Political Stabilization by an Independent Regulator

• Published date: 01 April 2007
• DOI: https://doi.org/10.1016/S0573-8555(06)82015-8
• Date: 01 April 2007
• Pages: 383-415
• Author: Antoine Faure-Grimaud,David Martimort

CHAPTER 15

Political Stabilization

by an

Independent Regulator

Antoine Faure-Grimaudaand David Martimortb

aLondon School of Economics, FMG and CEPR

bUniversité de Toulouse,and IUF

Abstract

This chapter analyzes the relationship between elected partisan political princi-

pals with partisan objectives and their regulators when those regulators can be

captured by an interest group, namely the firm they are supposed to regulate.

Independence of the regulator stabilizes regulatory policies and avoids much of

the fluctuations induced by an exogenous political uncertainty on the electoral

outcomes: a stabilization effect. However, independence also increases the cost

of preventing regulatory capture: an agency cost effect. Even when both effects

are taken into account, regulatory independence still increases ex ante social

welfare. We also investigate how the independence of the bureaucracy affects

electoral outcomes when political uncertainty is endogenized by modeling the

decision of forward-looking voters who compare the policy platforms offered

by competing political principals. Endogenizing political uncertainty reinforces

the stabilization effect.

Keywords: regulation, bureaucracy, organization of the government, partisan

politics

JEL classifications: D82, L51

15.1. Introduction

There seems to be a consensus, not only amongst political scientists but also

between politicians and the electorate, to recognize that economic policies re-

main relatively stable even though political parties alternate in power. This phe-

nomenon also appears to be more pronounced in areas where political powers

engage in substantial delegation to more or less independent agencies. Delega-

tion is indeed prevalent in many countries for a wide range of economic deci-

sions. A large body of empirical literature focusing on the American case has

documented extensive delegation from Congress to administrative non-elected

CONTRIBUTIONS TO ECONOMIC ANALYSIS © 2007 ELSEVIER B.V.

VOLUME 282 ISSN: 0573-8555 ALL RIGHTS RESERVED

DOI: 10.1016/S0573-8555(06)82015-8

384 A. Faure-Grimaudand D. Martimort

agencies.1Out of a data set of 257 important pieces of legislation between

1947 and 1990, which on average comprehend 45 major provisions, Epstein

and O’Halloran (1997) computed for instance delegation ratios by counting the

percentage of provisions of these bills which one way or the other involved

delegation of authority by Congress. Eighty-one percent of all laws delegated

authority to at least one cabinet department, thirty-eight percent to indepen-

dent regulatory agencies, sixteen percent to independent commissions. Fifty-two

percent of the laws in their sample created at least one new agency, board or

commission to whom substantial authority was delegated. If the European Cen-

tral Bank remains the most spectacular example of delegation to a new European

institution, the European Union has also created a dozen of independent agen-

cies over the last thirty years or so. Those agencies differ greatly in terms of the

procedural requirements, the membership of their management board and the

role of member states in the nomination of directors. For instance, in the field

of merger control, the European Commission was delegated the competence to

regulate mergers under the 1989 Merger Control Regulation.2

This chapter provides an explanation for a possible link between the degree

of independence of a regulatory agency and policy stabilization. The first key-

ingredient of our model is the possibility of capture of these agencies by the

interest groups they regulate. Indeed, in a world of informational asymmetries,

regulators accumulate information about the welfare effect of different policies

and they can be bribed for manipulating information. The second ingredient is

political uncertainty. The extent to which an agency is affiliated to a political

party affects the likelihood that this particular regulator remains in place as po-

litical powers alternate in office. The status of the regulatory agency changes

thus significantly the collusive opportunities between this regulator and the in-

dustry.

In a nutshell, we argue below that regulatory independence, although it stabi-

lizes policies and avoids unnecessary fluctuations due to pressures by politically

elected principals, also increases the scope for regulatory capture. Nevertheless

we can identify circumstances where the first effect dominates, justifying the

constitutional choice of this institutional mode.

Interestingly, this set of results might shed new light on one of the most doc-

umented episode of merger regulation in Europe: the De Haviland case. After a

complete two-stage investigation, the Merger Task Force recommended in Sep-

tember 1991 that the merger between the Franco-Italian ATR and De Haviland

1See Kiewert and McCubbins (1991).

2Another well-known example of delegation is given by environmentalpolicy. When passing the

Clean Air Act, the U.S. Congress did not establish exactly what concentrations of harmful substances

are permissible in the air but, instead, delegated to the EnvironmentalProtection A gencythe author-

ity to do so. In the U.S., other examples of independent agencies include The Interstate Commerce

Commission made independent in 1889, the Federal Trade Commission (1914), the Federal Com-

munication Commission (1934), the Securities and Exchange Commission (1972), the Consumer

Product Safety Commission (1972) and the Nuclear Regulatory Commission (1974).

Political Stabilization by an Independent Regulator 385

be rejected on grounds that it would create a dominant position for the combined

firm in the relevant market. Even though this stance was heavily criticizedat the

time by both the French and Italian governments, the mechanisms available to

these countries to control bureaucratic drift3were weak and Commissioner Leon

Brittan could argue that “nobody wanted a system that only served the interests

of the countries that shout loudest”4signifying thereby the degree of formal in-

dependence of the Merger TaskForce vis-à-vis member states and advocating de

facto its depolitization. However and beyond the De Haviland case, other critics

of the European Union merger control have also documented the lack of trans-

parency of its decisions and its lax enforcement,5two ingredients which clearly

point at the possible capture of this agency.

In our model, we examine two possible status for regulatory agencies, captur-

ing stylized views of real life institutions. First, the regulator may be affiliated

to his political principals and be removed from office each time a new political

principal is elected. Second, the regulator may be independent and keeps office

whoever political principal gets elected.6This difference in the independence

degree of an agency affects the set of collusive agreements which can be signed

between the interest group and his regulator. In the first case, collusion can only

occur ex post, once political principals have been elected. In the second case,

collusion can also occur ex ante, before the political principals get elected. This

latter possibility enlarges the set of collusive agreements and thus increases the

agency cost of capture.

In this context, our main result is that an independent regulator plays a stabi-

lizing role in the political process. When collusive side-contracting suffers from

some transaction costs which are convex with the size of the bribe exchanged,7,8

the independent agency wants to smooth the possible bribes it may receive over

the possible electoral outcomes. The agency cost of capture depends now on an

average between the regulatory stakes that either principal would like to imple-

ment. To react to the threat of capture of this independent regulator, political

3On this issue see Weingast and Moran (1983) and McCubbins et al. (1987, 1989).

4Quotation taken from Pollack (1996, p. 294).

5See Wilks and McGowan (1995).

6Most of the time, in the real world, the board of regulatory agencies is made of a fixed number

of members with staggered terms who are removed by the elected government once those terms

are finished. Our dichotomy can also correspond to policy implementation by different types of

agencies. As documented before, there is some variance in the delegation ratios according to the

type of agency to whom policies are delegated. For instance, there is no doubt that the cabinet

departments are less independent from political powers than the Federal Trade Commission.

7For instance, these transaction costs can be viewed as a reduced form for the possibility that

collusion may be detected which could result in collusive partners facing heavy penalties. Our as-

sumption is then that the probability of detection increases with the size of the bribes exchanged.

Implicit in this formulation is the idea that bigger frauds are more easily detected by the principal.

8In Faure-Grimaud and Martimort (2001), we offer microfoundations for this assumption. Here,

we take as given the technology of side-contracting and do not derive it from more fundamental

assumptions. This short-cut turns out to be necessary to introduce political uncertainty in a simple

way.