The lie behind the Right’s attack on wages

Written By: - Date published: 7:39 am, February 12th, 2010 - 35 comments
Categories: minimum wage, national/act government, roger douglas, wages - Tags:

Paying a person doing the same work as another person less money because of their sex or religion or ethnicity or any other grounds prohibited under the Human Rights Act is illegal and abhorrent. Yet, the Right wants to do just that with a private member’s bill from Roger Douglas reintroducing a lower minimum wage for 16 and 17 year olds (Age is a prohibited ground for discrimination when the person is over 16, HRA s21(1)(i)).

Why do they want to do this? The real reason is because they hate the minimum wage. If you aren’t in the position to bargain for a living wage then that’s tough sh*t for you, is the Right’s position. But that’s not the official line, of course.

No, they wheel out some supposed “academic research” that supposedly shows the abolition of the youth minimum wage on April 1 2008 has caused unemployment among 15-19 year olds to rise faster than for other age groups. The “academic research” is a post on an obscure libertarian blog and the evidence is false. Abolishing the youth minimum wage hasn’t caused 15-19 unemployment to rise. 15-19 unemployment has risen because of the recession.

15-19 unemployment is always higher than that of other age groups. Yes it has risen more than the unemployment of other age groups but it has risen in proportion with the unemployment of other age groups. If the increase was due to the abolishment of the youth minimum wage then we would expect the increase in 15-19 unemployment to be disproportionate. It hasn’t been disproportionate, in has been in proportion. Just like Maori unemployment has risen higher but in proportion with unemployment in other ethnic groups. If the Right genuinely think that abolishing youth rates increased 15-19 unemployment to be consistent they have to claim it also increased Maori unemployment, or find another explanation specific to Maori, or propose a lower minimum wage for Maori, and I don’t see them doing any of that.

Look, I go swimming with a mate sometimes. He’s a bit faster than me. He’s done 10 laps by the time I’ve done 8. If we were to do more and he reaches 20 when I’m still at 16 would we say ‘Jesus, what happened? How did you beat me by 4 laps when usually you beat me by 2? Did you have some extra weetbix for brekkie?’ No. The gap is proportionally the same, it’s just we’ve gone further. Likewise, the unemployment rate for the 15-19 year age group (only 2 years of which was covered by the youth minimum wage, remember) hasn’t risen out of proportion to the unemployment rate of other age groups,  the gap is bigger because total unemployment is higher:

There was long-term trend through the 2000s of the ratio of 15-19 unemployment to general unemployment rising (because unemployment between older people was falling more sharply) but there is no change associated with the abolishment of the youth minimum wage on April 1 2008. If the youth minimum wage was causing youth unemployment, the ratios should have jumped. The relationships between 15-19 unemployment and unemployment in other age groups have remained basically the same. They have all risen due to the recession, and they have risen in proportion to each other. No other explanation needs to be made up…. unless of course your real agenda is to undermine the wages of working New Zealanders.

35 comments on “The lie behind the Right’s attack on wages ”

  1. Has there been a proportional increase in crime amongst teenagers and polynesians ( inclusive of maori) to match the jobless stats with possibly an increase in drug use too ?

    just wondering cos they’ll need a ready market to keep the jails full and well, “n*ggas gotta eat and do something to take they minds of being poor…nahmsayin ?”

    …ties in nicely with the unwillingness for drug refom.

    • Marty G 1.1

      yup. unemployment goes up, so does crime, especially burglaries.

      saw that for the june 2009 stats, and will likely see it when the dec 2009 stats come out too.

      those numbers don’t break down by ethnicity, it’s just recorded offences. there are numbers ethnicity of people getting convictions, but i wouldn’t rely on them to give an accurate pic of the ethnicity of people committing crimes.

      • Paul Walker 1.1.1

        Actually there are some interesting stats coming out the both the US and the UK showing that crime has not jumped in the way you would expect given the recent recession. I have commented on this here and here.

        • RedLogix 1.1.1.1

          A skim through the comments thread of that WSJ article would suggest that the academic rigour of those ‘stats’ is about as firm as their usual standard of article on global climate change… ie fairly floppy.

          • Paul Walker 1.1.1.1.1

            The US stats come from the FBI’s Uniform Crime Reports. Can see why they should be wrong.

            • Marty G 1.1.1.1.1.1

              Don’t know if it’s escaped your attention, Paul. But we’re not in America.

              There have been studies here linking crime rate and unemployment rate. I’ve shown there is a very high correlation between the two – http://www.thestandard.org.nz/might-be-taken-the-wrong-way-by-some/

              • Don’t know if it’s escaped your attention, Paul. But we’re not in America.

                And so? The lack of a positive relationship between unemployment and crime in the recent data from the US and the UK is still interesting.

                Note also what Papps and Winkelmann write “The unemployment-crime relationship is an old issue. No consensus has been reached by economists during the last three decades, nor does one seem likely to emerge in the near future. (Papps and Winkelmann p. 68)”

                Papps, Kerry and Winkelmann, Rainer. ‘Unemployment and Crime: New Evidence for an Old Question’, New Zealand Economic Papers, June 2000, v. 34, iss. 1, pp. 53-71.

              • Marty G

                Paul. Try to keep up. I asserted that crime goes up when unemployment does in NZ. I’ve supplied some data. You’ve supplied a link to a rightwing website giving its interpretation of the situation in the US.

                Not that you’re one to talk about consensuses among economists on issues but I know there isn’t a consensus among economists on this. Why should there be? This is an issue you want to ask criminologists about.

                If economists don’t have a consensus on relativity should we doubt that too? There’s this weird thing going on in rightwing economics, it’s trying to take over the social sciences, tidying up those ‘messy’ disciplines with its nice numbers and pretty formulas.

              • felix

                Interesting indeed Paul, but far more interesting if you’re more interested in the US.

                If you’re more interested in what’s happening here in NZ, then the interesting NZ studies are more interesting than the interesting US ones.

                Any comment on Marty’s interesting correlation or are we just ignoring that?

              • Not that you’re one to talk about consensuses among economists on issues but I know there isn’t a consensus among economists on this. Why should there be? This is an issue you want to ask criminologists about.

                It is a well studied issue. Its just that different studies come up with different results and thus as Papps and Winkelmann put it “The unemployment-crime relationship is an old issue. No consensus has been reached by economists during the last three decades, nor does one seem likely to emerge in the near future. (Papps and Winkelmann p. 68)’

              • Marty G

                You’ve got to love the arrogance of neoliberal economists.

                Is unemployment a cause of crime? Does rising unemployment lead to rising crime?

                Shall we ask an expert on crime and criminal behaviour?

                No, we know the answer “economists say there is no consensus”

                see, felix, we don’t need any of the other social sciences. Economics tells us all we need to know.

              • felix

                Ask a magician, get a trick answer.

            • Paul Walker 1.1.1.1.1.2

              If you’re more interested in what’s happening here in NZ, then the interesting NZ studies are more interesting than the interesting US ones.

              Any comment on Marty’s interesting correlation or are we just ignoring that?

              Its a correlation. Unfortunately correlation don’t tell us that much. There is a high correlation between rainfall and inflation, for example. But you don’t learn anything from that. Which is why serious studies use more advanced statistical techniques to analysis the data.

              The relationship between unemployment and crime could be positive. It is possible that an increase in unemployment will increase crime and studies have found such a relationship. But it is also possible that there is little relationship between the two, and other studies have supported this idea. Thus we don’t know. That explains the Papps and Winkelmann comment.

              • felix

                Without wanting to be rude can I make a wee suggestion? Do you think you could indicate when you’re quoting someone with some quote marks or tags?

                It’s just that it gets a bit confusing sometimes figuring out what you’ve said vs what you’ve quoted.

  2. jcuknz 2

    All your arguments do not alter the fact that it is silly to expect an untrained person to receive the same wage as an experienced worker.

    • Marty G 2.1

      And I’m not arguing that they should.

      Same pay for same work.

      An apprentice or other untrained person obviously isn’t doing the same work as a trained person.

      But it’s nice to see you resorting to arguing against a strawman straight away. That means you’ve got nothing to refute the actual position of the post.

  3. For the work involved in minimum pay i’d expect minimum training and in cases involvng physical labour i’d go for the young un…

  4. prism 4

    I remember being a youth worker in the days of union affected wages. There was a fairer deal to employers than everybody getting minimum wage from the start. Youth started on a lower level which rose with your age as you continued with the same employer. There was extra paid for your qualifications.

    It still had its unfair side though. As the worker became experienced and was given more responsibility the wage still kept pace with age rather than tasks. Young people were often given senior tasks but still paid the same as others.

    I have also been an employer. I think there should be a lower wage paid at the start, for three months say, and that allows for time to learn the job and for the extra work of employer and others checking and passing on needed knowledge. Then a wage rise, and recognition for responsibility and qualifications.

    Some present employers can be shocking. One I heard of was a woman managing a main street chain dress store receiving less than $10 per week over staff rate, despite the extra tasks – the target requirements for turnover, overview of staff prowess and honesty, upkeep of shop and care of valuable stock, security etc.

    • Daveo 4.1

      I have also been an employer. I think there should be a lower wage paid at the start, for three months say, and that allows for time to learn the job and for the extra work of employer and others checking and passing on needed knowledge. Then a wage rise, and recognition for responsibility and qualifications.

      There’s provision for that in the existing law passed by Labour. It’s called the new entrants minimum wage.

      What is the New Entrants Minimum Wage?

      A new entrant is a worker who is 16 or 17 years old except if

      * they have completed three months or 200 hours of employment, whichever is shorter, OR
      * they have been supervising or training other workers, OR
      * they are subject to the training minimum wage.

      http://www.ers.dol.govt.nz/pay/newentrant.html

  5. MartyG,

    I am not going to spend a whole day arguing with you here as my colleague Paul Walker does. But please read my follow-up post on the topic. You’re right: my first post just looked at the difference between the two without considering the possibility of a multiplicative effect; but yours dismisses the possibility of a constant term mattering.

    Paul and I had a beer over the weekend and wondered whether it’s best to think about the relationship as a level shift or a multiple effect (your ratio argument) and could see good cases for both. I also worried that my Friday post could have been wrong, and I wanted to put up an update fixing things if it were. So, I asked the data to tell me, as I always do when I haven’t a strong theoretical reason to reckon it has to be one or the other. I took a very simple statistical technique that lets me see what the relationship between the youth and adult minimum wage looks like over time. The correct relationship (according to the data) is a mix between a level shift and a multiple. The very simple model predicts the youth unemployment rate very well, from 1986 through early 2008. There’s noise in the model, but the model is never more than a couple of points above or below the actual youth unemployment rate.

    I took the very least restrictive model possible. The OLS model was simply:

    youth unemployment rate = constant term + (coefficient)*adult unemployment rate.

    Ordinary Least Squares regression tries to fit a line through the data that minimizes the sum of squared differences between the line and the observations. It’s a bog-standard statistical technique: the baseline model most folks will fit for most things. And, I didn’t complicate it up by throwing in a whole kitchen sink of things either. I wanted the simplest model possible to see what was going on in the error terms (the residual).

    If the relationship between the two, from ’86 to present, were just the ratio, the constant term would have come up insignificant and the coefficient would have been large and significant. Instead, it’s a mix of the two. Both the constant and the coefficient matter. I thought going in that it would mostly be the constant, but the data told me it was the mix; I go with what the data tells me. The best fit simple model combines a constant term and a multiple of the adult unemployment rate. And that makes sense. Suppose that the economy is running full bore as hard as it can go: everyone who ever wanted a job has one, and the only unemployment we have is frictional (the amount of time it takes to find a new job after quitting an old one or after joining the labour market for the first time). We’d expect in that case that youth unemployment will be whatever the natural frictional rate is: say 12% (it takes time for new young workers to get their first job). And adult unemployment will be whatever its natural frictional rate is: say 3%. When all is going full bore, youth unemployment will just be 9 points higher than the adult rate. If the economy starts to slow down, adult unemployment will rise but youth unemployment will rise by more. That part of the relationship is multiplicative. The very simple statistical model lets both of these happen when tracing out the relationship.

    Starting late in 2008, the actual youth unemployment rate spikes up dramatically compared with the simple model’s predicted unemployment rate. Is it the recession? Well, youth unemployment diverged from the model’s prediction in the last recession as well, but only by a couple of points. This time, it’s ten points higher than the simple model would predict. So the increase in youth unemployment this time around is five times greater than we’d have expected. Something’s caused that. Is it changing demographics: more folks in the 15-19 year old cohort? Well, from ’86 to ’96 the number of people in that age group was dropping while the residual (the amount by which actual youth unemployment exceeds the model’s prediction) was increasing; from 2001-2011, the number of people in that group has been increasing. I can’t think of any obvious reason why a drop in the number of youths would increase youth unemployment from ’86 to ’96 but an increase in the number of youths would massively increase youth unemployment from 2001 onwards with a big spike in 2008.

    There’s lots of stuff that could be wrong in my regression: it’s a very simple model and there’s lots of stuff for which I’ve not corrected. But that residual plot looks like a smoking gun to me, and to all the other economists here at Canterbury that I’ve shown it to.

    Sincerely,
    Dr. Eric Crampton
    Senior Lecturer
    Economics
    University of Canterbury

    • Bright Red 5.1

      oh lolz, and then there’s this:

      “If the relationship between the two, from ‘86 to present, were just the ratio, the constant term would have come up insignificant and the coefficient would have been large and significant. Instead, it’s a mix of the two. Both the constant and the coefficient matter.”

      The strawman here is the assumption that the ratio is constant over time. If you assume a constant ratio then you can pin anything inconstant on the coefficient. But clearly, if you look at the graph above there are long-term demographic and economic factors moving the ratio over time. Which invalidates your constant ratio assumption.

    • BLiP 5.2

      blah blah blah . . . there is no consensus . . . blah blah blah

  6. Bright Red 6

    libertarians = politics’s clowns

    You know how the clown has that serious look on his face, maybe he’s trying to stack a bunch of plates and he seems to have a logical method, then he slips on a banana skin and goes crashing over, knocking everything down? That’s libertarianism in a nutshell, it comes with its own banana skins

    Eric, why would an increased number of youths increase the youth unemployment rate? Surely you’re not confusing absolute numbers and rates?

    • Bright Red 6.1

      And I’m really really looking forward to you explaining the jump in Maori unemployment.

      Pray tell, was that the minimum wage too?

      In fact, other than a chronological coincidence based in dodgy statistical practice, have you any evidence that the end of youth rate caused the increase in youth unemployment?

      For example, are there any reports or studies or even anecdotes of 15-19 year olds (most of whom would be outside the youth minimumw age age anyway) saying they can’t get work because employers say they’re too expensive? Or any employers saying $12.50 is too much for a 16 year old so they’re firing them?

  7. StephenR 7

    He indicated that he might not hang around BrightRed. There is a dialogue at his blog though.

    Out of curiosity what, to you, would constitute proof that equal wages were harmful to youth employment? Just hypothetically, not necessarily using the stats provided.

    • Bright Red 7.1

      proof can come in many forms but to link a specific policy change to a specific supposed outcome I would want to see

      a) the outcome following the change, which is what Marty is disputing – ie. Marty shows there has been no relative increase in 15-19 year olds unemployment
      b) explanation for other similar outcomes that cannot have arisen from this change (eg. the increase in Maori unemployment) because if no explanation is forthcoming it suggests a different common cause
      c) an explanation of the mechanism of the causation. How did the policy cause the outcome? Preferably with some real world, rather than purely theoretical, evidence – like employers saying they had to fire 16 and 17 year olds because they were too expensive.

  8. Samuel Konkin 8

    Hi,

    Just to clarify, what is the graph showing? As in, what is the meaning of “Ratio of % unemployment for 15-19 year olds”?

    • Draco T Bastard 8.1

      It’s showing that the ratio of unemployment between age groups remained the same after the minimum wage was applied to under 18s.

  9. Samuel Konkin 9

    No, I mean, the literal figure, how is it calculated?

    • Bright Red 9.1

      Do you mean ‘what’s a ratio’?

      Take the unemployment rate of 15-19 year olds and divide it by the unemployment rate of the other age group that will leave you with a number: the ratio of 15-19 unemployment to the other group.

      So, if 15-19 unemployment was 10% and 20-24 unemployment was 5% the ratio would be 2.0

      the unemployment rates you can get from Stats NZ.

      I reckon that making a graph like the above would take about 10 minutes, there’s no trickery to it.

  10. Paying a person doing the same work as another person less money because of their sex or religion or ethnicity or any other grounds prohibited under the Human Rights Act is illegal and abhorrent. Yet, the Right wants to do just that with a private member’s bill from Roger Douglas reintroducing a lower minimum wage for 16- and 17-year-olds (Age is a prohibited ground for discrimination when the person is over 16, HRA s21(1)(i)).

    You’re missing one very important caveat:

    While s 22(1)(b) of the Human Rights Act makes it illegal to pay someone less than someone else for the same work because of “their sex or religion or ethnicity” or based on any of the other grounds in section 21(1) – which you reference, and which includes age – you’ve ignored section 30(2):

    Nothing in section 22(1)(b) of this Act shall prevent payment of a person at a lower rate than another person employed in the same or substantially similar circumstances where the lower rate is paid on the basis that the first-mentioned person has not attained a particular age, not exceeding 20 years of age.

    If you’re paying someone 16, or 17, or 18, or 19 less money for the same work, based solely on their age, it’s not illegal now, and Roger Douglas’s bill doesn’t change this one way or the other.

Links to post

The server will be getting hardware changes this evening starting at 10pm NZDT.
The site will be off line for some hours.