Research says low paying jobs damage future employment prospects

While unemployment is viewed as a bad signal by prospective employers, economists have speculated that being in a low-quality job may well be an equally bad signal. Professor Stewart has investigated this hypothesis and looks at how the overall employment prospects of people in low paid jobs compared on the one hand with those on higher rates of pay and on the other hand with those who are unemployed.

Professor Stewart looked at data on 4739 individuals over six years in the British Household Panel Survey (BHPS) from 1991 to 1996 (chosen to be prior to the introduction of the National Minimum Wage). He identified those earning less than £3.50 hour (in 1997 terms) as being low paid.

His study found that employees in a low wage job are 2.7 times as likely to be unemployed a year later as those who were higher paid. Professor Stewart also found that the probability of reentering unemployment for someone who gets a low-wage job after a spell of unemployment is twice that for someone with the same characteristics who manages to get a higher paid job after the unemployment spell.

His research also established that being in a period of low waged employment had almost the same detrimental affect on future employment prospects as a period of actual unemployment.

Professor Stewart said:

"Low-wage jobs act as the main conduit for repeat unemployment. The results in this paper suggest that not all jobs are ‘good’ jobs, in the sense of improving future prospects, and that low-wage jobs typically do not lead on to better things. If unemployed individuals’ future employment prospects are to be permanently improved, they need to find jobs where they can augment their skills (for example through training) and move up the pay distribution. Low paid jobs typically do not provide this."

The research paper, entitled "The interrelated dynamics of unemployment and low-wage employment", was published in the Journal of Applied Econometrics.

Citation: J. Appl. Econ. 22: 511–531 (2007). DOI: 10.1002/jae.922

Source: University of Warwick