Is the Affordability Crisis a Crisis?

I think so. Here's why and what we can realistically do about it.

How a Slow Burn Becomes a Crisis

I was asked the titular question of this post last week by Jenn White, host of the NPR show 1A. I answered that “crisis” seemed fair to me, even though affordability OG Annie Lowrey, who was also on the show, had been writing about the issue for years, focusing on housing, health care, child care, and higher education. But the case for moving from slow burn to crisis comes from the figure below.

The figure plots the price indexes for goods, excluding food and energy, and for groceries. I’ve long argued that affordability stress is driven by the price level, not the rate of inflation. Well, these two series show how sharply the level jumped as snagged pandemic supply-chains collided with strong demand for manufactured goods and groceries. The sudden shock to the price level—and I’m just using these two as an egregious proxy; the key fact is that the overall price level spiked—even over a period when incomes were boosted by savings and fiscal support, triggered the intense dissatisfaction with prices that has persisted ever since.

The fact that, post-spike, inflation in these series settled back down to earth never made people feel much better, because they were left with the elevated price levels. Economists focus much more on inflation—the rate at which prices are changing—than price levels, but if we want to better understand how people experience an economic (not to mention health) shock, this is too limiting.

The key point is that inflation doesn’t have a memory but the price level does. And you know who else has a memory? People! Let me unpack that a bit. Suppose some crazy shock sends inflation to 10% one month and then back down to 2% the next month. Economists and the Fed will “look through” the shock, declare supply and demand to be adequately realigned, and go about their merry business. But people qua consumers will be like “wait…what??” as their price levels are all the sudden much higher (and their incomes haven’t had time to catch up).

That’s not all. If you closely scrutinize the ends of the lines in the above figure, you’ll see that Trump’s tariffs are bending them up a bit, especially goods, which makes sense since many of them are directly targeted by his sweeping import taxes. The technical term is “adding insult to injury.”

Surely this crisis can’t persist forever. As long as wage and income growth surpass price growth, and we avoid spikes (and tariffs!), people will eventually acclimate to the higher price levels. But a) that hasn’t happened yet, and b) I wouldn’t count on acclimation to occur in healthcare, childcare, housing…areas where structural in-affordability persists, and where affordability policies are urgently needed (more on this below).

You Need Price Levels, Not Inflation, To Explain Consumer Sentiment

A very common refrain these days is that consumer sentiment is unusually negative given the state of the economy, as in low unemployment, rising equity markets, inflation way down from the 2021-22 rates. A good way to show this, as Cummings et al do here (Figure 2, based on earlier work by David Wilcox), is to build a simple statistical model wherein inflation, unemployment, real spending, and the S&P predict the UMich Sentiment Index. If you run the model through 2019 and then forecast forward based on actual values of the independent variables, you generate a fine fit for most of the sample but a clear gap over the forecast period, as shown in the “basic model” green line below.

You call that the vibes gap, as in the econ data say sentiment should be a lot higher than it is, but people are not feeling it.

But what if we try to incorporate the insight above re shocks to the price level? For technical reasons you can’t just plug the price level into the model so I did something a bit different intended to capture the impact of the price spike (see ftnt).¹ The prediction from that model—the red line—generates even worse sentiment than the actual data! People are downright upbeat in the sense that the predicted value that incorporates the price spike is below the actual read on sentiment!

Caveats abound. This a very simple model. There are some issues with the underlying UMich data. My spike variable (described in the ftnt) makes some judgement calls (e.g., I assume people expect inflation to trend at the rate it has over the past three years—though I fooled around with this assumption and it didn’t make much difference).²

But I’m confident this analysis is telling us something truthful, something that is consistently misunderstood by critics of affordability angst, who look at real wages, incomes, GDP, net worth and declare “Everything’s fine! Stop complaining!” The fact is that shocks to the price level have a much longer psychological shelf life than many imagine. FTR, I strongly suspect that a big reason for this is social media, which is a massive negative force multiplier in this space, but that’s just a supposition. Someone needs to write a dissertation chapter about that.

What Happens Next?

The main solution to the economic and psychological damage from a shock is that enough time must pass for the impact of the shock to fade. But that’s not the only solution. Here’s what else has to happen while time rolls along:

—Inflation must move back to target and stay there. Additional aftershocks to inflation will be very upsetting to people, as their acclimation to the new levels will be interrupted. This is one reason why Trump’s tariffs are not just counterproductive but terribly timed. The “insult-to-injury” point above is important in that it’s really jamming the acclimation function.

—Real incomes and wages must rise so people can get back their buying power. I’ve underscored the essential nature of this part of the solution, but not nearly as compellingly as Heidi Shierholz from EPI:

Today’s affordability debate, however, focuses almost entirely on prices, as if the only way to make life affordable is to make things cheaper. But that approach misses the bigger picture. Affordability depends on both prices and wages. The roots of today’s affordability crisis actually lie not in recent price spikes, but in the long-term suppression of workers’ pay.

—You know the Affordability Serenity Prayer, right?

Keynes, grant me the serenity to accept the prices I cannot influence,
the courage and policy acumen to influence the prices I can, and the wisdom to know the difference.

Policy makers must continue to plumb affordability economics in this spirit. For the vast majority of goods and services, markets work pretty well and care should be taken to use them to consumers’ advantage. In other—often very important—cases, markets either fail or inadequately provide the supply that’s needed at an affordable price point.

Let me explain (using shorthand and leaving out important nuances): childcare is on one side; housing and healthcare are on the other. As I understand it (and Chloe Gibbs et al should correct me if I’m not evincing “the wisdom” called for above) the market for childcare is quite elastic/responsive. The problem is that the people who need it most, as in young parents starting out in life, don’t have the necessary income to support a high-quality sector in this space. So, subsidizing their purchase makes sense (there are abundance-style barriers here too that must be removed).

But if you apply that subsidizing strategy in housing without building out the necessary supply, however, your subsidies will just end up being capitalized into the costs. Healthcare is another (non-)market where the “rents” being hoovered up by vested interests create a strong risk of throwing good money after bad, and as such, it is ripe for affordability reforms, as Neale M and I presented earlier this week.

I used to think that groceries were mostly an area wherein there wasn’t that much policy could do to help, beyond forcing less concentration in food production. But some creative folks have been convincing me that there may be some helpful ideas here too. More to come on that.

So, what happens next is, as usual, up to us, and by “us” I mean those willing to roll up our sleeves and implement the actions in the prayer. In today’s politics, that means the subset of Democrats who want to do what the taxpayers are paying them to do—i.e., address their problems—the academics and think tanks working on this, and the media reporting on it.

Not to mention the aging Substackers/policy fellows who view the affordability crisis as a great and compelling issue to work on in their dotage!

1. I calculated a smooth trend of the CPI level. I then took the 3 year change in that trend from 2016-19, to pick up what people were expecting inflation to keep doing. I then grew the trend variable at that rate from 2020-25. The variable in the regression is then the log difference between the actual CPI and this simulated, un-spiked trend.

2. More problematically from an econometric perspective, the spike variable is “non-stationary.”