The above title could also be applied to any social science (e.g. Psychology--the art of conflating correlation with causation). But when people who help others through government, they rely on inferences drawn from this fallacy. I was watching this video of Senator Jim DeMint talking with reason.tv regarding welfare spending. Both reason hosts caught DeMint in a contradiction. He states that the federal government should address the problems that lead to high school dropouts, including drug use, unwed mothering of children, and other socially conservative targets. In the previous breath, however, he wants to get the government out of education. As impressive as that cognitive dissonance is, what struck me (and should strike you, if you have the 30 minutes to watch the piece) is the way in which policymakers talk about social issues.
I suppose that one nice thing about an endless series of correlates is that policymakers at least understand that social issues are interconnected. That's a nice social work value put into action. But people are not made of correlates, they are made of determinants. The notion that the social problem of high school dropouts should be addressed with legislation (something congressmen of both parties lauded in the President's soon-to-be-forgotten State of the Union address) is often not justified on its own merits but on what dropping out is correlated with.
Of course it is better to graduate high school, but action is only justified when correlates such as illegal drug use, crime, unwed motherhood, lack of insurance, poverty, etc. are paraded as though they are directly caused by the decision to no longer attend high school. Dropouts do not justify intervention to remedy whatever antecedents are found by a biased policymaker. There is a strong case to make that dropouts are, at most, a local issue to be dealt with by schools, parents, and local governments.
Two examples from the above arguments:
1)
-dropouts often come from homes with single parents
-the government must institute policies to promote marriage
-this will reduce overall dropout rates
2)
-dropping out of school is bad and is also associated with poverty, drug use, and mental illness (for instance, not necessarily true)
-therefore, the government must institute policies from #1 to reduce dropouts
-this will reduce poverty, drug use, and mental illness
But correlation is how social policy is made. Take a policymaker's biases, combine them with those of various interest groups and think tanks, and you will find statistics to make any social problem seem like the pivot point for all the world's social ills. It's as if social problems were viewed through a kaleidoscope. In a kaleidoscope, you know all of the images you see are deeply related to one another. However, any sense of where one belongs in relation to another is indecipherable. The myth of perfect knowledge tricks us into thinking that if only we could put our finger down on one part of the scrambled image, it would begin to form a rational picture.
Surely, anyone would see that my example with dropouts is specious (though Sentaor DeMint might disagree). A policy program that was implemented by smart, non-ideological people with the public interest at heart would be able to actually craft an intervention that would address the antecedents of dropping out. A few reasons why that is foolish:
1) No one is motivated by altruism. People are usually rational and self-interested, whether they sit in government or private enterprise. If the correct course for a policymaker is to do nothing, his incentives are to do the opposite. He is beholden to the interests of his superiors, including their vested interest in their own careers (risk-aversion), and moreso to his constituents, or in the case of regulators, the industry being regulated. In addition, he is beholden to the organized labor interests of that industry. He is influenced by how media will cover this regulation, the political effects of his decision, and its place within the guttoral moral reasoning of society. He is not of a singular mind or noble, but a deeply conflicted man beholden to various stakeholders.
2) No one is omniscient. When policymakers decide to intervene in an industry such as education, they look only at the main effects of their policy (and not very hard, either). They do not look at what the effects of their policy will be on other areas. Those costs cannot be factored into the models evaluating this policy, as they can never be known. More importantly, it is rarely ever demonstrated that the policy intervention will actually have its intended effect. We would need perfect knowledge of the entire social system to actually know whether this policy will have anywhere close to the effect we want. In our sea of correlates, we know that the variables being manipulated are related in some way, but we do cannot know the full story of how they interact. Policy is awash in reciprocal determinism.
But no one, not even those who ostensibly want to curtail the size of government, truly wrestle with these issues. So, we are left with people like Senator DeMint talking out of both sides of his mouth. Promoting policies that favor marriage so we can reduce dropouts and in turn reduce crime and drug use. Or, with those same data points, policies that discriminate against single or same-sex couples, imprison youth in sclerotic schools, and do nothing to address larger social problems. The hubris inherent in claiming to know the effects of policy is extraordinary. It dehumanizes the process and serves only to insulate the decider further from the actual main effects and especially the unintended effects she has caused.
Policy is made for cyphers, not people. Cyphers are being of potential. They have probabilities, but no cause. They are not logically related. To them, people are like electrons in the quantum realm. With public health, we risk making our world like the quantum world-- an endless mindfuck of unending correlates stuck in a theoretical vacuum of no causation.
No comments:
Post a Comment