Are employers using part-time work to hang onto their workers?

So far during the current recession, the drop in employment hours has been much greater than the drop in employment. Some have described this as evidence that firms are seeking to hang onto their skilled workforce by reducing work hours rather than laying people off. Julia Gillard, for example, interpreted this as reflecting the fact that many employers, working co-operatively with employees and trade unions, are striking innovative arrangements to keep people attached to work during these difficult days (SMH, Aug 7, 2009). Is this a reasonable interpretation? Are employers moderating the impact of the recession by moving valued workers to part time work instead of laying them off?

I think the answer is ‘no’. First, the experience of part-time work in this recession appears to be similar to the experience in the last recession – though it is too early to be sure. Second, even though part time employment has increased in relative terms during the recession, this has not been because more people are moving from full-time to part-time employment. Rather, there has been a fall in the number of part-timers taking up full-time jobs.

Figure 1 shows overall employment and full-time employment, relative to the working age population, for the last 30 years (ABS trend estimates). Figure 2 shows the percentage of employment that is part-time.

Fig1 Employment and full-time employment

Fig2

Employment and full-time employment both fluctuate significantly with the economic cycle – with significant dips in the early 1980s, early 1990s and in 2009. The proportion of employed people working part-time has steadily grown over most of the period – except for between 2002 and 2008 where it was relatively steady. The recent decrease in average hours worked per employed person can be seen as a blip upwards at the very end of the period (denoted by C). However this pattern does not look particularly unusual, there were similar blips upward in the two previous recessions (A and B on the figure). It will be some time before we know if the pattern this time is different.

Moreover, how should we interpret these increases in part-time work during recessions? Is it evidence of a new flexible arrangements where workers reduce their hours rather than face unemployment? Or does it simply reflect the casualisation of the workforce, with more workers in positions where the employer can easily vary the work  hours from week to week?

Figure 3 suggests the latter rather than the former interpretation. About 80% of people in the ABS labour force survey are in the survey in successive months. The figure shows the number of people moving between full-time and part-time, as a percentage of the number of people employed in this sub-sample. Because there is much seasonal volatility, I show a 12 month moving average of the trends. That is, the July 2009 estimate is the average monthly movement over the previous 12 months.

Fig3 Monthly flows between full and part-time employment (12 month moving average)

On average, around 3 per cent of the employed population moves from full to part-time status each month. This has been growing over time as part-time work has become more important. The fraction moving to full-time is slightly larger than the fraction moving to part-time despite the fact that the share of part-time employment has been growing. The growth in part-time employment stems from the fact that more people are moving from non-employment to part-time than to full-time employment (not shown here).

There is no evidence that the flow from full to part-time has increased in the last year. Quite the opposite. The labour market slow-down has been associated with a general slow-down in movement between these two employment states. Moreover, the greatest change has been in flows from part-time to full-time, which have dropped significantly over the past two years. People who would have moved from part to full-time status are now staying in part-time employment.

This seems much more consistent with a story of a general growth in labour market flexibility than a story of arrangements being made between employers and workers to reduce hours during the recession so as to retain skills. Rather than telling full-time workers that they have to work fewer hours, employers are telling part-time workers that they cannot move (or move back) to full-time work. In some respects, this is still a good news story it may mean that fewer workers are being made unemployed. On the other hand it suggests that the burden of reduced work is still being concentrated on the most disadvantaged in the labour market. The part-time stepping stone between unemployment and full-time employment might simply have moved a little closer to the former.

This entry was posted in Economics and public policy. Bookmark the permalink.

9 Responses to Are employers using part-time work to hang onto their workers?

  1. derrida derider says:

    I have never understood lefties’ decrying of the wider spread of working hours we’ve seen in the last 20 years. If they’re not complaining about people being underemployed they’re complaining about them being overworked.

    There’s abundant evidence that the long-run driver of this widened distribution is changes in labour supply rather than a plot by slave-driving employers. And as for the shorter run, if labour demand is deficient isn’t it far better to have the deficiency spread around lots of people? Underemployment might suck, but its a damn sight better than unemployment.

  2. Tel_ says:

    Nice chartwork.

    I’d like to see Fig 1 stratified by number of hours worked per week. Part time work is a bit vague, could mean only one hour a week. Also part time workers regularly go up and down in the number of hours worked so plenty of changes are happening, ignored by the stats.

    Interesting that for the past 15 years we have seen steady growth in part-time employment, with almost no growth in full time employment. But the “flow” from PT to FT has at all times been higher than the reverse. This makes me think that a very important flow exists that is not being charted here.

    Another point, I note that the presumption is that employers make all the decisions. I’ve always seen employment (as with any trade) as a two way street. Why is it difficult to believe that employees find part time work convenient? I mean, if your hourly rate is good, the incentive to work yet another hour diminishes along with the nonlinear tax curve (made more nonlinear by medicare steps). If your hourly rate is bad, it makes sense to at least focus your time into some manner of self improvement, and only earn the minimal amount that you need to survive (plus some buffer I guess).

  3. Don Arthur says:

    Theres abundant evidence that the long-run driver of this widened distribution is changes in labour supply rather than a plot by slave-driving employers

    DD – I’m not convinced that changes in labour supply are the only thing driving this.

    If you create a chart like Bruce’s figure 1 (full-time employed by civilian pop)for males only you’ll see the proportion in full-time employment ratcheting down every recession. It never completely recovers.

    You might think it’s got something to do with early retirement or younger men going part-time to study or help care for children. But if you break the data down by age and marital status, it’s seems unlikely that either of those explanations really captures what’s going on.

    In recent years the proportion of older men in full-time work has increased. And since the downturn the proportion of 25 to 34 year olds in full-time work has fallen.

    I suspect that this has something to do with changes in the demand for labour rather than a change in the work preferences of workers.

    I’m curious about whether there’s been a change in the skills demanded by employers offering full-time work — perhaps a decrease in the demand for physical skills and an increase in the demand for interpersonal skills.

  4. Tel: Yes, it would be good if the ABS released more hours information. Another interesting question is how much change in hours there has been within the full-time population (eg decreases in paid overtime). BTW I should have added the ABS definition of part-time. It is “Employed persons who usually worked less than 35 hours a week (in all jobs) and either did so during the reference week, or were not at work in the reference week.”

    Re your second point. The flows from non-work to part-time are stronger than those from non-work to full-time. This is why part-time has relatively grown.

    All: The demand/supply balance over hours involves both employee and employer decisions. In periods of stable economic growth, the hours decision is more likely to be driven by employee preferences. However, decisions to reduce hours in times of recession are more likely to come from the employer – they are the ones facing falling demand for their product.

    I agree with DD that it is better to spread the hours reduction around. However, the point of my original post is that this is not being spread around as much as some people have suggested.

  5. If you are looking for more comparisions of hours worked in the last recession compared to this one see Bill Mitchell

  6. Tel_ says:

    Re your second point. The flows from non-work to part-time are stronger than those from non-work to full-time. This is why part-time has relatively grown.

    For three states, there should be six flows (three positive, and three negative where positive means increasing worker activity) or three nett flows (combining pos/neg flows in pairs). For simplicity, let’s only consider the nett flows:

    PT => FT is always positive
    UE => PT is almost always positive (based on my visual inversion of Fig 1)
    FT => UE seems generally positive but is the most variable nett flow.

    If you are looking for more comparisions of hours worked in the last recession compared to this one see Bill Mitchell

    I think Billy is barking up the wrong tree comparing 1991 with 2009 because the sort of deep changes that the USA is going through now, don’t have a counterpart for 1991 (e.g. the US dollar was not seriously under threat in 1991, it is now). However, his idea of a phase shift existing between the various flows could be interesting. Would it be possible to morph Fig 3 (perhaps even Fig 1) into state-space for comparison?

  7. Bruce Bradbury says:

    Tel: Bill Mitchell (and this blog) is referring to Australia, not the US. (Re-reading his blog, I can see how you could easily miss this. On the internet we all erroneously assume that everyone else knows where we are writing from. At least club troppo has .au in the web address ;-) ).

    For information, here is the latest gross flows data for Australia (000s). (Not E includes both unemployed and not in the labour force).


    June 2009 | July 2009
    | FT PT Not E

    FT |5847 270 115
    PT | 317 2182 184
    Not E | 101 194 4879

    So the net flows are

    PT=>FT 47
    FT=>Not E 14
    NotE => PT 10

    Unfortunately the flows data which can be downloaded from the ABS website dont go back to the last recession. I havent had time to dig more deeply. See also Jeff Borlands paper on labour market trends in recessions summarised here

  8. Tel_ says:

    Bill Mitchell (and this blog) is referring to Australia, not the US

    I did recognise this, but I automatically presumed a strong coupling between Australian and American economies. Not entirely unreasonable, given they are 10 times larger than we are, and highly influential in global economic and military affairs.

    Is there any analysis of the 2009 downturn possible without some reference to the subprime crisis and US banking practice?

    See also Jeff Borlands paper on labour market trends in recessions summarised here.

    Sounds interesting:

    The flows between employment and unemployment in Australia have followed a regular pattern during recessions and upswings. In particular, both tend to increase during recessions. It should not be surprising that the inflows to unemployment from employment increase during a recession, as this is a time when we expect many workers to lose jobs due to a reduction in demand for firms output. However, it might seem curious that the outflows from unemployment to employment also rise during a recession. How can it be that more people are finding jobs in recessions than upswings?

    To me this only reinforces a belief that recessions are times of restructuring, but no doubt plenty of people disagree with me on that particular point.

    I got interested enough to slap together a state space of the basic PT vs FT employment (ignoring the multidimensional flows that are too hard to chart). I’m using ABS pub 6202001 series: A183810W, A183808K, A183809L, A183823J (I convert all to population ratio so it makes sense to me). Spline fitting courtesy of gnumeric, red annotations by the Gimp. My conclusion is that inflections tend to happen around the end of a financial year. Bill Mitchell’s phase shift theory shows some legs in the small counter-clockwise loops at the lowest ebb of full time employment (MAY1983, AUG1993, JUN2003 — precisely one every 10 years).

    http://lnx-bsp.net/ABS_6202001_state.png

    [img]http://lnx-bsp.net/ABS_6202001_state_small.png[/img]

  9. Tel_ says:

    Hmmm, now I stare at this harder, a further conclusion comes to mind. Watch for the time when part-time employment (as population ratio) starts to go DOWN and that’s when you know the recession is over.

Leave a Reply to Bruce Bradbury Cancel reply

Your email address will not be published. Required fields are marked *

Notify me of followup comments via e-mail. You can also subscribe without commenting.