Recent Posts

Wednesday, October 29, 2014

For Fiscal Policy, The Future Is Now

The problem with modern analysis of fiscal policy is that it has fallen into the trap of following a certain conventional wisdom: it has to be very forward looking, in fact it is meant to have an analysis horizon longer than the expected lifetime of the Universe (depending upon the cosmological model). Upon reflection, I think the correct philosophy was the one expressed by the famed American Football coach George Allen: The Future Is Now.


Upsetting The Conventional Wisdom In Football



In order to understand the expression as it was originally used, we first have to understand the conventional wisdom in (North American) football. Teams that are not dominating their divisions are assumed to be in "rebuilding" mode, and they often will trade away veteran players in exchange for younger ones, or even for draft picks. The idea is that the proven veteran player will only have a few good seasons left, but they need younger players "for the future", when the team plans on becoming dominant.

George Allen, a coach who turned around the mediocre Washington Redskins, took advantage of the bias created by this conventional wisdom. He assembled a team of veterans (the traditional description is "cagey veterans") who were able to form a team that could  win now. This was possible as the conventional wisdom led other coaching teams to undervalue their veterans, creating a trading environment that could be arbitraged.


How This Works In Fiscal Analysis


This article is only an introduction, but I will summarise the philosophy here as:

  • the financial and business cycle effects of current fiscal policy have extremely limited impact on the macroeconomic trade-offs that matter on a "long" horizon (more than a couple of business cycles);
  • the real effects of fiscal policy can matter on a longer horizon (30 years at most), but these are almost invariably divorced from the associated "financial" analysis. 

A (possibly not obvious) implication of the first point is that debt/GDP ratios do not matter for countries that control the currency that they borrow in.

Other than the motto "borrowed" from a football coach, this is just Functional Finance (primer), but with an emphatic rejection of the conventional wisdom that we need to look at the effect of fiscal policy on a long horizon (in particular, the inter-temporal governmental budget constraint). Very simply, we cannot do so, and it is foolish to try.

For Financial Analysts, This Is Obvious


If you are a rates analyst or investor, this philosophy should not be a surprise. The only quibble one can have is the definition of "now". In football, "now" refers to the current football season; nobody would trade away players solely to make it easier to win one exhibition game. For a rates analyst, "now" is the current cycle, which implies working with what is forecast to happen on the longest sensible forecast horizon - at most 2-3 years. Once you are beyond that horizon, we are out of the realm of forecasts, and into the realm of convention. (For example, a believer in secular stagnation might expect that growth will be below average over the next 10 years, but it would be silly to believe one could use this structural bias to forecast the exact path of quarterly GDP growth rates for the next 40 quarters.)

I realise that 2-3 years is an eternity for many in finance and politics. But since the usual worry in this area of analysis is "how will our grandchildren deal with the debt we are leaving them?", that is an eye-blink.

With that out of the way, if I am wearing my rates analyst hat, it is clear that I do not care what the long-term impact of fiscal policy is. In fact, if I recommend buying the bonds of a country within a floating exchange rate regime, it is the precise opposite of a vote of confidence in the growth prospects of that country. Low interest rates are likely to be a sign of bad fiscal policy (such as ill-timed austerity policies), not good policy. (Of course, when you have a fixed foreign exchange regime, like the euro, default worries start to matter, which invalidates this rule of thumb.)

The classic example is that of Japan. A great deal of capital has been destroyed on the basis of analysing the Japanese government as a corporation, or obsessing over the question "How will Japan pay back its debt?"  The correct mode of analysis was to ask: "Will Japanese inflation rise enough so as to force rate hikes over the next 2-3 years?" (So far, the answer to that question has been "no".)

In other words, a financial analyst has no reason to care about the long-term fiscal picture.

Not So Obvious For Policy Makers


This philosophy is probably only novel for policymakers or concerned citizens. The approved way of looking at fiscal policy is to look at the long-term implications of policies.

In the United States, the Congressional Budget Office prepares 25 year budget forecasts, as well as even more uncertain 75 year plans. It is the accepted convention to put suggested budget plans into that format. The advantage of this standardisation is that competing budget plans suggested by politicians within a similar format, reducing the confusion in debates. This is particularly needed in the United States, where the political system allows individual politicians considerable latitude in advancing their agendas. By contrast, a country like Canada has very strong party discipline, and debates generally involve a few well-defined alternative policies. However, there is a side effect, as analysis outside the political process uses these projections as the baseline framework, whether or not projecting budget numbers on such a long horizon makes sense.

Just after I started in finance, the U.S. Federal Government ran a surplus, which was expected to continue. One of the non-intuitive properties of government debt dynamics under the usual condition of steady nominal GDP growth is that the debt-to-GDP ratio rapidly drops if the government runs a small surplus, while the debt-to-GDP ratio is stable even if runs similarly sized deficits. (This is because the denominator of the ratio – GDP – is steadily growing.)

Although it seems hard to believe now, there was a worry that these surpluses would lead to the Federal Government paying down all of its debt. In fact, the Fed Chairman at the time, Alan Greenspan gave a speech in 2001 discussing the potential impact on monetary policy and financial markets of the disappearance of the Treasury market. This was viewed as a big deal in markets, as Treasury securities were heavily used in interest rate hedging, which is needed by investment dealers as part of bringing corporate debt securities to market. This triggered a brief flurry of reports by market researchers on the time, most of which treated the idea of the disappearance of the Treasury market as being likely. I believe I did some minor number crunching on one of those reports, but I hope my name was not on the report. (I believe that I was skeptical about the disappearance, but I was too junior for anyone to care about my opinion.)

Producing wildly inaccurate forecasts is not the only risk associated with long-term budget forecasting. The other is that one picks up the notion that it is necessary to reduce the debt-to-GDP ratio to some magical level over the long term. And in particular, it is necessary to make changes now, so that some hypothetical problem 30 years out is avoided.

Another application of this philosophy is towards the analysis of the question:
Is government debt a burden on future generations?
The debate over this question is a perpetual source of material for articles on the economics blogosphere; Nick Edmonds provides a summary and comments on the latest eruption of the argument here.

With a philosophy of "The Future Is Now", it easy to guess that my one word answer to that question is "No." But to be intellectually honest, one has to look at the arguments of the other side. Nick Rowe has described one argument against debt being a burden as a zombie idea, and he has written more than half a dozen articles proving that debt can be a burden on future generations. Since Nick Rowe is a trained economist that used to believe in Functional Finance, and my views are just warmed over Functional Finance, I cannot easily dismiss his arguments. (UPDATE: I have modified the wording slightly regarding Nick Rowe's views on this question; his view is more conditional than was implied by my first version.)

I will have to look at this is in later instalments. But I will first note that this is a bit of a trick question, so there actually is no way to say what is the correct answer.
  • The question is vague. What is a "burden"? How do we define a "generation"? A mathematical model - or even a self-consistent verbal model of an economy - can only answer a precise question. This is changing the original question to a new one, and people may disagree about how to pose those new questions.
  • Even if you agree on a precise formulation of the question, it is possible to question the relevance of the model used to answer it. I would have to substantiate this elsewhere, but I have my doubts about the relevance of the overlapping generations models (such as those used by Nick Rowe) to policy makers in a modern welfare state.
  • Even if you agree on the model, you can disagree about the implications. Many people question the value of paying primarily wealthy bond holders coupon interest; if interest rates were locked at 0%, that would create room in the budget for other objectives. Although this may reflect my previous employment as a rates analyst, I think that interest payments actually serve a very useful social function within a mixed economy. I do not have space to cover this here, but I would note that falling interest rates have not been associated with increasing equality of incomes and wealth.

Once again, I expect to flesh this out in upcoming articles.

See Also:



(c) Brian Romanchuk 2014

8 comments:

  1. Yes, it is a bit of a trick question. If you answer "no", you appear to be missing the point; if you answer "yes", you appear to be buying into the idea that governments should strive to reduce debt.

    ReplyDelete
    Replies
    1. It's a bit of a trick question because the whole thing is a trick. Nobody's house is worth what they think it is. If they all sell at once the house is worth nothing. Nobody's savings are worth what they think they are. If they all check out at once they can't buy anything with it.

      But the point is that doesn't happen. The wiggles offset the waggles in the dynamic flow.

      It's all a useful illusion that keeps human beings feeling good and chugging away.

      Sometimes you can over do the analysis and get stuck in the edge cases.

      Delete
  2. Brian: thanks. (I think my previous comment got lost)

    My view is that the debt MAY be a burden on future generations. It depends. My simple little models are just to demonstrate a counterexample to the "we owe it to ourselves so it CANNOT be a burden" view.

    Some examples where it is NOT a burden:

    If the interest rate on the bonds is less than the growth rate of GDP, forever, then the government can rollover the debt+interest forever too, without ever increasing taxes, and this makes all future generations better off. (Nothing is forever in the real world, but the "sometimes" case is too hard for me to work out).

    If old people literally bequeath their bonds to their kids (as a freebie) then there is no first-order burden on the kids. Or, you could say there is a burden, but it's 100% offset by the bequest. Barro-Ricardo Equivalence.

    If the government borrows to build schools, etc., then the kids get the asset of schooling that can more than 100% offset the burden of the debt.

    It depends.

    ReplyDelete
    Replies
    1. Hello,

      Thanks for the response. I modified the wording in my text. Although my text describing your views was deliberately vague, I made the conditional nature more explicit. I expect that I may discuss your counter-example in an upcoming post, and the comments I made here were foreshadowing.

      In any event, the next article I expect to write on this topic is my take on how government spending can create a "burden" for future generations. Although I do see it as possible, I think that it is not something that is going to create a measurable impact.

      Delete
    2. The 'burden of debt' is a misnomer. It can only ever be a burden of spending and demand, which is resolved in the same way as any other excess spending episode.

      Debt is saving which is voluntary taxation. Because it is voluntary it can be reversed by an individual's decision, which can become a group decision if enough people change their mind for some reason. At which point government steps in to resolve the distributional conflict by making that voluntary taxation compulsory, or by banning the change of mind or by some other method.

      The key issue is that it can be resolved *at the time* should it arise. You don't have to worry about it now because it'll probably never be a problem - just like everybody trying to sell their house at the same time.

      We have a principle in systems engineering called YAGNI - you ain't gonna need it. Don't do something until it is necessary to do so.

      Delete
    3. There is a technical usage of "debt burden" within economics that would be more neutrally described as "debt service costs" - that is, interest payments. This usage is what is referred to in the classic 1944 paper by Evsey Domar - "The 'Burden of the Debt' and the National Income". I am interested in a more generic usage of the word "burden", but I am currently writing a multi-part essay that has a title that an homage to that paper. The first part is a discussion of the Domar paper, and I hope to publish it here "soon". The later parts will look at the more generic "inter-temporal drawbacks" associated with government debt. Since I will be writing hundreds (hopefully not thousands) of words on the topic very soon, I don't want to give away the punch line in my comments here. :->

      Delete
    4. Your recent posts have been excellent. Looking forward to your essay on Domar.

      Delete
    5. Thanks! I thought the Domar comment would catch your eye; that instalment is now up.

      Delete

Note: Posts may be moderated, and there may be a considerable delay before they appear.

Although I welcome people who disagree with me, please be civil.

Please note that my spam comment filter appears to dislike long "anonymous" posts. I get no warning about this, and only go through my "spambox" infrequently. The best bet it to keep comments short, and if you think the spam filter struck, let me know with a short comment.