Human resource (HR) measures to improve job satisfaction can prolong working lives, yet there is little evidence about the size of such effects. We provide a comprehensive literature review and an analysis of the link between the length of the remaining working life of older workers and a number of indicators of job satisfaction for 12 European countries, using the panel dimension of the Survey of Health, Ageing and Retirement in Europe (SHARE).

It is a well-recognized fact that European societies are ageing and that this creates challenges for traditional welfare systems. In many countries, the financial crisis of 2007-08 and the subsequent recession exacerbated a political and financial crisis within pension systems at a time when large numbers of the post-WWII baby-boomer generation started to retire. A remedy promoted throughout Europe, and by the European Commission in particular, is to extend working lives and promote work beyond the statutory retirement age.

Poor working conditions and low job satisfaction tend to increase workers’ propensity to retire (Andersen et al., 2007; Borella & Mascarola, 2009; D’Addio et al., 2010; Thorsen, 2012). In this paper, we estimate the additional time in work before retirement that results from greater job satisfaction. Company policy, we argue, can affect job satisfaction, which makes exploring the latter’s potential for extending working lives a worthwhile endeavour. This said, we do not claim that all working conditions seemingly connected with job satisfaction are under the purview of company policy. Many are subject to extraneous factors, such as the job insecurity resulting from economic crisis. Nonetheless, there are aspects of job satisfaction that organisations can shape, perhaps significantly, for instance, by adopting progressive age management strategies or working arrangements that promote physical and emotional wellbeing in the workplace. By way of shorthand, we refer to such practices as human resource (HR) policies.

Our data source is the Survey of Health, Ageing and Retirement in Europe (SHARE). We use data from wave 1 (2004), 2 (2006/7), 4 (2010/11), 5 (2013) and 6 (2015) for 12 countries that were covered in multiple waves.1 We use ten questions (EP026-EP035) asked on job satisfaction and working conditions in each of the four waves. The longitudinal dimension of SHARE is important for determining the link between working conditions and future retirement. Our sample is constructed by linking working conditions in the wave when an individual is first observed to the retirement status of the same individual in the wave in which the person is last observed. If the person retired in the last wave of observation, we further observe the year of retirement. This procedure leaves us with a data set that is derived from longitudinal data but has the structure of a cross-sectional data set with one observation per individual containing (among other variables) year, age when first observed, number of years worked between first and last wave the individual was observed and an indicator of whether the observation is top coded (i.e. the individual is still working when last observed). In econometric terms, we use a modified tobit approach (Wooldridge, 2009; Roodman, 2009) which adjusts the estimates for the observations that are top coded.

In examining the link between HR policy, job satisfaction and the act of retiring, our analyses first show a very high level of job satisfaction among older workers. While there are differences across countries, they are minor, as are variations between the different years for which data are analysed. Second, we find that job satisfaction is closely associated with the positive assessment of a range of workplace and working conditions typically within the purview of HR and business management, such as workload or pay. Third, our econometric estimations find that greater job satisfaction contributes to a delay in retirement, but the added time in work before retirement is relatively short, averaging around three additional months. In light of this evidence, we conclude that there may be very limited scope for HR policy to contribute directly to extending working lives: its main tool for doing so, namely the creation of working conditions associated with high levels of job satisfaction, appear to be already well deployed; greater job satisfaction among (older) workers would – on average – add little in terms of delaying retirement.

Table 1. Effects on length of working life from optimal HR policies
Source: Authors’ calculation based on country specific regressions (detailed results available upon request). Age group refers to age when first interviewed. ‘Estimated’ denotes the estimated average number of working years for the relevant age group given observable characteristics. ‘Counterfactual’ refers to the predicted number of years worked if all individual are assigned ‘very satisfied’ with their job

This paper has presented new evidence of the link between workplace and working conditions typically under the purview of HR policy on the one hand, and job satisfaction and associated retirement behaviour on the other. Its sobering conclusion is that, while job satisfaction among older workers is at a high level, the currently typical timing of retirement among highly and less job-satisfied older workers suggests little scope for extending working lives through higher job satisfaction.

Retirement decisions are the results of complex and multifaceted considerations, options and opportunities, as well as unforeseen events, ranging from the personal (e.g. suddenly deteriorating health) to the collective (e.g. business closure). Moreover, retirement decisions continue to be largely framed by legislation, the presence and the extent of the ‘enforcement’ of statutory retirement ages. A loosening of such regulations, such as the abolition of the default retirement age in the UK and previously the US may, however, lead to change in retirement and, importantly, employer hiring behaviour. Raising the statutory retirement age may be a more effective measure for extending working lives.

Estimates of this kind are necessarily diverse and uncertain, but suggest that the potential effects of HR measures are, given a choice of alternative tools, comparatively small. Of course, this should not deter efforts to create better and more age-friendly workplaces.

The article is based on:

Mikkel Barslund, Jürgen Bauknecht, Andreas Cebulla, Working conditions and retirement: How important are HR policies in prolonging working life? in: mrev management revue, Seite 120 – 141 mrev, Jahrgang 30 (2019), Heft 1, ISSN print: 0935-9915, ISSN online: 0935-9915,

About the authors:

Mikkel Barslund, Centre for European Policy Studies, Brussels, Belgium.

Jürgen Bauknecht, Fliedner Fachhochschule Düsseldorf | University of Applied Sciences, Düsseldorf, Germany.

Andreas Cebulla, SA Centre for Economic Studies, University of Adelaide, Adelaide Australia and National Institute of Economic and Social Research, London, United Kingdom.