Representation as a median mandate: Taking cross‐national differences seriously

Abstract

The extent and ways in which popular preferences influence government policy are absolutely central to our understanding of modern democracy. Paul Warwick's discussion of these in the European Journal of Political Research in 2010 puts itself at the heart of the debate with its critique of the median mandate theory of McDonald and Budge, proposing an alternative ‘bilateralist’ concept of representation. This article questions whether this concept has much to add to our theoretical understanding of representational processes. However, Warwick's further conceptual points deserve serious consideration. These concern the time horizons within which representative processes work, and the status of the median position given multi‐motivated voting. At the evidential level, Warwick argues that survey‐based measures of voter and party left–right positions fail to produce the correspondence between median and government policy positions that median mandate theory would have us expect. However, survey‐based measures of median voter and party placements obscure important cross‐national variation. Using the Comparative Study of Electoral Systems (CSES 2007), as Warwick does, this article shows that survey respondents norm their own and their country's party positions to their national context. The consequence is to make the political centre in all nations appear similar. Allowing for the relevant cross‐national differences brings the relationship between the median voter and government position back in line with expectations.

Jettisoning Illusions About the Median Mandate

“…For example, while the CSES has the Norwegian and Australian Labour parties located in virtually the same position, the CMP‐associated analysis shows Norwegian Labour located more than 2.5 units to the left of Australian Labour on the 0–10 CSES metric (see Best , 44). After adjusting both the CSES's MV and party positions by adding the missing cross‐national differences and rerunning an earlier Warwick responsiveness analysis, we found the relationship between government and MV left‐right positions is statistically indistinguishable from one‐to‐one (Best, Budge, and McDonald ). In short, the median mandate's theoretical expectation of a one‐to‐one relationship holds up in the survey data once the missing cross‐national variation is taken into account.…”

“…If the reasoning behind the one‐to‐one relationship is plausible, why does Warwick's survey‐based evidence not reveal it? As we have explained elsewhere (Best 2012; Best, Budge, and McDonald ), survey data implausibly norm most countries' political centers so they record something close to the same MV position in every country. Warwick tells us as much; he reports that the median citizens in most countries have a score of “5” (Warwick , 449).…”

“…To examine the influence of the norming on the relationship between voter and party positions in the CSES survey data, we have elsewhere leveraged the Comparative Manifesto (CMP, aka MARPOR) data to retrieve the missing cross‐national differences. We regressed the CSES‐recorded party positions on the CMP‐recorded party positions plus a set of country dummy variables (Best, Budge, and McDonald ). The coefficients on the country dummies provide an estimate of missing CSES cross‐national differences.…”

Responsive to the People? Comparing the European Cognitive Maps of Dutch Political Leaders and their Followers

“…This seems to be the case both in terms of the pro/anti-European positions, left/right ideological leanings as well as the prioritisation of issues. However, within this general pattern, national, partisan, and individual-level variations in the strength of the party/voter connection do exist (Best, Budge, & Mcdonald, 2012;Ray 2003;Steenbergen et al, 2007): the congruence between leaders and the public is stronger in proportional representative systems. Moreover, the likeliness of voters adopting party elites' positions is increased when the voter is highly attached to the party (Lindeboom, 2012;Ray, 2003).…”

Parties' Influence during Government Policy Negotiations: Parliamentary Dynamics and Spatial Advantages in the First Italian Republic

“…The latter theory argues that, whatever the dynamic of competition, the party supported by the median voter will be included in the government and pull government policy towards its own policy position thanks to its strong bargaining power (see also Best et al, 2012). Note, however, that the median parliamentary party in the McDonald and Budge model is estimated along a one-dimensional space.…”