The Washington PostDemocracy Dies in Darkness

What’s (not) up with productivity growth, part 1: Why this key variable is slowing across advanced economies

Source: Jason Furman, Council of Economic Advisers
Placeholder while article actions load

Off the top of my head, I’d say we’ve got five big economic problems:

1. High levels of income inequality and low economic mobility;

2. Large swaths of households/neighborhoods/countries facing stagnant or eroding living standards and severely constrained opportunities;

3. Persistently slack job markets;

4. Some really bad policy responses to the above;

5. Slowing productivity growth.

I don’t mean to be that guy — the downer economist — and am fully capable of coming up with positive attributes as well, especially for the U.S. economy. It has shown itself to be exceedingly resilient in the face of large, negative shocks and long-dysfunctional federal politics (and some policies, especially the quick reaction to the crisis in late 2007 by both fiscal and monetary forces, helped bring about this outcome; see here for details).

As someone who has worked for politicians, I learned that you “don’t bring problems unless you’ve got solutions.” My book, The Reconnection Agenda, purports to go at bullets 1-4 above.

But today, I’d like to begin to talk about the fifth problem — slowing productivity growth. I say “begin” because a) this stuff is dense and I don’t want to try your patience, and b) there’s a lot to say about it. More accurately, while it’s true that economists say a lot about what drives productivity growth, it’s also true that our understanding is limited.

So I’ll open this series with a compelling, albeit disconcerting, picture.

First, definitions. Productivity is simply output per hour. If your factory generated $100 worth of widgets per hour last week and $110 this week, your productivity grew 10 percent. Scaling that up to the macroeconomy, we begin to see why productivity is such a critical metric. It’s the way we create the potential for higher living standards. Faster productivity growth means we can invest and consume more per hour of work, or, we could work less without giving up much by way of consumption (Americans tend to take their faster productivity growth in more consumption; Europeans, in time off).

BTW, referring back to the first bullet above, given our high levels of inequality, that word “potential” in the last paragraph is important. There was a time when the typical, or median worker, could count on getting a boost from faster productivity growth. But that’s not been the case for a while now (see Figure A here).

So faster productivity growth is not by itself sufficient to raise the living standards of all who help to generate it, but it surely is necessary.

Which is why the figure above is of great concern to economists far and wide. Productivity growth has slowed significantly, and not just here are there, but in most of the advanced economies, including the U.S.

Moreover, we’re not talking about a blip, but a trend that’s persisted for a few years. Double moreover, it’s happening in enough places that you might conclude there’s some sort of global productivity virus underway. This is a very big deal.

If we’re really stuck in the productivity doldrums, and there’s at least one reason I’ll mention as to why we might not be, our ability to reap the benefits of faster growth will be severely constrained. When the pie grows this much more slowly, the fights over the slices get that much more intense. And given the inequality-induced skew in the bargaining and political clout of capital versus labor, I don’t think we’ll soon start to see a lot more output-per-hour find its way into middle-class paychecks.

Like I said, this is episode 1, so let me just tick off a bunch of suspects which I will revisit more fulsomely in subsequent posts, each with a qualitative grade of sorts:

We’re all still in the throes of the Great Recession: Demand and what Keynes called economic “animal spirits” just hasn’t yet recovered from the downturn. Economic problem #4 above — bad public policy — is implicated, especially in Europe. [Grade: I’ll bet this is in the mix.]

Technology, a historical driver of faster productivity growth, has stalled out. Productivity expert Robert Gordon hypothesizes that the productivity boosting innovations of the past, like electricity, plumbing, air conditioning, were far more transformative than the wares of the IT revolution (his paper on this is called Is U.S. Economic Growth Over? [Grade: maybe, but really hard to see around these corners.]

Technology hasn’t stalled, but it’s failing to “diffuse.” In a really interesting but dense new report from the OECD, they argue that there are still highly productive firms, but for some reason, their innovative prowess is not trickling down to lesser firms. [Grade: Interesting and plausible, but I can’t tell from their data if such diffusion contributed more to faster economy-wide productivity growth in the past.]

The quality of our workforce is deteriorating. Given that labor quality is an important input into productivity growth, this is a reasonable hypothesis. The problem is that the data, compiled by economist John Fernald, do not show such a deterioration. [Grade: It doesn’t seem to be in the data.]

We’re failing to invest in new, productivity-enhancing capital equipment. Along with labor quality and technological innovations, deeper capital investment is the third leg of the productivity growth stool. Here we’re onto something, as capital investment has slowed sharply in all of these countries, e.g., it explains the almost two-thirds of the decline in U.S. productivity growth. [Grade: this is clearly a factor, but it’s no less a head-scratcher. Now we have to explain why investment has trailed off. Weak demand is surely in the mix here, as above.]

We’re doing better than it looks (i.e., our productivity measures are biased down): It’s hard to measure both output and inputs and systematic measurement error has long been a candidate in this debate. Recent analysis by Goldman Sachs economists argues the quality improvements in IT and the proliferation of free digital content could be adding 0.5 percent to GDP growth per year that we’re not picking up. [Grade: That’s big, but since we’re talking about growth rates here, not levels, for this critique to stick, the bias has to keep growing over time. Also, they don’t count quality deterioration, like being stuck in a phone menu when you call some service provider … or flying these days.]

After all that, let me leave you on a more upbeat note. One thing we know about productivity growth is that it is not “near-term autoregressive.” Feel better? Sorry — by that I mean you cannot reliably predict this year’s productivity growth by looking at last year’s, or even that of the last five years.

In fact, those same Goldman Sachs researchers found that they had to take 40-year averages of past growth rates to get the best predictions for the next five-year average rate. I’m not really sure what to make of that other than this: based on the historical record, the fact that we’re stuck in the productivity doldrums now doesn’t necessarily mean we’ll be stuck there for long.

More to come on this in future editions of our new series.