It is all very curious. Once it became clear that the Republicans had won power on November 8, 1994, it took only seven days for the serious thinkers in the liberal press to rediscover the issue of poverty. “Scapegoat Time,” wrote Bob Herbert of the New York Times , while Anna Quindlen evoked the left’s favorite image for Republicans with “The Politics of meanness.” “More for the rich, less for the poor,” Mark Shields declared a week later.

Yet just a month earlier, when the October poverty statistics revealed quite clearly that the number of poor people had risen to an all-time high under Bill Clinton, the liberal columnists never said a word. When poverty rises under Democrats, it is a nonevent. It does not fit the prevailing view of the two parties. To the mainstream press, Democrats are never mean, only compassionate. For the press, “mean,” “greedy,” and “inequality” are stencils whipped out only when Republicans are in power.

Those religious people who normally do not specialize in economic statistics depend heavily on the press. They need to be on guard against partisan abuse of numbers. In an earlier column in First Things (“The Rich, the Poor, and Reaganomics” [April 1991]), I predicted that the number of poor people, after declining under Reagan, would rise under Bush despite the latter’s “kinder, gentler” rhetoric. It usually surprises people to see what actually happened, because the press, having blamed poverty on Reagan, paid little attention to the problem after he left the scene. What actually happens to the poor is a matter too important to leave to journalists.

At the end of the first year of the Clinton Administration, the Census Bureau’s Income and Poverty Report for 1993 found 39.3 million poor persons, the highest number since 1962. There are now 6.9 million more poor persons than in 1989, Reagan’s last year. The percentage of the population that is poor is also much higher than in Reagan’s last year: 15.1 percent versus 13.1.

Yet Mark Shields, Bob Herbert, Anna Quindlen and their colleagues have seldom expressed a worry about the poor under Democratic administrations, even on those occasions when Democrats presided over a large increase in poverty. For example, the greatest rise in poverty since World War II occurred under a Democrat, Jimmy Carter. Between 1977, when Carter took office, and 1981 when he left it, more than seven million people were added to the poverty rolls, the most ever in a presidential term. Meanwhile, the poor are used by the press as a club to beat Republicans.

In his column, Mark Shields described “the rich” as the top 20 percent in annual income. Yet Mark Shields-and his major newspaper and television colleagues-have incomes in the top 5 percent (above $106,000). Just the same, Mark (a great guy, by the way) and his friends have a way of using the expression “the rich” as if to distance themselves from it. When they say “the rich,” they don’t mean themselves and they don’t even mean exactly “rich,” they mean the non-liberal rich.

Moreover, Mark Shields writes about the poor in 1969 and the poor in 1994, twenty-five years later, as if nothing had changed during that period. Yet in the meaning of the term “income” and in the composition of those included in the poorest 20 percent of households great changes actually took place. During that quarter century Great Society programs for the poor increased exponentially. Welfare spending on the poor totaled more than $5.3 trillion and included every sort of noncash income that the new poverty industry could think up (Medicare, Medicaid, housing assistance, food stamps, Head Start, etc.), along with direct cash grants. If the effective income of the poor in 1994 was not far greater than it had been in 1969, a great deal of money was misspent.

Indeed, so much government money has been spent trying to improve the condition of the poor since 1965 that it would actually have been more efficient to mail every one of the poor families, households, and single persons in the United States enough cash to lift all of them above the poverty level every year. The cost-calculated each year as “the poverty deficit”-would today be about $50 billion, a small fraction of the money expended on means-tested programs.

In addition, the composition of the people officially classified as poor has changed dramatically since 1965. Proportionately fewer of the poor today are fathers of families or even single men. Far fewer, too, are in the labor force. The tide of immigration increased, and more of the poor are immigrants “passing through” poverty, soon to be poor no longer. Above all, and sadly, many more are a type of poor person new to America, healthy young people accustomed to staying dependent on welfare year after year.

Between 1967 and 1992, the number of households headed by persons over sixty-five grew by 9.2 million. This is because considerably more people are living to old age than was the case in 1967, and especially because the financial condition of older people has improved so much that nearly all of them now prefer to have households of their own. Most elderly householders have paid-up mortgages, and all benefit enormously from Medicare. Not only are more of them living longer, they are healthier and they look and feel younger than people their age did in the past. Even though many of the elderly are still to be found in the bottom 20 percent of household income, since they are retired and no longer earn income through employment, they are much better protected financially than their predecessors were in 1967. For the elderly, the poverty programs of the last twenty-five years have really worked.

For those under sixty-five, the picture has become far more bleak than it was. Crime rates (especially violent crime) have shot up some 700 percent. Illegitimate births have climbed more than 600 percent. Those of us who were in favor of the poverty programs never predicted such outcomes. On the contrary, our intention was that there would be less crime, less illegitimacy, and stronger families. Poverty programs may not have caused these unintended consequences, but they certainly did not achieve our intentions.

The picture for the bottom 20 percent by income looks like this. There were nearly 19.3 million households in this bottom fifth in 1992 (the last year for which data are available), and only one in five of these households is headed by a married couple, while three in five are headed by single females. Contrast this with 1960, when nearly two-thirds of households in the bottom fifth were headed by persons who worked. By 1991 the percentage of heads of households in the bottom fifth who worked (even part-time) had fallen to around one-third, and only 11 percent worked full-time throughout the year.

This sea change in household composition during the past twenty-five years explains why the old maxim “A rising tide lifts all boats”-used so effectively by John F. Kennedy in 1962-no longer seems to work. A hidden presupposition of this maxim is that household composition remains fairly uniform at all income levels. This is much less true today than it was twenty-five years earlier. In higher brackets today married- couple families predominate, with their capacity to generate two incomes; only 20 percent of households in the bottom fifth are married- couple households.

Moreover, in the bottom fifth today a majority of heads of households are not only unemployed but not even in the labor force (i.e., not even looking for work). Many are over sixty-five, retired, living on savings, pensions, and social security. Many are husband-absent single females with small children for whom to care. (These are the two most rapidly growing categories in the bottom fifth.)

These are the major reasons the real income of poor people has been declining in recent decades. Even an economic recovery (such as the great boom of the last six Reagan years, as well as the more modest boom that began in 1992) cannot lift incomes for those who are not in the labor force earning income.

The unemployment rate for married men is just over 3 percent (and for married women not much higher), which is why relatively few married- couple families in 1992 had incomes below $16,960, the cut-off point defining the bottom 20 percent. Employment matters. But it is not an option for those who, for various reasons, are not in a position to work full-time-the elderly, single mothers with small children, and others.

The press makes much of the “growing income inequality” between the poor and the nonpoor. But since most “in-kind” government subsidies are not counted in computing income, and since most of those in the bottom fifth are not earning income by employment, it is very hard to see what would raise the incomes of the poor. (Those in the bottom fifth report spending three times more than they report as income, but part of this can be accounted for by the elderly and others drawing on their savings.)

Age profile, marital status, and lack of paid employment are serious factors affecting the income of the bottom 20 percent of the American population. It helps no one to imagine that those in the bottom 20 percent are in the same position with respect to these factors as those with higher incomes. The composition of the bottom 20 percent is sui generis.

A final note. The percentage of the “officially poor” is, as noted, at the highest level in many years, 15.1 percent. Obviously, the “bottom 20 percent” includes a somewhat higher range of incomes. The poverty threshold in 1992 for a nonfarm family of four was $14,335, and lower for smaller families and persons living alone. But the cut-off point for the bottom 20 percent of families was $16,960.

What is behind “statistics”? My father and mother, perhaps like the parents of most readers of this journal, lived into their eighties and had an income that placed them well down in the bottom 20 percent. But they owned their home in Pennsylvania outright, as well as a condominium in Florida (with a small mortgage), and their expenses were few. They would have been astonished to have been counted as poor. My father thought he was rich; my mother marveled at how much better they were living than they had ever expected to. “Statistics” often give a false impression of uniformity masking the variety of realities they cover.

The good news is that, after having been ignored by the press during the first two years of the Clinton Administration, the poor are in for an explosion of attention from the media now that the Republicans have won the House and the Senate. The bad news is that too much of this attention will be based on false conceptions, and too much of it will be used to score partisan points.

The first moral obligation, as Pascal wrote, is to think clearly.

Michael Novak holds the George Frederick Jewett Chair in Religion and Public Policy at the American Enterprise Institute.