What the economy needs is a little inflation. Instead, prices in the U.S. seem to be headed still lower over the next few quarters. And that raises the troubling potential of actual, Japanese-style deflation and economic stagnation.

That's my summary of the just-released minutes from the Sept. 24, 2002, meeting of the

Federal Reserve's

Federal Open Market Committee, the body that decides to raise or lower short-term interest rates. It cut rates 50 basis points one week ago.

But while the Federal Reserve is worried about the potential for deflation across the entire U.S. economy, many individual companies are already coping with actual deflation. Mix this "local" deflation with high levels of corporate debt and you create potential time bombs ready to blow up on investors.

Economywide, deflation may or may not arrive. But local deflation is already very much with us, and investors can see it at work in companies as varied as McDonald's, General Motors and Cisco Systems. Consider the following three industry examples.

Fast-food Deflation

Anybody who's been to a

McDonald's

(MCD) - Get Report

or a Burger King lately knows what I'm talking about. In September, Burger King, a unit of

Diageo

(DEO) - Get Report

, started a price war when it began selling 11 menu items, including its bacon cheeseburger, for 99 cents each. McDonald's responded almost immediately by creating its Dollar Menu of eight items. It also upped the stakes by including two of its heftiest sandwiches, the Big 'N Tasty and the McChicken, which had previously sold for $1.99.

You want to argue that this is evidence of a coming deflationary cycle, be my guest, but I see two tired companies that have been unable to differentiate their products engaged in a price war because they can't think of anything else to do to revive sales. Even cutting prices in half doesn't seem to have done much for demand. McDonald's same-store sales climbed a whopping (pun intended, please) 1.3% in October after dropping 2.8% in the September quarter.

But the real problem for McDonald's is a saturated market for a product that has remained essentially unchanged for years -- and a management that insists it can fix things by tinkering with the menu and trimming a few underperforming restaurants.

The Sinking Feeling in Detroit

There wouldn't seem to be much similarity between making cars and flipping burgers, but the two industries share a basic business problem. You've got to pay for factories and burger stores, pay workers and meet debt payments, even if nobody is in line for a Whopper or kicking tires in the auto showroom.

This explains why

General Motors

(GM) - Get Report

and rival car makers started to offer zero-money-down, 0% loans and zero-payments-for-90-days deals as soon as 0% financing offers showed signs of fading appeal. Even these incentives might not be enough. Sales of cars and light trucks fell 27% in November from a year ago. Granted that sales in October 2001 set an all-time record; still, the trend isn't reassuring.

The automakers face a one-two punch: Deflation in their industry means they can't raise prices (the average selling price of a GM car is down 2% from 2001), and a lack of inflation increases the burden of the liabilities they carry.

Take GM, for example. The company has 460,000 retirees -- about 2.5 for every active worker -- all drawing pensions and health care benefits, whether the company sells any cars or not. GM pension liabilities now stand at about $75 billion, and at the end of 2001, the company's pension plan was underfunded by about $9 billion. So far this year, GM has put about $2.2 billion into the plan.

But that pension obligation grows with every turn of the deflationary screw. At an 8% return, the company will wind up having to pony up another $13 billion through 2007, estimates UBS Warburg. If rates of return stay roughly where they are now -- with 10-year Treasury notes yielding 3.5% -- then GM would have to put as much as $17 billion into its pension plan by 2007. And that would essentially eat up all the company's cash flow during the period.

That's why Detroit's back-door effort to raise prices this year is so critical. While the automakers are offering their zero deals, they've also been raising prices on the most popular models of cars and trucks. The Ford F-150 pickup sold for an average of $27,079 in September 2002, up from $26,308 in September 2001.

Technology Eats its Own

I remember way back in 2001 when one of the reasons to buy chipmaker

Xilinx

(XLNX) - Get Report

was that its programmable logic chips were being increasingly incorporated into products by electronics makers who wanted to cut costs by getting new products to market faster. The logic then was that companies could simply reprogram chips rather than take the time to design and then contract for a new custom chip. Xilinx and competitor

Altera

(ALTR) - Get Report

were selling more and more chips to

Cisco Systems

(CSCO) - Get Report

, for example.

Well, in this technology market, costs can never be low enough, and on Nov. 6 Cisco told Wall Street analysts that it would be increasingly using exactly those custom chips (ASICs) in its product to get costs down even further. It seems that custom chip makers have succeed in cramming enough functions on a single chip so that they've leapfrogged the time-to-market advantages of programmable chips. That's not a minor shift, given that Cisco Systems uses $130 million of programmable logic chips each year, according to estimates by Wachovia Securities.

Is this deflation, though, or just business as usual in the technology industry? Falling prices have been the historical norm -- with improved manufacturing efficiencies enabling companies to make money at lower prices, and a constant flood of new products providing a supply of high-margin new business.

But there are troubling trends that support an argument that something fundamental has changed in the technology industry. New products do seem to have shorter lives as high-margin goods before competitors leap in to send margins falling. The flat-screen controller market and

Genesis Microchip

(GNSS)

are a case in point. Genesis didn't have much time as the dominant provider before Asian competitors started to drive down prices. And we're witnessing the proliferation of a new kind of business in Asia that specializes not just in commodity production of another company's design but in creating that design, too.

Why the Big Deal?

Why is falling inflation such a big problem to the Federal Reserve? Most of us who can remember the high inflation of the 1970s and 1980s reflexively think that inflation is a bad thing.

The Federal Reserve's thinking goes like this, in the governors' own words: "Indeed, the members did not rule out the emergence of appreciably lower inflation. In this regard, some observed that a significant decline in inflation from current levels could imply an unwelcome tightening of monetary policy in real terms. In addition, further sizable disinflation that resulted in a nominal inflation rate near zero could create problems for the implementation of monetary policy through conventional means in the event of an adverse shock to the economy that called for negative real policy interest rates."

This fear of further declines in an already-low inflation rate is a likely explanation of the Federal Reserve's stronger-than-expected move last week. Current levels of inflation give the central bank room to use interest rate cuts to get the economy moving. With inflation at an annual 2.2% in September (as measured by the Core Consumer Price Index, anyway), any bank that borrows from the Federal Reserve at 1.25% is getting money for less than nothing; the future dollars it has to pay back are worth 2.2% less, but the bank is paying only 1.25% to secure them.

The real interest rate that a bank pays to borrow from the Federal Reserve at this point is actually a negative 0.95%. With those kinds of numbers, banks have a real incentive to borrow money, which increases the money supply, makes credit easier to get, and, in general, greases the wheels of the economy so that it can run faster.

But if inflation sinks further, interest rate cuts increasingly lose their power. If inflation falls to 0.5%, for example, and the fed funds rate stays at 1.25%, banks are borrowing at a real interest rate of 0.75% instead of -0.95%. They've lost 1.7 percentage points of incentive to expand their borrowing.

Inflation Math

And if inflation sinks to 0, then the real rate is the same as the 1.25% charged by the Federal Reserve.

The real trouble happens, though, as the Federal Reserve or any other central bank cuts interest rates to 0% in a 0% inflation environment. As the Japanese banks have proved beyond a reasonable doubt, in this situation there's no incentive to increase borrowing from the central bank or to increase lending. At that point, the central bank's monetary tools have lost all power to move the economy.

Does all this mean that the Federal Reserve knows that an economywide deflation is coming? I don't think so. The Federal Reserve certainly sees economywide deflation as a possibility -- a remote one, in fact, if we can believe the public statements of Federal Reserve members.

In my opinion, the Federal Reserve's actions are founded not on a belief that deflation is coming, but out of a conviction that this is the time to exert maximum leverage. There are two parts to that conviction. First, there's the Federal Reserve's study of the Japanese economic quagmire, concluding that the big mistake was waiting to cut rates until the deflationary momentum was fully established. It's almost impossible to stop deflation once the trend is in place, so it's best to act early and boldly, the Federal Reserve has concluded. It's better to err on the side of excessive cutting than to let deflation get established.

Second, by cutting rates now and maybe again early in 2003, the Federal Reserve gets to make maximum use of whatever inflation still lingers in the economy. Why wait until inflation falls even lower to cut interest rates since that decreases the leverage of the cuts that come from a big spread between real and nominal interest rates? The ability to create negative real interest rates is a powerful tool that can be put to use now and that might not be available if the central bankers wait.

Keeping an open mind on the likelihood of deflation isn't easy right now. The media are ready to supply a steady diet of deflationary gurus predicting

Dow

3600. Rather than be stampeded into panic or blindly ignore the deflationary scenario completely, try to evaluate the evidence as evenhandedly as you can.

The Bottom Line

As for the three industries outlined above, I see mixed outcomes ahead. Clearly the auto industry is in for a long period of falling prices because of excess global supply, and that trend will put the pinch on U.S. auto producers. But that doesn't strike me as evidence of deflation across the entire economy any more than the decline of the U.S. steel industry did in the 1960s and 1970s. Products and industries mature, and falling prices almost always come from that.

Similarly, I don't think McDonald's and Burger King prove there is some general decline in the price of restaurant food. Demographics seem to have simply moved the market away from these companies, and the companies haven't responded particularly well.

The one example that does suggest the possibility for long-term deflation is that of Xilinx (and many other technology companies). That possibility hinges on the rise of China as a world-class manufacturing base. My colleagues at

CNBC

are doing a series on China all this week. In my contribution to that series on Business Center tonight, I'll be taking a look at "Deflation: Made in China."