With Modern Monetary Theory (MMT) making inroads in the public policy debate, some New Keynesians have transitioned from ignoring or dismissing the approach to engaging with it. This is healthy for both sides. There has been a tendency, though, to make “we’ve known it all along” type statements. A comprehensive response to the “nothing new” claims is provided by Bill Mitchell in a recent three-part series of posts (part 1, part 2 and part 3). My focus here is narrower and concerns a view (for example, expressed in a considered response here) along the lines that MMT has nothing new to say when the economy is at full employment.
Leaving aside that MMT and New Keynesian Economics (NKE) define full employment differently, this claim obscures a fundamental difference between the two theories. In NKE, as with neoclassical approaches more generally, there would be an automatic tendency to full employment in the long run if there were no imperfections or rigidities to impede the supposed adjustment process (including, in the case of NKE, no impediment to appropriate interest-setting monetary policy). For example, the favored justification for a fiscal response to unemployment in the wake of the global financial crisis was that the economy was in a liquidity trap, taken to mean that the real rate of interest was stuck above its so-called ‘natural’ rate due to monetary authorities confronting the zero lower bound. Absent this rigidity and any other rigidity (sticky wages, sluggish revision of expectations, etc), NKE implies that there will be full employment in the long run.
MMT is not neoclassical. It comes out of a different, broad theoretical tradition (which includes Post Keynesian economics and related Kalecki and Keynes influenced approaches) in which there is no general tendency for the economy to tend toward full employment even in the absence of imperfections and rigidities. In these theories, complete wage, price and interest-rate flexibility would not imply spontaneous adjustment to full employment over any time frame. Nor could policy administered interest-rate adjustments be relied upon to generate full employment, even when the economy was not at the zero lower bound, in contrast to the conclusion of NKE. The notion of a natural rate actually makes no sense in MMT or in the theoretical tradition from which it emerged.
The rejection of natural rate reasoning, and by extension appeals to a liquidity trap in the sense this term has been used since the global financial crisis, is theoretically informed by the capital debates (Matias Vernengo provides an informative introduction) and various arguments of Keynes (Bill Mitchell has a long series of posts entitled ‘Keynes and the classics’ beginning here). It is also consistent with the paucity of empirical support for a significant inverse relationship between the rate of interest and private investment upon which neoclassical spontaneous adjustment to full employment depends (see, for example, papers by Robert Chirinko, Jamus Lim and Steve Sharp and Gustavo Suarez).
Why is this relevant to the present comparisons between MMT and NKE?
It is relevant because it means the policy implications of MMT are in fact different to NKE over any time frame. In MMT there is no presumption that the economy will be at full employment in the long run. (This is so unless a job guarantee is in place, in which case there will be what MMTers refer to as ‘loose full employment’ at all times. This will be full employment in the sense that anyone who is willing and able to take a job at the job-guarantee wage will be employed.) In the absence of a job guarantee, there is no presumption in MMT that the economy tends toward full employment. An implication is that larger government deficits now do not necessarily imply smaller deficits or higher taxes in the future. It may be that full employment requires deficits in the future just as it does now. True, if it turns out in the future that full employment can be achieved with a smaller deficit, MMT will call for a smaller deficit. But it could go the other way.
If, as MMT prescribes, a job guarantee is put in place, full employment will not even be the relevant benchmark. With a job guarantee implemented, loose full employment will be achieved at all times irrespective of the overall level of demand. The relevant benchmark will then be the “non-acceleraiting buffer employment ratio” (NAIBER), meaning the smallest fraction of total employment that can be located in the job guarantee sector while still maintaining low, stable inflation. On the basis of MMT, there would be no spontaneous tendency for the economy to move toward the NAIBER, just as there is no tendency in the absence of a job guarantee for the economy to move toward full employment.
The premises and logic of MMT lead to the job guarantee as the means of simultaneously achieving and maintaining both full employment and price stability. Unless this is also the conclusion of NKE, based on its premises and logic, it cannot be said that MMT “has nothing new to say at full employment”. Clearly it is not the obvious conclusion of NKE. If it were, New Keynesians would have drawn the conclusion long ago. Within the framework of NKE there need to be rigidities (e.g. a liquidity trap) to justify fiscal demand management. Even under conditions of unemployment, there is no place for fiscal demand management in NKE unless a rigidity can be identified.
It is perhaps also worth noting that a related claim – that MMT can be reduced to NKE with the policy tools switched (with fiscal rather than monetary policy assigned to demand management and monetary rather than fiscal policy geared toward controlling the level of interest on public debt) – is similarly unfounded. The observation so far as it distinguishes MMT’s preferred assignment of policy tools from NKE’s preferred assignment is a good one in my view (discussed previously here). However, the notion that MMT can be characterized simply by assuming within the New Keynesian framework that the interest rate can be geared solely toward keeping interest rates on public debt low relies on monetary authorities being able to choose the interest rate without regard to the so-called natural rate of interest and to do so indefinitely. This notion is unproblematic in MMT since within this framework there is no valid reason for supposing the existence of a natural rate of interest. The interest rate, in the MMT framework, is a policy variable, including in the long run. But in NKE, the existence of a natural rate follows from its neoclassical premises (including its marginalist microfoundations). So to impose, within the New Keynesian framework, the arbitrary restriction that the interest rate must be set according to public debt considerations is not only ad hoc but violates the constraint on policymakers that is implied by the existence of a natural rate, an existence that follows from the premises of the theory.
In short, MMT is not merely a policy position on government deficits that can be faithfully and fully represented by placing arbitrary restrictions on New Keynesian models. It is a theory built upon non-neoclassical foundations in which even under complete wage, price and interest-rate flexibility there would be no tendency for output to move to any particular level independently of fiscal policy, over any time frame. One implication of this – but only one implication – is that there is no reason to suppose that the fiscal stance called for today is in any particular relationship to the fiscal stance called for in the future. Fiscal surpluses today will not necessarily call for tax cuts in the future and fiscal deficits today will not necessarily call for higher tax rates in the future.
The foregoing argument is not intended to deny that there has been considerable common ground between MMT and NKE when it comes to the appropriateness of fiscal deficits in the aftermath of the global financial crisis. Economists of both persuasions have quite rightly been supportive of expansionary fiscal policy. The argument concerns the theoretical justification for this policy stance in the short term and the implications of this policy stance in the long term.
Balancing the Budget Over the Cycle
Institutions, Monetary Operations and a Demand-Led Global Economy
Vernengo on Keynes vs Neoclassical Synthesis
More on Keynes vs the Neoclassical Synthesis