Employers and Active Labour Market Policies: Typologies and Evidence

Among scholars and practitioners, there is a growing recognition of the important role of employers in the success of active labour market policies in Europe. However, there is a lack of systematic evidence about why and how employers engage in active labour market policies. In this article, the preferences and behaviour of employers towards active labour market policies are untangled. A typology of four types of employers is constructed for analytical and empirical analysis. By distinguishing positive and negative preferences from participation and non-participation, four types of employers are identified: the committed employer, the dismissive employer, the sceptical employer and the passive employer. The utility of the typology is tested with survey data on employer engagement in Danish ALMPs. The findings indicate that only a minority of Danish employers can be classified as ‘committed employers’, and the majority are either ‘dismissive’ or ‘passive’ employers. In the final section, this finding and the usefulness of the typology for analytical and empirical research is discussed.


Introduction
The active labour market policies (ALMPs) in Europe, since their introduction in the early 1990s, have mainly focused on increasing labour supply and reducing unemployment by improving the motivation and qualifications ('employability') of the unemployed and other disadvantaged groups in the labour market (European Commission, 1993;OECD, 1994). The role and functions of employers (the demand-side) in ALMPs have been less debated and acknowledged both in policy design and implementation as well as in labour market and social policy research.
However, the engagement of employers is an important precondition for implementing effective employment services and successful activation strategies. Effective ALMPs require employers that are willing to recruit unemployed persons in unsubsidised as well as subsidised employment. Programme evaluations of ALMPs indicate that activation programmes which are implemented at the workplace level (e.g. wage subsidies and work experience programmes) are more efficient than activation programmes implemented by the public authorities themselves (Kluve, 2006(Kluve, , 2010Card et al., 2010;Bredgaard, 2015). However, studies also indicate that the majority of employers tend to avoid the Public Employment Service (PES) and use other non-public and informal recruitment channels. 1 Some employers have negative perceptions of the PES and perceive jobseekers referred by the PES as less motivated and trustworthy than other jobseekers (OECD, 2015;Hasluck, 2011;Larsen and Vesan, 2012). The article proceeds by presenting an analytical framework for understanding the role of employers in ALMPs and reviews the small but growing literature on the subject. On the basis of this literature, a typology of four different types of employers is constructed. In contrast to existing typologies, it makes a clear distinction between preferences (attitudes) and behaviour (participation). The four types of employers are labelled: (1) the committed employer (positive attitudes and participation), (2) the dismissive employer (negative attitudes and no participation), (3) the sceptical employer (negative attitudes but participation) and (4) the passive employer (positive attitudes but no participation). The typology can be used for comparing employer engagement in ALMPs by country, sector or historically. It also allows for better targeting of employers in political decisionmaking by identifying strategies that enhance or deepen the engagement of each type of employer.
In the second part of the article, the applicability of the typology is tested with data from a nationally representative workplace survey from Denmark. In the comparative literature, Danish employers are considered more engaged and active in ALMPs than employers in most other countries (cf. Martin, 2004aMartin, , 2004bMartin and Swank, 2012;Thelen, 2014). However, the empirical analysis finds that the majority of Danish employers are in fact not positive and do not participate in ALMPs. The data indicate that the majority of Danish employers are either 'dismissive' or 'passive' in ALMPs. In the concluding section, explanations for this finding are discussed.

Analytical framework
It is useful to distinguish between three ALMP approaches: the supply-side approach, the matching approach and the demand-side approach (cf. Table 1).
The supply-side approach has been and remains the dominant approach in ALMPs in Europe. The objective is to make jobseekers ready to (re)integrate into the ordinary labour market by addressing their (assumed) lack of qualifications or motivation, otherwise known as human capital and work-first approaches (cf. OECD, 1994;Lodemel and Trickey, 2001;Peck, 2001). Education and training, job search assistance and benefit sanctions are typical examples of instruments used in this approach to improve the qualifications or motivation of jobseekers.
The matching approach aims at matching labour supply (jobseekers) and labour demand (employers) and is one of the core functions of the PES. The basic role of the PES is to break down information asymmetries and redress the lack of credible information on both sides of the labour market by providing information on vacancies and assisting in matching vacant jobs with jobseekers. The matching approach also reflects an increasing acknowledgement that activation programmes connected to local employers are more effective than public sector programmes (Bredgaard, 2015). Some of the typical instruments in this approach are vacancy databases, job fairs, wage subsidies and work experience programmes.
The third policy approach to ALMP, the demand-side approach, is often taken for granted and not explicitly addressed in policy design or implementation (cf. Ingold and Stuart, 2014;Aa and Berkel, 2014). However, in recent years there has been an upsurge in the interest on the role of employers in ALMPs among scholars, policymakers and practitioners. For instance, this is the case in the UK (Fletcher, 2004;Gore, 2005;Farnsworth, 2006;Ingold and Stuart, 2014;Ingold and Valizade, 2015), the Netherlands (Aa and Berkel, 2014), France (Salognon, 2007), Norway (Mandal and Osborg Ose, 2015) and Denmark (Bredgaard, 2014;Bredgaard and Halkjaer, 2016). The demand-side approach to ALMPs is targeted at the recruitment practices and personnel policies of employers, and aims at preventing the direct and indirect discrimination of 'disadvantaged' jobseekers (cf. Becker, 1957) as well as persuading more employers to engage in ALMPs. The demand-side approach may use obligatory instruments (such as quota schemes and anti-discrimination laws) as well as voluntary instruments (such as campaigns and persuasion) (Aa and Berkel, 2014;Bredgaard, 2015).
It is important to note that employers play an explicit and active role in both the matching and demand-side approach to ALMP, while they are implicit and taken for granted in the supply-side approach. The three approaches to ALMP are not mutually exclusive but may under certain circumstances function as complementary strategies. For instance, training and education programmes may be more effective if they are connected to private enterprises with a need for qualified labour. In other situations, anti-discrimination regulations may be necessary to open up the labour market for qualified and motivated unemployed jobseekers (e.g. refugees and disabled persons).
There are different levels and forms of employer participation in ALMPs. Employers may use the PES to recruit new employees (either directly by contact with PES officers or indirectly through the use of public job databases). There are two basic types of recruitment options through the PES: ordinary recruitment or participation in activation programmes. In the first case, the PES performs a job matching function and assists the employer in finding the right candidate for the vacant job. In the second case, the employer participates in different types of subsidised or unsubsidised activation and employment programmes (e.g. work experience programmes, subsidised employment, wage subsidy schemes, vocational counselling, training schemes, job rotation, apprenticeships).

Typologies for employer engagement in ALMP
In the literature, there are different typologies for categorising and measuring the level and character of employer engagement in ALMP. Ingold and Valizade (2015) make a useful distinction between typologies that (1) focus on categorising the type of employer participation (behaviour and attitudes) or (2) focus on categorising the reasons for participation (motives).
In this article, I focus on categorising the behaviour and attitudes of different types of employer, and less on their motives for participation. In the literature, there is a tendency to conflate attitudes and behaviour. For instance, Martin (2004aMartin ( , 2004b) used a scale with different degrees of participation and combined it with the ideological position of the employers. 2 Similarly, Nelson (2013) used a five-point scale to compare employer participation in ALMPs in Denmark and Germany: ideological opposition and nonparticipation, no opposition and non-participation, minimal participation, more extensive participation and full-blown social partnership. The problem is that these scales conflate attitudes (ideology) and behaviour (participation) on a single scale rather than separating them. This is problematic in cases where employers participate in ALMPs even if they are ideologically or politically opposed to activation programmes but derive other benefits from participation (e.g. cheap labour, economic subsidies, good reputation) or in cases where employers do not participate even if they are supportive of ALMPs (e.g. due to lack of awareness or contacts with the PES). In a recent study, Ingold and Valizade (2015) compare degrees of actual participation ('employer engagement') in the UK and Denmark and do not include indicators on attitudes or motives. They distinguish between employers that are 'instrumentally engaged' (one-off and ad hoc activities such as vacancy placement) and 'relationally engaged' (deeper, repeated and sustained engagement in ALMPs).
In the second group of typologies, which categorise employers' motives for participation in ALMPs, there is recognition of other types of motivation beyond shortterm economic self-interest. For example, Aa and Berkel (2014) identified three types of employer motivation in the Netherlands: (1) employers who wished to recruit new workers through a public recruitment channel, (2) employers predominantly interested in reducing their wage costs and (3) employers wishing to recruit disadvantaged workers due to corporate social responsibility (CSR). In the British context, Coleman et al. (2014) identified four types of motivation for recruiting unemployed young people: (1) altruistic motive (wanting to help young unemployed persons), (2) opportunistic motive (motivated by the financial incentive and/or convenience), (3) responsive motive (candidate just came along) and (4) strategic motive (part of a recruitment strategy). Bredgaard and Halkjaer (2016) derived six theoretical explanations for employers' participation in the implementation of ALMPs (neo-classic economic theory, theory of collective action, power resource theory, varieties of capitalism, institutional theory and CSR theory) and used them to explain why Danish employers recruit unemployed persons through the wage subsidy programme.
In Figure 1, a typology is proposed that distinguishes between attitudes and participation. This typology illustrates that employers' attitudes and participation in ALMPs are more complex and varied than often assumed. The idea is that positive or negative attitudes do not necessarily produce a certain type of behaviour (e.g. positive attitudes may also correlate with non-participation, whereas negative attitudes may correlate with active participation). There may be different reasons for this type of behaviour, e.g. a strong sense of corporate social responsibility (Aa and Berkel, 2014;Bredgaard, 2014), the type of jobs and company-specific skills at the workplace (matching the profiles of jobseekers) (Hall and Soskice, 2001) or persuasion of member companies by peak employer associations (Martin, 2004a(Martin, , 2004b. (2) The dismissive employer has negative attitudes and does not participate in ALMPs. This type of employer can be understood within a neo-classical economic framework in which employers are motivated by profit maximisation and cost reductions (Friedman, 1962) and only participate in ALMPs if it serves their short-term economic interests, e.g. by reducing total labour costs (Ibsen, 1999). (3) The sceptical employer has negative attitudes to ALMPs but nevertheless participates.
This type of employer is more 'puzzling' than the two above, and also more susceptible to change as participation may be withdrawn (shift towards the dismissive employer) or attitudes may become more positive (shift towards the committed employer). There seems to be two basic and related explanations for this type of employer: strategic accommodation (i.e. employers participate in ALMPs because they expect to reap favours on other issues, such as access to contracts with public agencies, or contain a less favourable course of action, such as quotas) (Hacker and Pierson, 2002;Paster, 2013) or selective incentives (i.e. employers participate in ALMPs because they get access to services that are exclusive to participants, such as free recruitment assistance and judicial guidance) (Olson, 1971).
(4) Finally, the passive employer is basically positive towards ALMPs but does not participate actively, which may be due to a lack of knowledge or a perception that ALMPs are irrelevant to the functioning of the company. In the first case, effective persuasion strategies by the PES may convince the employer to participate in ALMPs and imply a shift towards the committed employer. In the second case, persuasion strategies will probably not work, since the recruitment of unemployed jobseekers is not regarded as relevant to the functioning of the company (cf. Hall and Soskice, 2001).
The typology can be applied for comparisons of the distribution and characteristics of different types of employers in national labour markets as well as across national labour markets. In the following, the usefulness of the typology as an analytical tool will be tested on the empirical case of Danish employers and ALMPs.

Danish employers and active labour market policies
The Danish case of employers and ALMPs is relevant for two reasons. First, comparative studies indicate that there is a relatively high proportion of employers with positive preferences and active participation in Denmark (i.e. 'committed employers'). Second, Denmark has been a frontrunner in the development of ALMPs and the engagement of employers in the governance of labour market policies (cf. Schmid, 1998;Auer, 2000;Schmid and Gazier, 2002). In the following, the two rationales for selecting Denmark and the context for employers' engagement in ALMPs is described.
Comparative studies indicate that Danish employers participate actively in ALMPs compared to, for instance, British and German employers (cf. Martin, 2004aMartin, , 2004bMartin and Swank, 2012;Nelson, 2013). In the early 2000s, Martin (2004b: 130) found that 68 per cent of Danish employers participated in ALMPs compared to 40 per cent in the UK. Martin also suggested that the most important reasons for Danish companies to participate was their sense of social responsibility (51 per cent), that programmes involved companies in a new way (50 per cent) and that employers gained access to subsidies (38 per cent). In the UK, the most important reasons for participation were political pressure and a desire to please the new government (31 per cent), i.e. strategic accommodation (cf. Hacker and Pierson, 2002;Paster, 2013).
Comparing Denmark and Germany, Nelson (2013: 40) found that although roughly the same percentage of companies participated in ALMPs, the participation of Danish companies was on average much more intense than in German companies. Nelson suggested that Danish companies participated in ALMPs to gain access to skilled labour, while German companies were motivated to gain access to cheap labour (Nelson, 2013: 43).
In a recent comparison of employer engagement in Denmark and the UK, Ingold and Valizade (2015) found that employer participation in Denmark was slightly higher than in the UK. Ingold and Valizade identified vacancy placement as the minimum level of employer participation and found that it was almost similar in the two countries. Participation in ALMPs is considered a deeper and more committed type of engagement between employers and the PES (jobcentres). Ingold and Valizade found that almost three out of four employers in Denmark and the UK participated in at least one ALMP programme (defined as funded programmes, work placements and specialist schemes), although participation in 'funded programmes' (such as wage subsidy, young person apprenticeships or one-off and on-going payments) was higher in Denmark (50 per cent) than in the UK (16 per cent). They concluded that Danish employers were more 'relationally engaged' in ALMPs, while British employers were more 'instrumentally engaged'.
In sum, the comparative evidence suggests that a relatively high proportion of Danish employers can be categorised as 'committed employers' (positive preferences and active participation in ALMPs). One possible explanation for this level of employer engagement may be the institutional history of the governance of ALMPs (Martin, 2004a(Martin, , 2004bMartin and Swank, 2012).
Denmark was among the pioneers in the introduction of activation policies and ALMPs (Torfing, 1999;Auer, 2000). The 'paradigm shift' occurred in 1994 with a labour market reform that combined macro-economic stimulus of the economy and new modes of governance and implementation of labour market policies. The reform introduced ALMPs in a balanced combination of 'human capital' and 'work first' approaches which were labelled the 'right and responsibilities' approach (Jørgensen, 2002). The 'rights' of persons on unemployment benefits were improved by giving them access to individual action plans, activation programmes and longer-term education and training programmes. The 'responsibilities' (conditionality) were tightened by reducing the duration of unemployment benefits to seven years, stricter availability criteria and repealing the right to re-earn eligibility to unemployment benefits through participation in activation programmes. The role of employers in the implementation of the new activation strategy was not explicitly addressed in the legislation, but the governance of labour market policies was decentralised to regional labour market councils in which employer associations and trade unions were represented. The underlying idea was that employer associations would take ownership and responsibility for regional labour market policies and implementation and assist in informing and persuading their member companies to participate in the implementation of ALMPs (Bredgaard, 2011).
The international literature referencing the Danish case on ALMPs tends to focus on this history of the labour market policy in the 1990s (Martin, 2004a(Martin, , 2004bMartin and Swank, 2012;Thelen, 2014). However, in the 2000s important changes were made in the governance and implementation of labour market policies with implications for the participation of employers in ALMPs.
Firstly, labour market policies were renamed 'employment policies' to signal that implementation should focus on 'work first' rather than 'human capital' approaches. This change of policy coincided with a reduction in open unemployment and a political desire to increase the economic incentives of the unemployed (e.g. 'making work pay'). The shift towards 'work first' remained focused on the supply-side of the labour market. The demand for labour and the then role of employers was taken for granted and not explicitly addressed (cf . Table 1).
Secondly, the governance of employment policies was decentralised to new local jobcentres but under central supervision and control. In this process, the social partners, including employer associations, were gradually marginalised from the implementation of employment policies. This led to a gradual decline in the feeling of ownership and responsibility towards the implementation of ALMPs by employer associations (Bredgaard, 2011).
In 2008, the global financial crisis hit the Danish economy and labour market hard and led to a prolonged economic recession. In the aftermath of the economic crisis, decision-makers increasingly realised that the jobcenters had to improve their role in matching labour supply and labour demand. Danish employment policies gradually shifted towards the 'matching approach' to ALMP (cf. Table 1). The motivation to engage employers in ALMPs was supported by impact evaluations showing that activation programmes delivered by (private) employers performed better in terms of participants' subsequent employment rates and income than activation programmes carried out by the local jobcentres or other public institutions (Rosholm and Svarer, 2011;Danish Economic Council, 2012). The financial incentives for local jobcentres to prioritise a local activation strategy with local employers instead of municipal activation projects were also enhanced.
This led to an increase in employer engagement, indicated by an increase in the use of wage subsidies and work experience programmes. In spite of the economic recession, local jobcentres succeeded in engaging more employers in participating in wage subsidy and work experience programmes. As an indication, the share of participants in wage subsidy and work experience programs increased from 26 per cent of all unemployed in 2007 to 45 per cent in 2011. Currently, about 40 per cent of all the unemployed participate in wage subsidy or work experience programmes.
However, employers are less inclined to use the local jobcentres for ordinary recruitment of employees. A survey from the Danish Ministry of Employment (from spring 2016) shows that while 28 per cent of all companies cooperate with a jobcentre on wage subsidies, work experience programmes, adult apprentices or job rotation, only 13 per cent of the companies cooperate with a jobcentre on ordinary recruitment of employees.
In a comparative perspective, the labour force surveys of the European Union also indicate that the 'market share' of the Danish PES is lower than the European average. Data from 2012 on people recently employed (who have started their job within the last twelve months) shows that the PES was only involved at any moment in finding the present job in 7 per cent of the cases in Denmark compared to 9 per cent on average in EU countries (OECD, 2015: 137).
In summary, the history of Danish ALMPs and the employers show a mixed trend. On the one hand, employer associations have been marginalised in the governance of employment services. On the other hand, ALMPs are increasingly focusing on engaging employers in the recruitment of unemployed jobseekers and on delivering activation programmes at the workplace level. This mixed trend makes it interesting to study empirically how employers are actually engaging in ALMPs.

Empirical analysis
In the following, the level of engagement of Danish employers in ALMPs and the disbursement of different types of employers in the typology is examined. The data originate from a nationally representative survey collected by Statistics Denmark in the autumn of 2013 among a random sample of Danish companies with more than five employees. Of the 4,500 private and public workplaces that were contacted, 1,499 workplaces completed the questionnaire (response rate = 33 per cent). The respondents were the persons making decisions on personnel management at the workplace level (e.g. HRM manager, owner or director). The data were then weighted by industry and the number of employees in order to ensure that the sample was representative for the population of Danish companies.
The questionnaire included questions designed to capture employers' attitudes towards ALMPs and questions concerning participation in ALMPs. Employers were asked to what degree they had confidence that the jobcentres would help them recruit labour. Almost half of the respondents (47 per cent) had confidence that jobcentres 'to a great degree' or 'some degree' would help them recruit labour, while slightly more than half (53 per cent) had little or no confidence in the jobcentres. 3 The companies that had recruited new employees within the last year (72 per cent of total sample) were then asked about the type of recruitment channels they used for recruiting new employees. The majority of employers rarely or never use the formal and public institutions for recruiting staff (i.e. jobcentres) and instead use informal or nonpublic methods for recruitment. Of the respondents, 50 per cent in the survey never use the jobcenters, 25 per cent rarely use them, 18 per cent sometimes use them, 5 per cent often use them and the remaining 2 per cent always use jobcenters when recruiting staff. In our survey, the most popular channels for recruiting new staff are instead private and public job search databases on the internet.
In the survey, there were three different indicators of participation in ALMPs: (1) whether employers have recruited one or more employees who were formerly long-term unemployed, (2) whether employers had recruited one or more employees in 'flex-jobs' and (3) whether employers had recruited one or more employees through the wage subsidy scheme.
Of the respondents, 13 per cent had recruited employees who were formerly longterm unemployed (defined as more than two years of unemployment); 23 per cent had recruited employees in 'flex-jobs'; and 31 per cent had recruited employees through the wage subsidy scheme. The main difference between flex-jobs and wage subsidies is that the former are permanent positions for persons with permanently reduced working ability, while the latter are temporary positions (maximum one year) for unemployed persons.
The indicators on attitudes and participation in ALMPs can now be combined to test the distribution of Danish employers among the four types of employers identified in the typology. The attitude variable was recoded to positive attitudes (great degree or some degree of confidence) or negative attitudes (low degree or no degree of confidence). Similarly, the first participation variable (how often the workplace recruited staff through the jobcentres) was recoded to either 'yes' (always/often) or 'no' (rarely/never). The three remaining participation variables (long-term unemployed, 'flex-jobs' and wage subsidies) were already categorised as either 'yes' or 'no'. In Table 2, the attitude variables and the participation variables are combined.
Contrary to the initial assumption, we see that the 'dismissive employer' is the most common type of employer, regardless of which indicator is used. For example, 54 per cent of employers are not confident that jobcentres will help them with the recruitment of staff (negative opinion) and do not recruit staff through jobcentres (no participation). There are also a high proportion of employers with negative attitudes who do not recruit the long-term unemployed (48 per cent), participate in recruiting employees with wage subsidies (43 per cent) or recruit employees in flex-jobs (39 per cent).
Approximately one third of the employers can be classified as 'passive employers'. They have confidence that the jobcentres will help employers recruit labour (positive opinion) but do not participate actively in ALMPs or recruit employees from the jobcentres (37 per cent, 39 per cent, 34 per cent and 30 per cent respectively for each of the four participation indicators in the table). This also implies that there is considerable potential for jobcentres to improve the awareness and relevance of the services they are providing and potentially turn 'passive' employers into 'committed' employers. Only a minority of the employers can be categorised as 'committed employers'. For example, only a small minority of the employers have confidence that jobcentres are helping employers to recruit staff (positive attitude) and actually recruit employees through the jobcentres (6 per cent) or have recruited long-term unemployed people (8 per cent). A slightly higher proportion of the employers have confidence that jobcentres will help them to recruit staff (positive attitude) and do actually participate in ALMP programmes (13 per cent participate in wage subsidies and 18 per cent participate in 'flex-jobs').
Finally, the 'sceptical employer' is the most uncommon type of employer. This type of employer does not have confidence that jobcentres will help in recruiting staff (negative attitude) but nevertheless participates in recruiting staff from jobcentres (3 per cent), recruits long-term unemployed people (5 per cent) or participates in wage subsidies (11 per cent) or 'flex-jobs' (14 per cent). Even if the sceptical employers currently participate, they may withdraw participation and become a dismissive employer instead.
In Figure 2, the four participation indicators have been grouped together, and the average distribution of employers have been calculated.
The figure shows that almost half of the workplaces can be categorised as 'dismissive employers' (46 per cent). Approximately one third of the workplaces can be classified as 'passive employers' (35 per cent), while the remaining workplaces can be classified as either 'committed employers' (11 per cent) or 'sceptical employers' (8 per cent).

Discussion and conclusions
There is a small but growing literature on employers and ALMPs and this article has contributed to that literature by proposing and testing a new classification of different types of employers. In the empirical literature on employer engagement in ALMP, attitudes and participation of employers are often conflated when they should be separated to allow for measurement and better targeting of ALMP interventions. A typology with four types of employers is put forward for national and cross-national comparisons of employer engagement in ALMPs. The typology contributes to a more nuanced understanding of why and how some employers participate in ALMPs while others do not, and thus allows for better targeting of employment and activation programmes towards different types of employers.
The utility of the typology was tested on the Danish case of employer engagement in ALMPs. The empirical findings did not support the initial assumption that the majority of Danish employers would support and actively participate in the maintenance and development of ALMP. Although the engagement of Danish employers in ALMPs is high in a comparative perspective, it is found that employers mainly participate in (subsidised) activation programmes rather than in ordinary (unsubsidised) recruitment. The findings indicate that there is considerable scope for improving the outreach activities of local jobcenters and for improving the participation and attitudes of employers in ALMPs.
A more systematic understanding of the preferences and behaviour of different types of employers is an important precondition for more effective and targeted ALMP programmes and interventions on the demand-side of the labour market.

Notes
1 In twenty-one countries where data are available in the Labour Force Survey of the European Union, on average only 9 per cent of individuals who have recently started their new job indicate an involvement of the PES (OECD, 2015: 136-7).
2 Martin (2004aMartin ( , 2004b compares employer participation in social policies in (corporatist) Denmark with (pluralist) UK. The empirical analysis draws on fifty-five Danish and fifty-two British interviews with human resource officers or CEOs conducted between 2000 and 2001. The sample was randomly selected from 200 companies drawn from the Danish newspaper Børsen top five hundred list of employers and the British Financial Times list. There was by implication some selection bias in the sample towards larger and probably more active companies. Making statistical generalisations about the entire population of Danish or British employers is also difficult due to the small sample size.