Bank Supervision and the Limits of Political Influence over Bureaucracy

Abstract

The relationship between elected officials and unelected bureaucrats is one of recurring interest to students of government. It is widely asserted (albeit incorrectly, see Rabin and Bowman, 1998; Svara, 1998) that the first generation of administration scholars (e.g., Wilson, 1887; Goodnow, 1900) believed that bureaucrats could and should be absolutely controlled by elected officials. Politicians made policy, and civil servants applied their neutral, nonpartisan expertise to carry out the wishes of their elected superiors in the most efficient manner possible. This politics/administration dichotomy, however, later gave way to decidedly less optimistic theories about the ability of elected officials to the bureaucracy. According to this literature, bureaucrats enjoyed considerable, if not almost complete, independence from the macropolitical system. Congress as a whole, the president, and appointed officials played minor roles, at best, in the day-to-day operations of the bureaucracy (Cater, 1964; Freeman, 1965; McConnell, 1966). Moreover, bureaucrats were primarily interested in maximizing their budgets (Downs, 1967; Niskanen, 1971; 1975), and they possessed numerous advantages (e.g., longevity, superior information and expertise, civil service protections, and over the flow of information) relative to elected officials such that they had substantial discretion to pursue their own interests (Redford, 1969; Rourke, 1976; Heclo, 1977; Loci, 1979). Meanwhile, the literature on Congress argued that the members of that institution were more concerned with satisfying their electoral constituencies than with monitoring the bureaucracy. Congressional oversight, where it occurred, was criticized as uncoordinated, episodic, ad hoc, and ineffectual (Bibby and Davidson, 1972; Derthick, 1990; Dodd and Schott, 1979; Ogul, 1976; Scher, 1960; Sundquist, 1981). Thus the policy process had come to be dominated by an iron triangle of bureaucrats, interest groups, and congressional subcommittees that generated policies that benefitted themselves and ignored the concerns of the at large (Lowi, 1979; McConnell, 1966; see also, Ripley and Franklin, 1991). By the 1980s, however, theories of political of the bureaucracy had come almost full circle, back to a view that espoused great faith in the ability of political officials to the bureaucracy. A growing body of theoretical and empirical research--much of it grounded in the principal-agent theory of political/bureaucratic relations--claimed to demonstrate numerous instances of effective political of the bureaucracy (e.g., Moe, 1984; Wood and Waterman, 1991). According to this public literature, methods of included the creation of carefully structured procedural and reporting requirements, direct intervention with an administrative agency, or the granting of legal standing to select interest groups. At the same time, a number of empirical studies found that agency outputs were associated with such factors as the activities of oversight committees, the political ideology of the oversight committees, the shared powers of political appointments and the budget, the ideology of the president, and court decisions. (For a good review of the choice literature referenced above, see Bendor, 1990; Wood and Waterman, 1991; Worsham, Ringquist, and Eisner, 1997.) The empirical evidence for political of the bureaucracy appeared so robust that Wood and Waterman (1991, 822) concluded, We believe this evidence for active political is so strong that controversy should now end over whether political occurs. Future research should turn toward exploring the determinants of political control (emphasis in the original). A final explanation for political/bureaucratic relations is the so-called bureaucratic culture model, which assumes that bureaucrats possess substantial policy-making discretion, and that they use this discretion in a manner that best meets their prior construction of what good policy should look like. …