If this be a recession, it is not like any we've seen.

Its epicenter lies closer to New York than Indianapolis.

Its onset was announced not by a buildup of steel and clothing in warehouses and loading docks, but by a buildup of office buildings with nobody to rent them.

And its severity is likely to be measured not simply by the length of the unemployment line, but also by the sharp drop in prices for stocks, junk bonds and real estate.

Until late this summer, most economists had been reluctant converts to the recession scenario. Their optimism finally gave way after a doubling of oil prices stoked inflation, sent stock markets reeling and sapped consumer confidence. The consensus view now calls for a mild "supply shock" recession that will generate its own recovery by 1992.

"The slump should be short and shallow," predicted Jerry Jordan, chief economist at First Interstate Bancorp in Los Angeles, speaking for most of his professional brethren.

Or will it? While most economists continue to view the current downturn in the context of the past 50 years of economic data stored in their computer models, others question if they are not like the infamous generals who have outfitted themselves splendidly to fight the last war. Is it not logical to question, they say, why a U.S. economy that has shifted so dramatically from blue-collar to white, that has taken on unprecedented levels of debt and that has been dragged kicking and screaming into a newly competitive and interrelated world economy -- why this economy would behave just like it did before?

"The faster the world is changing, the less accurate the computer models are likely to be," argued Lester Thurow, dean of the Sloan School of Management at the Massachusetts Institute of Technology and something of an renegade among academic economists. "And the world is now changing very quickly."

Unlike many past recessions, this downturn is not being molded on the factory floor.

Although the economy as a whole has lost as many as 600,000 manufacturing jobs since January 1989, it is merely an acceleration of a process that began 10 years ago. Manufacturing employment has dropped from 22.3 percent of the work force in 1980 to 17.6 percent today. Few doubt it will continue its long-term employment decline. But what is most notable about the last year or so is how those losses are no longer offset by gains in "the service sector" -- retail stores, restaurants, hospitals and medical labs, TV and movie production companies, travel agencies, software developers, banks and insurance companies, to name a few.

Consider the states at the cutting edge of recession. Among the nine states that reported virtually no growth in jobs or actual year-over-year job losses in August, seven were clustered in the Northeast: all six New England states, plus New York (Michigan and the District of Columbia were also on the list). These were among the places where jobs and income grew faster than the nation as a whole during the 1980s, largely on the strength of a real estate boom and spectacular growth in service employment.

Even as those economies began to unwind three years ago, with the cooling of the construction and real estate markets, unemployment rates remained remarkably low. Only when the service industries began to decline did the unemployment rates begin to move up -- and, in many cases, the reversals were stunning and dramatic.

Connecticut, for example, which had been adding 10,000 jobs a year in insurance and other financial services, by this summer was adding none. Retail and wholesale trade, used to 15,000-a-year job growth, actually began shrinking. Meanwhile, construction and manufacturing, which had begun to slide much earlier, were each down roughly 9,000 jobs.

In New York State, the retail and wholesale sectors had been adding 38,000 jobs a year on average since 1982. This year: down 30,000. Financial services, good for 18,000 new jobs each year during the '80s, were down 12,000.

The slowdown looks a lot different in the states of the Rust Belt. In Ohio and Indiana, for example, construction employment was higher this August than last, while manufacturing jobs were down only fractionally. Meanwhile, service employment was still growing -- 80,000 in Ohio, 40,000 in Indiana.

Based largely on such state-by-state jobs numbers, the economic forecasting firm DRI/McGraw-Hill now calculates that a dozen or more states are already in a recession; economist Allen Sinai of the Boston Co., a unit of American Express Co., said it's closer to two dozen. But in almost every state at or over the edge, softness in the service sector has tipped the balance toward recession.

"We've had a long period of decline in manufacturing and construction," explains Sinai, "but what distinguishes the last few months is that the service sector is showing up weak and losing momentum... . Considering how large the service sector has become, I don't think we can really have a full-fledged recession until service industries begin to slide."

This notion that the service economy might actually lead a recession and operate relatively independently of the goods-producing sectors (manufacturing and construction) is still somewhat foreign to many economists. Traditional theories about the business cycle discount it. And the data on which economists rely -- reports from purchasing managers, figures on inventory buildups, capacity utilization and the length of the average workweek, the performance of the Dow Jones industrial average, construction starts, the rise and fall of interest rates -- shed little light on the performance of service companies that account for two-thirds of the country's gross national product.

"We don't have a set of accounts to measure the service economy very well," admits Sinai, who has been studying it more closely for the Washington-based Coalition of Service Industries as well as his own employer, service-sector giant American Express. "And analytically, we as economists haven't come to grips with it yet."

There is a growing body of evidence, however -- some of it statistical, much of it admittedly anecdotal -- that the rhythms and cycles by which the economy grows and contracts do not always track those of the industrial sector.

Commerce Department data reveal that service companies rely on manufacturing and construction for only about a quarter of their sales, while relying on other service companies for roughly half. Put another way, if you're wondering who all those accountants and lawyers and marketing specialists and consultants and computer programmers are doing business with, as likely as not it is with each other. That helps explain why the recession appears to have "spread" from New England along the Amtrak line as far south as Washington -- and why it is hasn't "spread" to the Midwest or Southeast, where the service sector is less predominant, or parts of the Pacific West.

Obviously, there are a number of service industries that are closely tied to manufacturing -- railroad companies that move finished goods and retail outlets that sell them. But these days it also works the other way around. Among the fastest growing manufacturing industries are some that depend heavily on service-sector customers: aircraft manufacturers that rely on the airlines; medical equipment makers that sell to hospitals and labs; computer makers, whose largest customers include all those banks and insurance companies. Government data confirm that the percentage of business investment in plant and equipment made by the service sector has been increasing steadily.

The composition of the unemployment line today also bears out the general impression that this is more a white-collar recession than we've seen before. Back in September 1982, at the height of the last recession, roughly half of those listed as without jobs were classified as white-collar employees or service industry workers. In September 1990, that figure was 60 percent. By contrast, only 23 percent of the unemployed today fall into the category generally thought of as assembly line and factory workers; eight years ago, it was 34 percent.

Economist Robert Gordon at Northwestern University suggests one additional reason recessions may be less driven by ups and downs of the manufacturing cycle -- namely, those ups and downs are becoming less severe. The reason: overseas sales that remain strong even when U.S. sales are weak. Because of a declining dollar and a decade of factory modernization that has made U.S. products more competitive, estimates are that exports now account for 11.2 percent of U.S. manufacturing shipments, up from 8 percent during the high dollar days of 1983.

"To the extent that manufacturing these days is protected by {rising exports} and the economy relies more on the behavior of the consumer, we could have what you might call a service-sector recession," said Gordon.

"The possibility of a service-sector recession cannot be precluded," agreed Victor Zanowitz, a professor emeritus at the University of Chicago Business School and one of the leading scholars on the business cycle. But he quickly added, "I don't think we have seen one yet."

If there were a man-on-the-street explanation for the current economic situation, it might go something like this: The U.S. economy came to be dominated during the 1980s by financial manipulators who made huge fortunes by using borrowed money to run up the price of stocks and real estate so high that it eventually had to come crashing down, bringing the savings and loans and the banks and maybe even the whole economy down with it.

By the end of the decade, public and private debt in the United States rose to the highest point since the Great Depression: 1.9 times the gross national product versus 1.4 times at the beginning of the decade. Nearly half of all the working capital financing American business was borrowed from bondholders and banks, up from one-third 10 years ago. The debt held by the average American household has reached 88 percent of after-tax income, up from 72 percent at the beginning of the decade.

Much of this debt was used to pay for a buying spree that helped fuel the economy of the 1980s, in the process driving up the prices of businesses (stocks and junk bonds) and real estate. Now, as all can see, the bubble has burst.

Nationwide, the price of the average single-family home rose less than 1 percent per year after a decade of spectacular double-digit increases, and prices are actually falling in the Northeast and West. Harder hit has been the value of commercial real estate: According to one widely used index, the value of office buildings in the last three years has fallen 15 percent nationwide. Stocks are off 20 percent from their highs this past summer. And junk bonds -- bonds that have become the moral equivalent of stocks because they are so risky -- as a group have lost 35 percent of their original value, most of it in the last year.

Most forecasters are at something of a loss to say precisely what this steep decline in asset values means for what is called revealingly, the "real" economy. Few would say that the run-up in debt and asset prices actually caused or triggered the current downturn. And while a number of analysts say that the high level of debt and the rapid fall of asset prices could make a recession longer and deeper, they admit they have no good way to quantify that notion or incorporate it into their computer models. That is one reason that most are standing by their prediction of a recession that is short and shallow.

Not Gary Shilling, who as far back as 1987 was predicting a sharp decline in asset prices and serious recession. Shilling said the problem with most economists is that they keep looking for signs of an "income statement recession" -- one in which profits and sales dip for a few months -- when they should be tracking the coming "balance sheet recession" in which investors and lenders simply get wiped out, companies fall into bankruptcy and banking is strained to its limits.

"We don't have any experience with this kind of recession {since the Great Depression} -- that's why those guys with the computer models can't pick it up," Shilling said.

"Maybe it will start when a money market fund -- which most people think is like a bank account but isn't -- actually declines in value because some Japanese bank can't roll over its commercial paper. And then a couple of big LBOs go belly up just as a few more Third World countries forfeit on their loans and the {bank insurance fund} finds itself in need of a bailout. People are going to say, 'Good grief, when is this all going to end?' Confidence just evaporates -- and then watch out."

Just last week, economist Henry Kaufman warned that it would be folly to view the recession that he believes is now underway as just a garden variety version because of the high levels of debt and the extreme "fragility" of the U.S. financial system. In other words: Don't count on short and shallow.

While Shilling and Kaufman are viewed as doomsayers by many economists, more mainstream forecasters admit they haven't factored any sort of financial sector shock into their computer models or published economic projections.

Earlier this year, the Brookings Institution published a paper by economists Ben Bernanke and John Campbell, whose computer model showed that corporate debt payments now consume such a large share of corporate cash flow that even a moderate recession could send 20 percent to 25 percent of big U.S. manufacturing firms into insolvency.

While not predicting that will happen, Bernanke, a professor at Princeton University, said the high level of debt and the inflated values at which some assets are being carried on company books will certainly make a recession shock "significantly worse."

One way that the problems of high debt and declining asset values are transmitted to the rest of the economy is through the banking system, and clearly that has already begun to happen.

Largely as a result of falling real estate values and tougher accounting rules imposed by federal regulators, the value of problem loans and foreclosed property on the books of the nation's banks and thrifts has gone from about $60 billion at the end of 1985 to roughly $150 billion today (including the $45 billion or so of those loans now in the hands of the government).

To deal with the crisis, many banks have had to trim staff and reduce the amount of lending they do -- not just to real estate developers, but increasingly to other kinds of businesses that are reporting difficulty in securing what used to be routine loan approvals. The Federal Reserve recently reported that overall bank lending had declined $5 billion so far this year, to $315 billion, and Chairman Alan Greenspan has said that the central bank is worried that a credit crunch may be developing that could trigger a recession.

But declining asset values also can work their way into the "real" economy by way of the consumer. "For the first time since anyone can remember, housing values are falling," said MIT's Thurow, a reversal that "strikes right at the middle class, which, in effect, has used the family home as a savings account all these years."

Geoffrey Moore, who heads the Center for International Business Cycle Research at Columbia Business School and is among the most respected students of business cycles, conceded it is difficult to predict how the problems of the financial sector will work their way through to the real economy.

"There comes a time when the speculative excesses have to be corrected by a recession or even a depression," he said. "It happened that way in 1929, of course, and during several of the great panics of the 19th century, and it could happen this time, too.

"But right now," he was careful to add, "I'd have to say it looks more like an average type of downturn."