| The previous tests of public capital effect on multifactor productivity using Solow's growth accounting technique yield differing estimates of this relationship. I reconcile these results and show why they differ. I highlight the possible problems when Solow's technique is used in empirical work and the problems with previous works. To solve these problems, I use the Malmquist Index to estimate multifactor productivity growth which, in turn, I use to estimate the output elasticity of public capital. My results show that, contrary to recent work, the effect of public capital on output is positive and significant. However, the effect of public capital on each industry is smaller compared to its effect on total industry. I find that the magnitude of this effect of each type of public capital varies across the different types of public capital. In regional analysis between the Sunbelt and Snowbelt, the magnitudes of effects are significantly different from each other in some cases while some other cases are not. These results show that the effects of public capital in the Sunbelt are not always larger than those in the Snowbelt. |
