Is unemployment really 6.4%? II

So yesterday Steve asked the question about our unemployment rate, “Is it really 5.3% or is it something closer to 6.4%”? He made the important point that unemployment by itself isn’t as informative as we’d like to think. The participation rate is important information too. So what is the unemployment rate given the participation rate?

So what Steve did was add back the change in participation rate to the published unemployment rate to get an estimate of a participation-adjusted unemployment rate. But that assumes that everyone at the higher participation rate would be unemployed. That isn’t necessarily a bad assumption if you think that the participation rate has declined entirely because people have become discouraged and dropped out of the labour market. That may well happen from time to time. But in general it is too strong an assumption. Part of the change in the participation rate must to attributed to changes in the unemployment rate and part of it to employment.

So first thing I downloaded the seasonally adjusted unemployment rate and seasonally adjusted participation rate from the ABS and graphed them.

Over that period the participation rate has slowly trended up from about 60% to about 65%. Similarly, unemployment has trended down from the early 90s with an untick during the GFC. I don’t expect unemployment to return to pre-GFC levels due to the current labour market regulations – but that is another story.

So then I do something really naughty – I run an OLS regression with the unemployment rate as the dependent variable and the participation rate as the independent variable. Naughty because that regression suffers from serial correlation and the variables themselves have unit-roots. I’m just interested in the fitted values. This month (actually November 2012) the seasonally adjusted unemployment rate was reported to be 5.2% and the participation rate was reported as being 65.1%. The fitted value for November 2012 is 5.9%. So not 6.4%, but somewhat higher than 5.2%.

Econometricians will be horrified by the regression and others can argue that the various changes to the IR regime over the period 1978 – 2012 means that there might not be a stable relationship between unemployment and participation and so on.

Update 1: GrantB asks the question: “Would a plot of both normalised to their Feb 1978 values show the trends better?” Here it is.

Update 2: Grey trolls, “If you actually had an confidence in it having any meaning at all you would show the line of the “true” unemployment rates since 1978 so we could have a good laugh.”

Not a bad story there – actual unemployment rises above participation-adjusted unemployment during recessions. Got driven very low during the Howard era and then reverts in 2008 (GFC effect? or Fair Work?).

Update 3: Seems our friend Grey can’t tell the difference between fitted values and residual values. So I’m posting a picture of the residuals.

This entry was posted in Uncategorized. Bookmark the permalink.

21 Responses to Is unemployment really 6.4%? II

  1. Grant B

    Would a plot of both normalised to their Feb 1978 values show the trends better?

  2. Grey

    There are so many reasons why this is daft.
    If you actually had an confidence in it having any meaning at all you would show the line of the “true” unemployment rates since 1978 so we could have a good laugh.

  3. Grey

    Hmm, I must be using a different data set from you because the scatter plot looks rubbish for getting a linear regression equation and an r squared value of 0.25 indicates that.

  4. Sinclair Davidson

    0.247 on my calculation.

    When I add in a quadratic term the adj-R2 rises to 0.366, the fitted value is 5.4 – but I can’t think of a good reason to add the quadratic.

  5. Skuter

    Sinc, it is not necessarily daft. Are the two series cointegrated? If so the long run relationship could reasonably be described by such an equation. Also, you mentioned serial correlation. There is good reason to think that the discouraged worker effect operates with a lag

  6. Grey

    Very nice {update 2], Paul Keating would be delighted, it turns out it was the recession we didn’t have to have – and in fact didn’t have.

    You are far too smart not to know this is complete numerical non-meaning. And you call me a troll!

  7. Alice

    Just heard from a friend that the ambulance service is no longer “training its own” and is sending all paramedics to uni first (at great cost to the young person). Along with the police force which no longer “trains its own”…along with nursing which no longer “trains its own” on the job…along with?

    and we fucking wonder why youth unemployment is 20%??
    Check in in five years – it will be higher.

    Doesnt this stupid government realise not all of our youth can afford to be students on campus for three years notwithstanding hecs?

    Im pissed off about this. This doesnt make good sense to me so why are they trying to shuffle every kid to uni. What a bad outcome is this for the youth of our country (trying to be trained on the job not sent off to waste their time and money on campus??)

    Who or what idiot is thinking up these pathways for our young?

  8. Sinclair Davidson

    Grey – what do you mean? Non-meaning is relative to what you’re looking for. From your comments you’re looking for a model that explains unemployment. This post does no such thing. It has the very (very, very) limited goal of producing a participation-adjusted unemployment rate. Now that figure may be of little interest to the world, as you suggest – but that is another story. (I can imagine a variable like this being thrown into an asset-pricing model – in the 80s and 90s they often were.)

    In the meantime looking at the picture in update 2 we can see a sluggish job market recovery after the recession “we had to have” with actual unemployment much higher than participation-adjusted unemployment. Of course we all knew that happened anyway (but as my old PhD supervisor used to say “It’s good when you prove the obvious”). To my mind more evidence of how terrible Keating actually was – although now there is a lot of nostalgia for those days.

  9. Alice

    No its not Fair work that has done it.

  10. Alice

    Its GFC obviously (come on a couple of regulations versus the GFC??) Of course its the GFC.
    As for Keating..the know it all who set us on the road to ruin? The big lets all get deregulated king? The big lets all go global king? The big lets all do super king?
    That man is responsible for more damage than most people think but Howard didnt undo it either.

    No-one did.

  11. Grey

    What nonsense, a proper measure would accentuate unemployment during recessions as people drop out of the market and reduce unemployment during the boom times as people flood back into the market.

    What you have done, purposefully or not, is provide an excellent topological profile of Belgium – only not quite as hilly.

  12. Sinclair Davidson

    a proper measure would accentuate unemployment during recessions as people drop out of the market and reduce unemployment during the boom times as people flood back into the market.

    hmmmm – you do know the difference between a fitted value and a residual? The residuals follow the very pattern you have described. Let’s keep your humiliating lack of knowledge between ourselves and not tell the others – they might mock you unmercifully.

  13. Grey

    No the residuals do not follow the pattern I describe because a linear regression is not the statistical tool to capture what the original bloke was trying to describe.
    The reason your fitted line comes up with meaningless values is because the R2 value is 0.24 – which is the common outcome of scatter plot that looks like a pair of bunny ears on its side. When you get a scatter plot shaped like a pair of bunny ears it is generally not a good idea to try and fit a line to it.

  14. Sinclair Davidson

    Oh no! The participation rate only explains 24% of the variation in the unemployment rate. Well duh! How much did you think it would explain?

    If you want a model of the unemployment rate you’d have to throw in things like lagged GDP or other measure of economic activity, then business expectations etc.

  15. Grey

    “The participation rate only explains 24% of the variation in the unemployment rate”
    ” Let’s keep your humiliating lack of knowledge between ourselves and not tell the others – they might mock you unmercifully.”

  16. Sinclair Davidson

    You wasting my time Grey. Do you know anything about regressions or have you just passed an UG course?

  17. Skuter

    Sinc, upon further reflection, why are you regressing the unemployment rate on the participation rate? Shouldn’t it be the other way around, i.e. changes in the unemployment rate cause changes in the participation rate through the (offsetting) discouraged and added worker effects? I am struggling to see the logic behind the original exercise.

  18. Sinclair Davidson

    Skuter – I’m trying to drag out an estimate of the unemployment rate adjusting for participation. If I wanted to argue that people dropping out of the labour market drove unemployment then I’d do what you suggested. You are wanting to test a theory – I just want to create a variable to use for some other purpose. I think that is what is troubling Grey too.

    So for example, I could imagine this variable (actually the residuals) being labelled “excess unemployment” and being thrown into an asset-pricing model or used in an APT test, or a public choice paper and so on.

  19. Grey

    ” I am struggling to see the logic behind the original exercise.”

    No kidding.
    Actually if I was trying to do what Steve originally was asking – and I had the chutzpah to decide that the ABS being composed of professional statisticians hadn’t already generated the appropriate criteria for the job – then it would be something like this.
    Detrend the participation rate (since there is a social phenomena driving greater participation since 1978) then estimate a baseline of maximal participation – say the lower 95 % percentile of the detrended participation rate and report that as an adjustment to the unemployment rate.

    Would it be a very good measure? No, because discouragement is not the only thing driving the participation rate, for example recessions will probably add people who had taken early/semi retirement or people who in good times were happy to rely on their partner’s incomes.

    I personally think this regression exercise is without value completely, but if it has value then it isn’t for what Steve was suggesting.

  20. Sinclair Davidson

    Detrend the participation rate (since there is a social phenomena driving greater participation since 1978) then estimate a baseline of maximal participation – say the lower 95 % percentile of the detrended participation rate and report that as an adjustment to the unemployment rate.

    Well do it and I’ll post it up.

    If the ABS has estimated a participation adjusted unemployment rate – as you suggest – can you provide a link to where they have published it? I can’t find it on their site.

  21. John A

    Alice, you said on 28 Dec:

    “Its GFC obviously (come on a couple of regulations versus the GFC??) Of course its the GFC.
    As for Keating..the know it all who set us on the road to ruin? The big lets all get deregulated king? The big lets all go global king? The big lets all do super king?
    That man is responsible for more damage than most people think but Howard didnt undo it either.

    No-one did.”

    Are you not conflating some opposed forces there? Deregulation (which was actually an all-party policy shift) versus fiddling with superannuation?

    In terms of general principles, what is so damaging about less regulation, more competition, and the international discipline of a floating exchange rate?

Comments are closed.