Vote buying or (political) business (cycles) as usual?

We study the short-run effect of elections on monetary aggregates in a sample of 85 low and middle income democracies (1975 - 2009). We find an increase in the growth rate of M1 during election months of about one tenth of a standard deviation. A similar effect can neither be detected in established OECD democracies nor in other months. We argue that the cycle is related to systemic vote buying which requires significant amounts of cash at election times. The finely timed increase in M1 is consistent with this and it cannot be, fully, accounted for by alternative explanations.