1. Stay not leave: Retaining talent at all ages

Growth in the labour force is now declining in many countries as lower fertility reduces inflows of young people into the labour force while outflows have increased because of the baby-boom generation reaching retirement (Bloom, Canning and Fink, 2010[1]). The population aged 20-64 – the prime working age population – as a share of the total population has been trending down on average across OECD countries since 2010. It has declined from an average of 60% in OECD countries in 2010 to 58% in 2021 and is projected to fall to 53% by 2060 (Figure 1.1).

The COVID-19 pandemic led to an unprecedented rise in labour shortages in many countries and companies, but shortages pre-date the pandemic. Labour shortages have been rising steadily since the Global Financial Crisis (Figure 1.2, Panel A). In 2019, prior to the pandemic, about 55% of employers in a survey of more than 40 000 employers across all industry sectors in 40 countries were reporting talent shortages, compared with 35% in 2013. Employers in Finland, Hungary, Slovenia, Sweden and the United States reported the largest increase in labour shortages in 2019 (Manpower Group, 2020[2]). Skilled trades, for example electricians and mechanics, sales and marketing, and technicians were at the top of the list of occupation groups that were scarce (Manpower Group, 2020[2]). Following the COVID-19 crisis, the share of employers reporting shortages rose to 69% in 2021, and as countries start to recover remains high at 75% in 2022. Health care professionals also entered the top ten most in-demand roles reflecting ageing populations and the growing need for health care. Unfilled vacancies have also grown considerably over the last decade, and particularly since the COVID-19 pandemic (Figure 1.2, Panel B).

In this report, job tenure – the length of continuous time spent with the same employer – is used as a measure of job retention. Retention and labour turnover are key concepts for analysing the relationship between businesses and employees (Box 1.1). They are two sides of the same coin. Retention, measured at the firm level, tells companies how attractive their company is in holding onto talent, while turnover is a measure of the recruitment efforts that are needed because of employees leaving the company each year (voluntarily or involuntarily). Retention and turnover measure the strength of the attachment between employees and employers, and the extent to which they are too high or too low has implications for productivity, inequality and well-being. Some turnover is inevitable and desirable, but high or unwanted worker turnover (“excessive job turnover”) can have serious consequences for business including higher staff costs, the loss of skills and experience, and the erosion of workplace diversity.

Labour turnover or mobility because of either changing jobs or moving in and out of the labour market is a normal part of how labour market’s function. The flow of workers between firms and sectors over time facilitates adjustment to economic shocks (for example to product demand, technology, or costs) and structural change. This reallocation process is an important driver of productivity growth as workers move away from lower productivity firms towards higher productivity firms.1 Job-to-job changes are the main component of labour turnover, and when voluntary can be a key driver of employee pay, and career progression (Topel and Ward, 1992[3]; Hahn, Hyatt and Janicki, 2021[4]). Job changes tend to be more frequent while young, and gradually decline over a workers’ lifetime.

Although there are benefits associated with workers moving between jobs, there are also benefits to longer job tenure (Figure 1.3). The benefits of longer job tenure and the costs of labour turnover create a trade-off at the aggregate level between job stability and labour mobility. There is no optimal level of job stability or labour mobility, this depends on factors such as technology, firm size and social preferences.

Excessive job turnover can hurt employees, businesses and society by adversely affecting productivity and inclusiveness. Productivity can be affected through multiple channels. Firms suffer from the loss of workers with job-specific skills and experience, disruption to the production or services they provide, and incur the costs of hiring and training new workers. The loss of key employees can have a particularly damaging impact in small and medium enterprises (SMEs), as departing workers may have particular skills or knowledge that are difficult to replace. At the firm level there is robust evidence that turnover can have a negative effect on productivity (Kuhn and Yu, 2021[5]).

Differences in the rate of retention of different groups within firms can erode workplace inclusiveness (Griffeth, 2000[6]) and lead to poorer overall organisational performance if particular groups, such as older workers, are less likely to be employed thereby depriving an organisation of critical knowledge. Certain groups of workers such as the low-skilled, people with poor health or disabilities, ethnic minorities, and older workers are less likely to be employed and have lower hiring rates (OECD, 2020[7]). This contributes to greater inequality in income and well-being within firms and in society at large.

If it is well managed, an age-diverse workforce can bring about several benefits, chief among them, improved productivity and lower turnover for companies and higher wages for workers. Greater age diversity in firms has been found to be associated with higher productivity and lower turnover (OECD, 2020[7]). Complementarities between workers of different ages can lead to higher productivity over and above that of the sum of individual workers’ productivity. For example, older workers can provide advice to younger workers and can draw on firm-specific knowledge accumulated over time, while younger workers might bring knowledge of new technologies for example.

As countries recover from the COVID-19 pandemic, firms are starting to rehire or replace workers, or hire workers with a different mix of skills in the context of a generally very tight labour markets (Domash and Summers, 2022[8]). The additional costs due to the pandemic induced disruption and labour market reallocation come on top of the already considerable time and effort that is required to recruit staff to fill vacancies or expand the workforce.

There are significant business costs to replacing employees which are typically larger in “thin” markets and when skills are specialised (Jäger and Heining, 2022[9]). Using German data on worker exits from approximately 34 000 small firms,2 Jäger and Heining (2022[9]) estimate that the marginal replacement cost ranges between EUR 65 000 and EUR 84 000, or between 2.3 and 3 annual salaries of an incumbent.3 These estimates are much higher than previous firm survey based estimates such as Manning (2011[10]). However, Jäger and Heining (2022[9]) argue that their estimates capture costs missed in firm surveys, including the higher costs of retaining incumbent employees who become more valuable because of co-worker turnover. They also capture the loss of firm-specific human capital acquisition that is not embedded in worker training, i.e. tacit knowledge that takes time to acquire. Finally, they include the costs of replacing incumbent workers where the match between worker and firm is very good – something that takes time to be revealed.

Employee turnover can also be contagious, which means that the spread of employee turnover through a workplace or work group can result in costs that exceed a single instance of turnover (Porter and Rigby, 2021[11]). Felps et al. (2017[12]) found that an employee’s decision to voluntarily leave an organisation is influenced by the attitudes and behaviours of their co-workers. In particular, how well employees feel they fit in with their job and the community (“job embeddedness”), and the job search behaviour of co-workers were important predictors of voluntary quits.

Employees can benefit from stable employment as it allows opportunities for career advancement through training and the accumulation of valuable job or firm-specific skills. Yet investment by firms in innovation and employee training arguably requires long-term relationships. Firms invest in training programmes to enhance the human capital of employees with the aim of raising firm productivity and profitability. If turnover is unnecessarily high, firms will not capture the full benefits of this investment. Likewise, for workers, too many job changes and periods of unemployment may limit opportunities for such training, leading to losses in human capital and earnings potential.

If turnover is high, firms have an incentive to take advantage of the training investment of other firms and therefore under-invest in their own employees (Acemoglu, 1997[13]). This effect could depress overall training levels in the economy, thereby reducing productivity. This effect is more likely to be found in high turnover sectors where general training is more important than more advanced firm-specific training. Training can also encourage workers to leave through improving their skills and therefore their outside options, but it may also encourage them to stay by raising their pay and performance.

An ageing society – by 2060 the number of people in OECD countries over the age of 80 is expected to reach 151 million, up from 68 million in 20224 – leads to concerns about declining GDP growth per capita as a result of the decline in the working age population (Maestas and Zissimopoulos, 2010[14]; Bloom, Canning and Fink, 2010[1]) and has implications for public finances in the face of rising pension and health costs (Aksoy et al., 2019[15]; Cooley and Henriksen, 2018[16]). By supporting longer working lives – where possible – and boosting the retention and productivity of older workers, governments can help ameliorate the possible negative effects of population ageing and longevity on economic growth and government budgets. However, working longer is not necessarily possible for all workers in the absence of improvements to job quality and health.

Older people who are displaced or voluntarily leave the labour market often struggle to find jobs and face much larger earning losses relative to younger workers if they do re-enter employment (OECD, 2018[17]). This has direct fiscal implications for governments from increased social welfare payments and reduced income tax receipts. The health status of older workers who leave the labour market prematurely often declines, with an increase in the incidence of depression, alcohol-related conditions and higher mortality, which also affects government budgets (OECD, 2018[17]).

This section looks at pre-pandemic (up until 2019) trends in average job tenure across OECD countries for which there is available data. At the beginning of 2023 it is not yet clear whether the pandemic has had a temporary or permanent impact on trend changes in tenure and turnover.

In 2019, the average length of a job held by the same worker was nine years and 10 months for women and 10 years and six months for men across 30 OECD countries. Between 2012 and 2019, overall job tenure fell in most OECD countries,5 after taking into account changes in the composition of the workforce.6 After adjusting for changes in socio-demographic structure, job tenure decreased by about 7.5% (or around 8.8 months) for men and women on average across OECD countries (Figure 1.4) between 2012 and 2019 (throughout this chapter changes between 2012 and 2019 refer to changes between the average of 2012/14 and 2017/19). Over this period, only Iceland, Hungary, Australia, Korea and Latvia saw an increase in average tenure for men; and only Korea saw an increase in average tenure for women. For women, average tenure declined by more than 12% in Lithuania, Ireland, Sweden, Portugal, Slovenia and Greece. Average tenure decreased by more than 12% for men in Ireland, Spain, Portugal, Luxembourg, Sweden and France (declines of between 12 to 16 months).

Job tenure naturally varies across age groups reflecting greater time spent in the labour market for older workers. For young workers (aged 15-29, excluding students), average job tenure in 30 OECD countries in 2019 was two years and five months.7 For those aged 30-54, the average length was 10 years and one month, and for mature workers aged 55-64 the average was 18 years and 10 months.

On average across 30 OECD countries, average job tenure has fallen by 9.5% for young workers (aged 15-29), 7.2% for prime-aged workers and 5.7% for mature workers (Figure 1.5) between 2012 and 2019. Young workers in Spain, Portugal, Greece and Slovenia experienced the largest declines in job tenure. All countries in the sample, except for Korea and Latvia, saw a decrease in average tenure for prime-aged workers. The average mature worker in 24 out of 30 countries experienced a decline in job tenure. Job tenure rose slightly for mature workers on average in Korea, Czech Republic, Austria, Poland, Latvia and Iceland.

The decline in average job tenure between 2012 and 2019 was larger for low-educated workers (i.e. with less than upper-secondary education) than for other education groups (Figure 1.6). This was the case for all age groups and both men and women. Over two-thirds of the OECD countries in the sample saw reductions in tenure for low-educated workers. In some countries, the reduction in tenure was large, exceeding 10% in Lithuania, Ireland, Luxembourg, Estonia and Sweden. In contrast, Latvia and Italy saw workers without an upper-secondary qualification experience an increase in tenure of 4% and 0.2% respectively. Job tenure has also declined substantially for those with high education in some countries. Lithuania, Spain, Ireland, Luxembourg, France and Finland saw declines of over 10% for those with high education. Average tenure also declined substantially in some countries for those with medium levels of education (Finland, Sweden, the Slovak Republic, Slovenia, Lithuania and France), which might reflect the hollowing out of middle-skill occupations.

Job tenure also varies by sector, occupation and firm size (Figure 1.7). Many sectors such as public health or social care struggle to retain staff compared to other sectors. Across broad sectors, average tenure tends to be highest in Agriculture, forestry and fishing at around 16 years and lowest in Accommodation and food services at about six years (Figure 1.7, Panel A). Differences such as these will partly reflect the business models and technology of different industries. For example, the skills required by people working in accommodation and food services will be relatively low on average compared with workers in manufacturing. This pattern can also be seen at the occupation level. Skilled agriculture forestry and fishery workers have average tenure of over 16 years, although this reflects that many of these workers will be older and self-employed (Figure 1.7, Panel B). Relatively unskilled elementary occupations have average tenure of about eight years.

Large firms generally find it easier to retain staff and this is reflected in longer average tenure for workers in these firms (Figure 1.7, Panel C) (see also Chapter 2).

Average tenure data are censored – the data only include the length of tenure for current jobs, and it is not known for how long these jobs will last. An alternative is to consider the survival probability for retention of a particular cohort, such as 55-64 year-olds, i.e. the likelihood of workers being employed in the same job in five years’ time. On average across the OECD, just over half of working 55-59 year-olds leave their employer by the time they are 60-64, compared to only about 30% of 40-44 year-olds who leave by the time they are 45-49 (Figure 1.8). Although older workers have relatively long average tenure, by the time they reach 55-59, the likelihood that they will continue in the same job falls dramatically. Further, there are stark differences across countries; in Korea, Luxembourg, Slovenia, and France the share of 55-59 year-olds who remain in the same job five years later is less than 40%. In contrast, for 55-59 year-olds in Germany, Portugal, Ireland, Norway, the Netherlands and Iceland, the likelihood that they are still in the same job five years later is greater than 65%. As the following section shows, workers in this age group are more likely to exit the labour market than change jobs. Displaced older workers have much lower re-employment rates than younger workers; therefore retaining older workers – when they can continue working – can be beneficial.

The flip side of the decline in average job tenure is rising labour market turnover (the sum of movements into and out of jobs over the course of a year). People can change jobs, become unemployed or leave the labour market (inactivity) for a variety of reasons, which can broadly be defined as either voluntary or involuntary. Involuntary reasons include being laid off or a fixed term contract coming to an end. All other reasons are considered voluntary and are discussed in Section 1.5. On average across 27 OECD countries, the labour market turnover rate was just under 22% in 2019 (Figure 1.9).8 Approximately 42% of this is accounted for by workers changing jobs, while the rest is split between workers shifting to non-employment (unemployment or inactivity) (23%) or being hired from non-employment (35%). There is wide variation in turnover across countries: the annual turnover rate is over 25% in Denmark, Finland, Australia, Lithuania, Iceland, Sweden, Estonia and Spain, compared with about 15% in Greece and the Czech Republic.

Over the period 2012-19 job-to-job flows have increased in all but two of 27 OECD countries, after making adjustments for the demographic composition of the population and the business cycle (Figure 1.10, Panel A).9 Although there has been an increase in job-to-job transitions for all age groups over this period, mature workers are less likely to change jobs compared to younger workers (Figure 1.10, Panel B). This likely reflects older workers’ preferences for fewer job changes as they are more likely to have settled into a good “match” with their employer and have more stable life circumstances. It could also reflect reservations on the part of employers for hiring older workers.

Low-skilled mature workers are more likely to change jobs relative to higher skilled mature workers in 18 out of 27 OECD countries (Figure 1.10, Panel C), however in most countries (except Hungary, Poland and Finland) the differences in rates of job-to-job change among different age groups are very small and not statistically significantly different from each other. There is no evidence that high skilled older workers are more likely to change jobs compared to low and middle skilled workers.

In the immediate aftermath of the pandemic quit rates rose to record highs in the United States, and there is evidence of quit rates also rising in the United Kingdom and France (OECD, 2022[20]). In 2021, over 47 million Americans voluntarily quit their jobs.10 While monthly quit rates remained higher than usual in the first six months of 2022 in the United States, arguably these quit rates have been broadly in line with the pre-COVID trend of a 0.1 percentage increase every year since 2009 (Fuller and Kerr, 2022[21]). This uptick in quits appears to be driven primarily by workers changing jobs, rather than leaving the workforce altogether (Barrero et al., 2021[21]).

Although the rate at which mature workers (aged 55-64) change jobs is low relative to younger workers – the rate at which mature workers voluntarily leave a job for either unemployment or inactivity is higher relative to younger workers in most OECD countries (Figure 1.11, Panel A). With four exceptions (Australia, Estonia, Iceland and Latvia), the rate at which mature workers voluntarily leave a job for non-employment is higher relative to prime-age and young. This largely reflects early retirement which could be due to a change in preferences and having the financial ability to retire early, or ill-health, for example. The rate at which older workers voluntarily leave a job is the highest in the Czech Republic, Luxembourg, France, Austria, Lithuania and Latvia. It is relatively low in Sweden, Iceland, Italy and the Netherlands (Figure 1.11, Panel A).

The rate at which mature workers are laid off (involuntary separations) and enter unemployment or inactivity is generally at or below the rate for younger workers (Figure 1.11, Panel B). Further, after adjusting for demographics and the business cycle, the likelihood of becoming involuntarily unemployed or inactive fell on average between 2012 to 2019 across the 25 countries for which there is available data. This decline in the risk of being laid off (or a fixed term contract ending) was felt across gender, age groups and education groups (on average across 25 OECD countries).

However, for mature workers there are considerable differences across education level in the likelihood of ending up unemployed or out of the labour force. In some OECD countries (including Czech Republic, Latvia, Lithuania, and Estonia), the rate at which mature workers leave employment is much higher for low educated workers compared with higher educated workers (Figure 1.11, Panel C). This is primarily due to low-skilled mature workers being laid off or a temporary contract coming to an end, rather than for voluntary reasons. However, in the majority of countries there is very little difference between the separation rates of low, middle and highly educated workers.

Labour force survey data give an indication of the broad reasons why people leave work (job or business) voluntarily. These include for personal health reasons, to pursue education or training, to take care of other family members, retirement, or other reasons.11 Among workers aged 55-64, retirement is the main reason for quitting a job across OECD countries for which there is available data (Figure 1.12, Panel A). Illness is also a major reason for 55–64 year-olds quitting a job, 21% of 55-64 year-olds quit due to illness or disability, on average across OECD countries. Because the survey question only allows a single response, a respondent may select the retirement option even if illness is the reason for leaving a job. Therefore, it is possible that leaving due to illness is understated among older age groups. Among workers aged 50-64, 25% quit because of illness or disability. For workers aged 30-54, the majority quit either for family reasons (35%) or for “other” reasons (34%). A further 24% quit because of illness or disability (Figure 1.12, Panel B).12 These data illustrate the importance of illness as a significant factor in driving people to leave their job.

The response category “other” in reasons for leaving employment in labour force survey data will include changing job, which could be for a variety of reasons. While it is not possible to explore these reasons further using labour force survey data, the 2022 AARP Global Employee Survey conducted in 12 countries in 2022 sheds light on the reasons workers might be leaving their jobs (Choi-Allum, 2023[22]).

The AARP Global Employee Survey finds that 65% of respondents have made one or more employment related changes (starting a new job, starting a business, being laid off, or retiring) in the last five years, and 35% have made no change. Out of those who have made some employment-related change, almost half (45%) have started a new job. The survey shows that overall, low pay (29%), feeling undervalued (27%) and lack of advancement in their job (23%) were the top three reasons why people have switched to a new job in the last five years (Infographic 1.1). To follow a new passion or career (21%), lack of professional development in their job (17%) and poor working conditions (16%) were also important reasons for starting a new job or taking steps to find a new job.

Men and women report similar reasons for having left or taken steps to leave over the last five years, but there are significant differences by age group. For older and prime aged workers, low pay, feeling undervalued, and lack of advancement in their job were the top three reasons given for those who have switched jobs (Figure 1.13, Panel A). For young workers the top three reasons for starting a new job were low pay (38%), feeling undervalued (30%), and to follow a new passion or career (26%).

The overall ranking of reasons is similar between people with different levels of education (Figure 1.13, Panel B). Among people who switched jobs, those who had not completed upper secondary education report low pay (29%), feeling undervalued (22%), and poor working conditions (20%) as the top three reasons, closely followed by lack of advancement in their job (19%). For workers with a tertiary education, feeling undervalued (29%), low pay (27%), and lack of job advancement (26%) were the top three reasons. Respondents with low and medium levels of education were more likely to report health reasons as a cause of changing job, 15% and 13% respectively, compared to only 9% of respondents with a tertiary education. Those with a tertiary education were much more likely to cite following a new passion or career, or lack of professional development as a cause of making a change or looking, compared to those with low education.

Among respondents aged 50-64 years who retired, apart from reaching or being close to retirement age (27%), the main reasons retirees gave were experiencing health problems (25%), caregiving (23%), low pay (22%) and concerns about COVID-19 (20%) (Figure 1.13, Panel C). This illustrates the importance of health and flexibility concerns for workers deciding whether or not to retire.

References

[13] Acemoglu, D. (1997), “Training and Innovation in an Imperfect Labour Market”, The Review of Economic Studies, Vol. 64/3, p. 445, https://doi.org/10.2307/2971723.

[15] Aksoy, Y. et al. (2019), “Demographic Structure and Macroeconomic Trends”, American Economic Journal: Macroeconomics, Vol. 11/1, pp. 193-222, https://doi.org/10.1257/MAC.20170114.

[26] Albagli, E. et al. (2021), “Productivity Growth and Workers’ Job Transitions: Evidence from Censal Microdata”, National Bureau of Economic Research, Cambridge, MA, https://doi.org/10.3386/W28657.

[28] Autor, D., A. Dube and A. McGrew (2022), The Unexpected Compression: Competition at Work in the Low Wage Economy, https://bcf.princeton.edu/wp-content/uploads/2022/10/Combined-Slides.pdf (accessed on 12 December 2022).

[21] Barrero, J. et al. (2021), “COVID-19 Is a Persistent Reallocation Shock”, AEA Papers and Proceedings, Vol. 111, pp. 287-91, https://doi.org/10.1257/PANDP.20211110.

[1] Bloom, D., D. Canning and G. Fink (2010), “Implications of population ageing for economic growth”, Oxford Review of Economic Policy, Vol. 26/4, pp. 583-612, https://doi.org/10.1093/OXREP/GRQ038.

[22] Choi-Allum, L. (2023), Retaining Older Workers Through the COVID-19 Pandemic and Beyond: A Global Perspective, AARP Research, Washington, DC.

[16] Cooley, T. and E. Henriksen (2018), “The demographic deficit”, Journal of Monetary Economics, Vol. 93, pp. 45-62, https://doi.org/10.1016/J.JMONECO.2017.11.005.

[23] Davis, S., R. Faberman and J. Haltiwanger (2012), “Labor market flows in the cross section and over time”, Journal of Monetary Economics, Vol. 59/1, pp. 1-18, https://doi.org/10.1016/J.JMONECO.2011.10.001.

[25] Davis, S. and J. Haltiwanger (2014), “Labor Market Fluidity and Economic Performance”, National Bureau of Economic Research, Cambridge, MA, https://doi.org/10.3386/W20479.

[8] Domash, A. and L. Summers (2022), “How Tight are U.S. Labor Markets?”, https://doi.org/10.3386/W29739.

[19] Farber, H. (2010), “Job Loss and the Decline in Job Security in the United States”, in Abraham, K., J. Spletzer and M. Harper (eds.), Labor in the New Economy, University of Chicago Press, Chicago.

[12] Felps, W. et al. (2017), “Turnover Contagion: How Coworkers’ Job Embeddedness and Job Search Behaviors Influence Quitting”, Academy of Management Journal, Vol. 52/3, pp. 545-561, https://doi.org/10.5465/AMJ.2009.41331075.

[30] Fuller, J. and W. Kerr (2022), The Great Resignation Didn’t Start with the Pandemic, https://hbr.org/2022/03/the-great-resignation-didnt-start-with-the-pandemic.

[6] Griffeth, R. (2000), “A meta-analysis of antecedents and correlates of employee turnover: update, moderator tests, and research implications for the next millennium”, Journal of Management, Vol. 26/3, pp. 463-488, https://doi.org/10.1016/S0149-2063(00)00043-X.

[4] Hahn, J., H. Hyatt and H. Janicki (2021), “Job ladders and growth in earnings, hours, and wages”, European Economic Review, Vol. 133, p. 103654, https://doi.org/10.1016/J.EUROECOREV.2021.103654.

[24] Hyatt, H. and J. Spletzer (2013), “The recent decline in employment dynamics”, IZA Journal of Labor Economics, Vol. 2/1, pp. 1-21, https://doi.org/10.1186/2193-8997-2-5/FIGURES/6.

[9] Jäger, S. and J. Heining (2022), “How Substitutable Are Workers? Evidence from Worker Deaths”, Working Paper, No. 30629, NBER, Cambridge, MA, http://www.nber.org/data-appendix/w30629.

[27] Kline, P. et al. (2019), “Who Profits from Patents? Rent-Sharing at Innovative Firms”, The Quarterly Journal of Economics, Vol. 134/3, pp. 1343-1404, https://doi.org/10.1093/QJE/QJZ011.

[5] Kuhn, P. and L. Yu (2021), “How costly is turnover? Evidence from retail”, Journal of Labor Economics, Vol. 39/2, pp. 461-496, https://doi.org/10.1086/710359/SUPPL_FILE/19260DATA.ZIP.

[14] Maestas, N. and J. Zissimopoulos (2010), “How Longer Work Lives Ease the Crunch of Population Aging”, Journal of Economic Perspectives, Vol. 24/1, pp. 139-60, https://doi.org/10.1257/JEP.24.1.139.

[10] Manning, A. (2011), “Imperfect Competition in the Labor Market”, Handbook of Labor Economics, Vol. 4/PART B, pp. 973-1041, https://doi.org/10.1016/S0169-7218(11)02409-9.

[2] Manpower Group (2020), Closing the Skills Gap: What Workers Want, Talent Shortage 2020, https://workforce-resources.manpowergroup.com/white-papers/closing-the-skills-gap-know-what-workers-want#main-content (accessed on 25 August 2022).

[29] Molloy, R. et al. (2016), “Understanding Declining Fluidity in the U.S. Labor Market”, Brookings Papers on Economic Activity, Spring 2016, https://www.brookings.edu/bpea-articles/understanding-declining-fluidity-in-the-u-s-labor-market/.

[20] OECD (2022), OECD Employment Outlook 2022: Building Back More Inclusive Labour Markets, OECD Publishing, Paris, https://doi.org/10.1787/1bb305a6-en.

[7] OECD (2020), Promoting an Age-Inclusive Workforce: Living, Learning and Earning Longer, OECD Publishing, Paris, https://doi.org/10.1787/59752153-en.

[18] OECD (2019), OECD Employment Outlook 2019: The Future of Work, OECD Publishing, Paris, https://doi.org/10.1787/9ee00155-en.

[17] OECD (2018), OECD Employment Outlook 2018, OECD Publishing, Paris, https://doi.org/10.1787/empl_outlook-2018-en.

[11] Porter, C. and J. Rigby (2021), “The turnover contagion process: An integrative review of theoretical and empirical research”, Journal of Organizational Behavior, Vol. 42/2, pp. 212-228, https://doi.org/10.1002/JOB.2483.

[3] Topel, R. and M. Ward (1992), “Job mobility and the careers of young men”, Quarterly Journal of Economics, Vol. 107/2, pp. 439-479, https://doi.org/10.2307/2118478.

Notes

← 1. However, not all job to job transitions will entail workers moving to more productive firms – the fraction of “productive moves” will vary enormously across countries and sectors – therefore a highly fluid labour market is not necessarily a sign of economy-wide productivity gains (Albagli et al., 2021[26]).

← 2. They use unexpected deaths of workers as a source of exogenous variation. Studies trying to estimate replacement costs face the problem that employee turnover and productivity might be correlated not because turnover reduces productivity, but because low productivity causes employees to leave.

← 3. These results are similar to those found by Kline et al. (2019[27]).

← 4. OECD Population projections database: http://stats.oecd.org//Index.aspx?QueryId=88954.

← 5. This is consistent with previous OECD research (OECD, 2019[18]), upon which this section builds. To avoid showing changes based on a single year, changes are based on the average of three years, for example the change between the average value of 2012-14 and the average value of 2017-19.

← 6. The adjustment is obtained by means of regression analysis that identifies changes in job tenure between an average of 2012-14 and an average of 2017-19, controlling for workers’ age, gender and level of education. Changes in the composition of the workforce need to be taken into account as they may obscure how job tenure is changing for specific groups of workers by age, gender and education. For example, job tenure typically rises with age and so average tenure will tend to increase as the workforce ages with baby boomers moving into the older age groups. Average job tenure rises because there are fewer workers at younger ages with lower tenure and more workers at older ages with higher tenure. Similarly, tenure rises with educational attainment and so average tenure will mechanically increase as the education level of the workforce rises. Thus, average tenure without any adjustment for compositional changes may increase even if tenure is falling for workers at each age and education level.

← 7. Job stability can be measured in several ways, for example the average (or median) elapsed tenure of all current jobs, retention rates for five-year age groups, or the probability that a particular job will continue to exist. No single measure gives a perfect indication of the degree of job stability. Average or median tenure is the most common measure, but because it is based on all existing jobs, we do not know how long these jobs will last for.

← 8. Alternative data for example from the LEHD (Longitudinal Employer-Household Dynamics) and JOLTS (Job Openings and Labor Turnover Survey) shows that turnover is significantly higher. Total turnover from LEHD for 2017-2019 (average) was 61% made up of job-to-job flows of 22%, separations to non-employment of 18% and hires from non-employment of 20%.

← 9. The available measures of job-to-job transitions capture movements between jobs across consecutive calendar years. Transitions from one year to the next may in fact conceal one (or several) transitions in and out of employment that occur over the course of the reference year. For example, a worker who was employed in the previous year may have experienced a period out of employment before finding the current job but will be recorded has having experienced a job-to-job transition since job status is recorded yearly. The measure, therefore, should not be interpreted literally as an indicator of direct transitions from a job to another, but rather as a measure of transitions that involve short periods of non-employment between jobs. In the United States, the rate at which workers change jobs has declined since the 1980s, at least up until the early 2010s (e.g. Davis, Faberman and Haltiwanger (2012[23]), Davis and Haltiwanger (2014[25]), Hyatt and Spletzer (2013[24]), Molloy et al. (2016[29]). Job-to-job flows data from the LEHD (Longitudinal Employer-Household Dynamics) show that job-to-job flows have increased by 3.5 percentage points between 2012/14 and 2017/19. Autor, Dube and McGrew (2022[28]) also show an increase in job-to-job flows since pre-pandemic trends.

← 10. U.S. Bureau of Labor Statistics.

← 11. These data from the European Union Labour Force Survey show the main reason for leaving, hence a respondent can only select one response category. Therefore for example, someone who retires before the state pension age may select early retirement as the reason for leaving their current job, but they may also have health problem(s) which may in fact be the real reason for retiring.

← 12. In France, education and other reasons are not separated.

Disclaimers

This work is published under the responsibility of the Secretary-General of the OECD. The opinions expressed and arguments employed herein do not necessarily reflect the official views of the Member countries of the OECD.

This document, as well as any data and map included herein, are without prejudice to the status of or sovereignty over any territory, to the delimitation of international frontiers and boundaries and to the name of any territory, city or area.

The statistical data for Israel are supplied by and under the responsibility of the relevant Israeli authorities. The use of such data by the OECD is without prejudice to the status of the Golan Heights, East Jerusalem and Israeli settlements in the West Bank under the terms of international law.

Photo credits: Cover © VectorMine/Shutterstock.com.

Corrigenda to publications may be found on line at: www.oecd.org/about/publishing/corrigenda.htm.

© OECD 2023

The use of this work, whether digital or print, is governed by the Terms and Conditions to be found at https://www.oecd.org/termsandconditions.