The Fed, Presidential Elections, and Coincidence – Part 5

Over the past few days I’ve had a few posts looking at whether the Fed acts to improperly influence elections. For example, the following chart,

reproduced from an earlier post shows that the increase in real M2 per capita every month is higher in years in which a Presidential incumbent from the same party as the Fed chair runs for re-election than in other years. By contrast, when an incumbent is running for re-election but he is from a different party than the Fed chair, the growth in real M2 per capita is not only lower every month than in other years, it is also negative.

I’ve also produced tables (see the same link as above) showing that the fed funds rate, which the Federal Reserve controls directly, has similar very clear-cut behavior: when a “friend” is running for re-election, the fed funds rate is unusually low, and when an “enemy” is running for re-election, the fed funds rate is unusually high.

A cynic might conclude from this that the Fed helps its friends by making money cheap and easy to benefit its friends, yet tight and expensive to hurt its enemies.

Now, some readers have objected. They point out this can be coincidence. The Fed may well be reacting properly to conditions in the economy. Let us consider that argument in more detail. If the Fed is behaving impartially, to get the clear-cut results we see in real M2 per capita and fed funds rate, the Fed needs a reason to slow down the economy when an incumbent from the other party is running for office, and it needs a reason to try to jump start the economy.

Put another way, for this to be a coincidence, for the Fed to be operating properly, when an “enemy” is running for re-election, the economy has to be growing quickly, and when a “friend” is running for re-election, the economy has to be faltering. (Take another look at the table above – the table doesn’t exactly brook a middle ground.)

On the other hand, a cynic such as myself would disagree. The cynic would say – the loose money supply in the last column… that will cause rapid growth. Similarly, the tight money supply in the second to last column… that will cause growth to slow.

This makes for a testable hypothesis, H0: Fed acting properly v. H1: Fed acting improperly. So… what does happen with growth? The table below shows the growth rate in real GDP per capita per quarter. (Unfortunately, the data is not available monthly).

What do we see? Well, in the three quarters leading up to an election, real growth per capita is fastest precisely when the Fed’s friends can use the help. And in two of the three quarters leading up to an election, growth is slowest precisely when the Fed’s enemies can use the extra weight dragging them back. In the quarter, growth is almost at its slowest, missing by a mere 0.02%!

Note also… when growth is fastest in column 4, it is usually quite a bit faster than in other columns – there is little room for confusion on the Fed’s part, and simply insisting that the Fed was working with preliminary data seems a stretch.

What about in the quarter in which the election takes place? What explains that? Well… elections take place about a third the way through that quarter. Thus, the Fed has the time to reverse course. In fact, after at least a year of improperly tight money, it has to loosen up to prevent a recession, and after at least a year of improperly loose money, it has to tighten up to prevent runaway inflation.

Sure, the data is limited, but when the same results are obtained several ways, appealing to coincidence gets less and less likely. I’ll be honest, I managed to convince myself last week. But if any of you still need convincing, in the next day or two, I’ll look at the coincidence question again, from a different direction, this one involving both cause and effect of Fed behavior.


Monthly M1 and M2
Monthly CPI
Quarterly population data which I then linearized into monthly
Quarterly real GDP per capita


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