With all the talk of excess inventory and flood of foreclosures, the idea of a looming housing shortage sounds unrealistic if not fanciful.

After all, the most recent data from the National Association of Realtors (NAR) out last week showed a 5.1% decline in existing home sales in June. Meanwhile, total housing inventory increased 2.5% to four million homes available for sale, an 8.9-month supply, up from an 8.3-month supply in May.

Foreclosures, too, are an issue with a vast backlog of distressed properties and underwater loans sitting just below the surface, according to RealtyTrac, an online foreclosure marketplace. The company forecasts that more than three million properties will get hit with foreclosure filings by the end of the year.

But if you take a step back from the current doom and gloom of foreclosures and declining sales and focus on the low construction levels over the past few years, some economists say a housing shortage might be in the offing. A 2009 report by Massachusetts Institute of Technology economics professor William Wheaton says that despite the glut of existing homes, with current depressed levels of construction, there might be excess demand for newly constructed homes.

In the past seven years, housing starts first exceeded but then fell short of the historical norm of 1.6 million, according to the NAR, with a deficit forecasted to grow into 2011. The chief economist of NAR said last month that the big drop in home construction suggests a shortage could become an issue later.

Longer-term demographics support this theory, says Ross DeVol, executive director of economic research at the Milken Institute, an independent think tank based in Santa Monica, Calif. We re only adding about 600,000 new housing units a year now, and the long-term growth in new households is 1.3 million to 1.4 million per year, says DeVol.

That household formation rate has fallen off somewhat because of the recession. But that decline is misleading because college graduates have chosen to live at home with their parents while they find their financial footing, and people defer getting married for a year or two. But long term, that household growth says that if we build substantially less than that amount, which we re doing now, in four, five or six years, if we don t ramp up housing starts, we could see a shortage, DeVol says.

There s a tendency in any market that when you overshoot on the upside which we did through 2007 in real estate you undershoot on the downside, DeVol says. But underlying growth in population demographics namely, how many people will be entering the work force is somewhere in that range of 1.3 million to 1.4 million, he says. One risk is that so many home builders leave the field during the current downturn that there could be capacity constraints in the long term as the U.S. population continues to grow, says John Vogel, professor of real estate at the Tuck School of Business at Dartmouth.

Consider that at the peak of housing bubble in 2005 nearly 2.1 million new units were built. In 2006, that number dropped to 1.81 million; in 2007, as the bubble deflated, new units fell to 1.34 million. By 2009, only 550,000 new units were built, says DeVol.

There won t likely be constraints in overbuilt places like Las Vegas, Phoenix, Riverside, Calif., or Miami and Ft. Lauderdale. But if the pace of home construction doesn t pick up, we are going to begin to see some tightness in some areas of the country that didn t have the boom and bust occur, DeVol says.

The regions most likely to be undersupplied by mid-2012 are those where supply and demand are now in balance, says Celia Chen, senior director of housing economics at Moody s. Chen includes states like Washington, Oregon, New Mexico and Utah in this group. This is where strengthening demand, combined with construction that will remain below trend, would result in undersupply, she says.