A normative analysis of public capital

Revisiting the Optimal Stationary Public Investment Policy in Endogenous Growth Economies

“…Indeed, the next proposition shows that the ratio cannot exceed (1 ). This threshold is the ratio that maximizes the steadystate growth rate in the related model of Marrero and Novales (2005), in which is positive.…”

“…For small values of , i.e., below 0.05, the model-based optimal ratios are close to current public investment ratios for the OECD and the US. Although some empirical papers (e.g., Holtz-Eaking, 1994, Hulten and Schwab, 1991, and Tatom, 1991) obtain estimates of that are close to zero, a recent consensus has developed that suggests that this elasticity is between 0.1 and 0.2 (see, e.g., Ai and Cassou, 1995, Cassou and Lansing, 1998, and Shioji, 2001. Hence, either current public investment policies are suboptimal or existing models omit relevant factors and hence o¤er misleading policy prescriptions.…”

“…Lynde and Richmond (1992) and Ai and Cassou (1995) account for non-stationarity in the data and estimations are smaller but still signi…cant: the former estimates = 0:2 using time series techniques, while the latter estimates ' between 0.15 and 0.2, using a GMM estimation process. In a more recent paper, Shioji (2001) uses dynamic pannel techniques and estimate the elasticity of output with respect to infrastructure to be somewhere around 0.1 and 0.15.…”

“…While total public outlays have represented meaningful shares of GDP during the last four decades, public investment-to-output ratios have generally declined over the same period. 1 Nevertheless, this ratio has remained relatively stable since the 1990s. By 2000, public investment represented 3.7% of total real GDP in the OECD and 3.1% in the US.…”

“…Rather, we revisit the optimal choice of public investment in a more general and plausible framework than those mentioned above. Our framework allows for i) long-lasting private and public capital stocks (Ai andCassou, 1995, andLansing, 1998), ii) a lower depreciation rate for public capital than for private capital (Ai and Cassou, 1995), iii) an elasticity of intertemporal substitution below unity (as suggested by Prescott and others) and iv) assumes that there is a need to …nance a non-trivial proportion of public services in output in each period Lansing, 1998, andMarrero, 2005). This economy allows for a gradual transition between di¤erent steady states, which Futagami et al (1993) found to be an important factor in determining optimal public investment policy in a Barro-type framework.…”