How GDP Numbers Show That The Trump Tax Cuts Are Working: A Response To James Pethokoukis

Posted: Aug 27, 2018 11:42 AM
How GDP Numbers Show That The Trump Tax Cuts Are Working: A Response To James Pethokoukis

Last quarter's GDP growth was an impressive 4% (as of this writing, there are revisions in the pipeline).

Coincident indicators for the third quarter, which we're in now, also are pointing towards higher growth rates than we've typically seen during the past ten years of lackluster growth. I'll show that in an upcoming article.

Predictably, reactions to this high growth rate fall along political lines. The tax cuts engineered by President Trump and Speaker Ryan and Majority Leader McConnell are seen as the cause of the growth spurt by republicans; whereas, critics of the administration tend to dismiss the tax cuts as a factor. For example, my friend James Pethokoukis, a conservative columnist from The American Enterprise Institute and critic of the administration, wrote a piece for The Week entitled “America's economy is booming. Trump has nothing to do with it.,” referring to claims that the tax cuts caused the spike as 'hot take fever.' Though in fairness to my old friend, he acknowledges that perhaps the tax cuts helped but he says it’s too soon to tell.

Warnings about 'hot takes' on short-term data are worth remembering, but new data can be under-emphasized as easily as it can be overemphasized. To use statistical theory jargon, there is a danger of both 'type 1 error' and 'type 2 error.' I think of it this way, we already had old data, and we don't have future data, so the only data which introduces new 'signal' by which to test theories is new data.

So, what does the new data indicate? That the tax cuts had an effect. First of all, the economy grew unusually quickly during that quarter, significantly higher than the average of the last ten years. 

Pethokoukis, argues from growth data from the Obama administration:

"For one thing, it doesn't necessarily tell you a whole lot about where the economy is heading. There were eight quarters of 3 percent growth or faster scattered across the Obama presidency, including four of 4 percent or faster and one of 5.2 percent."


The argument doesn't really seem compelling. Think of it this way, Obama administration had 32 quarters. Four of them were 4% or better. If that's the baseline, and making no prior assumptions about the timing of them, there's a 1/16th chance of getting a plus 4% quarter if they are randomly occurring.  Trump has had 6 completed quarters, and one plus 4 quarter. That means that 1/6th of his quarters is a plus 4%. It beats the Obama frequency by more than 2 1/2 times. So, using a frequency approach, which Pethokoukis does of binary outcome (less than 4 or not less than 4), shows that Trump's frequency of hitting that target is much better.

Of course, if we're trying to measure the effect of tax cuts, not just of the comparative presidencies, which presumably is the point at issue, then we would look only at the plus 4s after the tax cut. The tax cut kicked in at the beginning of the year, so we only have two quarters reporting since then. That means that 1/2 of the quarters is a plus 4 (and it will be 2/3rds if expectations of another 4 percenter occur this quarter.  If Obama admin is the baseline frequency expectation for the alleged 'new normal' then you've got a 16% chance of hitting that growth target. Since the tax cut, we've hit it 50% of the time. That’s a pretty big coincidence if the taxes had nothing to do with it.

Plus, there's the confounding variable of monetary policy. Look at the Fed Funds rate over the same time period:

One of Obama's plus 4s comes on the quarter immediately following one of the most massive series of cuts in interest rates ever recorded. So, if you are looking at plus 4s under Obama which did not occur following massive monetary stimulus, then you're at 3 out of 32, or a 9.4 percent probability of getting a plus 4 randomly choosing quarters during Obama administration which do not immediately follow a massive cut.

Of course, and the probability math gets a lot more complex here), the plus 4 after the tax cut this year occurred after a large increase in interest rates, so instead of a monetary tidal wave like the first 4 plus under Obama, or a monetary high tide as in the case of the other 3 plus 4s under Obama, this one occurred during a monetary tide outflow.

The rate hikes started in Nov. of 2015 and Obama had no plus 4s at all after that, and that quarter itself had a very low, less than 1/2 a percent, growth rate.

Not that I recommend analyzing policies by counting plus 4 quarters and comparing them. It decreases the signal and lowers the resolution, but since Pethokoukis used this approach, I thought it would be useful to pursue a more rigorous version of it. Using that approach to try to disassociate the last quarter's spike from the tax cuts doesn't seem to have worked.

If one wanted to turn the resolution back up and look at the actual growth rates, not the beat the hurdle rate, we end up with average growth rates like this:

After tax cuts


Trump Years


Obama Years


Ten Years


Remember that almost all of the ten-year period and the overwhelming majority of the Obama years occurred in an ultra-low rate environment, whereas all of the Trump years occurred in a rising rate environment.

But there's something else to take into account, a completely different coincidence to account for if the recent growth spurt had nothing to do with the tax cut - the slowdown before the tax cut kicked in.

GDP growth fell to roughly 2 1/4 during the quarter when the cuts were expected to pass, but had not yet been implemented. As I predicted during that time (and what better test of a theory is there than to use it to predict an event, and then wait to see if the event occurs?), business did the rational thing and deferred revenue-generating (meaning also GDP-generating) activity until the expected future cuts.

Time will, of course, provide us with more data, but so far, the data overwhelmingly support the notion that the tax cuts had an effect on this latest growth spurt.

But Pethokoukis is right to question the sustainability of those rates, not because the tax cuts didn't matter, but rather because they did. The evidence, as I've discussed elsewhere (LINK to tax cut videos on Vident if they're there) is that since part of the reason for growth spurts is that deferred activity kicks in, eventually that deferred activity all occurs and the effect wears off and growth rates return to a more sustainable level. 

That's a look at GDP. Other maco-economic metrics such as Gross Output and Profit tell the same story, which we'll see later.