We proudly celebrate July 4 as the date our country was founded. But most historical accounts suggest that the lights of Independence Hall in Philadelphia were dark on that day in 1776, with the pivotal decisions leading to the creation of our republic having occurred two days earlier. Should we cancel our July 4 holiday and start celebrating July 2?

Some version of that question is faced by the statisticians in the Federal government every time they get new information on the economy. Sometimes it is new data that arrive after a longer passage of time. In other instances, it is a new method of combining and assembling the data already on hand. Whatever the source, the new information shifts the sands of our economic landscape, making historical growth appear and disappear in the nanoseconds it takes for a computer screen to refresh. Do we admit to the world that the economy is really not the way we said it was yesterday?

With a few prominent exceptions, the answer from the Federal statistical agencies to this question is yes. Thus we received the news from the Bureau of Economic Analysis that the U.S. recession of 2001 was not the one-quarter wonder that we thought it was last month. In fact, according to the restated figures on Gross Domestic Product released by the BEA, the U.S. economy actually suffered three consecutive quarters of declining output last year, beginning with the first quarter.

Moreover, the BEA now says that the worst quarter for the economy last year was actually the second quarter, and not the third, as originally stated. Data revisions turned what was once said to be an anemic, but positive, 0.3 percent rate of expansion during April-June of 2001 into a 1.6 percent rate of contraction. That's a nearly two percentage point revision in growth, representing about $175 billion worth of output. Or, if you like, about 46 times as large as the restatement WorldCom made over a span of years.

It's hard to imagine a worse time to make these kinds of adjustments. With this year's second quarter growth figures coming in at a scant 1.1 percent, and the first quarter's growth revised downward from 6.1 to 5.0 percent, the U.S. economic recovery is looking a lot more fragile today than it did only a month ago. News that the recession itself was longer than previously thought is hardly welcome at this point.

But we ignore the truth at our own peril. Its may be harmless to celebrate Independence Day on the fourth of July, but acting as if the economy is safely out of the woods when the facts say otherwise is clearly dangerous.

But what exactly are the facts? The revisions our statisticians make to economic history are truly a double-edged sword. They may get us closer to the truth, but they also remind us that no one really knows what the truth is. And the unpleasantness and approximations that one must make to even attempt a task as daunting as adding up the entire American economy is something that most economists and business people would rather not know.

It’s a task that's not going to get any easier. Much of what we produce in our economy, like entertainment, legal advice, or health care, can't be measured and weighed like the widgets of old economic textbooks. When different parts of the economy, especially computers, grow rapidly, it can expose ugly flaws in methods we thought were working. And, to make matters worse, spending on defense and other priorities may crowd out funding increases for the agencies that feed all of our appetites for economic information.

There's an old saying that economists like to compute growth out to two decimal points just to prove they have a sense of humor. In the wake of these revisions to our economic history, and in anticipation of those that will certainly follow, the grim truth in that joke is quite apparent.