Evan Soltas
Apr 3, 2015

The Madre of All Bubbles

Why is Spain in the terrible position it is? A quarter of its workforce is unemployed. It is producing much less than it did in 2008. Tens of thousands are protesting in the streets against the status-quo political order.

One explanation might be fiscal recklessness. Yet this explanation simply doesn't work. Spain's government had run large primary budget surpluses for the decade leading up to the crisis. And, if you look at the so-called "structural" budget deficit -- which takes into account the effects of booms and busts on the government's spending and tax revenue -- Madrid was still running a surplus.

Another explanation is that the eurozone is broken. A common currency without a fiscal union or a great deal of labor mobility is a recipe for disaster, because whenever one country's economy diverges from the rest, there is no way to throw it a life-raft -- you can't cut interest rates just for it, the country's government can't run large budget deficits without causing a financial crisis, and the country's workers can't all pick up and move.

There's a lot to that explanation. But the most important one, I think, is debt. Lots and lots of household debt.

It's well known that Spain experienced the madre of all debt-fueled housing bubbles from 1995 to 2008. For comparison, it was vastly larger than the American housing bubble: Real home prices more than tripled in Spain, whereas they doubled in the U.S. (You can see so for yourself on The Economist's neat house-price tool.)

When prices came crashing down, Spanish households were left with enormous mortgage debts. And their response was to slash their consumption spending, plunging the Spanish economy into depression. (For more on this explanation, see Atif Mian and Amir Sufi's book, House of Debt.) Add to that the fact that, in Spain, mortgage debts are "full-recourse" -- which means that, even if you default on your mortgage and walk away, the bank can still come after you for the rest of the money.

Had Spain avoided the debt-fueled housing bubble, the design flaws of the eurozone would have been a much smaller problem. And the bubble would have been a problem, euro or no euro -- look at the experiences of the United States or Japan after their respective property bubbles.

In fact, in some research I've been working on, I find that all of the Spanish boom and bust in home prices is explained by mortgage debt. Had real mortgage debts just grown at a constant rate, instead of ballooning as they did in the 2000s, home prices would have been 30 percentage points lower at the peak of the bubble.

How can we say that? Well, Spain is broken up into fifty different provinces. In some of those provinces, mortgage debt soared. In other provinces, it didn't. By comparing the rise of home prices in provinces where debts are exploding and provinces where debts are prudent, we can estimate the effect of mortgage debt on home prices.

I've found quarterly province-level data on home prices (measured in euros per square-meter) and mortgage-debt issuance from the Spanish government's Ministry of Housing and National Institute of Statistics from 1995 to present.

For every 10 percentage point increase in monthly mortgage-debt issuance in a given province, home prices increase about 1.5 percent. The graph below displays that result, using the last few quarters of debt growth and fixed effects for quarter and province.

If we interpret that as causal -- more on that in a moment -- then we can back out what would have happened to Spanish home prices had not mortgage debt not boomed. Specifically, the counterfactual scenario is: What if province-level mortgage debt issuance only grown at the rate of inflation? Here it is.

What you should notice is that the rise in real home prices is much slower. By the peak of the bubble, I find a 30-percentage-point gap between bubble and the no-bubble home price. And the pullback of mortgage-debt issuance explains all of the massive decline in Spanish home prices. Real home prices would have never declined without the debt bubble.

Also, the increase in home prices in the counterfactual is fully explained by the increase in real Spanish GDP -- that is, richer Spaniards spending more on homes. That's why I say that the expansion of mortgage debt explains all of the Spanish housing bubble.

What about the idea that home prices drove up mortgage debt? (That is, people had to take out bigger ones to cover the higher home price.) Atif Mian and Amir Sufi investigated that possibility in the U.S. and contradict that idea. Debt still rose in cities that were so flat and well-zoned that construction could keep home prices down. That suggests debt causes home prices to rise and not the other way.

This is probably just the beginning of my work on the Spanish housing bubble. It may be the topic of my senior-year thesis. (Yes, the recent posts on rents and the Internet were me brainstorming.) We'll see.

Update (4/14): It seems as though this will be my senior thesis. Atif Mian will be my advisor.