Keynes vs Hayek and cognitive macroeconomics
This evening Paul Mason leads a debate for LSE, the Adam Smith Institute [Update: I made this assumption based on the debate being listed on the ASI's "Events" page. A BBC editor has kindly emailed me to correct my mistake. Thanks!] and Radio 4 on Keynes versus Hayek. It’s a good day to talk about it, with UK growth figures coming in at a disappointing 0.2% and a lot of uncertainty about whether public sector austerity is slowing economic growth, or conversely enabling more private investment which will improve things in the long term.
[Update 3 August 2011: The programme will be broadcast this evening at 8pm UK time, and will be available after that from this address]
On Hayek’s side, Jamie Whyte and George Selgin are speaking; for Keynes, Duncan Weldon and Lord Skidelsky. Do come along if you can, though the event is likely to be full so arrive early (5.30 is suggested). If you don’t get in, come to the Old Bank of England pub on Fleet Street afterwards for drinks and email (leigh@inon.com) or tweet (@leighblue) me if you want to meet up.
The Keynes versus not-Keynes debate (in this country at least) is often seen as a straight left-versus-right argument. It’s not as simple as that. There is certainly a broad identification of Keynes with the centre-left (I’ll explain why below). But the right’s arguments fall into three distinct camps: neoclassical free-market economics, monetarist models (confusingly, the “New Keynesian” school is mostly a monetary argument) and Austrian school, whose intellectual leader was Hayek.
Actually, Keynes, Hayek and monetarism are closer together than we usually think. They all differ from neoclassical economics in a similar way, but their interpretations and responses are distinct.
First I’ll outline the neoclassical argument and explain its weaknesses. Neoclassical economics is in fact the textbook microeconomic argument: a market is made up of many people, each of whom is aware of all the goods and services available, knows which ones they want and in what proportions, and can see how much they have to pay for them. The theory also makes a value judgment about Pareto efficiency, which effectively means that all trades should be voluntary, and any market transaction will only happen if it makes both parties better off. From these assumptions one can build the whole theory of competitive markets and general equilibrium, which says that everyone will become as well off as they can possibly be through private trades.
The assumptions are clearly artificial – not everyone knows all the goods that are available, there are lots of transaction costs when we trade with each other, and most markets are not fully competitive. The question is: is the real world close enough to neoclassical assumptions so that free markets alone will still achieve the best outcomes?
The answer seems to be no. In a neoclassical world there can be no recessions, no unemployment and no inflation. Yet all three of these phenomena are present and worrying right now, and throughout most of recorded economic history. The whole field of macroeconomics has arisen out of this answer, and from the corresponding questions: how does the world differ from those simple assumptions, and what should we do about it?
This is where the three schools of macroeconomic thought diverge – though parts of their answers to those questions are still similar.
The fundamental difference between the real world and the neoclassical world is that there are limits on our brains. We cannot access all the information about all the products in a market, the different people and companies that make them, or even the prices that they are sold for. We cannot even fully process the information we have, because cognitive processing takes time, so we use heuristic shortcuts to reach approximate answers.
The three intellectual traditions differ in what they do with this insight.
The Keynesian school observes that – for various reasons – people sometimes become more risk-averse. Emotions are a factor in the economic decisions we make; so are news stories and peer pressure; and these can lead to a widely held concern about the future. In this circumstance, people want to save more money than before – which they do by spending less on buying things from other people. However, if everyone wants to save money at the same time, you get an effect which was once described to me as “mattress economics”. If everyone puts their money under the mattress at the same time, nobody ends up with any more money; you will save the money you would have spent hiring me to cut your hair, but I will also save the money I would have spent buying the apples you’ve grown. We both end up with the same amount of money, but with no haircuts or apples.
This problem would be solved if we could all cut our prices at the same time – if haircuts suddenly cost half as much, and apples cost half as much, then my savings would be worth twice as much in real terms, and I would feel it was safe to spend some of it. But we can’t do this – both because we have a cognitive bias towards keeping things the same, and because it is very hard to get a clear picture of what’s happening in the market. Maybe this is just a temporary disruption because of the snow, and you’ll come for a haircut tomorrow. Maybe I have a new banana pie recipe, so I just don’t feel like apples today. Market insight is hard to acquire and takes a long time, so prices take a long time to adjust.
The Keynesian response: government spending. If the government borrows money from us all collectively, it can soak up all the savings under the mattress. And it can then buy your apples (and maybe give them to the poor) and hire me to give haircuts to the army – or give the money to poor people who haven’t had a haircut in years (insert joke about investors in Greek debt).
So that’s Keynes.
Hayek agrees with Keynes about one thing – markets are not neoclassical, because we don’t have full knowledge of everything that is going on. Hayek’s work is partly based on that of von Mises, who wrote a lot about the limits of human cognition and knowledge – he invented a whole field to study it, called praxeology. However, Hayek disagrees on the consequences – and the remedy.
The Austrian model says that things change over time, in an unpredictable way (another difference from the neoclassical models, where everything is either fixed for all time or highly predictable). If I stop buying apples, it might be because someone has discovered strawberries. Or because scientists found that we shouldn’t eat so much fruit after all. Or the fashion has just changed, and apples are now less popular than blackberries.
If this happens, it’s inevitable that apple farmers will become unemployed. It’s a sad fact of life but there’s no point subsidising someone to just keep growing a fruit that nobody wants. Apple prices will fall (maybe not right away, but soon enough), the farmers won’t be able to keep trading – or they will find that they can make more money by ploughing up the orchards and planting strawberries, or building cottages.
Government should accept this and let it happen. There’s no way for a government – sitting in Washington DC or London, thousands or hundreds of miles from the orchards and the supermarkets – to know whether apple demand has fallen permanently or temporarily. Only the farmers, supermarkets and consumers can know that – or at least, even if they don’t know it yet, they will find out by experimenting with new prices or different ways to promote apples.
If the government does try to intervene, it will just cause more trouble – either by buying millions of apples that nobody wants, or by giving money to poor people who will not buy apples either, but will instead try to spend it on strawberries – driving the price of strawberries up and causing inflation.
And that’s Hayek.
I’ll briefly mention monetarism because it’s an important strand of thought, even though not part of this evening’s debate. Monetary policy starts from roughly the same place as Keynes – there is an overall fall in economic demand because people want to save more money. But, instead of the Keynesian approach of saying “OK, we’ll let them save more money (by lending it to the government) and we’ll spend it for them”, monetarists say “if people want to save more money, we’ll give them more money directly”. This is done by, in effect, printing money at the Bank of England or the Federal Reserve. With this money, they buy government bonds or something similar, so that the money finds its way out into the market and people can spend or save it as they wish. This in turn reduces interest rates (unless they’re already close to zero, like now) and makes more money available for investment. Investment increases GDP and helps to end the recession.
Austrian economists disagree with this too, because they say the central bank can’t know whether interest rates are too high or too low – it is for the market to decide. Keynesians tend to be broadly accepting of monetary policy, though some argue that it stops working when interest rates are too low and the banking system is in danger – like now.
So the technical differences between the schools, in summary: Keynesians say that mass psychology drives falls in demand across all sectors of the economy, and needs to be fixed with government intervention; Hayekians say that technological or other changes drive falls in demand in some sectors, and that the market is the only way the economy can adjust to this.
There are philosophical differences too. Keynesians tend to see the economy as one big aggregate – the goal is to generate the maximum amount of wealth across the whole of society. This might mean taxing some people to spend on others – if the net effect is positive, that’s fine. Hayekians are usually more individualistic. It’s a bad thing to forcibly take money from people, even if it makes the whole population better off (and anyway, they say, it usually doesn’t).
This is why the two schools largely correspond with left and right wing politics respectively. Keynesians are OK with government spending in general (which by its nature, nearly always implies some degree of taxing the rich to spend on the poor) – so they believe government spending is a legitimate tool to cure economic problems. And if they can use the opportunity to get some more schools or hospitals built, or more transfers to poor inner-city children, that might not have been passed otherwise, so much the better.
Hayekians don’t like the government taking anyone’s resources, and they also believe that doing so is not effective at improving the economy. In fact, they believe it’s actively counterproductive, because it slows the natural process of economic change and progress, as well as providing the wrong incentives for people to invest and generate wealth.
Both beliefs may be a little self-serving: it’s not that common to find someone who believes in government spending and redistribution but thinks Keynesian policies are economically damaging; nor someone who doesn’t like government spending philosophically but believes Keynesian intervention is effective anyway. But that’s life – our beliefs about which solutions are effective often match our values about what kind of solutions are morally legitimate. It’s hard to know which causes which.
My final question is this: Is there really an unbridgeable gap between these two philosophies?
Personally, I believe in herding, and in the effect of emotions on economic decision-making, because I’ve seen them both regularly in my research. And I believe the Keynesian diagnosis that increased desire for savings can directly cause an economic slowdown across all sectors.
And yet, I also believe that there are big inter-sectoral changes. It’s hard to know whether a slowdown (such as today’s slow growth) arises from overall fear and lack of demand, or from snow, royal weddings and a growing desire to use twitter instead of going to the cinema.
There is truth to both models – the economy is made of thinking, feeling, cognitively biased people who do not act like rational economic agents; and it also changes quickly and unpredictably because of new technology, countries getting richer or older, and many other pressures.
Are there policy solutions that can handle both of these problems? Yes, there are.
We should work with the grain of human knowledge and cognition instead of against it.
Use cognitive and psychological research to understand when the economy is suffering an overall fall in demand and not just a sectoral change.
Use monetary policy to increase the amount of money in the economy in a targeted way – which will cause mild inflation (the good kind) and reduce the effect of sticky prices.
Use deficit spending – but in targeted ways which will help the economy to make its transitions instead of hindering it. Education and job training are the most important of these – guided especially by the needs of newly emerging industries. More unemployment insurance, but partly tied to the recipient’s willingness to try new things and absorb new ideas.
Market testing of new services; randomised testing of new policies. A financial system that is structured to encourage diversity of investments and reduce herding – for example capital reserve requirements which are higher if the bank’s assets are closely correlated with those of other banks.
And freedom for individuals and companies to come up with good ideas and try them out; minimal regulations and openness of government to working with private or voluntary service providers in all areas.
In short, subsidise the production and testing of new ideas; let the market flourish; and if people start herding, all heading the same way and reducing diversity, disrupt the flow and shake things up.
This is what a cognitive macroeconomics would look like – and I think both Keynes and Hayek might like it.
[Update 3 August 2011: The programme will be broadcast this evening at 8pm UK time, and will be available after that from this address]
On Hayek’s side, Jamie Whyte and George Selgin are speaking; for Keynes, Duncan Weldon and Lord Skidelsky. Do come along if you can, though the event is likely to be full so arrive early (5.30 is suggested). If you don’t get in, come to the Old Bank of England pub on Fleet Street afterwards for drinks and email (leigh@inon.com) or tweet (@leighblue) me if you want to meet up.
The Keynes versus not-Keynes debate (in this country at least) is often seen as a straight left-versus-right argument. It’s not as simple as that. There is certainly a broad identification of Keynes with the centre-left (I’ll explain why below). But the right’s arguments fall into three distinct camps: neoclassical free-market economics, monetarist models (confusingly, the “New Keynesian” school is mostly a monetary argument) and Austrian school, whose intellectual leader was Hayek.
Actually, Keynes, Hayek and monetarism are closer together than we usually think. They all differ from neoclassical economics in a similar way, but their interpretations and responses are distinct.
First I’ll outline the neoclassical argument and explain its weaknesses. Neoclassical economics is in fact the textbook microeconomic argument: a market is made up of many people, each of whom is aware of all the goods and services available, knows which ones they want and in what proportions, and can see how much they have to pay for them. The theory also makes a value judgment about Pareto efficiency, which effectively means that all trades should be voluntary, and any market transaction will only happen if it makes both parties better off. From these assumptions one can build the whole theory of competitive markets and general equilibrium, which says that everyone will become as well off as they can possibly be through private trades.
The assumptions are clearly artificial – not everyone knows all the goods that are available, there are lots of transaction costs when we trade with each other, and most markets are not fully competitive. The question is: is the real world close enough to neoclassical assumptions so that free markets alone will still achieve the best outcomes?
The answer seems to be no. In a neoclassical world there can be no recessions, no unemployment and no inflation. Yet all three of these phenomena are present and worrying right now, and throughout most of recorded economic history. The whole field of macroeconomics has arisen out of this answer, and from the corresponding questions: how does the world differ from those simple assumptions, and what should we do about it?
This is where the three schools of macroeconomic thought diverge – though parts of their answers to those questions are still similar.
The fundamental difference between the real world and the neoclassical world is that there are limits on our brains. We cannot access all the information about all the products in a market, the different people and companies that make them, or even the prices that they are sold for. We cannot even fully process the information we have, because cognitive processing takes time, so we use heuristic shortcuts to reach approximate answers.
The three intellectual traditions differ in what they do with this insight.
The Keynesian school observes that – for various reasons – people sometimes become more risk-averse. Emotions are a factor in the economic decisions we make; so are news stories and peer pressure; and these can lead to a widely held concern about the future. In this circumstance, people want to save more money than before – which they do by spending less on buying things from other people. However, if everyone wants to save money at the same time, you get an effect which was once described to me as “mattress economics”. If everyone puts their money under the mattress at the same time, nobody ends up with any more money; you will save the money you would have spent hiring me to cut your hair, but I will also save the money I would have spent buying the apples you’ve grown. We both end up with the same amount of money, but with no haircuts or apples.
This problem would be solved if we could all cut our prices at the same time – if haircuts suddenly cost half as much, and apples cost half as much, then my savings would be worth twice as much in real terms, and I would feel it was safe to spend some of it. But we can’t do this – both because we have a cognitive bias towards keeping things the same, and because it is very hard to get a clear picture of what’s happening in the market. Maybe this is just a temporary disruption because of the snow, and you’ll come for a haircut tomorrow. Maybe I have a new banana pie recipe, so I just don’t feel like apples today. Market insight is hard to acquire and takes a long time, so prices take a long time to adjust.
The Keynesian response: government spending. If the government borrows money from us all collectively, it can soak up all the savings under the mattress. And it can then buy your apples (and maybe give them to the poor) and hire me to give haircuts to the army – or give the money to poor people who haven’t had a haircut in years (insert joke about investors in Greek debt).
So that’s Keynes.
Hayek agrees with Keynes about one thing – markets are not neoclassical, because we don’t have full knowledge of everything that is going on. Hayek’s work is partly based on that of von Mises, who wrote a lot about the limits of human cognition and knowledge – he invented a whole field to study it, called praxeology. However, Hayek disagrees on the consequences – and the remedy.
The Austrian model says that things change over time, in an unpredictable way (another difference from the neoclassical models, where everything is either fixed for all time or highly predictable). If I stop buying apples, it might be because someone has discovered strawberries. Or because scientists found that we shouldn’t eat so much fruit after all. Or the fashion has just changed, and apples are now less popular than blackberries.
If this happens, it’s inevitable that apple farmers will become unemployed. It’s a sad fact of life but there’s no point subsidising someone to just keep growing a fruit that nobody wants. Apple prices will fall (maybe not right away, but soon enough), the farmers won’t be able to keep trading – or they will find that they can make more money by ploughing up the orchards and planting strawberries, or building cottages.
Government should accept this and let it happen. There’s no way for a government – sitting in Washington DC or London, thousands or hundreds of miles from the orchards and the supermarkets – to know whether apple demand has fallen permanently or temporarily. Only the farmers, supermarkets and consumers can know that – or at least, even if they don’t know it yet, they will find out by experimenting with new prices or different ways to promote apples.
If the government does try to intervene, it will just cause more trouble – either by buying millions of apples that nobody wants, or by giving money to poor people who will not buy apples either, but will instead try to spend it on strawberries – driving the price of strawberries up and causing inflation.
And that’s Hayek.
I’ll briefly mention monetarism because it’s an important strand of thought, even though not part of this evening’s debate. Monetary policy starts from roughly the same place as Keynes – there is an overall fall in economic demand because people want to save more money. But, instead of the Keynesian approach of saying “OK, we’ll let them save more money (by lending it to the government) and we’ll spend it for them”, monetarists say “if people want to save more money, we’ll give them more money directly”. This is done by, in effect, printing money at the Bank of England or the Federal Reserve. With this money, they buy government bonds or something similar, so that the money finds its way out into the market and people can spend or save it as they wish. This in turn reduces interest rates (unless they’re already close to zero, like now) and makes more money available for investment. Investment increases GDP and helps to end the recession.
Austrian economists disagree with this too, because they say the central bank can’t know whether interest rates are too high or too low – it is for the market to decide. Keynesians tend to be broadly accepting of monetary policy, though some argue that it stops working when interest rates are too low and the banking system is in danger – like now.
So the technical differences between the schools, in summary: Keynesians say that mass psychology drives falls in demand across all sectors of the economy, and needs to be fixed with government intervention; Hayekians say that technological or other changes drive falls in demand in some sectors, and that the market is the only way the economy can adjust to this.
There are philosophical differences too. Keynesians tend to see the economy as one big aggregate – the goal is to generate the maximum amount of wealth across the whole of society. This might mean taxing some people to spend on others – if the net effect is positive, that’s fine. Hayekians are usually more individualistic. It’s a bad thing to forcibly take money from people, even if it makes the whole population better off (and anyway, they say, it usually doesn’t).
This is why the two schools largely correspond with left and right wing politics respectively. Keynesians are OK with government spending in general (which by its nature, nearly always implies some degree of taxing the rich to spend on the poor) – so they believe government spending is a legitimate tool to cure economic problems. And if they can use the opportunity to get some more schools or hospitals built, or more transfers to poor inner-city children, that might not have been passed otherwise, so much the better.
Hayekians don’t like the government taking anyone’s resources, and they also believe that doing so is not effective at improving the economy. In fact, they believe it’s actively counterproductive, because it slows the natural process of economic change and progress, as well as providing the wrong incentives for people to invest and generate wealth.
Both beliefs may be a little self-serving: it’s not that common to find someone who believes in government spending and redistribution but thinks Keynesian policies are economically damaging; nor someone who doesn’t like government spending philosophically but believes Keynesian intervention is effective anyway. But that’s life – our beliefs about which solutions are effective often match our values about what kind of solutions are morally legitimate. It’s hard to know which causes which.
My final question is this: Is there really an unbridgeable gap between these two philosophies?
Personally, I believe in herding, and in the effect of emotions on economic decision-making, because I’ve seen them both regularly in my research. And I believe the Keynesian diagnosis that increased desire for savings can directly cause an economic slowdown across all sectors.
And yet, I also believe that there are big inter-sectoral changes. It’s hard to know whether a slowdown (such as today’s slow growth) arises from overall fear and lack of demand, or from snow, royal weddings and a growing desire to use twitter instead of going to the cinema.
There is truth to both models – the economy is made of thinking, feeling, cognitively biased people who do not act like rational economic agents; and it also changes quickly and unpredictably because of new technology, countries getting richer or older, and many other pressures.
Are there policy solutions that can handle both of these problems? Yes, there are.
We should work with the grain of human knowledge and cognition instead of against it.
Use cognitive and psychological research to understand when the economy is suffering an overall fall in demand and not just a sectoral change.
Use monetary policy to increase the amount of money in the economy in a targeted way – which will cause mild inflation (the good kind) and reduce the effect of sticky prices.
Use deficit spending – but in targeted ways which will help the economy to make its transitions instead of hindering it. Education and job training are the most important of these – guided especially by the needs of newly emerging industries. More unemployment insurance, but partly tied to the recipient’s willingness to try new things and absorb new ideas.
Market testing of new services; randomised testing of new policies. A financial system that is structured to encourage diversity of investments and reduce herding – for example capital reserve requirements which are higher if the bank’s assets are closely correlated with those of other banks.
And freedom for individuals and companies to come up with good ideas and try them out; minimal regulations and openness of government to working with private or voluntary service providers in all areas.
In short, subsidise the production and testing of new ideas; let the market flourish; and if people start herding, all heading the same way and reducing diversity, disrupt the flow and shake things up.
This is what a cognitive macroeconomics would look like – and I think both Keynes and Hayek might like it.
Comments
Economists still don't have a good idea how we got into this mess.
They can't tell us how to get out of it.
And most importantly they can't tell us where we will be in a year's time, three years, ten years...
The sine qua non of a valid science being that it makes good predictions.
The only fiscal intervention proven to work is control of money supply, but governments are most unwilling to do that.
In a time of inflation, capping bank lending would hurt bank profits, and politicians are owned by (among others) the banks.
To quote Scrooge, "Humbug!"
Nothing will bridge the gap between the Keynesians, who are answering the question "How can we produce widespread prosperity?" and the Hayekians, who are saying "How can we keep the Keynesians from producing widespread prosperity?"
This country has completely lost its manufacturing base and many small businesses and yet the economists answers lie in the virtual world of investment banking.
The only answer in the real world is the John Lewis format for companies to invest in the future and their staff, protected from the predation of private equity. It does require the support of Government to create the necessary incentives such as reducing corporation tax to Zug levels and enhance the pension entitlements.
We should look to the Mondragon group of coops in Spain as an example but it will take many decades to establish.
The short debate could not really go into much depth, of course, but I was interested to hear that the main Hayek point was not to have a boom in the first place. Since free marketeers (private banks) generated most of the credit for speculation (derivatives etc), I am not sure how they can blame anyone other than "the free market" for the speculative booms of recent years.
How could this blog skip this whole argument? This is perhaps the most applicable Austrian argument today, and it directly ties into what the writer is supposedly talking about.
I see the most fundamental difference between Keynes and Hayek as follows. When things change in the economy, as they always do, should we:
(Keynes) collectively help cushion the adjustment by supporting employment and investment? This will result in some things being kept away from their natural market-clearing price, but will reduce short-term suffering and may increase total economic welfare.
or (Hayek) let prices adjust as quickly as possible to the new reality so that people redirect their labour and assets to new more productive uses? This will arguably result in a more efficient economy more quickly, but at the cost of volatility and short-term pain.