Wednesday, 25 June 2014

Wren-Lewis and Portes versus MMT.

This recent paper by Simon Wren-Lewis and Jonathan Portes is defective. It is entitled “Issues in the design of fiscal policy rules”. Plus the paper is according to the authors a “discussion paper”: so here’s my inspired(?) contribution to the “discussion”.
They attempt to work out a simple rule to govern the amount of fiscal stimulus suitable in a recession. As the authors put it, there is “no simple rule to guide fiscal policymakers” and “this paper is about the search for” such a rule. Indeed, they claim that such a rule may never be found.
The truth is that MMTers thought up a rule long ago which has all the simplicity of E-MC2. The MMT rule is also advocated by Positive Money and Prof. Richard Werner.
The MMT rule is along the following lines (in green).
In a recession, the government / central bank machine should run a deficit by creating and net-spending enough new money to bring the recession to an end. The term “net-spend” simply alludes to the fact that the deficit can be increased by raising government spending and/or cutting taxes, a choice which is PURELY POLITICAL. I.e. unlike many economists, MMTers and Positive Money supporters are able to distinguish between ECONOMIC and POLITICAL matters.
Or if you want the MMT rule in fewer words, there is Warren Mosler’s version, his so called “Mosler’s Law” which says “There is no financial crisis so deep that a sufficiently large tax cut or spending increase cannot deal with it.” (See sentence in yellow at the top of Warren’s site.)
Other supporters (sort of) of the MMT rule include Keynes and Milton Friedman. Keynes in a letter to Roosevelt in the 1930s said that a way of dealing with recessions is simply to have government print and spend. See 5th paragraph here. And Friedman supported the MMT policy IN THAT he argued that government should not incur any debt. On the other hand, Friedman was an opponent of DISCRETIONARY government intervention to deal with recessions (i.e. adjusting government net-spending so as to deal with recessions), so to that extent, he disagreed with MMT, and indeed the large majority of present day economists.

Why “smooth” taxes and government spending?
Wren-Lewis and Portes’s paper starts with the claim that taxes should be smoothed. They say “Theory suggests that government should as far as possible smooth taxes and its recurrent consumption spending, which means that government debt should act as a shock absorber, and any planned adjustments in debt should be gradual.”
I smell a self-contradiction. The latter passage seems to say that taxes and government “current consumption” should not be adjusted too violently. Then they say that the only alternative adjustment (the debt) shouldn’t be adjusted too much either. So what are they saying? I’m baffled.
Anyway, the next sentence of the abstract reads, “This suggests that operational targets for governments (e.g. for 5 years ahead) should involve deficits rather than debt, because such rules will be more robust to shocks.”
Now does that actually mean anything? If I say I have a “target” for my household deficit (or surplus), what does that tell you? It doesn’t tell you whether I’m aiming to have the deficit / surplus stay relatively constant or gyrate wildly. And it doesn’t tell you whether I’m aiming for a surplus or a deficit. It tells you nothing.

Taxes are “distortionary”?
Then at the start of their section 2, entitled “Optimum debt policy”, they say “Government debt, which is a stock concept, does not directly influence social welfare. Instead it is the things that influence debt that matter, like taxes and government spending. This leads to perhaps the most fundamental and most well-known principle behind optimal debt policy, which is sometimes called tax smoothing. We can illustrate it in a somewhat stylised way as follows. Suppose social welfare declines as taxes rise, because taxes are distortionary.”
Now hang on: why should taxes be “distortionary”? Obviously taxes CAN BE distortionary, but they don’t HAVE TO BE. For example if government grabbed an extra X% of everyone’s income, that would be very distortion-free. Or if government placed a sales tax on every single item sold, that would also be very distortion-free.
Of course in the real world no set of taxes will be totally distortion-free, but if the authors’ point is to hold, they have to show there’s enough distortion to have a welfare reducing effect of the size they imagine or imply.

Monetary policy is also distortionary.
And even the fact that taxes involve a finite amount of distortion doesn’t get the authors’ argument very far because monetary policy is also distortionary, and for the following reasons.
Suppose resources are being allocated to every type of economic activity in an optimal way. Now suppose a recession occurs because of a fall in consumer confidence. (Incidentally, note that that is not an unrealistic “supposition”: the recent crisis was basically caused by a cut in spending by underwater households according to Mian and Sufi.)
Plus suppose that government deals with that recession by cutting interest rates. Well that’s clearly a move away from the optimum. That is, the recession was caused by a drop in consumer spending, and far from boosting consumer spending, government tries to boost INVESTMENT spending. That’s what I call a “distortion”.
And another distortion involved in monetary policy is that those funding monetary policy are the asset rich or cash rich: a small section of the population. To illustrate, if interest rates are cut, lending and investment supposedly rise, and those funding that increased investment are the asset or cash rich.
In short, the authors’ “distortion” argument does not stand inspection, and given that the rest of their paper relies on that “distortion” argument, all subsequent arguments in the paper collapse, far as I can see.

The reasons behind the simple MMT rule.
The MMT rule, to repeat, is simply to print money and net-spend it in a recession. Plus occasionally, and when inflation looms in a serious way, there will be a need to do the opposite, that is raise taxes and “unprint” the money collected.
That rule involves a number of elements which need to be justified, and one is the implicit merging of monetary and fiscal policy. That is, pure fiscal policy, i.e. having government borrow and spend goes out of the window. So what’s the justification for discarding pure fiscal policy? Well it’s simple, and as follows.
Governments normally have no idea what causes recessions. To illustrate and repeat, the recent recession has been attributed to bank irresponsibility and with some justification. But Mian and Sufi have also pointed to excess household borrowing. So how much of the blame lies with banks and how much with households? The answer is: “God knows”.
That uncertainty as to what causes recessions is a good reason for NOT WEIGHTING the response in the direction of either fiscal or monetary policy. That is, the latter uncertainty is a good reason to mix fiscal and monetary policy, which is what the MMT response to recessions involves.

Why not cut interest rates in a recession?
A popular way of dealing with recessions is to cut interest rates (a form of monetary policy). Unfortunately that idea is riddled with flaws and as follows.
First, the EVIDENCE seems to be that interest rate adjustments don’t have much effect. See here and here.
Second, I listed a string of defects in interest rate adjustments here.
Third, Richard Werner listed even more defects here.

PS (26th June 2014). Another reason for not biasing stimulus in the direction of fiscal or monetary policy is that there is much argument as to which is the more effective and quicker to have an effect: monetary or fiscal. Given that uncertainty, why not forget about the latter arguments and uncertainly and just implement fiscal and monetary policy at the same time? That’s guaranteed to work (unless of course both monetary and fiscal stimulus are totally useless, and no economist makes that claim, far as I know).
MMT implies no national debt.
An apparent defect in stimulus MMT style is thus. If the government / central bank machine constantly feeds or spends base money into the private sector, but doesn’t feed government debt into the private sector (which is what fiscal stimulus involves), then thanks to inflation, the national debt will eventually wither to near nothing. And that conflicts with the conventional view that a finite national debt is desirable.
The answer to that is that the merits of national debt are actually very debatable. Milton Friedman advocated a system where there was no debt. See his p.250, para starting “Under the proposal….”. Warren Mosler (a leading MMTer) explicitly opposes any debt. See his 2nd past para. And I threw cold water over the whole national debt idea here.

Keynes supported MMT?

My above claim to that effect is perhaps cheeky. It would be more accurate to say that MMTers stand on the shoulders of a giant called “Keynes”. Indeed, Dean Baker said he couldn’t see the difference between Keynes and MMT.


  1. Great post.
    Please clarify what you mean at the end by "monetary policy" and how it is "merged| with fiscal policy.

    You say that the idea of interest rate changes is "riddled with flaws".
    And we know from recent experience that flooding the economy with liquidity (quantitative easing) is ineffective when interest rates are very low.
    That leaves only elements of monetary policy such as bank regulations, exchange controls, etc. which have only minor effects on demand.

    So perhaps you would agree that fiscal policy is the main tool to be used for demand management? The role of monetary policy is then merely that of keeping interest rates stable or consistent with with international interest rates and exchange rate stability?

    1. Hi KK,
      Should have defined my terms! By “fiscal” I mean government borrows £X, gives bonds worth £X to those it has borrowed from and spends the £X. The private sector’s stock of base money does not change there.
      By “monetary” I mean one or more of the following. 1, interest rate adjustments, 2, the central bank prints money and the CB / government machine buys government debt (as under QE and indeed as under interest rate adjustments), or the “machine” buys other assets from the private sector. In all those scenarios, the private sector’s stock of base money changes.
      Now I bet those definitions aren’t 100% consistent with my above post. Feel free to pick holes in my definitions!
      Re “merge”, what I mean is that the “machine” prints money and spends it on the usual public sector items (health, education, etc) and/ or taxes are cut. In that case the private sector’s stock of base money rises, so that’s the monetary element. Plus there is expenditure on that public sector stuff (or taxes are cut) and that’s “fiscal”.
      Re your last paragraph, I think you are right to suggest there is little role for monetary policy, other than in a supporting role for fiscal policy. That is, as I argued above, if one goes for PURE fiscal policy, there’s the well known “crowding out” problem. That is the deflationary effect of the extra borrowing might negate the stimulatory effect of the extra spending. And there is uncertainty as to how big a problem that is. So my reaction is: “why borrow – i.e. why not just print money instead?”.

  2. "Merge" is a new word to me in this context. Most dictionaries of economics define fiscal policy as what you here call "merge"!
    And what you call "fiscal policy" would more conventionally be regarded a fiscal policy plus monetary policy open market operations.
    If I am right on this, maybe it would help if you (and Positive Money) used conventional terminology.

    However, these definitional confusions don't detract from the excellent substance of your post.


Post a comment.