Neither housing nor hotel: The emergence of “medium-term rentals” in post-Covid Canadian cities

TL;DR

This paper analyzes platform-mediated “medium-term rentals” (MTRs) in post-Covid Canadian cities, arguing they occupy a regulatory grey zone between municipally regulated short-term rentals and provincially regulated long-term tenancies. It combines quantitative platform data from Montreal, Toronto, and Vancouver with a qualitative case study of Ontario Landlord and Tenant Board rulings, concluding that municipalities should regulate “type of stay” rather than “length of stay.”

Key findings

• MTRs grew substantially on Airbnb/Vrbo across all three cities. In Montreal, MTRs’ share of active listings ranged from 22.1% to 24.4% in 2023, about four times the 2017 range of 4.6% to 6.6%. In December 2019, MTR shares of active listings were 11.6% in Montreal, 6.6% in Toronto, and 20.4% in Vancouver; one year later they rose to 24.0%, 23.5%, and 35.6%, respectively.
• Dedicated housing units operating as MTRs increased sharply as a share of total dedicated STR/MTR activity. In April 2018, MTRs accounted for 4.8% (Montreal), 2.1% (Toronto), and 5.4% (Vancouver); by April 2023, these had climbed to 20.4%, 46.2%, and 16.7%—a 3x increase in Vancouver, 4x in Montreal, and 10x in Toronto.
• MTRs also appeared on long-term rental platforms. By end-2022, furnished Craigslist/Kijiji ads reached 38.7% in Montreal, 31.6% in Toronto, and 39.2% in Vancouver. MTR prices generally sat between STR and LTR prices; in Montreal, nightly MTR prices on Airbnb/Vrbo correlated with monthly MTR prices on Kijiji at ρ = 0.84.
• The Ontario LTB case study found no single test for whether the Residential Tenancies Act applies to MTRs. In *Porter v. Ning (Wong)* (2023), a family booked a Toronto Airbnb for a 10-month stay but was evicted after the LTB found the rental exempt under section 5(a), despite the tenants’ efforts to establish residency.

Methods

• Quantitative analysis combined two datasets: Airdna web-scraped Airbnb/Vrbo STR data for Montreal, Toronto, and Vancouver from January 2015 to May 2023 (269,374 unique properties; 122,949,354 data points), and weekly Craigslist/Kijiji long-term rental scrapes from April 2020 to December 2022 (967,171 unique properties; 3,296,469 data points).
• The paper used descriptive and spatial analysis in R, including MTR shares of listings/activity, housing units converted to dedicated STRs/MTRs, furnished and short-term LTR ad shares, and price comparisons.
• Qualitative case study examined 16 Ontario Landlord and Tenant Board cases from 1993 to 2023 involving non-standard tenancies, selected for relevance to medium-term rentals and potential influence on future decisions.

Limitations

• The paper states that reliable estimates of the MTR market are nearly impossible because it is not tracked by Statistics Canada, CMHC, or other government agencies; Airdna’s machine-learning estimates of reserved versus blocked nights are described as the only significant potential source of error, mitigated by using large aggregates.
• The LTR platform data are only rental advertisements, with no availability calendars or confirmation of actual tenancies, so they are used only as supply-side indicators.
• The LTB case study is not a formal limitations section, but the text makes evident that findings are based on 16 selected Ontario cases and may not generalize to other provinces or all MTR arrangements; the quantitative analysis covers only Montreal, Toronto, and Vancouver.