The mainstream economics profession is not saying much during the crisis apart from some of the notable interventions from conservatives and a few not-so conservative economists. In general, what can they say? Not much at all. The frameworks they use to reason with are deeply flawed and bear no relation at the macroeconomic level to the operational realities of modern monetary economies. Even the debt-deleveraging (progressives) use such stylised models which negate stock-flow consistency that their ability to capture sensible policy options are limited. This blog discusses New Keynesian theory which is a current fad among mainstream economists and which has been defended strongly by one of its adherents in a recent attack on Paul Krugman. The blog is a bit pointy.
In our recent book Full employment abandoned, we have a section on the so-called new Keynesian (NK) models and we argue that they are the latest denial of involuntary unemployment, in a long list of efforts that the mainstream has offered over the years. Each effort fails but they never give up.
I was reminded of New Keynesian economics this week when I read the vituperative reply to Paul Krugman’s New York Times article How Did Economists Get It So Wrong? The attack on Krugman came from prominent Chicago University New Keynesian John Cochrane. Krugman’s article came out on September 2 while the reply was dated September 12, 2009.
I may write a specific blog about this dispute but many readers have asked me to comment on where New Keynesian models might fit into modern monetary theory of macroeconomics. So I thought I might briefly provide some ideas on that theme in this blog.
The short answer to the question is: Nowhere. New Keynesian models are irrelevant to anything useful.
But I am sure you will want some substance to back up that conclusion. So here it is.
Cochrane opens with this salvo. He says of Krugman’s piece that:
Most of all, it’s sad. Imagine this weren’t economics for a moment. Imagine this were a respected scientist turned popular writer, who says, most basically, that everything everyone has done in his field since the mid 1960s is a complete waste of time. Everything that fills its academic journals, is taught in its PhD programs, presented at its conferences, summarized in its graduate textbooks, and rewarded with the accolades a profession can bestow, including multiple Nobel prizes, is totally wrong. Instead, he calls for a return to the eternal verities of a rather convoluted book written in the 1930s, as taught to our author in his undergraduate introductory courses. If a scientist, he might be a global-warming skeptic, an AIDS-HIV disbeliever, a stalwart that maybe continents don’t move after all, or that smoking isn’t that bad for you really.
If you think about it modern monetary theory is essentially making the same statements about mainstream macroeconomics as Cochrane constructs Krugman as saying. Most of the macroeconomics written about, taught in universities, thought about is inapplicable to a fiat monetary system. It is a concoction of classical theory of value and prices, late C19th marginal theory and monetary theory, and more recent add-ons (Cochrane calls them “frictions” – to the free market models of the classics). Gold Standard reasoning (or the convertibility that followed) is deeply embedded in the framework.
It assumes a government budget constraint works as an ex ante financial constraint on governments analogous the textbook microeconomic consumer who faces a spending constraint dictated by known revenue and/or capacity to borrow. It then imposes on this fallacious construction a range of assumptions (read: assertions – assumptions can usually be empirically refuted) about the behaviour of individuals in the system and generally concludes, that even with frictions slowing up market adjustments, free market-like outcomes will prevail unless government distortions are imposed.
That body of analysis and teaching is a disgrace and I would hardly judge the veracity of an idea by the fact that the proponent holds a Nobel Prize in Economics. The awarding institution is hardly an unbiased arbiter of truth and reason.
As an aside, if you go to the Nobel Prize home page you will see the following headings – Nobel Prize in Physics, Nobel Prize in Chemistry, Nobel Prize in Medicine, Nobel Prize in Literature, Nobel Peace Prize, Prize in Economics.
Did you spot the missing attribution in the case of Economics? To understand why you have to go back to the original arrangements that were made in the will of Alfred Nobel. Here is what the organisation tells us:
In his last will and testament, Alfred Nobel specifically designated the institutions responsible for the prizes he wished to be established: The Royal Swedish Academy of Sciences for the Nobel Prize in Physics and Chemistry, Karolinska Institute for the Nobel Prize in Physiology or Medicine, the Swedish Academy for the Nobel Prize in Literature, and a Committee of five persons to be elected by the Norwegian Parliament (Storting) for the Nobel Peace Prize. In 1968, the Sveriges Riksbank established the Sveriges Riksbank Prize in Economics in Memory of Alfred Nobel. The Royal Swedish Academy of Sciences was given the task to select the Economics Prize Laureates starting in 1969.
So the economics award was not even part of the original deal but came from the Swedish central bank who must have been upset that my profession was considered less worthy. The fact is that the economics profession is not worthy of this sort of accolade given the state of its theorising and worth to humanity. Mainstream economists hinder human potential and human progress although they allow some individuals to become extremely wealthy at the expense of millions of others.
Anyway, Cochrane’s last sentence above is classic mis-association. The mainstream has never come to terms with Keynes’ General Theory (that is the “rather convoluted book” he is referring to above). While I do not think much of the General Theory, it does successfully expose the irreconcilable flaws in the existing macroeconomic theory of the day (1930s). I should add that the elements of this discredited theory now form the mainstream core of economic reasoning again.
But then to associate that with global-warming denial, AIDs-denial, etc is poor logic and is suggestive of worse to come. And it surely comes.
Cochrane then continues:
Most of the article is just a calumnious personal attack on an ever-growing enemies list, which now includes “new Keyenesians” such as Olivier Blanchard and Greg Mankiw. Rather than source professional writing, he plays gotcha with out-of-context second-hand quotes from media interviews. He makes stuff up, boldly putting words in people’s mouths that run contrary to their written opinions. Even this isn’t enough: he adds cartoons to try to make his “enemies” look silly, and puts them in false and embarrassing situations. He accuses us literally of adopting ideas for pay, selling out for “sabbaticals at the Hoover institution” and fat “Wall street paychecks.” It sounds a bit paranoid.
Okay, without going word-for-word against Cochrane I thought I would show you, by examining the “source professional writing”, why New Keynesian models are not worth considering. Some of what follows is at the pointy end of my blogs. Some of it is taken from my recent book with Joan Muysken noted in the introduction.
I have already written several blogs showing the association between the 1994 OECD policy agenda and the abstract and flawed NAIRU models developed by various economists in the later 1980s and beyond. You will find a host of links HERE
The OECD Jobs Study articulated a microeconomic reform agenda aimed at the supply-side of the labour market based on the false presumption that unemployment was a attribute of individual failure (poor attitutes or skills and policy distortions) rather than a systemic failure (not enough jobs due to deficient aggregate demand).
The accompanying macroeconomic policies that emerged in the 1990s (after the monumental failure of the Milton Friedman-inspired Monetarist “monetary targetting” experiment in the 1980s), were in the form of inflation targeting which have seen monetary authorities narrowly focusing on inflation and largely ignoring the consequences of this obsession for the real economy.
In taking such a narrow view of macroeconomic policy governments have eschewed the use of fiscal policy as the best weapon for reducing unemployment. They have increasingly advocated the virtues of budget surpluses, even if in some cases, cyclical events have proven their views to be wrong.
The NAIRU paradigm that was laid out in the late 1980s has dominated this area of economic literature and provided the authority to policy makers to pursue the supply side activism. However, the mounting empirical anomalies and theoretical critiques seriously dented its image of respectability within the mainstream profession, particularly in the USA.
However, the orthodox economics paradigm has shown considerable flexibility when confronted with empirical anomaly, somewhat like the Lernean Hydra.
You might find David Gordon’s 1972 book Theories of poverty and underemployment; orthodox, radical, and dual labor market perspectives (Lexington Books) interesting – he traces how orthodoxy keeps reinventing itself when confronted with an anomaly that exposes the theoretical structure to rejection. It is one of my favourite books from my student and postgrad days.
In this context, while the NAIRU paradigm has struggled to survive the policy failures that it had motivated (persistent unemployment and rising poverty), a new theoretical edifice, the NK approach, has emerged.
The NK approach has provided solace to an orthodoxy that continues to deny the existence of involuntary unemployment and instead wishes to reassert the flawed prognostications embedded in Quantity Theory and Says Law.
The NK approach is the most recent orthodox effort to attempt reconciliation between macroeconomic theory and what is alleged to be microeconomic rigour (markets clear to give optimal outcomes based on decentralised decisions by rational and maximising individuals).
I note that in general, the literature that has aimed to develop “microeconomic underpinnings” of extant macroeconomic theory, typically aims to hijack any non-orthodox macroeconomic ideas back into the orthodox market-clearing, long-run neutral framework. The money neutrality framework asserts that in the long-run fiscal policy has no real benefits but causes inflation. It is a highly flawed framework.
The NK theory is a quintessential expression of this tradition. Importantly, from a policy perspective, the NK approach is also the most recent theoretical structure to be co-opted by orthodox policy makers to justify inflation targeting.
Despite the fact the NK approach is fast becoming an industry in academic and policy making circles it has received very little critical scrutiny in the literature.
Fellow Post Keynesian and modern money sympathiser, Marc Lavoie’s 2006 article (‘A post-Keynesian amendment to the new consensus on monetary policy’, Metroeconomica, 57(2), 165-192) is an exception.
There are three major conclusions to be drawn from this literature:
- The so-called microfoundations of New Keynesian models are not as robust as the various authors would like to claim;
- The so-called Keynesian content of the models should be taken with a grain of salt;
- The rationale these models provide to justify their claim that tight inflation control leads to minimal labour market disruption is highly contestable. The only reasonable conclusion is that the approach has no credibility in dealing with the issue of unemployment and cannot reasonably be used to justify aggregate policy settings.
But to understand these conclusions you need to examine the approach in more detail.
New Keynesian models
The NK approach has provided the basis for a new consensus emerging among orthodox macroeconomists. A typical representation of this approach is found in Carlin and Soskice’s 2006 book Macroeconomics: imperfections, institutions and policies, published by Oxford University Press, where you read in their preface that:
Consensus in macroeconomics has often been elusive but the common ground is much wider now than has been the case in previous decades … There is broad agreement that a fully satisfactory macroeconomic model should be based on optimizing behaviour by micro agents, that individual behaviour should satisfy rational expectations and that the model should allow for wage and price rigidities … The three equations … [summarising the model] … are derived from explicit optimizing behaviour on the part of the monetary authority, price setters, and households in imperfect product and labour markets and in the presence of some nominal rigidities.
NK theory thus attempts to merge the so-called Keynesian elements of money, imperfect competition and rigid prices with the real business cycle theory elements of rational expectations, market clearing and optimisation across time, all within a stochastic dynamic model.
Simpflifying NK theory is easy despite its deliberate complexity. I am reminded of a beautiful section in a book by American economist (and Marxist) Paul Sweezy who wrote in 1972 in the Monthly Review Press an article entitled Towards a Critique of Economics.
He argued that orthodoxy (mainstream) economics in recent times had:
… remained within the same fundamental limits” of the C19th century free market economist. He said they had “therefore tended … to yield diminishing returns. It has concerned itself with smaller and decreasingly significant questions … To compensate for this trivialisation of content, it has paid increasing attention to elaborating and refining its techniques. The consequence is that today we often find a truly stupefying gap between the questions posed and the techniques employed to answer them.
He then cites a wonderful example of mainstream written reasoning which the modern NK economists would be proud off. It is taken from Debreu’s 1966 mimeo on Preference Functions. Here is is for some light relief:
Given as set of economic agents and a set of coalitions, a non-empty family of subsets of the first set closed under the formation of countable unions and complements, an allocation is a countable additive function from the set of coalitions to the closed positive orthant of the commodity space. To describe preferences in this context, one can either introduce a positive, finite real measure defined on the set of coalitions and specify, for each agent, a relation of preference-or-indifference on the closed positive orthant of the commodity space, or specify, for each coalition, a relation of the preference-or-indifference on the set of allocations. This article studies the extent to which these two approachas are equivalent.
In my so-called economics “education” I have read countless articles like this one – saying nothing about anything that will be of any benefit to humanity. I like playing chess. I always thought of my so-called education in economics with all the mathematics that came with it to be playing chess although a boring version.
Anyway, NK has all this sort of complexity and obtuseness and more.
But we can simplify it to three basic equations that purport to capture the essential nature of the economic system.
First, the New Keynesian IS equation. This is the relationship that brings investment (I) and saving (S) together to ensure there is full capacity utilisation in the long-run (so Says Law holds).
What is not always recognised is that in most NK models, the micro foundations of the IS curve allow neither savings nor investment to play any role. This is usually motivated by the fact that in real business cycle models the capital stock is typically ignored because any flux in investment and resulting changes to the stock of productive capital to externally imposed “productivity shocks” (which are the way RBC theories claim business cycles occur) actually has little bearing on the dynamics of their models.
As a consequence of this glaring omission, the so-called intertemporal IS relation (the across time conjunction between investment and saving) is derived using intertemporal utility maximising behaviour by consumers, who face a trade-off between consumption and leisure.
The nominal rate of interest then equates the nominal intertemporal marginal rates of substitution in consumption, such that consumption can be smoothed out over an individuals’ life time.
It is assumed that individuals can always borrow and lend at the prevailing interest rate to implement their life-time consumption plan. Thus, while savings and investment may take place at the individual level, they are assumed to cancel out at the aggregate level because all income is assumed to be consumed. So there are no capital market constraints on anyone.
To be consistent with this approach, bonds are issued for one-period only and the role of money in this approach is only to facilitate transactions. In other words, the NK approach takes us back to the pre-Keynes, Quantity Theory era where money is used only as a means of payment and a unit of account. Classic contributions in this regard come from Buiter (2006 ‘The elusive welfare economics of price stability as a monetary policy objective. Why new Keynesian central bankers should validate core inflation’, ECB Working Paper Series, no 609) and Woodford (2006 ‘How important is money in the conduct of monetary policy?’, Department of Economics Working Papers No 1104, Queens University, Canada).
The reality is that none of the NK models handle money in a way that remotely corresponds to the dynamics and operational realities of a modern monetary economy based around a fiat currency.
Getting pointy, the NK IS relation is derived from an approximation of the Euler condition for intertemporal optimal consumption around a zero-inflation steady state. It is usually presented in terms of deviations from natural levels and implicitly defines the stabilising interest rate – that is, the Wicksellian natural rate of interest – as the rate r* that equates aggregate demand to the natural level of output y*.
So the Austrian influence creeps in here.
Problematic is the fact that when there is a permanent demand shift in this model r* changes which means that as a consequence a temporary demand shock, has a different impact compared to a permanent demand shock, since the latter leads to a change in r* and hence has an impact on monetary policy (Marc Lavoie, 2006 makes this point too).
Second, the New Keynesian Phillips curve is important. The NK Phillips curve bears a close resemblance to the Expectations Augmented Phillips curve, the latter being based on natural rate theory inherited from Friedman and Phelps. So there is no long-run trade-off between inflation and unemployment (and maybe very little trade-off in the short-run) and any attempts by the government to use fiscal policy to enforce a trade off if they consider unemployment is too high will only cause inflation.
However, as a result of the NK Phillips curve being derived from so-called optimising behaviour, its coefficients have a specific interpretation. Firms are assumed to employ so-called Calvo price-setting, which has become the standard NK approach.
Accordingly, under monopolistic competition only a fraction of firms set their prices in the current period. The remainder of firms keep their price at the level of the previous period. Optimal consumer and producer behaviour implies that the (log of) the deviation of marginal costs plus mark-up on prices from its normal level is proportionally related to the (log of) deviation of output from its natural level.
All this means is that there are adjustment lags imposed on the normal natural rate story and which allow short-run trade-offs between inflation and unemployment to occur.
But Calvo price-setting does not allow lagged inflation to influence current inflation which was basic in the original Friedman conception. Carlin and Soskice (2006: 608) aptly observed the:
NKPC brings back rational expectations into the inflationary process, but it throws out the baby (the empirical fact of inflation inertia) with the bath water of non-rationality.
As a result of this anomaly, ad hocery enters the fray. NKs quickly recognised that in the applied world of macroeconomics there is usually a lagged dependence between output and inflation taken into account. The primary justification is empirical.
But trying to build this in to their model from the first principles that they start with is virtually impossible. No NK economist has picked up this challenge, and instead they just introduce lagged inflation anyway. So like most of the mainstream body of theory they claim virtue based on so-called microeconomic rigour but respond to anomalies that are pointed out when that “rigour” fails to deliver anything remotely consistent with reality, with ad hoc (non rigourous) tack ons.
So at the end of the process there is no rigour at all – using rigour in the way they use it which is, as an aside, not how I would define tight analysis.
Anyway, it follows from the (ad hoc) model that if you stabilise inflation then you automatically stabilise the output gap. The proponents of the NK approach claim virtue from this constructed logic by asserting that this outcome is also efficient from a welfare perspective because their model is underpinned with optimising microfoundations.
But while Blanchard, O. and J. Gali (2005) (‘Real wage rigidities and the new Keynesian model’, NBER Working Paper Series, no 11806) claimed that this was a “divine coincidence” in the NK model it only occurs as a result of the absence of imperfections. Blanchard and Gali (2005: 3) then stated that the:
… optimal design of macroeconomic policy depends very much on the interaction between real perfections and shocks … [and] … Understanding these interactions should be high on macroeconomists’ research agendas.
So you quickly get the pattern of reasoning. When confronted with a problem the NK economists bring out another ad hoc solution as was proposed by Blanchard and Gali (2005) in the form of real wage rigidities, which clearly also eliminates the divine coincidence at the same time.
Third, the New Keynesian monetary rule completes their system.
Without attempting to understand how central banks actually operate, New Keynesians derive their monetary rule (which is just an interest rate setting reaction function) by assuming that the central bank minimises a loss function in which both deviations of inflation from its target value and deviations of output from its natural level play a role, subject to the Phillips curve discussed above.
They also assume that the central bank can control aggregate demand using the interest rate, through the IS relationship. So they have a sort of Taylor rule such as the real interest rate set equals the natural interest rate plus some function of the inflation gap plus some function of the output gap (output out of sync with the potential).
So the models are always represented in real terms, whereas the central bank can only set the nominal interest rate. To get around that problem they presume that the central bank can observe both the natural output level and the natural rate of interest correctly.
It follows analytically that in the NK models, the central bank will only achieve its target inflation rate when it correctly estimates both the natural output and the natural rate of interest levels correctly – a tall order one would suspect.
The deficiencies of the New Keynesian approach
The alleged advantage of the NK approach is the integration of real business cycle theory elements (intertemporal optimisation, rational expectations, and market clearing) into a stochastic dynamic macroeconomic model.
But it is obvious that notwithstanding the air of rigour, the NK results are still always conjunctions of abstract starting assumptions and ad hoc additions to make any traction with reality.
This indicates an important weakness of the NK approach. The mathematical solution of the dynamic stochastic models as required by the rational expectations approach forces a highly simplified specification in terms of the underlying behavioural assumptions as we have already indicated in our description of the standard model.
But the ability of these models to say anything about the actual operations of central banks is severely compromised by the highly simplistic behavioural assumptions employed, notwithstanding Friedman’s long-standing appeal to empiricism.
The empirical credibility of the abstract NK models is questionable. This holds, in particular, for the NK Phillips curve and its potential to represent real world inflation dynamics.
In their survey of the literature on inflation dynamics in the US economy, Rudd and Whelan (page 4) observed:
… the data actually provide very little evidence of an important role for rational forward-looking behavior of the sort implied by these models. (Rudd, J. and K. Whelan (2005) ‘Modelling inflation dynamics: a critical review of recent research’, FEDS working papers¸ no 2005-66)
Further, after finding similar results for the Euro area, Paloviita (page 858) concluded that the
… results obtained suggest that NKPC can capture inflation dynamics in the euro area if the rational expectations hypothesis is not imposed and inflation expectations are measured directly – we find evidence that lagged inflation seems to be needed to properly explain the persistence of European inflation. (Paloviita, M. (2006) ‘Inflation dynamics in the euro area and the role of expectations’, Empirical Economics, 31, 847-860)
There are many similar studies that have exposed these sort of weaknesses in NK models.
Clearly, the claimed theoretical robustness of the NK models has to give way to empirical fixes, which leave the econometric equations indistinguishable from other competing theoretical approaches where inertia is considered important. And then the initial authority of the rigour is gone anyway.
This general ad hoc approach to empirical anomaly cripples the NK models and strains their credibility. When confronted with increasing empirical failures, proponents of NK models have implemented these ad hoc amendments to the specifications to make them more realistic. I could provide countless examples which include studies of habit formation in consumption behaviour; contrived variations to investment behaviour such as time-to-build , capital adjustment costs or credit rationing.
But the worst examples are those that attempt to explain unemployment. Various authors introduce labour market dynamics and pay specific attention to the wage setting process. One should not be seduced by NK models that include real world concessions such as labour market frictions and wage rigidities in their analysis. Their focus is predominantly on the determinants of inflation with unemployment hardly being discussed (for example, Blanchard and Gali, 2005).
Of-course, the point that the NK authors appear unable to grasp is that these ad hoc additions, which aim to fill the gaping empirical cracks in their models, also compromise the underlying rigour provided by the assumptions of intertemporal optimisation and rational expectations.
The NK approach is another program of theoretical work designed to justify orthodox approaches to macroeconomic policy, in this case the virtues of inflation targeting. In the orthodox tradition, it also denies the existence of involuntary unemployment. However, it categorically fails to integrate its theoretical structure with empirical veracity.
So after appreciating that, you will be better placed to read and dismiss Cochrane’s response to Krugman. His characterisation of what economists have been doing for the last 30 years is as follows:
Macroeconomists have not spent 30 years admiring the eternal verities of Kydland and Prescott’s 1982 paper. Pretty much all we have been doing for 30 years is introducing flaws, frictions and new behaviors, especially new models of attitudes to risk, and comparing the resulting models, quantitatively, to data. The long literature on financial crises and banking which Krugman does not mention has been doing exactly this bidding for the same time.
The literature that has been produced bears no relation to the modern monetary economies that most of us live in and also categorically failed to see the crisis coming.
A friend calls this literature La-La land. Not a bad description.