ESSAYS IN APPLIED INVESTMENT THEORY (MICROECONOMICS, PRODUCTION)

The thesis is an empirical investigation into the economics of costs of adjustment.;Chapter II empirically assesses the adjustment costs approach. It emerges that, at least at the total manufacturing level, costs of adjustment are "well-behaved". This important result is obtained with no assumptions made about specific optimal decision rules for investment itself. The evidence also shows that in a profit maximizing environment, flexible accelerator models are inappropriate.;Chapter III deals with the fact that, even if costs of adjustment agree with the a priori of investment theory, structural Q equations suffer from the problem that Q is not exogenous relative to investment. To be sure, whenever an observable average Q is employed such problem always arises. Only for a particular shape of adjustment costs is this difficulty avoided and the structural interpretation of the model retained. An empirical model of this type is estimated and tested using a new procedure in the literature. The results show that such model is misspecified. The conclusion is therefore that empirical Q equations are not likely to be structural models, given the underlying theoretical setting currently adopted in the literature.;Chapter I develops the notion that adjustment costs are crucial in today's most popular models of investment behavior: the rational flexible accelerator and the neoclassical version of Tobin's Q theory. For these specifications to be structural, that is for them to stem from a well-defined underlying dynamic optimization problem, adjustment costs must exhibit properties consistent with the solution of that optimum problem. Yet the converse is not true: costs of adjusting quasi-fixed inputs of production may affect the firm's operations regardless of specific optimal decision rules for investment. The Chapter expounds a number of appealing results in investment theory which hold if adjustment costs behave the way the theory demands.