Tuesday, 15 February 2011

Explaining quantities without prices

However, the peculiarity is the way young people seem to be so discriminated against in the statistics. Our adult rates fall back to just 4.5 per cent once youth are washed out of the figures.

There appears to be a structural flaw, an age fault line, that has lain concealed in the New Zealand economy during better times and which the recession is now bringing to light.

So why is New Zealand at the extreme when it comes to the difficulties its young now seem to face? The answer turns out to be unusually complicated.
Here's the list of candidate explanations in John McCrone's lengthy feature article:
  • A growing group of "disaffected 15 to 24-year-olds who are "not in employment, education or training""
    • But if they're not in the labour force, they won't be in the unemployment stats.
    • Youth labour force participation rates are well down - this means the problem is worse than the unemployment stats would suggest. Youth labour force participation rates have been below fifty percent for five of the last eight quarters; they'd previously dropped below fifty percent once in 1999 and in four quarters between 1992 and 1994.
  • The recession has a bigger effect on marginal groups like youths and especially minority youths
    • But the effect has been far larger in this recession as compared to prior recessions - youth unemployment has gone up by rather more relative to the adult unemployment rate this time around.
  • Skill matching problems: graduates with good skills nevertheless have a hard time matching their skills with employer demand during tight labour markets
    • True. But unemployment in the 15-17 year old cohort is very very high and won't fall into that category.
  • A glut of kids has come onto the labour market due to demographics
    • Oh yeah? HLFQ.SAZ3AA says otherwise. That series gives total labour force (employed and unemployed) aged 15-19. The average number of persons of that age category in the labour force, March quarter 2008 through September quarter 2010, is 164k. The average for the entire series going back to March quarter 1986 is 161k. Let's look to the period when unemployment was lowest for 15-19 year olds: September quarter 2004 through December quarter 2005. During that period, there were on average 163k kids in the labour force. There is no way that changes in the number of 15-19 year olds in the labour force is what's driving current youth unemployment rates. Maybe there's something there if you consider the slightly older group. But it's the 15-19 year olds who are having the roughest time on the labour market [the unemployment rate for 20-24 year olds is less than half of that for the 15-19 year olds].
  • It's a shift in industry demand: firms now demand tech-savvy workers for service employment
    • Did that happen all of a sudden part way through 2008? No?
  • Firms may be more "risk adverse" in taking on young kids.
    • Now we're starting to get somewhere. But why - WHY - might firms have just suddenly started becoming so "risk adverse" starting around third quarter 2008 - and to a far greater extent than in prior recessions? Can we think of anything? Anything at all?
  • Firms just aren't hiring as much - they're keeping on existing workers
    • Ok, if this is happening, then it would more greatly affect folks coming into the labour market for the first time. But why is it worse this time for folks in the 15-19 bracket than it was in any prior recession going back to 1986?
  • Industries hit worst by the recession - retail, hospitality and construction - tend to employ more youths
    • Aren't those the industries that usually get hit in recessions? And this time around's been rather worse for 15-19 year olds
  • Wait, this looks promising:
    Finally, O'Reilly says, there is probably the most frequently cited reason for soaring youth unemployment - Labour's 2008 abolition of youth rates.

    Employment laws were changed so that even 16 and 17-year-olds had to be paid the adult minimum wage.

    O'Reilly says this removed any incentive for traditional youth employers, like supermarkets and fast- food restaurants, to take on young staff.
    The abolition of the differential youth minimum wage.
    • Finally! The first three times I read the story, I missed this part. I just can't see how any of the other tendered explanations fit the data.
Here is the picture that a candidate explanation has to be able to, well, explain. Again, and as before, this is a plot of the residuals of a dead simple OLS regression where the youth unemployment rate (15-19 year olds) is the dependent variable and the adult unemployment rate (20 and up) is the independent variable. This one's now updated with December quarter 2010 results. The residual in December quarter was 8.5 (9.3 if we restrict the model to quarters prior to the regime change). The highest value prior to the abolition of the youth minimum wage was 2.5 in June quarter 2003. There's a six point difference. Something happened in 2008 that resulted in current youth unemployment rates six points higher, relative to adult rates, than was the trend through prior recessions. I can't see how any of the other provided explanations explain why things are so much worse this time around.

O'Reilly and Paul Dalziel both point to government programmes to fund industry training courses for youths as potential solution. We might also want to consider not pricing them out of the market.


  1. "we also might want to consider pricing them out of the market". I worry that the youth minimum wage was "pricing out" adult low productivity workers.
    Is there any data that can tell us how the impact of the removal of the youth minimum wage has been distributed between a reduction in total jobs and a shift from youth workers to adult workers?
    As you say, the adult unemployment rate is at a very low 4.5 percent.

  2. I don't know that for the current period in NZ. But there's a US estimate; I noted it here.

  3. Alfred, if you are worried about the youth rates pricing adults out of the market, then how would youth 'skilling' adults out of the market be any different. Training the youths will make them more attractive as employees than some adults.

    Youths would be hired instead of adults, but because their value has increased, not because their price is lower. Why is one better than the other?

  4. Thanks Eric!
    Kimble, yea umm "worry" was the wrong word, I was "worried" about the interpretation of the data rather than the underlying issue (Eric's link moves some way towards resolving these "worries"). I'm quite agnostic on the whole issue and was merely looking for more info. I would always support the hiring of a more productive youth over a less productive adult--at the same price.
    To see that perhaps half of these youth jobs were lost rather than given to full minimum wage workers definitely supports the reintroduction of the youth minimum wage.
    I have no comment on the inherent educational benefits of youth work--or on the benefits of adult employment--to future productivity of these workers, however I'd be happy to read any material available.