HOME prices in the United States have been falling for nearly three years, and the decline may well continue for some time.

Even the federal government has projected price decreases through 2010. As a baseline, the stress tests recently performed on big banks included a total fall in housing prices of 41 percent from 2006 through 2010. Their “more adverse” forecast projected a drop of 48 percent — suggesting that important housing ratios, like price to rent, and price to construction cost — would fall to their lowest levels in 20 years.

Such long, steady housing price declines seem to defy both common sense and the traditional laws of economics, which assume that people act rationally and that markets are efficient. Why would a sensible person watch the value of his home fall for years, only to sell for a big loss? Why not sell early in the cycle? If people acted as the efficient-market theory says they should, prices would come down right away, not gradually over years, and these cycles would be much shorter.

But something is definitely different about real estate. Long declines do happen with some regularity. And despite the uptick last week in pending home sales and recent improvement in consumer confidence, we still appear to be in a continuing price decline.

There are many historical examples. After the bursting of the Japanese housing bubble in 1991, land prices in Japan’s major cities fell every single year for 15 consecutive years.

Why does this happen? One could easily believe that people are a little slower to sell their homes than, say, their stocks. But years slower?

Several factors can explain the snail-like behavior of the real estate market. An important one is that sales of existing homes are mainly by people who are planning to buy other homes. So even if sellers think that home prices are in decline, most have no reason to hurry because they are not really leaving the market.

Furthermore, few homeowners consider exiting the housing market for purely speculative reasons. First, many owners don’t have a speculator’s sense of urgency. And they don’t like shifting from being owners to renters, a process entailing lifestyle changes that can take years to effect.

Among couples sharing a house, for example, any decision to sell and switch to a rental requires the assent of both partners. Even growing children, who may resent being shifted to another school district and placed in a rental apartment, are likely to have some veto power.

In fact, most decisions to exit the market in favor of renting are not market-timing moves. Instead, they reflect the growing pressures of economic necessity. This may involve foreclosure or just difficulty paying bills, or gradual changes in opinion about how to live in an economic downturn.

This dynamic helps to explain why, at a time of high unemployment, declines in home prices may be long-lasting and predictable.

Imagine a young couple now renting an apartment. A few years ago, they were toying with the idea of buying a house, but seeing unemployment all around them and the turmoil in the housing market, they have changed their thinking: they have decided to remain renters. They may not revisit that decision for some years. It is settled in their minds for now.

On the other hand, an elderly couple who during the boom were holding out against selling their home and moving to a continuing-care retirement community have decided that it’s finally the time to do so. It may take them a year or two to sort through a lifetime of belongings and prepare for the move, but they may never revisit their decision again.

As a result, we will have a seller and no buyer, and there will be that much less demand relative to supply — and one more reason that prices may continue to fall, or stagnate, in 2010 or 2011.

All of these people could be made to change their plans if a sharp improvement in the economy got their attention. The young couple could change their minds and decide to buy next year, and the elderly couple could decide to further postpone their selling. That would leave us with a buyer and no seller, providing an upward kick to the market price.

For this reason, not all economists agree that home price declines are really predictable. Ray Fair, my colleague at Yale, for one, warns that any trend up or down may suddenly be reversed if there is an economic “regime change” — a shift big enough to make people change their thinking.

But market changes that big don’t occur every day. And when they do, there is a coordination problem: people won’t all change their views about homeownership at once. Some will focus on recent price declines, which may seem to belie any improvement in the economy, reinforcing negative attitudes about the housing market.

Even if there is a quick end to the recession, the housing market’s poor performance may linger. After the last home price boom, which ended about the time of the 1990-91 recession, home prices did not start moving upward, even incrementally, until 1997.

Robert J. Shiller is professor of economics and finance at Yale and co-founder and chief economist of MacroMarkets LLC.