The political and economic context of this tragic development can be described simply but accurately. When the official war against poverty was declared in 1964, unemployment was declining and prices were stable. Above all, there was a euphoric belief that a judicious fiscal and monetary fine-tuning by the New Economists around the White House had ended the business cycle. As Lyndon Johnson stated that faith in one of his last messages to Congress, “No longer do we view our economic life as a relentless tide of ups and downs.” Why not skim off just a bit of this perpetually expanding wealth and use it to rescue the people of the other America? Happy assumptions about the economy were the premises of the federal idealism of the 1960s.
The unofficial war against the poor of the 1970s takes place under completely different conditions. There is chronic high unemployment—the worst since the Great Depression—and the chairman of the Council of Economic Advisers promises that it will continue for the rest of the decade at levels thought intolerable in the 1960s. The nation has just emerged from a severe round of inflation and the first, modest signs of recovery are enough to make people fear that prices are going to soar again. Above all, there is no confidence that anyone really knows how to get the economy out of the relentless tide of ups and downs in which it founders.
Someone has to be the victim of this new pessimism. If there were vigorous plans for full employment—including measures financed by redistributionist tax policies it would be possible to fight inflation, joblessness and the deterioration of the public sector simultaneously. But that would require a political consensus well to the left of where most politicians, including President Carter, locate the American people today. Instead, Carter proposes a moderate program of economic stimulus, while his chief economic adviser candidly tells the Congress that it will be 1979 before the unemployment rate might drop below six percent. Meanwhile, the liberal Democratic governor of New York presents a budget that catches most of the Republican opposition off balance since it is somewhat to their right. In the political economy of 1977, the poor have been volunteered to sacrifice on behalf of the rest of us.
Trouble is, no one wants it to be said that he is insensitive to the plight of our most vulnerable citizens. Therefore it would be convenient if the poor could simply disappear. A new study published in January by the Congressional Budget Office, called “Poverty Status of Families Under Alternative Definitions of Income,” takes a major step in turning the poor into non-persons. By an act of statistical legerdemain CBO eliminates the poverty of 3,628,000 families and doubles the rate at which the poor have been escaping from their fate since 1965. This remarkable news comes at a time when other government figures are reporting an increase of 2.5 million in the number below the poverty line. Obviously, the discovery is politically propitious.
The Supreme Court, Mr. Dooley said in that celebrated aphorism, follows the election returns. So do the statistics. I do not suggest that scholars simply juggle their numbers in cynical obedience to political manipulators (although that happens, too). It is more complex: the economic-political mood of a period focuses certain facts and blurs others. In the 1950s, most sincere and compassionate intellectuals could see clearly what was right with America; in the ‘60s, they could see clearly what was wrong. (Without getting into an epistemological thicket, I should simply say that I am not an agonistic relativist in such matters. The subjective view from the bottom of the society up is more objective than from the top down; the 1960s were more truthful than the 1950s.)
I first became aware that millions of poor people were about to vanish when I read the 1976 Report of the Council of Economic Advisers, a document issued by the Ayn Rand disciple, Alan Greenspan, in January 1976. One conclusion of this report was that the official definition of poverty neglected to count in the “in-kind” income received by the poor, i.e. the value of medicaid, housing subsidies, food stamps and the like. The council, while admitting that the data were not quite clear, said that these in-kind contributions total 130 percent of the gap between the cash incomes of the poor and the poverty line. The report hedged. It did not flatly say that all of the in-kind benefits actually could be attributed to the official poor—some of them went to people above the line-—but it strongly implied that a major revision of our estimate of the number of the impoverished was in order.
Writing in the summer 1976 issue of the neoconservative journal The Public Interest, Edgar K. Browning, an economist from the University of Virginia, was not astentative as the council had been. “When in-kind transfers . . . are counted as income,” Browning said, “the average poor family in 1973 had an income that was approximately 30 per cent above the poverty line. In terms of the average income of officially poor families, there is practically no poverty—statistically speaking—in the United States today, and, indeed, there has not been for several years. It only remains for our accounting procedures to be modified to record this achievement.”
By last fall, this good news was being communicated to the people by the press. On the Op-Ed page of The New York Times Harry Schwartz told of the “more realistic definition” of poverty developed by Browning and others. He chided the government for issuing statistics that were, for most people, “misleading.” Then in January of this year the Congressional Budget Office released its study, asserting much the same thing. Since the government had said in late September that the number of the poor had gone up by 11 percent during 1975, the news of our hidden achievement was especially surprising.
In fact, the statistical method used in these revisions could have been designed by Dr. Pangloss. The procedure was to take all of the factors—some real, some imagined—that might conceivably have led to an overcounting of the poor and to give them the greatest weight possible. Simultaneously, one carefully ignored all of the data that suggest that there has been an undercount of the poor. Not surprisingly, this tendentious exercise resulted in the inevitable conclusion that the number of poor people has been exaggerated. Harry Schwartz and others thought that they were champions of social scientific rigor. In fact, they were unwitting shills for some careless and prejudiced analyses. And they were helping to make some very real poor people invisible.
Some of the considerations that were excluded systematically from the new definitions are the subject of an extensive literature. It is shocking that the Congressional Budget Office, which heretofore has observed rather high standards, should have permitted a study as politically biased as this one. The kindest thing one can say is that perhaps it is innocently incompetent.
The CBO analysis takes the official poverty definition as its point of departure, and then criticizes it for overcounting. That definition, developed by Mollie Orshansky of the Social Security Administration in 1964, defines a number of annual-income poverty lines depending on family size, sex of the family head, farm or non-farm residence, and number of children. The figure usually cited in the press as “the” poverty line is the figure for a four-person urban family. It is arrived at by taking the cost of an “economy” plan food budget designed by the US Department of Agriculture for “emergency or temporary use when funds are low” and multiplying that figure by three. Since three-to-one is the assumed ratio between food and total budget cost. Since 1969, the figure has been increased by reference to the consumer price index rather than the cost of food in the economy plan.
The CBO carefully notes how this method does not take into account the value of in-kind goods and services. It just as carefully ignores the voluminous literature which criticizes the definition for being much too low, that is for undercounting the poor. One of the most prominent critics has been Mollie Orshansky herself. The three-to-one ratio between food cost and total budget is based on a 1955 study which there is reason to believe is out of date; the basic food component itself never has been changed. The consumer price index does not adequately reflect the impact of inflation on the poor: using a base year of 1967 as 100, the CPI stood at 158.3 as a national average in 1975—but it was 162.3 for the poor and 157.2 for the wealthy.
Last April the Department of Health, Education, and Welfare did a much more careful job of surveying the problems of defining poverty. (This report, of course, was available to the CBO researchers.) It concluded that if one raised the multiplier from three-to-one to five-to-one, and used “low cost” rather than an “economy” food plan—two changes that are “within the range of reasonable judgment”—then the official poverty lines would be twice as high and one-third of the American people would be defined aspoor. Orshansky considered several alternative definitions that would correct for undercounting; the result at one extreme (which is, however, quite plausible) was that the 1974 poverty line for a four-person family went from $5008 to $8118. This would increase the number of the poor in 1074 from 24.3 million to 55.4 million.
The CBO did not have to agree with Orshansky’s highest estimate. But when a supposedly scientific analysis of a statistical definition does not even mention the reams of serious critiques of that definition from another point of view, something is profoundly amiss. Harry Schwartz might be forgiven for not knowing about the HEW study, the CBO cannot be, and should not be. One becomes suspicious when there is such a glaring omission. It calls to mind Gunnar Myrdal’s remark that ignorance is never random. In this case, it serves the neo-conservative mood of many ex-liberals.
The CBO lingers over underreporting of the income of the poor, emphasizing, for instance, that only 74.5 percent of public assistance shows up in income statistics included in the government’s current population survey. But the same chart from which the CBO gets this information shows that only 44.2 percent of dividend, interest, rental and like income of the middle and upper class show up in the survey. The CBO does not emphasize this matter—one that becomes quite important when relative shares are at issue, since the missing poverty income is worth 2.8 billion dollars while the middle and upper class shortfall comes to 46.3 billion dollars.
There is no mention in the CBO report of an even more dramatic case of missing data: the census population undercount. It is generally thought that the government misses as many as 30 percent of the population in each census. Most of the missing are poor: people without regular addresses, phones or jobs, and illegal aliens who do not want to become any kind of statistic. The Carter administration estimated this month that there are four million people “whose names are not listed in federal files because most of them pay no taxes and receive no government benefits.” They are sometimes called the “poorest of the poor.” They are not mentioned by CBO.
In attributing the value of the in-kind goods and services as income to the poor, the CBO values these items at their dollar cost to the federal government. Thus part of the increased “income” of the other America is that cost of unnecessary, and even harmful, medical care at inflated prices in medicare mills; and the money which various real estate speculators rip off from housing subsidy programs. The CBO makes a large assumption here: that in-kind income received from the government is the functional equivalent of cash income. In the process, a remarkable negative fact about American society actually sneaks into a relentlessly positive analysis. The “natural” workings of the economy—incomes before any transfer payments are added in—leave about 25 percent of the people poor. Moreover, this shocking fact has changed only “slightly” (the CBO’s word) in the past decade (28 percent of families were below the poverty level before transfers in 1965, 24 percent in 1975). So one out of four citizens is dependent upon the government, not the private economy, for a crucial portion of its standard of living.
Is it fair to equate the in-kind income poor people receive with the pay checks of the other three quarters of the population? One obvious difference leaps to mind: cash income can be spent however a person wants; in-kind income cannot. Indeed, on the CBO assumptions people can enormously increase their income—and even enter the middle class by suffering a catastrophic illness at government expense. (This problem gets one sentence in the study.) A large part of medical in-kind benefits goes for the care of terminal illnesses. It is a gain that some of the poor can now have some decency in their final moments on earth, but I do not think that this fact should be counted as a major contribution to the elimination of poverty in America.
But let me confront the difficult case where important benefits are indeed conferred, and not just upon those about to die. Food stamps, which unquestionably represent a significant gain for the nation, are an obvious example. Can they be simply valued in terms of their money cost (or their market value)? I think not. Poverty is not simply a statistical phenomenon describing certain characteristics of an aggregate of individuals; it is social reality as well. That is, the poor do not merely lack income. They also go to jail for their wrongdoing, which the middle class, in most cases, does not; they get sick more often; they suffer from greater emotional and mental stress; and so on. Harry Caudill and, more recently, Kai Erikson have hauntingly evoked the demoralization which welfare dependency can bring. Would the CBO be happy if three-quarters of the people received statistically adequate incomes but were cut off from any meaningful work functions in the society?
If the CBO study had dealt with this problem, and the others I have cited, that would have been an earnest of good faith even if it were badly done. Most of these difficulties are not even alluded to; those that are get the most superficial mention; and nothing is allowed to interfere with the basic, and it seems to me preordained, conclusion. To be sure, there are references to forthcoming technical documents and one hopes that they will raise, and perhaps confront, some of these issues. But Alice Rivlin, chief economist of the CBO, claims that this study provides the basis for a congressional evaluation of welfare reform. Any congressman who reads it is likely to be seriously misled.
Finally, one other instrument of social analysis which this study ignores should at least be noted: the human eye. The CBO analysts work in Washington. While traveling through that city did they ever lift their eyes from the examination of mathematical equations and look out the window? Perhaps they take only the sanitized routes. Did Harry Schwartz of The New York Times invest a dollar in a round-trip subway ride to the South Bronx? If he had, he would not have written his column as he did. Has anyone looked at Detroit recently? St. Louis? The Appalachian hollows or the barrios of the Southwest?
I know all of the objections to personal and random observation as a means of social scientific analysis. I share some of them. But before we become a nation of computer cretins, we might just take a walk and see if poverty still exists, massively and cruelly. It does.