Annex 2: Some Evidence of Expenditure Needs and the Costs of Providing Public Services in the Territories

How is expenditure need measured?

There are a number of ways to measure expenditure need, ranging from a population proxy approach to a more complex representative expenditure system (RES) approach. Examples of methods used to measure expenditure need and revenue capacity around the world are the Barnett Formula in the United Kingdom, the RES in Australia, and the TFF in Canada’s three territories.

Introduced in 1978, the Barnett Formula automatically applies a share of any change in program spending in England compared with Scotland and Wales to the expenditure base for the two jurisdictions. There are two components to the expenditure base in Scotland and Wales: historical expenditures in place before devolution of responsibilities and incremental expenditure changes. The Barnett Formula determines only the second element of the expenditure base. The formula does not allocate expenditure on the basis of comparative need and does not take into account differences in the cost of providing services.¹ Negotiation and bargaining determine funding levels, primarily on the basis of population share, change in planned spending, and comparability of programs. This is not a comprehensive expenditure need formula.

The RES is the expenditure equivalent of the Representative Tax System (RTS) approach. The RTS measures a jurisdiction’s fiscal capacity by assessing the revenues that would be raised from a typical or standard revenue regime at average tax rates. Similarly, an RES approach measures the amount it would cost per capita to provide typical services of national average quality, assuming a national average rate of efficiency. For example, a standard basket of programs and services may include health, education, public welfare, infrastructure and housing. One would measure workload and costs in each of these expenditure areas to estimate the cost of providing the standard basket of services. If the amount of funding required is more than the national average, then the jurisdiction would be at a disadvantage compared with others and would receive a higher than average grant.

For example, the measurement of expenditure need for childcare, elementary, and secondary education would require first constructing a basket of education services representative of the average across Canada (i.e., teacher-pupil ratios, availability of special education services, etc.). Then the cost of providing these services across different provinces would be measured, taking into account the workload differences (e.g., proportion of school-aged children, proportion of children requiring special education, proportion of children with English or French as a second language, number of languages in which instruction and school books must be offered). Finally, the differences in the costs of providing this standardized basket of services would be taken into account (e.g., different wage and salary conditions, real estate prices, material costs and diseconomies of scale of serving remote or dispersed communities).

The Australian Commonwealth Grants Commission uses a rigorous RES approach to estimate state expenditure needs. Under the Australian system, there are eleven expenditure categories (education, health, law, order and public safety, welfare, concessions and other payments, culture and recreation, aboriginal community services, general public services, services to industry, transport, and economic affairs and other purposes) that are further subdivided. The equalization entitlement is determined by the following formula: standardized expenditure plus average budget deficit or surplus minus standardized revenue minus specific-purpose grants.²

There are a number of significant disadvantages with the RES approach, making it inappropriate for Canada’s territories. Among other things, a great deal of judgement is required to determine the average workload for an expenditure area. Also, it would be more difficult in Canada’s territories to assess economies of scale or to determine the impact of geography on costs.

Expenditure need in the territories is determined by using an historical expenditure measure that is escalated to reflect relative population growth in the territories compared with Canada as well as changes in provincial and local spending in Canada. It is not a comprehensive RES, but rather approximates expenditure needs over time. The historical expenditure measure, the Gross Expenditure Base (GEB), was established in 1985 and was based on 1982–83 expenditures in Yukon and the Northwest Territories. At that time, the GEBs of the two territories accurately represented territorial expenditure needs. Historical spending patterns can be reasonable indicators of expenditure needs and can continue to be relevant, with some adjustments.

This approach avoids using specific standards for services to determine funding levels, as in the case in Australia, and is much less complex. That said, the calculation of GEBs have become more complex over time due to adjustments that were made to reflect devolution of programs to the territories and the creation of a separate GEB for Nunavut. In the absence of an expenditure need study, the Panel believes that the current GEBs are a reasonable proxy for territorial expenditure needs.

In the previous TFF formula, the GEB escalator attempted to reflect more than population growth by including a measure of comparability with the provinces through the provincial-local expenditure portion. This assumed that the mix of public programs and services is comparable in the provinces and the territories, and that the growth rates in public expenditures are also similar. It is generally accepted that even though the current GEBs are based on 1982–83 revenues of the territorial governments, the GEB escalator is a reasonable reflection of changes in territorial expenditure need.

1. Twigger, Robert.(1998). The Barnett Formula. United Kingdom House of Commons Library, Research Paper 98/8, January 12, 1998; Edmonds, Timothy.(2001). The Barnett Formula. United Kingdom House of Commons Library, Research Paper 01-108, November 30, 2001; The Scottish Parliament.(2001). The Barnett Formula, Research Note no. 31, May 4, 2001, Edinburgh Scotland; Bell, David.(2001). The Barnett Formula. Department of Economics, University of Stirling, January 2001.
2. Government of Australia,Commonwealth Grants Commission. (2006). 2006 Update Report.

Table of Contents	Next