Why did the middle class shrink?

In his publication Is the Middle Class Shrinking?, Clive Hamilton writes ‘there does not seem to be any survey evidence on identification with class terms’. We all know Clive is no ordinary leftist, but it is remarkable that he became a leading ‘progressive’ thinker despite clearly having read very little about class. There is lots of research on this, going back nearly 60 years.

This research suggests that the long-term answer to Clive’s question is ‘yes’. Though the proportion of people identifying as ‘middle class’ has been trending upwards since the early 1990s, most of the surveys on class identification between 1949 and 1984 found more middle-class people than a 2005 survey. The one exception was 1965, but it was just 1% lower than the 2005 figure of 50% ‘middle class’.

The curious thing about the apparent shrinking of the middle class is that it occurred while many of the sociological markers of the middle class, such as education and professional or managerial occupations, were showing long-term increases. For example, in 1947, just before the first class survey I have, 12% of workers were in professional or managerial occupations. But in 1949, 54% of people thought that they were ‘middle class’. In the mid-1990s, 31% of workers were in professional or managerial occupations, but overall only 45% of people considered themselves to be middle class.

One possibility is that though compared to the past a higher proportion of people have professional occupations, university education and high income, the relativities have moved with them and so the middle class has not grown. Around 50% middle class seems right, with 25% higher and 25% lower. But very few people will describe themselves as ‘upper class’. Indeed, despite the huge increases in wealth at the top end of the income spectrum, less than 2% of Australians call themselves upper class. Despite the loss of blue collar jobs, the working class seems to have peaked at 53% of the population in 1993, though it has been declining since (41% in 2005). Could people have been objectively moving up but subjectively moving down?

It is possible to acquire sociological attributes of the middle class without feeling middle class. About 15% of people with university degrees call themselves ‘working class’. 18% of people in professional occupations are ‘working class’. The same proportion of those earning $78,000 to $104,000 a year see themselves as working class, despite having the money to live a comfortable middle-class lifestyle. (All 2005 figures.)

Though these statistics show the sociological markers of class are fairly reliable – the vast majority of people with degrees, in professional jobs and high incomes are ‘middle class’ – they also suggest that class is more complex than just these things. In her excellent book Australian Liberals and the Moral Middle Class, Judy Brett argues as her title suggests that the middle class saw itself in moral terms. In 2003 I summarised it this way:

She notes how the middle class has defined itself in moral terms, as displaying certain values that distinguish it from the working class. In the earlier years, this came from a sense of duty, and Brett persuasively shows how this was displayed in the middle class response to various national crises. This ethos, though never extinguished, has greatly weakened since the 1950s, for many reasons – peace and affluence reducing the need for it, the decline of the churches, and the rise of a new form of individualism based on self-expression rather than self-reliance.

On this account, there are cultural differences between the classes, which still manifest themselves in different patterns of tastes, attitudes and aspirations. These are perhaps more resistant to change than occupation and income – hence the nouveau riche and the genteel poor. Unfortunately, there seem to be very few questions in the surveys that ask about class that could serve as good proxies for differences in class attitudes.

If culture is inherited more than occupational and income status, one explanation for the period of middle-class decline could be demographic – they simply did not have enough kids to maintain their share of the population. Educated women have always had fewer children than uneducated women (though mostly because they were less likely to marry). That wouldn’t explain the last dozen years in which the middle class has been growing again though. Perhaps the long boom is causing sufficiently large changes in objective social position that subjective social self-perception is catching up.

22 thoughts on “Why did the middle class shrink?

  1. While knowing how people identify themselves in terms of class is interesting, I wouldn’t put too much weight into the answers. Certainly this is not a way to define classes. What university graduate with 100+ salary will define himself as working class? Probably a union boss.

    I don’t know what is upper class, but as you say, few people will openly say they belong to it. These are such distortions thta make the whole analysis extremely unreliable, as not only we are dealing with self identification, but on top of that, with what people are willling to say to a stranger.

    I am completely ignorant about the problem, but my humble view is that there ought to be an objective way to define class based on objective parameters such as occupation, education level and income. Occupation, I think, should be the key parameter.


  2. Boris – On the other hand, you usually don’t need to know a person’s education, job or income to guess their likely class. Ten minutes general chit-chat would let you hear their accent, detect aspects of their interests and worldview, and judge their clothes and manners. There will be a high correlation with education, job and income, but the causation is running in both directions, with middle classness both helping create and reinforcing the more objective sociological markers.


  3. “Ten minutes general chit-chat would let you hear their accent, detect aspects of their interests and worldview, and judge their clothes and manners”

    Hmmm … Kerry Packer, upper classs on any sensible definition of the term (inherited wealth, boarded at Geelong Grammar)

    Accent – broad Australian
    interests – Rugby League
    Worldview – f*ck off you f*ckin’ wanker
    Clothes – ??
    Manners – f*ck off c*nt


  4. Spiros – On the other hand, my rare encounters with the overclass suggest that conversation quickly involves names of prominent persons and wealth beyond the possibility of the most overcharging tradesman.

    No indicator is perfect, but would I rather have ten minutes with a person or ten minutes to read their census form? Probably the former.


  5. If people lose their high paying salary (Middle class term) and move to a low wage (Working Class) do they change class? Do members of the Melbourne Club lose their class (as distinct from clout) when they face bankruptcy? Or is it a cultural and personal thing as well? Class is not supposed to an useful indicator of voting behaviour. David Kemp research “proved” it in the 1970’s. Why can you drive through or take public transport through a suburb anywhere in Australia and easily predict voting patterns? Travelled on a train lately?
    ‘Labor View from Broome’


  6. Class and party ID are only moderately correlated. But the parties are changing as well as the class structure; the Labor Party is becoming more middle class on culture issues and the Liberal Party is becoming more working class on cultural issues.

    The Greens do well among upper and middle class people, but not so well among the working class. This may change as they are increasingly funded by radical left unions that will want something for their money.


  7. Andrew allow me to be provocative and suggest that class distinctions no longer matter anymore, the terms ‘working’ ‘upper’ and ‘middle’ class having become meaningless in the traditional sense.

    When we can’t identify members of any class according to their accent, residence, income, education or employment, or when those indicators conflict with each other (for example, in the case of Kerry Packer) then we can’t meaningfully allocate people among class categories: any such allocation would have to be so qualified as to render it unuseable for analysis.

    No wonder the numbers don’t seem to make much sense – the ‘Venn circles’ keep moving around and intersecting with one another!

    I remember seeing at school a photo of a crowd of Australian men, taken at the announcement of hostilities in September 1914: the ‘working classes’ wore flat ‘ats, the middle classes wore boaters and the upper class gents wore toppers. In those days, you could tell which class a person belonged to with a glance.

    Now, it might be better to restrict the indicators of class membership to incomes, and discussion of changes in income levels and distribution. That then might be a starting point for associating certain tendencies in attitudes and behaviours with different income levels, with the caveat that those tendencies are unlikely to be stable.


  8. Isn’t this class thing so passed its used by date though?
    Isn’t it more of an IQ factor now?

    No one gives a shit who or what you came from as long as you peform.

    Becoming CEO of BHP and family background has no importance whatsoever. It’s ability and peformance in previous jobs that matter when choosing a candidate.


  9. “family background has no importance whatsoever. It’s ability and peformance in previous jobs that matter when choosing a candidate.”

    True, the upward mobility is huge nowadays. Yet it does help to be a son of Kerry Packer or Ruppert Murdoch, doesn’t it?

    It think the trasdtional rigid class structure is outdated, and class boundaries blurred. But it does not mean mean that the word class, in cultural more than economic terms, still exists. However apparently everyone has his own idea of class. Maybe we need some sort of definition.


  10. What happened to the study that showed “middle class” splitting in two different directions? The “middle/cultural” and “middle/aspirational” (can’t remember exactly what they were). The divisions were reasonably predictable (i.e. living in a restored house rather than a new one, interest in “high” culture rather than mass consumption of “pop” culture). I doubt the aspirational middle class readily identify with the concept of “middle class”, and therefore don’t self report as such.

    As for Kerry Packer – he spent as much time as he could slumming. My guess was it was an overstated reaction to that upper class background where he didn’t really fit (father didn’t like him much, school achievement was poor). You could, however, place him in the upper class by his consumption.


  11. Boris


    Years ago you couldn’t be a stockbroker unless you had family background. There days you can’t be a stockbroker unless you perform.

    The same goes for all jobs and professions.


  12. Andrew Norton

    I think accent went out as a class signifier about 15 years ago. The old rough nasally “working class” accent is pretty rare among people under 25, and even the most expensively educated and well-groomed child will sound no different that some airhead from Neighbours.


  13. Also universities are so easy to attend nowadays that two generations ago half the current student body would either have been jockeys, chippies, check-out chicks, hairdressers, or in jail


  14. John, I have to disagree with you. There is a real difference in both accent and class between Melbourne Uni law students (three-quarters of whom I’ll bet come from private schools) and the average Broadmeadows teenager. While private school kids may not speak the Queen’s English, they have their own accent (Ja’mie from Summer Heights High is a corker of an example), which is quite different from that of less privileged kids. I’m not a speech therapist, but I think the private school accent involves stronger/clearer enunciation, representative of kids who are encouraged by their schools and middle-class parents to speak their minds.


  15. John – I don’t know enough under-25 people to judge this – there is definitely a new set of accents among younger people but I am not sure how uniform they are.

    It reminds me, there is a program tonight on the ABC on accents.


  16. Rajat

    Well I agree with you on the example you have given, but I think you have chosen probably the two most extreme outliers in Australia! 🙂


  17. No shearers left Yobbo – he’d be working at the mines making $100,000 instead. His accent wouldn’t change though.


  18. The ABC show on accents last night was pretty good, particularly on the role of kids in setting accents. People from all classes speak with what they called the ‘general’ Australian accent, but I think it would be most dominant in the middle class.


  19. Spiros, from my ancient experience (early ’90s), it would be 95% if you included the selective and sought-after high schools such as Melbourne, Uni, Glen Waverly, Balwyn, etc. I was friends with a girl who went to the non-selective Heathmont High, but her accent was as middle-class as they come.


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s