The Fed, Presidential Elections, and Coincidence – Part 3

I decided to follow up my morning post showing changes in real M2 per capita look exactly like they would if the Fed were trying to influence Presidential elections by looking at some of the reader comments.

Spencer, in particular, made some useful comments. I haven’t had the time to run through many of them, but here are two: 1) look at the one year (rather than the 3 months) change in real M2 per capita by month, and then 2) take the first difference of that to look at the rate at which the one year change in real M2 per capita itself changes.

The table below shows the 12 month percentage change in real M2 per capita (i.e., the change in M2 over and above (or under and below) the inflation rate and the change in the population) in each of the months leading up to and including November in years in which there is no presidential election, in years in which there is a presidential election but no incumbent. It also shows the years in which there is a Presidential election and an incumbent, differentiating between whether the incumbent and the Fed Chairman come from the same or opposing parties. Note that the category showing the fastest growth in real M2 per capita is colored green, and the category showing the slowest growth in real M2 per capita is colored orange.

With all due respect to those who have been insisting that the results are a coincidence or some random effect, I’m sorry but I just don’t see how that’s credible. This is precisely, with zero deviation, the exact behavior in real M2 per capita we would expect if the Fed were trying to influence elections. Output is rarely this cleancut.

Since data from the Fed became available in 1959, when there is a Presidential election involving an incumbent from the same party as Fed chairman, the year on year Money Supply (as measured by real M2 per capita) increases each and every month by more than when there is no election or when there is no incumbent. By contrast, when there is a Prsidential election involving an incumbent from a party other than that of the Fed chairman, the year on year Money Supply decreases each and every month. I don’t see how one can reach any conclusion other than that the Fed is meddling in Presidential Elections in a blatant and partisan way.

Spencer’s other suggestion doesn’t seem to help the story… I’m not sure what it says. Taking the one year percentage change of the one year percentage change (i.e., trying to see how the yearly percentage change in real M2 per capita is itself changing – looking for acceleration or deceleration), we get this:

I’m honestly not certain how to interpret this. The rapid growth in the third column in the months leading to an election, combined with the fact that in Table 1 the third column is always smaller than the other three, may indicate that a few months before the election, the Fed may start to worry it has overshot in its campaign of destruction against an incumbent from a different party. (See the –14,887% change in March, which is an order of magnitude beyond the other large changes, which occur in April and February.) But otherwise, I don’t think the second (percentage) difference is giving us much mileage here.

—————————————————

Data:
Monthly M1 and M2
Monthly CPI
Quartlery population data which I then linearized into monthly

—————————————————

As always, let me know if you want my spreadsheet.

—————————————————

Update… I’ve started cleaning data to look at interest rate movements around election time. Expect a post within the next few days…