# Bulletin – June 2022 Australian Economy Job Mobility in Australia during the COVID-19 Pandemic

## Abstract

The COVID-19 pandemic has led to large disruptions to the Australian labour market. Initially, workers were less likely to change jobs because of the uncertain economic environment, the decrease in advertised jobs and the JobKeeper program that helped workers remain attached to their employers. More recently, job mobility has increased as workers have caught up on planned job changes or been encouraged by the strong labour market to change jobs, particularly in high-skilled roles experiencing strong labour demand. This article reviews developments in job mobility in Australia through the pandemic, and compares these outcomes to other advanced economies. It also examines the potential implications for wages; a high rate of job mobility tends to be associated with higher wages growth in a tight labour market, as employers in sectors with high demand for labour compete for new staff or raise wages to retain staff.

## Introduction

Job mobility – the movement of workers between jobs – underpins the efficient operation of the labour market by matching people with jobs that better fit their preferences and skills. People switch jobs for a number of reasons. These can include higher wages, improved working conditions, different hours, workplace flexibility, job satisfaction or job security. To the extent that job mobility results in better job matching and increased labour reallocation to more productive firms, it can play a role in labour productivity gains. Job mobility is also a key feature of labour market flexibility that helps the economy adjust to economic shocks and structural change, which can affect the number and types of jobs available.

Prior to the pandemic, the Australian labour market had experienced a downward trend in job mobility for a number of decades, and job mobility was around historically low levels (Graph 1). While low job mobility is not necessarily problematic – some job changes are not voluntary and there can be benefits from longer job tenure – the decline in job mobility may have been associated with a general decline in business dynamism (Ellis 2021). In particular, it coincided with a slower reallocation of labour to productive firms, which contributed to lower productivity growth (Andrews and Hansell 2019). The rate of job mobility tends to move with the business cycle, with sharp declines experienced in the recessions of the early 1980s and 1990s and following the global financial crisis (GFC). The step down in mobility following the latter two downturns suggests that structural factors are also usually at play.

The onset of the COVID-19 pandemic led to large disruptions to the Australian labour market that corresponded with an initial decline in economic activity amid restrictions to contain the health effects of the virus. A significant number of people were stood down from their jobs or faced reduced hours during this period. There was also a sharp fall in job mobility, reflecting a reluctance to change jobs, fewer opportunities to switch jobs and the effect of JobKeeper in keeping workers linked to their employers. As economic activity rebounded in late 2020, so too did the labour market. As of early 2022, job mobility appears to be around its highest level in over a decade, returning to levels last seen prior to the GFC. The increase has been underpinned by the strength of the labour market, especially in particular sectors, and a backlog of job changes that had been deferred.

This article reviews developments in job mobility and turnover in Australia since the onset of the pandemic, and compares these outcomes to other advanced economies. It also examines the role of job mobility as a potential factor influencing wage outcomes, which can have implications for monetary policy.

## Measuring job mobility and turnover

Aggregate measures of job mobility and job turnover in Australia are published by the Australian Bureau of Statistics (ABS) on an annual basis as a supplement to the monthly Labour Force Survey (LFS) and are available from 1972. To examine job mobility and turnover on a timely basis through the pandemic, we made use of several quarterly measures in the person-level Longitudinal Labour Force Survey (LLFS) microdata, which covers LFS responses from 1982.[1] The dataset followed (anonymised) individuals for eight consecutive months, which allowed us to track job switching behaviour over that time. The key metrics of interest are:

• Job turnover: people that lost or left any job in the past three months.[2] Job turnover can be voluntary or involuntary. Voluntary job separations capture workers who leave a job with the intention of finding another job, as well as those who quit for life-cycle or other personal reasons. Involuntary separations include workers who lose a job due to economic reasons (such as retrenchments), dismissals, the ending of a temporary job and own illness.
• Job mobility rate (constructed by the Reserve Bank): the share of employed persons who have been working for their current employer/business for less than three months and were also employed in the previous quarter.[3] The job mobility rate is a narrower measure than job turnover; job mobility captures workers who left their jobs and are currently employed in a new one, including those who started a new job after a brief period of unemployment or absence from the labour force. The higher frequency measure of job mobility tracks relatively closely with the annual ABS data, despite conceptual differences between the measures; a person who switches jobs multiple times during the year would only be counted once in the annual measure.
• Expected job mobility rate: the share of workers who expect to change jobs in the next 12 months.

## Job mobility at the onset of the pandemic

The onset of the COVID-19 pandemic in Australia in March 2020 had significant effects on the labour market (Graph 2). Movement restrictions and lockdowns, as well as precautionary behaviour, meant many businesses closed or operated at reduced capacity. Corresponding to the decline in economic activity, a significant number of workers lost their jobs or had their hours reduced, while others rapidly adjusted to working from home. Job vacancies and advertisements also fell sharply. Government policy measures, including the introduction of the JobKeeper wage subsidy, helped to keep many workers attached to their employer; in the absence of JobKeeper, employment would have declined much further (Bishop and Day 2020).

These disruptions had a large impact on job turnover (Graph 3). The number of employed persons who lost their jobs rose sharply. This included workers who were retrenched or made redundant, and those who lost their jobs because their employer went out of business or had no work available. This was associated with a large increase in the number of people temporarily exiting the labour force, with a 3.3 percentage point drop in the participation rate in the three months to May 2020.

Job mobility also declined at the start of the pandemic. The share of employed persons who changed jobs fell to its lowest level in over two decades in mid-2020; only around 2 per cent of employed people changed jobs in the three months to August 2020 (Graph 1). Seeking out a better job/wanting a change became a much less common reason for leaving a job. This likely reflected workers' increased risk aversion to switching jobs amid heightened uncertainty, as well as fewer opportunities for workers seeking to improve their job match due to lower labour demand. In broad terms, this is consistent with the pro-cyclical nature of voluntary turnover. Voluntary turnover tends to fall during economic downturns when workers are less confident about switching jobs and employment opportunities are less plentiful; it tends to rise during periods of strong labour market conditions when workers are more willing to bear the costs and risks of changing jobs. However, this wasn't a typical cyclical event as the downturn was precipitated by lockdown measures to counter the health effects of the pandemic. The extended lockdowns also meant it was more difficult for workers to look for new jobs, and the implementation of JobKeeper kept many workers linked to their employers. With that said, the number of people who left their jobs due to ‘other reasons’ spiked during periods of lockdowns, which may have captured workers who left their jobs during the pandemic for a multitude of reasons, including wanting a better job, but felt that it was too difficult to attribute their situation to one category.

The decline in job switching was evident for both full-time and part-time workers, but more so for the latter group. This likely reflected the more adverse impact of lockdowns on contact-intensive industries that typically employ a higher proportion of part-time workers and casuals, such as accommodation & food services (discussed below).

## The recovery in job mobility

As health-related restrictions eased and the economic recovery progressed, the number of people working reduced hours was gradually unwound and more workers gained employment as firms resumed hiring (or workers resumed their previous jobs if they returned to their previous employer). Many workers who had left the labour force at the height of the pandemic have since returned. Consistent with the improvement in labour market conditions, the number of involuntary job separations has declined to be around more usual levels. Voluntary resignations have picked up, with the most common reason being to seek a better job or the desire for a change. This is also evident in the increase in the job mobility rate, which since mid-2021 has been at a high level relative to recent history. For example, the average job mobility rate over the past decade was 2.8 per cent; in February 2022, it was 3.3 per cent – nearly 440,000 people switched jobs in the three months to February. In contrast to the earlier stages in the pandemic when the mobility rate declined more for part-time workers, it is now at a similarly high level for both full-time and part-time workers.

Information from the Bank's liaison program also indicates that voluntary turnover rates have increased since mid-2021 to be above average, particularly in sectors or roles experiencing strong demand for labour. According to liaison, while this partly reflects some catch-up following two years of lower staff turnover during the pandemic, it is also due to increased competition for labour, with workers being enticed to join other firms for a higher salary. A simple counterfactual exercise – which compares the actual cumulative flow of job switches since February 2020 to what the cumulative flow would have been if the number of job switches had remained at its average level in 2019 throughout the pandemic – suggests there has been a sizeable number of job switches beyond a simple catch-up following the initial decline, accounting for around 1.5 per cent of total employment as of February 2022.

The speed of recovery in job mobility from the COVID-19 shock is in contrast to the GFC – job mobility did not recover to its pre-GFC levels and trended lower for the next decade. The natures of these events, however, were distinctly different. The recent downturn was precipitated by lockdown measures to counter the health effects of the virus. The rapid recovery in the labour market that followed the removal of these measures and ongoing underlying momentum supported by policy settings have enabled a faster rebound in job mobility, which quickly exceeded its pre-pandemic level. The labour market recovery was much slower after the GFC and aggregate demand remained low for many years. A similar post-GFC decline in job mobility has been evident in other advanced economies, suggesting that structural factors play a role – for example, this may be related to the ageing population, the decline in the share of startups and a rise in larger firms, and policy settings (Engbom 2019; Decker et al 2014; Decker et al 2020; Hermansen 2019).

## Characteristics of job mobility during the pandemic

The impacts of the pandemic on the labour market have been uneven across different groups of people and sectors in the economy. Activity restrictions and isolation requirements have more adversely affected industries that tend to be contact intensive, such as accommodation & food services. These industries also tend to employ a higher share of younger, part-time and casual workers. By contrast, employment has recovered rapidly to be well above pre-pandemic levels in other industries, such as health care & social assistance and professional services; this is partly due to the health response to the pandemic, strong underlying demand and/or the availability of remote or socially distanced work.

While the extent of job mobility generally varies across industries, the impact of the pandemic on job mobility has also been more concentrated in some areas of the economy than others. In general, industries with lower earnings, lower wages growth and a younger average worker age are typically associated with higher job mobility. This is consistent with workers having less incentive to move from their existing job or industry in which they have accumulated experience that adds to their earning potential (D'Arcy et al 2012). High-skilled jobs have experienced particularly sharp increases in job mobility since mid-2021; notably, the job mobility rate in professional services has increased to be around its highest level in over two decades (Graph 4). This is consistent with information from the Bank's liaison program that suggests that strong labour demand in professional services has encouraged more employees to seek higher rates of pay or new opportunities, and resulted in a very competitive job market for businesses seeking to hire workers, particularly those skilled in IT. Job mobility in the healthcare industry – where employment growth has been particularly strong since the start of the pandemic – has also increased to be at its highest level in over two decades. Likewise, job mobility in contact-intensive industries and construction has recently picked up, after being relatively low during the earlier stages of the pandemic.

Changes in job mobility during the pandemic have been evident across most age groups (Graph 5). At the onset of the pandemic, the decline in job mobility was most pronounced among younger workers (aged 15–19 years and 20–29 years), likely reflecting reduced demand for entry-level workers and the more adverse impact on contact-intensive industries that employ a higher share of young workers. Typically, younger workers have higher rates of job switching than older workers. This is because young people have less firm- and industry-specific human capital than more experienced older workers and so have more to gain by changing jobs and increasing the quality of a job match; an example of this is a university graduate who switches from casual employment to a full-time career in an industry related to their studies. Job mobility for these workers has since recovered, alongside improvements in youth labour market outcomes. In fact, job mobility has increased for most age groups in recent months to be above pre-pandemic levels.

Job mobility includes workers changing jobs within the same sector (referred to as ‘churn’) and workers switching to jobs in different sectors (Graph 6). In the early stages of the pandemic, there was a sharper decline in job churn than in workers changing industries, though the latter also declined. Since mid-2021, as job mobility increased with the strengthening labour market, the proportion of job switches involving changing industries as distinct from churn has returned to its historical average. However, there have been some differences across sectors. In the business services sector, the large increase in job mobility from mid-2021 was mostly driven by increases in job churn, with only a modest increase in workers switching jobs from outside the sector. The household services sector has experienced similar increases in job switches within and into this sector, while job mobility in the goods sector has been relatively stable throughout the pandemic. In addition, data on job switching by the skill level of the job shows that those workers who have changed jobs have mostly moved into a job requiring the same level of skill.

## Job switching expectations

Some insights on whether job mobility will remain elevated or return to previous levels can be obtained from data on future employment expectations. Actual job mobility has largely followed expected job mobility over the past two decades, although these have differed at times (Graph 7). Workers remain upbeat about the jobs outlook, with the share of workers expecting to change jobs/seek other employment within the next year being around its highest level since 2008. The increase in expected job mobility has been primarily driven by full-time workers and is more pronounced among female full-time workers. By contrast, the share of part-time workers expecting to change jobs has returned to its longer term downwards trajectory.

Higher skilled occupations have continued to report elevated levels of expected job mobility, while lower skilled occupations have remained more in line with the longer term average. By industry, expected job mobility for the year ahead was elevated in healthcare & social assistance and financial & insurance services, and subdued in accommodation & food services, arts & recreation and other services, and construction.

## Job mobility and wages growth

While workers may change jobs for a number of reasons, higher wages are often the motivation. Data from the Household, Income and Labour Dynamics in Australia (HILDA) survey show that people who move jobs generally gain higher-than-average wage increases (Graph 8). People who stay with the same employer have lower but more stable wage growth over time. Recent domestic and international research has shown that higher job mobility also tends to be associated with higher aggregate wages growth (Faberman and Justiniano 2015; Karahan et al 2017; Moscarini and Postel-Vinay 2017; Deutscher 2019). This can arise from two channels: directly, because workers typically experience a pay bump from changing jobs; and indirectly, because an employer may offer a pay raise to retain a worker in their current job due to competition for labour.

In Australia, higher job mobility has tended to be associated with higher aggregate wages as measured by the Wage Price index (WPI) (notwithstanding that other factors are also important drivers of wages, as discussed below) (Graph 9). The WPI measures the wages of a sample of jobs (rather than a sample of workers). In this way, it will capture the higher wages that an employer pays for a particular job in the face of high job mobility – for example, higher wages offered to poach or retain a worker.[4] While high job mobility tends to be associated with wages growth, the direction of causality could run either or both ways. High job mobility may lead to workers receiving a wage increase by switching jobs or through higher competition for workers in their industry. However, the reverse could also occur, whereby wages growth incentivises workers to switch jobs within their industry or it could lead to higher job mobility across industries – for example, higher wages growth in particular sectors may lead workers to retrain and switch jobs into those sectors, and/or encourage firms to consider workers coming from other industries even if their work experience is not seen as directly relevant.

As part of its set of models, the Reserve Bank uses the Phillips curve framework to consider the implications for wages growth based on a number of cyclical factors, including: spare capacity in the labour market; inflation expectations; a measure of changes in firms' output prices; and lagged wages growth (Bishop and Greenland 2021). Job mobility is not usually included in this model; however, using this framework, we explored whether job mobility provides additional information for future wages growth above and beyond the variables in the Bank's baseline wages model (see Appendix A for full model results).

The results indicate that job mobility can help to explain future changes in WPI growth in the near term. According to the model, the increase in the job mobility rate in the decade leading up to the GFC supported wages growth generally, and the trend decline in job mobility over the following decade is estimated to have been a drag on wages growth. With that said, the inclusion of job mobility in the model provides some but not much additional information to explain wages growth beyond the baseline model. The existing measure of spare capacity in the labour market in the model (the unemployment gap) is able to account for most of the variation in wages growth related to labour market tightness, including the effects of job mobility. This result is not too surprising as the unemployment gap and job mobility tend to move closely together.

Overall, the recent pick-up in job mobility in some sectors in Australia is expected to contribute over time to employers offering higher wages to retain workers or hire new workers as the labour market tightens. Some firms in the Bank's liaison program reported having increased wages in response to elevated turnover, with some of these firms paying out-of-cycle wage increases to some workers to meet the increase in market salaries for those roles and prevent further increases in turnover. In addition, some firms have paid more to attract new staff to fill vacant or new roles.

## Job turnover, mobility and wages growth through the pandemic: Comparing Australia with the United States and the United Kingdom

Similar to Australia, job turnover and mobility have increased sharply in recent months in the United States and the United Kingdom (Graph 10). In the United States, where data on job mobility are not readily available, the job quits rate – the share of workers leaving their jobs to take new jobs or to exit the labour force – has been around a record high since mid-2021. In the United Kingdom, the job mobility rate has increased to an elevated level. Wages growth has picked up sharply in both of these countries – notably more so than in Australia. Notwithstanding the structural differences in the US and UK labour markets, the extent of the recovery in labour supply, in the face of strong labour demand, appears to have played a role in the different wage growth outcomes so far. While labour force participation declined at the onset of the pandemic in all three countries, it has since increased in Australia to be at a record high, whereas the US and UK participation rates still remain below pre-pandemic levels (although it has been recovering recently in the United States) (Graph 11). COVID-19-related health concerns, high accumulated savings due to fiscal support, increased retirements and workers re-evaluating longer term personal and professional goals (perhaps as part of the ‘Great Resignation’) have likely contributed to people leaving the labour force in the United States and the United Kingdom (Agarwal and Bishop 2022). As a result, firms in these countries have been paying higher wages to retain and attract workers. Conversely, health concerns appear to have had less of an impact on labour supply in Australia – partly due to comparatively better health outcomes – and retirements have remained around usual levels throughout the pandemic.

## Conclusion

Job mobility in Australia has increased to a high level recently, underpinned by the strong labour market. This followed a sharp decline in job switching at the onset of the pandemic amid widespread disruptions to the labour market, and a multi-decade trend decline prior. The increase in job switching appears to be partly driven by strong labour demand for some high-skilled jobs, coupled with workers catching up on planned job changes that were put on hold during the pandemic. High levels of job switching tend to be associated with higher wages, both at the individual and aggregate level, in a tight labour market. While job turnover has also been high in the United States and the United Kingdom, labour force participation rates have not yet recovered to pre-pandemic levels; by contrast, Australia is experiencing a record high level of participation. This appears to have played a role in relative wage outcomes across the countries. The increased rates of job mobility in Australia in some sectors is expected to contribute to higher labour costs as firms find that they need to offer higher wages and other benefits to attract new workers or retain existing staff.

## Appendix A: The RBA Wages Phillips curve model with job mobility

We examined whether the addition of the job mobility rate to the RBA Wages Phillips curve model provides additional information for wages growth. We estimated the equation below:

$%∆{WPI}_{t}=\alpha +\beta %∆{WPI}_{t-1}+\gamma \left(\frac{{u}_{t-1}-{u}_{t-1}^{*}}{{u}_{t-1}}\right)+\tau ∆{u}_{t-1}+\phi \frac{{%{∆}^{ye}DFDdef}_{t}}{4}+\theta \frac{{Inflationexp}_{t-1}}{4}+\delta Jobmobilit{y}_{t-1}+{ϵ}_{t}$

Where:

 ${\mathrm{%}∆WPI}_{t}$ Quarterly growth in the private sector Wage Price Index (WPI) (hourly rates excluding bonuses); seasonally adjusted. ${u}_{t}$ Quarter-average unemployment rate; seasonally adjusted. ${u}_{t}^{\mathrm{*}}$ Non-accelerating inflation rate of unemployment (NAIRU) (described in Cusbert 2017); two-sided smoothed quarterly estimate. $∆{u}_{t}$ Quarterly change in the unemployment rate; seasonally adjusted. ${\mathrm{%}∆}^{ye}{DFDdef}_{t}$ Year-ended growth in the domestic final demand implicit price deflator; seasonally adjusted. ${Inflationexp}_{t}$ Trend inflation expectations estimated using a Kalman filter (described in Cusbert 2017); two-sided smoothed quarterly estimate. $Jobmobilit{y}_{t}$ Quarterly job mobility rate; seasonally adjusted.
Baseline model Job mobility 0.066*** (0.020) −0.362** (0.169) 0.425*** (0.109) 0.337*** (0.110) −0.533*** (0.156) −0.402** (0.193) −0.097 (0.073) −0.059 (0.073) −0.012 (0.086) −0.109 (0.102) 0.587*** (0.132) 0.771*** (0.153) – 0.142** (0.057) 0.74 0.77 (a) The dependent variable is the quarterly growth in the private sector WPI (hourly rates excluding bonuses); standard errors in parentheses; ***, **, and * denote statistical significance at the 1, 5, and 10 per cent levels, respectively. Sources: ABS; authors' calculations

## Endnotes

Susan Black is from Economic Analysis Department and Emma Chow completed the work while in Economic Analysis Department. The authors would like to thank Nalini Agarwal, James Bishop, Mark Chambers, Matthew Fink, Jonathan Hambur, Jess Meredith, Neya Suthaharan and Tom Williams for thoughtful advice and suggestions in preparing this article. [*]

The LFS is a monthly survey of Australia's resident population aged 15+ years. The LLFS data – the longitudinal data from this survey – contains responses to the monthly labour force survey from 1982 onwards, and include respondents' employment outcomes in the surveyed months as well as some data on worker characteristics. [1]

This captures individuals who indicate that they have lost or left any job in the previous three months, and not necessarily an individual's main job. For multiple job holders, this data may reflect changes in their main job or in a secondary job. [2]

We infer whether an individual switched jobs using the variable on job tenure, which records the number of months an individual has been employed with their current employer/business. Depending on how an individual is interpreting the question and when LFS samples were taken, a tenure of three months might not signify a recent switch. [3]

The WPI does not capture the pay bump that a worker might receive from switching to a role that is substantially different to their previous role. This would be included in broader measures of employee earnings growth like average weekly earnings. [4]

## References

Agarwal N and J Bishop (2022), ‘ COVID-19 Health Risks and Labour Supply’, RBA Bulletin, March.

Andrews D and D Hansell (2019), ‘Productivity‐Enhancing Labour Reallocation in Australia’, Treasury Working Paper No 2019-06.

Bishop J and I Day (2020), ‘How Many Jobs Did JobKeeper Keep?’, RBA Research Discussion Paper No 2020-07.

Bishop J and E Greenland (2021), ‘Is the Phillips Curve Still a Curve? Evidence from the Regions’, RBA Research Discussion Paper No 2021-09.

Cusbert T (2017), ‘Estimating the NAIRU and the Unemployment Gap’, RBA Bulletin, June, pp 13–22.

D'Arcy P, L Gustafsson, C Lewis and T Wiltshire (2012), ‘Labour Market Turnover and Mobility’, RBA Bulletin, December, pp 1–12.

Decker R, J Haltiwanger, R Jarmin and J Miranda (2014), ‘The Role of Entrepreneurship in US Job Creation and Economic Dynamism’, Journal of Economic Perspectives, 28(3), pp 3–24.

Decker RA, J Haltiwanger, RS Jarmin and J Miranda (2020), ‘Changing Business Dynamism and Productivity: Shocks versus Responsiveness’, American Economic Review, 110(12), pp 3952–3990.

Deutscher N (2019), ‘Job-to-Job Transitions and the Wages of Australian Workers’, Treasury Working Paper No 2019-07.

Ellis L (2021), ‘Innovation and Dynamism in the Post-pandemic World’, Speech to the Committee for the Economic Development of Australia, Webinar, 18 November.

Engbom N (2019), ‘Firm and Worker Dynamics in an Aging Labor Market’, Federal Reserve Bank of Minneapolis Working Paper No 756.

Faberman RJ and A Justiniano (2015), ‘Job Switching and Wage Growth’, Chicago Fed Letter No 337.

Hermansen M (2019), ‘Occupational Licensing and Job Mobility in the United States’, OECD Economics Department Working Paper No 1585.

Karahan F, R Michaels, B Pugsley, A Şahin and R Schuh (2017), ‘Do Job-to-Job Transitions Drive Wage Fluctuations over the Business Cycle?’, American Economic Review, 107(5), pp 353–357.

Moscarini G and F Postel-Vinay (2017), ‘The Relative Power of Employment-to-Employment Reallocation and Unemployment Exits in Predicting Wage Growth’, American Economic Review, 107(5), pp 364–368.

HILDA Disclaimer