Help the Government of Canada organize its website!

Complete an anonymous 5-minute questionnaire. Start now.

Skip booklet index and go to page content

Data Sources and Methods for the Residential Water Use Indicator

Caveats and limitations

The MWWS assumes metered households are on volume-based fee structures. This is not necessarily the case for all metered households. In some communities, metered households may be on flat-rate fee structures similar to unmetered households.

Any difference in water use between metered and unmetered households cannot be explained solely by the difference in water pricing structures. Although household water use is affected by pricing, it is also affected by location, climate and socio-economic variables.

The substantial decrease in per capita household water use from 2006 to 2009 cannot be explained solely by the increase in household metering rates over the same period.Footnote [1] In 2009, for Canada as a whole, temperatures were below the 1971-2000 normal and below 2006 temperatures from May through July, and were only very slightly above normal in August. Rainfall was higher than both the 2006 level and the 1971-2000 normal in July and August 2009. The lower temperatures in June and July and higher rainfall in July and August could have contributed to the lower per capita water use in 2009 compared to 2006, particularly in the residential sector where lawn watering constitutes an important end-use of water in the summer months.

The representativeness of the sample with respect to community size has changed through time. From 1983 to 1999, the MUD survey collected data only from municipalities with a population greater than 1000 residents. In 2001 the MWWS was expanded to sample municipalities with populations less than 1000 people. In 2009 the survey was further expanded to include a larger sample of municipalities with less than 1000 residents. Although improvements have been made, municipalities with small populations remain under-represented in the MWWS, while municipalities with large populations are over-represented. Data remain comparable among surveys, as the effect of small municipalities on the aggregate statistics is small due to population weighting.

Both the MUD survey and the MWWS exclude municipalities on federal lands and First Nations municipalities.