Background and Motivation

Living longer can be considered one of the main achievements of modern societies. In developed countries, longer life expectancy was, however, accompanied by a significant fall in the effective age of retirement, partly driven by older workers leaving the labour market following recessions in the 1970s, 1980s and 1990s (see Fig. 1). Many developed countries have since introduced reforms and participation rates of older workers have increased, especially amongst those countries at the forefront of reform (Sonnet et al. 2014). These reforms have encompassed changes in employment regulations, recruitment and retention practices, social insurance, and tax and pension systems. Together such reforms are helping to prolong working life and their wider diffusion and adoption will be a powerful tool in tackling the challenges faced in rapidly aging societies.

Fig. 1
figure 1

Average effective age of retirement in OECD countries, 1970–2016 (by gender). Note: The average effective age of retirement is calculated as a weighted average of withdrawals from the labour market over a 5-year period. Source: OECD – Ageing and Employment Policies: Statistics on average effective age of retirement, available at: http://www.oecd.org/els/emp/average-effective-age-of-retirement.htm

Full size image

In this context, obvious questions concern the identification of measures that are effective in encouraging workers to work longer, and employers to offer jobs to older workers. In other words, what reforms encourage workers to stay in employment: More flexible work arrangements? Better skills? Better employment services? And similarly, what will encourage employers to keep the elderly on the job: Subsidized training of older workers? Promotion of life-long learning, to increase the learning ability of older workers? Information campaigns aimed at changing outdated attitudes?

The objective of this paper is to review findings on the design and effectiveness of selected employment-related measures aimed at increasing the employment of older workers in EU and OECD countries, and on that basis, identify best practices. The paper focuses on measures which help overcome worker- and employer-based barriers to the employment of older workers. Among the measures addressing worker obstacles, the paper concentrates on the promotion of training, on adjusting employment services and active labour market programs (ALMPs) for older workers, and on promoting better working conditions; among the measures addressing employer obstacles, it focuses on adjusting employment protection rules for older workers, on subsidizing wages of older workers, and on challenging employers’ negative perceptions. In addition, to offer a perspective into more holistic approaches that countries have applied, the paper provides three case studies: the United Kingdom’s New Deal 50 Plus, Germany’s Perspective 50 Plus, and Canada’s Targeted Initiative for Older Workers.Footnote 1

The organization of the paper is as follows. We first review the experience with approaches and initiatives addressing obstacles for increasing employment on the side of older workers themselves (Section 2). We proceed with a similar review of the obstacles that can be attributed to the side of employers (Section 3). We then provide a more holistic experience with employment programs targeted on older workers by presenting three case studies, featuring the recent experience in United Kingdom, Germany and Canada (Section 4). The final section concludes by providing conclusions and policy recommendations (Section 5).

Addressing Worker Obstacles

Some of the impediments for late-life employment engagements can certainly be attributed to older worker themselves. These impediments include low skill levels, reduced cognitive abilities – and perhaps low motivation – that may affect the effectiveness of training and employment services provided to older workers, productivity-wage gap that could be increasing with age, and preferences of older workers to work under flexible work arrangements.Footnote 2 Below we discuss these issues under three headings: promoting training of older workers, adjusting employment services and ALMPs for older workers, and promoting better working conditions.

Promoting Training of Older Workers

There are several market failures that may lead firms to underinvest in training. First, training may increase a worker’s value not just in the firm that paid for the training but in other firms as well. Consequently, a firm that invests in training risks losing its investment to its rivals. The better strategy may be for the firm to free ride on its rivals’ investments by attracting their trained workers without having to invest themselves (the poaching argument). The free-rider problem causes all firms to reduce their investments in training. Second, if training is subject to returns to scale, individual firm training investments will be costlier than those provided by a single provider. Third, the under-provision of training may be related to informational imperfections and asymmetries (UK Commission for Employment and Skills 2012). In the absence of reliable information on the returns to training, firms may prefer to be cautious and provide a sub-optimal level of training. Furthermore, even when they are confident of positive returns to training, financing such training may be difficult due to credit constraints: potential lenders who may be needed to finance the training may lack the necessary information to calculate the training payoff and may thus be hesitant to lend.

Broadly speaking, there are three approaches addressing the tendency to underinvest in training by firms: subsidies, training funds, and individual learning accounts. Subsidies – cash grants and tax incentives – provide additional funds or reduce obligations of firms, conditional that firms increase the volume of their training. Sectoral training funds allow all the firms that are using a particular type of skills to agree to pay equally for training through the payroll tax and then share in the benefits of trained workers. This coordination forces all firms to internalize the training externality arising from labour mobility.Footnote 3 Individual learning accounts provide another modality to incentivise individuals to undertake training, the one that may leave more room for their discretion. These approaches are considered below; in addition, we also summarize two studies focusing on the impact of retirement policies, in particular changes in retirement age, on the incidence of training of older workers.

Tax incentives

Customized versions of tax incentives aimed at increasing firm-based training are used by many countries (according to CEDEFOP (2009a), such taxation schemes exist in 15 EU member states). For example, in the Netherlands firms can claim 120% of their training expenditure as a tax deduction – and, for workers who are at least 40 years old, an additional 20% percentage points. Similarly, in Austria firms can deduct actual training costs from their tax base for a wide range of training activities (OECD 2004) and are, in addition, given either a direct subsidy of 20 percent of training cost or a tax credit amounting to 6% of training cost. Certain strings are attached to these benefits (in terms of the choice of implementation agencies and overall size of training). A tax credit applies also for France’s small and medium size firms, where it covers ten hours of training per employee at an hourly cost of up to 75 EUR and with a ceiling cap of 5000 EUR for a firm within a two-year period. Similar tax incentive was ongoing between 1999 and 2001 in the Belgian Flemish region, where firms’ training expenditures were deducted from the property tax payment, subject to certain conditions (chief of them being that training expenditures increased in comparison to the preceding year).

While rigorous evaluations of tax incentives to stimulate employer-based training are virtually non-existent, in a rare example of such a study Leuven and Oosterbeek (2004) find only little if any increase in training, partly on the account of substitution effects.Footnote 4 As mentioned above, in addition to claiming 120% of their training expenditure as a tax deduction, Dutch firms can increase their tax deduction by an additional 20 percentage points for training workers who are at least 40 years old. Exploiting this discontinuity, Leuven and Oosterbeek show that the Dutch scheme increased training rates of workers aged just above 40 by 17% while, simultaneously, reducing training rates of those aged just below 40 (their training rate dropped, although insignificantly, by 4.7%). Thus some of the positive impact is based on substitution or postponement of training for workers aged just below 40, which led the authors to question whether the overall effect of training provision is positive, a result they describe as “discouraging” for policy makers. The authors could not confirm any effect of training on wage rates.

Summarizing the evidence, Müller and Behringer (2012) claim that tax incentives are not an effective means of stimulating additional training, including training of older workers, and are accompanied by deadweight and substitution effects. Moreover, they argue that tax schemes often fail in targeting specific groups of firms or workers as well as in stimulating aggregate demand for training. It is also noteworthy that in several countries tax incentives have failed because they were poorly utilized by firms (for example, in Brazil and Chile, see Johanson 2009).

Direct subsidies via financial grants or vouchers

To stimulate training, many countries subsidize training via financial grants to firms. For example, Ireland subsidizes firms that prove the need for training. In the 1990s, the subsidy (Training Support Scheme) was graduated by firm size. Firms with less than 50 employees could receive a grant of 65% of training costs, firms with 51–200 employees up to 50%, and firms with 201–500 employees up to 25%. The evaluation of the scheme reported by Müller and Behringer (2012) found that the scheme was popular among the participating firms both because firms could choose the training providers and because they could adjust these grants to suit their needs. While 55% of small firms stated that they would have offered no or less training in the absence of subsidies, only 27% of large firms stated the same, indicating that large firms may have profited from the scheme more than proportionally.

In Eastern Europe training grants were carried out by Lithuania and Poland. In Lithuania, grants for wide range of disadvantaged groups existed until 2008. While employers – recipients perceived the program positively (above all, they claimed that the newly acquired skills could be applied immediately), less desirable aspects of the program included (i) disproportional amounts of grants benefitting large firms and workers with higher education, (ii) failure of the program to influence career and salary of trainees, and (iii) subsidized training substituting the unsubsidized training.

Training grants implemented in Belgium, the United Kingdom and the United States appear to have stimulated firm training. In Belgium, public funding leveraged by far larger firm-based funding of training programs and resulted in enhanced employability of workers. In the United Kingdom, several pilot projects – consisting of, for example, supporting a “Training Champion” and financing part of training costs – were well-received among the participating firms. Similarly, training matching grants programs in the United States (Massachusetts and Michigan) resulted in increased level of training, although with no longer-term effects.

In a rare example of rigorous evaluation, Singer and Toomet (2013) analysed German government subsidized job training program for workers older than 45 years that paid the full amount of training costs. The results show that two years after beginning the program, participants had on average 2.5 percentage points higher probability to be in paid employment than non-participants.

Training vouchers are another way of paying training subsidies. Possible advantages of this instrument are lower bureaucratic burden, small threshold needed for the firm to qualify for subsidized training, and better targeting capacity. Training vouchers for employers are available in several European countries, including Belgium, Germany, France, Italy and Nederlands. Görlitz and Tamm (2016) analyse the impact of vouchers on employment and gross monthly income in the German voucher program that reduced training costs by up to 50% but only up to 500 EUR per work-related training course. They find that the programs had an insignificant effect on gross monthly income, on employment status and on the duration of employment and unemployment. Similarly, Hidalgo et al. (2014) investigate the effect of vouchers in the Netherlands and report a negligible impact on wages and mobility but significantly positive on training participation.

In summary, as argued by Müller and Behringer (2012), direct subsidies may be more effective than tax incentives and may facilitate the targeting of specific groups of firms and employees. They may also entail smaller deadweight losses. On the other hand, systematic targeting may be administratively complex, so reducing deadweight losses may entail increasing administrative burden. Müller and Behringer (2012) also summarize policy lessons as follows: (i) the program must be communicated effectively, (ii) the focus should be on small businesses since they are usually less likely to train; (iii) administrative burden must be reduced to attracting firms to participate, (iv) incentives should be substantial and well-tailored to employer needs and, last but not least, firms themselves should co-finance training to reduce deadweight effects (public resources paying for training the firms would undertake in the absence of the subsidy).

Training funds

Training funds are another approach to address the under-provision of training resulting from a free-rider problem consisting of poaching trained workers from other firms. Typically, training funds collect their resources via a levy (that can be mandatory or agreed by collective agreements) and use these resources to stimulate training and/or other activities contributing to skills development, typically via partial reimbursement of training costs. For example, the sectoral training agreements allow all the firms using a particular type of skill to agree to pay equally for training through the payroll tax and then share in the benefits of trained workers. The coordination forces all firms to internalize the training externality arising from labour mobility. To stimulate (additional) training, funds are reimbursed in proportion to the level of training employers provide to their workers. In his review of training funds, Johanson (2009) lists 62 countries with training funds worldwide.Footnote 5

Options for the design of training funds are numerous. Recipients of funds may be employers or workers; funds may be available for training of all workers or targeted to specific groups (including older workers); and funds’ coverage may be universal or sectoral. The mechanisms used to stimulate training are also diverse: a cost reimbursement scheme involves reimbursement of the cost incurred for training, usually in a fixed proportion; a levy-grant scheme provides resources conditional on meeting certain criteria (such as adopting a systematic training approach), and a levy-exemption scheme allows firms to reduce or eliminate their levy obligations by the amount of training provided.

One of the most well-known examples of training funds is France, where such funds date back to the 1920s. As agreed among social partners, unless spending a required amount on training themselves, firms pay a training levy that is used in a variety of ways to enhance skill formation of various actors, organized on regional and industry levels. Training funds exist also in the Netherlands, where in 2009 there were 135 funds under around 200 sector collective agreements. The levy varies from 0.5–2%, and in 2010, it was on average 0.41% of the total wage bill. Despite a large amount of funds there is no evidence about higher training levels in the sectors with funds than it is in sectors without funds (Müller and Behringer 2012, Kamphuis et al. 2010). In contrast, training funds in United Kingdom were abolished. A levy-grant scheme existed during from the 1960s to the 1980s, but due to the dissatisfaction of employers, the program was abolished in most sectors because, according to Müller and Behringer (2012), employers perceived the government as interfering unduly and the levy as an additional cost. However, the government introduced an apprenticeship levy for all employers in April 2017 (UK Department of Education 2017).

It is startling that no rigorous evaluation of “cause-impact” type has been carried out on training funds, so whether these programs “work”, that is, whether they increase the volume of training, is an open question. Existing review papers admit that the evaluation of training funds is patchy and that most evaluations simply contrast outputs and targets without assessing the net impact. Thus these reviews either fail to provide an assessment about the effectiveness and efficiency of training funds (Müller and Behringer 2012; Johanson 2009) or provide a positive answer without accurate justification (Dar et al. 2003). We are aware of only one analytical study of training funds – the study of Kuku et al. (2016) on Mauritius. The study provides evidence that, in the case of Mauritius, the training subsidy does increase training for the smallest firms, but that for medium and large firms, the adverse effect of the levy on training outweighs the incentive effect of the subsidies.

Being left with the casual evidence, three conclusions can be reached. First, large firms benefit more than small firms. That large firms use training funds more intensely than small firms is a well-known fact (established for France, Mauritius, Netherlands, and Singapore); however, the study on Mauritius points out that the program atypically benefits the smallest firms (with the smallest firms receiving more benefits than they are paying in as levy), because they can claim proportionally much larger shares of the training costs. Second, administrative hurdles can be overbearing – and, indeed, prohibitive. Müller and Behringer (2012) claim that the bureaucratic burden diminishes the funds available for reimbursement, leads to deferral of training and, particularly in the case of small firms, deters firms from filing reimbursement claims. And third, success factors are hard to pinpoint. Unsurprisingly, given the lack of proper, “cause-effect” evaluations of training funds, it is difficult to derive general success factors of training fund programs, and often the importance of country-specific conditions is emphasized. Nonetheless, employer buy-in is often cited as imperative for the success of these programs.

To summarize, although the idea of training funds is appealing, the evidence shows that the implementation of this idea encountered numerous obstacles. Dar et al. (2003, p.6) thus conclude that “while these schemes have, in general, had a positive impact on increasing in-service training, they have been inequitable – large employers have benefited to a greater extent than small or medium-size employers. Employer reaction to these schemes has been mixed, with most (especially the smaller ones) feeling that the levy is an additional tax that has been imposed on them unjustifiably. Problems associated with administering the fund and problems of non-compliance abound, especially with such schemes in developing countries.”

Individual learning accounts

Whereas other schemes focus mainly on encouraging firms to invest in their workers training, Individual Learning Accounts (ILAs) are targeted directly towards individuals and aim to give them more autonomy in their upskilling decisions. While implementations vary across countries, ILAs generally work as follows: eligible individuals open a bank account, generally at a government entity created specifically for such purposes, with the goal of saving financial assets for the purpose of training. To stimulate their savings, the government provides incentives in the form of tax deductions, subsidized interest rates, or direct financial top-ups. The government generates a list of authorized providers of training or education, sometimes limiting the fields of study or training based on perceived labour market needs. Individuals are generally given considerable discretion in the use of funds in their ILAs within the list of pre-approved providers. However, the government may choose to influence the choice of training directly, e.g. by employing counsellors who can veto an individual’s enrolment in a particular program, or indirectly, through additional financial incentives for utilizing certain training programs.

ILAs first emerged in the late 1990s and have been implemented in several countries, either as full-fledged programs or as pilot projects, including Austria, Belgium, Canada, Ireland, Italy, the Netherlands, the United Kingdom, and the U.S. When the UK implemented its ILA scheme in 2000, each individual, upon opening a learning account with a minimum deposit of 25 GBP, received a government grant of 150 GBP (Schuetze 2007). In a trial program in the Netherlands, each participant received a grant of 450 EUR, to be spent in 18 months (CEDEFOP 2009b). In Sweden, individuals receiving training in lasting five days or longer received a tax reduction and a grant of 1000 SEK, or approximately 100 EUR (Schuetze 2007; CEDEFOP 2009b).

The available evidence shows that ILAs can boost training participation (Austria, Netherlands) and that they also may improve employment outcomes (United Kingdom, U.S.); in contrast, in both the United Kingdom and U.S. earnings impacts of such schemes have been found insignificant or small. Below we summarize two studies – on the U.S. and the Netherlands – that use methodologically sound approaches.Footnote 6

Under the ILA program that was introduced in the U.S. in 1998, jobseekers can select training from a wide array of state-approved programs and providers, and the state can design the program with varying intensity of counsellor interference in the choice of training and with various designs of the reward structure. A Perez-Johnson et al. (2011) study of a controlled experiment under this program provides interesting policy insights on fundamental choices in designing individual learning accounts. First, mandatory, intense counselling discourages participation in training, but for those that participate in training, such counselling tends to raise their wages. One interpretation is that intense counselling is successful in discovering and building upon comparative advantages of jobseekers and thus in directing jobseekers to types of training that lead to higher paying jobs. Second, when counselling is voluntary, few customers request it, but their likelihood of training is higher than under mandatory (and intense) counselling. But perhaps as a consequence of lack of counselling, such customers are associated with higher likelihood of working in an occupation for which they have not trained. And third, program design does not affect employment status following the completion of the program.

Renkema (2006) studies the impact of ILA in Netherlands on the learning intentions of older workers. The study uses quasi-experimental design to assess the effect of ILAs on learning intentions among worker in small- and medium-size firms in the technical installation and elderly care sectors (sectors with a large share of low-skilled and older workers) in the Netherlands. Individuals learning accounts were found to be associated with significant impact on learning intentions for older workers and experienced workers in the technical installation, but not in the elderly care sector. One of the reasons for such findings could be in the fact that technical installation sector is also more subject to innovations than elderly care sector.

To summarize, ILAs can be a useful policy tool to encourage lifelong learning and are particularly suited for targeting workers in non-traditional or less stable forms of employment. The measures discussed in previous sections – tax incentives, subsidies or training funds – generally require participation of the employer, which means that they are less likely to help certain types of workers, such as the self-employed or those on fixed-term contracts who frequently switch employers. Other active labour market policies which target individuals often require the individual to be registered as unemployed. ILAs, by contrast, can be administered directly to individuals who are working but who would benefit from additional training. As workers increasingly are engaging in such forms of work, ILAs are an option that warrants consideration.

Raising the retirement age and the incidence of training

Raising the retirement age has shown to be instrumental in increasing the incidence of training among older workers. In 2006, the Netherlands abolished its early retirement option for public sector workers born in 1950 or later. Montizaan et al. (2010) studied the effect of this reform on training participation of older workers. The control group consisted of those who retired at age of 62 years and 3 months and had a pension replacement of 70% (under the pre-reform regulations), and the treatment group of the same age individuals in the post-reform period. Had the treatment group retired at that age, their replacement rate would be 64% and to receive a 70% replacement rate, they would have to postpone their retirement until the age of 63 years and 4 months. Montizaan et al. (2010) find a significantly positive increase of the training participation of 5.7% for workers in large firms and insignificant difference for workers in small firms. Moreover, the effect was larger for low-educated workers (they were 19.5% more likely to participate in training relative to the control), and non-significant for workers with higher education. Interestingly, the reform also significantly increased participation in long training courses (those lasting more than 48 h).

In line with the above results, generous early retirement policies are found to discourage training of older workers and flexible retirement policies to stimulate it. Fouarge and Schils (2009) investigate training differentials in employed individuals on a threshold of 55 years (groups 25–44, 45–54, and 55–64) in 13 European countries based on underlying retirement schemes.Footnote 7 Calculation of generosity indicator is based on implicit taxes on continued employment and replacement rate while calculation of flexibility indicator is based on strictness to entitlement (minimum age and minimum contribution period). The results show that training participation of workers in the age group 45–54 is on average lower by 3.2% and for workers in the age group 55–64 the participation is lower by 5% relative to workers in the age group 25–44. Additionally, older workers with lower education participate in training significantly more often relative to workers with average education. Generosity of retirement scheme does play important role. An increase of generosity indicator by one point will reflect in decrease in training participation for workers in the age group 45–54 on average by 4.5% and in training participation of workers in the age group 55–64 on average by 5.2% relative to the 25–44 age group. In contrast, flexibility retirement scheme does have positive impact on training participation. An increase of flexibility indicator by one point will on average increase training participation by 3.8% for workers in the age group 45–54 and on average by 6%, although not statistically significant, for workers in the age group 55–64.

Adjusting Employment Services and ALMPs for Older Workers

Given the specific needs and constraints faced by the older workers, OECD (2006) argues that may be scope for specialized ALMPs for older workers. Our literature search yielded only limited studies along these lines. Below we summarize a line of research on how training of older workers should be adjusted to suit the needs and abilities of older workers, and on the type of ALMPs found to be most effective for older workers. Further evidence on how employment services and ALMPs are adjusted for older workers, in the context of more complex, holistic nation-wide programs, is provided in cases studies below.

Training of older workers is undoubtedly one of the ways to foster their productivity and thus employability. Indeed, Picchio and van Ours (2013) showed for the Netherlands that older workers who receive on-the job training were more likely to retain employment. But the question arises: given the inevitable decline in cognitive skills that accompanies aging, how effective is training of older workers? Research on learning of older adults indicates that they are able to acquire new skills – but that training has to be adjusted to the needs and capabilities of older workers.

According to Picchio (2015), experiments show that older workers get lower test scores and take more time to complete training than younger workers. Age differences in training performance are more pronounced when people are trained in new tasks; when training content requires more problem solving; when training is more abstract and formal; and when training is on information and communications technology.

But research shows that older workers, including those experiencing cognitive decline, can learn new skills with long-lasting training effects. Picchio (2015) thus claims that training methods for older workers should be tailored to their learning needs and abilities. First, older learners should have ample time to process new knowledge. Second, integrating multiple instructional methods (combining lecturing with active participation, for example) can improve the performance of older workers. Third, older workers’ training should be self-paced, allowing enough time to compensate for the decline in cognitive abilities, internalize the importance of training, and understand and master the training content. And fourth, older adults perform better if training builds on pre-existing knowledge, that is, if it is on-the-job, applied, or based on communication and management skills.

As for selecting the most effective type of ALMPs for older workers, a meta-study of Kluve (2017) yields very interesting and relevant results. Consistent with the above finding of Picchio (2015) about lower training performance of older workers, Kluve (2017) finds that job search assistance programs are the most suitable type of ALMPs for older workers. Moreover, Kluve (2017) reports that the impact of ALMPs is larger and more likely to be positive in the medium and long run (the result that applies for all age groups), and that among older workers, the impact is the largest for long-term unemployed.

Promoting Better Working Conditions

Promoting flexible work arrangements, including part-time jobs, is a way to both extend the length of working lives as well as to increase workers’ well-being. Graham (2014) argues that by providing social contacts and interactions, personal growth, autonomy, and a sense of purpose, voluntary late-life work has positive effects on well-being. But she stresses the importance of choice. Retiring sooner than one would like, and also remaining employed longer than one would like, has negative effects on well-being. The benefits of late-life work are linked to a person’s ability to choose to remain in the labour force and under conditions considered to be acceptable. Such benefits can be achieved, among others, by promoting part-time work by older workers, whereby they can gradually reduce their working hours while remaining in the labour force.

In line with the above reasoning, Širok (2011) argues that the extension of working career in Slovenia is importantly linked to flexible work arrangements. He finds such flexibility underlying all stages of the process of transition of Slovenian workers to inactivity, from early withdrawal from the labour market to retirement and to continuation of work activities in retirement. Based on the analysis of successive labour force surveys from January 2005 to June 2006, he finds that (i) work activity of older workers, including of retired persons, takes place predominantly in informal types of employment, as well as in sectors that allow informal employment (such as agriculture), (ii) conversely, the lack of work flexibility of blue collar workers is reflected in their retirement at early age as well in the abrupt (»overnight«) fashion of their retirement, and (iii) early reduction of work activity is associated with inflexible forms of employment and low integration in the employment market. He also finds that flexibility of work arrangements importantly determines the speed of transition to inactivity.

What are the options for promoting flexible work arrangements? OECD (2006) argues against retirement conditions that discourage part-time work, in particular “cliff-edge” patterns of retirement in which individuals shift instantly from working full-time to working not at all. The report also notes that barriers to part-time work may arise if there are substantial fixed costs for each person employed irrespective of the number of hours they work which would make it more expensive for an employer to hire two part-time workers rather than one full-time worker. Both Austria and Germany have introduced an old-age part-time employment scheme which allows workers to switch from full-time to subsidised part-time work over a specified period. But OECD (2006) notes that these schemes had mostly been used as de facto early retirement schemes rather than as gradual retirement schemes, as workers chose to work full-time over the first half of the period and then stopped working altogether over the second half of the period.

Addressing Employer Obstacles

The other side of the equation of increasing employment of older workers relates to employer obstacles. Employer may resent hiring older workers for various reasons, and below we address the international experience in dealing with the following three issues: employment protection rules that favour older workers, subsidizing wages of older workers, and challenging employers’ negative perceptions about older workers.

Adjusting Employment Protection Rules for Older Workers

Policies hindering lay-offs of older workers may have positive effects on keeping currently employed workers but may have side effects on workers that are just before the threshold and may hinder the access to jobs by the unemployed. In 1987, France introduced such a policy requiring firms to pay a tax to the unemployment insurance system if they fired a worker aged above 50 years. The amount of the tax was a function of previous monthly gross wages. In 1992, the policy was changed and workers that were older than 50 years at the time of hiring were exempted from the policy. Behaghel et al. (2008) studied the effects of the 1992 change of the law and found that the exit rate from unemployment to employment for those just below the age threshold decreased by 53% for males and by 42% for females (although the latter was not statistically significant). However, the effect is much smaller for transitions under the fixed-term contract. The effect for men with same specifications transition rate decreased by 6.7% and for females by 11.3%, for both groups the effect was insignificant.

In line with the above results, abandoning the lay-off prohibition of older workers in Slovenia positively affected their probability to be hired under the permanent contract both from unemployment and from another permanent job. With the goal of increasing employment opportunities for older workers, the 2013 Employment Relations Act stopped granting the dismissal protection to workers who, at the time of hiring, already past the threshold of protection dismissal (however, protection dismissal is kept by workers who conclude a new contract by forfeiting present employment). In a multivariate setting, Vodopivec et al. (2016) find that the new law increased the relative probability of transition from a permanent to another permanent contract with a new employer for older workers. While under the new law the probability of transition from a permanent to another permanent contract with a new employer in general decreased by 8.9%, the relative probability of accessing another permanent contract with a new employer increased by 32% for older workers (those older than 55 years). Similarly, under the new law, the relative probability to exiting from unemployment to permanent employment, as opposed to fixed-term employment, for older workers increased by 62%.

To summarize, special employment protection (“positive discrimination”) of older workers has important negative side effects. While such protection may have positive effects on keeping currently employed workers, the evidence shows that it leads to substitution between workers of different ages and that it hinders the access to jobs by the unemployed. Very tellingly, abandoning the lay-off prohibition of older workers in 2013 in Slovenia positively affected their probability to be hired under the permanent contract both from unemployment and from another permanent job.

Subsidizing Wages of Older Workers

In overcoming the reluctance of employers to hire and/or keep older workers, wage subsidies offer another alternative. In general, there is a persuasive evidence about the effectiveness of the program – that is, that wage subsidies increase employment of the target group, at least in the short run (note wage subsidies can be granted temporarily – to increase hiring of workers, or permanently – to increase the level of employment of the target group). Not enough, however, is known about some other important effects of the program – above all, about their deadweight, substitution, and displacement effects, as well as their long-run effects. Several countries have introduced wage subsidies to stimulate employment of older workers and below we summarize the findings of the studies examining the effects of such programs.

The rationale for wage subsidies is simple: wage subsidies lower the cost borne by the employer, and can thus bridge the gap between the productivity and the pay of the worker. How successful such subsidies are in increasing employment, however, depends on the responsiveness of both labour demand and labour supply to wages. For example, if lowering the wage paid by employers to older workers does not stimulate employers to hire more workers of this group (that is, if labour demand for older workers is inelastic), the subsidy will be ineffective.

For older workers the gap between the productivity and the pay of the worker often arises under inflexible wage regulations that prevent wages to adjust to changes in worker’s productivity over the lifecycle. Such is the case in Slovenia. Wages in Slovenia are subject to seniority pay system – there is a long-standing tradition (imposed both by the law and collective agreements) that the worker’s base pay increases by 1% for every two years of work experience, regardless the length of employment by the current employer. This pay system thus imposes higher mandatory costs of employment of older workers – regardless of changes in their productivity over the lifetime. Using matched employer-employee data covering the universe of labour market participants, Vodopivec (2014) finds that for older workers higher relative wages compared to younger workers are indeed not justified by their lower productivity. Despite earning a premium of roughly 20% compared to their counterparts under 30 years of age, his estimates show that a productivity of workers over 50 is almost 10% lower than the productivity of their younger counterparts.

The effectiveness of hiring subsidies for older workers has been studied on the example of Germany. The program was introduced in 1998 and paid subsidies to older unemployed workers – those over 50 years of age – upon reemployment, for up to 70 months and covering up 70% of labour costs. Using rigorous, difference-in-differences methodology and taking advantage of changing rules about program eligibility, Boockmann et al. (2012) find that the probability of exiting unemployed for the “treatment” group (those just over 50 years of age and thus eligible to hiring subsidy) was slightly increased while it was reduced for the control group (those just under 50 years of age and thus ineligible for the subsidy). However, this effect was found statistically insignificant. Moreover, the authors also find that the subsidy program stimulated transitions to subsidized jobs and reduced transitions to non-subsidized jobs, hinting on strong deadweight effects – the fact that the workers hired under the subsidy program would have been hired even without the subsidy.

Huttunen et al. (2013) show that the wage subsidy paid to employers failed to increase employment rate of older workers, wages or working hours. To stimulate employment of older workers and to provide incentives against early retirement, in 2006 Finland introduced a wage-subsidy program for all low-wage workers, both previously employed and unemployed, who were 54 years old or older. The subsidy lasted for five years and amounted to up to 44% of wages (at a declining scale, for workers with a monthly wage between 900 and 2000 EUR). The study finds that a 10% decrease in labour costs due to wage subsidy increases employment only by 0.67%, with the effect only marginally significant. In other words, the results suggest that demand for older workers is inelastic, that is, non-responsive to lowering of the wage of older workers. Moreover, the study finds that subsidy had little effect on hourly wage and hours worked, but that it did help to increase full-time employment of older workers (because the subsidy was only available to full-time workers).

Results of the wage subsidy program in Belgium also suggest low elasticity of employer demand for older workers, although the program was effective in reducing early retirement in some settings. In 2002, Belgium reduced social security contributions by 400 EUR per quarter for workers older than 58 and working at least 80% of full-time working time. The subsidy was reduced for workers working 33 to 80% of full-time hours, and was not available to those working less than one third of the full time. Studying this reform and using a triple-difference approach, Albanese and Cockx (2015) find that a 10% increase of subsidy increases the employment rate for men by 1.1 and for women by 0.5%, implying a very low elasticity. However, for some groups -- notably for workers in firms where a large share of workers had taken the early retirement option before the subsidy was introduced – the program found relatively elastic labour demand, with a 10% reduction of the labour cost increasing the probability of being employed by 9%, The study also finds that the reform caused a shift towards larger hours worked – it reduced the number of men employed for less than one third of the full time by 13%, and by 4% for women from the same group. In contrast, the effect on hourly wage was statistically insignificant.

To summarize, the above evidence suggests very limited effectiveness of wage subsidies in encouraging the hiring and retention of older workers. All three reviewed studies find that the elasticity of demand for older workers is low. Even if the program take-in is large, it may happen through the substitution of subsidized for non-subsidized work, with little overall gain in employment of older workers (as shown by the German study). In his evaluation of wage policies, Boockmann (2015) thus suggests that “To achieve the policy goal of increasing employment among older workers, policymakers need to consider other options. Where subsidies are meant to counteract incentives for early retirement, the focus should be on removing such incentives as well as on increasing the employability of older workers instead of subsidizing wages” (p. 7).

Challenging Employers’ Negative Perceptions about Older Workers

A large number of countries have introduced awareness raising campaigns to tackle negative perceptions about older workers. These countries include Australia, Denmark, Finland, Germany, Ireland, Luxembourg, Netherlands, Norway, UK, and U.S. (OECD 2006). Some governments, together with employers and trade unions, also issued employment guidelines propagating best practices in employing older workers.

One of the most recognized such initiatives is the UK Age Positive campaign.Footnote 8 The campaign started in December 1999 to promote and prepare employers and individuals on the new age discrimination legislation coming into force in 2006. The main purpose of campaign is to raise awareness of age discrimination in workplace among employers and individuals. The campaign highlights the strengths, benefits and opportunities of age diverse workplace from employer’s as well as from individual’s point of view. Additionally, the campaign provides guidelines for employers about good practices covering different areas of employment (OECD 2006).

The campaign used various channels to encourage employers to make decisions about recruitment, training and retention that not discriminate individuals because of their age. The main channel is the website where research and statistical information publication, as well as links to other sources, can be found. The campaign also uses publications, press, events and awards to spread the message. The campaign issued guidelines of good practices in six areas – recruitment, selection, promotion, training and development, redundancy, and retirement – and offered indicators with which conformity with these guidelines can be assessed. In addition, in order to promote best practices, campaign is using the so-called Age Positive Champions, a group of individuals, business, academic institution, research and lobby organisations (Age Positive Alliance) who are serving as examples to other employers, through their research and own employment practices as well as working with the campaign team (OECD 2006).

Case Studies on Promoting Older Worker Employability

Many OECD countries have introduced employment programs targeted specifically at improving the employment prospects of older workers. Such programs have had mixed results and are often short lived, but findings may have positive lessons for the delivery of mainstream services. This section considers the design, impact and lessons from successful targeted programs that have provided more holistic support to older unemployed workers in the United Kingdom, Germany and Canada.

Case Study 1: The New Deal for the Over-50s – United Kingdom

There has been a marked increase in the employment rate of older British workers prompted by major reforms to age discrimination legislation, benefit eligibility, pension rules and employment programmes. Older unemployed people now typically participate in mandatory programmes targeted at a wide range of unemployed and disability benefit claimants but, between 1999 and 2010, they were also able to participate in a voluntary programme targeted specifically at meeting their needs – the ‘New Deal 50 plus’. Evaluation findings from this programme give insights into how best to design employment services for the older unemployed.

An overall lesson from the evidence from the New Deal 50 plus experience is that for older people who volunteer for a programme of return-to-work support, providing personal advice and guidance, combined with tailored services and in-work financial incentives, are most valued. In-work financial incentives facilitated entry into work, and there is evidence that the way such payments are made has implications for take up rates, with simplicity and payments directly to the individual favoured by participants. There was evidence also that exposure to a job through short, 3-to-5-day work trials, was effective in encouraging employers to hire older workers, particularly among the long-term unemployed and those with little direct experience of the job sector.

The New Deal 50 plus (ND50+) was designed as follows. Individuals aged 50 or over and who had claimed out of work benefits for at least six months were eligible to participate. The programme was voluntary and was delivered through Jobcentre Personal Advisers (PA) who could provide support for up to six months to those who chose to stay on their ‘case loads’. The PA helped design and agree upon an action plan with each participant; helped with job-search and job applications; organised possible training opportunities and provided ongoing personal support. The PA could refer participants to other mainstream employment programmes, such as short ‘work trials’, but ND50+ participants were eligible also for a time-limited in-work financial incentive and a targeted in-work training grant. Participants were also provided financial incentive for reemployment – »Employment Credit« (see below). The program started in October 1999 in nine pathfinder areas, and was rolled out nationally from April 2000.

ND50+ achieved commendable outcomes. The program supported over 150,000 older unemployed people between April 2000 and December 2004. Evaluations indicated the job entry rate was 31%. The National Audit Office (NAO) estimated that in 2005–06 the programme generated additional net job outcomes of over 2200 and the cost per job entry was 2370 GBP. The net cost per additional job was 6370 GBP but after allowing for the benefits that flowed back to government, through reduced benefit expenditure and greater tax revenue, the direct net cost to the Government was 50 GBP per job outcome, far lower than most comparable programmes (NAO 2007, p. 41). Evaluation evidence suggested that ND50+ job outcomes tended to be sustained, although employment was less likely to be sustained by those entering part time jobs.

Most of the evaluation evidence on ND50+ is of a qualitative nature and was undertaken in the early phases of the programme. Evidence from PAs indicates that many felt that participants received significant and useful help which did much to offset loss of self-confidence, to reorient them in the labour market, and to help them overcome employer age discrimination. Since many participants did not take up the offer to receive continuing PA advice and guidance, the resources provided were sufficient to deliver a high-quality programme for those that did want to take advantage of the support. The main positive feature of this ‘case loading’ (from the perspective of ND50+ participants) was to establish a sustained participant-adviser relationship where the adviser was recognised to be personally committed to the participant. Despite the value placed on case loading, there was no statistical correlation with likelihood of securing employment. This was most likely explained by the fact that the most job-ready participants did not enter case loading.

The ND50+ ‘Employment Credit’ (EC) – the financial into-work support – was originally a tax-free wage supplement providing 60 GBP a week for those working full time (40 GBP for part time, working between 16 and 30 h a week) for a year, provided the recipients’ annual income was estimated to be less than 15,000 GBP.Footnote 9 From the outset, PAs regarded the EC as the ‘key feature’ of the ND50+ and a major incentive for participants. In 2003, however, the EC was replaced by the ‘50+ return to work’ element of the general in-work Working Tax Credit (WTC). The advantage of the EC was that it was simple, highly visible to participants but ‘invisible’ to employers who might otherwise adjust salaries. By contrast, the 50+ element of WTC is more complex: payment is made in arrears and it is targeted at the poorest workers and based on household not individual income. Data on the take up of WTC by age is not available, but survey-based evidence suggested that this design change had a detrimental effect on ND50+ participation. Participants felt that the WTC provided less of an incentive to take work and PAs found it more difficult to persuade participants that work was a viable option.

Evaluation evidence from the earlier period suggests that the EC was important in facilitating and encouraging, but not decisive, in the return to work; just under half would have accepted their job without EC. What the EC (and WTC) does is make a lower wage more attractive/acceptable to recipients. There is evidence that EC was most effective in areas of low wages and low living costs - suggesting that in such areas it addressed a ‘benefits trap’. Evidence suggests that job retention in the year following the end of EC entitlement was high, with a survival rate of about two-thirds by the end of the second year. Only a minority of recipients had replaced a substantial amount of the credit after a year. Coping strategies in the face of reduced income included ‘surviving’ by cutting back on household spending and increasing hours of work.

The ND50+ also offered ‘Training Grant’ (TG), but the intake was low. The TG was initially set at 750 GBP and was available for expenditure on job-related training for a period of one year to all those who found work through the programme. In 2002 the amount of the TG was raised to 1500 GBP and the period covered was increased from one to two years. 1200 GBP was available for training relevant to the job and 300 GBP for wider career-related training. Take up of the TG was low, despite high levels of awareness. Amongst the reasons cited by ND50+ participants for this were that they were too old to train or that they had the skills needed for the relatively low skilled jobs they obtained. A further reason cited for low take up was that participants had no experience of buying training for themselves; little knowledge of what training they needed; of what the money might buy; or where would be a good place to get it. It relied too heavily on the older person to be proactive, at a time when individuals’ key concerns were with entering and settling into work. By contrast, the TG was valued by participants moving into jobs with prospects for advancement and the self-employed (who used it to attend local courses in administration and marketing). The TG was also valuable in some small firms where it had been possible to integrate it with employer spend on training.

The importance of attitude emerges strongly from the evaluation evidence. The success of ND50+ has been attributed to the voluntary nature of the scheme and the commitment to work of participants (i.e. those coming forward tended to be younger and were most committed to working and most disliked being on benefits). After 2010 this voluntary approach was dropped and replaced by mandatory programmes targeted at all the very long-term unemployed and at ‘inactive’ people claiming sickness and disability benefits.

Case Study 2: Perspective 50plus: Employment Pacts for Older People in the Regions - Germany

The German Government has introduced complex, holistic reforms to increase the employment rate of older workers and prevent long-term unemployment. This process has comprised changes in benefit entitlements and pension arrangements alongside institutional changes and the development of new employment programmes. In 2005 this included the merger of long-term unemployment benefit and social assistance into a single ‘basic income support for jobseekers’ (BIS) which is delivered through a national network of co-managed or solely controlled municipal Jobcentres.

As part of the wider ‘Initiative 50plus’ of the German Federal Ministry of Labour and Social Affairs, ‘Perspektive 50plus’ was introduced in 2005 and was designed to enable the new Jobcentres to tackle the employment barriers of long-term unemployed more effectively. In contrast to previous policies that emphasized public sector employment, the program aims to place older workers to the unsubsidized employment in the private sector based on regional employment pacts (among employers and business networks, non-profit organizations, health provides or volunteer networks). This process attempts to mobilize resources at the local level, to advance creative solutions, and to induce a competition for best practices, based on federal funding. The program typically consists of the assessment of the jobseekers’ skills, counselling, job search assistance and coaching, training as well as interventions to improve health and mobility. The programme was delivered in three phases until it ended in 2015.

By focusing on the reintegration of the longer term unemployed in regular labour market jobs rather than providing them with limited ‘make work’ employment as under the previous schemes, the programme represented a significant change in policy. The programme was delivered through local ‘employment pacts’ which brought together Jobcentres, employer organisations and other partners to provide individualised support for the unemployed. The pacts also raised awareness of, and promoted the recruitment of older workers amongst employers. In each area employment pacts promoted a cross-sector approach including labour market, employment, social and health organisations.

The program was implemented in two phases. Only 93 out of 438 Jobcentres were selected to participate in the first phase of the programme but by 2012 there were 78 regional employment pacts covering most of the country. The initial networking focus concerned Jobcentres being incentivised to liaise with employers and trade unions, and their respective organisations. In the second phase of the programme new applicant Jobcentres had to collaborate with a Jobcentre or group of Jobcentres already in the programme thus creating learning networks. Each pact or network had a coordinating unit based either with a Jobcentre or external partner.

The Ministry, Jobcentres and other tiers of government co-financed employment pact budgets. Total spending in the second phase amounted to 787 million EUR, of which 56% was provided by the Ministry. Project management was largely decentralised with the Ministry principally steering the system through agreed targets for placements and sustained employment in jobs subject to social insurance contributions. Failure to meet targets would reduce the Ministry contribution giving local Jobcentres strong incentives to meet their targets. Day-to-day management of the programme was contracted out to a private non-profit organisation which acted as the ‘Programme Managing Agency’. The Agency did not have a direct contractual relationship with the pacts but was responsible for reviewing proposals and making recommendations to the Ministry; providing support for individual networks; organising knowledge transfer activities; and running the programme database.

Participating Jobcentres were free to use the targeted funding to employ additional staff (usually on fixed-term contracts) and/or procure services from external service providers. Some of the additional funding could be used to directly support participants (e.g., meet travel costs or self-employment start-up grants) or provide wage subsidies to employers. Evaluation findings suggest most of the funding was spent on additional Jobcentre staff, followed by payments to external providers.

Descriptive statistics on the program are as follows. In 2010 more than 280,000 people participated in ‘Perspektive 50plus’, representing more than 20% of the stock of unemployed workers aged over 50. Over 70,000 of the participants entered employment. Participants had low levels of education and basic skills and few were older than 60, with an average age of 54 years. The average time out of employment was more than two years and 13% had not been employed for over ten years. Around one-quarter suffered from health conditions restricting their working ability. A high proportion lived on their own and reported high levels of depression and social isolation. During the evaluation period, from January 2010 to March 2011, one-third of those who had actively participated in the programme took up employment fully covered by social insurance or self-employment. Almost 70%, or 24% of all participants, were in sustained employment lasting at least six months. This compared with an integration rate of 19% for the same age and benefit group in mainstream services. In terms of ending benefit recipiency, however, outcomes were modest: only 27% of those who took up work stopped claiming the benefit. This was a consequence of the design of BIS which can be retained by those entering low paid work, especially those who could only work part time because of their health circumstances. There was evidence that the programme also had a positive impact on the subjective perceptions and well-being of participants and their satisfaction with service delivery.

A comparison between 2010 expenditure on Perspektive 50plus and standard operations found that spending on the programme was lower both per participant and per sustained employment outcome, the results to be credited to more effective counselling and innovative institutional arrangements. This cost efficiency was attributable to higher success rates in terms of re-employment, and less use of costly wage subsidies or make-work schemes. The relative success of the programme was considered to rest on individualised counselling and the intensity of the treatment (frequent interactions and taking the time each participant needed) and the flexible combination of measures, including more investment in short-term training and more self-employment start-up support. It was attributed also to intensified cooperation between Jobcentres and other local service providers and the flexibility that pacts had to tailor services to the needs of their labour markets. The programme pioneered new approaches in improving participants’ subjective health conditions, their personal management of health problems and their health-related life styles. In rural areas with poor public transport services the pacts also sought innovative solutions for clients’ job-related transport problems, like support in re-qualifying for a driver’s licence, or interest-free loans for buying a second-hand car. The main success factor of the programme seems, however, to have been lower caseloads achieved by engaging more frontline staff, either within the Jobcentres or employed by external providers.

The Ministry has now chosen to integrate successful components of Perspektive 50plus into regular labour market programmes and the approach to networking and individualised coaching has shaped a new ten-year programme targeted at the long-term unemployed. Over 1000 Jobcentre counsellors who previously delivered Perspektive 50plus are now employed in Jobcentre ‘activation centres’ which offer intensive employment counselling and complementary health and social welfare support to the very long-term unemployed.

In their rigorous evaluation, Boockmann and Brändle (2015) find substantial positive effects of the program. To identify the treatment effects, they exploit regional variation in program participation. They find that the program increased placement into unsubsidized employment by five to ten percentage points, but due to substantial lock-in effects program participants have a higher probability of remaining on public welfare benefit receipt for up to one year after the start of the program. Another evaluation concluded that the approach was more ‘innovative, effective and cost-efficient’ than the standard programmes otherwise available for BIS claimants (Knuth 2014).

Case Study 3: The Targeted Initiative for Older Workers - Canada

The Targeted Initiative for Older Workers (TIOW) is a federal-provincial cost-shared programme that helps unemployed workers aged 55 to 64. It is a small-scale programme which complements fiscal reforms and other employment services that aim to increase the labour market participation of older workers. The programme is targeted at smaller communities, of fewer than 250,000 people, that have high unemployment and/or where a single industry or employer has downsized or closed especially in traditional sectors, such as forestry, fishing, mining, textiles and clothing. Projects are designed to improve the employability of older unemployed workers by delivering services that increase their ability to find work, reintegrate back into employment and ensure that they remain employable while their communities undergo economic adjustment.

TIOW was introduced in 2006, initially as a temporary measure but has since been extended and expanded through three-year inter-governmental agreements. The federal Government contributes a maximum of 70% of total TIOW costs per region, with the participating province or territory contributing a minimum of 30%. The most recent agreement covered 2014–17 and involved a federal investment of 75 million CAD. It also broadened eligibility criteria to support small communities experiencing unfulfilled employer demand and/or skills mismatches. The federal Department - ‘Human Resources and Skills Development Canada’ - establishes the broad framework and eligibility criteria for participants and communities. Provincial and territorial governments have flexibility to tailor TIOW design and delivery to local needs and they identify and select local projects and providers.

The TIOW model was based on best practice and lessons learned from earlier older worker programmes. Individual projects involve a combination of activities which must include job-search assistance, interview techniques and job counselling, supplemented by at least two other employability improvement activities (such as improving basic skills, vocational skills training, work experience, direct marketing to employers and preparation for self-employment). Although service assistance is individually tailored, TIOW projects include group-based activities that are designed to foster peer support amongst participants. Most projects are delivered through contracts with service providers and community-based organisations.

Applicants are assessed to determine who will benefit from participation and whether some applicants may benefit from referral to other services. Amongst eligible unemployed people priority is given to those most in need, such as those who have been unemployed for over a year, those not receiving Employment Insurance benefits, and/or those with low levels of education or redundant skills. Participants must commit to at least 25 hours per week for a minimum of four weeks of employment assistance. Employability improvement activities, such as training and subsidised employment, may last for different periods with the total duration of TIOW activities and services ranging from 8 weeks to a maximum of 43 weeks.

An important feature of TIOW is that participants receive income support via allowances and subsidised wages. Project sponsors pay allowances to participants who are engaged primarily in developmental activities, such as training and competency assessment. Participants are paid wages if they are working for a community-based organisation or a private sector employer.

Between 2007 and 2015 TIOW provided re-employment support to over 35,000 unemployed workers for an average cost per programme participant of 7000 CAD. Evaluation studies suggest the programme was well-targeted. Over 90% of projects were delivered in communities that had been affected by a significant downsizing or sectoral closure, with 60% being delivered in areas with unemployment rates over 10%. Half of TIOW participants had not attained more than a high school diploma (including 24% not having completed high school), while half had some post-secondary education. Approximately 20% of participants had been out of the paid labour market at least two years prior to their participation and a further 31% had experienced job loss in the year prior to starting the programme.

Post-programme surveys indicated that the group learning and peer support activities had a particularly positive impact on participant learning, confidence and motivation (especially as most of the older unemployed had not been engaged in formal learning settings for considerable periods of time). Furthermore, surveys of programme participants found that 75% found employment in their local area during or after their participation and at least 70% felt more employable as a result of their participation. 44% were still employed eight months after programme participation and 41% employed after 18 months. Participants who had been unemployed for over a year before participation were less likely to find work and, on average, were employed for shorter periods. 75% of those who worked after the programme experienced a change in their work setting (50% changed sectors, 25% changed employers). More than 80% indicated that they were satisfied with the support and training they received.

A comparison with unemployed people who participated in a more limited (and cheaper) employment assistance programme found that the employment rate of TIOW participants was 6% higher. Furthermore, there was no differential impact on earnings or the use of Employment Insurance or social assistance benefits (although poor data collection procedures hampered the ability to fully assess impacts). The evaluation results showed also that whilst employed TIOW participants experienced a drop in income relative to their previous earnings, this decline was less than that experienced by other older workers suggesting that programme participation may have mitigated potentially more negative income impacts.

What differentiates the TIOW model from many other Canadian labour market programmes is its flexibility: provinces and territories can customise and tailor their projects to best meet participant needs and to respond to evolving labour market opportunities and challenges at the local level. Another distinctive feature is the efforts made to learn from and share best practice, including the identification and dissemination of a compendium of successful and innovative TIOW projects across jurisdictions. Promising practices highlighted in this compendium signalled a number of lessons learned, including (but not limited to) the following:

  • The importance of implementing dynamic recruitment and advertising strategies to ensure potential clients and employers are aware of the TIOW and the benefits it yields;

  • The value of facilitating strong partnerships and networking opportunities among and between TIOW project sponsors, clients, employers, and local organisations to identify specific labour market needs and opportunities;

  • The merits of seizing viable opportunities for tailoring and/or customising projects to best support participant and employer engagement; and

  • The advantages of embracing technology and social media platforms to facilitate communication and/or the acquisition of new skills.

Conclusions

This paper reviewed international experience with efforts to increase the employment of older workers, focusing on initiatives and programmes that addressed worker and employer obstacles. The key lessons learned from the review are summarized below.

Training Promotion of Older Workers

Direct subsidies – financial grants or vouchers provided to firms – seem to be more effective than tax incentives to increase the level of training and in improving employment outcomes. They also appear to facilitate targeting of specific groups of firms and workers. Positive evidence on both financial grants and vouchers is supported also by several rigorous evaluations, showing that such programs can increase both training participation and employment and wage outcomes, although the effects may not be strong. On the negative side, the studies report deadweight and substitution effects. Training funds seem not to offer the best alternative. While they may increase training for the targeted group of firms, they are unlikely to stimulate training of specific groups of workers, including older workers, and they contribute to significant deadweight losses.

Individual learning accounts (ILAs) – the new kid on the block among the training schemes – remain a promising tool, although the full-fledged evaluation is still missing. The U.S. experience also provides useful insights on fundamental trade-offs in designing ILAs – above all, that mandatory, intense counselling discourages participation in training, but on the other hand, that such an approach is more successful in raising wages of training participants, possibly because it steers jobseekers to types of training that proves most beneficial to them (and to jobs for which they have trained). Interestingly, the U.S. experience also shows that when counselling is voluntary, few customers request it, and the likelihood of training is higher than under mandatory counselling. And the United Kingdom’s experience in the early 2000s provides a cautionary example of how a hastily implemented and poorly designed ILA scheme can be subject to abuse.

Very importantly, there is strong evidence that raising the retirement age increases training among older workers, and that flexible retirement policies also stimulate such training.

Adjusting Employment Services and ALMPs for Older Workers

Research shows that older workers, including those experiencing cognitive decline, can learn new skills with long-lasting training effects, but that training methods for older workers should be tailored to their learning needs and abilities. Among others, older learners should have ample time to process new knowledge. They also benefit from integrating multiple instructional methods and from building on pre-existing knowledge, for example, from on-the-job training as well as applied training based on communication and management skills.

Promoting Better Working Conditions

Promoting flexible work arrangements – including part-time jobs, whereby workers can gradually reduce their working hours while remaining in the workforce – can extend the length of working lives and increase workers’ well-being.

Adjusting Employment Protection Rules

The above review shows that special employment protection (“positive discrimination”) of older workers has important negative side effects. While such protection may have positive effects on keeping currently employed workers, the evidence shows that it leads to substitution between workers of different ages and that it hinders the access to jobs by the unemployed.

Subsidizing Wages of Older Workers

Our review shows that wage subsidies are a rather ineffective tool for encouraging the hiring and retention of older workers, as the elasticity of demand for older workers is typically low. Even if the program take-up is large, it may happen through the substitution of subsidized for non-subsidized work, with little overall gain in employment of older workers. It is thus easy to agree with Boockmann (2015) that the policy goal of increasing employment among older workers should be achieved via other options.

Challenging Employers’ Negative Perceptions about Older Workers

A large number of countries have introduced awareness raising campaigns to tackle negative perceptions about older workers, such as the UK Age Positive campaign. While there are no studies documenting their effectiveness in improving labour market outcomes of older workers, such campaigns – by highlighting the strengths, benefits and opportunities of age diverse workplace – undoubtedly importantly contribute to a climate conducive to employment of older workers.

Lessons from the Case Studies

  • United Kingdom’s experience with its New Deal 50 plus voluntary program for older people shows that providing personal advice and guidance, combined with tailored services and into-work financial incentives, can facilitate entry into work. Personal advice has proved important in offsetting loss of self-confidence by jobseekers, in reorienting them in the labour market, and in helping them overcome employer age discrimination. Exposure to a job through short, 3-to-5-day work trials was also effective in encouraging employers to hire older workers.

  • Germany’s experience with its Perspective 50plus shows that innovative institutional and reforms – changes in benefit entitlements and pension arrangements alongside institutional changes and the development of new employment programmes – can both improve labour market outcomes and reduce costs. The program typically consists of the assessment of the jobseekers’ skills, counselling, job search assistance and coaching, training as well as interventions to improve health and mobility. In contrast to previous policies that emphasized public sector employment, the program aims to reintegrate older workers to the unsubsidized employment in the private sector with the help of regional employment pacts. This process mobilizes resources at the local level, advances creative solutions, and induces a competition for best practices, based on federal funding.

  • Canada’s experience with its Targeted Initiative for Older Workers highlights the importance of flexibility in programmes designed to help displaced older workers re-enter employment. Such flexibility allowed the programme – in which unemployed workers aged 55 to 64 were immersed in in-depth job-search assistance, counselling, and training – to seize viable opportunities for customising projects to best support participants and employers. Furthermore, participant surveys show that for older unemployed workers, group learning and peer support activities have a particularly positive impact on participant learning.