THE DRAMATIC DROP in the unemployment rate last month is extremely heartening. From 6.9 per cent in November, it declined to 6.4 per cent last month. But that is certainly no reason for the Carter administration to change its economic strategy and abandon its plans for a $25-billion tax cut. The administration has got itself and the country into trouble before by changing its mind too quickly on the basis of a few encouraging statistics that never developed into a trend. Last spring Mr. Carter let himself be persuaded to drop the idea of a $50 rebate to every taxpayer, on the ground that the economy had expanded unexpectedly fast in the first quarter. The growth rate has been slowly declining ever since. Abandoning the rebate was clearly the wrong decision. Before people now begin revising fiscal policy on one month's unemployment figure, they need to think again about the nature of that number.

To be useful, first of all, the unemployment rate has to be adjusted to the season. In January, for example, unemployment drops sharply as Christmas jobs end and cold weather cuts construction. The statistician's job is to try to discern the underlying patterns from figures that dance around with the calendar. It was never a simple computation, but with the very heavy unemployment since 1974 it has become vastly complex. There has never been any hint of political manipulation in these figures. The trouble is that no one formula deals entirely successfully with all of the wildly changing circumstances of the last few years. Often the basic trend can be discerned only long after the event, and that is why the statistics are constantly being revised. That is also why it's unwise to make policy on the evidence of one or two months' figures.

The unemployment rate is also, in a genuine sense, an opinion poll. If you don't have a job, the question is whether you want one -- and want it badly enough actually to go looking for it. Suppose, for example, that a young woman marries and, following an old tradition, settles down to keep house for her husbank. She isn't counted as either employed or unemployed. She simply isn't figured in the labor force at all. Then suppose that she changes her mind and goes out to look for a job as, in fact, young women increasingly do. As soon as she starts looking, she is counted as unemployed. The economy hasn't otherwise changed, but the unemployment rate is up.

In recent years the number of people in the labor force -- the employed plus the unemployed -- has consistently risen from November to December. This year, contrary to the usual pattern, it fell. No one can say exactly why; the drop was not a large one. But if people had poured into the labor force last month as they did in previous Decembers, the unemployment rate would have been two-tenths of a percentage point higher than it was -- a significant difference. If the growth of the labor force should slacken over any sustained period, the administration would need to run a less stimulative fiscal policy. But, again, it would be foolish to shift basic policy because some people decided not to look for a job in December. Perhaps they only put it off a month.

Unemployment is disproportionately an affliction of the young. Of the 6.3 million people who are unemployed in this country, nearly half are under 25 years old; nearly a quarter are under 20. To design remedies for unemployment it is essential to remember that there are in fact dozens of rates describing dozens of very different categories of unemployed people. The 17-year-old youngster in central Washington, looking for this first job, is unemployed for a different reason from that which has the Ohio steelworker with 20 years' experience out of work. The unemployment rates for blacks are still more than twice as high, category by category, as the rates for whites. The rate for black teenagers -- and remember that the rate counts only the people who are actively seeking work -- is still 37.3 per cent.