Sir Josiah Stamp, an official of Great Britain's Inland Revenue Department between 1896 and 1919, once had these words of profound wisdom:

"The government are very keen on amassing statistics. They collect them, add them, raise them to the n th power, take the cube root and prepare wonderful diagrams. But you must never forget that every one of these figures comes in the first instance from the village watchman, who just puts down what he damn pleases."

Well, things may have improved a bit since Stamp's days, but not much. Every time you see a statistic, you ought to be skeptical. How did it get there? Does it really mean what it seems to mean?

The inspiration for this caution comes from a short but intriguing essay by Joseph J. Minarik, a research associate at the Brookings Institution, who systematically demolishes one of the prominent statistical conveniences of our time: the concept of "inflation of the necessities."

Journalists are great consumers of and publicists for statistics, and though some of us have long questioned the worth of the "necessities" index, its survival powers far exceed such modest skepticism. There is a plausible explanation for this. Increases in the necessities index -- which embodies price increases for food, shelter, energy and medical care -- almost always outpace rises in the regular index.

This conveys a message people want to hear or, at any rate, that journalists think people want to hear: As bad as inflation is, it's worse for average families and the poor, who spend more of their income on necessities.

Anyone who wants to believe that the necessities index does, in fact, confirm this truth ought to stop reading here. Minarik convincingly demonstrates that it doesn't

First, he points out that many items measured in the consumer price index under the "necessities" category aren't really necessities at all. Is steak a necessity? Is dining out -- included as away-from-home meals -- a necessity? Is buying a house (most of the housing component) a necessity?

Next he reminds us of some of the statistical quirks that further blur the distinction between necessity and luxury. When the price of beef rises, for example, consumers buy more lower-priced items, such as chicken. But the consumer price index won't capture that change because it measures a fixed basket of goods and services.

More importantly, the housing component of the index is almost universally recognized to contain a heavy upward bias as a result of the inclusion of the prices of new homes and current mortgage interest rates. Most people aren't buying new homes or paying current interest rates. In the necessities index, this distortion is magnified further because the housing component constitutes a larger proportion of the total index.

Finally, the necessities index suspiciously omits clothing. Minarik wonders whether its absence reflects the fact that clothing prices have advanced much less than the total consumer price index: 4.1 percent annually between 1969 and 1979 against an over-all increase of 7.1 percent.

To show how these technical qualifications affect measurement of inflation of the necessities, Minarik constructed an alternate index. He added clothing, subtracted the away-from-home meals and gave the rent-component of the housing index the full share for shelter.

The following table gives the results. It lists the increases in the over-all consumer price index, the standard necessities index (as published, for example, by the National Center for Economic Alternatives in Washington) and Minarik's alternative.

Cynics might conclude that average families and the poor have, in fact, suffered proportionately less from inflation than other people. But Minarik doesn't conclude that, and you shouldn't.

Based on what we know, it's extremely difficult to generalize about who's helped and who's hurt by inflation. There is an enormous amount of arbitrariness in the inflation process. Consider two families with the same incomes, one which purchased a home 10 years ago and one which is buying today. The economic position of the first family may have increased substantially in the past decade, while the second family may have been clobbered.

As for the poor, poverty was a problem before inflation and probably will be after -- if there ever is an after. But inflation itself almost certainly has had a varied impact. Some programs that help the poor and elderly (Social Security, food stamps) rise automatically with price indexes; others (Aid to Families with Dependent Children) do not.

But, as Minarik warns, the necessities index shouldn't be used as evidence justifying subsides to offset inflation in the necessities. Not only do such policies have a dubious factual basis, but they are likely to be self-defeating.

"We delude ourselves if we suppose that policies centered on food, housing, energy and medical care, will justify subsidizing imported oil, tolerating higher wage settlements or running large federal budget deficits," he says, "We have an inflation problem in every sector, and the longer we deny it, the worse it will get." (TABLE) (COLUMN)Consumer(COLUMN)Standard(COLUMN)Minarik's (COLUMN)price index(COLUMN)necessities(COLUMN)necessities 1970(COLUMN)5.5%(COLUMN)5.5%(COLUMN)3.6% 1971(COLUMN)3.4(COLUMN)3.7(COLUMN)3.7 1972(COLUMN)3.4(COLUMN)4.1(COLUMN)3.8 1973(COLUMN)8.8(COLUMN)13.0(COLUMN)11.4 1974(COLUMN)12.2(COLUMN)12.9(COLUMN)10.4 1975(COLUMN)7.0(COLUMN)7.7(COLUMN)6.2 1976(COLUMN)4.8(COLUMN)3.7(COLUMN)3.6 1977(COLUMN)6.8(COLUMN)8.2(COLUMN)7.0 1978(COLUMN)9.0(COLUMN)10.9(COLUMN)8.6 1979(COLUMN)13.3(COLUMN)17.4(COLUMN)12.1 Annual average(COLUMN)7.4%(COLUMN)8.6%(COLUMN)7.0%(END TABLE)