Did Alan Greenspan Illegally Manipulate Presidential Elections?

Regular readers may recall I had a series of posts on how the Fed tries to influence elections. Results are summarized here. While the previous posts looked at the Fed’s behavior going back to 1959 (data on M1 and M2 on the Fed’s website goes back to 1959), today’s post deals with the Alan Greenspan years. I have a hunch that Greenspan was the kind of guy to blatantly meddle in elections.

I know, its crazy talk but humor me – put on your tin-foil hat and let’s pursue this hunch for a moment. What would it imply? Considering that Greenspan is a firm believer in Ayn Rand’s nonsense, we’d expect Greenspan to favor Republicans over Democrats – all else being equal, we’d expect to see unusually tight money when an incumbent Democratic president ran for re-election or a sitting Democratic VP ran for President, and unusually loose money when an incumbent Republican ran for re-election or a sitting Republican VP ran for President. However, not all elections are close, so the Uncle Alan would feel more of an urge to meddle when elections are close.

(As an aside, the purpose of the Fed is maintain the money supply just right – it issues enough new money to lubricate the economy, but not enough to cause a recession. If it sees inflation on the horizon, it tightens up; if it sees a recession, it pumps money into the economy.)

Therefore, in the five elections that took place under Greenspan’s watch, we’d expect the most meddling in 1992 and 1996 (when the Republican candidate clearly lost in a competitive election), followed by 2000 (when the Republican candidate lost the popular vote but only by a smidge). We’d expect the 2004 elections when the Republican candidate won by a few percentage points to involve less meddling. We wouldn’t expect much if any intervention in 1988 considering that the Democrats nominated Michael Dukakis who proceeded to euthanize his own campaign without any intervention from the Fed. Another reason not to expect unusual behavior in 1988 – Mr. G had only taken office the year before, and perhaps was still unsure of how far the envelope could be pushed.

And where we would look to see evidence of the Fed’s behavior? Well, the Fed could move the fed funds rate or it can move M1 (cash and its equivalents). M1 has several advantages over the fed funds rate: nobody really pays attention to M1 whereas every change in the fed funds rate is literally national news, the Fed can move M1 every day if it chooses, and its effect is more widely dispersed throughout the economy. For a subtle guy like Big Al, M1 is the tool of choice.

So what do we see? The graph below shows the percentage change in real GDP per capita and the percentage change in real M1 per capita in November of each year of the Greenspan tenure as Fed Chairman.

Let’s look at each election year in turn. 1988 is the year we don’t really expect to see the Fed doing anything unusual. And we don’t.

What about 1992? In 1992 not only was monetary policy extremely loose (data on Money Supply from the Fed’s website goes back to 1959 and 1992 was the second loosest year in terms of money supply in the entire sample!), it was loose for no particular reason. After all, the economy was actually growing faster in 1992 than at any other point in the GHW administration.

How about 1996? 1996 saw the single biggest decline in real M1 per capita in all the years going back to 1959. Sure, the economy was growing at a good clip, but it was only the 17th fastest growing year in the sample. One might argue that in 1996 the Fed was anticipating irrational exuberance, but then why it would ease up on the throttle in 1997 and 1998?

If 1996 saw the single biggest decline in real M1 per capita in all the years going back to 1959, the year 2000 saw the second biggest decline. And this despite a slowing in the economy.

Anyone seeing a pattern here? Three consecutive elections coincidentally having three of the biggest moves in the Fed’s most important tool ever observed. Alan Greenspan truly was a man of extremes around election times, at least when those elections didn’t favor Republican candidates.

What about 2004? 2004 doesn’t seem to be have particularly loose monetary policy, despite our cynical prediction. But perhaps it occurred to Mr. Greenspan that after 3 consecutive elections with extreme jumps in real M1 per capita, it might make sense to use another tool for once. And it just so happens that the real fed funds rate, the fed funds rate less inflation, was negative 1.59, and the third lowest in the sample. (Compare that to negative 0.77 in 2003, when the economy was actually struggling but there was no election.)

So what do we conclude? Under Alan Greenspan, the Fed’s behavior around the time of Presidential elections has been precisely what a raving conspiracy theory believing lunatic would predict. Its hard to believe the Fed Chair can behave like a James Bond villain, but this is my tenth post on the subject, and the numbers always seem to say the same thing, no matter what figures I use or how I look at the data. So I’m left with no choice but to quote a James Bond villain. As Goldfinger once told 007:

[T]hey have a saying in Chicago: “Once is happenstance. Twice is coincidence. The third time it’s enemy action.”

I wonder what Goldfinger’s friends in Chicago would say about the fourth and fifth time. But I do know that if we as a nation don’t hold Greenspan accountable for this, we should expect more of this from Bernanke in 2008. In fact, we will deserve it.


Monthly M1
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.



James K. Galbraith reminds me in e-mail, and Archer points out in comments, that the Fed’s actions in 1992 did have some justification. To quote Archer:

Actually, there was a very big reason: the banking industry.
The whole industry had taken a huge hit between the S&L crisis and LBO lending gone sour. The industry was badly undercapitalized.

In my mind, Greenspan’s finest (perhaps only) accomplishment was engineering and maintaining a very very steep yield curve in the early 1990s. It was a massive subsidy to the banking industry that went largely unnoticed at the time. Hence the reason for continued loose money despite economic growth.


Update 2. I’ve had several people point out that there are lags between changes in the money supply and movements in the economy. This is one of those urban myths, like supply side economics, that everyone knows is true but doesn’t hold up when one looks at the data. And in this case, I looked at the data not long ago in another post. Table 3 shows the correlation between changes in real M1 per capita and real GDP per capita. In general, there is an -87.5% correlation occuring in real time (i.e., a change in real M1 per capita and a change in real GDP per capita occur in the same quarter).

The table also shows… in election years, changes in the real money supply can lead changes in the economy. This may be because the Fed is pulling out all stops: jawboning, moral suasion or guidance at the monthly regional meetings, and perhaps even messing with the required reserve ratio.


Update 3. Corrected the link in Update 2. My bad. Also, rephrased update 2 to (hopefully) make it clearer.