Adam Gurri has raised a number of good points about Andrew’s post, and I want to offer some remarks of my own.  Some of these are versions of points I raised in the comments on Andrew’s first post on this topic three years ago. I also draw on some material from my forthcoming book on classical liberalism and the family.

First, let me defend Andrew’s position as a legitimate one in the context of classical liberalism. Assuming one is not an anarchist on principle, then one has to answer the question of what the state should do and where its limits are. If one believes that the state’s job is to prevent harm to third parties, and that’s a key part of Andrew’s argument, it is not a priori out of bounds to ask what the state could do to protect children from harm by parents and others. And once one asks that question, the notion that the state might require some sort of competence test for parents is not absurd. It might be a bad idea, and I’m going to argue that it is, but in principle, it’s not evidence of Andrew’s apostasy, particularly in the fairly nuanced way he developed the argument.

That said, I think it’s a bad idea. I think the argument Andrew makes is a perfect example of what can happen when philosophers (and others) make arguments about what government should do in the absence of serious consideration of the way in which the state actually works. The philosophical argument has an implicit premise on which much of the case rests:  “if the state worked just the way we philosophers say it should…” Economists were guilty of this same error for most of the 20th century (witness the failures of socialism, Keynesianism, and a whole variety of regulations as governments just wouldn’t or couldn’t do what our blackboard models said they should). The problem is that real-world political actors have incentives that lead them to act in ways other than those that philosophers and economists think they should.

Just as economists, including strongly free market ones like me, can point to all kinds of ways in which markets “fail” in the sense of not producing optimal outcomes, so can we point to all kinds of ways that parents do, or might, fail at raising their children well, or even neglecting or abusing them. But the existence of those market and parenting failures is not an ipso facto case for government intervention, even licensing. That is, government failure is just as real as market failure and parenting failure. The result is that the best laid plans of parenting licensers will inevitably fail.  And that means we have to ask the important question of “as compared to what?” when we see evidence of parenting failure.

Imagine a parent who fails the licensure test. If the next step is to remove the kids from the home, we have to ask “and do what with them?”  Is the alternative that the state will offer for the children really better, on net, than their current situation?  Suppose that alternative is foster care.  There is enough empirical evidence on the problems with foster care, especially short-term placements where the incentive to really behave as a steward for the child is weaker, to be skeptical that it would be an improvement.  When we account for the psychological effects on younger children of being taken from their parents and placed with strangers, the comparative analysis suggests the case for intervention is even weaker. And if the case for such intervention is weaker, the case for licenses is weaker.

Just as we might ask with apparent “market failures” whether there are ways to improve the rules of the game such that market outcomes might be even better, or whether there are other forms of voluntary cooperation other than traditional private property rights that might address the situation (think of the work of Lin Ostrom here), we can ask in the case of “parenting failure” whether there are other institutions that could be brought into play to help these parents perform better (e.g., a religious institution, a neighborhood group, extended family members, etc.). Rather than think immediately of state solutions involving licenses and potentially removing children, why not work to help parents become better by finding ways to strengthen the role of these other institutions?

In such cases, the sorts of civil society solutions noted above are far more likely to be appropriate than removing the kids from the home of what are otherwise well-intentioned parents. One of the problems facing state intervention is knowing all of the fine details of each particular case sufficiently to come up with a solution.  Andrew says children of parents who fail the license test should be “put up for adoption by someone licensed.” Again, this sounds good as stated, but the reality of the process is that this doesn’t happen immediately (there are no spot markets in parental rights). Is the child really better off in the state’s hands in the meantime?  In general, those closest to the family are in the best position to understand the problems at hand, imagine an effective solution, as well as having an incentive to act on that knowledge.  Bureaucrats with dozens of cases or more are unlikely to come close to the knowledge and incentives possessed by those in the family’s local sphere.

Beyond the comparative question, there’s another political economy question Andrew’s proposal overlooks. He seems to think we will gather philosophers and parenting experts to write the license exam making it an adequate measure of parenting ability. Assuming that we actually know what makes for competent parenting (and I’m more skeptical than Andrew that we have agreement on that), real-world licensing processes work are much more messy. If the state is creating licenses, the exam creation process will be a political one, and the opportunities for Bootleggers and Baptists problems will be enormous. In short, everyone with a pet issue or a financial gain to be made will be lobbying the process to see their particular concern added to the test of adequate parenting. One need only think of the controversy over Lenore Skenazy’s Free Range Kids and the recent stories of parents being arrested (arrested!) for letting their kids walk alone to a nearby park. The news is full of such stories, and you can bet that not just the “Baptists” who think that children require bubble wrap and that sexual predators are around every corner, but also the “Bootleggers” who stand to make big money depending on what counts as adequate parenting, will be frothing at the mouth to get that license exam look they way they want. That’s the reality of what it means for the state to issue parenting licenses.

That reality means we have to live with imperfection. As much as we all want to prevent harm to children, and there certainly are bad parents out there, we cannot eliminate all of that harm. In the end, we must compare the imperfect behavior of parents with the imperfections of the political process. I would argue that in the case of parenting licenses, the imperfections of politics raise a much greater danger than do the imperfections of parenting, especially when we get outside the market/state dichotomy and look toward civil society for help in improving parenting. The instrument of the state may look like a finely honed scalpel in the hands of philosophers, political theorists, and economists, but in the hands of the political actors who make the actual decisions it is more like a blunt, rusty saw that will not just fail to excise the problem, but also cause a whole bunch of new ones in the process.

Arguments that don’t take the reality of the political process into serious consideration end up with proposals like Andrew’s that look good, perhaps, on paper but turn out to be awful when implemented.  Parenting failure is bad.  Government failure is worse.