Knowledge that Transforms

Cramer, H., Mathematical Methods of Statistics (Princeton: Princeton University Press, 1946). Pesaran, M. H., Alternative Econometric Approach to the Permanent Income Hypothesis: an International Comparison: A Comment, this REVIEW, LV (1973), 259-261. Rao, B. B., Alternative Econometric Approach to the Permanent Income Hypothesis: An International Comparison, this REVIEW, LV (1973), 261. Singh, B., Maximum Likelihood Estimation of the Friedman Consumption Function, this REVIEW, LVII (Feb. 1975). Singh, B. and H. Drost, Alternative Econometric Approach to the Permanent Income Hypothesis: an International Comparison, this REVIEW, LIII (Nov. 1971), 326-334. Tukey, J. W. Components in Regression. Biometrics (1951). Willassen, Y., Specification and estimation of the original P.I.H. (Working Paper) 1972. Wold, H., Nonlinear Estimation by Iterative Least Square Procedures. Festschrift for J. Neyman. Wiley and Sons, 1966.

Edwin J. Elton, Martin J. Gruber, Paul R. Kleindorfer, A Closer Look at the Implications of the Stable Paretian Hypotheses, The Review of Economics and Statistics, Vol. 57, No. 2 (May, 1975), pp. 231-235

T HE Index of Consumer Sentiment is one of the oldest and most respected measures of consumer psychology. The contribution of the Index to the prediction of aggregate consumer durable spending was first examined by Mueller and Katona (1953, 1957). The Index has been used in several econometric models of the U.S. economy (Friend and Taubman, 1964; Duesenberry et al., 1965; Evans, Klein and Schink, 1967). However, the contribution of the Index is only marginally significant in most of these studies, if account is taken of stocks of durables and relative prices. Friend and Adams (1964) have noted that several cyclical variables the average work week in manufacturing, stock market prices, and the unemployment rate -appear to be close substitutes for the Index in automobile demand equations, yielding approximately-equal coefficients of determination when specified with similar lag structures. More recently, Burch and Stekler (1969) compared forecasts of consumer durable expenditure based on income and attitudes with those generated by naive strategies. The use of a serial correlation correction of the error term proved necessary to generate forecasts superior to an assumption of no change in durable sales. Even so, the best fitting equation they estimated was unable to forecast turning points. Thus, the recent evidence suggests to many econometricians that attempting to use expectational variables in large scale econometric models is more trouble than it's worth.1 This is especially true since more convenient variables, such as the unemployment rate, or the average work week in manufacturing, appear to substitute so well. Saul Hymans expressed this sentiment accurately:

The emphasis here is on a micro-economic derivation of the theoretical propensity to consume windfalls under limited borrowing capacity. This is done in the next section by means of a comparative static analysis employing a reduced intertemporal utility function. The final section reports a test whose results conform to the prediction of the analysis that the propensity to consume windfalls is negatively related to size of the windfall.

T HE standard neoclassical investment analysis postulates a fixed timing relationship between changes in the determinants of the desired stock of capacity and actual investment responses these changes elicit.' There are, however, good reasons to suppose that no such fixed timing relationship exists. This paper seeks to explore one class of variables which could influence firms' decisions concerning the time shape of the investment response to a given change in the desired stock of capacity. The influence of financial conditions on firms' investment decisions has long been recognized.2 What we hope to show is that firms will react differently to a given change in the desired stock of capacity if they are faced with differing financial constraints. Financial variables will play a dual role in the analysis. First, as will be evident below, standard neoclassical investment analysis requires a discount rate to determine the optimal equipment-output ratio. We believe that this discount rate is only mildly responsive to changes in financial market conditions. The relevant discount rate, which includes a substantial risk premium, should exhibit less variance over time than the short term cost of funds.3 Reasons for this lie in the ex post fixity of factor proportions and long term nature of investment in equipment. Financial variables should, however, exert an important impact on the time path chosen for the investment response to a given change in desired capacity. We assume that the cost of funds to the firm depends on the firm's ability to utilize retained earnings (cash flow) as well as the coniditions prevailing in the bond and equity markets. Cost is interpreted in a broad sense and definitely includes the risks which management may believe non-internal financing entails.4 Given this view of firm behavior, we expect that the firm will invest more slowly or postpone investment if it expects that either (a) the costs of not having needed capacity in place are outweighed by the gains to be had from making more extensive use of internal sources of finance during a slow expansion, or (b) the costs of outside sources of finance (or the opportunity cost connected with the investment use of retained earnings) will be lower during subsequent periods to such an extent that it is worth the delay. Firms may hasten their investment response even if this involves additional costs in terms of internal disruption, etc., if these costs are overcome by the present abundance of internal funds or low current finance costs. Section I integrates the effect of financial conditions into the standard analysis to produce an estimable investment relationship. Section II describes the method of estimation. Section III presents results and conclusions.

Aitkin, M. A., Miscellanea -Some Tests for Correlation Matrices, Biometrika, 56, 2 (1969), 443. Bartlett, M. S., and D. V. Rajalakshman, Goodness of Fit Tests for Simultaneous Autoregressive Series, Journal of the Royal Statistical Society, B 15 (1953), 107-124. , A Note on the Multiplying Factors for Various Chi Square Approximations, Journal of the Royal Statistical Society, B lt (1954), 296-298. Buttimer, D. J., Supply response in the Irish dairy and beef herds, 1953-1970: an econometric exercise, Irish Journal of Agricultural Economics and Sociology, 4, 1 (1972-1973). Farrar, D. E., and R. R. Glauber, Multicollinearity in Regression Analysis: The Problem Re-visited, this REVIEW, 49, (1967), 92-107. Haitovsky, Y., Multicollinearity in Regression Analysis: Comment, this REVIEW, 51, (1969), 486-489. Huang, D. S., Regression and Econometric Methods (New York: John Wiley & Sons, 1970). Kmenta, J., Elements of Econometrics (New York: The Macmillan Company, 1971). Murphy, J. L., Corrective Procedures for Selected Econometric Problems, Introductory Econometrics (Illinois: R. D. Irwin Inc., 1973). Phlips, L., Business Pricing Policies and Inflation: Some Evidence from EEC Countries, Journal of Industrial Economics, 1969, no. 1. Silvey, S. D., Multicollinearity and Imprecise Estimation, Journal of the Royal Statistical Society, Series B, vol. 31 (1969), 539-552.