A Quick Analysis of the Unemployment Numbers

I am no statistician. Ask anyone who knows me. Statistics was the only paper I struggled with when studying accountancy. But when I heard yesterday that the unemployment figures for the June 2020 quarter had actually fallen to 4%, I decided that the old adage is true. There are most definitely lies, damned lies and statistics.

How, you may ask, when we are into the worst recession for 160 years can unemployment have actually fallen? Is it possible that the numbers have been fudged somehow, to make the government look as if it is doing better with the economy than it really is? The answer is no… well, probably no. The devil as they say, is in the detail, and the details don’t really matter when we aren’t that worried about unemployment. But now that we expect the numbers to be much worse than they (apparently) are, maybe it is time to take a look at how these things are reported, and see if the truth matches expectations in any feasible way.

I had done some basic numbers myself. The number of unemployed for the March quarter was 116,000, which represented 4.2% of the working population. Paul Goldsmith, National’s finance spokesman, has recently stated that there are now 212,000 people unemployed. That, to me, represents approximately 7% unemployment… not 4%. But here is the first thing to think about. The number of unemployed is an average for the quarter, so 116,000 represented the average number during the March quarter – not the number at the end. That came as a surprise. So, the June quarter numbers are an average also, which means that, again, these are not the numbers at the end of the quarter. If the numbers spiked significantly in the last 2 weeks of the quarter, this would be flattened out by taking an average over the whole quarter.

Remember also that the June quarter included the entire lockdown period, where many employees were being propped up by wage subsidies, the first tranche of which didn’t end until early June. So you can see why the average might be distorted. Job losses would not really have got going until mid to late June, and the results are averaged out.

Remember also that this is the quarter ended 30 June 2020, and we are already into August. All layoffs since then are not included. The numbers are not particularly up to date in a fast moving economy – fast moving backwards, that is.

Now for another little twist. To be considered ‘unemployed’, a person has to be actively looking for work. How many people were doing that during lockdown, when most of us couldn’t actually go to work anyway? When the country was in lockdown, the unemployment figures actually fell to 3%, presumably because people could not actively search for work, but by the end of the quarter, the figure had risen to 5% – hence the 4% average.

So the lockdown itself has significantly distorted the figures.

Now, instead of just counting those who are actively seeking work, we need to look at the ‘underutilisation’ rate. This is much more meaningful. The underutilisation rate is a broader measure of spare capacity in the labour market,  including hours, worked. It includes both those who are out of work and those who would like to work more hours. For the June quarter, the underutilisation rate climbed from 10.4% to 12%, the biggest quarterly increase since 2004. As a result of businesses cutting people’s hours, the number of underemployed people rose by 33,000, and the number of potential job seekers rose by 18,900.

Now here is another twist. The number of people not considered in the labour force – people such as retirees and students – increased by 37,000 in the June quarter. I don’t have a full explanation for this, but it will include people opting to retire and people who do not qualify for jobseeker payments because of excess assets or redundancy payments. These people may have adopted a ‘wait and see’ approach during the lockdown, and fallen out of the employment statistics. After all, if you were laid off during lockdown, you could not be counted as actively seeking work, as no one was hiring at that stage.

When it is all added up, the drop in the number of people in the labour force fell to the largest amount since the global financial crisis.

The number of hours actually worked during the June quarter also fell, by approximately 10% – the largest decrease since records began.

We must not forget the effect of the wage subsidy on employment numbers. Employers were effectively bound to keep workers employed if they had accepted a wage subsidy on their behalf. This is another twist in the figures which, along with lockdown, grossly distorts the unemployment figures for the June quarter.

So let us not be fooled by the numbers. Unemployment is rising catastrophically and will only get worse in the next few weeks when the wage subsidy ends. I am standing by my estimate of 7% unemployment as of today, and most of the banks were expecting a June quarter figure of around 5.9%. There will be no more numbers before the election though, so the government can point to these statistics whenever they are asked about unemployment. As I said before, there is nothing like lies, damned lies and statistics.

If you enjoyed this BFD article please consider sharing it with your friends.