The minister insists he was not proposing to levy a new tax, but only to divert some of the revenue stream from the existing gas tax to help impoverished city governments pay for public transit and similar worthy causes: what once were known as humble public works, newly dignified as "infrastructure." So read his lips: No new taxes. Just the same old confiscatory taxes you know and love. "The bottom line," he said, "is that more money has to flow to deal with the problems of cities. However we do it is fine with me." I admire the Minister's flexibility, but that still doesn't make it a good idea. It's wrong in sum, and wrong in each of its supporting premises. The cities are not short of funds. The feds are not the proper source for them. And if they were, there would be no particular reason to earmark gas taxes to that end.

First, the revenue situation. At bottom, the agitation for a "new deal for cities," a campaign that seems largely confined to the Greater Toronto Area, or more particularly the circulation base of the Toronto Star, is mostly a demand for more money for cities.

The sorts of constitutional changes that some have proposed, putting cities on an equal footing with the federal and provincial governments, are simply not going to happen.

The argument, repeated in countless news stories and political speeches, is that cities are desperately short of funds, caught between the services their constituents expect them to deliver and the limited revenues at their disposal, owing to years of short-sighted cuts by their provincial paymasters -- who in turn blame years of short-sighted cuts by their federal paymasters.

What a relief, then, to report that no such revenue shortage exists. Taking 1990 as our starting point -- at the height of the last economic boom, just to avoid any accusations of sandbagging -- governments in general have enjoyed a nearly 60% increase in revenue.

That's still roughly 10% after inflation and population growth are factored in.

And of the three levels of government, which do you think has enjoyed the fastest growth in revenues? That's right: the cities. According to figures compiled using Statistics Canada's Financial Management System, which puts the various governments' books on a standard accounting basis, local government revenues grew some 3-4% more than their provincial and federal counterparts over the last 12 years.

Now, granted, much of that money was in the form of transfers from other levels of government: The cities are unusually dependent on handouts. But if the argument is that dependence has led to penury, it won't wash. It turns out city politicians make remarkably good beggars.

What they haven't been very good at is curbing spending. Privatization, which has made uncertain progress at best in other levels of government in Canada, is all but unknown at city hall. Jobs-for-the-boys union rules still hamstring the delivery of many municipal services. And city politicians seem uninterested in changing any of it.

But why should they? City governments are the least accountable of any, for the simple reason that no one in the public pays them the slightest attention. (Go on: I dare you.

Name your city councillor. Just try.) But then, why should they? Since city governments only raise a fraction of their revenues in taxes, voters devote more of their time to keeping tabs on their provincial governments, where the real action is.

The very worst remedy for this situation is what Mr. Collenette and his ilk have proposed: to transfer still more funds to the cities, from a still more remote source. John Manley, Mr. Collenette's paymaster at Finance, put it succinctly: "I don't favour the federal government collecting taxes on behalf of cities." If the provinces want to help -- and cities are, after all, their responsibility -- they can do so quite easily. They can convert their existing transfers to the cities into a defined share of the provincial sales tax. If they really want to help, they might even make this conditional on the abolition of municipal property taxes, which conform to no known principle of public finance, being neither fair (they are not good measures of ability to pay) nor efficient (they bear little relationship to services used).

Much of what is now paid for out of city budgets could be financed through user fees -- and privatized. Transit, in particular, is ripe for this sort of reform: It is not more subsidy that will improve service and increase ridership, but more incentive to compete for customers. If congestion is a concern, let cities do the right thing at last and charge road tolls, forcing drivers to pay for the scarce road space they consume. A modest municipal sales tax could pay for whatever remained.