Exposing prejudices against a bias: Three essays on short-run biased technical change in macroeconomics.

摘要

I explore the implications of short-run biased technical change for macroeconomics.; A large literature in macroeconomics is concerned with technical change in the short-run (annually or quarterly). However, this literature has focused exclusively on neutral technical change, i.e. that which does not affect factor elasticities.; I demonstrate that the neglect of short-run biased technical change is based on stylized facts of long-run economic growth and necessary conditions on permanent technical change for long-run balanced growth paths to exist in neoclassical growth models. However, observations from data are suggestive of short-run biased technical change, and balanced growth is consistent with short-run biased technical change.; In an SDGE model I find that (i) biased technical changes cause fluctuations that are qualitatively similar to those caused by neutral technical changes, and (ii) small biased technical changes can account output volatility matching the U.S. experience. In a growth accounting framework, using data from 35 U.S. sectors, I find that (iii) if biased technical change occurs, then it contaminates Solow residuals (even when time-varying factor shares are used), and, again, (iv) small biased technical changes can result in economically important effects on value-added growth.; Based on (i)--(iv), I analyze labor's share from the 35 U.S. sectors. Theory suggests that, given certain conditions, factor shares equal factor elasticities, so movements in labor's share may proxy for biased technical change. I find that labor's share movements are rejected as exogenous to instruments correlated with factor demand growth in fewer sectors than Solow residuals are. Also, after removing the portion of labor's share movements explained by growth in the real wage, the real price of capital, and the capital/labor ratio, the residual cannot be rejected as an exogenous impulse in a large majority of U.S. sectors (27 out of 35). The volatilities of these residuals are, on average, 29 percent as volatile as their constituent series.; I conclude that short-run biased technical change does occur; observed factor shares are a starting point to quantify this technical change; and the evidence suggests that short-run biased technical change may have important effects on economic fluctuations.