Those bad Keynesians are to blame

Today I have been working on a new book and have been deeply emeshed in paradigmic debates. The practical relevance, other than the work gives me another day’s pay to maintain my part in keeping aggregate demand growth moving, is that two Nobel prize winners (Phelps and Krugman) have had a recent paradigmic dispute about similar themes. One attack was implicit (Phelps on Keynesians), the other very direct and personal (Krugman on Phelps). Neither understand modern monetary theory (MMT) although Krugman is closer than Phelps. Phelps’s work, in my view, has been used by neo-liberals for years to undermine the employment prospects of millions of workers. It is also a primary IMF tool for keep less developing countries poor. Sounds like a topic to be discussed.

A fruitless clash of economic opposites by Edmund Phelps appeared in the Financial Times on November 2, 2009. The link I have provided will allow you to read it in full without disclosing any personal details (by way of registration).

As background, Phelps played a big role in my earlier life – given I was working on Phillips curves and wages and inflation as part of my masters coursework and thesis and again during my PhD work. He was the big wheel of orthodoxy that I was attacking at the time (ps: I have never stopped).

In yesterday’s New York Times column, Paul Krugman attacks Phelps for misrepresenting his opponents and “not listening” to what is being said around him. After quoting Phelps’s first paragraph which said “Keynesian economics, which had been nearly forgotten inside the macro field, has found new voices from outside. They take the position that fiscal ‘stimulus’ of all kinds is effective against slumps of all causes”, Krugman replies:

OK, no point in reading any further.

Nobody, and I mean nobody, holds that alleged position. The position held by Keynesians … is that fiscal stimulus is necessary only under certain special conditions. Namely, when you’re up against the zero lower bound, and conventional monetary policy is useless, fiscal stimulus may be your best option …

But this complete misrepresentation by Phelps – it doesn’t even rise to the level of caricature, since it bears no resemblance to what people like me are saying – is characteristic. For the most part, the opponents of stimulus just don’t listen; they have this image of the idiot Keynesian so fixed in their minds that they can’t be bothered to pay any attention to the actual arguments.

You might have thought that the worst economic crisis since the 30s, a crisis that should not have happened according to non-Keynesian models, would prompt at least a little intellectual curiosity. But no.

So what exactly did Phelps say? Phelps seeks to compare what he calls “two very bad kinds of theories” – Keynesian versus Neoclassical equilibrium theory – which he says “are driving out good theories” in the current policy debate. I won’t deal with the neo-classical approach – which in modern days remains the mainstream (neo-liberal approach).

Phelps says the fallacy of the “Keynesians”:

… is their premise that all slumps, all of the time, are entirely the result of “co-ordination problems” – mis-expectations causing a deficiency of demand. Having modelled the effects of expectations decades ago, I know they have consequences. I agree that companies appeared to underestimate the cutbacks and price cuts of competitors on the way down. That excessive optimism signalled deficient demand for goods and labour. So any stimulus then may have had a Keynesian effect. By now, though, such optimism has surely been wrung out of the system. To pump up consumer or government demand would force interest rates up and asset prices down, possibly by enough to destroy more jobs than are created.

So you see straight away that Phelps does not comprehend how the modern monetary system operates. He uses the type of macroeconomic reasoning that I will describe below as being “hydraulic”.

Why do interest rates rise when government demand rises? Well in the hard-core neo-classical models it is because there is a finite supply of savings (loanable funds doctrine) that are rationed via interest rates changes and if one cohort is borrowing to spend then that has to come at the expense of another borrower. Underpinning this nonsense is the view that the government is revenue constrained.

The alternative mainstream view that emerged out of Keynes thinks that money demand is motivated by transactions and the more transactions the more demand for money you would have. They also claim that the central bank controls the money supply (they believe in the money multiplier) and so if governments spend more, money demand rises, and for a given money supply there is an excess money demand. The only way this can be rationed is for interest rates to rise (and bond prices to fall).

But neither views are remotely accurate depictions of the way the fiat monetary system operates. Government spending creates new net financial assets (bank reserves) and increases income – which increases savings. If the government borrows to match its net spending it is just offering the holders of bank reserves (that the deficits created) an interest-bearing asset. But it is just borrowing the funds it created back.

Further, the central bank does not control the money supply and monetary growth is endogenous and reflective of the demand for loans by consumers and firms. The central bank sets the interest rate and supply reserves to the banking system as required. Loans are never deposit-constrained – in fact, loans create deposits.

So Phelps’s representation of what is likely to happen when governments spend more is sadly astray. But then what would you expect from a Nobel prize winner in economics.

Phelps thinks this is a more accurate depiction of the current economic problems:

As I see it, the poor state of balance sheets in households, banks and many companies augurs a “structural slump” of long duration. Employment will recover, quickly or slowly, only as far as investment demand will carry it. It is highly uncertain whether government spending on infrastructure would help, after taking into account the employment effects of the higher tax rates to pay for it.

So once again he fails to understand the operational characteristics of the monetary system that he is making so-called expert commentary about.

It is clear that if private investment growth is stronger then output and employment will grow. That is a basic premise of all “Keynesian-flavoured” understandings and also of modern monetary theory (MMT). As I explain below, we should always clearly distinguish MMT from the ides of Keynes and Keynesians.

Here we are also using investment in the way economists use it (not financial planners etc) – the building of productive capacity.

But his lack of understanding (he is stuck in gold standard logic) is betrayed in the last sentence, especially “higher tax rates to pay for it”.

Tax rates may rise in the future if the government wants the private sector to have less capacity to spend. This blog covers that in detail – Functional finance and modern monetary theory.

But rising tax rates will have nothing to do with “financing” the spending. Taxpayers fund nothing!

Further, with higher levels of economic activity, the automatic stabilisers (increased tax revenue (not necessarily rates)) will reduce the budget deficit anyway.

On the tax front, I read an interesting interview with David Card, who along with his co-author Alan Krueger published a famous study which showed that increasing the minimum wage somewhat has a negligible impact on employment levels.

Card said in reply to a question about the conventional mainstream assumption that higher taxes erode work incentives (reduce employment via a supply-side withdrawal) that:

… one would normally assume that higher taxes [mean] lower wages and lead to a bit more work. That would have been my starting presumption, to tell you the truth: that the long-run labor supply elasticity is pretty small, and probably negative.

As an aside, the interview also gives you some idea of how pernicious the conduct of the mainstream economics profession is. In the broad ranging interview he was asked about his path-breaking minimum wage work and among other things he said:

I think economists who objected to our work were upset by the thought that we were giving free rein to people who wanted to set wages everywhere at any possible level. And that wasn’t at all the spirit of what we actually said. In fact, nowhere in the book or in other writing did I ever propose raising the minimum wage. I try to stay out of political arguments.

I think many people are concerned that much of the research they see is biased and has a specific agenda in mind. Some of that concern arises because of the open-ended nature of economic research. To get results, people often have to make assumptions or tweak the data a little bit here or there, and if somebody has an agenda, they can inevitably push the results in one direction or another. Given that, I think that people have a legitimate concern about researchers who are essentially conducting advocacy work. I try to stay away from advocacy of any kind, but that doesn’t prevent people from being suspicious that I have an agenda of some kind.

I’ve subsequently stayed away from the minimum wage literature for a number of reasons. First, it cost me a lot of friends. People that I had known for many years, for instance, some of the ones I met at my first job at the University of Chicago, became very angry or disappointed. They thought that in publishing our work we were being traitors to the cause of economics as a whole.

So what is Keynesian?

The approach to policy called “Keynesian” is hard to pin down – although Krugman attempts to argue that “The position held by Keynesians … is that fiscal stimulus is necessary only under certain special conditions. Namely, when you’re up against the zero lower bound, and conventional monetary policy is useless, fiscal stimulus may be your best option”.

But that leaves the field of contenders fairly wide and suggests that monetary policy is the primary aggregate policy weapon of choice in all situations other than when a liquidity trap is encountered. That proposition is not consensual.

The 1983 book by the (sadly) late Alan Coddington Keynesian Economics: The Search for First Principles published by George Allen and Unwin, London is an excellent introduction to the different views that parade under the banner of Keynesian.

The major unifying theme stems from Keynes’ challenge of the neo-classical thought that dominated until the Great Depression. Keynes clearly considered that government intervention was necessary to ensure that under-full employment stalemates didn’t arise – due to lack of aggregate demand. The idea that you could have a steady-state situation with mass unemployment was anathema to neo-classical thought, which paraded models that essentially always assumed full employment.

Any unemployment was voluntary and an expression of a preference for leisure in the neo-classical scheme. Coddington characterised this line of thinking as “reductionist” – the body of thought being constructed from a priori theorising about individual maximisation and choice.

The consideration that what applies to an individual will apply to all individuals left the mainstream approach of the day open to the criticism that its failed to understand the importance of the fallacy of composition – you can read more about this issue via these blogs.

The point was that neo-classical thinking had no macroeconomics – that is, the ability to understand the relations between the aggregates. That understanding had to encompass the consideration that demand and supply curves were interdependent at the aggregate level (for example, a wage cut might reduce supply costs but it also reduces incomes and thus demand).

It further had to understand that an individual act will not be of sufficient gravitas to disturb the aggregate and even a within-firm change (such as a wage cut). But when you translate that action to the aggregate you get different outcomes – typically the opposite to what you would predict from an analysis of the micro (individual) world.

For example, a wage cut at the firm level might induce the firm to hire more workers because the lost income will not impact significantly on the firm’s revenue but if you cut all wages, then employment will likely fall because the income effects are so large.

Coddington then divides the development of “Keynesian” thinking into three different camps: (a) the fundamentalists; (b) the hydraulics; and (c) the reconstituted reductionists.

Briefly, the fundamentalists consider Keynes to be a complete break from neo-classical thinking. They do not believe that individual’s face endemic uncertainty and cannot make decisions with “stable and clearly specified objectives” but are rather influenced by herd-like behaviour (he uses an example of riot to show how moods can be group-derived). In the neo-classical world, there is no interdependence between “preference functions” – that is, each person is in it for themselves and is not influenced by what others are doing or thinking.

Further, the fundamentalists point to Keynes’ rejection of loanable funds theory and the substitution of liquidity preference as the determinant of the rate of interest (see these blogs for more on this).

Finally, the rejection of Say’s and Walras’ Law (that price and interest rate adjustments would guarantee continual full employment) was a feature of Keynes’s departure from the past thinking. In the same way that Marx had talked about effective demand, Keynes considered (even though he didn’t cite Marx as an inspiration) that aggregate demand determined employment and output levels.

This was in stark contrast to the neo-classical thinking which construed that all real variables (employment and output) were predetermined by technology (via marginal productivity) and the real wage rate in the labour market (which was the outcome of labour demand and supply). So the full employment level of output was already determined and prices just adjusted in the goods market to ensure it all sold. Compositional issues between consumption and investment demand and supply were quickly resolved via interest rate changes (a la the loanable funds doctrine).

The Great Depression showed empirically that the neo-classical theory was (and is) a total nonsense while the academic output of Keynes (and others) showed why it was nonsense.

Secondly, the hydraulics emerged in the Post World War II period as policy makers tried to come to terms with the messages of Keynes and text-book writers started to encapsulate it (simplify) it for students. The emergence of this line of thinking really started in the General Theory itself when in Chapter 2, Keynes left the door open for the neo-classical school to reassert itself.

I refer here to his use of marginal productivity theory of labour demand which was straight mainstream. Soon after the General Theory was published the mainstream set about tinkering with the body of idea and what emerged has been also called the Neo-classical synthesis. The so-called IS-LM model which simplified the ideas to the point of futility (it was a static model with no role for expectations etc).

But the hydraulics dominated the debate in the 1950s and 1960s and while retaining a role for discretionary government counter-stabilisation policies introduced all the nonsense about crowding out and debt crises. The static framework did away with the ideas that behavioural relationships (consumption, investment, demand for money etc) were subject to endemic uncertainty and instead assumed they were stable.

The “Keynesian” model was then mathematically expressed (all functions assumed to be stable) and policy analysis was then possible. Increase policy lever X and Y happens. A hydraulic relationship between spending, income and output (and employment) – push spending and the rest change in predictable ways.

Finally, the reconstituted reductionists emerged in the 1960s (mostly) and Robert Clower and Axel Leijonhufvud are mostly associated with this descriptor. This was a debate about what Keynes actually meant.

In the 1950s, particularly, there was a major debate proceeding concerning the microfoundations of macroeconomics. As noted above, the neoclassical paradigm hardly considered macroeconomics to be a separate conceptual structure to microeconomics. The emergence of Keynesian economics in the 1930s had coincided with the emergence of macroeconomics as a separate line of enquiry.

The break with neoclassical thinking came with the failure of markets to resolve the persistently high unemployment during the 1930s. The debate in the ensuing years were largely about the existence of involuntary unemployment. The 1930s experience suggested that Say’s Law, which was the macroeconomic component and closure of the neoclassical system based on the optimising behaviour of individuals, did not hold.

The neoclassical economists continued to assert that unemployment was voluntary and optimal but that some factors not previously included in the model prevented Say’s Law from working. Keynes, following Marx and Kalecki, adopted the distinctly anti-orthodox approach and refuted the basis of Say’s Law entirely.

The theoretical push to reassert Say’s Law by neoclassical economists was severely dented by the work of Robert Clower (1965) and Axel Leijonhufvud (1968). They demonstrated, in different ways, how neoclassical models of optimising behaviour were flawed when applied to macroeconomic issues like mass unemployment.

Clower (1965) showed that an excess supply in the labour market (unemployment) was not usually accompanied by an excess demand elsewhere in the economy, especially in the product market. Excess demands are expressed in money terms. How could an unemployed worker (who had notional or latent product demands) signal to an employer (a seller in the product market) their demand intentions?

Leijonhufvud (1968) added the idea that in disequlibrium price adjustment is sluggish relative to quantity adjustment. Leijonhufvud interpreted Keynes’s concept of equilibrium as being actually better considered to be a persistent disequilibrium. Accordingly, involuntary unemployment arises because the labour market is not in equilibrium and there is no way that the unemployed workers can signal that they would buy more goods and services if they were employed.

Any particular firm cannot assume their revenue will rise if they put a worker on even though revenue in general will clearly rise (because there will be higher incomes and higher demand). The market signalling process thus breaks down and the economy stagnates.

So you can see that it is not clear what we mean when we talk of a “Keynesian” position.

Where does Phelps fit in to this?

Prior to Phelps famous 1967 article on the Phillips curve, the Phillips orthodoxy was in the tradition of Keynes and saw price adjustment as a response to disequilibrium arising from the labour market. Unemployment in this type of model was usually considered to be involuntary.

Phillips himself (who the famous curve is named after – and not without some controversy) was articulating a process where disequilibrium in the real sector caused changes in nominal aggregates.

The neo-classical writer, Irving Fisher, who inspired Friedman and Phelps, wrote in the 1920s that price changes and employment (real activity) levels was cast in terms of the reverse causality and considered to be an equilibrium relation.

During the Phillips curve orthodoxy period (1950s to late 1960s) policy makers considered there was a trade-off between inflation and unemployment. So the “cost” of lower unemployment was some inflation. The aim was to get the lowest rates of each consistent with the expectations of voters. In general, nations enjoyed full or close to full employment during this period.

The emergence of Milton Friedman’ famous 1968 article and Phelps’s two articles in 1967 and 1968, respectively was really an expression of this neoclassical discontent with the lack of optimising microfoundations in Keynesian macroeconomics.

They reasserted neoclassical microfoundations and were then left to explain why Say’s Law did not work all the time. To overcome that problem they followed Irving Fisher and identified misperceptions of inflation as the factor that prevented Say’s Law from working according to the market-clearing model.

In other words, when prices rose faster than money wages (hence real wages fell), firms would employ more workers. But why would workers supply more labour at lower real wages (which was contrary to the mainstream model)? Because they thought the money wages were rising and hadn’t observed the inflation. As soon as they found at the truth, so Phelps argued, they would withdraw their labour supply and employment would drop again.

So the business cycle was characterised as swings in labour supply which was at total odds with the empirical evidence – but these guys never let the facts get in the road of their highly stylised mathematical models of outer space.

The concept of full employment – where there had to be enough jobs to match the preferences of the labour force lost meaning with development of the so-called expectations-augmented Phillips curve of Friedman and Phelps.

This model spearheaded the resurgence of pre-Keynesian macroeconomic thinking in the form of Monetarism. The embedded Natural Rate Hypothesis (NRH) outlined a natural rate of unemployment (NRU), where the inflation-unemployment tradeoff was allegedly a trade-off between unemployment and unexpected inflation. The NRU is now popularly called the NAIRU (although the two concepts are slightly different).

As workers gained more information the trade-off vanishes. At this point there is only one unemployment rate consistent with stable inflation – the NRU. So in the so-called long-run there was no stable trade-off between inflation and unemployment.

These developments represented a major theoretical break from the previous versions of the Phillips curve. The pre-Monetarist Phillips curve models were based on a disequilibrium notion of the relationship between inflation and unemployment in that they modelled the adjustment of prices and wages to some labour market imbalance between supply and demand.

The causality was strictly from the labour market disequilibrium to the price adjustment function. There was no presumption that full employment is inevitable or a tendency of a capitalist monetary economy.

The Friedman-Phelps story and the later developments under the rubric of rational expectations and the New Classical School are, instead, based on a market clearing relation and the causality is reversed.

Unemployment is considered to be voluntary and the outcome of optimising choices by individuals between work (bad) and leisure (good). In the natural rate world of Friedman and Phelps, the Central Bank can promote variations in the unemployment rate by introducing unforeseen changes in inflation, a temporary capacity allowed due to expectational inertia on behalf of the workers.

There is no theory in the natural rate hypothesis that changes in the unemployment rate cause changes in inflation.

Full employment is assumed to prevail (with unemployment at the natural rate) unless there are errors in interpreting price signals. The tendency is always to restore full employment by market mechanisms. There is no discretionary role for aggregate demand management.

The rise in acceptance of Monetarism and its new classical counterpart was not based on an empirical rejection of the Keynesian orthodoxy. It was just an argument based on highly abstract a priori theorising and reasserted the conservative ideology at the expense of liberal thinking.

It was not, in a word, a Kuhnian scientific revolution. However, the shift in the Phillips curve in the 1970s as the OECD economies began to fail was a strong empirical endorsement for the NRU theory, despite the fact that the instability came from the cost side (OPEC oil price hikes) rather than excessive spending.

Any Keynesian remedies proposed to reduce unemployment were met with derision from the bulk of the profession who had embraced the NRH and its policy implications.

The NRH was now the characterisation of full employment and it was asserted that the economy would always tend back to a given NRU, no matter what had happened to the economy over the course of time. Time and the path the economy traced through time were thus irrelevant. Only microeconomic changes would cause the NRU to change.

So Phelps’s work in developing the natural rate hypothesis allowed the conservative profession to reassert Say’s Law again and undermine active use of fiscal policy. It set the agenda that has been developed during the neo-libearl years.

Accordingly, the policy debate became increasingly concentrated on deregulation, privatisation, and reductions in the provisions of the Welfare State.

Also we haven’t returned to the full employment days since then and in my view Phelps has a lot of blood on his hands in this regard.

You can read a lot more about Phelps and this debate in my recent book with Joan Muysken – Full employment abandoned.

Keynes or Keynesianism is not MMT

While it is true that many proponents of MMT (not me) are very partial to the writings of Keynes, the main body of MMT is not derived from Keynes nor particularly influenced by his work (in my view).

To complicate matters further the term Post Keynesian is used – largely to distinguish modern progressive (non-Marxist – although that is even contentious) thinking from the Coddington taxonomy.

However, some Post Keynesians are fundamentalist Keynesians (in the Coddington sense) – for example, Paul Davidson.

There is even disputes about how Post Keynesian should be spelt – lower case P, lower case K, hyphenated etc. Ridiculous nonsense. The only sensible thing Henry Kissinger probably said was “University politics are vicious precisely because the stakes are so small”.

Anyway, all Post Keynesians agree that the orthodox unemployment buffer stock approach (NAIRU) to inflation control is costly and unacceptable. The neo-liberal solution to the resulting unemployment is to pursue supply-side policies (labour market deregulation, welfare state retrenchment, privatisation, and public-private partnerships) to give the economy “room” to expand without cost pressures emerging.

Post Keynesians, in general reject this strategy because the sacrifice ratios are high and the distributional implications (creation of under class and working poor and loss of essential services) are unsavoury.

However there is no alternative consensus. Some Post Keynesians, following closely the policy recommendations of Keynes himself advocate what I have termed “generalised expansion” in my academic writing – this is where the government ensures that spending is sufficient to purchase all available output.

In essence, this policy purchases at market prices or provides incentives to profit-seekers to create private employment expansion. Typically, public and private capital formation is targeted. This strategy ignores the role for a buffer employment stock policy, which allows the government to guarantee full employment using automatic stabilisers by purchasing at fixed prices.

Typically, Post Keynesians advocate generalised fiscal and monetary expansion mediated by incomes policy and controlled investment as a solution to unemployment.

However (indiscriminate) expansion in isolation is unlikely to lead to employment opportunities for the most disadvantaged members of society and does not incorporate an explicit counter-inflation mechanism. It also fails to address the spatial labour market disparities.

How can we be sure that the investment will provide jobs in failing regions? Upon what basis are the most disadvantaged workers with skills that are unlikely to match those required by new technologies going to be included in the “generalised expansion”?

Where is the inflation anchor in this approach – if spending is manifest in purchases at market prices.

An understanding of MMT, allows you to conclude that the State can resolve demand gaps which cause unemployment in two ways: (a) by increasing net spending via purchasing goods and services and/or labour at market prices as explained in the previous sub-section; and/or (b) by using its currency issuance power to provide a fixed-wage job to all those who are unable to find a job elsewhere.

The employment buffer stock approach is what I call the Job Guarantee. The JG is an effective strategy for a fiat-currency issuing government to pursue to ensure that work is available at a liveable wage to all who wish to work but who cannot find market sector employment (including regular public sector).

The government would provide a buffer stock of jobs that are available upon demand. The JG differs from a Keynesian expansion because it represents the minimum stimulus (the cost of hiring unemployed workers) required to achieve full employment rather than relying on market spending and multipliers.

The JG also provides an inherent inflation anchor missing in the generalised Keynesian approach.

Clearly, and emphatically, a mixture of (a) and (b) is likely to be optimal although (a) alone is not preferred.

The JG is juxtaposed with what the NAIRU approach which accompanied a regime shift in macroeconomic policy in the 1970s. The NAIRU approach is exemplified by tight monetary policy that targets low inflation, using unemployment as a policy tool rather than a target.

The countries that avoided high unemployment in the 1970s maintained an employer of last resort capacity which allowed them to absorbs the fluctuations in demand without significant unemployment consequences.

There is some reason why countries like Japan which has been through hell over the last 20 years or so still has relatively low unemployment.

While there may be a need for more generalised expansion – to boost public infrastructure investment which enhances the profitability of private sector investment and contributes to aggregate demand and employment – this approach is limited by real resource availability.

But whether there is a capacity to pursue a more general approach is moot. For MMT, it is independent of the need for a JG.

But as a policy preference I would ensure that the JG was accompanied by social wage spending to increase employment in education, health care and the like.

But, sole reliance on public sector investment to achieve full employment, would create considerable economic inflexibility. The ebb and flow of the private sector would not be readily accommodated and an increasing likelihood of inflation would result.

Further and crucially, public investment is unlikely to benefit the most disadvantaged workers in the economy. The JG is designed to explicitly provide opportunities for them. By way of example, during the golden age in Australia (1945-1975) when public capital formation and social wage expenditure was strong, full employment was only achieved because the public sector (implicitly) provided a JG for low skilled workers. This experience is shared across all advanced economies.

The JG is thus designed to ensure that the lowest skilled and experienced workers are able to find employment. The JG is a full employment policy and should be judged on those terms. It does not presume that JG jobs will suit all skills. For some skilled workers who become unemployed in a downturn the income loss implied would be significant.

The contention is that a fully employed economy with the JG workers paid minimum wages is a significant improvement, when compared to the current unemployment.

Further, environmental constraints militate against generalised Keynesian expansion. Higher output levels are required to increase employment, but the composition of output remains a pivotal policy issue. JG jobs would be designed to support local community development and advance environmental sustainability.

JG workers could participate in many community-based, socially beneficial activities that have intergenerational payoffs, including urban renewal projects, community and personal care, and environmental schemes such as reforestation, sand dune stabilisation, and river valley and erosion control. Most of this labour intensive work requires very little capital equipment and training.

In our past work we have denoted this form of spatially targeted employment policy as Spatial Keynesianism, in contrast to the bluntness of “Keynesian” tools which fail to account for the spatial distribution of social disadvantage.

How is that for a plethora of uses of the term “Keynesian”.

That is enough for today.