Realism on Infrastructure Investment

Keith Hennessey has an excellent post on government infrastructure investment. Here are his key points:

  1. Capital investment by government often pursues multiple policy goals, some of which conflict with maximizing productivity growth. If you’re investing for long-run growth you’ll invest differently than if you also have goals to maximize short-term job creation and to change the future balance of energy sources to reduce greenhouse gas emissions (for instance). The pursuit of multiple policy goals lowers the expected economic growth benefit of public capital spending.
  2. Geographic politics distorts and often dominates government investment in physical infrastructure. Highway funds and airport funds especially are allocated in part based on which Members of Congress have maximum procedural leverage over the spending bill. Even if you could somehow get Congress to stop earmarking infrastructure spending (good luck), and even if you could rely on the Executive Branch not to allow their own political goals to influence how they allocate funds, local geographic politics would come into play at the state level, since much federal infrastructure spending flows through State governments. This is where reality most falls short of a valid theoretical starting point for increasing productivity and long-term growth.
  3. Non-geographic politics can distort government capital spending. This is principally an Executive Branch concern, as we saw with the Obama Administration’s decision to throw good money after bad to postpone Solyndra’s failure. And rent-seekers come out of the woodwork, looking to leverage their connections to government officials to win infrastructure investment contracts.
  4. Once “investment” is favored, everything gets relabeled as investment. The Obama Administration has been particularly guilty of this; almost every spending increase they propose is an “investment” of some sort. We should allow them some rhetorical leeway, and we should recognize that government has other reasons to spend money than just to maximize future economic growth. At the same time, it’s misleading when they claim that increased government spending that serves other policy goals (some quite legitimate) also increases future economic growth.
  5. There’s a difference between government investments in the commons and government spending that primarily benefits individuals.  A new airport benefits all who use it. A scientific research grant benefits the researcher and society as a whole if his research advances our understanding. A subsidized student loan is an investment in human capital, but the return on that investment accrues mostly to the student and his or her family. That’s not wrong, it’s just having a more limited effect on increasing long-term growth for society as a whole.
  6. Government investment in physical infrastructure is slow. The Administration learned this as they tried to force money out the door in 2009 for “shovel-ready jobs” that turned out not to be there. This doesn’t mean you don’t build roads and improve ports and airports, it just means the short-term fiscal stimulus argument for this type of spending is weak.
  7. Government investment in physical infrastructure is intentionally expensive because of “prevailing wage” requirements, championed by construction labor unions, that mandate the government must pay more for workers than an aggressive private firm might be able to find in the labor market.
  8. We should evaluate the marginal productivity benefits of additional investment. The President sometimes argues that building the national highway system was good for growth, therefore his specific proposal to increase highway spending is good for growth, too. But those are different investments, and we need to examine the marginal benefits (and rate of return) on the specific incremental investments he is now proposing. The transcontinental railroad definitely increased national economic growth, but that doesn’t mean the feds should subsidize a costly California bullet train with questionable growth benefits.
  9. International comparisons of government infrastructure are silly. U.S. government capital spending should be determined based on what will most increase U.S. productivity without comparison to what other countries are doing. If American ports are clogged and that is harming our trade and slowing American economic growth, then we should upgrade our ports. We shouldn’t instead improve our airports because other countries have shinier ones. We have a different geography, a different economy, and different infrastructure needs than does China, or Japan, or Dubai or France. It is crazy to suggest that the U.S. should build bullet trains because China is doing so.
  10. Government investment faces no market discipline. Capital investment in a private firm can face some of the above challenges—a CEO, for instance, might want a new facility built in his hometown rather than where it will produce the highest rate of return. Or a firm might reject an investment that would maximize its’ workers’ productivity because that investment is inconsistent with the firm’s broader strategic goals. But these firms ultimately face the discipline of the market to curb their excesses. Government does not, and in some cases policymakers are rewarded by their election markets to distort infrastructure investment even farther from its growth-maximizing ideal.
  11. Government capital investment financed by raising taxes on private capital investment will slow long-term economic growth. While in theory there probably are government infrastructure investments with very high rates of return, all of the above reasons suggest that in practice the actual rate of return on government-directed investment is going to be lower than in the private sector. If you advocate raising capital taxes (on capital gains and dividends, for instance, as Senate Democrats appear poised to do) at the same time you argue for increased government capital spending, you’re shifting capital investment from the private sector to the public sector. That will slow long-run economic growth rather than increase it.

As Hennessey notes and as I second this is not a denial that “smart government capital investment can increase productivity and contribute to faster long-run economic growth.” Instead, it’s an argument for caution but also for more thought about how to make government investment smarter. See also Tyler’s related comments.

Comments

Comments for this post are closed