Wednesday 11 May 2011

Fiscal externalities of population

Bryan Caplan's changed my mind again. I'd always expected that the overall fiscal externality of having more children was small and, perhaps, slightly negative. Public schooling isn't cheap and neither are the per student subsidies for tertiary education. But I'd not thought too hard about it. I choose to mood-affiliate with Julian Simon and consequently tend to expect that increased population is generally beneficial; I'd just not expected that the effect on the government's coffers was particularly positive on average.

Bryan cites Wolf et al's finding that replacing the average non-parent with the average parent provides a $217,000 net fiscal benefit: total taxes paid by the children less total government services consumed by the children. Since lots of government services like national defence are non-rivalrous, having a bigger population reduces the average burden. The average additional child confers $83,000 in additional net fiscal benefit. The greater is the proportion of spending on rivalrous rather than non-rivalrous goods, the lower will be the fiscal benefit: the numbers for New Zealand would have to be lower since more of our spending is on transfer payments relative to defence.

Bryan argues for child tax credits to encourage greater fertility. Relatively small baby bonuses can push folks at the margin into having an additional child - Bryan says about $9000 would do it, citing work by Kevin Milligan. If folks whose children are expected to have lower than average positive fiscal contributions were more responsive to the tax incentive, the policy's consequences would not be what Bryan's expecting. But the Milligan paper finds the opposite: richer couples' fertility was much more responsive to the Quebec tax credit. There's reasonable heritability of income generating capacity; the baby bonus would increase the average positive fiscal effect over time rather than reduce it.

There's still something weird in the Wolf et al paper.
Nonparents pay about $327,000 more in taxes than they receive in benefits, while parents pay about $278,000 more in taxes than they receive in benefits. The offspring of the parents, in turn, pay over $266,000 more in taxes than they receive in benefits. In each case, the excess of tax payments over benefit consumption results partly from the exclusion of pure public goods from the benefit side of the balance sheet: whereas all tax payments represent a reduction in private consumption possibilities, only the “excludable” part of public expenditures belong in these accounts.
Suppose that the "non-excludable" part of public expenditures totals $300K per capita, giving a slight overall deficit. If those expenditures are invariant to new persons' entry, then we're right to exclude that portion since the benefits received by the new persons come at zero marginal cost. The bulk of the new entrants' taxes help defray the average cost of providing a fixed set of public goods.

But why should we expect that government expenditure on nonrivalrous goods is invariant to population? If spending instead is elastic, then we'd need to count the new spending on rivalrous goods for which the new person's entry were responsible. Depending on that elasticity, we could quickly get back into the world where the fiscal effect is ambiguous.

I'm generally very nervous about taxes intended to rectify fiscal externalities. If the numbers came out showing a negative fiscal effect of additional children, would Bryan recommend per-child taxes? Suppose that more disaggregated data showed that the negative fiscal externalities of children born to poor unwed mothers were large and that abating subsidies for such births weren't sufficient to counteract the negative effect. Taxes? Regulatory measures? Note that the Wolf paper shows a much smaller positive fiscal effect for parents with less than a high school diploma; the required responsiveness of government spending on nonrivalrous goods to increased population need not be large to there flip the sign of the effect.

The permissible boundaries of taxation, regulation, and subsidy effectively disappear if policy needs to correct fiscal externalities.

No comments:

Post a Comment