Explaining Reallocation’s Apparent Negative Contribution to Growth in Deregulation-Era Developing Countries

We investigate the findings in several recent studies using plantlevel data that report negative growth in the reallocation component of aggregate labor productivity growth. This empirical finding is also counter to almost all theoretical models of aggregate productivity growth, where inputs reallocate on average from the lower value to higher value activities. We attempt to solve this puzzle for Chile. We cannot explain it away by conditioning productivity growth on two types of labor and capital instead of just one labor type. We do find that the puzzle disappears when we define aggregate productivity growth in terms of its impact on aggregate final demand. Then labor reallocation contributes positively to economic growth in most years. We hypothesize that the negative covariance between labor inputs and measured productivity growth may arise from the price measurement error in productivity and the fact that a reduction in price occurs when plants increase output (inputs) and face downward sloping demands.