One puzzle these days is why Americans are so confident at the shopping mall and so glum in opinion polls. By many measures, the country's prosperity is broad-based. Families are buying and renovating homes at a ferocious pace. Sales of existing homes in 2005 are expected to reach a record 7.1 million units. Since mid-2003 the number of payroll jobs has increased by 4.2 million. The unemployment rate of 5 percent is low by historical standards. But in polls, Americans are downbeat. The University of Michigan's index of consumer confidence was 74.2 in October, a big drop from 96.5 in July. The three-month decline is the second-largest on record (the first occurred around the 1990 recession).

This and other surveys could signal an economic slowdown or recession. There are obvious grounds for anxiety. In October new car and truck sales plunged 14 percent. Although gasoline prices are falling, they're still higher than a year ago. Homeowners will also face bigger winter heating bills, reflecting higher energy costs. Economist Marc Levinson of J.P. Morgan expects average households to pay $900 to $1,000 more in the Northeast and $700 to $800 more in the Midwest than last year. The real estate boom could recede or even implode. Because many homeowners are borrowing against rising housing prices and spending the extra cash, that could hurt ordinary shopping. Should a recession actually occur, of course, the gap between today's strong economy and sour public opinion would disappear.

But until then, I have another theory to explain what's been a persisting disconnect between our mood and our behavior: the hangover from the 1990s boom. We subconsciously compare everything now with what happened then; and the comparison favors the past and disparages the present. Almost nothing looks as good as it did then. We were marching toward a carefree future. The Internet was everything -- and American companies dominated the Internet; the business cycle was dead or dying; interest rates and inflation were low; stock prices would rise forever; budget deficits were disappearing; and unemployment was low. The powerful U.S. economy could subdue almost any threat (say, the 1997-98 Asian financial crisis).

Not coincidentally, the Michigan confidence numbers reached unprecedented levels in the late 1990s; the historic peak occurred in January 2000 at 112. It wasn't simply that the economy did well. What was distinctive is that it did so well that it suggested we could take its future for granted. We called it the New Economy, which implied that the rules of the game had changed. There were explanations for all this bliss: new technologies; adoption of just-in-time inventory practices; the revival of entrepreneurship. These arguments were satisfying; they were also superficial. Alfred E. Neuman had become our chief economic guru: What, me worry?

The central fantasy was that we could dispense with uncertainty and anxiety. Now they've reasserted themselves with a vengeance. We fret about China, a housing "bubble" (remembering the stock and tech "bubbles"), huge trade and budget deficits, oil -- as well as terrorism, Iraq and possible pandemics. The return of worry partly accounts for the weakness of consumer-sentiment polls. People are less confident about the future. But what then explains the strength of actual consumer spending? The answer is that Americans' personal spending decisions depend less on their general view of the economy and more on their personal circumstances -- and these haven't shifted so dramatically.

Although our mood went on a roller coaster, changes in our well-being (income, wealth) were less erratic. In the late 1990s, some Americans did fabulously; but most simply did well. Since 2000 many Americans have done poorly; but most (with jobs, solid incomes and refinanced homes) still did well. In many ways, the economy is stronger now than it was then. Here are two examples. First, household net worth -- what people own minus what they owe -- is about $50 trillion, up from $42 trillion in 1999; gains from homes have more than offset losses on stocks. Second, per capita incomes (after inflation) grew almost 9 percent from 1999 to 2004. Living standards haven't stagnated.

We have a real economy and a rhetorical economy: what's actually happening and what we say is happening. The first is often more stable than the second. I wouldn't hang too much on this distinction. It certainly doesn't preclude an economic slowdown or even a recession. Levinson thinks consumer spending will weaken; so does Richard Curtin, director of consumer surveys at the University of Michigan at Ann Arbor. The recent declines in confidence reflect real events (hurricanes, higher oil prices, rising interest rates). But I would contend that the distinction helps explain the resilience of American consumers in the past five years. They've been unexpectedly resilient in part because they've never been quite as desperate as common wisdom -- buzz, chatter, commentary -- holds.