However, common sense doesn’t apply when it comes to poverty. The problem lies with the way ACOSS (and other poverty researchers) arbitrarily define the “poverty line” - as an income below 50 per cent of the median income.

But the median income (and hence the poverty line) shifts up and down as the economy booms or slumps. As growth slowed down, the median income fell, depressing the poverty line, and magically reducing the number of “poor” people below it. But in reality, nothing much changed.

Most poverty in Australia is also transitory. The 2.2 million people under the poverty line this month are not the same as those last month. People fall sick and their income falls below the poverty line, but then these people recover and go back to work. Women take time off work to have a baby but go back to work later. Students may have incomes below the poverty line, but not after they graduate and start working.

The incomes of people who are between jobs may fall temporarily until they get back on their feet (about two-thirds of people on Newstart Allowance are on the payment for less than a year before returning to work).

Arguing for reducing poverty by giving “the poor” more money is difficult when the people who make up “the poor” keep changing.

This preoccupation with relieving the symptoms of poverty through government handouts diverts attention from the real cause of poverty - the lack of paid work. The ACOSS figures show that more than 80 per cent of people under the poverty line do not have a full-time job, and more than 50 per cent aren’t in the workforce at all.

Households with at least one adult in full-time work rarely fall below the line.

This means there are two ways to reduce “poverty”: increase the value of welfare benefits faster than the value of wages, or move substantial numbers of people off welfare and into full-time jobs. Anti-poverty campaigners invariably emphasise the first option and neglect the second, but the first actually undermines the second.

Ross Gittins reviewed the latest ACOSS poverty research and concluded that the dole should be increased to raise the incomes of the unemployed beyond the poverty line. ACOSS too want more generous benefits.

What this would mean in practice is that people who don’t work and whose incomes are below the poverty line would have their incomes raised by more than people in low-paid jobs whose incomes are just above the poverty line.

Not only would this be unfair but it is also counterproductive, resulting in more people on welfare than before. Massaging poverty numbers by increasing welfare benefits makes the real poverty problem worse, for the main cause of poverty is joblessness, and people will be less inclined to take jobs as benefits become more generous.

In the United Kingdom (in an echo of Hawke’s famous promise to Australia in the 1980s) the Blair government promised to end child poverty. The way it went about it was to increase benefits. As a result, many families moved from just below the arbitrary poverty line to just above it, but the effective result of this huge splurge in public spending was that very little in their lives really changed.

The government in Britain has learned from this mistake. It now wants to broaden the poverty targets so policy solutions don’t focus exclusively on giving people more money. Instead, policy will focus on tackling the causes of poverty - things like truancy rates, drug and alcohol dependency, and chronic intergenerational unemployment.

Anti-poverty campaigners in Australia should learn from this. It’s easy to tweak the poverty numbers by increasing welfare spending, but such a policy is likely to prove self-defeating because it increases dependency and misses the principal cause of poverty, which is joblessness, not low benefits.

Peter Saunders is a Senior Fellow and Andrew Baker is a Policy Analyst at The Centre for Independent Studies.