Health changes spread the pain

THE government's plans for private insurance would increase inequality.

ONE of the more contentious bills left over from the previous government and soon to be re-introduced into parliament means-tests the private health insurance rebate and increases the Medicare levy surcharge payable by higher-income taxpayers who do not take out private health insurance (PHI).

Speaking last February, then prime minister Kevin Rudd claimed the measure, twice defeated in the Senate, would permit enormous savings, to a cumulative amount of $100 billion. In contrast, health funds and the opposition claimed the changes would reduce private insurance coverage and thus increase the burden on public hospitals.

These claims that coverage would fall were incorrect. But so are the claims made by the government. The reality is that the proposed changes, though not likely to reduce coverage, will be seriously distorting.

Start with the vast savings claimed by the government. No doubt, substantial amounts are spent on the rebate: in 2008-09, about $4bn. As PHI coverage rises with income, much of this spending goes to high-income households. It therefore seems a case of inefficient churn, in which higher-income earners pay tax only to get some of those payments back as benefits. All else being equal, it would be better to eliminate the benefits and reduce tax rates, getting a double dividend: improved incentives and lower administrative costs.

In this case, however, matters are not so straightforward. The rationale for the PHI rebate is to offset another distortion: that public hospital services are provided without a direct user charge. Private hospital cover competes, to some extent, with public hospitals; absent a matching subsidy, its price, as reflected in PHI premiums, would be inefficiently high.

As a result, taking the structure of our health system as given, an efficient subsidy would reflect the costs private cover allows the public system to avoid. Reducing the rebate could therefore distort choice between private and public hospitals and increase healthcare costs overall.

However, and this is the nub of the issue, the government, despite what it says, is not proposing to reduce the subsidy to PHI: rather, it is changing its form.

Specifically, while means-testing would reduce the on-budget benefit to high-income households, the proposed rise in the surcharge increases the benefit those same households get, in the form of avoided tax payments, by taking out private cover. The government, in other words, is replacing a spending carrot with a no smaller, and for many households larger, tax stick.

It is precisely because the implied tax slug is so large that the government and Treasury are right: a fall in coverage is unlikely. But two points follow.

First, to claim this is a saving is nonsense. It is no more a saving than would be a law that replaced direct government funding of roads with a requirement that all households give $500 to a private road building fund or pay a $1000 tax fine. Clearly, such a measure would not reduce the total volume of resources extracted from taxpayers; only smoke and mirrors could make it seem otherwise.

The smoke and mirrors here are that resources redirected to particular uses through such tax penalties are off-budget. And, to make matters worse, Treasury, when it reports tax expenditures, treats benefits and penalties differently. For tax benefits, Treasury calculates the amount of tax forgone, which reflects the change in behaviour associated with the benefit. But for tax penalties such as the surcharge, it calculates the revenue raised.

This bears an inverse relation to the change in behaviour the penalty induces, as the greater the behavioural change, the less revenue is collected.

Were the surcharge increased even further, for example, no payments at all would be made, so its measured cost would fall to zero: the exact opposite of economic reality.

But there is a second, even more important, point: the change from the carrot to a bigger stick is by no means neutral.

Its economic substance is to increase effective marginal tax rates, with all the distortions that creates. Additionally, the Henry report notes the proposed change would cause serious implementation problems and "increase administrative complexity for policy-holders, insurance providers and the ATO". On top of all that, increasing the surcharge will make high-income consumers even less responsive to the price of private health insurance. Put slightly differently, with the penalty for not having PHI increased, the demand for PHI insurance by high-income earners will become less price elastic. That reduced price elasticity, which the Treasury modelling seems to ignore, creates incentives for increases in PHI premiums, probably by 10 per cent to 15 per cent above the levels that would otherwise have prevailed.

In the long run those increases will be taken mainly in the form of higher incomes by health professionals in the private system, as increases in charges in that system will meet less price resistance. This will have two further effects.

First, as the public and private systems compete for health professionals, it will increase costs in the public system, likely reducing service availability. Second, it will make PHI even less affordable for low-income consumers, ensuring that we truly do end up with two healthcare systems: one for the better off, who will be forced to have private cover, and one for those with lower incomes, who will have to rely on public hospitals.

If this is what the government wants, it is playing with fire, for the resulting political economy puts the future of the public system under even greater threat. It is difficult to believe immediate political gains from some accounting illusions could justify so dangerous a course.

None of this is to endorse our healthcare financing system. That system, including the rebate, needs a comprehensive overhaul to be even tolerably sustainable. It is a pity the Rudd government's health reforms simply ignored private insurance and the role it should play. But messing it up even further, as these changes would, merely squares the error.