The pent-up demand question

Are we really headed for a spending spree?

I like using this newsletter to write big-idea pieces which, though they may incorporate a bit of economics, are really about broad questions relating to how society works. That stuff is fun for me to write and hopefully for you to read. But there are a bunch of simmering macroeconomic debates happening at the moment that I’m trying to think my way through, and I want to do some of that thinking here. What I’m saying is: if notes on inflation and such aren’t your cup of tea, I hope you’ll bear with me for a week or two—just roll your eyes and scroll past the email in your inbox if you like—until I get back to reflecting on, oh say: what the Fermi paradox tells us about the nature of human progress and the future of the market economy.

Today, I want to discuss pent-up demand. Around the internet, people are arguing about the likely trajectory of the American economy over the next year or two or three. Hopefully, as vaccinations continue, we get closer and closer to something like a normal state of affairs in which people can once again be close to lots of other people. As that happens, the kinds of economic activity that have been suppressed by the virus and by measures aiming to keep the virus under control should begin to resume. A really big question is: what happens then? The US is currently about 10m jobs short of the level of employment a year ago (and 12m short of where it likely would have been had there been no pandemic). Do we very quickly recover all those jobs as the economy resumes normal operations? (For context, from April to August of last year, the employment deficit dropped by about half: from the more than 20m jobs lost in the spring to about 10m by the end of the summer.)

The answer to this big question depends in large part on just what sort of effect fiscal policy has in coming months. Back in December Congress passed $900bn in new fiscal support for the economy, and it looks as though another $1.9trn bill is going to go through in the very near future. Much of that aid is intended to help speed up vaccination efforts and assist with other public health needs, and some of it is relief to households and businesses that are just trying to survive, but a fair amount looks more like garden variety stimulus, and at any rate all of the money stands to add to aggregate demand. It’s a good chunk of change, and in fact some people are now warning that it could be too much: that the amount of stimulus in the economy risks pushing spending up above levels the economy can accommodate, and it might thus cause an uncomfortably big jump in inflation.

One of the things the inflation worriers are concerned about is the fact that the government also provided quite a lot of fiscal help last year, and that as a consequence of this help household incomes did not fall by very much and savings rates soared. Savings, they argue, represent pent-up demand: deferred spending which will no longer be deferred once the economy is healthy again, and which will instead be added to the newly passed stimulus to create a torrent of cash chasing too few goods and services. 

What I’m wondering is: will it though? On its face, the notion seems to make sense. Savings did rise a lot over the past year; the personal savings rate as of December was nearly 14%, which is the highest it’s been since the 1970s. It’s not that surprising that savings are so high. Incomes didn’t fall that much last year, as mentioned above, but many parts of the economy were closed to consumers. People couldn’t buy what they wanted to buy so they saved their money, and it stands to reason that when they can buy things again they will, drawing down their savings to acquire the stuff they couldn’t in 2020. We even have an historical parallel to work with here: the second world war. Savings soared during the war because incomes were skyrocketing while production capacity was largely reserved for the war effort. Then after the war, flush American households went on a spending spree, buying up all the cars and appliances and what not that they hadn’t been able to before.

Look closer, though, and things don’t seem quite so straightforward. Consider the distribution of savings, for one thing. Higher-income households have been more able to save than lower-income households. Among low-income households, spending In January was actually quite a bit higher than it was a year ago, while spending among high-income households remains depressed. Now, it is possible that the well-off will go on a spending binge this year. But maybe not. Higher-income households have lower propensity to spend than lower-income ones, and many may be content to keep their savings in savings. (Our household is not rich by any means, but we’ve been fortunate to be able to save more over the past year than in prior years, and this has largely allowed us to get closer to our desired savings level than we’ve managed in the past; we don’t intend to spend our savings down, in other words, at least not until the kids go to college or the furnace dies or we retire.)

The comparison with the second world war is interesting in this respect. The conclusion of the war led to a sharp decline in the American savings rate. But it didn’t really give way to a period of pronounced dissaving, in which people spent down what they’d accumulated before. Indeed, after its initial drop, the savings rate climbed substantially from the mid-1940s to the early 1950s. Consumption did soar after the war, but that was in large part due to the fact that incomes were growing robustly, and that people were saving less out of the income they earned after the war than they saved out of their income during the war. Reflecting on this, it seems possible that a rise in consumption this year funded by the drawing down of accumulated savings is most likely to occur in a world in which incomes are not growing by very much—perhaps because the recovery remains weak—and in which households thus have little choice but to spend out of their savings. In the same way, a stronger recovery, in which income growth is healthier, is likely to yield less of a drawdown in accumulated savings.

In addition, it is worth noting that consumption spending has not been completely repressed. Here in the US, there have been some things which you haven’t been able to buy. There have been many other things which you’ve been able to buy provided you’re willing to accept some level of increased health risk. And there have been a lot of things that you can buy as much of as you can afford. We have seen much more of a shift in the composition of consumption than we have a shift in the level of consumption. The second world war again provides an illuminating point of comparison. In 1942, real GDP rose by a staggering 19%, but durable goods consumption fell by 37%. That’s wild to think about. In 2020, in contrast, real GDP dropped by 3.5%—but consumption of durable goods rose by 6.4%.

Personal consumption did decline overall, but the drop was entirely associated with a fall in spending on services. To what extent is pent-up demand for services a thing? How is it going to work? People will begin consuming personal services of various sorts again, but they’re not going to get two haircuts or two physicals to make up for the ones they didn’t get last year. If there is no lingering reticence to engage in activities in crowded places, then we probably should expect to see some bounce back in food services and similar activities. People will very probably go out to eat more often than they normally would have pre-pandemic (I’m pretty sick of cooking), see live entertainment more often, and so on. I certainly expect there to be increased demand for many recreational activities: holiday bookings and such.

To what extent will increased spending run up against capacity constraints? It’s honestly not clear to me. It’s possible that some of the changes in consumption habits we experienced over the past year turn out to be more durable than we anticipated, such that attendance at movie theaters, say, just doesn’t rebound quite as much as we might have thought it would. It’s possible that increased spending is about quality upgrading in addition to increased frequency of consumption: you go to the sit-down chain restaurant rather than the fast-food outlook, for instance, or you book a more expensive vacation rather than a second vacation (since it’s not like Americans have any more annual leave than they normally do). It’s also worth noting that prices for many services have plummeted over the past year, such that a big jump in prices in response to renewed demand would in part reflect a return to the pre-pandemic price level. And to some extent increased demand for services will be met by undoing some of the shift in the composition of spending which occurred last year; so, some people who’d worked delivering meals to houses in 2020 will shift into work at restaurants in 2021.

In terms of the risk of overheating, there are other things worth considering as well. How much slack remained in the economy prior to the onset of the pandemic? If people go nuts for meals out this year and there’s a crush of restaurant hiring, it’s possible some of that just pulls into the labor force more of the workers who never did return after the global financial crisis. It also remains to be seen whether there has been some finding of efficiencies in service industries as a consequence of the pandemic (thanks to increased use of telemedicine, for example).

But I think we should also be aware of the possibility that what looks like pent-up demand really isn’t—or isn’t to nearly the degree that the aggregate savings rate suggests. And furthermore, it is these kinds of uncertainties, alongside the asymmetry of risk associated with inflation relative to unemployment (the latter is way worse), which argue in favor of maintaining a dovish policy stance. It might be a little uncomfortable to see inflation rise to 4% or 5% this year and next (although I think we often overstate just how uncomfortable such things really are). That discomfort would come with certain compensating benefits, like really rapid declines in unemployment. But if we’ve guessed wrong, and demand comes in weaker than expected, and the recovery is slower than we hoped for, and the political moment for stimulus looks to have passed…? We really don’t want to be in that place.