Mapping the (mis)match of university degrees in the graduate labor market

Manuel Salas-Velasco

doi:10.1186/s12651-021-00297-x

Salas-Velasco, Manuel

2021-05-07 00:00:00

This paper contributes to the scarce literature on the topic of horizontal education‑job mismatch in the labor mar ‑ ket for graduates of universities. Field‑ of‑study mismatch or horizontal mismatch occurs when university graduates, trained in a particular field, work in another field at their formal qualification level. The data used in the analysis come from the first nationally representative survey of labor insertion of recent university graduates in Spain. By estimating a multinomial logistic regression, we are able to identify the match status 4 years after graduation based on self‑ assessments. We find a higher likelihood of horizontal mismatch among graduates of Chemistry, Mathematics, Phys‑ ics, Pharmacy, and Languages and Literature. Only graduates in Medicine increase the probability of being adequately matched in their jobs. It may be hypothesized that horizontal mismatch is more likely among those graduates in degree fields that provide more general skills and less likely among those from degree fields providing more occu‑ pation‑specific skills. Other degrees such as Business Studies, and Management and Economics Studies increase the probability of being vertically mismatched (over‑ educated). Vertical mismatch preserves at least some of the specific human capital gained through formal educational qualifications. However, some workers with degrees in Labor Rela‑ tions and Social Work are in non‑ graduate positions and study areas unrelated to their studies. The paper also shows that graduates in the fields of health sciences and engineering/architecture increase the probability of achieving an education‑job match after external job mobility. Keywords: Education‑job mismatch, Higher education, Horizontal mismatch, Job turnover, Multinomial logistic regression, Spanish university degrees JEL Classification: J24, J63, I21, C50 1 Introduction jobs. The (mis)match between the level of formal educa - In most economies, there is a connection between the tion and the level required for the job has been, indeed, educational attainment of the labor force and the jobs the focus of substantial research in the labor and educa- performed by the workers. In general, job titles are tion economics literature since the appearance of Free- defined in terms of educational requirements that coin - man’s (1976) book The overeducated American. See, for cide with the levels of formal education. Of particular surveys of the literature, Leuven and Oosterbeek (2011), interest is to analyze whether the tasks assigned to differ - McGuinnes (2006), and Sloane (2003); for a meta-analy- ent positions can be performed effectively with the quali - sis, Groot and Maassen van den Brink (2000). fications provided by the education system or, on the In this article, we focus on the labor market for univer- contrary, there is no connection between the contents sity graduates. The paper contributes to the understand - of the educational curriculum and the contents of the ing of the mismatch between professional (academic) degrees and labor market positions. Most theoretical *Correspondence: msalas@ugr.es In any case, it is not an easy duty to define which education is appropriate Department of Applied Economics, University of Granada, Campus for each job since the educational requirements of the positions differ among Cartuja, 18071 Granada, Spain companies and change over time. © The Author(s) 2021. This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http:// creat iveco mmons. org/ licen ses/ by/4. 0/. 14 Page 2 of 23 M. Salas‑Velasco and empirical studies of education-job mismatch have education-job (mis)match is important for the educa- focused predominantly on graduate over-education (e.g., tional policy given that higher education is highly subsi- Dolton and Vignoles 2000). Over-education (or vertical dized in Spain. The article is also novel in the sense that mismatch) appears when graduates work in non-gradu- it incorporates methodological improvements that we ate jobs. However, this article focuses on another type of comment below. Two well-cited papers by Robst (2007) education-job mismatch that has received less attention and Nordin et al. (2010), published in the same journal, in the literature: the unrelatedness of a worker’s field of already addressed the mismatch between the individual’s study to his or her occupation at their formal qualifica - field of education and his/her occupation (horizontal tion level, also referred to as horizontal mismatch. Rela- mismatch). Robst’s (2007) match/mismatch measure was tive to vertical mismatch, there are much fewer published based on subjective answers to the question of whether studies of horizontal mismatch—see Somers et al. (2019) the job the college graduate held was closely related, not for a recent systematic literature review. In the lat- related, or somewhat related to his/her highest degree ter paper, it is evidenced that, unlike vertical mismatch, field. In Nordin et al. (2010), the authors crossed 34 occu - there are still no theoretical models that explain the phe- pations with 29 different fields of education in a table nomenon. Nonetheless, the empirical evidence suggests and made the same classification. Nonetheless, both that the likelihood of horizontal mismatch is among papers present drawbacks. In Robst (2007), the author other things determined by the extent to which employ- used an ordered logit model which indicated whether ees possess general skills as opposed to occupation-spe- a major had a higher or lower likelihood of being hori- cific skills (Somers et al. 2019). In the labor market for zontally mismatched, but the author did not distinguish university graduates, the issue of horizontal mismatch is whether the undergraduates were occupying college- considerably less studied than vertical mismatch (or over- level occupations or they were filling typical high school education) mainly due to the lack of relevant data on graduate positions. The implications for educational fields of studies of university graduates. Horizontal mis - policy and the labor market are different. In the case of match (or field-of-study mismatch) occurs when gradu - Nordin et al. (2010), the authors only presented a table ates, trained in a particular field, work in another field with the fields of education and the shares of matched, at their formal qualification level. For example, a person weakly matched, and mismatched individuals (there is earning a degree in Mathematics working as a computer- no econometric model). In their percentages, they did aided design technician. Robst (2007) was one of the first not distinguish either whether the graduates were in papers devoted to the horizontal mismatch. In this study, positions typical of graduates or lower-level positions. some of the majors with the highest prevalence rates of Some results were striking. For example, 80% of men mismatch between work and degree fields included Eng - and 75% of women with a degree in Biology were mis- lish and foreign languages, social sciences, and liberal matched. In this last classification, among other occupa - arts. “Typically, these majors provide more general skills tions, the authors included teachers of upper secondary than occupation specific skills” (Robst 2007, p. 402). On education. However, according to the proposal we make the contrary, computer science, health professions, and in this paper, they would be well-matched because they engineering had low prevalence rates. “Most of these are occupying university positions in a related field, i.e., majors focus on skills that apply to relatively specific teaching Biology. Our paper contributes thus to improv- occupations” (Robst 2007, p. 402). The specific human ing the deficiencies of those publications by focusing on capital cannot be easily transferred to other sectors, and the Spanish labor market for recent university gradu- graduates in these fields are less likely to search for a job ates. In particular, the article aimed to determine which in other sectors. They are more likely to work in a job degree fields (narrow fields of education) were associ - that is directly related to their field of study in order to ated with being horizontally mismatched in the labor use specific human capital, which was accumulated dur - market for higher education graduates in Spain: when ing university studies. Graduates of these fields are there - graduates are employed in a graduate job that is not fore less likely to be horizontally mismatched. related to their field of study. By estimating the likeli - Because the number of empirical studies on horizon- hood of being horizontally mismatched (field-of-study tal mismatch among university graduates is limited, this mismatch), we also simultaneously estimate the probabil- paper contributes thus to the scarce existing literature on ity of being vertically mismatched (over-education), and the topic by providing the taxonomy of educational mis- full job mismatched (i.e., field-of-study mismatch and match in the labor market for university graduates and investigating its incidence among Spanish higher educa- tion graduates based on self-assessments. In addition, For example, in 2010, only 62 percent of U.S. college graduates had a job that the map of degrees done in this article according to the required a college degree (Abel and Deitz 2015). Mapping the (mis)match of university degrees in the graduate labor market Page 3 of 23 14 over-education). The taxonomy that we propose allows after external job turnover, on the other hand. Section 5 us to better identify situations of educational mismatch shows the results of the econometric analysis. Section 6 in the graduate labor market. Besides, the multinomial provides a discussion and some policy implications. Sec- logit model of the probability of education-employment tion 7 concludes the paper. matching that we suggest allowed us to draw a map of university degrees according to the type of (mis)match. 2 Empirical measurement This is also a novelty. Additionally, our article aimed to Job mismatch can be defined as the discrepancy between study external labor mobility that takes place in the early the qualifications that individuals possess and those that stages of graduates’ working lives. A good match between are wanted by the labor market. But when we talk about graduates’ degrees and their jobs will likely happen after qualifications, we can refer either to the formal qualifi - job turnover. cation (formal education) or to skills or competencies For the analysis carried out in this paper, we used indi- (European Centre for the Development of Vocational vidual-level data from the first survey of labor insertion Training, 2014). In the first case, formal qualification is of university graduates in Spain. The Encuesta de Inser - “the formal outcome (certificate, diploma or title) of an ción Laboral de titulados Universitarios (EILU 2014) is a assessment process which is obtained when a competent nationally representative random sample of Spanish uni- body determines that an individual has achieved learning versities and university graduates. A total of 30,379 grad- outcomes to given standards and/or possesses the neces- uates from the class of 2010 were surveyed 4 years after sary competence to do a job in a specific area of work” graduation. The survey asked workers directly whether (European Centre for the Development of Vocational their particular qualification was appropriate for the Training, 2014, p. 202). In the second case, the term work that they did. Many Spanish university graduates qualification refers to “knowledge, aptitudes, and skills were employed in jobs that neither required a degree nor required to perform specific tasks attached to a particular made use of expert knowledge learned at the university. work position” (European Centre for the Development of The degree of fit between the qualifications obtained by Vocational Training, 2014, p. 202). Skill mismatch arises graduates and their job characteristics can be considered when workers have higher or lower skills proficiency than one important performance indicator in higher educa- those required by their job. If their skills proficiency is tion. This latter is an expensive investment—it is highly higher than that required by their job, workers are clas- subsidized in Spain—and the highest return for society is sified as over-skilled; if the opposite is true, they are obtained when individuals are well-matched to employ- classified as under-skilled (Pellizzari and Fichen 2013). ers such that the knowledge and skills that were acquired Likewise, educational mismatch arises when work- through higher education are optimally utilized on the ers’ levels of formal education are higher or lower than labor market. Therefore, research on the study of the the required levels of education of their employment. labor market for graduates and their educational (mis) This mismatch is also known as a vertical mismatch. match is justified. In the discussion section of this article, Over-education (or over-qualification) and under-edu - the reader will find more arguments. cation (or under-qualification) are the two types of ver - The rest of the paper is organized as follows. Section 2 tical mismatch. Over-education exists when a worker outlines the empirical framework behind the measure- is employed in a job that requires a lower level of educa- ment of vertical and horizontal education-job mismatch tion than that possessed by the worker. Under-education in the graduate labor market. In Sect. 3, we describe the exists when a worker has a lower level of education than data set drawn from the National Statistics Institute of required for the job (e.g., Chevalier 2003; Duncan and Spain. We also identify four types of education-job mis- Hoffman 1981; Hartog 2000; Leuven and Oosterbeek match according to the most appropriate level of formal 2011; Mavromaras et al. 2013; Park 2018; Sicherman education and study area to perform a job, and we pro- 1991). In this regard, it should be noted that educational vide summary statistics on the incidence of mismatch mismatch can imply skill mismatch, but skill mismatch among Spanish higher education graduates. In Sect. 4, does not imply necessarily educational mismatch (Allen we introduce the econometric models of the probability and Van der Velden 2001). For example, when working of being (mis)matched in the first and current job, on the one hand, and the probability of being well-matched Although education is often used as a proxy for skills, the two terms have different meanings (International Labour Organization 2014). The sample in the EILU2014 was restricted to ISCED-97 5A level (Bachelors and Masters or equivalent) graduates. ISCED stands for International Stand- In practice, the terms over-qualification and over-education are used ard Classification of Education. interchangeably. The same for under-education and under-qualification. 14 Page 4 of 23 M. Salas‑Velasco in a position below one’s level of study, skills learned in graduate in a job requiring sub-degree level qualifications formal education may not be fully used; over-education (or no qualifications at all) was defined as over-educated. would be synonymous with being over-skilled. Let’s Results showed that 38 percent of all graduates surveyed think of a medical graduate working as a dental assistant. were over-educated in their first job. This proportion fell But, if this medical doctor works in a hospital as a sur- to 30 percent by the end of the survey period, 6 years geon but s/he says that would perform the job better if later (Dolton and Vignoles 2000). Over-educated gradu- s/he possessed additional skills, s/he would have a skills ates earned significantly less than peers in graduate jobs deficit, but s/he would not be under-educated. (Dolton and Vignoles 2000). Nonetheless, vertical mismatch of education (mis- More recently, in the 2012 and 2015 Survey of Adult match of the level of education and job) is not the only Skills (PIAAC), employed workers aged 25–64 reported form of educational mismatch. In this article, we sug- their level of educational attainment (formal qualifica - gested two other educational mismatches. On the one tion) and the level needed for the job. In the first case, the hand, the horizontal educational mismatch, when the survey question was: Which of the qualifications (ISCED- own level of education matches the requirements of 97) is the highest you have obtained (education that has the job but the type of education is not appropriate for been completed)? To identify vertical mismatches, the the job. For example, an economics major working as answers given to this question are compared with the an engineer might be considered to be working in a job responses to the question: Talking about your current unrelated to the degree field (Robst 2007; Tao and Hung job. If applying today, what would be the usual qualifi - 2014). On the other hand, vertical and horizontal educa- cations (ISCED-97), if any, that someone would need to tional mismatch, when the highest level of education held GET this type of job? Among workers with a university by a worker does not match the required level of educa- qualification (ISCED 5A or 6), 75 percent (OECD aver - tion for his or her job, and also the type/field of education age) reported being in a well-matched situation. How- is inappropriate for the job. However, the study of skill ever, over 34 percent of workers in England (UK), Korea, mismatches is beyond the scope of this paper and our Estonia, and Japan reported being over-qualified for their database, unlike surveys such as REFLEX, does not con- job (which means having qualification of ISCED 5A or tain detailed information on skills acquired and required 6 while working in a job needing ISCED 5B or below). by jobs. In the case of Spain, 24 percent of university graduates reported being in the latter situation (Organisation for 2.1 M easuring vertical education‑job mismatch Economic Co-operation and Development 2018). Over-education can be assessed subjectively by asking An alternative approach to analyzing the mismatch the respondent to give information on the minimum between education and jobs consists of determining the educational requirements of the job and then compar- educational requirements of the occupations from some ing this with the individual’s acquired education or by objective measurement. In particular, over-education simply asking the respondent whether or not they are can be assessed based on information either about the over-educated (McGuinness, 2006). Dolton and Vignoles average or modal education level within the occupation (2000) used data from the National Survey of 1980 of the worker (realized matches/statistical approach) or Graduates and Diplomates to measure the incidence of about educational requirements coming from an a priori over-education in the UK graduate labor market. They assumed correspondence between education and occu- concluded that a significant proportion of British gradu - pations such as ISCO or DOT classifications (job analy - ates were over-educated in the 1980s. The question used sis/normative approach) (Kupets 2016). For example, to measure over-education was: What was the minimum Rumberger (1987) obtained an objective measure of the formal qualification required for (entering) this job? A degree of educational mismatch once he converted the educational requirements of each occupational category of the DOT into equivalent years of schooling and com- pared the result with the years of schooling that work- Over-educated or over-qualified: an individual has completed more years of ers actually had in those occupations. Regarding the formal education than the current job requires. Over-skilled: an individual is unable to fully use acquired skills and abilities in the current job. See Quintini (2011). ISCO stands for International Standard Classification of Occupations (Inter - national Labour Office), and DOT stands for Dictionary of Occupational When the measurement is limited to university graduates, the group with Titles (U.S. Department of Labor). the highest level of education, under-education is not possible and vertical mismatch has the same meaning as over-education. However, in our analy- The author was discussing the United States. The DOT was last updated sis carried out in this paper, vertical mismatch (over-education) is a more in 1991, and it is rarely used. Today, occupations are classified using the restrictive concept in the sense that it includes university graduates whose Standard Occupational Classification (SOC) system—a United States gov - work does not require a university degree but is related to their field of ernment system of classifying occupations—and data are provided through study. the Occupational Information Network, known as O*NET. Mapping the (mis)match of university degrees in the graduate labor market Page 5 of 23 14 mode-based statistical approach, if an employee’s educa- (completely mismatched). College-educated workers in tional attainment is higher (lower) than the modal educa- jobs unrelated to their field of study earned less than their tion level of individuals working in the same occupation, well-matched peers (Robst 2007). However, a limitation he/she is classified as over-educated (under-educated) of Robst’s work is that the author did not exclude from (e.g., Kampelmann and Rycx 2012; Kiker et al. 1997). As the analysis undergraduates working in positions that to the mean-based statistical approach, over-educated only require a high school or less education. For example, workers are those whose educational attainments are PIAAC data revealed that 22 percent of U.S. workers with greater than one standard deviation above the mean a university qualification (ISCED 5A or 6) would be hold - within their specific occupation; workers whose edu - ing a position requiring less formal qualification (Organi - cational attainments are more than one standard devia- sation for Economic Co-operation and Development tion below the mean are defined as under-educated (e.g., 2018). Surely, the wage effects of mismatch by degree Groot 1993; Verdugo and Verdugo 1989). All of these field found by Robst (2007) would be different. studies were based on the total employed workforce. In Europe, using representative samples of European Focusing more recently on workers who had completed university graduates graduating in 2000 (REFLEX sur- tertiary education, Rossen et al. (2019) employed a vari- vey) and 2003 (HEGESCO survey), Verhaest et al. (2017) ant of the realized matches approach coding a person as determined the match status 5 years after graduation being over-educated if his/her highest educational attain- based on self-assessments. The vertical educational mis - ment level was higher than the benchmark education match was based on the survey question: What type of level of his/her occupation group at the two-digit ISCO education do you feel was most appropriate for this work? level. As a benchmark, they applied in their main analy- A graduate is considered to be over-educated if his/ ses the 80th percentile of the levels of education within her educational level exceeds the appropriate level. The each occupational group. They made use of the 2016 horizontal educational mismatch was based on the sur- wave of the European Labour Force Survey (EU-LFS) for vey question: What field of study do you feel was most 21 EU countries. Furthermore, the sample was restricted appropriate for this work? Possible answers were: (1) to respondents aged 20–34 years. Over-education as exclusively own field, (2) own or related field, (3) a com - a vertical inadequacy was about 28% in total. The high - pletely different field, or (4) no particular field. They con - est rates were measured for France, Austria, Italy, and sidered horizontal mismatch if they answered (3) or (4). Greece where more than 35% of workers were over-edu- By combining the two types of mismatches, they got four cated, whereas the lowest rates were observed for Esto- categories: pure match, mere vertical mismatch, mere nia, Belgium, and Latvia with rates below 20%. horizontal mismatch, and pure mismatch. On average, 74.2 percent of graduates were well-matched 5 years after 2.2 M easuring horizontal education‑job mismatch graduation. The average incidence of horizontal mis - Horizontal mismatch measures the extent to which match was just over 10 percent but close to 16 percent workers, typically graduates, are employed in an occu- in Poland and Estonia, and above 18 percent in the UK. pation that is unrelated to their principal field of study In Spain, the incidence of horizontal mismatch was 4.5 (McGuinness et al. 2018). In the subjective self-assess- percent. ment method, respondents are asked how closely their educational field is related to the work they do.2.3 Limitations In one of the first studies on horizontal mismatch, The different measures proposed in the literature to esti - Robst (2007) studied the relationship between college mate the required education for a job—based on worker majors and occupations in the United States. Using data self-assessment, realized matches, and job analysis—often from the 1993 National Survey of College Graduates, the give different results of the incidence of the over-educa - following question was used to examine the education- tion. Self-assessment methods may be biased because job match: To what extent was your work on your princi- they rely on the objectivity of respondents. But an objec- pal job related to your highest degree field? Was it closely tive approach is also surrounded by controversy. Since related, somewhat related, or not related? Fifty-five per - the objective measure reflects an average requirement cent of individuals reported that their work and field of associated with all jobs in a particular occupation, it may study were closely related, but 20 percent of the sample not reflect the requirement associated with the particular reported their field of study and work were not related job held by the respondent. Also, the statistical mode- based method suffers from the misclassification problem: Although the main advantage of this method resides in the fact that it requires little information, since it is enough to know the educational level of the workers, nevertheless the boundary of a standard deviation is quite arbi- The statistical method usually yields significantly lower estimates of over- trary. education (e.g., Leuven and Oosterbeek 2011). 14 Page 6 of 23 M. Salas‑Velasco over-educated workers may be classified wrongly as Table 1 Description of the sample by broad groups of university degrees (ISCED 5A programmes) well-matched if the number of higher educated work- ers in a given occupational group increased significantly Freq Percent and pushed the modal level of education up even in the Diplomatura 9,339 30.74 absence of changing job tasks/requirements. In the stand- Technical Engineering and Technical Architecture 3,700 12.18 ard deviation-based measure of over-education, the Licenciatura 46.26 boundary of a standard deviation is quite arbitrary. For Engineering and Architecture 14,053 7.74 a broad discussion of the advantages and disadvantages, 2,352 see for example Hartog (2000), Leuven and Oosterbeek Grado 880 2.90 (2011), and Verhaest and Omey (2006), among others. Other university degrees before Bologna 55 0.18 Even though the normative/statistical approach has its Total 30,379 100.00 limitations, it is more or less feasible to measure the ver- Source: author’s calculations from EILU2014 tical mismatch. But an objective approach would be too complex to measure the horizontal mismatch, that is, the discrepancy between the graduate’s field of study and Table 2 Description of the sample according to broad branches that most appropriate for the job. Despite the potential of knowledge disadvantage that employees’ perceptions of the hori- zontal (mis)match are by definition subject to self-report Freq Percent bias (Banerjee et al. 2019), a potential advantage of this Arts and Humanities 3,231 10.64 approach is that graduates’ field of study is directly com - Hard Sciences 2,955 9.73 pared with the content of their jobs. “The individual Social and Legal Sciences 13,458 44.3 assessments, while perhaps subjective, are expected to Engineering and Architecture 6,793 22.36 provide important information” (Robst 2007, p. 401). Health Sciences 3,942 12.98 This will be the approach taken in this paper. Total 30,379 100.00 Source: author’s calculations from EILU2014 3 Description of data and matching procedure Including grados in Building and Computer Engineering 3.1 EIL U2014 graduate survey In Spain, universities follow a career system, which means that students begin their studies with their major Institute of Statistics (INE). Using a combined method of already selected and take courses that are pre-assigned obtaining information—direct interviews (Web and tel- for their entire major, with only a few electives available ephone) and use of administrative data, approximately each year. In the educational curriculum prior to the 30,000 university graduates of the 2009/2010 academic Bologna reform of 2010, there were two basic types of year were interviewed. Specifically, 30,379 university university programs: short-cycle programs called diplo- graduates from Spanish universities were interviewed in maturas, which were more vocationally oriented and the Encuesta de Inserción Laboral de titulados Universi- lasted 3 years (e.g., Nursing); and long-cycle programs tarios (EILU2014): 86% had studied at a public univer- called licenciaturas, which lasted 4, 5, or 6 years (e.g., sity and 14% at a private university. By gender, 40.3% of Economics, Law, and Medicine, respectively). Also, other the graduates were men, and 59.7% were women. Table 1 degrees awarded were engineering degrees and Archi- shows the description of the sample according to wide tecture (5 years on average) and technical engineering groups of university degrees and Table 2 displays the degrees and Technical Architecture (3 years on aver- description of the sample according to broad branches age). A nationally representative sample of university of knowledge. graduates of these degrees was surveyed between Sep- tember 2014 and February 2015 by the Spanish National 3.2 The taxonomy of educational mismatch in the labor market for Spanish higher education graduates In practice, researchers use one method or another depending on the avail- Let us focus on the study of educational mismatches in able data. 13 the employment of the university graduates surveyed. Nordin et al. (2010) built 29 different fields of education and created 34 different occupations. They "subjectively" constructed a matrix of fields of education-occupations matching. Licenciaturas and engineering degrees/Architecture were equivalent to the Master’s degree in the American system of higher education. With the reform of Bologna, all the degrees (called grados) have a duration of four The database contains 30,379 responses from graduates interviewed only years, equivalent somehow to the American Bachelor’s degree. Some excep- once (a single cross-sectional dataset). This figure is the total number of tions are Architecture (5 years) and Medicine (6 years). observations in the raw data. Mapping the (mis)match of university degrees in the graduate labor market Page 7 of 23 14 No mismatch Horizontal mismatch (adequate match) e.g. BA in Sociology e.g. Graduate in working as director of Medicine working as a production and medical doctor operations Vertical mismatch Vertical and horizontal mismatch e.g. BA in Economics working as an e.g. Bachelor’s in accountingand Biologyworking as a bookkeeping clerk kitchen helper A different area (or no Own area of studies (or particular area) a related area) The most appropriate study area for work Source: author's elaboraon Fig. 1 Higher education graduates’ degrees and their jobs: the education‑job match The EILU2014 questionnaire contained an employee were asked to indicate: Q2. What do you think is, or was, self-assessment of the level and type of education most the most appropriate study area for this work? They had appropriate for the first job after graduation and the several options: B1. Exclusively the area of studies of my current job, that is, the job at the time of being surveyed degree. B2. Some related area. B3. A totally different area. (around 4 years after finishing the university studies). B4. No particular area. We developed two measures of job matching among Following Verhaest et al. (2017), we cross-tabulated the university graduates. For our first measure, we used the answers to the first question about whether employers following question to determine whether or not an occu- requested a university credential vs. a sub-degree level pation required a degree: Q1. What is, or was, the most qualification for the job, and the answers to the second appropriate level of education to carry out this work? question about whether graduates hold positions of their Respondents could select from the following education area of specialization vs. unrelated to their field of study. levels: A1. A university degree. A2. Tertiary vocational We identified four situations of educational mismatch in education. A3. High school. A4. Middle-high school. Fig. 1: adequate match (no mismatch), horizontal mis- Our second measure of matching assessed the quality match, vertical mismatch, and vertical and horizontal of the education-job match by determining whether or mismatch (double mismatch). First, graduates were con- not the field of study of the individual’s degree was related sidered well-matched (no mismatch) if they responded to the job that the interviewee was performing. Subjects A1, and B1 or B2. Second, we identified the horizontal educational mismatch when the type of university educa- tion was not appropriate for the job, but the level of formal The interviewees were asked to exclude occasional/sporadic employment. The appropriate level is preferable to the often-used alternative of the required level. The latter may partly measure formal selection requirements Figure 1 is a simplification to illustrate educational mismatch. We took real whereas the former is more likely to refer to actual job content (Allen and examples referring to the current occupation of Spanish university graduates Van der Velden 2001). four years after graduation. The most appropriate level of studies for the job No university studies University studies are necessary 14 Page 8 of 23 M. Salas‑Velasco Table 3 Distribution of educational (mis)match in the labor misallocated. Although the survey data (EILU2014) market for university graduates in Spain appeared to indicate that there was a slight reallocation of university degrees in the labor market 4 years after First job Current job leaving university, the reality is that the percentage of Freq Percent Freq Percent mismatched graduates in the labor market remains high and does not seem to have changed in the last 10 years Educational (mis)match (Fig. 2). This goes to point out that the educational mis - No mismatch 13,899 57.16 12,387 66.38 match is a structural problem in the Spanish labor mar- Horizontal mismatch 1,422 5.85 1,379 7.39 ket, with an ever-increasing number of graduates that is Vertical mismatch 3,166 13.02 1,725 9.24 not able to absorb an economy with a high rate of youth Vertical and horizontal mismatch 5,827 23.97 3,169 16.98 unemployment and a business environment character- Total 24,314 100.00 18,660 100.00 ized by small firms where graduates cannot make full use The sub‑samples analyzed include only wage ‑ earners workers. See footnote 19 of their university knowledge. However, the problem of for further details Source: author’s calculations from EILU2014 educational mismatch not only affects the Spanish case. It is also relevant in countries such as Estonia and the United Kingdom (Fig. 2). Some explanations: (i) supply education matched the requirements of the job (if they of educated labor exceeds demand (McGuinness 2006); responded A1, and B3 or B4). Third, the educational mis - or (ii) imbalances in composition (individuals studying match was measured as vertical when the acquired level in fields where there is little demand) (Ortiz and Kucel of education was higher than the level of education more 2008). suitable to perform the job, although the area of studies Nonetheless, an in-depth analysis of the reasons for was related to the university degree (if they responded A2 education imbalances in the Spanish labor market was or A3 or A4, and B1 or B2. Finally, the vertical and hori- outside the scope of this paper. Our objective was to zontal mismatch was considered when the attained level identify, in the first and current jobs, which univer - of education was lower than the appropriate, and the type/ sity degrees were more likely to fall in each of the four field of education was inappropriate for the job (if they squares in Fig. 1. Since all possible states are covered, responded A2 or A3 or A4, and B3 or B4). which are disjoint and at this level of analysis their order To provide a better sense of our matching classification, is irrelevant, an appropriate estimation method is offered Table 3 shows these measures of educational mismatch. by the multinomial logit model. We found that about 57–66% of graduates were adequately matched in their jobs in terms of formal (and type of ) uni- 4 Methodology versity education. Around 6–7% were horizontally mis- 4.1 A multinomial logit model of job matching matched. But a considerable percentage of graduates (37% A multinomial logit model (MLM hereafter), also known and 26%, first and current jobs, respectively) worked in as multinomial logistic regression, is suitable for our jobs that didn’t require a university degree. analysis of the educational (mis)match across Spanish Examination of the data in Table 3 revealed that educa- university degrees. Our response variable had four cat- tional mismatch is a significant phenomenon in the labor egorical outcomes that did not have an ordered structure: market for higher education graduates in Spain. Univer- appropriate match (no mismatch), horizontal mismatch, sity graduates accept jobs that do not require a univer- vertical mismatch, and vertical and horizontal mismatch sity degree and/or do not match their specialties. As a (j = 1,2,3,4, respectively). result, qualified human resources in Spain are severely The MLM considers the probability of a certain event j as (McFadden 1974) The sub-samples in Table 3 included only wage-earners workers. From the total sample of 30,379 graduates, self-employed workers were excluded ′ ′ prob Y = j = exp x β / exp x β j (1) (around 7% in the first job and about 10% in the current job). The important k reduction in the number of observations in the current job was mainly due to k=1 the fact that around 22% of graduates were still in their first job at the time of being surveyed and they were not asked questions Q1 and Q2. The rest of the This model provides the probability that an individual cases not considered was due to missing values (around 7% in the first job and with specific characteristics x is in group j. In this paper, about 4% in the current job), and individuals who were not asked questions the predictor variables used were university degrees (nar- Q1 and Q2 because they basically never had worked (around 6% in the first job and about 3% in the current job). row fields of education). Several control variables were In Table 3, to the question of what was the most appropriate study also included in the regressions. area for the job, the majority of horizontally mismatched graduates (77.6%/80.0%) stated “a totally different area” and 22.4%/20.0% “no particular The multinomial logit model is also described in Greene (2012). area” (first job/current job). They would be our explanatory variables of interest. Mapping the (mis)match of university degrees in the graduate labor market Page 9 of 23 14 No mismatch Horizontal mismatch Vercal mismatch Vercal and horizontal mismatch Source: Eurostat and author's elaboraon Fig. 2 Educational (mis)match in Spain and Europe in 2005, 5 years after graduation. Eurostat (Reflex project). Percentages predictor variables, we introduced university degrees. In The natural normalization in our case was β = 0 , with th 23 the survey, there were up to 123 different degrees, which the probability to j outcome be defined as were grouped into 27 categories (narrow fields of educa - exp x β tion) in the regressions. Besides, we considered gender prob Y = j = , if j = 1,2,3 and internship while studying as control variables for the 3 ′ 1 + exp x β k=1 first job; for the current position, gender, having a Mas - (2) ter’s degree, and age. Table 7 (Appendix) showed the And for the baseline category (vertical and horizontal descriptive statistics. mismatch), we would have 4.2 A binomial logit model of external labor mobility prob(Y = 4) = , if j = 4 As we have anticipated in the introduction, this article (3) 3 ′ 1 + exp x β k=1 also aimed to study the empirical relationship between educational mismatch and job mobility. According to However, if we wish to draw valid conclusions about the “job matching theory,” mismatched employees might the direction and magnitude of the relationship between try to improve their fit by changing jobs until an optimal an independent and dependent variable in an MLM, we match is reached (Jovanovic 1979). Jovanovic’s (1979) should calculate marginal effects (Bowen and Wiersema search-and-matching model of the labor market sug- 2004). The marginal effects are defined as the slope of the gested that employees change jobs more often at the prediction function at a given value of the explanatory beginning of their careers. The number of jobs (meas - variable and thus inform us about the change in predicted uring the number of times the individual has changed probabilities due to a change in a particular predictor. employer) is an indicator of job mobility in general, either In this article, we used as the dependent variable in voluntary or involuntary. The EILU2014 dataset contains the MLM the four categories of educational mismatch data on job turnover. We were able to identify whether already shown in Table 3, both in the first job (a varia - or not graduates who were mismatched to their jobs ble that we labeled as mismatchfirstjob) and in the cur - rent employment (labeled as mismatchcurrentjob). As Age was referred to December 31, 2014, and it was already in intervals in The probability of mismatch is compared to the probability of mismatch in the database. In relation to the Master’s degree, we do not know when it was the reference category. awarded, so we have chosen to use this information only in the current job. Switzerland Finland Czech Republic Austria Norway Germany France Italy Netherlands Spain Estonia UK 14 Page 10 of 23 M. Salas‑Velasco after graduation achieved an education-job match after true nature of the relationship between a predictor and moving to other positions in other companies (external the dependent variable in an MLM, we must acknowl- mobility). edge that coefficients […] are potentially misleading” To examine the factors that explained the job match- (Wulff 2015, p. 316). Instead, to be able to draw valid ing, we estimated a binomial logit model (or binary logis- conclusions about relationships, scholars must rely on tic regression). The reduced form for this model would be other interpretational devices such as predicted probabil- (McFadden 1974) ities and marginal effects (Wulff 2015). In this respect, Tables 8 and 9 (Appendix) show the estimated marginal x β effects in the first job and current employment, respec - prob[Y = 1] = i ′ 29 tively. And Tables 4 and 5 show the predicted probabili- x β 1 + e ties for some selected degrees. where Y is the dependent (dichotomous) variable; the Let’s focus first on the educational mismatch in the first x row vector contains the independent or explanatory job. Table 8 shows the estimated marginal effects. A clear variables (including a constant); and β is the vector of advantage of marginal effects is that they provide us with parameters to be estimated. Furthermore, it is assumed rich and intuitively meaningful information not available that the non-observed ɛ’s follow a distribution of logistic through the interpretation of coefficients. However, in probability. order not to tire the reader with the interpretation of all Our dependent variable was gotmatching which took marginal effects, Fig. 3 shows in the four quadrangles of a value of 1 if the graduate was mismatched in his/her education-job mismatch the university degrees for which first job and, after moving to another job (employer), the estimated marginal effects in Table 8 are positive s/he achieved the matching. It took the value of 0 oth- and show statistical significance at 5%. The results reveal erwise, that is, if the graduate was mismatched in the that occupations requiring more specific human capi - first job and after moving to another company was still tal exhibit a lower probability of educational mismatch. mismatched. We restricted the analysis to wage-earn- u Th s, we have three degrees that have the highest like - ers—in both, first job and current job. In relation to the lihood of obtaining an education-job match: Medicine, explanatory variables, and given that the sample for the Nursing, and Veterinary (Fig. 3). For example, having fin - analysis was reduced considerably, we included univer- ished Medical Studies increases the average probability of sity degrees according to broad fields of knowledge and being well-matched in the first job by 0.5364; or having types of degrees. Our explanatory variable of interest finished Nursing Studies is associated with an increase of was the number of different employers for whom the 0.1850 in the average probability of being well-matched university graduate had worked during his/her “short” in the first job after graduation (Table 8). These results working life. In addition, gender was included as a con- are in line with published works focusing on horizontal trol variable. mismatch among university graduates (e.g., Nordin et al. 2010; Robst 2007). In contrast, a horizontal mismatch 5 Results may find it harder to preserve any specific human capi - 5.1 E ducation‑job mismatch among Spanish university tal that is encompassed within a type of qualification, graduates though general human capital may have a role to play This section shows the results of the estimation of the here. We find that graduates in History and Philosophy, MLM. Two types of analysis have been carried out. The and Political Science and Sociology, increase the prob- first one corresponds to graduates’ initial job after leaving ability of being horizontally mismatched (Fig. 3). university. The second analysis corresponds to the edu - However, as seen in Fig. 3, the vast majority of gradu- cational mismatch in their employment at the moment ates occupy positions for which, according to them, a of being surveyed. However, the sign of the estimated university degree was not necessary. On the one hand, model coefficients does not determine the direction of we find that graduates with some degrees such as the relationship between an independent variable and the probability of choosing a specific alternative (Bowen The marginal effects in our research were calculated using the average mar - and Wiersema 2004). “If we are interested in inferring the ginal effects (AME) approach, which relies on actual values of the independ - ent variables (the covariates were all dichotomous). For the global contrast of the estimated models, the Chi-square test was The data collected did not allow us to distinguish between voluntary and used. The null hypothesis is that all the coefficients of the equation, except involuntary separations. Internal labor mobility (intra‐firm mobility) is out - the constant, are null. In the first job: Wald chi2(84) = 3228.82; in the cur- side the scope of this paper given the limitations of the database. rent job: Wald chi2(90) = 36,479.40. In both cases, the associated p-value A permanent job separation involves a change of employers for the was very low (less than 0.001). The result of this test allows us to reject the worker (Jovanovic 1979). null hypothesis accepting both models as good. 27 30 The estimates were made using the statistical program Stata/SE 15.1. In comparison with the reference category. Mapping the (mis)match of university degrees in the graduate labor market Page 11 of 23 14 Table 4 Predicted probabilities of educational mismatch in the first job for selected degrees No mismatch Horizontal Individual of reference 67% Individual of reference 7% Veterinary 82% Political Sc. and Sociology 17% Nursing 83% History and Philosophy 27% Medicine 96% Vertical Vertical and horizontal Individual of reference 6% Individual of reference 20% Labor Relations 16% Journalism 33% Business 27% Biology 33% Tourism 35% Fine Arts 45% The individual of reference is a man who did not do an internship during his studies and got a different qualification than those analyzed. The sum of the probabilities in the four situations is equal to 1 (100%) Source: author’s calculations Table 5 Predicted probabilities of educational mismatch in the current job for selected degrees No mismatch Horizontal Individual of reference 78% Individual of reference 2% Medicine 99% Journalism 14% Political Science and Sociology 15% History and Philosophy 25% Vertical Vertical and horizontal Individual of reference 6% Individual of reference 14% Management and Economics Studies 19% Labor Relations 40% Business Studies 28% Social Work 45% The individual of reference is a 30–34 years old man with no Master’s degree. The sum of the probabilities in the four situations is equal to 1 (100%) The odds practically do not change when considering women graduates Source: author’s calculations Engineering, and Management and Economics Studies, mismatch), graduates end up in non-graduate positions increase the probability of being vertical mismatched. On which contents are not related to their field of study. the other hand, other university degrees such as Biology, Table 4 shows the predicted probabilities of being Fine Arts, Journalism, or Social Work increase the prob- (mis)matched in the first job after graduation for some ability of being vertical and horizontally mismatched selected degrees of Fig. 3. For example, the probability (Fig. 3). For example, having finished Fine Arts is associ - that a Spanish graduate is adequately educated in his or ated with an increase of 0.2007 in the average probability her first job is 67%, but that it increases to 83% for Nurs - of being doubly mismatched (Table 8). Nevertheless, an ing Studies and up to 96% for Medicine. The probability important distinction between the two types of mismatch of being horizontally mismatched is 7%, but it rises to is that a vertical mismatch can preserve some of the spe- 27% for History and Philosophy. The probability of being cific human capital that is encompassed within a type of vertically mismatched is 6%, but it increases to 27% for academic qualification. The engineering or economics Business Studies. Finally, the probability of being vertical fields impart certain job-specific skills that are clearly understood in the job market. But in the case of the full job mismatch (i.e., over-education and field-of-study These probabilities have been calculated using the command margins in Stata/SE 15.1. 14 Page 12 of 23 M. Salas‑Velasco No mismatch Horizontal mismatch Medicine History and Philosophy Nursing Studies Political Science and Sociology Veterinary First job Vertical mismatch Vertical and horizontal mismatch Tourism Sports Fine History and Studies Social Business Science Philosophy Arts Work Studies Management Labor Tourism and Fine Arts Political Relations Studies Economics Science and Studies Sociology Teacher Studies Journalism Labor Biology Relations Technical Engineering Business Studies Psychology Engineering Chemistry Source: author's elaboraon Fig. 3 Mapping the (mis)match of university degrees for higher education graduates in Spain in their first job and horizontally mismatched is 20%, but it rises to 45% well-matched and how the double mismatch has also 32 34 for Fine Arts. been significantly reduced. Let’s focus now on the current job. As we said, the cor- First, workers with a degree in Medicine increase, rect way to interpret the effect of the explanatory vari - again, the probability of being well-matched in their ables on the probability of the different situations of job current jobs. The predicted probability of a perfect matching is to obtain the marginal effects of the regres - match is 99% (Table 5). It is also noteworthy that engi- sors which are shown in Table 9. Figure 4 shows the map neers and technical engineers, who were vertically mis- of degrees according to their educational (mis)match. matched in their first job (over-educated), are no longer It shows only degrees for which the estimated marginal in their current job. As discussed below, they increase effects in Table 9 are positive and show statistical signifi - the probability of achieving an educational match after cance at 5%. Finally, Table 5 shows, for the current job, job turnover. One likely mechanism behind the results the probability of being well-matched (78%), horizontally is the type of human capital individuals acquired dur- mismatched (2%), vertically mismatched (6%), and verti- ing their university education. Medical doctors and cally and horizontally mismatched (14%). It is remark- engineers have highly specialized skills which are to a able the important increase in the probability of being large extent occupation-specific and their transferabil - ity across jobs is limited. Although specialized majors earn a premium on average—specific majors’ graduates earn the most at almost every age (Leighton and Speer The probabilities estimated in Table 4 practically did not change when con- sidering women. Gender was not statistically significant in the estimates of the first job. As two reviewers point out, one of the limitations of self-assessment-based In parentheses, probabilities for the individual of reference. These prob- educational mismatch measurement is that matches could improve over time abilities change according to the degree (see Table 5). because people convince themselves that the match is better. Mapping the (mis)match of university degrees in the graduate labor market Page 13 of 23 14 Horizontal mismatch No mismatch History and Political Philosophy Science and Sociology Journalism Medicine Tourism Studies Languages and Pharmacy Literature Physics Sports Biology Science Chemistry Mathematics Current job Vertical mismatch Vertical and horizontal mismatch History and Social Business Sports Philosophy Work Studies Science Labor Relations Fine Tourism Fine Arts Tourism Arts Studies Studies Political Science and Management Sociology and Journalism Economics Biology Studies Source: author's elaboraon Fig. 4 Mapping the (mis)match of university degrees for higher education graduates in Spain in their current job 2020), a natural concern is that they may be riskier than "specific," actually produce graduates with highly versa - general fields. Skills that are valuable but not transfer - tile skills. For instance, a Bachelor of Mathematics aims able may leave a worker vulnerable to sector-specific to increase the student’s ability in analytical thinking, shocks or economic downturns and may reduce his/her quantitative reasoning, and problem-solving that is nec- probability of finding employment (Leighton and Speer essary for work in mathematically oriented careers (e.g., 2020). actuarial analyst, data analyst, game designer, or invest- Second, several degrees have gone from being cata- ment analyst). In fact, according to the REFLEX survey, loged as vertically mismatched to be cataloged as the most required competencies in the Spanish gradu- horizontally mismatched. There is still a resource misal - ate labor market are mainly transferable skills, “in other location of the human capital in terms of formal quali- words, skills learned in one context that are useful in fications; however, graduates are now carrying out jobs another” (Salas-Velasco 2014, p. 509). which demand a degree, although without requiring Third, Table 9 and Fig. 4 show that there are workers specific university specialties. Typically, as Robst (2007) in jobs not requiring a degree that remain mismatched suggested, those degrees provide more general skills 4 years after graduation. There are university graduates than occupation-specific skills. This would be the case who are still over-educated; this is the case, for example, of History and Philosophy, Journalism, Languages and of Business Studies (28%), and Management and Eco- Literature, Political Science and Sociology, Mathemat- nomics Studies (19%). In the case of Social Work (45%) ics, Pharmacy, Chemistry, or Physics (Fig. 4). For exam- or Labor Relations (40%), graduates are still vertical ple, the predicted probability of horizontal mismatch in and horizontally mismatched. The probability of being the current job is 25% for History and Philosophy, 15% mismatched is shown in parentheses (see Table 5). An for Political Science and Sociology, and 14% for Journal- ism (Table 5). Some of those degrees, usually considered https:// www. prosp ects. ac. uk/ 14 Page 14 of 23 M. Salas‑Velasco interesting result of our study is that some degrees that average grade of the academic record that could approxi- are often thought of as "broad," entailing general human mate the ability. In addition, as one of the reviewers very capital that can be used in different occupations, actually well points out, it is unclear a priori whether the educa- produce skills that are quite specialized (e.g., Bachelor of tional mismatch is a "good" or a "bad" thing for workers. Economics). To resolve this question, one should look at whether edu- Regardless of how much graduates and employers cational mismatch causes a wage penalty or increases the invest in job search, the initial match is unlikely to be risk of unemployment. However, these last two aspects perfect (Allen and Van der Velden 2005). As a result, the are outside the scope of this paper. We hope to give an adjustment mechanisms employed by agents are of great answer in future research, as long as there is information importance. One way of adjusting to initial mismatches that allows it. is by learning new and/or specific skills. In our study, the probability of getting an education-job match increases 5.2 Analysis of educational mismatch and external labor if a master’s degree was completed (Table 9). Also, the mobility probability of being (mis)matched relates to gradu- Many university graduates have likely changed jobs since ates’ age. Being under 30 years old is associated with an graduation, and labor mobility has allowed them to get increase of 0.0502 in the average probability of being an education-job match. Thus, turnover patterns can be well-matched in the current job (Table 9). On the con- informative on the nature of the matching of workers to trary, the probability of being horizontally mismatched jobs. A binomial logit model of external labor mobility relates to graduates 35 years of age or older. Therefore, was presented in Sect. 4.2. The estimated marginal effects the mismatch is increasing in age. This is a result also are shown in Table 6. The results indicate that keep - found in the literature (Somers et al. 2019). In general, ing everything else constant, the greater the number of it seems that the lowest rates of mismatch do happen at employers for whom a graduate has worked, the higher young ages (Bender and Heywood 2011). Younger Span- the probability of achieving a job match. The coefficients ish graduates are most likely to make the transition from associated with gender do not show statistical signifi - a state of mismatch to a state of a match in the early cance in both regressions (Models I and II). However, in stages of their careers. comparison with hard science degrees, graduates in the Lastly, we would like to point out that the role that abil- fields of health sciences and engineering/architecture ity and other unobserved individual characteristics play increase the probability of achieving an education-job in the matching process remained to be tested. “Control- match after job turnover. Conversely, individuals gradu- ling for unobserved heterogeneity might be important ating with arts and humanities degrees—also social and if the probability of educational mismatch is correlated legal sciences degrees—reduce the likelihood of achiev- with innate ability” (Bauer 2002, p. 222). We know that ing the job match after job mobility (Table 6, Model I). In some degrees such as Medicine and STEM degrees (col- particular, having a university degree in the field of health lege programs in science, technology, engineering, and sciences represents an increase of almost 18 percentage mathematics) attract students with higher average ability points in the probability of achieving an education-job and the dispersion around the mean is lower. Therefore, match after external labor mobility. This probability also as was predictable, they are occupying typical gradu- increases appreciably if the individual is an engineer/ ate positions (high-skilled jobs) 4 years after gradua- architect (4.3 percentage points). On the other hand, the tion; and the well-match vs. horizontal mismatch will probability of obtaining a good fit is significantly reduced depend on the relative specificity of college majors and if the worker obtained a degree in the field of arts and the transferability of skills across occupations. However, humanities (decreases almost 15 percentage points), and there are many other degrees where the heterogeneity of if he/she obtained a degree in the field of social and legal the students admitted by universities is much higher, and sciences (decreases by about 5 percentage points). If we some of our results could be a result of ability differences focus on the typology of university studies, we see in between individuals. For example, in Fig. 4, a degree in Table 6 (Model II) that engineering degrees and Archi- Sports Science increases the probability of being both tecture, also technical engineering degrees and Technical horizontally and vertically mismatched; a degree in Polit- ical Science and Sociology increases the probability of In any case, the questionnaire asked the salary (in wide intervals) only for the first job. But this information is not available in the database made public. being both horizontally and completely mismatched, and For the global contrast of the estimated models, the Chi-square test was a degree in Tourism Studies increases the probability of used. The null hypothesis is that all the coefficients of the equation, except being in the three boxes of educational mismatch. How- the constant, are null. In Model I: Wald chi2(6) = 393.15; in Model II: Wald chi2(7) = 265.35. In both cases, the associated p-value was very low (less ever, we could not investigate this issue in-depth due to than 0.001). The result of this test allows us to reject the null hypothesis the limitations of the database; it does not even have the accepting both models as good. Mapping the (mis)match of university degrees in the graduate labor market Page 15 of 23 14 Table 6 Logistic regression of the likelihood of achieving an education‑job match after external labor mobility Average marginal effects Model I Model II dy/dx Std. Err dy/dx Std. Err Number of different employers since graduation 0.0426** 0.0028 0.0435** 0.0028 Female (= 1) 0.0086 0.0111 0.0128 0.0112 Arts and Humanities − 0.1483** 0.0243 Hard Sciences reference Social and Legal Sciences − 0.0458** 0.0185 Engineering and Architecture 0.0426** 0.0206 Health Sciences 0.1768** 0.0272 Diplomatura 0.0138 0.0119 Technical Engineering and Technical Architecture 0.0692** 0.0164 Licenciatura reference Engineering and Architecture 0.1454** 0.0207 Grado 0.0389 0.0427 Other degrees before Bologna − 0.0760 0.1288 Delta‑method to compute the standard errors Model VCE: Robust Dependent variable: gotmatching [= 1 (30%); = 0 (70%)] Number of obs. = 7,471 Wage‑ earners both in the first job and in the current job ** p‑ value < 0.05 Source: author’s estimates Architecture (surveyors), increase the probability of unlikely that an average Spanish university graduate can achieving a job match after job turnover, compared to a change employer ten times in four years. Among other licenciatura. things, because employment opportunities are limited The results in Table 6 suggest that the relative specific - and labor mobility is relatively low in the Spanish labor ity of college majors is associated with a lower probability market. In fact, in the sample used in Table 6, the average of being mismatched after job turnover. But the question job turnover was 2.85. Therefore, educational mismatch that arises is: how many times does a university graduate likely becomes a permanent phenomenon in the job mar- have to change jobs to get a good match? Using the esti- ket for Spanish graduates. mates shown in Table 6, Tables 10, 11 (Appendix) show the probability of achieving the job match according to 6 Discussion the number of times the graduate changes employer. For The mismatch between the educational requirements example, in Table 10, the likelihood of obtaining a job for various occupations and the amount of education match if the individual changes only one time is 23.4%. obtained by workers is large and growing significantly But it would be necessary to "buy" ten jobs to have a high over time (Vedder et al. 2013). Countries that have a rela- probability (68.4%) of achieving the job matching (result tive over-supply of highly skilled workers show higher based on model predictions). The latter may be possi - levels of over-education for graduates (Verhaest and Van ble in an economy such as the United States where the der Velden 2012). This mismatch between education and labor market is characterized by significant flexibility employment has been the focus of substantial research and mobility, but not in Europe, and less in Spain. It is (e.g., Groot and Maassen van den Brink 2000; McGuin- ness 2006). More attention has been paid recently to the so-called horizontal mismatch as well, that is, the mis- “Job shopping refers to the period of experimentation with jobs and accom- panying high rates of mobility, which typically occurs at the beginning of the match between a worker’s field of study and the content working life” (Johnson 1978, p. 261). According to the “theory of job shop- of his/her job (e.g., Robst 2007; Verhaest et al. 2017). ping,” workers search for a high-quality match (e.g., Anderson et al. 1994). In connection with this idea, McGuiness and Wooden (2009), using Australian longitudinal data, identified mismatched workers (over-skilled in their study) as moving rapidly between jobs but also relatively unconfident of finding an The average number of different employers since graduation was 3.53 improved job match. among those workers who got a good education-job fit. 14 Page 16 of 23 M. Salas‑Velasco Education-job mismatches are almost inevitable in the attracts, although not always, students with lower aca- early years of the career of university graduates. New demic ability. graduates rarely have the exact skills employers require. The situations that perhaps should concern us the most This is not (necessarily) a reflection on the shortcomings are those of complete educational mismatch. Almost 17 of higher education. Some skills are best learned on the percent of Spanish graduates were in non-graduate posi- job, and higher education is expected to do more than tions unrelated to their studies four years after gradua- providing a narrowly described set of directly utilizable tion. From the point of view of educational production, competencies (Allen and Van der Velden 2005). Moreo- these situations constitute a clear (external) inefficiency ver, individuals having attended different undergradu - because their studies have been useless: “external effi - ate programs have different stocks of human capital that ciency implies that the results of educational processes can be differentially valued by employers resulting in an are desirable for society (social utility)” (Salas-Velasco initial mismatch for some university degrees. Also, the 2020, p. 163). These degrees may have a high component lack of work experience of recent graduates stops them of education consumption and/or are being demanded by from occupying positions of their educational level. It students with less academic ability. In these cases, per- is then likely that many fresh college students accept a haps better school guidance would be desirable for them position below their educational level because they can to pursue vocational training studies instead of univer- obtain practical skills and experience that can be used sity degrees that are more costly to society. Also, because in different higher-level positions or jobs. The “theory of they are in low-wage occupations, they will not be able to career mobility” already predicted that “it will be rational return to society via taxes that society gave them. There for some individuals to spend a portion of their work- is perhaps a "matching problem" here in the individual’s ing careers in occupations that require a lower level of choice of alternative educational paths. schooling than they have acquired” because “more edu- We cannot give magic recipes to improve the match- cated individuals are more likely to move to a higher-level ing of fresh graduates with their jobs in the Spanish labor occupation” (Sicherman and Galor 1990, pp. 177–178). market. In the first years of their professional careers, u Th s, (vertical) mismatch would be a temporary phe - the educational mismatch may be due to the fact that nomenon, which would greatly reduce the need for pol- they earned a degree but lack the skills or competen- icy intervention. cies that are needed to perform high-skilled jobs. Using In the case of Spain, according to the EILU2014 gradu- information from the REFLEX survey for Spanish higher ate survey, around 13 percent of university graduates education graduates, Salas-Velasco (2014) showed that were in non-graduate jobs just after leaving the higher non-cognitive skills are more demanded in job positions education institutions (HEIs), and just over 9 percent than cognitive skills. However, our graduate survey does remained in mismatched jobs four years after graduation. not contain information on competencies, unlike the They were indeed carrying out jobs related to their stud - REFLEX survey, so this aspect cannot be analyzed. The ies (over-educated but matched in the field of study). But, mismatch may also be related to the search activity of why offer subsidized university degrees if these jobs can recent graduates. University graduates with higher ability be carried out with, for example, higher-level vocational are, in general, more ambitious and involved individuals, training (post-upper secondary school level)? Surround- and search more or more efficiently. Getting a good edu - ing countries such as Switzerland, with a lower offer of cation-job match would thus be related to greater ability. university degrees and an excellent dual system of voca- But our survey also does not contain information on the tional education and training (VET), have a lower inci- ability of recent graduates, so we have not been able to dence of educational mismatch among their university explore this hypothesis either. graduates (see Fig. 2). According to the European Com- The optimal transition from university to employment, mission, the phenomenon of over-qualification in Spain in terms of speed and quality, is also influenced by vari - coexists with the need for more qualified workers mainly ables as important as the structure of the labor market, with a VET background (European Centre for the Devel- the productive model of the economy, and the business opment of Vocational Training 2015). Nonetheless, the cycle. In this regard, it is necessary to highlight the busi- Spanish secondary education system remains academic ness dimension of Spanish firms. In small and medium‐ and university-oriented. There have been attempts to sized enterprises (SMEs) and family businesses, an reform the formal VET system, but it is still less popular education-occupation match can hardly be achieved even (lower social recognition) than the Baccalaureate; and it four years after obtaining a university degree when work- ers have already gained skills from the labor market and/ or have learned to do a better job search. Medium and large companies are those that offer highly qualified jobs, Although it is also true that youth unemployment is much higher in Spain. Mapping the (mis)match of university degrees in the graduate labor market Page 17 of 23 14 and also possibilities for promotion through well-defined questions that remain are whether those non-monetary career ladders. Therefore, if the average business size in benefits outweigh the monetary returns and whether Spain does not increase in the following decades, situa- society is willing to subsidize investments in higher edu- tions of educational mismatch will continue to exist for cation from which a lower tax collection is expected—as many university degrees. In the case of physicians and graduates work in lower-skilled and lower-paying jobs— nurses, their good educational match is due not only to as well as a reduction in the GDP growth through the the fact that they have specific human capital (highly spe - waste of human capital and the implied reduction in pro- cialized skills which are to a large extent occupation-spe- ductivity (Organisation for Economic Co-operation and cific and their transferability across occupations/sectors Development 2016). is limited) but also because their “only” employer is a very large company: the public sector. Thus, we hypothesize 7 Conclusion that the education-job match is more likely in monopso- This paper examines the education-job (mis)match in the nistic labor markets; when there is only one employer of labor market for university graduates. The topic is rel - a certain type of work and the human capital demanded evant and pertinent given the amount of resources that is specific for the positions offered by the monopsonist— both individuals and society allocate to the production of together with a regulation for the access and exercise of highly qualified workers. As the main novelty, this arti - the profession. On the other hand, the business cycle is cle studies the horizontal mismatch which has been less also important. The unemployment of tertiary educa - studied in the literature, that is, when university gradu- tion graduates in Spain was 24 percent in 2014, the year ates hold jobs at their formal qualification level but not in which the graduates of our survey were interviewed. related to their field of study. The paper contributes to This should be noted in interpreting the importance of the existing literature on this topic by providing the the mismatch. In all likelihood, graduates surveyed had taxonomy of educational mismatch in the labor market no choice but to accept non-graduate jobs and/or discon- for university graduates and investigating its incidence nected from their fields of education. Hence, the mis - among Spanish higher education graduates based on match is involuntary. Future graduate surveys should be self-assessments. In addition, the map of degrees done used to check if a more favorable labor market in terms in this article according to the education-job (mis)match of employability improves the education-employment is important for the educational policy given that higher adjustment among graduates. education is highly subsidized in Spain. The article is also The map of degrees done in this article according to the novel in the sense that it incorporates methodological education-job (mis)match is important also for the edu- improvements on some already published papers. cational policy given that higher education is highly sub- In this work, we use a subjective self-evaluation of a sidized in Spain. We can raise some questions that can be sample of 30,379 Spanish university graduates from the answered in future research. Should we change the map class of 2010, surveyed four years after graduation. Grad- of university degrees offering only those that really allow uates inform us, on the one hand, whether or not their a good education-job fit? Is there a rationale for policies current (initial) positions need (needed) a university promoting access to higher education even in the pres- degree and, on the other hand, what is (was) the most ence of a mismatch? Should vocational education be appropriate study area or field of education for these posi - enhanced by guiding students properly about their edu- tions. Tabulating the answers to both questions, we iden- cational choices after completing compulsory education? tify four situations of educational mismatch: appropriate Is the horizontal mismatch acceptable? After all, gradu- match, horizontal mismatch, vertical mismatch, and ver- ates are occupying highly qualified positions although, tical and horizontal mismatch. By estimating a multino- in principle, they do not use the specialized knowledge mial logistic regression, we categorize university degrees gained in college. The answers will depend on the value in each of these four categories. Some results were that society places on higher education and its willing- expected. University degrees that entail specific human ness to pay for it. Some studies have found that there are capital (e.g., Medicine, Nursing, Veterinary, and engi- significant non-monetary benefits from higher education neering/architecture degrees) are more likely to match that accrue even to mismatched graduates, including bet- education-occupation. Other degrees that involve a gen- ter self-reported health, and external benefits for the rest eral human capital that has value across various occupa- of society (e.g., Green and Henseke 2016). However, the tions (e.g., hard science degrees such as Mathematics, Physics, or Chemistry, and liberal arts degrees such as According to Eurostat (https:// ec. europa. eu/ euros tat), unemployment rates in 2014 (second quarter) of tertiary education graduates (ISCED-97 levels 5 and 6) aged 25 to 29 years old were 37%, 24%, and 10% in Greece, Spain, and “For the economy as a whole, total output then depends on how workers the EU-28, respectively. are assigned to jobs” (Sattinger 1993, p. 831). 14 Page 18 of 23 M. Salas‑Velasco History, Literature, or Sociology) increase the prob- better match between their degrees and their jobs. Thus, ability of being horizontally mismatched. In this case, we turnover patterns can be informative on the nature of the do not believe there is a severe misallocation of human matching of workers to jobs. The estimation of a binary resources since workers are occupying graduate posi- logistic regression has allowed us to investigate this ques- tions. It is almost impossible to establish a one-to-one tion. The results indicate that an important percentage of relationship between the field of study and occupation graduates (30%) who were mismatched in their first job for those graduates whose degrees allow more flexibility become well-matched in their current employment after in terms of their careers. Other results are more worry- moving to a different firm. But the results also show that ing in terms of the "waste" of university educational out- a recent graduate needs “to buy” several jobs to achieve put. Some degrees (e.g., Business, and Management and an education-job match. Economics) increase the probability of being vertically An important question that arises in this paper is that mismatched (over-educated) in the first and current jobs. if workers with a Bachelor’s degree are over-qualified for The excessive production of graduates in business and their jobs and people with non-college education have economics at Spanish universities reflects this education- the same earnings as those with BAs in an occupation, it work mismatch. In these situations, workers use in some is hard to justify the time and costs of going to college. way the human capital acquired during their university But we should recognize that formal education, although education. We should ask ourselves whether it would not important, is only one aspect of job matching. Moreover, be better to promote vocational education and training in going to college has non-monetary benefits for individu - many of these cases. It is cheaper to produce vocational als in terms of better health, habits of life, open-minded- skills, and individuals are more likely to be well-matched ness, etc. that should also be taken into account in this in their jobs. The situation is even worse for workers in type of studies. non-graduate positions and study areas unrelated to their studies (e.g., Social Work). In these cases, it would be necessary to consider whether we really should produce Appendix this type of degree at the university. See Tables 7, 8, 9, 10, 11. The paper also shows that many university gradu - ates change jobs and job turnover allows them to get a Mapping the (mis)match of university degrees in the graduate labor market Page 19 of 23 14 Table 7 Descriptive statistics of the explanatory variables included in the multinomial logistic regression First job Current job Frequency Percent Frequency Percent Architecture 176 0.72 120 0.6 Biology 813 3.34 537 2.9 Business Studies 748 3.08 588 3.2 Chemistry 635 2.61 503 2.7 Engineering 1761 7.24 1523 8.2 Fine Arts 221 0.91 128 0.7 History and Philosophy 1178 4.84 841 4.5 Journalism 1253 5.15 867 4.6 Labor Relations 384 1.58 297 1.6 Languages and Literature 932 3.83 701 3.8 Law Studies 870 3.58 668 3.6 Management and Economics Studies 1511 6.21 1220 6.5 Mathematics 356 1.46 295 1.6 Medicine 708 2.91 696 3.7 Nursing Studies 2085 8.58 1506 8.1 Pharmacy 532 2.19 422 2.3 Physics 348 1.43 265 1.4 Political Science and Sociology 306 1.26 229 1.2 Psychology 928 3.82 710 3.8 Quantity Surveyors ( Technical Architecture) 567 2.33 402 2.2 Social Work 676 2.78 491 2.6 Sports Science 465 1.91 356 1.9 Teacher Studies 3054 12.56 2377 12.7 Technical Engineering 2727 11.22 2151 11.5 Tourism Studies 670 2.76 465 2.5 Veterinary 291 1.20 217 1.2 Other university degrees 119 0.49 85 0.5 Female (= 1) 14,817 60.94 11,275 60.4 Internship (= 1 yes) 15,852 65.20 Master’s degree (= 1 yes) 6271 33.6 Age (under 30 years old) 11,040 59.2 Age (from 30 to 34 years old) 4588 24.6 Age (35 years old or older) 3032 16.2 Observations 24,314 18,660 Source: author’s elaboration from EILU2014 14 Page 20 of 23 M. Salas‑Velasco Table 8 Educational mismatches in the first job after graduation. Only wage ‑ earners workers (excluding self‑ employment). Average marginal effects No mismatch Horizontal mismatch Vertical mismatch Vertical and horizontal mismatch dy/dx Std. Err dy/dx Std. Err dy/dx Std. Err dy/dx Std. Err University degrees (narrow fields of education) Architecture 0.0377 0.0583 − 0.0582 0.0382 0.0410 0.0490 − 0.0206 0.0530 Biology − 0.1630 0.0468 0.0030 0.0239 0.0509 0.0416 0.1091 0.0418 Business Studies − 0.2663 0.0472 − 0.0247 0.0251 0.1848 0.0407 0.1062 0.0421 Chemistry − 0.0607 0.0477 0.0044 0.0244 0.0828 0.0416 − 0.0265 0.0435 Engineering 0.0629 0.0460 0.0114 0.0231 0.1008 0.0404 − 0.1752 0.0424 Fine Arts − 0.3210 0.0549 − 0.0349 0.0313 0.1552 0.0438 0.2007 0.0464 History and Philosophy − 0.2702 0.0469 0.0910 0.0227 − 0.0323 0.0428 0.2116 0.0412 Journalism − 0.2000 0.0459 0.0328 0.0231 0.0606 0.0408 0.1065 0.0411 Labor Relations − 0.2700 0.0497 0.0146 0.0250 0.1144 0.0424 0.1411 0.0437 Languages and Literature − 0.0202 0.0473 0.0197 0.0234 − 0.0615 0.0437 0.0620 0.0420 Law Studies − 0.0794 0.0467 − 0.0010 0.0239 0.0386 0.0416 0.0419 0.0420 Management and Economics Studies − 0.1397 0.0456 0.0030 0.0233 0.1259 0.0403 0.0107 0.0412 Mathematics 0.0185 0.0516 0.0322 0.0245 − 0.0231 0.0472 − 0.0276 0.0467 Medicine 0.5364 0.0623 − 0.0686 0.0351 − 0.1667 0.0602 − 0.3012 0.0598 Nursing Studies 0.1850 0.0462 − 0.0349 0.0244 − 0.0052 0.0412 − 0.1449 0.0422 Pharmacy − 0.0633 0.0486 0.0422 0.0240 0.0777 0.0422 − 0.0566 0.0447 Physics 0.0872 0.0527 0.0159 0.0254 − 0.0456 0.0490 − 0.0575 0.0481 Political Science and Sociology − 0.2732 0.0515 0.0533 0.0241 0.0647 0.0445 0.1551 0.0447 Psychology − 0.1524 0.0467 0.0333 0.0236 0.0199 0.0418 0.0992 0.0417 Quantity Surveyors − 0.0108 0.0487 − 0.0372 0.0261 − 0.0048 0.0438 0.0528 0.0434 Social Work − 0.2353 0.0477 0.0085 0.0247 0.0475 0.0421 0.1793 0.0420 Sports Science − 0.2351 0.0492 0.0041 0.0256 0.1905 0.0412 0.0404 0.0443 Teacher Studies − 0.1937 0.0450 0.0075 0.0230 0.1209 0.0401 0.0652 0.0405 Technical Engineering − 0.1030 0.0450 − 0.0037 0.0229 0.1026 0.0401 0.0042 0.0407 Tourism Studies − 0.3350 0.0477 0.0325 0.0240 0.1671 0.0409 0.1354 0.0423 Veterinary 0.1734 0.0570 − 0.0619 0.0382 − 0.0361 0.0505 − 0.0753 0.0519 Other university degrees Reference Reference Reference Reference Control variables Female (= 1) − 0.0024 0.0068 − 0.0108 0.0033 0.0043 0.0047 0.0088 0.0060 Internship (= 1 yes) 0.0200 0.0073 − 0.0266 0.0036 0.0134 0.0052 ‑0.0069 0.0065 Dependent variable: mismatchfirstjob In bold italics, marginal effects that have a positive and statistically significant contribution to the probability of being well‑matched or mismatched in the first job at a significance level of 0.05 (5%). In italics, for a significance level of 10% Standard errors for average marginal effects are computed by the Stata margins command using the Delta‑method Model VCE: Robust Number of obs. = 24,314 Except for rounding errors, the sum of the marginal effects for the four categories must be 0 Source: author’s estimates Mapping the (mis)match of university degrees in the graduate labor market Page 21 of 23 14 Table 9 Educational mismatches in the current job. Only wage‑ earners workers (excluding self‑ employment). Average marginal effects No mismatch Horizontal mismatch Vertical mismatch Vertical and horizontal mismatch dy/dx Std. Err dy/dx Std. Err dy/dx Std. Err dy/dx Std. Err University degrees (narrow fields of education) Architecture − 0.0046 0.0811 − 0.0624 0.0801 0.0140 0.0524 0.0530 0.0576 Biology − 0.2496 0.0610 0.0971 0.0480 0.0478 0.0435 0.1047 0.0472 Business Studies − 0.2766 0.0609 0.0549 0.0485 0.1419 0.0424 0.0798 0.0472 Chemistry − 0.1004 0.0620 0.0963 0.0482 0.0582 0.0434 − 0.0541 0.0497 Engineering 0.0248 0.0603 0.0741 0.0475 0.0458 0.0423 − 0.1448 0.0481 Fine Arts − 0.3855 0.0690 0.0815 0.0527 0.1329 0.0455 0.1711 0.0516 History and Philosophy − 0.3392 0.0610 0.1822 0.0472 − 0.0401 0.0454 0.1971 0.0463 Journalism − 0.2792 0.0600 0.1283 0.0474 0.0429 0.0428 0.1081 0.0464 Labor Relations − 0.3321 0.0631 0.0848 0.0488 0.0757 0.0442 0.1717 0.0479 Languages and Literature − 0.1297 0.0614 0.1083 0.0477 − 0.0433 0.0455 0.0648 0.0472 Law Studies − 0.1508 0.0606 0.0584 0.0481 0.0524 0.0430 0.0401 0.0472 Management and Economics Studies − 0.1568 0.0598 0.0752 0.0476 0.0991 0.0421 − 0.0175 0.0468 Mathematics − 0.1036 0.0645 0.0959 0.0490 − 0.0028 0.0471 0.0105 0.0506 Medicine 2.0203 0.0841 0.1181 0.0801 0.1033 0.0632 − 2.2417 0.0590 Nursing Studies 0.0964 0.0606 0.0080 0.0485 − 0.0101 0.0431 − 0.0943 0.0476 Pharmacy − 0.0907 0.0628 0.1044 0.0483 0.0452 0.0440 − 0.0589 0.0503 Physics − 0.0195 0.0667 0.0975 0.0494 − 0.0075 0.0485 − 0.0704 0.0547 Political Science and Sociology − 0.3081 0.0642 0.1314 0.0483 0.0477 0.0458 0.1290 0.0492 Psychology − 0.1891 0.0606 0.0893 0.0477 0.0092 0.0436 0.0905 0.0468 Quantity Surveyors − 0.1492 0.0625 0.0528 0.0489 0.0058 0.0450 0.0907 0.0479 Social Work − 0.2775 0.0617 0.0504 0.0489 0.0339 0.0441 0.1932 0.0468 Sports Science − 0.2613 0.0627 0.0964 0.0487 0.1369 0.0429 0.0280 0.0496 Teacher Studies − 0.1665 0.0591 0.0540 0.0473 0.0699 0.0420 0.0426 0.0459 Technical Engineering − 0.1423 0.0593 0.0758 0.0473 0.0605 0.0420 0.0060 0.0462 Tourism Studies − 0.3774 0.0613 0.1129 0.0480 0.1261 0.0428 0.1385 0.0472 Veterinary 0.1795 0.0769 0.0179 0.0574 − 0.0707 0.0585 − 0.1268 0.0630 Other university degrees Reference Reference Reference Reference Control variables Female (= 1) − 0.0114 0.0074 − 0.0073 0.0043 0.0021 0.0046 0.0166 0.0061 Master’s degree (= 1 yes) 0.0620 0.0075 − 0.0006 0.0043 − 0.0232 0.0050 − 0.0381 0.0061 Age (under 30 years old) 0.0502 0.0080 − 0.0124 0.0049 − 0.0159 0.0049 − 0.0219 0.0065 Age (from 30 to 34 years old) Reference Reference Reference Reference Age (35 years old or older) − 0.0147 0.0105 0.0439 0.0052 − 0.0107 0.0067 − 0.0186 0.0085 Dependent variable: mismatchcurrentjob In bold italics, marginal effects that have a positive and statistically significant contribution to the probability of being well‑matched or mismatched in the current job at a significance level of 0.05 (5%). In italics, for a significance level of 10% Standard errors for average marginal effects are computed by the Stata margins command using the Delta‑method Model VCE: Robust Number of obs. = 18,660 Except for rounding errors, the sum of the marginal effects for the four categories must be 0 Source: author’s estimates 14 Page 22 of 23 M. Salas‑Velasco Availability of data and materials Table 10 Probability of getting an education‑job match (*) The data used for the analysis are available at the Instituto Nacional de Estadís- according to the number of company changes (Model I) tica repository: https:// www. ine. es/ dynt3/ ineba se/ es/ index. htm? padre= 2785& capsel= 2876 Margin Std. Err p‑ value Number of different employers since graduation Declarations 1 0.2339 0.0174 p < 0.001 Ethics approval and consent to participate 2 0.2752 0.0189 p < 0.001 Not applicable. 3 0.3206 0.0208 p < 0.001 4 0.3697 0.0228 p < 0.001 Consent for publication Not applicable. 5 0.4217 0.0251 p < 0.001 6 0.4755 0.0273 p < 0.001 Competing interests 7 0.5298 0.0293 p < 0.001 The author declares no competing interests. 8 0.5835 0.0309 p < 0.001 Received: 17 March 2020 Accepted: 19 April 2021 9 0.6352 0.0318 p < 0.001 10 0.6840 0.0321 p < 0.001 (*) In comparison with the individual of reference: a man who studied a hard science degree References Number of obs. 7471 Abel, J.R., Deitz, R.: Agglomeration and job matching among college gradu‑ Adjusted predictions. Delta‑method to compute the standard errors ates. Reg. Sci. Urban Econ. 51, 14–24 (2015) Allen, J., Van der Velden, R.: Educational mismatches versus skill mismatches: Model VCE: Robust Eec ff ts on wages, job satisfaction, and on‑the ‑job search. Oxf. Econ. Pap. Wage‑ earners both in the first job and in the current job 53(3), 434–452 (2001) Source: author’s estimates Allen, J., Van der Velden, R. (2005). The flexible professional in the knowledge society: Conceptual framework of the REFLEX project (REFLEX Working Paper Series No. 1). Research Centre for Education and the Labour Market (ROA). Table 11 Probability of getting an education‑job match Anderson, P.M., Meyer, B.D., Pencavel, J., Roberts, M.J.: The extent and conse‑ quences of job turnover. Brookings Papers Econ. Act. Microeconomics according to the number of company changes (Model II) 1994, 177–248 (1994) Margin Std. Err p‑ value Banerjee, R., Verma, A., Zhang, T.: Brain gain or brain waste? Horizontal, vertical, and full job‑ education mismatch and wage progression among skilled Number of different employers since graduation immigrant men in Canada. Int. Migr. Rev. 53(3), 646–670 (2019) Bauer, T.K.: Educational mismatch and wages: A panel analysis. Econ. Educ. Rev. 1 0.1928 0.0090 p < 0.001 21(3), 221–229 (2002) 2 0.2290 0.0095 p < 0.001 Bender, K.A., Heywood, J.S.: Educational mismatch and the careers of scientists. 3 0.2698 0.0106 p < 0.001 Educ. Econ. 19(3), 253–274 (2011) Bowen, H.P., Wiersema, M.F.: Modeling limited dependent variables: Guidelines 4 0.3149 0.0124 p < 0.001 for researchers of strategic management. In: Ketchen, D.J., Bergh, D. (eds.) 5 0.3638 0.0150 p < 0.001 Research methodology in strategy and management, vol. 1, pp. 87–134. 6 0.4157 0.0180 p < 0.001 Elsevier Press, Amsterdam (2004) Chevalier, A.: Measuring over‑ education. Economica 70(279), 509–531 (2003) 7 0.4695 0.0212 p < 0.001 Dolton, P., Vignoles, A.: The incidence and effects of overeducation in the UK 8 0.5240 0.0243 p < 0.001 graduate labour market. Econ. Educ. Rev. 19(2), 179–198 (2000) Duncan, G., Hoffman, S.: The incidence and wage effects of over ‑ education. 9 0.5780 0.0268 p < 0.001 Econ. Educ. Rev. 1(1), 75–86 (1981) 10 0.6301 0.0287 p < 0.001 European Centre for the Development of Vocational Training. (2014). Terminol- In comparison with the individual of reference: a man who studied a licenciatura ogy of European education and training policy (2nd ed.). Publications Office of the European Union. https:// www. cedef op. europa. eu/ files/ 4117_ en. Number of obs. 7471 pdf Adjusted predictions. Delta‑method to compute the standard errors European Centre for the Development of Vocational Training. (2015). EU Model VCE: Robust Skills Panorama: Spain analytical highlight. European Commission. Wage‑ earners both in the first job and in the current jobhttps:// skill spano rama. cedef op. europa. eu/ sites/ defau lt/ files/ EUSP_ AH_ Spain_0. pdf Source: author’s estimates Freeman, R.: The overeducated American. Academic Press, Cambridge (1976) Green, F., Henseke, G.: Should governments of OECD countries worry about graduate underemployment? Oxf. Rev. Econ. Policy 32(4), 514–537 (2016) Acknowledgements Greene, W.H.: Econometric analysis, 7th edn. Prentice Hall, Hoboken (2012) The author thanks the editor and three anonymous reviewers for their valu‑ Groot, W.: Overeducation and the returns to enterprise‑related schooling. able comments on a previous version of this manuscript. Econ. Educ. Rev. 12(4), 299–309 (1993) Groot, W., Maassen van den Brink, H.: Overeducation in the labour market: A Authors’ contributions meta‑analysis. Econ. Educ. Rev. 19(2), 149–158 (2000) As a sole author, I prepared and wrote the entire article. The author read and Hartog, J.: Over‑ education and earnings: Where are we, where should we approved the final manuscript. go? Econ. Educ. Rev. 19(2), 131–147 (2000) Funding Not applicable. Mapping the (mis)match of university degrees in the graduate labor market Page 23 of 23 14 International Labour Organization. (2014). Skills mismatch in Europe: Statistics Robst, J.: Education and job match: The relatedness of college major and work. brief. https:// www. ilo. org/ global/ stati stics‑ and‑ datab ases/ WCMS_ Econ. Educ. Rev. 26(4), 397–407 (2007) 315623/ lang‑‑ de/ index. htm Rossen, A., Boll, C., Wolf, A.: Patterns of overeducation in Europe: The role of field Johnson, W.R.: A theory of job shopping. Quart. J. Econ. 92(2), 261–278 (1978) of study. IZA J. Labor Policy 9(3), 1–48 (2019) Jovanovic, B.: Job matching and the theory of turnover. J. Polit. Econ. 87(5, Part Rumberger, R.W.: The impact of surplus schooling on productivity and earnings. J. 1), 972–990 (1979) Hum. Resour. 22(1), 24–50 (1987) Kampelmann, S., Rycx, F.: The impact of educational mismatch on firm produc‑ Salas‑ Velasco, M.: Do higher education institutions make a difference in com‑ tivity: Evidence from linked panel data. Econ. Educ. Rev. 31(6), 918–931 petence development? A model of competence production at university. (2012) High. Educ. 68(4), 503–523 (2014) Kiker, B.F., Santos, M.C., De Oliveira, M.M.: Overeducation and undereducation: Salas‑ Velasco, M.: The technical efficiency performance of the higher education Evidence for Portugal. Econ. Educ. Rev. 16(2), 111–125 (1997) systems based on data envelopment analysis with an illustration for the Kupets, O.: Educationjob mismat ‑ ch in Ukraine: Too many people with tertiary Spanish case. Educ. Res. Policy Pract. 19(2), 159–180 (2020) education or too many jobs for lowsk ‑ illed? J. Comp. Econ. 44(1), 125–147 Sattinger, M.: Assignment models of the distribution of earnings. J. Econ. Lit. (2016) 31(2), 831–880 (1993) Leighton, M., Speer, J.D.: Labor market returns to college major specificity. Eur. Sicherman, N.: Overeducation in the labor market. J. Law Econ. 9(2), 101–122 Econ. Rev. 128, 103489 (2020) (1991) Leuven, E., Oosterbeek, H. (2011). Overeducation and mismatch in the labor market Sicherman, N., Galor, O.: A theory of career mobility. J. Polit. Econ. 98(1), 169–192 (IZA Discussion Paper No. 5523). Institute for the Study of Labor (IZA). (1990) Mavromaras, K.G., McGuinness, S., O’Leary, N.C., Sloane, P.J., Wei, Z.: Job mismatch Sloane, P.J.: Much ado about nothing? What does the overeducation literature and labour market outcomes: Panel evidence on Australian university really tell us? In: Büchel, F., de Grip, A., Mertens, A. (eds.) Overeducation in graduates. Econ. Rec. 89(286), 382–395 (2013) Europe, pp. 11–45. Edward Elgar, Cheltenham (2003) McFadden, D.L.: Conditional logit analysis of qualitative choice behavior. In: Somers, M.A., Cabus, S.J., Groot, W., Maassen van den Brink, H.: Horizontal Zarembka, P. (ed.) Frontiers in econometrics, pp. 105–142. Academic Press, mismatch between employment and field of education: Evidence from a Cambridge (1974) systematic literature review. J. Econ. Surv. 33(2), 567–603 (2019) McGuinness, S.: Overeducation in the labour market. J. Econ. Surv. 20(3), 387–418 Tao, H.L., Hung, C.Y.: Vertical and horizontal educational mismatches of female (2006) graduates in Taiwan. Asian Econ. J. 28(2), 181–199 (2014) McGuinness, S., Wooden, M.: Overskilling, job insecurity, and career mobility. Ind. Vedder, R., Denhart, C., Robe, J.: Why are recent college graduates underem‑ Relat. 48(2), 265–286 (2009) ployed? University enrollments and labormar ‑ ket realities. Center for College McGuinness, S., Pouliakas, K., Redmond, P.: Skills mismatch: Concepts, measure‑ Aor ff dability and Productivity (2013) ment and policy approaches. J. Econ. Surv. 32(4), 985–1015 (2018) Verdugo, R.R., Verdugo, N.T.: The impact of surplus schooling on earnings: Some Nordin, M., Persson, I., Rooth, D.O.: Education–occupation mismatch: Is there an additional findings. J. Hum. Resour. 24(4), 629–643 (1989) income penalty? Econ. Educ. Rev. 29(6), 1047–1059 (2010) Verhaest, D., Omey, E.: Discriminating between alternative measures of over‑ Organisation for Economic Co‑ operation and Development. (2016). Skills matter: education. Appl. Econ. 38(18), 2113–2120 (2006) Further results from the Survey of Adult Skills. https://doi.org/https:// doi. Verhaest, D., Van der Velden, R.: Cross‑ country differences in graduate overeduca‑ org/ 10. 1787/ 97892 64258 051‑ en tion. Eur. Sociol. Rev. 29(3), 642–653 (2012) Organisation for Economic Co‑ operation and Development. (2018). Education Verhaest, D., Sellami, S., Van der Velden, R.: Differences in horizontal and vertical at a Glance 2018: OECD Indicators. https://doi.org/https:// doi. org/ 10. 1787/ mismatches across countries and fields of study. Int. Labour Rev. 156(1), eag2018‑ ‑ en 1–23 (2017) Ortiz, L., Kucel, A.: Do fields of study matter for over ‑ education? The cases of Spain Wulff, J.N.: Interpreting results from the multinomial logit model: Demonstrated and Germany. Int. J. Comp. Sociol. 49(4–5), 305–327 (2008) by foreign market entry. Organ. Res. Methods 18(2), 300–325 (2015) Park, K.: Vertical and horizontal mismatches and career mobility: Evidence from female college graduates. AsianP ‑ ac. Econ. Lit. 32(2), 109–125 (2018) Publisher’s Note Pellizzari, M., Fichen, A. (2013). A new measure of skills mismatch: Theory and Springer Nature remains neutral with regard to jurisdictional claims in pub‑ evidence from the Survey of Adult Skills (PIAAC) (OECD Social, Employment lished maps and institutional affiliations. and Migration Working Papers No. 153). OECD Publishing. Quintini, G. (2011). Over‑ qualified or under sk ‑ illed: A review of existing literature (OECD Social, Employment and Migration Working Papers No. 121). OECD Publishing.

http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png

Journal for Labour Market Research Springer Journals

http://www.deepdyve.com/lp/springer-journals/mapping-the-mis-match-of-university-degrees-in-the-graduate-labor-wW02qgcWST

Loading next page...

Publisher: Springer Journals
ISSN: 1614-3485
eISSN: 2510-5027
DOI: 10.1186/s12651-021-00297-x
Publisher site: See Article on Publisher Site

Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Mapping the (mis)match of university degrees in the graduate labor market

Mapping the (mis)match of university degrees in the graduate labor market

Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Mapping the (mis)match of university degrees in the graduate labor market

Mapping the (mis)match of university degrees in the graduate labor market

Abstract

Journal

Recommended Articles

There are no references for this article.

Our policy towards the use of cookies