It’s odd that President Barack Obama didn’t mention housing in his second inaugural address. After all, he spent his first term in the shadow of a housing meltdown, and remaking federal housing policy remains a central piece of unfinished business.

Let’s hope that Obama tells us in his State of the Union speech what his plans are for Fannie Mae, the Federal National Mortgage Association, and Freddie Mac, the Federal Home Loan Mortgage Corp. The Federal Housing Administration seems headed for a multibillion-dollar bailout.

The government was the housing bubble’s abettor-in-chief, subsidizing borrowing directly with the mortgage-interest deduction and indirectly through easy borrowing policies encouraged by the FHA, Fannie Mae and Freddie Mac. The subsidies must end and be replaced with regulations that limit aggressive real estate lending during booms.

The fragility of the housing market in Obama’s first term made major changes appear imprudent. Now, however, with the latest Standard & Poor’s Case-Shiller Home Price Index showing that housing values have gone up for almost a year, the government should reverse its approach to mortgage markets and stop using the tax code and government-sponsored enterprises to encourage real estate borrowing.

Because even unsubsidized lenders have repeatedly underestimated the downsides of real estate investment, the tax code should limit the interest deduction both in absolute size and as a share of total income. Regulators should require that real estate lenders hold even more capital than now required.

Unavoidable Failure

The U.S. has been a country of real estate speculators ever since King James I gave the Virginia Company property rights. Although real estate bubbles have been ubiquitous, however, they haven’t always led to financial crises. When banks and borrowing are conservative, bubbles come and go with little collateral damage. When the entire lending system depends on rising real estate prices, systemwide failure is unavoidable.

Property bubbles are dangerous and common because there are always reasonable explanations to justify boom-level prices. Economists in 2005 estimated that the cost of owning real estate was only 2.5 percent of the market price of the home in Orange County, California, and 2.4 percent in San Francisco. This low cost reflects expected appreciation and low interest rates. Those meager costs, which reflected low real interest rates and expected capital gains from future price growth, justify price-to-rent ratios of 40, which is what these areas had during the boom. When price growth slows, those price-to-rent ratios no longer seem sensible, prices collapse, and so do banks.

I recently finished a historical survey of U.S. real estate bubbles and found that calculations could similarly justify many boom-level prices seen in the past. During the land boom of 1818 that preceded the bust of 1819, farmers were selling cotton for more than 30 cents a pound. Paying $7, $70 or even $700 for an acre capable of growing 1,000 pounds of cotton annually seemed sensible. Iowa wheat farmers in 1910 could be excused for paying $100 an acre, when those acres produced 18 bushels each and wheat sold for $1 a bushel.

A marvelous 1930 book called “The Skyscraper” wanders through the minutiae of skyscraper building in New York in 1929. With office rents going for $3.50 a square foot, paying $200 a square foot for land (about what the Empire State Building’s developers spent) and building up at $12 a square foot was eminently reasonable.

Justifying Booms

Cross-city comparisons have also been used to justify bubble-level prices. On Feb. 9, 1887, during the flurry of an early real estate boom, the Los Angeles Times compared prices in Los Angeles with those in other cities, such as Cleveland, and concluded that “these figures should convince any one that considering the great natural advantages enjoyed by this city, prices of realty are not in so inflated a condition as is sometimes supposed by the less sanguine among us.”

During the most recent boom, Las Vegas prices seemed reasonable to buyers who were comparing them with prices in Los Angeles, which may explain the geographic spread of the bubble from the coast to nearby interior areas.

The big mistake that runs through real estate history is that buyers and investors forget the 19th-century economist Alfred Marshall’s dictum that “the value of a thing tends in the long run to correspond to its cost of production.”

Alabama and Iowa buyers in 1819 and 1919, respectively, were burned as agricultural supply increased and the prices of cotton and wheat tumbled. The “Skyscraper” builders lost fortunes when office rents fell to far less than $3.50 a square foot, partly because New York had erected enough tall buildings. Las Vegas buyers apparently forgot that there was essentially an unlimited supply of desert land.

Buyers sometimes combine the more excusable error of forgetting the power of supply with wild flights of fancy about future price growth. The economists Karl Case, Robert Shiller and Anne Thompson surveyed buyers in Orange County in 2004, who apparently expected 17.4 percent price appreciation in each of the next 10 years.

Some real estate bubbles, such as the Alabama boom of 1818, ended in financial chaos, while others, such as the Los Angeles boom of the 1880s, didn’t. When lending is limited, a bursting bubble leads to only localized damage from a bit of overbuilding. When borrowing is excessive, real estate crashes can take down the entire financial system.

Yet the government has long supported overleveraging. The mortgage-interest deduction makes it artificially inexpensive to borrow to buy or build bigger houses. To those who say it will never be politically possible to get rid of it, why not start by gradually reducing the current cap on it from $1 million to $300,000, and by limiting the size of the deduction relative to total income? If the deduction was limited to 15 percent or 20 percent of income, this would discourage borrowers from taking on too much housing-related risk.

Effective Signal

Changing the tax code is a far better way to discourage risk taking than the Dodd-Frank Act’s rule that makes banks liable for providing risky loans. I agree with my Bloomberg View colleague Caroline Baum that these rules will be tough to enforce and may have pernicious side effects. Tax reform could send a more direct and effective signal to borrowers.

Yet better banking regulations are needed. Land and housing prices typically go down after they go up. Banks that are heavily invested in real estate during a boom can threaten the system during a bust. Regulators can counter this risk by pegging real estate values not to current market prices but to longer-term average values. This would essentially increase the capital requirements during booms for real-estate-heavy institutions.

Finally, the government should stop offering borrowers an underpriced default option through subsidized mortgage insurers such as Fannie Mae, Freddie Mac and the FHA. We don’t need to shut down Fannie and Freddie, but they should be required to demand high premiums for mortgage insurance.

Promoting homeownership by subsidizing borrowing is as likely to lead to a foreclosure society as an ownership society. The next housing and financial crisis is just around the corner if government doesn’t stop fueling real estate speculation. President Obama has a chance to remake housing policy -- he should take it.

(Edward Glaeser, an economics professor at Harvard University, is a Bloomberg View columnist. He is the author of “Triumph of the City.” The opinions expressed are his own.)

To contact the writer of this article: Edward Glaeser at eglaeser@harvard.edu.

To contact the editor responsible for this article: Katy Roberts at kroberts29@bloomberg.net.