Abstract
The European Parliament is one of most prominent substantive applications of NOMINATE to the study of roll call voting outside the U.S., yielding tremendous insights into the voting patterns of the world’s most important transnational parliament. However, this body of research cannot facilitate comparisons of ideological shifts over time, because it exclusively employs scaling models that are static. In this paper, I produce dynamic ideal point estimates for the first six European Parliaments from 1980 to 2009 that can be compared over time. These estimates show a significant amount of ideological shifting for some Members of the European Parliament. I explain the measurement strategy, and compare cross-sectional estimates to existing measures as a validity check. I also offer three applications highlighting the types projects that scholars of the European Parliament might wish to use these dynamic measures to study further.
This is a preview of subscription content, log in via an institution to check access.
Similar content being viewed by others
Notes
Hix et al. (2009, pg. 25) write that “voting in the European Parliament is predominantly one-dimensional, and increasingly so. The W-NOMINATE scores on dimension 1 correctly predict approximately 85% of votes in the first elected European Parliament and approximately 90% in the fifth elected European Parliament, while the scores on dimension 2 only predict an additional 6% in the first Parliament and 2% in the fifth.”
A partial exception to this is Lo (2013a), who estimates dynamic ideal points for Irish MEPs using a roll call discontinuity design. This paper explicitly compares ideal points for Irish MEPs at three different points in time, and finds that their voting records shift to the right following the Irish electorate’s rejection of the Treaty of Nice, then shifts back to the left following its ratification in a subsequent referendum. To the best of my knowledge, this is the only study that has estimated inter-temporally comparable ideal points for the European Parliament.
I simulated linear trends from Monte Carlo simulations and estimated ideal points from them using the dynamic linear model here, and find that the ideal points are highly correlated even under those conditions.
Data from the 7th European Parliament are not available.
I drop Innocenzo Leontini, an MEP belonging to Forza Italia, from the data, due to his low frequency of voting.
At this point it is worth pausing to consider the sheer computation magnitude that estimation of dynamic EP ideal points using MCMC would entail. In this data set, \(N=2710\) MEPs express preferences on \(\sum J_{it} = 21{,}438\) roll call votes over \(T=59\) periods (explained below), resulting in 35,693 estimated ideal points. In comparison, the Martin–Quinn scores estimate approximately \(9\, ^*\, 77=693\) ideal points in total. Assuming that estimation runtime is proportional to the number of ideal points and bills, the EP data set estimates more than 50 times more ideal points with four times as many roll call votes. Based on the five-day runtime cited above, the implied runtime of the dynamic IRT model using standard MCMC is approximately 3 years—a runtime that is impractical.
Imai et al. (2016) also develop variational algorithms for a series of other ideal point models, including the basic two-parameter IRT model, an ordinal IRT model, a hierarchical IRT model, as well as scaling models for text and network data.
I do not discard these 10 votes, but instead include them as part of the voting record of MEPs in the first half of 1980.
There is no clear rule on what innovation variance to select, and Martin and Quinn give little guidance on this matter. For this application I made the determination based on examining different versions of Fig. 4 that used different variance parameters, and selected one that looked “reasonable”. Larger values will produce rougher estimates, and smaller values will produce smoother ones. In both my paper and in Martin and Quinn’s application, a key finding is that there is significant trending of legislators even when innovation variances are very small, and in this spirit, I also tested smaller innovation variances of \(\omega _x^2=0.001\). Notably, certain MEPs (such as Marco Panella in Fig. 3) still exhibit significant trending even when \(\omega _x^2\) is set that small.
Other right leaning groups set at this value include the European Democrats, European Right, and Independence/Democracy. For the purpose of this paper, when mentioning any EPG I include all EPG predecessors as coded in the ideal point data provided. For example, UEN refers not only to Union for Europe of the Nations, but also includes the Progressive European Democrats, European Democratic Alliance, and Union for Europe.
Notably, this includes members of Rainbow Group, Left Coalition, and Technical Group of Independents
These goodness-of-fit metrics are not implemented in either Martin and Quinn or Imai et al’s paper, but are available for NOMINATE, and the formulae to calculate them is in Poole and Rosenthal (1997).
Roll call data for the sixth EP were available on Simon Hix’s website and are used in this paper. However, they do not publish ideal point estimates for that EP, so I do not include that comparison here.
Some of the differences in Table 1 are also likely due to differences that result from using a quadratic utility function (Carroll et al. 2009a). As a test of this, I estimated static ideal points for EP5 using Clinton et al. (2004), which also uses a quadratic utility function but omits the dynamic linear model. This model also placed the Greens slightly to the left of GUE/NGL, as the dynamic IRT model does.
In cases where one wishes to break this autoregressive relationship when estimating inter-temporally comparable ideal points, a roll call discontinuity design is generally preferred. Researchers may prefer this approach in cases where they are modeling the effect of a sharp, exogenous shock on roll call votes. See Lo (2013b) for an application of this strategy.
Because of the “backwards sampling” component of the dynamic linear model, \(x_{it} \mid x_{i,t+1}\) as well. Thus, placing the UEN to the left of the EPP in the dynamic IRT model is even more likely because the UEN lies to the left of the EPP even in the NOMINATE estimates in session 3.
Comparisons of ideal points over time using W-NOMINATE either entail making the theoretical assumption that legislators never shift positions over time (if one estimates a constant ideal point model), or assumes that an ideal point of 0.7 measured using data in one legislature is exactly comparable to an ideal point of 0.7 measured using data in another legislature. Since the bills across legislatures differ, this assumption is highly unlikely to be true in most cases.
Marco Panella hops between Rainbow Coalition and Independents before settling with ALDE in the sixth EP, but his voting record shifts largely between the first and second EP.
This perspective is consistent with the finding that with national parties use list placements to reward MEPs on powerful committees (Frech 2016).
The six MEPs here are Friedrich-Graefe zu Baringdorf (Germany), Juan Mara Bandrs Molen (Spain), Nel van Dijk (Netherlands), Paul Sates (Belgium), Herman Verbeek (Netherlands), and Wilfried Telkmper (Germany).
Of the six treatment MEPs, only three remain in the EP after the third EP.
As an alternative, party positions coded using the Comparative Manifesto Project coding scheme with Euromanifestos are also available (Schmitt and Wüst 2012).
The Green EPG does not exist in either the first or second EP, so those cases cannot be included in the analysis here.
References
Achen, C. H. (1977). Measuring representation: Perils of the correlation coefficient. American Journal of Political Science, 21(4), 805–815.
Aldrich, J. H. (1995). Why parties? The origin and transformation of political parties in America. Chicago: University of Chicago Press.
Budge, I., Klingemann, H.-D., Volkens, A., Bara, J., & Tanenbaum, E. (2001). Mapping policy preferences: Estimates for parties, electors, and governments, 1945–1998. Oxford: Oxford University Press.
Carroll, R., Lewis, J. B., Lo, J., Poole, K. T., & Rosenthal, H. (2009a). Comparing nominate and ideal: Points of difference and Monte Carlo tests. Legislative Studies Quarterly, 34(4), 555–591.
Carroll, R., Lewis, J. B., Lo, J., Poole, K. T., & Rosenthal, H. (2009b). Measuring bias and uncertainty in dw-nominate ideal point estimates via the parametric bootstrap. Political Analysis, 17(3), 261–275.
Carroll, R., Lewis, J. B., Lo, J., Poole, K. T., & Rosenthal, H. (2013). The structure of utility in spatial models of voting. American Journal of Political Science, 57(4), 1008–1028.
Clinton, J., Jackman, S., & Rivers, D. (2004). The statistical analysis of roll call data. American Political Science Review, 98(02), 355–370.
Cooper, J., & Brady, D. W. (1981). Institutional context and leadership style: The house from Cannon to Rayburn. American Political Science Review, 75(2), 411–425.
Cox, G. W., & McCubbins, M. D. (2005). Setting the agenda: Responsible party government in the US house of representatives. Cambridge: Cambridge University Press.
Downs, A. (1957). An economic theory of democracy. New York: Harper and Row.
Epstein, L., Martin, A. D., Quinn, K. M., & Segal, J. A. (2007). Ideological drift among supreme court justices: Who, when, and how important. Northwestern University Law Review, 101, 1483.
Frech, E. (2016). Re-electing meps: The factors determining re-election probabilities. European Union Politics, 17(1), 69–90.
Han, J.-H. (2007). Analysing roll calls of the european parliament: A Bayesian application. European Union Politics, 8(4), 479–507.
Hare, C., Armstrong, D. A., Bakker, R., Carroll, R., & Poole, K. T. (2015). Using Bayesian Aldrich–Mckelvey scaling to study citizens’ ideological preferences and perceptions. American Journal of Political Science, 59(3), 759–774.
Harrison, J., & West, M. (1999). Bayesian forecasting and dynamic models (Vol. 1030). New York City: Springer.
Hix, S. (2001). Legislative behaviour and party competition in the European Parliament: An application of nominate to the EU. JCMS. Journal of Common Market Studies, 39(4), 663–688.
Hix, S. (2002). Parliamentary behavior with two principals: Preferences, parties, and voting in the European Parliament. American Journal of Political Science, 46(3), 688–698.
Hix, S. (2004). Electoral institutions and legislative behavior: Explaining voting defection in the European Parliament. World Politics, 56(2), 194–223.
Hix, S., & Noury, A. (2009). After enlargement: Voting patterns in the sixth European Parliament. Legislative Studies Quarterly, 34(2), 159–174.
Hix, S., Noury, A., & Roland, G. (2005). Power to the parties: Cohesion and competition in the European Parliament, 1979–2001. British Journal of Political Science, 35(2), 209–234.
Hix, S., Noury, A., & Roland, G. (2006). Dimensions of politics in the European Parliament. American Journal of Political Science, 50(2), 494–520.
Hix, S., Noury, A., & Roland, G. (2007). Democratic politics in the European Parliament. Cambridge: Cambridge University Press.
Hix, S., Noury, A., & Roland, G. (2009). Voting patterns and alliance formation in the European Parliament. Philosophical Transactions of the Royal Society B, 364(1518), 821–831.
Ho, D., & Quinn, K. (2010). Did a switch in time save nine? Journal of Legal Analysis, 2, 69.
Høyland, B. (2010). Procedural and party effects in European Parliament roll-call votes. European Union Politics, 11(4), 597–613.
Imai, K., Lo, J., & Olmsted, J. (2016). Fast estimation of ideal points with massive data. American Political Science Review, 110(4), 631–656.
Jordan, M. I., Ghahramani, Z., Jaakkola, T. S., & Saul, L. K. (1999). An introduction to variational methods for graphical models. Machine Learning, 37(2), 183–233.
Lo, J. (2013a). An electoral connection in European Parliament voting. Legislative Studies Quarterly, 38(4), 439–460.
Lo, J. (2013b). Legislative responsiveness to gerrymandering: Evidence from the 2003 Texas redistricting. Quarterly Journal of Political Science, 8(1), 75–92.
Lo, J., Proksch, S.-O., & Gschwend, T. (2013). A common left–right scale for voters and parties in Europe. Political Analysis, 22(2), 205–223.
Martin, A. D., & Quinn, K. M. (2002). Dynamic ideal point estimation via Markov chain Monte Carlo for the US supreme court, 1953–1999. Political Analysis, 10(2), 134–153.
Martin, A. D., & Quinn, K. M. (2007). Assessing preference change on the U.S. supreme court. The Journal of Law, Economics, & Organization, 23(2), 365–385.
Martin, A. D., Quinn, K. M., & Epstein, L. (2004). The median justice on the United States supreme court. North Carolina Law Review, 83, 1275.
McElroy, G., & Benoit, K. (2010). Party policy and group affiliation in the European Parliament. British Journal of Political Science, 40(02), 377–398.
McFadden, D. (1973). Frontiers of econometrics. In P. Zarembka (Ed.), Chapter conditional logit analysis of qualitative choice behavior (pp. 105–142). New York: Academic Press.
Noury, A. G. (2002). Ideology, nationality and Euro-Parliamentarians. European Union Politics, 3(1), 33–58.
Poole, K. (2005). Spatial models of parliamentary voting. Cambridge: Cambridge University Press.
Poole, K. T. (1998). Recovering a basic space from a set of issue scales. American Journal of Political Science, 42(3), 954–993.
Poole, K. T. (2000). Nonparametric unfolding of binary choice data. Political Analysis, 8(3), 211–237.
Poole, K. T. (2017). The scientific status of geometric models of choice and similarities judgment. Public Choice, 171(3–4), 245–256.
Poole, K., & Lewis, J. (2004). Measuring bias and uncertainty in ideal point estimates via the parametric bootstrap. Political Analysis, 12(2), 105–127.
Poole, K., Lewis, J., Lo, J., & Carroll, R. (2011). Scaling roll call votes with W-nominate in R. Journal of Statistical Software, 42(14), 1–21.
Poole, K. T., Lewis, J. B., Rosenthal, H., Lo, J., & Carroll, R. (2016). Recovering a basic space from issue scales in r. Journal of Statistical Software, 69(7), 1–21.
Poole, K. T., & Rosenthal, H. (1985). A spatial model for legislative roll call analysis. American Journal of Political Science, 29(2), 357–384.
Poole, K., & Rosenthal, H. (1991). Patterns of congressional voting. American Journal of Political Science, 35, 228–278.
Poole, K., & Rosenthal, H. (1997). Congress: A political-economic history of roll call voting. New York: Oxford University Press.
Poole, K. T., & Rosenthal, H. (2001). D-nominate after 10 years. Legislative Studies Quarterly, 26(1), 5–29.
Proksch, S.-O., & Lo, J. (2012). Reflections on the European integration dimension. European Union Politics, 13(2), 317–333.
Rohde, D. W. (2010). Parties and leaders in the postreform house. Chicago: University of Chicago Press.
Schmitt, H., & Wüst, A. M. (2012). Euromanifestos project (emp) 1979–2004. GESIS Data Archive, Cologne. ZA4457 Data file Version 1(0).
Author information
Authors and Affiliations
Corresponding author
Additional information
I gratefully acknowledge the helpful comments from my referees and editors, and especially Keith Poole, whose mentorship helped make this research agenda possible.
Rights and permissions
About this article
Cite this article
Lo, J. Dynamic ideal point estimation for the European Parliament, 1980–2009. Public Choice 176, 229–246 (2018). https://doi.org/10.1007/s11127-018-0551-3
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s11127-018-0551-3