Are we too cheap to test macro theories, or afraid of the answers we’d get?

David Glasner recently began a post with the following observation:

In my previous post, I suggested that Stephen Williamson’s views about the incapacity of monetary policy to reduce unemployment, and his fears that monetary expansion would simply lead to higher inflation and a repeat of the bad old days the 1970s when inflation and unemployment spun out of control, follow from a theoretical presumption that the US economy is now operating (as it almost always does) in the neighborhood of equilibrium. This does not seem right to me, but it is the sort of deep theoretical assumption (e.g., like the rationality of economic agents) that is not subject to direct empirical testing. It is part of what the philosopher Imre Lakatos called the hard core of a (in this case Williamson’s) scientific research program.

I’m going to stay away from the “equilibrium” dispute, mostly because I don’t know what people mean by the term, and also because I think many of these debates are merely about semantics, of no practical importance.  But the question of whether monetary policy can reduce unemployment is of great practical importance, and also incredibly easy to test.

If we had any serious interest in finding out the impact of monetary policy on real variables, we’d have the Federal government create NGDP and RGDP futures markets, and subsidize trading at a high enough level to create a deep and highly liquid market.  Then we’d watch NGDP and RGDP futures contract prices rise in response to new information on QE1, QE2, and QE3.  Because skeptics about the ability of monetary policy to lower unemployment are usually the same people who believe markets are efficient, we’d have that debate resolved very quickly and we could move on to more important things, like whether NGDP targeting or inflation targeting are superior.  (I think NGDP targeting is better, even if nominal shocks don’t impact real output.)

Instead we have an FOMC made of up people who agree with Williamson, and others who disagree.  And that committee makes decisions that move global real equity prices by TRILLIONS of dollars in a single day.  And they do so in complete ignorance of the answer to Glasner’s question, even though a few MILLION dollars spent on a prediction market (Robin Hanson could design it) would quickly answer this question, and allow for much more effective monetary policy.


Tags:

 
 
 

25 Responses to “Are we too cheap to test macro theories, or afraid of the answers we’d get?”

  1. Gravatar of Steve Steve
    17. December 2012 at 13:55

    http://www.forbes.com/sites/timworstall/2012/12/12/cftcs-intrade-shutdown-dont-they-have-better-things-to-mither-about/

    Meanwhile, the CFTC is busy trying to shut down Intrade. US users get kicked on next Sunday, December 23. Merry Christmas!

  2. Gravatar of Steve Steve
    17. December 2012 at 13:55

    kicked OFF

  3. Gravatar of Greg Hill Greg Hill
    17. December 2012 at 14:14

    “we’d have the Federal government create NGDP and RGDP futures markets.”

    Seems like these markets would also be useful for firms that wanted to hedge their hiring and investment decisions, a neat complement to Robert Shiller’s Macro Markets. Insofar as the success of most firms’ hiring and investment decisions are positively correlated with NGDP and RGDP, they hedge their bets by shorting these indices. But I suppose this might undermine the purpose of these futures markets from you point of view.

  4. Gravatar of Steve Steve
    17. December 2012 at 14:40

    Greg Hill,

    Maybe, but the best way for a company to hedge its hiring and investment decisions is by paying down debt or issuing shares. Optimal Capital Structure theory and such.

    It’s more likely that investors would take small positions in their “Alternative Asset” buckets, along side things like TIPs. But if it turns out the markets believe there is a link between TIPs breakevens and RGDP potential, that’s good knowledge to have, right?

  5. Gravatar of Major_Freedom Major_Freedom
    17. December 2012 at 15:11

    Are we too cheap to test macro theories, or afraid of the answers we’d get?

    Neither. Any macro “test” will fundamentally alter the structure of the economy, thus bringing about a new set of initial conditions, after which the “test” cannot be replicated, and thus any “confirmation” or “falsification” becomes impossible.

    You can’t observe counter-factuals.

    Temporal changes over time do not display causality in economics.

  6. Gravatar of JP Koning JP Koning
    17. December 2012 at 15:18

    “…skeptics about the ability of monetary policy to lower unemployment are usually the same people who believe markets are efficient.”

    As a believer in efficient markets, wouldn’t that make you a monetary policy skeptic?

  7. Gravatar of Doug M Doug M
    17. December 2012 at 16:05

    How does the government subsize the market without creating a price distortion of its own that would outweigh any price signal that the market might provide?

  8. Gravatar of Doug M Doug M
    17. December 2012 at 16:06

    CFTC is suing the makers of “prediction markets.”

  9. Gravatar of Greg Hill Greg Hill
    17. December 2012 at 16:13

    Steve,

    Thanks for the thoughtful reply. I’m not an expert, but I think, for some firms at least, shorting an NGDP index could be superior to paying down debt. Suppose future NGDP takes one of two states: “high” or “low.” Suppose the firm’s investment project does well in the “high state,” but poorly in the “low state.” By shorting the NGDP index, the firm hedges its investment risk, accepting a lower net return in the “high NGDP state” in exchange for a higher net return in the “low NGDP state.”

    Alternatively, a firm interested in managing its risk could reduce its outstanding debt before making this investment, thereby allowing the firm to absorb greater losses in the “low NGDP state.” It’s not clear, to me at least, that the debt-reduction alternative would always be superior to the alternative of shorting the NGDP index. Ditto for issuing shares. But, as I said, I’m not an expert.

    As an aside, and much more ambitiously, if 1) firms are reluctant to hire or invest because they’re uncertain about future demand for their products, and 2) if demand for their products is (variously) correlated with NGDP, and 3) if opinions about future NGDP differ, then Scott’s NGDP and RGDP futures markets could be the way out of our morass, though for a different reason.

  10. Gravatar of Tomasz Wegrzanowski Tomasz Wegrzanowski
    18. December 2012 at 04:58

    I’m still surprised neither you nor Robin Hanson made any attempt at setting up even low-level NGDP/RGDP/inflation futures. It might be as simple as calling Intrade explaining why such markets would see trading, and then advertising them everywhere so it really happens.

    Or even setting up website with toy money predictions. Old toy money Intrade was pretty good at predicting too.

  11. Gravatar of Saturos Saturos
    18. December 2012 at 06:01

    I still wish Scott would read the new Arnold Kling post, it’s highly relevant.

  12. Gravatar of ssumner ssumner
    18. December 2012 at 06:18

    JP, No, I’m the rare bird that believes in both Ratex and sticky prices.

    Tomasz, One of my commenters said he would try–I never heard back. That’s something for others to do—I don’t have the time or money. I can’t do everything. Every single day I get emails from people telling me I need to do this or that, or do some favor for them, or write some post, or read some article and give comments. It never stops.

    Saturos, I’ve read it and several days ago starting working on a reply–haven’t had time yet to finish it.

  13. Gravatar of Saturos Saturos
    18. December 2012 at 06:32

    Scott, I’m sorry for bothering you! Must be hard being you these days!

  14. Gravatar of JP Koning JP Koning
    18. December 2012 at 07:37

    Scott, thats sounds kind of New Keynesian to me, in that it starts with a rational expectations general equilibrium model, then adds sticky prices after.

  15. Gravatar of Major_Freedom Major_Freedom
    18. December 2012 at 08:02

    Doug M:

    How does the government subsize the market without creating a price distortion of its own that would outweigh any price signal that the market might provide?

    That’s just it. The very “stimulus” that Keynesians and Monetarists of all stripes talk about, depend on such price distortions, or else the stimuli would not have the intended effect. Conscious or not, if it weren’t for the Cantillon effect, then inflation could not do what Keynesians and Monetarists think it does.

  16. Gravatar of Doug M Doug M
    18. December 2012 at 09:06

    MF,

    Usually the government intrevenes in markets because it wants to move a price (interst rates, FX rates, and Oil most frequently). Distoring the price is the goal.

    But, in this case, our professor is suggesting that the government create a market because the prices will have important signals that will influence policy that will move the prices that the govenment is watching — feedback loop. And to subsidize it, creating noise, and subsidize in order for the market to reflect back some sort of signal. Financial sonar!

  17. Gravatar of Major_Freedom Major_Freedom
    18. December 2012 at 09:27

    Doug M:

    But, in this case, our professor is suggesting that the government create a market because the prices will have important signals that will influence policy that will move the prices that the govenment is watching “” feedback loop. And to subsidize it, creating noise, and subsidize in order for the market to reflect back some sort of signal. Financial sonar!

    Yeah about that silly NGDP futures market thing. Apparently, the futures contracts will be fixed in “price”. Sumner has said so. And yet, Sumner expects there to be speculators who will go head over heels trading in them. Apparently, buyers who think the price is too low, and sellers who think the price is too high, you know, a necessary component to all markets, will enter this futures “market.”

    I liken this NGDP futures market proposal to every other cockamamie central planner scheme the advocates of which believe they can create markets by state decree. Thou shall invest in a financial security that never changes in price, because thou shall find it lucrative!

  18. Gravatar of ssumner ssumner
    19. December 2012 at 06:09

    Saturos, Keep bothering me, just don’t expect quick replies.

    JP. Yup, that’s why I’m the only real New Keynesian left.

  19. Gravatar of PrometheeFeu PrometheeFeu
    19. December 2012 at 17:29

    @ssumner

    Why would you need both an NGDP and an RGDP futures market? Would’t an NGDP market + looking at TIPS tell you the same thing? Is that because TIPS are indexed to CPI and you prefer the GDP deflator?

    I’m also curious as to whose numbers you would go with for the NGDP market? And how would you handle revised numbers?

  20. Gravatar of Major_Freedom Major_Freedom
    19. December 2012 at 17:49

    PrometheeFeu:

    Another question is why would any investor invest in an NGDP futures contract that never changes in price. Sumner believes that if the Fed offers these contracts, and promises to purchase or sell as many as they want at a fixed and unchanging price, that somehow investors will want to give up liquidity for zero return. The precise reason why anyone would want to invest in a literal zero return security, is going to have to be explained by someone other than me, because it is beyond my comprehension.

  21. Gravatar of PrometheeFeu PrometheeFeu
    19. December 2012 at 18:27

    @Major_Freedom

    I don’t think Scott wants the Fed to conduct OMOs on NGDP futures. He just wants the Fed to target them. It’s highly unlikely the market in NGDP futures would have the necessary depth to support OMOs, especially in time of crisis.

    I have an inkling with regard to your second question, but I would rather see Scott’s answer.

  22. Gravatar of Major_Freedom Major_Freedom
    19. December 2012 at 19:20

    I don’t think Scott wants the Fed to conduct OMOs on NGDP futures. He just wants the Fed to target them.

    I recall him saying that the Fed would be buying and selling them, but I could be wrong.

    It’s highly unlikely the market in NGDP futures would have the necessary depth to support OMOs, especially in time of crisis.

    I don’t think it would have any depth, because there is no return to investing in them. I distinctly recall Sumner saying the prices of the futures would never change.

    I have an inkling with regard to your second question, but I would rather see Scott’s answer.

    I am interested in your thoughts about the matter.

  23. Gravatar of PrometheeFeu PrometheeFeu
    20. December 2012 at 15:33

    @Major_Freedom:

    “I recall him saying that the Fed would be buying and selling them, but I could be wrong.”

    I recall him saying both of those things, but more recently talking less about the Fed conducting the bulk of its OMOs in the NGDP contracts market. (To be honest, that would sound like blowing a hair drier on the thermometer.)

    My thoughts on the matter since Scott Sumner appears to be busy with important-people-related things. The contracts would trade at a discount of their expected settlement value because of 1) the time discount 2) the Fed is going to miss its target in every single period, creating a risk. If they do their job well, they’ll be close to the target, but this will offer some opportunity for speculation and risk management. Finally, Scott Sumner specifies that the NGDP futures contracts market would be subsidized. So you would probably want to trade in them for the subsidy.

  24. Gravatar of ssumner ssumner
    20. December 2012 at 17:38

    Prometheefeu, Yes an RGDP market would be better, partly for the reasons you give, although a TIPS estimate would certainly help.

    Revisions aren’t much of a problem as long as they are unforcastable. The forecast of the initial reading should equal the forecast of the revised number, at least if we are talking about 12 months out.

  25. Gravatar of Major_Freedom Major_Freedom
    29. December 2012 at 10:30

    Prometheefeu, Yes an RGDP market would be better, partly for the reasons you give, although a TIPS estimate would certainly help.

    Would a RGDP futures contract trade at a discount in order to attract investors? Or will they too remain fixed in price, thus making them unattractive for investors?

Leave a Reply