Should we adjust the population count? Dr. Eddie Williams, director of the Joint Center for Political Studies, recently urged the secretary of commerce to direct the Census Bureau to develop and implement a plan for adjusting the 1980 census counts. If the secretary does not act immediately, he argued ["Counting All Americans," op-ed, July 31], a lot of people will be short-changed for the next decade -- presumably by a lack of equity in the allocation of federal funds.The Post seconded his argument Aug. 6 in its editorial, "Accounting for the Uncounted."

Attention to this question is in order, but precipitate action is not.

Experience tells us that next April's decennial census will count almost everyone. The "almost" represents an estimate of under-count, which was about 2.5 percent, or 5.3 million people in 1970, even though we counted more than 203 million people. Our research showed that the undercount was about 1.9 percent for whites and 7.7 percent for blacks. Hispanics also probably were undercounted at a disproportionate rate.

The first thing we are going to do about an undercount in 1980 is to eliminate as much of it as we can, especially the differences between groups. Extraordinary efforts will be taken to count everyone. Our plans are being supported actively by elected officials, minorities and various organizations across the country.

The second action we are going to take is to develop more information on the undercount. We expect, for example, to have estimates of undercount for each state. After the census, we plan to evaluate our performance intensively. Having budgeted up to $40 million for these activities is a demonstration of our commitment. This research and evaluation will continue and the results will be disseminated, regardless of when a decision or what decision is made as to adjusting the counts .

Sometime in 1981, we should have a pretty good idea of any undercount rates -- but only at the national level. That is the earliest point at which an adjustment reasonably could be made, but there are serious questions as to its desirability then. As noted by the National Academy of Sciences' Panel that Dr. Williams cited, the adjustment method that could be used then might produce figures further from the true population for many local areas. Alternative methods relying on results obtained trhough post-census surveys and matching studies -- would be technically superior.

If a decision were made today to adjust the census counts, it could not be implemented for at least two years -- and this is the key point. No one will be short-changed if we use the months ahead to sharpen our research and to ensure that all affected parties become aware of the consequences of an adjustment procedure. Also, a pre-census announcement to adjust might give people the harmful impression that it is unimportant to fill out the census form.

On balance, we believe we have chosen the best path. Through systematic study and wide consultation, the Census Bureau will examine the undercount adjustment issue comprehensively. The decision then can be made on the best facts available.

Here are some of the questions that need more study:

What would be the consequences (legal and otherwise) if there were two sets of census figures: 1) basic counts for apportionment, and 2) adjusted counts for the distribution of federal funds?

How many dollars actually will be at stake and how would the distributions of funds increase or decrease by size and type of jurisdiction?

If census figures are not adjusted, are there satisfactory alternatives to compensate for any apparent inequities?

Is the question of equity a policy matter solely for the Commerce Department and the Census Bureau to resolve, or also a political question for Congress?

Would an early decision to adjust for undercount reduce the intensity of efforts to achieve as complete a count as possible?

Would a complex adjustment procedure be understood well enough to be credible?

We believe it would be unwise to slight such questions and imprudent to rush to judgment.