Lant Pritchett
Published: 1999
Total Pages: 46
Get eBook
May 2000 - Using the word capital to represent two different concepts is not such a problem when government is responsible for only a small fraction of national investment and is reasonably effective (as in the United States). But when government is a major investor and is ineffective, the gap between capital and cumulative, depreciated investment effort (CUDIE) may be enormous. A public sector steel mill may absorb billions as an investment, but if it cannot produce steel it has zero value as capital. The cost of public investment is not the value of public capital. Unlike for private investors, there is no remotely plausible behavioral model of the government as investor that suggests that every dollar the public sector spends as investment creates capital in an economic sense. This seemingly obvious point has so far been uniformly ignored in the voluminous empirical literature on economic growth, which uses, at best, cumulated, depreciated investment effort (CUDIE) to estimate capital stocks. But in developing countries especially, the difference between investment cumulated at cost and capital value is of primary empirical importance: government investment is half or more of total investment. And perhaps as much as half or more of government investment spending has not created equivalent capital. This suggests that nearly everything empirical written in three broad areas is misguided. First, none of the estimates of the impact of public spending identify the productivity of public capital. Even where public capital could be very productive, regressions and evaluations may suggest that public investment spending has little impact. Second, everything currently said about total factor productivity in developing countries is deeply suspect, as there is no way empirically to distinguish between low output (or growth) attributable to investments that created no factors and low output (or growth) attributable to low (or slow growth in) productivity in using accumulated factors. Third, multivariate growth regressions to date have not, in fact, controlled for the growth of capital stock, so spurious interpretations have emerged. This paper - a product of Poverty and Human Resources, Development Research Group - is part of a larger effort in the group to understand the importance of public sector actions for economic growth.