EITM: Electoral and Party Systemic Institutions, Structure, and Strategic Context: Empirical Implications of Theoretical Models of Effective Democratic Representation.

Regents Of The University Of Michigan - Ann Arbor, Ann Arbor MI

Investigators

Abstract

Political-economy theories note that policymakers have two goals: (a) to obtain and retain power (office seeking) and (b) to enact policies and foster outcomes they favor (policy and outcome seeking). In stable polities (i.e., excluding coercion), policymakers have four broad classes of policy they can direct toward these goals: public-good provision, broadly targeted redistribution, narrowly targeted distribution, and rent extraction. In democracies, finally, policymaking representatives use these tools to pursue these goals in the strategic context of partisan electoral competition. Substantively, this project will leverage and advance theories of comparative democracy and of comparative and international political economy to build powerful, estimable, and interpretable empirical models that directly, but manageably, reflect the multifarious and complex interactions of electoral institutions, of party and party-systemic structure, and of the strategic shape of particular partisan-electoral contests in determining the relative weight of these four classes of policy in policymaking output. Party- and electoral-system institutions, structure, and strategic context interact in quite complex ways to shape the effective representation of societal interests in the ultimate policies, but, as this project seeks to demonstrate, in ways that a nonlinear empirical-modeling strategy can encompass effectively. One venue where the effects of such complex contextual interactions in shaping effective representation should be more-observable is in the composition of policymaking activity, particularly in the redistributive (broadly targeted) versus distributive (narrowly targeted) share of budgetary activity. Institutional, structural, and strategic contexts that induce more partisan, interest-based representation foster greater redistributive emphasis, and those that induce more particularistic, geographically-based representation favor distribution. I have begun to outline such a comparative democratic political economy of budgeteering more fully elsewhere, arguing: (i) that the degree to which parties are able to act as strategic units (strategic party-unity) should determine the capacity of democratic policymakers to budgeteer, i.e., to manipulate the budget for political (i.e., electoral and partisan) purposes, (ii) that national and district-level electoral competitiveness should determine the magnitude of their incentives to budgeteer, and (iii) party-system polarization, electoral-system and district magnitude, and the degree to which parties receive their electoral support as units (representational party-unity) should determine the nature of the budgeteering that serves policymakers goals of gaining and retaining power and of producing their desired policies and outcomes. The combination of capacity and incentive size and nature, therefore, should determine the degree and character of the budgeteering I have also begun to outline an estimable nonlinear empirical model that could shed light on the multiple, complex interactions among these many causal factors in shaping fiscal policies. In this theoretical description of a substantive situation, (i) strategic capacity, c, times (ii) incentive magnitude, m, determines the amount of budgeteering we expect, while (iii) the nature, n, of the budgeteering that we expect in pursuing these goals depends on several party, party-system, and electoral-system factors that shape whether effective representation has more partisan-interest or particularistic-geographic basis and so whether policy has more redistributive or distributive nature. Thus, taking (for now) the ratio of redistributive to distributive budgetary activity, R, as the dependent variable, theory predicts that R=-+c(Xc)m(Xm)n(Xn)+-, and empirical models should reflect that proposition (EITM). In previous work, I showed how nonlinear empirical modeling can gain powerful yet interpretable leverage on the multiple, complex interactions implied by such specifications to the degrees that theory enables specification of the functions and their arguments, c(Xc), m(Xm), and n(Xn), and that these functions and/or arguments differ empirically in the empirical sample. I have so far begun to specify c(Xc), m(Xm), and n(Xn) in this new context as (i)-(iii) of the previous paragraph sketch, and I have demonstrated preliminary empirical plausibility and found some hope for empirical success by showing a role of partisan strategic capacity in shaping the relative (re)distributiveness of postwar US fiscal policy that supports the theory and showing a pattern of coefficients on partisan- and geographic-representation independent variables in explaining the policy mix in subsamples of developed democracies divided according to my own qualitative evaluation of conditions (i)-(iii) in those countries. This proposal would fund continuance of this project: (a) specifying further and in tighter connection to comparative-politics and comparative political-economy theories, c(Xc), m(Xm), and n(Xn), (b) gathering, operationalizing appropriately, and disseminating the data components of R, Xc, Xm, and Xn, (c) writing and disseminating the necessary software algorithms to estimate such models, and (d) estimating such models, and disseminating the theoretical, substantive, and methodological advances produced thereby. The intellectual merit lies both in advancing and demonstrating an inEITMlr modeling technique for bringing theory that previously evaded empirical evaluation due to its interactive complexity under direct empirical scrutiny and in whatever substantively and theoretically new light is thus shed. Given that complex multiple interaction is inherent in most classical and modern political economy and, indeed, political theory, empirical techniques that offer powerful, estimable, and interpretable leverage on such complexity are crucial. This particular approach should offer much promise for usefulness across the social sciences, where such complexly interactive causality is rife. The research project will also involve the next generation in these advances directly through RA-ships It also seeks to broaden their use in this and subsequent generations through the public provision of transparent software tools to implement such nonlinear modeling.