Published in the New Zealand Herald, 25 June 2026
The Accident Compensation Corporation exists to mend people. If you are hurt in an accident, the scheme pays for your care and some of your lost wages. In return, you give up the right to sue whoever caused it.
But the money was never the point. The point is to get you better and back to the job and life you had before.
So, one might expect the Accident Compensation Corporation (ACC) to check whether the people it pays for do recover and go back to work. It does, sort of. It stops your weekly payments and lets five weeks of silence pass, without ringing your employer or asking whether you have a job to go to.
On the ACC’s books, the day the money stops is the day you officially got better.
That might sound like harmless paperwork. It is not. The ACC is now leaning on that very number to tell the country its long crisis is over.
For about a decade, the scheme drifted and people took longer to recover. More of them got stuck on long-term support for sprains and broken bones that ought to heal, and the future cost of all those open injuries roughly doubled.
Then the Government acted, and fairly. It ordered a review, installed a new chair and made the ACC report its numbers every month. The long-term pool had been swelling by nearly 15% a year. By April this year, that growth had stopped dead. Good on everyone involved for getting that far.
The economist Charles Goodhart gave his name to a warning that fits the ACC’s return-to-work figure exactly. Once a measure becomes a target everyone is judged by, it stops reflecting the truth.
Because the figure treats the end of a payment as a return to work, it can be improved by ending more payments. That is close to what has happened. In the year to September, on Treasury’s figures, the ACC turned down cover or support almost 173,000 times, up 13.6%, and suspended weekly payments 80% more often.
In the year to June, a record of more than 8700 people came off long-term support. Yet on figures released under the Official Information Act, only 13 in every 100 went back to the job they had before, and for about half of the exits, the ACC could not say why.
The ACC is not the first corner of Government to manage numbers rather than the things those numbers measure. The country’s main gauge of surplus or deficit goes by the mouthful Obegal. When the ACC’s losses grew alarming, the Minister of Finance built a second version, Obegalx, where the x stands for “excluding ACC”. There was a respectable reason: the ACC’s swings do distort the books. But the timing, just as those losses turned awkward, was hardly coincidental.
Some of the tightening was overdue. The ACC had been paying for things that injury no longer explained, and trimming that is good housekeeping when it is using other people’s money. Still, the headline number cannot tell good housekeeping from simply shoving people out the door. Both look identical on paper.
But a worker who has not healed does not vanish from the state’s books. The cost just moves somewhere else, for example, to a hospital waiting list or a benefit. It could also be borne by the patients themselves, which is of course not what the ACC once promised.
Treasury looked at what drove the improvement, and the answer is awkward for the cutters. It was not, mainly, the cutting. The biggest lift came from a change made before the turnaround plan existed. The ACC brought back one-to-one case managers after years of call-centre queues. Real people who know your name and follow your case.
That change cost real money but it is the single most effective thing the ACC has done. The ACC’s external actuary reckons that getting recovery right could take $500-$800 million off the future cost of existing claims within two years.
Cutting people loose looks cheap. It is not. Spending properly on recovery looks dear, but it is where the real savings are. The decent thing and the prudent investment really line up in this case.
So here we are. If you decide that recovery means the day the payments stop, you can always report a recovery. You can hit every target and clear the long-term list. The one question that decides whether any of it lasts, whether the injured person actually got better, is the one your measure was never built to ask.
The ACC was created to get injured people back into working life. A number that does not reflect whether that happened says nothing about success. It only records relief at no longer having to pay.
For its own financial viability, and for the welfare of the people it is responsible for, the ACC can do better than that.
