Greater Toronto Area:
A Review of Alternatives
Report Prepared for the City of Toronto
Also at City of York
10 January 1997
This report compares two approaches to governance reform in the Greater Toronto Area
(GTA). The first is the Megacity model, as identified in the Ontario Government's City of
Toronto Act 1996 --- currently before the legislature. The Act would amalgamate Metro
Toronto and its six local cities into a single city of 2.3 million people. The second
approach is identified here as the Mayors' model --- a synthesis of reform proposals put
forward by various GTA mayors in which local governments are strengthened but limited
in population to well under one million.
The comparison draws on experience in the United States, where there is a wider range of
city sizes than in Canada. The experience is used to evaluate the models on the basis of
two key principles advanced by David Crombie's "Who Does What" advisory panel:
Democracy, Accountability and Responsiveness
The consideration of the "Democracy, Accountability and Responsiveness" principle
analyzes the difficulty larger local governments (over one million population) face in
sustaining viable democratic processes. The three US megacities over two million
population have faced particularly intractable problems. Two of the three (New York and
Los Angeles) are contending with advancing secession movements --- sparked by the
belief that remote city halls were ignoring local neighborhood needs. The third megacity
(Chicago) has lost almost one million residents,(1) who seceded "with their feet."
So far as the "Efficiency" principle is concerned, the US experience points to
substantially higher costs for cities in the population range over one million. These cost
penalties are typically in the 20 percent to 40 percent range and can exceed 100 percent
in the case of amalgamated municipalities. For residents and businesses of these larger
cities, cost premiums mean higher property taxes and/or lower service levels.
There appears to be no valid reason why GTA cities should not remain well within the
efficient urban population range (above 100,000 and below 1,000,000), a range providing
for high quality services with lower tax burdens. For services delivered more efficiently
over a larger area, the government's newly announced
GTA coordinating board (Greater Toronto Service Board) will provide an effective
mechanism to ensure the delivery of efficient, quality services over a wider area.
The existence of this Board makes the existing five upper-tier (regional) governments
redundant and invalidates any apparent case for megacities --- whether in the "416" area
or the "905" area.
Local and regional governance has been the subject of considerable attention over the last
year. Various proposals have been advanced, especially with respect to the Greater
Toronto Area. The debate has been narrowed to two basic approaches --- the Megacity
model and the Mayors' model.
1. Megacity model: The Megacity model has been proposed by the provincial
government. The six cities located within the Municipality of Metropolitan
Toronto (Metro) and Metro itself would be amalgamated into a single "megacity."
Legislation to implement the Toronto "megacity" was tabled on December 17,
1996.(2) Further, a Greater Toronto Area coordinating board, the Greater Toronto
Service Board (GTSB) would be established to coordinate regional services. The
structure of Greater Toronto Area local governments outside Metro would be
reviewed in early 1997. Some municipal amalgamations could be proposed, and
eventually a two-tiered governance structure would be established by abolishing
regional governments (Durham, Halton, Peel and York) and amalgamating
existing cities into larger, stronger cities.
Mayors' model: The Mayors' model consists of two separate but similar
proposals, The Governance Solution for the Greater Toronto Area(3) by Mississauga
Mayor Hazel McCallion and Change for the Better(4) by the Mayors of the six cities
within Metro. These two proposals would reform governance GTA-wide by
strengthening cities, holding population to a range under 1,000,000 each, and
establishing a Greater Toronto Area coordinating board to oversee regional
services. The coordinating board would be composed of elected officials from the
local governments. The present regional governments would be abolished.
Similarities and Differences
The two models are similar in their approaches to regional public services. Both the
Megacity model and the Mayors' model envision a greater Toronto coordinating body
with powers limited to coordinating specified regional services (GTSB). And the two
proposals share similarities insofar as cities would be strengthened. The key difference is
that in the Mayors' model, city sizes would be similar both within and outside Metro (the
"416" and "905" areas, respectively); in the Megacity model, there would be a major
imbalance between the inner city (2.3 million population) and the "905" cities (all well
under one million population).
In developing its proposal, the government established the "Who Does What" panel to
review alternatives.(5) The "Who Does What" panel employed the following four
principles in its deliberations. These principles are appropriate for evaluation of the two
1. Democracy, Accountability and Responsiveness: Municipal government is a
democratic institution fundamental to local political decision-making. Its structure
should be as understandable as possible to promote public access, participation,
and accountability. It should respect and accommodate diversity and be
responsive to the needs and preferences of communities.
2. Fairness: The structure should ensure that costs and benefits are shared fairly
across the entire community.
3. Efficiency: The structure should allow services to be delivered by the lowest level
of government that has the capacity to do so effectively. It should also be more
cost-effective than the current system, delivering maximum value with available
4. Coordination: The structure should encompass the interests of the entire
community. It should support the strategic coordination of certain key services
and foster an approach to decision-making which integrates economic,
environmental and social considerations.
This report analyzes the two models primarily in light of the first and third principles.
The second principle should be left to Toronto residents themselves to apply, since there
are many concepts of fairness. The fourth principle, coordination, can be achieved with a
wide variety of governance structures and would not appear to be critical to the
differences between the Megacity model and the Mayor's model. However, it should be
emphasized that the Megacity's potential to dominate Greater Toronto Area regional
processes would probably reduce the likelihood of "905" cooperation within the GTSB
(see further discussion of the GTSB, below).
DEMOCRACY, ACCOUNTABILITY AND RESPONSIVENESS
Other things being equal, smaller governmental units are more democratic than larger
Smaller governments are more accountable: Smaller governments tend to be
more accountable and responsive to their citizenry. They are more accountable to
citizens, because the individual citizen has more "voice" in a smaller
governmental unit. For example, a voter's voice in a city of 100,000 population is
10 times as strong as in a city of 1,000,000. Bigger government is more remote
from the electorate and is, by definition, less accountable and less responsive.
Smaller governments are more responsive. As governments increase in size,
processes and communications necessarily become more bureaucratized (more
rigid). As government processes become more rigid, they become less
understandable to the individual citizen. This discourages people from addressing
issues with their government. This less efficient and effective feedback process
often results in smaller problems escalating into crises, as it is only when
circumstances become unbearable that citizens have sufficient impetus to deal
with the overly complex processes typical of more remote governments. With a
less efficient and effective feedback system, the quality of government services is
likely to decline.
Larger governments are more susceptible to special interests: This is so for
three reasons. First; special interests have the financial resources to hire
professional advocates (such as lobbyists) to learn, understand and manipulate the
rigid processes of larger governments. Conversely, individual citizens and
neighborhood groups rarely have the financial resources to hire professional
advocates. Second; there are economies of scale with respect to political advocacy
--- it is simpler and less expensive for special interests to influence a larger
government than multiple smaller governments. Third; the more diffuse voice of
the electorate makes larger government more susceptible to special interest
Smaller local governments are more attuned to communities and
neighborhoods: Regional governments are necessarily more sensitive to broader
geographic issues than local, community or neighborhood issues. This is because
regional governments include a larger number of communities, which diminishes
the voice of each such community in the political process. Individual,
neighborhood, and community issues are likely to be less effectively addressed by
larger, rather than smaller governments. As a result, regional governments are not
appropriate for local governance.
Large governments are less controllable: Larger governments tend to be more
difficult for policy-makers to control. As governments become larger, elected
officials must rely to a greater degree on their staffs and are less well positioned to
effectively exercise their oversight function.
The municipal secession trend: The political disenfranchisement that occurs when
governments are too large is feeding a growing municipal secession movement. Strong
succession movements have begun in two of three U.S. megacities (cities with more than
two million population):
New York City's borough of Staten Island is well advanced in its process to
secede and establish itself as an independent city (population: 400,000).
Proponents predict a binding referendum before the end of the decade.
A strong secession effort has begun in the San Fernando Valley of Los Angeles,
which would create a new city of more than one million people, dropping the
population of Los Angeles to two million. State legislation to facilitate the
secession process was narrowly defeated in 1996 and will be reconsidered in 1997.
The proposed secession is also emerging as a major issue in the mayoral election
due later this year.
A consequence of diluted democracy is that governments in larger U.S. cities have
generally been less successful in delivering quality public services to their residents. This
has contributed, along with other factors, to a virtual population hemorrhage. Chicago,
the only U.S. city of more than two million population that is not facing a secession
drive, has lost nearly one million residents since 1950.(6)
The municipal secession movement should be considered in the context of its daunting
challenges. Municipal enabling legislation tends to require super-majority (more than 50
percent) referendum results and favorable municipal boundary commission decisions, in
addition to well financed advocacy campaigns by the larger government to preclude
Shall Metro residents have local government? In each community there are both local
and regional public policy issues. Local issues are most competently addressed by local
governments, and regional issues by regional coordinating bodies. But the Megacity
model would leave Metro residents without local government and with two levels of
government that are effectively regional in scale (megacity and GTSB). Local
government is required within Metro just as much as it is required outside Metro.
The stronger city governments proposed by the Mayors' model would be more
accountable and responsive to their electorates than the amalgamated regional Metro
government. Consequently, these cities would respect and accommodate diversity and be
responsive to the needs of communities and neighborhoods more than would one Metro-wide government. Cities averaging 400,000 population are, by definition, more
accessible to their citizens than an amalgamated regional government of 2.3 million.(7)
Regional public policy issues would be handled, under both proposals, by the GTSB.
The Megacity model approach to governance within Metro is anti-democratic. More
remote government is not more democratic government; to the contrary it is less
democratic. To paraphrase Abraham Lincoln:
Government of the people, by the people and for the people is government that is
closer to the people.
Controlling costs is a significant challenge in government. This is because competitive
incentives, which effectively regulate costs in the private sector, are largely non-existent
in government. Virtually all of the international evidence indicates that cost control
becomes more challenging as the size of government increases. For example:
Higher costs: Larger governments generally have higher unit costs than smaller
governments. According to the U.S. government's Advisory Commission on
Average costs are generally understood to have a u-shaped relationship to
scale of production. As production increases from zero, average costs
initially decline, then level off, and finally after a point begin to increase.(8)
U.S. cities of more than one million population spend 21 percent more per capita than the cities with 500,000 to 1,000,000 residents, 18 percent more per capita than cities with 100,000 to 500,000 residents and 13 percent more than the average city (Chart: US City Expenditures per Capita by Size).(9)
U.S. counties with more
than one million
population spend 42
percent more per capita
than counties with
300,000 to 500,000
residents and 30 percent
more than the average
U.S. primary and
secondary school districts
with more than 50,000
students spend 18 percent
per pupil more than
school districts with 15,000 to 25,000 students and six percent more than
the average school district.(11)
U.S. public transit systems operating more than 1,000 buses have costs per
service hour 26 percent higher than systems with 300 to 500 buses and 19
percent higher than the average public bus system.(12)
In all of the cases above, the cost of larger government units is greater than that of
both medium sized units and average sized units. There is no evidence that the
higher cost structures of larger cities is significantly related to factors such as
demographics or service levels , since many smaller US cities exhibit similar characteristics.
Diseconomies of management scale: As governments increase in size they
require additional layers of management and support personnel, further increasing
One important variant of economies of scale is Diseconomies of scale in
management. As the size of a provision unit increases, beyond some point,
scale economies attained as a technical matter of production may be offset
by management difficulties that multiply as the provision unit attempts to
organize more production in house.(13)
Greater resistance to innovation: Smaller governments tend to experiment more
than larger governments, and they are more innovative. A healthy competition
develops between local governments as the media and citizens take note of
comparative efficiency and effectiveness. As government increases in size, there
are fewer comparable governments against which to compare their performance.
In addition, the larger bureaucracy and stronger trade unions tend to more
successfully resist innovation. Internationally renowned urbanologist and Toronto
resident Jane Jacobs cited discouragement of innovation in expressing her
opposition to the Megacity model at a recent public meeting at Toronto City
Greater resistance to legislative reform: Larger governments are more difficult
for legislatures (such as the Ontario provincial parliament) to reform. Larger
governments are able to marshal considerable political and financial support to
maintain the status quo.
Government amalgamations do not produce lower cost government. The evidence is
clear that amalgamated governments are not more cost efficient. For example:
A number of city-county
amalgamations have occurred in
the United States. In each case,
the amalgamated government
represents only a part of the
metropolitan area (as would be
the case with an amalgamated
Toronto). U.S. amalgamated
cities of more than one million
population spend 112 percent
more per capita than the
amalgamated cities with 500,000
to 1,000,000 residents, 152
percent more than amalgamated
cities with 100,000 to 500,000
residents and 38 percent more
than the average amalgamated city (Chart: US Amalgamated City Expenditures
per Capita by Size).(15)
During the 1950s and 1960s, US primary and secondary school districts
consolidated on an unprecedented basis. In each of these decades, the number of
school districts was reduced by half. Costs per pupil rose from 45 percent to 80
percent more in the decades of consolidation than in the decades before and after.
During the 1960s and 1970s, US transit agencies went through unprecedented
consolidations. Transit costs per kilometer increased 42 percent per decade during
the consolidation period, compared to eight percent and 14 percent increases in the
decades before and after (inflation adjusted).(17) Again, the reduced cost escalation
following the "run-up" of costs during the consolidation period were on top of a
much inflated base.
The research demonstrates that municipal amalgamations do not save money, either in
Canada or elsewhere.(18)
There is no academic evidence to suggest that consolidation produces savings.(19)
Contrary to the claimed and unsubstantiated efficiency advantages of amalgamation, a
study of 48 US metropolitan areas found that "the level of expenditures will fall as the
number of jurisdictions increases."(20)
Increased cost escalation pressures: Where governments merge, unit costs in the
amalgamated government migrate to the highest unit cost pre-existing government cost
structure. This is illustrated by the largest component of municipal costs --- employee
wages and benefits. Successive collective agreements can be expected to increase the
compensation of municipal employees to the level of the highest paid workforce of the
pre-existing cities.(21) Downward convergence in labor rates is unprecedented. There
would also be a convergence of collective agreement work rules toward the least
productive provisions, which further increases the unit costs of government services,
though less obviously.
In the longer run, unit costs are likely to rise at a greater rate in amalgamated
governments. Again, employee compensation can drive this dynamic. With fewer or even
no comparable governments for bench-marking, government employee labor rates are
established through an arbitrary and political non-market process. In addition, labor
disputes are more disruptive to regional governments than local governments. The larger
number of residents impacted by strikes in larger jurisdictions increases the political
pressure on elected officials to settle, skewing the balance of power in the favor of trade
Moreover, municipal amalgamation's purported cost efficiency advantages relate almost
entirely to administrative costs. Administrative costs represent a relatively small
percentage of municipal budgets --- 10 to 15 percent. "Delivering maximum value with
available resources" requires even greater attention to the direct costs of service delivery
that constitute 85 to 90 percent of municipal operating costs. The comparative return is
illustrated by a KPMG study that estimated the maximum savings from amalgamation in
the Hamilton-Wentworth region at two percent. The savings from "alternative service
delivery approaches," which would not require amalgamation, were estimated at from 15
to 30 percent.(22)
But the research demonstrates that even the comparatively insignificant administrative
cost savings do not survive the transition from projection to reality. A frequently
advanced example contends that an amalgamated Metro government would employ a
single fire chief instead of the present six city fire chiefs. It is likely, however, that a new
"megacity" fire chief would be installed over the present city chiefs, who would become
area chiefs (if not immediately, in a relatively short period of time).
Despite the overwhelming evidence to the contrary, Metro has suggested that
amalgamation would save up to $208 million annually.(23) But a substantial portion of
Metro's savings would be achieved through such measures as achieving industry staffing
ratios, application of best practices, workforce flexibility, and competitive tendering.
Such productivity improvements are not the result of amalgamation --- they are rather
better management practices that could be achieved by any government, large or small
(But they are more likely to be achieved by smaller, rather than larger governments.)
The performance of the more remote Metro government reinforces the case for not
amalgamating the six cities into a larger megacity. The Metro Council faces a significant
budget shortfall in 1997. It has characterized its choices as simply two --- "cut service to
avoid a tax increase" and "combining cuts and tax increases."(24) The substantial
economies that can be achieved from "alternative service delivery strategies" are not
mentioned, despite their citation by Metro itself in its advocacy of amalgamation. There
is virtually no reasonable prospect that amalgamation of the six cities into a megacity
would be more cost-effective than the current system, and it surely would not deliver
maximum value with available resources.
GREATER TORONTO SERVICE BOARD
Both models propose a board to coordinate regional services in the Greater Toronto Area
(GTSB). GTSB, if wisely designed, would not only improve fairness and coordination,
but would also be more responsive, accountable and efficient. The public would be best
served if the GTSB design includes these features.
Accountability: A board that is accountable to the constituent municipalities.
Board members should be municipal elected officials, appointed by their
municipalities (and serving at the pleasure of their municipalities). A board
constituted in this manner would substantially improve the likelihood that GTSB
and municipal policies are coordinated. The alternative of a separately elected
board would create an uncoordinated public policy framework that would
encourage disagreements between elected officials at the two levels of
government. A separately elected board would also be more likely to seek
expansion of GTSB powers at the expense of the cities, re-establishing service
Limited powers and coordination: Provisions should be adopted to limit GTSB
authority to specific regional powers. This will reduce the potential for service
duplication to evolve. GTSB powers should be expanded only as agreed upon by
the constituent municipalities.
Limitation on administrative staff size: Provisions should be adopted to
establish and maintain the administrative staff size at GTSB to the minimum level
required to provide the required quality and quantity of service.
Provisions to guarantee efficiency: GTSB should be required to contract for all
of its core services and have no operating personnel. Contracts might be with
constituent cities, other publicly owned organizations (such as the Toronto Transit
Commission) private non-profit organizations or with private firms through
No direct taxing or assessment authority: GTSB administration and services
should be financed by user fees and by assessments ratified by (not imposed upon)
the constituent cities. This would ensure greater accountability and coordination of
regional and local policies.
TOWARD DEMOCRATIC AND EFFICIENT GOVERNMENT
Both the Megacity model and the Mayors' model effectively deal with the coordination
principle through the proposed regional coordinating body (GTSB). And while both the
Megacity model and Mayors' model would strengthen cities and abolish regional
governments outside Metro,
the Megacity model would
abolish cities and effectively
maintain regional government
("megacity") within, violating
the principle of democracy,
responsiveness. The Megacity
model would eventually
create a two tier GTSB-local
government structure outside
Metro (after an interim three
tier period) and a two tier
inside Metro. With its reliance
on a megacity (rather than a
more manageably sized local
government) in Metro, the Megacity model would dilute democracy, creating government
that is more remote and thereby less accountable and responsive to its electorate.
Moreover, the Megacity model is inconsistent with the efficiency principle, in that
amalgamated governments tend to be more costly than smaller governments.
The Megacity model approach with respect to governance inside Metro is not only
inconsistent with its approach outside Metro, but is at odds with the general world-wide
trend toward more democratic institutions, devolution, decentralization, market driven
government and customer oriented government.
In the longer run, implementation of megacity's diluted democracy within Metro is likely
to extend to communities outside Metro. A future provincial government, professing less
of a commitment to "smaller government," and less dependent electorally on
constituencies outside Metro could be expected to impose a similar model there, making
local government a thing of the past throughout the Greater Toronto Area.
Finally, any revision of local governance should be subject to a binding referendum in
each of the affected cities. There is no more appropriate issue for referral to the people
than the question of how they shall be governed in their local communities.
By pre-emptively implementing the Megacity model, an historic opportunity would be
lost. There is general consensus that the government's proposed Greater Toronto Services
Board is a much improved approach to coordination of regional public services.
Establishment of GTSB would provide the opportunity to design "right-sized" local
governments in both the "416" and "905" areas. There is no doubt that cities averaging
well below one million in population would be more democratic and more efficient than
1. From its peak population (1950).
2. City of Toronto Act, 1996.
3. The Governance Solution for the Greater Toronto Area, Mayor Hazel McCallion, City of Mississauga (November 5, 1996).
4. Change for the Better: A Framework for Restructuring Local Government, Mayor Frank Faubert, City of Scarborough; Mayor Barbara Hall, City of Toronto; Mayor Doug Holyday, City of Etobicoke; Mayor Mel Lastman, City of North York; Mayor Frances Nunziata, City of York; and Mayor Michael Prue, Borough of East York (November 1996).
5. WDW Panel Recommendations on Local Governance (December 6, 1996). Accessed from the Ontario Ministry of Municipal Affairs and Housing web page at http"//www.mmah.gov.on.ca/inthenews/backgrnd/120696ce.htm on December 9, 1996.
6. In 1950 Chicago's population was 3.62 million, and dropped to 2.72 million in 1995 (U.S. Census Bureau data).
7. The amalgamated cities that are to be created outside Metro are likely to average 400,000 population, or less.
8. Advisory Commission on Intergovernmental Relations, The Organization of Public Economies (Washington, DC: 1987).
9. Calculated from 1990-1 United States Census Bureau data. Education and government enterprise (utility) expenditures excluded (in most cities these functions are performed by other public agencies or privately owned utilities). Data does not include amalgamated city-county governments.
10. Calculated from 1991 United States Census Bureau data, education and government enterprises excluded (Does not include amalgamated city-county governments).
11. Calculated from United States Census Bureau data (1991).
12. Calculated from United States Federal Transit Administration data (1994).
13. Advisory Commission on Intergovernmental Relations, The Organization of Public Economies (Washington, DC: 1987).
14. December 2, 1996.
15. Calculated from 1990-1 United States Census Bureau data. Education and government enterprise (utility) expenditures excluded (in most cities these functions are performed by other public agencies or privately owned utilities).
16. Analysis of U.S. Department of Education data.
17. Analysis of American Public Transit Association data.
18. See Andrew Sancton, "Reducing Costs by Consolidating Municipalities: New Brunswick, Nova Scotia and Ontario," Canadian Public Administration (Fall 1996) for a summary of this research.
20. David A. Sjoquist, "The Effect of the Number of Local Governments on Central City Expenditures, " National Tax Journal (May 1982), cited in Sam Staley, "Bigger is not Better: The Virtues of Decentralized Local Government, Policy Analysis, Cato Institute (Washington, DC: January 21, 1992).
21. Only a relatively small increase in labor rates, less than three percent, would cancel out even the most optimistic projections for administrative savings, because employee compensation is a far larger item of expense than administration.
22. Cited in Sancton.
23. "Estimated Savings and Benefits from Integration in Brief," attachment to letter from Alan Tonks, chairman of the Municipality of Metropolitan Toronto to the Honorable Al Leach, Minister of Municipal Affairs and Housing (November 28, 1996).
24. The Municipality of Metropolitan Toronto web page (http://www.metrotor.on.ca.tax/budget_priorities.html) accessed December 10, 1996.
25. Competitive tendering is also called "public-private competition," in that all qualified organizations, both public and private, are allowed to compete on the same basis for contracts. Public agencies around the world have demonstrated a remarkable ability to reduce unit costs when they are subjected to competition, while maintaining or improving service levels. Competitive tendering of various public services has been legislatively mandated in the United Kingdom, Australia, Denmark, Sweden, Finland, New Zealand, Colorado, and elsewhere. In Australia, a nationally adopted program requires conversion of many public services (especially municipal and regional services) to competitive tendering and will be a requirement of federal funding assistance to state governments.
International public policy consulting firm with specializations in economics, public transit,
transportation, devolution, labor policy, public expenditure policy, and strategic planning.
In business since 1985
Completed projects in North America, Australia, New Zealand, Europe, and Africa.
Examples of projects:
Conducted research on governmental efficiency, performance, governance, and
Directed state legislative and policy program for the American Legislative Exchange
Council (Washington), 1992-1995; included oversight of programs, publications, and
Conducted performance audit of British Columbia Transit (Vancouver).
Drafted Michigan legislation to restructure public transit organization and governance in
Drafted amendments to principles adopted at the Federalism Summit (National Governors
Association, Council of State Governments, National Conference of State Legislators, &
American Legislative Exchange Council, Cincinnati, 1995).
Presented seminars on politics, economics & government services to legislators and public
administrators from 35 states, Canada, New Zealand, Australia, Slovenia, Belarus, the
Czech Republic, Russia, and South Africa.
Principal appointed to 3 terms on the Los Angeles County Transportation Commission by Mayor
Tom Bradley (1977-1985)
While on the Commission, chaired 2 American Public Transit Association Committees
Chaired Transportation Research Board (National Academy of Sciences) Energy
Contingency Planning Conference
Co-author of numerous publications, such as:
People, Markets and Government (State Legislators Guide to Economics)
Emerging Transit Organizational Structures: Options for Improving Customer Service