Friday, April 27, 2012

About infrastructure

This is based on a comment I made here.

Infrastructure (roads, power, water, rail, ports, airports) is typically a network good or a hub for networks. As such they provide benefits that are not necessarily captured by the network provider. Hence, private providers are likely to underprovide infrastructure. Indeed, infrastructure provision was a major feature of government expenditure up to around the middle of the C20th.

A notable long-term trend in public policy is that the state was better at infrastructure when it was not trying to do quite so much. The problem is, at least in part, that, with the huge expansion in government income support from the 1960s on, infrastructure is also competing with welfare, education, health and related spending much more than it used to.

Another wrinkle is that the rise of permit raj's in land use in the second half of the C20th undermine a fundamental incentive to publicly provide infrastructure. Much of the point of the state providing infrastructure is that (unlike the private sector) it can gain revenue from increased land value and production: it can tax the positive externalities infrastructure provides. Unfortunately, a much cheaper and easier way to generate increased land tax, etc revenue is simply to engage in land rationing (Krugman's Zoned Zone) and reap higher land/property/capital taxes. (And, not coincidentally, makes developers a great source of political donations.) This both raises the cost of infrastructure (by driving up land prices) and lowers the incentive to provide it.

Add in "infrastructure is evil" environmentalism, which turns NIMBY (not in my backyard) into BANANA (build absolutely nothing anywhere near anyone) and you have factors undermining infrastructure provision that are at least as important as any push for privatisation.

On which point, there is a long history of infrastructure provision wavering back and forth between public and private provision, depending on which set of policy pathologies have been most saliently problematic in the preceding period. Many countries have gone through a period when public provision/production pathologies were most salient, hence the swing to private provision, including of infrastructure.


  1. This comment has been removed by a blog administrator.

  2. > the state was better at infrastructure when it was not trying to do quite so much.

    Sure. Declining marginal utility explains that: positing even halfway competent civil servants, the very first projects undertaken will be those that have huge and compelling payoffs.

    By the time the State owns 30-50% of the GDP, only the low hanging fruit is left, and we get citizens taxed to pay for $10 million dollar studies of bike share programs for left-handed eskimos.

    1. You're a bad man, but I like you :)

    2. But my point is sadder than yours; there are high value infrastructure projects which get delayed or even not done because of interactive dysfunctions from too much government.