This paper investigates the linkage between small business bank lending and small business employment growth by testing two hypotheses: first, whether there exists any causal relationship between bank’s small business lending and the small business employment at the disaggregate level and, second, whether the evidence on the causality from the disaggregate data is consistent with the evidence from the aggregate data and the pooling data. The general hypothesis would be that the greater the level of regional aggregation, the more probable that we can find clear causality between bank credit and economic activity, but the empirical results from the aggregate data do not always represent the true causality existing from the disaggregate data when the aggregation bias exists. The findings of this study support both hypotheses: aggregating data does help in identifying the causality, and also the data aggregation sometimes misrepresent the true causal relationship of less aggregated sample. We found that the pooling regression is an effective way to identify the problem in aggregation bias.
Kang, Eungmin; Edwards, Mary; and Ratha, Artatrana, "Causality Between Bank Lending and Employment Growth: Evidence From Small Business Lending" (2007). Economics Faculty Working Papers. 14.