According to the Census Bureau, nobody lives in Tenney, Minn.

That report puzzles the town's five legal residents, all of whom hold public office there. The Census Bureau ought to be scratching its head, too. If nobody lives in Tenney, whom did the bureau think it was communicating with when it sent the mayor a letter informing him that the town's population supposedly was zero?Though this episode may sound funny, it's no laughing matter. Since the Census is used to determine the allocation of federal funds and the apportionment of representation in Congress, an area that is under-counted can be shortchanged in many ways.

Yet more than 5,300 communities have filed challenges to their preliminary 1990 census reports, claiming the census takers missed entire neighborhoods and in some cases even entire towns. The complaints from New York, which claims to have been undercounted by 1 million, have been particularly loud.

What's more, the critics can point to the discrepancy between the Census Bureau's pre-census estimate that the national population would come to 250 million and its preliminary report that the figure comes to only 245.8 million.

Though there should be plenty of red faces at the Census Bureau, this embarrassing situation is still no justification for current efforts to get the government to fudge on its census figures.

For one thing, most of the horror stories involving missed towns and missed neighborhoods can be corrected.

For another, many communitiess complaining about the census have their own citizens to blame for some of the purported miscounting. Don't forget that many Americans refused to cooperate with the census takers.

For still another, the census never has been exact. Many statisticians consider a discrepancy of, say, a couple of million in a population as big as America's to be an acceptable margin of error.

So let's turn a deaf ear to suggestions that the census be replaced with modern statistical sampling methods. That would be like replacing an election with a public opinion poll. In either case, the outcome would lack much public confidence.