METHODOLOGY FOR RENTAL MARKET SURVEY
Canada Mortgage and Housing Corporation (CMHC) conducts the Rental Market Survey (RMS) every year in October 1 to estimate the relative strengths in the rental market. The survey is conducted on a sample basis in all urban areas with populations of 10,000 or more, and targets only privately initiated structures with at least three rental units, which have been on the market for at least three months.
The survey collects market rent levels, availability (outside Quebec), turnover and vacancy unit data for all sampled structures. Data is collected using a combination of telephone interviews and site visits, and information is obtained from the owner, manager, or building superintendent. The survey is conducted during the first two weeks of October, and the results reflect market conditions at that time.
CMHC’s Rental Market Survey provides a snapshot of vacancy, availability (outside Quebec), and turnover rates and average rents in both new and existing structures. There also exists a measure for the change in rent that is calculated based on existing structures only. The estimate is based on structures that were common to the survey sample for both the previous year and the current Rental Market Surveys.
The change in rent in existing structures is an estimate of the change in rent that the landlords charge and removes compositional effects on the rent level movement due to new buildings, conversions, and survey sample rotation.
The rent levels in new and existing structures are also published. While the per cent change in rents in existing structures published in the reports are statistically significant, changes in rents that one might calculate based on rent levels in new and existing structures may or may not be statistically significant.
Difference between Percentage Change of Average Rents (Existing and New Structures) AND Percentage Change of Average Rents from Fixed Sample (Existing Structures Only):
Percentage Change of Average Rents (New and Existing Structures): The increase/decrease obtained from the calculation of percentage change of average rents between two years (example: $500 in the previous year vs. $550 in current survey represents an increase of 10 percent) is impacted by changes in the composition of the rental universe (e.g. the inclusion of newly built luxury rental buildings in the survey, rental units renovated/upgraded or changing tenants could put upward pressure on average rents in comparison to the previous year) as well as by the rent level movement (e.g. increase/decrease in the level of rents that landlords charge their tenants).
Percentage Change of Average Rents from Fixed Sample (Existing Structures Only): This is a measure that estimates the rent level movement. The estimate is based on structures that were common to the survey sample for both the previous year and the current Rental Market Surveys. However, some composition effects still remain e.g. rental units renovated/upgraded or changing tenants because the survey does not collect data to such level of details.
Availability: A rental unit is considered available if the existing tenant has given, or has received, notice to move, and a new tenant has not signed a lease; or the unit is vacant (see definition of vacancy below).
Rent: The rent refers to the actual amount tenants pay for their unit. No adjustments are made for the inclusion or exclusion of amenities and services such as heat, hydro, parking, and hot water. For available and vacant units, the rent is the amount the owner is asking for the unit. It should be noted that the average rents reported in this publication provide a sound indication of the amounts paid by unit size and geographical sector. Utilities such as heating, electricity and hot water may or may not be included in the rent.
Rental Apartment Structure: Any building containing three or more rental units, of which at least one unit is not ground oriented. Owner-occupied units are not included in the rental building unit count.
Rental Row (Townhouse) Structure: Any building containing three or more rental units, all of which are ground oriented with vertical divisions. Owner-occupied units are not included in the rental building unit count. These row units in some centres are commonly referred to as townhouses.
Vacancy: A unit is considered vacant if, at the time of the survey, it is physically unoccupied and available for immediate rental.
Turnover: A unit is counted as being turned over if it was occupied by a new tenant moved in during the past 12 months. A unit can be counted as being turned over more than once in a 12 month period.
Data Reliability Measures
CMHC does not publish a statistic if its reliability is too low or if publication of a statistic would violate confidentiality rules. The ability to publish an estimate is generally determined by its statistical reliability. As indicated earlier, CMHC currently uses the coefficient of variation (CV).
A letter code representing the statistical reliability (i.e., the coefficient of variation (CV)) for each estimate is provided to indicate the data reliability. CV of an estimate is defined as the ratio of the standard error of the estimate to the estimate itself and the CV is generally expressed a percentage. For example, let the average rent for one bedroom apartments in a given CMA be x̄ and its standard error be σx̄. Then the Coefficient of Variation is given by CV = σx̄ / x̄.
Reliability Codes for Proportions
CMHC uses CV, sampling fraction and universe size to determine the ability to publish proportions. The following letter codes are used to indicate the level of reliability of proportions:
- A — Excellent
- B — Very good
- C — Good
- D — Fair (Use with Caution)
- ** — Poor — Suppressed
The following two tables indicate the level of reliability of proportions:
If the proportion is zero (0) and the sampling fraction is less than 100% then the following levels are assigned:
|Sampling Fraction (%) range|
|Structures in Universe||(0,20]*||(20,40]||(40,60]||(60,80]||(80,100)|
|3 – 10||Poor||Poor||Poor||Poor||Poor|
|11 – 20||Poor||Fair||Fair||Fair||Good|
|21 – 40||Poor||Fair||Fair||Good||Very Good|
|41 – 80||Poor||Fair||Good||Good||Very Good|
|81+||Poor||Good||Good||Very Good||Very Good|
Otherwise, the following table is used to determine the reliability level of proportions:
|Coefficient of Variation (CV) %|
|Percentage||0||(0, 5]||(5, 10]||(10, 16.5]||(16.5, 33.3]||(33.3, 50]||50+|
|(0, 0.75)||Excellent||Excellent||Excellent||Excellent||Excellent||V. Good||V. Good|
|(1.5, 3)||Excellent||Excellent||Excellent||V. Good||Good||Poor||Poor|
|(3, 6)||Excellent||Excellent||V. Good||Good||Fair||Poor||Poor|
|(6, 10)||Excellent||Excellent||V. Good||Good||Poor||Poor||Poor|
*(0, 20] means sampling fraction is greater than 0% but less than or equal to 20%; others are similar
Reliability Codes for Averages and Totals
CMHC uses the CV to determine the reliability level of the estimates of average rents and a CV cut-off of 10% for publication of totals and averages. It is felt that this level of reliability best balances the need for high quality data and not publishing unreliable data.
CMHC assigns a level of reliability as follows (CV’s are given in percentages):
- A — If the CV is greater than 0 and less than or equal to 2.5 then the level of reliability is Excellent.
- B — If the CV is greater than 2.5 and less than or equal to 5 then the level of reliability is Very Good.
- C — If the CV is greater than 5 and less than or equal to 7.5 then the level of reliability is Good.
- D — If the CV is greater than 7.5 and less than or equal to 10 then the level of reliability is Fair.
- ** — If the CV is greater than 10 then the level of reliability is Poor. (Do Not Publish)
- 1 April survey was conducted for the last time in 2015.