Gig Worker Minimum Wage Australia: What the Fair Work Draft Order Really Says

On 8 July 2026, the Fair Work Commission published a draft order that would, for the first time in Australian history, set minimum pay rates for gig delivery workers. The headlines called it a 25% pay rise. A federal political party posted that drivers were “set for a pay rise to at least $31.30 an hour.” Several news outlets have reported it as though it is already law.

It is not law. It is a draft. Submissions on it close in a matter of days. And when you read the actual documents – the 30-page decision and the 17-page draft order, both of which are public – the story is considerably more complicated, and far more interesting, than any headline has suggested.

This guide is built entirely from those primary documents: the Expert Panel’s decision [2026] FWCFB 167 and the draft On-Demand Delivery Employee-like Worker Minimum Standards Order. Everything below is quoted or drawn directly from them. Where the Commission criticises the deal, we quote the Commission. Where the union defends it, we quote the union.

What Actually Happened

Australian law created a new category of worker: the “employee-like worker” (ELW) – someone who is not an employee, but is not really an independent business either. Delivery riders are the textbook case. The Fair Work Commission can set minimum standards orders for them.

The Transport Workers’ Union applied for one. But here is the part that is not widely understood: the proposal that reached the Commission was a consent deal negotiated with the platforms themselves. The TWU’s own covering letter said so:

The draft MSO has been prepared in conjunction with Portier Pacific Pty Ltd t/as Uber Eats and DoorDash Technologies Australia Pty Ltd t/as DoorDash. As such, the draft reflects a consent proposal developed between the parties.

TWU covering letter, quoted at [8] of the decision

So this is not a union victory imposed on unwilling platforms. It is a negotiated settlement between the TWU, Uber Eats and DoorDash, which the three of them then jointly asked the Commission to bless. The Commission held a five-day hearing (11-15 May 2026), heard objections, and has now published a draft – with significant criticisms of the very deal it is proposing to make.

The Expert Panel was Justice Hatcher (President), Vice President Asbury and Commissioner Connolly.

The Actual Rates

Almost every report has quoted a single figure: $31.30. The real table has four vehicle classes, and the rate you get depends on what you ride or drive. Straight from clause 14.2 of the draft order:

From 1 January 2028, and annually thereafter, the rates would be indexed to the percentage increase in the National Minimum Wage from the Commission’s Annual Wage Review (clause 14.7).

Note also who is not covered: clause 2.3 excludes anyone using a vehicle with a carrying capacity of more than 1 tonne. And note who is: coverage extends to the delivery of “consumables” – food, beverages and liquor – and grocery items. The order’s dictionary defines “collection” to include “shopping for and/or selecting and/or purchasing”, which means the people picking your Woolworths or Coles order off the shelf are covered too.

It Is Not an Hourly Wage. This Changes Everything.

Every headline has described this as riders “getting $31.30 an hour”. That is not what the draft order does. It does something quite different, and the difference is the whole story.

Clause 14.1 of the draft order says:

An ELW must be paid no less overall than the earnings floor over any earnings period set by the DLPO which may be up to 21 days… The “earnings floor” for an earnings period is the number of hours of engaged time… multiplied by the minimum hourly rate.

Draft On-Demand Delivery Employee-like Worker Minimum Standards Order, clause 14.1(a)

Unpack that and there are three separate mechanisms hiding inside a single sentence, each of which reduces what it delivers.

Mechanism 1: It is a top-up, not a wage

You are not paid $31.30 an hour. You continue to be paid per delivery, exactly as now – fares, surge, Quests, promotions, all unchanged. At the end of the period, the platform compares what you actually earned against the floor. If you earned more than the floor, you get nothing extra. If you earned less, they top you up.

The Commission is explicit about this at [92]: the system “does not involve any modification of the current arrangements whereby ELWs are paid per delivery… but it serves to place an income floor under these arrangements.”

Mechanism 2: "Engaged time" excludes waiting

This is the big one. The floor is calculated on engaged time, which the order defines as running from when you accept an engagement to when you complete it. Sitting in a car park logged in and waiting for an offer is not engaged time. The order expressly excludes “non-engaged time… or time during which the ELW is not engaged in providing the services”, along with breaks, breakdowns, and time after a customer cancels.

So a courier who is online for 20 hours but engaged for 14 of them has an earnings floor built on 14 hours, not 20 – even though all 20 hours were given up, and even though the platform controls how many offers arrive.

Work out what that means per hour of your actual life:

This is not our observation alone. It was put squarely to the Commission by the Rideshare Driver Network, which opposed the deal on exactly this ground. From the decision at [54]:

The RDN opposed the proposed MSO on the basis that its minimum rates only applied to “engaged time” rather than all time that an ELW is logged onto a digital labour platform, and thus enabled systematic unpaid waiting time under the control of the platform.

[2026] FWCFB 167 at [54]

That objection is on the public record, and the draft order was published anyway. Whether it is answered before 29 July is now up to the submissions.

Mechanism 3: The platform picks the averaging window – up to 21 days

The floor is not tested daily, or per shift, or even weekly. It is tested over an “earnings period set by the DLPO which may be up to 21 days”. The platform chooses.

Why does that matter? Because averaging hides bad periods inside good ones. A courier who has one strong week well above the floor and one dreadful week well below it will, over a 21-day window, average out above the floor – and receive no top-up at all, despite having worked a week for less than the minimum standard.

The shorter the earnings period, the more protective the floor. The longer it is, the less often it bites. And the draft lets the platform set it, up to the maximum. This is a genuinely significant design choice, and it has received no public attention whatsoever.

And the platform can strike out your engaged hours

Clause 14.4 sets up a three-strikes mechanism. If a platform “reasonably believes” one of your completed engagements included non-engaged time, it issues a first notification. If it believes it again, a second notification. After that, any further periods it reasonably believes include non-engaged time simply do not count as engaged time – which lowers your earnings floor.

To be fair, there are safeguards: the platform must tell you what it is relying on, must tell you the aggregate non-engaged time each period, the notifications expire after six months, and you can dispute. And there is an obvious anti-gaming purpose here – stopping people accepting an order and then going for a coffee. But the effect is that the platform, which controls the app and the data, also decides which of your hours count toward your own minimum.

Who this actually helps – and it does help someone

Let us be scrupulously fair, because the reform is not worthless. It is targeted, and it hits its target.

Courier B is exactly who this is for – and for them it is a substantial, meaningful gain. The Commission itself says at [92] that the order “is likely to substantially improve the earnings of a large number of ELWs“, and it is right.

But understand what that sentence means. It improves the earnings of the worst-paid. If you are a reasonably effective courier working the rush, this reform will very likely change your pay by nothing at all – and no headline has told you that.

What the Fair Work Commission Itself Says About the Deal

Here is the part that ought to have led every news report. The Commission is proposing to make this order. It is also, in the same document, eviscerating the rates in it. Both things are true, and the tension between them is the real story.

The rates are below the casual award rate for every single vehicle

The Commission built its own comparison table. It set the proposed gig rates against what an employed delivery driver doing comparable work is legally entitled to under the relevant Award.

The Commission’s conclusion, at [80], is worth reading twice:

The above table shows that the rates in the proposed MSO are below the Award casual hourly rate for all types of vehicles to be covered by the MSO, with no allowance as a result for recovery of any costs necessarily arising in the performance of the work… For a vehicle up to 750 kg carrying capacity, the Award labour component would be $31.64, leaving less than a dollar per hour for cost recovery.

[2026] FWCFB 167 at [80]

"No apparent underpinning mathematical logic"

The most damning line in the entire decision, at [93]:

Because the proposed rates of pay were the outcome of a negotiation between the TWU, Uber and DoorDash, they have no apparent underpinning mathematical logic.

[2026] FWCFB 167 at [93]

The rates were not costed. They were bargained. $31.30 is not the output of a model; it is the number three parties shook hands on. The Commission says so, in terms, in the document proposing to adopt it.

The union admitted it under cross-examination

At [77], the decision records the TWU’s witness, Mr Boutros, being cross-examined. He agreed the proposed rates were “a long way off full costs recovery”, and explained the union’s strategy: it was better to lock in a substantial lift and real protections quickly, and pursue full cost recovery later. The exchange ends:

So I think your answer is it’s better than what they’ve got but it’s not costs recovery? – That’s right.

Transcript, 11 May 2026, quoted at [77]

That is an honest answer, and the strategy behind it is defensible – we set out the case for it later in this guide. But it is a very long way from “delivery drivers are getting a 25% pay rise.”

The Commission rejected Uber and DoorDash's arguments outright

The platforms argued the rates do provide adequate cost recovery. The Commission’s answer at [78] is blunt: “We do not accept this position… it is plain, we consider, that the proposed rates do not provide for cost recovery.”

Two of the platforms’ specific arguments were dismantled, and both matter to any courier.

Argument 1: "Your car is a sunk cost, so it costs you nothing"

Uber and DoorDash argued, on expert evidence, that a courier’s vehicle should be assigned zero cost – because most couriers already own the car and mainly use it privately, so it is a “sunk cost”. The Commission, at [82]:

We do not accept this proposition. The use of a primarily personal vehicle for additional work purposes will necessarily hasten the depreciation of the vehicle, and that is recognised as a business cost by the tax system… the amount of vehicle depreciation for work use may end up being relatively small, but it is not zero.

[2026] FWCFB 167 at [82]

Argument 2: "You get tax deductions, so you already recover your costs"

This one is worth every courier’s attention, because it is a misconception a lot of drivers hold about themselves. Uber submitted that a courier “who in fact uses a pre-existing vehicle already obtains a measure of cost recovery through the tax system“. The Commission, at [83]:

That proposition is misconceived and is rejected. The tax system does not provide for the recovery of the cost of capital assets used in business; rather, it recognises that the cost of such assets is a business expense such that it ought not be included in taxable income. It remains necessary for the ELW to bear the expense of business use of the relevant asset.

[2026] FWCFB 167 at [83]

And the Commission found the work is worth just as much

The platforms argued gig delivery is fundamentally different from employed delivery work. The Commission accepted the business models differ – but on the actual labour, at [79]:

The actual labour performed by ELWs… is not in any significant way different to an employee engaged to perform metropolitan delivery work. No difference in work value is identifiable.

[2026] FWCFB 167 at [79]

And at [91] it goes further: if anything, gig delivery may have a higher work value, because “the work is more dynamic, involving tighter timeframes for delivery and a less regular and predictable flow of work” – and grocery delivery adds the task of picking and packing items off supermarket shelves.

So: the same work. Possibly harder work. Paid below the casual rate for it, with the vehicle to come out of that.

The Same Union Wants 72% More for the Same Vehicle

There are two other cases running alongside this one, covering “last mile” delivery – the parcel couriers who deliver for the likes of Amazon and Australia Post. The TWU has applied for minimum standards orders there too.

And what the TWU is asking for in those cases, for the same vehicles, is dramatically more than what it agreed to accept from Uber Eats and DoorDash. The Commission set the two side by side at [87]:

Note that the last-mile claim also carries a separate payment per kilometre – actual cost recovery, of the kind this deal does not have at all.

The Commission does the arithmetic at [88]. Assuming a conservative 25 kilometres travelled per hour, the last-mile rate for a petrol car up to 750 kg is 72 per cent higher than the equivalent rate in this deal. For a one-tonne van, it is 84 per cent higher.

To be fair to the TWU, the Commission also notes at [89] that the last-mile rates it is claiming are themselves significantly above the established owner-driver rates, and that the expert cost model behind them was criticised in ways that “have a degree of substance”. The truth may well sit between the two figures. But the gap is enormous, and it is the same union arguing both.

And this is not merely a curiosity. It is the reason the Commission is making this order interim only. Amazon, Australia Post and other last-mile businesses argued – reasonably, the Commission found at [90] – that they could end up “saddled with significantly higher minimum rates of pay than Uber and DoorDash”, putting them at a competitive disadvantage. The Commission calls this a potential “regulatory distortion of competition“, and it is why the whole thing will be reviewed once the last-mile cases are decided.

One more number, from the QUT research cited at [97]: average pay in the last-mile delivery sector is already more than $10.00 an hour higher than in on-demand delivery. The gap this deal is being celebrated for closing is smaller than the gap that already exists between two sets of couriers doing, in the Commission’s own words, work with no identifiable difference in value.

The Clause That Explains Why Uber and DoorDash Signed

It has puzzled a lot of people that the platforms did not fight this – that they co-proposed it. Uber Eats and DoorDash are not in the habit of volunteering to pay more.

The answer is probably sitting in clause 16 of the draft order, and nobody has written about it.

The set-off provision

Clauses 16.3 and 16.4 deal with what happens if a platform treats you as an “employee-like worker”, complies with the order – and then a court later finds you were actually an employee all along.

In that situation, the draft says that if any court or tribunal assesses a claim for employee entitlements, then

the full value of any and all benefits and payments provided by the DLPO to the deemed worker for the deemed work will be taken to be paid in satisfaction of, and set off against, any and all employee entitlements arising with respect to or in connection with the deemed work.

Draft Order, clause 16.4(a)

In plain English: if you later prove you were really an employee, the platform gets to deduct everything it has already paid you from what it owes you.

Misclassification litigation – the argument that gig workers are employees in disguise, entitled to wages, leave, superannuation and unfair dismissal protection – has been the single largest legal threat hanging over these platforms worldwide. Australian courts have so far found Uber Eats drivers are not employees (see Gupta v Portier Pacific [2020] FWCFB 1698, cited by the Commission at [79]). But that risk has never gone away.

It also switches off four state laws

Less remarked upon, but significant: clause 16.1 provides that where this order applies, these laws do not apply to the worker or the platform in respect of those services:

• Chapter 6 of the Industrial Relations Act 1996 (NSW)
• Owner Drivers Forestry Contractors Act 2005 (Victoria)
• Chapter 10A of the Industrial Relations Act 2016 (Queensland)
• Owner Drivers (Contracts and Disputes) Act 2007 (Western Australia)

These are the state-based regimes that have historically given contract carriers avenues to pursue minimum rates and unfair-contract relief. Chapter 6 in New South Wales is the source of the General Carriers Determination – the very instrument the Commission used earlier in its own decision to show what a properly-costed rate looks like.

So a federal order that pays below the casual award rate would also displace the state regimes that produce higher, properly-modelled rates. Whether that is a fair trade for the protections the order brings is precisely the sort of question submissions exist to raise.

The Rest of the Order: What It Takes, and What It Genuinely Gives

The pay clause has taken all the attention. But the order runs to seventeen clauses, and several of them matter a great deal – in both directions. Some codify burdens onto the worker in black and white for the first time. Others give riders rights they have never had.

Clause 4: every cost is yours, and it says so

This is a minimum standards order, and one of its clauses is a list of things you must pay for. It is worth reading in full, because it removes any ambiguity:

• 4.1 – you must keep your vehicle registered, at your own expense.
• 4.2 – you must acquire, maintain and repair your vehicle, at your own expense.
• 4.3 – you must pay all the running costs, at your own expense: “fuel, maintenance and parts, administrative costs and the costs of purchasing or leasing a vehicle”.
• 4.4 – you must pay for any licences, permits or eligibility checks the platform requires, at your own expense.

Set that beside the Commission’s own finding that the rate leaves less than a dollar an hour for cost recovery, and the shape of the deal becomes very clear. The order does not merely fail to fund your vehicle. It expressly places the whole of that burden on you, in the same instrument that sets your minimum.

Clause 8: your parking fines are your problem

Clause 8.1 provides that you pay any fines or penalties imposed on you, with no entitlement to reimbursement – unless you incurred the fine “because the ELW has followed an express direction of the DLPO”.

Consider what that means in practice. Delivery work in a dense city routinely involves stopping where you should not stop, because there is nowhere else to stop. The economics of the job depend on it. The platform benefits from the speed. But the ticket is yours, unless you can show the platform expressly directed you to park there – which, of course, it never will.

Clause 9: the insurance clause every courier must read

This one is genuinely important, and it cuts both ways.

On the other side, clause 9.2 requires the platform to take out and maintain personal accident insurance for you, at its expense, at “a reasonable minimum level of cover”. That is a real obligation and a genuine gain.

But read 9.3, which qualifies it:

• the cover “need not be held… at a level of cover equivalent to workers’ compensation schemes”;
• any benefit paid under it is set off against any workers’ compensation entitlement you might have;
• you cannot claim the same benefit under both.

So the order guarantees you accident insurance – while stating in terms that it need not be as good as the workers’ compensation an employee would get for the same injury doing the same work.

Clause 11: the "right to time away" is unpaid

Clause 11 is headed “Right to time away”, which sounds like leave. It is not. In full:

11.1 An ELW determines whether and when to accept engagements.
11.2 For the avoidance of doubt, an ELW can choose to take “time away” from the application by choosing not to accept engagements. Such time away will be without pay.

Draft Order, clause 11

In other words: you may stop working, and you will not be paid. That is not a new right. That is a description of what already happens to anybody who stops working. It is difficult to see what clause 11 adds, beyond making clear that nothing resembling paid leave is on the table.

What the Order Genuinely Gives You

We have been hard on the pay clause, because it deserves it. But it would be dishonest to leave the impression that the order is worthless. It is not. Several of its terms give riders things they have never had, and for many couriers these may matter more than a top-up they will never receive.

The Honest Case For the Deal

We have spent several thousand words on what is wrong with this order. Fairness now requires the other side, and the other side is stronger than the critics allow.

1. Right now, there is no floor at all

Whatever the shortcomings of $31.30 on engaged time, the alternative it replaces is nothing. Not a lower number – nothing. There is currently no minimum whatsoever for gig delivery workers in Australia. TWU surveys have reported riders earning as little as $12 an hour. For those riders, this order is transformative, and the Commission says so plainly at [92]: it is “likely to substantially improve the earnings of a large number of ELWs.”

2. The union made a considered strategic judgment

The TWU did not fail to notice that the rates fall short of cost recovery. It said so, under oath. Its position, as recorded at [77], is that after months of consultation with workers and members, it judged it better to secure a substantial lift and enforceable protections quickly than to spend years litigating for a perfect number.

That is a real argument, and the Commission itself half-accepts it at [98]: the TWU’s approach, treating these rates as “the first step towards rates which properly provide for cost recovery”, is “broadly consistent with the principles we have described“. Its criticism is narrower than it first appears – not that the strategy is wrong, but that the deal “does not offer any road map to the desired destination.”

3. Menulog is the argument nobody wants to make out loud

Here is the uncomfortable fact sitting underneath this entire case. Menulog closed its Australian operations on 26 November 2025, after 19 years, taking roughly a quarter of the market and around 120 jobs with it. Deliveroo left Australia in 2022. Foodora went before that.

Food delivery is a business with famously thin margins, and Australia has now watched three platforms leave. The Commission acknowledges at [92] that the agreed rates were tailored to the sector’s “capacity to commercially sustain the payment of higher minimum rates of pay.”

4. The protections are permanent; the rate is reviewable

Rates can be raised later. Dispute resolution, delegates’ rights, records, consultation and a mandatory information statement, once established, are structural. They change the balance of power in a way that a number on a page does not. And this order is expressly interim, to be reviewed once the last-mile cases are decided.

The Commission’s own verdict, at [92], is the fairest summary anyone has offered: the considerations for and against are “finely balanced.”

What This Means For You

You Can Still Change This. Submissions Close 29 July.

This is the part that makes reading all of the above worth your time. The draft is not final. The Fair Work Commission has invited workers, businesses and anyone else with an interest to make a submission – and it is genuinely open to riders.

Five things worth raising, if you deliver

1. Engaged time versus online time. The floor ignores the hours you spend logged in and waiting. Tell the Commission how many hours a week you actually give up compared with the hours you are “engaged”. Real numbers from real riders are exactly what is missing from this record.
2. The 21-day earnings period. The platform gets to choose an averaging window of up to three weeks, which lets a good fortnight bury a terrible one. Ask why the floor is not tested weekly.
3. Cost recovery. The Commission has already found the rates leave less than a dollar an hour for a car. Tell it what your fuel, servicing, rego and insurance actually cost you. A per-kilometre running rate – which the same union is seeking in the last-mile cases – would fix this.
4. Do tips count toward the floor? The order says you must be paid “no less overall” than the floor, without clearly saying whether customer tips count toward it. If they do, a customer’s generosity would subsidise the platform’s minimum obligation. That deserves a straight answer.
5. The three-strikes non-engaged-time rule. Clause 14.4 lets the platform stop counting hours it “reasonably believes” were not engaged. If you have been wrongly accused of anything by an app with no human involved, say so.

What happens next

• 29 July 2026, 4pm AEST – submissions close. The Expert Panel then decides whether to hold a further hearing.
• 10 August 2026 – the order commences, if the Panel decides to make it based on the draft.
• After that – the order is interim. Once notices of intent are published in the last-mile cases (MS2024/1 and MS2024/2), the Commission must begin a review within 21 days, and rates could change substantially.

We will update this page when the Expert Panel decides. In the meantime, if you take one thing from this guide, take this: the number in the headlines is not a wage, is not law, and for most couriers will not arrive at all. Read the documents. They are public. They are more honest than the coverage of them.

Frequently Asked Questions

Related Guides

• Uber Eats Driver Australia: Pay, Requirements and Real Earnings
• DoorDash Australia: Promo Codes, Fees, Groceries and Dasher Pay
• Food Delivery Jobs in Australia: Uber Eats, DoorDash and More
• ABN Work and Tax Guide for International Students