Rasier Operations BV v E Tū Incorporated
[2025] NZSC 162
•17 November 2025
| IN THE SUPREME COURT OF NEW ZEALAND I TE KŌTI MANA NUI O AOTEAROA |
| SC 105/2024 [2025] NZSC 162 |
| BETWEEN | RASIER OPERATIONS BV |
| AND | E TŪ INCORPORATED |
| Hearing: | 8–9 July 2025 |
Court: | Winkelmann CJ, Glazebrook, Ellen France, Williams and Miller JJ |
Counsel: | P F Wicks KC, K M Dunn, N L Walker and J A Tocher for Appellants |
Judgment: | 17 November 2025 |
JUDGMENT OF THE COURT
AThe appeal is dismissed.
BThe appellants must pay the respondents one set of costs of $50,000 plus usual disbursements. We allow for second counsel.
____________________________________________________________________
REASONS
| Para No | |
| Winkelmann CJ, Williams and Miller JJ | [1] |
| Glazebrook and Ellen France JJ | [146] |
WINKELMANN CJ, WILLIAMS AND MILLER JJ
(Given by Miller J)
Table of Contents
| Para No | |
| Introduction Provisions purporting to define relationships The operation of the platform in practice Onboarding Drivers’ capacity to refuse or cancel rides, and the Uber Pro rewards programme Disciplinary practices All digital platforms are not the same Bryson v Three Foot Six Ltd What weight should be given to contractual statements about the parties’ relationship? Inequality of bargaining power and s 6 Employee status in this case Does Uber supply transport services to passengers? Does Uber engage drivers to deliver the transport services it supplies to users? Does Uber engage drivers under contracts of service? Freedom to work for others Disposition | [1] [59] [78] [110] |
Introduction
On behalf of four Uber drivers the respondent unions sought a declaration that the drivers are Uber’s employees; in the language of the Employment Relations Act 2000 (ERA), Uber hired them to do work for hire or reward under contracts of service.[1] The drivers succeeded in the Employment Court and Court of Appeal.[2]
[1]Employment Relations Act 2000 [ERA], s 6. As noted below, we refer to all of the Uber entities as “Uber” unless we think it appropriate to distinguish among them. See also below at [25].
[2]E Tū Inc v Rasier Operations BV [2022] NZEmpC 192, [2022] ERNZ 966 (Chief Judge Inglis) [EmpC judgment]; and Rasier Operations BV v E Tū Inc [2024] NZCA 403, [2025] 2 NZLR 150 (Goddard, Ellis and Wylie JJ) [CA judgment].
Uber appeals by leave.[3] It denies that it hires drivers to work for hire or reward at all. Its case is that it supplies digital services to drivers and riders, separately, via its ridesharing platform, which they join via Uber’s Driver and Rider apps, and that these services enable drivers and riders to connect and form their own business relationships. It says that Uber drivers use their own cars to transport riders pursuant to agreements formed between them on a trip-by-trip basis, and it merely facilitates those arrangements by matching a rider who wants to go from A to B with a driver who is willing to take them there. It says that the rider pays the fare to the driver and Uber earns a service fee, paid by the driver, for its services. In fact Uber actually receives the fare and deducts the service fee, which it calculates as a percentage of the fare, before periodically remitting the balance to the driver, but it says that it does so in its capacity of facilitator.
[3]Rasier Operations BV v E Tū Inc [2024] NZSC 177 (Glazebrook, Ellen France and Williams JJ).
The respondents do not say that the four drivers are full-time Uber employees. They worked flexible hours and accept that they are employees only when working for Uber. They they are employees when they have signalled their availability for work by logging into the Uber Driver app.
If, as the Courts below and courts in some other jurisdictions have found,[4] Uber does contract with drivers and riders for the provision of passenger transport services, it says that they work not for hire or reward under contracts of service but as independent contractors. It says that the classical indicia of an employment relationship are absent; it does not control the drivers, who may work as little or as much as they like and are permitted to compete with it; it has not integrated them into its business; and in substance they are in business on their own account.
[4]See below at [94]–[95]; and, as an example, Uber BV v Aslam [2021] UKSC 5, [2021] 4 All ER 209 at [102].
Uber also operates its “Uber Eats” platform in New Zealand, connecting drivers, restaurants and members of the public for food delivery services. We need not consider that platform because it was common ground before us that the status of drivers on it is not relevantly distinguishable from that of drivers on the Rides platform.[5]
The questions on appeal
[5]The Courts below found that the one of the four drivers, who completed deliveries on the Eats platform, was also an Uber employee when he was logged on to the Eats delivery app: see EmpC judgment, above n 2, at [30] and [92]–[93]; and CA judgment, above n 2, at [236].
Under s 214 of the ERA an appeal lies by leave from the Employment Court to the Court of Appeal on the ground that the decision was wrong in law (provided that no appeal lies from a decision on the construction of an employment agreement). This Court held in Bryson v Three Foot Six Ltd that the ultimate decision of the Employment Court on a putative employee’s status is not ordinarily amenable to appeal because it is a question of fact.[6] However, the correct interpretation of the statutory definition of “employee” in s 6 of the ERA is a question of law, and a decision of the Employment Court will also be wrong in law if the Court’s findings of fact are insupportable.[7] If the appellate court finds that an error of law was made in the Employment Court, it may confirm, modify or reverse the decision.[8]
[6]Bryson v Three Foot Six Ltd [2005] NZSC 34, [2005] 3 NZLR 721 at [23].
[7]At [24]–[28].
[8]ERA, s 214(5).
In this case the Court of Appeal granted leave on three questions of law: whether the Employment Court misdirected itself on the application of s 6; whether the Court erred by misapplying the test (or, in the alternative, whether its conclusion was so insupportable as to amount to an error of law); and whether the Court erred by finding that joint employment may arise simply as a result of a number of entities being sufficiently connected and exercising common control over an employee.[9] Uber the third question.[10] With respect to the first, the Court held that the Employment Court had materially misdirected itself with respect to s 6.[11] Having reached that conclusion, the Court of Appeal assessed for itself whether the drivers were in reality employed by the Uber companies involved, and found that they were.[12]
[9]Rasier Operations BV v E Tū Inc [2023] NZCA 216, [2023] ERNZ 334 (Courtney and Mallon JJ) at [14].
[10]See CA judgment, above n 2, at [17].
[11]At [120] and [137].
[12]See at [141]–[142] and [236].
Uber sought leave to appeal to this Court on a number of grounds: the Court of Appeal ought to have accepted that the parties had expressed their intention with respect to employment status in their contract; the Court erroneously characterised provisions of the contract as “window‑dressing”; the Court erred in its application of the common law tests (which we discuss below); and the Court failed to stand back and consider the totality of the relationship.
The underlying question is whether the Employment Court erred in its interpretation of “employee” in s 6, as the Court of Appeal found.[13] That is a question of law for purposes of s 214(1) of the ERA.
[13]See Bryson, above n 6, at [24], [29] and [31].
If the Employment Court did err in law, a second question arises: whether this Court ought to reverse or modify the Employment Court’s decision.[14] Neither party asked that we remit the proceeding for further consideration in the Employment Court, as would be appropriate if the outcome turned on contested facts or called for that Court’s specialist expertise.[15] They invited the Court of Appeal to make its own findings about the drivers’ employment status[16] and took the same approach in this Court. For reasons which will become apparent, we think that is an appropriate course in the particular circumstances of this case. We approach this question accordingly by asking whether Uber has met the normal appellate onus of showing that the findings of the Courts below were wrong.
Overview of this judgment
[14]See at [25]–[28] and [40]; and ERA, s 214(5).
[15]ERA, ss 215–216.
[16]CA judgment, above n 2, at [141]–[142].
Uber’s terms and conditions state explicitly that drivers are not employees, and they describe relationships between Uber entities and drivers, between drivers and users, and between Uber entities and users, in which Uber’s role is that of the provider of matching and payment services.[17] Uber describes drivers’ use of their private cars to provide on-demand passenger transport services to consumers as “ridesharing”, a term which evokes the idea of carpooling.
[17]See below at [29]–[33].
The Courts below found that in reality Uber offers passenger transport services to riders and hires drivers to provide those services.[18] The Court of Appeal described some of the language in Uber’s non‑negotiable documents as “window‑dressing” and “fiction”, evidently designed to confer control over drivers’ labour while skirting the margins of employment law.[19]
[18]EmpC judgment, above n 2, at [51]–[52]; and CA judgment, above n 2, at [225]. See also Arachchige v Rasier New Zealand Ltd [2020] NZEmpC 230, (2020) 17 NZELR 794 at [53].
[19]See CA judgment, above n 2, at [157], [184], [186], [204] and [217].
These findings make it necessary to examine the Uber–driver relationship in some detail. We accordingly outline relevant provisions of the contracts between Uber and drivers and Uber and riders below from [28], before examining how the platform operates in practice from [49].
We then turn at [67] to the underlying question of law we have posed. We consider the statutory definition of “employee” in s 6, and this Court’s 2005 decision in Bryson. We did not understand Mr Cranney, for the respondents, to dispute that the Employment Court appears to have erred in its interpretation of s 6, so making an error of law for purposes of s 214. Before us, the parties focused on whether the Court of Appeal also misinterpreted both s 6 and Bryson. We also consider whether an employment tribunal may ignore contractual provisions which label the parties’ relationship as something other than that of employer and employee, and the role played by inequality of bargaining power when applying s 6.[20]
[20]We use “employment tribunal” to refer to the Employment Court or Employment Relations Authority.
We then turn at [94] to the second question, which has two parts: whether Uber does engage drivers for hire or reward; and if so, whether they work under contracts of service such that they are correctly classified as employees.[21]
An overview of the Rides platform
[21]ERA, s 6(1)(a).
Anyone who wants to drive with Uber may download its Driver app, but they must have an account with Uber-enabled driver features to use it. They and their vehicle must meet standards set by Uber.[22]
[22]We summarise these requirements below at [34].
Drivers signal their availability for work by logging onto the Driver app. It appears that from time to time they must verify their identity when on the app.[23] The app tracks their travels so long as they are logged on.
[23]The four drivers were sometimes “manually selected” for identity verification.
Consumers who have downloaded the Rider app, entered personal and payment information, and accepted Uber’s terms are described as users. (Because users cannot allow anyone else to use their accounts we use that term, as Uber appears to do, interchangeably with “rider”). They may request rides by opening the app and requesting a ride or “trip” to a specified destination. The app offers them a fare. The rider accepts it. The app then locates the nearest available driver who is logged on to the Driver app, and offers that driver the trip. The driver is given the rider’s first name,[24] rating and distance (in minutes and kilometres), and Uber also discloses the trip’s direction and duration if the driver has a sufficient rewards tier level and trip acceptance rate.[25] If the driver does not accept the trip within about 15 seconds the app offers it to the next closest driver.
[24]There is a conflict in the evidence about whether the driver is given the rider’s name when offered the trip or only upon accepting it, but nothing turns on it.
[25]See below at [26] and [60]–[62].
The driver having accepted the trip, they and the rider are put in contact through their respective apps so they can message about details of the pickup, such as the precise location. The rider can see the driver’s photograph and average rating along with the car’s registration number, model and colour. The driver and rider can monitor one another’s location on a map on their phones. They are not given one another’s mobile phone numbers.
Either party may cancel the trip before it starts, or during it, but a cancellation fee may be payable if the rider does so.
When the rider is collected and the driver starts the trip (by pressing a button on the Driver app), the app supplies the driver with the rider’s destination and a proposed route. The driver need not follow that route but Uber may retrospectively reduce the fare, at the driver’s cost, if the rider complains that the driver departed from it.
The fare is set by Uber’s algorithm using the proposed route. At peak times the algorithm may apply surge pricing, increasing the price to reflect demand. In major New Zealand centres Uber offers an “upfront” fixed fare, which is communicated to the rider when the trip is offered. Uber the right to depart from that fare in its discretion. Otherwise, it appears that the rider is given an estimated fare. In neither case is the fare communicated to the driver when the trip is accepted. The driver is given the fare information when they complete the trip. If the fare has not been fixed, Uber calculates it at the end of the trip. It appears that in practice the fare rarely differs from the estimate.
When the trip ends Uber BV debits the rider’s credit card and the rider is sent a receipt by email. It is from Uber, states that Uber hopes the rider enjoyed their trip and states that “[y]ou rode with [driver’s first name]”. A record of the trip is created by Uber and sent to the driver. It the rider’s first name but not their full name or contact details. It is described as an invoice to the rider, but it is not sent to them. Uber BV deducts a service fee of 28 per cent and periodically remits the balance to the driver.
After the trip Uber asks the rider and the driver to rate one another on a scale of one to five stars. We return to the operation of the Rides platform below.
The contracts and the parties to them
There have been several versions of Uber’s terms and conditions over time, and the four drivers joined the platform at different times.[26] The parties accept that for our purposes it is sufficient to refer to the following documents:[27]
(a)A Services Agreement which states that it was last updated on 1 December 2018, as supplemented by various terms and addenda.[28] It governs the relationship between Uber and drivers on the Rides platform. Under this agreement, Uber BV, a company incorporated in the Netherlands, licenses drivers to use the Driver app and Rasier New Zealand Ltd (Rasier NZ), a New Zealand company which is part of the Uber group,[29] supplies the matchmaking service between drivers and users.[30] Uber BV collects fares from users and facilitates payment, through Rasier NZ, to the drivers.
(b)User Terms last updated on 18 December 2019. Under this agreement, users of the Rider app contract with Uber BV, which provides them with access to the Rides platform and arranges rides with drivers and receives payment of the fares.
(c)Community Guidelines last updated on 17 January 2022 and incorporated into the Services Agreement and User Terms. These Guidelines regulate standards of behaviour and provide that anyone who does not follow them may lose access to their Uber accounts.
We refer to all of the Uber entities as “Uber” unless we think it appropriate to distinguish among them.
[26]See EmpC judgment, above n 2, at [30].
[27]These documents have been amended from time to time. We do not need to discuss the Eats documents: see above at [5]. The second and fourth appellants, Uber Portier BV and Portier New Zealand Ltd, are party to the Eats documents, not the Ride documents.
[28]These include a service fee addendum (updated 1 December 2018), Uber Pro terms and conditions (7 February 2021) and an Uber Pool addendum (8 November 2021). A new services agreement, dated 21 April 2022, added a clause relating to the GST treatment of Uber gift cards and vouchers.
[29]We were not given any information about Uber’s corporate structure. We do note that in Aslam, above n 4, at [1] and [3] the United Kingdom Supreme Court stated that Uber BV owns the rights to “Uber’s smartphone application”.
[30]From March 2016 to November 2018 this function was performed by the first appellant, Rasier Operations BV, and before that by Uber BV. See EmpC judgment, above n 2, at [29].
Uber also operates a rewards programme, now known as Uber Pro, for drivers, who are enrolled in it unless they opt out. Under Uber Pro terms and conditions drivers can earn points for completing trips. Trips completed during specified hours earn more points. There are four tiers of rewards (called Blue, Gold, Platinum and Diamond) based on points accrued over three-month periods. Drivers enjoy the status they have earned in a period A for the remainder of that period and the next period B, but points balances are reset at the end of each period, meaning that they must earn enough points during period B to retain their tier status during period C, and so on. The upper three tiers also require drivers to maintain a high rating from riders and not exceed a specified rate of trip cancellations. Benefits such as fuel and tyre discounts apply to all tiers, although the sizes of the discounts may vary. As noted above, drivers in the upper three tiers who also have a high acceptance rate are told, when offered a trip, what is the trip’s duration and direction. Drivers in the upper two tiers may be given priority airport rematches, meaning they get precedence over other drivers for rides requested from certain airports when they complete trips there.
Uber has the unilateral right to amend these documents and it does so from time to time. Drivers and riders must accept these amendments when they next log onto their respective apps.
Terms and conditions of the Services Agreement and User Terms
The argument focused on features of the contracts that are said not to reflect the real nature of the relationships among Uber, drivers and riders, and to evidence Uber’s control of drivers when they are logged on to the app. We describe relevant provisions briefly.
Provisions purporting to define relationships
The Services Agreement states that Rasier NZ provides drivers with “lead generation services” which allow each driver to act as “an independent provider of peer‑to‑peer passenger transportation services”, and that Uber BV “facilitate[s] payment of [f]ares”. Drivers acknowledge that the act of providing passenger transportation services “creates a legal and direct business relationship” between drivers and riders. Drivers agree that they alone are responsible for anything done by riders and for any liabilities arising from the provision of transport services. They also agree that neither Rasier NZ nor Uber BV shall be deemed “to direct or control you generally or in your performance” under the Agreement.
The Services Agreement expressly attempts to exclude employment law. It :[31]
This Agreement is not an employment agreement, and does not create an employment, independent contractor or worker relationship (including from a labour law, tax law or social security law or insurance perspective), joint venture, partnership or agency relationship.
Where by law the driver is nonetheless deemed to be an employee, worker, agent or representative of Uber, the driver indemnifies Uber against claims by anyone, including governmental authorities, based on such deemed relationship.
[31]A similar clause appears in the User Terms, purporting to exclude any “joint venture, partnership, employment or agency relationship” between Uber, riders and drivers as a result of the contract between Uber and riders or their use of the transport services.
The User Terms state that they establish a contract between the rider and Uber BV. Rasier NZ, which provides the matching services, is not party to the User Terms. The services provided comprise:
… the provision of a technology platform that enables you … to: (a) arrange and schedule transportation services or delivery services with independent third party providers of those services who have an agreement with [Uber]; and (b) facilitate payments [to such third party providers].
Users “acknowledge that Uber does not provide transportation or delivery services or function as a transportation carrier”. The User Terms provide that “all such transportation or delivery services are provided by independent third party contractors” who are not employed by Uber and whose services Uber is not responsible for.
The User Terms state that Uber facilitates payment to drivers as their “limited payment collection agent”. Payment to Uber is “considered the same as” payment made directly to the driver. Uber agrees that it will maintain a complaints management service and “manage [it] on behalf of” drivers in a reasonable manner.
Drivers’ obligations with respect to work
The Services Agreement provides that drivers must hold all required licences and permits to provide passenger transport services and must have the legal right to work in New Zealand. Their vehicle must meet Uber’s current standards and be authorised by Uber for use. It must also be insured. Drivers may be subject to background and other checks from time to time. Their right to use the Driver app is non‑transferable and non-sublicensable. New drivers must go through a driver account activation process, which requires in particular that they meet regulatory requirements (age, insurance, vehicle certificate of fitness and registration, and a P licence endorsement). To obtain a P endorsement they must verify their identity and pass a police check. Uber will validate and review these documents and confirm that the vehicle meets its minimum requirements.
There are no minimum hours of work, nor any obligation to work at times chosen by Uber. Drivers “retain the sole right” to “decide when, where and for how long [they] want to use” the Driver app or Uber’s lead generation services.[32] They are free to choose the route they take to deliver a rider to the destination.[33]
[32]By regulation, drivers are however subject to maximum work hours to manage fatigue. Uber limits the number of hours a driver may use the Driver app accordingly.
[33]This is subject to drivers’ obligation of due skill, care and diligence (see below at [37]) and Uber’s right to reduce the fare (see below at [40]).
The Services Agreement also provides that drivers “retain the complete right to engage in other business or income generating activities, and to use other ridesharing networks and apps” in addition to Uber’s. It goes on to give Uber broad powers to deactivate or suspend drivers at Uber’s discretion.[34]
[34]See below at [47].
Under the Services Agreement, drivers must provide their services to riders with due skill, care and diligence, and maintain high standards of professionalism, service and courtesy. They also agree they will choose the most effective and safe manner to complete trips.
The Community Guidelines apply to both drivers and riders. They insist generally on safe and respectful behaviour and provide that non-compliance may lead to loss of access to the Uber platform. This extends to conduct that may damage “Uber’s brand, reputation or business”. Under the heading “[s]treet hails and off‑platform pickups”, the guidelines state that off-app pickups are prohibited and assert that “[t]he law also prohibits street hails or touting” while using the Uber apps.
Fares, service fees and other payments
The Services Agreement provides that Uber will calculate a recommended fare that drivers “can elect to charge”. It also sets surge pricing in its discretion. The fare is described as a default fare payable in the event that the driver does not negotiate a lesser one. The driver cannot negotiate a higher fare. Nor can they solicit tips; rather, the Rider app invites the rider to tip after the trip is completed. Uber also controls the amount and payment of other charges, such as cancellation fees or cleaning charges.
Uber also reserves the right to adjust payment for reasons such as inefficient routes. The driver authorises Uber to accept payment of the fare (and any applicable taxes payable by the rider) and remit the balance to the driver after deduction of a fee payable to Uber and calculated as a percentage of the recommended fare. Drivers are responsible for meeting their own tax obligations, including income tax and GST, but Uber may remit taxes directly to New Zealand tax authorities on a driver’s behalf if required by law. It prepares and issues a receipt to the rider on the driver’s behalf. Uber may adjust the service fee and add other fees on notice to drivers.
From time to time Uber makes what it accepts are direct payments to drivers, to encourage them to (for example) take three consecutive trips. But the Services Agreement also provides that drivers are not entitled to any additional monetary amounts in connection with Uber’s attempts to increase user numbers and increase existing users’ use of the app.
Drivers’ capacity to refuse or cancel rides
Under the Services Agreement drivers “alone decide … when to try to accept, decline or ignore a User request”.
A rider’s request that has been accepted may be cancelled by the rider or the driver, subject to Uber’s then-current policies. The app requests a reason for cancellation and Uber may charge the rider a cancellation fee. Drivers are not charged cancellation fees.
Control of driver–rider communication
As noted above, drivers are given the rider’s first name and pickup location via the Driver app when they accept the rider’s trip request, but they are not given mobile phone numbers or other information they might use to contact the rider outside the app. The Services Agreement prohibits them from doing so, providing that “[y]ou shall not contact any User or otherwise use any of the personal information made available to you” by Uber except to fulfil in-app trips.
Terms on which riders are carried
The User Terms provide that riders agree not to cause “nuisance, annoyance, inconvenience or property damage” to drivers or anyone else. A series of disclaimers seek to exclude Uber’s liability for provision and quality of services to the fullest extent permitted by law.
Ratings
The Services Agreement and User Terms both encourage drivers and riders to rate one another and leave feedback, which Uber may share with others. The Community Guidelines explain that there is a minimum average rating for drivers for each city and they may lose access to the app if they fail to meet it. The rating is an average based on the driver’s last 500 rated trips (fewer if the driver has not done that many). The Guidelines do not state what the minimum rating is.
Discipline
Uber operates a complaints process under which it resolves complaints by riders or drivers. It may decide to cancel or reduce a fare in response to a complaint. Drivers may be restricted from using Uber’s app or services, or have their accounts deactivated, for any breach of the Services Agreement, or any contravention of an Uber policy, or for any reason at the sole and reasonable discretion of Rasier NZ:[35]
Rasier NZ retains the right to, at any time at its sole discretion, restrict you from using the Uber Services in the event of a violation of this Agreement or any relevant Uber policy, your disparagement of Rasier NZ, Uber or any of their Affiliates, or your act or omission that causes harm to Rasier NZ’s, Uber’s or their Affiliates’ brand, reputation or business as determined by Rasier NZ in its sole discretion. Rasier NZ also retains the right to restrict you from using the Uber Services for any other reason at the sole and reasonable discretion of Rasier NZ. Uber retains the right to, at any time at its sole discretion, deactivate or otherwise restrict you from accessing the identification and password key assigned to you by Uber (“Driver ID”) and/or the Driver App, in the event of a violation of this Agreement, any relevant Uber policy, including the Community Guidelines, your disparagement of Rasier NZ, Uber or any of their Affiliates, or your act or omission that causes harm to Rasier NZ’s, Uber’s or their Affiliates’ brand, reputation or business as determined by Uber in its sole discretion. Uber also retains the right to deactivate or otherwise restrict you from accessing the Driver ID and/or Driver App, for any other reason at the sole and reasonable discretion of Uber.
[35]Emphasis omitted.
Under the Services Agreement drivers do not have recourse to New Zealand courts. Disputes must be resolved by mediation or arbitration in accordance with International Chamber of Commerce rules and the proceedings must remain confidential.
The operation of the platform in practice
We turn to the evidence about how Uber’s platform operates in practice. We need not survey all of the evidence on this topic, as the Court of Appeal helpfully did in its judgment.[36] Much of that evidence reflects the contractual terms and is not in dispute; the issue is whether the evidence collectively establishes that Uber exercises control and we examine that topic below from [115]. In the present section of our judgment we confine ourselves to evidence that elaborates on certain provisions of the agreements or is said to show that some provisions do not depict the actual operation of the platform.
Onboarding
[36]CA judgment, above n 2, at [179]–[204].
The evidence about Uber’s processes for engaging drivers, a practice which it calls “onboarding”, is a little obscure. There is evidence that Uber solicits people to “drive with Uber”, nominating an average hourly rate earned by drivers in their area. At one time it appears new drivers had to attend an in-person induction, a practice which contributed in 2016 to a finding that Uber London drivers were “workers” for the purpose of English employment law.[37] Since 2020 the Uber New Zealand process is online and there is no interview.
Drivers’ right to compete with Uber
[37]Aslam v Uber BV England and Wales Employment Tribunal 2202550/2015, 28 October 2016 at [40].
We accept that, as the Courts below found, drivers are free to decide where and when they work on the Uber platform, which does not preclude an employment relationship when they are working.[38] The Employment Court found that none of the four drivers in this case worked full-time on the platform.[39]
[38]EmpC judgment, above n 2, at [53]–[54]; and CA judgment, above n 2, at [188]–[189]. See also Aslam, above n 4, at [91].
[39]EmpC judgment, above n 2, at [72]. We note that one driver, Mr Keil, appears to have worked more than 40 hours in some weeks.
However, the starting point is that Uber drivers are not free to compete with one another, as one would expect if they were autonomous sellers of transportation services (or participants on some other two-sided platform such as Airbnb). There is an element of competition for ratings and access to rewards, but drivers may not choose passengers based on fares, or compete with one another over price, or (since contact details are not shared and riders may not choose their driver when requesting a trip) build personal goodwill with passengers.[40]
[40]As noted above at [44], the Services Agreement prohibits off-app contact.
The provision in the Services Agreement that drivers are not entitled to additional payments from Uber as it expands its market share has the evident purpose of ensuring that Uber need not share with drivers the benefit of network effects, which occur when a transport platform such as Uber’s becomes ubiquitous: the more universal its service, the greater the platform’s utility to its customers. Uber’s evidence describes “[t]he network a platform business cultivates” as its “core asset”. Firms in markets characterised by network effects often compete for the market, not in it, and the winner may secure a substantial degree of market power.[41]
[41]Cyrille Schwellnus and others Gig economy platforms: Boon or Bane? (OECD, ECO/WKP(2019)19, 15 May 2019) at [35]; and Ania Thiemann and Pedro Gonzaga Big Data: Bringing Competition Policy to the Digital Era (OECD, DAF/COMP(2016)14, 27 October 2016) at [50].
There is evidence that other ridesharing platforms operate in New Zealand, but the scant evidence we have indicates that Uber’s share of ridesharing business as at December 2022 typically exceeded 90 per cent[42] and Uber has 11,000 drivers in New Zealand.[43] Uber plainly competes in a market that includes taxi services and other ridesharing platforms, but we have no information about that market. On the limited evidence we have, Uber’s authority vis-à-vis its drivers does not appear to be constrained by their right to work for its rivals.
[42]The evidence of Ms Rosentreter, a witness for the respondent unions, dated 6 December 2022 was that Uber typically supplies more than 90 per cent of platform-based delivery services, both passenger and food. In argument Mr Wicks KC conceded that Uber’s market share is likely significant but noted there was no basis for Ms Rosentreter’s figure and submitted that little weight be placed on it.
[43]This figure was given in an affidavit of an Uber witness, Emma Foley, dated 21 November 2022 in support of Uber’s application for leave to appeal to the Court of Appeal. The Court of Appeal referred to 6,000 drivers, apparently sourced from the respondents’ submissions.
Consistent with that, the Court of Appeal found that there is no reliable evidence of drivers actually competing with Uber when on its app.[44] The Court remarked that it is not easy to see how a driver could multi-app at the same time, given that Uber will sanction them if they do not respond to offered rides and do not accept consecutive rides.[45]
The driver rating system
[44]CA judgment, above n 2, at [196].
[45]At [195] and [202].
Uber’s case is that it does not engage drivers as employees but rather allows anyone to sign on provided they meet regulatory requirements and pass checks intended to ensure passenger safety. It relies on user ratings to ensure that drivers are providing the safe, professional and secure service that the Services Agreement requires of them.
The drivers’ case is that the rating system is integral to Uber’s tight control of their work. The Court of Appeal agreed, pointing out that the system does not work to inform users’ choice of supplier, as is normal for most platform businesses.[46] Riders the driver’s average rating but cannot choose an alternative driver based on ratings. It agreed with the United Kingdom Supreme Court in Uber BV v Aslam that the system operates as an internal management tool for Uber and as a basis for making termination decisions when ratings show that drivers are not meeting Uber’s performance levels.[47]
[46]At [197].
[47]At [197] citing Aslam, above n 4, at [99].
The evidence is that passengers rate drivers on a scale from one to five stars. If a passenger gives a rating of four or less, they are required to select from a list of problems such as “[a]ggressive driving”, “[d]river not polite”, “[i]gnored my requests”, “[b]ad map route” or “[v]ehicle cleanliness”.[48] The driver’s rating is an average over the past 500 rated trips.[49] Mr Cranney made the point that a low rating takes 500 rated trips to drop off the system, but that invites the response that a single low rating has no significant effect on an otherwise perfect set of 500.[50] The more relevant point, which we accept, is that it is challenging to maintain an average rating at the high levels that drivers must meet to avoid adverse consequences. The evidence is that the minimum driver rating below which a driver may lose access to the app was formerly 4.5 but it is now not specified. As we next explain, however, the more important minimum rating is that needed to attain and retain rewards under the Uber Pro terms.
Drivers’ capacity to refuse or cancel rides, and the Uber Pro rewards programme
[48]Uber does not count a rating towards a driver’s overall rating if it relates to fare price, app issues or pool problems.
[49]There is evidence that only about 50 per cent of riders leave a rating.
[50]A single one-star rating would reduce the average to 4.992.
We have explained that under the Services Agreement drivers are free to accept, decline or ignore trip requests. However, they are discouraged from doing so. Uber monitors drivers’ location when they are on the Driver app and monitors rides declined or ignored. A driver who does not accept three trip requests in a row will be logged off but may immediately log back on.[51] Drivers whose acceptance rate is too low receive warnings that being logged on is an indication that they are available to accept trip requests.
[51]The same applies if a driver cancels three trips in a row.
We introduced the Uber Pro rewards programme above at [26]. In practice, drivers must attain at least a 4.70 driver rating and a cancellation rate of no more than three per cent to achieve Gold, Platinum or Diamond status, which is necessary if Uber is to give the driver trip duration and direction. They must also maintain a trip acceptance rate of at least 85 per cent to access those features.[52] The high acceptance rate requirement must reduce the utility of the additional information of trip duration and direction because drivers would still need to accept almost all trip requests.[53] Other drivers do not know the rider’s destination or the fare, so cannot decline rides on those grounds.
[52]This requirement prevented some of the four drivers from accessing trip duration and direction, despite meeting the other criteria.
[53]As the Court of Appeal noted, “[o]nly those who rarely exercise the choice to decline rides (a prerequisite for maintaining high status) are given the information relevant to such choices”: CA judgment, above n 2, at [192(a)].
The distinction between the upper three tiers is the number of points required to qualify; at least 1,000 for Gold, 1,800 for Platinum and 2,800 for Diamond. Points are earned at one point per trip plus another four points per trip during hours of peak demand (essentially, a period before and after normal working hours on weekdays and Friday–Sunday nights). As noted above, points reset every three months, meaning that they cannot be banked. The Uber Pro rewards programme therefore creates powerful incentives to be available for work at Uber’s preferred hours, to accept and complete trips, and to maintain a very high driver rating.
Disciplinary practices
We have explained that Uber may restrict or terminate drivers’ access to the platform restricted for loosely specified reasons, including breaches of Uber policies.[54] The evidence indicates that the grounds include customer complaints, poor customer reviews, high trip cancellation rates, accepting cash from customers, and poor ratings. As noted earlier, Uber controls the complaints system. It also operates a three‑strikes regime for breaches such as accepting cash from customers.
[54]Above at [47].
As noted, Uber has not disclosed what is the minimum average driver rating that must now be maintained, although a driver may be disciplined for breaching it by having their accounts deactivated. Uber defended this practice by stating that drivers are not sanctioned for breaching the minimum rating without first being given warnings that they are approaching it. A driver who has been suspended for poor ratings must attend a training session, at their own expense, before they may return to the Driver app.
There is evidence that Uber disciplines drivers for infractions such as GPS manipulation or location disabling to improve their place in an airport queue. These appear to rest on Uber’s practice of monitoring drivers’ locations when they are on the app.
All digital platforms are not the same
The tenor of Uber’s evidence is that it is typical of platform businesses, which match buyers and sellers and minimise transactions costs, making possible transactions that might not otherwise happen.
But as the United Kingdom Supreme Court observed in Aslam and the Court of Appeal agreed in this case, all digital platforms which connect sellers and buyers of goods and services on a per-service basis[55] are not the same and it is instructive to consider the ways in which Uber differs. In Aslam Lord Leggatt SCJ, writing for the Court, explained that while booking agents for services such as hotel accommodation typically book on their standard terms and handle payment, in return for a service fee, they do not offer a standardised product which they define or set the price, customer ratings are used to help new customers choose a supplier, and the platforms do not try to restrict customers from dealing directly with suppliers in future.[56] None of this is true of Uber. The of Appeal offered a similar example in this case, in the form of platforms which merely connect tradespeople and homeowners who negotiate their own terms of service without becoming party to those arrangements.[57] The Court’s point, with which we agree, was that each platform must be assessed by reference to its own contractual terms and behaviour.[58] Some are not participants in the market for the retail services they facilitate, but others are.
The Employment Relations Act 2000
[55]An OECD Working Paper describes businesses such as Uber as “gig economy platforms” because they deal in labour; they match workers to customers on a per-service or “gig” basis. That distinguishes them from businesses that match suppliers and customers for other services (such as Airbnb) and also from business-to-consumer platforms such as Amazon. See Schwellnus and others, above n 41, at [5].
[56]Aslam, above n 4, at [103]–[104].
[57]CA judgment, above n 2, at [206].
[58]At [208].
The first object of the ERA is “to build productive employment relationships through the promotion of good faith in all aspects of the employment environment and of the employment relationship”.[59] The Act seeks to do this in a number of ways, including relevantly by “acknowledging and addressing the inherent inequality of power in employment relationships” and by “protecting the integrity of individual choice”.[60]
[59]ERA, s 3(a).
[60]Section 3(a)(ii) and (iv).
Employment status under the Act is the gateway to a number of important rights and obligations for employees and employers, including entitlements such as minimum wage, holidays, parental leave, sick leave, bereavement leave and the rights to unionise and bargain as a collective. Status is determined by deciding under s 6 of the Act whether a person is an “employee”. An “employer” is “a person employing any employee”.[61]
[61]Section 5 definition of “employer”. The definition includes a person engaging or employing a homeworker.
Section 6 provides so far as relevant:
6 Meaning of employee
(1) In this Act, unless the context otherwise requires, employee—
(a)means any person of any age employed by an employer to do any work for hire or reward under a contract of service; and
…
(2)In deciding for the purposes of subsection (1)(a) whether a person is employed by another person under a contract of service, the court or the [Employment Relations] Authority (as the case may be) must determine the real nature of the relationship between them.
(3)For the purposes of subsection (2), the court or the Authority—
(a)must consider all relevant matters, including any matters that indicate the intention of the persons; and
(b)is not to treat as a determining matter any statement by the persons that describes the nature of their relationship.
…
Bryson v Three Foot Six Ltd
The question in Bryson was whether a dispute over employment status raised a question of law that might be the subject of an appeal from the Employment Court under s 214(1) of the ERA.[62] This Court held that no error of law had been made, so restoring the decision of the Employment Court. The first alleged error of law had been whether s 6 of the ERA adopted the approach to employee status that had been taken under predecessor legislation, the Employment Contracts Act 1991, as interpreted by the Court of Appeal in TNT Worldwide Express (NZ) Ltd v Cunningham.[63] This Court was not prepared to confine s 6 in that way. It elected to answer what it described as the real question: whether the Employment Court Judge had correctly directed herself in accordance with s 6.[64]
[62]See Bryson, above n 6, at [17] and [23]–[24].
[63]See at [30] citing TNT Worldwide Express (NZ) Ltd v Cunningham [1993] 3 NZLR 681 (CA).
[64]At [31].
The Court quoted the definition of employee and recounted that under s 6 an employment tribunal must determine the real nature of the relationship between the persons who are said to be employer and employee under a contract of service. In deciding the real nature of their relationship the tribunal must consider all relevant matters, including any matters that indicate the intention of the persons, and is not to treat as determinative any statement by the persons that describes the nature of their relationship.
The Court went on to state that “all relevant matters” plainly include:[65]
(a)the written and oral terms of the contract between the persons, which will usually contain indications of their common intention;
(b)any divergences from or supplementation to those terms which are apparent in the ways in which the parties have actually behaved in practice; and
(c)features of control and integration, and whether the contracted person has been effectively working on their own account (the fundamental test). The Court remarked that the definition of employee reflected the common law and added that the control, integration and fundamental tests were important determinants of the relationship at common law.[66]
[65]At [32].
[66]At [31]–[32].
It will be seen that the tribunal must determine the real nature of the relationship and must consider all relevant matters, but the legislation does not prescribe that any one relevant matter is more important than another or impose a particular methodology for determining the real nature of a relationship. Nor did this Court prescribe a methodology. It confined itself to stating that it will not usually be possible to examine the relationship in light of the control, integration and fundamental tests until the tribunal has examined the terms of the contract and the way in which it operated in practice.[67]
Did the Court of Appeal misinterpret Bryson?
[67]At [32].
The parties to the present appeal did not suggest that Bryson was wrongly decided, but counsel did address issues of interpretation and methodology, to some extent revisiting both Bryson and TNT. Mr Wicks KC, for the appellants, argued that:
(a)The legislative history of s 6 shows that it was intended only to stop employers mislabelling employees as independent contractors.[68]
(b)Bryson establishes a methodological framework under which a court must look to the terms of the contract, any divergence from it in practice, features of control, integration and independence, and any other relevant matters.
(c)The Court of Appeal erred in this case by looking to contract and practice to identify the parties’ rights and obligations as a matter of reality, then deciding whether those rights and obligations amounted to a contract of service; it ought instead to have derived their “intention as to status” largely from the terms of the contract.
[68]Citing, in particular, Employment Relations Bill 2000 (8-2) (select committee report) at 5–6.
It will be apparent from what we have already said that each limb of this argument rests on a misunderstanding of Bryson. The Court did not hold that s 6 is intended only to stop employers mislabelling employees; it held rather that an employment tribunal must establish the real nature of the parties’ relationship. It did not prescribe a methodology or framework for analysis; it anticipated rather that an employment tribunal will normally, but not invariably, establish the terms of the contract and its workings in practice before deciding whether the relationship is one of employer and employee. It did not hold that the tribunal must ascertain the parties’ common intention as to status from the terms of the contract; it held rather that indications of their intention will normally be found there. This last point reflects the language of s 6(3)(a), which envisages that an employment tribunal will ascertain the parties’ intention from any relevant matters that evidence it.
The parties to an agreement to do work for hire or reward may commence their relationship by negotiating a written contract that accurately records all of their rights and obligations, and as a matter of fact they may not have diverged materially from it in practice. In such cases the matters that are relevant when ascertaining intention, which is an objective inquiry, will be found in the contract, read in its commercial context.[69]
[69]Ermogenous v Greek Orthodox Community of SA Inc [2002] HCA 8, (2002) 209 CLR 95 at [25] per Gaudron, McHugh, Hayne and Callinan JJ, which was adopted in McDonald v Ontrack Infrastructure Ltd [2010] NZEmpC 132, (2010) 8 NZELR 108 at [47]. See also Bryson, above n 6, at [20]; and Prasad v LSG Sky Chefs New Zealand Ltd [2017] NZEmpC 150, (2017) 15 NZELR 178 at [57]. See also CA judgment, above n 2, at [108] and [111].
Bryson and TNT both differ significantly from the present case in this respect. The contracts in those cases declared the plaintiff to be an independent contractor and it was necessary to decide what weight, if any, ought be attributed to the parties’ use of that term, but the contracts otherwise described with reasonable accuracy what happened in practice. That is why Cooke P was able to say in TNT that if the case were approached as a question of contractual interpretation there could be little doubt that the plaintiff was intended to be an independent contractor.[70] In this case, it is not merely Uber’s descriptions of legal relationships that are in dispute. It is a distinctive feature of the case that the language of its “take it or leave it” documents is said to actively disguise the reality of the parties’ relationship.
What weight should be given to contractual statements about the parties’ relationship?
[70]TNT, above n 63, at 684.
Section 6(3)(b) eliminates the difficulty which the Court of Appeal grappled with in TNT: in an agreement which is not a sham, may an employment tribunal discount or disregard the parties’ plainly expressed intention as to the status of their relationship?[71] The legislation states that the tribunal may not treat the parties’ characterisation of their relationship as determinative. It is implicit that the tribunal may attach some weight to such provisions.
[71]The question in TNT was to what extent the common law approach to this problem was applicable under the Employment Contracts Act 1991. In Bryson the Court of Appeal summarised the common law approach and examined the legislative history of s 6 of the ERA: see Three Foot Six Ltd v Bryson (2004) 2 NZELR 29 (CA) at [62]–[75] per William Young and O’Regan JJ. See also CA judgment, above n 2, at [96].
The question to which we now turn is whether statements in Uber’s documents about the nature of relationships among Uber, drivers and riders should be given any weight at all. The Court of Appeal discounted the contractual disclaimers of employee and independent contractor status as unilateral window-dressing by Uber.[72] The Court reasoned that an employer cannot exploit an imbalance of power to insist that an employee impliedly contract out of the ERA by agreeing to describe what is in truth an employment relationship as something else.[73]
[72]CA judgment, above n 2, at [217].
[73]At [7].
The same issue arose in Aslam. The United Kingdom Supreme Court held that to treat Uber’s terms as the starting point in classifying the parties’ relationship would be to allow Uber power to decide for itself whether employment legislation applied to its drivers. Lord Leggatt SCJ stated that:[74]
[76] … To do so would reinstate the mischief which the legislation was enacted to prevent. It is the very fact that an employer is often in a position to dictate such contract terms and that the individual performing the work has little or no ability to influence those terms that gives rise to the need for statutory protection in the first place. The efficacy of such protection would be seriously undermined if the putative employer could by the way in which the relationship is characterised in the written contract determine, even prima facie, whether or not the other party is to be classified as a worker. …
[77] This point can be illustrated by the facts of the present case. The Services Agreement … was drafted by Uber’s lawyers and presented to drivers as containing terms which they had to accept in order to use, or continue to use, the Uber app. It is unlikely that many drivers ever read these terms or, even if they did, understood their intended legal significance. In any case there was no practical possibility of negotiating any different terms. In these circumstances to treat the way in which the relationships between Uber, drivers and passengers are characterised by the terms of the Services Agreement as the starting point in classifying the parties’ relationship, and as conclusive if the facts are consistent with more than one possible legal classification, would in effect be to accord Uber power to determine for itself whether or not the legislation designed to protect workers will apply to its drivers.
[74]Aslam, above n 4.
In Construction, Forestry, Maritime, Mining and Energy Union v Personnel Contracting Pty Ltd, the High Court of Australia took the same view.[75] Kiefel CJ, Keane and Edelman JJ held that parties’ opinion as to the status of their relationship are generally irrelevant:[76]
The parties’ legitimate freedom to agree upon the rights and duties which constitute their relationship … does not extend to attaching a “label” to describe their relationship which is inconsistent with the rights and duties otherwise set forth. To do so would be to elevate their freedom to a power to alter the operation of statute law to suit themselves or, as is more likely, to suit the interests of the party with the greater bargaining power.
[75]Construction, Forestry, Maritime, Mining and Energy Union v Personnel Contracting Pty Ltd [2022] HCA 1, (2022) 275 CLR 165.
[76]At [58].
The Judges reasoned that because the true characterisation of an employment relationship is a question of law, the parties’ descriptions are unlikely to assist and will rarely even be relevant.[77] Gageler and Gleeson JJ were prepared to accept that such provisions may inform but cannot alter the court’s assessment of the “substantial relations” between the parties.[78]
[77]At [66].
[78]At [127]. See similarly at [184] per Gordon J, with which Steward J agreed.
In this case the Court of Appeal accepted that weight may sometimes be given to the parties’ characterisation of their relationship.[79] That will be so where there is no reason to think that such terms result from inequality of bargaining power. The Court instanced cases where there has been a genuine and legally informed negotiation:
[111] Where an agreement has been negotiated between a worker and a principal, each of whom has access to legal advice and a meaningful opportunity to engage on what the terms of their relationship should be, it may be appropriate to give some weight (albeit not determinative weight) to statements in the agreement about what sort of relationship they intend to commit to.
[79]See CA judgment, above n 2, at [111]–[112].
However, the Court held, the same is not true of contracts offered by the putative employer on a “take it or leave it” basis, even if the employee understands the legal significance of the employer’s terms:[80]
… in that context, it is much less likely that labels and similar terms genuinely reflect the parties’ intentions about the real nature of their relationship. Where an employer uses their superior bargaining power to include inaccurate relationship labels (or other window-dressing terms) in a contract that is in substance a contract of service, the fact that an employee who accepts that agreement has a good understanding of what the misleading term means is beside the point. An employee cannot expressly contract out of the protections of the ERA, however well informed they may be about those protections and however well they understand the provision excluding those protections, because of the risk of abuse by employers. Similarly, an employee cannot impliedly contract out of those protections by agreeing to an inaccurate label for their relationship, however well they may understand the provision and however informed they may be about the implications of doing so.
[80]At [112].
It is plain that terms which classify a putative employment relationship as something else cannot be permitted to determine the real nature of that relationship. That would be to allow the parties to decide for themselves whether the legislation applies. An employment tribunal’s task under s 6 is to determine for itself the real nature of the putative employment relationship, by reference to all relevant circumstances. The parties’ understanding of the legal significance of the terms they employ is but one such circumstance.
For the reasons given above at [75]–[76], we accept that terms which the parties use in a contract to describe their relationship may be afforded weight in normal circumstances. In this case, however, the Court of Appeal described some of the language in Uber’s documents as “window-dressing” and “fiction”.[81] We agree with the Court of Appeal that no weight need be given to contractual language which disguises the real nature of the relationship.
Inequality of bargaining power and s 6
[81]The Court of Appeal borrowed the term “window-dressing” from the reasons of Sedley LJ in Autoclenz Ltd v Belcher [2009] EWCA Civ 1046, [2010] IRLR 70 at [104].
It is axiomatic that an enactment must be interpreted in light of its purpose.[82] In her decision in this case, Chief Judge Inglis held that s 6 and the statutory entitlements that surround it “reflect a statutory recognition of vulnerability based on an inherent inequality of bargaining power”.[83] As we have noted, under s 3 the ERA pursues its object of building productive employment relationships by promoting good faith in all aspects of the employment environment and relationship in a number of ways, including by “addressing the inherent inequality of power in employment relationships” and by “protecting the integrity of individual choice”.[84]
[82]Legislation Act 2019, s 10(1).
[83]EmpC judgment, above n 2, at [8].
[84]ERA, s 3(a)(ii) and (iv).
The Court of Appeal held that while it is permissible to bear in mind the underlying protective purpose of the ERA, no substantial assistance could be obtained from s 3 when interpreting s 6.[85] The Court pointed out that the ERA is concerned not with inequality generally but with inequality in employment relationships; it followed that there is a risk of circular reasoning if inequality is permitted to determine status.[86]
[85]CA judgment, above n 2, at [117].
[86]See also at [124]–[125].
We agree that inequality of bargaining power may not be used to extend the statutory concept of doing any work for hire or reward under a contract of service. As Mr Cranney suggested, the Chief Judge may not have intended to go so far as to hold otherwise. She added immediately that the purpose of the ERA must be looked to when deciding whether a person is an employee, which is a different question.[87] However, her language does convey the opinion that inequality of bargaining power may enlarge the meaning of “employee” in s 6, and to that extent we agree with the Court of Appeal that she was in error.
[87]EmpC judgment, above n 2, at [9].
Indeed, it was not necessary to go so far. Inequality of bargaining power may account in the first place for statutory language that requires an employment tribunal to determine the real nature of the parties’ relationship, as the Court of Appeal held.[88] The Court cited Aslam for the proposition that it is the very fact that an employee frequently cannot influence the employer’s chosen terms that gives rise to the need for statutory protection in the first place.[89]
[88]CA judgment, above n 2, at [117] and [124].
[89]At [104] citing Aslam, above n 4, at [76].
As we have explained above, inequality of bargaining power may also inform the tribunal’s findings as to the real nature of the relationship in any given case. As the Court of Appeal held, it may explain the presence of contract terms “designed to make [a] contract appear to be something other than that which it in reality is”.[90]
[90]At [104].
In this Court Uber sought to build on the Court of Appeal’s conclusion that s 3 is not of substantial assistance when interpreting s 6. Uber argued that not until an employment contract is established does s 3 come into play. Only then, the argument ran, may inequality of bargaining power be taken into account. The argument cannot be reconciled with the statutory language. Inequality of bargaining power is relevant, almost by definition, because it may affect the terms of a bargain that has yet to be struck. It may explain, as a matter of fact, why a contract does not reflect the real nature of the parties’ relationship.
For these reasons, we answer the underlying question of law we have posed above at [9] and [14] accordingly: the Court of Appeal did not misinterpret s 6 or Bryson, as Uber contended in this Court.
Employee status in this case
In this case it is not merely the label that Uber attaches to the relationship with drivers (a supplier of lead generation and payment services to them) that is in issue. Uber resists both of the traditional labels—independent contractor and employee—because it insists that it does not engage drivers to supply it with transportation services at all. The Services Agreement says in substance that Uber is not in the business of delivering passenger transport services, and (with the other Uber documents) it claims to establish or recognise discrete relationships among riders and drivers, riders and Uber, and drivers and Uber.[91] The Courts below have found these claims inaccurate and misleading.[92] They have concluded that the carefully drawn and apparently comprehensive written contracts between Uber and drivers disguise the real nature of their relationship.
[91]See above at [29]–[33].
[92]See especially EmpC judgment, above n 2, at [37] and [79]; and CA judgment, above n 2, at [185] and [204].
Courts in some other jurisdictions have reached similar conclusions.[93] We recognise that these decisions must be treated with some care. Legislation in some European Union member states, the United Kingdom and Australia appear to serve similar purposes and employ similar tests, but it is evident from the reported decisions that there are differences in the legislation and regulatory settings, and sometimes in Uber’s documentation (which appears to have evolved over time). For instance, in Aslam Uber drivers were found to be “workers”, which in the United Kingdom legislation is an intermediate category between independent contractor and employee statuses.[94] And while Uber has usually lost in these jurisdictions, that has not always been the case. In Australia it prevailed in Gupta v Portier Pacific Pty Ltd and Nawaz v Rasier Pacific Pty Ltd (t/as Uber BV), among others.[95] It also prevailed in the New Zealand Employment Court in Arachchige v Rasier New Zealand Ltd.[96]
[93]See, for instance, those cited at CA judgment, above n 2, at [180].
[94]See Aslam, above n 4, at [38].
[95]Gupta v Portier Pacific Pty Ltd [2020] FWCFB 1698, (2020) 296 IR 246; and Nawaz v Rasier Pacific Pty Ltd (t/as Uber BV) [2022] FWC 1189, (2022) 317 IR 134.
[96]Arachchige, above n 18.
It is appropriate to approach the analysis by first considering whether Uber engages drivers for hire or reward to supply passenger transport services. If so, they are either independent contractors to Uber or its employees. It is then necessary to consider whether it engages them under contracts of service. If so, they are employees.
Does Uber engage drivers to work for hire or reward?
This question is appropriately divided into two further sub‑questions. The first is whether Uber supplies transportation services to users or merely provides them with a technology platform that they may use to arrange transportation services that are supplied by others. If supplies transportation services, the second question is whether it engages drivers to deliver them. The Courts below answered both questions in the affirmative.
Does Uber supply transport services to passengers?
There is no pre-trip contact between rider and driver, let alone a negotiation. In the absence of such contact, Uber has sought elsewhere to characterise itself as the agent of the driver and rider for the purpose of establishing a contract between them. The User Terms which were in issue in Aslam provided that “Uber UK” acted as disclosed agent for drivers and a rider–driver contract arose on the driver’s acceptance of the trip.[97]
[97]Aslam, above n 4, at [28]. In Aslam the United Kingdom Supreme Court found that there was no factual basis for Uber’s claim that Uber London (a subsidiary licensed to operate private hire vehicles in London) acted as agent for drivers when accepting trips: at [49]. Because the New Zealand User Terms differ, we need not review the Court’s reasons for that finding.
The New Zealand User Terms do not purport to create a contract between riders and drivers or claim that Uber acts as agent for them or for drivers for that purpose. The only agency relationship which the User Terms recognise expressly is a limited payment collection agency for drivers.[98] As we explained above at [30], the User Terms and Services Agreement both expressly exclude any other agency relationship.
[98]See above at [33].
Rather, Uber’s position before us was that the Services Agreement contemplates a rider–driver contract and such contracts arise implicitly on a trip‑by‑trip basis. As noted earlier, the Services Agreement asserts that the mere act of providing passenger transportation services creates a “legal and direct business relationship” between rider and driver.[99] The User Terms assert that Uber does not provide transportation services and such services are provided by independent third‑party contractors (the drivers).[100]
[99]See above at [29].
[100]See above at [32].
We have explained above at [18] that a prospective rider must first download the Rider app and create an account, adopting the User Terms by agreement with Uber BV. Users use the app by requesting a trip to a specified destination. Uber offers the rider a fare for the trip. The rider accepts the offered trip. The key terms of the trip have been agreed, but the only parties to the transaction at this point are Uber and the rider.
Rasier NZ then offers the trip to the driver who is nearest to the pickup point. In all other respects—notably, price, route and service quality—the Uber platform treats drivers as entirely fungible to this point. It does not offer riders the opportunity to choose from a selection of drivers. There is no reason to do so, because drivers are vetted by Uber (by initial screening and thereafter by monitoring customer ratings and other metrics), destination and price have been agreed with Uber, and the service is standardised under terms of service set by Uber.
Nor does the platform provide the driver with the trip duration or direction unless, as explained earlier, they enjoy Gold, Platinum and Diamond status and maintain an acceptance rate of at least 85 per cent.[101] The only information which the driver has about the rider is their first name, average rating and distance. But drivers who do not accept three consecutive trip requests are logged off of the app.
[101]See above at [60].
When the driver responds to the offered trip by accepting the trip on the app, Uber tells the rider that the trip has been accepted and discloses the driver’s present location, first name, rating and vehicle information. The only criterion which the rider might have to change driver at that point is the driver’s average rating, and the rider is not given that option. To change driver the rider would have to cancel the trip and re‑book, potentially incurring a cancellation fee.
When the trip ends it is Uber BV that collects the fare and pays the driver. The ’s reward is calculated as a percentage of the fare.
Users and drivers are practically anonymous vis-à-vis one another throughout the entire transaction. As explained above at [18]–[19], they are put in contact via the app but they have only one another’s first names and locations and drivers are prohibited from contacting users outside the app.[102] When the trip ends, Uber generates an invoice, purportedly from the driver to the rider, but as explained above the rider never sees this invoice and remains anonymous on the copy sent to the driver.[103] If any difficulty arises, in the form for example of a complaint about the driver’s conduct, the User Terms provide that it will be resolved by Uber.
[102]See also above at [44].
[103]See above at [23].
This anonymity would be unremarkable if Uber were a taxi service, but it is not consistent with the idea that the drivers and passengers form a contractual relationship through delivery of a service that a third party, Uber, has promised to the passenger. A passenger could not reasonably be expected to think they were contracting with the driver when they got into the car. They ordered the service from Uber and agreed to pay Uber for it.
The Chief Judge found that Uber is in the business of providing a transport service in which it controls marketing, price and terms of service for its own benefit.[104] The Court of Appeal agreed, stating that Uber’s core business proposition is to make a substantially homogeneous passenger transport service to riders.[105]
[104]EmpC judgment, above n 2, at [51]–[52].
[105]CA judgment, above n 2, at [225].
In our opinion these conclusions were plainly available to the Courts below. The bare account of Uber’s business model in this section of the judgment demonstrates that it would be remarkable had they found otherwise. It follows that we agree with the Court of Appeal that Uber’s contracts are “window-dressing” to the extent that they assert Uber is a mere facilitator of transport services between drivers and users.
Does Uber engage drivers to deliver the transport services it supplies to users?
The Court of Appeal reasoned that “drivers contract with Uber to provide transportation services to riders” whether or not there is also a rider–driver contract.[106] It relied on a concession made by Mr Wicks in argument that drivers provide services to Uber.[107] The Court found that “a driver contracts with Uber to provide transportation services to riders” and stated that this went no further than Mr Wicks’ concession.[108]
[106]At [185].
[107]See especially at [180].
[108]At [183].
In this Court Mr Walker, for the appellants, explained the concession as going no further than acceptance that drivers may provide Uber with some services, instancing referral fees that Uber may pay when a driver refers a new driver to the platform. Clearly that is not how the Court of Appeal understood it. However, there is some ambiguity about the language of the concession and we heard no argument on the point from Mr Cranney.
On the view we take of the case, nothing turns on the concession, if such it was. It must follow from the Employment Court’s finding above that Uber is in the business of supplying passenger transport services that Uber also engages drivers to deliver those services. Otherwise, as the United Kingdom Supreme Court put it in Aslam, Uber would have no means of performing its contractual obligation to supply passengers with trips.[109]
[109]Aslam, above n 4, at [56].
That is also how Uber earns its revenues; it charges riders for trips. Although it describes itself as a technology business, it does not make money by distributing its software. Its revenue takes the form of fares paid. It pays the drivers by remitting a percentage of the fare to them.
We conclude that the findings in the Courts below that Uber engages drivers’ services for hire or reward to supply its passenger transport services were similarly available.
Does Uber engage drivers under contracts of service?
That brings us to the second sub‑question posed above at [15]: whether Uber engages its drivers under contracts of service. If it does, they are employees for purposes of the ERA. If it does not, they are independent contractors.
Following Bryson, employee status should normally be addressed by reference to the traditional common law tests: control, integration and the “own business” or fundamental test. As noted earlier, Bryson establishes that while there is no prescribed methodology s 6 does reflect those tests, which were important determinants of employee status at common law.[110] We agree with the Court of Appeal that it was appropriate to apply them directly in this case.[111]
[110]See above at [72]–[73].
[111]CA judgment, above n 2, at [133].
The cases identify considerations which point toward independent contractor status.[112] They include expertise vis-à-vis the hirer, freedom to work for others at the same time, the right to subcontract performance, the provision of tools and equipment, carriage of a risk of loss and opportunity for profit, limitation of the hirer’s supervision to setting the parameters of the task, and payment on a per-task basis.[113] Considerations which point towards employee status include the payment of wages based on time worked, the employer’s control of the method of doing the work, the obligation to display the employer’s insignia, the use of training or disciplinary processes, and the employer’s ownership of any goodwill arising from the work.
[112]A number of them were drawn together in the judgment of Lord Wright in Montreal v Montreal Locomotive Works Ltd [1947] 1 DLR 161 (PC) at 169. See also Lee v Chung [1990] 2 AC 374 (PC) at 382 citing Market Investigations Ltd v Minister of Social Security [1969] 2 QB 173 (QB) at 184–185.
[113]See, for example, Lowe v Director-General of Health [2017] NZSC 115, [2018] 1 NZLR 691 at [43] per Arnold and O’Regan JJ; and Leota v Parcel Express Ltd [2020] NZEmpC 61, (2020) 17 NZELR 395 at [38].
However, it has long been recognised that the strength of the common law tests lies in their capacity to adapt to changing social conditions and new ways of working. Speaking of the control test, Mason J said in Stevens v Brodribb Sawmilling Co Pty Ltd that emphasis had moved from the exercise of control in fact to the right to exercise it:[114]
It is said that a test which places emphasis on control is more suited to the social conditions of earlier times in which a person engaging another to perform work could and did exercise closer and more direct supervision than is possible today. And it is said that in modern post-industrial society, technological developments have meant that a person so engaged often exercises a degree of skill and expertise inconsistent with the retention of effective control by the person who engages him. All this may be readily acknowledged, but the common law has been sufficiently flexible to adapt to changing social conditions by shifting the emphasis in the control test from the actual exercise of control to the right to exercise it, “so far as there is scope for it”, even if it be “only in incidental or collateral matters”.
[114]Stevens v Brodribb Sawmilling Co Pty (1986) 160 CLR 16 at 28–29 (citation omitted) quoting Zuijs v Wirth Brothers Pty Ltd (1955) 93 CLR 561 at 571 per Dixon CJ, Williams, Webb and Taylor JJ.
In this case the Chief Judge emphasised that control may be exercised in different, but no less effective, ways in an environment where the means of control are online:[115]
[64] The short point is that the way in which an employer exerts “authority” and “control” over an employee (classic hallmarks of an employment relationship) has acquired less direct, but equally effective, characteristics due to developments in technology. That makes them no less potent, or relevant, to the s 6 analysis.
[115]EmpC judgment, above n 2.
Similarly, in Personnel Contracting, Gordon J cautioned that workers may too readily be classified as “own business”, given the reality of modern working arrangements, or if they are in the gig economy or working in one or more businesses in the same week.[116] The same point can be made with respect to an apparent absence of integration. As a number of authorities have emphasised, when faced with new ways of working it is important to examine the relationship as a whole and apply all relevant criteria.[117]
[116]Personnel Contracting, above n 75, at [181].
[117]See, for example, Prasad, above n 69, at [91]–[93].
After examining these factors she concluded that each of the four drivers was in an employment relationship when carrying out driving work for Uber.[164]
Court of Appeal
[164]At [93].
The Court of Appeal identified two questions:[165]
(a)Did the Employment Court misdirect itself on the application of s 6?
(b)Did the Employment Court err by misapplying the test in s 6 or was the Court’s conclusion so insupportable as to amount to an error of law?
[165]CA judgment, above n 154, at [15]. Leave had also been granted on a third question relating to joint employment but the appellants had abandoned their appeal in respect of that question so the Court did not address it: at [17]. See also above at [7] per Winkelmann CJ, Williams and Miller JJ.
The Court of Appeal answered the first question in the affirmative. It considered that there was a material misdirection “whether one focuses on the specific aspects of the Chief Judge’s approach … or the overall thrust of the judgment”.[166] The Court of Appeal said Chief Judge Inglis’ approach arguably:[167]
… involved an expansion of the category of workers properly seen as employees, by reference to concepts of vulnerability and the need for protection, and who was working in whose interests”.
The Court considered that, “[a]t the least, her approach added glosses which have the potential to complicate, and distract attention from, the orthodox s 6 inquiry.”[168]
[166]At [137].
[167]At [136].
[168]At [136].
The Court did not consider it necessary, because of its answer to the first question, to consider the second question.[169] Nor did it refer the matter back to the Employment Court for reconsideration under s 215 of the ERA. The Court of Appeal was mindful of the direction in s 216 that it must on appeal have regard to the specialist nature of the Employment Court but considered that in this case it could consider the matter afresh, as it had been urged to do by the parties:[170]
…in the present case we are satisfied that we have the benefit of factual findings made by the Chief Judge, we have all the relevant evidence before us, there is no material contest between the parties about the factual circumstances in which the drivers carried out their work, and it is in the interests of justice that this question be determined without the need for a further hearing.
[169]At [139].
[170]At [142], and see at [141].
The Court began with what it said was the first stage of the s 6 inquiry: “inquiring into the real nature of the parties’ mutual rights and obligations”.[171] The Court of Appeal analysed the written agreement between the parties and concluded that “the substantive rights and obligations described in the driver agreement do not, on their face, appear to give rise to an employment relationship”.[172] But the Court questioned whether these aspects of the driver agreement could be taken at face value.[173] The Court reviewed the realities of the relationship and concluded that “many of the provisions designed to point away from employee status are window-dressing”.[174] For example, it held that fares are in reality determined solely by Uber and that “the apparent flexibility in relation to fares contemplated by the driver agreement is a fiction”.[175] The Court also held that the apparent flexibility to choose whether to accept or reject rides offered to them and to cancel rides differed in reality, largely due to the system of incentives and sanctions applied by Uber.[176] The same applied to the apparent flexibility to choose their own route and to engage in “multi-apping” (using both Uber and another ridesharing app simultaneously).[177]
[171]At [143], and see at [97].
[172]At [177].
[173]At [178].
[174]At [204].
[175]At [184].
[176]At [192]
[177]At [193]–[196].
The Court considered that the contractual matrix as a whole confers on Uber a high level of control over the way the drivers work when logged onto the app.[178] The Court did not find great assistance in industry practice because there is no single model of service provision through online platforms.[179]
[178]At [204].
[179]At [208].
With regard to intention, the Court held that the intention of the parties must refer to their “common intention about the substance of their mutual rights and obligations, objectively ascertained”.[180] The Court commented that the evidence in the Employment Court as to the subjective intention of both the drivers and Uber itself should therefore preferably not have been given.[181] The Court did not consider that:[182]
… there are any indications of the intention of the parties’ intentions that provide material assistance, other than the indications of intention implicit in the objectively ascertained mutual rights and obligations of the parties having regard to the driver agreement and related documents and the realities of the parties’ working relationship.
[180]At [210].
[181]At [211].
[182]At [217].
The Court then went on to consider the common law tests of control, integration and the “fundamental test”, saying that applying these tests to the real nature of the relationship between the parties to determine whether the relationship amounts to a contract of service was the second stage of the inquiry.[183] The Court assessed these tests taking into account only the situation when the drivers were logged into the Driver app.[184] The Court concluded that the drivers could not, when logged into the app, be said to be carrying on transport service businesses on their own account. Drivers have no opportunity to establish business goodwill of their own, nor to influence the quantity or quality of work they receive, nor, except to a limited extent, their revenue.[185] The right to work for other employers was discounted because employees can have that right.[186]
This Court’s decision in Bryson
[183]At [219]–[220].
[184]See at [234].
[185]At [234]. Much of the so-called freedom to decide whether or not to accept a job, for example, was in reality only available to those who had high status (meaning they would rarely avail themselves of that freedom): see at [192(a)].
[186]See at [233].
This Court in Bryson overturned the majority decision of the Court of Appeal holding that Mr Bryson was not an employee.This was because there had been no error of law in the determination by the Employment Court that he was an employee.[187]
[187]Bryson, above n 155, at [31]–[40].
Judge Shaw in her decision in the Employment Court had summarised the principles she considered to have been established by the Employment Court in earlier cases on s 6 of the ERA as follows:[188]
·The Court must determine the real nature of the relationship.
·The intention of the parties is still relevant but no longer decisive.
·Statements by the parties, including contractual statements, are not decisive of the nature of the relationship.
·The real nature of the relationship can be ascertained by analysing the tests that have been historically applied such as control, integration, and the “fundamental” test.
·The fundamental test examines whether a person performing the services is doing so on their own account.
·Another matter which may assist in the determination of the issue is industry practice, although this is far from determinative of the primary question.[[189]]
[188]Bryson v Three Foot Six Ltd [2003] NZEmpC 164, (2003) 2 NZELR 105 [Bryson (EmpC)] at [19].
[189]The Judge had rejected a submission made on behalf of Mr Bryson that, under the Employment Relations Act 2000 [ERA], evidence of industry practice should be completely disregarded. She said that would be contrary to the common law and would mean that the Court could not take account of matters which were important to the parties. She said that the ultimate decision in a case like that one depended upon the entire factual matrix: at [20]–[21].
This Court found that Judge Shaw’s list of principles “accurately state[d] what the Court must do and list[ed] the matters which are relevant”.[190] Significantly for this appeal, this Court considered intention relevant but found that Judge Shaw had given reasons for not being able in that case to find a common intention (meaning that there was no error of law).[191]
Submissions of the parties
The appellants
[190]Bryson, above n 155, at [32].
[191]See at [9], [32], [33] and [35].
The appellants submit that there is no need to depart from the framework in Bryson. In their submission, application of the common law tests should allow for consideration of modern ways of working. In this case it is submitted that the real nature of the relationship between Uber and the four drivers is not one of employment because:
(a)the Services Agreements between the parties, each read as a whole, indicate that employment was not intended;
(b)in practice, the drivers acted in a manner consistent with being self‑employed;
(c)Uber did not exercise relevant kinds of control over the drivers (that is, subordination);
(d)the drivers were not structurally or from a presentational perspective integrated into Uber’s business; and
(e)the drivers made business decisions about revenue and expenses and so were in business on their own account.
It is submitted further that, in this case, the Court of Appeal wrongly used a two-stage inquiry:[192]
(a)it looked to the contract and practice to identify the substance of the parties’ mutual rights and obligations as a matter of reality (identification of what is being classified); and
(b)it then determined whether those rights and obligations amounted to a contract of service (application of the criteria for classification).
[192]CA judgment, above n 154, at [97]–[98].
In the appellants’ submission, by dividing the analysis in this way, the Court of Appeal failed to look for indications of the parties’ intention as to the status of their relationship. Intention is (and has always been) a criterion of classification, not merely a tool of interpretation. In respect of the other common law tests (control, integration and the fundamental test), the appellants accept that a two-stage approach is implicit in the Bryson analysis. For those tests, the appellants submit that courts have always determined the rights and obligations before deciding whether those features are consistent with an employment relationship.
Looking at the relationship overall, it is submitted that it is not an employment relationship. The drivers chose whether to work, and if they did, when, where and for how long. They provided all required equipment, including a vehicle. They were not paid in wages. They worked for competitors whenever they chose. The only significant thing the appellants determined was the pricing algorithm but, if drivers did not like a price, they could decide not to accept a trip, cancel a trip while it was underway, or log off altogether.
The respondents
The respondents submit that the Court of Appeal correctly assessed the intention of the parties by assessing objectively the common intention. It then applied the control, integration and fundamental tests and, again correctly, concluded that the drivers were employees.[193]
[193]The respondents also submit that the Employment Court was right to conclude that s 6 has a broader role than that identified by Uber but they do not elaborate on that submission.
It is submitted that the Court of Appeal’s reasoning is consistent with decisions in other comparable jurisdictions with the exception of Australia, which is an outlier.[194] In the respondents’ submission, while the driver terms describe the relationship in a manner consistent with Uber’s preferred view, they do not reflect the real nature of the relationship.
[194]The respondents refer in detail to the United Kingdom Supreme Court decision in Uber BV v Aslam [2021] UKSC 5, [2021] 4 All ER 209. The appellants point out that this decision is against the background of a different statutory scheme under which there is an intermediate class of “workers” who are self-employed but who provide services as part of a profession or business undertaking carried on by someone else. “Workers” are entitled to some but not all rights of employees: see at [38]. The conclusion of the United Kingdom Supreme Court in Aslam was that the Uber drivers came within the third category, “workers”, which is not contained in our legislation. We accept that this diminishes Aslam’s significance, but the United Kingdom Supreme Court’s conclusions on certain aspects related to the common law tests are still of some assistance.
The respondents submit that, when all relevant matters are considered — including statements that describe the nature of the relationship, matters that indicate the intention of the persons, and divergences from and supplementation of the terms and conditions apparent from the way in which the relationship works in practice — the conclusion is clear. The drivers are employees because they are controlled by Uber, are integrated into Uber’s business and are not in business on their own account.
The respondents point out that all aspects of the asserted driver/passenger agreement are determined by Uber. Uber has complete control over it and the right to modify and add to it unilaterally. All performance standards and pricing are determined by Uber as are all other aspects of its service.
Issues
We now turn to consider the following questions:
(a)Did the Employment Court err in law (as the Court of Appeal found)?
(b)Did the Court of Appeal also err in law?
(c)What is the proper approach to s 6 of the ERA?
(d)Were the four Uber drivers employees?
Did the Employment Court err in law?
In our view the Court of Appeal was correct, for the reasons it gave, to hold that the Employment Court had misdirected itself in law.[195] The Court of Appeal was also entitled to decide for itself whether or not the four drivers were employees using the correct legal test, particularly as it was urged to do so by the parties.[196]
[195]See CA judgment, above n 154, at [118]–[137]. We differ from the majority as to whether the Chief Judge meant to suggest that inequality of bargaining power could extend the statutory concept of what constitutes an employee: see above at [89] per Winkelmann CJ, Williams and Miller JJ. We consider that the Court of Appeal was correct to read her judgment as holding that inequality of bargaining power may extend the statutory concept of employment: see CA judgment, above n 154, at [120(a)] and [124]–[125].
[196]See CA judgment, above n 154, at [142].
We accept the submission of the appellants that the proper test is the Bryson test without modification. Including vulnerability based on inequality of bargaining power would mean adding a dimension not included in Bryson, and we do not consider that it helps analytically with the operation of the tests.[197] For example, whether relevant control or integration arises from inequality of bargaining power or just because that was the agreement between the parties with equal bargaining power is irrelevant under most of the common law tests.
[197]Unlike the majority, we accept the appellants’ argument that it is not until an employment contract is established that s 3 comes into play: contrast above at [92] per Winkelmann CJ, Williams and Miller JJ.
We accept that vulnerability may be relevant to ascertaining actual common intention and in determining how much weight it should be given, but reference to s 3 is not needed for that as inequality of bargaining power is already part of the caselaw in that regard.[198] It may also, as the majority in this Court and the Court of Appeal both suggest,[199] account for the presence of contract terms designed to mask the true nature of the agreement between the parties. But again, recourse to s 3 is not needed as it is the true nature of the relationship that is at issue, whether or not any attempt to mask it arises because of vulnerability of one of the parties.
Did the Court of Appeal err in law?
[198]See Leota v Parcel Express Ltd [2020] NZEmpC 61, (2020) 17 NZELR 395 at [6], where the Employment Court noted that, where one party does not understand the implications of agreeing to the label suggested by the other, “the way in which the relationship operated in practice is likely to be more revealing of its true nature”. See also at [5].
[199]Above at [91] per Winkelmann CJ, Williams and Miller JJ; and CA judgment, above n 154, at [104].
We accept the submission of the appellants that the Court of Appeal in turn erred in law by effectively discounting intention.[200] As recognised by this Court in Bryson, the parties’ common intention as to the status of their relationship at the time of entering into the relationship remains part of the common law test, albeit not decisive.[201] As we understand the Court of Appeal decision, it held that the nature of the relationship is ascertained on the basis of the contract between the parties, and the reality of the relationship if it differs from that described in the contract, and it is then assumed that the parties must have intended that their relationship be of that nature.[202] This is circular and would always mean that actual common intention has effectively no role to play. While contracts are interpreted objectively (under the common law at least),[203] in this case what is required is the assessment of a relationship (not merely the contract) and the parties’ actual common intention at the start of that relationship. This has human dimensions not necessarily limited to contractual terms nor indeed how the relationship later operates in practice.[204]
[200]See CA judgment, above n 154, at [108] and [114].
[201]Bryson, above n 155, at [31]–[32].
[202]See CA judgment, above n 154, at [108].
[203]The position is different under civil law systems: see Helen Winkelmann, Susan Glazebrook and Ellen France “Contractual Interpretation” (2020) 51 VUWLR 463 at 473 citing Chartbrook Ltd v Persimmon Homes Ltd [2009] UKHL 38, [2009] AC 1101 at [39] per Lord Hoffmann.
[204]It follows that we differ in some respects from the majority’s view above at [75]–[76]. In particular, we do not agree that subjective intention is irrelevant in this context where it is evidence of actual common intention.
If both parties genuinely intended their relationship not to be one of employer/employee, that should at least be accorded some weight.[205] This may lead to a conclusion that factors pointing towards an employment relationship are outweighed by factors pointing the other way, at least in marginal cases. Instead, the Court of Appeal, having effectively discounted actual common intention at the first stage of its analysis, did not consider it at all in the second part of its analysis alongside the other common law tests. Intention is, however, one of the factors to be considered, as confirmed by this Court in Bryson.[206]
[205]Even if the parties are wrong and the real nature of the relationship is in fact one of employment when analysed in accordance with all of the common law tests.
[206]If intention was meant to be merely a function of the court’s ultimate view of the relationship, then it would not have been identified as a separate factor by Judge Shaw: see Bryson (EmpC), above n 188, at [19], as approved by this Court in Bryson, above n 155, at [32].
This Court in Bryson did not require the division of the inquiry into two stages.[207] In our view this division by the Court of Appeal was in itself an error of law, particularly as it led to the Court of Appeal wrongly discounting actual common intention as one of the relevant factors under s 6. The overall test is “the real nature of the relationship”,[208] meaning essentially how the arrangements operate in practice. We agree with the majority that it would usually make sense to start with consideration of the contract as that is the basis of the arrangement, but it does not have controlling effect.[209] Practice can override or supplement the contract. Whether or not the relationship is one of employer/employee is decided under the common law tests, including intention.
[207]See above at [73] and [75] per Winkelmann CJ, Williams and Miller JJ.
[208]ERA, s 6(2).
[209]See above at [75] per Winkelmann CJ, Williams and Miller JJ.
We consider too there is an issue (as the appellants point out) with only looking at the situation when drivers are on the Driver app. This may make sense for the consideration of the control test (although one indication of control causing a person to be an employee may be control outside of working hours), but it could certainly distort the consideration of whether a person is in business on their own account and we think could distort consideration of the integration test as well.[210] We consider this an error of approach on the part of the Court of Appeal that also amounts to an error of law in that it unduly restricts the relevant factors under the common law tests.
[210]For instance, because only considering the times when drivers are on the Driver app could suggest a higher level of integration into Uber’s business than is truly the case.
One of the traditional markers for an independent contractor relationship is the ability to choose whether or not to work (including for other parties) and the time and often place of work. Only to look at the time the drivers are on the Driver app risks effectively removing from consideration the fact that there are no minimum hours[211] and that there is no restriction on drivers driving for anyone else, including direct competitors.[212] This does not mean that these factors are controlling or that all gig workers would be independent contractors. It just means that these factors should be considered along with all other relevant factors under the common law tests.[213]
What is the proper approach to s 6?
[211]As the majority does above at [122].
[212]This ability to work for others has traditionally be seen as a strong indicator of independent contractor status: see, for example, Koia v Carlyon Holdings Ltd [2001] ERNZ 585 (NZEmpC) at [43]. While other ridesharing platforms may not have a significant presence in New Zealand at present, this could change, and taxi services are still competitors to Uber. Also the drivers are able to pursue work other than driving passengers or meal delivery and can engage in other activities as they wish.
[213]Contrast above at [122] per Winkelmann CJ, Williams and Miller JJ.
As noted above, in our view the proper approach to be applied is the Bryson test without modification. This involves ascertaining the real nature of the relationship between the parties using the common law tests, including their actual intention at the time of contracting. In this case the whole relationship is set out in the contractual documents (including policies).[214] It does not appear that practice departs from those documents, although the characterisation of the relationship in those documents is rightly challenged by the respondents as discussed below. The issue for this appeal is whether, judged on the correct test (which, as noted above, the Court of Appeal failed to apply), the real nature of the relationship, as outlined in those documents, is one of employer/employee.[215] We now examine those common law tests as they apply to the four drivers at issue in this appeal.
Were the four Uber drivers employees?
Intention
[214]See the majority’s description of these documents above at [25]–[48].
[215]Because we take the view that the Court of Appeal erred in law, our task is to consider for ourselves whether the respondents were employees, given we were not asked to remit the proceeding to the Employment Court. Contrast the majority’s approach above at [10].
The actual common intention of the parties is judged at the time of entry into the contract. Uber’s intention clearly was that the drivers not be employees. We consider that the intention of the four drivers in fact matched that of Uber: to run their own businesses as independent contractors. Mr Ang, for example, said that he had answered an advertisement on Trade Me and that what attracted him was the representation that he could earn $35 per hour at a minimum. He noted that “[t]he advertisement also stressed being one’s own boss and referenced flexibility, autonomy and independence.”[216]
[216]Whether a contract is presented on a “take it or leave it” basis is in our opinion irrelevant: contrast CA judgment, above n 154, at [112]. The issue is the actual common intention of the parties. For example, if franchise agreements to operate a franchise business are presented on “take it or leave it” terms, this would not mean that a franchisee who accepts the offer did not intend to operate their own business.
All of the four drivers could and did intend to fit Uber around other commitments, including family commitments. Mr Abdurahman, Mr Keil and Mr Ang all signed up to work on other ridesharing apps.[217] Mr Rama worked part-time as a chef and Mr Ang worked part-time as a massage therapist.[218] All four drivers signed a contract saying they were not employees[219] and presumably filed taxation returns in accordance with that status.[220]
Control
[217]Mr Ang stopped using the other app as he found it was not worth it financially: see above n 114 per Winkelmann CJ, Williams and Miller JJ.
[218]EmpC judgment, above n 154, at [72].
[219]There is no basis in the evidence for saying the four drivers did not understand that they were signing a contract containing that provision. One had, for example, been a human and workers’ rights advocate in his native country, had a degree from a New Zealand university and had worked in New Zealand as a union organiser and for a member of Parliament. Another had a master’s degree from a New Zealand university and was involved in the management of several community organisations. A third also had a degree from a New Zealand university and had worked for various government agencies. This is not to discount the evidence from some of the drivers that they had trouble understanding all of the details of the contracts.
[220]Including deducting from their assessable income the vehicle running costs related to earning their income from Uber. Employees, unlike independent contractors, are not entitled to deduct from their assessable income any employment-related expenditure: Income Tax Act 2007, s DA 2(4); and see John Prebble and Grant Pearson Fundamentals of Income Taxation (Thomson Reuters, Wellington, 2018) at [51.7].
We accept the submission of the respondents that there are very strong elements of control by Uber. Uber controls the price,[221] the terms of service for drivers and passengers (which can be changed by Uber at will),[222] the complaints and refund system, and the ratings system (which is not one that allows customers to choose drivers based on rating and so its only function is as a measure of control by Uber of the drivers).[223] There is limited interaction between the customer and driver which is largely limited to interaction while driving. Otherwise, communications are with Uber (either via the platform or via service centres, again part of control by Uber over the drivers).[224]
[221]The United Kingdom Supreme Court likewise considered this a significant factor in Aslam, above n 194, at [92] and [94].
[222]See at [95].
[223]See at [99].
[224]See at [100].
The rewards system incentivises platform entry at times of peak demand.[225] Drivers can decline rides (or cancel them) when on the system at all reward levels, but there are consequences.[226] Information on trip direction and duration and therefore fares is only available for those who have high reward status. This does allow more informed choice but refusal rates allowed are still low.[227] In our view, incentives can amount to control and do in this case.
[225]As the majority note above at [26] and [61], trips completed during peak hours receive more points.
[226]See above at [59]–[60] per Winkelmann CJ, Williams and Miller JJ.
[227]See above at [26] and [60] per Winkelmann CJ, Williams and Miller JJ. Although we note that allowing only low cancellation rates once a ride is accepted is understandable.
The fact that certain aspects of the control exercised by Uber relate to transport regulations, health and safety, and quality of service does not mean such measures can be disregarded,[228] although it is relevant the extent to which similar standards and regulatory compliance would be required of independent contractors.
Integration
[228]The United Kingdom Supreme Court in Aslam likewise found that the fact that some aspects of the way Uber operated its business were required for compliance with the regulatory regime was not a “reason to disregard or attach less weight to those matters in determining whether drivers are workers”: Aslam, above n 194, at [102].
There are strong indicators of integration into Uber’s business.[229] We agree with the majority, for the reasons they give,[230] that Uber is in the business of providing passenger transport services to users. We agree that Uber engages drivers to deliver the transport services it supplies to users.[231] This means that the drivers are fundamental to Uber’s business.
[229]Contrast CA judgment, above n 154, at [226].
[230]Above at [98]–[109].
[231]Above at [110]–[114].
The areas discussed above which Uber controls (including price, quality standards and communications) are also an indication of how integrated the drivers are in Uber’s business. The areas where drivers have some autonomy are limited to vehicle choice, expenses of running the vehicle, and deciding when and if to enter the platform. But the incentives ensure entry at a time which suits Uber’s business. That this also serves the interests of the drivers and the public does not diminish Uber’s interest in this.
It is significant that there is no ability for a driver to build their business by virtue of contact with Uber customers. This means the customers are Uber’s customers and not those of the driver. That the drivers are not in uniform and that the cars are not branded Uber is of limited significance, given that ordering through the platform means that users know they are dealing with Uber and that is the reason they use the platform. Other factors such as the complaints system and advertising also reinforce our conclusion.
The fact that the drivers own their vehicles and thus provide capital essential for the transport services may point against integration. On the other hand, Uber collects the fares and retains the whole fare for a period before passing their share onto the drivers. This means it has the use of the fares for working capital for a period. This points also to integration.
Fundamental test
The drivers own and manage their vehicles, and they choose when, where and how long to work. They are free to do other work including for other transport providers.[232] These factors are traditionally seen as strong indicators of independent contractor status.[233] In this case, however, these factors are overridden by the extent of the drivers’ integration into Uber’s business, as well as the drivers’ lack of control over the quantity and quality of the work they receive, the price paid for it and their inability to develop their businesses through contact with users.[234]
Conclusion
[232]Although, as the majority note, the Court of Appeal found that in practice using multiple ridesharing apps is not feasible: see above at [124].
[233]We also consider that it is an important advantage for independent contractors that they can deduct expenditure incurred in earning their assessable income, employees not being able to deduct any employment-related expenses: see above n 220. This advantage should be attached some weight in this case: contrast above at [128] per Winkelmann CJ, Williams and Miller JJ.
[234]See Aslam, above n 194, at [101] where the Court noted that these factors of control meant drivers had “little or no ability to improve their economic position through professional or entrepreneurial skill”.
Overall, the factors pointing away from employee status[235] are outweighed by those pointing towards it.[236] The four drivers were therefore employees.
Result
[235]The factors pointing away from employee status include intention, the drivers’ ability to choose their hours of work and to work for others (including Uber’s direct competitors), plus vehicle ownership.
[236]The factors pointing towards employee status include integration, control and the lack of realistic ability for the drivers to develop their own Uber businesses.
We agree with the majority that the appeal should be dismissed and we also agree with the costs order.
Solicitors:
Russell McVeagh, Wellington for Appellants
Oakley Moran, Wellington for Respondents
0
14
0