The structure of public problems


There are parallel conversations going on in social policy, says Matt Cowgill, "Values on one level, data another". How values and data interact is an interesting question.

A decade ago, I was researching the debate over poverty. In 2003 the Senate Standing Committee on Community Affairs held an inquiry into poverty and small government activists and conservatives like Peter Saunders at the Centre for Independent Studies (CIS) were keen to make sure there was no increase in income support payments or easing up on mutual obligation requirements.

The anti-anti-poverty activists on the right were playing a defensive game. They made sure their opponents had to fight every step of the way from definition of the problem to policy response. And in the process they helped lay bare the structure of public problems.

The first step is to define the problem. The CIS argued that "Any definition of poverty is arbitrary and reflects the value judgements of those who develop it". While many people will accept they should defer to experts on issues of fact, most do not feel any need to defer on issues of value. So with this move, the CIS is saying that non-experts are entitled to form their opinion on what poverty is.

This forces the debate to the next step — arguing about why we should care about poverty as social scientists define it. The CIS argued that the problem social scientists and the welfare lobby were really concerned about was inequality. That’s why social scientists defined poverty in terms of relative income. This forced social scientists to have an argument about values — an argument where they couldn’t appeal to their status as experts.

Anti-poverty campaigns usually rely on statistics that show the problem is bad and getting worse. So the next step is to discredit the measurement of poverty. The CIS’s Peter Saunders had a long list of reasons the data couldn’t be trusted. For example, he argued that the Australian Bureau of Statistics itself said that data on the bottom 10% of the income distribution was unreliable.

If social scientists and anti-poverty activists still refused to give up, the next argument was about the causes of poverty. The CIS’s Lucy Sullivan argued that the real cause of deprivation was the welfare state. Easy access to welfare handouts meant that too many children were growing up in jobless single parent households and growing up without learning norms of chastity, thrift, industriousness and personal responsibility. Researchers in universities tended to assume this issue had been settled early last century and were surprised to see scholarly looking publications from US, UK and Australian think tank researchers being taken seriously.

If a lack of personal responsibility and adherence to social norms was the cause of the problem this raised another issue for debate. Who was morally responsible for solving the problem? This was another values debate where social science expertise had little authority. So even if researchers could show large numbers of people living in hardship, policymakers and members of the public felt entitled to ask whether this was really the taxpayers’ responsibility.

So by the time we got to debating the policy response, it wan’t at all clear what the solution was. According to people like Peter Saunders, the best policy response was much tougher mutual obligation requirements on jobless welfare recipients. In time mutual obligations would extend to parents sending their children to school and debate would move onto issues like income management.

The war over poverty was fought with arguments about both facts and values. In many cases it difficult to prise them apart. If the definition of poverty was shot through with value judgements, then so too was the data and theories about causation. And for many policymakers the problem shifted from poverty to welfare dependence. That suggested a whole range of new policy responses.

This entry was posted in Politics - national. Bookmark the permalink.
Notify of
Newest Most Voted
Inline Feedbacks
View all comments
derrida derider
derrida derider
8 years ago

As someone involved in that 2003 debate, I only partly agree with this retelling.

It’s true that people on both sides of this debate are very prone to put forward “factual” arguments that embody hidden and eminently disputable value propositions (if anything the poverty activists were the worst offenders here), but Saunders’ critique of the data on relative income poverty as data was pretty bloody solid – in truth we can say little about how poverty (as most people would define poverty) moved in the 80s and 90s because the data had systematic weaknesses that shifted over time.

But that of course that is a two edged sword – you can’t “prove” the welfare state caused poverty if you can’t even measure poverty. And Lucy Sullivan’s contribution was far more lightweight at both conceptual and analytic levels, to a degree that should have embarrassed the CIS (who are ideologues but unlike the IPA are not usually crude shills).

Don Arthur
Don Arthur
8 years ago

DD – ‘Bad’ Peter Saunders had an effective response to the Smith Family/NATSEM report.

Look at this 7:30 Report interview with Saunders and NATSEM’s Ann Harding. Interviewer Maxine McKew keeps up the pretense that it’s a technical debate between experts.

But it’s not a technical debate. It’s a debate over values. Half the average income is half the average income. Half the median is half the median. Why would you pick one measure rather than another? That’s the line of questioning McKew didn’t pursue.

David Walker
David Walker(@d-w-griffiths)
8 years ago
Reply to  Don Arthur

A little off-topic, but can we dump the name ‘Bad’ Peter Saunders? I have long admired the other Peter Saunders, but ‘Bad’ Peter Saunders seems to make reasonable contributions to debate on a regular basis. Popularising a label like ‘Bad’ Peter Saunders just seems an unpleasant thing to do to someone who has done nothing to merit it. If I were him, I might wear it as a badge of honour – or I might also dislike it intensely.

I wouldn’t let my kids talk this way, and I sort of see Club Troppo as my intellectual house.

Is there anything wrong with ‘the conservative Peter Saunders’?

And eat your broccoli, please.

8 years ago

Andrew Leigh in ‘Battlers and Billionaires’ takes the ‘inequality’ side commenting that he couldn’t find a ‘single authoritative study on relative poverty (less than half the median income) over the past 40 years’ worth contesting. In arguing an increase in inequality in the book, Leigh quotes Roger Wilkins on inequality increasing with figures from 1993 to 2009 ie covering the Keating ‘recession’ and the GFC. The NATSEM 2012 expenditure survey covering the period 1984 – 2010 gets a thumbs down largely because of ‘an extremely large change (32%)’ in the bottom quintile incomes between 1984-88. Excluding the early Hawke Government figures, the author argues that overall inequality increased but expenditure surveys are still a bit sus compared with income – ‘the poor often borrow where the rich often save’.
Leigh points to Brazil under President ‘LuLa’ as one country that has reduced inequality through targeting the needy. Entrenched ‘very deep’ disadvantage of children in poverty – less than 50% of median household income – was more than just a problem of income according to a recent Productivity Commission paper.

derrida derider
derrida derider
8 years ago

The reason you worry about half-median rather than half-average is because the average is sensitive to extremes where the median is not. If Bill Gates came to live in Australia that would imediately place more people under the half-mean line but not the half-median line; a half-mean line is therefore a measure of income INEQUALITY while a half-median line may not be. The concept of (relative) poverty is about how poor people’s living standards relate to those of the “typical” (proxied by median) person in the broader society, not about how poor people’s living standards relate to Bill Gates’. Inequality may be a problem (I think it is – I read the subjective wellbeing literature), but it’s a different one at both an ideologic and practical level from poverty.

Perplexed, a lot of people think that actual consumption is a far better indicator of material deprivation, lack of capability, etc – things which correspond to what most people think of as “poverty” – than self-estimated current income. Indeed, there is a very long tradition of using changes in consumption patterns rather than just levels as an indicator of changes in poverty prevalence (google “Engel curve”). What we do know is that in the HES, in common with worldwide experience, a person’s current income is a poor predictor of their current consumption. As you should expect the things that drive a wedge between income and consumption (ie savings, borrowing and wealth) all vary radically among different demographic groups.

Of course all this is not and never was a purely technical debate. In a better world people would not obscure things by mixing the normative and positive questions up. But that doesn’t license you to ignore, select or falsify the technicalities – quite the contrary. Yes, it does matter if Australian kids are going without food, and it does matter if we are disincentivising the deserving rich by casting their taxable pearls before the swine of the lazy and immoral. But before you can decide how much you care about either you need to decide whether either problem is empirically large or small.

Don Arthur
Don Arthur
8 years ago

DD – Do you think inequality of subjective wellbeing is a problem? Or do you think it’s aggregate wellbeing that matters? (if there is such a thing)