Cause Ineffectual

The Lived-Experience Fallacy

Let's say I made the argument that smoking causes cancer, and that I backed up this argument with a mountain of scientific data. Now suppose that someone responded to all this with the following: "But my grandpa Bob smoked cigarettes all his life and never developed cancer! So smoking doesn't cause cancer after all!"

Would you be convinced by this reply? I hope not. Smoking is a contributory cause of cancer: those who smoke have a much higher likelihood of developing cancer than those who don't because the act of smoking contributes something toward that outcome, even though that outcome doesn't always happen. So just because some smokers don't develop cancer doesn't mean that smoking plays no role in causing it.

I frequently use this example when teaching causal reasoning in my logic and critical thinking classes. The point behind the example is that personal anecdotes do not invalidate statistical generalizations, which are by nature probabilistic. Most students have no difficulty seeing this point, likely because the link between smoking and cancer has been made abundantly clear to them. Yet students will often turn around and commit this error later on when talking about issues in which they might have a personal stake.

For example, in response to the claim that marijuana use increases the likelihood of developing certain mental illnesses,¹ students will sometimes cite the fact that they have personally used marijuana without developing mental illness. Yet these experiences are irrelevant. Even if it turns out that marijuana use isn't a risk factor for mental illness, citing one's personal experience with marijuana does absolutely nothing to show that. This is because we are dealing with statistical probabilities.

Another example: in response to the claim that children raised in single-mother households fare worse compared to those raised in two-parent families,² students will sometimes cite their own success stories being raised by a single mother. There is no doubt that these examples exist, but they do not falsify the statistical generalization that single-mother households on average fare worse. Affirming this does not detract from the dignity of these students or their parents.

In fairness to my students, it's an easy error to make when it concerns something you're invested in, which might explain why it's so widespread. These days, we see this kind of fallacious reasoning (which logicians call hasty generalization) at work in the appeal to "lived experiences" as a special source of knowledge.

Woke activists often use lived experiences as evidence of widespread injustice. Yet basing one's entire case on lived experiences is, quite simply, bad statistical reasoning. Why should one's personal experience of racism carry any special weight? Should the experience of the smoker who never developed cancer also carry special weight? What about the experience of the drunk driver who managed to get home safely? None of these experiences carry any authoritative weight, especially on matters of policy. As the old legal maxim goes, "hard cases make bad law." We can say the same when it comes to lived experiences: lived experiences make bad policy.

The point isn't that all experiences of racism are like those of the lucky smoker, nor is it to cast real experiences of racism in a negative light. Rather, the point is that one cannot prove or disprove generalizations simply on the basis of personal experiences. This is a basic rule of statistical reasoning that seems to have been lost on many people who should know better. 

Notes
1. Marie C. McCormick et al., The Health Effects of Cannabis and Cannabinoids: The Current State of Evidence and Recommendations for Research (The National Academies Press, 2017), ch. 12.
2. See Sara McLanahan and Isabel Sawhill, "Marriage and Child Wellbeing Revisited," The Future of Children 25:2 (Fall 2015), 3–9.