Thursday, 26 February 2015

The failure of macroeconomics

This is the text of a talk given at Manchester University on 26th February 2015 proposing the motion "This house believes that mainstream economics has failed". It was followed by contributions from Trina Watson of the Post-Crash Economics Society (supporting), Dr. Andrew Lilico and Dr. John Ashworth (opposing), and a lively debate. 

Since the financial crisis there has been growing criticism of “economics”. From the Queen’s famous question “why did no-one see this coming?” to the Occupy movement and now to the Post-Crash Economics society founded by students at this university, people have questioned the purpose of an economics profession that failed to see the disaster approaching and seemed to have little coherent idea what to do about it.

Nonetheless, a blanket condemnation of "economics" as having failed is I think too wide. I therefore wish to narrow the framing. There are many economists out there doing important work, both in industry and in academia, on labour markets, on the behaviour of firms and households, on trade dynamics, on market functioning. I do not by any means wish to suggest that these have failed. Microeconomics, as a discipline, is going from strength to strength. My beef is with macroeconomics.

Olivier Blanchard, the IMF’s chief economist, recently wrote:
We in the field did think of the economy as roughly linear, constantly subject to different shocks, constantly fluctuating, but naturally returning to its steady state over time. Instead of talking about fluctuations, we increasingly used the term “business cycle.” Even when we later developed techniques to deal with nonlinearities, this generally benign view of fluctuations remained dominant.
The models that macroeconomic practitioners developed reflected this essentially linear view. Blanchard went on to observe that although macroeconomists did not ignore the possibility of extreme tail risk events, they regarded them as a thing of the past in developed countries. Western governments had inflation licked because of inflation-targeting central banks. Bank runs had been solved by deposit insurance and central bank lender of last resort functions. Sudden disastrous reversal of capital flows and balance of payments crises were problems for emerging market economies, not for developed European economies. And anyway, central banks could prevent or stop market “panics” by flooding the place with liquidity. If you get the policy settings right, linear models will work.

Except that they won’t. And that is because these models are not realistic views of how the economy actually works. Representative agents aren’t actually representative of anyone. Rational expectations are driven as much by emotion as logic. Behavioural economics is still in its infancy, but we are now beginning to understand just how much humans are driven by instincts such as herding. And nowhere is this more apparent than in the finance industry.

The financial crisis drew to our attention – once again – the crucial role of the finance industry. No industry that can cause such havoc when it goes wrong should ever be regarded as irrelevant or superficial. On the contrary: financial institutions perform the functions of capital allocation and money transmission which are so vital to our economy. Disruption or interruption of these functions, even if only for a brief time, has terrible consequences.

Yet macroeconomists regarded the behaviour of financial institutions and the motivations of those who work in the finance industry as so unimportant that they could safely be ignored. Representative agent models, flawed though they are, at least attempt to explain the behaviour of households and firms: but the behaviour of banks, and things that don’t call themselves banks but do bank-like things, stayed under the radar until far too late. Macroeconomists described the finance industry as a “veil”, rather than as the beating heart and circulatory system of the modern monetary economy: linear models, if they included banks at all, portrayed them as passive intermediaries, rather than active agents whose rational expectations are not necessarily aligned with those of their customers or, indeed, with the best interests of the economy as a whole.

The failure of most macroeconomists to foresee the financial crisis grew out of their incorrect understanding of how money is created, and perhaps more importantly, how leverage builds up. “Loanable funds” models, which portray the role of the financial sector as intermediating existing funds, are not only wrong, they are dangerous. They do not show how exuberance in credit creation arising from the irrational belief that asset values can keep rising forever carries the seeds of its own destruction. And they encourage belief in exogenous factors as the cause of financial crises – the “Asian savings glut” springs to mind. The huge increase in broad money prior to the financial crisis did not come from Asia, or from Mars. It was created by American and European banks.

Leaving banks out of economic models, or – worse – modelling their money-creating function incorrectly, made it impossible for mainstream economists to understand the significance of the build-up of credit that led to the financial crisis. The warnings came principally from people outside mainstream economics, particularly the followers of Hyman Minsky. After the crisis, Minsky’s “financial instability hypothesis”, long consigned to a dusty shelf in a dark cupboard, suddenly became hot news. Unsurprisingly, since we had just lived through something that looked very like a "Minsky moment".

Clearly, the exclusion of the financial industry from models of the macroeconomy was a major omission. Equally clearly, the fact that most macroeconomists did not, and to a large extent still do not, understand the mechanisms by which money is created and circulated in the modern monetary economy, is a big, big problem. Central banks are now “adding” the financial sector to existing DSGE models: but this does not begin to address the essential non-linearity of a monetary economy whose heart is a financial system that is not occasionally but NORMALLY far from equilibrium. Until macroeconomists understand this, their models will remain inadequate.

But macroeconomists are not oracles. It is their job to identify trends, not to predict specific events; it is both unreasonable and dangerous of the public to expect them to play the prophet. Macroeconomists have been cast in the role formerly held by priests and shamans, a role they appear to have welcomed though they are ill-equipped to perform it. They have garbed themselves with the cloak of infallibility and the breastplate of omnipotence. The financial crisis stripped them of these trappings, revealing them to be, underneath, somewhat skimpily clad.

It is fair to say that the academic macroeconomists have done a lot of soul-searching since the financial crisis, and there are important signs that things are beginning to change. But some of the most influential people in macroeconomics have spent their lives developing theories and models that have been shown to be at best inadequate and at worst dangerously wrong. Olivier Blanchard’s call for policymakers to set policy in such a way that linear models will still work should be seen for what it is – the desperate cry of an aging economist who discovers that the foundations upon which he has built his career are made of sand. He is far from alone.

At the Bank of England's One Bank Research Agenda seminar yesterday, Deputy Governor Ben Broadbent commented:
Economists cling to old ideas in the face of overwhelming evidence that they are wrong, or they choose the evidence that suits their particular framing.
Macroeconomics has indeed failed: not because of an intrinsic inadequacy in the discipline itself, but because of confirmation bias and selection bias among macroeconomists. Who would have thought it?

Related reading:

When the Nile floods fail
Financial hurricanes


  1. And with this, I'm quite surprised that anyone takes Paul Krugman and his irk seriously when yet still now, some 7 years after the initial crisis began in August 2007, we get the same fools procastinating against those who actually saw the crisis coming, warned against it and urged action - only for these siren calls to be ignored by the mainstream as messages from fools and outliers. And still many ignore the wealth of empirical evidence and right off the heterodox school of thought as unworthy of their attention.

    Well I, as a non-economist- have a message for them, if the average person cannot make head nor tale of what you are saying and writing, then its a clear admission of most of what you stand for is erroneous, or worse, outright lies. And as an historian, its the likes of me who will expose you and denounce you for what you are: little more than shaman and soothsayers espousing a corrupting influence upon an ignorant population in awe of your supposed wonder and mastery of a subject you bring into disrepute.

    1. Bravo.

      More than historians. Any applicable knowledge. But, we, from many fields, are going further than denouncing but improving and more likely replacing. We know better. We should spend less time on the fallacious.

      I'm not sure macroeconomics has that many people convinced. But, their promoters can't save a bad product when better ones are already at hand.

    2. Speak for yourself. I am no economist but Paul Krugman's columns are very clear and consistent.

    3. No doubt Anton you may prefer to ignore this little exchange:

      Or the fact that even the Bank of England, no redoubt of heterodox economics, had this to say about the subject at hand:

      There are many other instances I can point out, I am after all qualified to teach history, however, I'll shrill for my camp with a clear conscious, namely one is always open to new ideas and new orthodoxies and as such, actually do recognise that the earth rotates around the sun and that macroeconomics is deeply flawed - if this were not the case the 2007/8 great financial crisis would not have occurred, but economic history dictates it has occurred.

    4. When you say "get the same fools procastinating against those who actually saw the crisis coming", how does Krugman fit into this criteria?

    5. There's no such word as procastinating. There's procrastinating, but you don't know what it means.

    6. As a qualified historian, Mr. Rogers might want to spend some more time with primary sources in the form of Krugman's own writings.

      Krugman may be "too mainstream", but he was openly worried about a Minsky moment well before the financial crisis.

      That's from 2001, unless Mr. Rogers wants to accuse the fool Krugman of hacking the Internet Archive to cover his shame.

      --GBR, Raleigh-Durham

    7. RE: History a good form of economics. +100 for Christopher Rogers.

      Professor Keen has said that economic history is very important and not much of it is required in economics programs.

      I think history is an excellent way to learn about economics. Especially economic cycles. After all it is cheap experience. And not many of us witnessed the roaring 20s and 1930.

      One of my favorite classes was, "Business Labor History", that also covered the robber barons.

  2. Excellent. A list of main reasons of the diminished utility of mainstream economics (inappropriate models, outlier Minsky, ...). But if I am allowed I would add: the economists ignorance of accounting both micro and macro. Remembering the ECB warning of fair value accounting...

  3. Do banker's have much more accurate knowledge and tools used in common practice than macroeconomic theories? Don't they also have histories of their industries and credit cycles? Histories of success and failure. I have read some of them.

    Accounting records and financial statements are quantitative histories and can span large amounts of time and are available from different eras.

    Shouldn't bankers know better than macroeconomics? Thus it is logical that they are responsibility for their business than nebulous economists.

  4. Correction:
    Thus it is logical that bankers are responsible for their business rather than nebulous economists.

  5. This comment has been removed by the author.

  6. Who won the debate? Do tell :-)

    I worry, slightly, that the basic premise of these debates is that mainstream economics is a set of fixed ideas that can be argued against by heterodox economists... the mainstream isn't that tangible.

    Having said that, can I suggest a few more failures to add to your list? I agree that finance, endogenous money, tail risk, liquidity and disequilibrium/multiple equilibria are missing from the Economics 101 explanations. That's one reason I prefer stock-flow consistent (SFC) approaches that are less constrained, except that there are no 'black holes - all flows must have an origin and a destination'. SFC approaches consider the impact of balance sheets - financial firms, central banks, even debt restructuring, profit retention, taxes/tax avoidance. The SFC approach is naturally pluralistic (different analyses depending on the problem and setting), history matters, and you can analyse differences in the institutional and legal setting.

    I don't disagree with you that behavioural economics is useful, especially at the micro- level (rules of thumb, biases and prospect theory rather than maximisation of expected utility) but at the macro-level the most important constraints are liquidity and solvency... which vary by sector and institution... so it doesn't really help to think about macro in terms of behavioural economics.

    Anyway, I'm a big fan of your work and I completely understand why you choose the areas you did... especially after the recent Twitterstorm about the fictional money multiplier.

  7. "But macroeconomists are not oracles. It is their job to identify trends, not to predict specific events; it is both unreasonable and dangerous of the public to expect them to play the prophet. Macroeconomists have been cast in the role formerly held by priests and shamans, a role they appear to have welcomed though they are ill-equipped to perform it."

    This is a very good point! It appears to me that many of the problems economics seems to face are coming from the fact that some influential economists some decades ago happened to be overconfident about their discipline's capacities to predict and control. This view has propagated and established the notion among many economists and the public that economics (including macro) can more or less accurately predict economic events and should be judged upon its success of doing so. But if one steps back just for a moment and considers that somebody is supposed to be able to predict the behaviour of millions of people (which the above claim basically amounts to), this seems to be an unfairly high and unreasonable bar to set for any discipline.

    This is not to say that economics (or macro for that matter) has no predictive power whatsoever, but it may be necessary to rethink the importance of the different criteria against which we measure the success and usefulness of such a discipline. From this point of view then, the negligence of the financial sector and money in macroeconomic models is even more startling.

    1. The fundamental problems with mainstream economics are many: all result from a desire in the profession to eradicate diversity and impose a standard version of "economic science". That initiative came mostly from the USA and to some extent the UK..

      Amongst its grand delusions are the following:

      (1) The study of economics can only be conducted by means of econometric modelling.
      (2) The use of mathematical models means that economics is a natural science, and there are standard laws that apply almost without exception.
      (3) It follows therefore that economics is capable of predicting future events; the only limitations are the large number of variables and acquiring accurate data for those variables.
      (4) Challenges to the underlying assumptions of neoclassical economics are trivial, and cannot be allowed to threaten the mainstream dogma.
      (5) The teaching of economics must be confined to mainstream thinking.
      (6) The professional recognition of economics research must be in the major US journals, which are almost exclusively neoclassical in approach.

      With this concoction of seriously erroneous assumptions and premises, the world of western economics established an institutional framework that more closely resembles an organised religion with both state and private sector funding. It arrogates to itself the determination of accurate knowledge, and controls the careers of all those who aspire to or engage in the study or practice of economics. It is not a serious scholarly discipline, because it eschews through political and financial power all the basic requirements of scholarship.

      I wish I could tell you that this problem was confined to economics. My view is that similar problems are now occurring with all tertiary education and research -- for very similar reasons. It suits those in power within the profession; it suits governments, who fear independent research; and it suits business, which has no time for non-conformist views of the human world. We are entering a new Dark Ages, yet the majority of people are fooled by the hi-tech lighting that is everywhere.

    2. Xenos,

      May I say I concur with your summation of what is wrong, not only in economics, but is subjects as diverse as history and sociology - too many rigid models and generalisations that ignore the reality.

      Indeed, I'd go as far to say all of what you have outlined, i.e., the "dogma" now apparent is due to the finacialisation of higher education, where intellectual curiosity and love of one's subject has been perverted, meaning higher education is but a means to employment, rather than humanist learning and pushing boundaries of thought - effectively, we have reached an intellectual cul de sac, which impedes, rather than promotes human development.

    3. @Christopher. Yes, these structural problems are explicitly linked with the financialisation of higher education. What baffles me, though, is exactly why and how the changed structure of financing led to a reduction in diversity of thought and disciplinary approaches. The study of this metamorphosis -- towards homogenization of thinking -- is surely long overdue.

    4. Teaching, and research, in finance has similar problems (heavy reliance on models and econometrics). The solutions are similar, too.

    5. Guest(xenos) said,
      "With this concoction of seriously erroneous assumptions and premises, the world of western economics established an institutional framework that more closely resembles an organised religion... "

      What locations is economics realistic? Where outside the west, and/or in some western. pockets? (Many parts of the non western world were excelling in science and math in the western dark ages.)

      History, accounting, business, engineering, many sciences, tradespeople, and successful investors seem to be resources of better economics understanding. Does any one else know some other relevant fields or people?

      Inside western economics, perhaps accurate economic data long enough to be historical data, stock flow consistent approach. Any others?

    6. I'm seeing this slogan on #chaneEconomics and other places.

      "Pluralism in Economics"

      Maybe more accurate and less ambiguous slogan should be used?

      Does this say successful ideas from the outside are welcome, and adapted? Multiplicity of ideas from outside of economics is needed for a better revision or revolution.

      Minor point:

      Plural means more than one in number. It could be two! So, it could describe the current state of economics. In that case new label same thing.

      The definition of pluralism is ambiguous. Some meanings might approach stated goals of students. Some might meanings might meet goals of entrenched insiders.

      "pluralism |ˈploŏrəˌlizəm|
      1 a condition or system in which two or more states, groups, principles, sources of authority, etc., coexist.
      • a form of society in which the members of minority groups maintain their independent cultural traditions.
      • a political theory or system of power-sharing among a number of political parties.
      • a theory or system of devolution and autonomy for individual bodies in preference to monolithic state control.
      • Philosophy a theory or system that recognizes more than one ultimate principle. Compare with monism .
      2 the practice of holding more than one office or church benefice at a time.
      Link was from Nile Lancastle's page

    7. I was wondering if Economics was better in Italy. Goodwin worked in Italy and I just saw Giancarlo Gandolfo's, "Economic Dynamics: Study Edition". He was a student of Richard Goodwin of the Goodwin growth model which is a dynamic cyclic growth model. Not a static model. It is covered in the book, internet, Goodwin's publications, etc. It is a prime example of dynamic math.

      Also, I wonder about Australia having better economists. Professor Keen often demonstrated Goodwin Cycles in his videos.

      The Goodwin cycle (System Model) is an excellent introduction to dynamic economics. My term is econodynamics. This is the way engineers and

      {I have actually looked at data of the variables in the Goodwin Cycle. To compare picture of reality reflected in the data compared to the dynamic model. There are real cycles but a little different with a twist. Of, course every one knows there are cycles through experience or reading history. I wonder if there is much interest in the data?}

    8. Should have said:
      This is the way engineers and quantitative scientist model dynamic systems. (Systems, means multiple parts thus systems of equations. Often an equation for each part.)

    9. FWIW, this rot and degeneration into religion is only happening in some fields. Economics is seriously rotten. Physics is rotten, but the good physicists renamed themselves "materials scientists" and kept working. (Non-human) biology is doing just fine. Medicine is totally rotten, but that's different.

  8. " system that is not occasionally but NORMALLY far from equilibrium. Until macroeconomists understand this, their models will remain inadequate"

    Do you mean that the deposit asset pair used in the financial system 's T account is not in equilibrium as the deposit is used as a value that has allready been realised against an asset whos value has yet to be realised. And this confusion comes from the erroneous loanable funds concept.

    1. To me that sounds like a making of a cyclic system. There are folks who talk abut a credit cycle?

    2. Yes, there are a number of cyclical processes in macroeconomics.

  9. A thoroughly sensible posting. With its 'identification of trends' it moves towards framing macroeconomics as an intelligent way of talking about the socio-economy that does not aspire to predict because far too many others factors come into play. Predicting that central banks can prevent bank runs refers to the behavior of a small part of the socio-economy, not to macroeconomics’ potential to explain the whole. Macroeconomics as intelligent talk takes us towards McCloskey's 'rhetoric of economics' - which also moves us towards recognizing that it is not the only way of talking. Economists are prone to hubris in this respect.

    One point startled me that microeconomics is going from strength to strength. I wonder what refers to. Winning Nobels or helping humanity?

  10. I still don't think that the endogenous money theorists have provided a solid case for the importance of money creation mechanics. It seems a bit foo foo circular to say that anticipated money base levels move nominal expectations, but that may be the state of the world bumping into imprecise methods of comprehension. Aggregate loans are still bounded by the money base, which adjusts according to revealed Fed reaction functions. Taking for granted the order of lending and reserves, the role of exogenous money boundaries should still influence market behavior through perceived reaction functions.

    1. I suspect with people like you ... that the desire to `sound clever`, precludes you ever actually `being clever`.

    2. Nobody said the CB can't influence market behaviour, but it's a question of exactly how it does this. Does it control the level of money or only interest rates? Mainstream theory says it can choose between the two; endogenous money theory says the latter.

    3. Walter,

      1) Confused precedence. Aggregate loans are bounded by the monetary base ex post, not ex ante.

      2) Fallacy of aggregation. In the absence of micro prudential measures limiting the lending of individual banks, there is zero evidence that banks take any notice at all of central bank reaction functions when making lending decisions. There are two reasons for this. Firstly, the provision of reserves pertains to settlement, not lending: it is simply not credible that a central bank would refuse to supply reserves to settle deposit withdrawals. To do so would put at risk the vital circulation of money in the economy. To my knowledge the only central bank that has threatened to limit reserve creation is the ECB, and I've pointed out elsewhere that this threat is not remotely credible when applied to an entire banking system. Secondly, even if a central bank did limit the provision of reserves to the system as a whole, individual banks would be unlikely to restrict their lending. Rather, they would seek to hoard reserves by attracting depositors.

    4. Sorry, I mean "Walnut" of course.

    5. If financial institutions believe that a central bank is downwardly biased against some metric, like PCE inflation over 2%, then that should imply the conditional credibility of reserve constraints. Shouldn't the credible threat of reserve action would appear, ex ante, through inflation expectations and presumed loss experience given the absence of monetary shocks? Unlearning Econ said that the Fed acts through interest rates, and you commented that base money moves after the fact, but it seems like institutions, and everyone else, have no choice but to anticipate those actions.

      For example, I read your sentence as an example of the linkage between ex post and ante:

      "To my knowledge the only central bank that has threatened to limit reserve creation is the ECB, and I've pointed out elsewhere that this threat is not remotely credible when applied to an entire banking system."

      That may be another way saying that the appearance of the threat is an outcome of poorly established expectations, despite the mechanical ability to support the threat. Or, the ECB is implicated in all such nominal expectations built upon its expected behavior.

    6. Walnut,

      You still aren't getting it. Base money moves after the fact IN RESPONSE TO the behaviour of banking institutions. Therefore all these institutions anticipate is that the central bank will respond to demand for reserves. Expectations relate to the price of reserves, not their availability.

      A threat to restrict reserves ex post is not remotely credible for the banking system as a whole, since it would mean settlement failures. Reserve restriction ex ante does work, but it works by influencing the price of reserves, not their availability - this after all is how the Fed Funds rate target is maintained in normal times (in these days of excess reserves it is maintained by explicitly setting the IOER rate, of course).

    7. Agree with Xenos& C Rogers!! University is no longer a platform for free thought and boundary pushing- students r encouraged to stick to the party(lecturers/ authority on subject) line not adopting a line of individual thought in anything particularly Journalism in my case??

    8. mainstream econ does not say the CB could control the level of money. It could, if it chose to, target base money. But as everybody knows that's quite a different thing. Everybody knows what happened when level targeting was attempted.

      Frances, I'd be interested to hear a bit more about what econ needs to know about money creation.

      Right now, we know that an interest rate targeting bank isn't in control of base money and that broad money is on the hands of the banks via credit creation. What else do you think is needed?

      You could append to this a minsky story about how, as credit expands, banks accumulate lower quality assets and get themselves into a suicidal position, exposed to modest declines in asset prices. I reckon thats a perfectly sensible dynamic to model, test empirically and all that. No objections from me. But I don't immediately see what else is required in terms of how money creation works, to make that theory fly. If indeed that's what you have in mind.

      P.s. for avoidance of misunderstanding, I agree banks lend first and finance later, rather than vice versa, so if that's what you think mainstream econ needs to incorporate, fine
      Although don't immediately see why it's a terribly important detail.

    9. Luis: yes, the Minsky dynamic is why this is important. Also see the "demonetization" of money market funds in 2008: this was the critical event which nearly broke the payments system. This can *only* be understood in the context of understanding how broad money is created.

  11. I have done some inflation modeling (web based so easy for others to play with). A key insight is that there are feedback loops that can make simple linear formulas result in non-linear behavior. For example, the velocity of money depends on the interest rate and the inflation rate. But the price level depends on the velocity of money. So even if when both of these dependencies are linear you can get non-linear results.

    1. Frances, "Base money moves after the fact IN RESPONSE TO the behaviour of banking institutions. "

      In my model, which I think is a reasonable reflection of reality, both are part of a feedback loop. The behavior of the individuals and banking institutions also moves in response to base money. Both happen. And feedback loops easily make for non-linear behavior.

    2. "Macroeconomics has indeed failed: not because of an intrinsic inadequacy in the discipline itself, but because of confirmation bias and selection bias among macroeconomists. "

      Well said. I hope that anyone looking at my model can set aside their bias. Just look at the formulas one by one and see if they are reasonable or if you can think of a better one. You can "clone my insight" to get your own copy so you can change the formulas to what you think they should be and then announce your version. But even if you don't think the formulas are exactly right, what about the basic result of feedback loops making linear formulas result in non-linear behavior?

    3. An example of selection bias among macroeconomists would be people who don't believe in inflation not being willing to look at a model of inflation.

  12. I came here looking for something specifc as to why a non-linear approach may be a good idea, but instead I find a lot of meaningless rants.

    I have looked into some of the models, and I find they are even more mathematical and confusing as linear ones. And the results? Are they driven by the technology or something realistic?

    1. Did you read the paragraph in italics at the top of this post? As this was a short speech to undergraduates, not all of them economics students, I was hardly going to go into details about the merits and demerits of non-linear models.


  13. its an interesting point .
    Why did central bank liquidity not stop the panic , and then bring the money back .
    Obviously there must have been a broad assumption here . Yes a panic could occur , and a short term funding problem , but it would only be a short term problem .

    But why ?

    As for this point that credit creation , as oppose to credit allocation , caused the crisis , well I've always thought >>> o k , show me the evidence .
    Lots of evidence savings in pension funds , insurance co , local govt treasuries were involved . And external savings .
    Any link to a model , descriptive , or something , that can suggest the created out of thin air money was the greater ?