The WSJ has a piece out this AM making an argument that’s become common from economists who think about growth, myself included: there’s no good reason to expect team Trump to achieve their 3 percent GDP growth goal. Many previous pieces have made this case, based on the limits we observe to the two additive components of GDP growth: labor force and productivity growth.
I’ll get to those in a moment, but to me, the important and timely line in the WSJ piece wasn’t so much the analysis of why 3 percent is probably an unrealistic goal (though good for the Journal for amplifying this point). It’s this one sentence, with a factoid you need to know (my bold): “If the economy expands at around a 3% rate over the next decade—a projection Mr. Mnuchin says the administration will make in its budget proposal later this month [next week, actually]—government revenue over the time should be $3.7 trillion more than currently forecast, according to estimates by economists at Goldman Sachs Group Inc.” And those dudes know their trillions.
So that’s the play here. They’re going to pretend to get the extra revenue they need to offset their big tax cut by making an unrealistic growth assumption. The official revenue scorekeepers won’t buy it, so eventually we’ll see a more plausible score with lots of red ink, but when you hear administration officials claim otherwise, you’ll know why they’re wrong.
That said, people who have very reasonably come to distrust economists’ proclamations about growth forecasts might well ask, “how do you know what’s possible re future growth rates?” Fair point. One of the key drivers, as noted, is productivity growth, and economists failed to forecast both the mid-1990s speedup and the subsequent slowdown.
On the labor force—the other growth component—we’re on much more solid ground. Our aging demographics are baked in the cake, so, absent a bunch more immigration, which is possible but unlikely, that’s going to account for 70 percent of the slowdown, as shown in the table below from CBO projections (1% of the 1.4% growth deceleration).
On productivity growth, like I said, we just don’t really have a great bead on what makes it speed up or slow down, so our best move is to assume the long-term trend persists. Over the past 50 years, productivity growth has been around 1.5 percent, about where CBO is in the table above. No one can be too confident in that prediction, but you can be sure that there’s absolutely nothing in the Trump agenda—or anyone else’s agenda—that would justify a predicted jump in that growth rate.
One final point designed to encourage a healthy dose of skepticism about these forecasting exercises. The WSJ includes the figure below, showing the potential growth rate, the same topline variable in the CBO table above. Note how it wiggles up and down over time, suggesting movements in maximum labor force and productivity growth.
While there’s definitely some rich analysis that goes into this measure—CBO does the best work one can do on it—it ends up being pretty close to a simple trend extraction. That is, for all our number crunching and forecasting of what’s going to happen, we’re pretty much looking in the rearview mirror and saying, “the future will be a lot like the past,” especially regarding productivity growth.
Let me show you what I mean. The figure below plots the same potential GDP growth line from the WSJ figure against a smooth trend extracted from actual GDP growth. They’re not the same but they’re close. Interestingly, they diverge significantly at the end, but even if the trend is right, it’s still just about 2 percent, which is, in fact, the underlying GDP growth rate since 2000.
Summing up, economists’ productivity predictions tend to be off the current trend, which is as it should be, but admits that we’re lousy at catching turning points. Certainly, if you’re budgeting for the future, that’s by far the only prudent play. Any assumption of significantly faster growth should be treated with as much respect as magic beans. Aging demographics, on the other hand, are slowing growth in a way we can somewhat reliably predict. And most of all, fully discount BS revenue projections off of phony growth forecasts.