Are the results inconsistent with Jayaratne and Strahan (1996)?
I do not consider my results inconsistent with those of the Jayaratne and Strahan (1996) study. On the one hand, my study shows that the bank branching deregulations that took place prior to 1985 were in general not followed by faster economic growth. As a matter of fact, these early liberalizers grew on average 0.12% per year slower than did their neighbours. On the other hand, the deregulations that took place after 1985 (i.e., the second half of the sample) were in general associated with large, positive growth effects.
Such a trend is consistent with the results documented by DeLong and DeYoung (2007) that there exists so-called "learning-by-observing" in the banking industry consolidation process, that the earlier M&As were typically less successful than the later ones. The same logic can apply to the results of my study, that the opportunities enabled by the deregulations were exploited more successfully as time went by and experiences accumulated through a learning-by-observing process.
There were five cases in the late 1980s (Louisiana, Michigan, Missouri, Oklahoma, and Tennessee) in which statistically significant faster economic growth followed the intra-state branching deregulations. I noticed that in all of these five cases, inter-state banking deregulations also took place before or at least at the same year of the intra-state branching deregulations. I believe that, in these five cases, the opening of market to not only in-state but also large, out-of-state competitors (mainly from New York /North Carolina/California) created greater competitive pressures than the earlier deregulations that typically allowed competitions from within the state only.
I would also like to emphasize that, all of the state-level deregulations are not the same, although all of them are called “bank branching deregulation.” (Similarly, in the context of the Card/Krueger study, minimum wage of $5.05 may not be binding in New Jersey or for skilled workers, but it may be strictly binding in the poorer Southern states).
The effect of a specific deregulation depends on where and when it takes place; and the level of local banking market competition before the de jure deregulation. For example, I noticed from the study's evaluation results that, in previously more concentrated (competitive) local markets, the branching deregulations were in general associated with larger (smaller) effects. The explanation is straightforward: if there was already sufficient high level of competition in the local banking market (in particular in large urban markets such as NYC) even if the competition arose from local players only, the exclusion of outsiders may not have been a binding constraint for the local economy.
I believe that a very natural next step for this line of policy evaluation studies would be to look into the heterogeneity of the evaluation results and study why some states/counties benefit more (or less) from a same policy change, and whether some sectors/segments of the economy or population benefit more than do others. Such results, which are more informative than the typical "average effect" results seen in many studies, could help shed light on many competing hypotheses regarding how a certain policy change affects the economy. Indeed, many economic policies have distributional effects, which are not evident if only the average level effects are examined and reported.
References:
Huang, Rocco. Evaluating the real effect of bank branching deregulation. Journal of Financial Economics,
Jayaratne, J., Strahan, P.E., 1996. The finance-growth nexus: Evidence from bank branch deregulation. Quarterly Journal of Economics 111, 639-670.
DeLong, G., DeYoung, R., 2007. Learning by observing: information spillovers in the execution and valuation of commercial bank M&As. Journal of Finance 62, 181–216.