Commuters on Adelaide's Gawler and Seaford rail lines have lived through two changes of employer without the timetable, fare or platform ever moving. Before 2021 the trains were run directly by the state. From 31 January 2021 the same services ran under a private operator, Keolis Downer, inside an eight-year contract worth around $2.1 billion (approx.), signed by the Marshall Liberal government. At 3am on the first Sunday of February 2025, control passed back to the state; the trams followed at 3am on the last Sunday of July 2025. Nothing about the railway itself changed on either date. What changed, twice, was who was accountable for it — and what each change cost.
That second change is this audit's subject. The Malinauskas Labor government presents the return of trains and trams to public ownership as a clean correction of a Liberal-era mistake — a 2022 election promise delivered largely on schedule, without paying the operators a cent in exit penalties. That is a genuinely creditable piece of negotiation. But "no exit penalty" is not "no cost," and "back in public hands" is not "fully public": the maintenance contracts behind both networks run to 2035 regardless of who controls the trains. The fairest starting point is what the original privatisation was meant to achieve, and whether reversing it holds up as well as the government's messaging suggests.
What the Marshall government privatised
South Australia's rail and tram network was run in-house by the state-owned TransAdelaide for most of the past three decades, while the metropolitan bus network was progressively contracted out from the late 1990s. The Marshall Liberal government, elected in 2018, extended that logic to trains and trams. On 2 July 2019, Transport Minister Stephan Knoll released a combined tender covering the tram network and a bus contract, with a rail tender to follow. Knoll's case rested on two claims: a private operator, free of public-sector constraints, could run the network more efficiently, and the resulting savings could be reinvested in services — echoing the case made for Victoria's earlier rail privatisation, which Knoll cited directly. Tram operations moved to a private consortium in 2020, and from 31 January 2021 train operations transferred to Keolis Downer Adelaide under the eight-year, roughly $2.1 billion (approx.) contract, with options extending toward twelve years. On the government's own modelling, the deal was projected to save the state around $118 million (approx.) over that twelve-year horizon.
What Malinauskas reversed, and what it cost
The 2022 election turned that projection into a live dispute. Peter Malinauskas's Labor opposition campaigned on returning trains and trams to public ownership and banning further privatisation, winning government on 19 March 2022. In November 2022, new Transport Minister Tom Koutsantonis released the department's own reassessment: rather than the promised $118 million saving, the arrangement was on track to cost the state up to $120 million (approx.) over the same twelve-year window. The stated cause was specific: more than 100 (approx.) rail staff who had not transferred to Keolis Downer in 2021 remained on the public payroll, unassigned — a standing cost not offset by any private-side saving.
On 2 April 2023, the government announced a negotiated deal with Keolis Downer Adelaide and tram operator Torrens Connect to end both contracts years early. The state avoided paying termination fees the operators had put at up to $94 million (approx.) combined — instead agreeing a phased handback, funded by around $30 million (approx.) over two years for trains and a further $3 million (approx.) for trams, close to $33 million (approx.) in total. Trains returned to state control at 3am on Sunday 2 February 2025; trams followed at 3am on Sunday 27 July 2025. Two things did not change on those dates. Customer service and security functions are not due to fully transfer until June 2027, four years after the deal was announced. And Keolis Downer keeps the rail maintenance contract, Torrens Connect the tram equivalent, until 2035 either way. Adelaide's bus network — six regional contracts re-let by the Marshall government from July 2020 for an eight-year term with a two-year extension option, running to at least 2028 (approx.) — was not part of the 2023 deal and remains privately run.
The strongest case
A contract renegotiated on the state's terms, not the operators'
Give the Malinauskas government credit for the mechanics of this deal, because governments that campaign against a contract usually pay heavily to exit it. Keolis Downer Adelaide and Torrens Connect held a strong legal position: an unexpired eight-year contract, only in its second year, with a termination clause the operators valued at up to $94 million (approx.) combined. A government chasing a quick political win could have paid that figure, cancelled outright, and declared victory within months. Instead, Koutsantonis's department negotiated a phased handback timed to when the state was actually ready to resume operations — avoiding the termination fee entirely and spreading the transition cost over two budget years rather than one large payout. That is a materially better outcome than the alternative the government had every incentive to choose.
There is also a coherent case for public control of an essential, single-network service, independent of which operator runs it more cheaply in a given year. A metropolitan rail and tram system has no real competitor on its own corridor — commuters cannot switch to a rival line the way they might switch retailers — so the usual argument for privatisation, that competition disciplines price and quality, does not apply in the same way. Direct government operation puts the incentive to invest in reliability in the hands of the body that is also politically accountable for the outcome, rather than a private operator answering primarily to shareholders and to service levels fixed years earlier.
The evidence
Set the claims of both governments against the public record, and a mixed picture emerges: genuine fiscal discipline in how the reversal was executed, alongside real costs, a slower timeline than the announcements suggest, and two competing dollar figures that neither government has had independently checked.
The clearest number in this dispute is the swing in the government's own modelling: a projected $118 million (approx.) saving became, on the incoming government's 2022 reassessment, a projected $120 million (approx.) cost over the same contract horizon — a gap of roughly $238 million (approx.) between two governments' estimates for the identical trains and trams. Neither figure is an independent audit finding; both are internal departmental modelling released by the government of the day to support the policy it had already chosen. South Australians have never seen a third, disinterested assessment of what privatisation actually cost or saved — only two self-interested numbers, three years apart, pointing in opposite directions.
The Department for Infrastructure and Transport's own annual reports show total public transport use rising 12.6% (approx.) in 2023-24 against the prior year, then a further 2.9% (approx.) in 2024-25 — the year trains, and from July, trams actually returned to public control (DIT Annual Report 2023-24; DIT Annual Report 2024-25). At first glance that looks like support for the government's claim that public control improves service. Look closer and the pattern cuts the other way: the larger increase came in 2023-24, a full year before a single train changed hands, and growth then slowed to well under a quarter of that rate in 2024-25, the year the handback happened. Both years also sit inside the nationwide post-pandemic recovery in transit use recorded across every mainland capital, making it hard to credit the ownership change specifically.
No independent verification of either central figure — the $118 million claimed in 2019 or the $120 million found in 2022 — appears in the public reporting reviewed here. Section 39 of the Passenger Transport Act 1994 (SA) requires the Auditor-General to examine passenger transport contracts once awarded, and the Auditor-General's office has separately noted its audit of the department's 2023-24 financial statements was delayed pending information from the department. Until a disinterested reconciliation of the full twelve-year comparison is published, South Australians have two competing numbers from two governments with an obvious interest in each result — not a settled figure for what privatisation cost or saved.
The critique
Judged only on the terms the Malinauskas government has set for itself — no termination fee, an on-schedule handback, rising patronage — the reversal looks like an unambiguous win. Judged against the fuller record, three qualifications matter.
First, the case for reversing privatisation rests on the same kind of unaudited, self-interested modelling the case for privatisation originally used. The $118 million saving Knoll promised in 2019 and the $120 million cost Koutsantonis announced in 2022 were both produced inside the department, released by a government that had already settled on its preferred policy, and neither has been tested by the Auditor-General in any report identified here. A masthead that holds governments to dated, sourced figures cannot treat the second number as more trustworthy than the first simply because it points toward the policy this article's steel-man supports on principle. Both figures deserve equal scepticism until an independent office reconciles them.
Second, "back in public hands" overstates what has actually happened. The transition has cost the state a genuine sum — around $33 million (approx.) over two years — on top of whatever the stranded-workforce cost identified in 2022 continued to run while the handback was negotiated. Customer service and security staff, the employees passengers actually deal with, are not due to move to government control until June 2027 — six years after Labor first promised the change, and more than four years after the deal was struck. The two private operators keep the maintenance contracts — the work of keeping trains and trams mechanically running — until 2035, regardless of who is driving them. A commuter told the trains are "back in public hands" is being given a simplified version of an arrangement split three ways: driving and planning to government by 2025, customer-facing staff by 2027, and maintenance to two private contractors for a decade after that.
Avoiding a $94 million exit fee is real fiscal discipline. It is not proof that public ownership costs less to run — that comparison has never been independently checked.
Third, the patronage growth cited as evidence does not, on close reading, support the government's claim. The larger increase — 12.6% (approx.) in 2023-24 — happened before a single train changed hands; growth then slowed to 2.9% (approx.) in 2024-25, the year the handback actually occurred, and both figures sit inside a nationwide post-pandemic recovery that has little to do with who signs the paycheques of Adelaide's train drivers. And the principle behind the reversal — that an essential, single-network service should not be run for private profit — has not been applied to Adelaide's bus network, which remains under six regional private contracts extended by the Marshall government and untouched by the 2023 deal. The government has floated buses as a future target without a timetable or a cost, meaning the same test applied here will need repeating whenever that announcement comes.
The risk to the future
A split-accountability model with a decade still to run
The practical risk is not that public operation will fail outright — several Australian capitals run their own networks without incident. It is that the current arrangement, split between government-run driving and planning, government-run customer service from 2027, and privately-run maintenance until 2035, creates exactly the accountability gap that makes it hard for anyone — government, opposition, or a commuter on a delayed platform — to say cleanly whose decision caused a given failure. A mechanical fault tracing back to maintenance still under a Keolis Downer or Torrens Connect contract lets the government that promised "public hands" point to a contractor it does not control. A service failure tracing back to planning or staffing now inside government leaves the department that spent around $33 million (approx.) and three years of political capital reversing privatisation with nowhere left to point.
The second risk is fiscal. Two competing, unaudited estimates — $118 million (approx.) claimed as a saving in 2019, $120 million (approx.) claimed as a cost in 2022 — are now part of the public record without either being reconciled by an independent office. If South Australia moves on buses next, as the government has signalled it might, voters will be asked to trust similar departmental figures for a much larger contract group, with no more independent verification than existed the first two times. The lasting lesson of this decade-long round trip is not which side of the privatisation argument was right. It is that neither side has been required to prove its numbers to anyone outside government — and until the Auditor-General closes that gap, every future transport contract decision here will be argued the same way: on the government's own modelling, unchecked.