Saturday, December 13, 2014

Real business cycle theory and the high school Olympics

I have lost count of the number of times I have heard students and faculty repeat the idea in seminars, that “all models are wrong”. This aphorism, attributed to George Box,  is the battle cry  of the Minnesota calibrator, a breed of macroeconomist, inspired by Ed Prescott, one of the most important and influential economists of the last century.
All models are wrong... all models are wrong...

Of course all models are wrong. That is trivially true: it is the definition of a model. But the cry  has been used for three decades to poke fun at attempts to use serious econometric methods to analyze time series data. Time series methods were inconvenient to the nascent Real Business Cycle Program that Ed pioneered because the models that he favored were, and still are, overwhelmingly rejected by the facts. That is inconvenient.

Ed’s response was pure genius. If the model and the data are in conflict, the data must be wrong. Time series econometrics, according to Ed, was crushing the acorn before it had time to grow into a tree. His response was not only to reformulate the theory, but also to reformulate the way in which that theory was to be judged. In a puff of calibrator’s smoke, the history of time series econometrics was relegated to the dustbin of history to take its place alongside alchemy, the ether, and the theory of phlogiston.


How did Ed achieve this remarkable feat of prestidigitation? First, he argued that we should focus on a small subset of the properties of the data. Since the work of Ragnar Frisch, economists have recognized that economic time series can be modeled as linear difference equations, hit by random shocks. These time series move together in different ways at different frequencies.

For example, consumption, investment and GDP are all growing over time. The low frequency movement in these series is called the trend. Ed argued that the trends  in time series are a nuisance if we are interested in understanding business cycles and he proposed to remove them with a filter. Roughly speaking, he plotted a smooth curve through each individual series and subtracted the wiggles from the trend. Importantly, Ed’s approach removes a different trend from each series and the trends are discarded when evaluating the success of the theory.

After removing trends, Ed was left with the wiggles. He proposed that we should evaluate our economic theories of business cycles by how well they explain co-movements among the wiggles. When his theory failed to clear the 8ft hurdle of the Olympic high jump, he lowered the bar to 5ft and persuaded us all that leaping over this high school bar was a success.

Keynesians protested. But they did not protest loudly enough and ultimately it became common, even among serious econometricians, to filter their data with the eponymous Hodrick Prescott filter.

Ed’s argument was that business cycles are all about the co-movements that occur among employment, GDP, consumption and investment at frequencies of 4 to 8 years. These movements describe deviations of  a market economy from its natural rate of unemployment that, according to Ed, are caused by the substitution of labor effort of households between times of plenty and times of famine. A recession, according to this theory, is what Modigliani famously referred to as a ‘sudden attack of contagious laziness’.

The Keynesians disagreed. They argued that whatever causes a recession, low employment  persists because of ‘frictions’ that prevent wages and prices from adjusting to their correct levels. The Keynesian view was guided by Samuelson’s neoclassical synthesis which accepted the idea that business cycles are fluctuations around a unique classical steady state.

By accepting the neo-classical synthesis, Keynesian economists had agreed to play by real business cycle rules. They both accepted that the economy is a self-stabilizing system that, left to itself, would gravitate back to the unique natural rate of unemployment. And for this reason, the Keynesians agreed to play by Ed’s rules. They filtered the data and set the bar at the high school level.

Keynesian economics is not about the wiggles. It is about permanent long-run shifts in the equilibrium unemployment rate caused by changes in the animal spirits of participants in the asset markets. By filtering the data, we remove the possibility of evaluating a model which predicts that shifts in aggregate demand cause permanent shifts in unemployment. We have given up the game before it starts by allowing the other side to shift the goal posts.

We don't have to play by Ed's rules. We can use the methods developed by Rob Engle and Clive Granger as I have done here. Once we allow aggregate demand to influence permanently the unemployment rate, the data do not look kindly on either real business cycle models or on the new-Keynesian approach. It's time to get serious about macroeconomic science and put back the Olympic bar.

31 comments:

  1. Roger, nice post. speaking of mass unemployment as an attack of laziness as per Modiglani, Blanchard attribute the much lower labour force participation in the EU since the 1970s to their greater preference for leisure. According to Blanchard:

    "The main difference [between the continents] is that Europe has used some of the increase in productivity to increase leisure rather than income, while the U.S. has done the opposite."

    An unusual left-right unity ticket emerged to explain the great depression in the 1930s and the depressed EU economies from the 1970s: the great vacation theory.

    ReplyDelete
    Replies
    1. Thanks for your comment Jim. I actually agree with Olivier on that point. Its important to distinguish labor force participation from unemployment and I am prepared to at least entertain the idea that participation is influenced by national characteristics.

      Delete
    2. Really? The preference for leisure is the explanation? The hours worked in EU basically collapsed between early and late 70's. They used to work more than Americans and all of a sudden they worked less. Substantially so. Within 6-8 years. Did productivity explode in the 70's?

      Delete
  2. A brilliant summary of developments in real business cycle theory is that the Minneapolis Fed by Douglas Clement at https://www.minneapolisfed.org/research/pub_display.cfm?id=5386

    ReplyDelete
  3. I guess we have a different definition of "brilliant". This is a survey of Minnesota apologetics.

    ReplyDelete
    Replies
    1. When official data better account for intangible capital, labor productivity figures are likely to cycle up and down with the broad economy, which is consistent with real business cycle theory both in the 1990s and after 2007.

      Delete
    2. Keynesian macroeconomics postulated that the economy slips into recessions for all sorts of reasons such as shifts and turns in the animal spirits and a loss of consumer confidence leading to a fall in autonomous investment and autonomous consumption. A collapse in autonomous investment and autonomous consumption is the Keynesian explanation for the great depression.

      Both Keynesian macroeconomics and real business cycle theories, at least at the outset couldn’t explain why there were recessions.

      Both attributed to them to causes they were yet to explain. Keynesian macroeconomics could not explain what drove the waves of optimism and pessimism that either sharply increased or reduced investment.

      At least Prescott and other real business cycle theorists accepted that they must eventually unpack productivity drops and name causes that can be explored further and perhaps found persuasive or perhaps wanting.

      As research progressed, real business cycles were viewed as recurrent fluctuations in an economy’s incomes, products, and factor inputs—especially labour—due to changes in technology, tax rates and government spending, tastes, government regulation, terms of trade, and energy prices.

      At bottom, Keynesian macroeconomics makes an unjustified assumption that technological progress unfolds at a relatively smooth rate.

      Delete
    3. Nope. You've been drinking Ed's cool aid. There is no presumption in Keynesian economics, (at least in MY Keynesian economics) that TFP is smooth. But is it an important component of low frequency movements in unemployment? I don't think so. I don't care about the wiggles. Bob Lucas convinced me that they don't matter. I care about the trend.

      Delete
  4. I suppose the ultimate proof of worth is if any of these theories has successfully predicted the future and helped us shape it rather than just made the past fit.

    ReplyDelete
    Replies
    1. Dead on. And the history of post-WWII macro is the triumph of Keynes in persuading politicians to shape the future. The disagreement is over whether their interventions improved our lived. My view? YES.

      Delete
  5. Roger: Is it true that..." the economy is a self-stabilizing system that, left to itself, would gravitate back to the unique natural rate of unemployment."?

    I have come to the conclusion that this question doesn't make sense.

    For a sufficiently silly monetary policy regime, there is no way the economy will be self-stabilizing. For a sufficiently sensible monetary policy regime, the economy will be self-stabilizing. The only thing worth arguing about is which monetary policy regimes are silly and which ones are sensible.

    Any idea that the economy is "inherently" self-stabilising, or not self-stabilising, if the central bank "does nothing", depends on how we define "doing nothing". Does that mean doing nothing with the nominal interest rate? Doing nothing with the money base? Doing nothing with the price of gold? Doing nothing with the unemployment rate target? Doing nothing with the inflation target? Or what? There are 1,001 different ways of "doing nothing".

    I think that sensible keynesians, and New Keynesians, would agree with what I have said above. (I think Simon Wren-Lewis, for example, who is a sensible NK, at least roughly agrees.)

    I expect there are some macroeconomists who would say that the economy is self-stabilising for *any* monetary policy regime, even extremely silly ones. And those economists, IMO, are totally wrong. But I don't think you can include any New Keynesians among them. New Keynesians, for example, will say that things will go very badly wrong if the central bank doesn't follow something like the Howitt/Taylor Principle.

    Other than that disagreement, I like your post!

    ReplyDelete
    Replies
    1. Thanks for you comments Nick. I agree with everything you said. Much of the history of capitalism is about the development of stabilizing institutions (you cite the central bank but there are many others) that prevent the market system from exploding.

      Delete
    2. The odds of any system of differential equations in which future quantities depend on past quantities and rates of change being stable is extremely small once you get past a few trivial cases. There's a reason electrical engineers put capacitors all over the place and mechanical engineers love shock absorbers.

      Unless economists are using some other kind of mathematics from the rest of us, assuming stability is ridiculous. Doing something or not doing something is just adding another piece or two to the system. The central bank or what have you is a part of the system.

      The best you can do is design something stable for a certain parameter space.

      Delete
    3. This. Mainstream economists need to stop relying on maximization-and-equilibrium and to start treating their subject as the Complex Adaptive System that it is.

      Delete
    4. Why? Is there any evidence that a model of a complex adaptive system performs better, in any dimension, than a linearized DSGE model?

      Delete
    5. @kahleberg The difference in economic systems is teleology. Humans are goal oriented. Electrons are not. If electrical engineers understood that distinction, they would be moving into economics with same frequency that Russian billionaiires are moving to London.

      Delete
  6. I think defining real business cycles= program as strictly about the stochastic Ramsey model with TFP shocks is very limiting. RBC models to me are models where only relative prices and real quantities are determined, and nominal quantities are indeterminate (i.e you can add various mechanisms afterwards to determine them without necessarily affecting real variables). This includes I think most of your models, since I don't see any nominal quantities, Angeletos et al's models of sentiment shocks, Rios Rull's et al's models of search frictions in product markets that lead to changes in matching efficiency that look like TFP shocks etc...The main innovation in all these models (that were to a large degree started by Kydland and Prescott's original RBC models) is the realisation that a lot of business cycles can be explained without focusing on monetary factors: real shocks and real/non monetary mechanisms (which can include search frictions, real wage rigidity due to say efficiency wages, credit frictions and various informational imperfections).
    Focusing on RBC model in the very narrow sense of 1980's Minnesota macro dilutes the maybe more revolutionnary message of this type of models that we should look for causes of inefficient recessions and crises beyond the neoclassical synthesis/IS-LM stories of disequilibrium and people who just refuse to lower prices and wages.
    p.s: the Prescott/McGrattan paper from 2014 has a wider interpretation that Minnesota apologetics. The general point is that the TFP/efficiency wedge has been mismeasured, and labour market frictions have probably been overmphasized. The matching indeterminacy you exploit in your models could also apply to product markets (e.g in bargaining between retailers and wholesalers or producers). In that case it could deliver a theory of endogeneous TFP movements. I'd be interested if you had a model like that.

    ReplyDelete
  7. @daniels First: on the REAL aspect of RBC. That, I think, was a simplifying assumption. I see Kydland Prescott as a watershed that ditched the Geneal Theory and went back to Pigou's 1928 (Industrial Fluctuations) view of business cycles. Pigou's theory was rich, but verbal. He had five or six sources of shocks including agricultural shocks, productivity shocks, monetary shocks, news shocks and confidence shocks. The RBC program was to redo Pigou using functional equations and dynamic programming. Over the next thirty years, economists, myself included, added back the shocks to Pigou's framework. In 2007, at the dawn of the Great Recession, we had a pretty good description of the macroeconomy that worked reasonably well at understanding post-war business cycles (see for example, Smets and Wouters). The Great Recession led many of us to question that framework. It's now time top reintroduce the Keynesian idea of multiple steady state unemployment rates, and to retool the Keynesian program using modern techniques, just as KP retooled the Pigovian model.

    As for Prescott-McGrattan. In my view, the idea of a 'wedge' is simply doubling down on a model that it is clearly wrong. To model business cycles we must understand unemployment, not labor force participation. When the model is falsified, throw out the model. Don't add a 'time varying wedge' which is just a metaphor for everything that is wrong with the theory.

    ReplyDelete
    Replies
    1. I think you are too fast to dismiss business cycle accounting. If you are a hard core believer that all models must be fully robust to the Lucas critique, then sure it dn't make sense to look at semi structural models such as a neoclassical growth model with wedges. But in terms of a framework for organising a lot of different models with many different frictions (yours being one of many out there right now competing for the confused applied macroeconomist's attention), and in a world where building a general model of all the possible interactions of all the possible frictions is impossible, having a semi structural business cycle analogue to Solow's growth accounting can be useful to see the links between different frictions and richer models, and to organise them through some sort of simplified core model.
      The point remains: models that focus on labour markets may have been overemphasized recently, especially once you look outside the US- e.g the Eurozone. Maybe more thought should be given to frictions that reduce the efficiency of product markets (and that again are not the typical sticky price friction). These product market distortions tend to map into an efficiency wedge (or how else would you call a friction that reduces output even if there were no unemployment?). Again, using bargaining indeterminacy, you could get what look like permanent TFP shocks.

      Delete
  8. i was reading a bio of AFL leader meany last night; he and the afl cconsistently advocated for gov't jobs programs and housing, on the grounds that it was the gov'ts responsibility to provide
    meany hated communism; consistency is indeed a hobgoblin

    PS: can you fix your code ? I typed in a much nicer version of above, then selected google login, then hit publish...you lost my comment ???

    ReplyDelete
    Replies
    1. The difference between Meany and a communist is that Meany wanted business to actually build the housing and provide the jobs, but it was the government's job to make sure people could demand enough pay to afford them. Right now, the government gives business everything it wants, but demands next to nothing in return.

      If you look at the last 30-40 years, the US is caught in the same trap as the USSR 30-40 years ago. It can't raise living standards without some sacrifices on the part of the ruling class. Read some good books on the economics of the old USSR. The Europeans wrote lots of them, so they weren't surprised when the USSR suddenly lost its empire like some central intelligence [sic] agencies I could name.

      Delete
    2. And Meany was a guy who fought for those things, and lead a union at a time when unions had substantial clout.

      Delete
    3. "meany hated communism; consistency is indeed a hobgoblin"

      I'm amazed at the number of people who don't know what communism is/was.

      Delete
  9. There was this recent study showing, purportedly, that there's little ideological bias in economics.

    First, and this is a big theme in academic economics, you're rewarded for producing the research (and with gigantic carrots and sticks), but little, if at all, for really understanding well it's implications for the real world and policy. Actually really understanding the economics that's out there and being able to apply it well to the real world and policy, to interpret it well, gets you nothing if you aren't publishing new papers. And smart explanation, interpretation to the real world and policy, is not a pub that gets rewarded. So, you often have horrible interpretation to reality and policy by super positioned and compensated economists.

    Second, the bias is not in typical political English language, what that study searched for; it's in things like this; the techniques and modeling that are permitted, which so often are things that make libertarianism/plutocracy look so much better, while so much that makes libertarianism/plutocracy look so harmful are not permitted, or severely penalized in publishing and positions and rewards.

    Great case in point -- When do you ever see models that include positional externalities and also just even let readers know, in a completely positive way, what the total societal utils optimum is. Because you do this, with sensitivity analysis, and suddenly libertarian/plutocratic policy looks nightmarish. And that cannot be allowed to be seen by so many who control prestige journals and departments. All that is usually allowed is the Pareto optimum, which is complete libertarianism -- the only policy that's permitted to be considered, the only option we can know about, is the, only with unanimous consent, pure libertarianism, option. And positional externalities, the pink elephant of economics, are ignored.

    It's this kind of censoring where the bias lies, and is incredibly powerful, not in overt political English language.

    ReplyDelete
  10. Roger this is really really good. Your willingness to tell the truth that everyone knows but no one else is willing to say is most impressive. Many thanks from millions of people,

    ReplyDelete
    Replies
    1. Thank you Frank
      I look forward to seeing your No Hesitations piece

      Delete
  11. On Monday or so I will post something in similar style on the No Hesitations blog. Totally different topic but same underlying theme of "everyone knows but no one else is willing to say." Thanks for being a fine role model. Really impressive.

    ReplyDelete
  12. Interstingly in the original paper, Kydland and Prescott were not wholehearedly supportive of a business cycle / growth dichotomy: "We also think it is inadvisable to start our economics from some statistical definition of trend and deviation from trend, with growth theory being concerned with trend and business cycle theory with deviations. Growth theory deals with both trend and deviations."

    ReplyDelete
  13. I am not an economist as a republican candidate for office would say, but the idea that all models are wrong can be extended to much of our human activities. All Economics Is Wrong. All Engineering Is Wrong. All Politics Are Wrong. All Medicine Is Wrong. All Art Is Wrong. All Literature Is Wrong. An extension of these ideas is left as an exercise for the student.

    ReplyDelete
  14. As the wave theory of light is a highly simplified model, it is just a co incidence that microscopes, telescopes and eyeglasses sort of work ?

    ReplyDelete
  15. But Roger, isn't RBC "well established?" .... ;D

    ReplyDelete

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