John C. Haltiwanger and Scott Schuh Haltiwanger is a Professor of Econom- ics at the University of Maryland and Chief Economist of the U.S. Bureau of the Census; Schuh is an Economist at the Federal Reserve Bank of Boston. The authors’ acknowledgments appear at the end of the article. Gross Job Flows Between Plants and Industries A remarkable feature of the current U.S. economic expansion has been its ability to shrug off the adverse effects of financial crises and economic slowdowns around the world for nearly two years. Employment has been expanding smartly and unemployment has con- tinued to edge down. Recently, however, foreign-sector developments have triggered a sizable shift in the sectoral composition of U.S. employ- ment. By early 1999, employment growth in the goods-producing sec- tor—the sector most exposed to international trade— had stalled, while employment growth in the service-producing sector was still humming along (Figure 1). Indeed, many forecasters have been citing weakness in the industrial sector as a serious threat to the expansion. Historically, substantial shifts in labor demand between sectors have been correlated with the business cycle. The gap between the rates of growth for goods and services employment, a cursory measure of these shifts, is closely correlated with GDP growth (Figure 2). A common view of this correlation is that some sectors, such as manufacturing, are more sensitive to the business cycle than others. Thus, shifts in labor demand are merely passive responses to fluctuations in aggregate demand (GDP) that temporarily alter relative employment growth rates among sectors. Another, complementary view is that these shifts are active, permanent responses to sector-specific events that generate increases in structural unemployment and reductions in aggregate demand. Neither view fully explains current developments. In contrast to the first view, the decline in the employment gap has not been accompanied by the usual decline in GDP growth, although the deviation is historically moderate, so far. In contrast to the second view, the unemployment rate has shown no hint of rising as a result of labor demand shifts. These unusual developments highlight the inability of macroeconomists to completely understand the connection between shifts in labor demand and the business cycle, or to synthesize these two views.