Nonsense from the New York Fed.

This short article by Andrew Haughwout of the Fed and friends was recommended by Mark Thoma, so I had a look. It’s a laugh a minute. It packs more nonsense in its 600 words than I could manage if I tried. The authors argue that expenditure on transport should be part of any stimulus package, and they give four bullet-pointed reasons, the first three of which are very defective, and which I’ll take in turn. Reasons (quoted verbatim) are in green. 

Transportation spending produces long-lived investment goods whose benefits at least partially offset the cost of the investment, regardless of the stimulative effect.

“Partially offset”??? What on Earth is the point of an “investment” the costs of which are NOT COVERED by relevant “benefits”? Put that another way, Hogwort (apologies to J.K.Rowling) is advocating investments where costs exceed benefits. Economists the world over will be fascinated to know what the net benefits of that sort of “investment” are. Please, please, Hogwort: can we have more of your pearls of wisdom on this subject?

Reason No.2 is:

By improving long-term productivity, public investments can raise long-term growth expectations, thus providing additional stimulus relative to nonproductive expenditures.

Hm…. So the purpose of an investment is to improve “long-term growth expectations” rather than long-term growth? If that’s the argument then we could possibly save loads of money by not actually investing in infrastructure, but rather by inventing some sort of new psychedelic drug which gives everyone the ILLUSION or EXPECTATION that growth will improve. If the drug gives them mind blowing sexual fantasies at the same time, so much the better: that’s what I say.

Plus, what’s the phrase “long-term” doing there (apart from padding out the article)? I mean any properly thought out investment will bring benefits in the SHORT TERM (i.e. from the moment the investment is completed) as well as the long term. That is, if the investment DOES NOT bring net benefits as from the moment it is completed, that’s strong evidence the investment should be delayed till such time as the investment DOES BRING net benefits from the moment its completed. Every small business owner understands that point.

Reason No.3 is:

Accelerating already planned expenditures, a hallmark of transportation stimulus packages, adds less to public debt than creating entirely new programs.

Now hang on. The amount “added to public debt” is directly proportional to the stimulatory effect: i.e. (roughly speaking) the number of jobs created. Put that another way, maybe you can spend less by “accelerating already planned expenditures”. But then the stimulatory effect is less. Of course the relationship between, 1, number of jobs created, 2, expenditure and 3, addition to public debt is not exactly the same for all types of spending. But absent detailed research which shows exactly what those relationships are, the best working assumption is that the relationship is the same for all types of spending. Indeed, the authors of the Fed article themselves advocate spreading stimulus spending widely precisely because we can’t be sure what the multiplier is for different types of spending.

Reason No.4 is:

Given the wide range of multiplier estimates for various types of spending, a diversified approach is valuable and construction spending can be one component of a diversified portfolio of stimulative measures.

Gosh: some sense at last. Also the authors do have one worthwhile suggestion, namely that we should aim for “accelerated spending during recessions”.  I.e. public sector construction projects and other forms of spending should have some sort of flexibility built in, so that the relevant public sector employers can take on more people come a recession. I’ll back that. Thought of course there is the problem of public sector unions kicking up a fuss when the number of public sector employees is reduced come the recovery.