How to Think Straight: An Introduction to Critical Reasoning

by Antony Flew

  6.21 Statistics about poverty can also be misleading without anyone having been or being deliberately and, so to speak, straightforwardly dishonest in either their compilation or their employment. For a start there is the possible and often-realized confusion about the actual composition of the set of the poorest 10 percent, whether it is individuals or households which are being counted. We must never assume, because it is never true, that the membership of either set in one year will be exactly the same even in the immediately following year. But both the nature and the extent of the membership changes will differ from time to time and from place to place. Only if a large proportion of the members are found to remain members for many years will there be any reason to speak of a class of the poor.

  6.22 But before we can sensibly discuss whether there is or is not a class of the poor we must have a working definition of “poverty.” All accepted definitions are no doubt to some extent relative to the actual condition of the better-off in the society in question. But this modest relativity must be sharply distinguished from that of the definitions of those who identify poverty more or less completely with inequality and consequently define the poor as those each enjoying an income of less than such and such a percentage of whatever is the local average income.

  6.23 The introduction of such definitions is one way to discover more poverty for welfare states to combat. Another is simply to ignore some people’s actual sources of income in cash and/or in kind. In the United States, for instance, official statisticians have enormously exaggerated the actual extent of deprivation by counting only cash incomes and ignoring Aid to Families with Dependent Children (AFDC), food stamps, subsidized housing, free medical care, and other forms of tax-financed and state-supplied income in kind. Indeed the official census data are based on samples in which subjects are not even asked whether they receive food stamps, live in public housing, or are eligible for Medicaid. Furthermore, if we are genuinely concerned to discover the actual standard of living enjoyed or suffered by the clients of a welfare state we need to inquire without prejudice how many of those clients contrive to acquire incomes in cash and in kind larger than those they receive in a variety of tax-funded handouts. Such investigations have been made in the United Kingdom, and have revealed that the actual incomes are on average significantly larger. In part these additions represent casual earnings in the black economy, in part gifts from friends and relations, and in part what can only be described as miscellaneous. (Here I am reminded of a New Yorker cartoon in which a taxpayer is shown expostulating to an IRS official. The caption read: “What do you mean ‘clarify miscellaneous’? What do you think that word is for?”)

  6.24 Consider, to go back a bit, yet another kind of unfair financial comparison. Suppose that a large sum of money is mentioned. Before responding we should take care to ask ourselves to what in the present context it is relevant to relate this sum? If we are to respond sensibly we must have benchmarks to guide us. We have perhaps seen a headline screaming that some company has accumulated profits running into hundreds of millions of dollars. Before remarking sourly and suspiciously that that is where the money goes, we should ask what these millions amount to as a percentage return on capital employed or how they compare with the company’s wage bill or with its total sales. These are the comparisons which are relevant if we are thinking of the profits as booty available for a shareout among either the company’s shareholders or its employees or its customers, or—of course—the IRS. Nor should we ever forget that in such cases the IRS always gets two cuts, one directly out of the company’s profits and then another out of the shareholders’ income from shares in that company. And, if the subject of our discussion were the U.S. federal budget, we should have to recall that famously wry remark of Senator Everett Dirksen: “A billion here and a billion there, and soon you are beginning to talk of real money.”

  6.25 When we ask the right questions about figures and actually do relevant rough calculations, we may be surprised by what we find. We certainly shall be when we turn from questions about income to questions about wealth. The distinction is important. It is as easy as it is common to think of all wealth as available for spending. If any individual comes into a bit of money from a legacy or wins a cash prize in a lottery or is given some special bonus, then there is little to stop the spending of the whole lot in one gigantic spree. But it is a very different story if we are talking about the wealth of a nation. Most of this is in the form either of the physical means of production or of the residences of the population and the miscellaneous equipment of its daily living. We cannot collectively sell off any of these assets and spend the proceeds without thereby prejudicing our collective prospects of earning and enjoying a tolerable living in the future.

  6.26 We have here another instance of a general truth exemplified differently in chapter 5 (see especially paragraphs 5.22–5.23). We are not entitled to assume either that what applies to the whole of some group taken collectively applies equally to any and every individual member of that group or the other way around. Consider another example: Anyone can park here at any time, but it does not follow that everyone can do so simultaneously or even successively. The crucial practical difference is that while not everyone wants to park here, or at any rate not at the same times, almost everyone would like to lay their hands on some of someone else’s wealth and to spend it as their own income.

  6.27 In many countries, though perhaps less frequently in the United States than elsewhere, people are forever coming up with figures supposedly showing that some percent of all the wealth in that country is owned by a relatively tiny proportion of the total population. Before entering into any discussion of what, if anything, needs to be done about this alleged fact we need first to ask both where these figures come from and upon what principles they were compiled. Unless we have some satisfactory answer to the first of these two preliminary questions, we have no guarantee that the figures were not simply invented.

  6.28 Of course here and now and in this matter it is unlikely that such figures would be invented. But it is well to remember that both the U.S. Central Intelligence Agency (CIA) and some leading U.S. economists were apparently misled by their study of official Soviet statistics to believe, until nearly the time when its collapse became notorious, that the Soviet economy was healthy. We may also recall the story of the very senior judge in the former British Indian Empire who told a young (British) civil servant: “When you are a bit older you will not quote Indian statistics with that assurance . . . what you must never forget is that every one of those comes in the first instance from the chowty dar [village watchman], who just puts down what he damn well pleases.”

  6.29 It is important to press the second question too. Users of statistics, unlike users of television sets, need to know something about how the constituent materials were originally gathered and later cobbled together. Without this information we may easily draw unwarranted conclusions. If the figures referred to the United Kingdom, for instance, we should need to ask whether they were compiled from the Inheritance Tax returns and, if so, what if any allowance was made for the fact that most estates are too small to attract this tax? The sum of all these individually small estates could very well be, relative to the sum of all the biggest estates, large. Whether or not it in fact is, is the main part of precisely the point which is here in question.

  6.30 Again, what allowance, if any, was made for pension rights and life insurance policies? The former escape the Inheritance Tax net completely, whereas the latter, insofar as they are caught, are bound to come up tagged with a higher value than any that could have been imputed earlier. Yet pension rights and life insurance policies, usually backed by investments held in the name of some corporation, must during life constitute a significant proportion of the wealth, such as it may be, of most middle-class people.

  6.31 The Economist, a journal published in London but having much larger circulation outside the United Kingdom than in it, in 1973 made and published some
calculations about the distribution of wealth in an imaginary country, not Ruritania this time but Egalitaria. These calculations are relevant to us in three ways. First, they constitute another and even more impressive example showing how unexpected may be the results revealed by still not intolerably difficult calculations. Second, they underline the curiously neglected importance of the fact that people progress through life cycles from infancy to old age. We can scarcely expect, in either sense of “expect” (see paragraphs 5.14, 5.19, and 6.18), to be in the same financial situation at every stage in our particular life cycle. This is one of the reasons why the membership of the set of the poorest 10 percent in one country in one year will not be exactly the same as its membership in the following year and in the year after that. Third, they spell out some of the unnoticed implications of one set of egalitarian ideals. They thus provide those who nourish such ideals with a very necessary frame of reference, enabling them to assess more accurately how far and where the actual situation falls short of their own aspirations.

  6.32 According to the Economist, all Egalitarians are educated in public schools up to the age of twenty-one, with no opportunity at this stage to earn enough to save. All men then work for the same wage till sixty-five, when they retire on full pay. Women work for only twenty non-childbearing years but in those years get equal pay, with the same pension rights, as men. Inheritance is forbidden, but all earners and all pensioners save exactly 10 percent of their incomes, investing their savings in state bonds yielding 10 percent compound interest. This high rate perhaps compensates for the absolute embargo on all capital appreciation. The roundness of the figure also simplifies the arithmetic, as do two further stipulations: the net reproduction rate has been unity for the past eighty-five years and everyone dies on their eighty-fifth birthday. So how much of the privately owned wealth in Egalitaria is owned by the richest 10 percent of the whole population? “The answer seems to be that the wealthiest 10 percent of Egalitarians (who by definition are all the men aged from sixty-eight to eighty-four inclusive) must now own about 74 percent of the privately owned wealth . . .” (May 26, 1973, pp. 16–18).

  6.33 In many, perhaps all, countries the tale is told of a stranger asking a villager for directions and receiving the response: “If I wanted to go there, I would not start from here.” The question of the baseline from which they start is often crucial for appreciating the actual implications of a set of statistics. A once well-known study of crime in the United States set out to refute the notion that Supreme Court decisions of the Earl Warren era caused the subsequent soaring rise in the rates of crime. Charles Silberman argued that at the time of his writing, the courts “prosecute, convict and incarcerate a larger proportion of those arrested for a felony today than did courts of the 1920s” (Silberman 1978, p. 261, emphasis added). But the era of the Warren Court began not in the 1920s or even in the 1930s but in 1953, and its landmark criminal law decisions are largely from the 1960s—Mapp v. Ohio (1961), Gideon v. Wainwright (1963), Escobedo v. Illinois (1964), and Miranda v. Arizona (1966). Crime statistics based on the actual chronology of the Warren Court show that if the thesis Silberman was so eager to refute is to be refuted at all, then the materials for that refutation will have to be found somewhere else, as, in a way, Silberman himself seems to have appreciated.

  6.34 When Earl Warren became chief justice in 1953, the homicide rate in the United States was 4.8 for every one hundred thousand of the population—lower than it had been for four decades. The total number of criminal homicides in American urban centers of twenty-five thousand or more people was no higher in 1953 than it had been in 1937, even though the population of such urban centers had grown by more than 19 percent between 1940 and 1950 alone. In short, the Warren Court inherited not only a low homicide rate but one of which the overall trend was not rising. Yet a sharp rise in homicide began in the 1960s. The rate more than doubled from 1963 to 1971. (See, for instance, Sowell 1986a, chapter 9; cf. Wilson 1975.)

  6.35 A more recent and indisputably decisive example of misleading by the choice of a particular baseline is provided by the report of the generously financed and staffed Joseph Rowntree Foundation, Inquiry into Income and Wealth (Hills et al. 1995). These researchers started by so defining “poverty” that in any country a household must be accounted poor if it receives less than half of whatever are the average earnings in that country—a definition which, they assure us, is generally accepted and employed by social scientists. Using that definition they were able to find that between 1979 and 1992—note these dates—the proportion of the U.K. population to be accounted poor had risen from 7 to 25 percent. The publication of their report was rewarded by banner headlines. For lay persons, unaware of the social scientific definition of “poverty,” were bound to misunderstand it as having claimed to have discovered, in a country which is by world standards unusually prosperous, no less than fourteen million people whom such lay persons would account to be poor. But our concern here is with the second main finding in this report. That was that the formidable increases in vandalism, burglary, and violent crime that occurred between 1979 and 1992 are to be attributed to parallel increases during the same period in unemployment, in inequality (the gap between the highest and the lowest incomes), and in poverty, as we are told that poverty is defined by social scientists.

  6.36 The stated reason for adopting 1979 as the baseline was that 1979 was the year in which the first of a new series of official British Crime Surveys was published. It is obvious that for these researchers a further reason was that 1979 was the year in which, as a result of a general election, the Conservative party led by Margaret Thatcher entered office. The researchers, of course, had every right to abhor Thatcher, her administrations, and all their works. But as professed social scientists they ought to have been especially thorough and conscientious in their testing of a hypothesis which they so very strongly wanted to believe was true. Instead they revealed a breathtaking indifference to what surely were undisputed and decisively falsifying facts.

  6.37 For example, their own report shows that the period 1961–1979 was one of rising incomes, declining absolute and relative poverty, and, for most of the time, uniquely low unemployment. It was also a period of rapidly rising recorded crime, riot, and drug abuse. Yet these researchers do not attempt to explain how and why these facts consist with their hypothesis. Nor do they attempt to test that hypothesis by reference to earlier periods during which there was far more unemployment, more absolute poverty, and more inequality between the best off and the worst off than between 1979 and 1992 and during which there was also far less crime than at any time between 1961 and 1992. (For an exhaustive critique of this disgraceful report, see Dennis 1997.)

  6.38 Where there clearly is some historical trend, before-and-after studies can be especially misleading—even to those less eager to be misled! Suppose, for instance, that in some country there has been a long-term decline in the annual total of fatal automobile accidents. Then, if some piece of automobile safety legislation is introduced, it will be difficult or impossible to determine what part if any of any subsequent decline in the accident rate was produced by that legislation. Understandably, the backers of the legislation will claim that the decline was all their own work. But how, if at all, can they show that they are not committing the Whatever-follows-must-be-the-consequence Fallacy (see paragraph 1.53) and that the decline would not have occurred even without the legislation?

  6.39 A main theme of the present chapter is that what is needed first, most, and all the time is not technical expertise in statistics or anything else, but an unspecialized critical alertness. The way to acquire this or any other disposition is by practice. To make clear that it is a lay rather than an expert business it may help if some of the practice is on specialists, and these specialists on what they see as their best behavior as specialists. For example, many years ago Edmund Leach, who was in his day one of the leading British anthropologists, was invited by the British Broadcasting Corporat
ion (BBC) to deliver the lectures in a prestigious annual series. These lectures were afterward printed in the BBC’s weekly journal The Listener. In one of them Leach asserted, very confidently, that although “Admittedly the statistics show a numerical increase in the incidence of crime . . . this is a measure of police efficiency, not of the moral state of the nation. Crimes are created by Parliament: it needs a policeman to make a criminal” (November 30, 1967).

  6.40 Certainly it is true, true by definition, that in the legal sense of the term “crime,” what is and is not here and now a crime is determined by what the law says. Private homosexual relations between consenting men over twenty-one years old ceased to be criminal in England on the day when the corresponding change in the law took effect. So there is a sense in which British criminals are made (not by British policemen and policewomen but) by the British Parliament. But from this definitional premise we are not entitled to deduce the massively substantial conclusion that either the British legal system is or the British policepersons are in fact so counterproductive as actually to induce people to perform actions of the kinds proscribed as criminal. The point simply is that the British Parliament determines what kinds of actions are to be proscribed as criminal, not that it makes people perform such actions. Leach was misled into uttering this piece of nonsense by his desire to ally himself with numerous social-scientific colleagues who were at that time attempting to persuade the lay public that what we believed to be (and now know to have been) a rising tide of crime was in truth the merely imaginary product of a “moral panic.”