A Review of Gjerstad and Smith, Part 4

I thought I’d have one more go at Gjerstad and Smith’s assertion that home prices should be figured into consumer price trends tracked by the Fed. I am using CPI for all items here, which is a bit noisy. Core CPI is a little more stable. But, since all items was the index they referenced, that’s what I’m using.

Inflation with Home Prices

In Figure 1, black is the 12 month change in CPI, as measured. The red line is the change suggested by Gjerstad and Smith - using home prices rather than owner-equivalent rent. The blue line is the inflation rate that would be required of the traditional CPI index in order to counteract home prices. In other words, at the end of 2004, Gjerstad and Smith pointed out that reported CPI inflation was 3.3%, but inflation was really 6.2% using home prices as they suggested. What the blue line here is saying is that in order to keep their preferred index at 3.3%, the traditional CPI would have had to have been 0.4%. I’m not forcing the adjusted CPI to the 2% target here, but simply imagining a counterfactual where Gjerstad and Smith’s CPI followed the same trend over time that the traditional CPI did.

In Figure 2, I imagine a similar, simple all-else-equal counterfactual in nominal GDP growth. Here, the idea is that home prices aren’t part of the GDP measurement, so if the Fed had targeted a 2.9% lower inflation rate at the end of 2004, and there were no secondary effects except for lower prices, then nominal GDP growth in 2004 would have been 3.8% instead of 6.4%.

In general, it looks like targeting home prices as part of inflation produces a relatively stable NGDP growth pattern. It would have been around 4% at the peak of the 2000s housing boom instead of 6-7%. And traditional CPI inflation would have been slightly positive from 2003 to 2005, which, under the conditions of the time, probably would have been fine.

Of course, not all else is equal. Some of the lower growth would have been lower real growth rather than simply lower price inflation. And, some of that lower real growth would have probably meant lower residential investment and fewer new homes. That would worsen our housing supply problem. But, maybe that wouldn’t be the worst thing, if it meant avoiding 2008.

Not as wealthy as we think we are

I tend to push back against the idea that “we weren’t as wealthy as we thought we were”, which generally acts as a sort of apologist’s position to excuse the negative wealth shock of 2008. It is based on the idea that residential wealth in 2005 was just paper wealth because of faulty bubble models that overstate the role of credit markets.

I think that approach is an example of attempting to make us nominally poor enough to fit into our inadequate housing stock. In practice, that’s what it led to. It is the imposition of a negative nominal income shock because of model error. And, that ended up having devastating real consequences.

I do think that there is a way in which “we weren’t as wealthy as we thought we were” is true. And that is that the wealth we measure as “real estate value” is elevated because of the expectation of future excess rents. Those future excess rents are a liability that matches the excess wealth of real estate ownership. But, the liability of future excess rents does not appear on current conventional measures of aggregate net worth.

So, we are aren’t as wealthy as we think we are because our lack of adequate housing will create an ongoing liability to rentiers. But, the solution to that mismeasurement isn’t to create paper losses for the real estate owners. Those future excess rents still exist, no matter how hard you work to beat down real estate prices. That’s exactly what we did in 2008. That’s what people who have a bubble-focused model of housing thought we needed to do. That is what was unsustainable. And the way that it ceased to be sustained was the future excess rents increased even more until real estate wealth recovered. So, now we are even less wealthy than we think we are, because our excess rent liabilities are even higher. But it has jack all to do with credit bubbles.

Counterfactuals are hard

Anyway, back to the charts, if targeting a CPI that used home prices had been consistently followed, it would have moderated nominal GDP growth in both boom and bust. That is curious, because normally, I prefer to simply remove housing from the CPI, which I also think provides a more relevant guide for monetary policy.

But, maybe in an economy that is so tied up by its political housing constraints, tracking the value of economic rents as a sign of cyclical variability makes more sense than ignoring them.

I prefer nominal GDP growth targeting to inflation targeting, and 5% seems like a reasonable number to me. But, clearly there is a lid on our growth potential. In 2004 and 2005, nominal GDP growth was a bit over 6%. This was lower than in any previous post-World War II expansion, but some of our largest cities - the Closed Access cities - have crippled their ability to grow at much more than 3 or 4% plus inflation. And, so, those nominal growth levels were associated with mass migration out of the most housing constrained cities. That migration destabilized the places that the migrants moved to.

The growth rate of some destination cities of Closed Access housing migrants was tested again after Covid, though some of the destinations were different than they had been in 2005. Maybe urban housing constraints are binding on real GDP growth, and the signal of overheating is the number of residents moving away from housing deprived cities. That’s the new, context dependent, American macroeconomic process.

The displacement of poorer families from those cities is inevitable, given their current land use regulatory governance. That’s a microeconomic problem (Build more houses.). But when it gets fast enough that other cities can’t easily take them in at the pace they are being displaced without taxing their own local housing growth capacity, then I suppose it becomes a macroeconomic problem (Don’t print more money than an economy saddled with these frictions can handle.).

Maybe, even though the bubble folks are wrong, their solution of slowing down the economy in 2004 and 2005, when home prices were rising and migrants were pouring out of the Closed Access cities, may have some benefits. Even the idea that we are building too many homes is relevant, even if it is wrong in the broader, unbound sense. Those extra homes were largely being built for the migrants. And, even within the housing deprived cities, more housing production is associated with depopulation and displacement. So, it is the case that faster real growth is destabilizing.

If a building boom will inevitably be associated with mass displacement and surges of regional migration, then, in some perverse macroeconomic sense, an economy saddled with a dysfunctional lack of well placed housing could “build too many homes”. Closed Access housing makes good things into bad things. Maybe on this issue, bubble models are stopped clocks - they recognize the bad things even if they have the causality wrong.

We have to slow real growth enough so that the disrupted lives of the displaced families don’t create a contagion on everyone else. That displacement will inevitably happen during what appears to be a “housing bubble”. That was the case in 2005 and in 2022. Home prices through the roof. Homebuilders hitting regional capacities. Closed Access cities depopulating. It’s all one big crappy package. And, if everyone incorrectly refers to that crap-show as “housing bubbles”, why worry about labels if, in either case, all of those processes become destabilizing as long as we have so many frictions in housing production.

And, as long as there are policymakers who are guided by these wrong bubble models, then the risk remains that they will continue to turn into liquidationists whenever housing booms peak. If nominal GDP growth had been kept at 4% in 2004 and 2005, Gjerstad and Smith, and so many like them, would probably have had the will to speak up against a collapse in nominal spending in 2008.

In the 2009 that actually happened, they dropped their suggestion of stabilizing inflation and home prices once they became deflationary, but if the economy had remained slow enough in 2005, then maybe their commitment to price stability would not have been tested.

The futility of natural experiments

I suppose the Covid recession provides a natural experiment. Covid provided a real shock, and rent inflation spiked sharply. The method for estimating the rent component of the CPI has a bit of a lag. So, the traditional CPI measure was sort of like a CPI measure that excludes housing, for a while, at least.

I think this was good, because as I have written, I think it was good for the Fed to hold off and to not push back too hard on temporary inflation. Real rates and forward economic potential have risen as a result of that good policy choice.

A CPI adjusted with home prices would have been above 2% by September 2020, and it would have been up to 9% by June 2021. That would have put much more pressure on the Fed to raise rates into the Covid shock.

I think the fact that inflation came down even before much of the rate hikes, and forward inflation expectations have remained well controlled, shows that being slow to react was the right choice. But, at this point, bubble-focused skeptics would hardly be convinced. And, certainly, again, bubble-focused pundits will not be pushing for Fed stimulus now that the CPI adjusted with home prices is lower than the traditional CPI measure.

So, maybe the Covid experience argues, after all, for ignoring all housing inflation rather than amplifying it by tracking volatile home prices. I think, so far, it does. Remind me to update that conclusion if the data turns against me.

And, maybe J-Pow is doing the thankless work of ignoring the liquidationists, so that we have little to worry about on the downside this time, either.

I am starting to worry again that rate cuts will be coming too slowly. I suppose if rate cuts come soon enough, and we avoid recession, this will be a counterpoint to 2008. I doubt that many 2008 liquidationists will ever be won over, but maybe tomorrow’s policymakers can be.

If rate cuts are tardy, then the misguided models that suppose high home prices come from excess money and that they are followed by inevitable recessions will have yet another fake empirical point.

I hope J-Pow is up to the task.