Working papers

Abstract: Specialized knowledge-generating jobs comprise close to one fifth of employment and one fourth of the wage bill in French manufacturing firms. These jobs gained importance over the period 1999- 2015 at the cost of production-related jobs. Conditioning on firm size and shares of management workers, their higher shares in employment at the firm level are correlated with more innovation and intangible capital, greater product complexity, higher total factor productivity and profitability. This suggests that firms use specialized knowledge workers to generate within-firm knowledge and create firm capabilities. Consistently with empirical regularities, we model firms as organizations where efficient production of higher-value added, complex goods requires information acquisition by within-firm knowledge workers to develop capabilities beyond those created by management and hierarchies.

Abstract: In recent decades, middle-paid jobs have declined, replaced by a mix of high and low-paid jobs. This is labor market polarization. At the same time, initially skilled and typically larger cities have become even more skilled relative to initially less skilled and typically smaller cities. This is the great divergence. We develop a theory that links these two phenomena. We draw on existing models of polarization and heterogeneous labor in spatial equilibrium, adding to these a sharper interaction of individual- and city-level comparative advantage. We then confront the predictions of the theory with detailed data on occupational growth for a sample of 117 French cities. We find, consistent with our theory, that middle-paid jobs decline most sharply in larger cities; that these lost jobs are replaced two-to-one by high-paid jobs in the largest cities and two-to-one by low-paid jobs in the smallest cities; and that the lost middle-paid jobs are concentrated in an upper tier in the large cities and a lower tier in the smaller cities.

Abstract: We use US interstate banking deregulations to identify the bank finance-trade channel while controlling for state-country bank links. A 1% increase in banking integration between states caused a 0.23% increase in the state-country level foreign exports/domestic shipments ratio between 1992-1996. The observed effect is due to banks with foreign assets, while the US expansion of banks with only domestic assets has no impact on exports/domestic shipments ratio. Our findings support the bank finance channel of international trade.