Mathematics and Statistics as a Tool of Propaganda
This article argues that pure scientific research and its results can be presented either in a way that misguides or presents garbled information or, in the worst case, lies. The success of such practices lays in people's boundless credulity that can be characterized by the opinion that everything that comes from mathematics, statistics, or any other exact science is irrefutable and thus acceptable. It appears that it is not only non-democratic societies that deal with such a phenomenon, but even in everyday life of a democratic society. The article points to examples that are in the interest of groups persuading others with misguiding arguments. It also shows examples of such misleading action performed without suggestion of discussing another view of a particular problem.
Webster's Universal Dictionary and Thesaurus (1993) defines propaganda as follows: the organized spread of ideas, doctrines, etc. to promote a cause; the ideas, etc. so spread.
This definition characterizes history's most preferred type of action
fulfilling individual will of manipulating people or getting them on
someone's side. It lasted until now and developed itself a huge amount
of instruments for achieving its goals such as leadership, taking advantage
of mass psychosis, national sentiment, then sententious public appearance.
Many times, the facts presented were not so important. However, in today's
self realizing society, people cannot be fed with empty words, they
have to be given facts. We all can remember the old statement: If I
had one chicken for lunch and my neighbor had nothing, together each
of us ate on average one half of a chicken.
Exact and thus right
The role of exactness comes up. Mathematics including statistics, is a tool that most people cannot argue with. "I am not lying, the numbers show this." is a good example of alibism that certifies honesty.
However, if we assume that mathematics is an exact discipline and say that the concrete data cannot be presented in obviously contrary ways, a question comes up. Where does the difference come from? The answer is that mathematics actually doesn't always behave exactly. When automatic calculators and then computers developed, people could process more data than ever. In our century, various mathematics areas were in focus, among them statistics and also e.g. the theory of probability. These didn't bring direct inexactness, but gave more freedom to statistician's hands. It lay in his opportunity to choose initial conception. Thanks to statistics' complication for the general public, it turned out to be very easily abused.
Neal Koblitz mentions one interesting event. Several years ago, an
equation appeared in a TV discussion called the "Tonight Show". It happened
during an interview with Paul Ehrlich, author of a book about the population
bomb and a propagator of controlled natality (birthrate) as a guide
to solving all global problems. In that time, people were starting to
be concerned about ecological problems and justified that the solution
is in regulation of natality. Johnny Carson, the host of the show, was
in his best condition, so he could control the direction of the discussion
at any time, for example if his guest would step into close details
or into too serious discussion. However, Ehrlich solved it perfectly.
He took a piece of pasteboard and wrote for the viewers with capital
D = N x I
In this "equation," he explained that "D means total damage caused to environment, N is number of inhabitants and I the influence of each individual on environment." This equation shows that more people mean bigger pollution. He said: "We cannot get environmental pollution under control, if the number of inhabitants would uncontrollably grow". Johnny Carson looked at the equation, scratched his head, remarked something about how he never understood math and that it all seems quite impressive.
Who could contend with an equation? An equation is always exact, undoubted. Neal Koblitz hit the nail on the head when he said that to declaim someone who can support his statements with an equation is the same as to go against a mathematics professor. So nobody asks any questions. But did Ehrlich say that "I" is the same in case of a president of a chemical factory and for example you or me? The viewer is too terrified by a mathematical equation to ever use his common sense.
Mystification, discouragement, invocating an impression of precision and soundness - the effects of using equations in previous example. This approach can be found at universities, too. Neal Koblitz gives a magnificent example where an effort ends up in total absurdness.
Professor Robert W. Fogel from the Department of Economics, Harvard University, was using quantitative methods in history of economy. In 1974, he and Stanley Engermann made a sensational outcome using statistic arguments based on a wide amount of data processed by computer that showed (or sounded like) that the slavery system in the South was not only more humane, but also economically more effective than the system of free manlabor that existed in that time in the North. Although this thesis was controversial with the conclusions of all important traditional historians, the publication was accepted with cheers.
Initial applause lasted for quite a long time. Then the historians of the slavery era and specialists for using quantitative methods in history started to analyze the work. They found such an accumulation of big errors, doubtful assumptions and unqualified usage of statistics that the whole project became worthless. The work said that an average slave on a plantation owned by a farmer named Barrow got 0.7 whip lashes per one year. In the first place, the number is too low, because it was based on a wrong number of slaves, as well as the number of lashes Barrow made on a slave. However, more importantly, this number does not represent the most significant influence that the lashing had. Lashing, as well as lynching, was a tool of social discipline with the aim of impression not only on the actual victim, but also on all that were actually witnesses of such act or have at least heard of it. So the question really is: How often did Barrow's slaves have to look on lashing of one of their mates? The answer is: once in 4.5 days. Once again, the form in which the data are expressed gives them this or that meaning. If we would express the frequency of lynching like Fogel Engermann did for lashing, we'd get that in 1893, only 0.00002 lynches occurred per one black per one year. It is clear that an uninformed reader will never understand the historic consequences of the lynching of 155 blacks that really occurred in 1893.
Delusion of average
Let's return to the present and point out the danger of distortions and suppressing the importance of facts that still can happen. Imagine a company ABCD, Inc., with 500,000 shares which are distributed like this: One majority owner owns 378,900 shares, 10 smaller owners have 10,000 shares each, 100 owners have 100 shares each, 1000 people possess 10 shares each and finally 1100 people somehow got 1 share each. We can honestly declare the following fact: the average stockholder owns 226.14 shares. (He is quite "in the business" - at least the arithmetic average shows this.)