William Watson: Andrew Sancton, what goes into your thinking about how big a city government should be?

Andrew Sancton: I don’t think we have any data, or even a theory, that suggests what the ideal or optimal size of a municipal government is. The fundamental problem is that there are different cost curves or levels of economies of scale for different municipal services. Some services, like water supply systems and sewage systems, do have economies of scale and people benefit from having larger organizations provide those services. But there’s a whole range of other municipal services for which there is certainly no evidence of economies of scale, and some evidence that there are diseconomies of scale.

William Watson: Where does that lead you? To a conclusion that “anything goes”?

Andrew Sancton: It leads me not far off that position, to the view that we’re probably better off, in most large metropolitan areas, to have quite a wide array of municipal governments, some relatively large, some quite small. It certainly leads me to oppose big reorganization schemes that are designed to save money.

William Watson: Fernand Martin, what’s behind your general thinking in this area?

Fernand Martin: I agree with Sancton that economists have discovered no optimum size for cities. What we have is a rule of how cities should grow. And that rule is that cities should grow as long as their negative externalities are not subsidized. If cities’ bad effects are somehow subsidized, they may become too big. So our rule is: don’t subsidize negative externalities. Transport is a case in point. We know that there is congestion in large cities, but if you constantly subsidize infrastructure to reduce congestion and don’t make users pay for it, so they are not conscious of the costs of this congestion, then cities may become “too big.” In other words, what you need is the user-pay approach.

Andrew Sancton: I think we have to distinguish between optimal size for urban areas, or cities, with a small “c”, and the optimal size of municipalities. These are different questions. It’s certainly true that government subsidies that encourage people to live in huge metropolitan areas might be dysfunctional, but that’s a different question from the size of the municipalities within those metropolitan areas.

Fernand Martin: I’m no expert in government. However, I’ve been reading my former colleague Albert Breton for a while, and also about the European experience on this. The idea here is of course the principle of subsidiarity: those governments which are nearer to the people should have the responsibility either of producing, or at least organizing the production of these public services.

There are two ways to go here. The first is the public-choice approach, for which Mr. Sancton is a propagandist, and which I think is OK. When it comes to returns to scale, the small municipalities he has in mind are usually complemented by recourse to special districts. In other words, for police and other services where there are large economies of scale, you can group a lot of municipalities into a common jurisdiction. So public choice plus special districts does the job.

At the other end of the spectrum, you have the monopoly approach, as in the amalgamations undertaken by the Quebec and Ontario governments. But here in Quebec, where you have a monopoly, they have tried, through decentralization in terms of boroughs, to get back to the people. Although this is a rough summary, I presume that the two systems converge. Eventually the monopoly is diluted by a lot of decentralization, and the local government that emerges may be satisfactory. And the public-choice approach achieves the economies of scale through special districts.

William Watson: I sense Andrew would have a lot to say about that but before he does, can I just ask Fernand whether the corollary to not subsidizing the negative externalities is that you should subsidize the positive externalities?

Fernand Martin: Yes, you should subsidize positive externalities. It’s like trying to figure out the optimal size of a private firm. How big should a firm be? It depends on the type of production it’s involved in. With all the R&D it does, Microsoft can achieve things smaller entities can’t. It’s the same with cities. Some cities, like New York or other large metropolises, can play host to financial activities, say, that cannot be done on a small-town basis. Some activities that exhibit the “economies of agglomeration” are needed for financial centres and they cannot be done easily—they can be done, but not easily—by a lot of small cities. Of course you’ll mention Silicon Valley. But Silicon Valley is usually regarded as a special case, not reproducible elsewhere.

William Watson: Just one more detail on that: Does that line of thinking move in the direction of having a single municipality, or is it simply that some other government has to look to these externalities and make sure that they are taken account of?

Fernand Martin: If you want to get the right output, you may not be able to co-ordinate it without a dictatorial or powerful government at the top. For instance, in Montreal we have three transport commissions: the South Shore, the Island of Montreal and Laval. But they are not co-ordinated. Eventually they’ll have to come to some kind of a superior power to coordinate their activities. In other words, even if they can each exploit the economies of scale, it would be even better if their activities could be co-ordinated.

Andrew Sancton: As I understand this issue, it’s clearly the case that large financial organizations—banks and other financial institutions— appear to need large metropolitan areas, large cities, in order to get all the services they need so as interact with each other. And you’re not likely to set up headquarters of a major Canadian bank in Wiarton, or some similar place in Ontario, because it’s just too small. But I’m not aware of any literature that argues that there is a connection between the size of the municipality in which these financial institutions operate, and their success or lack of success. Fernand pointed to New York City, which is a large municipality, although it’s surrounded by hundreds of suburbs. But let’s look at London, England. There is the classic case of a world centre of finance located literally in the City of London, which is one square mile and is a tiny municipality. So I just don’t think there is a connection between size of municipality and health of financial institutions. There is a connection, of course, between the size of cities and financial institutions.

Fernand Martin: I lived in London for a while, and it’s very hard to compare the English unitary system of government with what goes on in Canada. In Canada, we may need a mayor to have certain things happen. In London, things happen through their own system, but they have a way of bringing about the same results. Which is why I would like to study the convergence between the two systems. I agree with you that with simple regressions we cannot connect size of municipality and the associated growth of financial centres, at least not for the moment. But there might be a connection. In France, certainly, they think they need a central power in order to achieve these things.

Andrew Sancton: Banking and finance obviously are extremely important to Toronto and to Canada as a whole. But the fact that the City of Toronto is now two million people instead of six hundred thousand is, I think entirely irrelevant to the health of the Canadian banking industry.

William Watson: Andrew, what’s your opinion of Fernand’s proposition that decentralization within a single government will eventually converge to the solution that would be produced by the co-ordination of lots of small governments?

Andrew Sancton: This is an argument that has always accompanied proposals for amalgamation. It’s almost inevitable that people who say we’ve got to amalgamate municipalities also say that we can both amalgamate to capture the benefits of amalgamation and at the same time decentralize to capture the benefits of small organizations. This argument was made in the Winnipeg and Halifax amalgamations, and it has been made elsewhere. But these decentralist institutions have great difficulty surviving in a centralized system. It’s true that the proposal for Montreal goes further in this direction than others have, but we’re still looking at unified work forces and a single corporate entity. I think it’s going to be very difficult for any city to have both amalgamation and serious political decentralization at the same time.

All kinds of American metropolitan areas have dozens, even hundreds of small municipalities and things get done. Why would you need to amalgamate everything and then try to recreate the benefits of smallness when to me the obvious solution is to go the other way: keep the institutions that are small and work out arrangements for capturing economies of scale?

William Watson: Fernand, I’m not sure you’d necessarily disagree with that.

Fernand Martin: No, certainly not, but when you have a lot of small municipalities, their co-ordination is eventually done by a superior government—by a provincial or a federal government. There is a transaction cost there. In other words, the choice is whether you do the co-ordination locally—and I agree with Professor Sancton that there might be some problem here—or do it, in our case, in Quebec City. Which is more costly? That’s the crucial point. I’m not defending amalgamation, nor proposing it. I haven’t studied it seriously enough. But that’s the nature of the choice we have to make.

Andrew Sancton: I agree with that, but let’s be clear. There are huge transaction costs within large organizations. And that’s where some of the diseconomies of scale come from: when distant parts of large bureaucracies try to communicate with each other.

Fernand Martin: But, look, the same problem arises in companies in the private sector. I have studied the pharmaceutical industry in great detail. It has small firms and big firms, real giants. And over time some companies transform from one into the other. There are cycles where they downsize, and when they grow big again. Economists are not married to a unique size of firm. It depends upon a thousand circumstances. And it’s probably the same thing in municipal affairs.

Andrew Sancton: I tend to agree with that. Let’s revisit the public transit issue in Montreal, which you raised. Here we’re going through a gigantic municipal amalgamation—actually two, in the Montreal area, if you include the South Shore—and yet it doesn’t solve the problem that you raised about public transit.

Fernand Martin: I agree with you. One criticism that I did not put in my report, because it wasn’t my job, is that you have to eventually go all the way, or do nothing. Because, in both cases, only half of the population is organized into an amalgamated city.

William Watson: Albert Breton and Tony Scott’s book on economic constitutions suggests that the optimal constitution is continually a trade-off between economies of scale and coordination costs. Are we just aware that these costs are out there and we have an intuitive feeling of what they may be? Or do people actually go out and measure them?

Andrew Sancton: Yes, people have measured them. It so happens that most of the people who have done the measuring, in particular in the US, have been advocates of the public-choice position. Now Professor Martin calls me a propagandist for public choice, which I don’t really mind him saying, though the only reason I fell into this area (and I did almost fall into it) is that people were making all kinds of claims about how amalgamations save money. If you go into the academic literature, you find that there has been research on this, and that the savings that are supposed to result from size don’t appear to be there. As I say, most of the people who have done the measuring come from the public-choice school and they have a theoretical explanation about why those savings aren’t there—they either don’t exist in the first place or they’re eaten up by rent-seekers—and that makes some sense to me.

Fernand Martin: But the literature you are reading is American. And of course there is a bias in the United States toward competition, so it’s no wonder that public choice is in favour. It’s perfectly normal that most of the literature would be about public choice and the merits of public choice. But if you go to Europe, it’s a different story. I recently heard a seminar on how things are done in France, how they operate with les Communes and how they produce all sorts of arrangements, without ever doing an amalgamation. The large units, like Paris, stay what they are. So the acceptability of this kind of government depends upon the world you’re operating in.

Andrew Sancton: Yes, but there are studies in other European countries, in England, for example, that certainly do not show that there are major savings from amalgamation. In fact, most of them tend to show that costs actually go up with the amalgamation. So the result is not restricted to the US.

William Watson: Fernand, you were asked to do a study of the Toronto case for the Montreal transition team.

Fernand Martin: On the condition that I would not have to defend amalgamation or its rival! They wanted me to find out what has happened.

William Watson: What’s your view of what’s happened?

Fernand Martin: That it’s a non-event because nothing has changed. My study was on the budgetary effects of the amalgamation. On the budgetary side, I can’t see what has happened because in constant dollars expenditure hasn’t changed since 1993. So, for me, not much has happened.

William Watson: But is amalgamation about total budgetary spending or is it about services delivered per dollar spent? The proposition is that this ratio would go up, isn’t it?

Fernand Martin: But we cannot measure the services because there is no way to develop an index number for them, since the output is not sold in the market. Municipalities are multi-product firms. When they are private firms, that is when their output is sold on the market, we can sum up these things and find out whether the output has increased or decreased. But in the case of municipalities, the output is usually not sold in the market and its cost of production may have something to do—or nothing to do— with its value, as appreciated by the people who consume it. If we cannot measure that, we are in the realm of opinions and anecdotes.

Of course, all sorts of people say “My services have gone down,” but there are others that may have gone up. The official papers that the City has produced—but which Professor Sancton has challenged in his recent article in the Montreal Gazette—list all sorts of new services that they have put in. I wouldn’t be surprised if they have cancelled services, but these may have been replaced by other services, or may not have been used much to begin with. In Montreal, for a long time, we had a policy of building baseball parks for kids. Well, apparently, in Montreal, because of the immigrants, baseball is not as popular as it was. Now kids want soccer fields. So we have closed some baseball fields and put in soccer fields. If you look only at one side of the thing, you say “Hey, the new Bourque administration is cutting recreation services because they have shut down baseball fields.” But the other side of the coin says “Yes, but they have put in new things.” How do you balance that? Is a soccer field equivalent to a baseball field? I don’t know.

Andrew Sancton: I agree that the problems of measurement are immense. They really are extremely difficult. But on the subject of soccer fields vs. baseball fields, I would have thought this is a perfect example of why it’s better to have smaller organizations that can be more sensitive to the demands of people in a particular area because people’s tastes for soccer and baseball are going to vary territorially. I know the argument can be made that maybe the borough councils in Montreal are going to be able to make those kinds of provisions, but that remains to be seen.

On the issue of Toronto, notwithstanding the difficulties of measurement I just want to make two points. First, the Toronto amalgamation has been justified in terms of saving money. That’s what the provincial government was trying to do. That’s the absolute prime motivation for implementing this policy. And Professor Martin’s characterization of it as a nonevent—which is justifiable looking at the numbers that now exist—is a very peculiar kind of conclusion, not from his point of view, but from the point of view of the government: We’ve gone through all this and there has been no change? From the Ontario government’s perspective, if we accept Professor Martin’s analysis, that means the policy has been a failure.

But my second point is that there is all kinds of other evidence, outside the official numbers, that also point to failure. The fact is that the complete harmonization costs of services, and particularly wages and benefits, are not in the official figures yet. On the service side of harmonization, it will be very difficult to get figures that everybody will agree on. On wages and benefits, it’s still difficult but a little more manageable.

William Watson: Is that because the harmonization hasn’t been completed yet?

Andrew Sancton: It’s still in process and there are significant costs related to the harmonization of the benefit packages. At the same time, of course, people are getting normal increases associated with inflation. So you have to separate out those kind of across-the-board increases from the harmonization increases, which is very difficult. A significant amount of money is involved in that but the bigger point is that almost everybody acknowledges that things are dramatically wrong in Toronto. It’s not just the provincial downloading, the extent of which is in dispute. It’s not just Mayor Lastman’s recent political difficulties. The problems stem from the fact that a massive organization is still—four years after the fact—being integrated and turned into an organization that works, and every step along the way there are problems. They still haven’t worked out where they are going to put people. There are still big issues about whether the Metro hall is going to be sold and where people’s offices are going to be.

Fernand Martin: Yes, but it doesn’t show up in the costs of running the City.

Andrew Sancton: It shows up in the sense that this massive bureaucracy can’t do anything because it’s constantly in turmoil about simple things—about getting itself established and going. I don’t think you can separate the fact that so many top officials have either been released from their job or have resigned. That is connected to the problems of amalgamation.

Fernand Martin: Yes, but it happens in private firms also. In the case of Nortel, or whatever, there are sometimes shake-ups in the management and yet things continue. What the shareholders are interested in is the profitability of the firm. Whether they are fighting at the top, as long as it doesn’t show up in losses, is immaterial.

Andrew Sancton: But a year or so ago, I would have acknowledged that most of the so-called shareholders in Toronto—the voters, the residents— were fairly happy. They thought amalgamation was kind of a non-issue. It happened; things were going along. But in the last year, the wheels have fallen off. Everything seems to be going wrong. There is a financial crisis. It’s true there is a debate about what the origins of the financial crisis are. But Toronto is a disaster. And the fact that anybody would be looking at Toronto as some kind of model to follow is beyond my comprehension.

Fernand Martin: OK, I agree with you, at least in a sense. But still it doesn’t show up in the budgets. That’s the only thing they asked me to look at.

Andrew Sancton: OK. Fair enough.

Fernand Martin: On the other hand, Professor Sancton argues that the official figures might be wrong. I don’t know if he could explain.

Andrew Sancton: It’s simply that Toronto has put out figures about the costs of the downloading and then the province put out figures of its own that are different.

Fernand Martin: But I presume that external auditors look at the books of the City.

Andrew Sancton: Toronto is in such bad shape now that they have not been able to prepare their financial statements for last year to be externally audited. There is very considerable chaos in…

Fernand Martin: Are we dealing with a case like Bre-X?

William Watson: There is a tax base in Toronto. There was no gold in Bre-X’s hills.

Andrew Sancton: But it is difficult to get numbers, even related to the downloading question. My only point about your study was that it didn’t acknowledge that the numbers were contested. I don’t know what the right numbers are, really. All I know is that it’s very much in the City’s interest to say that the problems have resulted from the downloading because Mayor Lastman and the majority’s position on Council now is that amalgamation was a good thing, and they have made the right decisions about harmonization and this sort of thing. Now you can say that, because of my express public pronouncements on this kind of thing, it’s in my interest to say that the problems have resulted from the amalgamation. But the point is that all of these numbers are contested, and on the downloading, there are clearly two separate sets of numbers.

William Watson: If I could go back to the general problem of evaluating the success of the amalgamation for a moment, what we would need is an index that would include all the services that it’s possible to offer?

Fernand Martin: As I said in my study, this thing is set out in the political market, not in economic terms. So in the next election, it’s up to the opposition party to raise these issues and find out what the people think about these things.

William Watson: Is it a satisfactorily responsive market, however?

Fernand Martin: Look, if the economists cannot do anything there, as I showed in my work, then it’s up to the political scientists to step in and say: this is a political problem, not an economic problem.

Andrew Sancton: But let’s just look at what the constraints are on a political response. And I know this is an issue in Quebec, as well. I’ve been asked this question many times: if one finds that an amalgamation isn’t working, can it be undone? Let’s assume there is agreement that there have been problems with the Ontario amalgamations and that they should be undone. I’m not even sure that I would advocate doing that, because we’ve just spent huge amounts of money on transition costs to amalgamation, and you’re going to have the same kinds of transition costs back again. So these are immense decisions and they cannot easily be un-made.

Fernand Martin: No, but a new administration can probably solve the problem of co-ordination and the fight at the top in running the business. This is surely not an impossible management problem, to run a city like Toronto. Bigger cities are run by bureaucracies too, in Europe and elsewhere, and I think Toronto could do that. Are you aware, just in passing, that Jean Charest in Quebec has promised to undo all this?

Andrew Sancton: I am aware. That is why I raise the question.

Fernand Martin: I live in Saint-Bruno, and last week the Municipal Council passed a resolution holding Charest to his word that if he’s elected, there should be a referendum here, and we should get out.

Andrew Sancton: In some respects, the most important aspect of the borough system in Montreal might be that it at least provides some kind of institutional shell that could allow a form of de-amalgamation. But anywhere else, including Toronto, the costs of de-amalgamation would be immense—although everybody should know that people are debating de-amalgamation in the United States. There’s the separation of the San Fernando valley from Los Angeles, which is very much on the political agenda there. A similar movement in Staten Island in New York seems to have gone into quiescence. But it is interesting that Americans, for whatever reason, seem to be spending more attention on making their bigger municipalities smaller, while we’re concentrating on making them bigger.

Fernand Martin: Yes, but it’s a matter of social climate. Here we are more in support of redistribution than they are in the United States. In many states in the United States, the constitution does not allow forced amalgamation.

Andrew Sancton: Speaking of redistribution, I think that is obviously a crucial issue in Montreal. It was not an issue at all in the debate in Toronto, although now the amalgamation is being justified, after the fact, in terms of redistribution. But to me creating these large organizations is a very peculiar way of implementing redistribution. There are other ways of doing it, if that’s what the objective is.

William Watson: That raises the question of what are the main motivations behind the Montreal amalgamation? What do you think are the principal motivations?

Fernand Martin: Well, frankly, I have no idea! I can’t see any reason. There are reasons, no doubt, but not reasons that concern efficiency.

William Watson: If the stated motivation is efficiency, you would say it’s unlikely that there are any efficiency gains to be made.

Fernand Martin: I have only one case, you know. I cannot generalize my study. If I can talk as a scientist, I have not much ground from which to extrapolate and say what will happen to Montreal. But three-quarters of the activities of the city are already amalgamated, somehow. There is not much that can happen in the last quarter. The last quarter has not been amalgamated because probably there was nothing there. Either in terms of economies or diseconomies.

William Watson: When you talk about economies of scale already having been realized you’re referring to the Montreal Urban Community?

Fernand Martin: I’m referring to the police, transit and things like that. The big things have already been amalgamated. So, as an economist, I say the rest is probably politically motivated. Which might be a very good reason. But I am not an expert to judge that.

Andrew Sancton: This is one of the great public policy mysteries. I can understand why, earlier in this decade, in Halifax for example, people might have thought they were doing something innovative and might save money with amalgamations. But there is no evidence that any of these jurisdictions are learning from each other. Except that I do think the Montreal amalgamation would not have taken place if the Toronto one hadn’t happened. It’s not that the policy-makers in Quebec City necessarily thought Toronto was a big success and it had to be emulated in Montreal if Montreal was going to compete. I think they thought the amalgamation was a good thing, possibly for reasons of redistribution, or administrative neatness, and the Toronto amalgamation somehow provided political cover. Because they did it in Toronto, nobody can say that— to be frank about it—nobody can say that we’re attacking anglophones in Montreal by abolishing their municipalities. The other thing, of course, is that it’s happening in many other parts of Quebec as well.

William Watson: Which makes it a high price to pay to eliminate the anglophone communities.

Andrew Sancton: Apart from that, it’s very hard to find any reason for it. You could look at redistribution, which I think is a serious issue, but already, certainly by North American standards, as Professor Martin said, a huge amount of integration of the main services had already taken place. Central-city mayors in the United States would be thrilled to have suburban municipalities paying for the cost of a police force covering a wide area. They might not want to give up control over their central-city police force, but they would be thrilled to have that suburban money flowing in for it. They already have that in Montreal. So the redistribution that we’re talking about here is on a relatively small scale.

You have to wonder whether this a symbolic policy? You know, “We just think that everybody should pay the same taxes and have the same level of service.” If that’s the argument, then, of course, we also hear about how the boroughs are going to be there and enable people to have different levels of service. So the mystery deepens!

William Watson: Let me thank you both for participating in this discussion of what for Montrealers has become a very heated question.

Photo: Shutterstock

William Watson
William Watson is an associate professor in the Department of Economics at McGill University, where he served as chair from 2005 to 2010.
Fernand Martin
Fernand Martin is Professor of Economics at the  Université de Montreal and a specialist in regional economics.
Andrew Sancton
Andrew Sancton is Professor of Political Science at the University of Western Ontario and specializes in urban politics and government.

You are welcome to republish this Policy Options article online or in print periodicals, under a Creative Commons/No Derivatives licence.

Creative Commons License