As part of his well-deserved early career award from the Academy of the Social Sciences, Andrew Leigh was asked to write a paper for their journal Dialogue. As he explains on his blog, he chose to write on something a ‘bit provocative’, the possible negative effects of ethnic and linguistic diversity. One of these possible negative (sic) effects is reduced support for the welfare state.
For this hypothesis, he draws on the work of Alberto Alesina and Ed Glaeser, who argue that one of the major reasons for the much smaller welfare state in the US compared to Europe is that the US is more racially diverse. Or to put it more bluntly, the wealthy white majority isn’t too keen on giving money to the poor black minority. Extrapolating from this, Andrew notes that Australia’s welfare state is small compared to Europe’s, and that our linguistic diversity is higher than either the US or Europe, and therefore ‘our high level of linguistic diversity helps explain Australia’s relatively small social welfare sector’.
I doubt it. Indeed, you only need to keep reading Andrew’s paper to find at least one reason for doubt. Using answers to a question in the Australian Election Survey about whether people agree or disagree with the proposition that ‘income and wealth should be redistributed’ he finds that only in Queensland is there are a statistical relationship between disagreeing with the proposition and levels of local ethnic diversity. This he puts down to the relative success of ‘racially-driven politics’ in that state, with One Nation its most public manifestation. But what about all the other states? They, after all, contain the vast majority of seats in the Australian Parliament.
My CIS colleague Peter Saunders has argued that this analysis of the comparative welfare states misses important cultural differences between the ‘Anglo’ countries and Europe. The Anglosphere countries have much older and more powerful traditions of individualism than Europe. Alan Macfarlane wrote a well-known book on this, The Origins of English Individualism, tracing it back many hundreds of years. In particular, the Anglo countries have a much greater belief in self-reliance.
Andrew L cites some survey research consistent with this – Americans and Australians are much less likely than Europeans to think the poor are trapped in poverty (ie it is not just impersonal social forces, you can personally do something to improve your lot) and much more likely than the Europeans to think that the poor are lazy. I don’t know a lot about European opinion, but there are numerous Australian surveys pointing to a moralised view of welfare among those capable of working.
In the 2005 Australian Survey of Social Attitudes, for example, only 5% of people disagreed with the proposition that ‘too many people these days rely on government handouts’. Only just over a fifth disagreed that ‘most of the unemployed just don’t want to work’. In a 2004 Saulwick Poll a mere 13% disagreed with the unemployed having to work for their welfare benefits.
Yet this isn’t exactly opposition to income redistribution as such. That Saulwick Poll found that most people were against cutting the unemployment benefit, and in the 2005 AuSSA two-thirds agreed that ‘cutting welfare benefits would damage too many people’s lives’. Polling has always shown strong support for the aged pension, and nearly half agree that ‘all families deserve payments from the government to help with the costs of raising children’. The Australian public wants to support people in need, but does not think it is fair or right that some people don’t pull their weight.
This scepticism about handouts to the ‘unworthy’ long predates any significant ethnic diversity. As John Roskam describes it in a chapter in Liberalism and the Australian Federation, pensions in the early 20th century were strictly means tested and restricted to the ‘deserving’ poor. In NSW, those who had been to prison or deserted their spouse were not eligible. When Labor expanded the welfare state after World War II, Liberal MPs expressed concerns that it would undermine self-reliance.
If we were looking at simple correlations, we would argue the opposite of Andrew’s thesis: in the post-war period Australia has become much more diverse and has spent much more on welfare. But I think this is largely coincidence – migration has put some added burdens on the welfare system, but there has been no discernible change in the public’s basic view of welfare (Richard Grant’s PhD thesis has the history).
The main differences between the Australian and European welfare states are that our system is much more heavily based on need (though there are partial excpeptions, such as the family benefits I have been complaining about) and means testing. In many European countries, when you go on benefits you receive a fairly large proportion of your previous income, but not here – that would encourage malingering and bludging, and discourage self-reliance. And that would be unfair to the people paying your bills.
To the extent that there is any relationship between ethnic diversity and welfare attitudes, it could be the reverse of Andrew’s argument. It’s not that attitudes to race drive attitudes to welfare, but attitudes to welfare drive attitudes to race. As Pauline Hanson puts it in her folksy way:
I am fed up to the back teeth with the inequalities that are being promoted by the government and paid for by the taxpayer under the assumption that Aboriginals are the most disadvantaged people in Australia. I do not believe that the colour of one’s skin determines whether you are disadvantaged. … We do not want a society in Australia in which one group enjoy one set of privileges and another group enjoy another set of privileges. Hasluck’s vision was of a single society in which racial emphases were rejected and social issues addressed. I totally agree with him, and so would the majority of Australians.
… I have done research on benefits available only to Aboriginals and challenge anyone to tell me how Aboriginals are disadvantaged when they can obtain three and five per cent housing loans denied to non-Aboriginals.
The majority of Aboriginals do not want handouts because they realise that welfare is killing them. This quote says it all: “If you give a man a fish you feed him for a day. If you teach him how to fish you feed him for a lifetime.”
She’s not upset about Aborigines because they are a different race. She is upset because they are getting benefits they don’t deserve (in her view) and which aren’t doing them any good, because they are not creating self-reliance.
My research on racial attitudes suggests that a smallish minority of Australians are ‘traditional’ racists, in that they believe some groups are inferior. But a much larger group has concern about ethnic groups ‘fitting in’, which probably means in part conforming to prevailing attitudes about work. This is probably partly why the Chinese, with their reputation for hard work, are fitting in pretty smoothly, while the Muslim Lebanese, with high rates of welfare dependence, are struggling to gain acceptance.
I don’t disagree with Andrew’s substantive policy conclusions. In general immigration is a good thing, but we should not (like some left-liberal academics do) try to conceal the problems that do exist. Where we differ is that I don’t think migration has had any significant historic influence on Australian welfare policy, and doubt that this will change in the future.
Andrew Norton 
This is the first honest representation of Hanson I have seen in an Australian blog. I predict a huge backlash from the left for this post.
LikeLike
“To the extent that there is any relationship between ethnic diversity and welfare attitudes, it could be the reverse of Andrew
LikeLike
Alternatively, if levels of local redistribution were linked to levels of racial diversity (ie non-Anglos were disproportionately on welfare), the hypothesis could be that being surrounded by welfare dependents makes people oppose redistribution (as well as racial diversity). This means that survey respondees are not necessarily inherently racist, but the direction of causation is still working in the direction AL proposed.
LikeLike
Rajat – I think ‘ethnic diversity’ is operating at too high a level of generality. None of the research on prejudice shows that people think in simple us/them terms; they have views about particular groups. I haven’t seen the detail of Andrew’s Queensland stats, but I’d say the ‘diversity’ here is Aborigines, where the perception that they are given unfair benefits plus the not-unreasonable conclusion that ‘sit-down money’ has had catastrophic effects on Indigenous communities would surely maximise hostility to the welfare state. People in Queensland witness drunken, destroyed Aborigines wandering the street, sights people in southern states are largely shielded from.
LikeLike
Andrew, thanks for a most thoughtful post. Past discussions with you have helped me think through this argument a bit better, but clearly I still have a way to go before you’re convinced.
As to the origins of Australia’s ‘mid-Atlantic welfare state’ (not as generous as Europe’s, not as stingy as America’s), I think Peter Saunders’ explanation doesn’t get to the core of it. Perhaps it’s a disciplinary difference. As a sociologist, he seems to regard attitudes as an answer. As an economist, my question is ‘where do the attitudes come from?’. Alesina and Glaeser’s answer – ethnic diversity (from the time of the Gold Rush, we’ve been more diverse than Europe) and the absence of proportional representation – seems pretty sensible to me.
As to the post-war trends, you’re right that diversity and the welfare state grew together. But I’d argue that the welfare state grew for reasons unrelated to immigration. Thanks to Milton Friedman (he acknowledged the irony), deduction of tax payments at source made possible a large revenue stream. The question for the 1950s is why we didn’t develop as generous a welfare state as Europe. My answer would be ‘more migrants, no prop rep’.
LikeLike
Andrew – Even on PR, I am not convinced. The US analogy breaks down because they don’t have strong party discipline and never had a strong labour movement to organise a political party, when we had both. PR would only make a difference is there was a significant constituency for a European-style welfare state, when there is no evidence that this has ever been the case. For much of the 20th century not even Labor was particularly in favour of it, because they wanted to use the IR system instead, which fitted with Australian culture: it was important to the male breadwinner that he have a sense that he was providing for his family, and not the state. Probably not coincidentally, the welfare state has grown rapidly as more and more households came to have no connection with the labour market, rendering this an ineffective strategy for assisting the poorest Australians.
LikeLike
In case people are interested in the consequences of welfare in addition to wondering about the motivation for providing it (or not providing it) it helps to appreciate the way that the New Deal and more recently the Great Society initiatives of the 1970s helped to generate a permanent underclass, especially in the black population of the US.
In the Australian context it would help to understand how the IR system generated perverse outcomes (especially regarding productivity), like the mass unemployment of the Great Depression.
One of the reasons for opposing the welfare state is the belief that it is not the best longterm strategy to help people in need while maintaining freedom and prosperity. As Andrew N hinted, that belief is more widespread in the US than elsewhere in the west. That probably provides a clue to a more productive research program than the Andrew L process of trawling for correlations to test hypotheses that don’t really get at the core issues even if they gain some plausibility from the analysis.
In other words, what difference does it make in the debate about welfare policy if the diversity hypothesis is true or false?
LikeLike
The abolition of slavery in the US was a labour movement that became a pivotal issue for both parties and helped to create today’s democratic party (with its southern carpetbaggers who are to the “right” of northern republicans on many issues), a civil war and the enduring intellectual dominance of the nothern states who opposed slavery (partly because slavery was of more use in the rural south and its presence would contribute to the south’s economic supremacy).
Australia’s welfare system in its current state is a major employer, probably the main employer in many suburban shopping areas where there are a number of job network offices around a large centrelink. This is all adding to the cost of labour downunda and makes the Aus system unique like the southern ocean.
LikeLike
Andrew, the key to understanding proportional representation is to think about the incentives it gives to politicians. Under PR, it’s much easier to support a class-based interest than it is under a constituency-based system. For example, I suspect that NZ never would have implemented the reforms of the 1980s if it had the PR system it has today. And I predict that NZ will remain more left-wing than Australia for the next 50 years, because of the difference in voting systems.
LikeLike
Andrew – I agree that PR creates opportunities for political entrepreneurs. But it could also be argued that the system used in Australia, the UK and NZ creates strong governments that can push through radical reforms from the left (eg the Attlee nationalisations and NHS establishment after WW2) or the right (your NZ example, ironically under a Labour government). Once established, welfare systems are highly resistant to winding back. The mid-1990s US reforms are the only case I can think of where policy has triggered a significant reduction in the number of welfare recipients. Overall, I think PR is a force for conservatism (in the maintaining the status quo sense, not the ideological sense) because it makes it harder to introduce radical changes.
LikeLike
Slightly off topic, but I was wondering if the comparisons of the level of welfare across countries are done on a net benefit basis, or by looking at payments alone?
Over the years, Australia has shifted a long way toward cash payments and away from tax based concessions. This can change the apparent level of welfare paid and degrees of dependency, even though theoretically people may have exactly the same net incomes as they would by not paying as much tax (and getting less cash payment).
Cross country comparisons that don’t factor this in seem to me to be misleading.
LikeLike
Spog – The large tax deductions family men used to be able to claim does mean that some comparisons with earlier times (such as mine between Howard and Menzies) understate the effective redistribution going on.
But even the terminology we use these days, such as ‘family tax benefit’, is a nod to the ideology of self-reliance, trying to water down the ‘welfare’ aspect and play up the own earnings angle.
LikeLike
Andrew – family tax benefit is a smelly mix of welfare and tax concepts that should have been kept apart.
I think tax deductions were the best method of recognising kids in the tax system that we had, and the process since (deductions turning to rebates, then cashing them out) really reflects the end of Western Civilisation As We Know It.
Bring back tax deductions.
But to my original point, do cross country (or even cross-time within one country) comparisons exist that are done on a net benefit basis? I have seen some through-time material that looked at families in this way (showing significant net gains in real terms under Howard), but not for other household types.
LikeLike
Spog – These things are taken into account in measures of inequality, but I am not sure that tax expenditures are counted as part of welfare spending. I doubt it but I am not sure. There is one huge tax expenditure that would have to be counted in Australia, which is the low taxes applying to superannuation. Effectively, this is current expenditure that is the equivalent of future payments of the aged pension.
LikeLike
Spog, Alesina and Glaeser devote a pretty large share of their book to thinking about issues such as this one (it’s even harder to compare social security payments across countries). I was pretty convinced that they’d thought hard about it, though I’d feel even better if I knew that someone like Peter Whiteford, who spends all his time on such problems, also liked their methodology.
LikeLike
I tend to agree with Andrew N on this one.
It seems to me that associating ethnic diversity with lower welfare spending is an unnecessarily long bow when underlying cultural differences (which show up clearly in surveys on ‘individual’ and ‘collective’ values) are so clear.
Furthermore – as argued, if it were true, surely welfare would have declined as we got more ‘ethnically diverse’. If anything, the last 35 years (the era of ‘multiculturalism’ as a buzzword) have seen a relentless expansion in the Australian welfare state which wasn’t seen at any stage of the previous 70 years.
Ireland too might be cited as an example – in most relevant areas (tax, welfare, income distribution etc) it lines up with the Anglosphere countries, but – though I don’t have figures to hand – I should think that it is far less ethnically diverse than Australia.
There are negatives associated with immigration, but I don’t think downward pressure on welfare budgets is one of them. If anything, since NESB voters have tended to back Labor more than ever in recent years, the bias could be in the other direction.
LikeLike
I just posted this on Andrew Leigh’s blog, so I thought that I’d put it here too, as it addresses a few of the questions raised above:
“I read your paper and found it very interesting, but I’m not convinced of the argument that the superficial similarity between social spending levels in the USA and Australia means that we really have similar attitudes to redistribution.
Before expanding on this, I agree that there is a lot of evidence of a relationship between immigration, diversity and access to social welfare.
In the first half of the 20th century in Australia this actually worked in the opposite way – the White Australia policy was intended to make sure that we had the
LikeLike
“But overall, it appears to me that whatever attitudes the majority of Australians have to different ethnic groups, the way we have structured the welfare system is actually particularly generous to the people we supposedly disapprove of.” (Peter Whitford)
Precisely. And the way public welfare works, there will always be an underclass, perpetuated by the system and not emancipated by it. People who really care need to explore other ways to help so that able-bodied people can proudly support themselves and their dependents.
LikeLike