The growth of Airbnb

In talking about Airbnb’s impact on housing and the questions of whether and how governments should intervene to protect housing supply and affordability for residents, two facts are often raised: that Airbnb is not the only short-term rental (STR) platform and that STRs existed before Airbnb launched the first version of its website in mid-2008. These facts are typically used to support the argument that there’s nothing really unique or new about Airbnb, so there’s no need for governments to act or to change their current approaches to STR regulation.

It’s true that the practice of listing STRs online has existed since the mid-1990s and the offline (print advertising and word of mouth) segment of the market for much longer than that. VRBO, an early market leader in online STRs, was founded in 1995 (and is now owned by HomeAway).[1] The U.S.-based Vacation Rental Managers Association (VRMA) has existed since 1985.[2] However, traditionally the industry has been concentrated in a limited number of tourist hotspots, resorts, and rural or remote locations.

Obviously, that’s no longer true. In the last five to ten years, the availability of new consumer information technologies (such as smartphones, portable global positioning systems and social media) and the shift in cultural values toward “collaborative consumption” and “sharing” (triggered in part by the recession that followed the recent global financial crisis) have combined to produce a sharp increase in the popularity of STRs. (I have little use for the term “sharing economy,” which I find vague and misleading, but I won’t go into that here. I instead recommend reading critiques such as this one, by Tom Slee.)

The increase in the popularity of STRs is evidenced by Airbnb’s rapid growth, and also the fact that the growth has not come at the expense of HomeAway, arguably its biggest rival. In 2015, the publicly traded HomeAway, which calls itself “the world leader in vacation rentals,” reported a 29 percent increase in its 2014 annual revenue, to a total of $446.8 million.[3]  That followed an annual revenue increase of 24% the previous year ($347 million in 2013).[4] The company attributed its 2014 revenue growth in part to an increase in paid listings (17% higher than 2013).[5] These growth trends were also reflected in the VRMA’s membership, which increased by a record 20 percent in 2014 (to include 473 manager members).[6] Meanwhile, the National Association of Realtors reported that 2013 sales of U.S. vacation homes were up by 47 percent since 2011 (to 717,000), with 89 percent of buyers planning to rent out those new homes within a year.[7]

If this STR growth had somehow stayed within the confines of beach and ski resorts, rural retreats, and a limited number of high-priced urban hotspots, it might not have become a urban housing affordability and equity issue, as I believe it now is (and has been for a while already). But that has not been the case. There’s been massive growth in cities, as well as expansion to new communities not previously viewed as tourist destinations, such as Winnipeg and Edmonton in Canada, and Spokane, Washington and Grand Rapids, Michigan in the United States.[8]

Listings estimates and growth 

There were no Airbnb listings before Airbnb launched its website in mid-2008, obviously, so the starting point was zero. It now claims more than one million listings (as does HomeAway) in 34,000 cities and 190 countries. I realize that many people who listed properties on Airbnb when it first launched were already renting out those properties and advertising them using other methods, so it’s not that every Airbnb listing represents a net increase in STRs. It’s also true that an unknown percentage of people currently list their properties on more than one STR platform. HomeAway and Airbnb target different segments of the STR market (HomeAway apparently being less interested in the urban and more so in the family getaway segment), but there is still an unknown amount of overlap.

Nevertheless and in spite of HomeAway’s tagline, based on Airbnb’s size and reach, as well as its public profile (venture capitalists are apparently fed up with hearing new start-ups described as the “Airbnb of X”), I think it has a defendable claim to being the most popular STR platform, certainly in the urban context. So, for the sake of better understanding the growth of Airbnb listings from zero to where they are now, I thought it would be useful to compile a list of estimates for various places. You’ll find them for about 30 locations, with dates and sources, below.


I emphasize that these are estimates. Airbnb’s own website limits search results to maximum of 1,000, making it difficult-to-impossible for an ordinary person to tell how many Airbnb listings exist in a given location if that number exceeds 1,000.

These estimates come mainly from media accounts, which do not always cite sources. Others are from researchers who have scraped Airbnb’s website, or from Airbnb itself, though it doesn’t often provide such figures. Note that in some cases the figures are for hosts (who may have more than one listing) and in others for the listings themselves. Given the rapid increase in listings and also the seasonal aspect of tourist-driven STR bookings, it’s also important to take note of the date of each estimate. The dates of these estimates range from 2010 to a couple days ago. Where I’ve found more than one estimate (NYC for example), I’ve listed them in date order. There are a few notable differences in figures for around the same date (e.g. Paris) and I can’t really explain those, except to point to differences in what’s being counted, geographic scope (what’s included in the boundaries of Paris, for example), definitions and methodologies. It’s also true that a few places (Amsterdam, Paris and New York among them) have taken much more active approaches to enforcement than others, so quantity fluctuations in those places could result from that. There is no explanation of methodology in most articles and reports, but the work done and published by researchers such as Tom Slee and Murray Cox is a welcome exception, as are earlier stories and studies published by Skift and the San Francisco Chronicle. Also notable is the 2014 report by the Attorney General for the State of New York, which is based on subpoenaed information.

Another caveat is that most of these figures are just for total number of listings. It’s more interesting to know the distribution of those listings by neighbourhood, type of listing, nightly rate, etc. Some of that information can be found for a limited number of cities in the work I’ve noted above.

This also seems like a good place to note some common problems with media accounts of Airbnb’s size and influence, which frequently quote Airbnb’s own so-called “economic impact studies” for various cities. Those studies in turn frequently commit the journalistic and scientific sin of reporting percentages without including the whole number on which they’re based. This article on a 156% increase in Airbnb guests in Cardiff, Wales is an example of that type of story. With only few possible exceptions, all I’ve seen Airbnb put out when it releases one of its “studies” is a news release or blog post containing a list of figures consisting mostly of percentages (e.g. 80% of New York hosts are or say X) and an undated infographic, such as this one. They don’t release definitions of terms (such as “visitors,” “economic impact,” or “local spending”), methodology or raw data that would allow others to check their math or evaluate for themselves. I’ve seen very little comment on this opacity in media reports, so I’m gathering that reporters are either not asking for or have not been successful in obtaining this info.

What’s the point?

All these qualifications and caveats aside, I think the larger and main point supported by the growth that these figures indicate is that there actually is something very new and different about what’s happening with Airbnb in cities around the world in the last five to seven years, including Vancouver. Given these numbers, especially where the STR listing growth has occurred in cities with shortages of affordable housing that pre-date Airbnb’s existence (San Francisco, New York and Vancouver, for example), I think it’s hard to make the case for a status quo approach to regulation and enforcement. I’ve provided examples of how some cities are changing their approach in these posts.

Unfortunately, the table is too long to show in its entirety here, but this is what the top of it looks like.

2015.05.30 Airbnb listings and hosts table

You can see the rest and download the table (as a spreadsheet) here.

*Apologies for the mix of footnotes and links in this post. Some of what’s above is adapted from a paper, thus the footnotes.

[1], “VRBO Company Snapshot,” Bloomberg Business, n.d.,

[2]VRMA, “About VRMA,” VRMA, n.d.,

[3]HomeAway, “HomeAway, Inc. Reports Fourth Quarter and Full Year 2014 Financial Results,” February 24, 2015,

[4]HomeAway, “Fact Facts,” HomeAway, n.d.,

[5]HomeAway, “HomeAway, Inc. Reports Fourth Quarter and Full Year 2014 Financial Results.”

[6]VRMA, “Vacation Rental Managers Association Celebrates Landmark Membership Growth,” VRMA, September 17, 2014,

[7]HomeAway, “Vacation Home Sales Spike Thirty Percent in 2013 According to New Research from HomeAway and National Association of Realtors,” HomeAway, April 2, 2014,

[8]Ryan Tumilty, “Home-Sharing Website Airbnb Now in Edmonton,” Metro, January 11, 2015, Colin Fast, “Winnipeggers Embrace Airbnb as World Cup Rounds the Corner,” Metro, January 5, 2015, Nicholas Deshais, “Spokane City Council Takes Step to Preserve Short-Term Rental Businesses – The Spokesman-Review,”, December 2, 2014, Sidorowicz, “Airbnb Licensing to Pick up Speed in GR, City Manager Promises Enforcement,” Fox17, November 13, 2014,

If a city wants to change its approach to regulating short-term rentals, what might it do?

Elsewhere on this blog, I’ve provided an alphabetical list of jurisdictions that have taken some sort of recent regulatory action on the issue of short-term rentals (STRs). It’s gotten out of date over the last couple months as I’ve focused on my course work and because so many municipalities (and a few counties and states) are taking steps in this direction these days. I do plan to update that list, but in the meantime it struck me that it might be at least as useful to re-organize it in a way that focuses more on the types of actions being taken.

Some common regulatory elements are definitely apparent, some of which may be adaptable to Vancouver – though it’s important to keep in mind that U.S. cities, which comprise the bulk of the list, often have wider powers and bigger budgets than their Canadian counterparts of equivalent sizes. Still, if Vancouver or any other city does want to adjust its approach to regulating STRs, it doesn’t have to re-invent the wheel. Instead, it can look to, and hopefully learn from, the many other jurisdictions that are doing the same (48 as of my last count, on March 23).

Scanning the reports on local government efforts to regulate STRs, it’s clear that few are trying to stop the practice all together. It’s reasonable and predictable, however, that local governments would want to understand and manage the extent of STR activity within their boundaries and the potential effects of that activity on housing supply, demand for city services, the local tourism sector, and residents’ quality of life. So, here are some of the main ways they’re trying to do that. Most of the actions below seem to have been implemented through the creation of a new bylaw or ordinance, and/or amendments to the local zoning regulations. The zoning changes are required where, as seems to be the case in many communities, renting out accommodation for less than 30 days is considered a commercial activity and is prohibited in residential zones (except by licensed bed-and-breakfast operators).

Ten common requirements of STR regulations

  • Registering and/or obtaining a permit
  • Inspections
  • Insurance
  • Presence of host
  • Local contact person
  • Notification of neighbours
  • Limiting the number of STRs neighbourhood or building
  • Limiting the type of housing used for STRs
  • Limiting the number of rental days per year
  • Paying taxes

I’ll discuss each of these briefly below and provide links to some jurisdictions that use, or are considering using, these tools to manage STRs. The fact that a jurisdiction isn’t mentioned or listed in one of the sections below doesn’t necessarily mean that it doesn’t have that requirement; It could mean I don’t have a good link to that information or that I’m relying on a media account and it doesn’t discuss that aspect of the regulation. For now, my goal is to provide examples of where these requirements are in place, not an exhaustive list. I’m also working on putting this information into table format. As usual, please bring any errors you find to my attention via Twitter or my contact form.

I’ll be commenting on these measures, and the enforcement issues I’ve seen arising from some of them, in a separate post shortly.

Registering or obtaining a permit

Setting up some kind of registration or permit system seems to be the lowest common denominator of STR regulatory programs – what everyone does regardless of the other measures they decide to implement. And since establishing, monitoring and enforcing a permit program takes staff time and resources, there is typically a fee for obtaining an STR permit. In some cases, the permit takes the form of a specific type of tax or business certificate, to which conditions are attached. In jurisdictions where permits are required, it’s also common to require the operator to provide the permit number in all advertisements and to display it in a prominent place inside the STR unit.

Permit Costs

Permit costs vary considerably. In some cases, the fee is as low as $25 (per property), as in Palm Desert. At $50 for two years, San Francisco’s fee is also low. City staff there have said it doesn’t begin to cover costs of administering the new rules. In Austin, the annual cost is $235 (April, 2015), plus a one-time $50 fee for first-timers to cover the costs of notifying neighbours. Savannah’s fee is $150 (per property) the first year and $50 to renew. In Portland, the cost varies by type of unit. For units in detached dwellings, the initial permit costs $178. Renewal is $62. For properties in multi-unit buildings, the initial fee is $100 and $62 to renew. The higher first-time fee is meant to cover the cost of the bedroom inspection, which is not required at renewal. In late 2014, the City of Auburn was considering a fee of $57. An STR licence costs $287 in Grand Rapids, Michigan. The Province of Quebec’s “tourist home” permit costs $240. Chicago’s vacation rental licence costs $500 for two years. St. Helena charges $1,075. Applicants may also have the planning department prepare a mailing list and labels (for notification of neighbours within 300 feet) for a fee of $200. Amsterdam’s is the highest charge I’ve seen: €772 as of 2012.

Who else is doing this? Some of the other jurisdictions where a permit requirement is in place or under discussion include Aspen, Massachusetts, Madrid, Nashville, and Sonoma.


Inspections are a fairly common requirement, but jurisdictions vary as to what type, their purpose, and which department carries them out. Maui County, Portland and St. Helena all require some kind of inspections prior to approval. Austin requires either a certificate of occupancy or an inspection carried out by a certified third-party inspector using this STR checklist.


Proof of insurance is a common condition of permit approval, but jurisdictions differ as to what types of insurance are required and how much. Insurance requirements are complicated by the fact that many homeowner policies do not cover short-term rental activity. This fact came up recently in connection with the Calgary incident. See: Austin (proof of property insurance); Chicago (homeowner’s fire, hazard and liability, as well as commercial liability insurance with at least $1 million in coverage per occurrence); Nashville (property insurance covering $1 million per occurrence); Quebec (civil liability covering at least $2 million per claim); Savannah; and San Francisco (at least $500,000 in liability, or proof that any hosting platform used provides at least that amount).

Local contact person

The media has reported various horror stories about vandalism, theft and raucous parties in Airbnb listings, a Calgary incident resulting in $75,000 in property damage being the most recent at time of writing. These examples have highlighted how people sometimes act differently – less responsibly – when they’re on vacation than when they’re at home, where the consequences for anti-social behaviour can be more immediate. This is less of a problem with guests in a hotel that has 24-hour staff and security cameras than it is when visiting partyers let loose in a residential neighbourhood and there’s no front desk to call. It’s also much less likely to be an issue when the host of the STR is actually home than when a whole house or apartment is rented out. But listings for entire houses or apartments (i.e. when the host is not present) make up a huge percentage of Airbnb listings, at least in many cities (Los Angeles, New York and San Francisco, for example). That’s why some local governments that allow the rental of entire homes (not all do) require permit applicants to identify a local contact person who can be available onsite to troubleshoot within a designated period of time. In Maui County, for example, the manager of an STR must be available to answer calls at any time and able to be present at the property within one hour of a request from a guest, neighbour or county agency. The manager must also have an office or residence within 30 driving miles of the property. Chicago requires the owner of a vacation rental or a designated contact person to have a home or office in the city. Aspen, Palm Desert, Savannah and St. Helena have similar provisions. Some jurisdictions go even further. New York State’s multiple dwelling law prohibits rentals of less than 30 days unless the person renting out the property is present. (See also the information Airbnb provides to New York City residents).

Notification of Neighbours

Some local governments require permit applicants to notify neighbours within a certain radius of the property of their plan to offer STRs. In some cases, the notice is a courtesy only. That’s true in Portland, but on the other hand, the notice letter the city has created for applicants to use is quite detailed. It includes city contact information and a summary of the regulations. Neighbours on all sides of the property and directly across the street must be notified, as well as the relevant neighbourhood association. Elsewhere, if enough neighbours object to the application, they can quash it. The city of Auburn, in California, was considering this as as of October 2014, but was also going to allow those whose applications were denied the option of paying a $480 fee to appeal to the city’s planning commission. Maui County allows neighbours within a 500-ft radius to have input into the permit process. In St. Helena, if 30% of property owners within a 300-foot radius object, the application goes to a planning commission hearing. As of April 23, 2014, the council was considering making it easier for neighbours to reach that threshold.

Limiting the number of STRs in neighbourhood or building

Some jurisdictions limit where STRs can operate by neighbourhood or zone. In Savannah, a zoning amendment that took effect on January 1, 2015, allows for short-term rentals in parts of the city that currently allow for inns, as well as three other mixed- use areas. STRs are not allowed in Savannah neighbourhoods that are primarily residential. While STRs (as opposed to licensed bed and breakfasts, which the city also refers to as short-term rentals) are banned in New Orleans, a proposal under consideration would see them allowed in some areas, but not the French Quarter. Sonoma allows vacation rentals in some zones, but not in “higher-density residential districts” or within certain agricultural areas. Sonoma also doesn’t allow STRs in “second dwelling units,” “farm family,” “agricultural employee,” or “farmworker” housing units.

In other cases, the limits are quantitative. As of October 2014, Grand Rapids began issuing STR permits that allow for the rental of one room in an owner-occupied residence, but the city will issue only 200 such permits annually. As of March 2012, St. Helena was prepared to issue only 25 permits on a first-come, first-served basis. Austin has created three categories of STRs. Type 1 is an owner-occupied dwelling, whether detached, duplex, or multi-family. Types 2 and 3 are not owner-occupied, with the former being a single family dwelling or duplex, and the latter being multi-family. For types 2 and 3 (not owner-occupied) the city will issue permits for no more than 3% of the single-family detached units in each census tract and no more than 3% of the units in each building in a non-commercial zone. Nashville also limits the number of STRs by census tract to no more than 3% of single-family and two-family units. Amsterdam limits the number of permits to 10% of all units in the private sector and as a result, stopped issuing permits in early 2014.

Limiting the type of housing used for STRs

Amsterdam residents who live in subsidized (renting for less than €699.48) or public housing, which makes up almost 50% of the city’s housing stock, are not allowed to offer their homes as STRs. This helps protect the supply of affordable housing for residents. The government states, “The presence of short stay residents in the city must not be at the expense of the quality of life in neighbourhoods, nor may it impact the availability of affordable subsidised apartments.” Similarly, Aspen has a housing authority that manages housing (both owned and rented or leased) that is meant to be affordable for the local workforce. It does not allow STRs in that housing. Some regulations, such as Portland’s, limit STR rentals to property that is used as a “primary residence.” Hamburg, Germany, allows residents to “occasionally” rent out their primary residence” without seeking government permission or obtaining a licence. A licence is required, however, to use property that is owned, but not lived in, as an STR. The advocacy group Share Better San Francisco wants to see the city’s existing rules changed to ensure that below-market rate units and in-law suites are not rented to tourists. Grand Rapids and Quebec do not allow individual rooms to be used as STRs.

Limiting the number of annual days or nights of STR use

Another way local governments are trying to manage STRs is by restricting the number of nights that a property can be rented – either at all, or when the host is not present. Amsterdam residents can use their primary residence as an STR for up to 60 days a year. In San Jose, hosts are limited to 180 stays per year when they are not present. In San Francisco, hosts can rent for an unlimited number of nights per year when they are present and a maximum of 90 nights per year when not present. However, in March 2015, the planning department said the  90-night limit was “unenforceable” due to lack of data (that Airbnb won’t provide). An amendment dealing with this issue, among others, has been proposed by the mayor and Supervisor Mark Farrell. It would see a 120-day limit on all STRs, regardless of whether the host is present.

Requiring minimum stays

In order to reduce turnover and the negative impacts on the conventional tourist accommodation sector, some jurisdictions specify a minimum length of stay for STRs. In Amsterdam, it’s seven nights (and no longer than six months). It seems the Amsterdam government sees STRs mainly as a suitable option for business travellers who may be staying in the city longer than a typical tourist, but not as long as a resident. The purpose of its STR policy is “to provide sufficient housing to accommodate (mainly international) employees and to enhance Amsterdam’s appeal as a business location.”  Shorter stays are only allowed in hotels and bed-and-breakfasts. In Madrid, the minimum STR stay is five days. Palm Desert requires at least 3 days and two nights.

Paying taxes

Many local or regional governments (including Vancouver’s) levy some kind of tax on tourism accommodation, often using that revenue to offset the costs of the city services tourists use, and/or to fund tourism marketing efforts. Hotel and bed-and-breakfast operators in many cities have complained that they are put at a competitive disadvantage on pricing by the fact that Airbnb hosts (generally) don’t pay this tax. Meanwhile, local governments are concerned about the foregone revenue. In 2012, Aspen estimated that it was losing about $100,000 per year in foregone sales tax (11.3%) and licensing fees. Sonoma conducted an audit in 2014 and estimated that it was losing between $500,000 and $1.3 million in taxes on illegal short-term rentals.

In an effort to even the playing field as well as capture revenue, many local governments are now requiring STR hosts to agree to collect and remit these taxes as a condition of obtaining a permit. Beyond that, in a handful of cities the process has been streamlined through formal agreements between Airbnb and the local government. In those cases, Airbnb is automating the collection and remission of tourist taxes on behalf of registered hosts. Such arrangements are now in place in Portland (as of July 1, 2014); San Francisco (as of October 1, 2014); Amsterdam and San Jose (as of Feb. 1, 2015); Chicago and Washington, D.C. (as of February 15, 2015); and most recently, Malibu (April 20, 2015). Malibu’s tax collection deal with Airbnb follows the city’s issuing of subpoenas to about 60 short-term rental sites in May 2014. At the time, STRs were already legal, as long as owners registered with the city and paid the 12% transient occupancy tax. But the city had found there were about 400 online ads for STRs in Malibu, while only 50 properties were registered. The city sent the subpoenas to determine the true number of STRs within its boundaries, and to help it follow up on unregistered properties and collect transient occupancy taxes from those hosts.

However, even in cases where Airbnb is cooperating, at least to some extent, on the tax issue, local governments are finding that imposing taxes on STRs and actually collecting what they’re legally entitled to are two different things – and that Airbnb’s cooperation has some very critical limits. In Amsterdam, Airbnb is collecting tourist taxes from visitors, but the city doesn’t have access to information on who’s paid, and so can’t know the rate of compliance. Under Portland’s new rules, Airbnb began collecting the transient lodging tax on bookings as of July 1, 2014 and remittng that revenue to the city monthly. Here again, though, it is difficult for the city to verify because so far only a small portion of city hosts have actually registered, and Airbnb does not want to provide the city’s Bureau of Development Services, which runs the permit program, with hosts’ names and addresses. According to this news account, Airbnb’s public policy director, David Owen, suggested the city’s pursuit of that info was similar to “the National Security Agency mining personal data on Americans from phones and the Internet,” and that “the city shouldn’t get ‘unfettered access’ to its hosts’ personal information without getting a court subpoena.” In January 2015, council responded by empowering its Bureau of Revenue to demand that all STR brokers provide it with the names and addresses of their Portland hosts. “The industry has made it clear they do not intend to play ball with us on the enforcement side,” said City Commissioner Nick Fish. “What we have been told by the industry is ‘butt out.’” As of early May 2015, the bureau’s new powers seemed to be going unexercised.

As mentioned above, I’ll comment on these measures and some enforcement issues that have arisen from them in a separate post, soon.