New Directions in Analytical Political Economy.

Some thirty years or so ago I read an article in the American Economic Review on interest rate theory, whose contents appeared to me to be both obvious and already known, seemingly published because of its attractive mathematical exposition. Although in that case the mathematics added nothing, in general mathematical formalism provides a useful set of tools. In the volume under review mathematics is used constructively to advance inquiry into certain topics and variables largely absent from conventional neoclassical economics, suggesting that the problem with the current practice of neoclassicism is not its mathematical formalism but how much is ruled out of disciplinary bounds by the adoption of certain limiting paradigmatic or particular assumptions.

The contributors to this volume present as their version of "normal science" [p. xiii] a "left structural model" which includes dynamic (evolutionary), institutional and structural variables in macroeconomic and microeconomic contexts and eagerly use mathematical formalism as one mode of expression and analysis. The contributors object to neoclassical optimization analysis, in part because the latter seeks unique results rather than an array of results each specific to different structural and distributional, as well as evolutionary, contexts; and in part because the latter is limited to optimizing behavior as such. They are willing to include, first, what neoclassicists would call social, political or anthropological variables; and second, models and themes from classical, Marxian/radical, Keynesian, Kaleckian, structuralist, and "institutional, evolutionary, behavioural and other political economy approaches" [p. 14]. The "analytical" aspect of analytical political economy refers to their use of mathematical formalism but "without giving it an absolutely privileged status over other kinds of analysis" [p. 11], though with an eye, "given the rhetorical importance of mathematical formalization in economics," to "getting wider attention and perhaps acceptance" [p. 12].

After a valuable introduction by the editor, thirteen essays (all so complex that it is impossible to do them justice; and, moreover, I am by no means expert in all the areas covered) are distributed among four parts. Part I, with essays by Alan G. Isaac, Peter Skott and Marc Jarsulic, treats unemployment, finance, and growth cycles. These papers deal with how the (natural) rate of unemployment is affected by structural and policy variables; and how the introduction of financial fragility and the interaction of two real factors, unemployment (a reserve army mechanism) and effective demand, can lead to cycles.

Part II, with essays by Dutt, Jong-Il You and Panagis S. Liossatos, is concerned with issues of accumulation, distribution, and technical change. These themes include the interaction among inflation, income distribution, accumulation and technological change; how the economy is neither inherently stable nor unstable; the role of innovation and technological diffusion; and so on, with an obvious emphasis on structural conditions, evolutionary consequences, and the existence of multiple equilibria and cycles.

Part III, with essays by Philip Mirowski, Malcolm Sawyer, and William Darity Jr. and Sumana Dhar, is concerned with value and price. These deal with a social theory of value/price, including money as a social institution; Post Keynesian pricing theory in relation to effective demand, accumulation and distribution; and a theory of price determination and terms of trade as between North and South.

Part IV, with essays by J. Mohan Rao, Jaime Ros, Edward J. Amadeo, and Bill Gibson, continues the foci of the final chapter in Part III (by Darity and Dhar), exploring economic development. These deal with multiple class-based models of rent and labor prices and their consequences for growth following W. Arthur Lewis's lead; multiple uses of investment-constraint gap models, in regard to inflation, growth and exchange rates; inflation and over-indexation as generated by social conflict; and, with regard to Nicaragua, something of a leftist version of public choice theory which deals with the relative power of capitalists and government.

At least two authors, Rao and Amadeo, show how neoclassical optimization (and comparative statics) analysis can be usefully combined with Marxist, Post Keynesian, institutionalist and structuralist analysis. Mirowski, who pushes the analysis into a social theory of price, reflects the neoclassical focus on price (vertical axis) rather than resource allocation (horizontal axis). Most contributors concentrate more on the consequences of structural variables than on their formation. So the book as a whole has a mixed relationship to neoclassical economics, though on the whole it is constructive.

The shame is that the authors' broadly Post Keynesian themes are omitted from or finessed in conventional neoclassical economics. The lines of reasoning explored in these essays conflict with, inter alia, the neoclassical quest for unique determinate, non-institutionally driven results. But they often are important and certainly deserve both the attention of open-minded economists of all schools and a recognized place within the practice of "normal science" in economics. Alas, the likelihood of their "wider attention and perhaps acceptance" is slight. Perhaps these talented and imaginative authors should, while trying for more, be content with doing their own thing and avoid the disappointment consequent to unrealistic expectations - something easier said than done.

This is to the manifest discredit of the members of a discipline which has become increasingly narrow, stylized, and insensitive. The problem is not with mathematical formalism, but with the deficiency of intellectual virtuosity and academic freedom as practiced in the discipline. So much more could be accomplished if economists in general were less interested in generating ostensibly unique determinate results and analyzing the operative factors and forces which actually govern results and doing so in an open-ended general equilibrium context. To take one example, if, given a change in prices, there can be a change in excess real balances and/or a change in liquidity preference, then what happens in the real world is a function of the relative weight of the two tendencies, an empirical matter with actual results varying over time, and with interactions; one need not focus exclusively on an all-powerful Pigou effect or Keynes effect.

Altogether this is a solid collection which is likely to receive less attention than it deserves. May I be proven wrong!

Warren J. Samuels Michigan State University