Winner of the New Statesman SPERI Prize in Political Economy 2016

Thursday 2 March 2017

A self-fulfilling expectations led recession?

The only two lectures on Oxford’s core undergraduate macro course that I still teach, and which I have just taught, are the last two on fiscal policy. I use the privilege of the last lecture to end on a reflective note. I acknowledge that macro rightly got a lot of stick by largely ignoring the role of finance, but I also point out that the poor recovery has involved a vindication of the core macro model: austerity is a bad idea at the ZLB, QE was not inflationary and interest rates on government debt did not rise but fell.

So far so familiar. But I end by showing them my this chart.

And I say that we really have no idea why there has been no recovery from the Great Recession, so there are plenty of mysteries left in macro. The puzzle is sharpest in the UK because the pre-crisis trend is so stable, but something similar has happened in most places. I think it is a suitable note of humility (and perhaps inspiration) on which to end the course. 

A mechanical way to explain what has happened is to bend the trend: to suggest that technical progress has been slowing down for some time. This inevitably means that the pre-crisis period is transformed into a boom. I have been highly skeptical about that story, but I have to admit part of my skepticism comes in part from traditional ideas about what inflation would do in a boom.

However another explanation that I have always wondered about and which others are beginning to explore is that perhaps we remain in an extended period of demand deficiency. Keynesian theory is very suggestive that such a possibility could occur. Suppose that firms and consumers came to believe that the output gap was currently zero when it is not, and that they erroneously believed that the recession caused a step change both in potential GDP but also possibly its growth rate. Suppose also that unemployed workers priced themselves into jobs by cutting their (real) wage or disappearing by no longer looking for work. The former could happen because firms could choose more labour intensive production techniques: scrapping the car wash machine for workers with hoses.

In that situation, how do we know that we are suffering from demand deficiency? The traditional answer in macroeconomics is nominal deflation: falling wages and prices. But because workers have already priced themselves into jobs, nothing more will come from the wages route. So why would firms cut prices?

If the pre-crisis trend still applies, it means that there are a large number of innovations waiting to be embodied in new investment. With this new more efficient capital in place, firms would either increase their profits on selling to their existing market or try to expand their market by undercutting competitors. We would get an investment led recovery, accompanied by rising productivity and perhaps falling prices.

But suppose the innovations are just not profitable enough to generate an increase in profits that would justify undertaking the investment, even though borrowing costs are low. Maybe a far more dependable motivator for embodied technical progress to take place is the need to satisfy an expanding market. The firm needs to install new capacity to satisfy growing demand for its product, and then it is obvious to investment in equipment that embodies new innovations. The accelerator remains a very successful empirical model of investment. (On both points, see this discussion by Caballero.) But if beliefs are such that the market is not going to expand that much, because firms believe the economy is ‘at trend’ and trend growth has now become pretty small, then the need to invest to meet an expanding market largely goes away.

This idea goes right back to Keynes and animal spirits of course. Others have more recently reformulated similar ideas, such as Roger Farmer. This is a little different from the idea of adding endogenous growth to a Keynesian model, as in this paper by Benigno and Fornaro for example. I’m assuming in this discussion that potential output has not been lost, because innovation has not slowed, but it is simply not being utilised.

It is this possibility which is the reason that I have always argued central banks and governments should have been much more ambitious about demand stimulation after the Great Recession. As I and others have pointed out, you do not have to attach a very high probability to the scenario that demand will create supply before it justifies a policy of ‘testing the water’ by letting the economy run hot. Every time I look at the data above, I ask whether we have brought this on ourselves by a combination of destructive austerity and timidity.


  1. I mentioned Ireland before when you were last saying austerity doesn't work, but the recovery appears to have happened in the face of much harsher austerity than the UK:-

    Also, you are looking solely at the UK, and the UK isn't a closed system. Is it possible that it has just gone somewhere else?

    1. Ireland's recovery from austerity induced recession has been very rapid, but there are a number of specific Irish economy factors in that, as well as the expected impact of greater competitiveness. Most countries have experienced something similar to, if in the most part less extreme, than the UK.

    2. ... most notably Ireland runs a healthy current account surplus (providing external funding to their economy). UK runs a current account deficit which has increased even as the stimulus spending tapered off, thus the two movements canceled each other out.

  2. Slightly off topic this, but: “austerity is a bad idea at the ZLB..”.

    The GDP maximising rate of interest is the free market rate. Any attempt by the state to artificially raise that rate will therefor reduce GDP, and the state artificially raises rates by selling government debt to all comers at below its current price. That rise in interest rates is ENTIRELY ARTIFICIAL, since that borrowing does not fund anything: e.g. infrastructure spending does not automatically rise when the BoE raises interest rates.

    As Warren Mosler said, the natural rate of interest is zero. Demand should be adjusted purely by fiscal means: i.e. printing money and spending it into the private sector.

    1. Printing money is monetary policy.

    2. Printing money is monetary policy.

    3. See my Australia instance, below

  3. Interesting post. I agree that a purely exogenous decline in productivity growth, that just happened to coincide with the recession, looks implausible. Too big a coincidence.

    But I *think* that your implicit model could only work with an implausibly high labour supply elasticity. (Roger gets around this by deleting the real wage equation from his search model, which I also find a bit implausible.)

    I want to suggest a variant: investment depends (negatively) on the (subjective) *variance* of Aggregate Demand. When the great recession hit, firms revised their beliefs about that variance, because they realised that the business cycle had not been tamed. So investment fell.

    I expect I should do a blog post on this.

    1. Could you explain your point about labour supply?

    2. Nick, I think Simon's implicit model actually works with a low labour supply elasticity - or even with inelastic labour supply. Workers really don't want to be unemployed, so they price themselves into jobs by working for a lower real wage. There is an equilibrium in which firms expect low demand and substitute towards labor and away from capital/intermediate inputs.

    3. It's not a very clear thought, but I will try:

      I mean "elastic" both in the standard sense (very responsive to changes in W/P), but also "elastic" in the sense of very responsive to changes in the existence of job opportunities. Yes, if nearly 100 workers permanently withdraw from the labour force whenever employment temporarily drops by 100 ("very elastic" labour supply) then we could get self-fulfilling expectations equilibria.

  4. "we have brought this on ourselves by a combination of destructive austerity and timidity"
    Yes, and more generally by lack of recognition and taking responsibility for performative economic parables, by both economists and politicians.
    See my "Economics as meant" presentation at the ETF in Leuven last Saturday.
    Lans Bovenberg's keynote speech about "Economics as discipline of hope" is also very relevant, but not yet online; can send it though.
    The challenge for economists is to take responsibility for the performativity of their theories and to use that for good.

  5. First congratulations for looking at a trend of GDP per head instead of total GDP. It might have been even better GDP per working age person or GDP per actual worker.

    However that trend is very tendentious, because it closely hugs the reported GDP index in the 35 years, 1982-2007, in which the UK was a net exporter of oil, and at the same time and not coincidentally "enjoying" both long term "Barber" consumption booms and "Lawson" property booms thanks to debt growth rate way higher than report GDP index growth rates.

    That of course was "above trend" growth, not trend growth. The line from the level in 1976 to the level in 2016 would show most of the 1982-2007 period as "above trend" indeed.

    F Coppola has pointed out that the rise and fall in "productivity", which ultimately drives GDP per head, over the past 35 years, is in large part attributable to the mining and energy sectors:
    «Notice when productivity started to slump. It was much earlier than 2008. In fact the data (which ONS have helpfully provided in Excel) show that output per hour started to fall in Q4 2006.»
    «The UK's massive productivity growth from 1990-2006 was due to oil, not financial services. Even energy utilities downstream from North Sea Oil had a greater productivity rise than financial services. And both oil and utilities suffered a massive collapse.»

    And that is before taking into account the effects of the stupendous debt boom which is the consequence of the asset-stripping policies favoured by New Labour and Conservative leaders over the past 35 years to buy the votes of property speculating baby boomers.

  6. As to the effects of oil "productivity" and debt growing faster than incomes, the younger Tony Blair in 1987 wrote very lucidly:
    «Mrs Thatcher has enjoyed two advantages over any other post-war premier. First, her arrival in Downing Street coincided with North Sea oil. The importance of this windfall to the Government’s political survival is incalculable.
    It has brought almost 70 billion pounds into the Treasury coffers since 1979, which is roughly equivalent to sevenpence on the standard rate of income tax for every year of Tory government. Without oil and asset sales, which themselves have totalled over £30 billion, Britain under the Tories could not have enjoyed tax cuts, nor could the Government have funded its commitments on public spending.
    More critical has been the balance-of-payments effect of oil. The economy has been growing under the impetus of a consumer boom that would have made Lord Barber blush.»
    «Instead, oil has earned foreign exchange and also produces remittance payments from overseas investments bought with oil money. The situation is neither stable nor healthy in the long term: but in the short term it allows the living standards of the majority to rise rapidly, even though the industrial base, the ultimate foundation of a successful economy, is still only achieving the levels of output of 1979.
    The fact that we have failed to use oil to build a productive and modern industry for the future is something historians will deplore.»

  7. The same tendentious trend line drawing is very energetically used by B DeLong and other Economists to make the same argument as to the overall (not per head) USA reported GDP index, and this article by a retired Fed economist shows clearly how tendentious that is, particular graphs 4 and 5:
    «The relevant question is this: Is the historical increase in real GDP from the mid-1990s up to the financial crisis due to an abnormal increase in potential output because of a sharp rise and subsequent sharp fall in labor productivity for reasons unknown, or is it due to the confluence of several events, some of which are well known? This is difficult to answer. But whatever the answer, the period of high growth looks like an anomaly.
    Figure 5 shows real GDP, the CBO’s estimate of potential output, and a trend line estimated from quarterly real GDP over the period 1949Q1 through 1994Q4, just before the sharp rise in output growth. The trend line and CBO’s estimate of potential are nearly identical up to the mid-1990s. Moreover, real GDP before the mid-1990s and after mid-2009 (the end of the recession) fall nicely along the trend-line estimate of “potential.” Hence, relative to the trend line, the behavior of real GDP during the 1995-to-mid-2009 period is an anomaly.»

    And that even without taking into account the asset stripping of the Alaskan oil fields or the Bakken "tight" oil areas.

  8. «This inevitably means that the pre-crisis period is transformed into a boom. I have been highly skeptical about that story, but I have to admit part of my skepticism comes in part from traditional ideas about what inflation would do in a boom.»

    As to those "traditional ideas", it does not make sense under them that a large mature economy or rather several, would be the target of large capital inflows from developing countries, allowing those large mature economies to run at the same time asset and employment growth and high immigration, while having large trade deficits for decades.

    As many central banks have repeatedly claimed very happily, large influxes of imported goods and immigrant labor from low wage economies have had a big effect in pushing down wage inflation, at the same time as large influxes of capital from largely the same economies have made it easy to finance a very loose credit policy while keeping a strong exchange rate; helping central banks and treasuries to focus on their favourite policies aiming to redistribute resources from "low productivity" and "inflationary" workers to "wealth creator" and "non-inflationary" rentiers.

  9. «the scenario that demand will create supply»

    In the UK the large (but shrinking) government deficit of the past 10 years, the large (and probably growing) trade deficit, and the large growth in borrowing, and the large increase in immigration seem to contradict the idea that there is a scarcity of *aggregate* demand. There seem to be rather concurrent "Barber" consumer credit booms and "Lawson" property credit booms. Probably "sustainable" GDP is about 10% lower than it is currently, and GDP growth half the current level.

    What to me seems to be happening is that while *aggregate* demand is being artificially boosted, in an attempt to bring back the feel of the 1990s, Treasury and BoE policy is aimed at upward redistribution. That seems to be a better explanation than an imaginary demand gap for falling median real worker wages and top and booming asset incomes.

    And consider the counterfactual: would the pull of higher demand lead to the supply of a new elephant oilfield in scottish waters? :-)

    It is a bit like for Greece: would the pull of higher demand in Greece lead to a greater supply of 25%-of-GDP hard-currency fiscal transfers "disguised as loans" to the greek government? :-)
    Would the pull of higher demand in Greece lead to a greater supply of greek-made Porsche Cayennes, Sony iPads, Samsung TVs, Versace jackets, for 25% of GDP? :-)

  10. Essentially, SWL is saying that he and other economists don't know much. So following his recommendations is little different from betting at Ladbroke's. That's hardly enough of a basis to blame politicians, esp. in Britain, which has a large fiscal and a large trade deficit. Since there no free lunches, every economic decision has tradeoffs - no advantages without disadvantages. Which ones one prefers depends on one's preferences. That is why politicians try to guess what the majority of voters' preferences are and act accordingly. In Britain, voters voted against expanding public debt. That is why the Tories won and the people SWL sympathizes with lost. So the population has to bear the consequences of voters' decisions. That's democracy. That does not mean that the majority is always right. Democracy just means accepting majority decisions (if they do not unduly infringe individual rights). The next election will let a number of people change their minds and so the governing majority. Economists are unhappy with that because they believe themselves to be in possession of the truth. As soon as they give up that belief - as SWL does - they must compete with all other political groups - Corbynites, Brexxiteers, Remoaners, Trumpists, Schaeubles etc. who believe to be in possession of the truth. And SWL and his ilk have not that much to show for themselves.

    1. We have a representative democracy, and the representatives are supposed to take expert advice. When they ignore that advice, they alone should be blamed.

    2. Take advice from people who don't know much and just advise "Borrow, borrow, borrow, spend, spend,spend" just to see what happens when the economy runs hot?

      Be serious.

    3. Would you buy a used car from such ignorant "experts"?

      Well, yes, if the price is low enough to take care of all risks. But economists don't have the money to reimburse the damages their advice can cause. Since decisions usually have to be taken in incomplete knowledge of the facts, it is your preferences which lead to your decisions. So if economists are incapable of giving convincing answers, what else can you do?

  11. I believe there are three secular trends which are now asserting themselves and will provide a substantial influence on growth in the coming years: demographics; innovation (which you mention) and adverse trends in energy costs.

    As regards demographics these trends are clear and lower consumption is not just for the retired but those in the 55+ age which is growing rapidly. This is a major issue.

    As regards innovation I find Robert Gordon's arguments very persuasive. They are in fact common sense in a way; there is no reason whatever to expect a constant stream of technological progress supporting growth and living standards; we appear to be heading into a fallow period, notwithstanding the growth of robotics and AI which may reverse some of these trends.

    As far as energy costs are concerned I don't know if you are familiar with the acronyms ECoE (Energy cost of energy) and ERoEI (Energy return on energy invested) as part of surplus energy economics. This shows that adverse trends in the cost of energy are, and will, make a substantial difference to both growth and living standards in the coming years.

    What I fail to understand is not these issues but how we can continue to project growth at historical levels in the light of them; it makes no sense to me. It seems that policy via targetting aggregate demand is targetting a false and unattainable position.

    1. Maybe you are right (there are strong counter-arguments), but what about my point about the implicit pre-Recession boom?

    2. I'm not sure about that but one other point did occur to me which interrelates the items I mentioned.

      The demographic shift which I mentioned will almost certainly result in lower aggregate demand. But you say, quite reasonably, that if aggregate demand shifts to a lower trend, clearly implicit here, then investment may fall in the light of that. But will it?

      If you take robotics as an example if the savings in labour costs are large, which they are in many cases, then, even in a situation of quite significant downtrends in demand it can pay to invest. Furthermore, with robotics, it is not just labour costs that are affected here; you need less land (car parking); less ancillary personnel etc.

      At the macro level there may be a continuing functional relationship between aggregate demand and investment (common sense) but at the sector level (which will not longer be just manufacturing but the much larger service sector) that picture may be quite different and it may be large enough to affect the overall relationship.

  12. Simon but isn't the elephant in the room here the role that credit expansion- primarily through mortgage equity release- played in keeping growth on that trend line as wages growth stagnated?

    When the crisis hit credit issuance dried up, property prices dived and business investment followed. The credit motor which has driven the rise in growth just stopped.

    You can see the long period of stagnation since then as partly a consequence of long term deleveraging and the recent pick up as evidence that as credit issuance surges again so growth picks up.

    This of course has been the position of many heterodox economists for years but even IMF economists using DGSE models have also made these arguments.

    If is the case then the solution to restoring sustainable growth has to be increasing the bargaining power of Labour (as the IMF paper says) as well as changing the structure of the economy so that we have more high skill, high paying jobs.

    1. But the impact of a credit boom on inflation is normally clear - see the Lawson boom at the end of the 1980s. So why nothing similar before the financial crisis? To the extent that there was a credit boom (clear in US, much milder in UK), it seemed to involve assets (housing) rather than goods and therefore GDP.

    2. Maybe because other forces helped to keep down inflation? Some people argue that the entry of a billion Chinese workers into the Global economy had a big deflationary impact. Others have suggested that high immigration and weak Labour unions meant that inflationary pressures were damped down.

    3. «But the impact of a credit boom on inflation is normally clear - see the Lawson boom at the end of the 1980s. So why nothing similar before the financial crisis?»

      As the BoE and others were very pleased with, massive imports from low-wage economies, and massive immigration from low-wage economies had a strong deflationary impact.

      «To the extent that there was a credit boom (clear in US, much milder in UK), it seemed to involve assets (housing) rather than goods and therefore GDP.»

      But of course a credit driven asset boom (just like a credit driven import boom) drives up GDP: it both encourages the supply of assets, that in the case of land and buildings take labour and resources to prepare and build, and also credit allows the cashing in of capital gain via for example remortgages. Some recent numbers:
      «About 28,900 loans were granted to home movers in October, a 20% decrease on a year ago and 8% down month on month. Home movers borrowed £5.9bn, down 9% on a the previous month and 18% on a year ago.[ ... ]
      Remortgage activity totalled £6.1bn, up 11% on September and 7% on a year ago. There was also a rise in the number of loans to 34,700, up 10% month on month and 5% on a year ago.»
      «Homeowners are increasingly using their properties like a bank account by remortgaging to free up cash and make life easier. The typical remortgage loan topped £170,000 for the first time in July – and applications were up by more than a third year on year.»

      Plus historical data:
      «Under Thatcher, this exploded to over £250bn across her premiership – a staggering 104% of GDP growth. ... But Blair did his homework and let loose – as did Thatcher – a wave of cheap credit, financial deregulation, house price inflation and an equity withdrawal-led consumption boom. Withdrawals under Blair’s leadership totalled around £365bn, that’s a full 103% of GDP growth over the same period,»

      This allows then cashing in of capital gains without even selling the asset.
      A lot of remortgage cash goes to buying goods and services, for example new furniture, school fees, a new car, a conservatory.

    4. And silly me I forgot a significant but hard to quantify aspect of asset price booms on spending: someone who gets a large yearly capital gain usually don't need to get a remortgage to turn it into cash: they can just cut their savings. Someone on a gross income of say £30,000 and gets a £10,000 tax-free capital gain every year for the past 35 years may feel no need to save for a pension, for example, and thus spend more of their income. I'll also quote an ex-member of the FOMC:
      «What the Fed did, and I was part of that group, we frontloaded a tremendous market rally starting in march of 2009. It was sort of a reverse Wimpy factor. Give me two hamburgers today for one tomorrow. [...] Once again, we frontloaded, at the federal reserve, an enormous rally in order to accomplish a wealth effect.»

  13. Given that Government spending has multiplier impact on GDP - how much of the gap can be directly explained by Government imposed austerity and reduced interest payments going back into circulation?

    1. One way of putting my argument is that the multiplier on this occasion was much greater than normal, because long term expectations changed.

  14. When you say "we really have no idea" you are speaking for the mainstream. Those of us in the heterodox or structuralist tradition see the failure to recover as not surprising and in fact exactly what our theories predict. We see growth as demand- or capital-constrained and are skeptical of the neoclassical belief in the full employment (of labor) of a pre-existing labor force.

    1. I think maybe the problem here is that these heterodox economists are clearly wrong on the up-side (booms lead to inflation), but maybe right on the down-side once interest rates hit zero and we have austerity.

    2. Well, no. The 1990s boom in the US did not lead to inflation. It led to high employment, a pause in the rise of inequality, and growth. I think there is a lot of agreement that hysteresis is a real possibility or even likelihood on the down-side. Why not on the up-side? (For readers who are curious, heterodox economists would be people like Lance Taylor or Marc Lavoie who you can can google. Marc has a piece at INET worth reading.)

  15. when capital costs are so low, businesses invest in capital at the expense of labor, thereby employing fewer workers. A reduction in employment further reduces demand. It's not just automation that is eliminating jobs, but the cost of capital, because higher interest costs require a greater ROI.

  16. "Every time I look at the data above, I ask whether we have brought this on ourselves by a combination of destructive austerity and timidity."

    The short answer is Yes. Infrastructure spending by the new Tory government, rather than austerity, would have had a positive impact on growth. The timidity was from Labour politicians not pointing this out.

  17. No post-GFC recession in Australia! In 2009, Commonwealth (Labor) government gave each household $900 as direct stimulus, and engaged in a country-wide all-schools (both public & private) buildings program, irrespective of whether the schools needed the extra buildings. This stimulus and decreases to the RBA's bank rate to the current 1.5% seem to have done the trick.

  18. “The traditional answer in macroeconomics is nominal deflation: falling wages and prices. But because workers have already priced themselves into jobs, nothing more will come from the wages route. So why would firms cut prices?”

    My usual take: why don’t progressive economists ever think outside the box to (or at least include in) the alternate possibility that labor in the US could (in the abstract at least) be at German level (or at least highest historical US level) union density — which would reset every other economic (and political) calculation?

    Which leads to a wider (I think wholly PSYCHOLOGICAL question — the way the human mind does not work):
    Why can economists have no trouble (could not possibly leave out) citing Euro debacle v. never happened Euro debacle — European austerity v. never was European austerity …

    … yet don’t just as automatically bring up US disappeared labor union density v. never disappeared US labor union density …

    … in the same automatic, inherent fact/counter factual way?

    Must I verbally emphasize automatic — why must I emphasize? Labor union disappearance is the root in almost all the rot in our society: political as well as economic.

    Just a (I think wholly) psychological thought. Something about being blinded by (trapped in?) the culture we live in. ???

  19. A path dependence argument that seems stretched. By this logic, the Great Depression produced a series of miserly people (think Sewell Avery of the 1930s chain Montgomery Ward, read this background for you UK readers) who never really ever wanted to 'waste money' and spend again, after being traumatized by the 1930s. Certainly that explains a certain class of people of that generation, but to say an entire country is under this spell after a short 2 year Great Recession seems stretching it. And even the USA recovered to trend after the Great Depression. So a structural, non-Keynesian argument seems more plausible.


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.