\documentclass[letterpaper,headrule]{foils} \usepackage{graphicx} \usepackage{amsmath,amsthm} \usepackage{graphicx} \usepackage{color} \usepackage{sgame} \usepackage{threeparttable} \usepackage{latexsym} \usepackage{threeparttable} \setlength{\foilheadskip}{-0.5in} \newcommand\vm[1]{% Vector or matrix \bm{\mathrm{#1}}} % Centered 2-column heads \usepackage{dcolumn} \newcolumntype{d}[1]{D{.}{.}{#1}} \newcommand{\DHEAD}[1]{\multicolumn{1}{c}{#1}} \newcommand{\HEAD}[1]{\multicolumn{2}{c}{#1}} \title{Political Ambition and Legislative Behavior in the European Parliament} \author{Steve Meserve $\quad$ Dan Pemstein $\quad$ Bill Bernhard \\ \small{University of Illinois at Urbana-Champaign}} \date{April 13, 2006} \MyLogo{} \begin{document} \maketitle \foilhead{Studying Career Ambition in the European Parliament} \begin{itemize} \item A legislator's expectations about future office affect the choices she makes while serving in her current position. \item The European Parliament (EP) houses legislators from across the EU, hailing from states with a variety of national political institutions, party systems, and political opportunity structures. \end{itemize} \foilhead{Career Ambition and Legislative Behavior in the European Parliament} \begin{itemize} \item Scarrow (1997) and Stolz (2001) find evidence for two EP career paths. Some MEPs pursue long-term careers within the EU while others seek to return home. \item When group leadership and national constituency preferences clash, nationally ambitious MEPs are more likely to vote against their group leaders than MEPs pursuing European careers. \item Nationally ambitious MEPs will focus their defections on bills that deal with the scope and pace of European integration. \end{itemize} \foilhead{Determinants of Career Ambition: Individual Level} \begin{itemize} \item Age \begin{itemize} \item We expect a curvilinear relationship between age and legislative behavior. \item Middle-aged MEPs will work to strengthen EU institutions. \end{itemize} %\item Gender % \begin{itemize} % \item Women typically have fewer opportunities to attain % national office than men (IPU 2006). % \item We expect women to defect less from group votes and to % support pro-European legislation. % \end{itemize} \end{itemize} \foilhead{Determinants of Career Ambition: National Party Level} \begin{itemize} \item Party Size \begin{itemize} \item MEPs from large parties wield comparatively more power at home and may value national careers more than small-party MEPs. \item Small party MEPs have higher within-party visibility, increasing their relative probability of obtaining national nomination. This increased expectation of success should trump the payoff difference. \end{itemize} \item Party's Presence in Government \begin{itemize} \item The career incentives generated by party government strength are similar to those produced by size. \item Parties who serve in government participate in the European Council and have a change to vet many bills that come before the EP, reducing intergroup conflict. %XXX this doesn't really make sense \end{itemize} \end{itemize} \foilhead{Determinants of Career Ambition: National Level} \begin{itemize} \item National Elections \begin{itemize} \item Nationally ambitious MEPs will vote against their groups more frequently as national elections approach, especially on legislation that expands EU power. \end{itemize} \item Electoral System \begin{itemize} \item Home-bound MEPs must please both parties and constituencies in majoritarian systems, but primarily parties under PR. \item MEPs facing majoritarian national elections will have more frequent reason to vote against their group leaderships, especially on EU-expanding legislation. \end{itemize} \end{itemize} \foilhead{Data} \begin{itemize} \item We collected roll call data from the 5th term of the EP, consisting of yes-votes, no-votes, and abstentions by 875 MEPs on 5778 votes. \item We cross-referenced roll call data with bill codings, MEP biographies and with member state and national party level data. \item Time-varying data is at day resolution. \item We control for MEP ideology, voting rule, and RCV sponsor. \item The observation level is the MEP-vote. \end{itemize} \foilhead{Data: Group Vote Defections} \begin{itemize} \item We classified group votes as those roll calls where at least two thirds of the group leadership attended and at which at least 90\% of the attending leaders voted in unison. % Note that we dropped non group votes from analysis in talk \item On simple majority votes we code only those MEPs that directly voted against their group leaders as defectors. \item MEPs defect on absolute majority votes when they vote yea on a group nay or when they vote nay or abstain on a group yea. \end{itemize} \foilhead{Method} \begin{itemize} \item The data are characterized by a complicated hierarchical panel structure. \item We include dummy variables to model the marginal effects of EP group, national party, and country on MEP deviation probability. \item To model the data's panel structure, we fit a crossed random effects model of the form: \begin{equation*} \begin{split} \text{Pr}(y_i = 1 | \vm{\beta}, \zeta_{m(i)}^{(m)}, \zeta_{v(i)}^{(v)}) &= \Phi\left[ \vm{x}_i \vm{\beta} + \zeta_{m(i)}^{(m)} + \zeta_{v(i)}^{(v)} \right], \\ \zeta_{m(i)}^{(m)} &\sim \text{N}(0, \sigma_m^2), \\ \zeta_{v(i)}^{(v)} &\sim \text{N}(0, \sigma_v^2) \end{split} \end{equation*} \item We use Bayesian estimation techniques. \end{itemize} \foilhead{Results: Overview} \begin{center} \begin{tabular}{lccc} Age & - \\ Age Squared & + \\ Young & & + & + \\ Old & & 0 & 0 \\ Natn'l Party Seat \% & - & - & - \\ Natn'l Party Cabinet \% & 0 & - & 0 \\ Natn'l Election Time\tnote{d} & - & - & - \\ Majoritarian & & & + \\ Integration Bill & + & + & 0 \\ Young $\times$ Natn'l Election Time & & - & - \\ Young $\times$ Integration Bill & & + & 0 \\ Old $\times$ Natn'l Election Time & & 0 & 0 \\ Old $\times$ Integration Bill & & - & - \\ Majoritarian $\times$ Integration Bill & & & + \\ Left-Right Ideology & + & + & + \\ Integration Ideology & - & - & - \\ Left-Right Distance & + & + & + \\ Integration Distance & 0 & 0 & 0 \\ Gender & 0 & 0 & 0 \\ Group Leader & - & - & - \\ Absolute Majority & + & + & + \\ Group RCV Sponsor & + & + & + \\ \end{tabular} \end{center} \foilhead{Results: Age} \begin{figure} \begin{center} \includegraphics{../papers/figures/ageyrs} \end{center} \end{figure} \foilhead{Results: Age, Bill Type, and National Election Timing} \begin{figure} \begin{center} \includegraphics{figures/colordummyints} \end{center} \end{figure} \foilhead{Conclusions} \begin{itemize} \item There is a link between national parties' candidate nomination strategies and the pace of European integration. \item Young MEPs have the greatest potential to return to national politics and are most likely to contravene group leaders. \item Ambitious MEPs focus their defections on legislation that determines the relative strengths of EU and national institutions. \item Timing matters: MEPs deviate more frequently as national elections near. \item Aspects of the political opportunity structure such as the size and governing status of national parties and the electoral system shape ambitious legislative behavior. \end{itemize} \foilhead{Future Research} \begin{itemize} \item Directly trace career paths. \item Investigate incentives for agenda manipulation generated by MEPs' career-oriented behavior. \item Jointly estimate ideology and ambition. \end{itemize} %%% Beyond this point are QA slides \foilhead{Descriptive Statistics} \begin{center} \begin{small} \begin{tabular}{ld{-2}d{-2}d{-2}d{-2}} \hline\hline & \DHEAD{Mean} & \DHEAD{Std. Dev.} & \DHEAD{Min} & \DHEAD{Max} \\ \hline Age & 53.07 & 9.21 & 23.12 & 79.31 \\ Gender & 0.34 & 0.47 & 0.00 & 1.00 \\ Left-Right Ideology & -0.08 & 0.48 & -0.81 & 0.82 \\ Integration Ideology & 0.26 & 0.53 & -1.00 & 1.00 \\ Left-Right Distance & 0.05 & 0.06 & 0.00 & 0.85 \\ Integration Distance & 0.14 & 0.21 & 0.00 & 1.31 \\ Group Leader & 0.07 & 0.36 & 0.00 & 1.00 \\ Natn'l Election Time & 860.99 & 462.73 & 1.00 & 1824.00 \\ Majoritarian & 0.27 & 0.44 & 0.00 & 1.00 \\ Natn'l Party Seat \% & 26.22 & 15.63 & 0.00 & 63.30 \\ Natn'l Party Cab \% & 24.02 & 36.14 & 0.00 & 100.00 \\ Integration Bill & 0.23 & 0.42 & 0.00 & 1.00 \\ Absolute Majority & 0.13 & 0.33 & 0.00 & 1.00 \\ Group RCV Sponsor & 0.19 & 0.39 & 0.00 & 1.00 \\ \hline\hline \end{tabular} \end{small} \label{tab:descriptive} \end{center} \foilhead{Results: Age, Bill Type, and National Election Timing} \begin{figure} \begin{center} \includegraphics{../papers/figures/dummypanel} \end{center} \end{figure} \foilhead{Results: Model Fit} \begin{center} \begin{tabular}{lccc} \hline\hline & Model 1 & Model 2 & Model 3 \\ \hline No Random Effects & 180,136 & 180,110 & 179,941 \\ MEP Effects Only & 176,763 & 176,740 & 176,567 \\ Vote Effects Only & 159,011 & 159,004 & 158,845 \\ Crossed Effects & 155,353 & 155,349 & 155,190 \\ \hline\hline \end{tabular} \label{tab:DIC} \end{center} %%% Not sure what to do with this... \foilhead{Data: MEP Characteristics} \begin{itemize} \item We recorded the age, gender, nationality, group memberships, leadership positions, and national party memberships of each MEP. \item We generated W-NOMINATE scores to measure legislator ideology (Poole \& Rosenthal 1985). \begin{itemize} \item Current research indicates their are two ideological dimensions in the EP: left-right and integration (Hix 2001, Hix, Noury \& Roland 2006). \item We split the dataset randomly, using half the votes to estimate ideology and the other half to test our theory. \item We consider both absolute scores and distance from median group members. \end{itemize} \end{itemize} \foilhead{Examples} \begin{itemize} \item \includegraphics[height=1in]{figures/ingo} Ingo Schmitt (Germany, CDU) \begin{itemize} \item 1981--1999 Lower house representative in Berlin. \item 1991--1998 Deputy minister in Berlin. \item 1999--2005 Member of the European Parliament. \item 2005--~~~~~~ Member of the Bundestag. \end{itemize} \item \includegraphics[height=1in]{figures/bob} Robert Goodwill (U.K., Conservative) \begin{itemize} \item 1992 Lost Redcar general election race. \item 1994 Lost Cleveland and Richmond EP race. \item 1997 Lost Leicestershire North West general election race. \item 1998 Lost South Yorkshire by-election to EP. \item 1999 Won EP seat in Yorkshire and the Humber. \item 2005 Elected to the House of Commons for Scarborough and Whitby. \end{itemize} \end{itemize} \end{document}