Endogenous Preferences And Delegation In The European Union

• Author: Simon Hug
• DOI: 10.1177/0010414002239371
• Published date: 01 February 2003
• Date: 01 February 2003
• Subject Matter: Articles

10.1177/0010414002239371ARTICLECOMPARATIVE POLITICAL STUDIES / February-March 2003Hug / ENDOGENOUS PREFERENCES & DELEGATION

ENDOGENOUS PREFERENCES

AND DELEGATION IN

THE EUROPEAN UNION

SIMON HUG

Universität St. Gallen

The interplay among intergovernmental and supranational actors is a defining feature of the

institutional life of the European Union (EU). Too often, however, these actors are considered

independent of each other, and their autonomy is assessed in a vacuum. This is problematic

becauseif there is such a thing as “endogenous preferences” in the EU, it appears exactly through

this interdependence of intergovernmentaland supranational actors. None of the institutionalist

approaches to the EU has come to grips with this fact yet. Based on some very simple game-

theoretic ideas, I offer in this article a rationale for “endogenous preferences” and discuss their

impact on issues of delegations. Some cursory empirical evidence supports the claims that the

preferences of supranational actors are related to those of the actors who select or appoint them.

Similarly,the analyses presented here suggest that preferences over delegation to supranational

actors are influenced by differences in policy views between principals and agents.

Keywords: EuropeanUnion; delegation; Commission; endogenous preferences; appointment;

European Parliament.

The process of European integration is shaped by the interplay of inter-

governmental and supranational actors. Most scholars engaged in anyof

the various strands of institutionalist research of the European Union (EU)

would probably agree to as much. Which set of the two actors predominates,

however,is largely in dispute. Although recently the increased attention paid

to the various principal-agent relationships have heightened our understand-

ing of this debate (e.g., Garrett, 1995; Jönsson & Tallberg, 1998; Kelemen,

41

AUTHORS’NOTE: This chapter is a revisedversion of a paper presented at the second confer-

ence, “Institutionalism and the Study of the European Union: Progressand Prospects for Inte-

gration” (University of Washington,Seattle, May 4-5, 2001), and draws on an earlier draft pre-

sented at the first conferenceheld at the University of Washington, Seattle, March 24-25, 2000.

The excellent comments by the editors Jim Caporaso, JeffCheckel, and Joe Jupille, those of the

other participants at the two conferences, and those of Lars-Erik Cederman proved extremely

COMPARATIVE POLITICAL STUDIES, Vol. 36 No. 1/2, February/March 2003 41-74

DOI: 10.1177/0010414002239371

© 2003 Sage Publications

2001; Lupia, 2000; Pollack, 1997; Tallberg, 2000a, 2000b; Tsebelis &

Garrett 2001), there are still a few pieces missing.

The main missing piece, in my view,is related to the selection or appoint-

ment process of the supranational agents in the EU. All strands of

institutionalist approaches obfuscate this important element of the EU insti-

tutional setup. Thus, they miss the fact that by institutional design, some cen-

tral actors have endogenous preferences. Endogenous preferences are nor-

mally taken as phenomenon impossible to address in a rational choice

approach and thus requiring a constructivist approach (e.g., Adler, 2002;

Risse, 2000). As I argue in this article, endogenous preferences can easily be

accounted for in a rational choice institutionalist approach. Hence, at first

sight, it might appear as though rational choice institutionalism could deal

easily with endogenously determined preferences. However, as I argue,

endogenous preferences or, more precisely,changes in basic preferences are

only problematic at the individual level. To capture this problem, however, a

clear definition of preferences is required. Although constructivist

approaches may very well provide “added value” (Adler,2002) with respect

to changing preferences, the claim that endogenous or changing preferences

are impossible to address in a rational choice perspective is untenable.

Thus I present in this article an argument for why some supranational

actors’preferences are best understood as endogenous. I also present empiri-

cal evidence supporting this claim for the preferences of the Commission. In

addition, I analyze how the endogeneity of preferences affects the delegation

of tasks to supranational actors. More precisely,I show how preference diver-

gences between member states and supranational actors affect preferences of

member states over delegation. Thus, based on simple game theoretical

notions, I derive two implications dealing with important aspects of the

principal-agent relationships in the EU. In testing these two implications

with information on policy positions adopted by member states and suprana-

tional actors before the adoption of the Amsterdam treaty,I find some support

for endogenous preferences and their effect on delegationdecisions. The first

empirical test uses implicitly as competing model the widely used assump-

tion that the Commission, either because of socialization or through a neo-

functional argument, is more integration-prone than all the member states.

This assumption, proven problematic here, is adopted by both constructivist

42 COMPARATIVE POLITICAL STUDIES / February-March 2003

helpful. Thanks arealso due to Thomas König for letting me use his data set for the analyses pre-

sented here, to GeorgeTsebelis for insightful e-mail exchanges, and to Mel Hinich for method-

ologicaladvice. Partial financial support througha special research grant fromthe University of

Texasat Austin is greatlyappreciated. Correspondence concerning this article may be sent to the

author at Institut für Politikwissenschaft, Universität St. Gallen, Dufourstrasse 45, 9000 St.

Gallen, Switzerland; e-mail: simon.hug@unisg.ch.

and rational choice scholars. The second empirical model, however, uses as

null-model the hypothesis that decisions about delegation are unrelated to the

preferences of actors making the decision to delegate.

In the next section, I briefly discuss the basic definition of and assump-

tions about preferences as normally understood in rational choice

institutionalism. This leads in my view to a much clearer statement of what

endogenous preferences are and what preference changes are problematic in

a rational choice approach. In Section 3, I show how and whythe preferences

of supranational actors can be considered as endogenous. Especially if these

actors are envisioned as part of a principal-agent relationship, the nature of

appointment and delegation makes the endogeneity of their preferences

clearly visible. In Section 4, I present empirical evidence in support of these

claims. The main thrust is that contrary to the most commonly held views, the

Commission’s preferences seem to relate to the preferences of its princi-

pals—namely, the member states. This appears especially when the Com-

mission’s preferences are compared to those of the European Parliament

(EP), a supranational actor who is not appointed by the member states’ gov-

ernments. Section 5 presents evidence supporting the view that delegationto

supranational actors is affected by the divergence of preferences between

principal and agent, and Section 6 concludes.

TAKING ENDOGENOUS PREFERENCES SERIOUSLY

Jupille and Caporaso (1999) quite accurately stated that a major differ-

ence between various strands of institutionalist approaches to the EU is

whether preferences are considered to be exogenous. Perhaps, more pre-

cisely, rational choice institutionalism has very little to say about where the

fundamental preferences of actors come from and can do little with individu-

als whose fundamental preferences change.1Institutionalist approaches in

the rational choice tradition subscribe largely to Becker’s(1986) maxim that

anthropology may give us indications about a human’spreferences, and with

these preferences as given, the economic approach will explainthe outcomes

in particular situations. Thus Stigler and Becker’s (1977) adage “de gustibus

non est disputandum” forcefully applies.

Hug / ENDOGENOUS PREFERENCES & DELEGATION 43

1. von Weizsäcker(1971) derived some implications for the case in which an individual’s

preferences over consumption in one period depend on the consumption in the previousperiod.

A well-behaved utility function, implying no change in tastes, derives,however, from the more

general choice set. Similar apparent changes in preferences are discussed by Strotz (1956). In

this author’s model, the apparent changes appear, however, because of a last period effect in a

model of consumer spending with finite horizon.