You Can’t Trust Official Statistics

Recently by Jack D. Douglas: American Paradise Lost

All official, statist statistics are only about the subjects and categories, and use only operational definitions and procedures to construct those statistics, which are determined ultimately by the politicians who control the state. All official statistics are, thus, inherently biased by the powerful individuals who run the states, so official stats. vary wildly among nation-states and other political groups. In America, for example, there are massive official stats. on the deaths the state categorizes as due to drug uses defined by the state as “illegal,” almost entirely those of poorer people. There are no official slats. on the deaths due to prescription drugs the state defines as legal, even though non-state studies show prescription drugs kill many times more Americans than the illegal drugs. Big Pharma. Corps. and doctors and the politicians they pay-off massively want to control non-pharmaceutifical drug uses to end them, not the immense drug uses of Big Pharma drugs. Only a tiny part of human experience is the subject of official stats., those parts that the state wants to know about in detail to control in some way for statist purposes. There are no official stats. on “paying officials bribes to get profitable legislation passed for the payers.” There are massive official stats. on the knowledge of k-12 students. Where are the official stats. on “Ignorance of Officials.”?

I started my Ph.D. study of suicides by going back and studying the history of official statistical data because almost all social science studies of suicide used the official, statist statistics on suicides uncritically [which no real scientist would do with any statistics]. I went back to the ancient world briefly [there is little info. on that] and began more detailed work on the tax roles from Medieval England, the moral statistics from the 17th century on in major states, and up to the modern states. When you study the origins of official stats. it is perfectly obvious they are created and collected for statist purposes and not for scientific purposes. They are obviously inherently biased by political power. Modern citizens and slaves who live inside these massive, bureaucratic states are controlled massively in a total-wrap-around way by the statist concepts and data the people of power create and use to control the citizens and slaves. As long as people live within those official data, they are being sub-consciously controlled by the people of power running the states.

My earliest book, The Social Meanings of Suicide [originally my Ph.D. thesis at Princeton, published by Princeton UP], reports on my findings from historical and comparative studies about the invalidities and unreliabilities of official reports and statistics on suicide in general. [“Valiidity” concerns the “truthfulness” of definitions of categories in statistics. “Reliability” concerns the degree of agreement among all the people using the definitions to construct the actual statistics.] I then did a study of coroner and medical examiner categorizations of suicide as a cause of death in the counties of New York State and showed the wild differences in definitions and methods used, making them totally unusable as any kind of “scientific data.” This was reported in various of my works, such as American Social Order and Investigative Social Research.

The coroner of Buffalo was very blunt. It was and is a largely Catholic county, so official categorization of cause of death as “suicide” is very controversial most of the time because it leads to not being buried in the Catholic way, etc. He said no death was categorized as “suicide” officially unless a note in the victim’s handwriting was found. This was totally different from the med. exam. system in NY City where the med. exam. has total power. Buffalo rarely did autopsies, but in NYC any unattended death was generally followed quickly by a full-scale autopsy, used partly as “practice” for med. students and interns and residents at NYU at that time, where I was invited to take part in one such autopsy to show me how they inferred suicide. It was very hot that day and I was already queasy, so I begged off. A med. examiner just north of the City told me you could throw all the possible “suicide” reports up and count as suicides those which stuck to the ceiling and you’d be as right as the official categorizations. And she was not laughing. She took it for granted as true after all her experience.

I and my students and many other colleagues did massive studies of official stats on crimes, etc., and showed the same things.

I tend to take this for granted and rarely bother writing about it for any data, especially economic data, though I’ve done that massively decades ago, as in The Myth of the Welfare State. There are no official stats. on social phenomenon that are “scientific.” Even population stats., counting heads, is inherently only roughly approximate when vast money and work is put into it. Fred Stephan, at Princeton, my first thesis chair, had been Pres. of the American Statistical Association, but he agreed with me that population stats. vary wildly, especially in other countries. My tutor from Harvard undergrad. days, Ivan Vallier, was doing some research in Argentina back then and asked to see the official pop. data for a recent period. They gave him the only official book of those stats. and he kept it under his bed in his hotel until returning it. In America that would seem insane to the statisticians. Fred had had similar experience in other countries,

“Suicide” is officially undefined or defined as “the intentional killing of one’s self” almost everywhere, “Intention” is inherently very problematic, then the operational procedures used are wildly different, so the stats. are inherently very unscientific. The individual categorizations vary wildly in their degree of uncertainty. For example, did Marilyn Monroe commit suicide? Unless someone killed her on purpose and comes forward and gives conclusive proof, we cannot know. We can only argue empirically and common-sensically to infer subjective probabilities.

All of these things are true, to widely varying degrees, about all the economic stats. Unemployment data is notoriously invalid and unreliable in the U.S. and everywhere. They change the small-print, hidden operational definitions frequently, normally in part to keep the official rates down. The mass of Labor Dept. statisticians have nothing to do with this. The few top political appointees and “expert” committees they appoint make the decisions, as is usually the case with official stats. If you read the fine print of the “experts,” you can follow these things. Who does? The top “experts” working with the political leaders of all these official statistical organizations can vary all kinds of “small variables” hardly noticed by non-experts — “seasonal adjustments,” periods considered, the wordings and personal characteristics of telephone pollsters, decisions to recall or not, etcetc.

All serious pollsters learn early in their training that face-to-face and telephone and written poll statistics can be changed — biased — widely by the wordings of questions, the tones of voice, the introductory messages, the accents of the pollsters talking on the phone, etcetc. There is a whole literature on this and any good and honest pollster knows that. Seymour Martin Lipset once told me that a good, serious pollster knows that he must know the general situation he is trying to get details on from the poll before he begins the polling, obviously from other, more reliable information about the general situation. So they do pretesting of how different groups interpret different words used in the polling, etc. I always thought Marty was one of the best precisely because he studied the real world with all good data and used polling in that context to get more precise about population proportions, distributions, changes over time, etc.

Huge masses of “data” come in to such agencies and they differ wildly in their apparent qualities. Some is thrown out as obvious trash, mistakes, or even hoaxes. How you define what is acceptable and what is not can make a big difference.

The rules of inclusion and exclusion of data in general, the times for beginning a study [e.g., Jan., 2012, versus Feb., 2012] and ending it, and on and on, make big differences in the official stats. produced by these bureaucracies.

There are good books on how to lie with statistics of this sort and how to defend against the liars. Reporters and the overwhelming majority of Americans obviously never read such books, unless they want to do the official statistical lying and then they always deny these obvious truths and pretend those books and essays like this one do not exist — they disappear the truth from the public, as usual.