The Impact of Financial Flows on U.S. Investment, 1948-1992: An Empirical Model of Institutional Investment Theory
Carrier, David, Marsh, Lawrence, Journal of Economic Issues
The role of investment in contributing to weak and negative GDP growth is well established by Keynesian theory. There are substantial differences among theories of the determinants of investment behavior, however. Post Keynesian models rely on capital costs and financial constraints in a flexible accelerator framework, but recent experience suggests that other institutional factors may play a greater role. Despite the lowest capital costs and most stimulative monetary policy since the 1960s, the rate of new investment in the early 1990s was lower than ever experienced during the postwar period.
The institutional view is that investment is responsive to economic conditions and financial flows, which are based in the institutions and practices of commerce and banking. Hyman Minsky distinguishes two types of changes in financial institutions that lead to investment fluctuations: legislated and evolutionary. The former occur because of the perceived necessity of reform, often in response to adverse evolutionary institutional change. Evolutionary change involves the financing practices of firms and the banking system and occurs because of profit opportunities that arise during business cycle expansions, when intense economic activity and credit demand lead to an expansion of the supply and the cost of money and near-money.
Specific evidence of the mechanism whereby financial difficulties translate into investment and GDP declines is described and then followed by an empirical specification of the historical pattern. The model links firms' investment behavior to relevant measures of financial conditions using a semiparametric method of nonlinear estimation. Linear econometric methods have come under attack in recent years for simplifying assumptions and inaccurate representation of economic processes. Nonlinear technique is demonstrated to be better suited for describing evolutionary processes and for analyzing the impact of economic policies and institutional reform on investment behavior and the business cycle. Institutional theory provides a rich verbal description of the pattern of behavior by which investment and output fluctuations follow from changes in the financial flows of firms and the banking system. Quantification of these relationships is generally avoided because of the lack of modeling and estimation techniques that retain the complexity of the actual relationships. An example is provided here of an evolutionary pattern model, which presents the dynamic element in terms of changes in institutional arrangements, conditions, and behavior through time. The decision process of economic agents is modeled, and since these agents look to changes in economic indicators such as stock and bond prices and credit availability for the information they need to project expectations into the future, quarterly flows of these variables are used in the estimation procedure.
This method stands in contrast to the usual static approach that is concerned with objective functions of more abstract elements removed from historical time and conditions. Traditional methods of quantifying economic relationships and of econometric estimation have been questioned recently, for very good reason. Mathematical properties, such as homogeneity, convexity, and transitivity, that may hold at the individual level do not necessarily apply in the aggregate, so that there is no macro analog to the microeconomic foundations of a theory. Overly simplified assumptions and conditions, such as the existence of a single market clearing price, may be specified that do not accord with real-world experience. These conditions can also vary from one theoretical problem to another as in game theory where results are often sensitive to varying initial assumptions, making general conclusions impossible.
These issues have led to an overhaul of the theory and mathematical representation of markets. Duncan Foley's  statistical market equilibrium is a recent example of a radically different conception of bow markets work. …