Report on Workshop 1, “Winners and Losers in the Elections of Eastern Europe”

prepared by Joshua Kjerulf Dubrow, Associate Professor at IFiS-PAN and Principal Investigator of the Research Project

October 22, 2013

Introduction

The first Workshop for the research project on Electoral Control in Eastern Europe was held October 18-19, 2013, at the Institute of Philosophy and Sociology, Polish Academy of Sciences (IFiS PAN) in Warsaw, Poland. The research project is designed to build an international and interdisciplinary scientific team focused on the use of data on candidates and parliamentarians in Eastern Europe to address critical issues in representation, accountability and political inequality. This first international workshop, titled “Winners and Losers in the Elections of Eastern Europe,” brought together PhD students and young and established scholars from sociology and political science from the U.S and Europe. The workshop introduced the data collection effort for the Eastern European Parliamentarian and Candidate Database (EAST-PaC), a dataset on parliamentarian candidates for all elections matched across time since 1989 in Poland, Ukraine and Hungary.

Funding and Organizational Support

This event was funded by a grant from Poland’s National Science Centre and with additional support from the Institute of Philosophy and Sociology (IFiS) and the Graduate School for Social Research (GSSR) at PAN, and organizational support from Cross-National Studies: Interdisciplinary Research and Training Program (CONSIRT) at The Ohio State University and PAN.  We thank Prof. Andrzej Rychard, Director of IFiS PAN, who delivered welcoming remarks. We are also grateful to GSSR for providing help with personnel and allowing us to use their facilities to accommodate the larger number of participants than initially registered.

Participants

Participants of the Workshop come from institutions in the U.S., Hungary, Ukraine, Romania and Poland.  The Workshop featured discussions from well-known scholars in the field, including Kazimierz Slomczynski, Mikolaj Cesnik, Henryk Domanski and, via video conferencing, Jakub Zielinski. This research project also intends to engage PhD students and encourage them to use these data for their dissertations. Of the 17 registered participants, five are PhD students, three of whom come from GSSR, one from Central European University and one from Corvinus University of Budapest.  Discussions with attendees from CEU and Romania allowed us to explore possibilities of collecting matched candidate data in Romania.

The Electoral Control project is, at this stage, geared primarily to sociologists and political scientists. Attendees of the Workshop came from departments of Political Science and of Sociology in various universities, as well as from the Institute of Political Studies at PAN.  Attendees from the Institute of Political Studies were keen on research collaboration and suggested other colleagues who could open new avenues for research with the Workshop data.  One intriguing possibility is collecting more data from 1985 via archival work and matching it with the existing data.

Structure of the Workshop: Information, Discussion, Networking

Structurally, the Workshop was designed to present substantive and methodological issues of the EAST-PaC data, encourage discussion, and facilitate networking.  As such, the Workshop opened with a presentation of the main thrust of the research project that generated creative new ideas among participants on the scientific use of these data.  Afterwards, participants presented, in short form, their research biographies with regard to the subjects of this Workshop; we learned that participants come from a variety of research experiences and interests, and all of them are interested in building EAST-PaC into their current research agendas.

Presentations in the Workshop were of either substantive topics related to, or methodological issues in, political candidates data.  A stand-out was the presentation by Prof. Sawinski. The efforts he and his team put in collecting, matching and cleaning the Poland candidates data that span 1985 to 2011 has not only led to a very high-quality and unique dataset, but to building the foundations for a cross-national standard of collecting matched data of this type.  We are looking forward to Professor Sawinski’s methodological article based on his presentation.

The Workshop also featured hands-on training in managing EAST-PaC data, led by Prof. Kazimierz M. Slomczynski and Dr. Irina Tomescu-Dubrow. The training was held in the Computer Lab of IFiS PAN and involved 10 participants.  The training featured STATA as a statistical software tool to manage such a complicated, large dataset (N > 30,000) that spans eight elections.

The end of the Workshop featured a video conference of Jakub Zielinski, live from Boston, Massachusetts.  Dr. Zielinski, along with Prof. Slomczynski and Prof. Goldie Shabad of the Department of Political Science at OSU, is an intellectual founder of this research project.  Dr. Zielinski spoke for an hour with Workshop participants on substantive issues of electoral accountability and what future research should do to address these issues.  The video conferencing achieved the goal of a fruitful discussion with a member of the research team who could not physically attend the Workshop; video conferencing will likely be a regular feature of future Workshops and conferences, so facilities that enable this type of communication are essential to have.

Products of the Workshop

A main goal of the Workshop was to bring people together who have common interests and expertise, and this goal was met.  The research project continues and will create two more such Workshops and more collaborative opportunities.  We are also building materials designed to assist researchers in creating scientific output based on these data on this topic.  Based on ideas that emerged from the discussions, we are constructing a bibliography on electoral control and accountability, along with country reports on elections, that will provide subject and methods contexts to produce scientific articles.  The materials will be made available on the project website, electoralcontrol.org.  Since this is an international research team, it is vital that team members have access to the latest available data and be able to communicate across time and space.  The Workshop generated useful ideas on how to manage these issues, including the use of online scientific management software and websites.  We will use this software, along with the main website, to facilitate research among the team and for future users.

Selected Workshop Abstracts

Presentation: Collecting Candidate and Contextual Data in Ukraine, by Dr. Natalia Pohorila, SOCIS – Center for Social and Marketing studies, Kiev, Ukraine

From 1990-2012, there have been seven elections in Ukraine. This presentation concerns the issues in data collection on candidates, parliamentarians and elections in Ukraine for this period.  Especially troubling Issues are those connected to data availability of the candidates’ personal information for the poorly documented 1990-1994 elections.  Complicating matters are the changes to the Electoral Law in-between almost all elections. Obtaining reliable official statistical data is also challenge due to the political bureaucratic structure of modern Ukraine, owing no small portion to the differing standards of disparate bureaus in charge of these data collection.   Providing contextual data on the candidate’s local environment has thus many problems needing manageable solutions. This presentation opens the discussion on the solution of these problems.

Presentation: Methodological Issues in Collecting and Cleaning Polish Candidate Data 1985 – 2011, by Prof. Zbyszek Sawinski, Educational Research Institute, Poland

Download PDF of this presentation, Methodological Issues in Collecting and Cleaning Polish Candidate Data 1985 – 2011 by Prof. Sawinski

This presentation is on the matching of candidate data through time in Poland, the structure of these data, and on introducing new variables based on characteristics of candidates.  A number of problems arise in the matching process, which may at first glance seem mundane but actually has very substantial effects on the quality of the data, and thus we addressed them. Since these data have been culled from various computerized formats that existed from the 1990s to now, problems include standardization of Polish characters and early limits on software imitations for string variables; this meant that the corrections of names was a big effort.  From elections 1985 to 2007, for example, we corrected 13,888 family names, 9794 first names and 5915 middle names.  Another major problem was “duplicated candidates” that, to solve, entailed a series of detailed decisions. In short, we created a decision tree intended to differentiate between candidates who, through the matching process, (a) were recorded as two separate people, but are actually the same person and (b) were recorded as two people that shared many similar characteristics but were, in fact, two different people.  Another problem is that, for the first time, in 2011 Year of Birth is not reported by Poland’s National Electoral Commission, for reasons that are not known to us.  In this presentation we present the fully cleaned data 1985 – 2011, which also allows us to compare and discuss the consequences of improperly matched data. We also introduce interesting new variables based on candidates’ characteristics.

Presentation: The role of nominal level legislative careers in explaining constituency service in parliament under mixed-member electoral rules. The Hungarian case, by Zsófia Papp, Corvinus University of Budapest, Hungary

The study of mixed-member electoral systems concentrates on the effects of mandate type and nominal level candidacy on representational roles and constituency orientation. The first contributes to the research on the mandate divide, while the second (candidacy) is the key component of the debate evolving around contamination between several different tiers of the electoral system. Single member district (SMD) representatives are expected to be more constituency-oriented than those elected from party lists. However, previous research demonstrates that list MPs often appear as the “shadows” of the SMD members, and engage in constituency service just like their SMD counterparts. The explanations are mainly built on the actors’ positions. This paper aims to contribute to the literature by considering the function of habits in members’ attitudes and behaviour. More elaborately, it investigates how the MPs SMD experiences interact with mandate type on the one hand and SMD candidacy on the other. Constituency orientation is perceived as the dominance of local aspects in the legislators’ parliamentary questioning. I hypothesize that members with more dominant constituency relatedness are more concerned about local issues and this career factor will also positively influence the established effect of mandate type and SMD candidacy. In the quest of demonstrating the interrelatedness of these variables I use Hungary as case study. The analysis rests upon written and oral questions in parliament submitted between May 2010 and January 2013. Due to the special features of the Hungarian electoral systems (candidates can run on multiple tiers) and the structure of public service provisions (county-centeredness) localism is defined as reference in the question to any particular area smaller than the country but not larger than the county. As the unit of analysis is the individual MP, raw data were aggregated in order to obtain information on the dominance of localism with regards to the individual members’ questioning behaviour. The number and the proportion of local questions submitted for oral and written response will be analyzed separately to highlight substantial differences between the two types of inquiry. Since both the dependent and independent variables rest upon easily accessible information, the whole population of MPs serving during the 2010-2014 electoral term will be placed under investigation.