Friday, May 30, 2014
The Plural of Anecdote is Not Data
NOTE: These seventeen essays were reproduced from my now defunct former blog, Random Notes, as they are going to be cited in an upcoming essay. For the most part they deal with three subjects, "common sense" and pragmatism, organics and GMO foods, and the belief in the inherent purity and superiority of all things "primitive". A few are on other topics, but I think those three cover most of them.
In my last post, "Interesting", I wrote a short anecdote about my wife's pregnancy in response to a rather shoddy documentary arguing that the obstetrics "industry" was centered on either profit making or the convenience of doctors. I was thinking about it today and wanted to make clear, my anecdote was not intended as evidence that doctors behave in any particular way, far from it, all I wanted to do was offer a single counter-example to show that their contention was hardly universally true.
However, after deciding to explain that, I realized that the error about which I was concerned, the worry that others might mistake an anecdote for proof, was in fact the whole foundation of their "documentary", in fact a failing of many documentaries. At their best, documentaries record historical events, or maybe capture a moment in time accurately. At their worst, they degenerate into propaganda, using a few dubious numbers, a handful of experts and a smattering of anecdotes as substitutes for any theoretical basis for their argument.
Sadly, the former type of documentary, that is one which actually documents an event, as the name would suggest, has fallen out of favor among our activist film makers, and we have been left with nothing but the propaganda form of film. And equally sadly, people have come to accept a handful of anecdotes as a substitute for theory, proper proof, or real argument.
We can see this most clearly in political campaigns, not the electoral campaigns so much, as the media campaigns pressing for government actions in a single issue such as nationalized medical care, or the efforts in 2007 and 2008 to make villains of the oil companies. Oh, the media will trot out a few dubious numbers, such as the "fact" that other nations are more pleased with their health care systems, or have longer life spans ("Poverty and Lifespan", "Shocking Numbers", "Misusing Numbers ", "Mathematical Deception", "Best of the Web Imitates Me XXII", "Bad Economics Part 1", "Lifespan", "Why the Numbers Don't Matter"), but mostly they find a dozen sob stories, to "put a face" on the issue, and then treat those stories of "abusive" practices or impoverished seniors as if they proved that nationalize medicine would cure our ills.
The problems with such approaches are legion. First, they lend themselves to the foolish "pragmatics" ("The Shortcomings of Pragmatism", "Pragmatism Revisited", "Pragmatism Revistied, Again") which characterizes so much of our political scene. That is, politicians find a problem, propose a short sighted "solution" and assume that is all they need to do. They never question the cause, never look for a theoretical cause. Never examine the effects of their remedy. They simply grasp at the nearest answer that seems to make sense and think that is the answer. Even when the problem is the result of an earlier "solution", that is never seen as an indictment of this approach, instead it just calls for more myopic "fixes".
The remaining issues are all related to the first. This approach, this use of sob stories with built in "villains" ("Life Without Villains", "Enemies Into Villains", "Rethinking My Earlier Position", "Authoritarian Oil Talk", "Antibiotics, Automobiles and the Free Market", "In Defense of Speculators", "Those Darn Speculators", "Speculators Again?") tends to prevent us from looking for a real cause. Fired up by populist nonsense, we assume the nearest parties, often the richest, are the ones to blame (""Greed Versus Evil"", "Fear of the "Big"", "Beware Populist Deception", "The Wrong Reform"). It is very similar to the way tort law looks for the nearest, richest party. And it has the same bad track record. Most often, despite the stories of "abuses", the people being blamed are actually doing good, providing goods or services, even remedying other problems, but they make easy targets. And so, we end up appealing to envy ("Envy Kills", "Envy And Analogy", "He's Bad So He Must Be Wrong", "Sour Grapes as a Political Philosophy") and letting it blind us to the real causes, which makes it even less likely we will ever identify, much less fix, the real problem.
Nor does the problem stop there. The appeal to emotions also often causes us to forget to ask whether the government should be doing anything at all, or whether it is a problem that is better remedies through some private action. ("Doing Something", ""Doing Something" Revisited", "What We Deserve", "Don't Blame the Politicians", "Who Is To Blame?", "What is Wrong with Us", "The Single Greatest Weakness", "The Difficulty of Principle", "Damn the Torpedoes!") In fact, the sob stories and the "human faces" on the problem can keep us from asking whether there is even a problem at all, or if ti might not be one of those cases of misfortune which occur in life from time to time ("Utopianism and Disaster", "Principles Versus Outcomes", "Life Is Not Fair - And Trying To Make It So Makes Things Worse"), or may even be the fault of the supposed victim ("Subsidizing Irresponsibility and Poor Planning", "The Cycle of Compassion"). By making an emotional appeal based on a handful of sad tales, this approach handily avoids all manner of uncomfortable, but essential questions.
And that is my problem with anecdotes.
Yes, viewed in one sense, all empirical evidence is nothing but anecdotes, so enough anecdotes may eventually constitute data. On the other hand, data alone does not tell us what to do any more than anecdotes do. Without a theory, without an integrated approach telling us what to do, how to act, we cannot do anything but flail about blindly. And in that case, if we are going to act without theories, without an understanding of what our actions are supposed to do, then we don't even need anecdotes. We can act randomly and produce results every bit as good as acting blindly in emotional reaction to sob stories.
And so, in the long run, anecdotes are either misleading, or pointless. And thus they serve no purpose but to allow those who would avoid reasoned argument to do so.
As I mentioned the health care debate, I will list my articles on that topic. They can be found by following the links found in my posts "Redefining Insurance... To Actually BE Insurance", "The Insurance Sham", "Government Efficiency", "High Cost of Medical Care", "Medical Reform, An Overview", "My Health Care Plan", "True Insurance Reform", "A Different Look at "Health Care Reform"", "Of Wheat and Doctors" and "You Gotta Have Faith".
For those familiar with my older posts, the pattern I describe above, blindly picking a solution, suffering unexpected consequences, and then blindly trying to fix those as well, is the same pattern I described in "The Endless Cycle of Intervention". It is also similar to the pattern of escalating welfare programs I described in "The Cycle of Compassion".
Originally posted in Random Notes on 2010/06/05.