republica) on December 10, 2005, Portuguese President Jorge Sampaio called
parliamentary elections for February 20, 2005, more than one year earlier
than scheduled. He had criticized the conservative government led by
Prime Minister Pedro Santana Lopes of the Social Democratic party (PSD),
who resigned on December 11, 2004, for having failed to cope with the
economic slump. Economic issues dominated the election campaign in this
country of 10.3 million people, known as one of western Europe's poorest
nations. Faced with an unemployment rate that reached 7.1 percent at the
end of 2004, the highest in the last six years, the Socialists insisted on
the need for more investment in technology and training, while PSD argued
that productivity should be increased, and suggested raising the retirement
age from 65 to 68. Socrates, of the opposition Socialist Party (SP),
argued for the need to modernize the economy. He had opposed the U.S.-led
intervention in Iraq and called for a closer foreign policy coordination
with the EU Member States. The election results were marked by a major
victory for the socialists and a crushing defeat for the conservatives.
For the first time in its history, SP secured an overall majority by
obtaining 121 seats out of 230. On the contrary, PSD recorded its worst
result since 1983 with only 75 members elected, down from 148 in 1987 and
102 in 2002. Its coalition partner PP managed to win 12 seats, two fewer
than in the previous election. The two partners in the Unitary Democratic
Coalition (CDU) won a total of 14 seats: 12 for the Portuguese Communist
Party (PCP) and 2 for the Green Ecological Party (PEV). The Left Bloc (BE)
gained three more seats, or eight members in total.

(2) The Portuguese post-election, face-to-face survey was conducted between
March 5 and May 8, 2005. The sample was stratified by region (7 regions
in the mainland: North, Centre, Lisbon-Lisbon district; Lisbon-Set�bal-
district, Lisbon-other districts, Alentejo and Algarve) and habitat (three
categories of parishes by number of registered voters: less than 3000;
3000-9999; more than 10000). Then, for each cell within the region and
habitat frame, parishes were randomly selected with probability
proportional to size, in order to ensure that the number of interviews
would be proportional to the number of voters in each cell, since there
was a previous decision of making the same number of interviews in each
parish. The process of selection was systematically repeated until the
electoral results of the 5 major political parties in those parishes,
taking into account the intended number of respondents, were less than
1 percent different from the general results of mainland Portugal in the
2002 legislative election. The number of parishes selected and interviews
conducted in each region/habitat stratum was proportional to the number of
voters registered in each stratum. In each parish, random route was used
to select the households. Individual respondents were selected using the
following criteria: last person living in the household that had his/her
birthday, with 18 or more years. The response rate was 31.2%.

(3) The demographic weight was constructed on the basis of Census 2001
information about distribution of residents 18+ years of age in
Continental Portugal on the basis of sex (2 strata), age (3 strata), and
education (3 strata). The political weight was constructed on the basis
of the 2005 election results in order to weigh the results of the vote
recall question.

The demographic and political weight variables were originally deposited
with a number of cases having missing data. In the CSES data file,
these cases are coded "0" (215 cases in B1010_2 and 428 cases in B1010_3).
These are cases for which at least one relevant demographic and/or
political variable was missing for the respondent. The collaborator
preferred to assign a weight of "0" to these cases given that the
demographic and/or political information was not complete for the case.


(1) On 28 November 2004 elections were held to choose the Head of State
(the first round), and both Houses of the Romanian Parliament. According to
the constitution, outgoing President Iliescu was not allowed to seek
another term as President and instead he ran for a seat in the Senate under
the banner of the ruling Social Democratic Party. A total of 12 candidates
ran in the presidential elections but two favorites were outgoing Prime
Minister Adrian Nastase (PSD) and the Justice and Truth candidate,
Bucharest Mayor Traian Basescu. Both were fervently in favor of Romania
joining the EU in 2007, and of maintaining close ties with the United
States. They differed on domestic issues. The opposition Justice and Truth
alliance said it would fight widespread corruption within the
administration and introduce a flat 16 per cent tax on personal income and
profits to crack down on the country's widespread illegal economy. The PSD
countered by saying it alone had the experience to lead the country.
Analysts said that voters could turn to Justice and Truth's blaming the
ruling party for widespread corruption and low living standards but also
said that many voters also credited the ruling party for bringing Romania
into NATO and for boosting the country's economic growth.
(Source: PARLINE database:

(2) This post-election, face-to-face survey was conducted between December
14 2004 and January 7, 2005. The sampling procedure was based on
localities, which were randomly selected from each strata of the population
(defined by the region and size of locality). In each selected locality,
two starting addresses were drawn at random by the principal investigators.
Further addresses were selected by a random route procedure - as every
fifth address from the initial address. In each household, one respondent
was chosen using the "first birthday method". In the case the selected
respondent could not be interviewed at the moment, the interviewer tried
three more contacts with him (her). Response rate: 70%.


(1) The 14 March 2004 election was the third election for President since
the dissolution of the Soviet Union in 1991 and the adoption of the 1993
Constitution of the Russian Federation. The office of President of the
Russian Federation embraces exceptional and wide-ranging executive powers.

The previous four years in office of the incumbent President, Vladimir
Putin, were chiefly characterized by a reviving economy, a consolidation
of state power at the centre, and above all by the perception of stability
that was broadly endorsed by the population at large. The seemingly
overwhelming popularity of the incumbent - as evidenced by his consistently
high approval ratings in public opinion polls before and during the
election campaign - produced a sense of predictability in the outcome of
the 2004 Presidential Election.

The presidential election came barely three months after the December 2003
elections to the State Duma, the lower house of the federal parliament.
Those elections had produced a significant shift in the configuration of
Russian parliamentary politics. The association of the incumbent with
United Russia had contributed to that party�s appeal to voters, and United
Russia had emerged from the elections with a two-thirds State Duma
majority. The established opposition parties had experienced a dramatic
decline in their support, and were left preoccupied with the loss of a
federal parliamentary presence or weakened party machinery.

In combination, the factors of an incumbent in an apparently unassailable
lead, and a weakened party political opposition, directly impacted on
the selection of candidates to run against the incumbent. He faced a field
of opponents who commanded little apparent public support. The better known
of them, Irina Hakamada and Sergey Glazev, ran without the backing of their
respective political party/bloc, and the rest had little profile
nationally, including the two State Duma party-nominated candidates,
Nikolay Kharitonov of the Communist Party (CPRF) and Oleg Malyshkin of the
Liberal Democratic Party (LDPR) who were not the leaders of their
respective parties.

In the course of the campaign none of the rival candidates articulated any
expectation of being able to defeat the incumbent, whilst one, Sergey
Mironov, openly supported him.

OSCE/ODIHR Election Observation Mission Report, Political Background

(2) The survey was conducted as the second wave of a panel of election
studie in Russia. The first wave was conducted after the Duma election
in 2003. The second wave, which included the CSES module, was a
post-election, face-to-face study focused on the March 14, 2004 Russia
presidential election. Interviews were conducted over the period April 4
to May 11, 2004. The sampling frame covered 95% of the total eligible
population, excluding those who lived in the far eastern depopulated
regions of Kamchatka Oblast or Sakhalin Island, those who lived in
areas of conflict such as the Republic Chechnya, and those who lived in
separated areas such as Kaliningrad Oblast. Moreover, institutionalized
persons in prisons, hospitals and the armed forces were excluded from

The sampling design was based on the following formula. First, 2029
consolidated raions (akin to counties in the US) were stratified into 38
strata based largely on geographical factors and level of urbanization,
but also based on ethnicity. Moscow city, Moscow oblast and St. Petersburg
city were selected for certainty as self-representing stratum. Other
raions were allocated to 35 strata of roughly equal size and randomly drawn
using probability proportional to size as primary sampling units. Each
selected PSU was stratified into urban and rural substrata, and targeted
sample size was allocated proportionally to the two substrata. In the
rural substrata, villages served as the SSUs. On the other hand, in the
urban substrata basically the census enumeration districts were employed
as the SSUs with little supplemental information if the district's
information was not available. After SSUs were selected with PPS, an
enumeration of dwelling units was made by visual inspection and recourse
to official documents. Finally, the required number of dwellings were
selected systematically starting with a random address in the list. The
Kish procedure was used to select one eligible adult from each household.

After three contacts without response the household was treated as
non-sample. After at least three unsuccessful visits to a household, the
household was treated as non-interview/refusal. A maximum of seven contacts
were made to persuade respondents to be interviewed. In the first wave, the
interviewers succeeded in completing 1648 interviews out of 3347 sample
lines (51.4%). Out of 1648 samples interviewed at the first wave, 1492
(90.5%) of the respondents were re-interviewed in the later wave.


(1) Voters in Slovenia went to the polls to elect the 90 members of the
National Assembly in the first elections since the country joined the
European Union in May 2004. Public opinion polls had suggested that the
ruling coalition, led by Prime Minister Anton Rop's Liberal Democrat Party
(LDS), would retain leadership although polls also indicated that half the
electorate was dissatisfied with the work of the Government and the
Parliament. The LDS, which had been in power for most of the past 12 years,
had suffered a setback on 13 June 2004, during the first elections to the
European Parliament (EP) held in the country. During the electoral
campaign, Prime Minister Anton Rop vaunted the results he had achieved as
head of government including on the economic front. Slovenia's GDP per
inhabitant is 72% of the European average i.e. a level equal to that of
Greece and Portugal. Mr. Rop promised that, if elected, his mandate would
focus mainly on education and research, while Mr. Jansa promised to improve
access to healthcare and education, to develop environmental policy and to
ensure a decent income for all.

The elections resulted in a surprise victory by the conservative Slovenian
Democratic Party (SDS) over the ruling Liberal Democracy of Slovenia (LDS)
by a narrow margin. The SDS won 29.1 per cent of the vote, i.e. 13.3 points
more than during the 2000 general elections. This party obtained 29 seats,
while its coalition partner, the Nsi won 8.9 per cent of the vote and nine
seats. Together both parties obtained one seat more than the three parties
of the outgoing coalition together. The Liberal Democrat Party (LDS) won
22.82 percent of the vote (23 seats), while the United List of Social
Democrats (ZLSD) won 10.20 per cent (10 seats) and the Democratic
Pensioners' Party (DeSUS) obtained 4.04 (4 seats) just enough for the 4 per
cent threshold, necessary to be represented in the National Assembly.

(2) This study was conducted as a face-to-face post-election survey over
the period March 17th, 2005 to April 25th, 2005. The study sampling frame
corresponds to the Central Register of Population (CRP) and includes
all residents with permanent address, including both citizens and
non-citizens (in principle at least 99% of the population). The study
employed a stratified two-stage probability sampling method in which the
primary sampling units (clusters of enumeration areas, CEA) were selected
using the following procedure: first, the entire country was divided into
about 9000 CEA�s and stratified according to 12 regions x 6 types of
settlements. After this, fixed numbers of CEA were selected inside each
strata with probability proportional to size of CEA. In total 136 CEA were
selected. Then 12 persons in each primary sampling unit were chosen by
simple random sampling. Persons selected from Central Population Register
were identified by name and address. Five contacts were made with the
household before declaring it non-sample. The response rate was 64.0%.


(1) "The general election in Spain on March 14, 2004 was held only three
days after the most lethal terrorist attack in the history of Western
Europe. In the wake of the massacre in Madrid, electoral participation
was unusually high. Against most expectations and survey polls, the
People�s Party (PP), which had been in government for eight years under
the leadership of Jose-Maria Aznar, was defeated by the Socialist Party
(PSOE), led by Jose-Luis Rodriguez Zapatero." (Colomer, 2005, pp. 149-50).
The initial electoral campaign focused on the battle against ETA (the
Basque separatist group). After the terrorist attack, it seemed that if
the voters accepted the government's assertion, putting blame on ETA (an
accusation ETA denied), Aznar's party would be rewarded at the polls. But
if people saw al Qaeda's hand in the carnage, voters might view the attacks
as retribution for Aznar's Iraq policy and punish his party when they
voted. Spaniards had reacted furiously in 2003 when Prime Minister Jose
Maria Aznar aligned himself with U.S. President George W. Bush and British
Prime Minister Tony Blair to support the invasion of Iraq. The government
had to back away from its initial certainty that ETA was behind the attacks
as news emerged that a van with detonators and an Arabic-language tape had
been found in a town where three of the bombed trains originated. A larger
than expected 77 per cent of the electorate turned out to vote. The
government was voted out of office as voters gave the Socialist Party 42.6
per cent of the vote. The very first day of the new government, Zapatero
ordered all Spanish troops to be withdrawn from Iraq.

(2) The CSES survey was conducted in the period March 15-20, 2004,
immediately after the parliamentary elections of March 14, 2004. The
survey is based on face-to-face interviews, on a random sample of
respondents. The sample frame covered approximately 99% of the total
population, although it excluded two North African cities: Ceuta and
Melilla. The interviews were distributed among the 17 Autonomous Regions
in proportion to their population and to community size within each region.
Municipalities with more than 500,000 inhabitants were all included, while
the remaining units were randomly selected. Once the number of interviews
to be done was been established (by size of community and autonomous
region), a computerized system to randomly extract municipalities and
electoral sections was applied. The number of electoral sections randomly
selected was related to the total number of interviews to carry out in the
municipality. A random route system was applied for household selection. In
case of refusal or non-contact, the interviewer went to the next household.
Age and sex quotas were used for respondent selection. These quotas were
established in each sampling point according to the ration between size of
community and age and sex at the national level and regional level in the
regions of Catalu�a, Andaluc�a, and Pa�s Valenciano. In the other regions,
the distribution was proportional.

(3) The original demographic weight variable adjusts for sex and age.


(1) The national election was held on September 15, 2002. Local and
regional elections were held at the same time. The electoral campaign
was dominated by immigration and the future of the large public sector,
as voters had to decide whether they wanted to go on paying among the
world's highest taxes to finance their welfare state, or if they preferred
a rightist recipe of tax cuts, privatization, and deregulation. Prime
Minister G�ran Persson, who leads the ruling Social Democrats, campaigned
for more public spending on such welfare state cornerstones as health and
care of the elders, education and  security. A promise to cut taxes seemed
to have played a role in the narrow lead in opinion polls of the opposition
bloc, formed by four center-right parties. One important issue that both
sides sidestepped was whether the country should join the European Union's
Economic and Monetary Union and the common currency, the Euro, which would
entail sweeping economic change for Sweden. At 80.11%, turnout was lower
than in previous elections. The Social Democrats, together with the ex-
communist Party of the Left and the Greens, won over 53% of the votes and
obtained 191 seats in Parliament. The opposition Moderate Party turned in
its worst performance since 1973, obtaining 15.26% of the votes and 55
seats, while the Liberal Party increased from 4.7% in 1998 to 13.3% in
2002. After the election, the Social Democrat Party leader, G�ran Persson,
began a third term as Prime Minister in a minority government as he had
refused to enter into any formal coalition arrangement with the Party of
the Left or the Greens, largely because of their opposition to EU

(2) The Swedish election study was separated into two samples, one
pre-election sample and one post-election sample. The CSES Module 2 was
included in the post-election, face-to-face part of the study. Due to
Swedish data laws the respondents in the Swedish election study 2002 were
asked if they agreed their answers would be a part of international
data set accessible on the Internet. Among the respondents there were 6
percent (70 respondents) who did not want to be included in the dataset.


(1) Elections were held for all the seats in the National Council at the
end of the members' terms of office. The electoral campaign focused on the
issues of the old-age and survivors' insurance scheme and the future of
social insurance. The Swiss voiced their concern at rising unemployment
(3.7%) and pension reform (the prospect of a retirement age of 67 years
was vehemently opposed). Other more peripheral issues included the conduct
of economic policy during a recession (attack by the Socialist Party on
the Christian Democratic People's Party) and immigration policy (dual
initiative pending from the Swiss People's Party on asylum and
naturalization) as well as environmental issues.

The issue of European Union membership, meanwhile, proved so unpopular
that it was dropped by all the parties with the exception of the
Socialists. The Swiss People's Party (SVP/UDC), led by Christoph Blocher,
became the largest party in the National Council. The SVP/UDC, a right-wing
populist party, ran an advertising campaign in the run-up to the elections
portraying asylum seekers and refugees as criminals, thereby earning itself
a rebuke from the Office of the United Nations High Commissioner for
Refugees (UNHCR). According to the official results, the SVP/UDC now holds
55 out of the 200 seats in the National Council, a gain of 11 seats since
the last election. Just behind is the Socialist Party (SP/PS) with 54
seats, an increase of three. The gains made by the People's Party were
mainly at the expense of the Christian Democratic People's Party (CVP/PDC)
and the Radical-Democratic Party (FDP/PRD), both of which lost seven seats,
with respective tallies of 28 and 36. The Greens won 13 seats, compared
with 9 in 1999.

(2) This is a post-election study, focused on the Swiss parliamentary
elections of October 19, 2003, conducted between October 20 and November 2
of the same year. It was conducted in three languages (German, French and
Italian), as a combination of telephone and mail-back surveys. The CSES
module was conducted as a supplementary postal survey completed by all
respondents from the telephone sample willing to answer a supplementary
self-administered questionnaire. Total panel attrition was 31.8%.
The sample weight variable adjusts for the differential attrition rate.
Sample for the telephone survey was drawn from fixed subscribers to
conventional telephone networks (households were randomly selected
from the Swisscom directory). Approximately 2-3% of the population were
excluded from the sample frame because of not possessing a telephone or
other reasons.


(1) This is the first election since the introduction of a unicameral
(or, at most, semi-bicameral) legislative system in April 2000. The
National Assembly is a now a non-standing body, convened on special
occasions only: "within three months of the expiration of a six-month
period following the public announcement of a proposal by the Legislative
Yuan to amend the Constitution or alter the national territory, or within
three months of a petition initiated by the Legislative Yuan for the
impeachment of the President or the Vice President." It is to have 300
delegates 'nominated by political parties on the basis of proportional

(2) This is a post-election, face-to-face election study, focused on the
Taiwanese parliamentary election of December 1, 2001. It was conducted
between January and April 2002. The sample frame includes approximately
98% of the eligible population. The primary sampling units were randomly
selected - boroughs and villages at the township-level and neighborhoods
at the urban level. A principal component analysis and cluster analysis
were applied to divide the 329 township-level units into 9 strata.
Subsequently, according to the probabilities proportional to size (PPS),
approximately 12 respondents from each sampled borough or village and 4
respondents from every selected neighborhood were randomly selected. After
three visits without success, the sampled respondent was replaced by the
supplemental sample, according to the identical socio-demographic
characteristics (area, sex, and age) of the first-sampled respondent.
The response rate was 35.39%.


(1) The main candidates for the 2004 presidential election were affiliated
with two large coalitions, respectively. For the Pan-Green coalition, led
by the DPP, the incumbent President Chen Shui-bian and Vice-President
Annette Lu ran for a second four-year term. For the Pan-Blue coalition,
the KMT�s chairperson, Lien Chan, ran for the presidency, with James Soong
(PFP) as vice-president. The Pan-Green coalition consisted of the two pro-
independence parties, DPP and Taiwan Solidarity Union (TSU), while the Pan-
Blue coalition consisted of the KMT, PFP, and the New Party (NP) which were

Established in 1986 during the last years of President Chiang Ching-kuo�s
administration, the DPP started as a a rag-tag coalition of opposition
stalwarts to what is now known as a well-organized electoral party.
The DPP�s main issue position was that Taiwan is a separate entity from
mainland China, which distinguishes it from the position taken by the
parties within the Pan-Blue alliance. The KMT was the dominant party
since Chiang Kai-shek moved from mainland China to set up his government-
in-exile. Since 1986, however, the KMT has undergone substantial change,
including "Taiwanization",  under Lee Teng-hui. Since 2000, the KMT has
become, internally, a more democratic party, in contrast to the Leninist-
type party of the past (Tan, 2002). The most obvious difference among
Taiwan�s parties was on the issue of national identity. The most notable
feature of the election campaign, however, was the failed assassination
attempt on President Chen on the eve of polling day.

This was one of the closest elections that Taiwan has witnessed since
the ushering in of democratic elections. The incumbent, Chen Shui-Bian
of the DPP, and his running mate, Annette Lu, won 50.1% of the vote; the
opposition, the Pan-Blue alliance with the KMT leader Lien Chan and PFP
leader James Soong won 49.9%. With an increase of almost 11 points on
its vote share in 2000 (39.3%), the DPP team won just marginally more
than half the vote. This increase represents two very significant
developments in Taiwanese electoral politics: the continuing growth of
the DPP, and the decline of the Pan-Blue alliance (especially the

(2) This is a post-election, face-to-face election study focused on
the March 20, 2004 Taiwanese presidential election. Interviews were
conducted between June 2004 and September 2004. The sample frame included
approximately 98 percent of the eligible population. The sample design was
based on official information provided by the Interior Department. A
cluster analysis was employed to divide the 359 township-level units into
9 strata. Subsequently, and according to probabilities proportional to size
(PPS), approximately 2 to 14 townships were randomly selected from each
strata, and then 4 to 28 boroughs or villages from each selected township.
Finally, 10 to 16 respondents were sampled from every selected borough or
village through systematic sampling. After three visits without success,
the sampled respondent was replaced by the supplement sample according to
the identical socio-demographic characteristic (i.e., area, sex, and age)
of the first-sampled respondent. The response rate was 27.2%.

(3) The Taiwanese 2004 election study includes one weight variable.
The demographic weight uses gender, age, education level, and area (based
on the level of socio-economic development). The aggregative indexes of
gender, age, and socio-economic development are based on official
documents released by the Interior Department. In addition, since the
information about education released by the official document may be under-
estimated, the index of education comes from the adjusted estimations of
Professor Yung-tai Hung of Political Science, National Taiwan University.


(1) In the November 2, 2004 elections US citizens voted for President as
well as the entire membership of the House of Representatives and one third
of the members of the Senate. Incumbent President George W. Bush's main
challenger was the democratic candidate, Senator John F. Kerry from
Massachusetts. The campaign was hotly contested with the war in Iraq being
a critical issue. President Bush's approval rating had been high,
especially when Saddam Hussein was so quickly ousted from power with the
U.S.-led military intervention on Iraq. However, as a result of the
continuing insurgency against U.S. and Iraqi government forces, the failure
to find the weapons of mass destruction that the Bush administration had
used to justify the war, the revelation that some U.S. soldiers had abused
Iraqi prisoners, and growing concern that the war might have created more
terrorist enemies than it had defeated had adversely affected his
popularity. But the Iraq issue was also a difficult one for Senator Kerry,
as the public wondered whether he would have handled it better, in light of
his support for the congressional resolution that had authorized the war
and his subsequent vote against funding the Iraq operation. Regarding the
domestic agenda, President Bush set said he would tackle education, health
care, energy and the economy, putting emphasis on limited government
spending, individual responsibility and the power of markets. The
Organization for Security and Co-operation in Europe (OSCE), at the
invitation of the U.S. State Department, sent some 100 observers to the
country at the beginning of October to observe the final weeks of the
campaign, election day and counting procedures. In its report, the OSCE
declared that the elections had "mostly met" standards for freedom and
fairness and reflected a "long democratic tradition". The turnout was the
highest since 1968. President Bush won re-election, gaining another
four-year term and a clear endorsement from a majority of voters. In the
Congress, the Republican Party increased its representation in the Senate
by four seats. It will retain control of the new Senate with 55 seats
against 44 for the Democratic Party and one seat held by a
Democratic-leaning independent. In the House of Representatives, the
Republican Party obtained a net gain of three seats, taking its tally to
231 members, the most since 1946.

(2) The 2004 American National Election Study was conducted on a fresh
multi-stage area probability sample intended to be representative of the
United States. Identification of the 2004 ANES sample respondents was
conducted using a four stage sampling process: a primary stage sampling of
U.S. Metropolitan Statistical Areas (MSAs) or New England County
Metropolitan Areas (NECMAs) and non-MSA counties, followed by a second
stage sampling of area segments, a third stage sampling of housing units
within sampled area segments, and concluding with the random selection of a
single respondent from selected housing units. To be eligible for
interview, respondents had to be United States citizens who were 18 years
of age or older on Election Day. Pre-election interviews, averaging 70
minutes in length, were conducted September 7 through November 1, 2004. No
interviewing was conducted on election day, November 2. Post-election
interviews, averaging 65 minutes, were administered November 3 through
December 20, 2004. Randomization, employed for selection of half-samples
to reduce overall interview length and for question order within
batteries, was implemented by the CAI instrumentation. The total sample
included 1,833 eligible persons and produced 1,212 pre-election interviews
for a response rate of 66.1%. In the post-election study, 1,066 persons
granted re-interviews for a reinterview rate of 88.0%. The CSES Module 2
questionnaire was administered to the 1,066 post-election respondents, at
the end of their post-election interview.


"2002 General Election - Official Result." Official Election Results
for New Zealand. Retrieved on August 2, 2004 from:

Abazov, Rafis (In pres). The parliamentary election in Kyrgyzstan,
February/March 2005, Electoral Studies, Available online 9 August 2006

Adam Carr's Electoral Archive. Retrieved from:

Agpalo, Ruben. Comments on the Omnibus Elections. Rex Printing Company,
Inc, Quezon City, 2004

Albanian Constitution

Albanian Central Election Commission

Al�ingiskosningar 2003. Election results 2003 from:

Anckar, Carsten and Dag Anckar: Finland. In Dieter Nohle (ed.), Elections
in Europe. A Data Handbook. Oxford University Press, forthcoming.

Andersen, J�rgen Goul (2003). The general election in Denmark,
November 2001. Electoral Studies 22, 153-193.

Andersen, J�rgen Goul (in press) "The parliamentary election in Denmark,
February 2005."  Electoral Studies, Available online 28 November 2005.
Assembly of the Republic of Portugal. Retrieved from:

The Australian House of Representatives:

The Australian Senate:

Balinski, Michel and Victoriano Ram�rez Gonz�lez (1997). "Mexican electoral
law: 1996 version." Electoral Studies, Volume 16, Issue 3, 329-40.

Balinski, Michel and Victoriano Ram�rez Gonz�lez (1999). "Mexico's 1997
apportionment defies its electoral law." Electoral Studies, Vol. 8, 1,

"Basic Law for the Federal Republic of Germany." Retrieved November
17, 2004 from:

The Basic Law of the HKSAR:

The Belgian Chamber of Representatives:

The Belgian Senate:

Bernas, Joaquin. The 1987 Constitution of the Republic of the Philippines:
A Commentary. Rex Printing Company, Inc, Quezon City, 1996

Biroul Electoral Central (Romanian Election Commission)

Brazilian Electoral Data (1982-2002) Electronic Edition.
Retrieved February 22, 2005:

Brazilian Electoral Data (1982-2002). Senate Elections. Electronic Edition.
Retrieved February 22, 2005:

The British Election Study. Retrieved:

"Bundestag election 2002." Official election results for German Bundestag.
Retrieved on November 17, 2004 from:

Bustani, Camilla (2001). The 1998 Elections in Brazil. Electoral
Studies 20, 305-339.

Butler, David and Gareth Butler (2006). British Political Facts
Since 1979. London: Palgrave.

Calvo, Ernesto and Juan Manuel Abal Medina (2002). "Institutional gamblers:
Majoritarian representation, electoral uncertainty, and the coalitional
costs of Mexico's hybrid electoral system." Electoral Studies, Volume 21,
3, 453-471.

The C�mara dos Deputados (Chamber of Deputies):

Carty, R. K. (2005). Canadian politics and the general election of 2004.
European Journal of Political Research, 44, 968-974.

Center Vlade za Informatiko (Government Center of Information, Slovenia).

Central Election Commission of the Russian Federation

Chilean Constitution. From:

Chilean Electoral Service (SERVEL), Retrieved from:

Cho, Wonbin (2005). The general election in South Korea, April 2004.
Electoral Studies 24, 511-551.

Clarke, Harold D., David Sanders, Marianne C. Stewart and Paul Whiteley
(2004). Political Choice in Britain. Oxford: Oxford University Press.

Code of the Kyrgyz Republic On Elections in the Kyrgyz Republic

Colomer, Josep M. "The general election in Spain, March 2004"
Electoral Studies, Volume 24, Issue 1, March 2005, Pages 149-156.

The Commonwealth of Australia Constitution Act:

Congreso de los Diputados:

The Constitution for the Republic of Korea:

The Constitution of Belgium. Retrieved on April 7, 2005, from:

Constitution of the Czech Republic:

The Constitution of Finland:

The Constitution of Ireland. [Bunreacht na h�ireann.] Retrieved
July 30, 2004:

The Constitution of Japan. Retrieved from:

The Constitution of the Kingdom of Norway. Retrieved July 08, 2004

The Constitution of the Kingdom of the Netherlands 2002. Retrieved:

The Constitution of the Kyrgyz Republic:

The Constitution of Mexico. Retrieved August 8, 2004 from:

The Constitution of the Portuguese Republic. Retrieved July 28, 2004

The Constitution of the Republic of Bulgaria. Retrieved July 30, 2004

The Constitution of the Republic of the Philippines. Retrieved from:

The Constitution of the Republic of Poland. Retrieved July 14, 2004

Constitution of Romania

The Constitution of Spain:

The Constitution of the Swiss Federation. Retrieved August 10, 2004

The Constitution of the United States:

The Constitutional Act of Denmark of June 5, 1953:

"CNE Resultados." Election results for Portugal. Retrieved July 28,
2004 from the Official site of the National electoral Commission (CNE):

Coakley, John and Michael Gallagher (eds.) (2004). "Politics in the
Republic of Ireland." Routledge.

Constitutional Affairs Bureau, HKSAR:

The Danish Parliament - The Folketing (People's Diet):

Dardanelli, Paolo (2005). "The parliamentary and executive elections in
Switzerland, 2003." Electoral Studies, Vol. 24, 123-160.

Direzione Centrale dei Servizi Elettorali, Dipartimento per gli Affari
interni e territoriali.Ministero dell' Interno, Italia.

Downs, William M. and Raluca V. Miller (2006). "The 2004 presidential
and parliamentary elections in Romania". Electoral Studies, 25, 393-415.

Dutch Electoral Council

Elecciones Generales de 14 de Marzo de 2004. (Election results, Total and
District). Retrieved February 1, 2005 from the Central Electoral Commission

"Election 2002 for the Chamber of Deputies of the Parliament of the Czech
Republic." Retrieved on November 29, 2005 from

Election Results Australia: Australian Election Commission. Retrieved on
May 4, 2005 from:

"Election results." Elections for the 16th Knesset, Israel. Retrieved
August 3, 2004 from:

Elections Canada On-line. Retrieved October 28, 2005 from:

"Elections for 39th National Assembly." Election results for Bulgaria 2001,
retrieved July 30, 2004, from the official site of the National Assembly of
the Republic of Bulgaria:

Elections in the National Assembly of he Republic of Slovenia, 3rd October
2004, Statistical Office of the Republic of Slovenia. Retrieved from:

Electoral Affairs Commission�s Report on the 2004 Legislative Council

"Electoral alliance and votes cast per electoral alliance by constituency
in Parliamentary elections in 2003" in p. 216-217 in book Parliamentary
elections 2003. Official Statistics of Finland. Elections 2004:1.
Helsinki: Statistics Finland, 2004.

The Electoral Code Of The Republic Of Albania

The Electoral Commission. United Kingdom. Retrieved:

"Erst-und Zweitstimmen in allen Wahlkreisen im csv-Format f�r eine
Weiterverarbeitung in einem Tabellenkalkulationsprogramm [data file]."
Retrieved April 4, 2003 from:

Federal Election Commission of the United States. Retrieved from:

Federal Election Commission of the United States. Retrieved April
26, 2005 from:

Federal Electoral Institute of Mexico (source for election results,
retrieved August 10, 2004):

Federal Law No 175-FZ of December 20, 2002. <<On the Election of Deputies
of the State Duma of the Federal Assembly of the Russian Federation>>.
Retrieved from:

Federal Law No FL-19-FZ of December 27, 2002.<<On the Election of the
President of the Russian Federation>>.
Retrieved from:

The Finnish Parliament Statistics 2003. (February 4, 2005)

Finnish Statistical Office website:

Fowler, B. (2003). "The parliamentary elections in Hungary, April
2002." Electoral Studies, Vol. 22, No. 4, 799-807.

Galen, Irwin A. and Joop J.M. Van Holsteyn (2004). "The 2002 and 2003
parliamentary elections in The Netherlands."  Electoral Studies 23,

Gallagher, Michael, Michael Marsh and Paul Mitchell (eds.) (2003). "How
Ireland Voted 2002." Hampshire, UK: Palgrave.

Garry, John (2004). "The general election in Ireland, May 2002." Electoral
Studies, Vol. 23, No. 1, 155-160.

Geddis, Andrew (2004). "The general election in New Zealand, July 2002."
Electoral Studies, Vol. 23, No. 1, 149-155.

General Election 2005. Research Paper 05/33 (17 MAY 2005). House of Commons
Library. Retrieved January 8, 2006 from:

Grzybowski, Marian and Piotr Mikuli (2004). Poland. In "Handbook of
Political Change in Eastern Europe", 2nd revised and updated edition,
ed. by Sten Berglund, Joakim Ekman, and Frank H. Aarebrot. Cheltenham:
Edgar Elgar.

Guidelines on Election-related Activities in respect of the
Legislative Council Elections:

Hardarson, �lafur Th. Parties and voters in Iceland. A study of the 1983
and 1987 Althingi elections. Social Science Research Institute-
University Press, University of Iceland, Reykjav�k 1995.

Hardarson, �lafur Th. "Iceland" in G. E. Delury (ed.) (Third Edition edited
by D.A. Kaple) World Encyclopedia of Political Parties and Systems. Facts
on File Publications, New York (1999), Vol. II, pp. 473-482.

Hardarson, �lafur Th. "The Icelandic Electoral System 1844-1999" in A.
Lijphart og B. Grofman (eds.) The Evolution of Electoral and Party Systems
in the Nordic Countries New York:Agathon Press, 2002, pp. 101-166.

Hardarson, �lafur Th. and Gunnar Helgi Kristinsson, "Iceland", European
Journal of Political Research 43: 1024-1029, 2004.

Hardarson, �lafur Th. and Gunnar Helgi Kristinsson (2001). The 1999
Parliamentary Election in Iceland. Electoral Studies, Volume 20, Issue 2,
pp. 325-331.

Hardarson, �lafur Th. and Gunnar Helgi Kristinsson, "Iceland" in European
Journal of Political Research, 38:3-4, 408-419, 2000.

Harper, M. A. G. (2003). "The 2001 parliamentary and presidential elections
in Bulgaria." Electoral Studies, 22, 335-344.

Hazan, Reuven Y., and Abraham Diskin (2004). "The parliamentary elections
in Israel, January 2003." Electoral Studies 23, 329-360.

Helms, Ludger (2004). "The federal election in Germany, September 2002."
Electoral Studies, Vol. 23, No. 1, 143-149.

Homepage of the federal parliamentary elections on May 18, 2003. Retrieved
on April 6, 2005 from:

House of Councillors: The National Diet of Japan. Retrieved from: and

House of Representatives Directory, 2005 Edition (Philippines).

House of Councillors (Upper House). Retrieved from:

Ignazi, Piero (2006). "Italy". European Journal of Political Research,
45, 1143-1151

Indridason, Indridi H. "A Theory of Coalitions and Clientelism: Coalition
Politics in Iceland 1945-2000". Forthcoming in the European Journal of
Political Research.

Japan: Central Election Management Council. Retrieved from:

Jesse, E., 2000. The electoral system: more continuity than change. In:
Helms, L. (Ed.), "Institutions and Institutional Change in the Federal
Republic of Germany". Macmillan, London, pp. 124-142.

Karasimeonov, Georgi (2004). Bulgaria. In "Handbook of Political Change in
Eastern Europe", 2nd revised and updated edition, ed. by Sten Berglund,
Joakim Ekman, and Frank H. Aarebrot. Cheltenham: Edgar Elgar.

Kavanagh, Dennis and David Butler (2005). The British General Election of
2005. London: Palgrave.

Kiesraad. Elections in the Netherlands. Retrieved from:

Kiesraad. The Dutch Elections Act. Retrieved from:

Klesner Joseph L. (2002). "Presidential and congressional elections in
Mexico, July 2000." Electoral Studies, Volume 21, Issue 1, 140-147.

K�r�s�nyi, Andr�s (2002). "Government and Politics in Hungary." Budapest:
CEU Press.

Kristinsson, Gunnar Helgi, "Parties, States and Patronage" in West European
Politics, 19:3, 433-457, 1996.

Kristinsson, Gunnar Helgi, Farmers' Parties. A Study in Electoral Adoption.
Reykjav�k: F�lagsv�sindastofnun H�sk�la �slands, 1991.

The Kuk Hoe (National Assembly):

Kuvendi i Republikes se Shqiperise

Kyrgyzstan Election Commission -

Law Library of Congress. Election Law:

LeDuc, Lawrence (2005). The federal election in Canada, June 2004.
Electoral Studies, 24, 303-344.

"Legislativas 2002." Results of elections in Portugal retrieved July 12,

Legislative Council Ordinance (Chapter 542 of the Law of Hong Kong):

Leonard, Dick and Roger Mortimore (2001). Elections in Britain: A Voter's
Guide. 4th ed. London: Palgrave.

Ley 180603 Org�nica Constitucional de los Partidos Pol�ticos, Republic of
Chile. Retrieved from:

Ley 18.700 Org�nica Constitucional sobre Votaciones Populares y
Escrutinios, Republic of Chile. Retrieved from:
McAllister, Ian (2005) The Australian federal election, October 2004.
Electoral Studies 24, 511-551.

McAllister, I. (2003). "The federal election in Australia, November 2001."
Electoral Studies, Volume 22, Issue 2, Pages 381-387.

Millard, F. (2003). "The parliamentary elections in Poland, September
2001." Electoral Studies, Vol 22, 367-374.

"Ministere de l'interieur: Election presidentielle 2002." Elections France.
Retrieved April 30, 2004 from:

Ministerio del interior. Spanish election results. Retrieved June 14:

Nicolau, Jairo (2004). The October 2002 elections in Brazil. Electoral
Studies 23 (2004) 329-360.

Nolledo, Jose. The Omnibus Election Code of the Philippines   National Book
Store, Inc, Metro Manila, 1996

Nurmi, Hannu and Lasse Nurmi (2004). The parliamentary election in Finland,
March 2003. Electoral Studies 23, 557-565.

Office of the Official Register. U.S. National Archives and Records
Administration. U.S. Electoral College:

Official Website for the 2004 Legislative Council Election:

Oficina Nacional de Procesos Electorales, Republica del Peru.

Paloheimo, Heikki: The Rising Power of Prime Minister in Finland.
Scandinavian Political Studies, vol. 26(2993): 3, p. 219-244.

"Pa�stwowa Komisja Wyborcza: Protok�l Pa�stwowej Komisji Wyborczej z dnia
26.09.2001 r. o zbiorczych wynikach glosowania na okr�gowe listy
kandydat�w na pos��w." Election results, Polish Parliamentary Elections,
2001. Retrieved July 14, 2004:

The Parliament of the Commonwealth of Australia:

Parliament of Finland. Presentation Pages:

The Parliament of the Republic of Iceland (Althingi):

Parliamentary Elections and Election Administration in Denmark:

"Parliamentary Elections in Hungary 2002." Retrieved April 20, 2003

Pasimio, Renato. The Philippine Constitution (Its Evolution and
Development) and Political Science. National Book Store, Inc, Metro Manila,

Plecit�-Vlachov�, Kl�ra and Mary Stegmaier (2003). "The Chamber of
Deputies election, Czech Republic 2002." Electoral Studies, Vol. 22, No. 4,

Poslaneck� Sn�movna (The Chamber of Representatives):

Portugal 2005 election results: Assembleia da Republica. Retrieved

"Project on Political Transformation and the Electoral Process in
Post-Communist Europe. Poland - Constituency Data (2001)." Retrieved April
3, 2003:

Registration and Electoral Office, HKSAR:

"Republic of Bulgaria National Assembly: Elections for the 39th National
Assembly."  Retrieved January 1, 2003 from:

Republic of the Philippines Commission on Elections COMELEC. Retrieved

Republic of the Philippines Commission on Elections COMELEC. The
Constitution of the Republic of the Philippines. Retrieved from:

"Results for the 29th Dail." Elections Ireland. Retrieved April 4, 2003

Schafferer, Christian (2003). "The legislative Yuan election, Taiwan 2001."
Electoral Studies, Vol. 22, No. 3, 532-537.

The Senado Federal (Federal Senate):

Sen�t (The Senat):

Sitio Historico Electoral, Ministerio del Interior, Gobierno de Chile .
Retrieved from:

Statistics Finland site. Parliamentary election results. Retrieved February
8, 2005:

Statistics Iceland. Election results 2003. Retrieved March 21, 2005:

"Stortingsvalget 2001." [Storting Election 2001.]. Statistisk sentralbyr�
[Statistics Norway], Oslo-Kongsvinger.

"Storting Election 2001" Norwegian Election results, 2001. Retrieved July

Suomen Eduskunta  (Finnish parliament)  -

"Sveriges Officiella Statistik Statistiska Meddelanden: Riksdagsvalen
1973 - 2002." Statistics Sweden. Retrieved February 3, 2004 from:

Swyngedouw, Marc (2004). The general election in Belgium, May 2003.
Electoral Studies 23, 545-571.

Swyngedouw, Marc (2002). The general election in Belgium, June 1999.
Electoral Studies 21, 105-154.

Swyngedouw, M., Billiet, J., Goeminne, B., Dewinter, L., Frognier A-P.,
Baudewyns, P., Berck A-S., "ISPO-PIOP 2003 General Election Study Belgium.
Codebook: Questions and Frequency tables" Leuven/Louvain-la Neuve, 2004.

Szajkowski, Bogdan (200&). The parliamentary election in Albania,
July-August 2005. Electoral Studies, 26, 227-231.

"Taiwan 2004 elections". Election results. Government Information Office,
Republic of China (Taiwan). Retrieved from:

UK Electoral Commission.

Vaalit Val. Parliamentary Elections:

Valen, Henry (2003). "The Storting election in Norway, September 2001."
Electoral Studies 22, 179-185.

Vander Weyden, P., D'Hondt and Alternative D'Hondt for Two-Tier
Districting Systems. The Belgian Electoral System'. IPSoM-Bulletin
2001/5, Brussels.

Vander Weyden, P. Het Belgische Kiessysteem: de klassieke en alternatieve
methode D�Hondt. Res Publica, VolXLIII ,2001(4), 595-616.

"Taiwan 2001 elections". Election results, retrieved June 22, 2004 from:

T�ka, G�bor (2004). Hungary. In "Handbook of Political Change in Eastern
Europe", 2nd revised and updated edition, ed. by Sten Berglund, Joakim
Ekman, and Frank H. Aarebrot. Cheltenham: Edgar Elgar, pp. 289-336.

"�bersicht Nationalrat Schweiz." Swiss Election results, 2003. Retrieved
August 12, 2004:

Whiteley, Paul, Harold D. Clarke, David Sanders and Marianne C. Stewart
(2005). "The Issue Agenda and Voting in 2005."  In Pippa Norris and
Christopher Wlezien, eds. Britain Voters 2005. Oxford: Oxford University