The Bank of England's chief economist has admitted his profession is in crisis having failed to foresee the 2008 financial crash and having misjudged the impact of the Brexit vote.
Andrew Haldane, said it was “a fair cop” referring to a series of forecasting errors before and after the financial crash which had brought the profession's reputation into question.
Blaming the failure of economic models to cope with “irrational behaviour” in the modern era, the economist said the profession needed to adapt to regain the trust of the public and politicians.
The problem for Haldane is that the problem is systemic. A few exceptions apart (the E3ME model of Cambridge Econometrics, where I am a director, being an example) most macroeconomic modelling is done on the basis of general equilibrium analysis. This assumes that perfect competition prevails; the people are rational; and that markets adapt (or ‘clear') perfectly to changes in economic stimuli. They also fail to treat money appropriately. The result is that of course the models fail to predict real world irrational behaviour. That is precisely because they are programmed not to do so, since those using them assume it does not exist.
Haldane could overcome his problems. He could use models that do not use the assumptions inherent in general equilibrium analysis. The E3ME model of Cambridge Econometrics, of which I am a director, does not do so, for example. It does not require rational behaviour. It does not require that markets clear. Sub-optimal outcomes can be predicted because they happen in the real world. And modern monetary theory can be built in.
That is the answer to Haldane's problem. He and the Bank of England have to look out of their windows (of which there are few in the Bank) and realise that the world is not the one they're modelling. Then their forecasts might be relevant. Right now they are forecasting for a world that does not and never will exist. Of course they get things wrong. But that's wholly their own fault.
Thanks for reading this post.
You can share this post on social media of your choice by clicking these icons:
You can subscribe to this blog's daily email here.
And if you would like to support this blog you can, here:
I would disagree with one part of your otherwise excellent summary, Mr Murphy
I do not accept that people are not rational, for the most part. They are.
The problem is indeed systemic. That is because behaviour which is “rational” depends on circumstances and possibilities. They are not the same for the comfortably off as they are for those in poverty. At some level those who punt this economic model know that: it is demonstrated in their conviction that rich people need more money to incentivise them to work hard: but poor people need less to encourage the same end. So it is not really ignorance, though many do not stop to join the dots.
This is seen in the ongoing project to pretend that poor people’s interests are the same as those of the rich. This is not just a false narrative: it is translated into policy choices.
An example is housing. People have been persuaded that owning a house is the route to prosperity and a minimum expectation for everyone. How has this been done? I would argue that it is true that for well off people owning property produces security and legacy. They therefore assume that is universally true, or they pretend to. But they did not sell this idea solely through persuasion and argument. They ensured that there is no alternative: by eroding pensions; establishing low interest rates on savings; reducing employment rights and social security etc they engineered a situation where there really is no option to house ownership if you aspire to even minimal financial security over a life time. It is therefore rational to take on that home ownership, with the inescapable consequence that house prices rise beyond the means of many: the failure to build social housing and the sell off of what existed was part of that strategy.
It is therefore entirely rational that poor people bust a gut to try to get onto what is tellingly called “the property ladder”: but what works for the wealthy does not actually work for the poor, and by now the middle income sector. Poor people do actually know that: but TINA. It is really quite clever, because mortgage commitments actually really do result in an alignment of interest between rich and poor at least insofar as the low interest rates become essential to the poor and debt is normalised.
The same argument applies to education. Again what is good for the wealthy, ie tertiary education and the premium that commands in wages, has been asssumed to be a universal good. So we encourage expansion of that sector, but not as a good in itself, as was previously believed. Rather young people have been persuaded it leads to higher income: ignoring the fact that increased supply of graduates will, by their own theory, necessarily reduce the price paid for them. So we see graduates flipping burgers and burger flippers on the dole. And neither can live on the wage; and graduates are resentful as promises are not fulfilled. As someone said, by now a degree is simultaneously expensive, essential and entirely useless in securing a high income. I am not arguing that education is not a good thing: but it is not a private good to be bought and rationed by money: rather it is a public good as it was previously characterised. that we accepted that shift in perception is one of the most astonishing things to happen in the last 50 years
We are not all in this together. But rational behaviour is different for different groups. Lumping all together is an error of monumental proportions, and is at the root of much nonsense, as I see it.
I like your argument
I agree everything except the human being rational bit. Recent philosophy has us capable of “bridled irrationality”, psychology that we go mad without emotions and biology has each of us as a collective of micro-organisms with two inter-connected brains, subject to such as parasite influence (the classic is toxoplasma gondii). If we were rational, advertising would disappear and politicians like May following the sloganeer recommendations of the most evil man in history would be arrested. Of course, we have problems with what rationality denotes and connotes – but it is safe to say we have made general use myth and excluded science from it. The rest is what we should be looking into and constitutes the ground for genuine empirical adequacy in economics. I go as far as thinking economics (and lots of other tradition) creates disabling societies.
A lot of people can have a stab at making rational decisions if given the right information (although the extent to which decisions are truly rational can be over-stated).
The more challenging aspect is the amount of information that would need to be acquired and processed to make rational decisions. The example in “Debunking Economics” where it considered the permutations of buying/not buying from a selection of 200 products to “maximize utility” produced a number of permutations that were simply impossible for the human mind to deal with – and this was even after introducing major simplifications of the choices available.
A simple example from my own life: I have an old Apple 4S phone which I am quite happy with and which was originally bought on a contract. I must have had that phone for over 2 years after the minimum contract period had expired before I switched it to a sim-only deal that saved me £15/month. I know about minimum contract periods, sim-only deals etc. but it just took that long for the job of switching to come to the top of my list.
It may be that widespread adoption of switching/money management tools driven by data analysis and AI will make the consumer more “rational” but until that happens inertia and ignorance will remain very important.
I am wholly aware of my own buying irrationality
I am sure I could save a great deal with a bit of effort
And maybe I am lucky that I am not forced to count every penny – I accept that
But even tho use who have to have trade offs to make
And we do make them
The result is that at the very least we will rarely appear rational because non-measured facts impact decisions
And very often we are just not rational
E3ME does strip out some of the worst assumptions. Friedman’s (1953) essay ‘The Methodology of Positive Economics’ claimed that assumptions like ‘firms maximise profits’ could be wrong AND make no difference to the reliability of the theory. Theories should be judged solely on the accuracy of their predictions. I see Sooty waving his magic wand here, yet Friedman’s paper was very influential and beyond the pale. Policy debates were disputes of the consequences of alternatives and positive economics could resolve them. This is very much 19th century positivism with “science” becoming authority, no better in principle than Jordan (1930 ish) trying to link proto-quantum biology to fascism or Enfantin (1870 ish) science to free love.
We should be modelling – this is primarily what science does. E3ME has a 20 year pedigree and is freely available for scrutiny. Economics has long split itself from such as the theory of inequality and poverty measurement, welfare economics, the theory of social choice, the theory of bargaining and of cooperative games, and the theory of fair allocation. There is a need to keep in mind that any positivism (there have been hundreds – we’d need a book to discuss those in Marxism alone) is normative and hermeneutic in some underlying way. It is possible to model in the modellers and express outcomes against green assumptions and best use of resources and so on. Each label in the model is likely to be mythical in respect of what it is supposed to mean in real lived experience.
Poverty (absolute or relative) is not just about finance, but also a signal for reasons to strive and to feel good because you aren’t in it for some. Given that gene patterns for intelligence (crudely) exist across poverty-wealth, continued poverty is a substantial loss of human resource. I’m only guessing but feel modelling might need a big twist away from monetary understandings deeply rooted in myth that may rely on disabling mechanisms we are failing to model. We’ve long been able to do this in such as social mice.
I well remember the economist who once threatened me for saying his work was normative.
He said it was positive because he accepted all the assumptions of neoclassical economics
He seemed unaware as to why I rolled around laughing……
I thought that Mervyn King had been aware of the on coming crisis in 2008 and had aired his worries both here and in the US (to closed minds mostly). Were his realisations sufficiently late in the day not to constitute a forecast?
Of course given the systemic nature of the problem at the BoE it might be that the signs were there but that so much had been invested in particular economic models that they paid no heed to their own information. A bit of a Titanic suitation.
I think David Blanchflower would beg to differ re King
Richard – do I know how much the GOTBOE is paid? Of course I do. Only joking. He gets more than even University CEOs. When I look at these salaries I see an annual income greater than the combined income of my entire working life. I find that offensive because I was reasonably well paid and I live a pretty good lifestyle even as a pensioner. I do wonder why people need/want so much money. I think it reveals a psychological deficiency in that money is used as a scoreboard to show others how well they’re doing rather than the satisfaction of doing a job well for the benefit of society and family. Maybe i was luck, like your Dad, in that i worked for a nationalised industry at a time when the mixed economy was appreciated. I remember when my union lodged a pay claim in 1989 and we were told that it wasn’t acceptable because a Principal Engineers pay was linked to the pay of the Chairman of the CEGB whose pay was linked to some civil service grade. At the time the Chairman, Walter Marshall, was earning something like £70k for running the entire electricity supply industry and a Principal Engineer earned about half of that. A few years later after fragmentation and privatisation the pay of the CEOs mushroomed out of all proportion compared to that of highly skilled engineers. I could go on but I’m sure you talked to your Dad about these things.
My father worked for the nationalised elcectricyrt industry and was horrified by pay escalation for a few after privatisation
He retired to avoid what he saw as corruption
Sorry to be the spoiler here but “rationality” in this context is restricted to an assumption that we, as individuals, always act out of self-interest in order to “maximise utility” (gain optimal satisfaction and best value for money etc.).
To that end, however, we are NOT generally rational and there is (thankfully) a school of economics devoted to the study of that irrationality. Its called Behavioural Finance and it is well worth knowing about. It is premised upon a series of key concepts, ‘heuristics’ and anomalies in human behaviour that are pretty easy to understand.
Some of them: The gambler’s fallacy, Prospect Theory, Cognitive Dissonance, Herd Behaviour, Loss Aversion and Mental Accounting may be familiar in their own right.
For those who are unfamiliar of this I tried finding a good, brief overview on the net and didn’t have much luck so I settled for this piece that is written for investors. You’ll see that it has a list of bullet point links under the main title. Links 3 to 11 provide, decent, brief explanations of Behavioural Finance’s main concepts, well most of them, enough to give you some idea. Its worth a look.
https://www.investopedia.com/university/behavioral_finance/
Thanks
@Marco Fante.
Thanks for the behavioral (sic) finance link.
A classic example of how academics with different ideas argue from faith positions defending their pet theory….discussing rationality !! The irony makes my head hurt.
“A few exceptions apart (the E3ME model of Cambridge Econometrics, where I am a director, being an example) most macroeconomic modelling is done on the basis of general equilibrium analysis.”
You do know that E3ME is a computational general equilibrium model, right?
It just uses a different parameter for the determination of market clearing rates, introducing what is basically a drift parameter for it into the model behavior and regulation.
There is a big element of the pure sales pitch from Cambridge Economics. Most models out there – certainly all the good ones – don’t make the basic assumptions CE claim they do.
CE heavily pitch this model to public and regulating bodies – so it’s not exactly surprising that the model is saying “look how regulation can improve output!”, which the E3ME model is geared towards showing.
E3ME is a pretty good model, but isn’t alone out there as one and would have still more than likely got the crash totally wrong (I wasn’t using it back then, but given it uses a lot of historical data for it’s assumptions, more than likely as I say). It just swaps one set of assumptions for another.
Oh, and as I say, it is still a general equilibrium model.
It is not a model built on what most would call GE bases – which I clearly specified
It uses a Keynesian formulation that is profoundly different
Fords and Bentleys are both cars
They are different
People can spot them apart
I’d suggest GE and E3ME can be spotted apart
I’m interested: how familiar are you with the E3ME model?
Enough
Why?
Well, I ask because you seem to be making some rather strange claims about it.
The E3ME model is a sectoral model (for energy/emmissions), and doesn’t tell us much of anything about the economy as a whole. E3ME uses things like inflation, interest rates, exchange rates as inputs, and doesn’t make any claim to forecasting those general economic indicators.
Which makes me wonder why you are comparing it to the models the Bank of England use. Those models have to be very different because they are trying to forecast very different things.
E3ME is what would normally be described as a GE type model (some would call it a partial equilibrium model). It just swaps out the standard GE assumption of demand/supply equilibrium and exchanges it for a different set of assumptions and estimates – in this case ones about the behavioral aspect of demand/supply optimization (in so far as demand and supply might not always be optimized).
Now there is nothing wrong at all in doing this and it works fairly well for sectoral analysis like E3ME. But it is very important to be aware that you are exchanging one set of assumptions for another.
No-one would use a basic GE model for sectoral economic modelling any more, but if you are looking at an economy as a whole the assumptions made at every sectoral level all have their own error functions, and they all interact with each other. Errors on errors, in effect. With all those estimates, assumptions and unknowns the errors you would likely see in a whole economy version of E3ME would likely be larger than that of a basic GE model – where at least the simple equilibrium assumption can be more easily calibrated and give you a more reasonable error function.
Nobody says GE models are in any way perfect, and not one economist of any note or experience thinks the equilibrium assumption is factually true (it is usually described as an approximate equilibrium, and lots of work goes into characterizing tit’s properties), but that does not mean GE models are useless either. Most GE models have refined the simple equilibrium assumption in various ways anyway, and I wouldn’t say most economics is done using GE models, as you are.
Which is why I am a bit confused about your blog article. You seem to be saying that you have an inherent knowledge of how the Bank of England do their economic modelling (which I can’t see how you do, from your background), that the modelling they do is based on GE only and is somehow flawed as a result, and that they should be using models like E3ME, which you seem to claim is superior to their models. Yet E3ME, despite being a type of GE model, doesn’t even claim to cover the same economic forecasts that would have lead the BoE to spot the financial crash.
It’s a very odd set of claims you are making. Almost as if you are comparing Fords and Bentleys. And apples, just to make sure.
Hmmmm
Now I wonder what you know about E3ME
Because you seem to think it something that is very definitely not what I see it as
Maybe you would like to email me?
At this point I would like to drop in a brief mention on behalf of those of us who contend that any and all attempts at “modelling” entire markets and economies are ill-conceived, in vain and unnecessary.
I won’t go on other than to mention that the great identities and ideas in economics (from Adam Smith through to Keynes) had no reliance on computers and no need of inferential statistics. It’s also worth wondering if the relative scarcity of great ideas and identities since Keynes has anything to do with the undue attention paid to grand models and superfluous math.
I think most economics requires little maths
And no regressions at all
“Now I wonder what you know about E3ME”
I have come across it in the course of my work, where we had an in house course from Cambridge Econometrics themselves. I probably still have the handbook somewhere, and am pretty certain you used to be able to download it from their site. I don’t think there is any secret as to what it is and how it works.
“Because you seem to think it something that is very definitely not what I see it as”
What do you see it as then? What does it do and what is it used for? Please enlighten me.
“Maybe you would like to email me?”
Why?
Because I strongly suspect you do not know how it has developed
“Because I strongly suspect you do not know how it has developed”
Again, please enlighten me. You said you were well acquainted with the model so I would be very interested to hear how it has developed…..from a sectoral economic model of energy and emissions model to a….full spectrum macro economic model??
I am beginning to suspect that you don’t really have the foggiest about E3ME and what it does and doesn’t do though.
“I think most economics requires little maths
And no regressions at all”
I think your friends at Cambridge Econometrics would disagree, given their E3ME model relies heavily on both. But if what you say is true, what does economics require?
Or is it just the economics you do that requires no maths?
I think that your approach suggests an antagonism that is unhealthy for the purposes of debate Bob
And I am not interested in that
Nor do I consider it healthy
I’m not trying to be antagonistic. But you keep avoiding answering any of my questions properly. I either get questions on questions in return, rhetoric or waffle.
I certainly haven’t had a straight answer as to what you say E3ME actually is – which should take you one or two sentences at most.
So yes, I am getting more than a little frustrated. In my experience people unwilling to answer simple questions are usually hiding something – normally a lack of knowledge.
I made you an offer
You declined it
There you go avoiding the question again.
I declined your very kind offer as frankly I don’t want to begin a private discourse with you, not least because I would rather this conversation stayed in public for all to see. If private, you could easily dissemble away.
I would have thought that you would also want to be able to show to the world, or at least the readers of this blog, that you actually can back up what you are saying and claiming.
After all, it should not be too hard. A couple of sentences explaining what E3ME is and is used for is all I am asking.
Easy
https://www.camecon.com/how/e3me-model/
Which reads remarkably like my comments
I had always assumed that when Keynes said that markets might work efficiently (and rationally) provided everyone was in possession of perfect information, he was, in his his way, explaining why markets don’t work efficiently.
Marco’s point about the irrationality of ‘players’ is covered extensively by game theory. Indeed it is rather what game theory is about. A bird in the hand and two in the bush and all that…… and more. Human beings are not cash registers.
I tend to think of wrong by design rather crudely. There are so many league tables around now being gamed under broad headings of quality and targets, I see little reason to believe any economic models at face value. Scientific macro and micro have to match up, at least through techniques of normalisation and re-normalisation (cosmic with quantum). Excuses like biology not reducing to chemistry are not true. It remains difficult to establish enough of the communication nodes of the system we try to study and manipulate. Even with existing models we could do much better than politicians claiming such as ‘universal credit is working well, as job centre staff I talk to say this when I ask’. Most models, if not as inane as this, are not relational enough with concepts and people they exclude.
Not really on topic anymore, but Mr Carney got GBP881,574 in 2017/2018, according to P54 of the last annual report: https://www.bankofengland.co.uk/annual-report/2018