I was on CPAC Sunday evening, on a roundtable with two other economists: David MacDonald (who coordinated the alternative budget for the Canadian Centre for Policy Alternatives); and Glen Hodgson (Chief Economist for the Conference Board). [You can watch it here – SG.]

The topic was Canada's projected $50 billion federal deficit; should we be concerned? I said I wasn't particularly concerned about the deficit, though I would have liked to have seen a more aggressive monetary policy replace part of the fiscal stimulus. David was a little more fiscally dovish, and Glen a little more fiscally hawkish (ignoring lots of more subtle differences of opinion). But none of us three economists wanted a balanced budget now.

But it was a phone-in show, and many of the callers were against any deficit, large or small.

If there were ever any time the government should run a deficit, now would seem to be the time (outside of a major war). At first sight, the idea that the government should never run a deficit seems totally incoherent. You can't always run a surplus. If you did, the national debt would go to zero, and then go negative, we would owe the government money rather than vice versa, and the government would eventually own everything. As I said before, fiscal conservatism, taken too far for too long, eventually leads to communism.

But perhaps fiscal conservatism is more about debt than deficits; it's the level of debt, not the rate of change of debt. "If you have any debt, it's never OK to run a deficit; but if you have a stock of savings, it's OK to spend part of that stock of savings in bad times".

I care about both debt and deficits; the level and the rate of change. One of the reasons I am relatively sanguine about the projected $50 billion deficit is that Canada's debt/GDP ratio is about 30%. That's low compared to recent history, and to other countries. If we had a much higher debt/GDP ratio, I would be more concerned about a $50 billion deficit.

But is a 30% debt/GDP ratio low enough? Why shouldn't it be 0%, or even negative? "Just because all the other countries do something stupid, and you used to do things that were even more stupid, doesn't make it sensible!"

What is the optimal debt/GDP ratio? Let's face it, economists don't seem to have a very good answer to that question.

Robert Barro gave a sort of answer. If the debt/GDP ratio is predicted to grow over time, eventually you will need to raise tax rates to avoid default. If the debt/GDP ratio is expected to fall over time, eventually you will need to cut tax rates to avoid communism. And it's not good to plan to have either rising or falling tax rates. That's bacause taxes distort incentives, and the marginal deadweight cost of those distortions increase with the marginal tax rate. So you minimise the present value of those deadweight costs by planning to keep tax rates constant in the future. This "tax smoothing" argument means that if times are currently about average, you set tax rates at whatever level is needed to keep the debt/GDP ratio the same in the future as it is now.

That's why I call Barro's answer a "sort of answer". His tax smoothing argument does not define an optimal long-run debt/GDP ratio. If unexpected bad times cause the debt/GDP ratio to rise, you don't raise taxes enough to bring it back down to where it was; you just raise taxes enough to match the extra interest payments and stop it rising further. And if unexpected good times cause the debt/GDP ratio to fall, you don't cut taxes enough to bring it back up to where it was; you just cut taxes enough to match the lower interest payments and stop it falling further.

I'm not happy with Barro's answer. Some years ago I wrote a paper with my Carleton colleague, Vivek Dehejia. We argued that there is a limit to the government's taxing capacity. There is a Laffer curve out there somewhere. If you try to raise tax rates too far, eventually you hit that capacity, where the economic activity you are taxing falls so much in response to higher tax rates that the tax revenue falls. That means that not only does the marginal deadweight cost of tax revenue increase with the tax rate (as Barro assumed); eventually it must increase at an increasing rate. This must happen, because when you approach the limit of the Laffer curve the marginal deadweight cost of taxation approaches infinity.

So Vivek and I argued a precautionary motive for government saving, and slowly paying down the debt. This didn't mean you could never run deficts in bad times; but it did mean you should aim to slowly reduce the debt/GDP ratio on average.

But then our argument didn't really answer the question either. Sure, it made sense that the government should aim to keep the debt/GDP ratio well below the maximum sustainable amount, just in case any really bad time came along, or a long string of smaller bad times. You should aim to reduce the debt/GDP ratio if it got too high. But how low is low enough? When should the government stop trying to save as a precaution against future emergencies?

Is an unused borrowing capacity a large enough war chest to handle possible future emergencies? Or should the government pay off all the debt? Or should it have an actual war chest, with a positive net asset position, so the debt/GDP ratio is negative?

I don't have a good answer to that question. And I don't know if anyone else does. And until there is a good answer to that question, it's hard to argue with anyone who thinks that deficits are a bad thing, even in bad times, because a 30% debt/GDP ratio is still far too high. Or that deficits are a good thing, because 30% is far too low.

(For a time I thought that the optimal debt/GDP ratio might be positive, because there was a strong demand for safe, liquid, government bonds, even at low interest rates. But that is an argument for a positive gross debt, not net debt, or debt minus assets owned by the government. It could be satisfied by the government acting like a financial intermediary — both borrowing and lending.)