BNB SUB 160A Rec'd 10/04/2019

Economics and Industry Standing Committee Legislative Assembly of Western Australia Inquiry into Short-Stay Accommodation

Airbnb Supplementary Submission Thank you for the opportunity to make a supplementary submission to the Inquiry into Short-Stay Accommodation following the public hearing on 20th February 2019.

This supplementary submission provides further information which addresses issues the Committee has considered throughout the public hearing process.

Economic Impact of Airbnb in Western Australia Airbnb’s submission outlined the economic impact of Airbnb in Western Australia, referring to a report by Deloitte Access Economics — Economic E!ects of Airbnb in Australia : Western Australia — which found that in 2015-2016 Airbnb guests who stayed in Western Australia spent $155.1 million, which supported 780 jobs, including 180 jobs in areas outside Perth, in brick-and-mortar businesses such as cafes, restaurants, and retailers, and contributed $99.7 million to Gross State Product.1

Airbnb understands that the Committee is interested in more recent data to inform their understanding of the economic contributions of home sharing to Western Australia. Unfortunately, the scope and detail of the report referred to above and in our original submission is not able to be replicated by Airbnb at this stage.

To inform the Committee’s understanding however, this supplementary submission includes some further information on Airbnb’s economic activity utilising Airbnb’s internal data. The Committee should note that

1 Deloitte Access Economics 2017, Economic E!ects of Airbnb in Australia, pp.55-57.

!1 the information included in the report by Deloitte Access Economics and the "gures below utilise two di#erent data sets - both temporally and geographically. The Deloitte Access Economics report utilises data from 2015/2016 and Western Australia, whilst the Airbnb guest survey information below utilises data from 2017 and Australia.

As noted in Airbnb’s original submission, there are 12,500 active Airbnb listings in Western Australia and in the past twelve months to 1st January 2019, our Homes community welcomed over 594,000 guests across the state — both domestic and international travellers.2

Airbnb regularly surveys our community — both hosts and guests — including on areas related to economic activity. The most recent survey of guests (n = 1,088) in Australia was conducted in 2017, which found that a typical guest spent $168 per day with 54% of spending by guests occurring in the neighbourhood in which they stayed. Below in Table 1 is a breakdown of the guest survey data from 2016 and 2017. The 2016 data relates to guests who visited Western Australia (n = 123) whereas the 2017 data relates to guests who visited Australia (n = 1,088).3

2 Airbnb internal data.

3 Airbnb internal data.

!2 Table 1.

Category 2016 (Western 2017 (Australia) Australia)

Food (restaurants, 32% 33% cafes, bars)

Groceries 11% 14%

Shopping 14% 16%

Cultural activities 10% 10% (museums, historical sites, local festivals)

Leisure activities 11% 10% (entertainment)

Transportation 19% 15%

Other services 3% 2%

The broad results are largely similar between both data sets, with the highest spending category being spending on restaurants, cafes, and bars followed by transportation. Whilst it is di$cult to extrapolate and directly compare the economic impact "gures in the Deloitte Access Economics report without a full economic analysis, Airbnb notes that the number of guests staying in Western Australia has increased from 171,500 guests in 2015/20164 to 594,000 in 2018. It follows that as the number of guests has increased, the total economic impact — driven by both what hosts earn, and what guests spend when they visit — has also increased.

Airbnb in context of the housing market Airbnb understands the Committee’s interest in understanding more about the short-term rental market and its relationship with the broader

4 Deloitte Access Economics 2017, Economic E!ects of Airbnb in Australia, pp.18.

!3 property market, and speci"cally its impact on housing availability and a#ordability. Broadly, Airbnb is simply not a primary factor in most housing markets in Australia, including Perth and Western Australia. The Airbnb community represents less than one per cent of both the Perth and Western Australian housing markets. Holding less than one per cent of the available market responsible for housing a#ordability is simply not credible. Further, these arguments distract from much larger macro factors such as enabling planning systems to provide adequate social and a#ordable housing options to match population growth, demographic changes, and taxation — including, notably the proposed changes to negative gearing from 1 January 2020 by the current Federal Opposition — to name a few.

A number of organisations have considered Airbnb’s impact on the housing market in Australia, and at the public hearing Airbnb made note of reports which may be of utility to the Committee’s considerations. The following research reports are attached to this supplementary submission.

The Grattan Institute, Peer-to-peer pressure: Policy for the sharing economy The Grattan Institute considered the impact of short-term rentals on the housing market and found that “any rent increases caused by the rise of short-stay rentals are likely to be localised or small. Short-term use of housing is a small fraction of the city-wide housing stock.”5

Tenants’ Union of NSW, Belonging Anywhere: Airbnb and renting in Sydney The Tenants’ Union of NSW conducted an in-depth analysis on the relationship between Airbnb listing activity (utilising scraped data) and the rental market in Sydney.

5 The Grattan Institute 2016, Peer-to-peer pressure: Policy for the sharing economy, p. 26.

!4 There were two main "ndings from this report: • “We found no link between particularly high numbers of Airbnb listings and rises in rent.”6 • “For all of Sydney there was no discernible e#ect on the vacancy rate.”7

The report concluded that “even if Australia is the most penetrated market there is actually not yet so much activity on Airbnb that would impact the private rental market.”8

SGS Economics and Planning, What Impact Does Airbnb have on the Sydney and Melbourne Housing Markets In 2018 Airbnb commissioned SGS Economics and Planning — a trusted, respected, and independent research and analytics "rm — to examine the impact of Airbnb on the housing markets in Australia's two largest cities — Melbourne and Sydney. Prior to the release of this report, no detailed analysis had been completed in Australia utilising Airbnb’s own data sets to better understand how Airbnb’s community is placed in the context of housing markets, speci"cally housing a#ordability. The report found that: “Broadly, in Sydney and Melbourne, it is not "nancially bene"cial to host a property on Airbnb instead of renting to a long term tenant”. The report concluded that: “The impact of Airbnb on the housing markets in Sydney and Melbourne appears to be minimal.”9

Airbnb notes that the Committee has also heard from the peak body for local government in Western Australia, the Western Australian Local Government Association regarding this issue, with WALGA noting they have no evidence “anecdotal or otherwise” on the relationship between short-stay accommodation and rental a#ordability.

6 Tenants’ Union of NSW 2017, Belonging Anywhere: Airbnb and renting in Sydney, p.14.

7 Tenants’ Union of NSW 2017, Belonging Anywhere: Airbnb and renting in Sydney, p.13.

8 Tenants’ Union of NSW 2017, Belonging Anywhere: Airbnb and renting in Sydney, p.16.

9 SGS Economics and Planning 2018, What Impact Does Airbnb have on the Sydney and Melbourne Housing Markets, p.5.

!5 Finally, for many Airbnb hosts the income they earn from sharing their home is vital to their livelihood and economic wellbeing. As noted in our original submission, 1 in 2 hosts (49 per cent) in Western Australia report that hosting helps to a#ord staying in their home, whilst 1 in 3 (31 per cent) report that host earnings go towards paying the cost of housing (mortgage or rent payments).

Vested Interests Airbnb would like to note the participation of the Australian Hotels Association (WA Branch) (AHA WA) at the recent Contrabnb, or Reformbnb, conference held in New York in 2018. This conference was organised by the Hotel Association of New York City, paid lobbyists, and other international hotel associations, the purpose of which was to discuss “Winning the Fight Against Short Term Lets". Shortly before attending this conference, the AHA WA released its regressive plan to ban holiday homes, ban holidays less than two weeks long, and only just stop short of a ban on almost all home sharing in Western Australia outright.

At a time when tourism in Western Australia is in dire need of a progressive, visionary policy agenda to grow, not inhibit the number of visitors travelling in the state and spending money in local businesses, this short-sighted agenda from a vested interest such as the AHA WA demonstrably fails the test of what’s good for tourism and the visitor economy in Western Australia. Furthermore, when this agenda is combined with rhetoric that the thousands of ordinary Western Australians who share their homes to visitors from around the world are part of “diseasing an industry and destroys the jobs of quali"ed and trained employees”,10 the Committee can well question the intent of vested interests to genuine policy-making in the public interest. It is notable that the AHA WA has cited the wellbeing of its employees as the reason for its policy agenda on short-stay accommodation. This sits incongruently with the AHA WA's consistent advocation for reductions to award conditions and penalty rates. Attached to this supplementary

10 https://www.afr.com/news/politics/airbnb-slams-tourism-australia-director-bradley- woods-for-conflict-of-interest-20181204-h18paa.

!6 submission are media releases and communiques issued by the AHA WA and Tourism Accommodation Australia.11

We look forward to working with the Economics and Industry Standing Committee and the Western Australian Government in relation to these issues, and would be pleased to engage in discussions and provide additional information which would be helpful for the Committee’s deliberations.

Sincerely,

Brent Thomas Head of Public Policy, Australia and New Zealand Airbnb

Attachments • The Grattan Institute 2016, Peer-to-peer pressure: Policy for the sharing economy • Tenants’ Union of NSW 2017, Belonging Anywhere: Airbnb and renting in Sydney • SGS Economics and Planning 2018, What Impact Does Airbnb have on the Sydney and Melbourne Housing Markets • AHA and TAA Media Release 5 June 2017, Australia’s peak hospitality and tourism accommodation bodies welcome ‘certainty’ in FWC transitional arrangements decision on penalty rates reform • AHA WA Media Release 23 February 2017, Penalty Rates Changes a Win for Public • AHA WA Essentials Bulletin 23 February 2017, Fair Work Commission decision on penalty rates

11 Tourism Accommodation Australia is a wholly-owned subsidiary of the AHA. Mr Bradley Woods was recently appointed interim CEO following the departure of TAA CEO, Ms Carol Guiseppi.

!7 April 2016

Peer-to-peer pressure Policy for the sharing economy

Jim Minifie Peer-to-peer pressure

Grattan Institute Support Grattan Institute Report No. 2016-7, April 2016 This report was written by Jim Minifie, Grattan Institute Productivity Growth Founding Members Program Support Program Director. Trent Wiltshire provided extensive research assistance and Higher Education Program made substantial contributions to the report. Joseph Moloney, Anisha Kidd and Grace Anthony also provided valuable contributions.

We would like to thank the members of Grattan Institute’s Productivity Program Reference Group for their helpful comments, as well as numerous industry participants and officials for their input.

The opinions in this report are those of the authors and do not necessarily Affiliate Partners represent the views of Grattan Institute’s founding members, affiliates, individual Google board members reference group members or reviewers. Any remaining errors or Origin Foundation omissions are the responsibility of the authors. Medibank Private Grattan Institute is an independent think-tank focused on Australian public policy. Senior Affiliates Our work is independent, practical and rigorous. We aim to improve policy EY outcomes by engaging with both decision-makers and the community. PwC The Scanlon Foundation For further information on the Institute’s programs, or to join our mailing list, Wesfarmers please go to: http://www.grattan.edu.au/

Affiliates This report may be cited as: Ashurst Minifie, Jim & Wiltshire, Trent, 2016, Peer-to-peer pressure: policy for the sharing economy, Grattan Institute Corrs ISBN: 978-1-925015-83-6 Deloitte GE ANZ All material published or otherwise created by Grattan Institute is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 3.0 Unported License Urbis Westpac

Grattan Institute 2016 Peer-to-peer pressure

Overview

Peer-to-peer platforms use online technology to help strangers increase disability service funding. Only recent purchasers of interact and do business. If you have booked a holiday rental, a licences who are in hardship should be compensated. car ride, or even a tradesperson to replace a broken window in recent years, it’s increasingly likely you have done so on a In peer-to-peer accommodation, states need to do more to get platform. Two of the world’s best-known platforms, Airbnb and the balance right between short-term use of property and the Uber, enable millions of users to find cheaper and more amenity of neighbours. They should give owners’ corporations convenient accommodation and travel. Others host markets in more power to limit disruptions caused by short-stay letting and everything from art to freelance work to finance. streamline dispute resolution. Councils should prohibit short-stay rentals only as a last resort. The prize for getting the peer-to-peer economy right is likely to be large. Ride-sharing businesses such as Uber can cut more than Peer-to-peer platforms will mostly improve an already flexible $500 million from Australian taxi bills. Other platforms are already labour market. Governments should not create a new labour boosting employment and incomes for those on the fringe of the category for peer-to-peer contract workers. But they need to labour market, and putting thousands of underused homes and strengthen rules that prohibit employers misclassifying workers as other assets to work. contractors. Some platforms should be obliged to provide work safety insurance, much as labour-hire firms are today. Some say peer-to-peer platforms bring hidden costs by risking work standards, consumer safety and local amenity, and eroding Today’s laws are mostly adequate to address competition and the tax base. These worries are not groundless, but they should consumer challenges posed by platforms. Regulators need to not be used as excuses to retain policies, such as taxi regulation, monitor platforms’ market power, and ensure they inform users of that were designed for another era and no longer fit. Governments their rights and responsibilities and deal fairly with users. should not try to hold back the tide to protect vested interests. Peer-to-peer platforms can boost the economy, but tax as a Yet governments do have a role to play in ensuring the peer-to- proportion of output may fall. Tax rules must be tightened to peer economy can flourish. In transport, other states should ensure international platforms pay enough tax. follow the lead of New South Wales, the Australian Capital Territory and Western Australia and legalise ride-sharing. They If governments act fast, Australia can make the most of the peer- should mandate safety checks and insurance for ride-sharing. to-peer economy. Not all traditional industries will be happy – but They should cut annual taxi licence fees and may need to consumers, workers, and even the taxpayer can come out ahead.

Grattan Institute 2016 1 Peer-to-peer pressure

Main recommendations

All state governments should legalise ride-sharing, following the Competition and consumer laws are mostly fit to deal with the lead set by ACT, NSW and others. (Chapter 2) peer-to-peer economy. (Chapter 5)

Governments should allow ride-sharing to operate but set The Australian Competition and Consumer Commission should minimum safety requirements. Governments should remove adapt existing competition law principles to the peer-to-peer restrictions on taxi licences (or cut their prices) and deregulate economy. Regulators will need to ensure platforms do not abuse pre-booked fares, but retain maximum fares for rank-and-hail the power they acquire as their user bases grow. The Australian trips. Disability service funding models will need to be adjusted. Consumer Law will apply to all peer-to-peer suppliers and Any compensation should be limited to people who bought taxi platforms. licences recently. Taxation laws should be tightened to ensure that the tax take Laws concerning short-stay accommodation need to do more to does not fall as the peer-to-peer economy grows. (Chapter 6) help people limit noise and loss of amenity. (Chapter 3) The peer-to-peer economy may pay a low tax rate under today’s State governments should give owners’ corporations more powers rules. The Commonwealth should limit tax minimisation by to control short-stay rentals, possibly even the power to ban multinational firms and should oblige platforms to share with the continuous, whole-premise short-stay rentals if agreed to by Australian Taxation Office information about the taxable activity of members. Local governments should focus on controlling their users. disruptions and protecting amenity, not primarily on limiting short- stay rentals.

Governments should adapt labour regulations to help people participate at low risk on platforms. (Chapter 4)

The Commonwealth should tighten ‘sham contracting’ provisions in the Fair Work Act, and require platforms to supply peer-to-peer workers with more information about the risks and responsibilities of being a contractor. States should ensure peer-to-peer workers in riskier occupations have workers’ compensation coverage.

Grattan Institute 2016 2 Peer-to-peer pressure

Table of contents

Overview ...... 1

Main recommendations ...... 2

1 How to build a better market ...... 4

2 Urban transport: point-to-point, peer-to-peer ...... 8

3 A room of one’s own: peer-to-peer accommodation ...... 22

4 Putting platforms to work: the job market...... 33

5 Competition and consumer regulation ...... 46

6 Tax and the peer-to-peer economy ...... 53

7 Conclusion ...... 58

8 References ...... 59

List of legislation ...... 71

List of cases ...... 71

Grattan Institute 2016 3 Peer-to-peer pressure

1 How to build a better market

Peer-to-peer platforms help strangers to connect and do 1.1 From market fairs to smartphones business.1 They can improve on existing markets and foster new ones by solving the three problems any market faces. They: Market fairs, stock exchanges and intermediaries have long brought buyers and sellers together. They still perform vital bring together a ‘critical mass’ of sellers and buyers, functions, but the online revolution is transforming how people connecting thousands or millions of participants; interact. The peer-to-peer service trade platforms that are the subject of this report follow two earlier generations of online make it easy to find a match and establish a price; and platforms. The first generation created marketplaces for products. Its leaders launched in the 1990s and are now major make it safe to do business, by verifying identities, pre- global businesses: screening suppliers, and providing rating and payment systems and even insurance. eBay launched as an online auctioneer in 1995. It was the first large firm to combine the search, review and transaction tools By hosting big markets that are easy to navigate and safe to use, that all peer-to-peer platforms now use. peer-to-peer platforms help people obtain more and cheaper services, and find work that suits them. Amazon launched as a book retailer in 1994. Nearly half of the products purchased on Amazon today are sold by third Platforms also help people access under-used assets, leading party sellers. It is valued at about US$200 billion. some to herald the rise of ‘the sharing economy’. In reality, transactions on platforms are usually on commercial terms. But Alibaba, a Chinese company, launched in 1999 as a business whether they are used for sharing or commerce, peer-to-peer marketplace for international trade. It is valued at about platforms can increase productivity and incomes. US$150 billion.

The second generation of online businesses are the social media platforms that emerged in the mid-2000s. They enable billions to communicate and share content. Their business models are not 1 The term ‘platform’ denotes the bundle of technologies that let people find funded by peer-to-peer exchange but mainly by advertising. partners and complete transactions. The term ‘peer-to-peer’ describes any platform that lets consumers and smaller businesses transact with each other. Today’s giants include: See Roth (2008) on critical mass, congestion and safety in market design.

Grattan Institute 2016 4 Peer-to-peer pressure

Facebook, the leading social media platform, launched in Although none have yet had the transformative impact of Uber 2004. It has more than 1.5 billion active users, and earns and Airbnb, platforms in other sectors are growing. They include: advertising revenue of US$4 billion a quarter. Its market value has risen to over US$300 billion. Finance: The revenues of the largest peer-to-peer platforms, Lending Club and Prosper, are reported to be doubling each Twitter launched in 2006. It has more than 300 million users year. Early in 2015, the top six global peer-to-peer finance and revenue of US$700 million a quarter. Its market value is firms were valued collectively at more than US$15 billion 4 about US$10 billion. though valuations have dropped sharply since then.

The third generation of online platforms are the peer-to-peer Work platforms are smaller and slower-growing still. Listed service trade platforms that have emerged in the late 2000s, Australian-based Freelancer claims almost 19 million mostly thanks to the smart phone.2 These are the focus of this registered users and revenues of $40 million. Freelancer’s report. An astonishing range of platforms has emerged. Table 1.1 US-based competitor Upwork maintains that its 10 million summarises the size and growth of some major platforms. The registered workers earn US$1 billion through the site.5 most successful are the transport and accommodation platforms: More than a thousand other platforms have sprung up to serve Transport: Uber Technologies, a ride-sharing service, niches from clothing and tool rental to bandwidth, food deliveries 6 launched in 2009 and now handles over 5 million trips each and pet services. To take art and design: Etsy (a handmade art day in about 70 countries. It has privately raised capital at a and craft marketplace) is valued at about US$500 million; valuation of over US$60 billion.3 While its revenue is Australian firms Redbubble (a design marketplace that also estimated to be growing at over 250 per cent each year, handles logistics), 99 Designs (a graphic design marketplace), Chinese ride-sharing firm Didi Kuaidi claims even more users and Envato (which operates marketplaces for website templates and more rapid growth. and other code products) are all growing fast.

Accommodation: Airbnb, launched in 2008, lists more than two million premises around the world. It is privately valued at about US$25 billion, higher than all but one hotel chain, and 4 Bruene (2015) its revenue is reported to double each year. 5 Australian firm hipages (which specialises in home improvement jobs) hosted over $1 billion of work in the 12 months to March 2015 and was recently valued at over $160 million (Tucker (2016)). hipages’s providers are skilled 2 Botsman and Rogers (2011) tradespeople but it could be considered a ‘peer-to-peer’ platform. 3 Shontell (2015) 6 AngelList (2016)

Grattan Institute 2016 5 Peer-to-peer pressure

Table 1.1: Estimated global size, growth and value of selected leading peer-to-peer service platforms, 2015 Buyers Sellers Revenue Revenue growth Value Industry Platform (million) (million) (US$ million) (year-on-year, per cent) (US$ billion) Transport Uber - 1 1,800** 260 62

Didi Kuaidi - 10 450* 1000+ 20

Accommodation 60 2 900 90*** 25

- 1 500** 10 3

Finance 1 - 430 100 3

2 (both sides) 200 150 2

Work 19 (both sides) 30 50 0.5

Notes: Many companies are not publicly listed; reporting is not as comprehensive as for listed firms. Non-US dollar figures converted to US dollars using 29/03/16 exchange rate (of 1 AUD = 0.76 USD). Buyers and sellers are latest available, global, and rounded to nearest million; some may not be active. For accommodation platforms, sellers represent the number of property listings, (the number of users will be lower as some users list multiple properties). Revenue is full-year 2015, rounded to nearest $5m (except * full-year revenue extrapolated from January-May 2015 revenue; ** full-year revenue extrapolated from Q1-Q3 revenue 2015). Revenue growth is full-year 2015 compared to 2014 (except *** two year annualised growth from 2013 revenue estimate), rounded to nearest 10 per cent. Finance revenue is net interest income. Valuations are latest available, from private transactions, or market capitalisation on 29/03/16; rounded to nearest billion, except Freelancer. Sources: media articles; company websites; annual reports

Grattan Institute 2016 6 Peer-to-peer pressure

1.2 Why platforms matter to policymakers some of the challenges for competition and consumer regulation posed by the rise of peer-to-peer platforms. Policymakers need to pay attention to peer-to-peer platforms for at least four reasons. First, some regulations obstruct people who Finally, platforms pose tax challenges. By growing the economy want to use platforms, and that limits productivity and income and handling payments electronically, they can grow the tax base. growth. Chapter 2 focuses on transport, and proposes that state But multinationals may not pay much local company tax and governments should rewrite taxi regulations to help realise the smaller peer-to-peer providers may not pay much GST. Chapter 6 benefits of peer-to-peer transport. recommends changes to the tax system.

Second, some peer-to-peer platforms can lead to tensions as the 1.3 Scope of this report market grows. Chapter 3 focuses on accommodation, and examines how owners’ corporation and local councils should The peer-to-peer world is developing fast. This report does not respond to short-stay accommodation platforms. Platforms can cover all of it. Important areas not covered include the opportunity also circumvent regulations that governments put in place to in peer-to-peer finance and the emerging frontier of ‘peers without achieve social goals, such as in employment. Chapter 4 proposes platforms’, using fully decentralised computing approaches such 9 adjustments to labour market regulation in light of peer-to-peer as blockchains. work platforms. Instead, the report focuses on the main challenges posed by Third, peer-to-peer platforms are only the most visible part of a peer-to-peer platforms to policymakers in the worlds of transport, broader shift towards platforms across the economy.7 Consumer, accommodation and work. It then considers broader competition, civil society and business functions – from navigation to consumer, and tax challenges posed by the peer-to-peer communication and finance, as well as increasingly sophisticated economy. computing – are increasingly managed on platforms. Because the value to a platform user may increase as the number of other users increases, a leading platform may acquire a strong competitive position. But by the same token, platforms may compete strongly with one another, to the benefit of consumers.8 Data access and privacy also become complex challenges when more work is done through online platforms. Chapter 5 examines

7 Evans and Gawer (2016) 9 A blockchain is a software technology that permits transactions to be verified 8 Weyl (2010) by many distributed peers.

Grattan Institute 2016 7 Peer-to-peer pressure

2 Urban transport: point-to-point, peer-to-peer

Peer-to-peer ‘ride-sharing’ platforms such as Uber and Lyft are Box 1: Australia’s point-to-point transport market revolutionising city transport around the world, challenging taxi industry incumbents and regulators alike.10 Australians spent about $5.5 billion on 230 million taxi trips in 2014, at an average fare of $24. There are about 21,000 taxis Ride-sharing platforms connect commercial drivers and people licensed across the country. seeking transport. Passengers use a phone app to estimate their fare, request a ride, see the driver’s rating summary, monitor the In the taxi rank segment, taxis pick up passengers at a driver’s approach, confirm the route while they are in progress, designated point. The hail segment involves pick-ups off the pay, receive a receipt, and rate the driver’s performance. Drivers street. If a customer orders a taxi for ‘first available’ pick-up, this is use an app to guide them on where jobs are likely, to see their called a ready-to-ride order; a taxi ordered for a later time is a passenger’s rating, get directions, rate the passenger, and pre-booked ride. About 70 per cent of taxi trips in metropolitan monitor fares and earnings. areas are rank or hail trips. Network companies take bookings and dispatch taxis. There are about one-quarter as many hire- Australian consumers have enthusiastically embraced ride- cars as there are taxis. Some states impose vehicle standards or sharing, even in states where drivers are breaking the law by minimum fares on hire car licences; hire cars may not do rank- offering the service (Box 1). Many customers say they find ride- and-hail work. sharing cheaper, more reliable and convenient than taxi travel.11 UberX served about 6 per cent of the Australian point-to-point Ride-sharing began in Australia when Uber launched its UberX transport market by August 2015, at a rate of about 15 million service in Sydney in May 2014. The ACT and NSW governments rides per year, and is growing fast.12 Drivers can also benefit from legalised ride-sharing in late 2015, and the South Australian and ride-sharing, as Chapter 4 discusses. Uber has more than 20,000 Western Australian governments’ reforms are scheduled to start monthly active Australian drivers.13 in July 2016. Other states are still considering regulatory options.

Notes: Sources: Australian Taxi Industry Association (2014); NSW Government (2015b); Taxi Services Commission (2016); Roads and Maritime Services (NSW) (2015)

10 The term ‘ride-sharing’ is in widespread use, though the rides are not shared in any meaningful sense. 11 Castle (2015); Law and Ma (2015) 12 Deloitte Access Economics (2016) 13 Uber (2015e)

Grattan Institute 2016 8 Peer-to-peer pressure

2.1 Ride-sharing is likely to lower fares, cut waiting times dollars in rental payments to licence owners.17 The restrictions and improve service also make it difficult to manage variations in demand, which can be significant in the point-to-point transport sector. State Many studies have shown that taxis in Australia are more governments issue peak-period taxi licences, but these do not expensive than they need to be because numbers are limited, accommodate unpredictable peaks (due to weather, for example), 14 licence fees are excessive and competition is restricted. and the peak-period taxis are costly to provide because they are Waiting times can also be excessive, particularly at peak times. off the road much of the time.

State governments have long restricted the number of perpetual Legalising ride-sharing is likely to cut average point-to-point and long-term taxi licenses in major cities. Most states also sell transport fares and waiting times. Operators will not pay inflated annual licences at a fixed price, or auction a number of them each licence rents; under-used private vehicles can be put to work; and 15 year. Restrictions on taxi licences make high taxi fares viable, part-time and occasional drivers can drive at times that suit them and increase the fees taxi operators pay to governments or and when there is most demand. Ride-sharing fares today are licence owners (see Box 3). usually lower than average taxi fares. Taxi fares in Canberra in late 2015 – after ride-sharing was legalised – were over 30 per Governments also prevent hire cars from competing with taxis. cent above standard UberX fares (Figure 2.1). A study for Uber Some states restrict hire car numbers by selling perpetual hire car found average taxi fares in Australia were about 25 per cent licences for tens of thousands of dollars; others cap the number of higher than average UberX fares in August 2015.18 annual hire car licences they issue. Some prescribe a minimum 16 fare or minimum vehicle cost. Such policies keep the hire car At peak times, though, ride-sharing fares can exceed peak taxi fleet small and push up car costs and fares. fares, sometimes by a large margin (Figure 2.1). Uber increases fares at times and in locations where waiting times would Together, these restrictions push up taxi fares by almost ten per otherwise be unacceptable at standard prices. Uber claims that cent, on average. The average taxi fare of $24 includes about two

14 For example, Taxi Industry Inquiry (2012) and Abelson (2010). 17 Licence leasing costs are conservatively $450 million per year, assuming a 6 15 New South Wales auctions a quota of annual taxi licences. Victoria sells per cent yield. About $2, or 8 per cent of the fare, goes to licence holders. annual taxi licences for about $23,000. Australian Taxi Industry Association (2014). Labour supply of drivers is likely 16 WA charges $170/year for hire car licences and sets a minimum fare of $60, more elastic than taxi travel demand, so licence rental costs mostly push up but new rules are set to be introduced in July 2016 (WA Government (2015)). fares rather than cutting driver incomes. Victoria sells perpetual hire car licences for $40,000. 18 Deloitte Access Economics (2016), p.30

Grattan Institute 2016 9 Peer-to-peer pressure

Figure 2.1: UberX and taxi fares in Canberra be lower for ride-sharing at peak times in more mature US Civic to Parliament House, 6 minutes, 3.3 kilometres, dollars markets.20 In NSW, average waiting times for an UberX are lower 16 21 Card surcharge than for a taxi. 14 Time Distance Some argue that ride-sharing fares will rise. A platform may seek Flag fall to raise fares if it achieves a high market share and drivers must 12 pay licensing costs and may face higher insurance premiums in cities where ride-sharing is legalised (see Box 2). But legalising 10 ride-sharing will also attract new drivers and platforms.22 That may 8 push fares down and cut waiting times, attracting more passengers. It is not yet clear how these competing forces will 6 play out.

4 2.1.1 Legalising ride-sharing is likely to improve reliability, service and customer satisfaction 2 Many taxi drivers and booking companies provide good or 0 excellent service today, but regulators and the industry have long Taxi UberX Taxi - weekend UberX - 1.5x struggled to meet consumer expectations of quality, timeliness, and night surge reliability and availability:

Note: Time estimates includes one minute of waiting time ($52/hour at speeds under 10 km/h for taxis); card surcharge is 10 per cent plus GST (and is optional). Weekend/night fares apply before 6 am or after 9 pm Monday to Friday and all day on a Saturday, Sunday or public holiday. All inclusive of GST Sources: Grattan analysis of posted fare data on Uber website and ACT Road Transport Authority website, accessed 23 March 2016. surge fares mobilise drivers, and that fewer than 10 per cent of 20 Uber trips are at surge fares.19 Average waiting times do tend to Uber (2014); Hall, et al. (2015) 21 Deloitte Access Economics (2016), p.5) calculated that the average waiting time for an uberX ride was 4.46 minutes, compared to 7.79 minutes for taxis. 19 Hall, et al. (2015) find that surge pricing increases the supply of cars. Chen, et 22 In early 2016 the operators of Australian taxi app GoCatch released a ride- al. (2015) find evidence that surge pricing limits demand but does not increase sharing platform, GoCar; international platforms such as Lyft and Didi Kuaidi the short-run supply of drivers as the ‘surge’ is too brief for drivers to respond. may enter the Australian market.

Grattan Institute 2016 10 Peer-to-peer pressure

In Victoria, the 2012 Taxi Industry Inquiry found ‘widespread Taxi service can be improved in much the same way if rides are customer dissatisfaction with the quality, reliability, cost and booked by smartphone apps such as GoCatch and Ingogo. But availability of taxi services’.23 relying on them alone to improve point-to-point transport service will short-change consumers. Many people want features that In NSW, satisfaction with taxi services is generally below taxis do not offer, including better vehicles and a more diverse satisfaction with public transport.24 Customers are most driver group. Taxis will always struggle to match the flexibility of dissatisfied with the price of taxis and with payment system ride-sharing in responding to peaks in demand. And competitive surcharges. pressure from ride-sharing will force taxi operators to improve service.28 In Western Australia, 84 per cent of respondents to a 2014 survey had at least one negative association with a taxi in the 2.1.2 Smartphone systems improve safety, but regulations previous twelve months. Only 34 per cent of survey are still needed respondents had a positive association.25 All Australian states regulate the taxi and hire car industry to Ride-sharing apps and platforms offer improved service. The apps make it safer. Taxi and hire car drivers must pass police and themselves permit a passenger to estimate fares and car arrival medical checks. Vehicles must pass roadworthy inspections and times, view the approach of a driver, monitor actual versus taxis must be fitted with duress alarms, cameras, and GPS advised routes, streamline payments, and review each trip’s locators. Many states require taxi network companies to maintain route, time, driver, and fare.26 But more importantly, the apps give trip records. drivers and passengers strong incentives to behave well: they know they will be rated after each trip, and that prospective ride Yet even with these requirements, taxi passengers and drivers partners will see their ratings before the next one. It is not have safety concerns. A Western Australian survey found that only 41 per cent of women feel safe catching a taxi alone at surprising that many consumers prefer ride-sharing to taxi travel, 29 and report that they find it more comfortable and more reliable.27 night. On the other side, taxi driving is one of the most

23 Taxi Industry Inquiry (2012), p.40. There has been an increase in satisfaction in taxi services in the past two years (Carey (2015)). 24 NSW Government (2015a), p.3 25 Cousins and Fels (2014), p.21 28 Wallsten (2015) finds that Uber’s rise in New York has been associated with a 26 Uber (2015d) decline in consumer complaints about taxis, on a per trip basis. 27 Castle (2015); Law and Ma (2015); IPART (2015) 29 Bruce (2013)

Grattan Institute 2016 11 Peer-to-peer pressure

dangerous occupations in Australia – perhaps 15 times as Ride-sharing may offer another safety benefit: some people who dangerous as the average job.30 would otherwise drink and drive may take a ride instead. People may already be doing so in some US markets: according to one Ride-sharing apps and platforms can help make transport safer: US study, ride-sharing in California may have cut drink-driving passengers and drivers are not anonymous; feedback may help deaths by about 5 per cent.34 Drink-driving deaths in Virginia fell weed out riskier drivers and passengers; the app tracks location; 20 per cent as Uber and Lyft grew rapidly in 2015, even while and cash transactions are not allowed.31 Platforms also screen overall traffic deaths rose.35 ride-sharing drivers and cars before they are allowed to transport passengers, though in some cases their screens may be inferior 2.1.3 Ride-sharing may improve disability access if the to those used for taxis.32 Platforms typically do not require drivers subsidy model is adapted to install hardware such as in-car cameras or duress alarms. Wheelchair Accessible Taxis (WATs) make up 5 to 10 per cent of There is little data on how the safety of taxis and ride-sharing Australia’s taxi fleet.36 State governments discount WAT licences, compare. But if ride-sharing drivers pass the same background allow WATs to charge an extra fee or higher fares when checks, app-booked ride-sharing seems unlikely to be less safe transporting wheelchairs, and subsidise fares for some people than pre-booked taxi work. More Uber drivers than taxi drivers are who have disabilities.37 WATs are typically vans, so operators women, perhaps because women see ride-sharing as safer than may earn the bulk of their income serving non-wheelchair regular taxi work.33 customers. Regular taxis also participate in disability access schemes.

Many users with disabilities are dissatisfied with the point-to-point 30 Frontier Economics (2014), p.4. In Victoria, there were 51 assaults in taxis in services on offer, yet have few alternatives. They say WATs do 2013/14 (Victoria Police (2014), p.31). Rank-and-hail work is particularly unsafe, not arrive reliably and promptly, and that some drivers are not as drivers can do little to screen out risky and anonymous passengers. 31 Uber does permit drivers to use cash in some non-Australian markets where card usage is low. 34 Greenwood and Wattal (2015) 32 Uber requires drivers to supply ID, undergo a police check, have a clean 35 Smith (2016) driving record, have held an unrestricted driver’s licence for at least one year and 36 Cousins and Fels (2014), p.75; Australian Human Rights Commission (2013). drive a vehicle less than nine years old (Uber (2016)). The driver-screening 37 For example, Victorian travelers with a disability receive subsidised fares of up processes used by Uber and Lyft in the US has been criticised at times. For to $60 per trip, and WAT drivers receive a ‘lifting fee’ of $16.70 (paid by the example, Uber in the US has been criticised for not using fingerprints and so not government) for each wheelchair passenger picked up (Taxi Services picking up some people with serious criminal records (Dougherty (2015)). Commission (2015)). In Queensland, a disabled passenger receives a subsidy of 33 Hall and Krueger (2015) finds that in the US women comprise 14 per cent of half the total fare, up to a maximum of $25 (TransLink (2015)). WATs often have Uber drivers, versus 8 per cent of taxi and limousine drivers. conditions attached, such as prioritising wheelchair jobs.

Grattan Institute 2016 12 Peer-to-peer pressure

well-trained.38 Some users instead make their own booking are removed, the subsidy model for WATs will need to be arrangements with individual drivers rather than central booking changed, as it will no longer be possible to offer WAT operators services.39 licence fee discounts.

Regulators may be able to include ride-sharing in schemes that 2.1.4 Ride-sharing may be a stepping stone to new offer cheaper and more reliable point-to-point transport for people transport models with disabilities.40 Some disability advocacy groups say they believe ride-sharing can help to improve services.41 Others, by Another reason to legalise ride-sharing platforms is that they may contrast, believe waiting times for WATs have increased since the pave the way for other new point-to-point transport models. Three arrival of Uber due to drivers preferring to drive for Uber rather are being developed. First, ride-sharing companies have than drive a WAT, and that WAT numbers may fall as ride-sharing developed real-time car-pooling functions. The apps enable expands.42 travellers to share a car for part or all of a trip. Uber and Lyft introduced their respective real-time car-pooling options, But ride-sharing may not cut wheelchair-accessible fares much. UberPOOL and Lyft Line, in a few cities in the US in 2014.43 Wheelchair-accessible vehicles are more specialized, so there may be fewer underused vehicles in circulation for ride-sharing to A second model is hybrid public-private transport, with routes tap into. And because WAT licences are already discounted, that adjust to passenger pick-ups and destinations. Hybrid public- licence rentals do not contribute as much to fares as they do to private transport services have begun operating in some US 44 45 standard taxi fares. Moreover, if entry restrictions to taxi operation cities. They could complement public transport. Third, ride-sharing platforms may use or develop driverless 46 38 In Western Australia in March 2015, more than 40 per cent of wheelchair- cars. Some consumers are enthusiastic about the technology, accessible taxi service were not delivered within the prescribed wait times (Department of Transport (Western Australia) (2015)); Taxi Industry Inquiry (2012); Department of Economic Development Jobs Transport and Resources (Victoria) (2015), p.14. 39 Australian Human Rights Commission (2013) 43 UberPOOL operates in 29 cities and 100 million UberPOOL trips have 40 Uber offers UberASSIST, which has specially trained drivers to help with reportedly been taken (Manjoo (2016)). passengers with special needs and cars are able to fit foldable wheelchairs and 44 Loup and Chariot operate in San Francisco. Uber has trialled Smart Routes in other mobility devices. Uber recently started a trial wheelchair accessible vehicle San Francisco and a minibus service, UberHop in Toronto. service in Brisbane (Silva (2015)). The scale of these efforts is not publicly 45 Uber research suggests that 64 per cent of UberX trips started or finished in available. an area underserved by public transport (Uber (2015e)). 41 Pro Bono Australia (2015); Uber (2015e) 46 Uber is said to be developing one (Gilbert (2015)), as are Google and various 42 Willingham (2016) car makers.

Grattan Institute 2016 13 Peer-to-peer pressure

but others have concerns about safety.47 It is unclear how soon to grow strongly in the years ahead, this consumer benefit may fully autonomous cars will be ready for widespread use. grow.

2.2 Peer-to-peer transport platforms will generate Drivers can also benefit from work on ride-sharing platforms. consumer and broader economic benefits Large increases in average hourly incomes are unlikely because platforms will probably not need to pay much above today’s going Peer-to-peer transport platforms increase productivity in the point- rates to find drivers. However, some drivers value the flexibility to-point transport market by improving matching efficiency and ride-sharing driving offers (see Chapter 4 for more information on helping drivers spend more of their working time with paying work on peer-to-peer platforms). passengers. In the US, UberX drivers typically spend a higher fraction of their time, and drive a higher share of miles, with a Benefits to consumers and drivers are partly offset by losses passenger in their car than do taxi drivers.48 They can also cut incurred by taxi licence owners. About $2 of the average taxi ride waiting times and improve service for passengers. Some of these covers payments to licence owners. Assuming that about one in productivity benefits accrue to passengers, some to the platform three ride-sharing trips would not otherwise have been a taxi ride, and some to drivers.49 then the average loss in taxi licence rental income is about $1.30 per ride-sharing trip, or about a quarter of the consumer benefit In a recent Deloitte study commissioned by Uber, consumer estimated for ride-sharing. benefits of Australian ride-sharing in 2015 were recently estimated at about $5.50 per trip on average, or $80 million per year in 2.3 What policymakers should do total.50 About 40 per cent of the benefits were attributed to lower prices for people who would otherwise have taken a taxi, and Following the lead set by the ACT, NSW, Western Australia and about 60 per cent to service improvements on those trips, and to South Australia (Table 2.1), other state governments should the value of additional rides induced by price cuts and service legalise ride-sharing, reduce annual taxi licence fees and remove improvements.51 If Uber and other ride-sharing services continue unnecessary taxi regulations. Consumers will benefit from improved service and lower cost point-to-point travel. Recently 47 Schoettle and Sivak (2014) announced reforms in the ACT and NSW largely meet the 48 Cramer and Krueger (2016) objectives set out below (see Table 2.1 for details on the ACT and 49 Benefit shares are determined by the market power of the platform and the NSW reforms).52 Western Australia has announced that it will price elasticities of consumers and drivers. 50 Deloitte Access Economics (2016). Results based on an estimated 14.5 million UberX trips per year, based on 1.2 million UberX trips in August 2015. 51 This estimate ranged from $54 million to $127 million, depending on the price 52 The Competition Policy Review identified taxis and ride-sharing as a priority elasticity of demand estimate used (ibid.). area for review (Recommendation 8) (Harper, et al. (2015)).

Grattan Institute 2016 14 Peer-to-peer pressure

regulate ride-sharing, but has not released details of its 2.3.2 Retain most taxi safety regulations and set safety approach.53 standards for ride-sharing services

2.3.1 Deregulate fares for pre-booked taxi and ride-sharing To regulate safety in point-to-point transport, state governments services should:

State governments should partially deregulate point-to-point fares. Require all taxi, hire-car and ride-sharing drivers to pass the They should: same criminal history and driving history checks and to have zero blood-alcohol concentration Allow providers to set ride-sharing, hire-car and pre-booked taxi fares. Fares are visible before a booking is made, and Require all ride-sharing vehicles to undergo an initial operators compete, so there is little case for regulation. roadworthy inspection and appropriate follow-up inspections.

54 Continue to set maximum rank-and-hail taxi fares. Continue to mandate that taxis have in-car cameras, duress alarms, and other technologies to promote passenger and Regulators could consider mandating that ride-sharing providers driver safety. Taxi work remains risky because rank-and-hail share much the same information as taxi companies about their work is anonymous and drivers must accept and carry cash. pricing (including surges), and other performance measures, such as waiting times, in order to allow periodic industry reviews. Taxi State governments should also set minimum insurance and ride-sharing companies would remain subject to standard requirements for ride-sharing drivers (see Box 2). Governments competition and consumer protections (see Chapter 5). should:

Retain existing compulsory third party (CTP) injury and third- party property insurance, or public liability insurance, requirements for taxis and hire cars.

53 WA Government (2015) 54 Require ride-sharing drivers to hold CTP injury and third-party The main reason for regulating rank-and-hail fares is that it can be inefficient property insurance. for customers and drivers to negotiate or shop around at ranks or on the street (Productivity Commission (1999a), p.36). The New Zealand experience, however, shows that deregulation of rank-and-hail taxi fares can generate good Mandate that platforms verify that their drivers have outcomes if customers can easily distinguish taxis operated by different appropriate CTP injury and third party property coverage. companies (ibid., p.14).

Grattan Institute 2016 15 Peer-to-peer pressure

Box 2: Insurance and ride-sharing

Compulsory third party insurance indemnifies vehicle owners who Uber also requires its drivers to have third party property damage are liable for causing personal injury. States generally include third insurance. Uber also states that it provides coverage of up to $20 party insurance in vehicle registration fees. Most states also million, through the Australian insurer CGU Insurance, for its drivers require taxi and hire-car operators to hold third party property for third-party personal injury and property damage. There does not and/or public liability insurance. appear to be a publicly available product disclosure statement for the coverage, but Uber states publicly that it covers drivers during an Third party insurance premiums are usually much higher for taxis UberX trip in the event that the driver’s own insurance has been than they are for private vehicles, reflecting that taxis are driven exhausted or is not valid. further each year and have more occupants. Hire car third party premiums are also usually somewhat higher than those for private Participating in ride-sharing in states where it has not been legalised vehicles. may void a personal policy (ride-sharing is generally not covered by personal insurance policies if ride-sharing has not been legalised or People injured by a ride-sharing driver have injury coverage due to non-disclosure of commercial activities). As a result, ride- through compulsory third party schemes and possibly through the sharing drivers in states where Uber has not been legalised may not ride-sharing platform. Uber’s Australian driver agreement requires have valid personal third-party property insurance coverage at all that drivers hold compulsory third party insurance. Passengers and times. Uber has stated in private communications that its contingent other road users injured in a ride-sharing vehicle remain covered insurance policy covers drivers year-round if the insurance company for personal injury through state third party schemes even if the declines coverage due to driving for Uber. Regulating ride-sharing will driver holds only personal third party insurance. help enable drivers to obtain appropriate third-party property insurance that covers part-time ride-sharing. But if a driver is found to not have appropriate compulsory third- Notes: In NSW, ‘green slips’ are sold by private insurance companies but conditions are party coverage, the insurer may try to recover premiums and regulated by the state. All states, except Queensland and Tasmania, require taxis to hold public impose penalties (for example, in NSW the CTP insurer can claim liability or third party property. Hire cars are required to hold public liability or third party property up to $2000 from a policyholder if an incorrect premium is in NSW, ACT, SA and not required in WA and Tasmania. The ACT government introduced a new ride-sharing CTP category on 1 April 2016. deliberately paid and the insurer pays out a claim). Sources: ACT Government (2015); NSW Government (2015b); Uber (2015b); Uber (2015c); Uber (2015d)

Grattan Institute 2016 16 Peer-to-peer pressure

These rules will encourage insurers to offer coverage to ride- 2.3.4 Maintain oversight of point-to-point transport safety sharing drivers in states where it is legal, perhaps through standards, but allow providers to implement them platforms on a pay-as-you-drive basis (using GPS technology included in the ride-sharing app).55 The pay-as-you-drive model is Governments should classify ride-sharing platforms with other hire implicit in the coverage ride-sharing platforms offer today (see cars as a separate category to taxis, as the ACT has done with its Box 2). Transport Booking Service (TBS) category.

Regulators should monitor the emerging safety record of ride- Governments could allow TBSs to register and check drivers and sharing and may need to adjust regulations in light of it. vehicles, but should retain oversight of platforms’ driver and vehicle safety screening processes, operations, and outcomes. 2.3.3 Adjust the disability access regime to ensure it works That may reduce costs for governments and drivers, and speed when ride-sharing is legalised up the onboarding process for new drivers. Governments should require platforms to maintain a record of trips. All taxis and ride- State governments will need to change their disability access sharing trips should be subject to GST (Chapter 6). arrangements for point-to-point transport.56 As taxi licences become less valuable and average fares fall, discounted licence Governments should remove luxury car requirements and uniform fees become less effective in attracting and retaining wheelchair- requirements from hire cars. Fees to become licensed as a ride- accessible taxi (WAT) operators. Governments may need to sharing driver should be set no higher than an amount to cover adjust all the main elements of their disability access schemes, administrative costs. including ‘lifting fees’ and fare subsidies (paid to taxi drivers or passengers on a per-trip basis), and subsidies for the purchase of Governments should remove unnecessary driver training wheelchair-accessible vehicles. requirements for taxi drivers (for example, extensive geographic knowledge tests). They could require new ride-sharing drivers to Governments could involve ride-sharing operators directly as undergo a short training course, to ensure they understand new disability service providers. The ACT has obliged all taxi and ride- regulations. sharing firms to forward any disability-related requests to a registered provider. 2.3.5 Compensate only people who purchased licences recently or are experiencing financial hardship

The proposed regulatory changes will cut the market value of 55 NRMA Insurance (2015) has indicated it is prepared to do so. 56 perpetual taxi and hire car licences. How far licence values fall The Victorian and NSW governments are currently undertaking reviews of their transport subsidy schemes. depends on whether entry to taxi operation remains restricted

Grattan Institute 2016 17 Peer-to-peer pressure

and, if so, how profitable it remains to operate taxis. If, as we would still be positive for any perpetual Victorian taxi licence propose, state governments price annual taxi and hire-car purchased before about 2000, even if licence values and rents licences very low, perpetual licence values will be worth very little. drop to zero now (Figure 2.2).61

Should governments compensate licence owners? States seem to An ethical case may be made for offering partial compensation to be under no legal requirement to do so. Governments often taxi and hire-car licence owners who bought licenses recently and change regulations in ways that cut employment, wages, profits or who face financial hardship.62 If governments decide asset values in some sector of the economy, yet governments are compensation is warranted, it could consider when the owner not typically obliged to compensate losers.57 Victoria made this bought the licence (and the average price prevailing at that time), explicit when it inserted a statement in legislation in 1983 that no the number of licences held, and, if feasible, the owner’s financial compensation to taxi licence owners would be payable for position, including their ability to access other forms of changes to regulations.58 government support.63 The case should be considered in light of hardships experienced by other community members who may The ethical case to compensate all taxi licence owners also receive no extraordinary support. seems weak. 59 They have long been in a position to appreciate the risk that regulations could change. Taxi licence yields (rents Some governments have offered compensation to affected divided by the current licence value) have long been above the licence holders. The NSW Government’s reforms included a yield on risk-free assets, suggesting that many owners did not see compensation fund of $250 million (Table 2.1). $108 million is to them as risk-free (Box 3). Taxi industry associations have even pay perpetual taxi licence holders $20,000 for each of up to two lobbied to have governments make maintenance of licence values licences, and $142 million is for those in financial hardship. NSW an explicit policy goal. That suggests that association members has also decided to retain quotas on taxi licences, which could have long been aware that policy affects licence values.60 help ensure perpetual licences retain some value if the market segments that are taxis’ exclusive preserve (rank-and-hail, and Even if liberalisation reduces licence values and rents to zero, perhaps airport pick-ups) remain valuable to operators.64 many licence owners will still have earned positive returns. Victoria serves as an example: the after-inflation licence return

61 Grattan analysis of data from the Taxi Industry Inquiry (2012), assuming past 57 Productivity Commission (1999b) yields averaged 6 per cent of contemporaneous capital values; and post-2010 58 Taxi Industry Inquiry (2012), p.14; Transport (Compliance and Miscellaneous) from the Victorian Taxi Services Commission. Act 1983, (Vic) s 90 62 OECD (2007); Productivity Commission (1999a), Chapter 4 59 Productivity Commission (1999a), p.XI 63 Productivity Commission (1999a) p.XII-XIII 60 IPART (2015), pp.18-19 64 NSW Government (2015c); NSW Government (2015d)

Grattan Institute 2016 18 Peer-to-peer pressure

Figure 2.2: Full liberalisation in 2016 would result in negative 40 per cent, or about $500 million in all, and were accompanied returns on Melbourne metro licences purchased after 2000 by a hardship fund of just $4 million.66 The ACT’s 2015 reforms Real annualised returns, percent of starting licence value, by licence offered no compensation to licence holders. purchase year: 1975-2015 40 NSW will fund compensation with a $1 levy on all point-to-point trips for five years.67 The levy delays the benefits of reform to 20 customers, many of whom have already paid fares to support the same licence holders for years. Funding compensation from the 0 broader state taxation base may be more efficient, though state taxation also transfers income from taxpayers and creates 68 -20 inefficiencies.

-40

-60

-80

-100 1975 1980 1985 1990 1995 2000 2005 2010 2015 Notes: Returns is the internal rate of return of the cash flows: purchase price; annual rent from purchase to 2015; actual licence values and rents to 2015, deflated by the consumer price index; zero rents and licence in 2016 and after. Sources: Taxi Industry Inquiry (2012)

66 NSW’s compensation is arguably too broad: some licence holders Victorian Government (2015) 67 NSW Government (2015d). The relative elasticities of taxi demand and supply who have made substantial returns over many years will receive 65 will determine how much of this $1 levy is paid by the consumer and how much compensation. Victoria’s 2012 reforms cut taxi licence values by is paid by the supplier. 68 Productivity Commission (1999b). More generally, state governments will experience a loss of revenue from annual licence sales. They also may need to 65 Over 90 per cent of licensees who own only one or two licences will receive increase disability service funding, because they may will no longer be able to compensation payments (NSW Government (2015c), p.3). get wheelchair-accessible taxis on the road by offering discounted licences.

Grattan Institute 2016 19 Peer-to-peer pressure

Box 3: Taxi licence values and entry restrictions

Where governments restrict entry to the taxi industry, taxis are in Figure 2.3: Capital city perpetual taxi licence values have fallen sharply demand even if fares are high. As a result, taxi operators are since the arrival of ride-sharing prepared to pay to rent taxi licences. Annual taxi licence rents vary $000s, 2015 dollars, annual 600 widely across cities, because some state governments have Melbourne imposed tighter entry restrictions than others. In the most restricted Brisbane cities, annual licence rents have at times exceeded $30,000. The 500 market value of perpetual taxi licences, in turn, reflects expected future rents; they have traded for over $500,000 (Figure 2.3). Sydney 400 The ‘rental yield’ on taxi licences (annual rents divided by the Adelaide market price of perpetual licences) has often exceeded the yield on Perth safe assets like government bonds. For example, Sydney taxi 300 licence yields exceeded Australian Government bond yields by 1.5- 4 per cent in the 6 years to 2005 (Transport for NSW (2005)). The Canberra high yields suggest that licence owners believed that there was a 200 risk that governments would relax restrictions in future. Hobart Some governments have relaxed taxi entry restrictions, triggering 100 large falls in perpetual licence prices. In 2013 the Victorian government made an unlimited number of annual taxi licences for 0 metropolitan Melbourne available at a fixed price ($22,703 in 2005 2007 2009 2011 2013 2015 2015). The value of Melbourne perpetual licences fell from about Notes: All data points except late 2015/early 2016 data points are from December each year. $500,000 to just under $300,000. Latest data points may be unreliable as they reflect a very small number of transactions in some cases. Latest data points for Perth and Hobart are from November 2015. Latest data point for Preliminary data for late 2015 and early 2016 (based on the small Melbourne is an average of January and February 2016. Latest data point for Sydney is from February 2016. Latest data point for Adelaide is from December 2015. No ACT perpetual taxi number of reported transactions) shows that licence values have licences have been traded since the taxi industry reforms were implemented in September 2015. fallen across Australian capital cities, even in states that have not Deflated using the consumer price index. Sources: Australian Taxi Industry Association (2014); WA Department of Transport (2016); NSW yet changed regulations. Market participants appear to expect Department of Transport (2016); Queensland Department of Transport and Main Roads (2016); regulations to change in those states. SA Taxi Council (2016); Taxi Services Commission (2016).

Grattan Institute 2016 20 Peer-to-peer pressure

Table 2.1: Point-to-point transport reforms in NSW and the ACT Reform NSW ACT Fares Only taxis permitted to undertake rank-and-hail work. Only taxis are permitted to undertake rank-and-hail work. Regulated maximum fares for rank-and-hail to remain. Fares for Pre-booked services must be able to provide a fare estimate. pre-booked services to be de-regulated. Booking services must be Surge pricing may not occur during a formally declared emergency. able to provide a fare estimate. Safety and Ride-sharing drivers subject to a medical test, must hold a hire car Ride-sharing drivers required to purchase a licence ($100 for one insurance driver’s authority ($45 application fee) and business registration, year, $400 for five years), undergo a police and driver history check, undergo a background check, have a blood alcohol concentration annual vehicle inspection, have a zero blood alcohol concentration below 0.02 and hold CTP and third-party property insurance. Drivers required to hold CTP and third-party property insurance Vehicles must be inspected once per year, and a qualified A Transport Booking Service (TBS) will be required to provide mechanic must undertake vehicle maintenance. workers’ compensation insurance for a driver if the TBS restricts a driver from accepting bookings from a competitor. Disability access Review of Taxi Transport Subsidy Scheme to be undertaken Wheelchair accessible taxi system unchanged. Wheelchair accessible vehicle driver incentive payment to increase from $7.70 to $15 and passenger subsidy cap increased from $30/trip to $60/trip. Interest-free loans for WAT purchases. Licences and In Sydney, no new perpetual taxi licences to be issued for four Annual taxi licence fees reduced from $20,000 to $10,000 on 30 compensation years. For the rest of NSW, annual taxi licences will be issued. October 2015 and to $5,000 in October 2016. A $250 million industry adjustment package for taxi and hire-car Hire car licence fees cut from $4,600 to $100 per annum. licence holders, financed by a $1 surcharge on each ride-sharing No compensation to licence holders. and taxi trip for five years. Compensation of $20,000 for each licence holder (up to two licences, total $108 million), a $142 million hardship fund and a buyback of perpetual hire-car licences. Administration Geographic knowledge, English language, service, presentation All taxi network operators and ride-sharing apps classified as a TBS. and other training requirements removed. Ride-sharing and taxis to operate through a TBS. Independent taxis Establishment of a new regulator to audit and oversee transport and traditional hire cars do not need to be affiliated with a TBS. companies. Transport companies to oversee service standards. For ride-sharing apps, an annual accreditation fee of $20 per driver. Removal of some training requirements for taxi drivers and the removal of regulations, such as the obligation to wear a uniform. Some minimum training requirements for ride-sharing drivers. Notes: There is only minimal publicly available information on the Western Australian and South Australian government’s interim reforms scheduled to start in July 2016. Sources: Grattan Institute; NSW Government (2015c); NSW Government (2015d); Han (2015); ACT Government (2015); WA Government (2015).

Grattan Institute 2016 21 Peer-to-peer pressure

3 A room of one’s own: peer-to-peer accommodation

Peer-to-peer companies such as Airbnb and Stayz not only make should only restrict short-stay rentals to manage noise or loss of it easier to find places to stay – a fold-out couch in someone’s amenity that affect neighbours. Concerns about rents are lounge room, an entire home or anything in between – they overblown and are not a valid basis for regulation. change, even revolutionise, the commercial accommodation experience. Travellers gain the opportunity to stay in private Local governments should respond where disruption from a homes and meet local people they would never have otherwise specific property is troubling neighbours. They should allow met. These accommodation platforms expand the range of occasional or single-room short-stay rentals, and only restrict choices for travellers, provide extra income for hosts, and can put continuous whole-premise short-stay rentals if there is strong otherwise idle real estate to valuable use. They can boost tourism evidence they are damaging neighbourhood amenity. and make it easier to manage temporary surges in accommodation demand – such as for a major sporting event or a State governments can also play a role by allowing owners’ natural disaster.69 corporations to be more effective in managing short-stay rentals. They could enable owners’ corporations to hold owners liable for But short-stay rental platforms pose challenges, too. The short- disruptions caused by their guests, and consider allowing owners’ stay rentals can affect neighbourhood amenity, divide members of corporations to control continuous, whole-premise short-stay owners’ corporations and displace longer-term renters.70 They can rentals as they see fit. make it easy to circumvent zoning and other regulations. 3.1 Accommodation platforms create benefits for travellers More regulation is needed to help secure the benefits of this new and hosts, but can negatively affect neighbours market without imposing costs on neighbours. Governments New accommodation platforms, particularly Airbnb, have expanded the traditional holiday house rental market to include 69 Airbnb listings and occupancy increased during the FIFA World Cup in Brazil homes, apartments and rooms in properties across Australia, and the Sochi Winter Olympics in 2014 (Feeney (2014); The Economist (2016)) particularly in inner-city areas of major cities. and the Cricket World Cup in Australia in 2015 (Dow (2015)). Airbnb has partnered with the Victorian government to help provide emergency accommodation in the event of natural disaster (Airbnb (2014b)). 3.1.1 Most properties on accommodation platforms are near 70 ‘Owners’ corporations’ go by different names across the states: owners’ the beach and in inner cities corporation (Victoria, ACT); strata corporation (South Australia); strata scheme (NSW, Western Australia, Tasmania); body corporate (Queensland, Northern Accommodation platforms are well established in Australia. Territory). Airbnb, the biggest global accommodation platform, began

Grattan Institute 2016 22 Peer-to-peer pressure

operating in Australia in 2009 and has more than 66,000 listings Figure 3.1: Airbnb is growing rapidly (including rooms and entire properties).71 Stayz has more than Thousands of bed spaces, New South Wales, six-monthly 40,000 entire-property listings.72 The approximately 28,000 100 Sydney bed spaces listed as available on Airbnb in Sydney in Motels and guesthouses January 2016 may be as much as 12 per cent of commercial beds across NSW (Figure 3.1).73 Thousands 80

Accommodation platforms have probably taken some market Hotels share from hotels, bed-and-breakfasts, and serviced apartments. 60 Hotels in some US cities have cut prices in response to local peer-to-peer competition.74 But peer-to-peer accommodation is also a distinct product that expands the choices available and 40 encourages people to travel.75 It appears to be of particular interest to tourists. Serviced apartments

Most people travelling to Australian cities want to stay near the 20 beach or in the inner-city, and that is where most Airbnb listings Airbnb (Sydney only) are. In Sydney, for example, the top 20 Airbnb suburbs (accounting for about 40 per cent of listings) are in locations within 0 2008 2009 2010 2011 2012 2013 2014 2015 5-10 kilometres of the city or near the beach. Notes: Latest Airbnb data is from January 2016. Non-Airbnb premises do not include those with under 15 rooms, so is likely an underestimate of total commercial accommodation. The number of people a listing accommodates is used as a proxy for bed spaces, and 71 Correspondence with Airbnb; Inside Airbnb data (Inside Airbnb is a website availability in the year ahead as at January 2016 is used to adjust listings for availability. that compiles publicly available Airbnb data for major cities). Sources: ABS (2014b), Inside Airbnb 72 Stayz website. Some Stayz listings may also be listed on Airbnb; some Stayz listings are commercial accommodation. In Sydney’s Bondi Beach area, about 9 per cent of residences are 73 Airbnb bed spaces include those in entire premises, private rooms and shared listed on Airbnb for at least part of the year, and a further 4 per rooms, adjusted for availability in the year ahead. NSW commercial accommodation with 15 or more rooms total 202,000 bed spaces in medium and cent of residences list rooms (Figure 3.2). In other inner Sydney large hotels, motels and serviced apartments (ABS (2014b)). The occupancy beachside suburbs, up to 5 per cent of residences are listed on rate for commercial accommodation is likely to be significantly higher than for Airbnb. Airbnb listings. 74 Zervas, et al. (2014); Freed (2015) and Truong (2015). 75 Airbnb (2013a)

Grattan Institute 2016 23 Peer-to-peer pressure

Figure 3.2: Airbnb listings are in tourist areas and in the inner-city 3.1.2 Most Airbnb bed-nights are in non-primary premises Dwellings in a Sydney suburb listed on Airbnb as % of total dwellings listed for the majority of the year 20 18 One of the distinctive aspects of peer-to-peer accommodation is 16 that it permits guests to stay with locals in their own homes, or in a unique home while the owners are away. 14 Private room 12 Yet much Airbnb activity is not in primary residences, even if a Entire house/apartment 10 primary residence is defined, generously, as any premise 8 available or occupied for short-stay use less than 120 days per 6 year (Figure 3.3). Prior-year occupancy shows that while only 10 per cent of premises are non-primary residences, 40 per cent of 4 bed-nights are in non-primary residences (left panel of Figure 2 3.3).76 But that measure probably understates the true frequency 0 of non-primary residence listings, because it includes listings that commenced partway through the year.

Using, instead, coming-year availability data, almost 40 per cent of listed Airbnb premises are non-primary residences, and over 60 per cent bed-nights are in non-primary residences (right panel of Notes: There is a minor upward bias as Airbnb listings in 2016 are compared to 2011 Census data. Some hosts may list multiple private rooms in a single property, or list both Figure 3.3). That measure probably overstates the true frequency, rooms in a property and the entire property. Not all listing are available full-time. The because some people list properties as available that they then do suburb of Bondi Beach is included in Bondi. not rent out. So while primary residences and rooms do comprise Sources: Inside Airbnb; ABS (2011) the majority of Airbnb’s listings, they may be less than a third of 77 Airbnb’s business.

76 ‘Bed-nights’ refers to the number of people a property accommodates multiplied by the number of nights listed available in the year ahead as at January 2016, or occupied in the year to March 2016 (Airbnb (2013b); Schetzer and Battersby (2016)). 77 Others studies also find that non-primary residences make up a significant proportion of Airbnb revenue (Schlesinger (2016); O'Neill and Ouyang (2016)).

Grattan Institute 2016 24 Peer-to-peer pressure

Figure 3.3: Non-primary residences make up 40 per cent of Figure 3.4: Multi-premise operators offer 25 percent of entire- occupied bed-nights and 60 per cent of available bed-nights premise listings on Airbnb Per cent of total listings/occupancy/bed-nights Hosts/entire dwellings listings, Melbourne and Sydney 100 100 Non- primary residences 80 80

60 60 Primary Hosts with: residences 10+ properties 40 3-9 properties 40 2 properties 20 1 property Rooms 20

0 Premises Bed-nights Premises Bed-nights 0 Previous year occupancy Coming year availability Hosts Properties Notes: Non-primary residences defined as those occupied or available for more than 120 Sources: Inside Airbnb; Grattan analysis days per year. Coming year availability data is from Inside Airbnb, and is Melbourne and Sydney only, as of January 2016. Previous year occupancy is estimated using data supplied by Airbnb, Australia-wide as of the year to March 2016. Occupied bed-nights is Many ‘non-primary’ residences are listed by people who have occupancy multiplied by the average number of people a listing accommodates. Coming more than one listing. In Melbourne and Sydney, just seven per year available bed-nights is the number of people a listing accommodates multiplied by cent of hosts list almost a quarter of all entire-property listings. availability in the coming year; excludes already-booked nights and may include nights listed as available that owners do not plan to fully book. Previous year occupancy data may and hosts with more than 10 properties list almost 10 per cent of incorrectly include non-primary residences as primary if they joined the platform partway properties (Figure 3.4). through the year. Sources: Airbnb; Inside Airbnb; Grattan analysis

Grattan Institute 2016 25 Peer-to-peer pressure

3.1.3 Impacts on rents are likely to be modest But even big rent increases would not provide much of a case for cracking down on short-stay rentals. Market prices provide signals Some are concerned that short-stay renters displace long-term about the most valuable use of assets, so restricting short-stay 78 residents and push up rents. Long-term tenants have clearly rentals to keep local rents down reduces incomes in aggregate. been displaced from the inner-city beachside suburbs where (in a Concerns about income distribution should be addressed through few cases) up to 15 per cent of homes are now listed as available the national tax and transfer system. for at least part of the year for short-stay rentals on platforms (Figure 3.2), even if some of the listings may not otherwise have 3.1.4 Short-stay guests can disrupt neighbours been available for long-term rent.79 Short-stay peer-to-peer rentals can, however, directly impose But any rent increases caused by the rise of short-stay rentals are costs on neighbours. The imposts seem most acute in apartment likely to be localised or small. Short-term use of housing is a small complexes, where neighbours are in close proximity and share fraction of the city-wide housing stock. The 25,000 Sydney ownership and use of common areas and facilities such as lifts, bedrooms – including those listed part-time – listed on Airbnb gyms and pools. 70 per cent of whole-premise Airbnb listings in comprise about half of one percent of Sydney’s four and half Melbourne and Sydney are apartments. In some apartment million or so bedrooms, and about 2 per cent of Sydney’s rental buildings in Melbourne, more than a quarter of apartments are housing capacity.80 A demand increase of that size is unlikely to rented out short-term.82 cause rents to rise much across the city. One US study – commissioned by Airbnb itself – estimates that Airbnb increased Residents of some complexes report that disruptions from short- 83 the monthly rent for an average one-bedroom apartment by $6 a stay tenants rented through peer-to-peer platforms are severe. month in New York City and by $19 in San Francisco, or under Long-term residents mostly complain about noise and ‘anti-social 1 per cent.81 behaviour’, but some also note increased wear and tear on common areas, security issues, overcrowding and a ‘hotel-like’ atmosphere.84 There have also been occasional more serious

78 Thomson (2016); Kusisto (2015); Hill (2015). 79 Prosper Australia (2015) notes that many apartments are empty year-round. Some landlords, attracted by short-stay rental’s flexibility, may now be renting 82 Independent Panel on Short-Stay Accommodation in CBD Apartment out properties they would otherwise have kept empty. Buildings (2015), p.23. 80 Includes bedrooms available only part of the year. Assumes the Australia-wide 83 Many submissions to the NSW Parliamentary inquiry into the regulation of rental housing share of 25 per cent applies in Sydney (RBA (2015)). short-term holiday letting describe amenity disruptions due to short-stay rentals. 81 Kusisto (2015) 84 Owners Corporation Network (2015), p.6; Watergate (2015).

Grattan Institute 2016 26 Peer-to-peer pressure

Figure 3.5: Short-stay rentals attract more complaints from It is unclear how widespread the problems are. In one small neighbours survey of large apartment complexes, short-stay renters attract Complaints about resident behaviour, per 365 nights complaints about three times as often as long-term renters do, but 1 the absolute rate still seems quite low at just under one complaint per year per apartment (Figure 3.5).86 Short-term rentals of detached dwellings through peer-to-peer platforms can also 0.8 Other attract complaints about noise from party houses, car parking Anti-social behaviour congestion, rubbish and anti-social behaviour, particularly in 87 Noise holiday areas at peak periods. 0.6 Many short-stay operators and their guests pay nothing for the noise and disruption they foist on neighbours, so there is almost 0.4 certainly much more of it than there should be.

3.1.5 States make it difficult for neighbours to manage 0.2 repeat disruptions from short-stay tenants All states have environmental laws and regulations aimed at managing neighbourhood disturbances such as from excessive 0 noise. If the laws are breached, local council officers or the police Short-term Long-term have the power to issue fines.88 Notes: Data is from six owners’ management companies for Melbourne CBD and inner-city apartments complexes, adjusted for estimated occupancy rates of long-term and short- term residents. Data may be subject to response bias. Short-term residents staying with an But in many states, the process for making an application is owner are not included. cumbersome, expensive and can take time to enforce. It may be Sources: Independent Panel on Short-Stay Accommodation in CBD Apartment Buildings (2015); Grattan analysis. difficult to prosecute short-stay guests, and the owner of the

85 incidents. Short-term use may violate building codes, with 86 implications for the insurance coverage of owners’ corporations. Independent Panel on Short-Stay Accommodation in CBD Apartment Buildings (2015), Annexure 2. 87 Bibby (2013); Dobrohotoff v Bennic [2013]; submissions to the NSW Parliamentary inquiry into the regulation of short-stay holiday letting. 88 For example, the Environment Protection (Residential Noise) Regulations 85 The Daily Telegraph (2015). 2008 (Vic) and Protection of the Environment Operations Act 1997 (NSW).

Grattan Institute 2016 27 Peer-to-peer pressure

property may not be accountable for disruptions caused by Box 4: Local governments’ approaches to short-stay rentals occupants.89 For example, in NSW, if a person is disturbed by unreasonable noise by a neighbour, the police or the local council. Some local governments restrict short-stay rentals or apply can issue a warning or a noise abatement direction to the onerous conditions on hosts. Randwick City Council in Sydney offending persons, which can remain in force for up to 28 days.90 has threatened peer-to-peer hosts with significant fines for Neighbours can also seek ongoing noise control orders, but must running even occasional unauthorised short-stay accommodation. go to court or to a tribunal to obtain one. Neither remedy may Waverley Council in Sydney (covering Bondi and surrounding have much effect on the behaviour of subsequent short-stay suburbs) requires that hosts obtain planning approval and occupants.91 demonstrate compliance with building codes. The WA cities of Busselton, Margaret River and Fremantle require planning As a result, noise and environmental regulation does not offer approval and/or registration, and restrict the number of guests. sufficient recourse against disruptions from peer-to-peer accommodation. And councils, with limited resources, may not be Other local governments are more liberal. Kiama, Gosford and able to manage the volume of complaints they receive. In Shoalhaven in NSW generally allow short-stay letting without response, some councils impose onerous conditions on short-stay requiring planning permission. The City of Hobart states that ‘one- rentals; others take a more liberal approach (Box 4). off, occasional usage’ by short-stay guests does not usually require council approval, but any changes to a dwelling to Queensland’s ‘party house’ legislation, introduced in 2014, accommodate short-stay guests require planning approval. enables local governments to require some or all ‘party house’ owners to obtain permits, which can include conditions such as Notes: The Shire of Augusta-Margaret River limits short-stay rentals to properties close to recognised ‘major tourist attractions’ and coastal settlements. Sources: McKenny (2014); Sansom (2015); WA Government (2012); City of Fremantle (2016). The Shire of Augusta-Margaret River (2014); City of Hobart (2015).

occupancy limits and noise controls.92 Registration also allows 89 The law on this is somewhat unclear. In a 2011 NSW case, a noise abatement order was served on a landlord due to the noise created by tenants (Jean local governments to quickly identify the property that is the Whittlam v Sara Hannah & John Hannah [2011]). Nonetheless this is a subject of a complaint and to contact the operator, and to impose cumbersome process. escalating penalties, up to bans, on landlords if they breach 90 NSW Environmental Protection Authority (2015). A breach of this can lead to a conditions repeatedly. fine of $200. 91 Alternatively, neighbours may take legal action for a potential breach of the common law concept of ‘nuisance’. If successful, a resident can receive a court order stopping the noise and/or receive compensation. This is also a time consuming and expensive process. 92 Larkins and Kane (2014)

Grattan Institute 2016 28 Peer-to-peer pressure

3.1.6 States make it difficult for owners’ corporations to Table 3.1: Many states do not permit owners’ corporations to manage short-stay rentals prohibit short-term leasing Ability of owners’ corporations to State/territory Owners’ corporations are important in the management of make rules about leasing disruptions from short-stay accommodation. They manage common property in a multi-dwelling complex, and define and Rules restricting leasing expressly New South Wales prohibited under state legislation enforce rules that affect the use of properties.93 Some states South Australia oblige people to try to resolve disputes through their owners’ Uncertain if owners’ corporations can Victoria corporation. For example, in NSW, members of owners’ make rules restricting short-term leasing Queensland, Northern corporations must seek a solution through mediation or through Territory the owners’ corporation and the state administrative tribunal Western Australia before contacting their local government for assistance. Can set rules requiring any lease of a Tasmania, ACT property to be a minimum of up to six Limitations imposed by state governments on owners’ months Source: State legislation; Independent Panel on Short-Stay Accommodation in CBD corporations’ powers to make rules make it difficult for them to Apartment Buildings (2015). manage short-stay rentals. Cease orders that may be effective with long-term residents are not effective with short-stay tenants. owners from leasing their properties.95 Tasmania and the ACT, by Owners’ corporations may not be able to hold apartment owners contrast, allow owners’ corporations to require any lease of a accountable for breaches of rules by their short-stay guests.94 property to be at least six months long.96 The law is unclear in Victoria, Western Australia and Queensland.97 Moreover, many state governments appear to prohibit owners’ corporations from making rules to ban short-term leasing (Table 3.1). New South Wales and South Australia expressly prohibit owners’ corporations from making rules restricting apartment 95 Independent Panel on Short-Stay Accommodation in CBD Apartment Buildings (2015). For example, s 49(1) of the Strata Schemes Management Act 1996 (NSW) states that ‘No by-law is capable of operating to prohibit or restrict the devolution of a lot or a transfer, lease, mortgage, or other dealing relating to a lot’. See also section 180(4) of Body Corporate and Community Management 93 Act 1997 (Queensland) and Community Titles Act 1996 (SA) s 37 1(a). Strata Community Australia (2015); For example, the Strata Schemes 96 Management Act 1996 (NSW) and Body Corporate and Community Strata Titles Act 1998 (Tas) s 91; Unit Titles (Management) Act 2011 (ACT) s 109 Management Act 1997 (Queensland). Rules are called ‘by-laws’ in some states. 97 94 Independent Panel on Short-Stay Accommodation in CBD Apartment The Victorian Civil and Administrative Tribunal (VCAT) recently ruled that a Buildings (2015), p.23; Lee v Owners Corporation No. 501391P; Benson (2015) body corporate did not have the power to prevent owners from renting their apartments for less than 30 days (the owners’ corporation is appealing this

Grattan Institute 2016 29 Peer-to-peer pressure

Box 5: Legal disputes about short-term stays in Victoria Box 6: NSW strata reforms streamline dispute processes

A recent Victorian Civil and Administrative Tribunal (VCAT) Reforms to strata laws recently passed by the NSW Parliament decision ruled that an owners’ corporation did not have power to increase maximum penalties for breaches of rules and will make it prohibit an owner from letting out his properties for short stays. easier for owners’ corporations to enforce rules, particularly for The owners’ corporation had a rule that prevented the owner of an repeated breaches. The changes are expected to take effect from apartment from using it for any business. VCAT held that the mid-2016. owners’ corporation could not use this rule to ban short-term rentals. The owners’ corporation is considering an appeal. Penalties for breaches of rules will double, with the new Act legislating a maximum penalty of $1100 for each offence. A In another case involving the same apartment complex, repeat offence within a year will warrant up to $2200 in fines (and Melbourne City Council tried to stop short-stay letting on the basis substantially higher for breaches of occupancy limits). For that it did not comply with the Building Code of Australia. The continued breaches, owners’ corporations can apply directly to the building’s classification allowed it to be used for ‘dwellings’; Tribunal without issuing a fresh notice to comply. Council argued that the term ‘dwelling’ excluded short-stay accommodation. The short-stay rental operator ultimately won a The NSW strata reforms do not empower owners’ corporations to ruling in the Victorian Court of Appeal on the basis that the term pass rules limiting short-stay rentals. The NSW parliamentary ‘dwelling’ was not confined to long-term usage. inquiry into short-stay holiday rentals, due to report this year, may consider that option. Sources: Johanson (2015); Watergate (2015); Independent Panel on Short-Stay Accommodation in CBD Apartment Buildings (2015) Notes: The Law Society of New South Wales (2015))recommended that the legislation expressly permit this. However, even a clear prohibition on short-term leasing would not Sources: NSW Fair Trading (2015); The Law Society of New South Wales (2015) prohibit short-stay rentals if courts do not consider the rental 98 prohibition on the running of a business from a dwelling, or its use arrangement a lease. Some owners’ corporations have sought 99 to stop short-stay rentals by relying on laws such as the for non-residential purposes. They have not always succeeded, as Box 5 shows.

The process for handling breaches of owners’ corporation rules decision); Owners Corporation PS501391P v Balcombe (Owners can be slow and unreliable. Even if owners are legally liable for Corporations) [2015] VCAT 956 (29 June 2015)); Good and Prudden (2015). 98 A recent Victorian VCAT decision found that long-term tenants had not subleased to their short-stay guests (Robb (2016)). 99 Johanson (2014)

Grattan Institute 2016 30 Peer-to-peer pressure

breaches by their short-stay guests under owners’ corporation Local governments should allow short-stay rentals but control rules and liable for fines, practically it can be difficult for owners’ disruptions. They should: corporations to penalise owners, as identifying guests and gathering evidence can be difficult and pursuing a case can be Freely permit continuous letting of spare rooms when the costly. Recent New South Wales strata reforms made it easier for landlord (or long-term tenant) is also occupying the property, owners’ corporations to enforce rules, particularly for repeated and freely permit all occasional short-stay whole-premise breaches (Box 6), but did not change rules relating to short-stay letting. use of property. Respond to disturbances in a timely way, applying a 3.2 What policymakers should do progressive sanction regime on owners whose short-stay tenants are disruptive. Sanctions might include fines and Appropriate regulation of short-stay accommodation will require prohibition for a period. All parties should retain recourse to changes to local government practices and to state government state tribunals and courts. limits on owners’ corporations powers.100 Finding an acceptable balance among competing concerns can be difficult, but laws Prohibit continuous whole-premise short-stay rentals only should help people limit unreasonable noise and loss of amenity, 101 where less restrictive approaches have been trialled and while not unduly restricting the use of property for short stays. shown to give insufficient protection to residents.

3.2.1 Local councils should focus on controlling State governments should support local governments by disruptions, not on limiting short-stay rentals streamlining their tribunal complaint and dispute resolution processes. State and local government regulations governing short-stay rentals need to be changed. In some areas, rules are too 3.2.2 Governments should empower owners’ corporations restrictive. In others, neighbourhood amenity is not protected. to hold short-stay rental operators accountable

100 The Victorian government established an independent panel in February State governments should give owners’ corporations greater 2015 to recommend ways to regulate short-term rentals in CBD apartment powers to hold owners liable for disruptive behaviour by their buildings; the government has yet to respond to recommendations. The NSW short-stay guests: Government started an inquiry into the regulation of short-term holiday letting in in September 2015. The Harper Review identified planning and zoning as a priority area for review (recommendation 8) (Harper, et al. (2015)). Governments should permit owners’ corporations to make 101 Independent Panel on Short-Stay Accommodation in CBD Apartment property owners liable for fines if short-stay guests break rules Buildings (2015)

Grattan Institute 2016 31 Peer-to-peer pressure

that cause disruption to other residents.102 3.2.3 State governments could permit owners’ corporations to limit full-time short-term letting They could also empower owners’ corporations to levy charges for short-stay use of dwellings, to cover the extra The reforms proposed above might not, in practice, sufficiently usage of common property by short-stay guests or even to reduce breaches by short-stay tenants. And they may not go far reasonably cover less tangible costs of disruption to long-term enough for some apartment complex residents whose amenity is residents. The owner should have recourse to dispute the affected by short-stay operations even if the guests are well- charge to a state tribunal.103 behaved. Some residents are disturbed by a hotel-like atmosphere in their complex, worried about having a succession Any such changes to owners’ corporation rules would need to be of strangers in their shared public spaces, or are inconvenienced supported by the standard majority of owners (often 75 per cent). by cleaning crews. They would like their owners’ corporation to have the power to ban short-stay rentals, at least full-time ones.104 State governments should also make it easier for owners’ corporations to enforce rules, particularly for repeated breaches Members of some owners’ corporations seem to disagree strongly (as has been done in NSW: see Box 6). When there are on whether to ban or otherwise limit short-stay rentals. Giving persistent breaches, state tribunals should be empowered to bar power to owners’ corporations to vote on whether to ban short- short-stay rentals in that property for a period of time. stay rentals would benefit some people, while denying them that power would benefit others. Empowering owners’ corporations in this way should improve amenity for residents of multi-dwelling complexes. Hosts would But if the reforms proposed above are judged by governments to probably try hard to ensure their guests adhere to owners’ be insufficient, governments should permit owners’ corporations corporation rules. They may be able to require a bond at the time to vote on whether to prohibit full-time short-stay letting of of booking, and to recover owners’ corporation fines from short- properties not lived in by the owner.105 State governments should term tenants. ensure that owners’ corporations cannot prevent owner-occupiers from occasionally renting out their properties short-term, or from 102 Owners’ corporations may already have this power in some states. Owners’ renting out single rooms in their properties short-term while they corporations could also be empowered to issue fines directly without the need to are living in them. apply to a state tribunal to levy a fine, as is permitted in South Australia (Community Titles Act 1996 (SA) s 34; Strata Titles Act 1988 (SA) s 19). The 104 person can appeal a fine to the Magistrates Court. As noted in Table 3.1, Tasmania and the ACT already permit owners’ 103 Owners’ corporations can already require payment for damage to common corporations to require any lease of a property to be a minimum of up to six property. Payments for less tangible costs may need to be capped by legislation. months, though it is not clear that short-stays would be considered a lease. 105 Alternatively, the tribunals could be so empowered. As proposed for NSW by The Law Society of New South Wales (2015).

Grattan Institute 2016 32 Peer-to-peer pressure

4 Putting platforms to work: the job market

Peer-to-peer platforms will, in the main, make workers better off. ensure that workers are well-informed when they elect to do peer- They make it easier to find the right person for a job or task, and to-peer work and do not unwittingly take on risks. help workers to find flexible work. Government should monitor how platform work develops before Platforms are likely to create work, but may create few jobs. At implementing other policies. It should not yet change the rules present, only nine per cent of Australian workers are independent that oblige contractors to obtain Australian Competition and contractors. That proportion is likely to grow, since many workers Consumer Commission authorisation to collectively bargain. It on platforms will be independent contractors, not employees. The should not yet broaden unfair dismissal laws to include platform consequences for workers’ entitlements and their relationships contractors, or require that platforms provide the means for with employers are significant and are examined in this chapter. workers to contribute to superannuation or buy insurance, or Many workers will only make the switch if they actively prefer require them to allow workers to export their customer reviews. As platform work. But platforms can, under some circumstances, more is learnt, some or all of these steps may be warranted. circumvent labour regulations and undercut firms and workers that adhere to the rules. 4.1 It is too early to tell how widespread platform work might become Policymakers should not, however, reclassify platform workers as employees, or create a new ‘platform contractor’ category, as The amount of work transacted on peer-to-peer platforms is still some have proposed.106 They should instead take just a few small. It seems likely that fewer than half of one per cent of adult safeguarding steps, while monitoring how platform work develops. Australians (80,000 people) work on peer-to-peer platforms more Governments should encourage platforms to provide information than once a month. About 20,000 people drove with Uber at least about risks and responsibilities to workers. They should tighten once in the four weeks to December 2015.107 About 70,000 ‘sham contracting’ provisions that prevent employers from tradespeople are registered on hipages, an Australian platform for misclassifying genuine employees as independent contractors in home-improvement, but it is not publicly known how many work 108 order to avoid paying them entitlements, and increase penalties through the platform in a given month. Airtasker, an Australian for breaches of these provisions. They should require platforms in ride-sharing and a few other riskier occupations to pay workers’ 107 Correspondence with Uber; Deloitte Access Economics (2015a), p.14 compensation insurance premiums. All these steps will help provides estimates that imply about 0.4 per cent of NSW adults worked on peer- to-peer platforms in 2015, including all tradespeople on hipages. 108 Tucker (2016). Hipages customers post 80,000 new jobs each month, worth 106 Hagiu (2015); Harris and Krueger (2015) well over $1.5 billion, or close to 2 per cent of all Australian trades jobs. Some

Grattan Institute 2016 33 Peer-to-peer pressure

odd-jobs platform, says that ‘many thousands’ regularly work Table 4.1: Estimates of income earning participation on platforms through the site, though the total value of jobs posted each month Measure of income US Australia is currently just $3.5 million, enough to support fewer than fifteen earned on platform (per cent of adults) (per cent of adults) hundred workers full-time at the minimum wage.109 There is little public information about how many Australians are active on other Actively working 0.4 <0.5 late 2015 in a single month in 2015 sites such as Freelancer, Expert360, 99Designs and Etsy. (labour only) There is, by contrast, credible data on how many Americans work Actively earning 1.0 0.7 through platforms. Bank payment data suggest that about 0.4 per late 2015 in a single month. in 2015 cent of US adults did paid work on a platform in September (labour & rentals) 2015.110 Other credible studies come to similar findings.111 Ever earned income 4 – 18 5 (labour & rentals) ever ever A further 0.6 per cent earned income that month by renting out an Sources: Actively working: US - 0.4 per cent (JPMorgan Chase Institute (2016), 0.3 per cent (0.5 per cent of workers; Katz and Krueger (2016)), 0.8 per cent (McKinsey Global asset on a platform. About one per cent of adults had worked on a Institute (2015)), 0.2-0.8 per cent (Harris and Krueger (2015)), Australia - 0.4 per cent (0.6 platform at some point in the three years ending in September per cent of working age; Deloitte Access Economics (2015a)). Actively earning income: US - 1.0 per cent (JPMorgan Chase Institute (2016)), Australia - 0.7 per cent (1 per cent of 2015, and three percent had rented out assets. working age; Deloitte Access Economics (2015a)). Ever earned income: US - 18 per cent (Burson-Marsteller (2016)) , 7 per cent (PwC (2015)), 4.2 per cent (JPMorgan Chase Table 4.1 summarises these and other estimates of participation Institute (2016)), Australia - ~5 per cent (8 per cent of employed people; Airtasker (2015b)). in platform work and earning more broadly. It also includes some estimates from other studies that are much larger because they Work on platforms, however, is growing fast. The number of include people who are not frequently active, or are based on self- active Australian Uber drivers more than tripled in 2015. Global reported participation from surveys (not bank payment data). site Freelancer reports that the number of jobs posted has grown 25 to 30 per cent a year.112 And US banking data also suggest that the number of people finding work on platforms there has also grown fast: ten times as many people earned income on platforms in September 2015 as three years earlier.113 However, rapid growth may not last: in the US the number of ride-sharing more narrow definitions of ‘peer-to-peer’ work might exclude hipages, as the drivers, the largest single category of platform worker, is already workers are all skilled tradespeople. 109 Correspondence with Airtasker and Airtasker website. 110 JPMorgan Chase Institute (2016) 111 Katz and Krueger (2016) find that 0.5 per cent of US workers (~0.3 per cent 112 Freelancer (2016) of adults) were employed through peer-to-peer labour platforms in late 2015. 113 JPMorgan Chase Institute (2016)

Grattan Institute 2016 34 Peer-to-peer pressure

close to the number of taxi drivers.114 A growth slowdown in ride- Figure 4.1: A third of workers are not in ‘traditional’ employment sharing employment is likely as the market matures, as has Per cent of employed persons 100 already occurred in some major markets such as New York. Other business owner/operators 4.2 Platforms create contractor roles, with entitlements that Independent contractors ‘Non-traditional’ differ from those of employees 80 Labour hire employment Under today’s regulations, most platform workers are likely to be Casual classified as independent contractors. They have more flexibility 60 than employees do, but may lack many protections that Part time employees enjoy.

4.2.1 Platforms shift the balance towards independent 40 Full time ‘Traditional’ contracting employment The nine per cent share of the workforce held by independent 20 contractors has changed little for fifteen years (Figure 4.1). As platforms grow, many of them are likely to host independent contractors exclusively, increasing the share of independent 0 contractors in the workforce. While platforms will grow in part as 2001 2014 existing contractors are attracted to the search and transaction Notes: ‘Traditional’ employment includes part-time and full-time workers with paid-leave entitlements. Casual workers are defined as employees without paid leave entitlements environments they offer, those advantages will also grow the pool Labour hire workers are people paid by a labour-hire firm and are all considered casual of independent contractors. workers. The number of independent contractors in 2001 is an estimate from Productivity Commission (2006), Table A.6. Other business owner/operators are owner/managers of unincorporated and incorporated enterprises, less the number of independent contractors. Platforms appear to have taken root in sectors and occupations in Sources: ABS (2013); ABS (2014a); ABS (2015); Productivity Commission (2006) which independent contractors are already common (though many employees also work in them). Almost two-thirds of in those sectors.115 Platforms have been successful in commercial independent contractors work in construction, administration and driving (Uber and Lyft), household repair and construction support services, arts and recreation, professional and technical 115 services, and transport. Only a quarter of employees work ABS (2015). Independent contracting is more evenly spread across occupations than across industries. It is more common among technicians and 114 tradespeople, machinery operators and drivers, labourers, and professionals and Hall and Krueger (2015) managers than among other occupations.

Grattan Institute 2016 35 Peer-to-peer pressure

(hipages), household services and errands (Airtasker), writing, Table 4.2: A ‘multi-factor’ test determines whether workers are website design, IT services and data entry (Freelancer, Envato). independent contractors or employees Suggests That is because platforms suit tasks that the customer only needs Suggests Factor independent employee infrequently and do not involve complex teamwork with colleagues contractor or deep knowledge of a specific workplace. Does the worker have autonomy? No Yes Platforms appear to have gained less share where work is • Can the work be delegated? required on a continuous basis, involves complex teamwork, or • Does the worker choose hours and decide the method of work? requires deep knowledge of the workplace. There are few large platform workforces in manufacturing, retail and wholesale trade, • Does the worker set standards? health care, or financial services. Traditional employment Does the worker operate like a business? No Yes relationships, whether continuing or casual, seem to prevail for • In a position to make a profit/loss? now in these industries. • Own public liability/accident insurance? 4.2.2 Workers are defined as independent contractors if • Issues an invoice? they have autonomy and operate like a business • Meets own GST obligations? • Pays own expenses and supplies own Courts apply a ‘multi-factor test’ in deciding whether a worker is equipment? an employee and therefore covered by national employment • Responsibility for defective work? standards, minimum wages and other workplace entitlements • Paid by result rather than hours? (Table 4.2).116 Is the worker able to work for other No Yes principals? Courts consider the ‘totality of the relationship’ between the Notes: This is not an exhaustive list of all the factors that a court could consider Sources: Fair Work Ombudsman (2015); Fair Work Building and Construction (2015); ATO parties. They typically find that a worker is an independent (2015f). contractor if he or she has autonomy and operates like a business. Workers on a platform that merely serves as a Table 4.2 sets out some factors that courts have considered, and matchmaker are unlikely to be found to be employees of that Box 7 considers some factors a court may consider if it is asked to platform. Yet a platform that exercises tight control over a worker determine whether Uber drivers are employees of Uber. could be found to be an employer.

116 The multi-factor test is also used to establish vicarious liability of the employer for actions of the employee and tax obligations of the employer.

Grattan Institute 2016 36 Peer-to-peer pressure

Box 7: Are UberX drivers independent contractors? 4.2.3 Independent contractors and employees have different entitlements Australian courts seem more likely to classify an Uber ‘driver- partner’ as a contractor than as an employee, but much will Platform contractors, like other independent contractors, are not depend on how much control Uber exerts over how drivers work. entitled to many conditions that must be provided to employees, unless they negotiate to have them in their contracts. Some factors suggest that Uber drivers are contractors. They choose their work hours and can make a profit or loss. They are Independent contractors are not entitled as of right to a minimum paid per trip, supply their vehicles and (usually) their phones. wage, maximum weekly hours, paid leave, notice periods after They must manage their own income tax payments and have their dismissal, redundancy payouts and unfair dismissal protections. own insurance. They can recover GST on expenses. They can form or join a union, but cannot collectively bargain unless the Australian Competition and Consumer Commission The Australian Taxation Office considers Uber drivers to be (ACCC) authorises such action as being in the public interest.118 contractors. Taxi drivers are generally not considered employees. Uber does not regard its ‘driver-partners’ as employees, though Hirers of an independent contractor are not obliged to make courts look beyond what is agreed by the parties.117 superannuation contributions if the contractor is not hired wholly or principally for the labour of the person. Contractors may not be Other factors would support a finding that Uber drivers are covered by government-run workers’ compensation insurance. employees. They cannot delegate the work or negotiate fares, Employers may be obliged to withhold income tax and pay payroll and they are discouraged from rejecting fares. They must use the tax on behalf of some independent contractors. Uber app to transact and have their vehicle approved by Uber. They may at times be paid a minimum hourly rate, are covered by Independent contractors do enjoy some legal and policy Uber’s contingent third-party insurance policy and must follow protections. Work health and safety laws apply to them as they do certain procedures when transporting passengers. to employees. They have the same protections against discrimination as employees do, under the Fair Work Act and anti- Notes: The Fair Work Commission found that a taxi driver was not an employee of the discrimination legislation. Their contracts can be set aside if the vehicle owner for the purposes of unfair dismissal laws (Voros v Dick [2013]). Sources: Foulsham & Geddes (2015); ATO (2015j); Battersby (2015b).

118 Part IV of the Competition and Consumer Act 2010 (Cth). For example, 117 Battersby (2015b) ACCC (2014a) and Stewart and Alderman (2013).

Grattan Institute 2016 37 Peer-to-peer pressure

terms of the contract are considered ‘harsh’ or ‘unfair’, including if evidence to date that such direct undercutting is occurring at large they are paid less than an employee performing similar work.119 scale.

Australia’s social safety net applies equally to independent 4.3.1 The flexibility of platform work suits some workers contractors and employees. Contractors are entitled to universal public health coverage, income support via unemployment On average, workers in ongoing, independent contracting and benefits, family support, disability support, and the age pension. casual roles all report similar levels of job satisfaction (Figure 4.2). Table 4.3 summarises the entitlements of employees and People in casual and independent contracting roles report lower independent contractors. satisfaction with job security, but higher satisfaction with job flexibility. Some people may accept higher risk for potentially 4.3 How workers win and lose on platforms higher rewards and greater flexibility. Still, for some people in insecure work, any positives fail to outweigh the negatives. For Platform contractors lack many of the protections of employees, example, men in the middle of their working lives who work as and may have less predictable income. Some therefore argue that casuals (including labour hire) report lower overall satisfaction and 120 the rise of platforms could make many workers worse off. especially low satisfaction with their job security.121

But platform work does offer benefits to many. Peer-to-peer There is evidence that people who work on platforms value the workers seem to find that their roles suit them. Platforms can flexibility it provides. About 85 per cent of US Uber drivers say actually reduce income variability: some people work temporarily that a reason they work for Uber is ‘to have more flexibility in my on platforms to fill gaps in their income from other sources. And schedule and balance my work with my life and family’, and many others, for whom conventional employment is a poor fit, can find say they prefer it to full-time regular employment.122 Respondents lasting work on platforms. to a survey of US adults cited flexibility as the second most appealing reason for sharing economy providers to work on Still, platforms may undercut firms whose employees benefit platforms.123 An Australian study (sponsored by Airtasker) found today from regulation or collective bargaining. Still, there is little that three-quarters of Australians surveyed believe that the

121 Wooden and Warren (2004); Productivity Commission (2015b), p.100; 119 Independent Contractors Act 2006 (Cth) s 15. Unfair terms in standard form Wilkins and Warren (2012). Low satisfaction remains after controlling for contracts with independent contractors are void providing the contract does not industry, hours, risk preference, tenure, age and education (Buddelmeyer, et al. exceed $300,000 for contracts shorter than 12 months, or $1 million for contracts (2015) longer than 12 months (Pereira and Cunynghame (2015)). 122 Hall and Krueger (2015) 120 Oliver (2015) 123 PwC (2015), p.20

Grattan Institute 2016 38 Peer-to-peer pressure

flexibility offered by working for themselves makes up for the lack 4.3.2 Platforms help some workers manage risk of job security.124 Some people use platform work to earn extra money when their Figure 4.2: Independent contractors, ongoing employees and income from other sources dips. They reduce the volatility of their casuals report similar overall satisfaction income and shorten spells of ‘frictional’ unemployment.125 Average self-reported job satisfaction scores, by employment type, 2014 10 There is evidence this is already happening. A third of active Uber Permanent/ongoing Casual Independent contractor 9 drivers in the US are also looking for a ‘steady, full-time job’. Three-quarters of US Uber drivers say that a reason they drive 8 with Uber is ‘to help maintain a steady income because other sources of income are unstable/unpredictable’.126 US banking 7 data also suggests (though not conclusively) that many workers 6 may use platform work to smooth out variations in income from other sources.127 The banking data study showed that in months 5 where peer-to-peer workers earned money on labour platforms, it 4 comprised about 15 per cent of their income, and their total income was almost identical, on average, to their income in 3 months when they did not work on platforms. 2 4.3.3 Platform work suits some workers 1 Some people on the fringes of the labour market are finding work 0 on platforms. About 2.6 million Australians want to find work or Overall job satisfaction Job security Flexibility work more hours (Figure 4.3). Of these, 1.6 million are un- and Source: HILDA wave 14 under-employed; of this group, about 800,000 say they are

125 The term ‘frictional unemployment’ refers to people in-between jobs actively looking for work. Approximately 1½ per cent of the labour force (about 180,000 people) are unemployed for one month or less, or left their job voluntarily, Ballantyne, et al. (2014). 126 Hall and Krueger (2015), p.11-12 124 Airtasker (2015a), p.9 127 JPMorgan Chase Institute (2016), p.26

Grattan Institute 2016 39 Peer-to-peer pressure

constrained by ‘unsuitable hours or location’, ‘skills/experience’, Figure 4.3: Over two million Australians want work or more hours ‘considered too old’, ‘no vacancies in line of work’, ‘ill health or Millions of people disability’, among other reasons. A further million people would 20 like to work but are not actively looking because they have caring 18 18.6 responsibilities or health issues, are studying, are discouraged, or 16 face age discrimination. There is evidence platforms are already 14 5.3 helping some people on the fringes of the workforce to get 0.4 0.8 0.4 1.0 started, or get back into work: 12 10 10.6 Inexperienced workers can gain experience and demonstrate their value on platforms. In one study, customer reviews on a 8 platform increased the likelihood of future employment for an 6 inexperienced worker by 20 per cent, and boosted future 4 wages. 128 2 0 Three per cent of US Uber drivers were ‘retired’ before starting Employed Unemployed Underemployed Underemployed Marginally No attachment Population work with Uber, two per cent were ‘stay-at-home parents’ and - looking - not looking attached to labour force (15+) seven per cent were students (of which two percentage points were not working).129 Labour force Not in labour force Notes: Marginally attached workers includes people who want to work but are not actively Eight per cent of current US Uber drivers were unemployed looking and people who want to work but cannot start immediately. just prior to joining Uber.130 Many Australian UberX drivers are Sources: ABS (2014b); Grattan Institute analysis from areas of high unemployment: in Brisbane and the Gold 4.3.4 Platforms may put pressure on some workers Coast, half of drivers come from the twenty postcodes with the 131 highest rates of unemployment. Yet platforms could push some people into contracting even if they would prefer to be employees. Today’s labour market regulations (including minimum wages, casual loadings, and 128 Pallais (2013) penalty rates) and collective bargaining can benefit workers but 129 Hall and Krueger (2015), p.10 can also increase costs to their employers. Platform contract work 130 Ibid., p.10; Zipkin (2015). 12 per cent of Australian UberX drivers are over 55 (Uber (2015a)). could displace employment even if all the contractors would prefer 131 Uber (2015a) to be employees and even if the platform work arrangements are

Grattan Institute 2016 40 Peer-to-peer pressure

not otherwise as productive as traditional employment. In other 4.4 What governments should do words, the platforms might circumvent labour regulations. Governments should adjust labour regulations to help realise the But it is not clear how many jobs might be at risk. As Section 4.2.1 efficiency gains and employment growth potential of platforms. notes, many successful platforms host tasks that were already largely done by independent contractors (such as home repairs, They should proceed with caution, though. It is early days. There driving, odd jobs, and graphic design). Many workers who benefit is quite strong evidence that many people have benefited from from minimum wages, casual loadings, or penalty rates work in platform-based work, and little evidence of harm. industries such as retailing, hospitality, or manufacturing, where work appears to be less conducive to peer-to-peer platforms. There are limits to what government can and should do. For example, it is difficult to regulate hours of work or payments per 133 Moreover, if labour regulations contribute to unemployment or to a hour when many services are not defined by the hour. Platform reduction in hours worked, then platforms that circumvent work and workers are diverse, with few common characteristics regulations could benefit current labour market outsiders even as that would support a common set of entitlements across they make some incumbent workers worse off. platforms.

Existing rules do prohibit sham contracting (the misclassification For now, policymakers should focus on mitigating the risks of of genuine employees as independent contractors). Employers platform work. They should take steps to ensure that workers are who misclassify employees as independent contractors are liable well-informed and are not inadvertently exposed to work-related for unpaid tax obligations, superannuation, leave entitlements, risks. This section sets them out. In time, other steps may prove unfair dismissal claims and breaches of award conditions, and warranted. could also be liable for a fine for sham contracting.132 Strengthening and enforcing the sham contracting prohibition may 4.4.1 Provide workplace information to new workers help deter employers from directly replacing employees with platform contractors. But it would not stop a platform with genuine Governments should encourage – and if needed, require – contractors winning market share from a traditional employer. platforms to supply new workers and their hirers with information about their legal, tax and insurance obligations. The information pack would help workers understand the responsibilities and manage any risks associated with working through the platform. It

132 Fair Work Act 2009 (Cth) ss 357-359. The Fair Work Ombudsman and Fair Work Building and Construction can investigate and prosecute instances of 133 Exceptions include work that is continuously controlled through a phone app, sham contracting. such as ride-sharing.

Grattan Institute 2016 41 Peer-to-peer pressure

could be modelled on the Fair Work Information Statement and labour hire firm for the purpose of workers’ compensation other information employers are obliged to provide new insurance coverage. employees. New South Wales considers that outworkers and firefighters are Governments should begin by encouraging work platforms to employees for workers’ compensation purposes.135 The ACT taxi provide such information, but could make it mandatory if there is reforms of late 2015 mandates that a driver of a Transport evidence that people are not well-informed. Booking Service who is not allowed to work for a competitor is to be covered by workers’ compensation insurance. 4.4.2 Expand workers’ compensation schemes to include some platform workers Design of such schemes will require careful work. Platform workers work widely variable hours, face widely variable risks, State governments should bring some platform workers into and may work through multiple platforms, so risks and income workers’ compensation schemes. Platforms whose workers are replacement rates may be difficult to estimate. otherwise likely to be underinsured would be candidates. Jobs could include meal delivery, ride-sharing, and household odd- 4.4.3 Tighten the sham contracting test jobs, where the workers do not come from a background as established independent contractors. Qualifying platforms would The Government should tighten the sham contracting provisions be obliged to pay premiums on behalf of workers (or equivalently, in the Fair Work Act, to deter misclassification of legitimate to deduct premiums from payments to workers). employees as independent contractors. It could follow the recommendation of the Productivity Commission to change the Governments should consider this because workers’ ‘recklessness’ defence within the sham contracting provision of compensation can be administratively less costly than individually the Fair Work Act to a test of ‘reasonableness’.136 That would purchased insurance. It can pool risks efficiently. And it provides discourage firms from improperly reclassifying an employment coverage, where some workers would otherwise be underinsured. relationship. Government could also increase penalties for sham contracting.137 State governments already deem some contractors to be employees for this purpose. In general, workers’ compensation schemes cover contractors who contribute little other than their 135 Workplace Injury Management and Workers Compensation Act 1998 (NSW) 134 136 Fair Work Act s 357; Productivity Commission (2015b), p.813-815 labour and are not ‘in business’ in the usual sense of that term. 137 Most states deem labour hire workers to be employees of the The maximum penalty for each contravention of the sham contracting provisions in the Fair Work Act is $54,000 (Fair Work Ombudsman (2015)). Organisations in breach of this provision may also be subject to other 134 Stewart (2013) undertakings.

Grattan Institute 2016 42 Peer-to-peer pressure

The Commonwealth should enforce other safeguards aimed at and where to participate could be found to be employees, as preventing the exploitation of independent contractors, including discussed in Box 7. unconscionable conduct provisions and unfair contract terms in the Australian Consumer Law and the ‘harsh’ or unfair’ protections It is too early to define a new ‘platform worker’ category. Any such in the Independent Contractors Act. The ACCC and Fair Work worker might operate more like a business than an employee, but 140 should provide guidance to platforms and contractors about these have less autonomy than a contractor. Another dimension provisions. considered could be whether the worker does or can work for other employers.141 But defining such a third category of worker 4.4.4 The employment test and platform workers would not be straightforward, and should only be attempted if it becomes clear that a meaningful group of workers would be Some have proposed changes to the ‘multi-factor test’ that is used covered by it. to determine whether workers are employees or independent contractors. They have proposed redefining some current Another reason to wait is that any new worker category would contractors as employees, or creating a new form of contractor also have to define a bundle of rights and responsibilities that that would sit between an employee and an independent apply to a sufficiently large and uniform class of workers. At this contractor.138 Such ‘intermediate’, ‘platform’ or ‘independent’ stage, is hard to specify such a bundle of benefits; platform roles workers would receive some of the benefits or protections are diverse and the sector is developing fast. Any such change received by employees. will have to wait until the shape of platform work becomes more evident. There is no need to change the multi-factor test. It can be hard to apply, and there is uncertainty about the true classification of a worker until a case is heard before a court.139 But the test is flexible enough to allow courts to consider new factors relevant to a platform/worker relationship, such as the extent of control over the work that platforms exercise. For example, people who work on a platform that sets prices, defines important work practices, and limits the autonomy of the worker to merely choosing when

140 Italy has an intermediate worker category that is determined by the extent that the principal can coordinate the activities of the worker (Casale (2011)). 138 Proposals in the US context include those of Hagiu (2015) and Harris and 141 Canada, Germany and Spain have ‘dependent contractor’ categories that Krueger (2015). focus on ‘exclusivity’ (Hagiu (2015); Schein (2014); Lee (2012); Weiss and 139 Productivity Commission (2015b), p.813; Stewart and Alderman (2013) Schmidt (2008), p.47-48).

Grattan Institute 2016 43 Peer-to-peer pressure

4.4.5 Possible further work reforms

In time, peer-to-peer platform work may become more established. It will become clearer whether some platform workers need extra protections or benefits that current laws do not provide. Reform options that may prove appropriate include the following:

The Commonwealth could amend competition and labour laws to allow some platform contractors to collectively bargain without ACCC authorisation.

Peer-to-peer platforms could be required to have an appeals process for a worker who believes he or she was unfairly removed from a platform. The government could implement this requirement following large numbers of complaints about unfair dismissal from platforms.

Governments could require platforms to provide peer-to-peer workers access to their ‘rating’ when they leave a platform. This may become necessary if the importance of ratings for future employability increases and if platforms do not make it easy for workers to export their ratings data.

Provide an opt-out or opt-in system for part of a worker’s platform earnings to be allocated to a package that deducts tax and superannuation.

Ultimately some of these options might fit some platforms. If many platform workforces have similar characteristics, the options could be packaged into a ‘platform worker’ category in labour law.

Grattan Institute 2016 44 Peer-to-peer pressure

Table 4.3: How employment status is decided for different conditions and benefits Condition Example Test Employee entitlements Independent contractor entitlements Work Minimum wage, leave, Multi-factor test All non-casual employment contracts must meet Independent contractors are not subject to benefits, maximum weekly conditions specified by the National Employment prescribed employment conditions. However, a security hours, notice period, Standards and the national minimum wage (casuals are contract for services can be set aside if ‘harsh’ or and redundancy, unfair entitled to a more limited range of benefits1), including: ‘unfair’. conditions dismissal, collective Minimum wage and maximum weekly hours bargaining Leave: parental, annual, carers, compassionate, long service; public holidays; notice and redundancy pay Fair Work Information Statement. Superannuation Commonwealth Employers must contribute superannuation for all Principals are not obliged to make superannuation legislation employees and contractors paid ‘wholly or principally for contributions for contractors if not paid ‘wholly or their labour’ aged over 18 who earn over $450 / month. principally for their labour’. Risk and Workers’ State and ‘Workers’ are covered by government-run workers’ If not covered by the definition of a ‘worker’, njury compensation Commonwealth compensation schemes. Some legislation deems some contractors are not covered by government-run legislation contractors under the multi-factor test to be ‘workers’. workers’ compensation schemes. Vicarious liability Multi-factor test An employer is vicariously liable for any wrongful acts of Employers are not vicariously liable for the actions an employee acting in the course of their employment. of an independent contractor. Work health and State and Organisations have a primary duty of care to ensure the health and safety of ‘workers’, a definition which safety Commonwealth includes contractors and employees, while they are engaged at work. Workers also have statutory obligations legislation under work health and safety legislation, such as taking reasonable care for their own health and safety. Taxation Income tax Multi-factor test Employers required to withhold income tax on behalf of Contractors liable to pay their own income tax. and employees (multi-factor test). Employees must pay the Contractors that pass the personal services Commonwealth personal income tax rate and cannot claim business- business test (Commonwealth legislation) can legislation related tax deductions. claim business tax deductions and pay the company tax rate (if incorporated). Payroll tax State legislation Employers liable to pay payroll tax for all employees. Principals are liable to pay payroll tax for most contractors that supply services. GST Multi-factor test Employees do not have GST obligations. Contractors must meet GST obligations. Social Unemployment Australia’s social safety net does not distinguish between employees and independent contractors safety net insurance, health care Notes: 1Casuals have a right to request unpaid parental leave after being employed for 12 months on a regular and systematic basis. Casuals are entitled to two days’ unpaid carers leave and unpaid days off on a public holiday. Casuals are protected from unfair dismissal if they have an ongoing employment contract that the employer has terminated. Sources: Grattan Institute analysis; ATO (2005); ATO (2014); ATO (2015e); ATO (2015i); Fair Work Ombudsman (2015); Stewart and Alderman (2013); Productivity Commission (2015b)

Grattan Institute 2016 45 Peer-to-peer pressure

5 Competition and consumer regulation

The peer-to-peer economy is subject to competition laws, which Box 8: Prohibition on the misuse of market power are intended to help Australians enjoy the benefits of competition: low prices, high quality, and wide choice.142 Section 46 of the current Competition and Consumer Act seeks to stop firms from misusing market power for an anti-competitive Peer-to-peer platforms may attract the attention of competition purpose. In its current version, section 46 prohibits an regulators in two main ways.143 First, incumbent firms, worried organisation with substantial market power taking advantage of it about maintaining their profits, may try to limit competition from for the purpose of eliminating or substantially damaging a platforms. Regulators will need to consider the range of tactics competitor; preventing entry of a competitor; or deterring or incumbents might use and be ready to react to rapidly changing preventing competitive conduct. market boundaries. The 2015 Competition Policy Review (Harper Review) Second, peer-to-peer platforms themselves may acquire market recommended changes to section 46 to put a greater focus on the power. Today’s competition and consumer laws appear adequate effect of conduct on competition. It proposed a revised section to prevent abuses, but in applying the laws regulators need to be that would prohibit any firm with substantial market power from aware of how platforms operate and compete. conduct that has the purpose, effect or likely effect of substantially lessening competition. Under the proposal, courts would have to 5.1.1 Existing competition laws will apply to anti- consider both the pro- and anti-competitive effects of a firm’s competitive conduct by incumbents and platforms conduct. Courts may be able to consider consumer benefits of a firm’s behaviour even if it strengthens the market position of the Section 46 of the Competition and Consumer Act (CCA) prohibits 144 firm. The Australian Government recently announced it plans to firms from misusing market power (Box 8). change section 46, as the Review recommended.

Notes: The Harper Review also recommended that the ACCC be empowered to authorise 142 The purpose of competition law is to ‘improve the economic welfare of conduct that could otherwise be found to be in breach of section 46. Authorisation would Australians by stopping conduct that would otherwise substantially lessen likely be on the basis of a net public benefit test, which would also permit consumer competition’ Sims (2015). Platforms must comply with industry-specific benefits to be considered where they may offset a reduction in competition. regulation as well (as in the case of ride-sharing and taxi regulation – see Sources: Harper et al. (2015); Australian Government (2016). Chapter 2). 143 The competition implications of collective bargaining by independent contractors are considered in Chapter 4. 144 Competition and Consumer Act 2010 (Cth) Part IV

Grattan Institute 2016 46 Peer-to-peer pressure

Firms are specifically prohibited from ‘predatory pricing’ (pricing 5.2 Network effects are critical in applying competition law below cost to deter or damage competitors).145 Section 47 to platforms and their competitors prohibits firms from ‘exclusive dealing’ (refusing to supply, or attempting to prevent someone else from supplying a In applying competition laws to the peer-to-peer economy, customer).146 All of these might be attempted by a platform or its regulators and courts must understand how ‘network effects’ competitors in seeking to gain advantage. influence the behaviour of platforms and their competitors, and how the interests of consumers may be promoted or harmed. A Competition law also prohibits anti-competitive mergers, platform exhibits network effects if the cost per user falls, or value arrangements and cartel conduct.147 Under section 50 of the Act, per user rises, as the number of users grows. the ACCC is able to stop mergers if they are likely to have the effect of substantially lessening competition in any market.148 The The ACCC and the courts will need to consider the pro- and anti- Australian Competition Tribunal can authorise a merger on the consumer aspects of network effects in applying competition law. grounds that it provides a net public benefit.149 The ACCC can ACCC and courts already have plenty of experience dealing with also authorise proposed joint-venture agreements if the public network effects in different industries, including cases in credit 150 benefit outweighs any public detriment. cards, energy and telecommunications networks.

Misuse of market power: in applying the misuse of market 145 Predatory pricing is prohibited by the general misuse of market power power test, regulators should observe that platforms compete provision and by sections 46(1AAA) and (1AA) of the CCA. The Harper review aggressively to build and then profit from network effects. For recommended the predatory pricing subsections be repealed if the new effects example, a platform may price aggressively to build scale, test is used. 146 Exclusive dealing may include a platform or firm using its market power to benefiting consumers in the short run but possibly reducing induce suppliers or customers to use or buy integrated services, limiting competition and harming consumers in the longer run. If a competition, Deloitte Access Economics (2015b). The Harper Review platform does attract a large user base, it may seek to prevent recommended repealing section 47; the Government has yet to make a decision users from operating on competing platforms. The current on this recommendation, Treasury (2015a). prohibition on the misuse of market power (Box 8) deals with 147 Platforms are not allowed to make or give effect to arrangements that contain cartel or exclusionary provisions (CCA s 4D and cartel prohibitions) and are also these opposing forces by considering the purpose of any prohibited from making or giving effect to arrangements that have the purpose, behaviour and its effect on competitors. It remains to be seen how effect, or likely effect of substantially lessening competition (CCA s 45). this situation would be assessed under the proposed new version 148 Similarly, firms are not allowed to collude or make ‘anti-competitive of this prohibition. agreements’ with other firms if this has the purpose, or has or is likely to have the effect of substantially lessening competition (section 45). 149 CCA ss 95AT and 95AZH. The Australian Competition Tribunal hears merger 150 authorisation cases and reviews of determinations of the ACCC. ACCC (2013c)

Grattan Institute 2016 47 Peer-to-peer pressure

Exclusive dealing: platforms would be prohibited from preventing Box 9: ACCC authorises the ‘ihail’ taxi booking app suppliers using a competing platform if to do would be likely to substantially lessen competition.151 Exclusive dealing provisions The ACCC recently granted conditional authorisation for a joint may also prohibit platforms or other firms from requiring their venture agreement between multiple taxi companies and suppliers to use additional services (such as a payment system). Cabcharge to create ‘ihail’, a taxi booking and payments platform. The ACCC has considered closely related issues in cases involving supermarkets, mining, petrol stations and The draft ACCC determination had rejected the proponents’ initial telecommunications.152 For example, Visa was fined for ihail proposal. The proponents than made a number of changes. preventing merchants using its payment platform from offering They allowed passengers to select their preferred taxi network on their customers competing payment platforms.153 the app and to pay cash rather than through the app if they wished. They allowed drivers to ‘opt-in’ to the booking app rather Mergers: the ACCC and the courts are also familiar with than opting-out, and allowed taxi networks that are not members evaluating network effects when assessing the effects on of the joint venture to sign up to the app. competition from proposed mergers, joint ventures and other agreements.154 For example, the ACCC rejected a proposed A factor in the ACCC decision may have been that the legalisation acquisition of Trading Post by Carsales on the basis that it was of ride-sharing in NSW and elsewhere provided additional likely to result in a substantial lessening of competition in competition for ihail. In its draft determination, the ACCC had automotive classified advertising.155 Similarly, the recent ACCC noted that competitive pressure from ride sharing was not authorisation of a joint venture between incumbent taxi operators assured. shows that it considers network effects when analysing how platforms compete (Box 9). Sources: ACCC (2016); ACCC (2015b); ACCC (2015a)

More generally, access by competitors to peer-to-peer platforms infrastructure has long been an important part of the Competition may become an important issue in competition policy. Access to and Consumer Act (Part IIIA, and the industry-specific regimes for telecommunications and energy). In deciding to authorise the joint venture to create and operate taxi booking app ihail, the ACCC 151 Exclusive dealing may also fall under the general provision prohibiting misuse appears to have considered access: the joint venture parties of market power, CCA s 46. agreed to allow taxis from competing taxi networks to operate on 152 ACCC v Metcash; Fortescue Metals Group; Australian Gas Light Company their platform. In time, it is conceivable that regulators may need (AGL) v ACCC; ACCC (2013a); ACCC (2003). to develop an approach to mandating access by competitors to 153 ACCC (2015d) 154 ACCC (2012) major peer-to-peer platforms. 155 Ibid.

Grattan Institute 2016 48 Peer-to-peer pressure

5.3 Consumer protection and peer-to-peer platforms despite limited knowledge of their operations, by providing a summary of previous buyers’ assessments of the seller.158 The emergence of the peer-to-peer economy has led to concerns about consumer safety. Some consumers are concerned that that Ratings systems are, however, imperfect. Users can post false or service quality may be low, that safety may be compromised or misleading reviews, or unconscious bias can influence reviews.159 that it will be difficult to know who is responsible if something goes Reviews on platforms can seem overly positive.160 Some wrong. The Australian Consumer Law (ACL) applies to both customers review positively in the hope of a reciprocal positive platforms and suppliers who are in business. Industry-specific review.161 Customers who had positive experiences are more regulation also applies in some cases.156 likely to leave a review. But despite these imperfections, the rapid growth of transactions on platforms suggests that ratings systems 5.3.1 Platform reviews encourage good service work tolerably well.162 Whether they succeed or fail, consumer law applies. If there is a single innovation that is most responsible for the success of the peer-to-peer platform model, it is the peer review 5.3.2 Services supplied on platforms are protected by system. Suppliers know that they will be reviewed and that the consumer law review will be visible to potential customers. That provides a strong incentive for suppliers to be trustworthy and reliable. The ACL covers consumer protection and fair trading for all Platforms, in turn, have strong incentives to design review consumers and businesses operating in Australia.163 The ACL systems that work well to control peer misbehaviour such as applies in principle to both ‘peers’ and to platforms. Most sections misleading advertising, failure to deliver, or fraud.157 The rating in the ACL relevant to the peer-to-peer economy require a person and review systems offered by most platforms help screen out or business to be in ‘trade or commerce’.164 Some small, part-time bad actors, reward good service, and provide detailed feedback peer-to-peer suppliers may therefore not fall under the ACL if they on what consumers value. Review systems help to radically do not meet this threshold. expand the range of possible trading partners: eBay and Amazon ratings systems enable buyers to trust sellers across the world, 158 Thierer, et al. (2015) 159 Aral (2013) 160 Zervas, et al. (2015) 156 Deloitte Access Economics (2015b) 161 Fradkin, et al. (2015) 157 For example, Airbnb changed its review system in 2014 so that users publish 162 Ibid. reviews only after the ‘host’ and the guest have both written their review (or after 163 Schedule 2 of the Competition and Consumer Act 2010 (Cth). 14 days, when the review period ends). Before this change, reviews were 164 The term, defined in section 2 of the ACL as, ‘trade or commerce… includes published immediately. Airbnb made this change to counteract retaliatory rating, any business or professional activity (whether or not carried on for profit)’ has which may have inflated reviews (Airbnb (2014a)). been given a wide definition by the courts (Nguyen and Oliver (2013)).

Grattan Institute 2016 49 Peer-to-peer pressure

Platforms that are more involved in controlling how a service is reviews, disclose commercial relationships, and not edit or omit advertised and delivered will typically be found to have greater reviews if to do so would be misleading. Review sites are not responsibility for it. For example, if a platform implies that their obliged, however, to substantiate or verify legitimate reviews by background checks and feedback make using their service safe, actual consumers. Because peer-to-peer platforms also host they are likely to be found responsible for ensuring that those reviews, they are likely to be subject to the same obligations as checks are effective. online review sites. Further judicial decisions may be needed before a clear set of responsibilities can be set out.168 Elements of the ACL that are relevant to the peer-to-peer economy include misleading and deceptive conduct, unfair terms Depending on the circumstances, a platform may not be liable for in contracts, service quality, unconscionable conduct and misleading information posted by users. The ACL provides for a compliance with safety standards (Table 5.1). If suppliers breach ‘publisher’s defence’ to a claim of misleading and deceptive conditions set out under the ACL, they may be subject to fines conduct. It protects information providers from being accountable and liable to refund the customer or pay damages. for representations if they were not aware that the representation was misleading.169 In Google v ACCC, the High Court held that First, ‘conduct that is misleading or deceptive’ is prohibited by Google did not engage in misleading or deceptive conduct when it section 18 of the ACL. For example, a peer-to-peer displayed ‘sponsored links’ that were considered misleading or accommodation host who states in an advertisement that the deceptive, as it did not author the links or endorse the misleading beach is a five-minute walk, when it is actually thirty minutes, representations.170 would be in breach of section 18, and can be liable for fines or damages for engaging in misleading or deceptive conduct.165

Platforms may be found responsible for claims made by both suppliers, and by customers in posting reviews. 166 For example, the ACCC’s 2013 guidelines for online review websites seek to 167 ensure that consumers are not misled by reviews. The guidelines state that websites are required to remove fake 168 A platform might even be sued by a supplier for defamation if an untrue review is posted and the platform does not remove it when made aware of it. 169 CCA s 19 165 Dynamic pricing, a feature of some platforms, may be subject to claims of 170 Google v ACCC (2013) 87 ALJR 235; King (2013). This contrasts with the misleading or deceptive conduct if the reason given for a rise in price is not decision of the Court of Justice of the European Union in the ‘Right to be justified or if it is implemented in a way that misleads customers. Forgotten’ case (Google Spain v González), where the court held that Google 166 Deloitte Access Economics (2015b) must remove links to personal information even when it is not the publisher of 167 ACCC (2013d) the content.

Grattan Institute 2016 50 Peer-to-peer pressure

Table 5.1: Provisions in the ACL that apply to the peer-to-peer Other ACL sections are also relevant to the peer-to-peer economy suppliers and some platforms economy, as set out in Table 5.1: Applicable to Provision Description platforms? Suppliers must not engage in ‘unconscionable conduct’ when dealing with customers; and platforms must not do so with Misleading and Platforms and suppliers must Yes; publishers’ suppliers.171 Unconscionable conduct is not precisely defined, deceptive not mislead or deceive defence may apply conduct s18 customers. but relevant factors include the relative bargaining strength of the parties, whether the behaviour was harsh or oppressive Unconscionable Platforms and suppliers must and whether the parties acted in good faith.172 conduct not engage in ‘unconscionable Yes ss20-22 conduct’. A supplier may be fined if they fail to supply a good or service A supplier cannot include ‘unfair that a consumer has paid for. A platform may also be liable if it Unfair contract contract terms’ in standard form terms contracts; if a term is found to Yes is involved in the supply of the good or service, although the ss23-28 be unfair it will be considered extent of this liability is unclear. void by a court. A supplier is liable for a Sections 23-28 of the ACL prohibits unfair contract terms, and Failure to supply pecuniary penalty if they fail to applies even to ‘standard form contracts’ that might be used s36 supply a good or service that a A platform may be by global peer-to-peer platforms. consumer has paid for. liable if it is Due care and skill Someone supplying a service involved in the Section 60 of the ACL requires that someone supplying a s60 must use ‘due care and skill’. supply of the service use ‘due care and skill’. Suppliers, and platforms that good or service. Product safety Suppliers of goods and are construed as supplying a service, must meet this ‘product-related services’ must standards standard. comply with relevant product ss106-107 safety standards. Sources: Competition and Consumer Act 2010 (Cth); ACCC (2015c); Beaton-Wells (2015); ACCC (2013b)

171 ACL ss 20-22. For example, Coles was ordered to pay $10 million in penalties and costs plus provide refunds for engaging in unconscionable conduct with its suppliers (ACCC (2014b)). 172 ACCC (2015c); Beaton-Wells (2015)

Grattan Institute 2016 51 Peer-to-peer pressure

Suppliers of goods and ‘product-related services’ must comply platforms; informing platform users of their legal rights and with relevant product safety standards.173 Some peer-to-peer responsibilities; and clearly stating how use of the site may affect platform suppliers may supply a good or service that must the validity of other contracts (such as insurance policies). comply with mandatory safety standards set by the ACCC. Platforms should assist their workers to comply with these 5.4.2 Competition policy laws. A platform may be liable under safety provisions if a platform worker did not meet safety standards if the platform The ACCC should seek to apply competition laws in light of the was involved in the supply of the service. network effects that operate on peer-to-peer platforms. It could also provide guidelines on how competition laws apply to peer-to- The platform or the ‘peer’ provider or both could be found to be a peer platforms. ‘supplier’, depending on what aspects of the supply they control. The Commonwealth should encourage the states to undertake 5.4 What policymakers should do reforms such as in taxi regulation (as discussed in Chapter 2). It should also ensure resourcing of the ACCC is adequate to enable Competition and consumer policy for the peer-to-peer economy it to investigate and prosecute anti-competitive behaviour. can improve consumer satisfaction and productivity. Governments should take the following steps. 5.4.3 Consumer policy

5.4.1 A peer-to-peer platform code of conduct The ACCC should develop guidelines for platforms – similar to the guidelines for online review websites – about their responsibilities Creation of a voluntary industry code of conduct for peer-to-peer in publishing and maintaining ratings systems; including guidance platforms could help platforms and suppliers to comply with in how the publisher’s defence may apply to platforms. It should competition and consumer laws, and reduce their compliance also encourage platforms to inform suppliers (including small costs. The ACCC can guide industries in developing voluntary ‘household’ peer providers) about their obligations under the ACL. codes of conduct. The Treasury has been charged with a review into consumer law The code of conduct should include: guidelines for dealing with and policy. The review should consider the scope of the disputes, including disputed reviews; privacy and handling data; publishers’ defence for platforms; the appropriateness of the protecting the interests of users who wish to use multiple current trade or commerce test for peer-to-peer providers; and should develop principles on how consumer law responsibility

173 should be allocated between peers or platforms. Product-related services include the installation, maintenance, repair, cleaning, assembly or deliver of consumer goods (ACL s 106-107).

Grattan Institute 2016 52 Peer-to-peer pressure

6 Tax and the peer-to-peer economy

Australian tax laws cover peer-to-peer economy participants as economy in mind, but they apply to it. Personal income tax and they cover other individuals and companies operating in Australia. capital gains tax laws are set appropriately for this new economy, Peer-to-peer suppliers’ income is subject to income tax, and but GST and company tax laws may need to be adjusted. personal tax avoidance is difficult. Additional economic activity on peer-to-peer platforms may boost tax revenues. 6.1.1 Suppliers are liable for income and capital gains tax on their peer-to-peer activities Yet under current laws, the rise of peer-to-peer economy might cut the GST and company tax the government collects as a share Most income of suppliers in the peer-to-peer economy is taxable of total income. That is because, first, platforms will help some income, and avoidance will be difficult as most transactions are small suppliers, many of whom are likely to fall below the GST electronic. registration threshold of $75,000, to displace larger firms. In the case of accommodation, short-stay suppliers using platforms are Income earned by platform providers is assessable income for tax purposes if the Australian Tax Office (ATO) considers the activity exempt from charging GST, while the hotels (and some serviced 174 apartments) with whom they compete are not. As the peer-to-peer to be a business and not a hobby. The ATO looks at whether economy grows, these revenue losses could become significant. there is a registered business name, an intent to make a profit and whether the activity is planned, organised and carried out in a 175 Secondly, the government faces a tax challenge as global business-like manner. If an activity is neither a hobby nor a electronic services become more important to the economy. While business, the person undertaking it is defined as an employee, overseas-based peer-to-peer platforms are just one of many such and is subject to taxation rules on employment income (the operators, governments have work to do to ensure they pay an employer withholds PAYG and applicable deductions, for appropriate amount of Australian company tax. Global example). agreements and unilateral action have begun to address the problem of base erosion and profit-shifting (BEPS) but they have The ATO is likely to consider most peer-to-peer economy a long way to go. providers as carrying on a business. It has stated that all income earned from renting out all or part of a home (through a platform 6.1 Peer-to-peer is subject to tax, but the tax share of income may fall 174 ATO (2015n) 175 ATO (2015a). A hobby is something done for pleasure or recreation. Australia’s tax laws were not designed with the peer-to-peer Business income may be subject to rules relating to ‘personal services income’ ATO (2015i).

Grattan Institute 2016 53 Peer-to-peer pressure

or not) is assessable income, as is income earned driving for a cent shift in GST applicable activity would reduce GST revenue by ride-sharing platform.176 ~$0.5 billion.

The peer-to-peer economy raises no specific capital gains tax Second, providers of short-stay accommodation through peer-to- issues. Short-stay rentals, like long-term rentals, can reduce the peer platforms do not have to charge GST. The ATO has advised capital gains tax exemption for the main residence.177 If a person that people renting out a room or their house through an rents out their entire primary residence, they lose the capital gains accommodation platform are charging residential rents and so are tax exemption for that period, unless they meet certain tests.178 If not required to charge GST (Table 6.1).181 Hotels and some a person living in their primary residence rents out a room, the serviced apartments, by contrast, must charge GST. If short-stay portion of a residence that is rented and the time it is rented for is accommodation providers take market share from them,182 GST subject to capital gains tax. revenue will be lower than it would have been otherwise.183

6.1.2 As a share of income, GST revenue may fall when Point-to-point transport, however, is an exception. Suppliers of peer-to-peer output grows taxi, limousine or ride-sharing services must register for GST even if their turnover is lower than $75,000.184 Peer-to-peer platforms fragment traditional business models, resulting in many small suppliers. Businesses with an annual revenue greater than $75,000 must register for GST.179 Many 181 ATO (2015m). The ATO may also seek to collect GST on service charges by peer-to-peer suppliers will fall under the threshold, so less GST platforms domiciled out of Australia. Payments for most imports of services (see will be collected than if the same activities were done by large Box 6.1) do not incur GST. If the local platform seller does not charge GST on 180 the entire consumer charge, the service fee of the platform would therefore not firms. GST revenue in 2014/15 was about $55 billion, so a 1 per include GST if it is treated as a ‘service import’ (Heinemann and Shume (2015)). 182 There is some evidence that hotels are losing market share to Airbnb. Zervas, 176 ATO (2015m) et al. (2014). 183 177 Ibid. Hotel revenue was $6 billion in (IBISWorld (2016)). Assuming hotel gross 178 If someone has a legitimate reason to move from their main residence (such value added is half of revenue, net GST raised from the sector would be about as for work), they can continue to receive the main residence CGT exemption $300 million. 184 even if they rent their main residence out, as long as they live in it for at least six Division 144 of the A New Tax System (Goods and Services Tax) Act 1999 months every six years (the temporary absence rule). Subdivision 118-B of (Cth). Taxi travel is defined as ‘travel that involves transporting passengers, by Income Tax Assessment Act 1997 (Cth); ATO (2015b). taxi or limousine, for fares’. The government implemented the exception to the 179 Exceptions include if the firm only sells ‘input-taxed’ goods and services (such $75,000 minimum turnover rule to avoid confusion that would have resulted if as financial services or residential rents) or if the business provides taxi, some taxi drivers had to charge GST and others did not; to enable business limousine or ride-sharing services ATO (2015k)). 18 per cent of businesses are users to claim an input tax credit on all taxi trips; and to enable consistency under the $75,000 threshold (Productivity Commission (2015a), p.188). across meters. Another policy motivation may have been to reduce tax evasion 180 Australian Government (2015) through cash fares (Rothengatter (2008), ATO (2015l)).

Grattan Institute 2016 54 Peer-to-peer pressure

Table 6.1: Who pays GST in the peer-to-peer economy? Box 10: The ‘Netflix tax’, a model for requiring peer-to-peer Service Liable for GST? Threshold? platforms to collect and remit GST?

Passenger transport Yes for taxis and ride- Government should consider requiring some peer-to-peer sharing $0 platforms to charge and remit GST. A possible model is the No for car pooling ‘Netflix Tax’, introduced in the 2015/16 Budget. The 'Netflix tax' is Accommodation Yes for hotels and some an amendment to the A New Tax System (Goods and Services serviced apartments No for residential rentals $75,000 Tax) Act 1999 (Cth). It expands the GST base to include including short-stay rentals purchases of digital goods and services from suppliers outside Local tasks Yes $75,000 Australia by Australian consumers. The operator of an ‘electronic distribution service’ must collect and remit GST when it controls Imported tasks (e.g. No N/A online designs) key elements of the supply such as delivery or payments. Imported platform No N/A The ‘Netflix Tax’ model could be extended to peer-to-peer services platforms operating in Australia. Government could require them Source: ATO (2015e) to collect and remit GST for the transactions they facilitate, Therefore, GST revenue will not change much if ride-sharing maintaining GST coverage as the share of small providers grows. operators take market share from taxis and hire-cars. Uber is Requiring platforms to collect and remit GST would remove the challenging in the Federal Court the ATO’s ruling that all UberX need for individual suppliers to do so (as Uber drivers must do at drivers must register for GST, but in the interim it has advised its 185 present). Yet suppliers would still need to record expenses to drivers to remit GST to the ATO. claim GST input tax credits. An alternative way to maintain GST revenues would be to lower the turnover threshold for GST registration. Yet this may impose higher administration costs on suppliers than the Netflix model would, as suppliers would have to remit GST individually.

Source: Treasury (2015b)

185 Battersby (2015a); ATO (2015d). If the Federal Court overturns the ATO’s determination on the definition of taxi travel, government could consider passing legislation defining taxi travel to include ride-sharing.

Grattan Institute 2016 55 Peer-to-peer pressure

6.1.3 Base erosion and profit shifting by multinational low-tax jurisdiction charge Australian customers. They then pay platforms offsets the benefits of peer-to-peer their Australian operations such a service fee.189 platforms If multinational-owned platforms pursue aggressive tax Ensuring that overseas-based companies, including peer-to-peer minimisation strategies while they displace Australian-based platforms that operate in Australia, pay an appropriate amount of companies, the loss in tax revenue could partly or even wholly company tax is a major challenge. It requires global cooperation outweigh the local productivity and consumer gains created by and Australian government action. The government is their services. For example, the operators of taxi networks pay implementing some recommendations made by the BEPS Project company tax. If ride-sharing providers capture market share at the of the OECD and the G20, and is making some domestic changes expense of taxis, and use BEPS strategies to pay minimal to tax laws. Yet further action is required to ensure tax revenue company tax, then domestic tax revenue could fall. losses do not offset the benefits from the peer-to-peer economy. Only multi-national action can fully address the BEPS problem. Multinational companies may use a variety of legal tax planning The OECD and G20 final report, released in October 2015, strategies to minimise company tax, resulting in what has come to outlines many strategies to combat BEPS. They include country- be called ‘base erosion and profit shifting’ (BEPS). The OECD by-country reporting of multinational enterprises, improved defines BEPS as ‘tax planning strategies that exploit … gaps and information exchanges among jurisdictions and new policies mismatches in tax rules to artificially shift profits to low or no-tax relating to the treatment of intellectual property.190 locations where there is little or no economic activity, resulting in little or no overall corporate tax being paid’.186 The Commonwealth Government has implemented the country- by-country reporting requirements proposed in the OECD/G20 BEPS covers a wide range of strategies to reduce tax.187 One is to plan.191 It has also made some unilateral changes in an attempt to send customers’ payments directly to an overseas parent located limit BEPS.192 In December 2015, it passed legislation in a low-tax jurisdiction.188 The parent company then pays the strengthening the general anti-avoidance rules to cover local subsidiary a service fee (for activities such as marketing) that just exceeds its costs, leaving the subsidiary with negligible taxable income. Overseas parents of Airbnb and Uber located in a 189 Ting (2015) 190 OECD (2015b) 191 The government signed a multinational agreement to share country-by- country reports with 30 other countries (ATO (2016a)). 186 OECD (2016) 192 Unilateral action may risk undermining global efforts to combat BEPS or 187 OECD (2015a) result in double taxation (Palmer (2015)). The OECD (2016) states that BEPS is 188 Khadem (2015) ‘a global problem which requires global solutions’.

Grattan Institute 2016 56 Peer-to-peer pressure

multinationals that artificially avoid having a taxable presence in Australia.193 The ATO is also scrutinising overseas marketing and procurement hubs.194

6.2 What government should do

The Commonwealth Government should continue to pursue a comprehensive global agreement on BEPS to ensure that firms operating in Australia pay a similar proportion of company tax to all other companies. The government should consider further unilateral policy action where appropriate.

The ATO should also require information from platforms where it would reduce the administrative burden for participants, and minimise tax avoidance. The ATO should also continue to provide guidance for peer-to-peer service providers to improve compliance.

If GST revenues are being eroded by growth of small peer-to-peer suppliers at the expense of large firms the ATO could require peer-to-peer platforms to collect and remit GST on behalf of suppliers (see Box 10). The government will learn lessons from the implementation of the Netflix tax.

193 Informally known as a ‘diverted profits tax’ or ‘Google tax’. Drysdale (2015); ATO (2015c); KMPG (2015); ATO (2016b). 194 ATO (2015g); ATO (2015h)

Grattan Institute 2016 57 Peer-to-peer pressure

7 Conclusion

Peer-to-peer platforms are building better markets. By cutting the yet, but some other reforms are needed. Platforms should be costs of searching and transacting with strangers, they help make encouraged to inform new workers about their responsibilities as a more trusting and productive economy. independent contractors. Selected platforms should be brought into workers’ compensation schemes. Sham contracting This report recommends a first round of reforms that will help provisions should be tightened to deter improper classification of position Australia to get the most from the peer-to-peer economy. workers.

First, governments should legalise ride-sharing. The NSW and Fourth, while the government did not develop competition laws ACT reforms provide good models for other governments to with the peer-to-peer economy in mind, existing laws are flexible follow. Consumers will benefit from lower prices and waiting times enough to capture anti-competitive conduct by platforms or and better service as ride-sharing services and taxis compete for incumbents. The main challenge for the ACCC is that platforms customers. Drivers will benefit from access to flexible work. Partial are likely to exhibit network effects that can benefit users but can compensation to taxi licence holders, if offered, should be limited result in strong market power for successful platforms. to people who bought taxi licences recently, and/or suffer severe financial hardship because of these changes. Similarly, consumer laws do not need major changes to function well in the peer-to-peer economy. Rating and review systems, in Second, peer-to-peer accommodation offers benefits to both the context of the broader protections offered by consumer law, hosts and guests. But it can cause disruptions to neighbours, and will encourage high-quality service and good behaviour by peers. so complaints need to be able to be managed quickly. The main challenge for consumer law is to define the Governments should make it easy for owners’ corporations and responsibility of platforms when their role is partly that of a local councils to hold owners accountable when their short-stay producer and partly an intermediary. tenants disrupt neighbours. Local governments should focus on protecting amenity, not banning all short-stay rentals when many Finally, most income earned by suppliers will be assessable do not cause problems. income for income tax purposes. But the tax take on economic activity will tend to fall as the peer-to-peer economy grows, Third, peer-to-peer platforms will mostly improve an already because small providers and multinational firms tend to pay less diverse and flexible – but imperfect – labour market. Platforms tax. The government should continue to address base erosion already provide work and income for many thousands of people. and profit shifting. Creating a ‘platform contractor’ category of worker is not required

Grattan Institute 2016 58 Peer-to-peer pressure

8 References

Abelson, P. (2010) 'The High Cost of Taxi Regulation, with Special Reference to http://registers.accc.gov.au/content/item.phtml?itemId=1117620&nodeI Sydney', Agenda: A Journal of Policy Analysis and Reform, p 41-70 d=b6ca5277d231313d2750fba9a5e401e7&fn=Carsales.com%20Limite d%20proposed%20acquisition%20of%20assets%20associated%20with ABS (2011) Census of Population and Housing, Australian Bureau of Statistics %20Trading%20Post%20from%20Telstra%20Corporation%20Limited% ABS (2013) Forms of Employment, catalogue number 6359.0, Australian 20-%20Advertising%20-%2013%20June%202013.pdf Bureau of Statistics ACCC (2013d) What you need to know about: online reviews - a guide for ABS (2014a) Labour Market Statistics, catalogue number 6105.0, Australian business and review platforms, Australian Competition and Consumer Bureau of Statistics Commission, accessed 15 March 2016, from https://www.accc.gov.au/system/files/Online%20reviews%E2%80%94a ABS (2014b) Persons Not in the Labour Force, Underemployed Workers and %20guide%20for%20business%20and%20review%20platforms.pdf Job Search Experience, catalogue number 6226.0.55.001, Australian Bureau of Statistics ACCC (2014a) 'ACCC proposes to grant authorisation to APESMA to collectively bargain on behalf of its translator and interpreter members', accessed 1 ABS (2015) Characteristics of Employment, catalogue number 6333.0, April 2016, from https://www.accc.gov.au/media-release/accc- Australian Bureau of Statistics proposes-to-grant-authorisation-to-apesma-to-collectively-bargain-on- behalf-of-its-translator-and-interpreter-members ACCC (2003) 'ACCC issues final information paper on bundling telco services', accessed 18 November 2015, from https://www.accc.gov.au/media- ACCC (2014b) 'Court finds Coles engaged in unconscionable conduct and release/accc-issues-final-information-paper-on-bundling-telco-services orders Coles pay $10 million penalties', accessed 2 December 2015, from https://www.accc.gov.au/media-release/court-finds-coles- ACCC (2012) 'ACCC to oppose Carsales acquisition of Trading Post', accessed engaged-in-unconscionable-conduct-and-orders-coles-pay-10-million- 18 November 2015, from https://www.accc.gov.au/media-release/accc- penalties to-oppose-carsales-acquisition-of-trading-post ACCC (2015a) 'ACCC proposes to deny authorisation of ihail taxi booking app', ACCC (2013a) 'Coles and Woolworths undertake to cease supermarket accessed 22 November 2015, from https://www.accc.gov.au/media- subsidised fuel discounts', accessed 18 November 2015, from release/accc-proposes-to-deny-authorisation-of-ihail-taxi-booking-app http://www.accc.gov.au/media-release/coles-and-woolworths- undertake-to-cease-supermarket-subsidised-fuel-discounts ACCC (2015b) Draft Determination: Application for authorisation lodged by ihail Pty Ltd in respect of joint venture arrangements between taxi ACCC (2013b) 'Court declares consumer contract terms unfair', accessed 24 companies and other particpants in the taxi industry to launch the ihail November 2015, from https://www.accc.gov.au/media-release/court- smartphone taxi booking application; Authorisation number: A91501, declares-consumer-contract-terms-unfair Australian Competition and Consumer Commission, accessed 10 March 2016, from ACCC (2013c) Public Competition Assessment - Carsales.com Limited - http://registers.accc.gov.au/content/index.phtml/itemId/1186905/fromIte proposed acquisition of assets associated with Trading Post from mId/278039/display/acccDecision Telstra Corporation Limited, accessed 25 November 2015, from

Grattan Institute 2016 59 Peer-to-peer pressure

ACCC (2015c) 'Unconscionable conduct', accessed 19 November 2015, from Airtasker (2015b) 'Future Of Work - September 2015', accessed 23 October https://www.accc.gov.au/business/anti-competitive- 2015, from http://blog.airtasker.com/airtasker-news/future-of-work- behaviour/unconscionable-conduct report/ ACCC (2015d) 'Visa ordered to pay $18 million penalty for anti-competitive AngelList (2016) 'AngelList', accessed 28 March 2016, from https://angel.co/ conduct following ACCC action', accessed 18 November 2015, from http://www.accc.gov.au/media-release/visa-ordered-to-pay-18-million- Aral, S. (2013) The Problem With Online Ratings, MIT Sloan Management penalty-for-anti-competitive-conduct-following-accc-action Review from http://sloanreview.mit.edu/article/the-problem-with-online- ratings-2/ ACCC (2016) 'ACCC grants conditional authorisation to ihail taxi booking app', accessed 22 March 2016, from https://www.accc.gov.au/media- ATO (2005) Superannuation guarantee: who is an employee?, Australian release/accc-grants-conditional-authorisation-to-ihail-taxi-booking-app Taxation Officefrom http://law.ato.gov.au/pdf/pbr/sgr2005-001.pdf ACT Government (2015) '2015 Taxi Industry Innovation Reforms', accessed 11 ATO (2014) 'Super for contractors', accessed 12 October 2015, from January 2016, from https://www.ato.gov.au/printfriendly.aspx?url=/Business/Super-for- http://www.cmd.act.gov.au/policystrategic/regreform/2015-taxi-industry- employers/In-detail/Contractors/Super-for-contractors/ innovation-reforms ATO (2015a) 'Business or hobby?', accessed 9 November 2015, from Airbnb (2013a) 'Airbnb Economic Impact', accessed 2 November 2015, from https://www.ato.gov.au/business/starting-your-own-business/business- http://blog.airbnb.com/economic-impact-airbnb/ or-hobby-/ Airbnb (2013b) 'New Study: Airbnb Community Contributes AUD $214 Million to ATO (2015b) 'Capital gains tax (CGT) and going overseas', accessed 4 Sydney and its Suburbs, Brings Tourists to New Neighbourhoods', November 2015, from https://www.ato.gov.au/General/Capital-gains- accessed 11 November 2015, from tax/In-detail/International-issues/CGT-and-going- https://www.airbnb.com.au/press/news/new-study-airbnb-community- overseas/?page=1#Main_residence_exemption_and_temporary_absen contributes-aud-214-million-to-sydney-and-its-suburbs-brings-tourists- ce to-new-neighbourhoods ATO (2015c) 'Combating multinational tax avoidance - a targeted anti-avoidance Airbnb (2014a) 'Building trust with a new review system', accessed 1 April 2016, law', accessed 9 March 2016, from from http://blog.airbnb.com/building-trust-new-review-system/ https://www.ato.gov.au/General/New-legislation/In-detail/Other- topics/International/Combating-multinational-tax-avoidance---a- Airbnb (2014b) 'Great News from Australia: Airbnb Partners with Victorian targeted-anti-avoidance-law/ Government to Prepare for Emergencies', accessed 4 November 2015, from http://publicpolicy.airbnb.com/great-news-australia-airbnb- ATO (2015d) Freedom of Information Request - ATO and Uber, accessed 19 partners-victorian-government-prepare-emergencies/ November 2015, from http://foi.iorder.com.au/downloadfile.aspx?filename=1-713IQ47- Airtasker (2015a) 'Future Of Work - January 2015', accessed 8 December 2015, P1.pdf&fromPage=FOIPublicationIentification.aspx&Prodid=1-713IQ47- from https://assets-airtasker- P1 com.s3.amazonaws.com/uploads/message_asset/message_attachmen t/32431/Airtasker_Future_of_Work_Research_Jan_2015.pdf ATO (2015e) 'How GST works', accessed 4 November 2015, from https://www.ato.gov.au/Business/GST/How-GST-works/

Grattan Institute 2016 60 Peer-to-peer pressure

ATO (2015f) 'How to determine if workers are employees or contractors', https://www.ato.gov.au/General/New-legislation/In-detail/Other- accessed 30 November 2015, from topics/International/Country-by-Country-Reporting-and-Transfer- https://www.ato.gov.au/Business/Employee-or-contractor/How-to- Pricing-Documentation/ determine-if-workers-are-employees-or-contractors/ ATO (2016b) 'Country-by-Country Reporting and Transfer Pricing ATO (2015g) 'In focus: Offshore marketing hubs', accessed 16 November 2015, Documentation', accessed 9 March 2016, from from https://www.ato.gov.au/Business/Large-business/In- https://www.ato.gov.au/General/New-legislation/In-detail/Other- detail/Business-bulletins/Articles/In-focus--Offshore-marketing-hubs/ topics/International/Country-by-Country-Reporting-and-Transfer- Pricing-Documentation/ ATO (2015h) 'In focus: Procurement hubs of Australian multinational enterprises', accessed 16 November 2015, from Australian Government (2015) Budget Paper No. 3, The Treasury, Canberra https://www.ato.gov.au/Business/Large-business/In-detail/Business- bulletins/Articles/In-focus--Procurement-hubs-of-Australian- Australian Government (2016) Fixing competition policy to drive economic multinational-enterprises/ growth and jobs, from http://sjm.ministers.treasury.gov.au/media- release/030-2016/ ATO (2015i) 'Personal Services Income', accessed 3 November 2015, from https://www.ato.gov.au/Business/Personal-services-income/ Australian Human Rights Commission (2013) Wheelchair Accessible Taxi Inquiry report, accessed 14 January 2016, from ATO (2015j) 'Providing taxi travel services through ride-sourcing and your tax https://www.humanrights.gov.au/our-work/disability- obligations', accessed 18 November 2015, from rights/inquiries/wheelchair-accessible-taxi-inquiry-report https://www.ato.gov.au/Business/GST/In-detail/Managing-GST-in-your- business/General-guides/Providing-taxi-travel-services-through-ride- Australian Taxi Industry Association (2014) 2014 State and Territory Taxi sourcing-and-your-tax-obligations/ Statistics, Australian Taxi Industry Association, from http://www.atia.com.au/wp-content/uploads/2014-State-Territory-Taxi- ATO (2015k) 'Registering for GST', accessed 1 April 2016, from Statistics.pdf https://www.ato.gov.au/business/gst/registering-for-gst/ Ballantyne, A., De Voss, D. and Jacobs, D. (2014) 'Unemployment and Spare ATO (2015l) 'Registration for ABN, GST and fuel tax credits', accessed 18 Capacity in the Labour Market', RBA Bulletin, September quarter 2014, November 2015, from https://www.ato.gov.au/Business/GST/In- p 7-20 detail/GST-issues-registers/Taxi-industry---issues-register/?page=2 Battersby, L. (2015a) 'Uber takes legal action over meaning of taxi in GST fight', ATO (2015m) 'Renting out part or all of your home', accessed 3 November 2015, Sydney Morning Herald, 4 August 2015, accessed 23 November 2015, from https://www.ato.gov.au/General/Property/Your-home/Renting-out- from http://www.smh.com.au/digital-life/digital-life-news/uber-tells- part-or-all-of-your-home/ federal-court-it-is-not-a-taxi-company-and-should-not-pay-tax- 20150804-gir69t.html ATO (2015n) 'The sharing economy and tax', accessed 25 October 2015, from https://www.ato.gov.au/Business/GST/In-detail/Managing-GST-in-your- Battersby, L. (2015b) 'UberX Australian drivers working as 'co-equals' to ride- business/General-guides/The-sharing-economy-and-tax/ share tech company', Sydney Morning Herald, 29 June 2015, accessed 25 February 2016, from http://www.smh.com.au/digital-life/digital-life- ATO (2016a) 'Country-by-Country Reporting and Transfer Pricing news/uberx-australian-drivers-working-as-coequals-to-rideshare-tech- Documentation', accessed 21 March 2016, from

Grattan Institute 2016 61 Peer-to-peer pressure

company-20150623-ghvjx1.html#ixzz3lyKhmBde demand/ondemand/press-release/ Beaton-Wells, C. (2015) We’ve already had consultation on the effects test: Carey, A. (2015) 'Taxi satisfaction levels rise in Melbourne as industry lifts more is just a political smoke screen, The Conversation, from game', The Age, 10 May 2015, accessed 15 December 2015, from https://theconversation.com/weve-already-had-consultation-on-the- http://www.theage.com.au/victoria/taxi-satisfaction-levels-rise-in- effects-test-more-is-just-a-political-smoke-screen-51246 melbourne-as-industry-lifts-game-20150508- ggx8ps.html#ixzz3uLGYL4TC Benson, A. (2015) 'NSW & ACT: By-Laws & Rules - What are they? Who do they bind? Can they be enforced?', accessed 22 March 2016, from Casale, G. (2011) The employment relationship: A comparative overview, Hart https://www.lookupstrata.com.au/nsw-act-by-laws-rules-what-are-they- Publishing, Geneva who-do-they-bind-can-they-be-enforced/ Castle, J. (2015) 'UberX vs taxi - which is best?', accessed 20 December 2015, Bibby, P. (2013) 'Loud parties caused 'psychological damage', father tells court', from https://www.choice.com.au/transport/cars/general/articles/uberx- Sydney Morning Herald, April 23 2013, accessed 1 April 2016, from vs-taxi-which-one-is-best http://www.smh.com.au/nsw/loud-parties-caused-psychological- damage-father-tells-court-20130423-2ic4p.html Chen, L., Mislove, A. and Wilson, C. (2015) Peeking Beneath the Hood of Uber, Proceedings of the 2015 ACM Conference on Internet Measurement Botsman, R. and Rogers, R. (2011) What's mine is yours: how collaborative Conference consumption is changing the way we live, HarperCollins, London City of Fremantle (2016) 'Short stay accommodation', accessed 22 March 2016, Bruce, D. (2013) Taxi User Survey: 2013 Statistics, WA Department of Transport from http://www.fremantle.wa.gov.au/business/your-business/short- Passenger Services Unit, from stay-accommodation http://www.transport.wa.gov.au/mediaFiles/taxis/taxis_user_survey_201 3_stats.pdf City of Hobart (2015) 'Change of Use of a Residence to Visitor Accommodation', accessed 22 January 2016, from Bruene, J. (2015) 'Six Alt-Lending Unicorns Worth Combined $15 Billion: http://www.hobartcity.com.au/Development/Planning/Change_of_Use_o Lending Club, Prosper, On Deck, Sofi, Avant, Funding Circle', accessed f_a_Residence_to_Visitor_Accommodation 31 March 2016, from http://finovate.com/six-alt-lending-unicorns-worth- combined-15-billion-lending-club-prosper-on-deck-sofi-avant-funding- Cousins, D. and Fels, A. (2014) Review of aspects of the Taxi Industry in Perth circle/ and Western Australia, Royal Automobile Club of Western Australia, accessed 8 April 2016, from Buddelmeyer, H., McVicar, D. and Wooden, M. (2015) 'Non-Standard http://rac.com.au/cs/groups/public/documents/internetcontent/racstg058 “Contingent” Employment and Job Satisfaction: A Panel Data Analysis', 159.pdf Industrial Relations: A Journal of Economy and Society, 54(2), p 256- 275 Cramer, J. and Krueger, A. B. (2016) Disruptive Change in the Taxi Business: The Case of Uber, NBER Working Paper 22083, accessed 8 April Burson-Marsteller (2016) 'Press Release: Forty-Five Million Americans Say They 2016, from http://www.nber.org/papers/w22083.pdf Have Worked in the On-Demand Economy, While 86.5 Million Have Used It, According to New Survey', accessed 1 April 2016, from Deloitte Access Economics (2015a) Review of the collaborative economy in http://www.burson-marsteller.com/what-we-do/our-thinking/on- NSW, Report for NSW Department of Finance, Services and Innovation, accessed 7 April 2016, from

Grattan Institute 2016 62 Peer-to-peer pressure

https://www.finance.nsw.gov.au/sites/default/files/Deloitte_Report_Revi survey, New York: The Center for Global Enterprise ew_Collaborative_Economy.pdf Fair Work Building and Construction (2015) 'Employee or contractor', accessed 6 Deloitte Access Economics (2015b) The sharing economy and the Competition November 2015, from http://www.fwbc.gov.au/employee-or-contractor and Consumer Act, accessed 8 April 2016, from https://www.accc.gov.au/system/files/Sharing%20Economy%20- Fair Work Ombudsman (2015) 'Contractors and employees – what’s the %20Deloitte%20Report%20-%202015.pdf difference?', accessed 5 November 2015, from https://www.fairwork.gov.au/how-we-will-help/templates-and- Deloitte Access Economics (2016) Economic effects of ridesharing in Australia, guides/fact-sheets/rights-and-obligations/contractors-and-employees- accessed 25 February 2016, from whats-the-difference#employeecontractor http://www2.deloitte.com/au/en/pages/economics/articles/economic- effects-ridesharing-australia-uber.html Feeney, M. (2014) 'Brazil Welcomes Airbnb Amid World Cup', accessed 17 November 2015, from http://www.cato.org/blog/brazil-welcomes-airbnb- Department of Economic Development Jobs Transport and Resources (Victoria) amid-world-cup (2015) Improving transport for Victorians with limited mobility: A better Multi Purpose Taxi Program: Discussion paper, accessed 31 March Foulsham & Geddes (2015) 'Are Uber Drivers Independent Contractors or 2016, from Employees?', accessed 20 November 2015, from http://economicdevelopment.vic.gov.au/__data/assets/pdf_file/0006/120 http://www.fglaw.com.au/are-uber-drivers-employees/ 4764/A-better-Multi-Purpose-Taxi-Program-Discussion-Paper.pdf Fradkin, A., Grewal, E., Holtz, D. and Pearson, M. (2015) Bias and Reciprocity in Department of Transport (Western Australia) (2015) On-demand Transport: A Online Reviews: Evidence From Field Experiments on Airbnb, discussion paper for future innovation, accessed 7 April 2016, from Proceedings of the Sixteenth ACM Conference on Economics and http://www.transport.wa.gov.au/mediaFiles/taxis/TAXI_P_OnDemand_T Computation, ACM, accessed 6 April 2016, from ransport_Green_Paper_discussion.pdf http://andreyfradkin.com/assets/reviews_paper.pdf Dougherty, C. (2015) 'Uber Missed Criminal Records of Drivers, Prosecutors Freed, J. (2015) 'Yes to Uber but no to Airbnb could be the future of corporate Assert', The New York Times, 19 August 2015, accessed 31 March travel', Australian Financial Review, 18 October 2015, accessed 7 April 2016, from http://www.nytimes.com/2015/08/20/technology/uber- 2016, from http://www.afr.com/lifestyle/travel/yes-to-uber-but-no-to- missed-criminal-records-of-drivers-prosecutors-assert.html?_r=1 airbnb-could-be-the-future-of-corporate-travel-20151015- gkaj20#ixzz3sRxyVSie Dow, A. (2015) 'Melbourne hotels booked out for Cricket World Cup final', The Age, 27 March 2015, accessed 4 November 2015, from Freelancer (2016) FY15 Full Year Results Presentation, Freelancer http://www.theage.com.au/victoria/melbourne-hotels-booked-out-for- Frontier Economics (2014) Uber regulated?, accessed 15 January 2016, from cricket-world-cup-final-20150325-1m8041.html#ixzz3sO75K9Ml http://www.frontier-economics.com.au/documents/2014/07/uber- Drysdale, D. (2015) 'Australia's new multinational anti-avoidance law', accessed regulated.pdf 2 December 2015, from https://www.icas.com/technical- Gilbert, B. (2015) 'Uber's CEO gave us a lot of clues about what the future looks resources/australias-new-multinational-anti-avoidance-law like', accessed 11 January 2016, from http://www.techinsider.io/ubers- Evans, P. C. and Gawer, A. (2016) The rise of the platform enterprise: a global future-is-driverless-cars-2015-9 Good, J. and Prudden, T. (2015) 'Litigation alert: VCAT rules Owners

Grattan Institute 2016 63 Peer-to-peer pressure

Corporations do not have power to prohibit short-term letting', accessed modernizing-labor-laws-for-the-gig- 5 November 2015, from economy/modernizing_labor_laws_for_twenty_first_century_work_krue http://www.brandpartners.com.au/component/content/article/82- ger_harris.pdf news/292-watergate Heinemann, F. and Shume, M. (2015) Uber, Airbnb, Netflix... Australia’s Steps to Greenwood, B. N. and Wattal, S. (2015) 'Show Me the Way to Go Home: An Tax the Sharing and Digital Economies, Tax Planning International Empirical Investigation of Ride Sharing and Alcohol Related Motor Indirect Taxes, from http://www.cgw.com.au/wp- Vehicle Homicide', Fox School of Business Research Paper No. 15- content/uploads/2015/07/Uber_Airbnb_Netflix_Sharing_Economy.pdf 054, Hill, S. (2015) The Unsavory Side of Airbnb, The American Prospect, accessed 8 Hagiu, A. (2015) Companies Need an Option Between Contractor and January 2016, from http://prospect.org/article/evictions-and- Employee, Harvard Business Review, accessed 25 November 2015, conversions-dark-side-airbnb from https://hbr.org/2015/08/companies-need-an-option-between- contractor-and-employee IBISWorld (2016) 'Hotels and Resorts in Australia: Market Research Report', accessed 1 April 2016, from Hall, J., Kendrick, C. and Nosko, C. (2015) The Effects of Uber’s Surge Pricing: http://www.ibisworld.com.au/industry/default.aspx?indid=1811 A Case Study, accessed 8 April 2016, from http://faculty.chicagobooth.edu/chris.nosko/research/effects_of_uber's_ Independent Panel on Short-Stay Accommodation in CBD Apartment Buildings surge_pricing.pdf (2015) Independent Panel on Short-Stay Accommodation in CBD Apartment Buildings - Final Report, Victorian Government Hall, J. V. and Krueger, A. B. (2015) An Analysis of the Labor Market for Uber’s Driver-Partners in the United States, Princeton University Working IPART (2015) Sydney taxi fares to apply and new licences to be released from Paper, from July 2015, Final Report, Independent Pricing and Regulatory Tribunal of http://dataspace.princeton.edu/jspui/bitstream/88435/dsp010z708z67d/ New South Wales, from 5/587.pdf http://www.ipart.nsw.gov.au/files/sharedassets/website/trimholdingbay/fi nal_report_- Han, M. (2015) 'Uber passengers to pay $1 extra for trips to compensate taxi _sydney_taxi_fares_to_apply_and_new_licences_to_be_released_from licence holders', Australian Financial Review, 17 December 2015, _july_2015_-_february_2015.pdf accessed 13 January 2016, from http://www.afr.com/technology/uber- passengers-to-pay-1-extra-for-trips-to-compensate-taxi-licence-holders- Johanson, S. (2014) 'Owners score legal win over Southbank short-stay 20151217-glq213#ixzz3x4vJIBmH operator', The Age, 21 May 2014, accessed 21 January 2016, from http://www.theage.com.au/business/owners-score-legal-win-over- Harper, I., Anderson, P., McCluskey, S. and Michael, O. B. (2015) Competition southbank-shortstay-operator-20140520-38mjd.html#ixzz3xvjX7hsc Policy Review - Final Report, accessed 8 April 2016, from http://competitionpolicyreview.gov.au/files/2015/03/Competition-policy- Johanson, S. (2015) 'Short-stay operators win tribunal decision over Watergate review-report_online.pdf building', The Age, July 15 2015, accessed 22 November 2015, from http://www.theage.com.au/business/shortstay-operators-win-tribunal- Harris, S. and Krueger, A. (2015) A Proposal for Modernizing Labor Laws for decision-over-watergate-building-20150714-gibv4k.html Twenty-First-Century Work: The “Independent Worker”, Brookings Institute, accessed 7 April 2016, from JPMorgan Chase Institute (2016) Paychecks, Paydays, and the Online Platform http://www.brookings.edu/~/media/Research/Files/Papers/2015/12/09- Economy Big Data on Income Volatility, accessed 6 April 2016, from

Grattan Institute 2016 64 Peer-to-peer pressure

https://www.jpmorganchase.com/corporate/institute/document/jpmc- December 2015, from http://focuslaw.ca/blog/employment/dependent- institute-volatility-2-report.pdf contractors-the-intermediate-category/ Katz, L. F. and Krueger, A. B. (2016) The Rise and Nature of Alternative Work Manjoo, F. (2016) 'Car-Pooling Helps Uber Go the Extra Mile', New York Times, Arrangements in the United States, 1995-2015 accessed 11 April 2016, 30 March 2016, accessed 7 April 2016, from from http://www.nytimes.com/2016/03/31/technology/car-pooling-helps-uber- http://krueger.princeton.edu/sites/default/files/akrueger/files/katz_krueg go-the-extra-mile.html?_r=0 er_cws_-_march_29_20165.pdf McKenny, L. (2014) 'Councils threaten home owners with $1 million fine for Khadem, N. (2015) 'Uber sends 25pc of fares to the Netherlands', Sydney renting rooms', Sydney Morning Herald, 26 September 2014, accessed Morning Herald, November 18 2015, accessed 10 December 2015, 10 November 2015, from http://www.smh.com.au/nsw/councils- from http://www.smh.com.au/business/the-economy/uber-sends-25pc- threaten-home-owners-with-1-million-fine-for-renting-rooms-20140926- of-fares-to-the-netherlands-20151118-gl1yqs.html#ixzz3sO90eSyC 10mchq.html#ixzz3sSGoRIxz King, S. (2013) 'Google – and everyone else – wins by High Court decision', The McKinsey Global Institute (2015) A Labor Market that Works: Connecting Talent Conversation, 8 February 2013, accessed 20 November 2015, from with Opportunity in the Digital Age, McKinsey & Company https://theconversation.com/google-and-everyone-else-wins-by-high- court-decision-12098 Nguyen, L. and Oliver, J. (2013) 'The Australian Consumer Law – another compliance obligation for universities?', accessed 7 April 2016, from KMPG (2015) 'Australia's Multinational Anti-Avoidance law and other tax http://www.minterellison.com/publications/australian-consumer-law- transparency measures', accessed 9 March 2016, from another-compliance-obligation-for-universities-HEF20130507/ http://www.kpmg.com/au/en/issuesandinsights/articlespublications/page s/multinational-tax-avoidance-bill-2015.aspx NRMA Insurance (2015) 'NRMA Insurance welcomes ACT ridesharing regulation', accessed 13 January 2016, from Kusisto, L. (2015) 'Airbnb Pushes Up Apartment Rents Slightly, Study Says', The http://www.nrma.com.au/nrma-insurance-welcomes-act-ridesharing- Wall Street Journal, 30 March 2015, accessed 7 April 2016, from regulation http://blogs.wsj.com/developments/2015/03/30/airbnb-pushes-up- apartment-rents-slightly-study-says/ NSW Environmental Protection Authority (2015) 'Noise', accessed 1 April 2016, from http://www.epa.nsw.gov.au/noise/ Larkins, D. and Kane, C. (2014) 'New laws target 'party houses' on Queensland's Gold Coast', accessed 15 March 2016, from NSW Fair Trading (2015) 'By-law reforms', accessed 24 November 2015, from http://www.abc.net.au/news/2014-06-03/new-laws-could-stop-party- http://www.fairtrading.nsw.gov.au/ftw/About_us/Have_your_say/Review houses-on-queenslands-gold-coast/5496588 _strata_community_scheme_laws/By-law_reforms.page Law, J. and Ma, W. (2015) 'UberX versus taxis: Why the time for taxis is over', NSW Government (2015a) Customer Satisfaction Index 2015, , Transport for accessed 12 January 2016, from NSW, accessed 13 January 2016, from http://www.news.com.au/finance/business/travel/uberx-versus-taxis- http://www.transport.nsw.gov.au/sites/default/files/b2b/publications/cust why-the-time-for-taxis-is-over/news- omer-satisfaction-index-2015.pdf story/cb1e2d9dc77aa29dd88646cbae1d5d4e NSW Government (2015b) Point to Point Transport Taskforce: Discussion Lee, G. (2012) 'Dependent contractors: the intermediate category', accessed 2 Paper, accessed 12 January 2016, from https://p2p-prod-

Grattan Institute 2016 65 Peer-to-peer pressure

assets.s3.amazonaws.com/p2p/s3fs-public/discussion-paper.pdf %20Inquiry%20Into%20the%20Adequacy%20of%20Regulation%20of %20Short%20Term%20Lettings%20151109_0.pdf NSW Government (2015c) Point to point transport. The NSW Government response to the taskforce report, accessed 11 January 2016, from Pallais, A. (2013) Inefficient hiring in entry-level labor markets, NBER Working http://www.transport.nsw.gov.au/sites/default/files/b2b/publications/poin paper 18917, accessed 3 December 2015, from t-to-point-nsw-government-response-to-taskforce-report.pdf http://www.nber.org/papers/w18917.pdf NSW Government (2015d) Response to the Taskforce Recommendations, Palmer, C. (2015) 'Experts caution Australia on unilateral ‘Google tax’', The accessed 11 January 2016, from Conversation, 22 March 2016, accessed 8 April 2015, from http://www.transport.nsw.gov.au/sites/default/files/b2b/publications/poin https://theconversation.com/experts-caution-australia-on-unilateral- t-tp-point-taskforce-recommendations-government-response.PDF google-tax-39794 O'Neill, J. W. and Ouyang, Y. (2016) From Air Mattresses to Unregulated Pereira, D. and Cunynghame, S. (2015) 'Federal Government passes new laws Business: An Analysis of the Other Side of Airbnb, Penn State School extending unfair contract terms protections to small business contracts', of Hospitality Management, accessed 11 April 2016, from accessed 2 December 2015, from http://www.ahla.com/uploadedFiles/_Common/pdf/PennState_AirBnbR http://www.holdingredlich.com/corporate-commercial/federal- eport_.pdf government-passes-new-laws-extending-unfair-contract-terms- protections-to-small-business-contracts OECD (2007) Taxi Services: Competition and Regulation, accessed 7 April 2016, from www.oecd.org/dataoecd/49/27/41472612.pdf Pro Bono Australia (2015) 'Uber Assists Passengers with Disability', accessed 13 January 2016, from OECD (2015a) Addressing the Tax Challenges of the Digital Economy, Action 1 http://www.probonoaustralia.com.au/news/2015/05/uber-assists- - 2015 Final Report, OECD Publishing, Paris passengers-disability# OECD (2015b) Reforms to the international tax system for curbing avoidance by Productivity Commission (1999a) Regulation of the Taxi Industry, accessed 8 multinational enterprises, accessed 17 November 2015, from April 2016, from http://www.pc.gov.au/research/supporting/taxi- http://www.oecd.org/tax/oecd-presents-outputs-of-oecd-g20-beps- regulation/taxiregulation.pdf project-for-discussion-at-g20-finance-ministers-meeting.htm Productivity Commission (1999b) Structural Adjustment — Exploring the Policy OECD (2016) 'About Base Erosion and Profit Shifting (BEPS)', accessed 21 Issues, Workshop Proceedings, accessed 29 February 2016, from March 2016, from http://www.oecd.org/tax/beps-about.htm http://www.pc.gov.au/research/supporting/structural- Oliver, D. (2015) The State of the Unions, Speech to the National Press Club, adjustment/structuralchange.pdf Australian Council of Trade Unions, from http://www.actu.org.au/actu- Productivity Commission (2006) The Role of Non-Traditional Work in the media/speeches-and-opinion/the-state-of-the-unions-address-by-actu- Australian Labour Market, Research Paper, accessed 7 April 2016, secretary-dave-oliver-to-the-national-press-club from http://www.pc.gov.au/research/supporting/non-traditional- Owners Corporation Network (2015) Submisssion to the Inquiry into the work/nontraditionalwork.pdf adequacy of the regulation of short-term holiday letting in NSW, Productivity Commission (2015a) Business Set-up, Transfer and Closure, Draft accessed 8 April 2016, from Report, accessed 7 April 2016, from http://www.ocn.org.au/sites/ocn.org.au/files/OCN%20Submission%20- http://www.pc.gov.au/inquiries/completed/business/draft/business-

Grattan Institute 2016 66 Peer-to-peer pressure

draft.pdf http://www.governmentnews.com.au/2015/05/coastal-nsw-councils-find- airbnb-groove/ Productivity Commission (2015b) Workplace Relations Framework, Draft Report, Schein, A. I. (2014) 'When Is A Worker A Dependent Contractor?', accessed 20 Prosper Australia (2015) 'Speculative Vacancies Defy Housing Affordability November 2015, from Pressures – Again', accessed 7 January 2016, from http://www.mondaq.com/canada/x/336198/employment+litigation+tribun https://www.prosper.org.au/2015/12/09/vacant-homes-up-22-despite- als/When+Is+A+Worker+A+Dependent+Contractor housing-crisis/ Schetzer, A. and Battersby, L. (2016) 'Airbnb: a numbers guide for Melbournians PwC (2015) The Sharing Economy, PwC, accessed 7 April 2016, from raking in the dollars', The Age, 17 January 2016, accessed 18 January http://www.pwc.com/us/en/industry/entertainment- 2016, from http://www.theage.com.au/victoria/airbnb-a-numbers-guide- media/publications/consumer-intelligence-series/assets/pwc-cis- for-melbournians-raking-in-the-dollars-20160114-gm5w9b.html sharing-economy.pdf Schlesinger, L. (2016) 'Airbnb under fire over commercial landlords', Australian RBA (2015) Submission to the Inquiry into Home Ownership: Proportion of Financial Review, 21 January 2016, accessed 7 March 2016, from Investment Housing Relative to Owner-Occupied Housing, Reserve http://www.afr.com/real-estate/airbnb-under-fire-over-commercial- Bank of Australia, accessed 6 April 2016, from landlords-20160121-gmb4h5 http://www.rba.gov.au/publications/submissions/inquiry-into-home- ownership/proportion-investment-housing-relative-owner-occ- Schoettle, B. and Sivak, M. (2014) A survey of public opinion about autonomous housing.html and self-driving vehicles in the US, the UK, and Australia, The University of Michigan Transportation Research Institute Roads and Maritime Services (NSW) (2015) 'Registration and licensing statistics', accessed 31 March 2016, from Shire of Augusta-Margaret River (2014) PS17 – Holiday Houses and use of http://www.rms.nsw.gov.au/about/corporate- Grouped Dwellings for Short Stay Planning Information Sheet, publications/statistics/registrationandlicensing/tables/table1116_2015q4 accessed 22 March 2016, from .html http://www.amrshire.wa.gov.au/library/file/2Services/09%20Planning/PS 17%20Holiday%20Houses%20and%20Use%20of%20Grouped%20Dw Robb, K. (2016) 'Airbnb: Victorian laws under pressure as landlord fails to evict ellings%20for%20Short%20Stay.pdf ‘sublet’ tenants', Domain, 2 April 2016, accessed 11 April 2016, from http://www.domain.com.au/news/airbnb-victorian-laws-under-pressure- Shontell, A. (2015) 'Uber is cleverly controlling its stock so it doesn't end up like as-landlord-fails-to-evict-sublet-tenants-20160401-gnv3p4/ Facebook', Business Insider, 14 December 2015, accessed 30 March 2016, from http://www.businessinsider.com.au/uber-is-about-to-raise-a- Roth, A. E. (2008) 'What Have We Learned from Market Design?', The giant-round-at-a-65-billion-valuation-heres-how-thats-even-possible- Economic Journal, 118(527), p 285-310 2015-11 Rothengatter, M. (2008) ''Taxing taxis’ - limits and possibilities for regulating tax Silva, K. (2015) 'Uber trials new Brisbane service for passengers with compliance behaviours of taxi-drivers: an Australian case study', PhD wheelchairs', Brisbane Times, 27 October 2015, accessed 14 January thesis, School of Arts and Social Sciences, University of Queensland 2016, from http://www.brisbanetimes.com.au/queensland/uber-trials- Sansom, M. (2015) 'Coastal NSW councils find Airbnb groove', accessed 15 new-brisbane-service-for-passengers-with-wheelchairs-20151027- March 2016, from gkk67d.html#ixzz3xBc2o5KK

Grattan Institute 2016 67 Peer-to-peer pressure

Sims, R. (2015) 'Opening Address: 13th Annual Competition Law & Economics services-airbnb-are-altering-economics-hotel-business-buffetts Workshop', accessed 3 November 2015, from https://www.accc.gov.au/speech/opening-address-13th-annual- The Law Society of New South Wales (2015) Submission to the NSW Strata and competition-law-economics-workshop Community Title Law Review accessed 15 March 2016, from https://www.lawsociety.com.au/cs/groups/public/documents/internetpoli Smith, M. (2016) 'Va. DMV: Uber, Lyft cut drunken driving deaths in 2015', cysubmissions/1045260.pdf accessed 29 February 2016, from http://wtop.com/virginia/2016/01/hold-sun-va-dmv-uber-lyft-cut-drunken- Thierer, A. D., Koopman, C., Hobson, A. and Kuiper, C. (2015) How the Internet, driving-deaths-2015/slide/1/ the Sharing Economy, and Reputational Feedback Mechanisms Solve the “Lemons Problem”, Mercatus Center, accessed 7 January 2016, Stewart, A. and Alderman, P. (2013) Independent contractors: a practical guide, from http://mercatus.org/sites/default/files/Thierer-Lemons-Problem.pdf CCH Australia Limited, Melbourne Thomson, J. (2016) 'Flat dwellers in Sydney’s ‘Airbnb belt’ hit hardest as the Strata Community Australia (2015) 'Understanding Strata - Short term leasing', sharing economy takes off', Domain, 15 January 2016, accessed 11 accessed 2 November 2015, from April 2016, from http://www.domain.com.au/news/flat-dwellers-in- https://www.stratacommunity.org.au/understandingstrata/short-term- sydneys-airbnb-belt-bear-the-brunt-as-sharing-economy-takes-off- leasing 20160115-gm3z3q/ Taxi Industry Inquiry (2012) Final Report - Customers First: Service, Safety, Ting, A. (2015) 'Explainer: how Uber and Airbnb are reducing their Australian tax Choice, accessed 2 April 2016, from bill', The Conversation, 1 December 2015, accessed 7 April 2016, from http://taxi.vic.gov.au/__data/assets/pdf_file/0019/20791/Taxi-Industry- https://theconversation.com/explainer-how-uber-and-airbnb-are- Inquiry-Final-Report.PDF reducing-their-australian-tax-bill-51111 Taxi Services Commission (2015) 'Travelling with a disability', accessed 13 TransLink (2015) 'Taxi Subsidy Scheme', accessed 14 January 2016, from January 2016, from http://taxi.vic.gov.au/passengers/travelling-with-a- http://translink.com.au/tickets-and-fares/concessions/taxi-subsidy- disability scheme Taxi Services Commission (2016) 'Taxi and hire car industry statistics', accessed Transport for NSW (2005) Benefit/Cost Assessment of Options for Reform of 31 March 2016, from http://taxi.vic.gov.au/about-us/overview/industry- Taxi Licensing - Final Report, accessed 30 March 2016, from statistics http://www.transport.nsw.gov.au/sites/default/files/b2b/taxi/Report- Benefit-Cost-Assessment-of-Options-for-Reform-of-Taxi-Licensing.pdf The Daily Telegraph (2015) 'Family’s Airbnb drug house nightmare as police raid holiday home and find hydroponic setup', The Daily Telegraph, 8 Treasury (2015a) Australian Government Response to the Competition Policy January 2016, accessed 21 January 2016, from Review, accessed 25 November 2015, from http://www.dailytelegraph.com.au/news/nsw/familys-airbnb-drug-house- http://www.treasury.gov.au/~/media/Treasury/Publications%20and%20 nightmare-as-police-raid-holiday-home-and-find-hydroponic- Media/Publications/2015/Government%20response%20to%20the%20C setup/news-story/ ompetition%20Policy%20Review/Downloads/PDF/Govt_response_CPR .ashx The Economist (2016) 'Buffett’s revenge' The Economist, 9 January 2016, accessed 11 January 2016, from Treasury (2015b) Tax Laws Amendment (Tax Integrity: GST and Digital http://www.economist.com/news/finance-and-economics/21685502- Products) Bill 2015: Exposure Draft Explanatory Material, accessed 6

Grattan Institute 2016 68 Peer-to-peer pressure

April 2016, from https://www.police.vic.gov.au/retrievemedia.asp?Media_ID=72176 http://www.treasury.gov.au/~/media/Treasury/Consultations%20and%2 0Reviews/Consultations/2015/Tax%20integrity%20Extending%20GST Victorian Government (2015) 'Taxi Reform Hardship Fund', accessed 29 %20to%20digital%20products/Downloads/PDF/EM_Tax_Integrity_GST February 2016, from http://www.vic.gov.au/news/taxi-reform-hardship- _and_Digital_Products.ashx fund.html Truong, A. (2015) 'Why the hotel industry isn’t afraid of Airbnb (yet), explained in WA Government (2012) Holiday homes local law 2012, Western Australian five charts', accessed 8 April 2016, from http://qz.com/551612/why-the- Government Gazette hotel-industry-isnt-afraid-of-airbnb-yet-explained-in-five-charts/ WA Government (2015) Green light for on-demand transport industry reform, Tucker, H. (2016) 'REVEALED: News Corp spent $40 million on its 25% stake in accessed 13 January 2016, from hipages, the tradie marketplace', accessed 31 March 2016, from https://www.mediastatements.wa.gov.au/Pages/Barnett/2015/12/Green- http://www.businessinsider.com.au/revealed-news-corp-spent-40- light-for-on-demand-transport-industry-reform.aspx million-on-its-25-stake-in-hipages-the-tradie-marketplace-2016-1 Wallsten, S. (2015) 'The competitive effects of the sharing economy: how is Uber Uber (2014) 'Three Septembers of UberX in New York City', accessed 1 March changing taxis?', accessed 7 April 2016, from 2016, from https://newsroom.uber.com/us-new-york/three-septembers- https://www.ftc.gov/system/files/documents/public_comments/2015/06/ of-uberx-in-new-york-city/ 01912-96334.pdf Uber (2015a) '15,000 Jobs, 1,000,000 Riders', accessed 20 October 2015, from Watergate (2015) '95.5% of voters say stop Watergate becoming a hotel', from http://newsroom.uber.com/australia/2015/08/15000-jobs-1000000- http://www.watergate.net.au/wg/index.php/news-a-announcements/107- riders-2/ 955-of-voters-say-stop-watergate-becoming-a-hotel Uber (2015b) 'Insurance', accessed 12 January 2016, from Weiss, M. and Schmidt, M. (2008) Labour law and industrial relations in http://www.ubersydney.info/insurance/ Germany, Kluwer law international Uber (2015c) 'Ridesharing with Uber: a safe, reliable and affordable transport Weyl, E. G. (2010) 'A price theory of multi-sided platforms', The American option', accessed 11 January 2016, from Economic Review, p 1642-1672 https://newsroom.uber.com/sydney/ridesharing-with-uber-a-safe- Wilkins, R. and Warren, D. (2012) 'Families, incomes and jobs, volume 7: a reliable-and-affordable-transport-option/ statistical report on waves 1 to 9 of the household, income and labour Uber (2015d) 'A Safe Ride from A to B', accessed 11 January 2015, from dynamics in Australia survey', Melbourne Institute of Applied Economic https://newsroom.uber.com/australia/ubersafeau/ and Social Research, Uber (2015e) 'uberX hits 10 million rides in Australia!', accessed 17 December Willingham, R. (2016) 'Disability groups fear Uber will kill off wheelchair cabs', 2015, from https://newsroom.uber.com/australia/2015/10/10millionrides/ The Age, 10 January 2016, accessed 14 January 2016, from http://www.theage.com.au/victoria/disability-groups-fear-uber-will-kill- Uber (2016) 'Join uberX', accessed 14 January 2016, from off-wheelchair-cabs-20151222-glthji.html#ixzz3xBeocZcr https://drive.uber.com/melbourne/uberx/join-uberx/ Wooden, M. and Warren, D. (2004) 'Non-standard employment and job Victoria Police (2014) Crime Statistics 2013/2014 accessed 7 April 2016, from satisfaction: evidence from the HILDA Survey', Journal of Industrial

Grattan Institute 2016 69 Peer-to-peer pressure

Relations, 46(3), p 275-297 Zervas, G., Proserpio, D. and Byers, J. (2014) 'The rise of the sharing economy: Estimating the impact of Airbnb on the hotel industry', Boston U. School of Management Research Paper, 2013-16, Zervas, G., Proserpio, D. and Byers, J. (2015) A First Look at Online Reputation on Airbnb, Where Every Stay is Above Average, Social Science Research Network, accessed 3 December 2015, from http://ssrn.com/abstract=2554500 Zipkin, A. (2015) 'The Sharing Economy Attracts Older Adults', New York Times, 25 September 2015, accessed 25 October 2015, from http://www.nytimes.com/2015/09/26/your-money/the-sharing-economy- attracts-older-adults.html?_r=0

Grattan Institute 2016 70 Peer-to-peer pressure

List of legislation

Commonwealth: State and Territory:

A New Tax System (Goods and Services Tax) Act 1999 (Cth) Body Corporate and Community Management Act 1997 (Qld)

Competition and Consumer Act 2010 (Cth) Community Titles Act 1996 (SA)

Fair Work Act 2009 (Cth) Environment Protection (Residential Noise) Regulations 2008 (Vic) Income Tax Assessment Act 1936 (Cth) Owners Corporations Act 2006 (Vic) Income Tax Assessment Act 1997 (Cth) Protection of the Environment Operations Act 1997 (NSW) Independent Contractors Act 2006 (Cth) Strata Schemes Management Act 1996 (NSW) Superannuation Guarantee (Administration) Act 1992 (Cth) Strata Titles Act 1988 (SA)

Strata Titles Act 1998 (Tas)

Sustainable Planning Act 2009 (Qld)

Transport (Compliance and Miscellaneous) Act 1983 (Vic)

Unit Titles (Management) Act 2011 (ACT)

Workplace Injury Management and Workers Compensation Act 1998 (NSW)

Grattan Institute 2016 71 Peer-to-peer pressure

List of cases

ACCC v Metcash [2011] FCA 967 [2011] FCAFC 151

ACCC v. Visa Inc & Ors (2013)

Australian Gas Light Company (AGL) v ACCC (No. 3) [2003] FCA 1525

Dobrohotoff v Bennic [2013] NSWLEC 61

Fortescue Metals Group Limited [2010] ACompT 2

Genco & Anor v Salter & Anor [2013] VSCA 365

Google v ACCC (2013) 87 ALJR 235

Google Spain v AEPD and Mario Costeja González (2014)

Hollis v Vabu Pty Ltd (2001) 207 CLR 21

Jean Whittlam v Sara Hannah & John Hannah [2011] Downing Centre Local Court 63913/11)

Lee v Owners Corporation No. 501391P

Owners Corporation PS501391P v Balcombe (Owners Corporations) [2015] VCAT 956 (29 June 2015)

Voros v Dick [2013] FWCFB 9339

Grattan Institute 2016 72 Belonging anywhere

Airbnb and renting in Sydney A Report by the Tenants’ Union of NSW March 2017 Introduction

Airbnb is an online short-term As a result, Airbnb has caused concern letting platform created in August amongst tenant representative groups 2008 in Southern California. It around the world. The perception from arrived in Sydney not long after, with these groups is that Airbnb & other similar the frst listings for Sydney appearing companies are encouraging landlords to by February of 2009. Airbnb was remove dwellings from the rental sector originally marketed as part of the due to higher returns in short term letting. sharing economy, where a spare room might be listed. But in 2017 Some owners and neighbours also listings are increasingly made for report being disturbed by short term entire homes. Compare below an letting guests. This is a problem that early ‘About page’ in February 2009 has been around for a very long time, to the About page currently on the but the rise of Airbnb has potentially site in March 2017’. exacerbated this issue.

Now and then: “About Airbnb” in their own words

February 20091 March 20172

What is AirBed & Breakfast? ... Airbnb is a trusted community marketplace for people to list, discover, AirBed & Breakfast is an online marketplace and book unique accommodation around for peer-to-peer traveling. the world – online or from a mobile phone or tablet. We enable people to earn money by renting out extra space, and offer travelers a viable Whether an apartment for a night, a alternative to hotels. castle for a week, or a villa for a month, Airbnb connects people to unique travel How it works experiences, at any price point, in more It’s simple. Nice folks, folks like you, list their than 34,000 cities and 191 countries... guest rooms, futons, and even couches on the site and set a price per night. Adventurous travelers looking for a place to stay can search the listings for an accommodation that’s just right.

1 http://web.archive.org/web/20090204194634/http://airb- 2 http://web.archive.org/web/20170330015713/https:// nb.com/home/about www.airbnb.com.au/about/about-us

Belonging anywhere: Airbnb and renting in Sydney Page 2 This paper seeks to give preliminary answers to questions concerning the impact of Airbnb on the private rental sector in New South Wales. We hope that it encourages the further exploration of these issues by academics, governments and advocates for tenants. “We hope that Primarily we want to know what it encourages the impact the presence of Airbnb in Sydney is having on private rents. further exploration Ideally this would be measured in two ways - by examining whether the stock of rental properties has been reduced, of these issues and whether the rents of those properties has increased. by academics,

We will approach this question by governments and examining the business case for removing a rental property from the advocates for market. In other words, will properties listed on Airbnb make more money tenants.” than simply renting the premises out to residential tenants.

We will also look to quantify how many properties will be affected by a potential change to council approvals that will require hosts to seek permission to use a property for short term lets for more than a certain number of days in a year.

In this paper we do not explore the issue of being disturbed by Airbnb guests or the impact on neighbours and community.

We have primarily used data supplied by AirDNA, a data scraping company, for this analysis. Unless otherwise noted this is the source for information about Airbnb listings in this report. Other data sources include Inside Airbnb (another data scraping site), the Department of Planning and Environment Population Projections and the Rent and Sales Report. More information on sources and methodology is at the end of the report.

Belonging anywhere: Airbnb and renting in Sydney Page 3 Airbnb in Sydney

Australia has been described by owners. As a result, we can only Airbnb as “the most penetrated speculate as to how many properties market in the world”1 with Sydney on Airbnb are listed by tenants or leading the country for listings. At by home owners. the time that fgure was quoted in June 2016 Airbnb claimed 18,000 This is an area which requires further listings in Sydney. An alternative use study in order to make any real claim of the phrase at the same time2 about the impact of short term lets described Sydney and Melbourne as upon the private rental market. It is 5th and 6th respectively in the world much harder to make the case that for ‘users’ of the site. This likely included Airbnb and other sites are having an travellers, not just local listings. impact if, for instance, much of the use are renters themselves merely Our estimate over the year to using the service to offset their September 2016 is that Airbnb usage already high rents. has grown to take in around $160 million in gross payments by guests. $130 million of that was for entire Airbnb listings across Sydney homes, and the remainder for rooms in larger premises, or even shared rooms. In comparison, the rent collected by the private rental market in the Sydney region for the same period is poorly measured, but would be well in excess of $10 billion.3

Airbnb does not ask its Australian hosts whether they are tenants or property

1 http://www.theaustralian.com.au/business/prop- erty/sydneys-eastern-suburbs-popular-with-aus- tralian-airbnb-users/news-story/44c628e64ec79e- 4d1a68d72e992888fc 2 http://www.theaustralian.com.au/business/austra- lia-key-to-airbnbs-plans/news-story/54b662a772eb6 d28fdd1061534371b0d 3 Our own conservative analysis of all bonds held by the Rental Bond Board in September 2016: https://data.nsw.gov.au/data/dataset/formal-gi- pa-access-application-2016-2017-fa-13/resource/ ca5c4a11-64f8-4583-91c9-d9436d38e2e9

Listings with more than 60 nights booked in a year Listings with 1-60 nights booked in a year

Belonging anywhere: Airbnb and renting in Sydney Page 4 The areas in which Airbnb is most heavily located are also some of the most affuent and established tourist- friendly areas, and are also close to Sydney’s main economic centres. They also historically have some of the highest private rents in Sydney.

The following page shows the proportion of rented dwellings and Airbnb listings in the Greater Sydney area and then the Waverley, City of Sydney and Manly local government areas. The initial listings number is not examined for recent activity or income – they have merely appeared on the Airbnb website and been captured by Inside Airbnb. The smaller, second circle is the number of entire homes How much extra rent? which have been frequently and The median rent in these recently booked, as measured by LGAs is much higher than Inside Airbnb.4 Greater Sydney Across Sydney 1.2% of Sydney’s almost 2 million properties were listed on Airbnb by September 2016. Even in popular Airbnb areas, only a tiny number were being frequently used.

We have illustrated the coverage on these three local government areas due to their level of Airbnb listing density and publicity as hubs of Airbnb activity.

4 http://insideairbnb.com/sydney/ Recently and frequently booked is defned as having received at least one review in the last 6 months and at least 90 booked nights in the last year.

Belonging anywhere: Airbnb and renting in Sydney Page 5 airbnb in Sydney Makeup of rented dwellings and Airbnb listings Sources: Insideairbnb, FACS, Planning & Environment

Greater Sydney All private 30.1% of private dwellings dwellings are rented (bonds lodged)

1.2% of dwellings have been listed on Airbnb

0.16% of dwellings are listed as entire homes and are recently and frequently booked on Airbnb

City of Sydney 53% rented dwellings 5.5% Airbnbs 0.9% entire homes, recently & frequently booked on Airbnb

City of Manly 34% rented dwellings 5.3% Airbnbs 0.8% entire homes, recently & frequently booked on Airbnb

City of Waverley 45% rented dwellings 14.2% Airbnbs 1.0% entire homes, recently & frequently booked on Airbnb Airbnb in Sydney

Listing Activity

Both supporters and opponents of which had more than 7 nights Airbnb cite large numbers for how booked in each month. many listings appear on Airbnb and how fast these numbers are We can see very clearly the summer growing.1 These numbers affect jump in casual bookings, but the the potential impact listings have growth in heavier use is much less on Sydney’s market. However the seasonal. This indicates a large number number of listings which appear to be of people are listing homes on Airbnb active at any one time is not as large. very infrequently, and only using the site at particular peak times. We have compared the number of “entire home” listings ever created The gap between listings ever created with the listings active in each month and those active in each month also from August 2014 to August 2016. suggests that many listings may have We have further broken down listings stopped being active. Intuitively this is obvious as the same host may have 1 http://www.abc.net.au/news/2017-01-31/airbnb- moved house. However the large booming-in-sydney-but-it-could-be-pushing-up- numbers also start to indicate that rents/8223900

Growth in Airbnb compared to monthly activity Greater Sydney 2014-16

Belonging anywhere: Airbnb and renting in Sydney Page 7 continued use may not be what hosts are seeking, and certainly aren’t what they are achieving.

We repeated the work at the postcode level for all the most densely listed Airbnb areas (compared to rentals) and found very similar stories. The peak time for Airbnb is over summer, particularly in the beach areas, but large numbers of listings over that period are for 7 or less days in the month, and do not continue after summer.

Where these summer listings are residents using their property while they have a week away, this would have little impact on the rents in the area.

However, these summer listings may also show a landlord using private rents during the cooler months and short term lets over summer. This would have a double impact on tenants - the frst set of tenants has to move, and when the dwelling is returned to the residential sector after summer the second set of tenants will be given a higher rent.

Without data about evictions in Sydney, it is impossible to know the extent to which these scenarios occur.

Belonging anywhere: Airbnb and renting in Sydney Page 8 Covering the rent

We used the AirDNA data to examine whether the incentive existed for a landlord to remove a property from the private rental market and move it to Airbnb.

Our analysis shows that across Sydney, the number of nights listed on Airbnb required to cover the median rent for the same size property in the same postcode is 194 nights each year. Of the 10 most populated airbnb areas in September 2016, the number of |days needed to cover the rent was in excess of 150 in every case. There are a few areas with very low Airbnb activity where the number required is below 60 days, but equally there are several areas where the number is well in excess of 365 days.

For both rents and listings income, we haven’t taken into account the costs associated. A busy Airbnb listing may come with higher costs than a private rental. Hosts may need to account for a Nights needed to cover the rent faster rate of wear and tear, as well as having to provide furniture, linen and other necessities.

The full interactive map is available at: https://tenants.org.au/tu/ AirbnbSydney2017

Belonging anywhere: Airbnb and renting in Sydney Page 9 Different property types do fnd it Entire home Airbnbs with 1 or more easier to cover the rent than others. nights booked, by number Where we look at the nights needed of bedrooms and nights needed to cover the rent by the number of to cover median rent bedrooms, it becomes very clear that Nights to larger properties fnd it much easier Airbnbs Beds cover to cover the rent. rent The majority of listings are for dwellings with 2 or less bedrooms. Of the nearly 6,540 208 15,000 entire home listings created that had any activity in the year to 5,217 190 September 2016, more than three 1,774 157 quarters (11,757) were studios, one bedroom, or two bedroom dwellings. 911 131

Of these smaller dwellings, the number 334 97 of nights needed to cover the rent is signifcantly higher than for larger bedroom properties. This means that a property owner may well be able to see beneft in utilising Airbnb over Entire home Airbnbs rented for residential tenancies in larger houses, greater than 60 days in the year, by though we note the median nights number of bedrooms and nights needed are still signifcant - 93 nights needed to cover median rent a year even for 5 bedroom properties. Nights to

A second strategy may also present Airbnbs Beds cover itself to an Airbnb investor. With a larger property, an investor can also rent list individual rooms at the same time as an entire home, so they can be 1,996 196 much more fexible with the revenue. 1,252 173 Looking at a bedroom breakdown of 348 158 those listings which actually are being 124 136 used for a signifcant number of nights a year, we can see that part of the 38 106 reason why these properties are being used more is that they have less of a burden to cover the same level of income as from a residential tenancy. The days needed are still signifcant, and the numbers of properties that reach even 60 days a year is small.

Belonging anywhere: Airbnb and renting in Sydney Page 10 It’s not easy covering the rent with Airbnb. Indeed most people just aren’t trying.

When we look at how many days a year it would take to cover the rent, we can see that the vast majority of listings have been booked for less than 60 nights in the year. In fact more than a quarter of listings that “It’s not easy covering had any activity in the year were booked for 10 nights or less. the rent with Airbnb. Very few properties with 60 or less Indeed most people booked nights actually made more than the median rent for that property just aren’t trying.” size in that postcode. It is much easier to do with larger properties as the price can vary.

It is not until we get to more than 100 booked nights a year that a signifcant proportion of properties are making more than median rent. This suggests that there is a cohort of hosts using the platform commercially, but the vast majority of users are not.

Does the listing make No more than the rent? Yes

Less than 60 booked nights per year

• 74.9% of entire More than home listings, 200 booked Between 60 and 100 • 0.1% of these list- nights per booked nights per ings make more than year. median rent year • 4.0% of • 10.9% of entire entire home home listings, listings, • 3.7% of these list- • 81.3% of ings make more than Between 100 and 200 booked these listings median rent nights per year make more than median • 13.0% of entire home list- rent ings, • 27.0% of these listings make more than median rent

Number of days booked in the last 12 months

Belonging anywhere: Airbnb and renting in Sydney Page 11 Are rents rising because of airbnb?

In trying to work out whether the market. In a balanced rental market presence of Airbnb has had an impact we believe we should be looking for on rents, we frst looked to see if it has at least a 5% vacancy rate.1 had an impact on supply of housing. If landlords were taking rentals off First we have taken the active the market for Airbnb we should see listings from the previous pages a drop in the vacancy rate. In these 6-7 and compared their growth to numbers, we are looking for a the vacancy rate for all of Sydney negative correlation – that is, when for the period August 2014 to August the listing numbers go up, the 2016, and for the postcode areas vacancy rate goes down. with greatest density of Airbnb. For all of Sydney there was no The vacancy rate measures the discernible effect on the vacancy rate. number of properties currently In fact there is a very weak positive available to rent at any one time. It is 1 Eric S. Belsky (1992) Rental vacancy rates: A generally recognised as one of the key policy primer, Housing Policy Debate, 3:3, 793-813, indicators of the health of a tenancy DOI: 10.1080/10511482.1992.9521110

Belonging anywhere: Airbnb and renting in Sydney Page 12 relationship. However, given how few active listings operate, and how concentrated Airbnb in Sydney is to the central and coastal areas this probably isn’t so surprising.

Looking at the three previously identifed areas, we start to get a better picture. For both Bondi and Manly, we can see a defnite seasonal effect on their vacancy rate with winter peaks and summer troughs. While active listings do follow the reverse of this seasonal effect, there appears to be very little impact on the vacancy rate. Manly had the strongest negative correlation, but this was still very weak at r= -0.26. Other correlations were between 0.1 and -0.1.

Belonging anywhere: Airbnb and renting in Sydney Page 13 Selected postcode vacancy rates compared to the Greater Sydney area (Greater Sydney = 0).

Active Airbnbs (forecast from September 2016)

The vacancy rate appeared not to have Sydney over the year to September been affected by Airbnb listings but 2016, and additionally the periods when we look at the longer trends we 2006-2011 and 2011-2016. We can see that compared to all of Sydney, found no link between particularly these areas have become increasingly high numbers of Airbnb listings tight over the last 12 years. and rises in rent.

For some areas this may be due to a Some of these areas did rise faster growing rental demand as house prices than the average across Sydney in take owner occupation out of reach, but the last year. However, a signifcant supply of rental properties hasn’t been number rose much more slowly than able to keep up. the average across Sydney, suggesting that Airbnb and other The rent is getting higher, short stay accommodation have not had a signifcant or consistent effect but how much? on rents in the area.

Finally, we look at the actual prices. Additionally, some areas saw rents rise The rent in Sydney continues to rise, comparatively faster in the period 2006- but given our observations so far, does 2011 than in the period 2011-2016. Airbnb appear to have had an impact Given the very small number of Airbnb on the rent? We examined the top 10 listings active in this period, this throws postcodes by the number of airbnb further doubt upon the specifc impact listing for each of 1, 2 and 3 bedroom of Airbnb may have had on rents in the entire home properties. The 1, 2 and 3 current period. bedroom listings in these 13 postcode areas account for 9951, or about a Particularly when combined with the third, of all entire home listings on low correlation to vacancy rates, it Airbnb across Sydney. seems unlikely that the prices of new rents went up in these areas more We compared those postcodes for the than they might have in the absence average increases in median rent for of Airbnb.

Belonging anywhere: Airbnb and renting in Sydney Page 14 Increases in median rents in Airbnb hotspots

2015-2016 2006-2011 2011-2016

Increase in Compared Increase in Compared Increase in Compared 1 bedrooms new rents to sydney new rents to sydney new rents to sydney Sydney Average 6.8% average 7.6% average 4.0% average 2107 - Avalon-Bilgola 12.2% 5.4% 4.5% -3.1% 5.9% 1.9% 2026 - Bondi 10.0% 3.2% 10.0% 2.4% 4.4% 0.4% 2030 - Vaucluse-Watsons Bay -10.4% -17.2% - - - - 2024 - Bronte-Waverley 14.6% 7.8% 10.6% 3.0% 5.0% 1.0% 2023 - Bellevue Hill 0.0% -6.8% 7.4% -0.2% 1.6% -2.4% 2095 - Manly 8.6% 1.8% 7.3% -0.3% 5.3% 1.3% 2034 - Coogee 4.2% -2.7% 8.1% 0.5% 3.0% -1.0% 2025 - Woollahra 7.3% 0.5% 10.6% 3.1% 1.7% -2.3% 2010 - Darlinghurst-Surry Hills 4.8% -2.1% 7.9% 0.4% 3.2% -0.8% 2029 - Rose Bay 8.9% 2.1% 9.3% 1.7% 3.1% -0.9%

2 bedrooms Sydney average 2.0% 10.0% 3.6% 2107 - Avalon-Bilgola -0.9% -2.9% 5.2% -4.8% 4.6% 1.0% 2026 - Bondi 6.6% 4.6% 11.6% 1.6% 4.2% 0.6% 2024 - Bronte-Waverley 4.6% 2.6% 5.9% -4.1% 4.7% 1.2% 2095 - Manly 14.3% 12.2% 9.0% -1.0% 5.4% 1.8% 2010 - Darlinghurst-Surry Hills 3.8% 1.7% 8.9% -1.1% 3.2% -0.3% 2025 - Woollahra 0.0% -2.0% 10.5% 0.5% 1.9% -1.7% 2021 - Paddington 1.4% -0.6% 9.3% -0.7% 2.5% -1.0% 2028 - Double Bay 3.7% 1.7% 7.6% -2.4% 4.1% 0.6% 2027 - Edgecliff-Point Piper 3.3% 1.3% 10.6% 0.6% 0.7% -2.9% 2034 - Coogee 4.8% 2.8% 9.7% -0.3% 3.6% 0.1%

3 bedrooms Sydney average 1.1% 10.0% 3.1% 2024 - Bronte-Waverley -7.1% -8.2% 8.9% -1.1% 0.0% -3.1% 2026 - Bondi 15.8% 14.7% 7.2% -2.8% 7.0% 3.9% 2107 - Avalon-Bilgola 6.9% 5.9% 5.7% -4.3% 4.3% 1.2% 2025 - Woollahra -25.9% -27.0% 6.7% -3.3% -2.0% -5.1% 2010 - Darlinghurst-Surry Hills 0.1% -1.0% 9.0% -1.0% 3.5% 0.4% 2031 - Randwick-Clovelly -1.1% -2.2% 10.1% 0.1% 3.5% 0.3% 2034 - Coogee 3.7% 2.7% 10.9% 0.9% 2.1% -1.0% 2095 - Manly 8.3% 7.3% 6.8% -3.2% 6.8% 3.7% 2041 - Balmain-Birchgrove -3.7% -4.8% 6.8% -3.2% 9.8% 6.7% 2029 - Rose Bay 50.0% 48.9% 14.5% 4.5% 6.8% 3.7%

Postcodes chosen are sorted by most active Airbnb areas for each bedroom size property. Each increase fgure shows the percentage increase in the price of new rents in the area over the relevant period. The increase is expressed as a yearly average for 5 year periods. The second column in each period shows the difference between the Greater Sydney average for the property type and the particular postcode.

Belonging anywhere: Airbnb and renting in Sydney Page 15 Conclusions and recommendations

We acknowledge the impact the Criticism of Airbnb exposes some of growth in short term letting has had the issues facing tenants in Sydney on communities in other parts of the and NSW generally - lack of security for world, but from our investigation the residential tenants, no practical cap on same does not appear to be true here. rising rents and limited ability to remain and set down roots in communities. In asking why this might be, we offer two observations. It is worth noting in passing that lower income renters in Sydney’s The frst may be obvious - even private market have mostly already if Australia is the ‘most penetrated’ been moved out of the areas where market, there is actually not yet so Airbnb is located. The almost total much activity on Airbnb that would absence of Airbnb, and particularly impact the private rental market. more commercially run Airbnbs, in This may change, and if it does the west of Sydney points to a then it means now is a good time continued concentration of property to establish a sensible regulatory wealth in the hands of the people regime. It means we can prevent who already have access to the impact on communities that properties in high-demand areas – other cities have reported. whether they are owned or rented. The second is less obvious. Australia has some of the weakest tenancy laws in the world, with landlords facing little to constrain their ability to maximise “Australia has some rents and capital gains, and directly subsidised by government in a number of the weakest of ways1. Compare this to New York with its million-strong cohort of rent controlled homes2, or San Francisco tenancy laws in and Europe with sensible restrictions on evictions. A landlord in Sydney the world, with may not perceive the same beneft in removing the property from the private landlords facing rental sector in order to utilise Airbnb as would a landlord in those other areas. little to constrain their ability to 1 Tenants’ Guide to Tax Reform, TUNSW, 2016. https://fles.tenants.org.au/policy/Tax-Reform- Guide-2016 maximise rents.” 2 Rent Stabilization in New York City Furman Centre 2012, http://furmancenter.org/research/publication/ rent-stabilization-in-new-york-city

Belonging anywhere: Airbnb and renting in Sydney Page 16 We support:

1. Further study into tenant evictions 5. A tenure neutral approach to in Sydney and NSW generally. regulation of short-term lettings. These studies should particularly Tenants experience signifcant extra include an examination of tenants hurdles to enjoy their homes in the who are being evicted for changing same way owner occupiers do. use of the premises to short term If Airbnb is accepted as a legitimate lettings. We simply do not know the use of premises for occasional reasons tenancies are ended in NSW personal use, then there should be apart from anecdotal evidence and no restriction placed upon a tenant occasional surveys. by landlords or strata corporations on tenants use of Airbnb which is not 2. Further study into the long term equally applied to owner-occupiers. impact of Airbnb on the way Sydney houses itself. We particularly encourage academia to negotiate the release of data directly from Airbnb themselves. “Now is a good

3. Ensure planning laws at least limit time to establish a the use of short term lets above 60 nights per year. sensible regulatory We are satisfed from the data presented here and previously that regime. It means commercial operators that may cause an issue for the supply of private rents we can prevent operate well above a 60 night per year level. We do not believe those using the impact on Airbnb or other sites a handful of nights a year are having any meaningful communities that impact on housing supply or skewing local markets. other cities have

4. The amendment of no grounds reported.” terminations to a list of reasonable grounds for termination in the Residential Tenancies Act 2010. This will prevent landlords from evicting tenants to vacillate between short term lettings and residential tenancy lettings, and allow an examination of the reasons given by the landlord.

Belonging anywhere: Airbnb and renting in Sydney Page 17 A note on Methodology

Sources Airbnb disputes the accuracy of the information derived from scrapes, but does We used a variety of sources for this not generally make data publicly available. report. Data concerning Airbnb’s listings Airbnb has since made an offer to provide in Sydney is based on AirDNA property data but this was not able to be included in listings to the end of September 2016. this report. AirDNA is a third-party frm that provides data and analytics on Airbnb listings and Information regarding private rental rates bookings. AirDNA is the only source of and bonds held by the Rental Bond Board information for short-term rental data are taken from the Rent and Sales Report which includes occupancy rates and published by Family and Community revenue data. Services. This records the number of bonds held, which is an undercount of We have also utilised data from private renters as not all renters pay InsideAirbnb, an explicitly activist project a bond. Rents recorded on the bond examining the publicly available data lodgment are the most accurate measure from Airbnb website. of the rent paid by tenants at the It is important to note that occupancy rates beginning of their tenancy. are modeled from direct observations of Information about vacancy rates in the listings1,2. We do not consider this to be Sydney were obtained from SQM Research an issue for the purposes of this report. Pty Ltd at: http://www.sqmresearch.com. Both of these sources create their data au/vacancy.php. through ‘scrapes’ of the Airbnb website, Department of Planning and Environment pulling down information about the listings (2014), 2014 New South Wales State and in an unauthorised manner. Local Government Area population and 1 https://www.airdna.co/methodology household projections. Sydney, NSW. 2 http://insideairbnb.com/about.html

Acknowledgments Disclaimer The Tenant’s Union of NSW has taken care in reporting accurately and providing analysis based on the publications of others. If an error exists, please get in touch below. This report was made possible with fnancial assistance from Shelter NSW, the Information on this publication state’s peak advocate for housing justice. remains the copyright of the Tenants’ Union of NSW, unless The Bélo symbol and Airbnb logo are stated otherwise. the property of Airbnb Inc. Use of these images is considered fair use and does not For further information about this indicate sponsorship or association with publication contact the Tenants’ Airbnb Inc. Union of NSW: (02) 8117 3700, Selected icons used in this report designed [email protected] by Freepik.com

Belonging anywhere: Airbnb and renting in Sydney Page 18 WHAT IMPACT DOES AIRBNB HAVE ON THE SYDNEY AND MELBOURNE HOUSING MARKETS?

PREPARED FOR Airbnb © SGS Economics and Planning Pty Ltd 2018

This report has been prepared for Airbnb with analysis of Airbnb and industry data. SGS Economics and Planning has taken all due care in the preparaton of this report. However, SGS and its associated consultants are not liable to any person or entty for any damage or loss that has occurred, or may occur, in relaton to that person or entty taking or not taking acton in respect of any representaton, statement, opinion or advice referred to herein.

SGS Economics and Planning Pty Ltd ACN 007 437 729 www.sgsep.com.au Ofces in Canberra, Hobart, Melbourne, Sydney Contents

Executve Summary 4

01 Introducton 6

02 Australian Housing Market 10

03 Sydney Housing Market 15

04 Melbourne Housing Market 21

05 Airbnb in Sydney 27

06 Airbnb in Melbourne 36

Conclusions: Airbnb in context 45

References 52 Executive Summary

Over the last two years, Airbnb has increasingly become SYDNEY AND MELBOURNE HOUSING MARKETS AIRBNB IN SYDNEY AND MELBOURNE the subject of debate regarding its impact on the housing market in Australia, partcularly in Sydney and Melbourne. Over the last decade, home ownership rates have declined Across Sydney and Melbourne, Airbnb listngs represent This debate spans media, private sector and university while rental rates have increased. Over the same period, a very small proporton of the total dwelling market. In research and is bundled in with a broader discussion around the proporton of Greater Sydney households living in 2017, there were approximately 25,000 listngs that hosted deterioratng housing afordability. To date, there has been detached houses has declined, while the proporton of guests in Sydney and 14,500 in Melbourne. Of these, 15,200 no research undertaken in Australia which has had access households living in apartments has increased. This growth were entre homes/ apartments in Sydney and 8,500 in to detailed Airbnb data. SGS Economics and Planning has in apartment dwellings has largely been clustered around Melbourne. This represents 0.9 per cent of all dwellings in been commissioned by Airbnb to analyse host data to beter inner city locatons, along public transport corridors, while Sydney and 0.5 per cent of all dwellings in Melbourne (refer understand the impact of Airbnb on the housing market in detached housing growth has occurred across the outer to Table 1). Sydney and Melbourne. suburbs. House and unit prices have rapidly increased across Greater Sydney over the last decade. Rents have increased TABLE 1: AIRBNB LISTINGS AS A PROPORTION OF TOTAL Australia’s populaton has grown rapidly over the last two at a slower rate compared to house prices, refectng trends DWELLINGS decades, with a large percentage of this growth being observed more broadly across Australia. However, in some GREATER GREATER concentrated in Greater Sydney and Greater Melbourne. locatons, rents have doubled. SYDNEY MELBOURNE Since 1990, dwelling prices have grown much faster than Total Dwellings 1,775,700 1,597,400 incomes as well as rental prices. There are a number of Similarly, home ownership rates have declined in Melbourne factors which have contributed to an increase in demand for while rental rates have substantally increased. Again, the Number of listngs that 25,000 14,500 housing and rising prices including strong economic growth, proporton of Greater Melbourne households living in hosted guests falling interest rates, the geographic shif of jobs, tax and detached houses has declined, while the proporton of Listngs that hosted 15,200 8,500 welfare setngs that encourage investment in housing. households living in apartments has increased. Apartment guests (entre homes/ growth has primarily been concentrated in Melbourne apartments) City, while detached housing growth has similarly been Listngs as a proporton 0.9% 0.5% concentrated in the outer suburbs. House prices have rapidly of all dwellings increased across Greater Melbourne over the last decade. However, median rents have remained relatvely stable for Source: SGS Economics and Planning, 2018 using Airbnb, 2018 and both houses and apartments. Australian Bureau of Statstcs, 2016

What impact does Airbnb have on the Sydney and Melbourne housing markets? 4 By way of comparison, the 2016 Census identfed 136,000 In 2017, of the listngs that hosted guests, the median nights Based on this analysis, the impact of Airbnb on the housing unoccupied private dwellings across Greater Sydney (7.3 hosted per listng in 2017 was 20 nights across Greater markets in Sydney and Melbourne appears to be minimal. per cent of all dwellings) and 167,500 unoccupied private Sydney and 66 nights across Greater Melbourne. In Sydney, The housing markets in Sydney and Melbourne have dwellings across Greater Melbourne (9.1 per cent of all almost 60 per cent of listngs host guests for between one experienced substantal growth in house prices, and to a dwellings)1. and 30 nights per year and only 10 per cent host guests for lesser extent rental prices, over the last two decades. This more than 180 nights per year. In Melbourne, over 35 per growth has been driven by a number of factors including The majority of listngs on Airbnb are entre homes or cent of listngs host guests for between one and 30 nights populaton growth, interest rates and tax incentves which apartments2. Airbnb listngs are spread across Greater Sydney per year and approximately 27 per cent of listngs host guests drive investment. and Greater Melbourne, however there is a concentraton for more than 180 nights per year. in the eastern half of Sydney, near the coast and CBD, and in Airbnb has grown in popularity over the last fve years. the inner city of Melbourne. AIRBNB IN CONTEXT However, Airbnb listngs are only likely to be rented 50 per cent of the tme that they are available, and it is considered The average nightly price for an entre home/ apartment Broadly, in Sydney and Melbourne, it is not fnancially to be more fnancially stable for a property owner to list a on Airbnb in Greater Sydney is $150, compared to $59 for a benefcial to host a property on Airbnb instead of rentng to a property on the long term rental market than on Airbnb. It private room and $39 for a shared room. This compares to long term tenant. In only a small number of cases, it is more is likely that Airbnb has a greater role in addressing a gap in $125 for an entre home/ apartment in Greater Melbourne, proftable to list a property of Airbnb full tme rather than on the tourist accommodaton market, more so than any impact compared to $49 for a private room and $34 for a shared the rental market. With the excepton of Melbourne City, it of the availability of rental accommodaton in Sydney and room. appears that rentng has higher expected takings over a tme Melbourne. period of a standard lease agreement. This does not take into The number of trips (i.e. guests staying in Airbnb propertes) consideraton the additonal costs associated with hostng on has increased, partcularly in Sydney Inner City3 and Airbnb including furnishing the listng, utlites and cleaning Melbourne City4. While popularity has increased, average which, once considered, would make hostng on Airbnb less nightly prices have remained steady. Median earnings of fnancially advantageous. Airbnb hosts has remained relatvely stable across Greater Sydney and slightly declined across Greater Melbourne. Hotel accommodaton occupancy rates are high, partcularly in Sydney and Melbourne, which potentally suggests there The hostng rate of Airbnb listngs in Greater Sydney and is a gap in supply across the tourism accommodaton market. Greater Melbourne are less than 50 per cent, on average. Airbnb nightly prices are, on average, cheaper than hotel This means that each listng has a less than 50 per cent prices. chance of being booked compared to availability.

1See htps://www.sgsep.com.au/publicatons/why-was-no-one-home-census-night 2An entre home/ apartment includes all fully self-contained spaces e.g. homes, apartments (incl. studios) and granny-fats This includes both primary places of residence as well as non-primary places of residence (such as a holiday home or secondary dwelling). 3Sydney Inner City (SA3 11703) incorporates Sydney CBD, Pots Point, Darlinghurst, Ultmo, Glebe, Camperdown, Chippendale, Surry Hills, Waterloo, Alexandria, Zetland and Rosebery 4Melbourne City (SA3 20604) incorporates Melbourne CBD, Southbank, East Melbourne, Carlton, North Melbourne, Parkville, Kensington and Flemington.

What impact does Airbnb have on the Sydney and Melbourne housing markets? 5 Introduction

Over the last two years, Airbnb has increasingly become the subject of debate regarding its impact on the housing market in Australia, particularly in Sydney and Melbourne.

What impact does Airbnb have on the Sydney and Melbourne housing markets? 6 CHAPTER 1: INTRODUCTION

RISE OF AIRBNB BENEFITS AND IMPACTS SGS Economics and Planning (SGS) has been commissioned ― Councils outside Greater Sydney having the power to by Airbnb to analyse host data to beter understand the decrease the 365 day threshold to no lower than 180 Over the past two decades there has been a debate around impact of Airbnb on the housing market in Sydney and days per year the causes and efects of deterioratng housing afordability Melbourne. This report focuses on identfying: ― Certain planning rules will apply to propertes on in Australia. Over the last two years, Airbnb has increasingly bushfre prone land; and become the subject of debate regarding its impact on the ― The share of propertes that are on Airbnb. ― Rules around neighbourhood amenity, including a housing market in Australia, partcularly in Sydney and ― The share of propertes which might not be available for mandatory Code of Conduct8. Melbourne. This debate spans media, private sector and private rental. university research. ― The impact on rent and home prices in areas with large This new policy has not yet been implemented. number of Airbnb listngs. The major argument is around the impact of Airbnb on ― The relatonship between additonal housing supply and THIS REPORT private housing market5 with concern that it is increasing the the presence of Airbnb. incentve for investment, reducing availability of long term ― The relatonship between the supply of Airbnb and hotel This report is structured as follows: rental propertes and pushing up rental prices. These debates and other forms of tourist accommodaton. have been observed worldwide. ― Secton 2 summarises some of the high level factors This research is tmely as the governments across Australia impactng both the rental and ownership markets in More recently, there has been discussion as to the profle of are considering policy responses to the rise of Airbnb. Australia over the last 10-15 years hosts, with concern that Airbnb hosts are property managers On 5 June 2018, the NSW Minister for Planning and the ― Secton 3 and Secton 4 explore some of the trends in with a portolio of propertes making signifcant profts6. Minister for Beter Regulaton and Innovaton announced both the rental and ownership markets in Sydney and new regulatons for short-term holiday letng across NSW. Melbourne over the last 10 to 15 years. Research contnues to focus on the potental impact of The new regulatons which are expected to be introduced ― Secton 5 and Secton 6 analyse Airbnb host data for Airbnb7, however this research has involved scraping of include: Greater Sydney and Greater Melbourne including the Airbnb website which has its limitatons. Scraped data number of listngs and where they are concentrated, includes Airbnb listngs that have never hosted guests and ― Allowing short-term holiday letng as exempt average prices and occupancy rates duplicate listngs. To date, there has been no research development 365 days per year when the host is present ― Secton 7 draws together the analysis of Airbnb data and undertaken in Australia which has had access Airbnb data. ― When the host is not present, a limit for hosts to rent trends in the housing market to beter understand the out propertes via short-term holiday letng of 180 days role of Airbnb in the housing market. in Greater Sydney, with 365 days allowed in all other areas of New South Wales

5See htp://insideairbnb.com/sydney/ and htp://insideairbnb.com/melbourne/ 6See htps://www.domain.com.au/news/australias-top-airbnb-host-made-53-million-in-the-last-year-20171124-gzsgg4/ and htp://mobile.abc.net.au/news/2018-02-23/entre-homes-commercial-listngs-have- surged-on-airbnb/9473368?pfmredir=sm&sf182827679=1 7See htps://cityfutures.be.unsw.edu.au/research/projects/technological-disrupton-private-housing-markets-case-airbnb/ 8NSW Department of Planning and Environment, 2018, htps://www.planning.nsw.gov.au/News/2018/Short-term-holiday-letng-plan-a-win-win

What impact does Airbnb have on the Sydney and Melbourne housing markets? 7 QUALIFICATIONS This research has focused exclusively on Greater Sydney and Greater Melbourne. This analysis has not covered propertes which are rented on HomeAway (formerly Stayz) or other sharing platorms and This analysis has been undertaken using data provided directly to SGS by Airbnb. does not capture other Australian cites. There are a number of unique terms related to this data that are referenced throughout this report. These are defned below. The report focuses on the trends across two SA3s9: Sydney Inner City and Melbourne City due to the concentraton of listngs in these locatons. These SA3s closely represent the inner city of both metropolitan areas: TERM DEFINITION Entre home/ All fully self-contained spaces which includes entre homes, ― Sydney Inner City (SA3 11703) incorporates Sydney CBD, Pots Point, Darlinghurst, Ultmo, Glebe, apartment entre apartments including studios and granny-fats. This Camperdown, Chippendale, Surry Hills, Waterloo, Alexandria, Zetland and Rosebery (refer to includes both primary places of residence as well as non- Figure 1). primary places of residence (such as a holiday home or ― Melbourne City (SA3 20604) incorporates Melbourne CBD, Southbank, East Melbourne, Carlton, secondary dwelling). North Melbourne, Parkville, Kensington and Flemington (refer to Figure 1). Trip A booking that has occurred. Not all bookings turn into trips due to cancellatons. Guest A person who stays at an Airbnb listng regardless of whether they’re an Airbnb account holder or not. Host An Airbnb user account that ofers accommodaton to guests. There are more listngs than hosts as hosts can have multple listngs, but a listng can only be ofered by a single host. Listng A single space ofered by a host. This can include entre homes or apartments, private rooms or shared rooms. Unique Hosts The number of Airbnb host user accounts that hosted any who Hosted number of guests for any number of nights. Unique Listngs The number of Airbnb listngs that accommodated any that Hosted number of guests for any number of nights. Guests Host earnings The total amount the host receives (excludes the service fee charged by Airbnb). Nights The total number of nights listngs had any number of guests. Average nightly Average over all listngs of the total paid by the guests for price trips.

9Statstcal Areas Level 3 (SA3) are geographical areas defned by the Australian Bureau of Statstcs. They generally have a populaton of between 30,000 and 130,000 people.

What impact does Airbnb have on the Sydney and Melbourne housing markets? 8 Each SA3 is unique and has a diferent narratve around the growth of Airbnb. There may be small pockets within each SA3 where the averages might not apply (ecological fallacy10).

Naturally, the analysis of data will not capture any issues relatng to the Airbnb guests and their potental impact on neighbours. These issues are the subject of proposed restrictons to be introduced across NSW and Victoria.

In NSW, the Government is seeking to introduce a regulatory framework to address instances of ant-social behaviour which impact on neighbourhood amenity by establishing a Code of Conduct. A “two-strikes-and-you’re-out” system is proposed11. In Victoria, the Government has recently established laws for strata buildings which includes a “three-strikes-and-you’re-out” system12.

FIGURE 1: SA3 BOUNDARIES FOR SYDNEY INNER CITY AND MELBOURNE CITY

Source: SGS Economics and Planning, 2018

10Ecological fallacy relates to the interpretaton of statstcal data where inferences about the nature of individuals are deduced from inference for the group to which those individuals belong. 11Refer to Fair Trading Amendment (Short-term Rental Accommodaton) Bill 2018 12Refer to Owners Corporatons Amendment (Short-stay Accommodaton) Act 2018

What impact does Airbnb have on the Sydney and Melbourne housing markets? 9 Australian Housing Market

Since 1990, dwelling prices have grown much faster than incomes.

Dwelling prices have also grown much faster than rental prices.

What impact does Airbnb have on the Sydney and Melbourne housing markets? 10 CHAPTER 2: AUSTRALIAN HOUSING MARKET

This secton summarises some of the factors impactng FIGURE 2: ANNUAL POPULATION GROWTH, GREATER SYDNEY AND GREATER MELBOURNE, 2002-2017 both the rental and ownership markets in Australia over the last 10-15 years including populaton growth, 140,000 change in house prices and rental prices and broader drivers of housing demand.

Figure 3 (overleaf) outlines some of the factors infuencing 120,000 housing demand, supply and afordability was adapted by the Natonal Housing Supply Council from a report prepared by the Productvity Commission. This highlights the complex and 100,000 dynamic nature of housing, demand, supply and afordability. Some of these factors are discussed below. 80,000 AUSTRALIA’S POPULATION HAS GROWN RAPIDLY OVER THE LAST TWO DECADES.

MUCH OF THIS GROWTH HAS BEEN Popula�on 60,000 CONCENTRATED IN GREATER SYDNEY AND GREATER MELBOURNE. 40,000 Underlying housing demand is largely a functon of populaton growth and demographic change. Populaton growth consists of natural increase, net overseas migraton 20,000 and net interstate migraton.

Australia’s populaton has grown from 19.3 million people in - - 2001 to 24.6 million people in 2017 (an additonal 5.3 million people in 16 years, or 330,000 people per annum). 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 Much of this growth has been concentrated in Greater Sydney and Greater Melbourne. During this period, Greater Greater Sydney Greater Melbourne Sydney has accommodated an additonal 1.0 million people, and Greater Melbourne has accommodated an additonal Source: SGS Economics and Planning, 2018 using Australian Bureau of Statstcs, 2017 1.3 million people. Annual populaton growth has been increasing (refer to Figure 2).

What impact does Airbnb have on the Sydney and Melbourne housing markets? 11 FIGURE 3: SOME FACTORS INFLUENCING HOUSING DEMAND, SUPPLY AND AFFORDABILITY

AFFORDABILITY

Finance costs Prices Quantty Type/ Quality

DEMAND SUPPLY

Demographics (number and type of households) Existng dwellings New dwellings

Economic circumstances Constructon costs of households (income, (labour, materials) employment, etc).

Taxes and Concessions (eg Investor demand (return on GST, First Home Owners Infrastructure costs alternatve investments) Scheme, Stamp Duty) (water, sewerage)

Consumer preferences Land availability (size, quality, locaton) (geography, zoning)

Land release and Rental prices and availability development processes including fees and regulaton

Source: Productvity Commission and NHSC, 2009

What impact does Airbnb have on the Sydney and Melbourne housing markets? 12 SINCE 1990, DWELLING PRICES HAVE DWELLING PRICES HAVE ALSO GROWN THERE ARE A NUMBER OF FACTORS WHICH GROWN MUCH FASTER THAN INCOMES. MUCH FASTER THAN RENTAL PRICES. HAVE CONTRIBUTED TO AN INCREASE IN Over the past 20 years, dwelling prices across Australia As illustrated in Figure 5, rents have increased broadly in DEMAND FOR HOUSING AND RISING PRICES. have risen much faster than incomes (refer to Figure 4). line with wages. Dwelling prices have risen faster than rents Housing need and afordability arise from the complex The Gratan Insttute (Daley, Coates, and Wiltshire 2018) due to low interest rates and tax changes which have made interacton of economic growth and demographic efects has identfed that rising dwelling prices are primarily due owning dwellings more atractve (Daley et. al 2018). (populaton and household growth) with the existng housing to rising land values, not constructon costs. Dwelling prices supply and the capacity of the housing market to respond to have grown much faster in areas closer to city centres due to FIGURE 5: HOUSING PRICES VS RENTS, 1997-2017 changing demand. the accessibility of these locatons to employment. 450 FIGURE 4: DWELLING PRICES VS FULL TIME EARNINGS, 1970- Housing Figure 3 illustrates some of the factors infuencing housing prices 2017 400 demand, supply and afordability. Additonal factors afectng housing afordability could also include: 350 400 300 ― Travel tme and the economic, social and environmental 350 costs related to accessing work, community services and 250 300 recreaton, 200 Rents ― Recurrent costs of operatng and maintaining the 250 Dwelling prices Wages housing, and CPI 200 150 ― The recurrent costs of housing infrastructure and housing-related service provision. 150 100 ee te enn 100 50 There is signifcant interplay between the demand and 1997 1999 2001 2003 2005 2007 2009 2011 2013 2015 2017 50 supply side of housing markets. Changing needs and tastes Source: Daly et al., 2018 can shape the type and quality of dwellings that are built. 0 Notes: Nominal index 1997 = 100. Nominal house price growth from The existng stock also infuences aspiratons and hence 1970 1975 1980 1985 1990 1995 2000 2005 2010 2015 Bank for Internatonal Setlements, 2017, ABS 2018 and ABS 2017. peoples’ willingness to pay for diferent types of housing and Wage price index excludes bonuses. Rents in the CPI are stratfed Source: Daly et al., 2018. according to locaton, type and size. for housing in diferent locatons. New supply is assumed to Real dwelling prices and full-tme weekly earnings index 1970 = 100. refect demand, but industry economics and government Data for 1970 to 2010 is from Yates, 2011. Data from 2010 is six incentves can also infuence the type of new housing monthly growth in residental property price index from ABS 2018, supplied. defated by the CPI. Earnings data is full tme ordinary tme earnings from ABS, 017, defated by CPI.

What impact does Airbnb have on the Sydney and Melbourne housing markets? 13 Similarly, Daley et. al (2018) have identfed factors that have ADDITIONALLY, THE NUMBER OF VACANT increased demand for housing and contributed to rising DWELLINGS OR UNOCCUPIED DWELLINGS house prices: HAS NOT INCREASED. ― Strong economic growth and rising household incomes Since the release of the 2016 Australian Census of Populaton ― Falling interest rates and more readily available credit and Housing data, the number of unoccupied dwellings ― The geographic shif of jobs – and therefore demand recorded on Census night has atracted a great deal of for accommodaton – towards large capital cites, atenton13. On Census night there were over one million partcularly their centres dwellings unoccupied across Australia. This is 11.2 per ― Tax and welfare setngs encouraging home-ownership, cent of all dwellings. Investors intentonally leaving homes including frst-home buyers assistance, and the exclusion unoccupied in capital cites and the rise of Airbnb were of family homes from capital gains tax, Age Pension blamed by some for this large number of unoccupied homes. assets test, and state land tax ― Tax setngs encouraging investment in housing, However, analysis by SGS Economics and Planning14 partcularly the capital gains tax discount and negatve identfed that the 11.2 per cent of unoccupied dwellings is gearing arrangements only 0.5 percentage points higher than the 2011 Census. The ― Tax setngs that discourage downsizing percentage of unoccupied dwellings has been trending up ― High rates of immigraton since 2001, but unoccupied dwellings have consistently made ― More foreigners responding to global economic factors up around 10 per cent of dwellings over the past 35 years. by investng in Australian housing ― A self-reinforcing cycle of increased prices and leverage. Unoccupied dwellings were also much more common in regional areas than in inner city locatons. For example, regional parts of South Australia and Tasmania, had over 30 per cent of dwellings unoccupied on Census night. While inner Sydney and Melbourne have rates below of 7.3 per cent and 9.1 per cent respectvely.

13htps://architectureau.com/artcles/one-in-six-new-homes-unoccupied-2016-census-suggests/ 14See htps://www.sgsep.com.au/publicatons/why-was-no-one-home-census-night

What impact does Airbnb have on the Sydney and Melbourne housing markets? 14 Sydney Housing Market

Rents have increased at a slower rate compared to house prices, reflecting trends observed across Australia.

What impact does Airbnb have on the Sydney and Melbourne housing markets? 15 CHAPTER 3: SYDNEY HOUSING MARKET

This secton explores some of the trends in both the THIS GROWTH IN APARTMENT DWELLINGS rental and ownership markets in Sydney over the HAS LARGELY BEEN CLUSTERED AROUND last 10-15 years including trends in dwelling mix and INNER CITY LOCATIONS, NEAR PUBLIC development, ownership and rental prices and trends TRANSPORT CORRIDORS, WHILE DETACHED in home ownership. HOUSING GROWTH HAS OCCURRED ACROSS THE OUTER SUBURBS. THE PROPORTION OF GREATER SYDNEY HOUSEHOLDS LIVING IN DETACHED HOUSES Analysis of housing growth between 2001 and 2016, IS DECLINING, WHILE THE PROPORTION OF highlights these paterns. Over this period, a high proporton HOUSEHOLDS LIVING IN APARTMENTS IS of new dwellings have been developed in the Sydney Inner INCREASING. City and these have primarily been apartments (refer to Figure 7). Across the easern half of Sydney, growth has The majority (56 per cent) of Greater Sydney households live primrily been in apartments. The more suburban areas have in detached houses, compared to just under a third (29 per experienced growth in detached houses. Sydney has more cent) who live in apartments. However, within the inner city dispersed paterns of medium and high density housing of Sydney, 76 per cent of households live in apartments, and compared to other Australian cites with new development only two per cent live in detached dwellings (refer to Figure spread across a wide area. 6).

FIGURE 6: DWELLING TYPE MIX GREATER SYDNEY VS SYDNEY INNER CITY, 2001-2016

GREATER SYDNEY

70% 90%

60% 80% 70% 50% 60% 40% 50% 30% 40% 30% 20% Popo�onolln Popo�onolln 20% 10% 10% 0% 0% paa a lao lln paa a lao lln ou apan ou apan

2006 2011 2016 2006 2011 2016

Source: SGS Economics and Planning based on 2001 to 2016 ABS Census data.

What impact does Airbnb have on the Sydney and Melbourne housing markets? 16 THIS GROWTH IN APARTMENT DWELLINGS FIGURE 7: DISTRIBUTION OF NET DWELLING CHANGE, SYDNEY, 2001-2016 HAS LARGELY BEEN CLUSTERED AROUND INNER CITY LOCATIONS, NEAR PUBLIC TRANSPORT CORRIDORS, WHILE DETACHED HOUSING GROWTH HAS OCCURRED ACROSS THE OUTER SUBURBS.

Analysis of housing growth between 2001 and 2016, highlights these paterns. Over this period, a high proporton of new dwellings have been developed in the Sydney Inner City and these have primarily been apartments (refer to Figure 7). Across the eastern half of Sydney, growth has primarily been in apartments. The more suburban areas have experienced growth in detached houses. Sydney has more dispersed paterns of medium and high density housing compared to other Australian cites with new development spread across a wide area.

Source: SGS Economics and Planning based on 2001 to 2016 ABS Census data.

What impact does Airbnb have on the Sydney and Melbourne housing markets? 17 HOUSE AND UNIT PRICES HAVE RAPIDLY INCREASED ACROSS GREATER SYDNEY OVER THE LAST DECADE.

The greatest increase in house prices has been observed in the eastern parts of Greater Sydney including the Eastern Suburbs (where the median dwelling price is almost $2.5 million) and Northern Sydney and Hornsby, which are the most accessible locatons to the Sydney CBD where employment is concentrated. The other locatons with partcularly high house prices are Northern Beaches, Inner West, Ryde and City and Inner South (refer to Figure 8).

Similar paterns have been observed across units with the eastern half of Greater Sydney having the highest prices and growing more rapdly (refer to Figure 9).

FIGURE 8: MEDIAN DWELLING PRICES ACROSS SYDNEY FIGURE 9: MEDIAN UNIT PRICES ACROSS SYDNEY

Eastern Suburbs 1.0 M Eastern Suburbs 2.5 M North Sydney and Hornsby Northern Beaches Northern Beaches 0.9 M North Sydney and Hornsby Inner West Inner West Ryde 0.8 M City and Inner South 2.0 M City and Inner South Sutherland Sutherland Ryde Inner South West 0.7 M Baulkham Hills and Hawkesbury Baulkham Hills and Hawkesbury Inner South West Parramata 0.6 M 1.5 M Parramata South West Blacktown Blacktown 0.5 M South West Outer South West Outer West and Blue Mountains Outer West and Blue Mountains 1.0 M 0.4 M Central Coast Central Coast

Median Unit Prices ($) Unit Prices Median Outer South West

Median House Prices ($) Prices House Median 0.3 M

0.5 M 0.2 M

0.1 M

0.0 M 0.0 M 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 Month Month Source: SGS Economics and Planning, 2018 using CoreLogic, 2017 Source: SGS Economics and Planning, 2018 using CoreLogic, 2017

What impact does Airbnb have on the Sydney and Melbourne housing markets? 18 RENTS HAVE INCREASED AT A SLOWER RATE COMPARED TO HOUSE PRICES, REFLECTING TRENDS OBSERVED ACROSS AUSTRALIA. HOWEVER, IN SOME LOCATIONS, RENTS HAVE DOUBLED.

The locatons where median rent prices have increased by the largest dollar value are locatons where house prices have increase at the highest rate, being Eastern Suburbs, Northern Beaches and North Sydney and Hornsby (refer to Figure 10). In these locatons, rents have doubled from around $500 per week to over $1,000 per week over the last decade.

Median rent prices for units have followed a similar patern, increasing from around $300 per week to over $600 per week (refer to Figure 11).

FIGURE 10: MEDIAN RENT FOR HOUSES ACROSS GREATER SYDNEY FIGURE 11: MEDIAN RENT FOR UNITS ACROSS GREATER SYDNEY

1.0 M Eastern Suburbs Eastern Suburbs Northern Beaches 600 Northern Beaches 0.9 M North Sydney and Hornsby North Sydney and Hornsby Inner West Inner West 0.8 M City and Inner South City and Inner South 500 Sutherland Ryde 0.7 M Ryde Baulkham Hills and Hawkesbury Baulkham Hills and Hawkesbury Sutherland 400 0.6 M Inner South West Parramata Parramata Inner South West 0.5 M Blacktown Blacktown South West 300 Central Coast Outer West and Blue Mountains Outer West and Blue Mountains 0.4 M ($) Unit Rent Median Central Coast South West Median Unit Prices ($) Unit Prices Median Outer South West Outer South West 0.3 M 200

0.2 M 100 0.1 M

0.0 M 0 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 Month Month Source: SGS Economics and Planning, 2018 using CoreLogic, 2017 Source: SGS Economics and Planning, 2018 using CoreLogic, 2017

What impact does Airbnb have on the Sydney and Melbourne housing markets? 19 HOME OWNERSHIP RATES ARE DECLINING AND RENTAL RATES ARE INCREASING.

The proporton of dwellings which are fully owned across Greater Sydney has declined over the last decade. While the proporton of dwellings being purchased has remained relatvely consistent, the proporton of propertes being rented has increased (refer to Figure 12).

FIGURE 12: TENURE OF DWELLINGS IN GREATER SYDNEY

40% Being purchased Fully owned Rented 35%

30%

Percentage of Dwellings Percentage 25%

20% 2006 2011 2016

Source: SGS Economics and Planning 2018, using Australian Bureau of Statstcs, 2006; Australian Bureau of Statstcs, 2011; Australian Bureau of Statstcs, 2016

What impact does Airbnb have on the Sydney and Melbourne housing markets? 20 Melbourne Housing Market

Median rents have remained relatively stable for houses and apartments.

What impact does Airbnb have on the Sydney and Melbourne housing markets? 21 CHAPTER 4: MELBOURNE HOUSING MARKET

This secton explores some of the trends in both the FIGURE 13: DWELLING TYPE MIX GREATER MELBOURNE VS MELBOURNE CITY, 2001-2016 rental and ownership markets in Melbourne over the last 10-15 years including trends in dwelling mix and GREATER MELBOURNE development, ownership and rental prices and trends in home ownership. 80% 90% 70% 80% SIMILAR TO SYDNEY, THE PROPORTION OF 60% 70% 60% 50% GREATER MELBOURNE HOUSEHOLDS LIVING 50% 40% IN DETACHED HOUSES IS DECLINING, WHILE 40% 30% THE PROPORTION OF HOUSEHOLDS LIVING 30% IN APARTMENTS IS INCREASING. 20% 20% Popo�onolln Popo�onolln 10% 10% The majority (68 per cent) of Greater Melbourne households 0% 0% live in detached houses, compared to only 15 per cent who paa a lao lln paa a lao lln live in apartments. However, within Melbourne city, 84 per ou apan ou apan cent of households live in apartments, and only three per 2006 2011 2016 2006 2011 2016 cent live in detached dwellings (refer to Figure 13). Source: SGS Economics and Planning based on 2001 to 2016 ABS Census data.

What impact does Airbnb have on the Sydney and Melbourne housing markets? 22 APARTMENT GROWTH HAS PRIMARILY BEEN CONCENTRATED IN MELBOURNE CITY, WHILE DETACHED HOUSING GROWTH HAS BEEN CONCENTRATED IN THE OUTER SUBURBS.

Analysis of housing growth between 2001 and 2016 across Melbourne has identfed that apartment growth has been largely clustered around inner city locatons near public transport, while detached housing remained concentrated in outer growth corridors. Apartment actvity over the tme period was largely concentrated around inner city suburbs.

Source: SGS Economics and Planning based on 2001 to 2016 ABS Census data.

What impact does Airbnb have on the Sydney and Melbourne housing markets? 23 HOUSE PRICES HAVE RAPIDLY INCREASED ACROSS GREATER MELBOURNE OVER THE LAST DECADE.

The greatest increase in house prices has been observed in the Inner East, Inner South and Inner Melbourne. Across the Inner East of Melbourne, median house prices have increase from $0.2 million to $1.4 million (refer to Figure 15).

Similar paterns have been observed across units (refer to Figure 16).

FIGURE 15: MEDIAN HOUSE PRICES ACROSS GREATER MELBOURNE FIGURE 16: MEDIAN UNIT PRICES ACROSS GREATER MELBOURNE

1.4 M Inner South 0.6 M Inner East Inner East Inner Inner South 1.2 M Outer East Outer East North East 0.5 M Inner North West North East 1.0 M South East South East West 0.4 M North West West 0.8 M Mornington Peninsula Mornington Peninsula 0.3 M 0.6 M Median House Prices ($) Prices House Median Median Unit Prices ($) Unit Prices Median 0.2 M 0.4 M

0.2 M 0.1 M

0.0 M 0.0 M 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 Month Month Source: SGS Economics and Planning, 2018 using CoreLogic, 2017 Source: SGS Economics and Planning, 2018 using CoreLogic, 2017

What impact does Airbnb have on the Sydney and Melbourne housing markets? 24 HOWEVER, MEDIAN RENTS HAVE REMAINED RELATIVELY STABLE FOR HOUSES AND APARTMENTS.

The locatons where median rent prices are highest are the locatons where house prices are highest, being the inner suburbs of Melbourne (refer to Figure 17). In these locatons, rents have increased from around $300 per week to over $550 per week over the last decade.

Median rent prices for units have followed a similar patern, increasing from around $250 per week to $400 per week (refer to Figure 18).

FIGURE 17: MEDIAN RENT FOR HOUSES ACROSS GREATER MELBOURNE FIGURE 18: MEDIAN RENT FOR UNITS ACROSS GREATER MELBOURNE

1.4 M 400 Inner South Inner South Inner East Inner 350 Inner East 1.2 M Inner Outer East Outer East North East North East 300 North West 1.0 M North West South East South East West 250 Mornington Peninsula Mornington Peninsula West 0.8 M 200 0.6 M

Median Unit Rent ($) Unit Rent Median 150 Median House Prices ($) Prices House Median 0.4 M 100

0.2 M 50

0.0 M 0 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 Month Month Source: SGS Economics and Planning, 2018 using CoreLogic, 2017 Source: SGS Economics and Planning, 2018 using CoreLogic, 2017

What impact does Airbnb have on the Sydney and Melbourne housing markets? 25 HOME OWNERSHIP RATES HAVE DECLINED WHILE RENTAL RATES HAVE SUBSTANTIALLY INCREASED.

The proporton of dwellings which are fully owned across Greater Melbourne has declined over the last decade. This has been accompanied by a similar increase in the proporton of propertes being rented (refer to Figure 19).

FIGURE 19: TENURE OF DWELLINGS IN GREATER MELBOURNE

40% Being purchased Fully owned Rented 35%

30% Percentage of Dwellings Percentage 25%

20% 2006 2011 2016

Source: SGS Economics and Planning 2018, using Australian Bureau of Statstcs, 2006; Australian Bureau of Statstcs, 2011; Australian Bureau of Statstcs, 2016

What impact does Airbnb have on the Sydney and Melbourne housing markets? 26 Airbnb in Sydney

The majority of listings hosted for 30 nights or less per year.

What impact does Airbnb have on the Sydney and Melbourne housing markets? 27 CHAPTER 5: AIRBNB IN SYDNEY

This secton analyses Airbnb host data for Greater Sydney including the number of listngs and where they are FIGURE 20: SHARE OF AIRBNB ROOM TYPES IN SYDNEY concentrated, average prices and hostng rates. Entre home/apartment 1% Private room AIRBNB LISTINGS REPRESENT A SMALL PROPORTION OF THE TOTAL DWELLING MARKET. Shared room 38% In 2017, there were, 25,000 listngs on Airbnb across Greater Sydney that hosted guests15, of which 15,200 were entre homes or apartments. The entre homes or apartment listngs represent 0.9 per cent of total dwellings across Greater Sydney.

THE MAJORITY OF LISTINGS ARE ENTIRE HOMES OR APARTMENTS

60% Entre homes or apartments represent 60 per cent of Airbnb listngs in Greater Sydney, compared with 38 per cent private rooms and one per cent shared rooms (refer to Figure 20). The following analysis generally focuses on entre homes/ apartments as these compromise the majority of listngs. Source: SGS Economics and Planning, 2018 using data provided by Airbnb, 2018

15The number of actve listngs on Aribnb is higher than the number that hosted guests as not all listngs host guests. For example, in December 2017, there were 31,600 actve listngs on Airbnb across Greater Sydney. What impact does Airbnb have on the Sydney and Melbourne housing markets? 28 AIRBNB LISTINGS ARE SPREAD ACROSS FIGURE 21: AVERAGE MONTHLY LISTINGS THAT HOSTED ON AIRBNB IN GREATER SYDNEY FOR 2017 GREATER SYDNEY, HOWEVER THE MAJORITY OF LISTINGS ARE CONCENTRATED IN THE EASTERN HALF OF SYDNEY.

The majority of Airbnb listngs which hosted guests in 2017 were concentrated in eastern Sydney including the near the inner city, inner west, eastern suburbs, lower north shore and northern beaches (refer to Figure 21). These are vibrant locatons where jobs and dwellings are located, as well as a number of tourist atractons including Sydney Harbour, Manly and Bondi Beach.

Source: SGS Economics and Planning, 2018 using data provided by Airbnb, 2018

What impact does Airbnb have on the Sydney and Melbourne housing markets? 29 THE AVERAGE NIGHTLY PRICE FOR AN FIGURE 22: AVERAGE NIGHTLY PRICE ENTIRE HOME/APARTMENT, GREATER SYDNEY. 2017 ENTIRE HOME/ APARTMENT ON AIRBNB IN GREATER SYDNEY IS $150, COMPARED TO P SHARED ROOM.

Average prices vary from $90 per night to $300 per night across Greater Sydney (refer to Figure 22). The locatons with the greatest number of listngs in Greater Sydney have an average price of $140 to $175 per night for an entre home/ apartment (refer to Figure 22).

Source: SGS Economics and Planning, 2018 using data provided by Airbnb, 2018 Note: where there are less than 10 listngs in an SA3, price data is not available.

What impact does Airbnb have on the Sydney and Melbourne housing markets? 30 THE NUMBER OF INBOUND TRIPS HAS FIGURE 23: INBOUND TRIPS (ENTIRE HOUSE/APARTMENT) – SYDNEY INCREASED, PARTICULARLY IN SYDNEY INNER CITY. 16K

The number of inbound trips (bookings which evolved into trips) has increased across Greater Sydney, with the greatest 14K increase in Sydney Inner City (refer to Figure 23). Each dot represents the results for a single month and SA3 (where data is available) with Sydney Inner City highlighted 12K in green. The grey lines show the average of all SA3s and its trend over tme while the green lines show the same but for Sydney Inner City only. 10K

8K Inbound Trips 6K

4K

2K

0K 2012 2013 2014 2015 2016 2017 2018 Month Source: SGS Economics and Planning, 2018 using data provided by Airbnb, 2018 Note: trends for Sydney Inner City SA3 are shown in green

What impact does Airbnb have on the Sydney and Melbourne housing markets? 31 WHILE POPULARITY HAS INCREASED, FIGURE 24: AVERAGE NIGHTLY PRICE (ENTIRE HOUSE/APARTMENT) – SYDNEY AVERAGE NIGHTLY PRICES HAVE REMAINED STEADY. 400

Since 2012, average nightly prices for entre house/ apartments has remained steady across Greater Sydney 350 (refer to Figure 24).

Each dot represents the results for a single month and SA3 300 (where data is available) with Sydney Inner City highlighted in green. The grey lines show the average of all SA3s and its trend over tme while the green lines show the same but for 250 Sydney Inner City only.

200

150 Average Nightly Price ($/night) Nightly Average

100

50 2012 2013 2014 2015 2016 2017 2018 Month

Source: SGS Economics and Planning, 2018 using data provided by Airbnb, 2018 Note: trends for Sydney Inner City SA3 are shown in green

What impact does Airbnb have on the Sydney and Melbourne housing markets? 32 MEDIAN ANNUAL AIRBNB HOST EARNINGS FIGURE 25: MEDIAN ANNUAL EARNINGS PER HOST THAT HAS REMAINED RELATIVELY STABLE. HOSTED GUESTS – SYDNEY

Median annual earnings of hosts who hosted guests (as 5K distnct from hosts who listed their property but did not host guests) across Sydney has decline from $5,100 in 2014 to $5,200 in 2017 (refer to fgure 25). 4K

3K

2K Typical Host Income ($AUD) Income Host Typical

1K

0K 2014 2015 2016 2017

Source: SGS Economics and Planning, 2018 using data provided by Airbnb, 2018

What impact does Airbnb have on the Sydney and Melbourne housing markets? 33 THE MAJORITY OF LISTINGS HOST GUESTS FIGURE 26: NIGHTS HOSTED PER LISTING THAT HOSTED GUESTS (ENTIRE HOUSE/APARTMENT) – SYDNEY P 60 60% The number of nights an Airbnb hosted guests per year across Greater Sydney is detailed in Figure 26. The median 50% nights hosted per year has increased from 18 nights to 20 50 nights per year. 40% In terms of distributon, of the listngs that have hosted 40 guests, almost 60 per cent of listngs host guests for between one and 30 nights per year. Only 10 per cent host guests for more than 180 nights per year. This is less than 2,500 listngs 30% 30 across Greater Sydney. 20% At an average rental price of $480 per week across Greater Hosted Nights Typical Sydney and an average nightly price of $150, an Airbnb listng 20 would need to be rented out for almost 170 nights per year 10% in order to draw in the same income as a rental property. This does not take into account any additonal expenses 10 associated with maintaining an Airbnb listng including 0% furnishing, utlites and cleaning etc. which would increase Pnon�o�nn�ol�nonl this number to more than 170 days. 0 2014 2015 2016 2017 > 331 1 - 30 31 - 60 61 - 90 91 - 120 121 - 150 151 - 180 181 - 210 211 - 240 241 - 270 271 - 300 301 - 330

Source: SGS Economics and Planning, 2018 using data provided by Airbnb, 2018 Note: Typical nights hosted is the median value for total nights hosted per listng during the one-year study period. Note: distributon includes only listngs with a year of tenure on the Airbnb platorm

What impact does Airbnb have on the Sydney and Melbourne housing markets? 34 THE HOSTING RATE OF AIRBNB LISTINGS IN FIGURE 27: HOSTING RATE – SYDNEY GREATER SYDNEY IS LESS THAN 50 PER CENT, Fairfeld ON AVERAGE. Blacktown Liverpool anu�na Auburn ou A hostng rate has been calculated comparing the number Carlingford 11 of actve listngs in each locaton and how ofen it is booked. Penrith 500 This analysis found that the hostng rate is less than 50 per Baulkham Hills Hurstville 1,000 cent across Greater Sydney. In locatons where there are the Campbelltown (NSW) 1,500 greatest number of propertes, this is lower (refer to Figure Bankstown 1,931 27). This may be due to competton in these markets, as well Canterbury as the quality of the listngs on ofer and availability. This is Merrylands - Guildford Parramata likely to vary in each locaton. Gosford Cronulla - Miranda - Caringbah Kogarah - Rockdale Dural - Wisemans Ferry Hornsby Sydney Inner City Pennant Hills - Epping Ryde - Hunters Hill Canada Bay Strathfeld - Burwood - Ashfeld Marrickville - Sydenham - Peter.. Botany Sutherland - Menai - Heathcote Leichhardt Pitwater Ku-ring-gai North Sydney - Mosman Chatswood - Lane Cove Manly Eastern Suburbs - North Eastern Suburbs - South Warringah 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 a�no�na

Source: SGS Economics and Planning, 2018 using data provided by Airbnb, 2018 Note: Bars are sized by average unique listngs that hosted guests.

What impact does Airbnb have on the Sydney and Melbourne housing markets? 35 Airbnb in Melbourne

The number of inbound trips has increased substantially in Melbourne City.

What impact does Airbnb have on the Sydney and Melbourne housing markets? 36 CHAPTER 6: AIRBNB IN MELBOURNE

This secton analyses Airbnb host data for Greater FIGURE 28: SHARE OF AIRBNB ROOM TYPES IN MELBOURNE Melbourne including the number of listngs and where they are concentrated, average prices and hostng Entre home/apartment rates. 2% Private room Shared room

AIRBNB LISTINGS REPRESENT A SMALL 35% PROPORTION OF THE TOTAL DWELLING MARKET.

In 2017, there were, 14,500 listngs on Airbnb16 across Greater Melbourne that hosted guests, of which 8,500 were entre homes or apartments. The entre homes or apartment 63% listngs represent 0.5 per cent of total dwellings across Greater Melbourne. Source: SGS Economics and Planning, 2018 using data provided by THE MAJORITY OF LISTINGS ARE ENTIRE Airbnb, 2018 HOMES OR APARTMENTS

Entre homes or apartments represent 63 per cent of Airbnb listngs in Greater Melbourne, compared with 35 per cent private rooms and two per cent shared rooms (refer to Figure 28).

16The number of actve listngs on Aribnb is higher than the number that hosted guests as not all listngs host guests. For example, in December 2017, there were 15,600 actve listngs on Airbnb across Greater Melbourne.

What impact does Airbnb have on the Sydney and Melbourne housing markets? 37 AIRBNB LISTINGS ARE SPREAD ACROSS FIGURE 29: AVERAGE MONTHLY LISTINGS THAT HOSTED ON AIRBNB IN GREATER MELBOURNE FOR 2017 GREATER MELBOURNE, HOWEVER THE MAJORITY OF LISTINGS ARE CONCENTRATED IN INNER MELBOURNE.

The majority of these listngs were concentrated in inner Melbourne and the surrounding suburbs (refer to Figure 21). There are some concentratons in the outer suburbs such as the Yarra Ranges in the east is a popular tourist destnaton. The Werribee SA3 also has a large number of listngs as it has a large number of dwellings.

Source: SGS Economics and Planning, 2018 using data provided by Airbnb, 2018 Note: data for Mornington Peninsula was not included in this analysis.

What impact does Airbnb have on the Sydney and Melbourne housing markets? 38 THE AVERAGE NIGHTLY PRICE FOR AN FIGURE 30: AVERAGE NIGHTLY PRICE ENTIRE HOME/APARTMENT, GREATER MELBOURNE. 2017 ENTIRE HOME/ APARTMENT ON AIRBNB IN GREATER MELBOURNE IS $125, COMPARED P A SHARED ROOM.

Average prices vary from $90 per night to $260 per night across Greater Melbourne (refer to Figure 22). The locatons with the greatest number of listngs in Greater Melbourne have an average price of $130 to $175 per night for an entre home/ apartment (refer to Figure 30).

Source: SGS Economics and Planning, 2018 using data provided by Airbnb, 2018 Note: where there are less than 10 listngs in an SA3, price data is not available. Note: data for Mornington Peninsula was not included in this analysis.

What impact does Airbnb have on the Sydney and Melbourne housing markets? 39 THE NUMBER OF INBOUND TRIPS HAS FIGURE 31: INBOUND TRIPS (ENTIRE HOUSE/APARTMENT) – MELBOURNE INCREASED SUBSTANTIALLY IN MELBOURNE CITY. 16K

The number of inbound trips (bookings which evolved into trips) has increased across Greater Melbourne, with the 14K greatest increase in Melbourne City (refer to Figure 31).

Each dot represents the results for a single month and SA3 12K (where data is available) with Melbourne City highlighted in green. The grey lines show the average of all SA3s and its trend over tme while the blue lines show the same but for 10K Melbourne City only. 8K Inbound Trips 6K

4K

2K

0K

2012 2013 2014 2015 2016 2017 2018 Month Source: SGS Economics and Planning, 2018 using data provided by Airbnb, 2018 Note: trends for Melbourne City SA3 are shown in blue

What impact does Airbnb have on the Sydney and Melbourne housing markets? 40 WHILE POPULARITY HAS INCREASED, FIGURE 32: AVERAGE NIGHTLY PRICE (ENTIRE HOUSE/APARTMENT) - MELBOURNE AVERAGE NIGHTLY PRICES ARE STEADY 400 Since 2012, average nightly prices for entre house/ apartments has remained steady across Greater Melbourne, with the prices trending downwards (refer to Figure 32). 350 Each dot represents the results for a single month and SA3 (where data is available) with Melbourne City highlighted in green. The grey lines show the average of all SA3s and its 300 trend over tme while the blue lines show the same but for Melbourne City only. 250

200 Average Nightly Price ($/night) Nightly Average 150

100

50 2012 2013 2014 2015 2016 2017 2018 Month

Source: SGS Economics and Planning, 2018 using data provided by Airbnb, 2018 Note: trends for Melbourne City SA3 are shown in blue

What impact does Airbnb have on the Sydney and Melbourne housing markets? 41 MEDIAN ANNUAL AIRBNB HOST EARNINGS FIGURE 33: MEDIAN ANNUAL EARNINGS PER HOSTS THAT HAS DECLINED. HOSTED GUESTS – MELBOURNE

Median annual earnings of hosts who hosted guests (as 5K distnct from hosts who listed their property but did not host guests) across Melbourne has declined from $4,000 in 2014 to $3,600 in 2017 (refer Figure 33). 4K

3K

2K

1K

0K 2014 2015 2016 2017

Source: SGS Economics and Planning, 2018 using data provided by Airbnb, 2018

What impact does Airbnb have on the Sydney and Melbourne housing markets? 42 P FIGURE 34: NIGHTS HOSTED PER LISTING THAT HOSTED GUESTS (ENTIRE HOUSE/APARTMENT) – MELBOURNE 60 60% The number of nights an Airbnb hosted guests per year across Greater Melbourne is detailed in Figure 34. The median nights hosted per year has increased from 42 nights 50 50% to 66 nights per year.

In terms of distributon, of the listngs that have hosted 40% guests, over 35 per cent of listngs host guests for between 40 one and 30 nights per year. Approximately 27 per cent of listngs host guests for more than 180 nights per year. This is 30% less than 4,000 listngs across Greater Melbourne. 30

At an average rental price of $420 per week across Greater Hosted Nights Typical 20% Melbourne and an average nightly price of $125, an Airbnb 20 listng would need to be rented out for almost 175 nights per year in order to draw in the same income as a rental 10% property. This does not take into account any additonal 10 expenses associated with maintaining an Airbnb listng including furnishing, utlites and cleaning etc. which would 0%

increase this number to more than 175 days. 0 Pnon�o�nn�ol�nonl 2014 2015 2016 2017 > 331 1 - 30 31 - 60 61 - 90 91 - 120 121 - 150 151 - 180 181 - 210 211 - 240 241 - 270 271 - 300 301 - 330

Source: SGS Economics and Planning, 2018 using data provided by Airbnb, 2018 Note: Typical nights hosted is the median value for total nights hosted per listng during the one-year study period. Note: distributon includes only listngs with a year of tenure on the Airbnb platorm

What impact does Airbnb have on the Sydney and Melbourne housing markets? 43 THE HOSTING RATE OF AIRBNB LISTINGS IN FIGURE 35: HOSTING RATE – GREATER MELBOURNE GREATER MELBOURNE IS LESS THAN 50 PER CENT, ON AVERAGE. Keilor anu�na Melton - Bacchus Marsh ou Melbourne City A hostng rate has been calculated comparing the number Knox 11 of actve listngs in each locaton and how ofen it is booked. Manningham - East 500 This analysis found that the hostng rate is less than 50 per Manningham - West 1,000 cent across Greater Melbourne. However, the hostng rate is Monash 1,500 Whitehorse - West 2,212 around 70 per cent for Melbourne City where there are the Whitehorse - East greatest number of listngs (refer to Figure 35). Whitlesea - Wallan Maroondah Yarra Ranges Nillumbik - Kinglake Maribyrnong Dandenong Essendon Yarra Moreland - North Banyule Tullamarine - Broadmeadows Stonnington - West Hobsons Bay Port Phillip Boroondara Brunswick - Coburg Wyndham Darebin - South Kingston Glen Eira Stonnington - East Darebin - North Bayside 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 a�no�na

Source: SGS Economics and Planning, 2018 using data provided by Airbnb, 2018 Note: Bars are sized by average unique listngs that hosted guests.

What impact does Airbnb have on the Sydney and Melbourne housing markets? 44 Conclusions: Airbnb in context

Broadly, in Sydney and Melbourne, it is not financially beneficial to host a property on Airbnb instead of renting to a long term tenant.

What impact does Airbnb have on the Sydney and Melbourne housing markets? 45 CHAPTER 7: CONCLUSIONS: AIRBNB IN CONTEXT

This secton draws together the analysis of Airbnb data FIGURE 36: ESTIMATED HOST EARNINGS COMPARED TO LOCAL RENT and trends in the housing market to beter understand the role of Airbnb in the housing market. nu�naou 10 700 1000 BROADLY, IN SYDNEY AND MELBOURNE, IT 2000 IS NOT FINANCIALLY BENEFICIAL TO HOST A 3317 PROPERTY ON AIRBNB INSTEAD OF RENTING 600 oop TO A LONG TERM TENANT. n�oapan Paoo Figure 36 shows a comparison of Airbnb earnings and rental aoo income for SA3s across Melbourne and Sydney. Each dot 500 represents a single SA3 for a single month sized by the number of listngs that hosted guests in that area at that tme. The intenton is to highlight those areas where it may be more proftable to remove the dwelling from the rental 400 market and list it on Airbnb full tme. These are shown below

the doted grey line. This does not take into consideraton the alloouannann additonal costs associated with hostng on Airbnb including 300 furnishing the listng, utlites and cleaning which, once considered, would make hostng on Airbnb less fnancially advantageous. 0 100 200 300 400 500 600 700 nou�aoloann

Source: SGS Economics and Planning, 2018 using data provided by Airbnb, 2018 and CoreLogic, 2017 Note: • Host earnings are estmated as expected host earnings over a weekly period whether the dwelling hosted guests or not. The provided dataset contained only earnings per host for hosts that hosted guests, and presumably there are a number of Airbnb hosts who are not successful in hostng guests each month. • One year of data was provided which gave the total number of actve listngs. Combined with the number of listngs that hosted guests, the “listng success rate” was estmated which represents the probability of hostng guests in any given month. Figure 36 uses a generous estmate taking the average plus one standard deviaton, the intenton being to overestmate potental Airbnb earnings. • The measure of rent used is the smaller of either the median house rent or the median apartment rent. The intenton being to underestmate potental rental earnings. • This over and under estmatng ensures that those dots above the line very likely belong above the line. It can be safely concluded that Airbnb is having litle to no impact on the rental market in these areas. Those dots below the line warrant further investgaton. • Finally, this comparison of costs ignores all other costs involved with each revenue stream including agent fees, bills, council rates, furnishing and maintenance.

What impact does Airbnb have on the Sydney and Melbourne housing markets? 46 THERE ARE A SMALL NUMBER OF AREAS FIGURE 37: HOST EARNINGS MINUS LOCAL RENT OVER TIME WHERE IT APPEARS TO BE WORTHWHILE Melbourne Sydney FOR SOME MONTHS OF THE YEAR. WITH THE EXCEPTION OF MELBOURNE CITY, IT APPEARS THAT RENTING HAS HIGHER EXPECTED TAKINGS OVER A TIME PERIOD OF

A STANDARD LEASE AGREEMENT. Knox Manningham - West City Melbourne - Tullamarine Broadmeadows - Whitehorse West Wyndham Auburn Penrith Inner City Sydney 300 The areas where it is considered to be more fnancially benefcial to host a property using Airbnb (compared to having a long term tenant) have been analysed to understand 200 how ofen this is the case (i.e. is it a seasonal trend or consistent). Airbnb and long term tenant income streams operate over diferent tme scales. 100 Figure 37 illustrates the diference between Airbnb earnings and local rent averaged over a quarter focusing only on 0 entre house/apartment listngs in areas that appear below the line in Figure 36. -100 With the excepton of Melbourne City, it only appears to be worthwhile hostng a property on Airbnb compared to rentng to a long term tenant for a couple of months per -200 year. Tenancy agreements generally guarantee a fxed income typically for six to 12 months (depending on the length of a lease) while Airbnb ofers no such stability. It is considered -300 unlikely that a landlord would take a property of the long term rental market unless it was more fnancially viable to list nou�aoloannnuoaln the property on Airbnb for at least six consecutve months of the year. It does not appear to be fnancially viable for a

landlord to rotate between rentng their property on the long 2016 2017 2016 2017 2016 2017 2016 2017 2016 2017 2016 2017 2016 2017 2016 2017 2016 2017 term rental market for six months of the year and listng it on Airbnb for the remaining six months. Source: SGS Economics and Planning, 2018 using data provided by Airbnb, 2018 and CoreLogic, 2017

What impact does Airbnb have on the Sydney and Melbourne housing markets? 47 HOTEL ACCOMMODATION OCCUPANCY FIGURE 38: AVERAGE HOTEL OCCUPANCY RATES RATES ARE HIGH WHICH POTENTIALLY Tullamarine - Broadmeadows SUGGESTS THERE IS A GAP IN SUPPLY Sydney Inner City ACROSS THE TOURISM ACCOMMODATION Botany Ryde - Hunters Hill MARKET. Melbourne City Hobsons Bay The locatons with the highest occupancy rates are located Cronulla - Miranda - Caringbah Stonnington - West near an airport (i.e. Tullamarine in Melbourne and Botany in Eastern Suburbs - South Sydney) or are a major tourist destnaton (Sydney Inner City, North Sydney - Mosman eastern and northern beaches in Sydney and Melbourne City Kogarah - Rockdale (refer to Figure 38). Port Phillip Manly Chatswood - Lane Cove Auburn Strathfeld - Burwood - Ashfeld Campbelltown (NSW) Monash Parramata Bayside Eastern Suburbs - North Dandenong Yarra Essendon Bankstown Glen Eira Richmond - Windsor Liverpool Gosford Brunswick - Coburg Boroondara Wyndham Penrith Kingston Maroondah Yarra Ranges 80% 0 10 20 30 40 50 60 70 80 90 Average Hotel Occupancy (%)

Source: SGS Economics and Planning, 2018 using Australian Bureau of Statstcs, 2013-14; Australian Bureau of Statstcs, 2014-15; Tourist Accommodaton, 2015-16

What impact does Airbnb have on the Sydney and Melbourne housing markets? 48 This is indicatve of a broader trend with high occupancy rates across a number of Australian cites (refer to Figure 39). Occupancy rates for hotel, motel and serviced apartments are highest for Sydney of the capital cites at 85 per cent occupancy.

City of Sydney has identfed a gap in the provision of three star visitor accommodaton across the City of Sydney local government area (City of Sydney 2018). This is considered to be a strong market for Asian tourists. It is possible that Airbnb is servicing this gap in the market.

FIGURE 39: OCCUPANCY RATES FOR HOTEL, MOTEL AND SERVICED APARTMENTS, 2016-17

90% 85% 80% 83% 83% 70% 77% 76% 78% 72% 70% 60% 50% 40% 30% 20% 10% 0%

BRISBANE CANBERRA MELBOURNE DARWIN SYDNEY HOBART PERTH ADELAIDE

OCCUPANCY NATIONAL AVERAGE Source: Tourism Research Australia, 2018

What impact does Airbnb have on the Sydney and Melbourne housing markets? 49 AIRBNB AVERAGE NIGHTLY PRICES ARE, ON FIGURE 40: AIRBNB NIGHTLY PRICES VS HOTEL NIGHTLY RATES AVERAGE, CHEAPER THAN AVERAGE HOTEL Gosford PRICES. Kogarah - Rockdale Strathfeld - Burwood - Ashfeld In Sydney Inner City, Airbnb nightly prices are, on average, Yarra Ranges $50 to $125 cheaper than hotel accommodaton (refer to Cronulla - Miranda - Caringbah Figure 40). In Melbourne City, Airbnb is on average $25 to Port Phillip $80 cheaper per night. While the nightly prices are cheaper, Bayside Airbnb ofers a range of accommodaton types including Brunswick - Coburg high-end operatons as well as more afordable shared Boroondara optons. Essendon Eastern Suburbs - South Wyndham In some locatons, Airbnb prices are higher than hotel nightly Chatswood - Lane Cove rates. This is likely due to tourist demand, the availability and Melbourne City quality of hotel accommodaton and where it is located, as Glen Eira well as the size of the Airbnb propertes compared to a hotel Manly room (e.g. entre home/ apartment versus hotel room). Yarra Botany North Sydney - Mosman Eastern Suburbs - North Auburn Parramata Hobsons Bay Stonnington - West Ryde - Hunters Hill Sydney Inner City Monash -150 -100 -50 0 50 100 150

Average Nightly Airbnb Price Minus Average Nightly Hotel Price ($)

Source: SGS Economics and Planning, 2018 using Airbnb, 2018; Australian Bureau of Statstcs, 2013-14; Tourist Accommodaton, 2014-15; Tourist Accommodaton, 2015-16

What impact does Airbnb have on the Sydney and Melbourne housing markets? 50 BASED ON THIS ANALYSIS, THE IMPACT OF AIRBNB ON THE HOUSING MARKETS IN SYDNEY AND MELBOURNE APPEARS TO BE MINIMAL.

Australia’s populaton has grown rapidly over the last two decades. With a large percentage of this growth being concentrated in Greater Sydney and Greater Melbourne.

The housing markets in Sydney and Melbourne have experienced substantal growth in house prices, and to a lesser extent rental prices, over the last two decades. This growth has been driven by a number of factors including populaton growth, interest rates and tax incentves which drive investment.

Home ownership rates are declining and rental rates are increasing in both Sydney and Melbourne. The development of apartments has rapidly increased, partcularly within the inner city of both Sydney and Melbourne.

Across Greater Sydney and Greater Melbourne, Airbnb listngs represent a very small proporton of the total dwelling market. In 2017, there were approximately 25,000 listngs that hosted guests in Sydney and 14,500 in Melbourne. Of these, 15,200 were entre homes/ apartments in Sydney and 8,500 in Melbourne. Entre homes/ apartments includes both primary places of residence as well as non-primary places of residence (such as a holiday home or secondary dwelling). This represents 0.9 per cent of all dwellings in Sydney and 0.5 per cent of all dwellings in Melbourne.

Airbnb has grown in popularity over the last fve years. However, Airbnb listngs are only likely to be rented 50 per cent of the tme that they are available, and it is considered to be more fnancially stable for a property owner to list a property on the long term rental market than on Airbnb. It is likely that Airbnb has a greater role in addressing a gap in the tourist accommodaton market, more so than any impact of the availability of rental accommodaton in Sydney and Melbourne.

What impact does Airbnb have on the Sydney and Melbourne housing markets? 51 References

Airbnb. (2018). Monthly listngs data for Sydney and Melbourne by SA3 provided to SGS. Australian Bureau of Statstcs. (2006). Tenure Type. Australian Bureau of Statstcs. (2011). Tenure Type. Australian Bureau of Statstcs. (2013-14). Tourist Accommodaton. Australian Bureau of Statstcs. (2014-15). Tourist Accommodaton. Australian Bureau of Statstcs. (2015-16). Tourist Accommodaton. Australian Bureau of Statstcs. (2016). Rent (weekly) Ranges. Australian Bureau of Statstcs. (2016). Tenure Type. Australian Bureau of Statstcs (2017), ERP by SA2 2001-2017. City of Sydney. (2015), Visitor accommodaton Acton Plan (Hotels and Serviced Apartments) CoreLogic. (2017). RP Data. Daley, J., Coates, B., and Wiltshire, T. (2018). Housing afordability: re-imagining the Australian dream. Gratan Insttute. Natonal Housing Supply Council, 2009, State of Supply Report 2008 SGS Economics and Planning. (2017), Why was no one home on Census night?, htp://www.sgsep.com.au/publicatons/why-was-no-one-home-census-night Tourism Research Australia. (2018), State of the Industry 2016-17.

What impact does Airbnb have on the Sydney and Melbourne housing markets? 52 Contact us

HOBART SYDNEY CANBERRA MELBOURNE PO Box 123 209/50 Holt Street Level 2, 28-36 Ainslie Avenue Level 14, 222 Exhibiton Street Franklin TAS 7113 Surry Hills NSW 2010 Canberra ACT 2601 Melbourne VIC 3000 +61 421 372 940 +61 2 8307 0121 +61 2 6257 4525 61 3 8616 0331 [email protected] [email protected] [email protected] sgsvicsgsep.com.au

Australia’s peak hospitality and tourism accommodation bodies welcome ‘certainty’ in FWC transitional arrangements decision on penalty rates reform

5 June 2017: Australia’s peak hospitality organisation, the Australian Hotels Association (AHA), and tourism accommodation body, Tourism Accommodation Australia (TAA), welcomed the certainty provided by the Fair Work Commission (FWC)’s announcement today on the transitional arrangements for penalty rates.

The FWC announced that the reformed penalty rate conditions for full time and part time hospitality employees working on Sundays would be phased in over three years, commencing from 1 July 2017. Sunday penalty rates would be reduced from 175% to 170% in 2017‐18, from 170% to 160% in 2018‐19, and from 160% to 150% in 2019‐2020. Sunday rates for casual employees will not change.

Reformed penalty rates for public holidays apply from 1 July 2017.

In a joint statement released by AHA CEO Stephen Ferguson and TAA Chair Martin Ferguson, they said that while both sectors would have preferred the reforms to have been phased in over two years, the decision would at least provide certainty to employers and employees, which will allow for better future planning.

“Hotels will now be able to make long‐term decisions about the future operation of outlets on Sundays,” said CEO of AHA, Stephen Ferguson. “The reform could lead hotels to increasing trading hours and services and employing more staff.”

Chair of TAA, Martin Ferguson, who was a former Labor Minister for Tourism and head of the ACTU said: “We realise that reforming penalty rates was a tough decision for the FWC, but ultimately this was essential and we are pleased that logic and long‐term vision won out in the end.

“The hotel sector is going through its largest‐ever expansion period and given that the industry is a 24/7 industry it was important that rates on Sundays and public holidays reflect the modern working environment.

“We have always supported workers being remunerated extra for working on weekends and public holidays, but the compensation needs to be fair and sensible. It’s now time for all parties to accept the FWC decision and transitional arrangements and ensure a smooth transition.”

Further information: Stephen Ferguson Carol Giuseppi Peter Hook CEO, AHA CEO, TAA Communications Manager, TAA T: 0412 654000 T: 0403 452177 T: 0407 462213 E: [email protected] E: [email protected] E: [email protected]

(tel:0893217701)

(/index.phtml) Penalty Rates Changes a Win for Public

Scoop.it (http://www.scoop.it) 23.02.17

The decision by the Fair Work Commission (FWC) on penalty rates is a win for the public as hospitality venues will be able to open more on Sundays and public holidays. The decision was announced on Thursday 23 February 2017.

Australian Hotels Association (WA) CEO Bradley Woods says while there are positives from the FWC’s decision, it does not go far enough.

“The decision does not go far enough to deliver changes that will see significant employment growth opportunities,” said Mr Woods.

“The FWC decision means that full and part-time workers under the Hospitality Award will have Sunday rates reduced from 175% to 150% and public holiday rates reduced from 250% to 225%.

“However casual worker rates were only changed slightly on public holidays - from 275% to 250% when we were seeking 175%.

“Whilst we did not get what we asked for, it is a step in the right direction. Along with opening for additional hours to service our hospitality economy, even more jobs could be created.

“Our case challenged the status quo which saw cost-prohibitive penalty rates on weekends and public holidays.

“The changes will encourage our members to offer more shifts for workers and longer hours for customers,” said Mr Woods.

A summary of the changes are as follows:

Sunday permanents reduced from 175% to 150% Sunday casuals remain unchanged at 175% Public Holiday permanents reduced from 250% to 225% Public Holiday casuals reduced from 275% to 250% Public Holiday changes commence 1 July 2017, transitional arrangements to be determined for Sunday rate changes

- Ends -

Further information: Kate McPherson, AHA(WA) Media and Communications Executive, on

(08) 9321 7701 or 0427 228 739

CORPORATE SPONSORS (http://www.aon.com.au/australia) () (http://www.foxsports.com.au/foxsportsven (tel:0893217701)

(http://www.diageo.com/) (http://www.hostplus.com.au/) (http://www.tweglobal.com/)

(https://www.facebook.com/#!/pages/Australian-Hotels-Association-WA/115032445195222) (https://twitter.com/AHAWA)

Home (/) | About AHA (/about_ahawa/) | Training (/training/) | News & Information (/news___information/) | Members Login (/members/) | Events (/events/) | Industry Suppliers (/industry_suppliers/) | Accommodation Hotels TAA & Statistics (/accommodation_hotels_taa/) | Hotel & Hospitality Links (/hotel___hospitality_links/) | AHA Products (/aha_products/) | Contact Us (/contact_us/) | Terms & Privacy (/terms___privacy/) RSA Online Course (/training/rsa_online_course.phtml) | RSA Certificate (/training/rsa_certificate.phtml) powered by SiteBox® (http://www.powercreations.com.au/) designed by JAZ (http://www.jaz.com.au/) Fair Work Commission Penalty Rate Decision View this email in your browser

23 February 2017 Dear AHA Member,

Re: Fair Work Commission decision on penalty rates

Today’s decision by the Fair Work Commission (FWC) on the Australian Hotels Association (AHA) application to reduce penalty rates for the Hospitality Industry (General) Award delivers important reductions in penalty rates that will help the hospitality industry create more jobs and offer more hours, but does not go far enough.

The FWC’s long overdue decision follows 3 years of representations by the AHA and our accommodation subsidiary Tourism Accommodation Australia (TAA) to the 4 year review of the Modern Award.

The AHA has led this fight on behalf of the industry, and has developed submissions and been involved in a number of hearings before the FWC in relation to this issue.

The AHA’s case challenged the status quo of cost­prohibitive penalty rates on weekends and public holidays. This battle on industrial relations was funded out of AHA reserves built up over many years and fighting fund fees were not required from membership to run this once in a generation fight.

This has been a long and expensive process, involving a large amount of work from the WA Workplace Relations team and our external legal advisers.

Forty­three AHA members had the courage to represent members by stepping into the witness box and being cross examined by Union barristers and the full bench of the FWC.

FWC Decision

In a 550­page decision following one of the most extensive evidentiary cases ever heard, the FWC found that existing penalty rates are restricting trading in hospitality businesses, lowering staffing levels and restricting the type and range of services provided to our community. The changes announced by FWC Commission President Iain Ross, who was appointed by Bill Shorten as Minister for Employment in 2012 have the potential to create new jobs and boost economic growth.

Commission President Iain Ross said that:

‘The evidence also supports the proposition that a lower Sunday penalty rate would increase service levels with a consequent increase in employment (in terms of hours worked by existing employees or the engagement of new employees). In particular, a reduction in Sunday penalty rates is likely to lead to: more stores being open on Sundays, increased Sunday trading hours, a reduction in hours by some owner operators, an increase in overall hours worked in retail stores’.

The changes to penalty rates awarded by the FWC will give businesses more capacity and new incentives to hire staff, more hours to existing staff but the benefits would have been greater if the reductions were more significant.

What are the changes?

The Commission resolved to make the following changes to the Hospitality Industry (General) Award:

When will the changes take effect?

The changes to the public holiday rates will commence on 1 July 2017.

It hasn’t been confirmed when new Sunday penalty rates will commence. The FWC is seeking submissions as to some ‘provisional views’ which have been proposed in respect of the change to Sunday penalty rates. Provisional views include ‘take home pay orders’ which would require employers to preserve current employees’ take home pay. Whilst we didn’t get everything we asked for, it is a step in the right direction.

Further information

A summary of the FWC’s decision can be found here.

Members will be updated as more information becomes available.

Please contact Libbi McLean or Madeleine Brown from AHA(WA)’s IR Team on 9321 7701.

Kind regards,

Bradley Woods CEO/EXECUTIVE DIRECTOR

Copyright © 2017 Australian Hotels Association (WA), All rights reserved.

unsubscribe from this list update subscription preferences