January employment increase makes new Australia record

Australia posted a 16k increase in employment in January 2018 which records the longest streak of job gains with a 16th consecutive positive print. The month’s increase was very close to the market and Westpac’s expectation for a 15k rise.

However, the composition of employment changes was quite stark with full-time employment down 49.8k and part-time employment up 65.9k. That comes after a strong 2017 for full-time employment growth and the annual growth rate is still an elevated 3.6%.

In the month, hours worked fell 1.4%. The annual rate of change moderated to +0.5% from +3.2% but the series is volatile and this month’s movements were largely due to base effects. If hours worked holds steady in February, the annual pace will lift to +2.1%.

By state, Queensland posted the highest jobs growth, up 19.7k, maintaining solid momentum but January's lift was driven by gains in part-time employment. On the other hand, a soft month was recorded in NSW (-21.1k) and WA (-8.9k), both coming after some solid gains in December.

Annually, jobs growth strength has been fairly broad-based across the states, with the mining states recovering and the south-east expanding. Yet, we will watch the pace of jobs growth in Victoria closely with total employment flattening out over recent months after a very strong run up.

Annual employment growth held steady at a brisk pace of 3.3%. Our Westpac Jobs index is still at a high level but the change in the index has moderated in recent months. Our outlook for employment growth in 2018 is positive with a forecast 2.2% annual rate an above trend pace albeit slower than 2017’s very strong performance. 

Overall, while January’s update continues the run of positive employment numbers, key questions remain for the labour market.

Firstly, there is uncertainty on how the unemployment rate will change given our solid jobs growth forecast.

In January, the unemployment rate fell slightly to 5.5% from 5.6% in December (revised up from 5.5%). This came with a decline in the participation rate to 65.6% from 65.7%.  Looking back over the past year, while the unemployment rate did fall from a peak of 5.9% in February 2017 to the current 5.5%, the drop was more muted than the gain in employment due to a rising participation rate. To examine how far the unemployment rate could continue to fall, we need to assess how far the participation rate could rise.

The first thing to note is that the uplift in the participation rate has been driven mainly by females joining, re-joining, or remaining for longer in the labour force. However it should be noted that this spike occurred alongside a rising male participation rate which had otherwise been in a downtrend. That compares to 2015’s spike in female participation, where the male participation rate held its level rather than increasing. This suggests that the most recent lift in participation rate has a cyclical aspect to it with stronger employment generally attracting more people to the labour force. In addition, subdued wages growth, high household debt and high house prices may be necessitating dual-income households. However, given the outsized gain in female’s participation there is also likely a structural gender dynamic occurring as well.

Converging male and female participation rates have been a long-term trend due to the large social change in women joining the workforce. Yet, the rate of convergence had been flattening out as the generational shift filtered its way through the labour market over time. Between the periods of 2009 and 2014, the female participation rate had remained relatively unchanged at around 58-59%. Perhaps this means that in aggregate, the recent increase in female participation is due to structural reasons outside of the generational shift in behaviour. A look at the type of jobs created over the past year suggests a changing composition of employment may have been a significant factor.

The graphs below show the annual jobs increase in the year to November 2017 (lhs) across different economic sectors broken down by gender. It is ordered by the female share of employment in the individual sector (rhs) with the top graph showing the sectors with the highest proportion of female employees, and the bottom graph the lowest. 

The striking aspect about the top chart is that the four sectors with the largest proportion of female jobs (health care, education, accommodation & food and retail trade) accounted for 242k of the total 387k jobs created over the year to November 2017 and 202k of those new employees in the sectors were female. That contrasts with the bottom chart where the only standout sector is the construction industry which has a very large male proportion and men accounted for 109k of the net 101k increase.

This suggests that the rise in female participation is a reaction to strong employment growth in sectors that typically have a higher proportion of female employment. These service sectors have been key to the transitioning economy and therefore strength is likely to continue. On the other hand, the construction sector accounted for 109k of the total 178k annual increase in male jobs. While the level of homebuilding remains at a high level, supply-demand factors look more balanced and slowing house price growth should restrain housing investment. The uncertainty here is whether similar jobs with transferrable skills will be created in other sectors.

Given the large transitional shift in the types of jobs created, we will likely see tightness in the labour market in some pockets before others. That leads onto the second question surrounding the unemployment rate in regards to where the NAIRU sits – the rate that we start to see wage pressures come through.

The RBA’s estimate of the NAIRU is around 5%. Yet, even though the unemployment rate has been approaching 5% over the past year, we are yet to see a reaction in wages. In a speech on Tuesday at the ABE Forecasting Conference, Assistant Governor Economic Ellis mentioned that “we are mindful that, as we approach that figure, there's a risk we find there is more room to come down before wage growth picks up in earnest”.

A possible reason for a lower NAIRU is that there is more slack in the labour market than that indicated by the unemployment rate alone. Indeed, the underemployment rate has held a high level with the last seasonally adjusted estimate at 8.3% in November.

It is difficult to draw a conclusion on what a high underemployment rate means for overall slack in the labour market. The ABS provides quarterly non-seasonally adjusted data on volume measures of unemployment and underemployment based on the hours of available labour. The data is consistent with the previous headcount measures in terms of directional movement. In terms of quantity, as at November, it shows underemployment at 3.1% and therefore a sizable proportion of the combined 7.2% of underutilised hours. But the readiness of the underemployed pool of labour to be drawn upon is uncertain given that it may require changing jobs rather than going from not employed to employed.

As such, finding the level of NAIRU may only be possible in hindsight - after we start to see wage pressures emerge. With that in mind, next Wednesday’s Wage Price Index data is all the more critical.