Does Money Matter? Quasi-Experimental Evidence from the End of Geographic Reserve Requirements

We study the end of the US Federal Reserve’s geographic system of reserve requirements in 1972, which eliminated geographic differences in the reserves that banks were required to hold against deposits. At the end of this regime, larger cities, which were designated Reserve Cities (RCs) under the geographic regime, experienced a 2 to 4.5 percentage point larger decrease in reserve requirements relative to smaller cities. We exploit the fact that some medium-sized cities were categorized as RCs only because they hosted a branch of their regional Fed Bank, even though they were much smaller than the average RC. We compare banks in these medium-sized RCs to banks in similarly sized non-RCs and find that both lending and deposits increased as reserve requirements declined. This increase was driven by banks subject to the Fed’s reserve requirements (member banks), while non-member banks showed no change. We find no effects of this monetary expansion on employment or wages. We aim to explain this apparent monetary neutrality by showing that changes in the money supply are only impactful when they affect interest rates. In our setting, these rates were likely unaffected by our local shocks and were largely pinned down by aggregates beyond the limits of our medium-sized cities.