Pattern of investment in Colombia: An econometric investigation using firm-level panel data

Government policies to achieve sustained growth through capital accumulation rest on firms' responses to uncertainties and to adjustment costs. Yet, very little is known about the adjustment costs of investment and the impact of the same on the pattern of investment in the developing countries. To shed light on the role of fixed adjustment costs on the pattern of investment at the firm level, this thesis estimates the probability of investment (hazard rate), and estimates the determinants of firm-level investment/disinvestment demand using balanced annual panel data for Colombia, covering the period 1977--1991.;The empirical model used here for estimating the hazard follows from Cooper, Haltiwanger and Power (1999), who found evidence for non-convex adjustment costs (i.e., an upward sloping hazard curve) using plant-level data for United States. This thesis extends that contribution by using the estimated hazard rate to model the demand for investment in a Heckman two-step procedure. The thesis also uses the Tobit V model to determine the role of fixed costs in a firm's decision to invest and estimates simultaneously the levels of investment and disinvestment demands. The econometric techniques employed address the problems of data censoring, unobserved heterogeneity and initial-conditions specification. The empirical results indicate a mildly downward sloping hazard for large and medium-sized firms and a mildly upward-slopping hazard for small firms. The shape of the hazard does not lend support to the non-convex adjustment cost hypothesis, and indicates that fixed costs of investment are not large compared to total investment costs so as to lead to jumps in the investment pattern. The Tobit V model reinforces the hazard results. The analyses also indicate that past levels of output, capital stock and the rate of purchase of machine and equipment are important determinants of both investment and disinvestment demand. The findings suggest that in order to enhance growth through investment, policymakers need to craft policies aimed at removing market distortions and introducing foreign competition to enforce efficient use of scarce investment resources. External evidence suggests that resources may have been misused due to the subsidized directed credit policy of the government during the study period.