Andrew Lilico: What Jonathan Portes gets wrong and why
Follow Andrew on Twitter.
Jonathan Portes is head of the National Institute of Economic and Social Research, an overtly and devoutly Keynesian thinktank greatly respected by UK economists. He has occasionally been accused of being biased in his economic analysis. That seems to me to be a deeply unfair and confused allegation, since it is NIESR's job to be "biased" if by "biased" you mean promoting a Keynesian macroeconomics perspective. He certainly isn't politically biased - it's not his fault if some on the Left support his views and policy recommendations, and it's worth noting that there have always been devoutly Keynesian right-wingers as well.
On Sunday he put up a blog entitled "The Austerity Delusion?" which challenged what Jonathan believes to be the claim of Fraser Nelson and John Redwood that there has been no "austerity". Jonathan frames his discussion in an "all serious people will agree with me and I'm being generous even giving airtime to the ludicrous ideas of Nelson and Redwood" way, and since he's wrong on every point, I think it's worth my rehearsing for you, Dear Reader, precisely why.
First, Jonathan frames his discussion with a piece of legerdemain. He tells us "it would be useful to have a definition of "fiscal stimulus" and hence of its opposite, fiscal consolidation or "austerity"". Whoah, there, cowboy! The non-technical definition of "austerity" is "n. 1. Sternness or severity of manner or attitude. 2. Extreme plainness and simplicity of style or appearance." In a political context, "austerity" is understood as implying a period of cutbacks of some sort, an age in which people have to simplify their living and seek to repay their debts. Jonathan may say: "Well, I mean "austerity" in a technical economists' sense." But politicians and political journalists are entitled to use the word "austerity" to mean what it means in a political context. And economists would rarely use the word "austerity" in any technical sense. They would use words such as "fiscal consolidation" to mean reducing the deficit. On any ordinary or political definition of "austerity" shifting from the largest fiscal stimulus of all time to merely 99.9% of that same fiscal stimulus would not be intelligently described as "austerity". Yet on Jonathan's definition it would.
Jonathan sets out five claims of Fraser and John that he thinks are wrong. The first is: "There has been no austerity, because the deficit remains large and the debt is rising." Jonathan says about the general case that "Many people will think that they are obviously talking nonsense and can simply be dismissed." It seems to me that, on the contrary, most people, in particular most politicians, political journalists, and especially voters, will think what this says (which I think is the view of John but not Fraser, though to be fair Jonathan is attacking a broader position rather than specifically attributing every argument to these two) obviously right. All the talk of "austerity" has led most voters to think the government is paying off debts, not accumulating more of them at a record pace.
But Jonathan wants to be more technical, and he's wrong on the technical point too. In a standard Keynesian model, of the sort many people involved in politics will be dimly familiar, there is an "injection" provided by the difference between government spending and taxation - the deficit. In such models, government deficit financing makes total GDP bigger, and the bigger the deficit, the bigger GDP. Most people in politics, since the 1930s, have used the terms "Keynesian stimulus" or "Keynesian deficit financing" or "Keynesian policy" to refer to this notion of bigger government deficits meaning bigger GDP. That occurs across the political spectrum - those that argue for a "Keynesian stimulus" almost always mean "running a bigger deficit."
Jonathan asks, incredulous, if Redwood thinks that cutting the deficit from 11% of GDP to 1% would mean that there was still a "Keynesian stimulus", just one of 1% not 11%. Yes, Jonathan, that's exactly what he means, that's exactly what the standard textbook Keynesian theory of the sort one learns at A level or first year university says, that's exactly how the terms have been used in politics for 75 years, and indeed that's exactly the concept behind things like "automatic stabilisers". Now you, Jonathan, may prefer to reserve the term "Keynesian stimulus" to describe something else, but I'm afraid that term's already been bagsied long ago. Similarly, consider this: Would deliberately raising doctor's salaries and increasing the deficit to pay for that be a "fiscal stimulus"? My guess is that Jonathan would say No, but virtually everyone in politics would say Yes.
Redwood is quite entitled to say that we are in a period of huge "Keynesian stimulus", not "austerity" by any normal (or indeed, sane technical) definition of those terms. What he means by that is that we've been running huge deficits, there's a theory that the bigger the deficit, the higher the growth, and recent events are pretty strong evidence that that theory is wrong.
Next, Jonathan challenges the claim that "There has been no austerity, because there have been no spending cuts". But if you define a period of "government austerity" as one in which government debts are paid down (or at the very least no more government debts are accumulated) and government spending is cut - which is a perfectly good definition - then if government spending hasn't been cut, the budget hasn't been balanced, and the deficit is rising not falling, then there isn't "austerity".
Jonathan moves on to the claim that "The government has abandoned its original plan and as a consequence deficit reduction has stalled." He claims the causality runs the other way - deficit reduction stalled and therefore the government abandoned its deficit reduction targets. But there's no economic causality there. The government wasn't obliged to abandon its deficit reduction targets because of stalling deficit reduction by anything other than politics. With full policy freedom, the government could have cut spending faster, to keep to its targets. Deficit reduction is stalling because the government has stopped trying to cut the deficit.
The next "error" to which Jonathan points is the notion that "austerity has failed because it was implemented by tax rises and not spending cuts" which he claims is a variant of the errant assertion that there has been no austerity because spending hasn't been cut. Here Jonathan moves from merely being unfair (as per definitions) or dissembling (as with the "Keynesian stimulus" point) to being outright wrong and at variance with orthodox mainstream theory and evidence. He asserts that "there is no economic logic in suggesting that austerity measures which take money out of people's pockets and reduce confidence in one way (like VAT or rail fare rises) cut demand, while measures which do so in other ways (cuts to tax credits and benefits, cutting EMA and the Future Jobs Fund) don't." That's just plain wrong, and Jonathan should know better.
First, it is well-established both in economic theory and in empirical studies that tax rises can damage growth. High-tax economies grow slower, over the medium-term, than low-tax economies. So as to see how well-established this point is I note that a classic recent mainstream econometric study of this was the 2008 paper "Government size, composition, volatility and economic growth" done for the European Central Bank by Afonso and Furceri, in which each additional percentage point of GDP of tax reduced annual GDP growth by 0.12 percentage points.
By contrast, there is no good economic theory according to which transfers increase aggregate demand, where by a "transfer" I mean a policy that takes money off one person and gives it to another. For example, if I impose a tax on rich people and give the proceeds of that tax to a poor person, there is no mainstream economic theory according to which that will make the economy grow faster. Conversely, there is no mainstream economic theory according to which cutting down on such transfers will make the economy grow slower. There are many such transfers. Benefits are a transfer. Another transfer would be over-paying public sector workers - so they produce the same outputs but get paid more than would be required to do so; that's just a transfer from non-public sector workers to public sector workers. There isn't the slightest reason in economic theory to believe that reducing over-payment of public sector salaries, if one can identify such over-payment, will cause a reduction in GDP.
Indeed, quite the opposite is the case. Transfers distort economic decision-making, thereby reducing economic efficiency and causing economies to grow slower. That doesn't mean all transfers are a Bad Thing - transfers allow us to do things we like, such as helping the poor or the old. But they make the economy grow slower, not faster.
Another kind of government spending also makes the economy grow slower - what is called "consumption spending". That means spending on things like doctors salaries, teachers' pay, consultants, IT support, and the like. (The EMA in Jonathan's list is an example of consumption spending.) In standard studies, conducted since the mid-1970s and now having achieved a high degree of academic consensus, each additional percentage point of GDP of government consumption spending reduces annual GDP growth by 0.1-0.15 percentage points.
These are key reasons why the orthodox advice on deficit reduction is that most of it (80% or more) should be delivered by cutting government consumption spending. The good way to cut a deficit is to cut consumption spending and do so early. The bad way - the way most government try and the way the Coalition has tried on this occasion - is by cutting government investment spending and raising taxes. The evidence from a large number of deficit reduction programmes around the world is that major deficit reduction is rarely, if ever, achieved by raising taxes and cutting government investment.
Jonathan knows this perfectly well, and I'd have expected more from him than to allege that there is no difference between tax rises and consumption spending cuts early in a deficit reduction programme.
Lastly, Jonathan attacks the claim that "Fiscal stimulus cannot work by definition, because "the money must come from the private sector"". He quotes the notion that fiscal stimulus can work by raising government spending and taxes at the same time. In the technical jargon, he is asserting that the "balanced budget multiplier" is greater than zero. That is a highly contentious claim - something Jonathan fails to point out - and if it were so "in general", as Jonathan asserts it is, the implication would be that we should all be Communists, since spending 100% of GDP would mean a larger economy than having any private sector at all. I can assure you that that is by no means an established orthodoxy amongst economists.
John and Fraser are perfectly entitled (as I do) to deny the existence of any general balanced budget multipler greater than zero. Whether output is raised or lowered by spending and taxing more surely must depend on what's done with the money. If the government taxes people and then spends the money on policemen hired to go around beating up anyone trying to set up a new business, I'm pretty confident output will fall. Some extra government spending may raise output; by the time we're already spending 50% of GDP (as now) most extra government spending will reduce output. It's just uber-statist ideology dressed up in thin economic garb to assert that the balanced budget multiplier is "in general" greater than zero. The implication of that would be that almost all governments, at almost all times, everywhere in the world, spend too little. If you believe that, you aren't John Redwood or Fraser Nelson.
Lastly, Jonathan attacks the "Treasury view" that increased public spending cannot, as a matter of accounting identity, increase output. He ascribes that view to Redwood. But Jonathan has totally misunderstood Redwood's contention. Read what Redwood actually says:
"Because the state needs so much tax revenue, other income and borrowings to feed it, it squeezes the rest of the economy. If the government decides on an extra pound of public spending paid for by a tax increase, that has no overall beneficial impact on the economy. The public sector grows by a pound, but the private sector shrinks by a pound. It is not a stimulus. If the state borrows an extra pound to spend, the private sector cannot spend the pound it lends to the government."
So Redwood's position is that, because the government has already raised taxes, other income (NI etc) so high and is already borrowing so much, there is no slack in any of these things for government action to exploit. That means that, under these conditions, the economy exhibits what is called "Ricardian equivalence" at the margin - each additional pound the government spends has the same impact, whether the government borrows or taxes to fund it, and fails to raise GDP.
Jonathan conveniently misses off the first part of Redwood's paragraph, in which he states quite clearly that it is the factors that squeeze the economy that make the rest of his proposition true - Redwood is simply not asserting the accounting version of the "Treasury view". How Jonathan fails to appreciate that obvious point I cannot say.
Overall, Jonathan repeatedly asserts uber-Keynesian ideology, of a particularly unreconstructed "more government is always better" type, in an attempt to patronise Fraser Nelson and John Redwood when their case is perfectly coherent and, in my view, perfectly correct. (Nothing wrong with patronising politicians or political journalists, of course - I do that all the time! Just be right when you do!) Jonathan is a smart chap, usually worth listening to and reading - and I strongly recommend his blog to you. But he can do better than this.