Childhood as an externality

09/12/2011 § 4 Comments

Bear with me for a moment.

Economic analysis is focused on agents/individuals who produce and consume goods and services. They can also make rational decisions based on some amount of information, expectations, etc.

Children don’t really fit. Yes, they consume stuff, but they are constrained by their parents’ preferences and don’t really face market prices. While there are reasons for their decisions, I don’t think they classify as ‘rational’. Their preferences certainly aren’t stable. They also can’t enter into contracts so are excluded from a lot of economic transactions.

So, childhood isn’t really part of an analysis of the economy. People become economic agents when they become adults or adult-ish. They sort of spring into the economy fully formed.

We could formalise this by treating childhood as an externality. Childhood is to some extent done to the child; we don’t choose our parents or their circumstances. Childhood is also somewhat a by-product or co-product of the economy. Whether a pre-schooler is in daycare, for example, is partly a result of the parents’ employment decisions.

The benefit of this approach is that it gives us a ready-made framework for thinking about the impacts of childhood. We can think about negative and positive externalities — childhood conditions that either tend to hold adults back or increase their opportunities.

And then, of course, we can think about what we can do about them. The market is the first-best solution assuming there are no externalities. Once we add in the externality that is childhood, what happens to that market solution? The presence of the externality isn’t sufficient reason to decide that the market solution is bad, but it means that we should examine it critically.

If we do decide to intervene, we could consider the mechanism to use. For example, agent-adults who are affected by negative externalities could claim compensation after the fact. They could assert that their adult lives have been polluted by childhood poverty, which was inflicted on them. I’m not sure who would pay the compensation. Parents are not solely responsible for a child’s situation; schools, the medical system, communities, etc. might all need to contribute. At any rate, if the parents also suffered negative externalities, then there is a chain of claims.

Alternatively, we could set up institutions to distribute the positive and negative externalities of childhoods amongst the agent-adults. Essentially, some portion of income taxes and welfare payments already represents this sort of transfer. We could more explicitly recognise the role of externalities in contributing to agent-adult outcomes, and thereby tag part of these transfers as a response to the externality problem. Another institutional approach would be to reduce negative externalities, subject to some assessment of the marginal cost. This approach could be an argument for quality public institutions from a strictly individualistic perspective.

Just some thoughts for a Friday. Thanks for reading.


Tagged: ,

§ 4 Responses to Childhood as an externality

  • I get damned nervous about approaches starting from an externality framework. Even leaving the Rotten Kid theorem aside (Coasean considerations making all within-family transactions bargains), how would you define an appropriate baseline level of care? It’s easy to see the jerk parent who beats his child as imposing an external cost that is policy relevant. But surely there are model parents who exceed everybody else’s standard of parenting quality. Do we all impose negative externalities on our kids for failure to meet that standard? Or do we view the kid as owing a big debt to the model parent, no debt to the modal parent, and being owed a debt by the malevolent parent?

    I’ve had to worry a bit about this stuff in thinking about intrafamily effects in alcohol policy work. I rule out pecuniary stuff because, if we allow harms by parents spending money on dumb stuff, where does it end? I took a bottom line of if it’s a crime, it’s a policy relevant cost; if it doesn’t meet the standard of criminality, it’s internalized by Rotten Kid / Coase. I’m not sure it’s the best line in the world, but I think it’s defensible. And I can’t see any obvious better one that doesn’t descend quickly into absurdities when applied to other consumption or behaviour areas.

    I sometimes entertain Veil ideas when working through these things. Imagine that society decides behind the veil that the minimum acceptable level of parenting is x-bar where potential parents have highly heterogenous cost functions for producing parenting quality. Post the veil, some go on to put in more effort, but others decide against having kids. If the Veil holds well, x-bar is set such that the gains in outcomes among those born exceeds the losses imposed on those who fail to be born and who otherwise would have preferred being born to low quality parents to not being born at all. Whenever we decide to up the standard for x-bar, nobody in the real world worries much about the costs on those who then fail to be born who would otherwise have been. That’s easy where x-bar is low enough that expected lifetime utility is zero or negative for those failing to be born, but there’s no way the bar is that low. I don’t know where that winds up taking us, but I worry a bit about it.

    I worry more about how efforts to improve experienced x among those with parents with poor parental quality production functions may wind up reducing the number of children produced by those with better production functions but who have a higher income elasticity of producing children.

    • Bill says:

      I understand your nervousness. Here, let me make it worse. What if we set the baseline at a theoretical level of perfect parenting? We could then penalise parents for the distance between their children’s actual performance and this theoretical frontier. More seriously, yes, I agree that there are issues with both measurement and application if childhood is an externality.
      I’m not sure about the Coasian story, though. My problem with it is that there is no initial transaction or agreement between the parent and child. The child is born into the arrangement. Therefore, nothing that follows can be treated as fully voluntary. Without that aspect to the arrangement, I don’t think we can just say that the effects are internalised.
      Furthermore, it isn’t just about the parent. Some effects of childhood are jointly produced by parents and society. For example, children tend to attend the local school. The local school is a function of the income or wealth of the neighbourhood. The parents’ choice of neighbourhood is a result of preference and their budget constraint. The child doesn’t get to choose the quality of the school; the quality is produced by a combination of the parents and society.
      From the child’s perspective, lifetime utility maximisation might entail moving some resources from the future to the present to produce better schooling (or medical care), raising total lifetime earnings and hence utility. The child doesn’t have that option. Demanding only that the parents don’t engage in criminal activity seems like too low a bar (or too low an x-bar).
      If we include in our analysis hypothetical or potential children with an option to exercise their right to existence, don’t we also have to endow actually existing children with an option to have had better childhoods?

      • Imagine Coase behind the veil though, Bill. There are certainly kids who wouldn’t be born if the parents were required to provide at least x_ effort, but who prefer being born to not for some range less than x_. Note further that the Rotten kid theorem generalizes to rotten parents so long as the kid can wind up being the benefactor in retirement.

        In my case, my children drive me to drink. A policy relevant external cost imposed on me by my kids? Doubt it.

        I suppose it’s not dissimilar to minimum wage analysis that puts weight both on improvement in outcomes among those employed and disutility for those who wind up being unemployed.

        Bottom line on the alcohol stuff, though, absent there being some consistent framework requiring identical standards across domains, it puts a thumb on the scales to talk about the extent to which parents’ alcohol consumption costs their children if we don’t have comparable figures on parents’ expenditures on sports cars, parents’ inadequate attention to the kids’ homework, or all the other margins on which parents might fail their children. I’m happy to take on a higher bar in a more complete framework applied consistently across domains. But we’d need some basis for setting that standard, and it’ll be contentious.

        On school quality, you’ve made a nice case for vouchers funded from general taxation and easy bussing to out-of-zone schools.

      • Bill says:

        Hmm, I obvious need to think more about this. I just thought I’d point out — I started down this track by trying to put together an individual-rights-based argument in favour of charter schools. It’s no wonder I’ve ended up making a case for vouchers.

What’s this?

You are currently reading Childhood as an externality at Groping towards Bethlehem.


%d bloggers like this: