The stimulus - spend, invest or incentivise?

Hal Varian in the WSJ (via Mark Thoma and Marginal Revolution) has touched on a topic I have been thinking about for a while: how is the fiscal stimulus best spent? On consumption or investment?

There are essentially two tools available for the stimulus: tax cuts and government spending. And there are five main sources of demand in the economy: private consumption, private investment [optionally divided into business and residential investment], government consumption, government investment and exports. I am not going to address all ten combinations, but focus on private investment - should we promote it, and if so, which are the best tools to do so? I am a priori neutral between tax cuts and spending; tax cuts are good because they let people allocate spending by efficient private choice; spending can be good if it achieves public goods that are not best purchased in the marketplace.

My intuition, like Hal's, is that private investment is important. But is there a clear argument for this? Here's an attempt to work it out.

Arguing for increased investment:

1. Private investment is likely to be artificially depressed at the moment. Many companies cannot raise long-term debt, which is the most common way of financing investment.
   
2. Investment should have the same multiplier effect on demand as consumption spending, but has the benefit of increasing long-term growth as well.

Against that, however:

1. One could argue for reducing investment in a recession if we agree that it's beneficial to smooth consumption. Crudely speaking, if you're starving it's OK to eat seedcorn.
2. Investment means resources are used for future instead of current consumption. Although investment might support employment (or allow it to be reduced by less), the recession might then show up in reduced living standards instead of unemployment.

Both of the 'anti' arguments apply if total demand and GDP are reduced below potential. If this year we can generate only 90% of the goods and services we did last year, then we may want to reduce investment in order to still spend the same amount on food and holidays. But if, by stimulus, we can maintain total demand at roughly its potential level, then there is no need to cut investment in order to boost consumption. If investment is at a normal rate, consumption will naturally be at a normal level too.

Government can't really choose an optimal level of investment - it depends on the time preferences of consumers as well as the existing stock of capital. Japan and Germany, for example, seem to have much better public infrastructure than the US - and so for them, it might make more sense to boost consumption than investment. Also, different populations have different preferences for future versus current consumption.

Therefore, it seems reasonable to take the average long-run rate of investment in each national economy as a target for the future. If actual investment is substantially below that, we can seek mechanisms to increase it. If it's at its normal level, then perhaps consumption should be boosted instead.

To help choose a mechanism, consider two reasons that the stimulus should work.

The main rational purpose of the fiscal stimulus is to directly bring unused resources into production, so that we avoid the deadweight loss of people and capital sitting idle for a year or two. But an important psychological argument for it is to give people and companies confidence that there will be future demand for their services, discouraging them from retrenching. In this way, agents choosing to buy things will create additional demand in the economy on top of the direct stimulus.

(Note that this argument has nothing to do with debt and savings rates. These are purely monetary effects which have no direct impact on resource usage or total production in the economy. Of course the existence and structure of debt and saving is one mechanism which affects how resources are used via signals and incentives; so it is important to understand it. But the amount of debt (public or private) is not relevant to the balance between investment and consumption - not least because every debt is an asset, so the total amount of debt in the world is zero.)

So an ideal mechanism for the US and UK (where investment does appear to be depressed below long-run levels) would do the following:

• Identify underemployed resources
• Identify investment opportunities with the greatest net present value
• Incentivise capital, or directly provide capital, to bring the chosen resources into use to implement the chosen opportunities.

Naturally, a free market is usually best at doing the above jobs. But if resources are underemployed (which is clearly true) and investment is not happening (which is probably true), then is the market doing its job?

There are two answers to this:

1. There are no profitable investment opportunities at present. Therefore the market is acting rationally in not investing.
2. There are profitable investment opportunities, but something is preventing them from being exploited by the market.

Answer 1 seems illogical. If the resources are unemployed, surely there should be a market price they can take, at which investment becomes viable. Sticky prices and minimum wages may possibly stop this from happening, but this would only be true if there had been so much overinvestment in the recent past that all profitable opportunities have been used up. But does anyone think we have had a splurge of overinvestment in either the US and UK this decade? On top of this, interest rates are so low that some investments which were not viable last year should have become profitable now.

Answer 2 is more convincing. Here are some factors that may stop the market from taking effect successfully:

1. Lemon projects. Famously, due to asymmetric information, the used car market can never achieve its potential. Buyers fear they will get a lemon and will only pay a discounted price; the seller knows the quality of the car much better than the buyer, so there is an adverse selection problem and only low-quality cars will be offered on the market, requiring a further discount. The same effect may be at play with potential investment projects. Although it is clear that some investments must be profitable, there is much uncertainty about which ones, and therefore investors are staying away altogether.
2. Uncertainty in capital markets. Although the total amount of money available in the market is bigger than ever, due to quantitative relaxation by central banks, nobody knows how long it will stick around. Banks are not renewing credit lines or extending new long-term loans because they can't know whether their own finance will be renewable. They must know that some money will be available somewhere; but nobody knows if they will have access to it, and so everybody reduces lending.
3. Transaction cost of reallocating resources. It could be that the underemployed resources are not the ones needed to implement the most profitable investments. Retraining takes a long time and is risky; resources may be in the wrong location or be otherwise unsuitable.

Probably all three of these factors contribute to the problem. So, once again, what mechanisms would help address them? Our candidates need to meet some or all of the following criteria:

1. Any solution that provides long-term predictability of new demand, either aggregate demand or demand for specific goods and services, should assist in resolving problem 1.
   
2. A solution providing guaranteed availability of long-term capital should help with problem 2.
   
3. Choosing the character of demand (under criterion 1) to be compatible with the underemployed resources in the economy will help with problem 3.
   

And, conveniently, there is a mechanism which can achieve precisely this, under certain conditions. The Obama "green stimulus" package.

However, I believe it needs to be backed up with certain complementary actions. Therefore, I propose the following structure for that package:

• Guaranteeing long-term demand by legislating for a future carbon tax at a known rate. This will provide incentives for the private sector by guaranteeing profitability of well-designed environmental investments. Joining the Pigou club, as Greg has it.
• Committing future carbon tax revenues to provide long-term investment finance. In the current environment, only government, central bank and a limited range of long-term private savings schemes (primarily annuities) can guarantee long-term finance. The central bank will prefer not to do so due to inflation risks, while private savings are already being mopped up largely by Treasuries. Government's ability to issue long-term debt may be constrained by its growing deficits, unless it credibly commits to future tax increases - the carbon tax could provide that commitment.
• Providing limited micro interventions to assist the incubation of the new sector. These need not be substantial, but there are certain public goods - such as education, or management and pricing of public commons - which can be most efficiently done by the state. Anecdotally, it seems that many resources which were employed in residential investment over the last seven years are among those which are now underused and would be ideal candidates to participate in this program.
• Direct some of the fiscal stimulus towards tax cuts and spending which directly supports this sector. A tax break for relevant investments is one option; spending directly on environmental goods is another. Note the balance between the stimulus, which is necessarily short term, and the carbon tax, which provides its long-term counterpart.

Undoubtedly there are other suitable mechanisms too. But this one - provided the carbon tax is credibly instituted - seems to meet the criteria so well as to be irresistible. And the logic of this package may provide the only politically credible chance to get the carbon tax in place for the next decade.

Update: Useful data from the Atlanta Fed about investment levels, with a somewhat similar analysis of the reasons investment might be falling.