The reason is as old as civilization, and as new as the Golden report: taxes. In 1697, a tax on landlords was introduced, there to remain until well into the 19th century. As a proxy for the rental value of the house, the tax was levied according to the number of windows it contained. Landlords responded in the same way that people have always done: They arranged their affairs to lessen their taxes.

Property taxes have a long history to answer for. Yet rather than address the fundamental absurdities to which any tax on property is inevitably prone, the Golden commission held fast to this antiquated method of municipal finance. While it may look like progress that property taxes in Metropolitan Toronto would be updated from the 1948 assessments on which they are currently based, that is only if you accept the commission's underlying assumption: that property values, whether you want to call them "market" or "actual," are a fair, efficient or even meaningful basis of taxation.

It is an assumption that the leading critics of market value assessment, including the four mayors from cities within the putative Greater Toronto Area who have banded together to oppose the commission's recommendations, also share. For the proposed alternative, unit assessment, is no less flawed than market value, if in slightly different ways. There is no more reason to tax people based on the size of their house, or the value of their house, than there is to tax them on the number of windows in their house.

At bottom, any property tax, however it is designed, is caught between two fundamentally unrelated purposes. On the one hand, it is supposed to reflect the municipal services the household consumes. On the other hand, it is also meant to represent the taxpayer's ability to pay. But in trying to do two things at once, it does neither very well.

Unit assessment is at least vaguely connected to the costs of the services the city provides the householder - the cost of laying sewers, for example, will be related to the street frontage each house occupies, although other services will show no such relation. But it fails miserably on the ability-to-pay score: a family that can afford a 10,000-square-foot house in the centre of town will be in very different circumstances than a family in a house of similar size in the suburbs, and should surely not expect to pay the same tax.

Market value assessment, as far as I can tell, is wholly unconnected with anything. Two houses may have the same resale value, but cost vastly different amounts to provide with municipal services: the one a monster home on the edge of town, the other a little pied-a-terre townhouse in Rosedale.

Yet it is no better at reflecting ability to pay: A house that's worth a bundle today might have been purchased long ago, when either the price was low or the householder was rich. But for all we know its inhabitant may now be a widow on OAS. How's she going to pay her property taxes? It is especially madness to be paying for education, as we do in part, with property taxes. For so far as property values are related to income, it means the richest districts get the best-funded schools. The greatest instrument of social mobility instead becomes a means of reinforcing social cleavages.

If making one tax do two different jobs doesn't work very well, then perhaps the answer is to split the job of raising revenues between two levies, each designed specifically for the task it is asked to perform. Where it is possible to assess the cost of providing municipal services for each house, and where all or most of the benefits of the service accrue to the individual household, why not simply charge for them directly, i.e. user fees? Services of more general benefit, like the police, or which it would be unwise or impossible to charge for directly, like education, could be financed through a municipal income tax.

Rather than spend millions totting up the value or the size of every property in the city, we could simply piggyback on the existing income tax system. It would just be a matter of adding on a few tax points, earmarked "municipal." We wouldn't have to replace all or even most of the revenues the property tax currently raises - $13-billion, in Ontario: user fees could pick up the rest. Indeed, given the current government of Ontario's plans to cut provincial income taxes by 30 per cent, a municipal income tax could be slipped in without anyone feeling a thing.