Estudios sobre la Economía Española - 2021/08

Evaluation of internship contract in Spain

Sara de la Rica

(ISEAK y UPV)

Lucía Gorjón

(ISEAK y Fedea)

Febrero 2021

fedea

Las opiniones recogidas en este documento son las de sus autores y no coinciden necesariamente con las de Fedea.

Evaluation of internship contract in Spain

SARA DE LA RICA (ISEAK - UPV)

LUCÍA GORJÓN (ISEAK - FEDEA)

ABSTRACT

The internship contract (IC) was originally designed to promote an appropriate employment transition for young people with higher education from their formative years to their working lives. Employers are supposed to invest in human capital and, in correspondence, they can benefit from lower wages and from tax reductions. This report evaluates the impact of this Active Labor Market Policy. The evidence shows that the use made of the IC is contrary to the spirit of the law. First, employment under the IC reduces the probability of being retained by the firm; second, for those who stay in the firm after the IC, their probability of signing an indefinite contract is significantly lower than that for those who sign another temporary contract. However, for those who move to another firm, the probability of signing an indefinite contract is higher if they come from an IC. These results suggest that firms do not use IC as an instrument for investing in training young people, but rather as an instrument for reducing labour costs and/or increasing contractual flexibility..

Sara de la Rica (ISEAK - UPV)

Lucía Gorjón (ISEAK - FEDEA)

The Youth employment partnerSHIP” project is funded by Iceland, Liechtenstein and Norway through the EEA and Norway Grants Fund for Youth Employment

Index

1. Introduction......4 2. Labour market context for youth in Spain......7 3. Internship contract: legislative issues......11 4. Evaluation......13 4.1. Dataset......13 4.2. Descriptive analysis......15 4.3. Empirical strategy......22 4.4. Results......23 5. Conclusions and recommendations......30 Bibliography......33 Annex – Complete estimations results......34

1. Introduction

In recent years, much of the debate on economic policy in Spain has revolved around the structural weaknesses of its labour market. From the industrial reconversion in the 1980s to the hangover of the Great Recession of 2008, our labour market has suffered fundamentally from two anomalies that drag down labour productivity in a substantial way: on the one hand, the high and persistent levels of unemployment, which reach 25% in recessive times and does not fall below 8% in times of great economic boom; on the other, the contractual duality, which causes the existence of relatively wel protected against unemployment workers due to the fact that their contracts are indefinite with high dismissal costs - the insiders - compared to another group - the outsiders - who are characterized by having temporary contracts, of short duration, which progressively make their employment situation more precarious. This job instability mainly affects young people, women and immigrants.

Successive governments have tried to tackle both structural problems, albeit with little success. The best tools available to combat these problems are Active Labor Market Policies (ALMP), interventions aimed at combating persistent unemployment and its chronification in certain groups. These policies make up all those interventions in which the competent administrations in labour matters work directly with unemployed people to improve their possibilities of labour insertion, with employed people to complement and improve their training and / or with companies for the continuous adaptation of their workers, as well as to motivate the hiring of certain groups at risk. They are, therefore, complementary actions to passive employment policies, which mainly consist in unemployment benefits. Institutions thus take an active role in the proper functioning of the labour market. Since the beginning of the 21st century, most European countries began to develop this type of support, guidance and improvement tools for the training of the workforce. However, as Kluever (2007) shows, Spain did not internalize its importance in the first decade of the 21st century, as did most of the surrounding countries. These tools did not begin to develop until 2010, when the Great Recession hit the labour market with particular violence, exposing the great problems that had remained during the economic expansion. Since 2012, Active Employment Policies have taken a central roll on the agenda of policy makers in our country, mainly at the regional level, since it is the Autonomous Communities who (since 2015) manage the ALMP exclusively.

Apart from the little importance given to the ALMP, Kluever (2007) documents by means of a European comparison that in our country, 40% of ALMP focuses on the subsidy for hiring certain groups. Contracting is thus encouraged by reducing or eliminating the contribution to social security for each person hired (even part of the remuneration is subsidized on certain occasions). This type of incentive clearly lowers recruitment costs, but its effectiveness in job creation is unproven. In fact, there is evidence (see García

Perez & Rebollo (2009a, 2009b)) about the ineffectiveness - dead weight - of these hiring incentive policies when creating jobs, since a large part of them would have been created also in the absence of the aid. In this way, they would be only working as a subsidy to the firms, but not increasing the job creation.

It should be noted, however, that there are many types of hiring subsidies, aimed at different groups. To assess their effectiveness, it would be necessary to proceed to a rigorous evaluation of each of them. The systematic evaluation of ALMP is undoubtedly a pending task in our country, partly due to the lack of adequate information to carry them out, and partly due to the lack of a generalized culture in favor of policy evaluation.

This document aims to alleviate this deficiency, providing knowledge about the impact of a specific type of subsidy: the internship contract. This contract is aimed at people under the age of 30 with higher education and was created with the purpose of offering an adequate transition for young people between the training stage and the work stage.

Internship contract's main goal is to increase labour stability in the area of studies developing professional practices in relation to the level and field of studies and giving incentives to firms to hire young workers once the internship period has finished. More precisely, the employer is obliged to provide the worker with training and effective work suitable for the purpose of the contract. The employer must also protect the development of the training process. The company offers young people monitored practices that allow them to adapt to the work environment, and in return, for the training investment made, the employer receives a bonus and the possibility of offering lower salary than the dictated by the collective agreement. The spirit of the law with regards to IC is to provide young workers with high educational attainment with the practical skills needed to be fully productive. Employers invest in specific human capital through the practical training, and in correspondence, can benefit from lower wages and from tax reductions. In principle, one would expect that this practical investment should lead to higher levels of employment stability for those who enter the labour market under the IC. The aim of the paper is precisely to test this hypothesis. Surprisingly, in despite of its long path, as far as we know, such evaluation has never been done before.

As we mentioned above, there is little previous research on the impact of ALMP in Spain, and especially on policies affecting younger workers. The study (Jansen and Troncoso, 2018) that is most directly related to our own examines the impact of the apprenticeship contract (AC) on the labour market of young people in Spain. The authors found that this type of contract is not frequently used, and that using the AC rather than other temporary contracts of similar length did not significantly reduce the level of employment precariousness of younger workers. Moreover, while the European Commission (2012) reviews the dual learning plans of 24 member states, the evidence provided of their effects was very limited. Thus, the present study is addressing an issue that has, up to now, not been adequately investigated.

This paper aims to help fill the existing gap in research on ALMP in Spain, and, in particular, on ALMP directed at younger individuals, by evaluating the impact of the IC in Spain. Since the aim of the IC is help young people working in an entry-level job achieve job stability thereafter, either at the same firm or at another employer, we evaluated the impact of the IC on two particular outcomes: i) the probability of staying at the same firm after the IC had expired; and ii) the probability of signing an indefinite contract after the IC had expired at the same firm (for those who stayed) or at another firm (for those who left). The gender perspective is taken into account throughout the analysis.

The report is structured as follows. Section 2 shows evidence on the degree of job instability that young people face in Spain. Section 3 describes the legislative aspects of the internship contract. Section 4 explains the dataset that we use in the empirical analysis; then it develops a descriptive analysis of the use of the IC by comparing the subsequent labour trajectories of younger workers who were employed under an IC with those of younger workers who were employed under a regular temporary contract, rather than the IC; third, Section 4 describes the empirical strategy used in the analysis and, lastly, results are presented. Once the diagnosis has been reached, the last section closes with some conclusions of interest and presents proposals for public policies to be considered.

Labour market context for youth in Spain

The Spanish labour market suffers from two main dysfunctions: The first of these dysfunctions is its high and volatile rate of unemployment. Given that the labour market adjusts to the economy through the extensive, rather than the intensive margin, the unemployment rate tends to increase when the economy faces an adverse situation, and tends to decrease greatly when the economy faces an expansion. Still, on average, the unemployment level in Spain remains higher than it is in other European countries. Figure 1 shows that 12 years after the start of the Great Recession in 2008, the level of unemployment is 14%. Thus, after Greece, Spain has the second-highest unemployment rate among the EU-28 countries.

Figure 1. Unemployment rate by country of eat EU-28 in 2019

As we noted above, the unemployment rate in Spain is not just high, it is volatile. While unemployment rates tend to be countercyclical, the patterns in Spain are exceptionally unstable. As Figure 2 shows, unemployment rates increase sharply during recessions, and decrease greatly during expansions. This high degree of volatility is particularly acute for young workers. At the peak of the last recession, the unemployment rate reached 43% for young workers (27% on average), whereas for workers aged 45 older, the rate was only half as high.

Figure 2. Unemployment rate in Spain by age

Source: Spanish Labor Force Survey, second term.

This pattern of extreme volatility in the unemployment rate is strongly connected to the second dysfunction of the Spanish labour market: its high temporary work rates. Because the Spanish labour market adjusts to shocks by hiring and firing, the percentage of temporary workers among those who are hired and fired is very high. Indeed, the rates of temporary employment in Spain are the highest in Europe, as Figure 3 depicts. Out of 100 salaried workers in Spain, 22 work under a temporary contract.

Figure 3. Temporary work rate by country of at EU-28 in 2019

Source: Spanish Labor Force Survey, second term.

These high temporary employment rates affect young workers in particular. For this reason, the Spanish labour market has been described as a dual labour market: one part of the workforce is made up of so-called insiders, who are typically workers over age 35 with indefinite contracts and strong protections against unemployment; while the other part of the workforce is made up of so-called outsiders, who are mostly younger workers with very unstable contracts, and almost no protections against unemployment. This duality in the labour market is the result of a labour market regulation enacted in 1984, which enabled employers to hire workers under temporary contracts without cause; i.e., even if the nature of the job was not temporary. Prior to 1984, employers were permitted to hire workers under temporary contracts only if the nature of the job was temporary. At that time, the Spanish labour market was suffering from high levels of rigidity, and the unemployment rate was high. The aim of the reform was to enhance flexibility and to facilitate hiring. The change in the law in 1984 led to a very strong and steady increase in the number of workers hired under temporary contracts. By the early 1990s, 30% of all new employment contracts were temporary.

Since the mid-1990s, a number of legal reforms have been implemented that incentivise employers to hire workers under indefinite rather temporary contracts. However, none of these reforms has been successful, and the dual labour market continues to affect 25- 30% of the Spanish workforce. This high rate of temporary employment means that there is more flexibility on the demand side, which is highly desired by Spanish firms. However, it also means that employers have fewer incentives to invest in training. In addition, if workers do not expect to be working at the firm that currently employs them over the long run, they may put in less effort. Thus, the career expectations of temporary workers tend to be low, and they are incentivised to look for better job opportunities at other firms.

Figure 4 displays some astonishing numbers. For example, in 2019, workers under age 20 had an extremely high temporary employment rate of 85%. The rate decreased with age, but almost half of workers between ages 25 and 29 were working under a temporary contract.

Moreover, the duration of these temporary contracts is often quite short, especially for young workers. Figure 5 shows the distribution of the first contracts signed by younger people in Spain between 2007 and 2017. The most frequent contract duration (27%) was between one and three months. However, 37% of younger workers entered the labour market through a contract shorter than one month. Thus, it is clear that younger workers in Spain are exposed to high levels of instability.

Figure 4. Temporary rate in Spain by age

Source: Spanish Labor Force Survey, second terms.

Figure 5. Duration of first contract signed in Spain

Source: Contracts administrative database, 2019

Both the high temporary rates and the short duration of the contracts severely damage the employment opportunities of young workers. Tied to this labour instability, during the first years in the market, lower salaries are charged and the incorporation into stable jobs with better working conditions is delayed, thus weighing down the entire labour career and even impacting on future retirement, the well-known scar effect.

There have been several policy interventions aimed at improving the entry of young people into the labour market. Valdés & Lahera (2010) offer a good overview of all the reform attempts since the 1990s to achieve greater employment stability, as well as their results. One of these measures is the proposal of the internship contract, which is presented and evaluated below.

3. Internship contract: legislative issues

Both the internship and the apprenticeship contract were established in Spain in 1998 as part of Royal Decree 488/98.¹ However, the legislation was later modified on several occasions, most fundamentally with Organic Law 5/2002.² The next major changes were enacted with the reform of 2015.³ The most recent reform was in 2018⁴, when the reductions in social security contributions for internship contracts were eliminated. In this section, we will delve into the details of the reform of the Workers' Statute of 2015 with respect to the internship contract and will outline the legal changes in the IC that have occurred since that time.

The aim of the IC is to enable young people who have finished their tertiary or professional training education to obtain the professional practice they need start their professional careers. This professional practice must be accompanied by investments in training by the firm that employs them. More precisely, the employer is obliged to provide the worker with training and effective work appropriate to the purpose of the contract. Moreover, the employer must also protect the development of the training process. The spirit of the law that implemented the IC is that employers are expected to invest in human capital by offering practical training, and will, in return, benefit by paying lower wages and taxes.

The eligibility restrictions refer to education and age: to qualify, the worker must have completed a bachelor’s or vocational training degree less than five years previously or be under age 30. The IC has a minimum duration of six months and a maximum duration of two years. Furthermore, up to two extensions of the IC of six months each (never shorter) are permitted. The position can be full or part time. After a worker has been employed under an IC, she or he cannot be hired under an IC at the same or at a different company for more than two years.

¹ Real Decreto 488/98, de 27 de marzo, por el que se desarrolla el artículo 11 del Estatuto de los Trabajadores en materia de contratos formativos

² Ley Orgánica de las Cualificaciones y de la Formación Profesional (Ley Orgánica 5/2002, de 19 de junio)

³ Artículo 11.1 del Real Decreto Legislativo 2/2015 de 23 de octubre por el que se aprueba el texto refundido del Estatuto de los Trabajadores. Las principales reformas intermedias son el Real Decreto 170/2004, de 30 de enero y la Ley 43/2006, de 29 de diciembre

4 Real Decreto-ley 28/2018, de 28 de diciembre

If a worker has reached the two-year maximum duration of the internship contract, and the company does not notify the worker that the employment relationship is being terminated, it automatically turns into an indefinite relationship. Thus, if the company wants to terminate the employment relationship, the employer must notify the employee of its intent when the contract expires. The worker has no right to compensation upon termination of the contract. In addition, if the contract has a duration of more than one year, a 15 days’ advance notification to the other party is compulsory in the event of termination. There is no such obligation for internship contracts with a duration of one year or less. Failure to give this notice will result in an indemnity equivalent to the corresponding salary. This obligation applies to both the employer and the employee; i.e., if the employee does not wish to renew the contract, she or he must notify the employer, and give notice if the contract is for more than one year.

The incentives for firms to use this type of contract instead of any other type of temporary contract are reductions in wages and reductions in social security contributions (eliminated in 2018). The wage reductions can be as high as 40% of the ordinary wage during the first year and 25% during the second year, relative to the salary set in the collective agreement for a worker who performs the same or an equivalent job. The intent of this wage reduction is to provide the employer with an incentive to invest in specific human capital, the cost of which should, according to standard economic theory, be shared between the employee and the employer. Moreover, the employer can claim tax reductions of up to 50% of social security contributions if (i) the worker had been previously registered as a jobseeker in the public employment system, and (ii) the worker is younger than 30 years old (or has finished the degree less than five years previously). Furthermore, there are bonuses for the conversion to an indefinite contract of €500 per year for males and €700 for females for the first three years.

There is another motivation for firms to use this type of contract: according to Spanish labour law, the maximum period a worker can be employed under a temporary contract at a given firm is two years, but this restriction does not apply to the IC. This means that after a young worker has been employed at a firm for two years under an IC, she or he can be hired for two more years under a regular temporary contract. Thus, employing workers under the IC can enhance a firm’s flexibility.

The main disincentive to hiring young people under the IC that firms in Spain face is that the duration of the temporary contract cannot be less than six months. As Figure 5 shows, less than 20% of young individuals enter the labour market with a contract of at least six months. It is a common practice in Spain for younger workers to sign several consecutive temporary contracts that are less than six months in duration each, rather than signing one longer temporary or indefinite contract.

4. Evaluation

4.1. Dataset

CONTINUOUS SAMPLE OF WORK HISTORIES

The dataset we use is called the Continuous Sample of Work Histories (CSWH), which consists of the social security records of a representative 5% sample of the Spanish population. The wave we use corresponds to the year 2017 (and thus refers to the period prior to the last legislative reform in 2018). This dataset includes the complete labour market histories of a representative sample of Spanish workers. This longitudinal dataset is very rich because it includes information on every employment, unemployment, and inactive spell the individuals in the sample had between the time they signed their first contract and 2017. Information on the individuals’ demographic characteristics, such as their gender, age, residence, and nationality, is also included. Moreover, the dataset includes information on labour market variables related to the individuals’ complete working lives, such as the sectors of activity and the sizes of the firms that employed them. Of particular interest for our purposes is that the dataset contains detailed information on the types and the durations of the contracts under which the individuals in the sample had been employed.⁵

CONSTRUCTION OF OUR DATASET

To develop the analysis of interest, we focus on all individuals who start working between 2002 (the first main legislative reform) and 2017 (prior to the last legislative reform). Given that the IC is a contract directed at younger individuals, we drop from the dataset all individuals who are older than 30 years old when they start working, mainly because they are not potential users of the internship contract. For the same reason, we restrict our analysis to individuals with higher educational levels. Since we want to measure the impact of the IC in the workers’ future labour market careers – in particular, whether their labour market stability increased – individuals who had signed only one contract are removed from the dataset. Consequently, we end up with a longitudinal dataset of a representative sample of young people with at least two employment episodes who are potential users of the IC over the 15-year study period (from 2002 to 2017).

⁵ For a detailed description of this dataset, see Duran (2007) and García-Perez (2008).

DEFINITION OF TREATMENT AND CONTROL GROUPS

To carry out the analysis, three interest groups are defined. The treatment group consists of those young people who sign an internship contract in the study period and meet the requirements specified above. Since the aim of the IC is to enable the worker to obtain professional practice appropriate to her or his completed level of (vocationa training or university) education, young people are expected to seek employment under an IC right after they graduate. However, the dataset provides evidence that many young workers sign an IC after having been previously employed, usually under another type of temporary contract. For this reason, we distinguish between two types of treated individuals: Treatment Group A includes all young individuals whose first contract is an IC; while Treatment Group B consists of young individuals who sign an IC after having previous work experience. These two groups of workers are both “treated” individuals, as both groups are employed under an IC contract. However, we make a distinction between them when testing the extent to having or not having previous work experience affected their subsequent job trajectories. In addition, as our aim is to conduct a clean analysis of the impact of the contract, we restrict Treatment Groups A and B to those individuals who signed a single IC.

However, in evaluating the effectiveness of the internship contract, we will pay particular attention to Treatment Group A – that is, those individuals whose first work experience is an internship contract – and compare their subsequent trajectories with those of a group with similar educational backgrounds and ages, but whose first work experience was under a standard temporary contract⁶ (and who never sign an IC). They will be defined as Control Group C: i.e., individuals who started their careers under a temporary contract lasting longer than three months. We eliminate from the control group those individuals who sign an IC in the future, because they belong to Treatment Group B. Thus, the three groups are exclusive. We restrict the minimum duration to three months (and we do not limit the maximum duration of the temporary contract) to ensure greater comparability with the group whose first work experience was under an internship contract, and whose expected duration is at least six months.

In order to check the robustness of the analysis, a similar evaluation is conducted that compares labour market entrants employed under internship contracts with labour market entrants employed under temporary contracts longer than six months, instead of limiting the duration to more than three months. We call this second comparison group Control Group D: i.e., young workers who started their careers under a temporary contract lasting longer than six months. For the main analysis, we focus on contracts longer than three months instead of six months because, as Figure 5 shows, not many young people entered their first job with a contract of more than six months. However, we expect that the results obtained when selecting one control group or another would be very similar.

All types of temporary contracts are included except the apprenticeship contracts and, of course, the internship contracts In addition some of the temporary contracts in Spain can be open-ended These are also included as long as their real duration is longer than three months.

4.2. Descriptive analysis

INITIAL DESCRIPTIVES

We begin by describing the use of the internship contract in Spain. Between 2002 and 2017, 21,768 individuals in the CSWH signed 26,212 internship contracts in Spain: 5,834 as a first employment episode and 15,934 after having previous labour experience. This means that 73% of the young workers who sign an internship contract in Spain over our study period has worked previously. Unfortunately, the database does not allow us to identify the moment the individuals completed their studies. Thus, we cannot differentiate between the jobs the individuals held before and after they graduate.

The disparity between the number of individuals in the sample and the number of contracts they signed indicates that it is not exceptional for young people to sign several internship contracts. Indeed, we find that 20% of the young workers has signed several internship contracts. Moreover, among the 15,934 individuals with previous experience who signed an IC, 1678 (10.5%) has previously signed an indefinite contract. In addition, 1,623 individuals (another 10%) have worked at the same company before they sign the internship contract. Finally, it should be noted that, on average, there is a gap of four years between an individual’s first work experience and the point at which she or he signed an internship contract. Excluding periods of unemployment or inactivity, these 15,934 individuals have an average of 561 days (roughly one and a half years) of work experience prior to being hired under an internship contract. Given these findings, it does not seem likely that the IC provided these young individuals with their first opportunity to sign a labour market contract, as we would have expected. We continue with the descriptive analysis in order to explore more deeply the use of this type of contract.

DURATION OF THE INTERNSHIP CONTRACT

It is worth remembering that internship contracts have a minimum duration of six months and a maximum duration of two years, with a maximum of two extensions of six months each. The CSWH includes information about the actual duration of each contact, not about its signed/expected duration. Consequently, we are able to observe the total duration of the IC, but not the extensions. However, we are able to identify how many contracts finished at the “expected” date or before it, i.e., with a final duration of six, 12,

18, or 24 months (as the legislation prescribes), or of durations in between. We include this category in the descriptive statistics because finishing or not the contract might have affected the individual’s future employment episodes. For example, some young individuals might choose to terminate their IC if they find a better job somewhere else.

As Figure 6 shows, it is a common practice for internship contracts not to be completed, and it is especially noteworthy that more than a quarter of those we examined were terminated before the minimum period established by law (six months) has ended. In total, more than 50% of the ICs are terminated before the minimum period established by law. In other words, less than half of the ICs are completed. In addition, we see that the likelihood of terminating the IC prematurely did not depend on whether the internship contract is signed at the beginning of an individual’s work career (Treatment Group A), or after the individual has worked previously (Treatment Group B).

Figure 6. Duration of internship contracts

LABOUR MARKET TRAJECTORIES

Next, we compare the labour market trajectories of the treatment groups to find out whether having or not having previous job experience affected their decisions and employment opportunities after signing an IC. As we explained before, given that entering into an IC entails, at least in theory, that the firm will invest in training, we would expect to observe that the individuals who started their labour trajectories through an IC enjoy higher levels of job stability thereafter. For this reason, our first variable is the staying rate; i.e., the proportion of individuals who stays at the firm that employed them under the IC. Additionally, we look at the type of contract the individual sign after being employed under the IC in order to determine whether the individual achieved stability (by signing an indefinite contract at the same firm) or did not achieve stability (by signing another temporary contract at the same firm after being employed under the IC). To complement this analysis and to ensure comparability, we also present the distribution of the contracts signed by those individuals who do not stay at the firm after the IC is terminated.

To sum up, Figure 7 provides the following evidence for these two treatment groups in any period between 2002-2017:

• Whether or not the IC is terminated before the expected completion date.

Whether or not the young person who is employed under an IC, (i) stay at the firm, (ii) move to another firm, (iii) move to unemployed or (iv) the IC employment period is the last observed employment spell.

• For those who stay at the firm or move to another firm, we display the type of contract.

Figure 7. Labour trajectories after the IC

Treatment Group A - without previous experience

Focusing first on those individuals who entered the labour market under an IC (Treatment Group A), we can see that 7% is still working at the end of 2017, which is the last observational period. Looking at those individuals who finish the IC on the expected date (except for those individuals who are not observed later on (5%)), we can see that approximately one-third continue working at the same firm, one-third move to another firm, and one-third become unemployed. However, the distribution was very different for those individuals who finished the IC before the expected date: of these workers, 73% move to another firm, 16% sign a new contract at the same firm before the IC finished, and 5% become unemployed. Hence, one important conclusion we can draw from this description is that a high percentage of young individuals move to another firm before the IC was completed. We consider this to be an unexpected stylized fact.

Next, we examine the type of contract the young individuals sign after being employed under the IC. Figure 7 shows that among those who stay at the same firm after the IC expires, only 23% sign an indefinite contract after being employed under the IC. This is also an unexpected result, as we would have anticipated the level of job stability to be higher after the training investment. However, most of these young individuals — 77%— continue working at the same firm with a temporary contract after finishing the IC. This indicates that firms are using IC as a way to reduce hiring costs, rather than as a training and probation time before providing the worker with a stable job. A third interesting result is that 27% of those individuals who move to another firm sign indefinite contracts, which is four percentage points higher than the rate of those who remained.

Focusing on those individuals who did not finish the IC (around 40%), Figure 7 shows that 16% of them stay at the firm, and out of this group, 18% sign an indefinite contract. This could be understood as a mechanism to retain particular workers. On the other hand, among those individuals who move to another firm, the incidence of indefinite contracts is highest, reaching 29% of all those who left. Therefore, it appears that for some of the young workers, the IC act as a safety net until they find a more stable job somewhere else.

Treatment Group B - IC users with previous job experience

We now describe the labour trajectories of individuals who sign an IC but have worked before. Figure 7 shows that the proportion young workers who has finished the IC is very similar to those who have entered the labour market by signing an IC. Among those who finish the contract, the patterns for the next labour market episode are also very similar: approximately one-third remain at the firm, one-third move to another firm, and onethird become unemployed. However, among the leavers (those who left before the IC was finished), 66% move to another firm (compared to 73% of those who has not worked before), 15% remain at the same firm under another contract (the same as in the previous group), and 11% become unemployed — a proportion that is twice as high as that found for Treatment Group A.

On the other hand, we again observe that those individuals who finish the IC before the expected date and move to another company are more likely to sign an indefinite contract in the following job episode: 29% of the young workers who follow this path made the transition to an indefinite contract, a share similar to that in in the previous group. However, the prospects of these young people if they finished their IC are notably worse than those of Treatment Group A, as they have a lower transition rate to an indefinite contract. It appears that having been employed under an IC do not have a positive impact on the subsequent employability of this group, as their levels of unemployment after the contract ended is significant. The young workers in both groups are most likely to have a temporary contract.

In conclusion, the patterns in the labour trajectories of the two treatment groups under analysis are similar. Having previous experience do not seem to affect whether a young worker has employment stability after being employed under the IC. On the other hand, having finished the IC before the expected date and moving to another firm is associated with an increase in the likelihood of signing an indefinite contract. Even so, the vast majority of the young people sign a temporary contract after the IC, regardless of whether they finish the IC or not, or of whether they change firms or not.

Up to now, we have documented that the internship contract do not enhance job stability, as most young workers who are employed under the IC later sign a temporary contract. However, it might be the case that the job trajectories of those who are employed under the IC are even more unstable. To check whether this is the case, we compare the employment trajectories of the young workers in Treatment Group A with those of the young workers who enter the labour market by signing a standard temporary contract instead of an IC — Control Group C.

Control Group (C) - Temporary contract users without previous job experience

Figure 8 presents the trajectories of Control Group C. Recall that this group includes all individuals whose first job experience is through a temporary contract that lasted more than three months but is not an IC and who would not sign an IC in the future. For this group, information on the end date is not presented, given that, as explained, the dataset does not include the expected end date, but rather the real end date. The first notable result of this analysis is that the proportion of individuals who stay at the firm after being employed under a standard temporary contract of at least three months is 30%, a share that is twice as large as the proportion for those who sign an IC. In addition, when we look at the type of contract, we see that the proportion of individuals who signed an indefinite contract is very similar to that for Treatment Group A.

Figure 8. Labour trajectories after the first temporary contract longer than 3 months

In sum, the descriptive analysis shows that those individuals who enter the labour market through a temporary contract longer than three months have higher overall stability levels than those individuals who enter through an IC. When we compare those individuals who sign an IC, we see that whether or not they have worked before is not associated with any significant variation in their subsequent labour market episodes. However, the evidence presented so far comes from descriptive analysis.

Heterogeneity between individuals in the different groups might explain the observed differences. In order to control for these potential differences, we develop econometric estimations to obtain conditional probabilities of different outcome results for the three groups under consideration. But before doing so, we present some descriptive regarding this potential differences among groups.

SOCIODEMOGRAPHIC DESCRIPTIVES

In order to provide an initial description of the potential differences among the three groups in some of the main variables of analysis, the following table is presented.

Tabla 1. Descriptives on sociodemographic characteristics

Table 1 indicates that there is a higher incidence among females in Group B, whereas there are no gender differences in Groups A and C at the start of the century (2000). However, by 2017, the incidence of females signing ICs in their first work experience (Group A) has increased. Looking at age, we can see that the 20-24 age group is the most frequent. However, it is remarkable that the incidence of very young workers in Group C (who sign standard temporary contracts and not ICs) decreases substantially over time. Focusing on the contract duration, we can see that by 2017, the most common contracts are those longer than 12 months. Finally, we observe that even though we restrict our sample to educated young individuals, their initial contribution groups are rather low, as most are hired under the low one.

4.3. Empirical strategy

Even though internship contracts have been available for 22 years, the labour market trajectories of those who have been employed under the IC have never been formally evaluated. Because this type of contract is associated, at least in theory, with the employer making investments in training, we anticipate that employment under an IC should lead to greater job stability than being employed under a standard temporary contract for which there is no such requirement. This report seeks to provide evidence on this question, thus filling the gap that exists in the evaluation of internship contracts.

More precisely, the aim of our empirical strategy is to test the impact of employment under the internship contract on different labour market outcomes. The first output we estimate is the probability of staying at the firm after the IC has ended in order to determine whether there is an association between signing an IC and the probability of staying at the firm, which is one of the purposes of the IC. To do so, we compare the probability of remaining at the same firm for young workers who are hired under an IC and have no previous work experience (Treatment Group A) and for individuals whose first employment episode is a temporary contract longer than three months (Control Group C). The dependent variable in this estimation takes three different values: specifically, i) remain at the firm, ii) move to another firm and iii) transition to unemployment (here, we include those individuals who are not seen later on). Methodologically, we develop a multinomial logit regression with these three different potential outcome variables.

The second step is to estimate a similar model, but in this case, we compare those individuals who enter the labour market through an IC (Treatment Group A) with those who sign an IC after having worked before (Treatment Group B). This allows us to test the extent to which the impact of the IC with respect to staying at the same firm differed depending on the workers’ previous labour market trajectories.

The next question we attempt to answer is what impact having been employed under the IC has on the next labour market episode of the young individuals. To do so, we compare the type of contract the individuals sign after the IC ended (for Treatment Group A) with the type of contract the individuals sign after being employed under a regular temporary contract longer than three months (Control Group C). For this purpose, we estimate separately the probability of signing an indefinite contract (as compared to a temporary one) depending on whether the individuals remain at the firm or move to another firm. In this second analysis, we exclude members of the treated and control groups who are unemployed after their first contract ended.

All estimations include a set of control variables related to the individuals’ demographic and labour market characteristics, including gender, age group when the contract is signed, duration of the IC or temporary contract, whether the contract ended at or before the expected date (when individuals who signed an IC are compared), size of the firm, contribution group (low, medium, or high) and sector of activity. Regional, monthly, and yearly fixed effects are also included in order to control for the economic cycle and regional variations.

In order to address the gender perspective, we repeat all regressions interacting the key variable of interest with a female dummy, so that we can test the extent to which there is a differential impact on males and females of the outcome variable for the treated and the control groups.

Finally, as a robustness check, we repeat the estimations in which Treatment Group A are compared to Control Group D; i.e., young workers who started their career with a temporary contract longer than six months instead of three months. This change makes sense given that the minimum duration of an IC is six months. By using standard temporary contracts with a minimum duration of six months, we make the treatment and the control group even more similar. However, as we explained, we have not done so from the very beginning of the analysis, as there are relatively few workers whose first standard temporary contract lasted more than six months. This robustness check allows us to see the extent to which the previous results depend on the contract duration of the control group.

4.4. Results

This section presents the results of each of the estimations, their interpretations, the gender perspective and results from the robustness checks. For reasons of conciseness, only the coefficient of the variable of interest is provided in each table, even though all estimations control for all the observable variables mentioned before. Nevertheless, the complete estimation results can be found in the annex. Marginal effects of the coefficients are presented to facilitate the interpretation of the results.

PROBABILITY OF STAYING AT THE FIRM AFTER AN IC

The results regarding the impact of employment under the internship contract on the probability of staying at the firm after the IC ended are presented in Table 1. As we explained in the previous section, we first compare the young workers whose first job experience is employment under an IC (Treatment Group A) with those who enter the labour market through a temporary contract other than an IC with a duration of at least three months, and, of course, meet all of the criteria for signing an IC (Control Group C). For both groups, we estimate through a multinomial logit the probability of moving to another firm, staying at the same firm, or transitioning to unemployment after employment under either an IC or a standard temporary contract (of at least three months’ duration).

Table 1 displays the results. The first notable finding is that, all other things being equal, the probability of staying at the same firm after employment under an IC is six percentage points lower than after employment under a regular temporary contract. In addition, the probability of moving to another firm is five percentage points higher after employment under the IC. Another interesting finding is that the probability of transitioning to unemployment is also higher after employment under the IC than after employment under a regular temporary contract.

Table 1. Estimation 1: Probability of staying at the firm Vs moving to other firm Vs moving to unemployment. Treatment Group A Vs Control Group C.

When we analyse the differential impact for women and men (Estimation 1.1), we find that the variable female x IC is statistically significant at the 5% level only for the unemployment transitions; meaning that females have a two-percentage-point lower likelihood than males of transitioning to unemployment after employment under the IC (compared to employment under a regular temporary contract). However, the probability of remaining at the same firm after employment under the IC is equal for men and women, and is lower than after employment under a temporary contract. It is also important to mention that the female coefficient is negative and statistically significant, which means that the women have lower probabilities of staying at the same firm than men (these results are based the previous estimation as well, which is provided in the annex).

Table 2. Estimation 1.1. Gender perspective: Probability of staying at the firm Vs moving to other firm Vs moving to unemployment. Treatment Group A Vs Control Group C.

As a robustness check, if we replace Control Group C with Control Group D (Estimation 1.2. in Table 3). We find that (i) the probability of staying at the same firm is roughly the same; (ii) the probability of moving to another firm is slightly lower; and (iii) the probability of transitioning to unemployment is not statistically significant (before it is one percentage point). The latter observation represents a change relative to our previous findings. However, given that the results of the impact of employment under the IC on staying and on moving remain the same, we conclude that our main results are robust to a change in the minimum duration required for the control group from three to six months.

Table 3. Estimation 1.2. Robustness check: Probability of staying at the firm Vs moving to other firm V moving to unemployment. Treatment Group A Vs Control Group D.

The next step is to compare Treatment Group A with Treatment Group B in order to measure whether employment under the IC affected the young workers differently depending on whether they have or have not worked before. Table 4 provides the results.

Table 4. Estimation 2: Probability of staying at the firm Vs moving to other firm Vs moving to unemployment. Treatment Group A Vs Treatment Group B.

The estimation results indicate that whether individuals have or have not worked before do not change the probability of staying at the same firm after employment under the IC. However, the probability of transitioning to unemployment increased slightly if the individual had worked before.

To sum up, these results indicate that, all other things being equal, the probability of staying at the same firm is higher after employment under a temporary contract (longer than three months) than after employment under the IC, regardless of whether the individuals have or have not worked before signing the IC.

PROBABILITY OF SIGNING AN INDEFINITE CONTRACT AFTER THE IC

Second, we want to test whether employment under an IC leads to higher job stability than employment under standard temporary contracts in the subsequent job episodes, either at the same firm or at another firm. A priori, given that an IC entails, at least in theory, that the employer invests in training, we would have expected that individuals who signed an IC have ongoing job stability at the same firm. This expectation is, however, not confirmed by the evidence. Regarding the impact of employment under the IC on job stability on other firms, we do not have a clear a priori expectation.

To test these assumptions, we first analyse whether individuals who stay at the same firm after the IC are more likely to remain under an indefinite contract than the workers whose first job experience is under a standard temporary contract longer than three months. This would be the expected result, given that the aim of the social security reductions associated with the IC is to promote indefinite contracts. Second, for those workers who left the firm after employment under the IC or the first standard temporary contract (longer than three months), i.e., the leavers, we estimate the probability of signing an indefinite contract versus a temporary one. For the two exercises, we focus on the comparison between Treatment Group A and Control Group C, given that neither of the two groups had previous job experience, which makes them more comparable. Table 5 presents both estimation results.

⁷ A potencial explanation for this finding is that individuals with previous job experience might have generated access to unemployment benefits

Table 5. Estimations 3 and 4: Probability of signing an indefinite contract Vs temporary contract. Treatment Group A Vs Control Group C. Estimation 3: stay at the firm. Estimation 4: move to another firm

The results shown in Table 5 indicate that, for stayers, the probability of signing an indefinite contract is 20 percentage points lower if they had signed an IC than if their first contract has been a regular temporary contract (longer than three months). We can conclude that employment under the internship contract is associated with less job stability at the same firm than employment under other temporary contracts, even though the IC is theoretically associated with investments in training by the employer. One possible way to interpret this finding is that, in reality, many employers that take advantage of the IC do not invest in training, but instead use such contracts exclusively to reduce their hiring costs through wage and tax reductions. If that is the case, we might conclude that the IC has not lived up to the expectations for which it was created.

Let us now comment on the results for the leavers; i.e., those young workers who left the firm after employment under the IC. Previously, we have seen that the probability of moving to another firm is higher for those with an IC than for those with a temporary contract. The interesting point here is that, conditional on moving to another firm, the probability of signing an indefinite contract at another firm is substantially higher for those who sign an IC . If we interpret this outcome as a supply-driven result (as a worker initiative), it seems to suggest that for some young workers, signing an IC provides them with temporary employment while they look for a more stable job not at the same firm, but at another one. Additionally, it might be the case that external firms interpret employment under the IC as a signal of worker productivity, which encourages them to offer these young workers stable contracts.

From a gender perspective, the results of Table 6 are mixed. First, we find that for stayers, the coefficient of female denotes the difference in the probability of signing an indefinite contract at the same firm for the control group; i.e., those whose first job experience is employment under a temporary contract longer than three months. Females are more likely to end up in an indefinite contract than their male counterparts.

Second, and again for stayers, when we decompose the impact of employment under the IC by gender, we find no significant differences for males or for females between employment under an IC or a temporary contract with regard to the probability of subsequently signing an indefinite contract. As before, the signs are negative for both males and females, but now none of the coefficients are significantly different from zero.

Table 6. Estimations 3.1 and 4.1: Probability of signing an indefinite contract Vs temporary contract. Treatment Group A Vs Control Group C. Estimation 3.1: stay at the firm. Estimation 4.1: move to another firm.

With respect to leavers, we find that as before, the negative coefficient for females reveals that for the control group – i.e., those whose first labour market experience is employment under a temporary contract longer than three months – males have higher probability than females of subsequently signing an indefinite contract if they move to another firm. Second, the positive coefficient of the IC indicates that male signers of an IC have a higher probability of subsequently signing an indefinite contract than similar male workers whose first labour market experience is employment under a standard temporary contract. However, for females, this effect is not observed. This tells us that the result we find when looking at the impact of employment under all ICs together is driven mainly by males, not by females.

The next step is devoted to the robustness analysis, in which we change the composition of the control group. As we mentioned before, we require that the job duration of the first temporary contract is not three months, but six months. The robustness check of the first exercise is presented in Table 7, which must be compared with the results shown in Table 5. Given that similar conclusions are reached, we can conclude that the result is robust, since for those individuals whose first contract is a regular temporary contract (longer than six months), the impact of signing an indefinite contract if they stay is almost 10 points higher – up to a probability of 29 percentage points lower – than if they have signed an IC. Regarding the impact of signing an indefinite contract for leavers, the results are exactly the same as those shown in Table 5.

Table 7. Estimations 3.2 and 4.2: Probability of signing an indefinite contract Vs temporary contract. Treatment Group A Vs Control Group D. Estimation 3.2: stay at the firm. Estimation 4.2: move to another firm.

Finally, in order to measure whether employment under the IC affects the young workers differently depending on whether they have or have not worked before, we compare those individuals who sign an IC with and without previous work experience, and develop the same estimations. We obtain the following results (Table 8). It is clear that the results are not statistically significant, meaning that we see no differences when we compare the treated workers who had and had not worked before.

Table 8. Estimations 5 and 6: Probability of signing an indefinite contract Vs temporary contract. Treatment Group A Vs Treatment Group B. Estimation 5: stay at the firm. Estimation 6: move to another firm.

To examine the potential changes in the impact of the IC in expansion and recession periods, the analysis has, when possible, been developed separately for different economic periods. However, for some estimations, the number of observations is not high enough, and the results might be questionable. For this reason, the results are not included in this report. However, in general, the results remain similar when we restrict the evaluations to a limited period of time.

To sum up, the results found in this study allow us to conclude that the IC, which was created to help young workers with higher education transition from a first temporary work experience to a more stable position, do not meet this aim if workers remain at the same firm. This is to some extent a failure of the design of the initiative: the firms that took advantage of the IC enjoy cost reductions through the reduction in social security contributions, but it appears that they do not use these reductions to provide young qualified workers with a stable future at the firm. On the other hand, and as an unexpected indirect effect, we find that young individuals who move to another firm after employment under the IC have higher chances of achieving stability than those whose first job experience is under a standard temporary contract (longer than three months).

5. Conclusions and recommendations

Spain is among the European countries with the highest unemployment and temporary employment rates. In response to the labour market precariousness that young individuals in Spain have faced for the last 25 years, the Spanish government developed the internship contract (IC), which is directed at highly educated young workers. Internship contracts were first implemented in 1998, and, according to the legislation, the aim of the IC is to provide young people who have already finished their university or vocational training education with professional practice at the start of their professional lives. This professional practice must be related to the level and field of studies, and must be accompanied by investments in training by the firms that employ the young workers. The spirit of the law that implemented the IC is that employers are expected to invest in human capital by offering practical training, and will, in return, benefit by paying lower wages and taxes.

The aim of this report is to evaluate the subsequent job trajectories of the young workers who are employed under an internship contract. In particular, we focused our evaluation on two specific outcomes: first, whether employment under an IC leads to an increase in the probability of staying at the firm after the IC has ended; and, second, whether employment under an IC increases the probability of signing a subsequent indefinite contract, either at the same firm (for those who stay) or at another firm (for those who left after the IC). For the two outcomes, the IC signers (treated) are compared with workers who meet the IC eligibility criteria, but whose first contract is a regular temporary contract of at least three months. In addition, control variables such as gender, age, sector of activity, size of the firm, and contribution group are included in the estimation to ensure that we are comparing very similar individuals.

The most important result of this paper is that employment under the internship contract reduces the probability of a young worker being retained by the firm after the

IC, when compared to that of a young worker whose first work experience is through a regular temporary contract (of at least three months). If firms had invested in training for these workers, we would have expected to observe that they had a higher probability of being retained by the firm. Hence, it may well be the case that in practice, firms do not invest in training these workers, and that they use the IC exclusively as a way of reducing hiring costs through wage and tax reductions.

The second main result of the study is that for those individuals who stay at the firm after the IC, their probability of signing an indefinite contract is much lower than that for those individuals who sign another temporary contract of at least three months. If firms had invested in training workers employed under the IC, we would not expect this result. Rather, it appears that the firms use these contracts to reduce hiring costs, and do not expect to offer these workers stable employment.

The last result obtained in the analysis is an indirect effect of the IC: those individuals who move to another firm after employment under the IC have a higher probability of securing an indefinite contract. Therefore, it appears that the IC is functioning as a temporary safety net for young individuals until they find a more stable job at another firm, and/or as a signal of high productivity for external firms.

Finally, in addition to the reduction in hiring costs as the rationale for firms to use the IC, we speculate about a second potential driving factor: under Spanish labour legislation, the maximum period a worker could be hired under a temporary contract at a firm is two years. This maximum period does not apply to the IC. This means that after two years of employment under an IC at a firm, young workers could be hired two more years under a regular temporary contract. This provides greater flexibility to firms that prefer to hire employees under temporary rather than indefinite contracts in order to avoid the much higher dismissal costs associated with the latter. Some Spanish firms clearly prefer to use temporary contracts (and they have used them in an abusive manner), which leads to low incentives for employers to invest in training their workers, and high incentives for workers to look for better job opportunities at other firms. It seems that the small incentive (€500 per year for men or €700 for woman) the firm would receive if they hired the workers under indefinite contracts is not enough to compensate for the loss of flexibility they face if they do so.

In light of these results, the main conclusion reached in this study is that the IC has not been an instrument to offer more stable labour market trajectories to highly educated young workers in Spain. The results seem to suggest that although the IC is, in theory, associated with investments in training by the employer, in practice, it is unlikely that the employers make such investments. Rather, many firms have used ICs exclusively for wage and tax reductions.

Based on this empirical evidence, our recommendation is to abolish the wage and tax reductions (the latter are already abolished) and to create a different internship contract that serves as an accredited training for highly qualified workers with no previous work experience. Given that receiving such training makes the most sense at the very start of a worker’s career, the contract should be restricted to young individuals who are entering the labour market. As it is currently structured, the IC has been shown to act as dead weight, as firms use it to reduce their costs of hiring young workers who probably would have been hired anyway, and do not comply with the requirement that they invest in training, which is the essence of the IC. Finally, the period of employment under an IC at a firm should be counted as time spent employed under a temporary contract, and should, therefore, be included in the maximum two-year period a worker can be employed under a temporary contract.

Bibliography

References

1. Durán, A. (2007). La Muestra Continua de Vidas Laborales de la Seguridad Social. Revista del Ministerio de Trabajo y Asuntos Sociales, n.o 1, págs. 231-240.

References

1. European Commission (2012), Apprenticeship supply in the Member States of the European Union, Directorate- General for Employment, Social Affairs and Inclusion.

References

1. García-Pérez JI (2008) La Muestra Continua de Vidas laborales (MCVL): una guía de uso para el análisis de transiciones laborales. Revista de Economía Aplicada XVI(E-I):5–28.

References

1. García- Perez, JI & Y. Rebollo Sanz (2009), Do wage subsidies affect the subsequent employment stability of permanent workers? The case of Spain” Moneda y Crédito, nº 228, pp. 65-103

References

1. García- Perez, JI & Y. Rebollo Sanz (2009), The use of permanent contracts across Spanish regions: Do regional wage subsidies work?, Investigaciones Económicas 33 (1), pp: 97-130

References

1. Jansen, M., & Troncoso-Ponce, D. (2018). El impacto de los contratos para la formación y el aprendizaje en la inserción laboral de las personas jóvenes (No. eee 2018-27). FEDEA.

References

1. Kluever, Jochen (2007), Active Labour Market Policies in Europe: Performance and Perspectives, Springer.

References

1. Valdés, Dal-Re, Fernando & Jesus Lahera (2010), La flexiseguridad laboral en España, Documento de Trabajo 157/2010, Fundación Alternativas.

Annex – Complete estimations results

In this annex, we present the results of the estimations explained throughout the report.

Clarifying note: in all the following regressions we use additionally as control fixed effects of Autonomous Community, month and year. However, for practical reasons, we do not include the estimated results in the tables. Their exclusion does not alter the principal conclusions that we extract from the estimations.

Table 1. Estimation 1 — Probability of staying at the firm Vs moving to other firm Vs moving to unemployment.Treatment Group A Vs Control Group C

Table 2. Estimation 1.1. Gender perspective — Probability of staying at the firm Vs moving to other firm V moving to unemployment. Treatment Group A Vs Control Group C.

Table 3. Estimation 1.2. Robustness check — Probability of staying at the firm Vs moving to other firm V moving to unemployment. Treatment Group A Vs Control Group D.

Table 4. Estimation 2 — Probability of staying at the firm Vs moving to other firm Vs moving to unemployment. Group A Vs Treatment Group B.

Table 5. Estimations 3 and 4 — Probability of signing an indefinite contract Vs temporary contract. Treatment Group A Vs Control Group C. Estimation 3: stay at the firm. Estimation 4: move to another firm.

Table 6. Estimations 3.1 and 4.1 — Probability of signing an indefinite contract Vs temporary contract. Treatment Group A Vs Control Group C. Estimation 3.1: stay at the firm. Estimation 4.1: move to another firm.

Table 7. Estimations 3.2 and 4.2 — Probability of signing an indefinite contract Vs temporary contract. Treatment Group A Vs Control Group D. Estimation 3.2: stay at the firm. Estimation 4.2: move to another firm.

Table 8. Estimations 5 and 6 — Probability of signing an indefinite contract Vs temporary contract. Treatment Group A Vs Treatment Group B. Estimation 5: stay at the firm. Estimation 6: move to another firm.

Fundación de Estudios de Economía Aplicada

Jorge Juan, 46

28001 Madrid

España

+34 91 435 90 20

www.fedea.net