Suchergebnisse

We sometimes observe situations in which governments - policy-makers - change policy and agents do not respond. One reason for the lack of response is the policy-maker's reputation and its effect on the credibility of the new policy. If a new policy is enacted in the face of a reputation that is exactly the opposite of the policy's intended effect or consequence, then the policy's low credibility leads to public skepticism about the new objective or threat. Agents will not alter their behavior since they believe the policy is temporary. This theory is tested by examining two recent changes in US macroeconomic policy that had implications for inflation expectations. The more general point, however, is that reputation and credibility hinder or enhance a policy-maker's effectiveness. In this particular example the public's nonresponse has the additional consequence of `prolonging the agony' of austerity policies designed to reduce inflation expectations or it can give policymakers `running room' to stimulate the economy, without a subsequent increase in inflation expectations.

This article addresses the lack of cohesion in econometric model building. This incoherence contributes to model building based on statistical criteria—correcting residuals—and not theoretical criteria. The models we build, therefore, are not valid replications of theory. To deal with this problem, an agenda for model building is outlined and discussed. Drawing on the methodological approaches of Hendry, Qin, and Favero (1989), Hendry and Richard (1982, 1983), Sargan (1964), and Spanos (1986), this agenda incorporates a "general to simple" modeling philosophy, a battery of diagnostic tests, reduction theory, and the development of models that include short-term and long-term parameters. A comparison is made between a model based on this agenda and a model based on corrected residuals. The findings show that the agenda-based model outperforms the residual correction model.

Part of a symposium on the future direction of political science research begins by citing some problems with formal models, case studies, & applied statistical modeling despite the fruit they have borne. In this light, the initiative to link formal & empirical analysis -- empirical implications of theoretical models (EITM) -- is described. The lax manner in which researchers typically operationalize their concepts & establish causation is addressed in terms of the scientific & social impacts of common political science research practices. Examples of EITM-type research include party identification, policy making & the Phillips curve, strategic interaction in international relations, & regulatory policy delay. 1 Appendix, 45 References. J. Zendejas