Wednesday, May 30, 2012

Government investment: timing vs. levels


It is often claimed that the governments of the United States, the UK, and Germany should spend more money because they can borrow at low rates, thus raising the present expected return on the investment considered as a whole. 
Maybe, but keep in mind that the interest rates on quality government debt are down, in part, because the risk premium is up...You might think the government investments are “low hanging fruit” in terms of quality.  Maybe yes, maybe no, but the low real interest rate doesn’t signal that, rather it signals merely that people expect to be repaid. 
In this argument for more government investment, the notion of government investments as low hanging fruit is doing a lot of the work.
I think that Tyler may be conflating an argument about levels of investment with an argument about the timing of investment.

Suppose the optimal level of government investment over the next decade is $2.2T over the next 5 years (Note: I did not make this number up). That still leaves open the question of timing. Should we do all $2.2T next year and then zero over the following 4 years? Or should we space it out evenly? Because capital (which you create by investment) is durable, we have a lot of flexibility in choosing when to make investments, even if how much investment we need is fixed.

The correct timing depends on the interest rate. Remember, in a world where government capital is not a perfect substitute for private capital, government bonds are net wealth, and there is no Ricardian Equivalence. So it's best to do your investing when government capital is cheap. In other words, when the interest rate on government debt is low. For the purposes of timing, it doesn't matter why the interest rate is low. If the interest rate is low because of risk premia, that's fine. It's still a great time to do whatever investing that you need to do. This is true even if the current average level of public investment is just right (or even if it's slightly too high!).

So the interest rate argument for government investment is all about timing, not levels. "More investment right now" does not necessarily mean "more investment overall". If you need to make some hay, the correct time to make it is when the sun is shining.

Update: Tyler writes to point out that rates are expected to stay low for quite some time, weakening the case for shifting investment forward. This is true, but I think there are good reasons to believe that rates will not go any lower; even if the yield curve is pretty flat right now, the low level of interest rates probably limits the downside risk.

19 comments:

  1. It drives non-economists bonkers when people like Cowen ignore half the engineering and financial equations to enable them to count angels on the head of a theoretical economic pin. There is a real cost to individuals and businesses from declining infrastructure that needs to be entered into the equation. It's way more expensive to wait until things collapse, or is it the assumption that we will do without certain bridges, tolerate excessive shipping delays etc?

    From the ASCE infrastructure report card: "Americans spend 4.2 billion hours a year stuck in traffic at a cost of $78.2 billion a year--$710 per motorist. Roadway conditions are a significant factor in about one-third of traffic fatalities. Poor road conditions cost U.S. motorists $67 billion a year in repairs and operating costs--$333 per motorist; 33% of America's major roads are in poor or mediocre condition and 36% of the nation's major urban highways are congested. The current spending level of $70.3 billion for highway capital improvements is well below the estimated $186 billion needed annually to substantially improve the nation's highways. (footnotes in link provided at http://www.infrastructurereportcard.org/fact-sheet/roads

    And, incidentally, why did you select a picture of a Grumman Tigercat fire bomber kit as the heading for your previous post. Perhaps to use a partially assembled plastic kit that looks like a tiny version of the real thing yet has no function other than decorative is an appropriate analogy for some economic models?

    ReplyDelete
    Replies
    1. add the locks on the ohio river, which are resulting in costly delays of bulk deliveries (ask american electric power) and the US texas ports, none of which will be large enough to receive cargo from asia after the panama canal expansion is completed..

      Delete
  2. Are you saying that the US/UK govts can now borrow for nothing in real terms? So they might as well so so ? I'm struggling to see why this is controversial. If Shell/Exxonn/Microsoft/ the very very big company could do so, they would. Why shouldn't govts? They only have to do something half sensible. (I'm leaving out multipliers and technical stuff.)

    ReplyDelete
    Replies
    1. Yes. But the de facto Fourth Branch of government in permanent session on K Street has an ideological objection to government spending on anything other than subsidies to big corporate interests. Obviously the bridge builders need to hire better lobbyists, or maybe just break class-solidarity with the other big lobbies.

      Delete
  3. Jamie Dimon8:07 PM

    Noah,

    I read your fantastic blog post. My recommendation is to take 25 Bn taxpayer dollars and bet it on red, err these CDO thingies I saw in a great i-bank presentation. guaranteed income!

    ReplyDelete
    Replies
    1. This is a good concept for a "Monk-E-mail". You'd just need to string along the joke a little while ...

      See for example:

      http://www.careerbuilder.com/monk-e-mail/default.aspx?mId=33630934.2&cbRecursionCnt=8&cbsid=63603d68101744c6a6b756ba69e4d332-334773940-RI-414:25:31

      Delete
  4. The assumption that investment should be done when rates are low appears to imply that the rate of return on available opportunities will be the same indefinitely. I would argue that depending on the type of investment being made, the rate of return might vary over time and therefore focusing on interest rates is only part of the calculation.

    Separately, investment has been supported on the basis of low interest rates for the past few years, yet rates are now significantly lower than even in the beginning of the year. At what level does the government determine rates are low enough to warrant investment?

    (I also could mention the MMT/MMR/Post-K view that sovereign currency issuers don't strictly need to sell bonds to finance investment. Treasuries are used to influence interest rates and largely represent a transfer of interest income to the private sector.)

    ReplyDelete
  5. Anonymous6:15 AM

    Another view is the opportunity cost of not investing when unemployment is high and borrowing is cheap.
    One concern is that BigG will crowd out the private sector. This simply won't happen in a labor market with 8+ percent unemployment. Considering the high social costs of unemployment (which translate into large economic costs borne by BigG) the downside of not investing is huge.

    jonny bakho

    ReplyDelete
    Replies
    1. dilbert dogbert1:41 PM

      Agree. It may be a good idea to stop the slide of cyclical to structural unemployment. Structural has a big long term cost.

      Delete
  6. I don't know what Tyler is thinking here. Yes investors are more risk averse, but if there were low risk private sector opportunities available, they would be being used and would drive up the interest rate. It is absolutely true that low interest rates should favour low risk, low return investment of all sorts. The real factor is as bakho writes - is there a danger of crowding out?

    ReplyDelete
  7. Tyler Cowen has no plan, design, or intent to be an honest broker of information.

    Everything the man says has only one purpose: to draw him attention.

    What you say contains so much common sense that no other conclusion about Cowen is possible

    ReplyDelete
  8. A picky point: "Remember, in a world where government capital is not a perfect substitute for private capital, government bonds are net wealth, and there is no Ricardian Equivalence."

    Ricardian Equivalence is about whether government spending is bond-financed rather than tax-financed (RE says it makes no difference). It's not about the effects of changes in the level (or timing) of government spending.

    I agree with your main point. We should think about fiscal policy as postponing and preponing government spending, leaving the level of "permanent government spending" (as in "permanent income") unchanged. We are talking about government spending "unsmoothing" (is that a word?).

    My old post, where this started to dawn on me too: http://worthwhile.typepad.com/worthwhile_canadian_initi/2011/03/the-preponed-government-spending-multiplier-may-exceed-one.html

    ReplyDelete
  9. Actually, now I think of it, the validity of the point that spending should be unsmoothed when real interest rates fluctuate over time works for consumption as well as investment, and for government consumption and investment as well as private consumption and investment. Governments have Euler equations too!

    ReplyDelete
    Replies
    1. That's a good point, and I hadn't thought of that!

      Delete
    2. And there are pretty easy rules of thumb to know "when" the gov't should be increasing spending:
      a) look at current unemployment
      b) look at average unemployment over last 5 years
      c) look at avg unemployment over last 10 years.

      If current unemployment (a) is higher than both (b) & (c), now is a good time for more fixing of infrastructure
      If (a) is lower than both (b) & (c), it would be a bad time.
      If (a) is between (b) and (c), it's time to be slowing increasing fixing (b > c) or reducing fixing (b < c).

      Of course, this would require gov't planning, and getting projects to have all the EPA and other paperwork mostly prepared so they're close to "shovel ready".
      And gov'ts are terrible about actual planning for spending money ... unless it's for projects for friends.

      Delete
  10. It's very frustrating to me that it is only the fetish for central bank independence that makes questions like this one apt in the first place. For one thing, to the extent that the debt issued by government is purchased by the central bank, the interest rate on the borrowing is irrelevant. Furthermore, the requirement that government debt financing by the central bank has to work through private sector intermediaries is a self-imposed legislative constraint that the national legislature could change at any time. It seems utterly absurd that a sovereign government faced with the need for both real infrastructure improvements that only public investment can achieve, as well as the need for spending in general to combat economic stagnation, should persist in tying its hands with these obsessive commitments to contingent and defeasible institutional arrangements.

    ReplyDelete
  11. MaxUtility1:09 PM

    @Woj - It's true that you have to include return on investment to thoroughly complete the equation. But it is pretty obvious that there are plenty of investment opportunities (infrastructure, et. al.) that have at least some positive return. On many of these, delaying work decreases the return on investment, so no time like "now".

    AS to what interest rate should trigger investment. We're investing all the time at whatever interest rate is current because there is always the need to build/repair our stock. So of course, there's no magic rate where you "start" investing. But given that real rates are now below zero, that seems like a pretty good time to front load your investment.

    Of course there are all the hazards of boondoggles, failure to pay down the extra debt later on etc. I think those are real concerns, but ones that could be mitigated by careful structuring of the extra investment, temporary tax changes, etc. But that's a real debate, not econo-quibbling about why interest rates are low and how will we know our exact rate of return on massive projects with multi-level effects.

    ReplyDelete
  12. Anonymous8:16 PM

    Noah,
    Another arguement for timing is that the construction sector has been hit hard, so bids will come in lower than at peak times, plus putting construction workers, engineers, architects, etc. back to work is a good thing.

    ReplyDelete
  13. If government could actually 'borrow' at these low rates, it might be an interesting conversation to have. But govt can't. No one is buying Treasury debt at these rates, but the Fed. And that's not borrowing. That's printing money.

    ReplyDelete