The divide between mainstream macro and MMT is irreconcilable – Part 3

This is Part 3 (and final) of my series responding to an iNET claim that Modern Monetary Theory (MMT) and mainstream macroeconomics were essentially at one in the way they understand the economy but differ on matters of which policy instrument (fiscal or monetary) to assign to counter stabilisation duties. In Part 1, I demonstrated how the core mainstream macroeconomic concepts bear no correspondence with the core MMT concepts, so it was surprising that someone would try to run an argument that the practical differences were really about policy assignment. In Part 2, we saw how the iNET authors created a stylised version of mainstream macroeconomics that ignored the fundamental building blocks (how they reach their conclusions about the real world), which means that they ignore important differences in the way MMT economists and mainstream macroeconomists interpret a given economic state. I will elaborate on that in this final part. Further, by reducing the body of work now known as MMT to be just ‘functional finance’, the iNET authors also, effectively, abandon any valid comparison between MMT and the mainstream, although they do not acknowledge that sleight of hand.

The series so far is:

1. The divide between mainstream macro and MMT is irreconcilable – Part 1

2. The divide between mainstream macro and MMT is irreconcilable – Part 2

MMT is not functional finance

The iNET authors write:

Our goal here is not to make an assessment of MMT as a whole, but to ask what is the relationship between the functional finance approach to government budgets and conventional economic analysis. Because we are interested in the logic of the functional-finance position rather than in MMT as a body of thought, we make only limited references to MMT literature here.

In other words, despite the title of their article, they are not really comparing MMT with mainstream macroeconomics at all.

While Modern Monetary Theory (MMT) proponents do make “the case for functional finance”, following Abba Lerner’s work in the 1940s and beyond, MMT cannot be distilled as being solely about functional finance.

Abba Lerner’s aim was to distinguish between what he called “sound finance” and “functional finance”.

I deal with that distinction in detail in this blog post (among others) – Functional finance and modern monetary theory (November 1, 2009).

‘Sound finance’ is a core component of mainstream macroeconomics. It is about fiscal rules – balance the budget over the course of the business cycle; only increase the money supply in line with the real rate of output growth; etc.

Lerner thought that these rules were based more in conservative morality than being well founded ways to achieve the goals of economic behaviour – full employment and price stability.

Lerner wanted to construct macroeconomic policy making in terms of the government using its “steering wheel” (policy tools) to stabilise fluctuations in the economy.

Fiscal policy was the steering wheel and should be applied for functional purposes – to achieve intended aims rather than become a self-referential exercise (controlling debt ratios etc).

He juxtaposed that conception with the laissez-faire (free market) approach which he considered was akin to letting a car zigzag all over the road.

He believed that if you wanted the economy to develop in a stable way you had to control its movement.

So his classic statement of functional finance was:

Government should adjust its rates of expenditure and taxation such that total spending in the economy is neither more nor less than that which is sufficient to purchase the full employment level of output at current prices. If this means there is a deficit, greater borrowing, ‘printing money,’ etc., then these things in themselves are neither good nor bad, they are simply the means to the desired ends of full employment and price stability …

MMT clearly draws on that statement although the language we use is different (we never talk about “printing money” as a short-hand for central bank monetisation) and we do not share Lerner’s view that taxation revenue and debt-issuance provides ‘funds’ to facilitate government spending.

But we do consider fiscal policy should be directed to advancing public purpose and the particular levels of resulting aggregates (for example, fiscal deficits/surpluses) are immaterial.

But in reaching that conclusion the core MMT team has drawn on many other historical sources and added some new ideas that put the support for Lerner’s functional finance vision into context.

One has to consider all the MMT components to understand why we think the fiscal aggregates, in themselves, are immaterial, contrary to the dominant role these aggregates play in mainstream macroeconomic narratives.

One has to understand the radical departure from the mainstream macroeconomics presentation of banking that MMT provides to understand why we think that there can be no financial crowding out in a modern monetary economy and why we were unconcerned about any inflationary consequences of the massive expansion of central bank balance sheets.

Etc.

To simply distill MMT down to being an application of functional finance is to ignore all the other components that give functional finance within MMT context.

Why ignore the mainstream core if you want to consider mainstream policy making?

The mainstream macroeconomics profession relies heavily on so-called Dynamic Stochastic General Equilibrium (DSGE) models.

Central banks and other forecasting agencies deploy to make statements about the effectiveness of fiscal and monetary policy.

I considered this type of model in several blog posts, including:

1. Mainstream macroeconomic fads – just a waste of time (September 18, 2009).

2. The myth of rational expectations (July 21, 2010).

3. Fiscal austerity damages real growth and prolongs the financial downturn (June 21, 2012).

4. Mainstream macroeconomics in a state of ‘intellectual regress’ (January 3, 2017).

5. Austerity is the problem for Britain not Brexit (January 9, 2017).

This is not meagre academic to-and-fro. The use of these models is prominent in current debates about Brexit.

In their quest to eliminate obvious differences between MMT and mainstream macroeconomics, the iNET authors write:

… we are focused on what might be called “orthodox policy macroeconomics” – the practical heuristics that guide policy makers and are reflected in undergraduate textbooks, as opposed to DSGE and related models of intertemporal optimization that are the basis of most current macroeconomic theory.

I found that extraordinary and dealt with it in Part 2 at some length.

But there was a matter that I didn’t get to yesterday.

If we consider the implications of their characterisation of the mainstream macroeconomics one would be led to assume that there is no correspondence between the core assumptions that characterise the mainstream approach (DSGE, intertemporal optimisation, etc) and the way policy is conceived, designed and executed.

That would be a very ignorant assumption.

Take some obvious examples.

A major player in the Brexit debate in Britain has been the National Institute of Economic and Social Research (NIESR), which has been pumping out ridiculous forecasts of doom should Britain leave the EU.

This organisation also advises politicians and has been influential in the design of British Labour’s fiscal rules.

On May 10, 2016, the National Institute of Economic and Social Research published its – Commentary – The economic consequences of leaving the EU – which said that:

… there is a degree … of consensus that leaving the EU would depress UK economic activity in both the short term (via uncertainty) and the long term (via trade).

This study was enthusiastically promoted two days later (May 12, 2016) by well-known British academic Simon Wren-Lewis who used it as authority to state “doesn’t everyone already know that nearly all economists think Brexit would have significant costs” (Source).

The NIESR uses a New Keynesian style, Dynamic Stochastic General Equilibrium (DSGE) model as a core approach to its modelling.

Its main model NIGEM has in their own words “many of the characteristics of a … DSGE model” including the use of rational expectations, which means their long-run solution is supply-determined.

But you need to understand the ‘supply-determined’ bit means that there is a dichotomy between the real and nominal worlds or in simpler language – money doesn’t matter in a long-run general equilibrium.

Wren-Lewis is also a defender of DSGE modelling as a tool for policy formulation.

In this blog post – Britain doesn’t appear to be collapsing as a result of Brexit (December 13, 2017) – I analysed the British Treasury report on the “immediate economic impact of leaving the EU”.

The Treasury produced rather stark forecasts in May 2016 (just before the Referendum), where they claimed that within two years of June 2016 the UK would have a GDP between 3.6 and 6 percent lower and the number of people unemployed would rise by as much as 820,000.

The findings were based on mainstream ‘economic’ and ‘econometric’ models’ of the type that the iNET authors choose to ignore in their comparison.

The models used by the British government are called computable general equilibrium (CGE) models, which are informed by gravity models, and – much like their better known cousins, dynamic stochastic general equilibrium (DSGE) models, and other macroeconomic forecasting models.

The models are notoriously unreliable and easily manipulated to achieve whatever outcome one desires.

These models are uniformly used by mainstream macroeconomists within and outside the academy and have produced ridiculous results over many years.

They failed to predict the GFC, even though it was obvious to Modern Monetary Theory (MMT) economists as early as the late 1990s that a crisis was brewing.

They then failed to predict the depth of the negative impacts of austerity – remember the mantra “growth friendly austerity” or “fiscal contraction expansion”.

The same models or derivations are used by the IMF and were used to design the so-called “growth-friendly austerity” that was part of the Greek bailouts. Disastrous consequences followed for Greece.

And the application of these models have catastrophically failed to estimate the short-run impacts of the Brexit vote.

We considered the Brexit nonsense in our recent Jacobin article (April 29, 2018) – Why the Left Should Embrace Brexit.

See also this blog post (with links within) – Mainstream macroeconomics in a state of ‘intellectual regress’ (January 3, 2017).

So it is quite remarkable that the iNET authors think it is reasonable to abstract from the importance of these models in the policy setting process.

Major differences in conceptions of the short-run

One of the core insights that MMT has inherited (absorbed) from the work of Keynes and Kalecki and others is the concept of the Fallacy of Composition.

I discussed the concept of the fallacy of composition in this blog post (among others) – Fiscal austerity – the newest fallacy of composition (July 6, 2010).

The Fallacy of Composition refers to errors in logic that arise:

… when one infers that something is true of the whole from the fact that it is true of some part of the whole (or even of every proper part).

So the fallacy of composition refers to situations where individually logical actions are collectively irrational.

These fallacies are rife in the way mainstream macroeconomists reason and serve to undermine their policy responses.

A simple Wikipeda example is:

If someone stands up out of their seat at a cricket match, they can see better. Therefore, if everyone stands up, they can all see better.

Now consider what this means.

It means that one cannot construct a macroeconomics from a simple aggregation of individual behaviour. What applies for the individual does not apply at the aggregate.

Before Keynes produced his General Theory in 1936, there was no discipline called ‘macroeconomics’. The mainstream neoclassical school talked about macroeconomics but they did it in the context of adding up individual relationships – demand and supply curves.

The problem that Keynes pointed out was that this aggregation was invalid because of the presence of compositional fallacies.

Now think about DSGE models.

Underlying them is the concept of general equilibrium, a Classical construct that still pervades mainstream thinking.

The modern restatement of the competitive general equilibrium model came from microeconomists Kenneth Arrow and Gérard Debreu in their 1971 book General Competitive Analysis.

In my post graduate studies this book was still one that students had to grind their way through with little salutory benefit.

What Arrow and Debreu thought they had established was that “under certain economic assumptions (convex preferences, perfect competition, and demand independence) there must be a set of prices such that aggregate supplies will equal aggregate demands for every commodity in the economy.”

In plain words, all markets will reach a long-run equilibrium (no excess supply or demand) if prices are flexible and the standard mainstream assumptions about the rationality and optimising behaviour of individuals hold.

So if there is an excess supply in one market (for example, unemployment in the labour market) then there must be an excess demand in another market (say, a product market(s)) and price rises will eliminate the unemployment (by cutting the real wage) and driving down demand for goods and services.

In other words, there is no unemployment in a flexible, competitive market.

Marx was the first to really criticise early versions of this idea (Say’s Law) saying that while the unemployed clearly wanted to consume more they didn’t have the income to allow that.

In more modern parlance, this highlighted the difference between notional excess demand (the unemployed wanting to consume more) and effective excess demand (demand backed by cash offers to purchase).

Accordingly, an excess supply in the labour market (unemployment) is 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?

In other words, even in a competitive market, price adjustment is sluggish relative to quantity adjustment. Accordingly, involuntary unemployment arises because 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.

Why is that important in the context of this discussion?

Mainstream macroeconomics relies on the concept of general equilibrium and the behavioural assumptions that underpin it.

In the early 1970s, there were four influential articles published which combined to demolish once and for all the concept of a unique competitive general equilibrium resulting from rational behaviour.

I am referring to what has become known as the – Sonnenschein Mantel Debreu theorem (SMD) – which demonstrated categorically that the so-called microfoundations (rationality assumptions) that dominate microeconomics have “no equivalent macroeconomic implications”.

What this means (in plain words) is that if an economy has a number of interdependent markets (places where supply and demand meets) and a number of consumers with different preferences for goods, then there is no guarantee that a unique equilibrium can be established where all markets clear at the same time and are consistent with the optimality outcomes that mainstream economics specifies (lowest cost production etc).

What it also meant was that what might hold for an individual’s demand curve (demand is inverse to price) cannot be shown to hold for an aggregate demand curve (the macroeconomic level).

The macroeconomic level demand curve can take any form even if the individual demand curve is typical (demand is inverse to price).

What this means is that there is no reliable way to build a macroeconomic approach from neoclassical micro foundations (rationality, optimisation etc).

The SMD proof was really a reflection of what Keynes and others considered to be the fallacy of composition in mainstream thinking.

The SMD theorem was an application of this concept to the core mainstream economic concept of equilibrium.

The way that mainstream macroeconomists sought to get around this problem was to assume away the fallacy of composition by introducing the so-called Representative agent – whether it be a firm or a consumer.

This means they try to get over the heterogeneity problem (that SMD demonstrated) by assuming that there is just one consumer and one firm who is typical – that is all individuals are the same – they act the same, think the same, make the same decisions – so we can just add them all up into one big macro agent.

So while New Keynesians claim their macroeconomics is ‘microfounded’ what they are really doing is solving an optimisation problem for this representative agent using microeconomic assumptions, which themselves are ridiculous.

The result of this optimisation exercise is, for example, an individual demand curve (quantity varies inversely with price) which they then claim is equivalent to the aggregate demand curve (or supply curve in the case of labour supply or output supply decisions).

So they get in a real state.

SMD showed that individual demand curves do not aggregate into a representative agent curve. So pretending that the macroeconomic functions have some correspondence with the microeconomic optimisation outcomes for an individual is deeply problematic.

As a result, we are back into the realm of fallacy of composition, a point that Alan Kirman demonstrated impeccably in his 102 article – Whom or what does the representative individual represent? (Journal of Economic Perspectives).

Kirman demonstrated that what holds for the ‘representative agent’ may not hold for any individual in the economy. Which means that if we invoke a policy shift that New Keynesians claim is good for the representative agent, all actual individuals in the economy may be damaged.

He concluded that the concept:

… deserves a decent burial, as an approach to economic analysis that is not only primitive, but fundamentally erroneous.

But in simpler terms, the single agents who have rational expectations and optimise their decisions through time (mostly assumed to have infinite lives) are devoid of any correspondence with reality.

Many interesting macroeconomic issues – such as income distribution and its effects on, say, household consumption – which MMT economists consider to be important – are just assumed away by the representative agent approach.

All of this is too complex to discuss in any more detail than that here.

But why is it important?

It is important because the iNET authors choose to ignore these foundational elements of mainstream economics in their zest to make the mainstream economists compatible with MMT.

By ignoring these foundational elements, they cannot tell us anything much, for example, about the way in which a mainstream macroeconomists constructs a short-run reality.

So the iNET article says:

On both sides, in other words, we are interested in the logic of the policy positions themselves.

But you cannot get to that logic unless you know where the policy positions have come from.

Their stylised “key assumptions” that they claim “orthodox policy macroeconomics and functional finance” share are:

1 In the short run, output is determined by the total desired spending of units in the economy (aggregate demand).

2. In the short run, unemployment is a decreasing function of the level of output, and inflation is an increasing function of the level of output.

3. There is a level of output such that the behavior of both inflation and unemployment is acceptable …

4. Economy-wide spending (aggregate demand) depends on, among other things, the interest rate set by the central bank and the budget position of the central government …

5. The evolution of the government debt-GDP ratio over time depends on the current fiscal position (the primary balance), the interest rate on outstanding public debt, and on the nominal growth rate of GDP …

At the level of description this depiction is true.

I have no issues with any of the five points although the language I would use is different.

But my interpretation of the short-run states and their relationship to policy settings is entirely different to the interpretation that a mainstream macroeconomist would have.

For example, a mainstream New Keynesian consider that short-run output is capable of policy manipulation because there are ‘sticky prices’. This is the Keynesian economics as a ‘special case’ of neoclassical theory argument.

Accordingly, they also claim that in the long-run no such manipulation is possible and policy interventions will only generate price effects (inflation).

An MMT economist doesn’t think that at all. For us, the long-run doesn’t exist independently of the short-run. Where you are is conditioned by where you have been.

And fiscal policy interventions are always able to manipulate demand and maintain full employment and that capacity has nothing to do with ‘sticky prices’.

So the way we see the short-run is not reconcilable with mainstream macroeconomics.

Further, while some New Keynesians are comfortable with short-run increases in the fiscal deficit (although usually only if monetary policy is deemed to be ineffective – the zero-bound interest case) to stabilise a recessed economy, they consider that the government must run surpluses in the upturn to ensure the debt ratio is stable or declining.

MMT economists don’t think that at all. Continuous fiscal deficits are typically required to allow the non-government sector to save overall.

Further, mainstream macroeconomists believe that if the central bank stabilises inflation, full employment (zero output gap) will follow as a market process – the restoration of the natural rate of unemployment.

MMT economists don’t think that at all.

There is no natural rate of unemployment that the economy converges to. In 1987, as a young academic, I wrote an article – The NAIRU, Structural Imbalance and the Macroeconomic Equilibrium Unemployment Rate – which was the basis of my PhD work.

It demonstrated that any steady-state relationship between unemployment and inflation was non-unique – that is, any level of unemployment could stabilise inflation by temporarily quelling the distributional conflict between labour and capital.

And that fiscal policy could manipulate that level by reversing structural imbalances that occurred over the cycle.

I had earlier considered the way in which a buffer stock employment approach could eliminate the Phillips curve trade-off between inflation and unemployment altogether.

Those efforts are now embedded in MMT in the form of the Job Guarantee, which proposes a completely different short-run dynamic than you will find in mainstream macroeconomics.

Further, mainstream macroeconomists construct the short-run position as saying that the fiscal deficits that may arise should be temporary because otherwise bond markets will increase their risk assessments and drive up yields on government debt.

MMT economists don’t think that at all.

We have pointed out that it is the government that can control yields whenever they desire and that means that MMT economists have none of the mainstream fears that continuous deficits will result in rising interest rates.

Further, MMT economists advocate that government stop issuing debt altogether. It is unnecessary and does not insulate the economy against the inflation risk that is embodied in spending decisions.

Conclusion

That is all I have time for.

I have made no imputations about the motivations of the iNET authors in writing this article. But it is a common strategy in the academy to head of challenges by claiming that there is nothing new to worry about.

The point is that we can present bland comparisons of different ways of thinking about reality. Or we can understand the differences and appraise them on their merits.

The iNET approach is, in my view, the former. And it doesn’t get us very far as a consequence.

If you are in Melbourne tonight and want to hear some original music …

Then my band – Pressure Drop – is playing at the Maori Chief Hotel, 117 Moray St, South Melbourne, from about 20:00 to late.

This is a great little inner city pub. And, what else is there to do on a Wednesday night in Melbourne anyway?

Lots of great dub, rock steady and reggae coming up tonight.

I can also discuss Modern Monetary Theory (MMT) during breaks in the sets!

That is enough for today!

(c) Copyright 2018 William Mitchell. All Rights Reserved.