Global is the "in" word today. It's hard to find a car company anywhere that isn't globalizing. Of course that means whatever any company wants it to mean. But there's no proof that it works.

Don't believe me, here's what a Booz-Allen team wrote: "...there is little evidence to date to indicate that success in globalization leads to real competitive advantages. In fact, there is even some perverse evidence to the contrary, e.g., the lowest-cost and most profitable U.S. vehicle manufacturer has for several years now been Chrysler Corp., arguably the most domestically focused and least global of all major producers."

So what's so hot about the idea, and why now? The why is simple: One@ The Japanese, and now the Koreans, built huge auto industries and began taking chunks of the American and European markets, as well as driving the Westerners out of traditional export markets.

Two: New and potentially huge markets have appeared in India, China, Southeast Asia and Russia. Globalism is seen as a way of winning some of the old business back and getting a start in new markets.

The premise is that lots of money will be saved by using common platforms around the world instead of starting autonomous car companies everywhere (like, say General Motors of Europe) or doing completely different vehicles for every region. The same with engines: Why build four or five different 4-cyl. engine families when one will do? And there is a belief that money will be saved by ordering parts from a few big guys rather than lots of smaller guys.

But that hasn't exactly worked yet. The best example might be Ford Motor Co.'s efforts to build a "world" car, which began in the early 1960s. The last try was the early '90s Ford CDW27 program, which produced the Mondeo in Europe and the Ford Contour and Mercury Mystique in the U.S. We'd have to call that less than a complete success because it cost $6 billion and took eight years to reach the U.S.

And while the car seems to be a success in Europe and may be profitable (although Ford's European operations lose so much it is hard to figure out if anything makes money), the U.S. models had problems from the start. Cramped rear seats and a blah 4-cyl. engine compared with the European version didn't help. They were priced so high that they pushed into Taurus territory and caused problems in Ford's price ladder. It's hard to believe that Ford couldn't have built a distinct U.S. model faster and at a lower cost. The CDW27 troubles probably led to Ford 2000.

The greatest chunk of cost in building a new car may well be in the tooling and the factory. The president of one Big Three company told me that this is where 60% of the money goes. If you are going to build the vehicle in Europe and the U.S., for example, the savings may be less than expected. You don't get two sets of tools for the price of one.

A former CEO of one Big Three company told me his people spent 25% of the time traveling back and forth on problems. The compromises also might mean that the vehicle built is not the best for each market. It's quite possible that separate vehicles might be better suited to separate markets, and without a big cost disadvantage.

Common engines also make sense unless a company gets stuck with a bad one around the world instead of just in one region. Why should an automaker have a clunker? Do you believe in the infallibility of auto execs? As for global parts suppliers, this means big-company global overhead.

What about the Japanese@ They started their assault on the world markets by building cars in Japan to sell at home and abroad. They also built other cars exclusively for the home market, and were quite successful. Then they went global, building plants in the U.S. and Europe, even setting up design studios to modify the sheet metal on the Japanese bodies. But they aren't gaining market share in the U.S. or Europe with all their global plants.

Today the Ford 2000 program involves designing platforms and engines from scratch for vehicles did will sell around the world. The first product will be the Escort due in a few years, so we'll have to wait to see if it works.

GM's program might be called "Whatever Lou Hughes Wants, Lou Hughes Gets." Mr. Hughes is GM's European-based president of International Operations, and he knows his business. The idea is that GM Europe will spearhead GM's global attack, and European-designed cars will be sold around the world. He has his cars in production in Europe, Mexico, Brazil, Asia and India. Somewhere it works, somewhere it doesn't, but it makes sense because Europeans have more experience building small cars than Detroiters.

At times, GM gets carried away with the philosophy. The 1999 mid-level Saturn will be built in the U.S. from a GM Europe platform. Yet it's reasonable to think (I have this on good authority) that the coming Saturn could have been done cheaper and faster without all the delays involved in using the European platform. I keep wondering how much trouble it is figuring out ways to hang plastic body panels on a platform designed for sheet metal.

GM is building a minivan in the U.S. for here and Europe. That is a good idea because it gives GM Europe a minivan without the cost of building a new factory. If there's a problem it's that Europe wanted a narrower van than GM would normally build for the U.S. So the new van looks a bit small compared to Chrysler's pacesetting minivans. We'll see if that makes a difference.

BMW and Mercedes are globalizing by spreading production out of Germany. Fiat is building its new world car in Brazil, but not in Italy. Chrysler has designed a little car it hopes can be produced for $6,000 in China. The problem with such cheapie cars is that there are millions of used cars running around the world. There's also the Indonesian problem, which is "Why let the foreigners make the money when we can build a national car here and keep the cash in the President's family?"

Still, globalization is in. It works for Coca-Cola. It may work with cars. But it won't be easy or simple. It's still "best car wins," global or not.

Jenny Flint is a Forbes columnist and a contributing editor for CNN FN's car show called The Most Toys.