Was the Financial Crisis Actually Caused by a Housing Shortage?

Economists and pundits have told us repeatedly that the financial crisis was all about excess: too much credit, too much money, too many homes. You’re saying that they have it exactly backward, that it was really caused by a shortage. What evidence is there that the conventional wisdom is off-base? And why did so many people get it wrong for so long?

There are three keys to recognizing the central role of supply in the financial crisis:

1. There are five metropolitan areas that differ from the rest of the country–New York City, Los Angeles, Boston, San Francisco and Silicon Valley, and (though smaller than the others) San Diego. I call them the Closed Access cities. In those cities, housing supply is politically obstructed so that the growth of new housing units is much more limited than in other metro areas. Those cities also are center points of the post-industrial economy and they provide ample economic opportunity for those that can move there. But since housing is so constrained, a bidding war has erupted for access to their economic opportunity. Since housing is the constraint, the bidding war is for housing. The bulk of the rise in real estate values during the boom was in the Closed Access cities, and the relative value of real estate in those cities is at least as high today as it was then, compared to the rest of the country. As long as the political barriers to entry remain, and the local economies provide opportunity, those real estate values will be persistently high.
    
2. There were also a few areas, like Florida, Arizona, and inland California, which generally have more generous local building rules. In those areas, the 2004-2005 period can be called a bubble. I call these the Contagion cities. Prices rose sharply, for a short time, probably to levels that would have been difficult to maintain, and then collapsed. But, even here, there was no excess building. Even here, it was Closed Access deprivation that was the key factor creating the bubble. These areas tended to experience a surge of migration in 2004 and 2005 from the Closed Access cities. Looking only at these cities, it seems clear that this was a period of sharply rising prices and construction. But, that new construction was only partly making up for the lack of building in the Closed Access cities. Many of the new buyers moving into these cities were sharply reducing their personal housing expenditures by moving out of the Closed Access cities. This was a bubble in compromise and exclusion. These cities also, briefly, had a shortage of housing because they couldn’t permit new building fast enough to make up for the migration surge.
    
3. An important factor to look at is the rental value of homes. Rent inflation has been high since the mid-1990s, especially in the Closed Access cities. Rent is a more pure indicator of housing supply and demand than home prices are. It is understandable that this important factor has been underappreciated. Most home buyers do not think of their purchasing decisions in terms of rental value. That is, in part, because we have lacked the tools to measure rental value in single family homes. Today, with new sources of data like Zillow.com, we can estimate rental values with much more detail. Mortgage lending will always tend to grow and to be more aggressive when home prices are high. It is tempting, then, to blame high prices on loose credit. Without focusing on the different types of local markets and the importance of rental value, it looked like loose credit was to blame, because, in total, US real estate increased in value compared to measures like household income. But, at the metro area level, home prices were highly sensitive to changing rents, and even during the boom period, rent was becoming a more important factor in the relative prices of homes across metro areas.

But weren’t Fannie Mae and Freddie Mac encouraging irresponsible homebuying?

No. There is no sign in Fannie and Freddie statistics of declining FICO scores or loan-to-value levels during the boom. In fact, because of political pressure against them, during the “hot” years of 2004 and 2005, they lost massive amounts of market share to the new private securitization markets. They only began to recover market share after the private markets began to collapse. That is part of their public mission. But, since a public consensus had developed which blamed high prices on loose credit, they were prevented from performing that mission. According to Fannie Mae regulatory filings, the average FICO score of borrowers on their total book of business in 2000 was 713. By 2005, it had increased to 721 and remained there until 2007. Then, credit was largely cut off to borrowers with lower FICO scores.

From 2008 to 2015, the Fannie Mae book of business grew by $487 billion lent to borrowers with FICO scores over 740 but declined by $378 billion to borrowers with FICO scores of 740 or less. During that same period–after the financial crisis–low tier housing markets in metro areas across the country declined sharply compared to high tier markets. The vast majority of foreclosures happened during that late period. It wasn’t irresponsible lending that led to the foreclosure crisis and the deep cuts to working class wealth. It was the federal clampdown on responsible lending in low tier markets after the crisis.

You spend a lot of time focusing on different kinds of cities: closed access cities, contagion cities, and open access cities. How would you define those terms to the average American, and can you give an example of each?

Closed Access cities have several common, related traits: vibrant local economies, strong limitations to housing growth, high local incomes, high rental costs relative to those high incomes, high home prices relative to those high rents, and a persistent out-migration of households with lower incomes, wealth, and education, because of the lack of housing. This includes New York, LA, Boston, San Francisco and Silicon Valley, and San Diego.

Open Access cities among the largest metro areas include Dallas, Houston, and Atlanta. Most smaller regional metro areas also are similar. Rents tend to be moderate in these cities. Home prices, at most, rose moderately during the housing boom because of low real long-term interest rates.

Contagion cities include most major metro areas in Florida, Arizona, Nevada, and inland California. The main characteristic that sets these cities apart from other cities is that they are the main destination of migrants moving away from the Closed Access cities. They tended to have strong rates of building during the boom, like many of the Open Access cities, but it wasn’t enough to meet the spike in demand created by the migrants. The housing markets collapsed in these cities in 2008 because the migration event collapsed.

In the closed access cities, housing prices increased dramatically in the run-up to the crisis. You write about how that led to moral panic and the wrong kind of policy responses. What triggered the panic, and why did the entire economy nosedive afterward?

For most of the “subprime bubble”, the homeownership rate was declining, and the number of first-time buyers was declining.

There were some types of deregulation in mortgage markets, which led to the rise of the private securitization market (home of the infamous Alt-A and subprime mortgages and “liar” loans) from 2004 to 2007. Surprisingly, those new mortgages did not lead to a surge of homeownership among families with lower incomes. In fact, there wasn’t an increase in homeownership at all. For most of the “subprime bubble”, the homeownership rate was declining, and the number of first-time buyers was declining.

It appears, in hindsight, that mostly what the looser terms of those loans were doing was allowing households with high incomes in the Closed Access cities to buy homes in markets where housing expenses are far outside the norm for both renters and owners. They were able to bid prices up in cities where rents had already been bid up. And, since those households could expand their consumption of housing with these new loans, but the Closed Access cities would not allow the supply of housing to expand with it, it was, ironically, a boom in lending to households with high incomes that triggered the housing migration event that overwhelmed the Contagion cities.

In the Closed Access cities, and really only in the Closed Access cities, during the boom, low tier home prices increased more sharply than high tier prices. This was also understandably taken as a sign that loose credit was the primary cause of rising prices. But, I have found that Price/Rent ratios tend to rise systematically up to a point, in every city, in every type of market, and during that time, the rise in Closed Access city home prices was so extreme that this systematic relationship between price and rent became noticeable for the first time. That systematic effect causes home price appreciation to moderate at very high price levels.

This was the reason low tier prices rose higher than high tier prices in the Closed Access cities, and it appears to have little to do with the effect of credit access to marginal borrowers. This is further confirmed by the fact that an insignificant portion of Closed Access housing is owned by credit-constrained households with low incomes, and those households were flooding out of the Closed Access cities by the hundreds of thousands. It just isn’t plausible that credit to unqualified borrowers with low incomes could have had any measurable effect on Closed Access home prices, let alone cause them to double or more.

As home prices rise, there will naturally be activity among speculators, aggressive lenders, investors, etc. That, combined with recent credit market deregulation and mistaken notions about who were the typical borrowers, led to the moral panic.

What, specifically, did policymakers do in response to that moral panic that they shouldn’t have?

First, the Federal Reserve began to tighten the money supply. They raised their target interest rate from one percent in June 2004 to 5.25 percent in July 2006. By early 2006, nominal GDP growth was declining, and while home prices remained relatively stable, new housing construction started to collapse.

By early 2006, residential investment and home construction were declining sharply. Federal Reserve officials saw this, and the related decline in GDP growth, as good news–a return to normalcy. But, since there weren’t too many homes–in fact, there weren’t enough well-placed homes–this was a very damaging error.

By August 2007, with the Fed Funds Rate still pegged at 5.25 percent, housing starts were into recessionary territory. Because of the moral panic, policymakers explicitly avoided policies that would have stabilized GDP growth, residential construction, or home prices, because they feared that stability would let speculators and reckless lenders off the hook. Pundits and policymakers frequently spoke of allowing panics to develop or prices to collapse in order to impose “discipline” on markets. In hindsight, with a new view of the causes of the bubble, this was off-base, but even by late 2007, it would have been reversible with monetary policy accommodative enough to maintain stable GDP growth.

There was a focus at the time on households who couldn’t afford their mortgages, and anecdotes filled the papers of families who had taken out loans to get through hard times who were now going into default. But, in hindsight, most of the defaults were an after-effect of collapsing prices. And, in fact, what was quietly happening in housing markets at the time was a backdoor exodus of qualified borrowers out of a market they increasingly saw as risky. While the country was focused on disciplining a market they thought was being brought down by subprime borrowers who couldn’t make their payments, the homebuilders had been receiving discipline for quite some time from qualified buyers who were getting cold feet before they even took ownership.

Collapsing housing starts had absorbed as much of the decline in homebuying as it could by the summer of 2007, and from that point on, the collapse was felt more acutely in collapsing prices.

The Federal Reserve quickly lowered the target interest rate to two percent, but then it left it there from May to September 2008, even though economic conditions were deteriorating. Then, in September, even though Fannie and Freddie were taken into conservatorship based on the determination that future defaults would be so high that they would be suffering losses for years to come, and then Lehman Brothers famously filed for bankruptcy, the Fed maintained their two percent rate target. This was, de facto, a tightening of monetary policy, and in fact, the Fed was never even remotely able to achieve a two percent rate level in the chaos that arose after their announcement.

The rate was so far above the rate they needed to target, that FOMC chair Bernanke was afraid that they would eventually have to sell all of their liquid assets (Treasury bills) in order to maintain it. Yet, the target rate remained at two percent while the country debated the various “bailouts” that the Fed and Treasury were proposing to stabilize financial markets. Rather than lowering the rate by buying Treasuries and injecting cash into the economy, the Fed began a policy of paying banks interest on their excess reserves – effectively offering to pay above-market interest rates to borrow cash from the banks to stick in its virtual vaults.

A collapse in financial markets and economic activity followed, along with rising unemployment. Also, the period after the September 2008 meeting was the worst period for rising mortgage delinquencies and collapsing home prices. The Fed responded, finally, by cutting rates to nearly zero in December 2008, followed by several rounds of Quantitative Easing, which finally managed to stabilize the economy.

Open access cities allowed housing supply to more closely match the actual demand people had for it. How did they fare once the recession hit?

Home prices in Open Access cities remained moderate during the boom, reflecting a combination of slightly higher values due to low interest rates which were moderated by declining rents. When housing starts declined in 2006, rent inflation in most cities moved significantly higher. After the recession hit, the main factor in Open Access cities was the newly strict control on lending. After a brief pause in the deepest part of the crisis, rent inflation has moved higher across the country. In low tier housing markets where potential buyers can’t get financing, prices are pushed too low to induce new building, and rents now are rising. In most Open Access cities, mortgage affordability as a percentage of income has been unusually low since the crisis, but rent affordability has risen to new highs.

Have closed access cities opened up since the crisis, or are they still pretty closed? Is there a chance that they could contribute to another economic crisis going forward?

Closed Access cities continue to reach a political limit to new building similar to their pre-crisis levels. So, housing starts in the Closed Access cities have generally recovered back to their pre-crisis levels. Those cities attract rich buyers, so the clampdown on lending hasn’t affected them as much as other cities. Prices have recovered more there, and homebuilding is back to its (low) political limit. In most other cities, building remains well below its pre-crisis level.

We are unlikely to experience a crisis similar to the one that developed in 2008 because the clampdown on lending has kept economic growth and residential investment growth low enough that migration has been more subdued. There remains the base-rate of migration out of those cities of households with lower incomes who are generally renters. But, the effect of restricted lending has kept prices low enough that homeowners in Closed Access cities are not selling and moving away as they did in 2004 and 2005.

Imagine there’s a policymaker in one of these closed access cities reading this right now. What’s the best, simplest thing they can do going forward to make housing in their city more responsive to the needs of local residents?

In general, the pendulum needs to swing back to where existing locals have less power to obstruct change within their cities. This may be something that is more appropriately implemented at a state or federal level. In the end, it is clear that rising costs and out-migration of poor residents is the result of lower rates of building. Building of all kinds will help.

On a more general point, it seems that higher property tax rates on existing properties are correlated with more affordable markets. This makes property less speculative, keeps prices lower, and helps to fund the infrastructure necessary to accommodate growth. Even in a revenue neutral scenario, property taxes are preferable to most other forms of local tax. Of course, maintaining the distribution of those taxes in a fair and equitable way across a metropolitan area is important.

But, more easily and more directly, it is important that we stop trying to solve a supply problem by creating a demand problem. Preventing the custodian in Topeka from buying a modest home doesn’t solve the problem of modest homes in Silicon Valley selling for a million dollars.