Question: why has employment growth collapsed in the first half of the year after recent claims by the Government that 60,000 jobs per year were being created?
The answer lies in statistical misunderstanding, Government spin and the failure of many commentators to read the numbers correctly. For the fact is that the 60,000 job-creation number was never real and the recovery in the labour market is sluggish at best. This post may get a bit involved but stay with me – for this is as much a story about how the recovery is being contrived as it is about bald numbers.
Last year, employment growth suddenly took off. In 2012 employment actually fell by 11,000 – and this was after a loss of nearly 300,000 since the start of the crisis. However, in 2013 everything changed. Employment grew on a full-year basis by 43,000 (this is consistent with claims by the Government who were using quarter-to-quarter figures).
This was quite a turnaround. The Government claimed their policies were working. For many commentators this was proof that recovery had returned. But there were a couple of problems.
- First, this employment growth took place while the economy remained in a domestic demand recession. Given that employment is sensitive to domestic demand, this didn’t make sense.
- Second, the usual pattern of an economy coming out of a recession is that employment growth lags. This is because if there increases in business output, the first beneficiaries are those already in employment; they get an increase in hours which had previously been cut.
- Third, the actual job numbers were throwing up some strange happenings. Self-employment (own-account workers) grew by over 10 percent and made up over half of the total employment growth. At one stage, self-employment was growing by nearly four times the rate of growth during the boom. This didn’t make sense – not with domestic demand stagnation. Agriculture employment showed a similar pattern.
These concerns were dismissed. Government policies were working and critics were just nit-picking. However, the CSO published warnings throughout all last year – warning people against interpreting growth trends. Why? Because they were re-aligning their sample base with the recent Census (don’t forget, the Quarterly National Household Survey is not a comprehensive head-count, just a sample; like a poll). This happens after every Census.
Did this account for the jump in employment? Was the growth statistical? A number of commentators said no; that the overall number was correct (a top-down survey); it was just the different categories that were being changed – such as manufacturing employment, or self-employment, or retail. However, the CSO constantly changes the overall number of people in work to keep it consistent with the Census – sometimes revising employment numbers in previous years (see here for a detailed CSO note which revised overall employment numbers going back to 2006).
This revision occurred incrementally from quarter 4 2012 to quarter 4 2013. So what happened to job growth during these quarters?
During the period of the revision, employment growth shot up. Before the revision period, employment was falling. After the revision employment growth was muted. This is strong evidence that employment growth in 2013 was inflated by the CSO’s statistical re-alignment but not conclusive. More evidence is needed – and this comes with a more stark comparison.
In the first half of 2013 employment grew by over 30,000 – during the period of the revision. In the first half of this year, employment growth was almost statistically insignificant.
There are some other indicators. For instance, Forfas reported that in 2013 employment growth in the internationally traded sector ( export sector) increased by only 3,400 above the 2012 increase while among IDA supported companies, the rate of employment increase in 2013 fell below the 2012 level marginally. So these sectors could hardly explain the issue.
And public sector employment didn’t contribute to any growth – it fell by over 6,000.
So did employment actually grow in 2013? Probably. But we are somewhat in the dark to establish where it happened. Let’s use the headline figures, leave aside the magnitude of growth, and see what sectors delivered the largest proportion of the growth.
What we find is that nearly half of the employment growth occurred in the lowest-paid sector of the economy – with the next largest being dominated by multi-nationals. If we can take these headline numbers at face value (and that’s a big if) then we find employment growth that is narrowly based and disproportionately located in the hospitality sector.
And what’s happened to all sectors in the first half of the year – a more reliable figure since it is post CSO revision?
The fastest growing sectors are construction and the public sector, sectors of public administration and education which, again, doesn’t indicate a broad-based growth. It is worth noting that the majority of sectors suffered a decline. Two notes: these are seasonally adjusted and exclude the not-stated category – so the total of the above sectors don’t add up to the total. And the agriculture sector – which experienced the biggest volatility during the CSO revision – may retain its volatility for a while.
So what are we left with?
- The 2013 employment growth was inflated by the CSO revision – though to what extent we can’t tell
- Growth in the first half of 2014 flat-lined
- What growth there is in the economy is not broad-based. In 2013 the low-paid hospitality sector could have made up 50 percent of the net gains while in the first half of this year most sectors declined
- The claims by the Government that 40,000 or 50,000 or 60,000 jobs were created are not substantiated. And commentators’ claims that the labour market has turned the corner should be put on hold for a little longer
Dan O’Brien was the only commentator that I have come across so far that has taken a sceptical view of the jobs figures:
‘Consider this: if the rate of employment growth in the first half of the year doesn’t strengthen, it will take until 2037 – another 23 years – to get the numbers at work in the Irish economy back to 2007 levels.’
A bit cheeky and hopefully this won’t be the case. But it is welcome tonic to the propaganda and misunderstanding and plain confusion that have been flying about for the last year.
The bottom line is that the Irish economy remains mired in stagnation. The CSO employment figures released yesterday confirm that.