The ACCC says this deal will raise premiums. Now it must say by how much

The number that isn’t there

The Australian Competition and Consumer Commission has told Western Australians that premiums would go up if IAG buys RAC’s insurance arm, and in the same breath that it has not worked out by how much. For a decision to block a merger, that is fine. The legal test asks whether competition is substantially lessened, and that is a judgement about market structure, not an arithmetic exercise. More than half of a state’s motor and home and contents insurance in one company’s hands is a structural fact you can see without a model.

But the file has now moved. IAG is lodging a public benefit application, which asks a different question: do the benefits of the acquisition outweigh the detriments? That is not a judgement about structure. It is a balance. And a balance needs two quantities in comparable units.

Why the asymmetry matters

Here is what will happen, because it always happens. The applicant’s submission will be costed. Scale economies in claims procurement, reinsurance purchased at national rather than state volumes, technology spend spread across a bigger book, the reduced risk of a single-state insurer carrying concentrated exposure to one catastrophe-prone coastline — all of it will arrive with numbers attached, produced by economists who were hired to attach numbers to it. Some of those benefits are real. Reinsurance scale in particular is not a fiction.

On the other side of the scale, unless the regulator changes its posture, sits the proposition that premiums would increase, magnitude unstated. A decision-maker confronted with $X million of documented benefit and an undocumented detriment has been handed a thumb on the scale, and it does not much matter which way the decision-maker is personally inclined. The form of the evidence does the work.

This is not a hypothetical asymmetry. The ACCC routinely interrogates claimed synergies in authorisation matters — it demands to know whether they are merger-specific, whether they would be passed through, whether the applicant’s own board papers support them. That interrogation requires the agency to be numerate about insurance cost structures. An agency numerate enough to audit IAG’s synergy estimate is numerate enough to put a range on the premium effect. Declining to do so is a choice.

The word ‘partnership’ is doing work

The two parties described a sale of RAC’s entire insurance operations as a twenty-year partnership. That is their characterisation, not mine, but it points at the thing that makes this acquisition harder for consumers to notice than most. The club began in 1905 and the brand is close to civic furniture in Perth. If the badge stays on the policy document while underwriting, pricing and claims decisions move east, the ordinary signal that a competitor has disappeared — the name coming off the shopfront — never fires.

Markets discipline prices when buyers switch. Buyers switch when they can tell two sellers apart. A structure that preserves the appearance of a local rival while removing the substance of one is precisely the case where you would expect the measured price effect to be larger than the textbook, not smaller, because the competitive constraint disappears faster than the perception of it. That belongs in the detriment column, and it can be estimated.

Who gets the benefit and who pays

There is a distributional question the authorisation process should force into the open. RAC is a member-owned club. The proceeds of selling its insurance arm accrue to that mutual and, through it, to its members. The detriment — higher premiums across Western Australia’s motor and home markets — falls on everyone who buys insurance in the state, members and non-members alike.

Those are not the same population, and a public benefit test that nets them off without saying so is doing distribution by stealth. The ACCC should require RAC to state, on the record, what it intends to do with the money, because a benefit that materialises as member services is a different public benefit from one that sits on a balance sheet, and neither is the same as a price cut.

The case against this argument

The strongest objection is that quantifying premium effects in general insurance is genuinely hard and often meaningless. Premiums move with catastrophe experience, reinsurance cycles, building cost inflation and regulatory capital, and a merger effect of a few per cent would vanish inside that noise. Publishing a number invites false precision, and false precision gets litigated. Better to make an honest qualitative judgement than a dishonest quantitative one.

I accept the difficulty and reject the conclusion. The answer to a noisy estimate is a range with stated assumptions, not silence. Regulators in other sectors publish exactly that, with confidence intervals wide enough to be embarrassing, and the discipline of writing the assumptions down is most of the value. And if the honest answer is that the effect cannot be bounded, then the honest thing is to say so in the determination and explain why the applicant’s benefit estimates deserve more credence than the agency’s own inability to measure harm.

Refusing to measure is not neutrality. In a balancing test, it is a position.