The distiction between spending and investment, or more properly between consumption and investment (even if you invest it, the money is still spent) has proved increasingly popular among finance ministers in recent years.

Some even suggest that money borrowed for investment should not really be counted as having been borrowed: hence the government of Quebec, for instance, which is preparing to announce that it has eliminated its operating deficit, that is, not counting capital spending.

But of course that's not quite right: even if the money you borrow is invested -- which is to say, spent -- it is still borrowed. What is more accurate to say is that governments should not borrow to finance consumption. They should borrow, if at all, only for investment -- that is, the kind of spending that offers benefits, financial or otherwise, not only at the time the money is spent, but years into the future. The thinking is that if the benefits of an expenditure are spread over several years, if not generations, so should the costs.

But is even this distinction valid? Look closely enough, and just about any expenditure will prove to have at least an element of investment in it.

Consider that most transient of household expenses, groceries. The food, it is true, is consumed that very week. But if you didn't eat properly, you'd get sick, which would cost you (or the government) more down the road. So buying groceries could be considered an investment in your future health.

Which means, according to a certain line of reasoning, it would be legitimate to go into debt to pay for the groceries.

By the way, did you notice that little sleight-of-hand I just slipped past you?

The returns to the investment in this case -- not getting sick -- are not actual benefits, in the positive sense, but merely the avoidance of some harm or cost that would arise in the future if the money were not spent. Once you learn to think this way, the concept of investment can be stretched still further. More than just another name for higher spending, it doesn't even have to be about spending at all If, for example, the government were to cut taxes, the forgone revenues could be thought of as a kind of expenditure. And if this resulted in some sort of ongoing, long-run improvement in the state of the nation, this, too, could be called an investment. In the same way, paying down the $588- billion national debt might be viewed as an investment -- the returns here are the interest charges that would have been paid on that portion of the debt that was retired.

In fact, you don't have to actually reduce the debt to earn that kind of return.

Merely by stopping the debt from growing as fast it otherwise would, restraint-minded governments in recent years have been, in a sense, investing. Which means that not only is every dollar in spending a kind of investment, so is a dollar left unspent!

What all of this confusion of terms should suggest is that it's not enough to say that something is an investment to justify allocating funds to it. Rather, the returns to society claimed for any proposed investment should be high enough to satisfy two conditions: one, that they exceed the costs of raising the money to pay for it, and two, that they exceed the returns that might be claimed for alternative investments of the same money.

One rule that should be observed in assessing these tradeoffs is that a return that can be quantified (as in, "x more dollars into student grants will produce y more graduates who will earn z higher salaries and thus pay q more taxes") is to be preferred to one that cannot: not because the latter are any less valid, but because they are less certain.

Yet more certain the return on an investment, paradoxically, the less it needs to be made out of public funds. The very capital investments that might justify public borrowing, because of their predictable long-term returns, ought for that same reason to be financed privately. The returns are typically measurable by the price charged to users: for example, the fee charged to motorists who use a toll highway. But the whole rationale for state intervention is to take into account those social returns that cannot be captured in prices: in this example, the cleaner air that results when fewer people use their cars. That calls for a discrete use of taxes and subsidies, rather than financing the whole project on the public ledger.

There is one investment of public funds, however, that offers a guaranteed return, and which only the government can make: paying down the national debt. With the average interest rate on outstanding government bonds at nearly 8 per cent, any alternative claim for public "investment," whether it takes the form of raising spending or cutting taxes, will be hard pressed to make its case.