John Armstrong writes that the working group's prescription for change may be blinkered by ideology such as time limits.

There it sits, a very large and seemingly very juicy statistic buried in the back pages of this week's discussion paper on welfare reform; a figure which screams for attention as indisputable evidence that the present benefit system must get the chop or, at a minimum, serious restructuring.

Or so the first offering of substance to emerge from the welfare reform think-tank established by Social Development Minister Paula Bennett would like the rest of us to believe.

The inclusion in its discussion document of a $50 billion estimate of how much those now on a benefit will cost the state over their lifetimes is a clear pointer to the direction in which Bennett's welfare working group - to give it its official name - is heading in terms of options for change.

Its recommendations are still some months away. The question is whether this advisory group will come up with something truly innovative to cut back the numbers of those who have spent a long time on the unemployment, domestic purposes, sickness and invalids benefits, or whether the group's prescription for change will be blinkered by ideology and it will recommend measures such as time-limited benefits which have had very mixed impact on beneficiary numbers in the United States.

The tone of the discussion document suggests ideology will win. While that may please some, though not necessarily all the present inhabitants of the Beehive, any such surrender to dogma will prove counter-productive in the long term if the Labour Party and the Greens cannot stomach the working group's recommendations.

The Government has accordingly sensibly ring-fenced some areas from the review - the level of payment that individuals get, for starters.

It is also seeking "practical" recommendations from the working group.

However, in setting up the group which is ably led by former chairwoman of the Commerce Commission Paula Rebstock, National has effectively shattered what was an informal consensus on welfare policy.

That consisted of National trying to look more caring towards beneficiaries after its track record in the 1990s and Labour trying to look tougher.

The net result has been the imposition of more obligations on beneficiaries to find work, backed up by increasingly tougher sanctions for non-compliance.

The working group will instinctively feel it has to be more radical than that to justify its existence. It could hardly serve up a centre-left package for a centre-right government. Likewise, its terms of reference requiring it to tackle "welfare dependency" effectively preclude it from endorsing the status quo.

It has accordingly been quick to denigrate the present benefit system as "outdated" and "unsustainable". It seems to have few scruples about the way it is going about trying to reinforce such a view.

The "future liability" of $50 billion in benefit payments is a case in point. Many news organisations covering Monday's release of the discussion document highlighted that figure. It gave a fresh angle on what was otherwise a rather long rehash of old arguments about the consequences of being dependent on a benefit and the desirability of finding paid work.

But the $50 billion figure is rather meaningless. You could add up the lifetime costs of paying someone state-funded superannuation, but that would not be a reason on its own for no longer paying it.

Beneficiaries are in a different political category to pensioners, however. The $50 billion figure has been concocted to paint the benefit system as an intolerable financial burden.

If this was part of a strategy to manipulate at an early stage what promises to be a humdinger of an argument once some reform options go on the table, then it worked even better than its proponents would have been hoping.

The $50 billion was widely misconstrued as being an estimate of the future size of the annual benefit bill. For that to be the case, the total amount budgeted for welfare payments would have to blow out by nearly eight times its present levels. The benefit system would have well and truly become unsustainable.

That is about the only funny note in a document that amounts to polemic masquerading as analysis. It is highly selective in citing statistics that suit its argument. It uses the experience of other countries to point to the relative failings of New Zealand's social assistance policies.

Perhaps worst of all, it makes assertions that are just plain wrong.

As part of the discussion paper's emphasis on getting beneficiaries into work, it cites Denmark, France, Germany, Sweden and Ireland as countries where, in contrast to New Zealand, sole mothers have higher employment rates than mothers with partners.

The OECD's family database tells a rather different story. In Ireland. mothers with partners outstrip sole mothers when it comes to paid employment - as they do in France and Germany, though not by much.

The spectacular fall in numbers of people on the unemployment benefit before the economy went into recession and the subsequent rise again as a result of that downturn would suggest the unemployment benefit is no longer a career choice, but remains a vital backstop when things get tough.

The working group seems to struggle with this, because economic recovery will bring another fall in the numbers on the dole, thereby removing part of its raison d'etre.

The group similarly tries to argue that the costs imposed on government by an ageing population will make the present benefit system unsustainable, even though the resulting demand for labour should see a significant and lasting cut in unemployment numbers.

To be fair, the group's focus here is more on the rising numbers on the domestic purposes, sickness and invalids' benefits.

The sickness and invalid categories are more problematic, but getting sole parents into work is not rocket science. The requirements are straightforward - availability of jobs, adequate training, adequate transport, adequate childcare and adequate pay to remove the large financial disincentives.

This costs money, lots of money.

In that light, the willingness to spend money in the hope of saving money down the track will be the true test of National's commitment to worthwhile welfare reform, regardless of what Rebstock's group recommends.