2:15pm 12 August 2026
Australia’s new minimum rate for food-delivery drivers has been celebrated as a “world-leading” victory for workers. It is also a revealing example of how the federal government approaches industrial policy. Announce the immediate benefit, but ignore the second-order effects, leaving consumers to discover the true cost later.
From 17 August, delivery platforms must ensure drivers receive between $31.30 and $32 an hour of “engaged time”, depending on the vehicle they use. The rates will rise by another 50 cents in January 2027.
The Albanese government legislated the new “employee-like worker” regime, empowering the Fair Work Commission to regulate contractors who had previously fallen outside the traditional employment system. The Commission then made the order following a joint proposal developed by the Transport Workers’ Union, Uber Eats and DoorDash.
$31.30 is below the $33.05 national minimum payable to casual employees once the 25 per cent loading is included. Gig workers also receive no paid annual or personal leave and must generally pay their own vehicle, fuel, maintenance, registration and third-party insurance costs. The Commission claims that the new rates fall below the relevant casual award rates and do not provide full recovery of drivers’ expenses.
However, ministers cannot take credit for providing “world-leading” protections while disclaiming responsibility for their consequences. The order is the product of the industrial-relations framework Labor deliberately created.
Its first problem is the misleading comparison with an ordinary wage.
Australia’s National Minimum Wage is currently $26.44 an hour. The new $31.30 bicycle-delivery floor is almost 18.4 per cent higher. A full-time employee earning the national minimum wage must ordinarily attend work at specified times, follow a roster and remain available to an employer for the duration of a shift.
A delivery driver accepts none of those constraints.
Gig workers decide whether to work, when to log on, which jobs to reject and when to go home. There is no manager refusing a request to finish early, insisting on a Saturday-night shift or asking why someone has not appeared for work. A driver can deliver for 40 hours one week, two hours the next and nothing the week after.
That autonomy has economic value. It is not an incidental perk that can be ignored when comparing the two forms of work, it is the central feature of the bargain.
This demonstrates why the government’s attempt to create an employment safety net without an employment relationship is so fraught. It wants drivers to retain contractor-style freedom while progressively acquiring employee-style protections. Each individual entitlement sounds attractive. The cumulative cost and distortion receive much less attention.
The Commission expressly accepted that drivers have no obligation to log on and, even when logged on, no obligation to accept an offered job. It nevertheless treated their labour as comparable to that of casual delivery employees because the physical task, collecting and transporting goods, is similar. That comparison considers what drivers do while completing a delivery but undervalues the freedom surrounding it.
The evidence before the Commission showed that supplemental income was the principal motivation for 70 per cent of drivers surveyed. Most “multi-app”, using more than one platform to find work. Uber described the labour pool as overwhelmingly part-time, supplementary and responsive to demand.
Some drivers undoubtedly depend on platform work to meet their basic needs, and there is a legitimate case for protecting them from exploitation. However, industrial policy designed around a minority of the most dependent participants can change the market for the much larger group that uses gig work intermittently precisely because it is open and flexible.
That is a crucial second-order effect, a policy intended to improve gig work may gradually make it less like gig work.
A higher rate does not guarantee a higher income
A statutory hourly floor also tells us little about what workers will earn each week.
The Australian order does not pay drivers $31.30 for every hour they are logged on. “Engaged time” generally begins when a delivery is accepted and ends when it is completed. Platforms can continue paying per delivery, using surge prices and incentives. They must provide a top-up only if a driver’s total payment over an earnings period of up to 21 days falls below the applicable hourly floor for recorded engaged time.
That distinction matters. If higher delivery costs cause platforms to offer fewer jobs, drivers may earn more for each engaged hour but spend longer waiting unpaid between deliveries. The government will point to a higher regulated rate. The driver may see little or no improvement in total income.
This is not theoretical. A recent National Bureau of Economic Research study of Seattle’s gig-work pay law found that the minimum standard increased base pay per task, however, highly active drivers completed fewer tasks, entirely cancelling out the increase in pay per delivery and producing no overall improvement in their monthly earnings. Drivers experienced more unpaid idle time and travelled farther between jobs.
The Seattle experience illustrates the basic problem with regulating a spot market for tasks as if it were an ordinary workplace. If the number of people who can log on remains unrestricted while the volume of deliveries does not rise, a higher price per task attracts or retains more available labour. Each worker can then receive fewer tasks.
Platforms may respond by limiting access, reserving work for their most productive drivers, introducing shifts or tightening performance management. Those responses may improve efficiency, but they erode the low barriers to entry and personal autonomy that attract people to gig work.
The policy may therefore produce fewer opportunities, more unpaid waiting and greater algorithmic control, all while being advertised as a victory for flexibility.
The unanswered multi-app problem
The new legislation also leaves an important practical question unresolved, what happens when a driver accepts work through more than one app at once?
The Fair Work Act expressly requires minimum standards to account for contractors working for multiple businesses, including work that may be performed simultaneously. The Commission would of course realise that most delivery drivers surveyed used multiple apps.
Yet the order does not contain an obvious mechanism for allocating overlapping time between platforms.
Where a driver accepts two deliveries through the same application, the order provides that the second period of engaged time generally begins after the first delivery is completed. However, a rival platform has no knowledge that its driver is already completing another company’s order.
Suppose a driver accepts an Uber Eats delivery and a DoorDash delivery travelling in roughly the same direction. Each platform records an engagement between acceptance and completion. Does each owe its hourly floor for the same 20 minutes? Must the time be divided between them? Does accepting the second job amount to delay, “non-engaged time” or fraudulent conduct? What evidence would establish that, particularly if neither delivery is late?
The order contains detailed procedures for platforms to exclude deliberately delayed time, but no express cross-platform offset or allocation rule.
This creates an obvious tension. If both companies can be required to count the same period, the worker may effectively receive two safety-net payments for one period of labour.
Before imposing a nationwide standard, policymakers should have answered this question. Instead, the legislation is coming into effect, and businesses, workers and ultimately courts will be left to resolve it.
This is another recurring feature of the government’s industrial policy, the first-order entitlement is prescribed centrally and benevolently, while the compliance problems and behavioural responses are pushed onto everyone else.
Someone will pay
The new labour cost will not disappear because the government calls the order a safety net. It must be borne through some combination of lower platform margins, increased restaurant commissions, higher delivery and service fees, fewer promotional discounts, reduced driver numbers and more intensive management of the remaining workforce.
Platforms may absorb part of the cost, though expecting multinational companies to permanently accept lower margins as an act of generosity is not an economic policy. Restaurants already operating on narrow margins are also poorly placed to absorb higher commissions. Consumers will inevitably carry at least part of the burden.
New York City’s own regulator projected that its delivery-pay rule could increase the average consumer cost of delivery by US$5.18 an order if platforms passed on the remaining labour cost. Seattle’s official post-reform data show that platform fees increased from 18.5 to 19.6 per cent of customer payments, even though delivery volumes continued to grow. New York City delivery-worker study; Seattle Office of Labor Standards
The international evidence is not uniformly negative. New York has recorded higher worker earnings and continuing demand, while Seattle’s official report presents a more favourable assessment than the NBER research. Nonetheless, even the more positive results reveal the trade-off of higher fees, greater productivity pressure and work concentrated among fewer or more active drivers.
In other words, higher regulated pay can be achieved, but not without changing who obtains work, how platforms manage them and what customers pay.
Those consequences are particularly relevant in Australia now. The ABS reports that meals out and takeaway food prices rose 4 per cent over the year to June, driven partly by wages and operating costs. The Reserve Bank says nominal unit labour costs (the measure of labour costs most relevant to firms’ prices) remain elevated. It has also revised its wage-growth forecasts higher partly because of annual wage reviews and other administered decisions. Australian Bureau of Statistics; Reserve Bank of Australia
This single order will not determine the national inflation rate, inflation is the accumulation of thousands of price and cost decisions across the economy. However, the Government cannot dismiss every intervention as too small to matter while continually adding new costs to labour-intensive services.
The pattern is familiar. A new entitlement is announced as a flag of progress. The costs are diffuse, delayed and transferred through supply chains, so they receive less political attention. When prices subsequently rise, businesses are blamed for profiteering and the government announces another subsidy or rebate to address the cost of living.
That is not a coherent economic policy. It is a cycle in which the government regulates costs upward and then spends public money trying to conceal the consequences.
A safety net for delivery drivers is defensible. Pretending it is costless is not.
At a time when Australians are already struggling with the price of everyday services, the federal government is again contributing to the cost-of-living problem through industrial policy that focuses on immediate beneficiaries while neglecting second-order effects. Higher delivery fees, fewer available jobs, and reduced flexibility will not appear in the ministerial press release, but they will appear in household budgets.
If the government wants credit for the wage increase, it must also accept responsibility for the bill that follows.
Leave a Reply