Wednesday, 7 December 2016

Scott Sumner opposes infrastructure spending and fiscal stimulus generally.

(Note: this article also appears on the Seeking Alpha site.)

Scott Sumner teaches economics at Bentley University, Massachusetts. He describes the Fed’s arguments for more fiscal stimulus as “bizarre”. As he put it a few weeks ago (I’ve put his words in green italics):

“I don’t use the term ‘bizarre’ lightly, as this stuff is not just wrong, or doubly wrong, it’s quintuply wrong.  It’s not even slightly defensible.”

Strong stuff! But I suggest it’s over the top. First, it’s not just the Fed that has started to sing the praises of fiscal policy: numerous other central banks and economists are doing the same. Thus Sumner must be a disappointed man. But let’s run thru his arguments, which are as follows.

1.  The Fed claims the economy does not need any more demand stimulus.  Indeed any boost to AD from more spending would be offset by tighter money.  So what’s the point?”

The point is that (as is now obvious), monetary policy is near out of options, or at least out of conventional options. There are of course negative interest rates and having central banks buy bonds or stakes in commercial enterprises, but taking commercial risks is certainly not what central banks were set up to do (quite rightly). So why not do some fiscal stimulus? If that forces central banks to raise interest rates, that means that come a recession, central banks can do conventional monetary stimulus: which is the position central banks want to be in. What’s the big problem there?

At least that’s an argument if you’re a fan of monetary stimulus. Personally I’m not: I favour having fiscal and monetary stimulus working in cooperation (i.e. in a recession, just have the state print money and spend it, and/or cut taxes). One reason for that is that come a recession there is no obvious reason why the problem is inadequate borrowing, lending and investment rather than inadequate spending on anything else. I expanded on that point recently here.

Conclusion: if interest rate cuts are a good way of regulating the economy (and Sumner thinks they are) then a bit of fiscal stimulus plus the consequent interest rate rise isn't a bad idea.

“2.  Even if fiscal stimulus were needed, it should be done via tax cuts.  It’s not efficient to vary spending for anything other than standard cost/benefit reasons, where benefits do not include demand stimulus.  And why should the central bank be telling Congress where to spend money?”

Let’s take that one sentence at a time. First: “Even if fiscal stimulus were needed, it should be done via tax cuts”.

The answer to that is that the question as to whether to expand or contract the public sector (i.e. raise or cut taxes respectively) is a PURELY POLITICAL DECISION. That’s a decision for politicians and voters at election time: not a decision for economists like Scott Sumner.

Re “cost/benefit reasons”, certainly infrastructure investments should be decided on cost/benefit criteria, like other investments, but the same doesn’t go for other types of government spending: defence, education, health care etc. Accurate measurement of benefits there is near impossible, thus decisions there are inevitably political.

“3. The Fed might respond that fiscal stimulus would not boost demand, but it would allow the current demand to be achieved with less monetary stimulus.  But why is that desirable?”

For reasons I gave above in response to point No.1.

“4.  The Fed might argue that it would prefer a higher trend rate of nominal interest rates, so that it hit the zero bound less often in future recessions.  But fiscal stimulus is an absolutely HORRIBLE way to achieve that objective:”

Prefixing something you don’t like with the phrase “absolutely horrible” is not an intelligent argument.

Next come two points prefixed with “a” and “b”.

“a.  Fiscal stimulus can only boost nominal interest rates by raising the global real rate of interest.  Just imagine how much fiscal stimulus it would take to boost the global real rate of interest by even 100 basis points.  (Hint: far, far beyond anything Congress would ever contemplate.)  Then think about how Japan did a massive amount of fiscal stimulus in the 1990s and 2000s, and ended up with some of the lowest interest rates the world has ever seen.”

Well at least fiscal stimulus enables us to increase demand without having to resort to bizarre stuff like negative interest rates and central banks accepting commercial risks by buying bonds of private corporations. Moreover, as Milton Friedman pointed out, stimulus dollars cost nothing in real terms. To put it figuratively, if we have to print a hundred tons of $100 dollar bills and dish them out to the population or spend the money on education, health care or whatever, what’s the problem, as long as that doesn’t cause excess inflation?

Also, Sumner suggests that astronomic amounts of fiscal stimulus would be needed to raise interest rates by any significant amount. But he gives no reasons. I’m not impressed by bald statements like that. I want to see REASONS given for that sort of claim or indeed any claim.

“b.  In contrast, monetary stimulus can easily raise nominal interest rates by 100 basis points, merely by raising the inflation target from 2% to 3%.”

I’m not totally clear what the argument is here, but I assume it’s the surprisingly common argument put by so called “economists” that a central bank only has to announce an inflation target of X%, and by some magic, inflation rises to that level. Apart from the obvious point that there is no clear cause / effect mechanism (i.e. no “transmission mechanism” to use the jargon), one has to wonder why inflation has fallen well below the 2% target and for years on end in numerous countries – e.g. Japan, as mentioned by Sumner himself just above? It’s clear that simply announcing a target does not of itself have much effect – a point which the average street sweeper has probably worked out.

5.  Economists agree, or used to agree that the US and other developed countries face severe long-term fiscal changes, due to an aging population.  The consensus is, or used to be, that now is a good time to start addressing these issues. (Remember Simpson/Bowles?).

The phrase “fiscal challenge” is a bit vague, but presumably Sumner means that increased spending on the elderly may lead to a rising national debt plus a rising rate of interest on that debt. Well the solution to that is to fund that spending from tax so that the debt DOESN’T rise (revelation of the century).

Plus there’s a very basic self-contradiction in Sumner’s latter point, as follows.

Assuming demand is inadequate even when the national debt is relatively high by recent historical standards, that means the private sector is hoarding money: it is prepared to hold state liabilities at low rates of interest (witness the near zero rates that have prevailed over the last five years or so).

But if the rate of interest (and in particular the REAL or inflation adjusted rate of interest) on the debt is zero (or even negative, which is where it has been recently) where’s the problem in upping the debt? Almost no interest (at least in real terms) will need to be paid.

Moreover, the debt:GDP ratio is nowhere near where it was just after WWII (or in the case of the UK, in the mid 1800s).

As to what to do if the rate of interest demanded for holding debt rises, that’s easy: pay off the debt when it matures and tell creditors to get lost. And if that results in excess inflation, then raise taxes and “unprint” the money collected, and/or raise interest rates.

As to a rise in the so called debt WITHOUT an accompanying rise in the rate of interest that needs to be paid on that debt (i.e. assuming the real or inflation adjusted rate of interest on the debt remains near zero), what’s the problem?

Sumner continues (his point No. 5):

It would be one thing if the Fed were proposing a short-term fiscal stimulus to boost demand right now.  But they aren’t, they don’t think we need more demand right now. Instead they are proposing a long-term fiscal stimulus, which would massively worsen the already worrisome long-term fiscal trends in America.  A decade ago, sensible economists (like Krugman) criticized these sorts of proposals as reckless, and they were right.  Japan has already shown that decades of fiscal stimulus do nothing to raise NGDP growth, and merely leave you with a higher debt/GDP ratio.  Why would we want to copy Japan’s failed experiment?  All they ended up with is lots of highway projects that are little used, and destroyed some of the once beautiful Japanese countryside.

That’s about the first valid point that Sumner makes: i.e. trying to forecast what fiscal stimulus (or indeed monetary stimulus) will be needed several years in advance is silly. No one knows what the economy will be doing then.

However, there is a straw man argument there as well: fiscal stimulus does not need to consist of uneconomic forms of infrastructure. As pointed out in point No.2 above, infrastructure investment should be done along commercial lines. I.e. the criterion should be: “does this investment pay for itself”. Assuming increased government spending is the order of the day, there are dozens of other forms of government spending that can be usefully increased (e.g. education, law enforcement, health care, etc).

Moreover, the decision to increase public spending rather than effect fiscal stimulus via tax cuts is a STRICTLY POLITICAL decision since it influences that proportion of GDP allocated to public spending, which itself is of course a political decision. Thus economists like Scott Sumner (to repeat) should not express opinions on that subject.

As regards the “Simpson-Bowles” point that a rising national debt is some sort of problem, that’s nonsense. To repeat, as long as the rate of interest on the debt is near zero, or below inflation, why should we worry about the debt? And if creditors start demanding more interest, just pay them off when debt matures and tell them to go away, which in effect equals QE. QE has not proved all that inflationary, but if it did, that inflation can be countered with various deflationary measures like interest rate increases (exactly what the Fed wants) or tax increases.

And another solution to the “worrisome” fiscal problem Sumner refers to is (as Keynes pointed out nearly a century ago) to fund fiscal stimulus from new base money rather than via debt.

“6.  What does improving education have to do with fiscal stimulus?  The US already spends more on public education than most countries, and education experts seem to agree that the real problem is poorly designed schools or bad home environment, not lack of money.  Would throwing a few more billions of dollars at the LA school system boost growth?  Would it turn the LA system into the Palo Alto system?  How?”

According to this Wiki site, the US is actually near the average so far as developed country spending on education goes. On that basis there is scope for more spending on education in the US if that’s what US voters vote for. But to repeat, that’s a political decision, not a decision for economists.

Incidentally, the Fed (i.e. Yelland) is also at fault in advocating a SPECIFIC TYPE of government spending, namely education. The decision as to how much to spend on education (or on defence, law and order, etc) is a POLITICAL decision: it’s a decision for politicians. In  contrast, if Yelland had just advocated more fiscal stimulus, that would have been OK since that is pretty much a strictly an ECONOMIC point.


Sumner’s description of the Fed’s ideas as being “bizarre” and “quintuply wrong” are way out.

No comments:

Post a Comment

Post a comment.