Youth rates, back on the agenda
The minimum wage for young people appears to be back on the political agenda. It shouldn’t be. Those wanting to bring back youth rates say they will slash youth unemployment. The evidence simply does not bear this out. There are much more important issues at stake including skills, education, productivity and living standards.
First, let’s look at the international research evidence. There have been numerous studies but attempts to link the level of the minimum wage to employment have been inconclusive. The findings of a 1998 International Labour Organisation survey (1) are still true: “All in all, it seems fair to conclude that the existing evidence supports both positions in the debate. Whether a minimum wage has a negative or a positive effect depends on many factors such as its relative level, the structure of the labour market and the country concerned.”
A major study by Dean Hyslop and Steven Stillman in 2004(2) looked at the effect of increases in the youth minimum wage in New Zealand the early 2000s. It found “no robust evidence of adverse effects on youth employment or hours worked. In fact, we find stronger evidence of positive employment responses to the changes.”
One of the most recent large‐scale studies was published in 2009 by Doucouliagos and Stanley(3). It re‐analysed 64 US minimum‐wage studies including 39 relating to teenagers and found not only bias in selection of published studies towards ones which show an adverse effect for employment, but once such effects were corrected for, positive effects between an increase in the minimum wage and employment. At worst, they found “no practically significant adverse employment effect”.
Research published in November 2010 by Dube, Lester and Reich(4) looked at low‐wage workers over a 16‐year period in all US counties bordering a county in another state with different minimum wages. They found that increases in minimum wages had no negative effects on low‐wage employment.
Eric Crampton argued in a recent column on this page that this last study did not apply to New Zealand because it studied restaurant workers who can’t be “outsourced” compared to call centre operators or machinists. But over a third of 15‐24 year old employees in New Zealand are in retail, accommodation and food services (which are the biggest users of minimum wages), and about three quarters are in industries that similarly can’t be easily outsourced.
Crampton asserts that the abolition of the youth minimum wage in 2008 resulted in the unemployment rate for 15‐19 year olds rising more steeply than that of other age groups. However minimum rates for 18‐19 year olds were raised to the adult minimum much earlier than that and as we have seen, Hyslop and Stillman found no employment problem. Only 16 and 17 year olds were affected in 2008 – well under half of the 15‐19 year old labour force. In addition, other events since 2008 may well have affected youth employment.
Some clues emerge by looking beyond unemployment rates to actual employment numbers. The number of unemployed 15‐19 year olds has risen slower between March 2008 and March 2011 than other age groups: by 56 percent compared to 80 percent. The unemployment rate has risen more quickly because it equals the number of unemployed as a proportion of a shrinking 15‐19 year old labour force. Over this period the 15‐19 year old labour force fell by 19,800. Many of these young people have gone back to school or into tertiary education. Secondary school rolls rose (contrary to Ministry of Education forecasts) as did 18‐19 year old participation rates in tertiary education.
This highlights the fact that the serious problem of youth unemployment is much more complicated than a simplistic wage theory. Many 15‐19 year olds are in secondary school, tertiary education or industry training – and more should be. Some may be unemployed while studying. Many are already raising families. But there is a core that is doing none of these. Rather than cutting wage rates, for which there is scant evidence of employment benefits, we should be improving education, training and employment pathways. Are caps on tertiary enrolments and policies discouraging tertiary institutions from increasing their level 1‐3 certificate enrolments (both also introduced in 2008) disadvantaging these young people? Are unclear school pathways to suitable vocational education and work‐based training letting them down? Are youth employment programmes effective and are there enough of them?
Finally, the minimum wage is part of a wider picture. Orthodox economics has a poor understanding of work and wages. That car you like in the car yard will not perform better if you pay more for it. But people do. The above studies complement evidence that productivity increases and turnover reduces if firms pay their employees well and help them increase their skills. One reason the minimum wage has assumed such significance in New Zealand is that – unlike Australia and many European countries – governments have jettisoned any other involvement in wage setting. During the 1990s, wage increases fell well behind increases in labour productivity.
Until we have a wage system that produces more sensible outcomes we will need a minimum wage that ensures that those at the bottom of the heap are treated with a modicum of decency.
Dr Bill Rosenberg is Economist at the New Zealand Council of Trade Unions New Zealand Council ofTrade Unions Te Kauae Kaimahi. Further details of the research quoted above, and other relevant research, can be found at http://union.org.nz/news/2011/minimum‐wage‐and‐jobs
1 Youcef Ghellab, “Minimum Wages and Youth Unemployment”, ILO, 1998.
2 Dean Hyslop and Steven Stillman, “Youth Minimum Wage Reform and the Labour Market”, NZ Treasury Working Paper 04/03, March 2004.
3 Hristos Doucouliagos and T.D. Stanley, “Publication Selection Bias in Minimum-Wage Research? A Meta-
Regression Analysis”, British Journal of Industrial Relations, 47:2, June 2009, pp. 406-428.
4 Arindrajit Dube, T. William Lester, and Michael Reich, “Minimum Wage Effects Across State Borders: Estimates Using Contiguous Counties”, The Review of Economics and Statistics, November 2010, 92(4): 945– 964.