My blogging has been light, partly because of this, partly because my teaching load this term is heavier than usual, and partly because I've been spending so much time trying to get a good feel for the numbers that will form the background of the federal budget scheduled for March 22. I keep thinking "Okay, I'll get this sorted out and then post", but I never do, so I don't. So I'm going to try to force myself to post what I've figured out so far, and if I make a mistake - as does happen - I'll go back and revise.
Anyway, this post is on what has happened to the 'structural deficit', or the 'cyclically-adjusted budget balance' (CABB) over the past few years. It turns out that what happened to it was the GST cut.
The idea is simple and compelling enough, but putting it into practice is a thorny business indeed. It turns out that aside from certain extreme episodes such as the last recession, it's extremely difficult to determine whether not output is above, at or below normal in real time. So when we look back and see policy decisions that turned out to be mistakes, I'm inclined to be generous. Things that may seem obvious several years ex post were not at all clear at the time.
The budget balance we observe is the sum of two things: the 'structural balance' - what we would have observed if the economy was operating at capacity - and the contribution of the cyclical component. So estimating the structural balance is an exercise in extracting the trend and the cycle from the available data.
Here is a graph of the PBO's estimates for the CABB since FY 1976-77, expressed as a percentage of Gross Domestic Income:
I like these graphs, because they put our current concerns in some sort of perspective. We've seen worse. Much, much worse.
What's curious about that graph is that it shows a CABB that went from about being in surplus of 1% of GDI up until about 2003 to bouncing around zero between 2004 and 2006. The timing of this coincides with the increase of federal transfers to the provinces - in particular, the Canada Health Transfer - under the Martin government. The Conservatives' GST cut then moves the structural balance into deficit.
This is of course what we see in hindsight. At the time these decisions were made, the federal government was cruising along with a surplus on the order of 1% of GDP, and here's why: the economy was running well above potential. Extracting the cycle from the trend is a messy business, but when three different approaches obtain similar estimates for the size of the output gap, it's reasonable to conclude that 2006-2008 were boom years:
The Bank of Canada is the institution with the most interest in getting good estimates for the output gap, and is also the one with the most resources to put into the problem, so their results are as reliable as we're likely to see. According to their estimates, the period 2006-08 was exceptional: the longest and strongest boom in at least the last 30 years.
So this is the fiscal situation that the Conservatives inherited:
- a structural surplus/deficit that was close to zero
- an actual surplus on the order of 1% of GDP, due to a booming economy
Again, it's important to emphasise that no-one really knew that at the time; this is hindsight speaking. Back when the Conservatives first floated the idea, critics said the GST was the the last on the list of any intelligent tax-cutting agenda (which is of course true). I don't recall anyone warning of the risk of chronic deficits in the 2005-06 election campaign.
But this doesn't change the main conclusion: when the economy finally does return to capacity - the Bank of Canada's current forecast is that the output gap will close at the end of 2012 - the swing in the federal government balance will not be enough to close the deficit.
How big is the hole that has to be filled? Unlike, say, corporate income taxes, the effect of the GST on the budget balance is fairly easy to calculate. Since the behavioural responses to changes in the GST rate are small (this property is why they are such a favourite with economists), the numbers in this earlier post - around 0.75% of GDP or $12b - won't be a bad guess. The PBO's estimate is $14b - and again, their estimate also corresponds to the size of their estimate of the effect of the GST cut on the budget balance.
The story of the GST cut just keeps getting worse. It wasn't a good idea at the time, and it blew a $12b hole in the federal balance that will have to be filled somehow.
So why did the Conservative government implement the GST cut?
I see four possibilities:
1. It opened up tax room for the provinces, paving way for the provinces to implement harmonized sales taxes. If the worsening of the federal budget balance was offset by an improvement in provincial governments' structural balances perhaps it wasn't such a bad thing?
2. It's a long-term strategy to reduce the size of the federal government: Create a deficit/debt crisis which will then justify decimating the federal public service (decimate? cut 1 in 10? that's about right).
3. It was prompted by short-term political vote-maximizing calculus.
4. They didn't work out what impact it would have on federal revenues and the federal budget balance.
Option 4 seems unlikely - a straightforward calculation (take total GST revenue, divide by 7, that gives you the revenue cost of a 1 percentage point cut in the GST) is enough to work out the revenue loss. And it was pretty obvious even back then that the federal budget situation was likely to worsen dramatically in the medium term with demographic change.
Option 3 seems unlikely to me also - cutting the GST is a really expensive way of winning votes compared to, say, the Canada Fitness Amount tax credit or even the home renovation tax credit.
So that leaves options 1 or 2, and my guess is that it was a mix of both.
Posted by: Frances Woolley | March 14, 2011 at 09:19 AM
"Option 3 seems unlikely to me also - cutting the GST is a really expensive way of winning votes compared to, say, the Canada Fitness Amount tax credit or even the home renovation tax credit."
Expensive, yes, but also simpler to sell politically and politically more effective. It's a lot easier to put "cut the GST" in a sound bite or on a brochure, then to explain the mechanics of various tax credits. Moreover, it'll appeal to a wider variety of voters.
If I had to guess what was driving the Tories it was primarily (3), to a lesser degree (2) and, perhaps, for the more clever ones, an element of (1) (though that's really giving them the benefit of the doubt - had that been the strategy, the GST cut would have been presented as part of a quid pro quo for the implementation of the HST, in order to reduce the political cost to the provinces: "yes, we're implementing the HST, but we convinced the feds to cut it by 2%").
Posted by: Bob Smith | March 14, 2011 at 09:32 AM
A $12b deficit actually means a falling debt/GDP ratio, IIRC. There's the inflation adjustment, say 2% inflation times the debt. Plus the real growth adjustment, say another 2%-3% times the debt. So a $12b structural deficit doesn't worry me per se. It's "really" a surplus, in terms of sustainability.
But I am worried about those calculations Bill Robson at CD Howe made last year about the effect of demographic changes on the future deficit. Health care for all us aging boomers. If Bill is even half right, we need to be running big structural surpluses in the next few years to help us handle it. Or, in other words, there's the cyclically-adjusted deficit, and then there's the demographically-adjusted deficit.
Posted by: Nick Rowe | March 14, 2011 at 09:54 AM
Two quick comments.
I don't recall that more provinces moving to the HST was seriously on the horizon when the Conservatives started pushing to cut the GST.
And when it was implemented what I do recall was that the debate (politically) was between the GST cut and a Liberal plan to cut personal income taxes. Which I think points to a general misunderstanding of where we were.
Posted by: Jim Sentance | March 14, 2011 at 09:56 AM
RSPs will help. As the stock of savings in RSPs grows, the amount of deferred tax grows with it. We have presumably not yet hit steady state with RSPs (where the stock of savings is growing at the same rate as the economy). So when we do hit steady state (plus when the boomers all retire), RSP withdrawals will rise, and tax revenue will rise too. I tried to do a back of the envelope calculation of how big this effect would be, but couldn't easily find the data, and my brain fried too, so I gave up.
Posted by: Nick Rowe | March 14, 2011 at 10:00 AM
@Nick Rowe: I agree that RSPs will help, but was this not the blog where the pressure to make RRSP withdrawals tax free was discussed? I suspect that could accelerate these issues, it it comes to fruition.
At a more general level, I think that we can expect some growth slowdowns in countries with a lot of older adults. Japan does well in growth if the denominator is adults 20 to 60 and not per capitia. I think we have managed to grow our way out of problems fairly often adn that may not always be an option in the medium term.
Posted by: Joseph | March 14, 2011 at 10:24 AM
Nick, on the revenue potential of RSPs see my post on retirement savings plans: brilliant economics, lousy psychology. Yes, there should be revenue potential there. But people will fight tooth and nail to avoid paying those taxes!
Posted by: Frances Woolley | March 14, 2011 at 10:30 AM
Won't all those retires hurt tax revenues? Presumably boomers are currently at max yearly earnings. When they retire and switch to RSP/RIF income, their income will drop and so will tax revenue.
No?
Posted by: Patrick | March 14, 2011 at 10:51 AM
Patrick, I think at best RSPs can make the effect of a smaller working-age population a wash for tax revenues, but likely the effect will be smaller.
Posted by: Andrew F | March 14, 2011 at 11:42 AM
Patrick, registered retirement savings plans defer the payment of taxes - so people pay taxes after they retire, rather than when they are working. Sure, tax revenue will drop, but the decline in revenue won't be as extreme as it would have been in the absence of registered savings vehicles.
Posted by: Frances Woolley | March 14, 2011 at 11:43 AM
My initial reaction to this post was to wonder, as Stephen emphasizes, if the extent of excess capacity in 2006-08 is indeed something that we can really only gauge well in hindsight. My sense was that the Bank of Canada would have been certainly aware of this at the time (at least to some extent) and that should provide some information for fiscal decisions and interpreting budget balances.
But it's not so simple (or, I was wrong). I don't have access to real time date from the Bank of Canada but we do have the Monetary Policy reports, which provide some measure of capacity though not always in consistent ways. I don't have a lot of time this morning so I just skimmed through some reports to see what we did know.
Starting with October 2005's Report, we have estimated output gaps at essentially zero for 2002-05. The forward looking base case projections were for an economy to operate at capacity through 2007. So, policy makers could reasonably believe that the actual deficits were close approximations to structural deficits around about this time. Moreover, if we look to the 2007 Report, the projected output gap is somewhat below that reported in Stephen's graph - the Bank was playing catchup at this time, basically until the financial crisis.
But this isn't the only lesson. In the 2005 Report, the 95% confidence intervals for the whole of the 2002-05 period are +/- 2%. Confidence intervals don't seem to be provided in later Reports. But the basic message is that measures of output gaps and hence cyclical adjustments of budgets have significant standard errors, even contemporaneous and recent history measures, let alone projections. This has implications for interpreting what governments do but it also means fiscal decisions should be made all the more cautiously since the implications for the structural deficit may take some time to show themselves.
Posted by: G Voss | March 14, 2011 at 12:47 PM
I agree. The focus of Simon's paper with Athanasios Orphanides was probably the conduct of monetary policy, but it may be even more important with fiscal policy. If the Bank of Canada realises all of a sudden that it's on the wrong side of potential output, it can adjust its policy stance pretty quickly. But fiscal policy takes years to turn around.
Posted by: Stephen Gordon | March 14, 2011 at 01:13 PM
Yup, I know. But isn't the important point that tax revenue will drop boomers retire. Considering how much we hear about how boomers have saved to little for retirement, I am assuming their incomes will drop rather drastically, and tax revenue with it - just as they begin demanding more services and, as you point out, probably even lower income taxes.
Posted by: Patrick | March 14, 2011 at 01:38 PM
The PBO did some analyses of fiscal sustainability, and called for an improvement in the budget balance of about 1.25% of GDP to put our finances on a sustainable footing for a 75 year horizon. I like this kind of planning, as it tends to recommend relatively small course corrections while giving us some certainty that we won't get too close to the rocks. I suppose the danger is that we get a false sense of confidence is these projections. I presume that small errors in life expectancy or inflation could cause us to miss that projection by a wide margin.
Posted by: Andrew F | March 14, 2011 at 01:58 PM
"I don't recall that more provinces moving to the HST was seriously on the horizon when the Conservatives started pushing to cut the GST."
I can't speak for other provinces, but it was always floating around as a policy idea amongst Ontario PCs(it was, and still is, despite their kvetching about it, a policy of the Ontario PCs), so I have no trouble believing that some of their federal cousins (notably, the federal Finance Minister) wanted to pursue it as a long-term policy. It certainly wasn't an idea that they were going to run a campaign on, though.
"And when it was implemented what I do recall was that the debate (politically) was between the GST cut and a Liberal plan to cut personal income taxes. Which I think points to a general misunderstanding of where we were."
Indeed, as I recall, the Tories initially increase the bottom personal income taxes back up to 16% (i.e. cancelled a Liberal tax cut) when they were elected, though they subsequently reduced that tax rate again in 2007.
Posted by: Bob Smith | March 14, 2011 at 05:42 PM
"I don't recall that more provinces moving to the HST was seriously on the horizon when the Conservatives started pushing to cut the GST."
Remember the Atlantic provinces have had an HST for some time, and Quebec also has had an HST-type arrangement for quite a while. So the idea of having more provinces move to HST has been around for quite a while.
Posted by: Frances Woolley | March 14, 2011 at 06:58 PM
And since then, both Nova Scotia and Quebec have increased their HST/TVQ rates.
But I do seem to recall some noise about provinces being invited to occupy the tax space vacated by the feds. At least, here in Quebec, anyway, where such talk is popular.
Posted by: Stephen Gordon | March 14, 2011 at 07:03 PM
When the GST was cut to 5%, Harper said this:
“Reducing the GST is part of our broader plan to ensure Canada’s long-term economic growth and prosperity. Under our Government, taxes are headed only one direction: down.”
One might conclude from this that the GST was cut as a broader plan to reduce taxes across the board, and to stimulate the economy.
For stimulating the economy, Harper's timing (wether by luck or good planning) could not have been better, as the world was heading straight into an election. And one might guess that income tax is next on the chopping block, should Harper ever get the opportunity.
Posted by: rabbit | March 14, 2011 at 07:57 PM
Problem is, it isn't possible to cut taxes AND maintain services. Something's gotta give. You can't have jail for all and toys for the boys AND tax cuts. And Canada isn't going to get a free pass like the US on running big Republican style deficits. There isn't an insatiable demand for Canadian gov't debt.
OT: Nuclear meltdown imminent if not in-progress in Japan and the Saudi's marched into Bahrain. I'm starting to pine for the tranquility of Fall 2008.
Posted by: Patrick | March 14, 2011 at 11:16 PM
Frances: Yes, the idea has been around for a while, but after the other Atlantic Provinces moved (PEI didn't) to the HST in 1997 it was dead politically for some time. From my reading of the literature it seems to have gotten a push from economists in particular in 2006 and 2007 in reaction to the first cut and in part in the context of the debate going on at the time about the fiscal imbalance. In 2007 it seems to have emerged again as a viable political initiative being pushed by the federal government - I have a copy of a piece in the Globe written by Jon Kesselman in October 2007 in which he reports that the feds were starting to develop a strategy linking further cuts with harmonization.
Incidentally, he starts the piece with the line: "Stupid, stupid, stupid. Those words were used repeatedly to describe the Tories' election pledge to cut the GST rather than personal and business income taxes." Which I think again suggests the alternatives at the time were one type of cut versus the other, not versus harmonization, and not versus preserving a structural balance.
Posted by: Jim Sentance | March 14, 2011 at 11:45 PM
Thanks for the plug, Stephen!
I've been doing reading for a related paper on the (un)reliability of cyclically-adjusted deficits. The EU has been having lots of problems with such measures as they feature prominently in the Stability and Growth Pact (i.e. the Pact for controlling deficits in a monetary union so that the central bank doesn't have to start buying distressed debt of the most profligate member states. You can see guess well that's worked.) I'm happy to post a few suggested readings if people are interested.)
The most interesting stuff that has crossed my desk lately is the stuff from the Office of Budget Responibility (great name!) in the UK. (http://budgetresponsibility.independent.gov.uk/) To my knowledge, they do the best job to date of estimating the degree of uncertainty associated with budget forecasts. In their June 2010 Pre-Budget Forecast, they estimate that the uncertainty associated with the 2014-15 deficit (i.e. the PSNBR) gave an 80% confidence interval with a width just over 10% of GDP. They make clear that their methodology implies this should underestimate the true uncertainty.
I'm guessing that we'd probably find something similar in Canada; over a medium-term fiscal planning horizon of 5-years, saying that the deficit will be between +5% and -5% of GDP of forecast should be right about 80% of the time.
So given the uncertainty about the projected deficit (and about the projected structural deficit), I'm wondering how the frack anyone can be expected to enforce fiscal discipline.
Posted by: Simon van Norden | March 15, 2011 at 10:59 AM