Winner of the New Statesman SPERI Prize in Political Economy 2016


Tuesday 17 March 2015

Radical macro lessons from the Great Recession

The relationship between NGDP targets, a higher inflation target and helicopter money

Just suppose that lower oil prices help generate a period of significantly above average growth in the OECD economies over the next five years. We avoid deflation, but despite more rapid growth modest increases in interest rates keep inflation at or below target. Even if this happens, the story of the Great Recession will not have been a happy one. If you compare where we are now to where we might have been without a financial crisis, there remains a huge gap. Even if we go a good way to closing that gap over the next five years, this very gradual recovery will have cost us dear. In some countries (most of the Eurozone) that cost is currently reflected in high levels of unemployment, while in others (the US) it manifests itself in real wage stagnation or decline. (The UK is now in the latter group: if you are bored with hearing this from me, see this from John Van Reenen.)

Are there lessons to learn from this? You can probably divide economists into two camps at this point. One group, the ‘supply group’ - which would include most of those setting monetary policy - tend to think that we have largely done the best we could under the circumstances. By circumstances, I mean two related things: a rather surprising increase in inflation during 2011, and an apparent decline in the ability of the ‘supply side’ of the economy to grow at the kind of rates we might have expected before the crisis. While the former is undeniable, the second is conjecture, because we cannot observe the key driver of long term growth, which is technical progress.

The second group of economists attribute more of the slow recovery since the financial crisis to deficiency in aggregate demand. I am in that second ‘demand’ group, and have argued that fiscal austerity is responsible for a great deal of the slow recovery. Implicit in such arguments is the idea that had demand been strong, any further increase in inflation around 2011 would have been modest and temporary, which with wise monetary policy need not have led to any increase in interest rates.

I think most of the demand group also share a view that it would be a large mistake to shrug off this bad experience as a one-off, or as something that only occurs every century or so. The ‘one-off’ story could focus on an unfortunate misreading of the Eurozone crisis: however, while this might explain the change in attitudes is some important institutions like the IMF, it is less plausible in explaining why policy makers around the world switched to austerity. The ‘every century’ idea is wrong because it fails to note the changes that have been brought about by the widespread adoption of 2% inflation targets, coupled with a view that the ‘natural’ real interest rate is also likely to remain low for some time.

Different members of the demand group have proposed three different and radical innovations in macroeconomic policy to help avoid this kind of mistake happening again. These are targeting the level of nominal GDP (NGDP), raising the inflation target, and some form of helicopter money. Are these innovations alternatives or complementary to each other?

There are some (notably market monetarists) who seem to argue that changing monetary policy to NGDP targets is sufficient. My own view is less optimistic. A clear advantage of NGDP targets (and not its only advantage) is that they would create expectations of a more expansionary policy during and after the recovery phase from a recession, but in my view this would not be enough to prevent liquidity traps happening. This is because I see the problem of the liquidity trap (nominal interest rates being unable to fall below some lower bound around zero) as central to why the Great Recession was so prolonged, and episodes where we experience a liquidity trap as becoming more frequent because the inflation target (explicit, or implicit within the NGDP target) is low.

Raising the inflation target is an obvious way of reducing the frequency of liquidity traps. If the natural real interest rate is 2%, for example, then with a 4% inflation target, the nominal interest rate has much further to fall before the lower bound is reached than if the inflation target was 2%. It is important to note that this argument does not preclude adopting a NGDP target, because any target path for NGDP includes an implicit inflation target. For that reason, you can view NGDP targets and a higher inflation target as either complementary or alternatives, where the latter is true only if you think one device does the required job by itself.

Helicopter money is essentially giving the central bank an additional instrument - a form of fiscal stimulus. In that sense it is rather different from NGDP targets or a higher inflation target, because it involves instruments rather than the objectives of monetary policy. For that reason, in principle it could be a complement to both the other radical suggestions. In a way helicopter money is best seen as an alternative to Quantitative Easing, and there is no reason in principle why QE is not compatible with NGDP targets or a higher inflation target. It is possible of course that if helicopter money was shown to be effective in dealing with liquidity traps, then it would make the case for other radical changes less compelling.

If I’m being realistic, I think that if the first sentence of this post turns out to be true, the chances of any of these radical changes being adopted before the next liquidity trap episode are very small. A period of strong growth will be sufficient for policy makers to pretend that the slow recovery from the financial crisis was either a one-off or the best that could be done in the circumstances. Instead I suspect that as the economy moves ever closer to its pre-crisis trend, the demand group of economists will convince more of the supply group that they were wrong. This will give greater credibility to the idea that radical changes to policy are required, and each alternative will receive greater analysis and probably greater support amongst economists by the time the next liquidity trap episode occurs. 


24 comments:

  1. The John Van Reenen paper reffering to a chart of GDP per capita states that

    “Figure 1 shows that unlike in the aftermath of other
    post-war recessions, there is no sign of recovering the output lost following the global financial crisis of 2008-09.”
    End quote

    The maths used there is wrong. The fact there is no return to trend is not extraordinary. The previous recessions are included in the data that created the trend line , but the last recession is not included. Therefore the last recession stands out below the trend line that has been drawn by the very nature of how it was drawn. The gdp per capita remains below the pre recession trend line for the recessions of 75 80 or 90 as well when the appropriate pre recession trend lines are drawn in. The gdp per capita always remains below the pre recession trend line for a long time on that chart after a significant fall, if returning at all, 10 years for 1990 to 2000, and so there is nothing extraordinary about 2008 to 2015.

    ReplyDelete
    Replies
    1. I suggest you try it using excel, and you will see you are completely wrong.

      Delete
    2. Because even if you estimate a trend up to and including 2014, you get 5/6 years with no clear tendency for output to return to trend.

      Delete
    3. Well If I estimate a trend line to 1990 then there is 10 years until an output return to trend.

      In order to compare this recovery to the previous recoveries it is required to draw the pre recession trend lines for each of the previous recoveries.

      Using the chart from your 6th March post , we have never recovered the notional lost output from the 1970 recession compared to the 1950 to 1970 trend line.

      Delete
  2. Simon Wren-lewis
    I am sorry to tell you that you are still supply sider group because you are using their theories and assumptions, at least with this post. But in some corners of theories you found inrefutable evidence that Agregate demand thinking is the only proper way to think about economy and you would like to belong to such group. I am sorry to tell you that you do not belong there. You argue for supply side.

    NGDP targeting and 4% inflation target are investment side arguments while helicopter money is not. Even tough in para where you try to conflate QE and helicopter money i would say that you do it using supply side arguments.

    It would do much good to look for causes of NEED for negative interest rate target.

    If you asume that monetary policy can be only constrictive, it can not function as expansionary. Then, it would be much clearer why ZLB is there. Monetary policy can only ease on constriction of economy which you and many other economists consider such easement of constriction as "expansionary".

    Why is NGDP and 4% inflation supply side argument and not demand side?
    They do not work on wage push inflation but on price pull inflation. Price pull inflation is stagflation. QE too.
    Helicopter money would be wage push inflation, but not when there is too much need for deleverage. There would be no inflation because a lot of money would go for debt payments. It would be destroyed.

    NGDP and 4% target are created with purpose to induce/force investors to invest more and then they will employ people. This is pure supply side argument.
    Investors want positive real interest rates for their investments which is a raoundabout way for saying that investors want profit.
    On the other side borrowers want negative real interest rates for their debt, otherwise they will go bust.
    How to solve this paradox of negative and positive real interest rates at the same time is what you should spend time about.

    Many solved this paradox long time ago. It is only by wage push inflation on fixed loan rates with help of interest tax deduction.

    And yes, 4% inflation rate target as lower bound would be helpful in the future but there is no use for it now. Monetary policy can not help expand. It can only ease on constriction.

    ReplyDelete
  3. We're discussion economic depressions. Which can be depressing, actually.

    ReplyDelete
  4. "Instead I suspect that as the economy moves ever closer to its pre-crisis trend, the demand group of economists will convince more of the supply group that they were wrong. This will give greater credibility to the idea that radical changes to policy are required, and each alternative will receive greater analysis and probably greater support amongst economists by the time the next liquidity trap episode occurs. "

    I hope you are correct.

    ReplyDelete
  5. Your criticism of NGDP Targeting as not being enough on its own seems to rest on the target being too low to have a high enough implied inflation target within it.

    Surely this is a criticism of the NGDP Target rather than NGDP Targeting per se? If the former, shouldn't you at least discuss the most common growth rate proposed by Market Monetarists? It is most often 5%. Wouldn't that be enough? Should it be higher, perhaps to make up for lost NGDP? Should it be a 5% level target?

    ReplyDelete
  6. You forgot one more alternative: prevent the US housing bubble from occurring.

    WHY IS HOUSING FINANCE STILL STUCK IN SUCH A PRIMITIVE STAGE? Robert J. Shiller January 2014 in which he offers:

    "This paper lists a number of barriers to housing finance innovation, and in light of these barriers, the problems of some major innovations of the past and future: self-amortizing mortgages, price-level adjusted mortgages (PLAMs), shared appreciation mortgages (SAMs), housing partnerships, and continuous workout mortgages (CWMs)."

    I know Krugman did not put it this way when comparing MIT men inside and outside institutional settings, but the response to this economic crisis, trying to clean up after central bankers had not take away the punchbowl in the mid-2000s now needs the taking away of central bankers' collective mirror of self-reflection as it seems to only give an image back of a sickly group satisfaction.

    ReplyDelete
  7. The faster the recovery, the greater the risk of overshooting. A slow recovery is bad, but an economy that rapidly oscillates is bad too. I think a lot of people (who matter) instinctively prefer a slow and steady recovery, and are willing to pay a high price for it. It doesn't matter what tools you give them; they are going to seek the same outcome regardless.

    ReplyDelete
    Replies
    1. A lot of people (who don't matter) would prefer to have an adequately paid job.

      Delete
    2. Prof. Wren-Lewis:

      What do you call the period from 1929 onward?

      Delete
    3. Whoops - text right, title wrong. Now corrected - thanks.

      Delete
  8. Put "average inflation" on the vertical axis, and "probability of hitting the ZLB" on the horizontal axis.

    Draw a downward-sloping trade-off for Inflation Targeting. The higher the inflation target, the lower the probability of hitting the ZLB.

    Now draw the downward-sloping trade-off for NGDP level targeting.

    I think the NGDPLT trade-off would be everywhere inside the IT trade-off. IT gives an inefficient menu of choices.

    ReplyDelete
    Replies
    1. What I can't get my head around is whether "you are allowed to use helicopter money at the ZLB" means a movement along the trade-off or a shift in the trade-off. Because helicopter money is only really helicopter money if it's permanent.

      Delete
    2. The existence of a liquidity trap is irrelevant. The propensity to hold cash instead of bonds because of the fear of capital loss if interest rates rise, is likewise questionable. There has been no lack of interest lately, in buying "risk free" government bonds, that pay the same coupon, regardless of the price they trade at, in the secondary market or at DMO auctions.

      Liquidity trap or not, the size of the monetary base (currency plus bank reserves) is largely irrelevant. It does not increase the risk of inflation. It does not increase the funds available for banks to lend. Banks do not lend reserves except to other clearing Banks, they are only used for interbank settlement. Reserves are the scoreboard for "fiscal assets" that are in the non-government sector of the economy including Commercial Banks. All government spending is "money" created out of thin air and starts life as "reserves" at its Central Bank.

      The current high level of reserves re-created by QE swapping Gilts back to cash (reserves), is forcing interbank rates to zero. The BoE is paying interest on reserves to keep its policy rate at 0.5%. It can change that rate up or down any time it wants; hence the term ZLB is meaningless to a Central Bank that pays interest on reserves.

      Please try and forget the term "helicopter money", stick with fiscal stimulus, which can only be performed by the currency issuer; the Treasury. The central bank (CB) can't create new money, only the Treasury can do that. It may get the CB to do the leg work for it, mind you. Like it did with the "Funding for Lending" scheme. A fiscal stimulus heavily disguised as a BoE monetary action.

      Delete
    3. Anon
      Banks do lend reserves to government too, not only to other banks.
      So the level of reserves might be important if mass defaults or deleverage starts. It reduces ammount of reserves which government borrows.

      Everything else you wrote i agree with. Thanks

      Delete
    4. The Fed already has a 2% PCE inflation target that they have happily failed. The market can interpret the casual fails as evidence of an asymmetric response to 2%+. Changing the targets, improving metrics, and focusing on levels doesn't seem particularly valuable when the Fed behavior can depart from explicit targets without challenge.

      What would prevent the Fed from "failing" revised targets, when they can fail the current one on a multiple year basis?

      Delete
  9. Simon,

    your photo has you aged many years. Is this the revenge of classical economics?

    Bloody good article by the way

    ReplyDelete
    Replies
    1. I hoped no one would notice! I think its one new white hair for every blog post.

      Delete
  10. I am no macro-expert, but I've been thinking that if we assume a downward sloping supply curve (with demand more inelastic than the supply - ie, demand crosses over the supply curve) due to input indivisibilities (ie, scale economies), a negative demand shock would lead to higher prices – ie, as firms struggle to recover their fixed costs with a smaller customer base. This would register as rising inflation, although there would basically be an over-supply – ie, the policy implications would be reverted. Admittedly, firms could refuse to raise prices to steal rivals' business, but that might lead to a war of attrition.

    This price increase would further squeezze disposable incomes, leading to a vicious cycle.

    Moreover, if you believe that under these circumstances, your payroll is mainly fixed - as it is made up mainly of knowledge economy people doing product developments - so that even in the downturn you cannot lay them off - given that continuous product innovation/improvement/personalisation is a key competitive dimension (ie, without it you are quickly obsolete, so better hang on to your market share with all your fixed costs in a sort of war of attrition) –you have an alternative explanation for lower productivity– ie, it is not about job-hoarding but about input indivisibilities (including payroll) in a knowledge economy.

    Overtime, as weaker firms exit the market, survival will be able to broaden again their customer bases by lowering prices (ie, in order to regain lost scale economies), which would boost real disposable incomes and revert the cycle thereof.

    Is this total nonsense? As to whether it is plausible to think in term of downward sloping supply-curves, nowadays there is a lot of talk about the the fact that digitalisation has penetrated all sectors (ie, not just hi-tech ones), whereby firms’ cost curves becomes front-loaded (high first-copy fixed cost and low marginal reproduction cost).

    ReplyDelete
  11. Professor Wren-Lewis,
    I am in agreement with your views. One question...
    In order to determine the NGDP price level, how many previous quarters or years would you use for averaging core inflation?

    ReplyDelete
  12. Sounds really interesting!! You know I am an accounting student and love to read such articles. My Professor Aloke Ghosh also writes such informative articles. His articles are written after lots of research.

    ReplyDelete

Unfortunately because of spam with embedded links (which then flag up warnings about the whole site on some browsers), I have to personally moderate all comments. As a result, your comment may not appear for some time. In addition, I cannot publish comments with links to websites because it takes too much time to check whether these sites are legitimate.