Scale and scope economies in higher education: The case of regional colleges of the University of Puerto Rico

This research examines the cost structure of the System of Regional Colleges of the University of Puerto Rico (SRC-UPR) in order to calculate the degree of scale and scope economies for alternative output bundles. A unique data set is constructed that contains pooled time-series/cross-sectional data across the six colleges of the SRC-UPR during the period 1977--78 to 1989--90.; When the academic programs are the unit of analysis, the data show that the production of all clusters of academic programs---Liberal Arts, Natural Sciences, and Engineering Technologies and Agriculture---regardless of the colleges to which they belong, is characterized by substantial economies of scale. In order to exhaust these scale economies, the size of the average academic program in each cluster of each college would have to increase far beyond observed sample values. This result implies that all academic programs are operating along the downward-sloping portion of the average cost curve.; When the colleges are the unit of analysis, the data show that the estimated cost function satisfies all regularity conditions except curvature in output space. In order to impose the proper curvature in output space, Cholesky factorization is used to reparameterize the estimated cost function. Virtually all previous studies in the field violate the curvature conditions, meaning that their results are incompatible with economic theory.; All colleges, at all considered levels of output, exhibit large ray (overall) economies of scale. At the same time, there are neither product-specific nor product set-specific economies of scale, irrespective of the colleges' composition. The production process of all colleges is characterized by wide global economies of scope, as well as by product-specific and pairwise economies of scope.; The data are consistent with the conjecture that colleges tend to over-invest in physical capital. The results are consistent with the hypothesis that colleges operate along an inefficient short-run cost function. This empirical finding may result from indivisibilities that make the stock of physical capital of the colleges unresponsive to small changes in output.