Australia’s unemployment rate may be back to where it was in the late 1970s but the structure of our labour market and our society is very different. For example, in the late 1970s almost 70 per cent of men aged 25 to 34 were married and working full-time. Today it’s less than 50 per cent.
Marriage and work are connected. A consistent theme in the research on marriage is that couples are more likely to marry when they feel economically secure. And the problem for marriage-aged men with low levels of education and skill is that secure full-time work is getting harder to find. According to Bob Gregory and Paul Frijters, 76 per cent of unskilled men aged 25-34 were employed full-time in 1981. But by 2001 this had fallen to 60 per cent
For less skilled, less educated men, the combination of marriage and full-time work is no longer the norm. By the early 1990s, a new pattern had emerged. In 1998 Bettina Arndt wrote about the way the decline in employment opportunities had affected men’s relationships with women and their role in families. Over the next decade, there was some improvement in the proportion of men in full-time employment. But in the recent downturn almost all the gains of a decade were lost. Much of what Arndt observed in 1998 continues today.
The unemployment rate conceals as much as it reveals. The chart below shows the changes in the labour market status of men aged 25 to 34. While the great majority are still working full-time, there appears to be a long term trend away from full-time employment. Watching the trends unfold is like watching tectonic plates shifting. Each recession is a seismic event where the proportion of men in full-time work suddenly falls. And in each recovery it fails to return to its pre-recession level.
Another way of looking at the trend is to graph full-time employment by age group. The chart below shows how full-time employment for older groups recovered during the boom while continuing to decline for younger men.
The shift is unlikely to be the result of voluntary downshifting, an increase in the number of househusbands or a collapse in the work ethic. Something has changed in the demand for labour. As the Productivity Commission’s Ralph Lattimore writes:
With technological change favouring higher skilled jobs and the growth of the service sector, the long-run structure of the economy has shifted away from jobs in which unskilled, less language-proficient, males specialise. For example, blue collar jobs accounted for 63 per cent of male jobs in 1971 and 46 per cent in 2000.
For men, full-time jobs are becoming more demanding — employers are looking for higher levels of skill. But there hasn’t been the same trend for part-time jobs. According to economist Alexis Esposto, the growth in part-time jobs taken by men, shows a pattern of deskilling.
For women this means a decline in the number of men of marriageable age who are able to offer the financial security of stable full-time work. More men are cycling between full and part-time work and some are leaving the labour market all together. In a 2002 paper, Bob Birrell, Virginia Rapson And Clare Hourigan suggest that changes in the labour market may increase the number of single parent families. They speculate that:
From the point of view of the mother, the option of staying single and surviving on the low but stable lone parent and family assistance welfare support package may seem preferable to partnering a man who cannot provide adequate financial support for the family.
The shift in the demand for labour is not about some kind of competition between men and women. When sub-groups of men lose access to secure full-time employment, there are women whose opportunities are also narrowed. For example, it increasingly takes two full-time incomes to save for a deposit and pay off a mortgage.
Policy makers are unlikely to address this problem by intervening in the labour market. What they are most likely to do is crack down on income support recipients (male and female) in an effort to force them into whatever jobs are available. And to help this along, there will be a great deal of discussion about job snobbery and the decline of the work ethic.
Over the past 30 years there has been a profound shift in the labour market — a shift that has implications, not just for men, but for the broader community. If you focus on the unemployment rate alone, you’ll miss the story.
Update
One suggestion raised in the comments thread is that "what might be happening is really an effective sharing of workloads" as women take on a greater role in earning household income.
If this is what’s going on, then why am I portraying the shift in men’s labour market status as a bad thing?
A more effective sharing of workloads would include not just paid work, but unpaid caring and household work. The amount of unpaid work in a household depends heavily on the age of the youngest child. So you’d expect to see coupled men with young children working part-time more often than those with older children and coupled men with children working part time more often than those with no children.
The charts below show that this isn’t the case.
When I read articles on the decline of marriage, my first thought is to ask about cohabitation. After all, people can have stable relationships without getting formally married.
But for reasons I’m happy to go into if anyone asks, I think marriage statistics are a useful proxy for statistics on stable relationships.
Feel free to go into it – for my benefit.
Meanwhile one thing you’ve not raised is the evidence that a substantial aspect of the skill deficit of men is not technical skill but social skills, which, alongside technical or ‘cognitive’ skills are also in increasing demand.
…the skill deficit of men is not technical skill but social skills…
Unfortunately we live in a world full of technical problems and their solutions. And while social skills might be great for expanding industries such as social services, they have marginal value in most industries. When you devalue technical know-how you get outcomes like the disaster in the Gulf of Mexico.
Nicholas – As I understand it, cohabitation is a less homogeneous category than marriage.
Most marriages begin with cohabitation. So a proportion of cohabiting couples represent an earlier stage of a relationship that becomes marriage (and ends with the death of a partner, separation or divorce).
There has been a trend towards marrying at later ages, so some of the apparent decline in marriage among 25-34 year old men reflects this. So can we treat statistics on cohabitation the same way we treat statistics on marriage?
I don’t think so. There is also a trend towards cohabitation that does not end in marriage. Among older cohorts, cohabitation was more frequently followed by marriage than in younger cohorts. I suspect that younger cohorts are more willing to enter into temporary cohabiting relationships (ie the couple does not see the arrangement as a ‘trial marriage’ or a pre-marriage stage in their relationship).
According to Ruth Weston, Lixia Qu and David de Vaus:
Ideally we’d have some way of distinguishing between different types of cohabiting relationships. But I don’t know if that’s possible.
The question I’d ask if I was reading this, was whether there’s been a reduction in stable relationships (married or cohabiting) among men in this age group. And because not all marriages are stable and not all cohabiting relationships are unstable, it’s hard to be sure.
Nicholas – On your second comment, this 2003 paper by Ross Kelly and Philip Lewis addresses the issue.
I’m interested in research on changes in the demand for different kinds of skill. If you have some references let me know.
Workers who are displaced by technological change need not be lacking in skill. It may be that skilled workers are pushed into unskilled work because of a fall in the demand for the skills they have (for example, in printing the shift from hot metal type displaced workers with trades qualifications).
The skills in increasing demand need not be ‘cognitive’ skills or skills imparted by formal education. As you say, they might be ‘interactive’ or ‘social’ skills.
Some jobs require a combination of skill types so that lacking one skill makes you unsuitable. For example, making coffee in a cafe requires numeracy, social skills and motor skills.
If there is a decline in demand for strength and physical skill then it may be that employers who do require these skills will demand far higher levels of punctuality, reliability etc than in the past.
This would create the misleading impression that a decline in the ‘work ethic’ was the cause of the problem rather than a shift in the demand for skills.
I’m sending you a couple of papers I was sent by someone in NILS in 2007 which may be of interest.
“For women this means a decline in the number of men of marriageable age who are able to offer the financial security of stable full-time work”
You need to look at it the other way also. For (silly) men, the tradition of finding a woman lower in status than themselves will be harder also.
I’m also not sure about what seems to amount to a lot of moralizing. It seems to me that the assumption with a lot of this is that men = work and women = childcare, breeding, and a bit of work. I’m sure that was great for the 1950s, but if people can’t get over marrying others of a different status to themselves and having slightly different family roles, then is anyone really supposed to care? If the answer to that is no, then I don’t really see the point of linking marriage and the employment status of lower class males. I also don’t see it as a problem if women have been smart enough to get better educated than men and hence displaced them in some jobs (especially because I’m sick of hearing the almost always low IQ male line of how useless university is). Good for them too. Why look at it as a problem when it can also be looked at as a success? It’s not like women are doing better in the job market than males on many measures, so to some extent, it’s really just making things more equitable.
I agree with Conrad.
It might be helpful to also plot the corresponding employment figures for women over the same period. After all, what might be happening is really an effective sharing of workloads. If that were the case, you would expect an increase in employment of women coinciding with the decrease in employment of men.
Would you then necessarily conclude that the shift were a ‘bad thing’?
In fact, in a progressive tax system, it might make a lot of financial sense for many families.
Conrad & Marks
I agree that if men were taking greater responsibility for the household’s caring and housework and women were taking more responsibility for income earning work then this wouldn’t be a problem — it would be a change in gender roles.
But I don’t think this explains the decline in male full-time employment — particularly the decline in the employment of less educated men.
Time use surveys still show a large gap in the amount of time men and women spend on caring and housework.
For example, see the ABS survey: 4153.0 – How Australians Use Their Time, 2006
Don,
if you look at education statistics, then what you’ll find is that girls are doing better at school than boys (women have pretty much caught up to boys in “boy” subjects like maths, and have always been better at all the rest), and women are doing better at uni also (almost 60% of graduates are female now). It’s therefore no surprise that poorly educated men are doing worse — because there are more of them, so any movement towards an increase in overall skills that is demanded by the workplace (which has no doubt happened given China would have removed lots of the no-brain jobs), is going to affect the group with the lowest levels of education, which is men.
A second factor is that, at least from the US data I’ve seen which I think was collected by one of their big government agencies (I’ll dig it up if you want — it’s not so simple from where I am now which is why I haven’t got the link), despite the rather glaring misconception people have that you need hardcore mathematics skills to get a good job, the skills employers think are most missing from the workforce are those to do with literacy and communication. Given these are the areas that men with low education seem to miss the most (try listening to unemployed men on the Frankston line speak to each other, or whatever your equivalent is), it’s no doubt another factor that’s causing problems in that group.
A third factor may be just psychological. It seems far more common for me to see older women doing jobs that don’t seem like much fun to me (i.e., sitting in offices doing boring paperwork, low level sales etc.). Perhaps some groups of men are just unwilling to bring themselves to do this type of stuff.
Don,
impressived effort to shopw the linkages. Love the second, spiky, graph. A lot of the change for younger men is of course due to the fact that they are full-time education meaning they work part-time.
I disagree with Conrad and Mark above. It is not the case that the wife now has the job the husband had formerly and that he hence simply has to suck it up. Rather, compared to the 50s, work has become more concentrated over households: in some households everyone works, whereas in others, none. And the one without a job falls apart. Of course the problems would disappear if all the single mums married rich guys and the single, low-skilled men, found sugar aunties, but how realistic is that? Blame the need for men to have self-esteem, the need for women to have a partner they can respect, or the penalty in welfare rates if the partnership stays together, but the phenomenon of unstable partnerships and low employability at the bottom is real.
What is probably also true (some old paper of Bob Gregory) is that the partnerships dont quite fail. Apparently, single mums keep showing up with the same guy to consultation hours. This leads some to say that the welfare system has eroded the stability of the low-income marriages.
The big if is whether the trend is indeed still down or that we have plateaud. If the trend continues, the welfare system must really be reformed, if not, we will probably go on as we are now. The age-participation graphs do look like a plateau has been reached for men, apart from the younger ones where it would seem likely increased education levels are to blame.
Conrad – Interesting points.
I think we agree that the decline in male employment is not caused by women moving into the labour force. If women were to withdraw, employers would not regard today’s jobless and underemployed men as a substitute.
Your first point raises an interesting question — if there’s an increased demand for skills, is this because of technological change (eg the introduction of ICT) or is it the result of globalisation (your point about China)?
I think you’re right about the second point (Nicholas made a related point at #2). Some researchers argue that there has been a far greater increase in the demand for ‘interactive’ skills than for ‘cognitive’ skills. Most agree that the demand for ‘motor’ skills has fallen.
The third I’m convinced about. I don’t have data on this. It reminds me of the recent growth in the number of security guards. This doesn’t look like a particularly exciting job but it’s dominated by men.
A common feature of many female dominated occupations is multitasking. For example, women who do admin work in offices juggle answering the phone, taking deliveries, dealing with people who come to the door, ordering office supplies, making coffee, supplying food for meetings, washing dishes that other staff leave lying around in the kitchen … etc etc etc.
Paul – You write:
I’m curious about this. One of the reasons I chose men 25 to 34 is because I thought it was safe to assume that most were no longer in education or training.
What does the data say on this?
Don,
in your graph the 24-35 group is only 2% down in 2010 on 2000. Its the 20-24 group that dropped almost 10%. Since in the same period uni increased a lot (there are now over a million students. The big growth has been international, but still a 1-3% increase of domestic students per year. On a 10 year basis, that’s a big increase).
The Melbourne Institute is the one that usually compiles the latest figures. I havent looked recently, but would guess the drop in full-time work in the young age range is mainly because of schooling coupled with part-time work (there’s been a big increase in that for men the last 10 years).
You’re right that there has been a profound shoft in the labour market over the past thirty years, but I don’t think you can hope to understand much about it by only looking at the time series for one very small minority group – married men aged 25-34. I reckon you need to do a lot more delving amongst the ABS Labour Force time series stats, especially 6224.0.55.001, 6291.0.55.001 and 62091.0.55.003, before you can do more than merely speculate.
Only 7.4% of Australia’s civilian population over 15 are now married men aged 25-34. But in 1979 it was over 13%. That means that changes outside the labour market – delayed marriage and population aging – have had a big impact on this group. Unless you can isolate out the impact of these then its dangerous to try and draw conclusions about changes in labour market conditions for them. Isolating these out is of course possible, but not by just looking at this age group alone (which in the historical time series represents very different cohorts of course).
BTW Don most cohabitation counts as “marriage” in the ABS stats. It’s only the prudish Yanks who insist on only counting de jure marriages in their figures – something to bear in mind if you compare US and Australian data on a wide range of issues.
Really good conversation to have.
My grandfather used to point out the big mistake of having women work. I though his views were out of date, but like most things I see he had a point hidden in there.
Women working is a great thing, but its disruption to the way scoeity works will take decades to settle down. The direct effects on things such as house prices (DINKS have distinct advantages, and more subtle effects where women no longer need male security but perhaps still think they do in some way…
any case work to do!
“The big if is whether the trend is indeed still down or that we have plateaud. If the trend continues, the welfare system must really be reformed, if not, we will probably go on as we are now”
I’d bet that the trend will still be down. However, I don’t think it is because of the welfare system (I think future increases will have almost nothing to do with it). It’s because of cultural values that are causing girls/women to dominate the education system (shared, incidentally, by many but not all Western countries, which means this is a testable hypothesis) — I just can’t see how there won’t be more unemployed males as long as females keep on increasing in their relative performance compared to males. Even 20 years ago, for example, I believe there were more males than females at university and women hardly ever enrolled in some courses (like engineering). All of that has changed now (except for IT!), but these changes are just starting to propagate through. You see women now, for example, in Engineering, which is the only course area along with IT that women arn’t the majority now. This means that every time a recession comes along and people that arn’t well educated become unemployed, more males will hit the scrap heap than females, and the bigger the discrepancy in education levels, the bigger the pool of unemployed males will become.
Paul – I agree, for younger men there’s a big fall in full-time employment associated with participation in education and training.
But what really interests me is the abrupt decline in full-time employment associated with recessions. This isn’t about 25-34 year old men suddenly deciding to drop out of full-time work and go to TAFE.
A lot of commentators seem to be saying there’s nothing to worry about this time — the full-time jobs have come back and men are getting them. But in the spiky graph you like, there’s another cliff.
Shouldn’t we be worried?
Don
You’re obviously correct in observing that full-time employment never recovered after the Howard recession of 1982 or the Keating recession of 1990-91 (the recesssion we had to have). But does that necessarily mean that it won’t recover this time? At present it’s too early to tell.
Are there any factors that might make the experience different this time? I don’t know. I suspect that labour market regulation is less rigid now than in 1990-91, even accounting for Fair Work rolling back some aspects of Work Choices. That might militate in favour of employers creating more full-time jobs now. It might also be that the resources/mining industry is a greater proportionate driver of growth now than in 1992 etc. At that time one would suspect that the services sector (where part-time employment is much more advantageous) was a major driver of the recovery, whereas now it seems to be largely driven by China, India and resources. I suspect that part-time employment is much less relevant/feasible in the resources industry.
Don,
I cannot see the point of the first graph. It plots the proportion of men who are married AND employed. So there are two trends in the same graph – employment and marriage rates, the latter of which has changed (i.e. declined) greatly over 32 years.
Could you give the reference or data source for the second graph? Perhaps even link to a spreadsheet?
The last graph also seems to show that that employment has moved strongly in favour of older age-groups and against younger age-groups over the past decade – not what I would have expected. Again, could you provide the data source (and save me trawling the ABS)?
Nice post.
Chris – The data is from 6291.0.55.001 Labour Force, Australia, Detailed
And yes, there are two trends in the first graph. The point I was trying to make is that something that was once the norm — the package of marriage and full-time work — now applies to less than half of men in this age group.
For less skilled men, the breadwinner role is clearly no longer the norm.
Don,
your spiky graph shows a big drop right at the start of each recession in full-time work, almost as if companies are waiting to shed some of their full-time workers at the first hint of a recession. I would be worried about this if the total number of men working was a lot lower from trough to trough by recession. However, squinting at this graph, full-time employment is similar to the dip of last recession and full-time+part-time is decidedly up on the last recession. That also indicates a bottoming out of the joblessness of men. We do seem to be hitting a plateau of about 850 full-time employment and close to 90% employment for this group. I think politicians and the population can live with that.
At the older age-ranges, we see a slight increase in participation. Its hence hard to see how these graphs would create enough momentum to get people seriously worried.
I think Paul that it is not so much that they are waiting. Rather, ‘companies’ (obviously I mean the people running them) are inherently human and thus in 99% of cases a bit lazy and with a strong preference for the feel-good path.
So by the time a recession arrives, there are a number of staff who probably should have been fired and haven’t. The impending recession creates a perfect storm – everyone else is doing it, it looks like good management, one avoids personal responsibility by blaming the ‘crisis’, etc. Although I am only reasoning anecdotally I think that a lot of retrenchments around the start of the ‘crisis’ were more trimming back the fat than cutting out the meat, or perhaps, ‘catch-up sackings’.
Of course, there is no worse time to be sacked! Maybe that helps explain part of the problem – a significant proportion may have found a new job if they had been sacked a year earlier?
More than that, if they sack someone they must give good reason for doing so. If they “downsize” the company then the jobs are redundant and no reason is required other than cashflow and economic management. The company may stay downsized for a couple of years before new people are hired to fill in the same jobs. Thus the law as it stands actually requires business to run in a cyclic manner.
Paul – I wouldn’t worry about this if:
1. The latest recession had been as deep as previous recessions; and
2. Full-time job losses were evenly distributed across the population.
An extended period of growth has not taken male full-time employment back to 1990 levels for men in the 30s and 40s. This suggests the shift has been permanent.
Has the shift plateaued? I’d be more confident about this if the recent recession had been as deep as previous recessions.
I stronly suspect there’s been a decline in full-time job opportunities for men. And this decline has hit certain groups of men much harder than others. One of the reasons it’s a problem is because policy makers have no intention of restoring labour market opportunity for these groups (they don’t how to).
What they are likely to do is reduce the value of income support (relative to community standards) and make it more conditional.
A few years ago you and Bob Gregory wrote:
Hi Don,
I agree that the latest dip might not be hard enough to see whether we have truly hit a plateau or not.
Yes, I still subscribe to the paragraph you quote. We are in a new world where full-time employment for all until a reasonable retirement age is a bygone dream. I think our society has learned to live with that reality and seems comfortable in carrying 20% of its population more or less permanently. That is more apolitical judgment than an economic judgment though. It seems to me that real reform of welfare system and the labour market is off the agenda at the moment.
But what do you think? Are Australians, and then particularly the politicians, upset enough about the dependency of a large proportion of its population to seriously address its policies?
No conceivable “reform of the welfare system” will change these trends, because they have never been driven by the welfare system.
Australia’s welfare system for working age people is one of the most tight-fisted in the developed world. It puts far harsher conditions on job seekers than almost any other system and its payment rates to those job seekers are quite miserly. That’s exactly why marginal people job try mightily to become single parents or disability pensioners rather than get classified as “job seekers”. “Reforming” this system without analysing and acknowledging the real drivers of social and economic change will simply cause misery to no purpose.
I find it hard to see the rise of part time work as anything but a great boon. There is no evidence at all that it is driven by increasing underemployment (ie increasing restrictions on labour demand); rather it is driven by changed labour supply behaviour.
People no longer face a binary choice between exactly 40 hours a week of work or nothing. Australia has the second highest rate of part time work in the OECD (after Paul’s country of birth), while average hours worked of those working full time (ie >30hrs/wk) are at or near the longest in the OECD. Put those two facts together and you must conclude that Australia has amongst the widest distribution of working hours in the developed world.
Yet both our overall average hours worked per worker and the proportion of the working age population in work are above the OECD average, and (not coincidentally) our unemployment rate below the average.
Paul,
My response to your comment about whether we’re ‘serious’ about reforming welfare was roughly DD’s. So I for one, and I guess DD, would be interested in what you mean by ‘reform’. Given that you’re not a ‘starve them out’ kind of guy, I’m genuinely interested.
DD, Nick,
I am not sure the welfare system has not been involved in some of these changes. It is hard to look at the instability of partnerships of couples without a job and not think of the economic incentives involved inherent in the welfare system. Also, the alternative to work is by and large the welfare system (to which I include single parents and DSP), and it is precisely the observation that we now work very close to the same hours we did in the 60s but nevertheless have transfered 20% of our population from dependency on family and community to the state, that makes it hard to ignore the role of the state in these changes.
What could potentially be done? Well, there’s a difference between what will be done and what should be done. Politically speaking the main course of action that would be on the table first in any situation with budget urgency behind it is to reduce the welfare rates or to further ramp up the entry barriers (time-limits). A move to in-work welfare in the form of earned-income credit is another front-runner amongst the things we might actually do.
My personal favourite in the set of things I think we should do is to see parents on welfare with kids as being state-employed parents whose job it is to look after those kids in a way society approves of: many families on welfare with more than 1 kid could probably not earn enough on the labour market to make professional child-care efficient, and hence the best allocation of their time is for them to be parents. A change of language and in reward structure would need to accompany that. Things you can then think of would include rewarding parents when kids show up at school, do their homework, and stay out of trouble with the police. Home-teaching packages could replace job-search as the thing expected of welfare dependents with kids.
Come on Paul, that’s far too vague. If we time-limited benefits what do you propose to do with those whose benefits run out and whom no employer will take? There really are people whose marginal product in modern workplaces would be negative, usually for reasons that have nothing to do with welfare dependency (though sometimes something to do with “modern workplaces”).
If your answer is “let them starve” then, apart from thinking your implicit social welfare function highly non-linear, I’d suggest you haven’t thought it through. You won’t be raising their employment rate much.
If we time-limited our existing system we would actually be much harsher than the US, who at least give food stamps. It would do nothing to enhance these peoples’ employability, and would have spillover effects to the wider society through crime, health and general harshness of attitudes – as indeed we see in the US.
We would be the only country in the world outside the US to put time limits on our “last resort” safety net (as distinct from things like unemployment insurance), and even in the US the “five years in a lifetime” time limit on TANF is mostly theoretical – it’s usually only applied selectively, as a mechanism for when the state wants to take the kids into care. And as I said, it’s not as though our system was exactly generous in the first place anyway.
Perhaps you propose workhouses?
DD,
hold your fire, we are probably on the same side!
Dont confuse my quick and dirty run-down on what might happen with what I personally advocate (the parenting thing).
I agree with Paul that policy makers could pay more attention to the parenting role. Another option is subsidised work.
Some of the policy makers who are the most insistent about the beneficial effects of employment on adults and their children are the most resistent to addressing demand side issues.
People who insist on the benefits of stable employment should be prepared to fund job opportunities. Pedro Carneiro and James Heckman write: