Thursday, January 1, 2015

Secular stagnation: a neo-paleo-Keynesian perspective

I first posted this piece back in January but it got deleted by from my blog by mistake. Since secular stagnation is back in the blogosphere with a vengeance: its time to repost it.

In a recent piece on his blog, David Beckworth has taken another swing at the secular stagnation hypothesis. Secular stagnation is a term coined by Alvin Hansen in a 1938 article in which he claimed that public expenditure might be required to maintain full employment.

Here is Alvin, as quoted by David...
"The business cycle was par excellence the problem of the nineteenth century. But the main problem of our times, and particularly in the United States, is the problem of full employment. ... This is the essence of secular stagnation— sick recoveries which die in their infancy and depressions which feed on themselves and leave a hard and seemingly immovable core of unemployment."
Hansen is writing in 1938, before Keynesian economics had been forever altered by Samuelson's bastardization of Keynes' key idea: that high involuntary unemployment is an equilibrium that can persist for decades. 

The secular stagnation hypothesis was resuscitated by Larry Summers.  In Larry's view
... it is increasingly clear that the trend in growth can be adversely affected over the longer term by what happens in the business cycle.
Larry supports his position with the following graph.

Figure 1: Downward Revision of Potential GDP 
Growth is the percentage annual increase in real GDP.  In terms of Figure 1, it is measured by the slope of the lines that represent estimates of potential GDP made in 2007 (the dark grey line) and in 2013, (the light grey line). Try as I may, I do not see compelling evidence of a change in the slope. 

What I do see in Figure 1, is evidence that a business cycle shock (the Great Recession) has caused a permanent downward shift in the level of GDP. And that is tragic because if Alvin Hansen is right, and I think he is, the gap between these two lines represents an annual loss of output of approximately one trillion dollars.  

Contemporary accounts of secular stagnation, beginning with Larry Summers, confound two distinct ideas. The first, and this is Hansen’s meaning, is that a market economy, in the absence of counter-cyclical fiscal and monetary policies, may experience prolonged periods of high involuntary unemployment. The second,  is that a market economy may experience a period of depressed productivity growth.

David is critical of the second of these two ideas. Here is David...
"Hansen’s article was of course spectacularly wrong as a guide to the next few decades. Instead of suffering through stagnation we entered an extended, broad-based, and massive economic boom. In hindsight we can see that his analysis, ... was unduly influenced by the depression he was living through, ... [which] was the result of specific policy mistakes rather than inexorable trends. Recent research by Alexander J. Field shows that the 1930s were actually a time of exceptionally high productivity growth. " 
David claims that postwar productivity  growth was high. That may be true. But it says nothing about secular stagnation, which is the idea that there has been a permanent long-term increase in involuntary unemployment. Back to David:
"The fact that Hansen was wrong does not prove that contemporary stagnationists are. In this case, though, history is repeating itself rather exactly. We do not pretend to know what the future path of economic growth in the United States will be. But the case for stagnation is weak—and, as in the 1930s, it is getting undue credence because of a long slump caused by policy mistakes."
Not so. David is confounding growth, the annual percentage increase in real GDP per person, with unemployment, the percentage of workers who claim to be looking for a job but are unable to find one. Growth remains a topic on which, as economists, we are spectacularly ignorant despite the fact that there is a large subset of  macroeconomists who have  studied very little else for the past thirty years. Unemployment, is a topic on which we have made considerable progress.

Were there policy mistakes?  Most certainly.  Would a more effective way of running monetary policy be to replace inflation targeting with a nominal GDP target as the market monetarists have claimed? I am skeptical.  

It is a premise of the monetarist position, that the real economy is self-stabilizing and that a rule based monetary policy is the most effective way to ensure both low inflation and maximum sustainable employment. 

Keynes claimed, in contrast, that the real economy can get stuck in a position of high unemployment and that permanent high involuntary unemployment can persist as an equilibrium phenomenon. See my earlier post on the neo-paleo-keynesian perspective. If Keynes is correct, and I believe he is, a single instrument, monetary policy, is not enough to hit two targets. Fiscal policy in one form or another, is an important second string to the policy maker's bow.


  1. Wouldn't the SWF you and Miles Kimball discuss help stabilize both growth and unemployment?

    1. So why don't we hear about this from other prominent economists? I don't see anyone else discussing it at all. No arguments for or against. It sounds like a great idea to me, but the economics discipline doesn't seem to know or care about the option.

  2. We will now hold our collective breaths waiting for our self-inflicted government to wake up and smell the stimulus. As if they gave a dam about the unemployed in the first place.

  3. Do you support a job guarantee program?

    1. Government has an obligation to maintain a low and stable inflation rate and to promote full employment. In contrast to much of recent macroeconomic theory, I do not believe that a market economy will generate jobs for everyone most of the time, in the absence of well designed fiscal and monetary policies.

      Two goals, full employment and low inflation, requires two instruments. One of these instruments is a monetary rule and I see no reason to think that an NGDP target will perform better at keeping inflation on track than inflation targeting, which was spectacularly successful in the period preceding the Great Recession.

      For the second policy instrument, I do not favor traditional fiscal policy. In my view, explained Here, fiscal policy should operate in the asset markets to eliminate inefficient swings in asset prices that are the root cause of financial crises.

  4. I'm not sure about NGDPLT either because even if it were theoretically optimal in some models there are practical issues. However there is a crucial piece of wisdom behind this change in target which is that inflation may be required to rise above 2% for periods of time for labour and investment markets to clear. A 3% or 4% percent Price Level Target would make a lot of sense (I think the MMs accept this as a compromise).

    In my mind this is due to the simple fact that there are not always enough investment available out there with a risk adjusted real return above -2%. We need the -3 or -4% investment to not be priced out of the market by money. The current alternative is for banks and people to store their wealth as idle fiat reserves which has a return to the economy of around -100% over the period that it is made idle.

    Of course cash holders don't get -100% return since if the central bank maintains prices below their target, cash savers get to keep their claim on other people's wealth even if they didn't participate in creating or maintaining this wealth. This means that later, when people want to draw down their cash savings and the idle money starts moving, central banks have two choices. They can let prices correct upwards, devaluing cash savings (which were not backed by any real value anyways) or they can, and this is usually their stated actual goal, maintain the illusion and instead use a tight monetary stance which devalues the savings of those who invested in the real economy.

    Fiscal stimulus alone may help restore demand, but you might need to do a huge mount for long periods of time if the monetary targets are not high enough and the money multiplier is not on your side. Plus the demand mix can get skewed, for example, toward consumption spending instead of investment spending if the government doesn't spend on things that are long lasting enough and doesn't match the consumption timeframe needs of cash savers. Temporarily higher inflation targets would allow a market for the negative return investments and you might get better resource allocations.

    It seems that sometimes people see negative returns as unnatural aberrations. This might be because of the fact that the 20th century was one of unprecedented technological and demographic growth which made low risk positive returns easy to find.

    The law of thermodynamics tells us that most things decay over time unless you input work and energy into them. Widely available low risk positive real return investments may be a thing of the past. We should not let overvalued fiat price out the type of investment that may be emerging as a crucial part of the savings market.

    1. Benoit. In my view, (see the link in my response to Random), high involuntary unemployment has nothing to do with inflexible prices and wages. We have plenty of evidence that wages and prices move both up and down.

      Market economies are very good at allocating the right number of left and right shoes. They are very bad at allocating credit from lenders to borrowers. Hayek was right to point to the fact that the seeds of the crash are in the boom. He was wrong to assert that there is no role for a government policy that can improve on an unregulated market outcome. The right way to implement a full employment policy is to use fiscal policy to stabilize asset prices both on the upswing and in the crash.

    2. You don't see a problem for a significant and growing portion of investments being priced out of the market by cash? When fiat money cuts off part of the market and makes it artificially uncompetitive, fiscal maneuvers will have relatively weak effects. Government will basically need to replace all the missing investments with its own having the right mix and right maturity structure matching future consumption needs of cash savers. Seems like a difficult task for politicians to get right compared to just unjamming the markets with a bit of temporary inflation.

  5. Some of us Keynesians want to try NGDP path level targeting and more fiscal policy. We want to try everything, it's not either/or.

    1. I am sympathetic to that position. But it's a good idea to have an understanding of the sickness before recommending a remedy. The sickness is in the asset markets. So is the cure.

  6. In a world with central banks making policy what does it even mean to say there are multiple equilibrium? the only thing that theoretically would matter is whether the economy is above or below the targets of the central bank and what the policy response will be.

    Doesn't anyone who supports the use of monetary policy at least over some period less than the long run agree that the economy will not return to 'equilibrium' which is just another way to say there is more than one equilibrium?

    Why not use fiscal policy to provide those things fiscal policy should provide? and simply do that at the level that maximizes potential gdp or exhausts all npv positive expenditures or some such measure? and leave the activity level of risk taking to be driven by interest rates to the level that uses available resources.

    the whole idea of 'fiscal policy' has a major flaw because it's unanchored. How can you call for fiscal policy without having some view of what structural level of fiscal spending is optimal or desired?

    Who has such a model?

    1. Dan
      There's a lot to respond to here. First, on multiple equilibria. A model has multiple equilibria: the real world does not. In the real world, people act on their beliefs. My use of multiple equilibrium models is a way to introduce the idea that animal spirits, confidence, self-fulfilling beliefs ( whatever you want to call them) are independent drivers of the unemployment rate.

      Fiscal policy, in the sense of government spending on roads and bridges, should be designed to satisfy sound cost-benefit principles. Who pays for those roads and bridges is another matter. By operating in the asset markets, the central bank and the treasury reallocate resources between generations and influence asset prices in every risk class. It is these operations that I refer to when I advocate fiscal policy as a second string to the bow.

    2. Perhaps I don't quite follow the idea of reallocation, but investment doesn't strike me as reallocation.
      doesn't every period win, obviously assuming the investment has a positive present value?
      The government can choose to invest today in educating children and the resources available in the future are larger than they'd be if the investment hadn't happened. But, by our definition here, the present value of that investment is positive, so the current generation is better off as well.
      current consumption is possibly being deferred but I suspect that might depend on how the investment is financed - money creation or CA deficit for domestic borrowing.
      I suppose as well there will be some impact on asset pricing, though I am not sure why you'd aim for that when the aim is to make some investment. It seems to me asset prices would feed into the calculation as an input not an output.

      according to Keynes the government doesn't play by the rules of the IS-LM (I think that was his insight) he concluded you could thus with monetary and fiscal policy influence the private sector to sort of push the IS-LM to some desired place. But, what rules are the government spending subject to?
      It may make sense in any case that government wouldn't be ruled by interest rates, because interest rates ration risk taking and resources in situations where the benefit can be claimed by the agent making the investment. While government makes many or mostly investment in public goods where the risk profile of the investor is irrelevant as the benefit can't be captured by the investor by definition.

    3. The present value of an investment depends on the market discount rate which fluctuates from one month to the next and is reflected in swings in asset prices. It is determined by the willingness of different groups of people to lend to each other and by the willingness of entrepreneurs to borrow money to create new buildings and machines.

      Classical theory holds that the discount rate fluctuates because rational forward looking people anticipate that the funadmanentals have changed. Keynesian theory (as I interpret it) holds that the discount rate fluctuates because different groups of people become more or less confident about the future. These swings in confidence have nothing to do with fundamentals. But they result in waves of construction and investment that have long-lasting consequences. There are too many houses in Spain because of a ten year boom in which Germany and China accumulated claims on Ireland, Spain and Greece.

      Swings in asset prices are inefficient because future generations, those who suffer the consequences of asset market instability, are not around to actively participate in the asset markets. By buying and selling assets on their behalf, governments have the ability (and I would add, the obligation) to prevent swings in market discount rates that result in the devastating consequences that we experience as financial crises.

    4. I see a lot of commentators think investment is always a good thing - referring to plant, equipment, machinery, etc. I think it is important to realize that one component of investment is inventories. When we talk about increasing investment we sometimes forget that this can happen by a drop in consumption. The consumption goods then remain in inventory. Since businesses didn't anticipate the drop in consumer demand they are stuck with too much inventory investment. In subsequent periods businesses rectify this by layoffs and cutbacks in production. If this becomes the new steady state, the economy becomes stuck in an undesirable equilibrium - as professor Farmer explained.

  7. Thanks for this, Prof. Farmer. Please also have a look at this contribution on the subject.

  8. Roger,

    I was surprised you didn't mention inequality as a source of secular stagnation. If the top 1% continues to receive 2/3 of the total increase in aggregate income, and they have a much higher than average MPS, then you're going to get either weak demand and unemployment, or unsustainable borrowing, a financial crisis, and unemployment. This becomes a secular problem, in part, because inequality tends to be self-reinforcing as poorer households lack the collateral necessary to borrow and invest.

    Happy New Year, Roger!

    1. Part of the inequality is due to the difficulty of entry into some economic groupings.

      The spectacular success of some economic groupings has led to a proliferation of groupings. The economy is in semi-stranglehold with a network of such groups. The result is a lack of competition and forced pricing for nearly all the necessary elements of modern life.

      It would be fiscal policy that could restore competition and correct this aberration.

    2. Greg
      Inequality is an issue largely because of the undue influence that is conferred by wealth on the political process. I do not believe that the distribution of resources across different quintiles of the income distribution has a lot to do with deficient aggregate demand; although it does determine the kinds of commodities we produce. In Victorian England, wealth was very unequal, but inequality in itself did not lead to mass unemployment. The rich provided jobs for the poor as servants, gardeners, and handsome cab drivers.

      Is that a society I would like to live in, given a decision made from behind Rawls' veil of ignorance? No. But a democratic society gives me the means to alter the income distribution through redistributive taxation. The question of whether a society should impose redistributive taxes is separate from the fact that market economies are unable to maintain full employment in the absence of stabilizing policies pursued by our elected representatives. Both are legitimate questions for government which is, after all, nothing but the coalition of the whole.

      And a very happy and prosperous New Year to you too!

    3. Roger,
      I want to make sure I understand your point about the effect of income distribution on the composition of output. Are you're saying that the production of goods for the wealthy requires labor just like the production of goods for the rest of the population? I don't disagree with that.

      The Paleo-Keynesian point I was trying to make is that, if the wealthy save rather than build mansions, aggregate demand will suffer. I believe Keynes, himself, made this point in the last chapter of the GT, arguing that the growth of capital no longer depends on the saving of the wealthy, and that a more equal distribution of income would increase the propensity to consume, which, in turn, would actually create more favorable conditions for investment. But perhaps I missed something.

    4. Greg
      No, I don't think you missed anything. I'm disagreeing with Keynes. I guess that's why I'm a neo-paleo-Keynesian and not just a regular paleo kind of guy.

    5. Distribution may not be a problem in the sense you are using if the market is complete. But markets are never complete (and cannot be). Even in advanced economies some 30 per cent or so of production is carried on ex-market. The threshold is set partly by the cost of entering the market at all, which is usually not trivial. In the 19th century, for instance, a significant fraction of the population were not employed because they were simply too malnourished to work. They could not afford the food to lift their physical condition to where they would be employable. In our times, a considerable number cannot afford the threshold costs (car, fuel, stable, housing, presentable clothes etc) to win jobs that would cover these costs. So really unequal distributions tend to push people out of the market - into the household economy, begging or the informal economy.

      This is another of Keynes directions of argument, IIRC. He thought that increasing state direction of investment would be needed to avoid both distribution issues and the tendency of private capital to focus on money rather than total human returns. It's a line of thought which looks very plausible when you consider environmental issues.

  9. Here's a distinction which I think is not being made quite clearly enough.
    The problem is being identified as a lack of aggregate demand, and then the underlying issue is being identified as private sector risk taking is too low.

    So some people look and say, well you can increase AD by either increasing C I or G so C and I aren't increasing so let's by policy increase G. Naturally that increases AD by definition.

    Then other people come along and say, well that's nuts C & I are too low so you increase G? how does that make sense, if C & I are the problem fix C & I. And we already know how to do that; use monetary policy - however implemented given the details of the moment.

    Then you come along and say, well hold on. you can't have AD without C I AND G so we need both monetary and fiscal policy.

    But here's the problem basically you have everyone talking past each other.

    I think you'd get more traction if you parsed the difference - even though the distinction is at best blurry.

    if you want to fix C & I: Use monetary policy
    if you want to fix G: use fiscal policy (!)
    and if you want to optimize AD: you need to optimize all three C I and G

    However, as far as I know, no one has ever tried to answer the question what is the right level of G to maximize C & I.

    In fact you get confused analysis like ricardian equivalence and intergenerational and intra generational 'redistribution' .

    But logically unless you want to argue that zero government is the right amount; you have some G which increase C & I and is not time shifting or distributive but is outright additive.

    But what level of G?

    1. Have you not followed the discussion of the ZLB? At the zero lower bound, monetary policy gets no traction and you can't fix C and I. G is one way to stimulate C and I at the ZLB.

  10. The are various ways in which inequality could affect the prospects of secular stagnation construed as a long run (many years) of unemployment, but I was thinking about a very simple aggregate demand problem.

    Imagine an economy where, in 1975, the top 10%’s income share was 30% of total income, but by 2015, the top 10%’s share had grown to 50%. Assuming the top 10% have much higher-than-average propensity to save, then, all else equal, the demand for consumption goods is going to be lower in 2015 than in 1975.

    In the classical story, interest rates will be lower and investment higher in 2015, so there’s no problem. In the Keynesian story, interest rates may be lower too, but, leaving technological progress aside, investment will only be higher if firms expect higher demand in the future. However, unless debt-financed public expenditures, debt-financed private expenditures, net exports, and/or some other variables are higher in 2015 than in 1975, an increase in investment sufficient for full employment seems unlikely.

  11. It's great post! Thanks for sharing!

  12. I can't get away from the thought that the demographics of an economy come into play when considering secular stagnation. Don't accumulating senior citizens become a counter balance to growth as they downsize out of their previous positions as consumers and productive workers? In which case it becomes harder and harder to create a self perpetuating stimulus because no matter how many roads or bridges you build, a growing percentage of the population will never have any use for them. Unless a surge of immigration or birth rates are injected into the economy in question, then as it ages, the economy will want to shrink rather than expand no matter how much stimulative intervention is used.

    If true, this raises enormous international policy questions regarding cultural identity and sovereignty. Do rentiers have a say in the long-term demographic prospects of an economic entity such as Japan (can they force the Japanese to expand immigration or fertility in order to unsure bond repayments), or must rentiers be forced to accept that the stimulus they have underwritten to date will never be repaid because the economy is destined to age and continue its habit of degrowth?

    1. Not sure why that posted as Unknown. Posted by William "BT" Hathaway, founder,


Note: Only a member of this blog may post a comment.