Party Positions in the EP -- p5
now a well-established track record in political science for estimating the policy positions of
political actors (see for example Castles and Mair 1984, Laver and Hunt 1992, Huber and
Inglehart 1995, Benoit and Laver 2005) although none has ever been used so far to estimate
directly the policy positions of party groups in the European Parliament. Expert surveys as a
research tool are often chosen for their economy: frequently a survey of experts represents the
quickest and least expensive way to gather data on party positions. Given their relative ease of
setup, it is a fairly simple matter to survey experts at any given time point, without the setup
costs of a huge data-gathering project, detailed document coding, time-consuming interviews,
or costly opinion surveys.
Besides the practical virtue of economy, expert surveys also have several compelling
substantive advantages in the context of measuring the policy positions of European party
groups. A first compelling advantage comes from the explicitly a priori approach to locating
policy positions of the expert survey strategy. The underlying assumption is that the key
substantive policy dimensions in the European Parliament can be identified in advance of the
location of party groups, based on substantive expert understanding of potentially salient EU
policy issues. The unknowns which experts are then asked to estimate are the locations of
each party group on these a priori dimensions. The estimates of party group positions are then
taken to be the statistically aggregated judgment of the experts, on each pre-defined
dimension. Unlike factor analytic scorings, constructed scaled measures, or locations in a
purely inductive space from multi-dimension scaling analyses, expert survey summaries
eliminate the need for subjective and often ad hoc, a posteriori interpretation of results in
terms of substantive policy scales.
A second reason to use expert surveys relates to their desirable statistical properties,
namely the property that according to well-understood statistical rules, we can represent our
uncertainty about our estimates of party group positions, on the basis of both the fundamental
variability of party positions as measured by differences in expert judgments, and the
estimation variability that is determined by sample size.
Finally, especially with regard to a rapidly evolving political institution as the EP, we
regard experts as the single best source of political information on European party groups.