Presidents, Tax Burdens, and the Subsequent Economic Growth

Presidents, Tax Burdens, and the Subsequent Economic Growth

Posted by Dan Crawford (Rdan) | 7/05/2011 02:52:00 PM

by Mike Kimel

Presidents, Tax Burdens, and the Subsequent Economic Growth


Over the years, I've posted variations of the graph below a few times:

Figure 1


The graph shows the change in the tax burden (i.e., current federal receipts / GDP) from the year before an administration took office to its second year in office on one axis, and the annualized growth in real GDP from its second year to its last year in office on the other axis. So, to use GW as an example, the graph shows the change in Federal Receipts / GDP from 2000 to 2002 on one axis and the growth rate in real GDP from 2002 to 2008 on the other axis.


JFK and LBJ are bunched together because LBJ took over during JFK's term, and the two together served eight years. Ditto Nixon/Ford. Truman, on the other hand, took over less than three months into FDR's fourth term, and then served just shy of eight years on his own. I figured that was enough to qualify as Truman's own administration. FDR's 12 years in office are broken into two administrations: through 1940, and 1941 - 1944. 1941-1944, being the World War 2 years, are sufficiently different in terms of focus and goals to qualify as a different administration.

The graph is rather busy. Removing the names of the Presidents (I've left 3 for which reasons which will become evident later), and using Excel to add a trendline may make things a bit more clear.

Figure 2.

The graph makes a point that would be controversial if people were aware of it, namely that in general, the more an administration reduced tax collections during its first two years in office, the slower the growth rate during its remaining years in office. This result contradicts the beliefs of most people who have taken an economics course in this country, and anyone who listens to Rush Limbaugh or watches Fox News. But the fit is surprisingly tight considering how few variables are involved. It doesn't matter who believes it, or whether we are happy with this result. I personally would much prefer to see lower taxes leading to faster economic growth than slower economic growth, but reality is what it is, not what we want it to be.

Now, if this is true, how does it work? Well, it could be that changes in the tax burden in years 0 to 2 economic growth in later years. (That isn't controversial, even if the direction that the data is.) But I suspect there is more at stake. The tax burden often moves, sometimes by quite a bit, even when there has been no change in marginal tax rates. I think a lot of the change has to do with regulation and enforcement. An administration that cuts tax burdens is reducing regulatory tax burdens and perhaps cutting enforcement, and not just at the IRS. A big part of that has to do with the ideology of the administration, which in turn affects who it places in key positions. The people GW brought into his administration are very different from the people FDR selected. And a tiger doesn't change its stripes... an administration staffed by people who believe in reducing regulation in year 1 is going to be staffed by people who believe in reducing regulation in year 7.

Sadly, I now have to repeat something that keeps coming up every time I make this point. I get comments and/or e-mails and/or even people telling me in person that what is happening is that slower growth is leading to lower tax collections. Returning to the GW illustration, if you really believe that mediocre growth from 2002 to 2008 caused the reduction in tax burdens from 2000 to 2002, you may have what it takes to write for the National Review or the Heritage Foundation. I'm not impressed by arguments based on time travel, nor am I in a position to hire.


Before we go on, some housecleaning. Nominal and real GDP come from the Bureau of Economic Analysis. GDP was first computed in 1929, so the first complete administration for which we have data is FDR I. Data on the Federal government's tax receipts comes from the Bureau of Economic Analysis' NIPA Table 3.2.

In addition to indicating that the administrations that produced faster growth tended to be the ones that started off by raising tax burdens, the graphs tells us a few things. Many of these things are about the media and the education system in this country, as what the data shows tends to be a surprise to most people. The fastest growth occurred during World War 2, which is also when the economy could best be described as a command economy. (Think tax rates above 90%, rationing, government directives, etc.) The years from 1932 - 1940, widely perceived as being a period of slow growth, actually had the fastest peace time growth for any period for which we have data. (The economic collapse of the Great Depression occurred before FDR took office.)

But is this all an anomaly? Some artifact of the data? Unfortunately, I don't think so. See, the graphs provide a bit of evidence that indicates the theory is actually more robust than it looks. Consider the following reasonable assumptions:

1. changes in the tax burden are indicative of behavior of the administration
2. the sorts of policies and behavior that affect of the economy don't change over the length of an administration (i.e., the tiger doesn't change its stripes)
3. the economy is affected by more than just the current administration's behavior
4. some policies can take a while to have an effect

Assuming all of that, one would think that the longer an administration was in office, the more likely it is that its growth would tend toward what "it should be" given its policies. Put another way, the biggest outliers should be the four year administrations. A glance at Figure 2 tells you immediately that this is, indeed, the case. FDR's WW2 years, Bush 1, and Carter are the outliers. And eliminating those outliers and focusing exclusively on the eight year administrations does, indeed, strengthen the model, as theory would suggest:

Figure 3.

So what we're left is increasing evidence that the more an administration cuts the tax burden in its first two years in office, the slower the growth it produces thenceforth. And again, you don't have to like this result (I personally don't) for it to be true.

All of which brings us back to Barack Obama, who is not on any of these graphs. Now, Obama inherited an economic disaster, but that's the past. He campaigned for the right to be the one making the decisions, and he got it. And what matters is whether he makes things better. FDR inherited a much worse economic mess, and went on to produce growth the likes of which haven't seen since. So... any chance Obama is going to pull an FDR? In a word, no. Whereas FDR raised the tax burden by more than any other President for whom we have data during his first two years in office, Obama has reduced the tax burden by about 1.2 percentage points. That puts him between Ike and Nixon/Ford when it comes to changes in the tax burden over the first two years. And both Ike and Nixon/Ford went on to produce subpar growth. So, barring some fortuitous change that has nothing to do with Obama, we can't expect much more than mediocre growth for the remainder of Obama's time in office. But I don't expect his eventual challenger to learn anything from the three figures in this post either. None of this bodes well for America.

As always, if you want my spreadsheets, drop me a line at my first name (mike) period my last name (kimel - with one m only) at gmail period com. I should also point out, you can find a lot more of this sort of analysis in Presimetrics, the book I wrote with Michael Kanell.