|
Property taxes are a large source of revenue for provinces and municipalities. In 2003–04, the last year in which the five-province standard was used, they amounted to $38 billion, or about $1,200 per capita. This was the second largest revenue source subject to Equalization, after personal income tax revenues ($47.5 billion). In that year, Equalization paid out $2.3 billion in respect of property tax, second only to the $3.8 billion paid out for personal income taxes.
The measurement of fiscal capacity arising from property tax revenues has consistently been one of the most controversial issues in the Equalization program. These controversies call into question the very nature of the tax. Is the property tax a levy on property values (wealth), as appearances suggest, or is it really a tax on underlying incomes (where property values are used to allocate the total income tax bill among property holders)? Is the property tax even a tax at all, or is it more like a user fee, a charge for benefits received?
Beyond the disputes about the fundamental nature of the tax, there are conceptual and measurement controversies arising from the unique features of property tax. This makes it extremely difficult to construct a robust Equalization tax base to reflect the taxing practices of provinces and their 3,500 municipalities.1
This annex provides a brief history of the property tax base, explains the development and justification of a new residential property tax base (based on actual taxing practices), and explains the Panel’s decision to
recommend that this new base be adopted.
Provincial property taxes were first included in Equalization starting in 1967. At first, only the provincial portion of property taxes was included and it was equalized on the basis of a weighted average of personal and business income taxes. Property taxes were equalized on the basis of what we would call today proxy bases, or economic indicators thought to be closely related to property tax revenue-raising capacity. Locally-levied property taxes were brought in gradually, starting in 1973–74.
In 1977, the property tax base was replaced by another proxy base, based on a multi-concept approach, which was in place until the 2004 Renewal. The multiconcept approach distinguished between three types of property: residential (about 57 percent of the total), commercial-industrial (42 percent), and the farm sector (one percent). It then used a mix of indicators to allocate fiscal capacity, including capital stock measures, provincial Gross Domestic Product (GDP), disposable income, agricultural land values, urbanization, and demographic changes.
It was a pragmatic approach, mixing together a number of concepts: income as a proxy for fiscal capacity, urbanization and demographic change as proxies for scarcity and market values, etc. Most notable in this base was the absence of the very measure which provincial and local governments actually tax: the market value of property. This reflected the fact that tax practices across provinces and municipalities were inconsistent and comparable property value data were not available.
Under the circumstances, the base represented perhaps the best that could be done. It was also periodically adjusted to reflect updated data and special circumstances. However, it was often criticized because its
conceptual underpinnings were very weak and because it was a major departure from the RTS philosophy.
The criticism of the multiconcept approach became stronger in the late 1990s as the adoption of market value assessment for residential property made it possible to use market value data for Equalization purposes.
The federal government, in the lead-up to the 2004 Renewal, invited all governments to work together to devise a better base for the residential property tax. About 50 meetings with ministers and senior officials were held between June 1999 and February 2004, 12 of which dealt in detail with the property tax base.
While a number of approaches were considered, the work focused on developing a property tax base that reflected actual taxing practices.
The structure of property tax rates (mill rates applied to the value of residential property) exhibits a consistent pattern. The rate of tax is systematically lower where property values are high. Every municipality sets a uniform mill rate for the property values in its area. However, those with higher average prices typically set a lower mill rate than those with lower property prices.
When tax rates of municipalities are aggregated for a whole province, there is an inverse relationship between individual property values and the tax rate they attract. In Ontario, for instance, a property worth less than $50,000 but located in municipalities with low average market values will have a mill rate in the three percent range, compared to rates under one percent for a property valued at over $200,000 in municipalities with high average property values.
Given that taxing practices had become consistent across provinces and municipalities, and property value data had become available, the federal government suggested the development of a stratified market value approach that would model actual taxing practices for residential properties.
Specifically, the federal government developed an equalization model that grouped municipalities according to brackets of average market values, and applied a representative average tax rate to each bracket. For example, if Canadian municipalities with average property values of $100,000 levied an average tax rate of two percent, the fiscal capacity for municipalities with such average property values would be the total market value of properties in these communities multiplied by two percent. If the average tax rate were 1.7 percent for municipalities with an average property value of $200,000, then the total market value of properties in these communities would be multiplied by 1.7 percent. This calculation would be repeated for communities with higher and lower property values, each with its corresponding tax rate. The sum would be the province’s fiscal capacity for the property tax base.
This approach is in keeping with the RTS approach. No value judgement is made as to whether the taxing practices are appropriate or not. They are observed and modelled in the same way as personal income tax revenues are modelled to reflect the average progressive rate structures that provinces have chosen. Similarly, sales tax revenue-raising capacity is modelled in Equalization to reflect the revenues that could be raised if average tax rates applied to goods and services that are actually taxed by provinces.
In June of 2003, provincial finance departments were invited to a conference where six academics were asked to judge the new proposal for stratifying market values. A second academic conference took place in November 2003 with ten leading academics, along with the provincial and federal governments.
Experts made varied and useful comments on measuring the fiscal capacity of the property tax base. One particularly helpful suggestion was that the national price elasticity of property tax rates with respect to nominal property values would be a simpler way of capturing the fact that municipalities tend to charge a lower rate of tax as nominal property values go up.
Using this suggestion, the federal government estimated the national elasticity of property tax revenues with respect to nominal property prices to be 70 percent. In less technical terms, it means that a municipality with twice the average market value would collect only 70 percent more tax revenue.
This elasticity approach was tested and found to yield results that were very close to those of the stratified market value approach. This simplified approach is less complex and avoids the need to define brackets (which runs the risk of stratification bias).
|