What is a (Eurozone) "structural deficit", again?

I don't have much to say about the latest EU "agreement" that hasn't already been said. To me, it looks just like the old "stability and growth pact" warmed over, and with a bunch of judges having to approve budgets. (Does this mean we are going to leave fiscal policy to the lawyers? Do they teach it in law school?)

But there's something other commenters have, I think, missed. The agreement says that countries can't have a structural deficit bigger than 0.5% of GDP. What exactly is a "structural deficit"? What does it mean? How do we measure it?

Read my old post, where I answered part of this question two years ago. Then let's continue from there.

First off, a "structural" deficit doesn't mean what you probably think it means. And it doesn't mean what politicians and lawyers (same people, usually) probably think it means. It doesn't mean "built in and hard to get rid of and unsustainable". Instead it means "what the deficit would be given the current set of taxing and spending rules if GDP were equal to Yn". (I'm coming back to what "Yn" means later).

Second, it is really easy to make the "structural" deficit anything you want it to be, without changing the actual deficit at all, just by fiddling with the laws. See my old post for examples.

Third, what does "Yn" mean, and how do we measure it? That's what I want to spend the rest of this post talking about.

Here is a list of different names for Yn, off the top of my head: the natural rate of output; potential output; full employment output; NAIRU output; LRAS output; normal output; trend output; etc.

Sometimes these are just different names for the same thing. But sometimes they aren't. Sometimes those differences in names reflect a difference in theoretical beliefs about macroeconomics.

Even when two economists agree on what they mean by Yn, they might disagree about the best way to estimate a particular number for Yn. Econometrically, it's not obvious how best to do it.

Even if two econometricians agree about how best to estimate Yn, they are agreeing on how to estimate past values for Yn. The "structural deficit" in a proposed budget is talking about a forecast for Yn in the coming fiscal year.

The Bank of Canada is no slouch, but IIRC, its real time estimates of current Yn are often very different from its own later revised estimates of past Yn, when it looks back with the benefit of hindsight. (Where are you Simon van Norden? Come in please!) If the Bank of Canada can't get current Yn right, even according to its own standards of what is right, how can a bunch of judges expect to agree on a forecast for Yn?

Look, economists and econometricians can't agree on this stuff. Does the Eurozone really want to let the lawyers try to figure it all out? Fiscal policy is a judgment call; always has been, and always will be. Best to let elected politicians make that call, with economists in a subsidiary role, trying to give the best advice they can.

On the other hand, I can see this as a nice little sideline for some macroeconomists. "You want to pay me to convince a bunch of lawyers that your actual deficit is not a structural deficit? OK."

Comments

[Dave]

Merkel needs a "prudency" cover to justify allowing the ECB to start printing - the flexible nature of "Yn" suite all sides.

[Livio Di Matteo]

Nick:
In this Euro definition of a "structural deficit" did you get any sense if countries would be allowed to vary the size of the deficit that can be incurred based on the size of the per capita incomes? For example, would Greece be allowed even smaller deficits as a share of GDP relative to say Germany because of its lower per capita GDP? One would think if the EU is now going to go into fiscal regulation it would specify rules to remove from weaker economies the temptation of incurring larger deficits.

[Frances Woolley]

The whole structural deficit concept is based on the idea of an economy on a stable long-run growth trend, with business cycles just being ups and downs on that long-term trend.

But structural breaks - i.e. changes in the long-run growth trend - happen. And it's hard to identify them, except in hindsight. E.g. as old-style large enterprises with employees are replaced with nimble striped down enterprises with self-employed contractors, it becomes far harder to collect tax revenues. (There are many more opportunities for tax avoidance on self-employment income; large corporations are great tax collectors - at least, great at collecting and remitting taxes on employees). This alters the relationship between government revenues and GDP, which in turn changes the structural deficit. But self-employment always increases during recessions, at least in Canada anyways - so how do we know if the current upwards trend is a fundamental change in the way businesses are organized resulting from technological and institutional changes, or just a cyclical blip?

(Still avoiding marking).

[Nick Rowe]

Dave: could be, and it would make sense. But the ECB isn't making the right communications to back that up.

Livio: the agreement didn't say anything about per capita income, unless I missed it. And i think the baseline assumption would be that everything scales up with population and per capita GDP, in debt/GDP ratios.

Frances: that is one of the necessary conditions, agreed. And it's one big problem, given likely changes in growth rates due to demographics, just for starters.

But even if everything follows a perfect trend, if you have a debt/GDP ratio of 100%, 2% inflation, 3% real growth, you can run a deficit, as conventionally measured, equal to 5% of GDP forever.

(Yep, I'm avoiding marking too.)

[K]

Nick: "you can run a deficit, as conventionally measured, equal to 5% of GDP forever."

That was my thought too. The only way I can square it is that they are assuming that "cyclical" deficits average 4.5% of GDP (in your scenario). Ie. "cyclical" means stuff you do in a recession but which is never counteracted during a boom. So "structural" in this case means what you naively hope for but totally fail to achieve on average.

[Patrick]

Regardless of why the UK stayed out (the claim is that it's to protect the city), I'm thinking they dodged another bullet. Score another point for small c conservatism.

[Ralph Musgrave]

Here is a brain teaser.

A structural deficit does not by definition involve stimulus. Therefor removing a structural deficit and/or the debt that arises because of a structural deficit, does not involve “anti-stimulus”. That is, none of that dreaded “austerity” is required to remove a structural deficit/debt.

That just leaves the question as to how to actually remove structural deficits/debt without an austerity effect. I’ll be awarding marks out of ten for answers submitted. Which is a bit of a cheek considering this blog is run by people who teach economics.

[nanute]

Ralph: I don't see how you can achieve your outcome without a central bank.
Nick: I thought under the latest agreement among EU members, deficits can't be more than 3% of GDP. This means more spending cuts or increased revenues in most of the countries, including France and Germany.

[Min]

Nick Rowe (earlier post): "Macroeconomists like to divide an actual deficit into two components: a cyclical deficit and a structural deficit. It reflects their distinction between automatic stabilisers and discretionary fiscal policy. I don't think that distinction is as clear and useful as macroeconomists think it is. And in ordinary language a "structural" deficit is one that is hard to get rid of. That's something quite different."

Indeed. If you want a neutral term to contrast with "cyclical", "secular" comes to mind. :) And, pardon me for saying so, but I suspect that the term, "structural", was chosen **because** of its non-neutral connotations, including the idea that it is hard to get rid of, with the further connotation that it is a flaw (Nick: "unsustainable"). Vide the 0.5% rule.

Nick Rowe (earlier post): "The distinction between automatic and discretionary fiscal policy, and hence between cyclical and structural deficits, is a political distinction, not an economic one."

Wisdom! :)

Nick Rowe (current): "The Bank of Canada is no slouch, but IIRC, its real time estimates of current Yn are often very different from its own later revised estimates of past Yn, when it looks back with the benefit of hindsight. (Where are you Simon van Norden? Come in please!) If the Bank of Canada can't get current Yn right, even according to its own standards of what is right, how can a bunch of judges expect to agree on a forecast for Yn?"

Well, hey, the US can project Social Security 75 years into the future. So there! Nyah, nyah, nyah. ;)(Or as I usually say, GIGO! GIGO! GIGO!)

[Nick Rowe]

nanute: "Nick: I thought under the latest agreement among EU members, deficits can't be more than 3% of GDP. This means more spending cuts or increased revenues in most of the countries, including France and Germany."

Yep, that's correct too. There are two limits: actual (planned? projected?) deficit can't be greater than 3% of GDP; and "structural" deficit can't be more than 0.5% of GDP. The 3% limit is problematic too, but I don't need to add anything there, since other bloggers have that covered. I'm just talking about the 0.5% limit on "structural" deficits, since the other bloggers seem to have missed it (I think).

[Nick Rowe]

Ralph: "A structural deficit does not by definition involve stimulus."

I disagree. If the distinction actually made any sense, I would be more inclined to say it's the other way around. An increase in the cyclical deficit, when the economy goes into recession, is normally thought of as an automatic stabiliser, while a deliberate policy of cutting tax rates or increasing spending, which increases the structural deficit, is normally thought of as fiscal "stimulus". (God I hate that word "stimulus"; it's such a knuckle-dragging way of talking about shifting the AD curve and the consequences of doing so on real output.)

Min: "Indeed. If you want a neutral term to contrast with "cyclical", "secular" comes to mind. :) And, pardon me for saying so, but I suspect that the term, "structural", was chosen **because** of its non-neutral connotations, including the idea that it is hard to get rid of, with the further connotation that it is a flaw (Nick: "unsustainable"). Vide the 0.5% rule."

"Secular" doesn't quite mean the same thing. Economists aren't very good at naming things. I can think of three places where the word "structural" is used in economics: structural deficits; structural unemployment; and structural vs reduced form equations. In all 3 cases it's used rather loosely, to mean "independent of *something*". But yes, the word has connotations that may or may not be valid, depending on context, and it can't carry the weight we put on it.

[Nick Rowe]

It might be useful to invoke some sort of version of Goodhart's Law here.

For any given country, where the government isn't trying to fiddle with the laws to make the "structural" deficit less than some arbitrary number, and where those laws about taxing and spending don't change much over time, the concept of a "structural" deficit is not totally useless. It can be a handy rule of thumb, if you don't expect too much from it. But if you make "structural deficit" a policy target, it's going to change its meaning. It might not mean anything any more.

[nanute]

Nick,
If a government continues, year after year, to consistently spend more than it brings in revenue wise, this isn't a structural problem? And it has just occurred to me: are they talking about RGDP or NGDP?
This exercise is bound to fail. Germany doesn't want any more borrowing by the major debtors, and yet is insisting that bondholders be guaranteed repayment. What am I missing here?

[Jon]

As a matter of political science, why should governments be allowed to borrow at all? It seems to me that allowing politicians to dispense benefits now and costs later is an incentive nightmare.

One of the big results in the last 30-40 years of economics is precisely that everything good a fiscal deficit can do can be done by the CB without the same incentive problems.

And as Nick put in a post a while back: plastic vs glass. The CB creates plastic; fiscal debts mint glass.

Note: I'm not saying the govt cannot act as an insurer but it ought to act as a proper insurer and have the necessary reserves.

[Tim]

Patrick:

I actually think from a Canadian perspective the UK rejection of a treaty amendment was the much more interesting story. From following the UK media over the weekend what happened vis a vis the EU and UK on Thursday night(I always consider the UK to be the EU's Quebec) was something akin to the night of long knives and the collapse of Meech Lake. It will be interesting to see in the weeks ahead to what the UK's future relations with the EU will be.

[Nick Rowe]

Tim: I hadn't noticed the Meech Lake analogy. I was thinking Dunkirk.

It's ugly, but I think the French and Germans are more pissed at the British than vice versa. Which is more like Dunkirk than Meech Lake.

nanute: if you divide both nominal deficit and nominal GDP by the price level, the deficit/GDP ratio is the same. But the correct inflation (as opposed to price level) adjustment for the deficit is to subtract the inflation rate times the level of debt to get the growth in real debt. Or, subtract the growth rate in nominal GDP times the debt/GDP ratio from the deficit to get the growth in the debt/GDP ration. Aaaah! Too much arithmetic for this evening.

Jon: well, government investment for the future should normally be deficit-financed. And there are good tax rate smoothing arguments for not having a balanced budget in all periods, even if you think monetary policy can handle the stabilisation question.

[Donald Pretari]

i can't imagine a Govt agreeing to use such a term that they couldn't fudge in the future when appropriate, at least plausibly enough to satisfy their own constituents or masters. Can you say, "Kabuki?"

[W. Peden]

Donald Pretari,

That almost wouldn't be so bad. But consider who gets to decide how the term is fudged: the central authorities in Brussels! Quite the concentration of rents.

These measures are less about making eurozone countries take control of their national finances and more about passing control of those national finances to the EU authorities.

[Jon]

"well, government investment for the future should normally be deficit-financed."

The way to deal with that is to use actuarial accounting not cash accounting.

"And there are good tax rate smoothing arguments for not having a balanced budget in all periods, even if you think monetary policy can handle the stabilisation question."

Again, you can make that argument but that doesn't mean the government needs to issue debt, as it could instead utilize a reserve fund for that purpose.

[Determinant]

Frances wrote:

The whole structural deficit concept is based on the idea of an economy on a stable long-run growth trend, with business cycles just being ups and downs on that long-term trend.

But structural breaks - i.e. changes in the long-run growth trend - happen. And it's hard to identify them, except in hindsight. E.g. as old-style large enterprises with employees are replaced with nimble striped down enterprises with self-employed contractors, it becomes far harder to collect tax revenues. (There are many more opportunities for tax avoidance on self-employment income; large corporations are great tax collectors - at least, great at collecting and remitting taxes on employees). This alters the relationship between government revenues and GDP, which in turn changes the structural deficit. But self-employment always increases during recessions, at least in Canada anyways - so how do we know if the current upwards trend is a fundamental change in the way businesses are organized resulting from technological and institutional changes, or just a cyclical blip?

(Still avoiding marking).

Big Enterprises are not only good tax remitters, they were turned into the delivery vehicle for the shadow welfare state by providing defined-benefit pensions and health (=drug) insurance to their employees, all by deliberate government policy. These policies contained a huge assumption that big enterprises were institutions and were long-lived. That turned out to be false. The model said "here are n big enterprises (institutions which deliver welfare-like programs), choose one. It will service you." We assumed that by making a choice from n that we had a "free market". It was a nice fantasy while it lasted.

The biggest problem of the Canadian welfare state is that one of its key vehicles, Big Enterprise, is no longer there. We have not come up with a new model to replace it.

I would love to see a thread on the policy implications for Welfare State design of the shift from a Big Enterprise to Small Business economy.

BTW this shift is at heart the reason I am a lefty. I know what policy changes I want to see to reflect the shift and those choices make me a lefty.

[Phil Koop]

"if you have a debt/GDP ratio of 100%, 2% inflation, 3% real growth, you can run a deficit, as conventionally measured, equal to 5% of GDP forever."

Yes - apparently, Europe is planning for a combination of low inflation and low growth! But what is cause and what is effect? ;-)

[Russil Wvong]

Jon: "One of the big results in the last 30-40 years of economics is precisely that everything good a fiscal deficit can do can be done by the CB without the same incentive problems."

I'm baffled by this comment. I'm an interested layman, not an economist, but I thought the current recession in the United States demonstrated the exact opposite! Nominal interest rates are zero, but unemployment is still high; in a liquidity trap, monetary policy is ineffective. Japan has been in the same situation since the 1990s.

So we still need counter-cyclical fiscal policy.

What am I missing? Is there some body of research that demonstrates the effectiveness of monetary policy in a liquidity trap? (Price-level targeting? NGDP targeting?) If so, what's prevented it from being effective in Japan or the United States?

[Nick Rowe]

Russil: that is precisely the question there's been a lot of debate about recently. Scott Sumner, at his blog Money Illusion, has been one of the strongest voices saying that central banks are not out of ammo. I agree. Most macroeconomists would disagree with the simple story that "interest rates are 0%, therefore central banks can do nothing more".

We're a bit burned out arguing about it though.

[Bob Smith]

Nick: "Does this mean we are going to leave fiscal policy to the lawyers?"

God yes! Oops, probably not the right blog for that!

Coincidently, for what it's worth, lawyers have been in charge of Canada's finances since 1993 and done reasonably well over that period (Martin, Manley, Goodale and Flaherty all being trained practioners of the dark arts). Maybe there's something to the suggestion (or maybe I'm just thinking a new field of "deficit" law would be just the sort of thing to pump up my billable hours).

In all seriousness, though, I think your assessment that it's just a rehashed "stability and growth pact" is probably right (hey, but that worked well, right?). At the end of the day, treaties are as much political documents as they are legal ones (indeed, given the general uselessness of many of the world's collections of treaties - i.e., generally anything with the words "human" or "civil" and "rights" in it isn't worth the paper it's written on - one could forgivably conclude that they are primarily political documents). Even in the EU, where EU treaties are theoretically legally binding on member states, at the end of the day, whether or not countries choose to comply with those treaties (or compel their neighbours to comply with their obligations - "sure, we know Greece is lying about it's deficit statistics, but hey, we are too so let's not get too boisterous about it") is a decision based on domestic political interests. Since the problem facing the EU is the large scale failure of political leadership (i.e., Europe is, or has heretofore been, led by a collection of moral midgets), introducing a new treaty doesn't resolve the problem.

[Jacques René Giguère]

@Ralph Musgrave:
"That just leaves the question as to how to actually remove structural deficits/debt without an austerity effect. I’ll be awarding marks out of ten for answers submitted. Which is a bit of a cheek considering this blog is run by people who teach economics."

This where we can see the dangers of the cyclical finance fallacy that was adopted to sell keynesianism to Biz Schools and Chamber of Commerce types." We will run deficits during recession but we will repay the debt during the next recovery". Of course, any surplus is contractionnary. Unless you really believe there will be crowding-in ( aka confidence fairy).

[anon]

Ralph Musgrave, if the central bank is unwilling to run an expansionary policy (but won't otherwise sabotage your effort) the only reasonable answer is to subsidize investments in order to attract capital inflows and stimulate bank lending. Or you could try to tax money balances indirectly, such as by redistributing income away from folks with high propensity to hold base money,

[Tim]

Determinant:
To follow your point about big enterprises one reason I have always suspected goverments favor the banking sector so much is that banks also play an important role in helping goverments enforce things such as tax and anti money laundering laws. I wonder what the effect of things such as prepaid payment cards and mobile payments will have on the traditional banking sector from both a regulatory perspective and an economic perspective especially as at least in my opinion I tend to see the traditional bank chequing account as going the way of the landline telephone for many people my age(under 30).

[Determinant]

I'm under 30 and I disagree on that, Tim. Canadian banks have been very proactive adapters of technology. The concentrated landscape that hasn't changed much in a century and clubby nature of the Canadian Payments Association has let it deploy technology easily, witness Interac, internet banking, etc.

Payments as negotiable instruments that need to be cleared will always need a bank. Banking requires capital and it's the chartered banks that have that capital.

Canadians can complain about many things with banks but Canadian banks have been very good adapters of technology to service client needs. We haven't been held back on that front.

[Tim]

Determinant:

I agree in Canada the chartered banks have tended to be much more at the forefront of innovation than the US or UK for example, having said I still think they will face competitiveness challenges in the future regarding payments. One of more interesting things I noticed recently was that Rogers has recently applied to form a chartered bank(something expressly not permitted in US but perfectly legal under our more enlightened regulatory structure). To be clear when taking about monetary policy and payments in Canada there are only a select group of institutions that "count" such as the Big Six, Laurentian, Desjardins, Alberta Treasury, HSBC, Credit Union Central, the BofC itself and a few foreign banks such as BofA which are collectively the "direct" clear members of the CPA. Rogers might get a bank charter but unless they become a "direct" clearer that arent't a "real" bank in a sense. In the US they let basically anyone clear through Fedwire with the Fed giving free daylight overdraft to even the smallest member. The real issue with mobile payments in Canada is both the carriers and the chartered banks have pretty cosy oligopolies and it would not seem to likely that they would want to cause problems for each other.

Overall I agree with your point about the decline of "Big Enterprises" and think is a huge problem for maintaining the social safety net especially in the US. Canada in my mind has avoided some of these issues through two ways. One is Canada has a partial but definately not full safety net seperated from enterprise. Second "big enterprise" has in many ways not declined to same extent in Canada as in other parts of the world. In many industries not just banking only a few players still hold most of the market share. In Alberta there might be a lot of smaller "wildcat" oil and gas companies but all of the refined product is produced by either Suncor, Shell, or Imperial; same with airlines, telecommunications etc. Now as some such as Stephen have pointed out many of these Canadian oligopolies have fairly low worker productivity so I guess there is a tradeoff.

[Aslak]

The Norwegian government actually does this on a regular basis. Norway has a fiscal rule that says the structural deficit not counting oil revenue can't exceed 4 percent of the sovereign wealth fund (which is the expected real rate of return on the oil fund) over the business cycle. So far politicians have basically kept to it, surprisingly enough

[Nick Rowe]

Aslak: interesting. My very quick Googling suggests that Norway's sovereign wealth fund is nearly 200% of GDP. Is that right?? (That seems really big.) If I'm right on that, then maybe a limit of nearly 8% of GDP for the structural deficit would maybe not be a binding constraint, in any case?

[Aslak]

It's actually closer to 130%, which admittedly is still very big. I do think you're right even so that the strucural deficit increasingly isn't the binding constraint, but rather the supply of labor. The government can't really spend more money without incurring Dutch disease (and even now we're at risk). Luckily, the civil service in our Finance Ministry is powerful and very Keynesian in the sense that they believe in countercyclical fiscal policy, including when times are good, and politicians generally respect their advice. Last I heard the oil-adjusted structural deficit was something like 3.2% and shrinking and unemployment is at about 3%. It's just very hard to keep an economy like Norway's from overheating, but then I suspect that perhaps Canada has a less extreme version of the same problem?

[Aslak]

Just for clarity, by 3.2% I meant 3.2% of the oil fund, not of GDP.

[Too Much Fed]

Frances Woolley said: "The whole structural deficit concept is based on the idea of an economy on a stable long-run growth trend, with business cycles just being ups and downs on that long-term trend."

http://blogs.telegraph.co.uk/finance/ambroseevans-pritchard/100013758/europes-blithering-idiots-and-their-flim-flam-treaty/

"Germany has kept the focus exclusively on fiscal deficits even though everybody must understand by now that this crisis was not caused by fiscal deficits (except in the case of Greece). Spain and Ireland were in surplus, and Italy had a primary surplus.

As Sir Mervyn King said last week, the disaster was caused by current account imbalances (Spain's deficit, and Germany's surplus), and by capital flows setting off private sector credit booms."

So would having a current account balance = 0, private debt = 0, and gov't debt = 0 lead to a more "stable path"?

[Jon]

Nick writes: "Russil: that is precisely the question there's been a lot of debate about recently. Scott Sumner, at his blog Money Illusion, has been one of the strongest voices saying that central banks are not out of ammo. I agree. Most macroeconomists would disagree with the simple story that "interest rates are 0%, therefore central banks can do nothing more"."

What is widely understood and agreed on... and is the the last 30-40 years to which I was referring is that if we are not faced with a liquidity trap, CB policy obviates the need for counter-cyclic fiscal measures. Many many economists would support the claim that we are now in a liquidity trap, but neither Nick nor Scott nor myself believe that to be so.

This is not a fringe position but it is a minority position.

[K]

Jon: How do you define "liquidity trap?"

[Russil Wvong]

Jon: "What is widely understood and agreed on... and is the last 30-40 years to which I was referring is that if we are not faced with a liquidity trap, CB policy obviates the need for counter-cyclic fiscal measures."

Thanks, that seems reasonable. A 1997 column by Paul Krugman, "Vulgar Keynesians":
http://web.mit.edu/krugman/www/vulgar.html

Returning to Nick's original post: what methodology does the Parliamentary Budget Office use to estimate potential GDP and the structural deficit? (The methodology I'm familiar with is that the unemployment rate -- or rather, the difference between the unemployment rate and the NAIRU -- is a measure of unused capacity.)

[Russil Wvong]

Looking at Annex C ("Long-Term Economic Projection") of the PBO's September 2011 Fiscal Sustainability Report, particularly Table C-1, the PBO appears to be projecting growth in potential GDP based on (a) the projected labour supply and (b) an estimated 1.2% annual increase in labour productivity (based on labour productivity growth over the last 30 years or so).
http://www.parl.gc.ca/PBO-DPB/documents/FSR_2011.pdf