What Job Instability?

29 June 2007 at 10:57 am Peter G. Klein 7 comments

| Peter Klein |

A truism among management scholars is that jobs, in the new, knowledge-based, hypercompetitive, deregulated, entrepreneurial, dog-eat-dog, Schumpeterian, long-tail economy, have become less secure. Perhaps my father or grandfather spent his career with a single firm and got a gold watch upon retirement but I constantly switch jobs, by choice or necessity, resulting in a loss of firm-specific or job-specific human capital, increased employee anxiety, and a deterioration of social bonds.

The data, however, suggest otherwise. In “The More Things Change, The More They Stay the Same: Trends in Long-Term Employment in the United States, 1969-2002,” Ann Huff Stevens finds in 1969, the average tenure for US men in their longest job was 21.9 years. In 2002, the figure is 21.4 years. The percentage of male workers working for a single employer for 20 years or more is the same was the same in 2002 as it 1969. By this measure, at least, jobs are as “stable” today as they were in the Good Old Days.

Entry filed under: - Klein -, Myths and Realities.

Thanks to Chihmao Spousal Combinations

7 Comments Add your own

1. Steve Phelan | 29 June 2007 at 12:07 pm

The data are actually for men 58-62 retiring in 1969 and 2002 and code for “time in longest job” rather than average tenure.

My father retired in 2002 aged 62. He also spent most of his life with one company but ended up “downsized” at the end. I would say the jury is still out for men who will be 60 in 2025 or later (like me). :-)

Didn’t Charles Handy start writing about “portfolio careers” around 1990? If he was reporting the start of a trend then we would not expect the effect of multiple jobs to start appearing in these data until after 2010.
2. Peter Klein | 29 June 2007 at 12:17 pm

Right, perhaps my phrasing above is awkward, but I meant the average duration, across men, of their longest job, not the average tenure per job per man.

It may indeed be too early to see the full effects of downsizing etc. by looking at today’s retirees, but one would expect at least some effect on that cohort, right?
3. Steve Phelan | 29 June 2007 at 10:21 pm

If 1990 marks the discontinuity then most men retiring today had a chance to have a job of 20+ years before the discontinuity thus the data will not reflect the discontinuity until 1990+20 = 2010.
4. Cliff Grammich | 30 June 2007 at 9:49 am

I’ll defer to the economists on this, but why would 1990 mark the discontinuity? Having (1) several uncles who left the hills of Kentucky for factory jobs in the North, and (2) done community research near the steel mills of Northwest Indiana, I’d say there was a much greater discontinuity (or more rapid transition to a “post-industrial” economy with supposedly more rapid job turnover?) in the early 1980s . . .
5. Cliff Grammich | 30 June 2007 at 10:16 am

Something I should have added before: CPS data reported by the BLS indicate the civilian unemployment rate was above 10 percent from September 1982 to June 1983–the only period of time since the data series began in 1948 that it was so high. So would this represent the peak discontinuity for this research to consider?
6. Steve Phelan | 30 June 2007 at 12:14 pm

I have no objection to arguing that the peak discontinuity occurred in 1982. The effects should then start to show in 2002. In that case, replicating the study with 2007 data should be very interesting.
7. David | 9 August 2007 at 1:42 pm

Here on a study showing job instability increasing. (Importantly, also ties in wage decline):

http://www.eric.ed.gov/ERICWebPortal/custom/portlets/recordDetails/detailmini.jsp?_nfpb=true&_&ERICExtSearch_SearchValue_0=ED422477&ERICExtSearch_SearchType_0=eric_accno&accno=ED422477