CHAPTER 15 DAMN LIES, STATISTICS, AND DOOMSDAY The previous chapter parades the rhetorical devices and logical fallacies commonly used to convince us that the environment is getting dirtier, resources are running out, and population growth will destroy humanity and the planet. This chapter arrays some of the statistical contrivances that are used to paint false pictures of various social issues. As case studies a) I'll use a widely-distributed 1993 pamphlet entitled The Index of Leading Cultural Indicators by a widely-known public figure in the United States, William Bennett, author of the best- selling The Book of Virtues, b) Clinton-administration Vice President Albert Gore, Jr.'s book entitled Earth in the Balance, and c) a Washington Post front-page story about blacks in the U. S.. And though these particular publications will be quite forgotten in a very few years, the material in these case studies will remain ever green because other publications in the future will use the same contrivances for the same dishonest purposes. THE BOOKLET OF STATISTICAL VICES My purpose is not to attack Bennett personally. Rather, I want to use this as an educational case study of the abuse of statistics. To help in the analysis I'll use as a cookbook a text entitled Statistics: A New Approach by W. Allen Wallis and Harry Roberts, written in 1956 and a landmark in statistics education for decades. One of its central features is instruction in how to avoid pitfalls in statistical thinking, and it lists many of the most common fallacies. Bennett says in his introduction (p. 1) that "we have experienced substantial social regression...over the last 30 years." He goes on to assert that "there has been a 560 percent increase in violent crime; more than a 400 percent increase in illegitimate births; a quadrupling in divorce rates..." and so on. Let's see the quality of the evidence he provides for his argument. 1. "Number of Crimes Committed" is the first topic Bennett addresses (p. 2). He presents the chart in Figure 15-1a showing "Millions of Violent Crimes" and another (Figure 15-1b) showing "Millions of Total Crimes". Here Bennett commits the first in Wallis' and Roberts' list of "Misuses of Statistics." These misuses are akin to the logical fallacies that false rhetoricians have used since Aristotle. Figure 15-1a and 15-1b The fallacy is that Bennett makes no adjustment for the growth of the country's population since 1960. Therefore crimes per capita - the relevant measure - have not grown as fast as Bennett's charts show. (Bennett notes this in his discussion, but that's like the small print in a contract. It's the dramatically rising curves in the graphs that hit the reader over the head.) 2. The next abuse the statistician notices is the choice of the period being charted in Figures 15-1a and 15-1b. Why since 1960? If Bennett had plotted a much longer period backwards, we would see that crime had been vastly higher at some times than in 1960. The date he chose coincides with his conclusion better than most other dates would. 3. Next we notice the really crucial issue for the issue at hand - the source of the data. Bennett's numbers come from the FBI. But informed social scientists know that FBI data refer only to reported crimes. The rate of reporting has increased greatly in recent decades, thereby biasing the FBI data trends. Therefore, since 1973 the U.S. Bureau of Justice Statistics has conducted "victimization surveys". These surveys show a mind-boggling result: no growth in violent crime (maybe even a decline) over the period of the surveys, with a large decline in non-violent crimes such as larceny, burglary, theft, and auto theft. But Bennett only presents the scary FBI series that seem to support his general claim even though the victimization data are available in the Statistical Abstract of the United States. 4. Bennett's second topic is "Median Prison Sentence for All Serious Crimes". His graph (Figure 15-2) shows a decline from 25 days in 1954 to 5 days in 1974, and then a modest rise to 1990, suggesting that the society has gotten soft on punishing crime. Figure 15-2 But ask yourself: Can five days in prison really be the "median sentence" for serious crimes? Of course not. So what do the numbers in Bennett's graph mean? When reading the fine print you find out that the numbers do not refer in any way to the "median sentence" that is in the title, but rather are a computation called "expected punishment" for the commission of a crime, a computation which includes the likelihood of being arrested, the chance of going to trial, the chance of being convicted, the chance of getting any prison sentence, as well as the sentence itself. Expected punishment is a very interesting and probably useful concept, but it has almost no connection with what Bennett claims to be talking about. This is an abuse of statistics akin to the non-sequitur fallacy in rhetoric. 5. Another common statistical misuse is combining apples and oranges when they don't belong together - the "fallacy of composition". Does "serious crimes" in the "median sentence" graph (Figure 15-2) have the same composition of crimes - and hence the same meaning - over the decades? Certainly not. Homicide and rape and theft have changed at vastly different rates, so lumping them together produces confusion at best. Figure 15-2 6. Why does Bennett start in 1954 whereas the first set of graphs (and most of Bennett's others) start in 1960? The obvious answer is that starting this graph at 1960 would make a less dramatic chart. Taken together, the graphs discussed above suggest that the presentation of crime data is rigged to get the worrisome effect that the author wants to achieve. 7. Bennett's third topic is "Juvenile Violent Crime Arrest Rates". This graph (Figure 15-3) Bennett at least puts on a "per 100,000" basis. But he does not tell us what the "100,000" refers to. Juveniles? Population? One cannot know what mischief may lurk behind the undefined number. Vagueness of definition is one of the most useful practices for the statistical doubletalker. Figure 15-3 contains all the fallacies in the earlier chart of total violent crimes, plus additional fallacies. Notice first that the vertical axis in the graph does not start at zero. This is a real meat-ax of a crude statistical trick, grade-school stuff. If the vertical axis had run all the way to zero, the rise in the curve would have been less dramatic to the eye. 8. Next we notice in Figure 15-3 that the subject is arrests rather than crimes, though the casual reader is not likely to notice the shift. Is it possible that juveniles have been getting arrested more frequently for given crimes than in the past? We don't know. But if so, the rate of arrests would give a misleading impression of the amount of crime. Shifting the definition is a most useful contrivance for portraying a false statistical picture. 9. The fourth topic in Bennett's pamphlet is "Children Relying on AFDC" (Aid to Families with Dependent Children). Again Bennett uses the now-familiar trick of showing (Figure 15- 4) total numbers of children without adjusting for population increase. 10. Bennett also mistitles the graph for AFDC as "Relying on", when the data refer to receiving AFDC. Who knows how many children get AFDC who do not rely on it at all? Indeed, how many children received AFDC half a century ago? How many children receive AFDC-type payments in Somalia? Zero in both cases, even though the children were and are more needy in those cases. The explanation, of course, is that Figure 15-3 may well show an increase in the generosity of the generous government programs rather than greater need or reliance. (Indeed, this is suggested by the rapid rise in AFDC receipt shown between 1960 and 1975, with constancy since then.) Figure 15-4 11. The rest of the booklet is depressingly similar. It also includes a howler. The rate of births per thousand teenage females is drawn in Bennett's graph as if it has risen from 15.3 in 1960 to 42.5 in 1990. But the unquestionable official data from the National Center for Health Statistics show that the birth rate for females age 14-19 fell from 89.1 in 1960 to 59.9 in 1990 (including an upturn since 1986). Bennett must have made some extraordinary error. *** No wonder the public distrusts statistical presentations. It is easier to mislead with numerical presentations than with words alone because there are more subtle tricks you can play. The only good that comes out of Bennett's booklet is that I have a wonderful demonstration of the statistical abuser's art for regular use in my elementary statistics course. A footnote about journalism: William Bennett probably would not be called a darling of the press. And journalists pride themselves on their "hard nosed" skepticism. But I have not read any criticism of Bennett's statistics in this booklet. And this is not because the subject would take tedious, difficult digging. I wrote the pages in this chapter in about three hours, with the assistance of only the Statistical Abstract of the United States, which should be on every journalist's desk. How come Bennett's falsities got a free ride? Is it because journalists simply share Bennett's general belief about "social regression" - a common belief that goes back to the Assyrians (see chapter 00), and therefore don't even subject the data to scrutiny? TRUTH IN THE BALANCE Clinton-administration Vice President Albert Gore, Jr.'s book is called Earth in the Balance. But it is truth that is in the balance, rather than our very durable planet. The book is as ignorant a collection of cliches as anything ever published on the subject. And there is lots of tough competition for that abysmal bottom spot. Just about every assertion in the book points in the wrong direction - suggesting that conditions are getting worse rather than getting better, which they are. Lest the reader accuse me of hunting-and-picking for errant soft targets, let's start with the very first topic in the book, soil erosion, and go from there. After the obligatory drama about how "eight acres' worth of prime topsoil floats past Memphis every hour", Gore says that Iowa "used to have an average of sixteen inches of the best topsoil in the world. Now it is down to eight inches". The first footnote in the book says only that his source was "conversations with the U. S. Army Corps of Engineers and the Iowa Department of Agriculture and Land Stewardship". One cannot check the Iowa situation with this anonymous quote. (Indeed, printed sources are generally scarce in the book.) But we do know the trend of increasing erosion for the country as a whole. If Gore had done his homework, he would have examined the data in the publications of the U.S. Department of Agriculture's Economic Research Service. He would have talked to Bruce Gardner, now Assistant Secretary of Agriculture for Economics, and to Gardner's teacher at the University of Chicago, Theodore Schultz, who has been watching soil erosion since his days as a farmboy in South Dakota in the 1920s, and who received a Nobel Prize for his work in agricultural economics and human capital. The senator would have read the articles by Schultz, and by respected agricultural economists Earl Swanson at the University of Illinois and Earl Heady at the University of Iowa. Gore would then have found that the facts are exactly the opposite of what he writes. The farms in the United States are becoming less rather than more eroded, on average. Decade after recent decade, fewer rather than more acres suffer from severe erosion. This emerges from comparison of Soil Conservation Service surveys done at intervals since the 1930s. This first item sets the pattern for all the rest. Gore alleges - in the first sentence of the book - that there is a "global ecological crisis", that conditions have been worsening. But in fact all trends pertaining to human welfare have been improving rather than deteriorating. The second specific item Gore mentions is DDT, "which became for me a symbol of how carelessly our civilization could do harm to the world". He gives no data and provides no references, though he later adds that DDT "can be environmentally dangerous in tiny amounts". A touch of research would have turned up tons of writings such as Mosquitoes, Malaria, and Man, by Gordon Harrison, who was director of the Ford Foundation's environmental program. Gore would have learned that with the aid of DDT, "India had brought the number of malaria cases down from the estimated 75 million in 1951 to about 50,000 in 1961. Sri Lanka...reduced malaria from about three million cases after World War II to just 29 in 1964". Then as the use of DDT went down, "Endemic malaria returned to India like the turnaround of a tide". By 1977 "the number of cases reached at least 30 million and perhaps 50 million". Does that suggest that DDT does harm to civilization, or does good? Gradually, it became clear, too, that DDT could be used quite safely. The scary scenarios in Rachel Carson's book, that Gore remembers troubled his mother, turned out to be without foundation. In 1971, amidst the fight that led to the banning of DDT in 1972, the president of the National Academy of Science - distinguished biologist Philip Handler - said "DDT is the greatest chemical that has ever been discovered". Commission after commission, top expert after top Nobel prize-winning expert, has given DDT a clean bill of health, as Gore could find out in Elizabeth Whelan's Toxic Terror and the host of references therein. But evidence on such matters has no place in Gore's book. The third item Gore mentions is Agent Orange, which he uses weasel words to describe as "the suspected cause of chromosomal damage and birth defects". Again no references in Gore's book. And with reason, because though Agent Orange (dioxin) was indeed "suspected" by Gore and his colleagues, it was pronounced innocent in the Federal courtroom when veterans brought suit. There simply is no solid scientific evidence of ill effects from dioxin. In August of 1991, the New York Times front-page headline was "U.S. Backing Away from Saying Dioxin is a Deadly Peril". The story continued, "Exposure to the chemical, once thought to be much more hazardous than chain smoking, is now considered by some experts to be no more risky than spending a week sunbathing". And the Centers for Disease Control now admits that the Times Beach evacuation was unnecessary. But Senator Gore has not gotten the word. Love Canal is next in the book. Gore seems unaware that the solid scientific consensus is that there was no observable damage to humans from living near Love Canal. So far we have only reached page 3. And the entire book is filled with this sort of environmental gossip, backed by no sources, and contradicted by solid data. Though the Senator undoubtedly cares sincerely about environmental and resource issues, his ignorance is willful rather than naive. He has been told in the past that his utterances on these subjects do not correspond with the facts. But he has chosen to ignore the scientific literature. Furthermore, the advisers Gore leans heavily on - Paul Ehrlich and Lester Brown - have been proven wrong in every one of the forecasts they have made in the past two decades, a truly astonishing record of consistency. Yet it is still their agenda that Gore puts forth, almost as if he is writing from handouts of the environmental movement. Moreover, Gore is suspicious about others' motives and behavior. He writes that "The statistics about forests can be deceptive too: although the United States, like several other developed nations, actually has more forested land now than it did a hundred years ago, many of the huge tracts...have been converted from diverse hardwoods to a monoculture of softwood". But the same U.S. Forest Service statistics that showed Gore that the total volume of trees is increasing also show that the volume of hardwood trees is going up, rather than being driven out by softwoods. Just who is deceiving whom? The reader may wonder who is to believed. One crude test is whether people will put their money where their mouths are. So here is my offer: I'll bet a week's or a month's pay with Senator Gore or anyone else that I've got the above matters right and he does not. And I'll go further: I'll bet that just about any broad aggregate trend pertaining to human welfare will improve rather than get worse - health, standard of living, cleanliness of our air and water, natural resource availability - you name it, and you pick any year in the future. First come, first served. It is not surprising that a U. S. senator does not have time for the kind of library digging that an academic researcher does. But that is no excuse for publicizing and acting on wrong facts, because it is not harmless - whether as a senator who has a large say in national policy on these matters, or as the President of the United States that some have Gore be. Gore would (among other measures) tax the use of new raw materials to force more recycling, establish higher mileage requirements for cars, require "efficiency standards throughout the economy" - all of which would raise costs and increase government intervention in people's lives. All this on the basis of beliefs he holds that are utterly contradicted by the solid scientific facts. RACIAL WORRY, FRUSTRATION, BUILT BY WASHINGTON POST The headline across the first and second columns of the front page of the Post on October 10, 1995 was "Worry, Frustra- tion Build for Many in Black Middle Class" and continued in the second and third paragraphs: "[M]any middle-class blacks are more jittery about their futures than they've ever been... The escalating anxiety among the black middle class was one of the findings of a national survey sponsored jointly by The Washington Post, the Kaiser Family Foundation, and Harvard University." The article reported the poll results, and even gave the "Methodology." But I found no backing in the poll data for the statements quoted above. There was nothing that showed trends over time or that there had been a change in any direction. On the only questions that refer to the past, the responses for whites and blacks are almost identical; in fact, blacks are slightly less convinced that things have gotten worse. The scores in the block of questions entitled "I believe, compared to 10 years ago" and then "I am farther away from attaining the American Dream" were: whites 62%, blacks 60%; "That it has become harder to get good jobs": whites 55%, blacks 53%; and "That it has become harder to find affordable housing:" whites 57%, blacks 52%. Hence whatever was true of blacks was true of whites, on this evidence. But the point of the story is that the experience of blacks is different - and worse - than whites. Furthermore, there is a large body of scientific research showing that questions comparing the past to the present are almost always invalidated by the "good old days" bias. Hence the reported data cannot reasonably be interpreted as any evidence at all about changes over the years in the thoughts of blacks or whites. I called writer Kevin Merida and asked if he could supply further data, but he said he could not. He said that the inter- pretations of trends were based on a variety of perceptions by various people. He also noted that as the writer he was not responsible for the headline. I also called Richard Morin, who is responsible for Post polling. He referred me to Merida and said he believed Merida was relying on data from a recent book by Jennifer Hochschild (1995, p. 57, Table 3-2, pp. 62-63). I said that I had spoken to Merida and Merida had not mentioned these sources. I also asked Morin to send me xeroxes of relevant Hochschild data - a reasonable request among social scientists - but he suggested I consult the book. I went to the trouble of examining the evidence in Hochschild's book, Facing Up to the American Dream. But there I found no support whatsoever for an increasing trend in "jitteri- ness." Hochschild says flatly on p. 57 that "Members of both races are as sanguine now as they were four decades ago." Fur- thermore, she shows at length in her Tables 3-1 and 3-2, refer- ring to exactly the same sorts of questions in the Post poll but asked in studies from the 1960s to the 1990s, that blacks are invariably more optimistic than whites. Indeed, the Post's own poll found just that, as cited above. All this is exactly the opposite of what the Post story suggested. I faxed this material from Hochschild's book to Mr. Morin on October 15, but I have not heard from him. If a politician or commercial enterprise made assertions such as the Post made in this story, either without supporting evidence or in actual contradiction to the evidence, the Post would pillory them and/or call for their prosecution. But no law protects the public from a newspaper doing this sort of thing, and indeed, the press always protests that any such legal action would "chill" a free press. Indeed, the press has a special responsibility for accuracy in these matters because it righteously takes objectivity and truth as its guiding standards, which neither politicians nor businesspeople claim for themselves. Yet the Post feels free to proceed as it does in this and in dozens of other similar cases of false bad news that I have collected and sent to editor Leon- ard Downie and to the ombudsman, the supposed protector of the public in such affairs, but with no response. Nor will another newspaper enter the fray and criticize its competitor in these respects; this code of silence closely resembles the behavior of the Los Angeles cops, so criticized by the Post in connection with the Simpson trial. As to non-journalistic writers, who among those who would like to have his/her work printed from time to time in the Post is suicidal or stupid enough to want to address the Post on such matters as these? And no advocacy organization has a stake in raising a howl about such misstatements; no one's interests is harmed except those who care about the plain, simple truth. IN CONCLUSION Where are the watchdogs of the truth? How do they make their barks heard? page 1 /article4 stats15m/November 6, 1996