AN ECONOMIC PARADOX ON INEQUALITY -Fred Argy

Comparing Australia with the rest of the developed world, we have a distribution of FINAL disposable incomes that is about average. Yet, measured the same way (i.e. using the GINI coefficient), inequality of MARKET incomes in Australia (the distribution of gross incomes before taxes and transfers among working age Australians) is not only distinctly higher than the OECD average it is the highest in the developed world.

 

            Final incomes          Market incomes      
              (GINI)                     (GINI and share of bottom quintile
                                                      in brackets)

Australia    30.5               42.1 (1.6%)

OECD av.    30.7             39.6 (4.5%)
It is not surprising that the gap between market inequality and final income inequality is wider in Australia than in other countries as we have one of the most effective, well targeted tax/transfer redistribution systems in the world. But why is underlying market inequality (which is the starting point for any analysis of inequality of   opportunity) so high in Australia compared with other countries?

The gap between top and low hourly wage rates (wage dispersion) and the gap between average earnings and minimum wage are not exceptionally high in Australia. These gaps are tending to widen (note especially the surging share of the top 1% of incomes documented by Andrew Leigh) – but the same is happening in other countries. So I don’t think this can offer an explanation.

One very partial explanation may be Australia’s low full-time employment to population ratio (we rely heavily on part-timers). Another small explanation may be the relatively high proportion of low-skilled occupations in Australia.

A third answer may lie in the distribution of hours worked by income quintiles, which seems to be very unequal relative to other countries. Why? Is it that, relative to the rest of the world, some working age Australians are working very long hours and some are acutely under-employed or jobless? If so, why exactly is this happening to such an inordinate extent in Australia?

Has anyone done some serious research on this issue and able to offer a definitive explanation? What about Peter Whiteford or Andrew Leigh?

This entry was posted in Uncategorised. Bookmark the permalink.
Subscribe
Notify of
guest
25 Comments
Oldest
Newest Most Voted
Inline Feedbacks
View all comments
backroom girl
backroom girl
15 years ago

Fred

My guess is that it is mostly due to levels of non-employment, supplemented perhaps by our high prevalence of part-time work.

Compared with most of the rest of the developed world, Australia has relatively low employment rates for the prime age population (ie those aged 25-54), both male and female. On the basis of 2001 OECD data (a bit old now, I know), we ranked 18th out of 30 countries for female prime-age employment and only 24th for male. (We do better in the rankings for youth, where we are in the top half-dozen countries and 55-64 year olds, where we are around the middle of the rankings.)

It is also fairly well-known that Australia has one of the highest rates of family joblessness in the developed world, largely due to the combination of our high rate of single parenthood (over 20 per cent) and their low rate of employment.

But I think it would be a mistake to presume that all of this joblessness is involuntary. For single parent households in particular, the income support system is arguably generous enough to make working optional (something that can’t be said for many other countries). And despite claims in some quarters to the contrary, single parent households have been the biggest winners from the expansion of family assistance under the current government, so it seems likely that incentives for single parents not to work (or to only work a little bit) are even stronger now than in the past.

On the issue of working hours, we may have an relatively unequal distribution relative to many other countries, but I don’t know that this is a bad thing. It is at least in part due to the fact that part-time work is probably more viable here than elsewhere (I would argue that part-time jobs in australia are ‘better’ jobs relative to full-time jobs than in most other countries) and that certain groups have a very strong preference for part-time work (including the single parents I spoke of above). And let’s not forget that while a significant minority of part-time workers would like to work more hours, the substantial majority (78%) report being quite happy with the hours they work.

derrida derider
derrida derider
15 years ago

This paradox has often worried those in the know (I remember Peter Whiteford talking about it in the early 90s). No-one knows which way the causality runs though – does the highly progressive tax/welfare system cause wide pre-transfer inequality by distorting labour supply, education and savings decisions, or does the wide pre-transfer inequality create political pressure for a progressive tax/transfer system? Or is it a vicious circle with both operating?

On part time work, what backroom girl said. A widening distribution in working hours should prima facie represent a welfare gain as people are able to make choices along the intensive (ie how much to work) rather than the extensive (ie work full time or not at all) margin. And even amongst the minority of part timers who want more hours, according to HILDA only a minority within that would actually prefer full time hours – most just want a few hours more a week. In other words, the overall welfare losses from part time workers being unable to get more work seem pretty small.

Nicholas Gruen
Admin
Nicholas Gruen(@nicholas-gruen)
15 years ago

Backroom Girl – more detail on this claim if you would please?

(I would argue that part-time jobs in australia are ‘better’ jobs relative to full-time jobs than in most other countries)

Patrick
Patrick
15 years ago

Well, in lower paid jobs part-time jobs in Australia are full time jobs in France ;)

backroom girl
backroom girl
15 years ago

Thanks for putting me on the spot, Nicholas. I should have known I wouldn’t get away with talking off the top of my head.

Well, unfortunately I can’t put my hands on any indisputable proof that part-time jobs are better here than elsewhere, but these are some of the arguments that I think are suggestive.

At last count (August 2005) about 40 per cent of part-time jobs in Australia carried access to standard annual and sick leave entitlements (ie on a par with permanent full-time work). Interestingly, women working part-time were significantly more likely to have such leave entitlements than men working part-time (45% vs 25%). I believe that in many more regulated labour markets such as in Europe, part-time work is by contrast much less likely to provide access to standard employment conditions.

Where part-time work is ‘casual’ (ie without leave entitlements), Australian awards have traditionally provided for a casual loading to compensate for that lack of entitlements, which means that casual part-time workers effectively receive a higher hourly rate for the same work than a permanent employee. I believe that this is a peculiarly Australian innovation, although we have yet to see whether it will survive the current changes in workplace relations.

My last argument is that, despite Australian women have relatively low labour force participation rates by world standards and a very high propensity to work part-time, we have one of the two or three highest ratios of female/male earnings on an hourly basis (ie one of the smallest gender gaps in pay). Given that the great majority of part-time workers in Australia are women, this also suggests that working part-time in Australia is not especially disadvantageous, at least in terms of hourly pay.

To get back to Fred’s initial conundrum, what we have I think are a labour market and a social security system that offer a wider range of individual choices than those in many other countries, and I think this may go quite a long way to explain our wider disparities in market incomes.

backroom girl
backroom girl
15 years ago

I could also add that it may be that jobless households per se are more financially viable in Australia than in other countries, or maybe just more socially acceptable. Where jobless people live with employed extended family, which they may be more likely to do in other countries, you will tend to get higher equivalised market incomes at the bottom of the household income distribution than in a country like Australia where as I pointed out earlier we have a relatively high incidence of jobless households.

Nicholas Gruen
Admin
Nicholas Gruen(@nicholas-gruen)
15 years ago

Thx BRG. Not trying to put you on the spot. Not after data. Just why you thought what you did – which is interesting. Thx for the elaboration.

Peter Whiteford
Peter Whiteford
15 years ago

I think that there are a range of factors and it is difficult to definitively determine what is going on.

However, to reinforce Fred’s point, take the recent study by Ann Harding et al. comparing the impact of direct and indirect taxes and benefits in Australia and the UK
see http://www.natsem.canberra.edu.au/publications/papers/dps/dp61/dp61.pdf
and look at table 1. For “original income” the ratio of the income of the highest quintile to the income of the lowest quintile in the UK is about 18 to 1 and in Australia it is 73 to 1!!!!

Apart from the interesting fact that we successfully redistribute more than the UK so that the distribution of final income is actually slightly more equal in Australia, you can also see that the difference is driven by differences at the bottom of the distributions. Let’s use a very rough conversion factor of 2.5 Australian dollars to one british pound, and you can see that in terms of original or market income the richest quintile in the UK are actually better off than the richest quintile in Australia (62,000 pounds – roughly $150k – vs $109 thousand in Australia), but the poorest quintile in Australia have much lower incomes than the poorest quintile in the UK (3,400 pounds – roughly $8500 – vs $1,500 in Australia).

So why do the Australian poor have such pitiful private incomes. Part of the answer is that the composition of the lowest quintiles differ. In Australia there are a relatively large number of aged people in the lowest quintile, while in the UK it is families with children. Aged people are not very likely to have earnings (but then neither are poor lone parents).

I also think that there is a phenomenon of aged Australians – and those close to retirement age – being income poor and asset rich. And assets are not measured directly in most income surveys.

Apparently poor Australians, particularly if they are aged have relatively high rates of home ownership and they also have relatively speaking extremely high liquid wealth. For example, quite a while ago I did a study of the impact of government non-cash benefits and also imputed income from owner-occupied housing (estimated from mid 1980s LIS). Imputed income from owner-occupied housing for the lowest quintile of the population was equivalent in value to 5% of cash disposable income in the UK but 25% in Australia (although this was 25% of a smaller amount). Australia has relatively high rates of home ownership, and because it is high on average it also means that it is fairly equally distributed.

Among older people imputed liquid wealth (estimated by assuming a constant rate of return across countries and grossing up from measured income from property and investments) was about twice as high in Australia as in the UK (and nearly 8 times higher than in the Netherlands, for example). The poorest 20% of older Britons had liquid wealth equivalent to just under a year of income, while in Australia the liquid wealth of the poorest 20% of the aged was equivalent to about 2 years and two months worth of cash income. In the Netherlands with a generous public pension, the imputed liquid wealth of the poorest quintile of the aged was equivalent to 2 weeks worth of income!

I think that there is more up-to-date confirmation of this. The recent ABS income survey includes information on net worth. see http://www.ausstats.abs.gov.au/ausstats/subscriber.nsf/0/CA2568A90021A807CA256F41007C8505/$File/6523055001_2003_04.xls

This shows that the poorest 20% of all Australian households have average gross incomes of just under $15,000 a year, but net worth on average of more than $280 thousand. This includes home ownership, but implies that the average wealth of the poorest 20% is more than 18 times their gross annual income. It is very difficult to make comparisons but what I have seen suggests that in an international context this is extraordinary.

For example http://www.federalreserve.gov/pubs/bulletin/2006/financesurvey.pdf shows that for the USA the bottom 20% by income has average incomes of just under US$11,000 a year (Table 1) and average net worth of around US$ 73,000, or not quite 7 times their income. PPP adjusted (roughly) this suggests to me that average gross incomes in the bottom quintile are about the same in Australia as in the USA, but net worth for the lowest quintile is more than 3 times higher in Australia.

So, first, what we have in Australia is an issue of apparently low income, but relatively high wealth (probably reflecting our habit of liking superannuation lump sums, and favouring home ownership through the exclusion of the family home from the pension assets test (which also applies to disability and lone parent pensioners). I think that on this particular dimension, Australia is an extreme case in the OECD.

Looking more at the working age population, as people have pointed out Australia has a relatively high rate of family joblessness, but this is despite a relatively low rate of individual joblessness. In southern Europe and Japan and Korea, jobless people tend to live in multi-generation households, while in Australia they are more likely to live separately, which will lead to a widening in measured income distribution. Correspondingly, part of the reason why the Australian welfare system appears more progressive than those in southern Europe is that higher income households in Italy, say, are more likely to receiving welfare payments because beneficiaries are more liekly to be living in multi-generation households.

Having said this, this is part of the explanation for Australia’s difference from the conservative southern european and Asian welfare states, but is unlikely to explain why we are different from the UK for example, or the USA or the Nordic countries.

It is possible that in a certain sense the Nordic countries “cheat” on their apparent employment performance. The Nordic welfare states look good because they are high employment societies. This will make their income distribution across the working age population look relatively equal; but the actual number of people at work on a given day is a lot less than the number counted as employed. (But their unemployment rates are not as alarming as one sometimes reads on websites or blogs critical of the Nordic model.)

For example, Sweden and Norway have the highest rates of sickness absence in the OECD
http://www.olis.oecd.org/olis/2005doc.nsf/43bb6130e5e86e5fc12569fa005d004c/41598c3dc5781f43c1257085002d4e2a/$FILE/JT00189736.PDF
These people are counted as employed and will make the distribution of earned income look higher and more equal. Having said this, sickness absences are more prevalent in Australia than in Denmark. In Sweden and Finland this sort of effect will also occur for people on some forms of parental leave

I also worry about measurement and the quality of data at the bottom end of the income distribution in Australia. I think that Peter Saunders (CIS) is right to be concerned about this, and the ABS has also said there is a problem, but why Australian data at the bottom of the income distribution should be worse than UK data, for example, is obviously difficult to say.

However, there are issues in some countries that may tend to understate income inequality. For example, the surveys used by the OECD and in LIS for Austria simply exclude the self-employed. In the case of the Nordic countries, the income data are based on tax and benefit and population registers. That is, they don’t actually ask people what their incomes are, but they take their tax returns and put in the income stated on the return, and similarly they add in the administrative record of how much they receive in benefits.

Last but certainly not least there is a significant conceptual problem related to employer social security contributions. The extreme case is France, where employer social security contributions amount to nearly one-third of total wage costs, but they are also high in Sweden, Italy, Austria and most other European countries (except the UK and denmark). Because employer contributions go from the employer to government they do not pass through households and so are not included in household income surveys. If it is assumed that employer contributions are incident on wages then in France, for example, you would have to increase gross wages by around 50% (i.e. 1/3 over 2/3) to get a measure of the full wage package. While there are variations, in most countries employer SS contributions are roughly proportional to wages over fairly wide earnings ranges, so inclusion would not affect measures of disparities in earnings. However, the gap between those with no earnings and those in employment would widen significantly. Of cours, in australia you should add in the Superannuation Guarantee into the wage package – but this only 9%, while in the US you should take account of employer-provided health care and pensions – which are more significant.

Fred Argy
Fred Argy
15 years ago

Thanks everyone. We have had some very interesting contributions. But I do think backroomgirl and (to a lesser extent) derrida are gilding the lily a little. No one disputes that part-time work offers flexibility and good conditions for many workers. The issue relevant to market inequality is not the quality of part-time work (which may be quite good in many cases) but the inequalities of employment opportunity.

Although only a ‘minority’ of workers wants to work more hours, these still amount to a significant figure overall. If you add the number of under-employed to the number of discouraged workers (who are not actively looking for work but would like to), you are talking about a number equal again to the official unemployment rate. And that’s a conservative estimate of mine

backroom girl
backroom girl
15 years ago

Look forward to that, Fred.

I’m not really arguing that there aren’t problems with equality of opportunity in Australia, just that things are not as black as some people (not you necessarily) might have us believe. I’m just trying to understand the shades of grey. But I would stand by my conclusion that the Australian labour market and social support systems offer a wider range of personal choice for many people than is the case in many other countries, and to the extent that this plays into wider (market) inequality that may be the price we have to play. (Though I’m sure Peter Whiteford is right in his assessment that many of the differences between countries in the lowest income quintile can be explained by looking at the situation of retirees.)

While Australian institutions may be a long way from the Scandinavian model, I also think they are a long way from the US one. And while I’m confident that most Australians do believe in a fair go of the (more) equal opportunity kind I’m not at all convinced that they would want to go too far down the Scandinavian path. And if we go too far in the US direction, I think the electorate would revolt against that too.

I do worry a lot about the longer term prospects of children growing up in households where parents either don’t work or have only an insecure attachment to the workforce. And I agree that something needs to be done to improve those prospects. I’m just not sure that it is possible to intervene for those children in an effective way without somehow enabling their parents to turn their lives around first.

In many cases this probably does need the combination of ‘help and hassle’ that you have spoken of earlier. But what we need are programs that really aim to build people’s capacity for true self-reliance (in all its aspects, not just financial), rather than programs that simply reinforce dependency by solving people’s problems for them. And unfortunately I don’t see all that much commitment to that goal from either side of politics here. (Noel Pearson is one exception, but look where that has got him.)

Tony Harris
Tony Harris(@tony-harris)
15 years ago

“It is not surprising that the gap between market inequality and final income inequality is wider in Australia than in other countries as we have one of the most effective, well targeted tax/transfer redistribution systems in the world. But why is underlying market inequality (which is the starting point for any analysis of inequality of opportunity) so high in Australia compared with other countries?”

As Backroom Girl pointed out, it is not surprising it is just what you might expect. The more attractive you make welfare, the more people will live with it for ever, thereby perpetuating relative disadvantage.

As for active labour market programs, what about starting with the elimination of the minimum wage and getting the right incentives in place?

Nicholas Gruen
Admin
Nicholas Gruen(@nicholas-gruen)
15 years ago

Backroom Girl,

I wonder if you’re right about the idea of help and hassle not being popular. I think it’s kind of there. Both Labour and the Coalition have made a big thing of programs of theirs – the Job Compact and Work for the Dole – but then not funded them properly. As you note, we’ve got Noel Pearson spruiking in the wings which I think is pretty influential.

My feeling is that we’re nearly there, and that we may well have been if the current government had been more interested in policy rather than symbols. But of course by ‘nearly there’ I mean in an ideological sense. It’s clearly not got through to funding.

On another point you raise, you talk about trying to fix up the lives of welfare families. I wonder if there’s anywhere that has done that successfully. Certainly the stylised fact that screams out of Heckman’s work is of interventions aimed at young kids working and pretty much everything else failing – at least in cost benefit terms.

Further, I expect the best way of intervening with dysfunctional families (at least where one doesn’t want to take the kids off them) is to work with them as best one can on, and on behalf of their kids.

backroom girl
backroom girl
15 years ago

I suppose my problem with the approach of both sides of politics to the issue of ‘help and hassle’ is that it tends to be authoritarian (ie do as I say, take whichever job or training program some bureaucrat decides you should, if it doesn’t work out come back and we will fix you up all over again), with fairly dire consequences for anyone who doesn’t want to just do as they are told. We also still have far too much a factory-line, one-size-fits-all approach (eg the current government’s solution is work first for everyone and Labor seems to favour training first for just about everyone, whereas the truth of the matter is that for some people work is the appropriate first step and for others it’s training).

I would rather see an emphasis on giving people the skills to make their own better-informed and considered decisions about what is best for themselves in the longer term (with appropriate help and guidance where necessary). This is not inconsistent with having a strong emphasis on people improving their capacity to support themselves and their families, it’s just that I think you will get better results in the long run if you allow people a little more choice about how they go about pursuing that goal.

But unfortunately I think you are right that neither side of politics is really interested in adequately funding that kind of active labour market policy.

Fred Argy
Fred Argy
15 years ago

Sorry Rafe I can’t buy that argument of yours and Backroom Girl that our welfare and related benefits are so ‘attractive’ that it encourages joblessness and can explain why we have greater market inequality among working-age Australians than other countries

In assessing how generous governments are to disadvantaged workers (with no individual bargaining power), you need to look at three things

backroom girl
backroom girl
15 years ago

But Fred the OECD also found that minimum-wage workers and their families are further above a 60% of median poverty line in Australia than in any other country bar none (including those Scandinavian countries) and that people on income support payments need to earn less in Australia to lift their incomes up to such a poverty line. Supporting people in low-paid work is actually something we do very well in Australia.

And it’s not really fair to compare Australian replacement rates with those that apply under social insurance schemes elsewhere, since income maintenance is a specific objective of such schemes and not of the flat-rate system we have here. It also depends on what your denominator is – if it’s average earnings, of course we will come out worse because our benefits are not tied to previous earnings, but if it is a low paid job we do much better.

In the end, I’m just not sure how instructive it is to compare Australian outcomes with those in other countries that over the past century have made many different policy choices than we have here. Our particular policy mix has its downsides, but it also has its advantages, and the same is true I would venture to say of any other country you might want to have a look at. What we need to do is to find the right Australian solution, I reckon – it’s something we’ve been good at many times before.

Fred Argy
Fred Argy
15 years ago

Backroom girl, your points are well made. I should of course have mentioned our high minimum wages. That was just an oversight but that fact is why I acknowledged the need to gradually reduce it relative to the median wage. In any case, in the context of the present argument, a high minmum wage works both ways – it costs employers more but encourages inactive jobless to enter the labour market (as the Fair Pay Commission pointed out).

I agree that international comparisons of welfare benfits are dicey but I was only trying to question the suggestion that you and Rafe made that our welfare system could explain why market inequalities are so high. The evidence on benefits may be hard to read but it surely gives no support to your hypothesis that our benefits are so generous that they encourage more voluntary joblessness than in other countries (except as I said for a few on Parenting Payments or Disability Pensions).

Again, thank you for your contributions.

backroom girl
backroom girl
15 years ago

Peter Whiteford may have a view on how our single parent benefits stack up against other countries – that is one area that he is an acknowledged international expert on.

The key problem when comparing levels of benefits across countries is how you can ensure that you are truly comparing apples with apples. When it comes to replacement rates, for example, while these make eminent sense when considering the situation of someone who has recently had a full-time job and become unemployed, I’m not sure how relevant they are to a single parent who may never have had a job or only worked irregularly and/or part-time.

In the end, Australia does have a much larger proportion of families that are jobless single parent families than just about any other country in the world. Most of those jobless single parents, by the way, are classified as ‘not in the labour force’ rather than ‘unemployed’, so they probably wouldn’t even show up in your hidden unemployed.

It stands to reason that the phenomenon is partly a result of labour demand and partly one of labour supply and I for one don’t know how you go about working out which is the predominant factor. But one thing I do know for certain is that when you consider the total value of the income support package for single parents it is worth a lot more now in real terms than it was 10 years ago, thanks to the efforts of the current government to throw money at families. (And that is still true, by the way, if you factor in the most recent ‘cut’, which you should remember only applies to a minority of single parents, those with a youngest child aged 8-15 who first claim benefit from July 2006).

Nice talking to you – have a good weekend.

Peter Whiteford
Peter Whiteford
15 years ago

Just to clarify my long previous post and respond to some subsequent points.

Do I believe that the distribution of private income is 4 times wider in Australia than in the UK, and that Australia has the most unequal distribution of private income in the OECD? No, I find this implausible, so I think that part – probably a large part – of the answer is related to the problems in data, measurement and concepts that I referred to. However, there are some reasons for thinking that there are also “real” differences. Determining how important these are is very difficult.

The case of generosity of benefits to lone parents is an interesting example of a real difference. I have argued elsewhere – and can send the reference – that replacement rates actually measure different things in different countries, and are not necessarily reliable measures of benefit adequacy. This is largely due to the role of employer social security contributions.

Benefit entitlements for lone parents in Australia are just above the OECD average as a % of the average wage, but Australia has about the second highest average wage in the OECD on a PPP adjusted basis. Adjusting lone parent benefits by PPPs Australia is about 40% above the OECD average, and exceeded only by Denmark, Japan, Luxembourg, Norway (just), Switzerland and the United Kingdom. Relative to the poverty line of 50% of median income, Australia has the second highest entitlements after Japan.

Now Japan, Switzerland and Luxembourg have theoretically generous benefit entitlements, but they don’t actually pay anyone (or almost no one) – lots of stigma, very harsh means tests and benefits are recoverable from your family and can in some cases become debts to be recovered once you leave benefits – so very few people actually claim social assistance in these countries.

Unemployment insurance payments and child care leave are more generous than social assistance in the Nordic countries and many other European countries, but entitlement is certainly not indefinite.

It is the eligibility conditions that to me are the most likely explanation for Australia’s high level of benefit receipt among lone parents and the relatively low level of employment. Under the “old system”in Australia a lone parent could get a benefit without a work test (i.e. no activation) until the youngest child turned 16. The only other countries where the systems are anything like this are the UK, NZ and Ireland. The next most generous was Norway where you could get a non-activity tested benefit up until the youngest child was 10 years. In the other Nordic countries social assistance benefits are ususally work tested from the age of 2-3 years (youngest child), but the municipality has to offer child care.

So if you factor in expected duration of receipt plus the relatively high benefit levels, in practice the 4 English-speaking countries – but certainly not the USA and Canada – have effectively the most generous benefit levels for lone parents in the OECD. And it is therefore not surprising to me that these 4 countries also have the highest proportions of their lone parent populations receiving benefits and the lowest percentages in paid employment.

Fred Argy
Fred Argy
15 years ago

Thanks for that Peter. It does clarify one issue very well – that is, relative lone parent benefit entitlements do help explain the high level of market inequality in Australia (and the UK). I was aware of this all along (indeed my estimate that ‘hidden unemployment’ was of the same order as official unemployment was in good part due to the low participation of single parents). However I suspected it had a secondary explanatory power. I think I was wrong. It could be quite significant (as backroom girl has been saying) – in which case the new eligibility requirements should significantly reduce market inequality by forcing many lone parents to work longer hours (if they can find appropriate jobs). We shall see.

That said, I suspect the mismatch in the market for unskilled labour (by region) is particularly marked in Australia and may also be having a significant influence on market inequality (by impacting on the distribution of hours worked by socio-economic groups). This problem does cry out for cost-effective labour market programs as well as wage flexibility. But that is an issue for another day.

Andrew Leigh
15 years ago

Apologies for coming late to a fascinating discussion. I can only agree with those who’ve pointed to part-time work as a likely cause. It’d be interesting to take out the differences in hours worked by looking at hourly wage inequality.

Here, our relatively high minimum wage must compress the distribution (either through firings truncating the bottom of the distribution, or wage rises pushing up the bottom of the distribution).

Peter Whiteford
Peter Whiteford
15 years ago

A few more caveats/points. While I think that high levels of joblessness among lone parents is likely to be a contributing factor, I don’t think it is a significant explanation of differences between Australia and the UK, which as pointed out above appear to be enormous. From OECD figures about 13.5% of all working age households in Australia contain no-one in paid work, but the figure for the UK is 13.4% (and more of these are lone parents). For a selection of countries, the corresponding figures are as follows: USA- 4.8%; Sweden – 6.0%; Denmark – 8.7%; Spain – 8.1%; Italy 9.9% and Poland – 16.6%.

So relative to the US and the Nordics, joblessness is likely to be important. In the case of Spain and Italy, individual joblessness is much higher than in Australia or the UK, so I think that the household income distribution is narrowed in these countries, because jobless individuals are more likely to share accommodation with someone who has a job.

But as I said differences in joblessness cannot explain the apparent differences between Australia and the UK, and of course both countries also have a very high share of part-time work – figures I have seen say 27% of all employees in Australia worked less than 30 hours pw in 2002 compared to 23% in the UK. I can’t see that this is a large enough difference to explain much. Also all studies I have seen suggest that wage inequality is greater in the UK and the minimum wage lower than in Australia.

So even if there is not a paradox to explain, there is still a puzzle.

Fred Argy
Fred Argy
15 years ago

May I add further to the statistical uncertainty by pointing out that according to Forster and d’Ercole 2005 (p.26), the UK, Australia and USA stand out with high market income inequality (in fact UK has higher inequality than Australia). These three English-speaking countries have an average GINI of 42 or more whereas the OECD average GINI is 39.6 and the Scandinavians are under 38! The three Anglos all have relatively high shares of part-time work and high jobless rates among working age households. They also all have relatively free labour markets. Is there a causal relationship here?

Despite your valiant efforts, Peter, there remains, as you say, a big puzzle as to what is happening here. It would be great if you and Andrew Leigh could do a little more number-crunching to try to sort out cause and effect here.

Fred Argy
Fred Argy
15 years ago

A correction to my earlier note. The USA does NOT have high rates of jobless households or high part-time employment rates. So the reason for its high market inequality must be the relatively large earnings differentials in that country – more so than in Australia or UK. So in the USA wage flexibility may be a key factor behind its high levels of market income inequality – whereas in Australia the explanation seems to lie more in the low job participation rates of low-skilled working age persons (large differences in working hours by socio-economic group).

Fred Argy
Fred Argy
15 years ago

To stir up a little more interest, I want add one further footnote on the causes of the low job participation rates of low-skilled working age persons (large differences in working hours by socio-economic group) in Australia and the policy implications.

Our low participation has been partly due to our greater wage rigidity and (until now) more open-ended welfare access than in the USA – but without the active labour market programs of the Scandinavians. That is, Australia may have been ‘falling between two stools’

Peter Whiteford
Peter Whiteford
15 years ago

Fred

I’m not sure that becoming more like the US is an imminent threat, so to speak. Australia and the US are virtually at opposite ends of the OECD spectrum in terms of benefit generosity to the poor and also in terms of the minimum wage.

OECD estimates for 2003 are that the take home income of a couple with children with one earner on the minimum wage was 73% of median equivalent income in Australia and 35% in the US (about 2 to 1). The income of an unemployed couple with children (not including rent assistance) was 52% of median income in Australia and 20% in the USA (about 2.5 to 1). For a single person net take home pay on the minimum wage was just more than twice as high relative to the median in Australia than in the US.

These differences mean that Governments would have to cut benefits or wages or both by very large amounts in nominal terms, or we would need relatively high inflation for a considerable periods of time to reduce the Australian safety net to the US level. Thus, to get to the US situation in terms of wage inequality will either take a relatively long time with change always going in the same direction, or would require some sort of seismic shift, which I think is less likely in the Australian electoral environment. To some extent, NZ did the seismic shift approach in the 1990s, but they then had first past the post voting and no upper house or states.

One of the interesting things is that despite these enormous differences in the basic welfare state settings the difference in unemployment rates is relatively modest at the moment – but that might be more a case of good luck than anything else.

I think – “from a “welfarist” perspective – that the evidence is that the most urgent social protection challenge facing Australia – apart from the circumstances of the indigenous population – is the relatively high share of jobless households, particularly those with children.

In one specific sense (and one only) I am less concerned about single people because if the unemployment rate in Australia is about the same as in the US, but if we have a higher share of single unemployed people living alone, I think this mainly means that we pay them enough so that they can afford to live separately. Having said this, further reductions in unemployment, particularly long term unemployment, are obviously a good thing and it is also fairly clear that high housing costs in capital cities can be a major problem for unemployed and low income workers.

For people with children there is an obvious danger of transmission of welfare receipt across generations, reflecting a variety of causal factors – this is the main reason why I would concentrate on lone parents and unemployed couples with children.

In a sense there is a greater similarity now in the US response to this challenge and the Nordic response than is sometimes acknowledged. The Nordic approach is built on high levels of employment and in particular on high levels of participation by mothers, supported by generous child care and parental leave. When the US made the welfare reforms in 1996 virtually all of the money taken out of AFDC was out into support services – in fact aggreagate spending may have increased. Child care assistance was dramatically expanded, training was expanded and health care coverage for the poor was also dramatically expanded. (There is a problem now because the Congress has to recertify legislation and administrative procedures, and I think that indirect assistance is being reduced).

In my view what Australia needs to do is shift expectations about whether mothers should be in paid work, both by changing work tests but also by increasing child care availability, keeping it affordable and also looking at the skills and education levels of some (not necessarily all) of those now on welfare benefits. This approach is consistent with both the US and the Nordic approaches. However, an expectation of very long hours at work probably does not help.

I also think that improving outcomes in terms of joblessness is more a matter of designing programmes carefully rather than some form of complete paradigm shift. That is, it is not a matter of choosing either the Nordic model or the US model (recognising that you are not saying this Fred) and heading down one of these two alternative paths, but of identifying which specific parts and programmes appear to produce better social outcomes and experimenting with these specific programmes adapted to Australian circumstances.