Institutional Design versus Reputational Effects on Bureaucratic Performance: Evidence from U.S. Government Macroeconomic and Fiscal Projections

Abstract

Institutional scholars often presume that the selection of a particular structural design necessarily shapes resulting policy outputs and outcomes. This phenomenon is an inherently political choice, yet we know little about its consequences regarding bureaucratic performance in a democracy. In this study, we examine whether the quality of agency outputs varies in relation to its institutional design. We empirically investigate this issue by examining the quality of current-year macroeconomic and fiscal projections generated by both presidential (Council of Economic Advisers and Office of Management and Budget) and congressional (Congressional Budget Office) support agencies, as well as an independent regulatory commission (Federal Reserve). Our statistical results uncover trivial cross-agency differences regarding bias and accuracy contained in this type of policy information. While varying degrees of political insulation might affect the durability of administrative agencies ex post, our empirical evidence rejects the politics of structural design thesis for short-term-oriented policy analysis. Specifically, bureaucratic task outputs are generally unaffected in an ex ante fashion by the extent to which an agency's institutional structure is insulated from political influence. Our empirical evidence suggests that these particular agencies' concern with reputational considerations are fairly homogeneous and thus outweigh the varying political pressures that they confront attributable to the institutional structure that they operate under.