METHODOLOGY FOR SECONDARY RENTAL MARKET SURVEY
Canada Mortgage and Housing Corporation (CMHC) conducts a survey of the Secondary Rental Market (SRMS) in September and October to estimate the relative strengths in the secondary rental market which is defined as those dwellings not covered by the regular RMS.
CMHC has identified the following dwelling components to be included in SRMS:
- Rented single-detached houses.
- Rented double (semi-detached) houses (i.e.. Two units of approximate equal size and under one roof that are situated either side-by-side or front-to-back).
- Rented freehold row/town homes.
- Rented duplex apartments (i.e. one-above-other).
- Rented accessory apartments (separate dwelling units that are located within the structure of another dwelling type).
- Rented condominiums (can be any dwelling type but are primarily apartments).
- One or two apartments which are part of a commercial or other type of structure.
The SRMS has three components which are conducted in selected CMAs:
- A Household Rent Survey of all households to collect information about rents.
- A Condominium Apartment Rent Survey of households living in condominium apartments to collect information about rents.
- A Condominium Apartment Vacancy Survey of condominium apartment owners to collect vacancy information.
All three surveys are conducted by telephone interviews. For the condominium apartment vacancy survey, information is obtained from the owner, manager, or building superintendent and can be supplemented by site visits if no telephone contact is made. For the other two surveys, information is collected from an adult living in the household.
All surveys are conducted in September and October, and the results reflect market conditions at that time.
CMHC publishes the number of units rented and vacancy rates for the condominium vacancy survey. For the condominium rent and household rent surveys, the average rent is published. A letter code representing the statistical reliability (i.e., the coefficient of variation (CV)) for each estimate is provided to indicate the data reliability.
Rented condominium apartments are surveyed in the following CMAs: Vancouver, Victoria, Calgary, Edmonton, Regina, Saskatoon, Winnipeg, Toronto, Ottawa, Montréal and Québec. Other secondary rental market units were surveyed in Abbotsford-Mission, Barrie, Calgary, Edmonton, Halifax, Montreal, Ottawa, Quebec, St. John’s, Toronto, Winnipeg, Regina, Saskatoon, Kelowna, Vancouver and Victoria.
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)