Tracking the new economy: Using growth theory to detect changes in trend productivity

The acceleration of productivity after 1995 prompted a debate over whether the economy's underlying growth rate would remain high. In this paper, we draw on growth theory to identify variables other than productivity-namely consumption and labor compensation-to help estimate trend productivity growth. We treat that trend as a common factor with two “regimes,” high- and low-growth. Our analysis picks up striking evidence of a return in 1997 to the high-growth regime, nearly 25 years after a switch from high- to low-growth. We find that both the common factor and regime-switching aspects of the model are important for identifying changes in trend productivity, and also show that the trend breaks are more difficult to detect with per capita (as opposed to per hour) based data because of persistent labor supply shifts. Finally, we argue that our methodology is effective in detecting changes in trend in real time: In the case of the 1990s, the methodology would have signaled the regime switch by 1999, or within roughly six quarters of when it occurred according to the full sample.