There are a lot of complications one can discuss about quality and cost of living, preferences, sources of rising costs or prices, complements and substitutions, absolute versus relative incomes, etc. etc.

At the core of the topic of this paper is that rent inflation has been very high, and very regressive. Forget for a moment all of the deeper implications, debates about causes, analysis of various models of real estate values. There is a simple fact at the center of this work, of a significant scale: American families have suffered a bout of regressive rent inflation of unprecedented proportions that has coincided with a period of tight mortgage access. Or, if you still feel inclined to be skeptical of all of this, it is a period which has coincided with very low volumes of mortgage originations to borrowers with low and medium credit scores.

I have developed a narrative framework that I think connects all the dots about real estate valuations, construction activity, rents, etc. But at a simpler level, there are these simple facts, and these simple facts probably don’t mix well with the other explanations you or I have or had favored for high real estate values.

And dealing with that simple fact - rent inflation has been significant and regressive - means coming to terms with the scale of economic struggle that the housing market has been creating for poorer American families.

Using adjusted Zillow rent data, rents in the 20% of ZIP codes with the lowest rents in 2015 have inflated by 38% in the last decade, after adjusting for general inflation. The average price of other goods and services has risen by about 30%. Rents in the most expensive ZIP codes are up by a bit under 40%, and the cheapest ZIP codes are up nearly 80%.

And, since the poorest families tend to live in the cheapest neighborhoods, they are families that tend to spend 1/3 to 1/2 or more of their incomes on rent. That combination is why this regressive inflation has lowered their actual incomes by so much more than what standard analysis says.

Rent inflation has been highly correlated with neighborhood incomes, and none of our standard ways of tracking incomes and cost of living are set up to capture this. Renters with the lowest incomes have had 15% of their incomes taken over the last decade in a way that is missed by aggregate statistical measures.

The BLS does try to measure cost of living differences for different families. They have a special report intended to measure inflation for poor families versus rich families. But, they only account for the issue of the consumption basket - that poor families spend more on rent than rich families do. They don’t have a different inflation rate for the two families. They would apply 24% excess rent inflation to both rather than applying 9% to the rich families and 38% to the poor families. As far as I know, there is no published public measure of well-being that accounts for this issue.

The reason is that Americans have managed to make it legally difficult to provide a basic human need, which is a historically peculiar and unsustainable thing to do, and it makes really weird things happen. The BLS isn’t set up to track this because, for Pete’s sake, it shouldn’t have to, and it has never had to before.

In a footnote in the special report about variances in experienced price inflation, they write, “Most goods and services use the geometric mean formula because consumers are generally able to substitute away from any particular item whose price is rising relative to others. Rent and Owner’s Equivalent Rent are calculated using a Laspeyres formula because consumers cannot easily move in response to changes in rent.”

They note that there are other categories that this applies to, like prescription drugs. But, when the price of medicine goes up 10%, it goes up 10% for everyone. The housing we consume is associated with our income levels, and housing shortages create this peculiar pattern of regressive inflation. Rents have gone up in poor neighborhoods by much more than in average or rich neighborhoods.

That point is why, (1) this problem can be described without some of the typical complications about the effect of substitutions over time, and (2) this problem is a really big problem.

The reason rent inflation has been so high and so asymmetrical is because there aren’t easy substitutes. And, since there aren’t easy substitutes, two families living just miles apart from each other can experience significant differences in rent inflation and cost of living trends.

In national statistics about income and cost of living, it is difficult to create comparative measures over time. If a 14” black & white TV took a week’s wages to buy in 1960 but today you can buy an 80” high definition color TV for a week’s wages, how much better off can we say today’s consumer is than the consumer in 1960? A TV better than the TV that the consumer purchased in 1960 could be picked up free of charge today. So, comparing today’s consumer to a 1960 consumer, should we say that today’s consumer gets a TV for free, or that both of them spend a week’s wages on TV?

Over long periods of time, all of these substitutions, preferences, and the scale of the incomparables make it really hard to compare one family to another with precision.

But, as the BLS staffers from that report note, rent over the period of a decade isn’t particularly complicated by all of that. Millions of families across the country are living in rental units where monthly rent has increased by something like 75% over the course of a decade. They are still living in them and paying the higher rents because there is no way around it.

Before 2008, this was a regional problem, mostly confined to New York City, Los Angeles, San Francisco, Boston, and San Diego. At that time, there was a substitution available, though not an easy one. Every year, hundreds of thousands of residents from those cities could choose to be regionally displaced - they could choose to move to a cheaper city. I should say that hundreds of thousands had to move to a cheaper city. There was no choice about what had to happen. The choice was about who had to do it. And, the high and regressive rents in those cities is a measure of how hard that substitution is - how hard the alternative had to get for some families to make the choice to be the ones to leave.

This is why I go on and on about the problems with the agglomeration model that economists tend to use. The decision to move to a place isn’t so dramatic. If the problem was that it was harder for Americans to move to some cities, these patterns in rents and prices would look very different.

If cities were expensive because they were growing, dense centerpieces of labor and innovation, there would be plenty of substitutions available. Make units smaller. Build higher. It would be sort of the opposite of the TV question above. If that was the binding issue in expensive cities, then families would be spending a stable portion of their incomes on rent, but over time, units might be getting smaller. And, we might ask, "Is today’s resident worse off than residents were 30 years ago? They both spend 25% of their income on rent, but the new resident lives in a smaller unit.” And, the agglomeration answer to that question might be “Sure, but the current resident also has access to an innovative industry that increases their income, so it’s a positive trade-off.” That answer could be the correct answer. It even applies at some scale to some residents of those cities today. Many people move in to New York City and live in a smaller unit in a bad local location in order to access those benefits.

The agglomeration question applies to people who live, and will live, in growing cities. Those cities exist. At some point, all cities were growing cities. Even if they aren’t growing, they might still have value related to agglomeration and urban amenities. That’s not the issue that is making this problem a centerpiece of the national economic conversation. It is true and irrelevant. That context does not produce 38% excess regressive rent inflation.

The topic of this paper, and the public conversation it relates to, is families that have or will have to move away from stagnant or shrinking cities and families just trying not to trade down in countless cities that are new to this phenomenon since the public policy disaster of 2008.

The reason families put up with 38% excess rent inflation is because the substitution mechanism is displacement. In cities like New York City and Los Angeles, the only poor families that haven’t moved away are those that either have some protection from rising rents or have such a strong connection to the location that they can’t leave. An increasing number of those residents are living in tents, cars, and shelters.

The families whose rents have gone up the most are in the lowest income quintiles. They are living in the lowest quintile of homes. There is nothing to trade down to for them.

And, now that it has become a nation-wide problem, the bottom rent for homes in every market has risen something like 75% over a decade. Recent cumulative rent inflation within each city can differ by 30%, but between cities, it has become surprisingly uniform. There is nothing to trade down to.

And, as far as I can tell, the economics academy just doesn’t know. There are examples like the Joe Stiglitz paper I discussed in the previous post. Papers get published using metro level rents that conclude that rising rents are caused by high demand from newcomers and when they come, locals move away easily as the economics move against them. Hosts of reviewers and mentors apparently aren’t aware of the empirical problem with that methodology. Economists and pundits push back against a sense of economic malaise because they show that the average home is bigger than it was 50 years ago and the average income adjusted for average inflation has risen over that time. I think they just don’t know. They must not know. It’s a correction they need to make in order to engage in the discussion with clarity and credibility. They have important things to say, but they are making an empirical error of a significant scale.

It may be helpful to think about the basket of services we are buying when we purchase housing - upgrades, extra space, shelter, basic personal infrastructure, or location (which could mean near Broadway shows, near a good job, near a menial job, or near family).

The rent inflation we have been accumulating is for nothing. It’s just for scarcity. So, the part of that bundle of services called “housing” that some families are most desperate not to give up is the part that is associated with rent inflation. It’s not the luxuries that are making housing expensive. It’s the things families aren’t willing to give up when they are bundled with expensive “nothing” that are raising the cost of housing. Being poor has become more expensive. It has become more expensive to exist in America.

This is why the solution to this problem isn’t to build more “affordable” housing - whether that’s smaller homes with fewer amenities or subsidized homes. I mean, if that was our only option, then building those homes would be better than not building them. But, the problem for American households is that when they purchase housing today, there is a fee for “nothing” that is the base price for what they purchase. They need the price of “nothing” to go down, because where “nothing” is bundled with a location close to important idiosyncrasies they value, know, and understand and won’t give up, then they will pay a lot for “nothing” to keep those other things.

We can’t reduce the price of “nothing” by increasing the supply of “nothing”. But we can increase the supply of houses so we don’t have to pay for “nothing” any more. And, it really doesn’t matter what houses are built.

That means that blocking “luxury” homes because we need “affordable” homes isn’t helpful. It also means that denying that there is a problem because the average home in 1970 was smaller than the average home today isn’t helpful. It also means that improving productivity or automation in construction isn’t very helpful in the current context. That won’t make “nothing” cheaper. We already have infinite supply of “nothing” and “nothing” is what we currently have to pay for when it is bundled with things we need. Families whose incomes are being swallowed by this problem aren’t paying for square footage. They are paying for “nothing” and “nothing” happens to be bundled with square footage close to grandma, an old job, etc.

This is a subtle issue. Let me be clear. Of course productivity is good, on the margin. My point is that currently, roughly speaking, the mean US home price is about $480,000. About $250,000 of that is for square feet. Maybe about $75,000 is locational amenities. And about $150,000 is for nothing. The cost of nothing in most of the country, as late as about 2015, was nothing. This is all a recent phenomenon that spread across the country because the construction industry was so hobbled that by 2015 its capacity was well below a sustainable pace of construction, and it would take years to return to a sustainable pace of construction. To get “nothing” back to free, we need a lot of square footage.

If we could reverse 20th century urban planning obstructions and put $3 trillion worth of square feet in locations with $3 trillion of potential locational value, that would be great. If we can’t do that, we need to put $6 trillion worth of square feet somewhere. That will come in the form of new homes, so that the average home will still have about $250,000 of structural value and $50,000 of locational value, and the “nothing” part will be free again. You won’t have to pay extra to avoid displacement.

That’s all true regardless of what productivity is. At current construction productivity levels, $6 trillion might produce 30 billion more square feet. Maybe if productivity was better, it would produce 35 billion more square feet.

It’s not particularly relevant to our current condition. Families aren’t paying elevated prices for more square feet. They are paying for nothing. 15 million 2,000 square foot homes will solve that problem almost as well as 15 million 2,300 square foot homes would.

You could argue that the mortgage crackdown was a hit to property management productivity because it locked millions of potential property owners out of the market, so that now rents are higher for any given structural value. That boost in productivity would probably create more value for renters than any achievable structural productivity improvements would. And, again, this is under current conditions. Of course manufacturing and construction productivity are good. It’s a sign of how screwed up we’ve made our residential real estate markets that it is not among our most binding problems.

This is also why I am comfortable referring to the issue as a housing “shortage”. Some economists bristle at the use of that word because economists have a pedantic definition of the word that laypeople frequently misunderstand. But, the 38% excess rent inflation for families with no available substitutions is for “nothing”, and paying for “nothing” is a sign of a shortage, in the pedantic sense.

PS. You might have noticed in Figure 1 that rent inflation has leveled out since 2022. I’ll have more on that for subscribers, but I will save it for another post.