Data and definitions

Box 1 Data source

This analysis is based on two data sources: the Municipal Water and Wastewater Survey (MWWS), 2004 (Environment Canada, 2007b) and the Households and the Environment Survey, 2006 (Statistics Canada, 2006).

Municipal Water and Wastewater Survey (MWWS) 2004, produced by Environment Canada, was based on the previous Municipal Water Use (MUD) survey. Compared to its predecessor survey, MWWS 2004 introduced several changes in the questionnaire and, for the first time, included returns from 660 rural municipalities and small incorporated towns with less than 1,000 residents (Environment Canada, 2007a). As indicated by Environment Canada (2007a), the total usable survey base for 2004 includes 1,418 municipalities, representing about 28.9 million Canadians (or about 90% of the estimated Canadian population for that year). However, almost half of these municipalities were included in the database only after imputation and adjustment of previous MUD data. The response rates of MWWS 2004 varied substantially by question (Environment Canada, 2007a).

The database used for the computations in this paper retained 1,009 municipalities, for which data on the variables of interest were available. Appendix Table A.1 shows the count and population distribution of these municipalities by type of region. Some additional observations with invalid responses on specific variables were further excluded from the multivariate analysis. The MWWS 2004 variables were weighted by the population served by the municipal system to obtain summary statistics.

The Households and the Environment Survey (HES) 2006, conducted by Statistics Canada, was based on the sample frame of the Labour Force Survey and had a sample size of 28,334 households. Households were categorized into Census Metropolitan Areas (CMAs), Census Agglomerations (CAs), and non-CMA/CA areas (Box 2). Rural households were defined as any household that is not within a CMA or CA. All HES 2006 variables used in this analysis were weighted using survey weights, both for summary statistics and for regressions analysis. Given the complex sampling design of the HES 2006, standard errors and levels of significance for the estimates were calculated using a bootstrap method of re-sampling (Box 3).

For further details on these data sources see Environment Canada (2007a and 2007b) and Statistics Canada (2006).

Box 2 Definitions: geography and regional types

This analysis uses the Rural and Small Town (RST) definition for rural areas, which distinguishes between Census Metropolitan Areas, Census Agglomerations and Rural and Small Town areas; this is complemented by a classification of rural areas into Metropolitan Influenced Zones (MIZ) (du Plessis et al., 2002). Incorporated towns and municipalities (Census Subdivisions or CSDs) are the building blocks for the delineation of CMAs, CAs and the MIZ zones. Hence, we applied this regional typology to the CSDs in the MWWS 2004 database.

It should be noted that MWWS 2004 extended the survey to rural municipalities with population of less than 1,000 individuals (the previous MUD included only municipalities with a population of over 1,000). However, in the database used for this analysis, these CSDs represent only about 5% the sample. In consequence, the statistics for rural should be interpreted primarily as Rural and Small Town municipalities with 1,000 or more residents and which are not part of a CMA or a CA. Appendix Table A.1 provides further details on the distribution of municipalities in the database by type of region.

A Census Metropolitan Area (CMA) has an urban core population of 100,000 and over. In this paper, CMAs have been divided into three groups based on population size: (1) Larger CMA, with population greater than 1.5 million (these are Toronto, Montreal and Vancouver); (2) Medium CMA, with population of 0.5 to 1.5 million people; and (3) Smaller CMA, with a population of less than 0.5 million people. A Census Agglomeration (CA) has an urban core population of 10,000 to less than 100,000. Both CMAs and CAs include all neighbouring municipalities where 50% or more of the workforce commutes to the urban core.

Rural and Small Town (RST) areas are towns or municipalities outside the commuting zone of CMAs and CAs. RST areas are disaggregated into four Metropolitan Influenced Zones (MIZ) based on the size of commuting flows of the workforce to any CMA or CA. The Strong MIZcategory comprises areas with a commuting flow of 30% or more. The Moderate MIZcategory comprises areas with a commuting flow between 5% and less than 30%. The Weak MIZcategory comprises areas with a commuting flow of more than 0% and less than 5%. The No MIZcategory comprises those areas where no individuals commute to a CMA/CA; this category is collapsed with the Weak MIZ group due to the small number of observations in the database.

To maintain consistency in the reporting of our results, we also present the results of the Households and the Environment Survey in terms of CMAs, CAs and MIZ zones.

We recognize that there are census urban residents (in settlements of 1,000 or more) and census rural residents (in the countryside and in settlements smaller than 1,000) within each CMA, CA and MIZ (du Plessis et al., 2002). Consequently, our use of the CSD as the geographic unit of analysis will sometimes include residents of a town with a municipal water provider plus nearby countryside residents who would obtain their water from a private source.

Definitions: water consumption and measurement

The key definitions concerning water use are indicated below. For more details see also Environment Canada (1999 and 2007b) and Statistics Canada (2006).

Water destination. Three destinations or "sectors" are shown in the MWWS database: domestic (i.e. all household users); commercial/industrial (i.e. all manufacturing and businesses uses); and other which includes system losses and unaccounted uses (which, according to Environment Canada (1999), is believed to be under-reported).The total water use is the sum of these three destinations.

Measure of water used. Average daily total water use per capita (also referred to as total water use per day, per capita) is calculated by dividing the total daily water flow in each area (for example, CMA or Strong MIZ) by the population served. Average daily domestic water use per capita (also referred to as domestic water use per day, per capita) is the water flowing through municipal water systems that is consumed by domestic or household users. These estimates are generated by dividing the domestic water flow of each municipal system by the population served.

Source of domestic water consumption. Municipal systems are water distribution centers that are run by municipalities and distribute water to residents who are connected to the system. Private wells are individual systems (usually paid for by the owner of the property) used by the private owner to obtain water.

Water metering. The degree of metering refers to the percent of total water flow that is metered for purposes of monitoring water use (used to determine the water use fee). It is estimated by municipal managers.  Areas classified as "high metering"are municipal systems with more than 90% of serviced households being metered. Areas classified as "some metering" reported metering for 10% to 90% of serviced households. Areas classified as "low metering" reported less than 10% of serviced households being metered.

Water treatment. the HES 2006 considers any activity undertaken by a household to treat their water supply to be water treatment. A filter or purifier on a household water tap is a device that attaches to a water tap for the purpose of treating water. A stand alone filter is any device that is not connected to a water system (municipal system or private well) that has the purpose of treating water. This can range from an ordinary simple charcoal filter to a more complex osmosis filtration system.