Recent work has shown that financial conditions faced by firms tightened during the Great Recession, impacting real outcomes. Additionally, house price declines have been shown to disproportionately affect small and young businesses during this period, although the mechanisms behind these reduced-form relationships have not been established. I connect these findings by providing evidence for a local bank channel through which house prices impact firms, focusing on how the effects vary with firm age and size. Combining confidential microdata from the US Census Bureau with publicly available house price and bank balance sheet data, I show that young businesses are sensitive to house prices, but the local bank channel plays a minor role. However, the local bank channel has a significant and sizable impact on businesses that are old and small. These results suggest that house price shocks work through different channels for young vs. old firms. To quantify the effects of house price shocks in a general equilibrium framework that is consistent with these findings, I develop a model where house prices work through a collateral channel and a bank credit supply channel to differentially impact heterogeneous firms. The model implies that the bank channel plays an important role in aggregate outcomes, while the collateral channel is critical for explaining the relatively poor performance of young businesses.
Macro and Micro of Productivity: From Devilish Details to Insights with Lucia Foster, Cheryl Grim, John Haltiwanger, and Zoltan Wolf
Firm-level revenue-based productivity measures have become ubiquitous in studies of firm dynamics and aggregate outcomes. One commonly used measure has increasingly been interpreted as reflecting “distortions” since in their absence equalization of marginal revenue products should yield no dispersion in this measure. A commonly used, but distinct, measure is the residual of the firm-level revenue function which reflects “fundamentals”. Using plant-level U.S. manufacturing data, we find these alternative measures are highly correlated, exhibit similar dispersion, and have similar relationships with growth and survival. However, the distinction between these alternative measures is important for quantitative assessment of aggregate allocative efficiency.