macro

Austerity and household debt: a macro link?

For some time now I’ve been arguing that not only does austerity have real effects but also financial implications.

When the government runs a deficit, it produces a flow supply of safe assets: government bonds. If the desired saving of the private sector exceeds the level of capital investment, it will absorb these assets without government spending inducing inflationary tendencies.

This was the situation in the aftermath of the 2008 crisis. Attempted deleveraging led to increased household saving, reduced spending and lower aggregate demand. Had the government not run a deficit of the size it did, the recession would have been more severe and prolonged.

When the coalition came to power in 2010 and austerity was introduced, the flow supply of safe assets began to contract. What happens if those who want to accumulate financial assets — wealthy households for the most part — are not willing to reduce their saving rate? If there is an unchanged flow demand for financial assets at the same time as the government reduces the supply, what is the result?

Broadly speaking there are two possible outcomes: one is lower demand and output: a recession. If growth is to be maintained, the only option is that some other group must issue a growing volume of financial liabilities, to offset the reduction in supply by the government.

In the UK, since 2010, this group has been households — mostly households on lower incomes. As the government cut spending, incomes fell and public services were rolled back. Unsurprisingly, many households fell back on borrowing to make ends meet.

The graph below shows the relationship between the government deficit and the annual increase in gross household debt (both series are four quarter rolling sums deflated to 2015 prices).

hh2

From 2010 onwards, steady reduction in the government deficit was accompanied by a steady increase in the rate of accumulation of household debt. The ratio is surprisingly steady: every £2bn of deficit reduction has been accompanied by an additional £1bn per annum increase in the accumulation of household debt.

Note that this is the rate at which gross household debt is accumulated — not the “net financial balance” of the household sector. The latter is highlighted in discussions of “sectoral balances”, and in particular the accounting requirement that a reduction in the government deficit be accompanied by either an increase in the deficit of the private sector or a reduction in the deficit with the foreign sector.

Critics of the sectoral balances argument make the point that the net financial balance of the household sector is not the relevant indicator. Most household borrowing takes place within the household sector, mediated by the financial system. Savers hold bank deposits and pension fund claims, while other households borrow from the banks. The gross indebtedness of the household sector can therefore either increase or decrease without any change in the net position. Critics therefore see the sectoral balances argument argue as incoherent because it displays a failure to understand basic national accounting. This view has been articulated by Chris Giles and Andrew Lilico, among others.

For the UK, at least, this criticism appears misplaced. The chart below plots four measures of the household sector financial position along with the government deficit. The indicators for the household sector are the net financial balance, gross household debt as a share of both GDP and household disposable income, and the household saving ratio. The correlation between the series is evident.

hh3

The relationship between the government deficit and the change in gross household debt is surprisingly stable. The figure below plots the series for the full period for which data are available from the ONS: from 1987 until 2017. With the exception of the period 2001-2008, where there is a clear structural break, the relationship is persistent.

hh1

Why should this be the case? One needs to be careful with apparently stable relationships between macroeconomic variables — they have a habit of breaking down. One reason for caution is that the composition of household debt has changed over the period shown: in the pre-2008 period most of the increase was mortgage borrowing, while post-crisis, consumer debt in the form of credit cards, car loans and so on has played an increasing role. Nonetheless, a hypothesis can be advanced:

If one group of households saves a relatively constant share of income — and this represents the majority of total saving in the household sector — then variance in the supply of assets issued by public sector must be matched either by variations in output and employment or by variance in the issuance of financial liabilities by other sectors. If monetary policy is used to maintain steady inflation and therefore relatively stable output and employment, changes in the cost of borrowing may induce other (non-saver) households to adjust their consumption decisions in such a way that stabilises output.

Put another way, if the contribution of government deficit spending to total demand varies and saving among some households is relatively inelastic, avoiding recessions requires another sector (or sub-sector) to go into deficit in order that total demand be maintained.

This hypothesis fits with the observation that the household saving ratio falls as the rate of gross debt accumulation increases. Paradoxically, the problem is not too little household saving but too much, given the volume of investment. If inelastic savers were willing to reduce their saving and increase consumption in response to lower government spending, then recession could be avoided without an increase in household debt. A better solution would be an increase in the business investment of the private sector: it is the difference between saving and investment that matters.

There is a clear structural break in the relationship between the deficit and household debt, starting around 2001. This is likely the result of the global credit boom which gathered pace after Alan Greenspan cut the target federal funds rate from 6.5% in 1999 to 1% in 2001. During this period, the financial position of the corporate sector shifted from deficit to surplus, matched by large rises in the accumulation of household debt. With the outbreak of crisis in 2008, the previous relationship appears to re-emerge.

Careful econometrics work is required to try and disentangle the drivers of rising household debt. But relationships between macroeconomic variables with this degree of stability are unusual. Something interesting is going on here.

EDIT: 22 November

Toby Nangle left a comment suggesting that it would be good to show the data on borrowing by different income levels. It’s a good point, and raises a complex issue about the distribution of lending and borrowing within the household sector. This is something that J. W. Mason and others have been discussing. I need another post to fully explain my thinking on this, but for now, I’ll include the following graph:

hh4

This is calculated using an experimental new dataset compiled by the ONS which uses micro data source to try and produce disaggregated macro datasets. Data are currently only available for three years — 2008, 2012, and 2013 — but I understand that the ONS are working on a more complete dataset.

What this shows is that in 2008, at the end of the 2000s credit boom, only the top two income quintiles were saving: the bottom 60% of the population was dissaving. In 2012 and 2013, the household saving ratio and financial balance had increased substantially and this shows up in the disaggregated figures as positive saving for all but the bottom quintile.

I suspect that as the saving ratio and net financial balance have subsequently declined, and gross debt has increased, the distributional pattern is reverting to what it looked like in 2008: saving at the top of the income distribution and dissaving in the lower quintiles.

Advertisements

Dilettantes Shouldn’t Get Excited

A new paper on DSGE modelling has caused a bit of a stir. It’s not so much the content of the paper — a thorough but unremarkable survey of the DSGE literature and a response to recent criticism — as the tone that has caught attention. The paper begins:

“People who don’t like dynamic stochastic general equilibrium (DSGE) models are dilettantes. By this we mean they aren’t serious about policy analysis… Dilettantes who only point to the existence of competing forces at work – and informally judge their relative importance via implicit thought experiments – can never give serious policy advice.”

The authors, Lawrence Christiano, Martin Eichenbaum and Mathias Trabandt, make a number of claims, most eye-catchingly: “the only place that we can do experiments is in dynamic stochastic general equilibrium (DSGE) models.” They then list a number of policy questions that are probably best answered using a combination of time series econometrics and careful thinking. After their survey of the literature, the authors conclude — without recourse to evidence — “… DSGE models will remain central to how macroeconomists think about aggregate phenomena and policy. There is simply no credible alternative to policy analysis in a world of competing economic forces.”

The authors seem to have been exercised in particular by recent comments from Joseph Stiglitz, who wrote:

“I believe that most of the core constituents of the DSGE model are flawed—sufficiently badly flawed that they do not provide even a good starting point for constructing a good macroeconomic model. These include (a) the theory of consumption; (b) the theory of expectations—rational expectations and common knowledge; (c) the theory of investment; (d) the use of the representative agent model (and the simple extensions to incorporate heterogeneity that so far have found favor in the literature): distribution matters;(e) the theory of financial markets and money; (f) aggregation—excessive aggregation hides much that is of first order macroeconomic significance; (g) shocks—the sources of perturbation to the economy and (h) the theory of adjustment to shocks—including hypotheses about the speed of and mechanism for adjustment to equilibrium or about out of equilibrium behavior.”

Stiglitz is not the only dilettante in town. He’s not even the only Nobel prize-winning dilettante — Robert Solow has been making these points for decades now. The Nobels are not alone. Brad Delong takes a similar view, writing that “DSGE macro has … proven a degenerating research program and a catastrophic failure: thirty years of work have produced no tools for useful forecasting or policy analysis”. (You should also read his response to the new paper, and some of the comments on his blog).

Back in 2010, John Mulbaer wrote that “While DSGE models are useful research tools for developing analytical insights, the highly simplified assumptions needed to obtain tractable general equilibrium solutions often undermine their usefulness. As we have seen, the data violate key assumptions made in these models, and the match to institutional realities, at both micro and macro levels, is often very poor.”

This is how a well-mannered economist politely points out that something is very wrong.

The abstract from Paul Romer’s recent paper on DSGE macro summarises the attitude of Christiano at. al.:

“For more than three decades, macroeconomics has gone backwards… Macroeconomic theorists dismiss mere facts … Their models attribute fluctuations in aggregate variables to imaginary causal forces that are not influenced by the action that any person takes. [This] hints at a general failure mode of science that is triggered when respect for highly regarded leaders evolves into a deference to authority that displaces objective fact from its position as the ultimate determinant of scientific truth.”

What is the “scientific” argument for DSGE? It goes something like this. In the 1970s, macroeconomics mostly consisted of a set of relationships which were assumed to be stable enough to inform policy. The attitude taken to underlying microeconomic behaviour was, broadly, “we don’t have an exact model which tells us how this combination of microeconomic behavours produces the aggregate relationship but we think this is both plausible and stable enough to be useful”.

When the relationships that had previously appeared stable broke down at the end of the 1970s — as macroeconomic relationships have a habit of doing — this opened the door for the Freshwater economists to declare all such theorising to be invalid and instead insist that all macro models be built on the basis of Walrasian general equilibrium. Only then, they argued, could we be sure that the macro relationships were truly structural and therefore not invariant to government policy.

There was also a convenient side-effect for the Chicago School libertarians: state-of-the-art Walrasian general equilibrium had reached the point where the best that could be managed was to build very simple models in which all markets, including the labour market, cleared continuously — basically a very crude “economics 101” model with an extra dimension called “time”, and a bit of dice-rolling thrown in for good measure. The result — the so-called “Real Business Cycle model” — is something like a game of Dungeons and Dragons with the winner decided in advance and the rules replaced by an undergrad micro textbook. The associated policy recommendations were ideologically agreeable to the Freshwater economists.

Economics was declared a science and the problems of involuntary unemployment, business cycles and financial instabilty were solved at the stroke of a pen. There were a few awkward details: working out what would happen if there were lots of different individuals in the system was a bit tricky — so it was easier just to assume one big person. This did away with much of the actual microeconomic “foundations” and just replaced one sort of assumed macro relationship with another — but this didn’t seem to bother anyone unduly. There were also some rather inconvenient mathematical results about the properties of aggregate production functions that nobody likes to talk about. But aside from these minor details it was all very scientific. A great discovery had been made: business cycles were driven by the unexplained residual from an internally inconsistent aggregate production function. A new consensus emerged — aside from sniping from Robert Solow and a few heterodox cranks — that this was the only way to do scientific macroeconomics.

if you wanted to get away from the Econ 101 conclusions and argue, for example, that monetary policy could have some short-run effects, you now had no choice other than to start with the new model and add “frictions” or “imperfections” — anything else was dilettantism. The best-known of these epicycle-like modifications is the “Calvo Fairy” — the assumption that not all prices adjust instantly following a policy change. This allowed those less devoted to extreme free-market politics to derive old favourites such as the expectations-augmented Phillips curve in this strange new world.

Simon Wren-Lewis describes this hard reset of the discipline as follows: “Freshwater created a revolution and won, and were in a position to declare Year Zero: only things done properly (i.e consistently microfounded) are true macro. That was good for a new generation, who could rediscover past knowledge but because they (re)did it ‘properly’ avoid any acknowledgement of what had come before.” The implication is that all pre-DSGE macro is invalid and, from Year Zero onwards, anyone doing macro without DSGE is not doing it “properly”.

This is where the story gets really odd. If, for instance, the Freshwater people had said “there are some problems with your models not fitting the data, and by the way, we’ve managed to add a time dimension to Walrasian general equilibrium, cool huh?” things might have turned out OK. The Freshwater people could have amused themselves playing optimising Dungeons and Dragons while everyone else tryed to work out why the Phillips curve had broken down.

Instead, somehow, the Freshwater economists managed to create Year Zero: everyone now has to play by their rules. For the next 30 years or so, instead of investigating how economies actually functioned, macroeconomists worked out how to get the new model to reproduce the few results that were already well known and had some degree of stability — basically the Phillips Curve. What they didn’t do was produce any new understanding of how economies worked, or develop models with any out of sample predictive power.

On what basis do Christiano et al. then argue that DSGE is the only game in town for making macro policy and,  more bizarrely, the only place where we can do “experiments”? One can certainly do experiments with a DSGE model — but you are experimenting on a DSGE model, not the economy. And it’s fairly well established by now that the economy doesn’t behave much like any benchmark DSGE model.

What Christiano et. al. are attempting to do is reimpose the Year Zero rules: anyone doing macro without DSGE is not doing it “properly”. But on what basis is DSGE macro “done properly”? What is the empirical evidence?

There are two places to look for empirical validation — the micro data and the macro data. Why look at micro data for validation of a macro model? The answer is that Year Zero imposed the requirement that all macro models be deduced — one logical step after another — from microeconomic assumptions. As Lucas, the leading revolutionary put it, “If these developments succeed, the term ‘macroeconomic’ will simply disappear from use and the modifier ‘micro’ will become superfluous. We will simply speak, as did Smith, Ricardo, Marshall and Walras of economic theory”

Is the microeconomic theory correct? The answer is “we don’t know”. It is a set of assumptions about how individuals and firms behave which is all but impossible to either validate or falsify.

The use of the deductive method in economics originated with Ricardo’s Principles of Political Economy in 1819 and is summarised by Nassau Senior in 1836:

“The economist’s premises consist of a very few general propositions, the result of observation, or consciousness, and scarely requiring proof … which every man, as soon as he hears them, admits as familiar to his thoughts … [H]is inferences are nearly as general, and, if he has reasoned correctly, as certain, as his premises”

Nearly two hundred years later, Simon Wren-Lewis’ description of the method of DSGE macro is remarkably similar:

“Microeconomics is built up in a deductive manner from a small number of basic axioms of human behaviour. How these axioms are validated is controversial, as are the implications when they are rejected. Many economists act as if they are self evident.”

What of the macroeconomic results — perhaps we shouldn’t worry whether the microfoundations are correct if the macro models fit the data?

The Freshwater version of the model concluded that all government policy has no effect and that any changes are driven by an unexplained residual. The more moderate Saltwater version, with added Calvo fairy, allowed a rediscovery of Milton Friedman’s main results: an expectations-augmented Phillips Curve and short-run demand effects from monetary policy. The model has two basic equations: aggregate demand (the IS relationship) and aggregate supply (the Phillips curve) along with a policy response rule.

The first, the aggregate demand relationship, is based on an underlying assumption about how households behave in response to changes in the rate of interest. Unfortunately, not only does the equation not fit the data, the sign of the main coefficient appears to be wrong. This is likely because, rather than trying to understand the emergent properties of many interacting agents, modellers took the short-cut of assuming that the one big person assumed to represent the economy would simply replicate the behaviour of a single textbook-rational individual — much like assuming that the behaviour of an ant colony would be the same as that of one big textbook ant. It’s hard to see how one can make an argument that this has advanced knowledge beyond what you could glean from a straightforward Keynesian or Modigliani consumption function. What if, instead, we’d spent 30 years looking at the data and trying to work out how people actually make consumption and investment decisions?

What of the other relationship, the Phillips Curve? The Financial Times has recently published a series of articles on the growing, and awkward, realisation that the Phillips Curve relationship appears to have once again broken down. This was the theme of a recent all-star conference at the Peterson Institute. Gavyn Davies summarises the problem: “Without the Phillips Curve, the whole complicated paraphernalia that underpins central bank policy suddenly looks very shaky. For this reason, the Phillips Curve will not be abandoned lightly by policy makers.”

The “complicated paraphernalia” Davies refers to are the two basic equations just described. More complex versions of the model do exist, which purport to capture further stylised macro relationships beyond the standard pair. This is done, however, by adding extra degrees of freedom — justified as essentially arbitrary “frictions” — and then over fitting the model to the data. The result is that the models are pretty good at “predicting” the data they are trained on, and hopeless at anything else.

30 years of DSGE research have produced exactly one empirically plausible result — the expectations-augmented Phillips Curve. It was already well known. There is an ironic twist here: the breakdown of the Phillips Curve in the 1970s gave the Freshwater economists their breakthrough. The breakdown of the Phillips Curve now — in the other direction — leaves DSGE with precisely zero verifiable achievements.

Christiano et al.’s paper is welcome in one respect. It confirms what macroeconomists at the top of the discipline think about those lower down the academic pecking order — particularly those who take a critical view. They have made public what many of us long suspected was said behind closed doors.

The best response I can think of once again comes from Simon Wren-Lewis, who seems to have seen Christiano et. al coming:

“That some macroeconomists (I call them microfoundations purists) can argue that you should model and give policy advice based not on what you see but on what you can microfound represents something that I cannot imagine any philosopher of science taking seriously (after they had stopped laughing).”

 

Strong and stable? The Conservatives’ economic record since 2010

In a recent interview, Theresa May was asked by Andrew Neil how the Conservatives would fund their manifesto commitments on NHS spending. Given that the Conservatives chose not to cost their manifesto pledges, May was unable to answer. Instead she simply repeated that the Conservatives are the only party that can deliver the economic growth and stability required to pay for essential public services. When pressed, May’s response was simple: ‘our economic credibility is not in doubt’.

Does the record of the last seven years support May’s claim?

The first statistic always quoted in such discussions is GDP growth. A lot has been made of the latest quarterly GDP figures, showing the UK at the bottom of the G7 league with quarterly GDP growth of just 0.2%. But these numbers actually tell us very little: they refer to a single quarter and are still subject to revision.

It is more useful to look at real GDP per capita over a longer period of time. This tells us the additional ‘real’ income available per person that has been generated. The performance of the G7 countries since the pre-crisis peak in 2007 is shown in the chart below, with the series indexed to 1 in 2007 for each country. (Data are taken from the most recent IMF WEO database.)

G7 GDP per capita, 2007-2016

GDP per capita in the UK only surpassed its pre-crisis level in 2015. By 2016, GDP per capita relative to the pre-crisis level was less than 2% higher than in 2007, putting the UK behind Japan, Germany, the US and Canada, slightly ahead of France, and well ahead of the Italian economy which remains mired in a deep depression. On this measure, the UK’s performance is not particularly impressive.

For most people, wages are a more important gauge of economic performance than GDP per capita. Here, the UK is an outlier. Relative real wage growth in the G7 economies is shown in the table below, alongside the changes in GDP per capita for the period 2007-2015.

Country

% change in GDP per capita, 2007-2015

% change in average real wage, 2007-2015

Canada 3.2 0.8
France -0.2 0.6
Germany 6.3 0.9
Italy -11.7 -0.7
Japan 3.0 -0.2
United Kingdom 0.7 -1.0
United States 3.7 0.5

Despite coming mid-table in terms of GDP per capita, the UK has the worst performance in terms of real wages, which have fallen by an average of 1% per year over the period. Even in depression-struck Italy, wages did not fall so far.

This translates into a fall of almost five percent in the real wage of the typical (median) worker since the crisis, as the chart below shows. This LSE paper, from which the chart is taken, finds that while almost everyone is worse off since the crisis, the youngest have seen the largest falls in income with 18-21-year-olds facing a fall in real wages of over 15%

Chart-3-LSE

With the value of the pound falling since the Brexit vote, inflation is once again eating into real wages and the latest figures show that, after a period of a couple of years in which wages had been recovering, real wages are now falling again and are likely to do so for the next few years. Average earnings are not projected to reach 2007 levels again until 2022 – by then the UK will have gone fifteen years without a pay rise.

A related issue is the UK’s desperately poor productivity performance. ‘Productivity’ here refers to the amount produced per worker on average. As the chart below from the Resolution Foundation shows, the UK has now experienced a decade without any increase in productivity — something which is historically unprecedented.

CHART-productivity

What causes productivity growth is a controversial topic among economists. Until recently, the majority view was that productivity is not affected by government macroeconomic policy. This position (which I disagree with) is increasingly hard to defend. As Simon Wren-Lewis argues here, evidence is mounting that the UK’s productivity disaster is the result of government policy: the Conservatives’ austerity policies have caused flatlining productivity.

Austerity — or, as it was branded at the time, the ‘Long Term Economic Plan‘ — was the central plank of Osborne’s policy from 2010 until the Brexit referendum vote in 2015.

As I and others have argued at length elsewhere, austerity was based on two false premises — ‘lies’ might be more accurate. The first was that excessive spending by Labour was a cause of the 2008 crisis. The second was that the size of the UK’s government debt posed serious and immediate risks that outweighed other concerns.

One thing that almost all macroeconomists agree on is that when recovering from a severe downturn such as 2008 — and with interest rates at nearly zero — the deficit should not be the target of policy. Instead, it should be allowed to expand until the economy has recovered.

Simply put, the deficit should not be used as a yardstick for successful management of an economy in the aftermath of a major economic crisis such as 2008. But since eliminating the deficit was the single most important target of the Conservatives’ so-called Long Term Economic Plan, we should examine the record.

In 2010, Osborne stated that the deficit would be eliminated by 2015. Two years after that deadline passed, the current Conservative manifesto states — in a passage that would not pass any undergraduate economics exam — that they will ‘aim to’ eliminate the deficit by 2025.

Even on their own entirely misguided terms, they have failed completely.

FIG-LTEP

While the dangers of the public debt have been vastly exaggerated by the Conservatives, they have had little to say about private sector debt. It is now widely accepted that the only remaining motor of economic growth is consumption spending. But with wages stagnant, continued growth of consumption cannot be sustained without rising levels of household debt.

This is the reason given when economists are asked why their predictions of post-referendum recession were so wrong: they didn’t anticipate the current credit-driven consumption burst. But this trend has been apparent for at least the last two years. It shouldn’t have been too hard to see this coming.

Chart-Credit-Cards

Just as the Tories tend to stay quiet on private debt, they also have little to say about the ‘other’ deficit — the current account deficit. This is a measure of how much the country is reliant on foreigners to finance our spending. The deficit expanded from 2011 onward to reach almost 5% of GDP. This is an important source of vulnerability for a country which is about to try and extricate itself from economic integration with its closest neighbours.

CHART-BoP- current account balance as per cent of GDP

Overall, the Tories economic record is far from impressive: stagnant wages and productivity, weak investment and manufacturing, rising household debt, and a large external deficit.

Now, a reasonable response might be that these are long-standing issues with the UK economy and are not the fault of the Conservatives. There is some truth to this. But if this is the case, Theresa May should identify and acknowledge these issues and provide a clear outline of how her policies will address them. This is not what she has done. Instead, she simply repeats her mantra that only the Conservatives will deliver on the economy, without providing any evidence to support her claim.

And then there is the decision to call a referendum on Brexit. It is hard to think of a more economically reckless move. Household analogies for government economic policy should be avoided — but I can’t think of an alternative in this case.

Following up on an austerity programme with the Brexit referendum is like sending the children to school without lunch money for six years and allowing the house to fall into serious disrepair in order to needlessly over-pay a zero-interest mortgage — and then gambling the house on a dice game.

Given this record, it is astonishing that the Conservatives present themselves, with a straight face, as the party of economic competence — and the media dutifully echoes the message. The truth is that the Conservatives have mismanaged the economy for the last seven years, needlessly imposing austerity, choking off growth in productivity, wages and incomes. They then called an entirely unnecessary referendum, gambling the future prosperity of the country for political gain.

Theresa May is correct — there is little doubt about the economic credibility of the Conservatives. It is in short supply.

Thoughts on the NAIRU

Simon Wren-Lewis’s post attacking Matthew Klein’s critique of the NAIRU provoked some strong reactions. On reflection, my initial response was wide of the mark. Matthew responded saying he agreed with most of Simon’s piece.

So are we all in agreement? I think there are differences, but we need to first clarify the issues.

Matthew’s main point was empirical: if you want to use a relationship between employment and inflation as a policy target it needs to be relatively stable. The evidence suggests it is not.

But there is a deeper question of what the NAIRU actually means – what is a NAIRU? The simple definition is straightforward: it is the rate of unemployment at which inflation is stable. If policy is used to increase demand, reducing unemployment below the NAIRU, inflation will rise until excess demand is removed and unemployment allowed to increase again.

At first glance this appears all but identical to the ‘natural rate of unemployment’, a concept originating with Friedman’s monetarism and inherited by some New Keynesian models – in particular the ‘standard’ sticky-price DSGE model of Woodford and others. In this view, the economy has ‘natural rates’ of output and employment, beyond which any attempt by policy makers to increase demand becomes futile, leading only to ever-higher inflation. Since there is a direct correspondence between stabilizing inflation and fixing output and employment at their ‘natural’ rates, policy makers should simply adjust interest rates to hit an inflation target. In typically modest fashion, economists refer to this as the ‘Divine Coincidence‘ – despite the fact it is essentially imposed on the models by assumption.

Matthew’s piece skips over this part of the history, jumping straight from Bill Phillips’s empirical relationship to the NAIRU. But the NAIRU is a weaker claim than the natural rate. As Simon says, all that is required for a NAIRU is a relationship of the form inf = f(U, E[inf]), i.e. current inflation is some function of unemployment and expected inflation. At its simplest, agents could just assume inflation will be the same in the current period as the last period. Then, employment above some level would causing rising inflation and vice versa.

More sophisticated New Keynesian formulations of the NAIRU are a good distance removed from the ‘natural rate’ theory – these models include imperfections in the labour and product markets and a bargaining process between workers and firms. As a result, they incorporate (at least short-run) involuntary unemployment and see inflation as driven by competing claims on output rather than the ‘too much nominal demand chasing too few goods’ story of the monetarists and simple DSGE models.

It is also the case that such a relationship is found in many heterodox models. Engelbert Stockhammer explores heterodox views on the NAIRU in a provocatively-titled paper, ‘Is the NAIRU Theory a Monetarist, New Keynesian, Post Keynesian or Marxist Theory?’. He doesn’t identify a clear heterodox position – some Post-Keynesians reject the NAIRU outright, while others present models which incorporate NAIRU-like relationships.

Engelbert notes that arguably the earliest definition of the NAIRU is to be found in Joan Robinson’s 1937 Essays in the Theory of Employment:

In any given conditions of the labour market there is a certain more or less definite level of employment at which money wages will rise … there is a certain level of employment, determined by the general strategical position of the Trade Unions, at which money wages rise, and at that level of employment there is a certain level of real wages, determined by the technical conditions of production and the degree of monopoly’ (Robinson, 1937, pp. 4-5)

Recent Post-Keynesian models also include NAIRU-like relationships. For example, Godley and Lavoie’s textbook includes a model in which workers and firms compete by attempting to impose money-wage and price increases respectively. The size of wage increases demanded by workers is a function of the employment rate relative to some ‘full employment’ level. That sounds a lot like a NAIRU – but that isn’t how Godley and Lavoie see it:

Inflation under these assumptions does not necessarily accelerate if employment stays in excess of its ‘full employment’ level. Everything depends on the parameters and whether they change … An implication of the story proposed here is that there is no vertical long-run Phillips curve. There is no NAIRU. (Godley and Lavoie, 2007, p. 304, my emphasis)

The authors summarise their view with a quote from an earlier work by Godley:

Indeed if it is true that there is a unique NAIRU, that really is the end of discussion of macroeconomic policy. At present I happen not to believe it and that there is no evidence of it. And I am prepared to express the value judgment that moderately higher inflation rates are an acceptable price to pay for lower unemployment. But I do not accept that it is a foregone conclusion that inflation will be higher if unemployment is lower (Godley 1983: 170, my emphasis).

This highlights a key difference between Post-Keynesian and neoclassical approaches to the NAIRU: where Post-Keynesian models do include NAIRU-like relationships, the relevent employment level is endogenous, due to hysteresis effects for example. In other words, the NAIRU moves around and is influenced by demand-management policy. As such, the NAIRU is not an attractor for the unemployment rate as in many neoclassical models.

Marxist theory also contains something which looks a lot like a NAIRU: the ‘industrial reserve army’ of the unemployed. Marx argued that unemployment is the mechanism by which capitalists discipline workers and prevent wage claims rising to the point at which profits and capital accumulation are depleted. Periodic recessions are therefore a necessary part of the capitalist development process.

This led Nicholas Kaldor to describe Margaret Thatcher as ‘our first Marxist Prime Minister’ – not because she was an advocate of socialist revolution but because she understood the reserve army mechanism: ‘They have managed to create a pool – or a “reserve army” as Marx would have called it – of 3 million unemployed … the British working classes have been thoroughly cowed and frightened.’ (This point is passed over rather quickly in Simon’s piece. In the 1980s, he writes, ‘policy changed and increased unemployment and inflation fell.’)

So we should be careful about blanket dismissals of the NAIRU. Instead, we must be clear how our analysis differs: what are the mechanisms which generate inflationary pressure at low levels of unemployment – conflicting claims or excess nominal demand? Is the NAIRU stable and exogenous? Does it act as an attractor for the unemployment rate, and over what time period? What are the implications for policy?

Ultimately, I think this breaks down into an issue about semantics. How far from the unique, stable, vertical long-run Phillips curve can we get and still have something we call a NAIRU? Simon adopts a very loose definition:

There is a relationship between inflation and unemployment, but it is just very difficult to pin down. For most macroeconomists, the concept of the NAIRU really just stands for that basic macroeconomic truth.

I’d like to believe this were true. But I suspect most macroeconomists, trained on New Keynesian DSGE models, have a narrower view: they tend to think in terms of a stable short-run sticky-price Phillips curve and a unique long-run Phillips curve at the ‘natural’ rate of employment.

There is one other aspect to consider. Engelbert Stockhammer distinguishes between the New Keynesian NAIRU theory and the New Keynesian NAIRU story. He argues (writing in 2007, just before the crisis) that the NAIRU has been used as the basis for an account of unemployment which blames inflexible labour markets, over-generous welfare states, job protection measures and strong unions. The policy prescriptions are then straightforward: labour markets should be deregulated and welfare states scaled back. Demand management should not be used to reduce unemployment.

While economists have changed their tune substantially in the decade since the financial crisis, I suspect that the NAIRU story is one reason that defence of the NAIRU theory generates such strong reactions.

EDIT: Bruno Bonizzi points me to this piece at the INET blog with has an excellent discussion of the empirical evidence and theoretical implications of hysteresis effects and an unstable NAIRU.

 

Image reproduced from Wikipedia: https://en.wikipedia.org/wiki/File:NAIRU-SR-and-LR.svg

Economics, Ideology and Trump

So the post-mortem begins. Much electronic ink has already been spilled and predictable fault lines have emerged. Debate rages in particular on the question of whether Trump’s victory was driven by economic factors. Like Duncan Weldon, I think Torsten Bell gets it about right – economics is an essential part of the story even if the complete picture is more complex.

Neoliberalism is a word I usually try to avoid. It’s often used by people on the left as an easy catch-all to avoid engaging with difficult issues. Broadly speaking, however, it provides a short-hand for the policy status quo over the last thirty years or so: free movement of goods, labour and capital, fiscal conservatism, rules-based monetary policy, deregulated finance and a preference for supply-side measures in the labour market.

Some will argue this consensus has nothing to with the rise of far-right populism. I disagree. Both economics and economic policy have brought us here.

But to what extent has academic economics provided the basis for neoliberal policy? The question had been in my mind even before the Trump and Brexit votes. A few months back, Duncan Weldon posed the question, ‘whatever happened to deficit bias?’ In my view, the responses at the time missed the mark. More recently, Ann Pettifor and Simon Wren Lewis have been discussing the relationship between ideology, economics and fiscal austerity.

I have great respect for Simon – especially his efforts to combat the false media narratives around austerity. But I don’t think he gets it right on economics and ideology. His argument is that in a standard model – a sticky-price DSGE system – fiscal policy should be used when nominal rates are at the zero lower bound. Post-2008 austerity policies are therefore at odds with the academic consensus.

This is correct in simple terms, but I think misses the bigger picture of what academic economics has been saying for the last 30 years. To explain, I need to recap some history.

Fiscal policy as a macroeconomic management tool is associated with the ideas of Keynes. Against the academic consensus of his day, he argued that the economy could get stuck in periods of demand deficiency characterised by persistent involuntary unemployment. The monetarist counter-attack was led by Milton Friedman – who denied this possibility. In the long run, he argued, the economy has a ‘natural’ rate of unemployment to which it will gravitate automatically (the mechanism still remains to be explained). Any attempt to use activist fiscal or monetary policy to reduce unemployment below this natural rate will only lead to higher inflation. This led to the bitter disputes of the 1960s and 70s between Keynesians and Monetarists. The Monetarists emerged as victors – at least in the eyes of the orthodoxy – with the inflationary crises of the 1970s. This marks the beginning of the end for fiscal policy in the history of macroeconomics.

In Friedman’s world, short-term macro policy could be justified in a deflationary situation as a way to help the economy back to its ‘natural’ state. But, for Friedman, macro policy means monetary policy. In line with the doctrine that the consumer always knows best, government spending was proscribed as distortionary and inefficient. For Friedman, the correct policy response to deflation is a temporary increase in the rate of growth of the money supply.

It’s hard to view Milton Friedman’s campaign against Keynes as disconnected from ideological influence. Friedman’s role in the Mont Pelerin society is well documented. This group of economic liberals, led by Friedrich von Hayek, formed after World War II with the purpose of opposing the move towards collectivism of which Keynes was a leading figure. For a time at least, the group adopted the term ‘neoliberal’ to describe their political philosophy. This was an international group of economists whose express purpose was to influence politics and politicians – and they were successful.

Hayek’s thesis – which acquires a certain irony in light of Trump’s ascent – was that collectivism inevitably leads to authoritarianism and fascism. Friedman’s Chicago economics department formed one point in a triangular alliance with Lionel Robbins’ LSE in London, and Hayek’s fellow Austrians in Vienna. While in the 1930s, Friedman had expressed support for the New Deal, by the 1950s he had swung sharply in the direction of economic liberalism. As Brad Delong puts it:

by the early 1950s, his respect for even the possibility of government action was gone. His grudging approval of the New Deal was gone, too: Those elements that weren’t positively destructive were ineffective, diverting attention from what Friedman now believed would have cured the Great Depression, a substantial expansion of the money supply. The New Deal, Friedman concluded, had been ‘the wrong cure for the wrong disease.’

While Friedman never produced a complete formal model to describe his macroeconomic vision, his successor at Chicago, Robert Lucas did – the New Classical model. (He also successfully destroyed the Keynesian structural econometric modelling tradition with his ‘Lucas critique’.) Lucas’ New Classical colleagues followed in his footsteps, constructing an even more extreme version of the model: the so-called Real Business Cycle model. This simply assumes a world in which all markets work perfectly all of the time, and the single infinitely lived representative agent, on average, correctly predicts the future.

This is the origin of the ‘policy ineffectiveness hypothesis’ – in such a world, government becomes completely impotent. Any attempt at deficit spending will be exactly matched by a corresponding reduction in private spending – the so-called Ricardian Equivalence hypothesis. Fiscal policy has no effect on output and employment. Even monetary policy becomes totally ineffective: if the central bank chooses to loosen monetary policy, the representative agent instantly and correctly predicts higher inflation and adjusts her behaviour accordingly.

This vision, emerging from a leading centre of conservative thought, is still regarded by the academic economics community as a major scientific step forward. Simon describes it as `a progressive research programme’.

What does all this have to with the current status quo? The answer is that this model – with one single modification – is the ‘standard model’ which Simon and others point to when they argue that economics has no ideological bias. The modification is that prices in the goods market are slow to adjust to changes in demand. As a result, Milton Friedman’s result that policy is effective in the short run is restored. The only substantial difference to Friedman’s model is that the policy tool is the rate of interest, not the money supply. In a deflationary situation, the central bank should cut the nominal interest rate to raise demand and assist the automatic but sluggish transition back to the `natural’ rate of unemployment.

So what of Duncan’s question: what happened to deficit bias? – this refers to the assertion in economics textbooks that there will always be a tendency for governments to allow deficits to increase. The answer is that it was written out of the textbooks decades ago – because it is simply taken as given that fiscal policy is not the correct tool.

To check this, I went to our university library and looked through a selection of macroeconomics textbooks. Mankiw’s ‘Macroeconomics’ is probably the mostly widely used. I examined the 2007 edition – published just before the financial crisis. The chapter on ‘Stabilisation Policy’ dispenses with fiscal policy in half a page – a case study of Romer’s critique of Keynes is presented under the heading ‘Is the Stabilization of the Economy a Figment of the Data?’ The rest of the chapter focuses on monetary policy: time inconsistency, interest rate rules and central bank independence. The only appearance of the liquidity trap and the zero lower bound is in another half-page box, but fiscal policy doesn’t get a mention.

The post-crisis twelfth edition of Robert Gordon’s textbook does include a chapter on fiscal policy – entitled `The Government Budget, the Government Debt and the Limitations of Fiscal Policy’. While Gordon acknowledges that fiscal policy is an option during strongly deflationary periods when interest rates are at the zero lower bound, most of the chapter is concerned with the crowding out of private investment, the dangers of government debt and the conditions under which governments become insolvent. Of the textbooks I examined, only Blanchard’s contained anything resembling a balanced discussion of fiscal policy.

So, in Duncan’s words, governments are ‘flying a two engined plane but choosing to use only one motor’ not just because of media bias, an ill-informed public and misguided politicians – Simon’s explanation – but because they are doing what the macro textbooks tell them to do.

The reason is that the standard New Keynesian model is not a Keynesian model at all – it is a monetarist model. Aside from the mathematical sophistication, it is all but indistinguishable from Milton Friedman’s ideologically-driven description of the macroeconomy. In particular, Milton Friedman’s prohibition of fiscal policy is retained with – in more recent years – a caveat about the zero-lower bound (Simon makes essentially the same point about fiscal policy here).

It’s therefore odd that when Simon discusses the relationship between ideology and economics he chooses to draw a dividing line between those who use a sticky-price New Keynesian DSGE model and those who use a flexible-price New Classical version. The beliefs of the latter group are, Simon suggests, ideological, while those of the former group are based on ideology-free science. This strikes me as arbitrary. Simon’s justification is that, despite the evidence, the RBC model denies the possibility of involuntary unemployment. But the sticky-price version – which denies any role for inequality, finance, money, banking, liquidity, default, long-run unemployment, the use of fiscal policy away from the ZLB, supply-side hysteresis effects and plenty else besides – is acceptable. He even goes so far as to say ‘I have no problem seeing the RBC model as a flex-price NK model’ – even the RBC model is non-ideological so long as the hierarchical framing is right.

Even Simon’s key distinction – the New Keynesian model allows for involuntary unemployment – is open to question. Keynes’ definition of involuntary unemployment is that there exist people willing and able to work at the going wage who are unable to find employment. On this definition the New Keynesian model falls short – in the face of a short-run demand shortage caused by sticky prices the representative agent simply selects a new optimal labour supply. Workers are never off their labour supply curve. In the Smets Wouters model – a very widely used New Keynesian DSGE model – the labour market is described as follows: ‘household j chooses hours worked Lt(j)’. It is hard to reconcile involuntary unemployment with households choosing how much labour they supply.

What of the position taken by the profession in the wake of 2008? Reinhart and Rogoff’s contribution is by now infamous. Ann also draws attention to the 2010 letter signed by 20 top-ranking economists – including Rogoff – demanding austerity in the UK. Simon argues that Ann overlooks the fact that ‘58 equally notable economists signed a response arguing the 20 were wrong’.

It is difficult to agree that the signatories to the response letter, organised by Lord Skidelsky, are ‘equally notable’. Many are heterodox economists – critics of standard macroeconomics. Those mainstream economists on the list hold positions at lower-ranking institutions than the 20. I know many of the 58 personally – I know none of the 20. Simon notes:

Of course those that signed the first letter, and in particular Ken Rogoff, turned out to be a more prominent voice in the subsequent debate, but that is because he supported what policymakers were doing. He was mostly useful rather than influential.

For Simon, causality is unidirectional: policy-makers cherry-pick academic economics to fit their purpose but economists have no influence on policy. This seems implausible. It is undoubtedly true that pro-austerity economists provided useful cover for small-state ideologues like George Osborne. But the parallels between policy and academia are too strong for the causality to be unidirectional.

Osborne’s small state ideology is a descendent of Thatcherism – the point when neoliberalism first replaced Keynesianism. Is it purely coincidence that the 1980s was also the high-point for extreme free market Chicago economics such as Real Business Cycle models?

The parallel between policy and academia continues with the emergence of the sticky-price New Keynesian version as the ‘standard’ model in the 90s alongside the shift to the third way of Blair and Clinton. Blairism represents a modified, less extreme, version of Thatcherism. The all-out assault on workers and the social safety net was replaced with ‘workfare’ and ‘flexicurity’.

A similar story can be told for international trade, as laid out in this excellent piece by Martin Sandbu. In the 1990s, just as the ‘heyday of global trade integration was getting underway’, economists were busy making the case that globalisation had no negative implications for employment or inequality in rich nations. To do this, they came up with the ‘skill-biased technological change’ (SBTC) hypothesis. This states that as technology advances and the potential for automation grows, the demand for high-skilled labour increases. This introduces the hitch that higher educational standards are required before the gains from automation can be felt by those outside the top income percentiles. This leads to a `race between education and technology’ – a race which technology was winning, leading to weaker demand for middle and low-skill workers and rising ‘skill premiums’ for high skilled workers as a result.

Writing in the Financial Times shortly before the financial crisis, Jagdish Bagwati argued that those who looked to globalisation as an explanation for increasing inequality were misguided:

The culprit is not globalization but labour-saving technical change that puts pressure on the wages of the unskilled. Technical change prompts continual economies in the use of unskilled labour. Much empirical argumentation and evidence exists on this. (FT, January 4, 2007, p. 11)

As Krugman put it:

The hypothesis that technological change, by raising the demand for skill, has led to growing inequality is so widespread that at conferences economists often use the abbreviation SBTC – skill-biased technical change – without explanation, assuming that their listeners know what they are talking about (p. 132)

Over the course of his 2007 book, Krugman sets out on a voyage of discovery – ‘That, more or less, is the story I believed when I began working on this book’ (p. 6). He arrives at the astonishing conclusion – ‘[i]t sounds like economic heresy’ (p. 7) – that politics can influence inequality:

[I]nstitutions, norms and the political environment matter a lot more for the distribution of income – and … impersonal market forces matter less – than Economics 101 might lead you to believe (p. 8)

The idea that rising pay at the top of the scale mainly reflect social and political change, … strikes some people as … too much at odds with Economics 101.

If a left-leaning Nobel prize-winning economist has trouble escaping from the confines of Economics 101, what hope for the less sophisticated mind?

As deindustrialisation rolled through the advanced economies, wiping out jobs and communities, economists continued to deny any role for globalisation. As Martin Sandbu argues,

The blithe unconcern displayed by the economics profession and the political elites about whether trade was causing deindustrialisation, social exclusion and rising inequality has begun to seem Pollyannish at best, malicious at worst. Kevin O’Rourke, the Irish economist, and before him Lawrence Summers, former US Treasury Secretary, have called this “the Davos lie.”

For mainstream macroeconomists, inequality was not a subject of any real interest. While the explanation for inequality lay in the microeconomics – the technical forms of production functions – and would be solved by increasing educational attainment, in macroeconomic terms, the use of a representative agent and an aggregate production function simply assumed the problem away. As Stiglitz puts it:

[I]f the distribution of income (say between labor and capital) matters, for example, for aggregate demand and therefore for employment and output, then using an aggregate Cobb-Douglas production function which, with competition, implies that the share of labor is fixed, is not going to be helpful. (p.596)

Robert Lucas summed up his position as follows: ‘Of the tendencies that are harmful to sound economics, the most seductive, and in my opinion the most poisonous, is to focus on questions of distribution.’ It is hard to view this statement as informed more strongly by science than ideology.

But while economists were busy assuming away inequality in their models, incomes continued to diverge in most advanced economies. It was only with the publication of Piketty’s book that the economics profession belatedly began to turn its back on Lucas.

The extent to which economic insecurity in the US and the UK is driven by globalisation versus policy is still under discussion – my answer would be that it is a combination of both – but the skill-biased technical change hypothesis looks to be a dead end – and a costly one at that.

Similar stories can be told about the role of household debt, finance, monetary theory and labour bargaining power and monopoly – why so much academic focus on ‘structural reform’ in the labour market but none on anti-trust policy?  Heterodox economists were warning about the connections between finance, globalisation, current account imbalances, inequality, household debt and economic insecurity in the decades before the crisis. These warnings were dismissed as unscientific – in favour of a model which excluded all of these things by design.

Are economic factors – and economic policy – partly to blame for the Brexit and Trump votes? And are academic economists, at least in part, to blame for these polices? The answer to both questions is yes. To argue otherwise is to deny Keynes’ dictum that ‘the ideas of economists and political philosophers, both when they are right and when they are wrong are more powerful than is commonly understood.’

This quote, ‘mounted and framed, takes pride of place in the entrance hall of the Institute for Economic Affairs’ – the think-tank founded, with Hayek’s encouragement, by Anthony Fisher, as a way to promote and promulgate the ideas of the Mont Pelerin Society. The Institute was a success. Fisher was, in the words of Milton Friedman, ‘the single most important person in the development of Thatcherism’.

The rest, it seems, is history.

What is the Loanable Funds theory?

I had another stimulating discussion with Noah Smith last week. This time the topic was the ‘loanable funds’ theory of the rate of interest. The discussion was triggered by my suggestion that the ‘safe asset shortage’ and associated ‘reach for yield’ are in part caused by rising wealth concentration. The logic is straightforward: since the rich spend less of their income than the poor, wealth concentration tends to increase the rate of saving out of income. This means an increase in desired savings chasing the available stock of financial assets, pushing up the price and lowering the yield.

Noah viewed this as a plausible hypothesis but suggested it relies on the loanable funds model. My view was the opposite – I think this mechanism is incompatible with the loanable funds theory. Such disagreements are often enlightening – either one of us misunderstood the mechanisms under discussion, or we were using different definitions. My instinct was that it was the latter: we meant something different by ‘loanable funds theory’ (LFT hereafter).

To try and clear this up, Noah suggested Mankiw’s textbook as a starting point – and found a set of slides which set out the LFT clearly. The model described was exactly the one I had in mind – but despite agreeing that Mankiw’s exposition of the LFT was accurate it was clear we still didn’t agree about the original point of discussion.

The reason seems to be that Noah understands the LFT to describe any market for loans: there are some people willing to lend and some who wish to borrow. As the rate of interest rises, the volume of available lending increases but the volume of desired borrowing falls. In equilibrium, the rate of interest will settle at r* – the market-clearing  rate.

What’s wrong with this? – It certainly sounds like a market for ‘loanable funds’. The problem is that LFT is not a theory of loan market clearing per se. It’s a theory of macroeconomic equilibrium. It’s not a model of any old loan market: it’s a model of a one very specific market – the market which intermediates total (net) saving with total capital investment in a closed economic system.

OK, but saving equals investment by definition in macroeconomic terms: the famous S = I identity. How can there be a market which operates to ensure equality between two identically equal magnitudes?

The issue – as Keynes explained in the General Theory– is that in a modern capitalist economy, the person who saves and the person who undertakes fixed capital investment are not usually the same. Some mechanism needs to be in place to ensure that a decision to ‘not consume’ somewhere in the system – to save – is always matched by a decision to invest – to build a new machine, road or building – somewhere else in the economy.

To see the issue more clearly consider the ‘corn economy’ used in many standard macro models: one good – corn – is produced. This good can either be consumed or invested (by planting in the ground or storing corn for later consumption). The decision to plant or store corn is simultaneously both a decision to ‘not consume’ and to ‘invest’ (the rate of return on investment will depend on the mix of stored to planted corn). In this simple economy S = I because it can’t be any other way. A market for loanable funds is not required.

But this isn’t how modern capitalism works. Decisions to ‘not consume’ and decisions to invest are distributed throughout the economic system. How can we be sure that these decisions will lead to identical intended saving and investment – what ensures that S and I are equal? The loanable funds theory provides one possible answer to this question.

The theory states that decisions to save (i.e. to not consume) are decisive – investment adjusts automatically to accommodate any change in consumption behaviour. To see how this works, we need to recall how the model is derived. The diagram below shows the basic system (I’ve borrowed the figure from Nick Rowe).

lf

The upward sloping ‘desired saving’ curve is derived on the assumption that people are ‘impatient’ – they prefer current consumption to future consumption. In order to induce people to save,  a return needs to be paid on their savings. As the return paid on savings increases, consumers are collectively willing to forgo a greater volume of current consumption in return for a future payoff.

The downward sloping investment curve is derived on standard neoclassical marginalist principles. ‘Factors of production’ (i.e. labour and capital) receive ‘what they are worth’ in competitive markets. The real wage is equal to the marginal productivity of labour and the return on ‘capital’ is likewise equal to the marginal productivity of capital. As the ‘quantity’ of capital increases, the marginal product – and thus the rate of return – falls.

So the S and I curves depict how much saving and investment would take place at each possible rate of interest. As long as the S and I curves are well-defined and ‘monotonic’ (a strong assumption), there is only one rate of interest at which the amount people wish to lend is equal to the amount (other) people would like to borrow. This is r*, the point of intersection between the curves. This rate of interest is often referred to as the Wicksellian ‘natural rate’.

Now, consider what happens if the collective impatience of society decreases. At any rate of interest, consumption as a share of income will be lower and desired saving correspondingly higher – the S curve moves to the right. As the S curve shifts to the right – assuming no change in the technology determining the slope and position of the I curve – a greater share of national income is ‘not consumed’. But by pushing down the rate of interest in the loanable funds market, reduced consumption – somewhat miraculously – leads to an automatic increase in investment. An outward shift in the S curve is accompanied by a shift along the I curve.

Consider what this means for macroeconomic aggregates. Assuming a closed system, income is, by definition, equal to consumption plus investment: Y = C + I. The LFT says is that in freely adjusting markets, reductions in C due to shifts in preferences are automatically offset by increases in I. Y will remain at the ‘full employment’ rate of output at all times.

The LFT therefore underpins ‘Say’s Law’ – summarised by Keynes as ‘supply creates its own demand’. It was thus a key target for Keynes’ attack on the ‘Law’ in his General Theory. Keynes argued against the notion that saving decisions are strongly influenced by the rate of interest. Instead, he argued consumption is mostly determined by income. If individuals consume a fixed proportion of their income, the S curve in the diagram is no longer well defined – at any given level of output, S is vertical, but the position of the curve shifts with output. This is quite different to the LFT which regards the position of the two curves as determined by the ‘deep’ structural parameters of the system – technology and preferences.

How then is the rate of interest determined in Keynes’ theory? – the answer is ‘liquidity preference’. Rather than desired saving determining the rate of interest, what matters is the composition of financial assets people use to hold their savings. Keynes simplifies the story by assuming only two assets: ‘money’ which pays no interest and ‘bonds’ which do pay interest. It is the interaction of supply and demand in the bond market – not the ‘loanable funds’ market – which determines the rate of interest.

There are two key points here: the first is that saving is a residual – it is determined by output and investment. As such, there is no mechanism to ensure that desired saving and desired investment will be equalised. This means that output, not the rate of interest, will adjust to ensure that saving is equal to investment. There is no mechanism which ensures that output is maintained at full employment levels. The second is that interest rates can move without any change in either desired saving or desired investment. If there is an increase in ‘liquidity preference’ – a desire to hold lower yielding but safer assets, this will cause an increase in the rate of interest on riskier assets.

How can the original question be framed using these two models? – What is the implication of increasing wealth concentration on yields and macro variables?

I think Noah is right that one can think of the mechanism in a loanable funds world. If redistribution towards the rich increases the average propensity to save, this will shift the S curve to the right – as in the example above – reducing the ‘natural’ rate of interest. This is the standard ‘secular stagnation’ story – a ‘global savings glut’ has pushed the natural rate below zero. However, in a loanable funds world this should – all else being equal – lead to an increase in investment. This doesn’t seem to fit the stylised facts: capital investment has been falling as a share of GDP in most advanced nations. (Critics will point out that I’m skirting the issue of the zero lower bound – I’ll have to save that for another time).

My non-LFT interpretation is the following. Firstly, I’d go further than Keynes and argue that the rate of interest is not only relatively unimportant for determining S – it also has little effect on I. There is evidence to suggest that firms’ investment decisions are fairly interest-inelastic. This means that both curves in the diagram above have a steep slope – and they shift as output changes. There is no ‘natural rate’ of interest which brings the macroeconomic system into equilibrium.

In terms of the S = I identity, this means that investment decisions are more important for the determination of macro variables than saving decisions. If total desired saving as a share of income increases – due to wealth concentration, for example – this will have little effect on investment. The volume of realised saving, however, is determined by (and identically equal to) the volume of capital investment. An increase in desired saving manifests itself not as a rise in investment – but as a fall in consumption and output.

In such a scenario – in which a higher share of nominal income is saved – the result will be weak demand for goods but strong demand for financial assets – leading to deflation in the goods market and inflation in the market for financial assets. Strong demand for financial assets will reduce rates of return – but only on financial assets: if investment is inelastic to interest rate there is no reason to believe there will be any shift in investment or in the return on fixed capital investment.

In order explain the relative rates of return on equity and bonds, a re-working of Keynes’ liquidity preference theory is required. Instead of a choice between ‘money’ and ‘bonds’, the choice faced by investors can be characterised as a choice between risky equity and less-risky bonds. Liquidity preference will then make itself felt as an increase in the price of bonds relative to equity – and a corresponding movement in the yields on each asset. On the other hand, an increase in total nominal saving will increase the price of all financial assets and thus reduce yields across the board. Given that it is likely that portfolio managers will have minimum target rates of return, this is will induce a shift into higher-risk assets.

Consistent modelling and inconsistent terminology

Image reproduced from here

Simon Wren-Lewis has a couple of recent posts up on heterodox macro, and stock-flow consistent modelling in particular. His posts are constructive and engaging. I want to respond to some of the points raised.

Simon discusses the modelling approach originating with Wynne Godley, Francis Cripps and others at the Cambridge Economic Policy Group in the 1970s. More recently this approach is associated with the work of Marc Lavoie who co-wrote the key textbook on the topic with Godley.

The term ‘stock-flow consistent’ was coined by Claudio Dos Santos in his PhD thesis, ‘Three essays in stock flow consistent modelling’ and has been a source of misunderstanding ever since. Simon writes, ‘it is inferred that mainstream models fail to impose stock flow consistency.’ As I tried to emphasise  in the blog which Simon links to, this is not the intention: ‘any correctly specified closed mathematical macro model should be internally consistent and therefore stock-flow consistent. This is certainly true of DSGE models.’ (There is an important caveat here:  this consistency won’t be maintained after log-linearisation – a standard step in DSGE solution – and the further a linearised model gets from the steady state, the worse this inconsistency will become.)[1]

Marc Lavoie has emphasised that he regrets adopting the name, precisely because of the implication that consistency is not maintained in other modelling traditions. Instead, the term refers to a subset of models characterised by a number of specific features. These include the following: aggregate behavioural macro relationships informed by both empirical evidence and post-Keynesian theory; detailed, institutionally-specific modelling of the monetary and financial sector; and explicit feedback effects from financial balance sheets to economic behaviour and the stability of the macro system both in the short run and the long run.

A distinctive feature of these models is their rejection of the loanable funds theory of banking and money – a position endorsed in a recent Bank of England Quarterly Bulletin and Working Paper. Partially as a result of this view of the importance of money and money-values in the decision-making process, these models are usually specified in nominal magnitudes. As a result, they map more directly onto the national accounts than real-sector models which require complex transformations of data series using price deflators.

Since the behavioural features of these models are informed by a well-developed theoretical tradition, Simon’s assertion that SFC modelling is ‘accounting, not economics’ is inaccurate. Accounting is one important element in a broader methodological approach. Imposing detailed financial accounting alongside behavioural assumptions about how financial stocks and flows evolve imposes constraints across the entire system. Rather like trying to squeeze the air out of one part of a balloon, only to find another part inflating, chasing assets and liabilities around a closed system of linked balance sheets can be an informative exercise – because where leverage eventually turns up is not always clear at the outset. Likewise, SFC models may include detailed modelling of inventories, pricing and profits, or of changes in net worth due to asset price revaluation and price inflation. For such processes, even the accounting is non-trivial. Taking accounting seriously allows modellers to incorporate institutional complexity – something of increasing importance in today’s world.

The inclusion of detailed financial modelling allows the models to capture Godley’s view that agents aim to achieve certain stock-flow norms. These may include household debt-to-income ratios, inventories-to-sales ratios for firms and leverage ratios for banks. Many of the functional forms used implicitly capture these stock-flow ratios. This is the case for the simple consumption function used in the BoE paper discussed by Simon, as shown here. Of course, other functional specifications are possible, as in this model, for example, which includes a direct interest rate effect on consumption.

Simon notes that adding basic financial accounting to standard models is trivial but ‘in most mainstream models these balances are of no consequence’. This is an important point, and should set alarm bells ringing. Simon identifies one reason for the neutrality of finance in standard models: ‘the simplicity of the dominant mainstream model of intertemporal consumption’.

There are deeper reasons why the financial sector has little role in standard macro. In the majority of standard DSGE macro models, the system automatically tends towards some long-run supply side-determined full-employment equilibrium – in other words the models incorporate Milton Friedman’s long-run vertical Phillips Curve. Further, in most DSGE models, income distribution has no long-run effect on macroeconomic outcomes.

Post-Keynesian economics, which provides much of the underlying theoretical structure of SFC models, takes issue with these assumptions. Instead, it is argued, Keynes was correct in his assertion that demand deficiency can lead economies to become stuck in equilibria characterised by under-employment or stagnation.

Now, if the economic system is always in the process of returning to the flexible-price full-employment equilibrium, then financial stocks will be, at most, of transitory significance. They may serve to amplify macroeconomic fluctuations, as in the Bernanke-Gertler-Gilchrist models, but they will have no long-run effects. This is the reason that DSGE models which do attempt to incorporate financial leverage also require additional ‘ad-hoc’ adjustments to the deeper model assumptions – for example this model by Kumhof and Ranciere imposes an assumption of non-negative subsistence consumption for households. As a result, when income falls, households are unable to reduce consumption but instead run up debt. For similar reasons, if one tries to abandon the loanable funds theory in DSGE models – one of the key reasons for the insistence on accounting in SFC models – this likewise raises non-trivial issues, as shown in this paper by Benes and Kumhof  (to my knowledge the only attempt so far to produce such a model).

Non-PK-SFC models, such as the UK’s OBR model, can therefore incorporate modelling of sectoral balances and leverage ratios – but these stocks have little effect on the real outcomes of the model.

On the contrary, if long-run disequlibrium is considered a plausible outcome, financial stocks may persist and feedbacks from these stocks to the real economy will have non-trivial effects. In such a situation, attempts by individuals or sectors to achieve some stock-flow ratio can alter the long-run behaviour of the system. If a balance-sheet recession persists, it will have persistent effects on the real economy – such hysteresis effects are increasingly acknowledged in the profession.

This relates to an earlier point made in Simon’s post: ‘the fact that leverage was allowed to increase substantially before the crisis was not something that most macroeconomists were even aware of … it just wasn’t their field’. I’m surprised this is presented as evidence for the defence of mainstream macro.

The central point made by economists like Minsky and Godley was that financial dynamics should be part of our field. The fact that by 2007 it wasn’t, illustrates how badly mainstream macroeconomics went wrong. Between Real Business Cycle models, Rational Expectations, the Efficient Markets Hypothesis and CAPM, economists convinced themselves – and, more importantly, policy-makers – that the financial system was none of their business. The fact that economists forgot to look at leverage ratios wasn’t an absent-minded oversight. As Oliver Blanchard argues:

 ‘… mainstream macroeconomics had taken the financial system for granted. The typical macro treatment of finance was a set of arbitrage equations, under the assumption that we did not need to look at who was doing what on Wall Street. That turned out to be badly wrong.’

This is partially acknowledged by Simon when he argues that the ‘microfoundations revolution’ lies behind economists’ myopia on the financial system. Where I, of course, agree with Simon is that ‘had the microfoundations revolution been more tolerant of other methodologies … macroeconomics may well have done more to integrate the financial sector into their models before the crisis’. Putting aside the point that, for the most part, the microfoundations revolution didn’t actually lead to microfounded models, ‘integrating the financial sector’ into models is exactly what people like Godley, Lavoie and others were doing.

Simon also makes an important point in highlighting the lack of acknowledgement of antecedents by PK-SFC authors and, as a result, a lack of continuity between PK-SFC models and the earlier structural econometric models (SEMs) which were eventually killed off by the shift to microfounded models. There is a rich seam of work here – heterodox economists should both acknowledge this and draw on it in their own work. In many respects, I see the PK-SFC approach as a continuation of the SEM tradition – I was therefore pleased to read this paper in which Simon argues for a return to the use of SEMs alongside DSGE and VAR techniques.

To my mind, this is what is attempted in the Bank of England paper criticised by Simon – the authors develop a non-DSGE, econometrically estimated, structural model of the UK economy in which the financial system is taken seriously. Simon is right, however, that the theoretical justifications for the behavioural specifications and the connections to previous literature could have been spelled out more clearly.

The new Bank of England model is one of a relatively small group of empirically-oriented SFC models. Others include the Levy Institute model of the US, originally developed by Wynne Godley and now maintained by Gennaro Zezza, the UNCTAD Global Policy model, developed in collaboration with Godley’s old colleague Francis Cripps, and the Gudgin and Coutts model of the UK economy (the last of these is not yet fully stock-flow consistent but shares much of its theoretical structure with the other models).

One important area for improvement in these models lies with their econometric specification. The models tend to have large numbers of parameters, making them difficult to estimate other than through individual OLS regressions of behavioural relationships. PK-SFC authors can certainly learn from the older SEM tradition in this area.

I find another point of agreement in Simon’s statement that ‘heterodox economists need to stop being heterodox’. I wouldn’t state this so strongly – I think heterodox economists need to become less heterodox. They should identify and more explicitly acknowledge those areas in which there is common ground with mainstream economics.  In those areas where disagreement persists, they should try to explain more clearly why this is the case. Hopefully this will lead to more fruitful engagement in the future, rather than the negativity which has characterised some recent exchanges.

[1] Simon goes on to argue that stock-flow consistency is not ‘unique to Godley. When I was a young economist at the Treasury in the 1970s, their UK model was ‘stock-flow consistent’, and forecasts routinely looked at sector balances.’  During the 1970s, there was sustained debate between the Treasury and Godley’s Cambridge team, who were, aside from Milton Friedman’s monetarism, the most prominent critics of the Keynesian conventional wisdom of the time – there is an excellent history here. I don’t know the details but I wonder if the awareness of sectoral balances at the Treasury was partly due to Godley’s influence?

The Fable of the Ants, or Why the Representative Agent is No Such Thing

Image reproduced from here

Earlier in the summer, I had a discussion on Twitter with Tony Yates, Israel Arroyo and others on the use of the representative agent in macro modelling.

The starting point for representative agent macro is an insistence that all economic models must be ‘microfounded’. This means that model behaviour must be derived from the optimising behaviour of individuals – even when the object of study is aggregates such as employment, national output or the price level. But given the difficulty – more likely the impossibility – of building an individual-by-individual model of the entire economic system, a convenient short-cut is taken. The decision-making process of one type of agents as a whole (for example consumers or firms) is reduced to that of a single ‘representative’ individual – and  is taken to be identical to that assumed to characterise the behaviour of actual individuals.

For example, in the simple textbook DSGE models taught to macro students, the entire economic system is assumed to behave like a single consumer with fixed and externally imposed preferences over how much they wish to consume in the present relative to the future.

I triggered the Twitter debate by noting that this is equivalent to attempting to model the behaviour of a colony of ants by constructing a model of one large ‘average’ ant. The obvious issue illustrated by the analogy is that ants are relatively simple organisms with a limited range of behaviours – but the aggregate behaviour of an ant colony is both more complex and qualitatively different to that of an individual ant.

This is a well-known topic in computer science: a class of optimisation algorithms were developed by writing code which mimics the way that an ant colony collectively locates food. These algorithms are a sub-group of broader class of ‘swarm intelligence’ algorithms. The common feature is that interaction between ‘agents’ in a population, where the behaviour of each individual is specified as a simple set of rules, produces some emergent ‘intelligent’ behaviour at the population level.

In ants, one such behaviour is the collective food search: ants initially explore at random. If they find food, they lay down pheromone trails on their way back to base. This alters the behaviour of ants that subsequently set out to search for food: the trails attract ants to areas where food was previously located. It turns out this simple rules-based system produces a highly efficient colony-level algorithm for locating the shortest paths to food supplies.

The key point about these algorithms is that the emergent behaviour is qualitatively different from that of individual agents – and is typically robust to changes at the micro level: a reasonably wide degree of variation in ant behaviour at the individual level is possible without disruption to the behaviour of the colony. Further, these emergent properties cannot usually be identified by analysing a single agent in isolation – they will only occur as a result of the interaction between agents (and between agents and their environment).

But this is not how representative agent macro works. Instead, it is assumed that the aggregate behaviour is simply identical to that of individual agents. To take another analogy, it is like a physicist modelling the behaviour of a gas in a room by starting with the assumption of one room-sized molecule.

Presumably economists have good reason to believe that, in the case of economics, this simplifying assumption is valid?

On the contrary, microeconomists have known for a long time that the opposite is the case. Formal proofs demonstrate that a population of agents, each represented using a standard neoclassical inter-temporal utility function will not produce behaviour at the aggregate level which is consistent with a ‘representative’ utility function. In other words, such a system has emergent properties. As Kirman puts it:

“… there is no plausible formal justification for the assumption that the aggregate of individuals, even maximisers, acts itself like an individual maximiser. Individual maximisation does not engender collective rationality, nor does the fact that the collectivity exhibits a certain rationality necessarily imply that individuals act rationaly. There is simply no direct relation between individual and collective behaviour.”

Although the idea of the representative agent isn’t new – it appears in Edgeworth’s 1881 tract on ‘Mathematical Psychics’ – it attained its current dominance as a result of Robert Lucas’ critique of Keynesian structural macroeconomic models. Lucas argued that the behavioural relationships underpinning these models are not be invariant to changes in government policy and therefore should not be used to inform such policy. The conclusion drawn – involving a significant logical leap of faith – was that all macroeconomic models should be based on explicit microeconomic optimization.

This turned out to be rather difficult in practice. In order to produce models which are ‘well-behaved’ at the macro level, one has to impose highly implausible restrictions on individual agents.

A key restriction needed to ensure that microeconomic optimisation behaviour is preserved at the macro level is that of linear ‘Engel curves’. In cross-sectional analysis, this means individuals consume normal and inferior goods in fixed proportions, regardless of their income – a supermarket checkout worker will continue to consume baked beans and Swiss watches in unchanged proportions after she wins the lottery.

In an inter-temporal setting – i.e. in macroeconomic models – this translates to an assumption of constant relative risk aversion. This imposes the constraint that any individual’s aversion to losing a fixed proportion of her income remains constant even as her income changes.

Further, and unfortunately for Lucas, income distribution turns out to matter: if all individuals do not behave identically, then as income distribution changes, aggregate behaviour will also shift. As a result, aggregate utility functions will only be ‘well-behaved’ if, for example, individuals have identical and linear Engel curves, or if individuals have different linear Engel curves but income distribution is not allowed to change.

As well as assuming away any role for, say income distribution or financial interactions, these assumptions contradict well-established empirical facts. The composition of consumption shifts as income increases. It is hard to believe these restrictive special cases provide a sufficient basis on which to construct macro models which can inform policy decisions – but this is exactly what is done.

Kirman notes that ‘a lot of microeconomists said that this was not very good, but macroeconomists did not take that message on board at all. They simply said that we will just have to simplify things until we get to a situation where we do have uniqueness and stability. And then of course we arrive at the famous representative individual.’

The key point here is that a model in which the population as whole collectively solves an inter-temporal optimisation problem – identical to that assumed to be solved by individuals – cannot be held to be ‘micro-founded’ in any serious way. Instead, representative agent models are aggregative macroeconomic models – like Keynesian structural econometric models – but models which impose arbitrary and implausible restrictions on the behaviour of individuals. Instead of being ‘micro-founded’, these models are ‘micro-roofed’ (the term originates with Matheus Grasselli).

It can be argued that old-fashioned Keynesian structural macro behavioural assumptions can in fact stake a stronger claim to compatibility with plausible microeconomic behaviour – precisely because arbitrary restrictions on individual behaviour are not imposed. Like the ant-colony, it can be shown that under sensible assumptions, robust aggregate Keynesian consumption and saving functions can be derived from a range of microeconomic behaviours – both optimising and non-optimising.

So what of the Lucas Critique?

Given that representative agent models are not micro-founded but are aggregate macroeconomic representations, Peter Skott argues that ‘the appropriate definition of the agent will itself typically depend on the policy regime. Thus, the representative-agent models are themselves subject to the Lucas critique. In short, the Lucas inspired research program has been a failure.’

This does not mean that microeconomic behaviour doesn’t matter. Nor is it an argument for a return to the simplistic Keynesian macro modelling of the 1970s. As Hoover puts it:

‘This is not to deny the Lucas critique. Rather it is to suggest that its reach may be sufficiently moderated in aggregate data that there are useful macroeconomic relationships to model that are relatively invariant’

Instead, it should be accepted that some aggregate macroeconomic behavioural relationships are likely to be robust, at least in some contexts and over some periods of time. At the same time, we now have much greater scope to investigate the relationships between micro and macro behaviours. In particular, computing power allows for the use of agent-based simulations to analyse the emergent properties of complex social systems.

This seems a more promising line of enquiry than the dead end of representative agent DSGE modelling.

On ‘heterodox’ macroeconomics

Image reproduced from here

Noah Smith has a new post on the failure of mainstream macroeconomics and what he perceives as the lack of ‘heterodox’ alternatives. Noah is correct about the failure of mainstream macroeconomics, particularly the dominant DSGE modelling approach. This failure is increasingly – if reluctantly – accepted within the economics discipline. As Brad Delong puts it, DSGE macro has ‘… proven a degenerating research program and a catastrophic failure: thirty years of work have produced no tools for useful forecasting or policy analysis.’

I disagree with Noah, however, when he argues that ‘heterodox’ economics has little to offer as an alternative to the failed mainstream.

The term ‘heterodox economics’ is a difficult one. I dislike it and resisted adopting it for some time: I would much rather be ‘an economist’ than ‘a heterodox economist’. But it is clear that unless you accept – pretty much without criticism – the assumptions and methodology of the mainstream, you will not be accepted as ‘an economist’. This was not the case when Joan Robinson debated with Solow and Samuelson, or Kaldor debated with Hayek. But it is the case today.

The problem with ‘heterodox economics’ is that it is self-definition in terms of the other. It says ‘we are not them’ – but says nothing about what we are. This is because includes everything outside of the mainstream, from reasonably well-defined and coherent schools of thought such as Post Keynesians, Marxists and Austrians, to much more nebulous and ill-defined discontents of all hues. To put it bluntly, a broad definition of ‘people who disagree with mainstream economics’ is going to include a lot of cranks. People will place the boundary between serious non-mainstream economists and cranks differently, depending on their perspective.

Another problem is that these schools of thought have fundamental differences. Aside from rejecting standard neoclassical economics, the Marxists and the Austrians don’t have a great deal in common.

Noah seems to define heterodox economics as ‘non-mathematical’ economics. This is inaccurate. There is much formal modelling outside of the mainstream. The difference lies with the starting assumptions. Mainstream macro starts from the assumption of inter-temporal optimisation and a system which returns to the supply-side-determined full-employment equilibrium in the long run. Non-mainstream economists reject these in favour of assumptions which they regard as more empirically plausible.

It is true that there are some heterodox economists, for example Tony Lawson and Ben Fine who take the position that maths is an inappropriate tool for economics and should be rejected. (Incidentally, both were originally mathematicians.) This is a minority position, and one I disagree with. The view is influential, however. The highest-ranked heterodox economics journal, the Cambridge Journal of Economics, has recently changed its editorial policy to explicitly discourage the use of mathematics. This is a serious mistake in my opinion.

So Noah’s claim about mathematics is a straw man. He implicitly acknowledges this by discussing one class of mathematical Post Keynesian models, the so-called ‘stock-flow consistent’ models (SFC). He rightly notes that the name is confusing – any correctly specified closed mathematical macro model should be internally consistent and therefore stock-flow consistent. This is certainly true of DSGE models.

SFC refers to a narrower set of models which incorporate detailed modelling of the ‘plumbing’ of the financial system alongside traditional macro Keynesian behavioural assumptions – and reject the standard inter-temporal optimising assumptions of DSGE macro. Marc Lavoie, who originally came up with the name, admits it is misleading and, with hindsight, a more appropriate name should have been chosen. But names stick, so SFC joins a long tradition of badly-named concepts in economics such as ‘real business cycles’ and ‘rational expectations’.

Noah claims that ‘vague ideas can’t be tested against the data and rejected’.  While the characterisation of all heterodox economics as ‘vague ideas’ is another straw man, the falsifiability point is important. As Noah points out, ‘One of mainstream macro’s biggest failings is that theories that don’t fit the data continue to be regarded as good and useful models.’ He also notes that big SFC models have so many parameters that they are essentially impossible to fit to the data.

This raises an important question about what we want economic models to do, and what the criteria should be for acceptance or rejection. The belief that models should provide quantitative predictions of the future has been much too strongly held. Economists need to come to terms with the reality that the future is unknowable – no model will reliably predict the future. For a while, DSGE models seemed to do a reasonable job. With hindsight, this was largely because enough degrees of freedom were added when converting them to econometric equations that they could do a reasonably good job of projecting past trends forward, along with some mean reversion.  This predictive power collapsed totally with the crisis of 2008.

Models then should be seen as ways to gain insight over the mechanisms at work and to test the implications of combining assumptions. I agree with Narayana Kocherlakota when he argues that we need to return to smaller ‘toy models’ to think through economic mechanisms. Larger econometrically estimated models are useful for sketching out future scenarios – but the predictive power assigned to such models needs to be downplayed.

So the question is then – what are the correct assumptions to make when constructing formal macro models? Noah argues that Post Keynesian models ‘don’t take human behaviour into account – the equations are typically all in terms of macroeconomic aggregates – there’s a good chance that the models could fail if policy changes make consumers and companies act differently than expected’

This is of course Robert Lucas’s critique of structural econometric modelling. This critique was a key element in the ‘microfoundations revolution’ which ushered in the so-called Real Business Cycle models which form the core of the disastrous DSGE research programme.

The critique is misguided, however. Aggregate behavioural relationships do have a basis in individual behavour. As Bob Solow puts it:

The original impulse to look for better or more explicit micro foundations was probably reasonable. It overlooked the fact that macroeconomics as practiced by Keynes and Pigou was full of informal microfoundations. … Generalizations about aggregative consumption-saving patterns, investment patterns, money-holding patterns were always rationalized by plausible statements about individual – and, to some extent, market-behavior.

In many ways, aggregate behavioural specifications can make a stronger claim to be based in microeconomic behaviour than the representative agent DSGE models which came to dominate mainstream macro. (I will expand on this point in a separate blog.)

Mainstream macro has reached the point that only two extremes are admitted: formal, internally consistent DSGE models, and atheoretical testing of the data using VAR models. Anything in between – such as structural econometric modelling – is rejected. As Simon Wren-Lewis has argued, this theoretical extremism cannot be justified.

Crucial issues and ideas emphasised by heterodox economists were rejected for decades by the mainstream while it was in thrall to representative-agent DSGE models. These ideas included the role of income distribution, the importance of money, credit and financial structure, the possibility of long-term stagnation due to demand-side shortfalls, the inadequacy of reliance on monetary policy alone for demand management, and the possibility of demand affecting the supply side. All of these ideas are, to a greater or lesser extent, now gradually becoming accepted and absorbed by the mainstream – in many cases with no acknowledgement of the traditions which continued to discuss and study them even as the mainstream dismissed them.

Does this mean that there is a fully-fledged ‘heterodox economics’ waiting in the wings waiting to take over from mainstream macro? It depends what is meant – is there complete model of the economy sitting in a computer waiting for someone to turn it on? No – but there never will be, either within the mainstream or outside it. But Lavoie argues,

if by any bad luck neoclassical economics were to disappear completely from the surface of the Earth, this would leave economics utterly unaffected because heterodox economics has its own agenda, or agendas, and its own methodological approaches and models.

I think this conclusion is too strong – partly because I don’t think the boundary between neoclassical economics and heterodox economics is as clear as some claim. But it highlights the rich tradition of ideas and models outside of the mainstream – many of which have stood the test of time much better than DSGE macro. It is time these ideas are acknowledged.

Economics: science or politics? A reply to Kay and Romer

Romer’s article on ‘mathiness’ triggered a debate in the economics blogs last year. I didn’t pay a great deal of attention at the time; that economists were using relatively trivial yet abstruse mathematics to disguise their political leanings didn’t seem a particularly penetrating insight.

Later in the year, I read a comment piece by John Kay on the same subject in the Financial Times. Kay’s article, published under the headline ‘Economists should keep to the facts, not feelings’, was sufficiently cavalier with the facts that I felt compelled to respond. I was not the only one – Geoff Harcourt wrote a letter supporting my defence of Joan Robinson and correcting Kay’s inaccurate description of her as a Marxist.

After writing the letter, I found myself wondering why a serious writer like Kay would publish such carelessly inaccurate statements. Following a suggestion from Matteus Grasselli, I turned to Romer’s original paper:

Economists usually stick to science. Robert Solow was engaged in science when he developed his mathematical theory of growth. But they can get drawn into academic politics. Joan Robinson was engaged in academic politics when she waged her campaign against capital and the aggregate production function …

Solow’s mathematical theory of growth mapped the word ‘capital’ onto a variable in his mathematical equations, and onto both data from national income accounts and objects like machines or structures that someone could observe directly. The tight connection between the word and the equations gave the word a precise meaning that facilitated equally tight connections between theoretical and empirical claims. Gary Becker’s mathematical theory of wages gave the words ‘human capital’ the same precision …

Once again, the facts appear to have fallen by the wayside. The issue at the heart of the debates involving Joan Robinson, Robert Solow and others is whether it is valid to  represent a complex macroeconomic system (such as a country) with a single ‘aggregate’ production function. Solow had been working on the assumption that the macroeconomic system could be represented by the same microeconomic mathematical function used to model individual firms. In particular, Solow and his neoclassical colleagues assumed that a key property of the microeconomic version – that labour will be smoothly substituted for capital as the rate of interest rises – would also hold at the aggregate level. It would then be reasonable to produce simple macroeconomic models by assuming a single production function for the whole economy, as Solow did in his famous growth model.

Joan Robinson and her UK Cambridge colleagues showed this was not true. They demonstrated cases (capital reversing and reswitching) which contradicted the neoclassical conclusions about the relationship between the choice of technique and the rate of interest. One may accept the assumption that individual firms can be represented as neoclassical production functions, but concluding that the economy can then also be represented by such a function is a logical error.

One important reason is that the capital goods which enter production functions as inputs are not identical, but instead have specific properties. These differences make it all but impossible to find a way to measure the ‘size’ of any collection of capital goods. Further, in Solow’s model, the distinction between capital goods and consumption goods is entirely dissolved – the production function simply generates ‘output’ which may either be consumed or accumulated. What Robinson demonstrated was that it was impossible to accurately measure capital independently of prices and income distribution. But since, in an aggregate production function, income distribution is determined by marginal productivity – which in turn depends on quantities – it is impossible to avoid arguing in a circle . Romer’s assertion of a ‘tight connection between the word and the equations’ is a straightforward misrepresentation of the facts.

The assertion of ‘equally tight connections between theoretical and empirical claims’, is likewise misplaced. As Anwar Shaikh showed in 1974, is it straightforward to demonstrate that Solow’s ‘evidence’ for the aggregate production function is no such thing. In fact, what Solow and others were testing turned out to be national accounting identities. Shaikh demonstrated that, as long as labour and capital shares are roughly constant – the ‘Kaldor facts’ – then any structure of production will produce empirical results consistent with an aggregate Cobb-Douglas production function. The aggregate production function is therefore ‘not even wrong: it is not a behavioral relationship capable of being statistically refuted’.

As I noted in my letter to the FT, Robinson’s neoclassical opponents conceded the argument on capital reversing and reswitching: Kay’s assertion that Solow ‘won easily’ is inaccurate. In purely logical terms Robinson was the victor, as Samuelson acknowledged when he wrote, ‘If all this causes headaches for those nostalgic for the parables of neoclassical writing, we must remind ourselves that scholars are not born to live an easy existence. We must respect, and appraise, the facts of life.’

What matters, as Geoff Harcourt correctly points out, is that the conceptual implications of the debates remain unresolved. Neoclassical authors, such as Cohen and Harcourt’s co-editor, Christopher Bliss, argue that the logical results,  while correct in themselves, do not undermine marginalist theory to the extent claimed by (some) critics. In particular, he argues, the focus on capital aggregation is mistaken. One may instead, for example, drop Solow’s assumption that capital goods and consumer goods are interchangeable: ‘Allowing capital to be different from other output, particularly consumption, alters conclusions radically.’ (p. xviii). Developing models on the basis of disaggregated optimising agents will likewise produce very different, and less deterministic, results.

But Bliss also notes that this wasn’t the direction that macroeconomics chose. Instead, ‘Interest has shifted from general equilibrium style (high-dimension) models to simple, mainly one-good models … the representative agent is now usually the model’s driver.’ Solow himself characterised this trend as ‘dumb and dumber in macroeconomics’. As the great David Laidler – like Robinson, no Marxist –  observes, the now unquestioned use of representative agents and aggregate production functions means that ‘largely undiscussed problems of capital theory still plague much modern macroeconomics’.

It should by now be clear that the claim of ‘mathiness’ is a bizarre one to level against Joan Robinson: she won a theoretical debate at the level of pure logic, even if the broader implications remain controversial. Why then does Paul Romer single her out as the villain of the piece? – ‘Where would we be now if Solow’s math had been swamped by Joan Robinson’s mathiness?’

One can only speculate, but it may not be coincidence that Romer has spent his career constructing models based on aggregate production functions – the so called ‘neoclassical endogenous growth models’ that Ed Balls once claimed to be so enamoured with. Romer has repeatedly been tipped for the Nobel Prize, despite the fact that his work doesn’t appear to explain very much about the real world. In Krugman’s words ‘too much of it involved making assumptions about how unmeasurable things affected other unmeasurable things.’ So much for those tight connections between theoretical and empirical claims.

So where does this leave macroeconomics? Bliss is correct that the results of the Controversy do not undermine the standard toolkit of methodological individualism: marginalism, optimisation and equilibrium. Robinson and her colleagues demonstrated that one specific tool in the box – the aggregate production function – suffers from deep internal logical flaws. But the Controversy is only one example of the tensions generated when one insists on modelling social structures as the outcome of adversarial interactions between  individuals. Other examples include the Sonnenschein-Mantel-Debreu results and Arrow’s Impossibility Theorem.

As Ben Fine has pointed out, there are well-established results from the philosophy of mathematics and science that suggest deep problems for those who insist on methodological individualism as the only way to understand social structures. Trying to conceptualise a phenomenon such as money on the basis of aggregation over self-interested individuals is a dead end. But economists are not interested in philosophy or methodology. They no longer even enter into debates on the subject – instead, the laziest dismissals suffice.

But where does methodological individualism stop? What about language, for example? Can this be explained as a way for self-interested individuals to overcome transaction costs? The result of this myopia, Fine argues, is that economists ‘work with notions of mathematics and science that have been rejected by mathematicians and scientists themselves for a hundred years and more.’

This brings us back to ‘mathiness’. DeLong characterises this as ‘restricting your microfoundations in advance to guarantee a particular political result and hiding what you are doing in a blizzard of irrelevant and ungrounded algebra.’ What is very rarely discussed, however, is the insistence that microfounded models are the only acceptable form of economic theory. But the New Classical revolution in economics, which ushered in the era of microfounded macroeconomics was itself a political project. As its leading light, Nobel-prize winner Robert Lucas, put it, ‘If these developments succeed, the term “macroeconomic” will simply disappear from use and the modifier “micro” will become superfluous.’ The statement is not greatly different in intent and meaning from Thatcher’s famous claim that ‘there is no such thing as society’. Lucas never tried particularly hard to hide his political leanings: in 2004 he declared, ‘Of the tendencies that are harmful to sound economics, the most seductive, and in my opinion the most poisonous, is to focus on questions of distribution.’ (He also declared, five years before the crisis of 2008, that the ‘central problem of depression-prevention has been solved, for all practical purposes, and has in fact been solved for many decades.’)

As a result of Lucas’ revolution, the academic economics profession purged those who dared to argue that some economic phenomena cannot be explained by competition between selfish individuals. Abstract microfounded theory replaced empirically-based macroeconomic models, despite generating results which are of little relevance for real-world policy-making. As Simon Wren-Lewis puts it, ‘students are taught that [non-microfounded] methods of analysing the economy are fatally flawed, and that simulating DSGE models is the only proper way of doing policy analysis. This is simply wrong.’

I leave the reader to decide where the line between science and politics should be drawn.