How can temporary work agencies provide more training?

Staffing agencies could play a more prominent role in the provision of training for the low qualified and previously unemployed

Alexander Spermann Spermann, A (University of Freiburg, Germany)

Elevator pitch

Temporary work agencies use training as a recruitment and retention argument when qualified labor is scarce. However, short job assignments present a major obstacle for employers and employees to increase investment in training. As temporary agency workers are mainly low-qualified and often previously unemployed, paid work in combination with training should lead to more sustainable employment. Adjustments in labor market institutions could make the provision of training more attractive for both staffing agencies and temporary agency workers.

Training plus temporary agency work
                        increase the employment probability for agency workers in Germany, 2011

Key findings


  • Scarcity of qualified workers and continuous pressure on updating skills drive demand for training.

  • Training serves as a self-selection and screening device as well as a recruitment and retention argument for staffing agencies.

  • Training allows agency workers to acquire competencies needed by user companies.

  • Temporary agency work allows workers to combine paid work with complementary off-the-job training.

  • Training for previously unemployed and low-qualified agency workers has positive external effects.


  • Return on investment in human capital cannot be captured by staffing agencies due to short job assignments.

  • Agency workers may quit after receiving training in favor of better paid jobs in a user company.

  • Training programs carry high administrative costs due to the large number of players involved.

  • Agency workers may be reluctant to invest in training due to short job assignments and insecurity regarding future labor market outcomes.

  • There is no guarantee that agency workers will receive higher pay after training, and agency training is no substitute for vocational or college education.

Author's main message

A scarcity of qualified workers with specific competencies drives demand for worker training. Staffing agencies contribute by allowing the previously unemployed to combine paid work with training. This training is predominantly offered in the form of short-term measures and e-learning to enable job assignments for low-qualified agency workers. The provision of training is associated with positive external effects. Therefore, governments should update labor market institutions in order to internalize these external effects. This will incentivize employers and employees to invest more in training.


Temporary agency work is characterized by a triangular relationship between agency workers, staffing agencies, and user companies. It consists, on the one hand, of an employment contract between temporary staffing agencies and temporary agency workers, and, on the other hand, an employee placement contract between temporary staffing agencies and user companies. Both contracts are preconditions for job assignments at user companies. In essence: Agency workers work for a user company but are employed by a staffing agency [2].

This business model has spread worldwide over the last decades. However, institutional arrangements differ enormously between countries. In some countries, such as Germany, staffing agencies act as employers for agency workers. According to German law, such agencies have to pay for sickness, vacation, and idle time. Conversely, in most other countries, staffing agencies are simply job placement agencies. According to French law, for example, they are only required to pay their agency workers in case of job assignments.

Although staffing employment has grown over the past decade, it has remained at very low levels in terms of overall employment share (the “penetration rate”). Figure 1 provides an overview of temporary agency work penetration rates between 2000 and 2013. In key markets, such as the US, Japan, and Europe, a maximum of four out of 100 employees are classified as agency workers.

Agency work penetration rates for selected
                        countries, 2000–2013

Agency workers are mainly young, male, low-educated, and often previously unemployed. They are predominantly employed in the industrial sector. As such, their job profiles change over time due to automation so that hard and soft skills have to be continuously developed. Agency work allows for on-the-job informal learning. Furthermore, training could also improve workers’ employability and increase their retention probability.

Discussion of pros and cons

This article explores the role of staffing agencies in the provision of training to low-skilled and unemployed workers. Relevant questions deal with the sufficiency of current training offered, the amount of investment being done by workers in their own human capital, and whether agency training actually benefits temporary workers. Furthermore, the role of governments with respect to agency training and issues regarding the provision of training during times of qualified labor scarcity are addressed.

The staffing agency’s perspective

Do staffing agencies offer enough training to their agency workers? According to standard human capital theory, firms will invest more in the human capital of their workers in proportion to a worker’s expected job tenure. In this way, the firm can profit from their investment over time. In today’s competitive world, firms will not offer general training because workers are free to leave. Firms have no guarantee that they will recoup their investment in general worker training. However, general training, such as for computer skills, might serve as a self-selection mechanism, i.e. more able workers will choose firms that offer training opportunities. General training might also be used as a screening device for worker ability as firms can learn more about workers’ productivity if workers demonstrate their skills by participating in general training programs [3].

Staffing agencies compete for appropriately skilled workers in an effort to fill vacancies at user companies. Matching labor supply to labor demand is at the core of their business model. If enough qualified workers are available, there is no need for training. However, if qualified labor is scarce, vacancies cannot be filled and revenues cannot be generated. Therefore, training can serve as an investment option to enable staffing agencies to fill this revenue gap. Furthermore, staffing agencies that offer training measures to their agency workers might be preferred by user companies due to assumed better pre-selection of candidates.

From a staffing agency’s perspective, training is useful, beyond the concepts of self-selection and screening, in the case of qualified labor scarcity. Training is an additional recruitment argument for those candidates that appreciate training as part of an overall employment package offered by staffing agencies. Training can also help increase the retention rate of incumbent agency workers. Nowadays, firms can no longer guarantee their employees’ long-term employment, and therefore might offer employability via training rather than job security. Thus, this implicit contract demands investment by employers in human capital. If this does not happen, the implicit contract is violated, resulting in declining employee commitment with the firm, and eventually, lower productivity [4].

Staffing agencies provide training in so-called monopsonistic markets, where firms exhibit extreme market power regarding hiring practices. This form of market power allows firms to pay workers less than what their productivity warrants. As a consequence, a gap between the output of the worker (increased by productivity enhancing training) and wages exists. Market-based firms pick the optimal level of training, where the additional costs of more training equal its additional benefits. Hence, in this setting, it is profitable to invest in the general training of their workers. Monopsony power may arise for two main reasons: (i) transaction costs related to matching and search frictions, and (ii) asymmetric information between the current employer of the worker and other firms. Labor immobility due to family obligations such as home care, high commuting costs (especially in rural areas), and high search costs (especially in regions with a high share of small- and medium-sized companies) might lead to high transaction costs. In the case of asymmetric information, the current employer is the only player that knows the exact training content offered to workers. Note that empirical evidence on monopsony power of firms is scarce [5].

Staffing agencies are further able to improve the impact of training by providing more relevant programs. In contrast to standard courses offered by training institutions, staffing agencies are very close to the companies that are in need of specific skillsets. Hence, they know exactly which skills are preconditions for agency workers to be assigned to user companies. Staffing agencies therefore have an incentive to provide relevant general and firm-specific skills to agency workers. An alternative option for staffing agencies is to hire workers with appropriate skills away from other firms by offering them higher wages. However, in times of qualified labor scarcity, this is not often a realistic option. Even if qualified workers could be hired away, the higher wages lead to increased prices for user companies. However, user companies may not want to rely on temporary agency workers if personnel services become too expensive. They may instead think about direct hiring. Therefore, staffing agencies will have to decide on how to meet user companies’ demands; training costs for agency workers and wage costs for hiring will be compared in a dynamic framework. Eventually, the user company will pay the price for filling vacancies, be that via staffing agencies or direct hiring.

The overall costs for training depend on the size of the firm. The smaller the firm, the higher the training costs per person relative to overall revenues will be. Training is typically provided via e-learning platforms that must be set up and maintained, or alternatively by external training providers that have to be selected and monitored by staffing agencies. If external funds via collective labor agreements or active labor market policy instruments are available, co-funding must be organized. Typically, staffing agencies have to organize the collaboration of several players such as agency workers, user companies, third-party funding institutions, and external training providers. These additional transaction costs may lead to prohibitive overall training costs, especially for small- and medium-sized staffing firms. Indirect costs may occur if agency workers quit their contract with the staffing agency after completing training and then continue working, either with a competitor or with a user company. In fact, this last point can be described as a positive external effect of training, which is captured by third parties. A small firm will be hit much harder by such behavior than a larger firm due to the higher relative costs.

In practice, staffing agencies have been well known for providing general and, to a smaller extent, firm-specific training. E-learning courses for office management software or foreign languages as well as forklift driver licenses are standard examples of general training. Internal training courses that enable agency workers to run specialized machines are examples of firm-specific training measures [2]. However, due to high transaction costs and positive external effects of training (i.e. potential lost returns on investment for the agency providing the training), staffing agencies do not offer enough training.

The agency worker’s perspective

Do agency workers invest enough in their own human capital through training? Temporary agency work has often provided opportunities for the previously unemployed to enter the labor market. While evidence suggesting that agency work is a stepping stone to regular employment has proven to be mixed (i.e. a bilateral rather than triangular employment relationship) [6], agency work is associated with work experience and on-the-job informal training.

Considering a decreasing half-life of knowledge (i.e. knowledge diminishes over time in an increasingly rapid manner), the relative importance of hard and soft skills compared to once-acquired qualifications has grown. Additionally, demographic challenges (e.g. in Germany and Japan, where the workforce is becoming increasingly older) make qualified labor scarcer than it has been in the past. This development suggests that temporary work agencies, which offer not only jobs but also more training for previously unemployed and low-qualified workers, should provide these workers with more stable employment.

From an agency worker’s perspective, paid work could be combined with off-the-job training (such as e-learning) so that further training is not associated with lower net income. However, as is so often the case, there is no free lunch; off-the-job training is associated with less leisure for agency workers. Therefore, even if employers do provide free training opportunities, workers have to invest their leisure time in order to take advantage of them. These represent opportunity costs for investment in the worker’s own human capital. Some agency workers will likely undervalue the potential benefits relative to costs. They may underestimate the effects of training on future labor market outcomes such as employability and sustainability of employment. But, it will be difficult for agency workers to understand why they should invest in training programs, particularly if job assignments are very short.

Furthermore, training measures such as e-learning or forklift driver licenses are certainly not a substitute for vocational or college educations. This type of agency training may help to remain employable in a dynamic labor market with fast-changing skill requirements, but the return on investment in human capital is very small in comparison with traditional schooling. Furthermore, it is not clear if agency workers experience an immediate increase in earnings after completing the training. It may even be—in the case of a firm exhibiting market power—that the workers will be paid below their productivity, even after training [5]. Overall, it is apparent that agency workers do not invest enough in their own training.

The state’s perspective

Should governments intervene to increase the amount of training provided by staffing agencies? General training is associated with positive external effects due to spillover effects for other companies, i.e. one firm invests in the human capital of its workers, but other firms that hire them benefit from that investment. Since staffing agencies do not account for these spillover effects, they will provide a suboptimal amount of training, or training incentives may disappear altogether.

One way to internalize external effects is to create funds at the national level, allowing firms to subsidize their training programs using state-supported means. Bargaining partners from several sectors have set up training funds in seven European countries: Spain, France, Italy, Belgium, Germany, Luxembourg and the Netherlands. A convenient example is Belgium, where bargaining partners in the staffing industry established a training fund for temporary agency workers in 2006. The fund is financed by staffing agencies, which contribute 0.4% of their total wage mass to a sectorial social fund, and an extra 0.1% for specific target groups. Training must take place during paid working hours. Both the cost of training and the wages of agency workers can be reimbursed, with the maximum recovery amount being €1,725 per agency worker [7].

However, most countries rely on company-based solutions, if any such support exists. In the US, training is primarily delivered as e-learning courses. The “big three” companies in staffing—Adecco, Randstad, and Manpower—offer thousands of online courses for their staff [4]. These activities reflect a world without internalized external effects, i.e. a suboptimal provision of training is observed. In Germany, however, short-term off-the-job training courses co-financed by public employment services via training vouchers are a common practice by big staffing agencies. Vouchers provide convenient instruments to internalize external effects. They are only issued for certified programs and exclusively offered by certified training institutions to avoid misuse. Participation is monitored and outcomes are evaluated on a regular basis to maximize impact [2].

Evidence on the effectiveness of training programs for temporary agency workers

Even if subsidies to correct market failures are theoretically justified, it is relevant to know whether training provided by staffing agencies has positive causal effects. Currently, the majority of available data on this topic comprises descriptive statistics based on surveys provided by federations and anecdotal evidence based on best-practice examples. Candidate training, ex-ante qualifications tests, forklift driver licenses, documentation of competencies, off-the-job training schemes, partial qualification schemes, part-time vocational education, full-time vocational education co-financed by public employment services, and certificates for e-learning courses represent just some of the enormous heterogeneity of training measures that can be found in practice [2]. However, it is difficult to determine the success of these programs as there are only a few studies that try to solve the fundamental evaluation problem, i.e. a person cannot be observed in two states simultaneously (participating in a program versus not participating in a program). Therefore, a comparison group is needed that serves as a proxy for the unobserved counterfactual behavior of participants, i.e. what would the outcome have been had the treated group not participated in the program. This is a precondition for measuring the causal effect of a training program on participants.

One field experiment from the Netherlands identifies the causal effects of training on productivity by randomly assigning workers to treatment and control groups. The field experiment was conducted in a multinational telephone company’s call center. The treated group, i.e. participants of the program, and the control, i.e. non-participants, were all call agents working for a multinational temporary staffing agency [8]. The treatment consisted of a one-week training course in conversation techniques for call agents. The main goal of the course was to decrease the average handling time of calls while maintaining quality. One hundred and seventy-nine agents participated in the field experiment, 86 were assigned to the treatment group, and 93 were assigned to the control group. The main key performance indicator was agent productivity, which was defined as the average handling time, i.e. the average time an agent needs to handle a customer call. It turned out that treated agents perform 10% better than non-treated agents [8]. This result provides hard evidence based on a robust evaluation method rather than descriptive or even anecdotal evidence. It highlights the efficacy of training for agency workers.

Another study used a pilot program for low-skilled unemployed youths in Germany to evaluate the joint effect of classroom training, individual coaching, and temporary agency work. The program was set up as a public-private partnership between the public employment services and a private temporary work agency. The staffing agency conducted individual profiling and skills assessment, followed by classroom training. After this training period, participants received an employment contract from the staffing agency and were assigned to user companies. The whole program lasted for up to 12 months.

To identify the causal effect on participants’ post-treatment employment probability, a comparison group was constructed. Due to capacity constraints, access to the program was denied to some eligible youths, which meant that interviewed case managers could describe a comparison group based on the participants’ observable and unobservable characteristics [1]. Administrative data on post-employment were available from the federal employment services. The outcome of interest was the long-term effect on employment probability after 18 months for those participants who stayed in the program for at least 6–12 months. The treatment effect found in this study is large (50 percentage points) during the first four months of the observation period, and remains significantly larger than zero for the whole duration. The authors conclude: “The program clearly and sustainably achieves the aim of increasing the employability of those participants that run through all of its components” [1]. This study highlights another successful example showing the type of positive impacts that training programs can have for low-skilled workers.

An empirical study based on survey data from 290 staffing agencies in Germany sheds more light on the advantages of training for staffing agencies and agency workers. It turns out that investment in training pays off in the form of higher fees and increased attractiveness for the agencies from both clients’ and workers’ perspectives. Training intensity is higher for more highly qualified agency workers. Furthermore, training is positively related to the competitiveness of staffing agencies [9].

Another recent study highlights the effect of on-the-job informal learning during temporary work [10]. The authors use representative OECD survey data for 24 industrialized countries in 2012, which allows measurement of the intensity of on-the-job informal learning, i.e. learning-by-doing, learning from others, and learning by keeping up-to date with new products or services. They find that workers in temporary jobs are more engaged in informal learning than permanently employed workers, although they are less likely to be involved in formal training activities. In addition, they find that training and informal learning are complementary activities [10].

If training for agency workers by staffing agencies leads to positive causal effects for the worker, how could staffing agencies be encouraged to provide more training to low-qualified workers in times of qualified labor scarcity? Overall, a major drawback for staffing agencies is the uncertainty regarding whether their return on investment in human capital will be positive in the case of short job assignments. Few staffing agencies interpret training as a component of their overall investment strategy, where the expected return may be captured by an intensified customer relationship with relatively more revenues and profits in the future. More recent business models have emerged, especially in the US, such as the so-called “managed service providers” and “blended workforce solutions,” which focus much more clearly on the long term. The big players in the staffing industry, who offer worldwide human resource solutions for international user companies, are mainly driving these developments. However, this trend is still not the norm for most staffing agencies.

Limitations and gaps

Empirical literature on the effects of training in connection with temporary agency work is still rare. Currently, selected surveys provided by federations and anecdotal evidence based on best-practice examples comprise the bulk of available data. However, some high-quality studies that take the fundamental evaluation problem into account highlight the positive causal effects of training provided by staffing agencies [1], [8].

Future research should include cost–benefit analyses. Even if training measures are effective, it might be that they are not cost-efficient. From a fiscal point of view, it is important to know if deadweight or displacement effects play a role. Furthermore, research periods are often too short. Additional knowledge about the long-term effects of training would be useful. Learning more about the sustainability of employment and higher pay as a result of training would also be of great interest. Closer collaborations between staffing agencies, user companies, and research institutions would help to collect more data, to develop more elaborate research agendas, and to apply more sophisticated research methods.

One further obvious gap is the lack of reliable sector-specific information on the extent of training. Comparisons among sectors are hard to find on a country-level basis, and are impossible across multiple countries. This is partly due to the triangular relationship at play, where training is often provided as a joint product by staffing agencies and user companies, sometimes even done together with other stakeholders such as the public employment services or external training institutions.

Summary and policy advice

In a world with a scarcity of qualified labor, temporary agency work can provide more than just on-the-job training for previously unemployed and low-qualified workers. On the one hand, job assignments are generally short in duration and overall costs for providing training are high, meaning that staffing agencies may not be able to capture the return on their investment in human capital. If user companies do not reward these efforts with higher fees, and if agency workers fail to express their appreciation through higher retention rates, then training will not be profitable for staffing agencies. On the other hand, training is associated with positive external effects, such as staffing agencies’ risk of lost return on human capital investment, and some evaluation studies find positive causal effects of agency work combined with training on the employment probability of agency workers. This result indicates that workers benefit from training programs, even if the staffing agency loses out.

New institutional arrangements could encourage staffing agencies to play a more important role in training. For example, if governments correct market failures that are linked to training’s positive external effects, then the modification of labor market institutions might facilitate more frequent training provision. Publicly co-financed training schemes such as the German training vouchers or “conditional on training” employer subsidies could also help foster increased training. In addition, more effective contractual arrangements could be designed to bind trained agency workers to the firm that provides this training, reducing the financial risk of providing training for staffing agencies. Convenient labor law modifications could make binding agreements more reliable without undermining workers’ rights. In sum, there are numerous ways in which staffing agencies could provide additional, effective training for low-skilled and previously unemployed workers.


The author thanks two anonymous referees and the IZA World of Labor editors for many helpful suggestions on earlier drafts.

Competing interests

The IZA World of Labor project is committed to the IZA Guiding Principles of Research Integrity. The author declares to have observed these principles.

© Alexander Spermann