Wednesday, April 25, 2012

Neither Real, nor Business, nor Cycles

It has often been said of the Holy Roman Empire that it was "neither Holy, nor Roman, nor an Empire." However, that joke has gotten a bit stale since Voltaire wrote it in the 1700s, so I think it's time for a new one. Real Business Cycle models, it turns out, are neither Real, nor about Business, nor about Cycles.

They are, however, the macro models that annoy me far more than any other (and I'm not alone). I'll explain the joke in increasing order of the things that annoy me.

First, "Cycles". The "business cycles" in RBC models are not periodic, like cycles in physics. But they are also not "cycles" in the sense that a bust must follow a boom. Booms and busts are just random shocks. The "business cycle" that we think we see, according to these models, is simply a statistical illusion. (Actually, RBC shares this property with New Keynesian and Old Keynesian models alike. Very few people dare to write down a model in which knowing you're in a boom today allows you to predict a bust tomorrow!)

Next, "Business". Businesses are called "firms" in economic models. But if you look at the firms in an RBC model, you will see that they bear very little resemblance to real-life firms. For one thing, they make no profits; their revenues equal their costs. For another thing, they produce only one good. (Also, like firms in many economic models, they are all identical, they live forever, they make all their decisions to serve the interests of households, and they make all decisions perfectly. Etc. etc.) In other words, they display very few of the characteristics that real businesses display. This means that the "business cycle" in an RBC model is not really the result of any interesting characteristics of businesses; everything is due to the individual decisions of consumers and workers, and to the outside force of technological progress.

Finally, "Real". This is the one that really gets me. "Real" refers to the fact that the shocks in RBC models are "real" as opposed to "nominal" shocks (I've actually never liked this terminology, since it seems to subtly imply that money is neutral, which it isn't). But one would have to be a fool not to see the subtext in the use of the term - it implies that business-cycle theories based on demand shocks are not, in fact, real; that recessions and booms are obviously caused by supply shocks. If RBC is "real", then RBC's competitors - Keynesian models and the like - must be fantasy business cycle models.

However, it turns out that RBC and reality are not exactly drinking buddies. I hereby outsource the beatdown of the substance of RBC models to one of the greatest beatdown specialists in the history of economics: the formidable Larry Summers. In a 1986 essay only slightly less devastating than his legendary dismissal of the Winklevoss twins, Summers identified three main reasons why RBC models are not, in fact, real:

1. RBC models use parameter values that are almost certainly wrong,

2. RBC models make predictions about prices that are completely, utterly wrong, and

3. The "technology shocks" that RBC models assume drive the business cycle have never been found.

I encourage everyone to go read the whole thing. Pure and utter pulpification! Actually, this essay was assigned to me on the first day of my intro macro course, but at the time I wasn't able to appreciate it.

So Real Business Cycle models are neither Real, nor about Business, nor about Cycles. Are they models? Well, sadly, yes they are...of a sort. You actually can put today's data into an RBC model and get a prediction about future data. But see, here's the thing: that prediction will be entirely driven by the most ad-hoc, hard-to-swallow part of the model!

Basically, here's how an RBC model works. You take every factor of production you can measure - capital, labor, inventories, etc. - and you factor it out, and then you're left with the part of production you can't explain, which is called the "residual". You then label this residual "technology", and you assume that it moves according to some sort of simple stochastic process - for example, an AR(1). The rest of the model is just a description of the ways in which the rest of the economy responds to that AR(1) technology "process". In RBC models, this response is usually as simple and uninteresting as possible; pretty much everything is driven by the uber-simplistic movement of "technology".

In other words, if I want to make a forecast using an RBC model, that forecast will be based on the assumption about tomorrow's level of "technology" - i.e. the part of the model that doesn't come from data we can directly measure - and that level, in turn, will be "predicted" by nothing more than the simplest stochastic equation imaginable! As you might therefore expect, RBC models do not do so well at forecasting the future (though, to be fair, a few disagree with that assessment).

(Note that this makes RBC a glaring example of what I call "Label-the-Residual Economics", in which the economist assumes that the part of the world that we can't measure is the Mysterious Force that drives everything, but that we can accurately predict the future behavior of this Mysterious Force.)

Now, some work has been done on improving RBC models since their inception - instead of technology, for example, some modelers try to tie business cycles to news about technology. But most of the macro profession has moved on to other types of DSGE models, especially new Keynesian models. And yet the RBC paradigm persists, especially at certain universities. Why? In a recent blog post, Simon Wren-Lewis lays out this case that's it all about the politics:

In RBC models, all changes in unemployment are voluntary. If unemployment is rising, it is because more workers are choosing leisure rather than work. As a result, high unemployment in a recession is not a problem at all...workers choose to work less and enjoy more free time... 
If anyone is reading this who is not familiar with macroeconomics, you might guess that this rather counterintuitive theory is some very marginal and long forgotten macroeconomic idea. 
You would be very wrong. RBC models were dominant in the 1980s, and many macroeconomists still model business cycles this way. I have even seen textbooks where the only account of the business cycle is a basic RBC model... 
One explanation [for RBC's popularity] is ideological. The commonsense view of the business cycle, and the need to in some sense smooth this cycle, is that it involves a market failure that requires the intervention of a state institution in some form. If your ideological view is to deny market failure where possible, and therefore minimise a role for the state, then it is natural enough (although hardly scientific) to ignore inconvenient facts. For the record I think those on the left are as capable of ignoring inconvenient facts: however there is not a left wing equivalent of RBC theory which plays a central role in mainstream macroeconomics.
Whether Wren-Lewis is right about this or not, I think the continued semi-popularity of RBC models definitely shifts the field of macro toward more politically conservative policy recommendations. It does this by shifting the "Overton Window". Without a strong RBC presence, macro might be primarily a debate between New Keynesians and Old Keynesians, or New Keynesians and complexity theorists, or New Keynesians and people who think we just don't know enough to really model the business cycle yet. In other words, it might be a debate between people who think that the economy can be managed effectively by central bank monetary policy, and people who think deeper government interventions are warranted. Instead, for the past two decades, academic macro has been primarily a debate between New Keynesians and RBC people - a debate between minimal-interventionists and those who oppose any sort of government intervention at all. In the "freshwater-saltwater" debate, supporters of things like fiscal stimulus were left high and dry.

At any rate, this bit about politics is a digression. The central point of this post is that Real Business Cycle models, whether politically motivated or not, are massively oversold as descriptions of the recessions and booms that we observe and live through. People should know that they contain neither Reality, nor Business, nor Cycles.

How long did it take the Holy Roman Empire to finally give up the ghost? Depressingly, it was more than half a century after Voltaire made his little joke...


  1. 1st. I'm really glad you wrote this. Like you, RBC models annoy me more than anything else. Others too. I've met many people who got into economics precisely because of macro topics...unemployment, GDP, recessions, booms, etc. only to get RBC theory forced down their throats in grad school and say, "Wait? What? THIS is what macroeconomics is? What a load of BS." (that many micro professors said they thought it was BS too didn't help.) Many students quit macro without seeing what else it has in store. That's a bit of a shame.

    2nd. I like the bit about the "residual." I was at an econ talk for laymen with my parents recently. The speaker kept talking about "technological progress" in vague terms. An audience member asked how this was measured. The speaker gave a strange roundabout answer that wasn't really much of an answer at all. It led to more confusion than clarification. The audience member obviously wasn't satisfied so I leaned over to him and told him the truth, "He won't describe it clearly because he doesn't want to admit that what he's talking about is the error term in a regression. It's not anything we can measure. It's precisely the opposite. It's all the things we can't explain." The guy looked at me like I was pulling his leg. I wish I was.

    1. Noah will no doubt be saddened to learn that he annoys you "more than anything else." Harsh!

  2. Given that Larry Summers wrote that beatdown in 1986, are we 26 years into the 50 year wind down? Pretty sad.

    As I said in my comment on Wren-Lewis' post, adherence to RBC indicates to me fairly fundamental dishonesty in the profession. To me the really shocking point in his piece is that RBC is often taught as the only account of business cycles.

    It now occurs to me that adherence to RBC might be a good touchstone to identify the ideologues. Economists who support RBC could be publicly shamed by those who want to actually do science. The public can understand something as concrete as claiming that the unemployed just don't want to work for the time being.

    The resulting acrimony won't be pleasant but the current pretense is worse.

    Alternatively, economics can hang out with RBC for another 24 years. I expect not much will be left of either by then.

  3. Yeah, you got a catchy hook there Noah. I don't know if RBC is the worst of the worst (modern portfolio theory and Black-Scholes-Merton may be even more useless), but certainly the number of totally bizarre assumptions (Wren-Lewis' examples are particularly funny) make RBC into dead wood.

    My view goes a stage further: I think all economic modelling is fundamentally problematic. Yes, all is a strong word, and some models are more equal than others (in Taleb's words: models of mediocristan work better than those of extremistan), but everything is weak to the problem that not all the influences and causes are identifiable. If you could construct models that take account of every causal effect, then they would work, but this is much more complicated than anything in physics (say) where the relevant causalities are much, much clearer, and therefore yield much less anomalous results.

    Economic models are interesting curiosities, but (living as we are in the fat tail of history) I don't think they are really anything more than heuristics.

  4. Good piece Noah however many of the criticisms apply to the whole class of Arrow-Debreu models and some criticisms to the RBC class of its children. The more fundamental criticism should be to the whole wider class otherwise the 'were New Keynsian and were different' defence of Mark Thoma can be made.

    I argue its the all A-D models that must go

    From feedback on that paper two issues have emerged - firstly some of the assumptions that all of neoclassical economics makes about the necessary conditions for equilibrium have been proven wrong.

    A good example is the assumption of homegenity - The spatial impossibility theory shows that in such a world there is no trade - you have to abandon the assumption of spatial homegenity to get trade.

    From this work follows that the neoclassical assumptions of constant returns to scale and perfect competition are also impossible, in the real world we see increasing returns and imperfect completion. Berliant and Ten Raa have proven that increasing returns is compatible with positive profits if you modify the Alonso land rent model to include production.

    But such a real spatial world only gives you islands of partial and temporary equilibrium. I have found no convincing mathematical proof, despite attempts by some, of prof of GE in such a realistically modelled world. All temporary equilibrium does is sent information signals to entrepreneurs that there is money to be made and equilibrium points to shift - you get non convexity from inevitably poor entrepreneurial decisions under uncertainty and that shatters the convexity assumption of A-D. You can only get GE in such a world with the Deux ex Machina Lucas assumption of 'rational expectations' which requires reqauires perfect foresight and psychic ability of every human being of every other human being till the end of time. As Buiter writes this Auctioneer at the End of Time assumption is a model of a centrally planned economy not a market one.

  5. Can I just kick in a little Nagarjuna in the spirit of your catchy hook?

    "There is neither existence, nor non-existence, nor both, nor neither."

  6. Phil Koop11:05 AM

    There is a theological approach to the Gospels (somebody help me out here - what is it called?) that assumes that the shared elements of each account represent a merely mortal consensus. To truly experience the word of God one must factor out these common elements!

    I confess to a certain amusement that the advocates of a mainstream branch of economics (one that makes particular and exaggerated claims to rationality) should be brothers in perversity to these not-very-mainstream theologians.

  7. I agree that RBC models are useless at best and wrong at worst on their own. And yet....

    With many models, the level of the predictions can be wrong, but they still give a useful delta. Black Scholes options pricing models are useless to price individual securities, but they are useful still for ranking relative values of securities.

    There is an analogous rule that is also true for RBC models. First, there are macro RBC models, and there are industry-specific RBC models. A comparison between industry RBC models for different sectors yields very useful results. You can get some very good information about relative performance this way. Even better you can get some great lag-lead information between industries, i.e. industry A leads industry B by ___ quarters. The relative values are better because a lot of the assumptions between models is effectively canceled out.

    1. Wait, "industry-specific RBC"? Do you just mean finding industry-specific Solow residuals and then doing time-series econometrics on those? Because the actual "model" part of the RBC model - investment, consumption, labor supply - wouldn't apply at the single-industry level. Just looking at industry-specific productivity is not "RBC"...

  8. ivansml12:34 PM

    Your post seems to imply that RBC models are still one of dominant schools in macroeconomics. Is that really the case? I'm just a graduate student so I can't claim to be any authority on the subject, but my impression was that RBC research program (in the sense of explaining macroeconomic fluctuations solely through frictionless competitive equilibrium models driven by exogenous technology shocks) is pretty much dead.

    Meaning that while methodological contributions of RBC literature became widely accepted, its economic message was dismissed in favor of models with frictions, such as sticky prices. So before the crisis, you had leading macroeconomists (Blanchard, or Woodford) speaking about "new synthesis", not about some ongoing epic fresh/salt-water struggle. Even proponents of RBC titled their Handbook of Macroeconomics survey "Resuscitating Real Business Cycles", which sounds pretty defensive to me.

    All that of course doesn't diminish arguments against RBC models (though there are probably better ones than nitpicking about meaning of words "business" and "real"). But non-economists may read your post and get the impression that RBC models are much more commonly accepted than they really are. There is enough ignorant and misinformed criticisms of macro going around, no need to add even more.

    1. Have you seen the DSGE models that are being made in labor, trade, and finance these days? Most of them rely on technology shocks as the main driver of business cycles. In other words, RBC may be out of favor in the core business-cycle field, but it has metastasized and is colonizing other areas of macroeconomics.

      The reason is that tech shocks and constant-returns-to-scale firms are easier to model than alternatives with nonzero profits and imperfect competition and demand shocks.

      Of course, the RBC approach to labor supply has been pretty much dropped everywhere.

  9. Good use of an old wisecrack, Noah. Another use of it, now increasingly passe, is that the long-ruling (now out of power) Liberal Democratic Party of Japan is neither liberal nor democratic nor a party (on the first, it is conservative, on the second it long was run by a handful of bosses and manipulated elections, and on the latter, it is just a collection of factions).

  10. Michael Harris10:43 PM

    Two macroeconomists are arguing in a bar.

    The New Keynesian slams down a beer glass angrily and exclaims "You real business cycle people have set the profession back twenty years!"

    "Ah," says the other, "so you DO believe in technical regress after all."

    1. This comment is full of WIN.

    2. Michael Harris11:51 PM

      Once Krugman uses it, my career highlight will be achieved.

      Otherwise, Yoram Bauman can have it.

  11. Business Cyclist6:01 AM

    Why are you so annoyed at the "cycles" bit? I don't see your point.

    1. It's not a 'cycle' because the booms and busts aren't actually linked. It's simply a stable economy occasionally buffeted by shocks.

    2. Because what they are doing is taking residuals, and pretending that they are actually the effects of a factor, rather than residuals from a model.

  12. Anonymous1:22 PM

    Why should we think they are exactly? What in the data tells you that an AR(1) process isn't sufficient? More than that, what in the data lets you know that they are linked?

  13. Anonymous1:42 PM

    Isn't the reason that RBC remains popular is that:
    1/ New Keynesian models/New Classical models are RBC models with added frictions (sticky prices, search frictions, etc).
    2/ When developing a new theory, it might be best to see how the particular friction you are interested in interacts with a model that is otherwise 'vanilla'? If you add a new friction (say) to a model with a whole load of other frictions in, it's not clear whether the change in the results you get is because of the new friction by itself, or because of the interaction of the new friction with the other frictions (eg. are uncertainty shocks in themselves important, or is it only because you add them to a model which assumes habit formation in consumption?)
    And of course:
    3/ RBC is the simplist microfounded model. Therefore it remains a useful teaching tool for first year grad students (or older undergrads).

    Just thinking aloud..

    1. A very good point.

      Have you read this post of mine?

  14. Anonymous3:56 PM

    A rule of thumb of mine is to be deeply suspicious of any theory that attempts to declare itself right in its title. This applies to 'Real' Business Cycles, 'Realist' IR theory, or 'Objective'-ism, all of which use daft reasoning to arrive at wrong conclusions which are nevertheless ideologically convenient.

    1. Anonymous4:04 AM

      I'm sorry, "real" isn't referring to its own correctness. It is a contrast to market failure arguments. But hey, who needs intellectual honesty when you have a political agenda to push?

    2. Anonymous6:42 AM

      Words have meanings beyond their formal etymologies. The implication is obvious and exists whether the namers intended it or not.

      Obviously those things didn't acquire their names as part of some cynical PR scheme, but the pattern (maybe) exists and is funny nontheless.

  15. Anonymous4:56 PM

    I think that "endowment effects" also play a role here. It must be immensely difficult to abandon a theory that you've worked on so hard. Also, the way you see the economy really depends on the lenses you look through. Economic data contains so much noise that you can rationalize it in an infinite number of ways, especially if the degrees of freedom are not too few (which they are not in an RBC kind of model).

  16. I'm not sure why people say RBC is microfounded since technology is so ad hoc...

  17. The big question is why the microeconomists allowed this.

  18. Two additional comments. First, I think that Paul Romer's work on endogenous growth is also a significant dagger into the heart of RBC, which relies on what I refer to as the technology tooth fairy, since the modeling of technological shocks bears a striking resemblance to children's stories. Secondly, it is worth noting, as I have done before, that in the investment management industry, where I have spent the last 35 years since my graduate school education, and where models to survive have to have some level of accuracy in their predictions, nobody, and I mean nobody, uses RBC models. For those of you who believe the efficacy of markets, you might consider that the actual marketplace of money, versus the marketplace of ideas, has spoken on this subject.

  19. Somehow in my ignorance and naivete I had imagined that RBC was model of how boom and (especially) bust was an inevitable consequence of the exponential model X= X0*(1+r)^N of the cost of debt.

    Question: Is there a formal theory of economic cycles that recognize that busts simply are a mechanism for (a) hitting the reset button on clearly unfeasible exponential growth, and (b) at the same time making sure that the general public pays the price of the reset through unemployment and low deposit interest rates.

    1. Not that I know of!

    2. Anonymous9:52 AM

      And yet we've had fairly stable long run exponential growth even taking into account a couple of big "resets". Maybe you could define "unfeasible" for us?

    3. Anonymous:

      I think what we have is a high exponential growth in money, and a low or near-zero-exponent exponential growth in actual value.

      Does that answer your question?


      I find it surprising if not a single economist has published a paper that says that recession is a reset for exponential debt, and a socialization of losses while private profits have already been taken, of course.

      Mind you, I'm not an Austrian, nor an Austerian. I'm more of a Paul Krugman fan, really.

  20. Jim Rose8:27 AM for Prescott’s response to summers. was that also assigned?

    Real business cycles are recurrent fluctuations in incomes, products and factor inputs, especially labor, that are due to non-monetary sources that sources include changes in technology, tax rates and government spending, tastes, government regulation, terms of trade, and energy prices.

    Kydland and Prescott’s 1982 showed that technology shocks, i.e., short-run variations around the positive growth trend for technology that makes economies grow in the long run, could be an important cause of output fluctuations.

    In today’s macroeconomic models, supply shocks typically play an important role alongside demand shocks.

    Extensions of Kydland and Prescott’s methodology included analyses relying on monetary shocks or international shocks to the terms of trade, and different propagation mechanisms, such as those implied by imperfections in credit, labor and goods markets

    1. Real business cycles are recurrent fluctuations in incomes, products and factor inputs, especially labor, that are due to non-monetary sources that sources include changes in technology, tax rates and government spending, tastes, government regulation, terms of trade, and energy prices.

      Recurrent business cycles due to changes in tax rates, regulation, and government spending? ;)

      In today’s macroeconomic models, supply shocks typically play an important role alongside demand shocks.

      Like, which one, for instance?

    2. Jim Rose7:49 PM

      thanks, but see Fiscal Sentiment and the Weak Recovery from the Great Recession: A Quantitative Exploration by Finn E. Kydland and Carlos E. J. M. Zarazaga at

      The U.S. economy isn’t recovering from the deep Great Recession of 2008-2009 with the anticipated strength. A widespread conjecture is that this weakness can be traced to perceptions of an imminent switch to a higher taxes regime.

      The main finding by Kydland and Zarazaga is that the fiscal sentiment hypothesis can account for a significant fraction of the decline in investment and labor input in the aftermath of the Great Recession, relative to their pre-recession trends.

      These results require a qualification: The perceived higher taxes must fall almost exclusively on capital income.

      This is not an unreasonable condition. It suggests that economic agents suspect that at times of stress, the tax structure that will be implemented to address large fiscal imbalances will be far from optimal.

      As agents realize that their capital income will be taxed more heavily in the future, they reduce their holdings of the capital stock by not completely replenishing the part of it that depreciates every period and by changing the composition of output in favour of consumption.

      The fiscal sentiment hypothesis can account for between three-fifths and the entire decline in private gross domestic investment, and one third to one half of the decline of labor input observed since the trough of the great recession

      In old Keynesian terms, a wave of pessimism is behind the weakness in investment and the weak demand for labour.

      Real business cycle provides grunt to the animal spirits view of investment. Real business cycle theory is unpacking the exogenous shocks central to old Keynesian macroeconomics.

  21. Jim Rose9:10 PM

    I should hav added Ellen R. McGrattan, forthcoming. "Capital taxation during the U.S. Great Depression," which found increased capital taxation during the U.S. Great Depression were small in the case of taxation of business profits, but finds large effects in the case of taxation of dividend income.

    • Tax rates on dividends rose dramatically during the 1930s and imply significant declines in investment and equity values and nontrivial declines in gross domestic product and hours of work.

    • The results are amplified if businesses make intangible investments which can be expensed from taxable capital income.

  22. Anonymous3:54 AM

    Instead of reading papers that deal with RBC models, I think it is far more advantageous to work out the 2nd edition of a book like Macroeconomics in Emerging Markets, by Peter Montiel (CUP, 2011).

    I am neither the author nor the publisher nor someone marketing the textbook. I am just a person that read the book and found it to be illuminating.

  23. You know who likes RBC? Tyler Cowen.

    I am trying to get people to realize the guy is a hack. Perhaps, perhaps this will be the final piece of evidence necessary.

    But then he's got all the Koch money coming in. What does he care?

  24. Nice piece. I am not a fan of RBC models but you have to be more precise. Three points.

    @Cycles: Since the papers of Frisch, Slutzky and Samuelson, "shocks" are modelled as kind of "fundamental" drivers of an economy which can be described using the comparison of a rocking chair. The economy tends to market clearing and equilibrium and then some forces drive you away from that point and the chair oscillates. Impulse and propagation. Therefore cycles -- boom and busts -- are not only driven by the shocks but by the structure of the economy. In the case of Samuelsons famous multiplier-accelator model it was the interaction between households and businesses which defines the propagation mechanism. So, even in simple RBC models we can have (quite regular) cycles.

    @ Businesses and Real: I completely agree. New Keynesian DSGE models added monopolistic competition and more detailed firm behaviour and a lot of channels to make the impact of demand shocks more prominent. However what is more a kind of problem in this models is the Frisch-Slutzky-Samuelson paradigm of an inherently stable market clearing process.

    @ Forecasting: This is a misunderstunding on the role of models in thinking about the world works. We build models to reduce complexity and focus on the central channels. We do not build models to forecast (you never try simple forecasting techniques with sophisticated models....)

  25. Anonymous7:07 AM

    From a heraldry amateur here; the double headed eagle on the picture is probably from one of the Habsburg emperors post 16th century. Hungary(silver double cross on red base on three green hills), Croatia(checkered white and red) and Dalmatia(three lion heads on blue base) were never part of HRE.
    HRE imperial banner was black double headed eagle on yellow and coat of arms was this:

    Nonetheless it's a good analogy.

  26. Anonymous1:45 AM

    Would you place Ricardian Equivalence in the same class as RBC? It seems to me to be in the same class of ideologically driven modelling, i.e. completely and utterly ignoring the facts. You put garbage in, you get garbage out. I probably don't need to mention that Ricardo himself disowned this idea quite vigorously... because that would be beside the point, we have enough modern day supporters of this idea