Bill Mitchell: immigration collapse reduced Australia’s unemployment by 2.7%

Last month, several Australian economists tried to argue that Australia’s plummeting jobless rate has been driven by stimulus and has little to do with the closure of Australia’s international border to immigration. Their arguments were ruthlessly destroyed by me and others (here, here here).

Now Bill Mitchell – Professor in Economics and Director of the Centre of Full Employment and Equity (CofFEE) at the University of Newcastle – has published extensive analysis showing that the collapse in immigration has reduced Australia’s unemployment rate by 2.7% versus what it would have been had immigration remained at it pre- pandemic level:

Essentially, the capacity of local firms to access cheap foreign labour attenuates against upward pressure on wages, in both specific segments of the labour market and more generally, and also undermines the necessity for firms to offer on-the-job training to ensure local workers are equipped with the capacities to undertake the vacant job opportunities…

The following graph shows Australia’s working age population (Over 15 year olds) from January 2015 to January 2022.

The red (dotted) line extrapolates the working age population from February 2020, based on the average monthly growth over the period January 2015 and February 2020.

It can be interpreted as showing what the working age population would have been had the pandemic not intervened.

The current gap (as at January 2022) between the actual working age population and the extrapolated aggregate is 556 thousand workers. That is how many extra people of working age there would have been had the pandemic not intervened.

The flattening of the population growth has forced employers to work harder to get workers as economy activity increases after the lockdowns and other restrictions.

It is one of the reasons that employers are lobbying the federal government hard to make it easier for workers on foreign visas to come back – they know as well as I do that this will weaken the bargaining power of local workers, particularly the unemployed and underemployed.

We could say that based on current employment growth and the most recent peak participation rate, the unemployment rate would be 6.9 per cent if the working age population had have followed the extrapolated trajectory, rather than the current official rate of 4.2 per cent.

That is, the estimate of the impact of the border closures…

The next graph shows the labour force from January 2015 to January 2022. The labour force is calculated by the statisticians as the working age population times the participation rate.

So it is the number of people from the working age population that are either employed (that is, working more than one hour per week) or unemployed (available and actively seeking work).

The red (dotted) segment is the extrapolation based on the estimated working age population in the first graph times the peak participation rate (66.3 per cent recorded in March 2021).

It shows what the likely labour force trajectory would have been had the pandemic not reduced the growth of the working age population.

The following points can be made:

1. The claim by government that the economy has recovered do not reflect the reality that the labour market is now smaller than it would have been had the pandemic not intervened.

2. The gap between the current labour force and what it would have been if the pandemic had not intervened is 394.9 thousand persons. This is higher than the gap in the actual working age population and the extrapolated aggregate because the participation rate is now higher.

3. Total employment is 259.6 thousand more than it was at the start of the pandemic (February 2020) while the working age population has only risen by 107.6 thousand. And, almost all of those extra jobs have gone to local workers.

Employment trends

The question then is what would employment growth had to have been to simultaneously achieve the current 4.2 per cent unemployment rate, in an environment represented by the extrapolated working age population – that is, without the pandemic-induced border closures?

The next graph shows the actual trajectory of employment from January 2015 to January 2022. The two major Covid waves post-February 2020 are shown.

The red dotted line post-February 2020 is calculated by assuming the working age population followed the extrapolated path as in the first graph above, and the participation rate was 66.3 per cent (the March 2021 peak – noting again that in January 2022 it was 66.2 per cent), and sustaining a 4.2 per cent unemployment rate.

The green dotted line (pre-pandemic) shows what employment would have had to have been to sustain a 4.2 per cent unemployment rate. It is based on the actual working age population and a participation rate of 66.3 per cent.

So, it shows how far the required employment growth had been pre-pandemic if a goal of a 4.2 per cent unemployment rate was in place.

The following points can be made:

1. The gap between recorded employment in January 2022, which delivered a 4.2 per cent unemployment rate, and what the employment would have had to have been if the working age population growth had not been flattened by the border closures to achieve that same unemployment rate is 377.3 thousand jobs.

2. The actual growth in employment between February 2020 and January 2022 was 2 per cent.

3. The growth that would have been required over the same period to achieve a 4.2 per cent unemployment rate if the working age population had have grown along the extrapolated path (and participation was at 66.3 per cent) would have been 4.9 per cent or an average of 0.21 per cent per month.

4. Between March 1978 and February 2020, the average monthly employment growth has been 0.15 per cent per month.

5. Without the border closure, a 4.2 per cent unemployment rate would have been very difficult to achieve given historical average employment growth trends.

The ‘What-if’ unemployment analysis

The final graph shows the evolution of the actual unemployment rate since January 1980 to January 2022 and the dotted line is the ‘What-if’ rate, which is calculated by assuming the peak participation rate, the extrapolated working age population and the actual employment since February 2020.

It shows what the unemployment rate would have been given the actual employment growth had the working age population trajectory followed the past trends.

So instead of an unemployment rate of 4.2 per cent, the rate would have been 6.9 per cent in January 2022, given the employment performance since the pandemic.

This finding puts a rather different slant to what has been happening since the onset of the pandemic.

But it also shows that the pandemic restrictions on external borders has allowed the unemployment rate to fall much more quickly than in past downturns…

This sort of analysis is rather tedious and time-consuming.

But it helps to substantiate the conclusions that the external border closures during the pandemic have benefitted local workers and the drive to increase the population growth again through schemes to attract cheap labour will worsen the outlook for local workers.

Professor Bill Mitchell has come to identical conclusions to me, Gareth Hutchens, Saul Eslake and Greg Jericho, who have each shown that Australia’s plummeting jobless rate has been driven by the closure of Australia’s international border to immigration.

Any economist claiming otherwise is either lying or stupid. The proof is in the data and it is undeniable.