Paul Davidson is America’s foremost post-Keynesian economist. Davidson is currently the Holly Professor of Excellence, Emeritus at the University of Tennessee in Knoxville. In 1978 Davidson and Sydney Weintraub founded the Journal for Post-Keynesian Economics. Davidson is the author of numerous books, the most recent of which is an introduction to a post-Keynesian perspective on the recent crisis entitled ‘The Keynes Solution: The Path to Global Prosperity’.
Interview conducted by Philip Pilkington
Philip Pilkington: Keynes famously claimed that the ideas of economists are extremely powerful and have huge influence on the way policymakers think. What struck me about your book The Keynes Solution was how well you related Keynes’ theoretical ideas to the problems the world is currently facing – and the proposed solutions. Before we talk in any detail about these ideas let me ask you this: to what extent do you think that Keynes was right about the ideas of economists?
Paul Davidson: He was absolutely right. All you need to do is look at the history of the US (and other developed countries) since World War II.
For the first 27 post war years, Paul Samuelson’s mistaken view of Keynes’ analysis reigned supreme as the Keynesian economic solution to all economic problems – and all politicians, whether Republican or Democrat, thought of fiscal stimulus as the correct response to recession. The Republicans via tax cuts and military spending, while the Democrats by more public spending.
The problem, as I indicate in my new book, was Samuelson got Keynes’ theory all wrong. Thus when inflation reared its ugly head in the early 1970s Samuelson had no idea as to Keynes’ solution to what Keynes called incomes inflation.
Instead Samuelson invoked the Phillips curve – which was not part of the Keynes analytical structure and was actually incompatible with Keynes’ General Theory – to advocate a trade-off between the rate of inflation and the rate of unemployment. When inflation rates increased as unemployment rates increased in the mid 1970s, i.e., stagflation – Samuelson’s Keynesianism was proved wrong and the door was open for Milton Friedman’s solution to all our economic problems; that is, the free market in goods and services, labor market free , and unregulated financial markets. Mrs. Thatcher and Ronald Reagan relied on this free market philosophy – as did to a large extent Clinton under the influence of Alan Greenspan and Robert Rubin.
As a result government regulations were reduced, labor unions were emasculated (remember Reagan and the air traffic controllers) the Federal Reserve aimed at targeting the rate of inflation (as Friedman’s quantity theory of money explained how the money supply did not affect real output and employment but merely affected the price level).
Milton Friedman and I had a debate about his theory vs. my view of Keynes in the Journal for Post-Keynesian Economics in the early 1970s which was then published as part of a book Milton Friedman’s Monetary Framework.
So we have two now defunct economists having the power to affect policies for more than 60 years through their academic scribbling – Samuelson for the first three decades after World War II and Friedman for the next three decades. Clearly this evidence of the power of economists – who Keynes claimed should be as humble as dentists.
PP: In the book you trace a lot of mainstream economics faith in markets back to what you call the ‘ergodic assumption’. Could you say something about the relevance of this concept to understanding mainstream theory – and why if this assumption is thrown out much of said theory has to go with it?
PD: Time is a device that prevents everything from happening at once. Economic decisions made today will have payoffs in days, weeks, months and/or years in the future. Mainstream economic theory presumes that decision makers can make optimal decisions regarding allocation of income and capital where the outcomes (payoffs) of such decisions are in the future. In order to make optimal decisions, these decision makers must ‘know’ future outcomes of all alternatives that they have to choose from. How do they ‘know’ these future outcomes?
To make statistically reliable estimates about future outcomes, the decision maker, following good statistical methodology, should draw a sample from the future and then calculate statistically reliable estimates about future probability of outcomes.
Since drawing a sample from the future is impossible, mainstream economists impose the ergodic axiom as a foundation of MAINSTREAM theory. This ergodic axiom states that the past, present and future economic events are all determined by a pre-existing and unchanging probability distribution. Therefore if one calculates probabilities on the basis of existing past market data, then , assuming the economic system is an ergodic system, this calculated probability distribution is the same as one would get if it calculated a probability distribution from a sample drawn from the future. In other words, the mainstream ergodic axiom presumes that data samples from the past are equivalent to drawing data samples from the future. Consequently the future is readily ‘knowable’ and decision makers will not make any mistakes.
Given this ergodic presumption individuals make the best choice possible to maximize their income and profits in markets free from government regulations and interference. This results in the rhetorical question that Ronald Reagan would often ask: “Why should bureaucrats in Washington know better than you do in how to spend your money?” The ergodic axiom is the basis of the laissez-faire philosophy that underlies mainstream economics.
[Note the risk management computer models used by large financial institutions such as Citibank, Lehman Brothers, etc. involved calculations of risk probabilities based on past financial market data to warn management about the future risks of any decision regarding future portfolio changes. These models, however, failed to warn these large financial institutions of the financial crash of 2007-2008. The result was that government had to bail out these large financial institutions in order to prevent a bigger economic disaster from occurring.]
On the other hand, if the future is uncertain as Keynes insisted is the case – i.e. the system is nonergodic – then there is a positive active role for government to produce market rules and regulations that prevent individuals from taking decisions that can cause markets to crash! [Similar to the government enforcing traffic rules to prevent auto accidents.] There is also a role for government to clean up the mess if the system does crash – via fiscal ‘stimulus’ policies and easy monetary policies!!
Accordingly, if one rejects the ergodic axiom as not relevant to our money using capitalist system, then all of mainstream theory is mere fiction.
[Note George Soros has argued that mainstream financial market theory is not applicable to real world financial markets because actions taken by market participants today can change the future outcomes in ways that are not predictable today – Soros calls this his ‘reflexivity’ concept. Thus Soros is in essence rejecting the ergodic axiom of mainstream theory.]
PP: Together with the ergodic axiom there is another key tenet of mainstream economics that you highlight in your book The Keynes Solution and that is the ‘neutral money argument’. Could you briefly explain this concept and why it is so important for mainstream economic theory?
PD: A second fundamental axiom of new and old classical mainstream economics is that “money is neutral”. For example, Milton Friedman asserted that money was always neutral – at least in the long run.
This neutral money presupposition of mainstream economics assumes that increases in the supply of money will affect neither the volume of goods produced nor the level of employment in the economy – even if there is significant unemployment and excess capacity to produce goods in the economy. In other words, the mainstream axiom of neutral money asserts as a universal truth that does not have to be proven that any monetary policy that increases the quantity of money available in the economy in order to increase market demand for the products of industry will have absolutely no effect on employment and production.
Consequently, the neutral money axiom is a presumption that any easy monetary policy is always inflationary. Thus when the Federal Reserve invoked what was called a ‘quantitative easing’ [often called QE policy] where the Central bank ‘printed’ money to finance an increase in market demand for goods, believers in mainstream classical economics insisted that the result of quantitative easing will be only to increase the inflation rate dramatically.
When inflation did not increase after QE and QE2 in 2010 and 2011 respectively, logically consistent mainstream economists said that the inflation rate increase was just around the corner and we will see it soon.
For Keynes and his Post Keynesian followers, inflation is not a matter of increasing the money supply. Inflation is the result of either speculation regarding stocks of existing commodities – what we call ‘Commodities Inflation’ and/or increases in profit margins and increases in wages relative to productivity increments – what we call ‘incomes inflation’. Consequently, anti-inflationary policies, as I explain in my book The Keynes Solution, involves a buffer stock policy for commodity inflation and an incomes policy for incomes inflation
Blind adherence to the neutral money axiom prevents mainstream theorists from recognizing that if there significant unemployment and idle capacity then additional government and private spending financed by the quantitative easing of the central bank will create profit opportunities for entrepreneurs to expand output and employment – and, of course, will not necessarily be inflationary if the government has a buffer stock policy and an incomes policy in action.
PP: Yet, Bernanke et al seem to be mainstream economists. Why is it that they occasionally advocate expansionary policies and yet continue to believe in their doctrines rather than scrutinising them in light of both their prescriptions and the evidence that emerges as a result of thee prescriptions?
PD: A sage once said: “Consistency is the mark of a small mind”. Nobody ever called Bernanke et al ‘small minded’. Yet there often appears to be an inconsistency with the policies Bernanke et al advocate and their logical theory.
Most mainstream economists such as Bernanke et al who call themselves [New] Keynesians never read Keynes’ General Theory. [Someone once said a ‘classic’ is defined as a book everyone cites but no one reads. For mainstream Keynesians Keynes's book The General Theory of Employment, Interest and Money is definitely a classic.]
Most post-WWII ‘Keynesians’ learned their views of Keynes – not from reading Keynes – but from what Nobel Prize winner Paul Samuelson said Keynes’ theory was. As I indicate in my book The Keynes Solution, in an interview in 1986 Samuelson specifically stated he found Keynes’ analysis ‘unpalatable’ and not comprehensible. Samuelson therefore stated “I finally convinced myself to stop worrying about it… I was content to assume that there was enough rigidity in relative prices and wages to make the Keynesian alternative to Walras operative.”
In other words, Samuelson merely assumed the system was a classical model but with [in the short run] rigidity in money wages – so that the classical market ‘solution’ only came about in the long run. This despite Keynes having an entire chapter in the General Theory entitled ‘Changes in Money Wages’ which analyzed how flexible wages (and prices) still did not automatically restore full employment in Keynes’ general theory. [Did Samuelson ever read Keynes' ‘classic’?]
Of course, 19th century economists already knew that if wages and prices were not flexible, unemployment could occur – so if Samuelson was correct in his interpretation of Keynes, then Keynes had nothing new to add to already well established classical theory! Consequently, if Samuelson was right, Keynes was a charlatan when he claimed to having provided a revolutionary new general theory.
Since the ‘master’ American post-war Keynesian, Paul Samuelson, was inconsistent between his short-run analysis and the long run, it is no wonder that his ‘Keynesian’ students and followers such as Bernanke, Krugman, Stiglitz, etc. are also inconsistent. As John Williamson [the originator of the so called ‘Washington Consensus’] once wrote to me: his view was that American Keynesians are too impatient to wait for the long run to solve the recession and unemployment problems. This ‘impatience’ is, I think, why Bernanke et al demand active government policies even though in the long run – according to their doctrine – no government intervention is better. I believe that is why Bernanke et al are inconsistent between their short run policy desires and the long run logical conclusion of their theoretical doctrine.
PP: Yes, that would explain it perfectly. While they find such solutions theoretically unpalatable, they nevertheless think them eminently practical. Interesting.
Moving on. You talk a bit in the book about commodity price inflations. What do you have to say about these and is there a Keynesian solution?
PD: Commodity inflation is identified with rising market prices of durable standardized commodities such as agricultural products, crude oil, minerals, etc. Typically these commodities are traded in public markets similar to the liquid asset security markets. These commodity markets tend to have prices associated with a specific date of delivery – either delivery today or at a specific date in the future. Markets for a future date delivery are limited to only a few months in the future.
Since most commodities tend to take a significant length of time to produce, the supply of commodities for any future date in these markets are fixed by already existing stocks plus any semi-finished product that are currently in the pipeline and are expected to be finished by the date of delivery. Accordingly, if there is any expectation of an increase in market demand for these commodities in the future, then people can speculate on these markets that the prices will rise.
Accordingly by buying a purchase contract in a market for future delivery the purchaser (speculator) believes he/she will make a large capital gain when he/she sells the product at the future delivery date in the spot market at a higher market price. In sum, any expected increase market demand will only inflate the price of these commodities today.
To prevent commodity price inflation requires the government to maintain an inventory of the commodity as a ‘buffer stock’ to prevent changes in demand and/or supply from inducing significant commodity price movements. A buffer stock is nothing more than some commodity shelf-inventory that can be moved into and out of the market to buffer the market from disorderly price disruptions by offsetting the previously unforeseen changes in demand or supply as they occur.
For example, since the oil price shocks of the 1970′s, the United States has developed a ‘strategic petroleum reserve’ where crude oil is stored in underground salt domes on the coast of the Gulf of Mexico. These oil reserves are designed to provide emergency market supplies to buffer the market price of domestic oil if there is a sudden decrease in oil supplies from the politically unstable Middle East. The strategic use of such a petroleum reserve means that the price of oil will not increase as much as it otherwise would if, for example, a political crisis broke out in the Middle East.
In other words, oil price inflation could be avoided as long as the buffer stock remained available to offset any immediately available commodity shortage. Thus, during the short Desert Storm war against Iraq in 1991, U.S. government officials made strategic petroleum reserves available to the market to offset the possibility of disruptions (actual or expected) from affecting the market price of crude oil. The Department of Energy estimated that this use of a buffer stock prevented the price of gasoline at the pump from rising about 30 cents per gallon during the brief Desert Storm period.
Use of buffer stocks as a public policy solution to commodity price inflation is as old as the biblical story of Joseph and the Pharaoh’s dream of seven fat cows followed by seven lean cows. Joseph – the economic forecaster of his day – interpreted the Pharaoh’s dream as portending seven good harvests where production (supply) would be much above normal causing prices (and farmers’ incomes) to be below normal. This would be followed by seven lean harvests where annual production would not provide enough food to go around while prices farmers received would be exorbitantly high. Joseph’s civilized policy proposal was for the government to store up a buffer stock of grain during the good years and release the grain to market, without profit, during the bad years. This would maintain a stable price over the fourteen harvests and avoiding inflation in the bad years while protecting farmer’s incomes in the good harvest years. The Bible records that this civilized buffer stock policy was a resounding economic success.
PP: There is a concern that, should governments pursue full employment policies we might see inflation arise due to wage-bargaining by workers. What do you think of this argument and how would you suggest dealing with this problem should it arise?
PD: Full employment might strengthen the bargaining power of labor sufficiently so that workers demand money wage increases which exceed the increase in the rate of productivity increase of labor, This will result in higher labor costs of production and therefore higher prices of products produced in firms throughout the economy. This increase in money wage rates that exceeds productivity increase is what Keynes called ‘incomes inflation’. Accordingly the government will have to institute some form of an ‘incomes policy’ to limit wage increases to not exceed productivity increases. A so called tax-based incomes policy or TIP policy would prevent this incomes inflation. TIP would use the corporate income tax structure to penalize the largest domestic firms in the economy if they agreed to such inflationary wage demands. If money wage increases could be limited to overall labor productivity increases, then workers end all other owners of inputs to the domestic production processes might willingly accept noninflationary money income increases.
PP: In your book you are sceptical about using a devaluation of the US dollar to decrease the trade deficit and raise employment. Yet many economists think that this would be a good strategy. Could you explain your scepticism and lay out briefly an alternative arrangement?
PD: There are several reasons for my scepticism that devaluing the US dollar is a desirable policy for reducing the US trade deficit by making imports sufficiently more expensive and making US exports cheaper to foreigners so that Americans reduce their total US spending on imports and foreigners increase their total US spending on US exports, until spending on imports equals what foreigners spend to buy US exports.
The first reason this is not a desirable plan is based on a technical condition called the ‘Marshall-Lerner condition’. Basically this condition says if the sum of the price elasticities for imports and exports is less than unity, then devaluation will increase the trade deficit. One need only note that in the 1980s when the US dollar exchange rate tended to decline rapidly, the US trade deficit increased! Consequently, from an empirical point of view, the Marshall-Lerner condition seems relevant to the USA trade balance problem then and therefore devaluation could worsen the US trade deficit, as it did in the 1980s. A sage once said: “those who do not study history are bound to repeat its errors”. The many economists you cite as thinking devaluation is a good strategy, obviously want the US to repeat the errors of the 1980s.
Secondly, those who argue that devaluation would end the US trade deficit are essentially arguing that if US exports become cheap enough to foreigners then US made products and their labor input would become ‘competitive’ with foreign produced goods’ labor costs. Since labor productivity in most manufactured goods is the same whether produced in the USA or in China, this means that the US labor costs per unit of output when measured in a single currency must be competitive with the labor costs , say, in China in order for US produced goods to be able to compete in the global marketplace.
In other words for US goods to be competitive, devaluation must reduce the value of labor income in the in the US to the equivalent of what is earned in China (where a recent New York Times article indicated workers in a Chinese factory were paid $15 a day for a 12 hour work day and 6 day work week). But do we really want a policy which reduces the average American worker’s annual income to less than $5000 per year? (The economists who say ‘yes’ really mean as long as we don’t reduce economists incomes!)
I should think that we don’t want to advocate a policy that reduces American workers income to the level of a Chinese ‘coolie’. Especially when there is a simpler way to correct the US trade deficit. This alternative is what I have called the International Monetary Clearing Union [IMCU] where the trading nations would essentially enter into an agreement to spend all their income earned on exports on the purchase of imports from foreigners. [The details of my IMCU plan can be read in my book The Keynes Solution]. Accordingly no nation could run a persistent trade imbalance. While at the same time no nation need depress their workers income.