Many of those who claim to be financial experts seem oblivious to the impact that a sea change in attitudes can have on the economy.

Yet the connection seems pretty straightforward: when people are confident about the future, they are willing to spend money and take on risks, such as a loan for a new house or car; when they are anxious and uncertain, they tend to postpone large purchases, reduce debt, and save more for a rainy day.

While it isn't always easy deciphering how most people feel, a welter of recent reports makes it clear that sentiment is moving in the wrong direction as far as the U.S. economy is concerned. In "Sniffles That Precede a Recession," economist and author Robert Shiller offers us his insights on this shifting social mood.

A recession has much the same pattern as the flu — starting with vague feelings of malaise and quickly building in misery until a patient’s activities are drastically curtailed. Then, all too gradually, comes an extended period of recovery, accompanied by lingering symptoms of discomfort.

With the unemployment rate up to 4.7 percent in September from 4.4 percent in March, the economy is feeling a chill. Is it descending into recession?

Most economists seem to be concluding that the current unpleasantness is a false alarm. They point to some good vital signs: the stock market is up, the dollar is cheap, the rest of the world is strong and the Fed is ready to respond.

But there are worrisome symptoms, and they bear close watching. The most important is a creeping sense of malaise that could turn into a general loss of confidence. The downturn in the housing market and the repercussions in financial markets are critical factors.

There have been only two domestic recessions in the last quarter-century — both of them also global recessions. According to the National Bureau of Economic Research dating committee, the first began in July 1990, the second in March 2001.

There were familiar warning signs for both of them — an initial sharp rise in unemployment, followed by slower increases that continued for a couple of years. In each case, as often happens with recessions, there was no agreement that a recession was under way until months after it started.

Diagnosis of a recession is hard because no single virus causes it. Instead, a recession seems to be a result of a confluence of many hard-to-measure factors. A decline in investment spending is typically one of them, and a recession is generally one of those rare events when residential and nonresidential investment both happen to decline together.

In some respects, the current situation looks a lot like the period leading up to the 1990 recession. We were coming out of a housing boom then, and the economy was emerging from an associated lending crisis — the savings-and-loan debacle. Now we are dealing with the subprime mortgage “crisis,” but so far, we have not seen the decline in nonresidential investment that occurred in 1990.

There are also some similarities to the 2001 recession, which likewise followed a huge speculative boom. The bursting of the Internet bubble brought a huge decline in corporate investment, and the 2001 recession helped to cleanse investors of their exaggerated hopes for the stock market, particularly for technology and the dot-coms. A similar cleansing of thinking appears under way regarding the housing market. But residential investment is not as big a component of gross domestic product as nonresidential investment; the decline in the housing market has apparently not yet been enough to push us into recession territory.

Consumer confidence indexes have not yet fallen as they did at the onset of the last two recessions. But confidence is a delicate psychological state, not easily quantified. It is related to the stories that people are talking about at the moment, narratives that put emotional color into otherwise dry economic statistics.

In August 1990, for example, a series of events in the Persian Gulf severely damaged business confidence, and that sequence seems to explain the timing of the 1990 recession. Saddam Hussein started his surprise invasion of Kuwait on Aug. 2, 1990, and the United States began sending jet planes to Saudi Arabia shortly thereafter; the Gulf War abruptly became a virtual certainty. Mr. Hussein asked Muslims around the world to join in a jihad against the forces opposing him. In the United States, people started canceling business trips. August was also the month when intense public conversation began about the economy’s weakness. In a sense, that was when the recession started, not the July date given by the bureau committee.

It is clear that salient, emotion-arousing narratives — those that capture the popular imagination and damage public confidence — are central to the etiology of recessions. As these stories gain currency, they impel people to curtail their spending, both in business and their personal lives.

Is this happening now? A disturbing narrative began to unfold in the last couple of months. People began talking of failed institutions — of the possibility that savings socked away in a money market account might actually be invested in subprime loans and so be lost. There has been fear of locked credit markets, of possible bank failures and runs on banks.

Some of these tales have faded — bank runs no longer seem a risk. But confidence in the economy remains fragile. More shocks are likely as an era of huge real estate speculation apparently ends, with the possibility of further surges in foreclosures and failures of financial institutions.

The narrative is still unfolding, and the extent of its virulence is not yet known.

Although I agree with Professor Shiller insofar as the link between a worsening social mood and its ultimately deleterious effect on spending, I would take issue with at least one of his assertions.

In my view, we've not seen anything yet as far as bank runs are concerned. Reports this weekend that numerous lenders are in behind-the-scenes talks to try and orchestrate some sort of coordinated bailout of bank-affiliated investment vehicles holding billions of dollars of unmarketable mortgage-backed securities suggest there is serious trouble afoot at one or more financial institutions.

If so, it won't be long before more than a few depositors and creditors at the most vulnerable institutions decide they would rather cut-and-run than hang in there and hope things work out OK. Once others catch on to what is happening, it won't be long before customers are lining up at the branches of at-risk banks all over town.