Overqualification in the German labor market

Contents

1. Introduction

2. Starting Point
2.1 Economic Context and Classification
2.1.1 Supply Side
2.1.2 Demand Side
2.1.3 Equilibrium
2.2 Methodology and Literature Overview
2.2.1 Measuring Overeducation
2.2.2 Overeducation Perspective
2.2.3 Upgrading View
2.3 General Findings
2.3.1 Overeducation Perspective
2.3.2 Upgrading View
2.4 Relation to Labor Market Theories

3. Background in and Findings for Germany
3.1 Annotations to German Labor Market and Educational System
3.2 Available Findings for Germany in Literature

4. Own Investigations
4.1 Data Set and Sample Restriction
4.2 Analysis from Overeducation Perspective
4.2.1 Method 1 after Gottschalk und Hansen: Description and Implementation
4.2.2 Method 1 after Gottschalk und Hansen: Placement and Annotations
4.2.3 Method 2 on the basis of Worker Self-Assessment (WA)
4.2.4 Method 2 on the basis of Worker Self-Assessment (WA): Placement and Annotations
4.3 Analysis from Upgrading View
4.3.1 Upgrading View on Method 1 after Gottschalk und Hansen
4.3.2 Upgrading View on Method 2 on the basis of Worker Self-Assessment (WA)

5. Combining and Interpretation of Findings

6. Conclusion
6. Reference List
6.1 Literature Sources
6.2 Online Sources

7. Appendix

1. Introduction

A strange course description of an adult education center created rumblings in the press last year. The title of the new announced course was: De-qualification for academics; accompanied by the explanation that an academic degree or even a doctorate can be a big hurdle for employment in certain jobs like construction helper. But there would be no reason to despair, since the necessary target group specific vocabulary, adequate clothing, and specifically attitude changes can be learned. Despite of these clichés and a scheduled starting date of 1st April, the most funny but also characteristic thing was that several journalists fell for this April fool. The subjective perception of higher education has obviously changed in parts (Karschnick 2012). In former times, people used to say that you have to study to be successful in your job. Is overqualification for university graduates in Germany really existing and increasingly becoming a problem?

However, this question can also be seen in another way. Directly related to this aggregated perspective is the decision making of an individual. If there are many university graduates working in an occupation, which does not require their education, it is obviously not very desirable. Is it worth to study and striving for an university degree? One additional aspect must not be forgotten as well: Also from the point of view of non- university graduates, the potential problem overeducation is very important, since they are directly affected by displacement effects of people with lower levels of education by university graduates.

The number of students in Germany is growing for years and one-time effects like the abolition of the compulsory military service and the conversion to a shortened upper school time from nine to eight years are only strengthening factors. Undisputedly, the tendency of young people to study has increased remarkably and this development will naturally be supported with statistics in the context of this thesis. This phenomenon is moreover not limited to Germany, but mattering in all industrial nations (Wiarda 2011). The European Union emphasized the importance of further supporting this political intended development as part of their Lisbon strategy. The future competitiveness would absolutely depends on an expansion of higher education (Figueirdo et al. 2011:1 et seq.). It can thus be assumed that this trend will continue and perhaps even accelerate. Of course, education has also non-material aspects that are out of question. A positive perceived influence on many different social outcomes is acknowledged (Barro/Lee 2001: 541). However, such aspects do not appear to be relevant here, because it shall be concentrated on material aspects of this topic.

If one considers the far-reaching consequences and the actuality, it is astonishing that this omnipresent topic in media is nearly always discussed in a quite unscientific and more ideological way. A glance at the special literature reveals that the underlying problem could be inherent in too little activity on this field, especially in Germany. In the course of this thesis, it shall be tried to give answers to these important questions connected to overeducation in a well-founded empirical analysis based on economic theory.

For this purpose, the starting point chapter 2 will form the basis for all further engagement by dealing with a topical introduction from the economic perspective. The next step in the following chapter 3 will be a rough sketch of crucial parameters for the German labor market and educational system. In chapter 4, it can finally be started with the analysis, which will be combined and interpreted by reference to everything previous in chapter 5. Completed is this thesis by a brief conclusion in chapter 6.

2. Starting Point

The empirical study of possible displacement effects of people with lower levels of education by university graduates is not new in economic and sociological literature. However, the most investigations take place in the wide range of overeducation and its relation to, for instance, earnings or job satisfaction, without drawing conclusions regarding the in the introduction mentioned accompanied problems of this phenomenon. Therefore, this mostly implicit considered connection of overeducation to the mismatch on the labor market is strongly connected to the used methods for determining and measuring overeducation. A quite large number of different approaches are used for investigation of inadequate qualification in general, because mostly not only the surplus of education is taken into consideration but logically the deficit of education as well. In the course of time several different measurement and estimation approaches were developed and discussed, according to different assumptions and focus areas. The crucial point is not just the fact that different methodology emerged, but more seriously that the results are very often depending on the used method and vary significantly. Therefore, it is essential to give an overview of the differing methodology and its development. This will be done in the second subchapter of this starting point chapter. Furthermore, general findings with regard to this different methodology, subchapter 3, and its relation to labor market theories, subchapter 4, serve as a general basis for this topic as well. First of all, though, a self-made basic thematic placement of education and hence skills into the economic context with particular regard to educational mismatch will be done next, which is the implicit joint initial point in the literature.

2.1 Economic Context and Classification

In the relevant literature, overeducation, also known as over-schooling or surplus schooling, is mostly defined by the level of education an individual has in excess of required education, which means the amount of education to perform adequately in a specific job. Vice versa, undereducation, also known as under-schooling or deficit education, is mostly taken into consideration as well and defined as the lack of education an individual has in relation to the required education of a job (Rubb 2003: 621; Leuven/Oosterbeek 2011: 9). In the following, it will always be used the term education and not schooling, because this denomination is more appropriate for comparing college and non-college graduates, which are here in the center of attention. According to the assumptions of the neoclassical perfect competitive labor market, the equilibrium emerges when the labor supply equals the labor demand. This equilibrium is “generating the competitive wage […] and employment”, whereby the wage “is the market-cleaning wage, because any other wage level would create either upward or downward pressure […]” (Borjas 2005: 164). However, the reality indisputably does not work in this simple way and the underlying requirement of production functions with a homogenous factor labor does not apply. Workers just cannot be aggregated into a single production factor labor, because “in fact, workers are very heterogeneous” (Borjas 2005: 105). This basic fact of a difficulty to determine equilibrium is the starting point of every systematic analysis regarding any kind of overeducation. Education is surely one of the most important characteristics to explain the heterogeneity of workers, the specific demand for them and thus overeducation. The problem of heterogeneity plays an important role in all relevant literature and will be discussed in detail later on several times.

Despite of the heterogeneity problem, the briefly introduced, fundamental neoclassical labor market framework is useful. The distinction between supply and demand of labor, or even better of labor with matching education, helps very much, because the mismatch can be explained by disagreement of supply and demand of adequately educated workers.

2.1.1 Supply Side

The supply side perspective was extensively presented in the introduction, and the easily observable development of onward expansion of higher education in advanced industrial nations at least since the 1960s is also a throughout accepted fact in the economic profession (Barro/Lee: 2001: 550 et seq.). It has to be taken into consideration though that adequate education is not overall the crucial characteristic of a matching worker. According to basic economic literature, it is more likely that the abilities of a worker in general are decisive (Borjas 2005: 247 et seqq.). It is useful to distinguish between abilities and skills, because skills are only the learned capacity to carry out pre- determined results whereas abilities additionally refer to things like aptitude, intelligence, and expertise. However, it is worth mentioning that abilities can also be improved by learning new skills (Ackermann 1988: 288 et seqq.). The ability describes in this context how good you are able perform a job, but this is very difficult to measure. Therefore, in the overall majority of the literature is concentrated on learned, ascertainable skills. These ascertainable skills in turn have two main sources; on one hand, the here focused on formal education like schooling, on the other hand, work-life experience (Tåhlin 2006: 11). Additionally, it has to be mentioned that also many more different secondary skill factors, like for instance on-the-job training, and field of education, play a significant role (Tåhlin 2006: 21). “[…] One must not forget the central point […]. Workers do differ in their abilities” (Borjas 2005: 249). Heterogeneity problems are because of these differences in measurability of skills and abilities virtually preprogrammed. In addition, it has also be taken into consideration that some people voluntarily choose a job, which does not correspond to their genuine skills, and they could easily work in a more challenging job (Gottschalk/Hansen 2003: 452). The heterogeneity is not only based on heterogeneous abilities but also on heterogeneous preferences. Hence, the analysis of overeducation on basis of any kind of educational certificate has to be conducted with caution.

2.1.2 Demand Side

On the demand side of labor, it is only beyond doubt that it has also an important influence. The scientific papers regarding overeducation mostly do not deal explicitly with this topic (Hartog 2000: 131), thus it is a “debate that is too often carried out exclusively from a supply-side perspective” (Figueiredo et al. 2011: 1). How it is possible to neglect this aspect is shown in the next subchapters. Nevertheless, it is widely recognized that the average skill level of job has risen (Korpi/Tåhlin 2007: 2). Gottschalk and Hansen (2003) are one of the very few authors who provide an useful analytical framework to explain the economic effects on this side. They only distinguish between two kinds of education and other characteristics of jobs are excluded. The overall idea is that it can be differentiated between college workers and non-college workers by defining them as perfect substitutes with different productivity. A concentration on a certain kind of skill type is done here, as it was described in the last subchapter. The college degree is seen as an indicator for skills to be able to draw conclusions to the actual decisive abilities. According to basic economic theory, the optimal combination of inputs for a profit-maximizing firm in the long run is made by choosing the optimal relation between quantity of labor and quantity of capital, the isoquant curve, in consideration of the price relationship of these both production factors, the isocost curve (Borjas 2005: 114 et seqq.). “[…] Alternative production techniques are believed to exist and firms are presumed to continuously adapt their production process to available input factors […]” (Tåhlin 2007: 2) Additionally, the firm has also the opportunity within the factor labor to weigh the number of college workers to the number of non-college workers. It is very likely that the firm will make its decision under compering the additional productivity of college workers with the additional costs they cause. For implementing this in the next step, Gottschalk and Hansen introduce a simplified two sectors model with different underlying production functions to their framework. In the first sector, the productivity of college graduates is significantly higher than those of non-college graduates, and in the second sector, the productivity differences are much smaller. Despite of concepts like time rates, the payment of workers should be somehow connected quite close to their productivity (Borjas 2005: 107 et seqq.). “A profit maximizing firm […] will choose to hire college and non-college workers depending on their relative price within the sector” (Cardoso 2004: 3). The demand for labor is therewith defined as the sector specific accordance between wage and productivity. Only college graduates in the vicarious first sector will earn significantly more than non-college graduates, because only here their productivity is higher than the productivity of non-college graduates. Vice Versa, overeducation is hence directly connected to an assumed lack of productivity in the vicarious second sector, since it does not make a difference for payment if somebody has a college degree or has not (Gottschalk/Hansen 2003: 451 et seqq.). However, the reduction to two production sectors is apparently too simple, because the reality encounters many different sectors. Even within an occupation, in the sense of field of activity like variously defined branches or sectors, are possibly different kinds of jobs, in the sense of specific activities. These activities provide in turn many different relations of productivity between college graduates and non-college graduates. This simple framework nonetheless points out the basic principle of demand for different skilled workers according economic theory.

2.1.3 Equilibrium

Now it should be possible to take supply and demand together for getting a rough overview of the equilibrium. The main focus is in both supply and demand on productivity. On one hand it is determined by somebody’s abilities, on the other hand it is the basis for calculating the relative price for hiring somebody, hence the basis for firms to set wages or adjust the production to the market-determined wage. If the productivity of a worker in a certain kind of sector is high, firms are willing to pay a higher wage concerning a better relative price of this worker. The other way round, a higher relative wage in contrast to other sectors will attract more workers with high productivity (Cardoso 2004: 4; Gottschalk and Hansen 2003: 451). “It may be noted that the supply of and demand for skills also affect each other” (Tåhlin 2006: 16). Supposing that a higher education leads to higher ability and therewith productivity, it can be assumed that the adequate education to develop the capabilities is strongly connected to the job characteristics. A significant productivity advantage compared with less educated workers can probably only be achieved in a challenging job. The uncontested view is that especially technological change but also globalization intensified and created these challenging jobs and allowed college-workers to increase the productivity considerable. This complies an increase in demand of educated workers and is often called skill-biased technological progress (Tåhlin 2006; Figueiredo et al. 2011: 23 et seq.; Cardoso 2004: 4). To mention the development on the supply side again for the sake of completeness, the politically initiated expansion has increased the supply of educated workers substantially.

In addition to the aforementioned logical interaction between supply and demand, more mutual influences are conceivable. It is very likely that pupils and therewith prospective workers adapt their plans for studying to changes in demand for college-workers and that governments act in the same sense by modifying their educational policies, for instance through greater financial support. In return, changes in supply lead probably also to a respond on the demand side, because the adjustment of production process in firms could be enhanced in extent and speed for input relation reasons as explained above. In other words, “technological change could endogenously become more biased towards high skills” (Tåhlin 2006:16 et seq.).

To conclude, the question, which is in center of attention in this thesis, is arising from theory as well. The economic context does not provide a clear explanation how the demand and supply for skills go hand in hand and if the problem of misallocation or in consequence overeducation is real. Nevertheless, the basic connections could be shown and several conclusions in the following will be possible on this basis.

2.2 Methodology and Literature Overview

A very useful distinction was applied by the two Swedes Korpi and Tåhlin (2007), who categorize the whole educational mismatch literature into two groups by implicitly considering the economic context and classification presented in the previous subchapter. This visualization in figure 1 will help to understand their argumentation, which will be stated and expanded at the end of the two subchapters, giving a brief overview of development and content of each major strand in literature. The schematic diagram tries to represent the complex relations of the underlying mode of action regarding skill demand and skill supply under special consideration of educational mismatch. It will be very beneficial to understand the distinction of Korpi and Tåhlin as well as further argumentations.

Abbildung in dieser Leseprobe nicht enthalten

Figure 1: Schematic diagram for inter-relations of skill matching

Source: Own illustration.

Assuming that the in 2.1 presented basic economic theory is applicable, these connections evolved out of this also have to be true. They have consciously no arrows, because it shall be emphasized that there is not always a clear causal relation between them but instead often, taken altogether, inter-relation. This simplified overview is naturally neither very detailed nor takes it all so far mentioned aspects into account. Nevertheless, it helps to identify where the reasons for mismatch of education can be found, because if all these connections would work effectively, overeducation would just not exist.

According to Korpi and Tåhlin, there are two major strands in the literature on educational mismatch, namely the so-called overeducation perspective and the upgrading view. The distinction between overeducation and upgrading in itself is not new. However, it is mostly applied in the context of skill change, skill matching, or skill utilization and often conducted in a very fundamental manner. Their direct application to examine the impact of educational mismatch and to view its results it in a greater context is quite unique. Unfortunately their theoretical explanations are concise, so that they will be expanded and work of further authors will be placed in context.

2.2.1 Measuring Overeducation

Before the announced distinction can be presented, it is absolutely necessary to finally answer the question, which should already implicitly be arisen, what kind of criteria can be applied to define what the often mentioned attribute “adequate” for education actually means. As previously written in 2.1, overeductaion is mostly defined as the difference between a worker’s attained level of education and the level of education required for the job the worker holds. The observing of attained education is quite easy, whereas the observing of required education is not. This measuring of the required level of education is needed for both overeducation perspective and upgrading view. In the literature of overeducation or more general educational mismatch, the required skills of a job are classically measured in three different ways. At heart and as already mentioned in chapter 2.2.1, the objective of measuring abilities is very limited. “Job requirements cover many dimensions of worker abilities and aptitudes, and required schooling [here better education] is only a single variable that necessarily compresses much information. Still it ’ s a step in the right direction” (Hartog 2000: 131). Therefore, all three ways of defining the job requirements are restricted to education as a certain kind of skill level. This is in the sense of this thesis, because exactly this relation between job requirements and academic or not-academic education is the object of inquiry.

The quite early-originated classification of Hartog (2000: 132 et seq.) will be used, who classified the measures into the following three categories. The division into three is undisputed and is acknowledged by nearly all publications, even though for instance Battu et al. (2000) or Verhaesta and Omey (2006) name these categories slightly different.

First, there is the possibility to measure required education by job analysis (JA). The basis for this is the information contained in occupational classifications. Every respondent’s employment is assigned for the job titles to an occupational classification and the required education is specified. The most well-known example is the US American Dictionary of Occupational Titles (DOT) as occupational classification, which includes with the General Educational Development (GED) scale an indicator for educational requirements. The advantages of this possibility are on one hand that the assessment is very objective and on the other hand that clear definitions allow investigations on the technology of jobs. However, the expensive necessity of updating, because of technological change for instance, and problems of translating these scale information into a certain level of education are serious disadvantages. The assumed neutrality is doubted as well, because it could be shown that the GED scale sometimes measures social standing of a job rather than educational requirements. Another very important shortcoming is that this method ignores variation in required education across jobs within an occupation, which is a heavy problem since the basis is the occupational classification (Leuven/Oosterbeek 2011: 11; Hartog 2000: 132).

Second, another possibility is to let workers answering survey questions to the adequateness of their education themselves and hence conducting a worker self- assessment (WA). This obviously subjective way of measuring required education can be done directly or indirectly by asking either which level of education is actually required or whether a different, higher or lower, level of education is actually needed. For a further deepening, it can be distinguished between questions to do the job (WAd) or to get the job (WAg). This means requirements to perform the job on one hand and recruitment standards on the other hand, an important differentiation (Verhaesta/Omey 2006: 2114 et seq.). In this case, the up-to-dateness and exactly goal-directedness towards the respondent’s job are the main advantages. The direct questioning possibly leads though to an overstating of the requirements of respondent’s jobs, because they want to emphasize the status of their position. This overstating could systematically be dependent of the respondent’s attributes like sex or education level itself. Furthermore, it is quite unlikely that workers know exactly about the performance of people in their job with a different kind of education or which education is precisely needed to get the job they hold (Leuven/Oosterbeek 2011: 10 et seq.; Hartog 2000: 132).

Third, it is also possible to consider the realized matches (RM) directly by deriving the required education for a certain job from the statements of the interviewed person. For this purpose, the mean or mode of the distribution of the respondent’s attained education for a specific job is set as its required education. It is mostly conducted on the basis of occupational classifications and only if the first two measures are not feasible. The justification of this is that the results are just market results, because “[…] the realized match is the result of demand and supply forces, and does not only reflect requirements” (Leuven/Oosterbeek 2011: 12). Moreover, the important problem regarding the ignoring of variation in required education across jobs within an occupation, as observed in the first method job analysis, occurs here as well. However, the conditions to required data are low. (Hartog 2000: 132 et seq.)

2.2.2 Overeducation Perspective

The first of the two main issues in traditional overeducation perspective literature was presented in the course of the last subchapter by describing the common possibilities to measure this phenomenon. Hereafter, investigations refer their follow-up question to things like earnings or sometimes job satisfaction, but usually only income or wage is directly compared with educational achievements. The first economic analysis in the field of overeducation in general and in the course of this approach were conducted after the onward expansion of higher education in advanced industrial nations at least since the 1960s (Barro/Lee: 2001: 550 et seq.), especially observed by a noticeable increase in the number of college graduates in the US at the beginning of the 1970s (Leuven/Oosterbeek 2011: 2). The initial situation hence is that there is an excess supply for high skilled workers in the labor market. In addition to also worth mentioning authors doing pioneering work like Harris in 1949, Eckaus in 1964, and Collins in 1979, it was particularly Richard Freeman with his book “The overeducated American” from 1976, who became famous for providing one of the first analyses of overeducation. For this purpose, he compared the average income of young people between 25 and 34 with college degree to young people of the same age without college degree, thus high school graduates. The comparison of these two aggregated groups allowed him to calculate an income premium of college graduates. His results of this macro approach were cataclysmic, because he detected in this way a decline of the college premium from 40 percent to 16 percent in only six years at the beginning of the 70s. His argumentation was that the supply outstripped the demand why college graduates have to enter less paid non-college jobs. However, only a few years later, Freeman’s findings, which led due to the extent and his pessimistic predictions to big attention, were relativized by a first improve of this methodology. By regarding a longer observation period and taking into consideration that high school graduates at this age have already longer labor market experience in contrast to college graduates, Smith and Welch could weaken the results considerably in 1978. Indeed they found also a decrease in relative earnings of college graduates, but shortly later, similar conducted investigations for the following years found in turn increases in the college premium. Therefore, a growth in interest arose to discover the reason behind developments relating to overeducation (Leuven/Oosterbeek 2011: 2; Battu et al. 2000: 82 et seq.).

Starting basis for the departure from the aggregated view on overeducation was the canonical Mincerian earnings function, named after its creator Jacob Mincer, one father of modern labor economics. The original or standard version of this multiple regression function relates logarithm of wage to attained years of education, years of work experience and its squared on an individual level (1). It is renowned for being the standard human capital earning function. Duncan and Hoffman (1981) adapted this function for the investigation of educational mismatch by distinguishing between an individual’s attained level of education and the education required for the job held. This specification allows separate estimations of returns for years of education required for the job, years of overeducation, and years of undereducation (2).

(1) Standard Mincerian earnings function:

(2) Educational mismatch specification after Duncan and Hoffman:

The standard Mincerian version in this educational context can be seen as literally taken from its British English meaning, because according to its American English equivalent as meat grinder it mixes up every amount of individual’s education and indicates its effect on wage in general. The possibly negative effects of overeducation are allocated with possibly positive effects of adequate education, whereas the educational mismatch version investigates separately overeducation’s influence on wage at an individual level by comparing workers with adequate education to those who have a surplus education, but ideally do not differ in other characteristics. The same holds true for undereducation. If the separation is conducted as described, “the specification establishes [to some extent] a link with the demand side of the labor market, with allocation processes and with immediately visible variations in reward for individuals with identical schooling operating in different positions” (Hartog 2000: 131). The splitted years of attained education are respectively the supply. This approach was hereafter frequently copied, modified and thus established as widely accepted as the separate sub-field of the economics of educational mismatch (Leuven/Oosterbeek 2011: 4 et seq.; Hartog 2000: 131 et seq.).

Another possibility or approach is to concentrate on the effects of overqualification by only considering whether the individual is overeducated or not, consequently in the form of a simple dummy variable (3).

(3) Overeducation as dummy variable after Patrinos:

In contrast to the approach of Duncan and Hoffman, the benchmark or default is not required but actual education. If the overeducation dummy has the value one, it means that your actual education is higher than the required, and this status of having too much education is directly compared to other ones with the same actual education but having a job requiring this actual education. Whereas a certain number for surplus years of education in the approach of Duncan and Hoffman is always directly related to required education and gives you the returns of these additional years in relation to the required ones. The earnings of individuals with the same job and associated required education are compared. Despite there were older similar approaches from, for instance, Verdugo and Verdugo in 1989, the idea for the alternative concept in this form goes back to Patrinos in 1997 and was criticized for one main reason: The non-consideration of the extent of overeducation is obvious and could lead to miss-estimations of its effects and wrong conclusions. However, it has the big advantage that the creation of a dummy variable is much simpler than identifying accurately the number of years of education (Leuven/Oosterbeek 2011: 9 et seq.; Battu et al. 2000: 85).

The theory of the overeducation perspective takes simultaneously the supply and the demand side of the labor market into consideration by observing variations concerning the wage for educationally matched and mismatched workers. It should become clear that the demand side in the whole presented overeducation perspective is seen as remarkably stable, because its temporal development is not explicitly taken into consideration. Despite all mentioned efforts to update the criteria for adequate education (2.2.1), the job requirements in this major literature strand are mostly assumed to be relatively inflexible and technologically determined. Only already changed production processes and not if firms would like to adapt their production process are taken into consideration. Duncan and Hoffman already emphasized this aspect. Overeducation is always defined at a specific date and not over time, independently from a current or possible future adaption process in demand to the change in supply. The supply is indeed adapted to changes in demand over time as well, but with respect to quite long education times and a strong influence of mostly slowly reacting governments these adaption effects are not so large and to some extent exogenous. Whereas firms are much more free to adapt their production process and by a strong business rivalry they are also forced to do so quite quickly (2.1.3). Especially the non-observance of this progress in the traditional production decision that firms will continuously take different production techniques and thus what kind of workers are available into consideration for their production has a special meaning (2.1.2). A non-consideration of the crucial time dimension cannot be legitimized by inflexible job requirements, because this runs totally counter to the traditional production decision. It is obviously much more likely that this non-consideration implicitly means a perfect adaption of the production process to available workers. “With such a flexibility on the part of firms a mismatch between job tasks and employee skills seems unlikely” (Korpi/Tåhlin 2007: 2 et seq.).

If additionally the connection between productivity and wage is not questioned, this conclusion allows a far-reaching implication. A discrepancy between educational achievements and wage can only be traced back to “unobserved heterogeneity across individuals” (Korpi/Tåhlin 2007: 3). In respect of the schematic diagram in 2.2, it can therefore be made a cut somewhere between educational achievements and abilities, which explains the appearance of overeducation. Any other connections in the overeduation perspective are explicitly or implicitly assumed to function effectively. According to the in 2.1.1 described importance of heterogeneity of workers, at the end of the day only deviation from abilities in relation to one kind of skill, named formal education, is measured. Overeducation is therewith from the start assumed, because workers are very heterogeneous, and this is the logical origin of the name of this literature strand. “All overeducation is [here] rather than real” (Korpi/Tåhlin 2007: 2 et seq.).

2.2.3 Upgrading View

As in the case of the overeducation perspective, the effects of overeducation on wage or income are in the center of attention. Furthermore and as already mentioned, the common possibilities to measure this mismatch phenomenon (2.2.1) take also place in advance of this investigation approach.

The first economic analysis in the course of this approach started significantly later than the overeducation perspective. The main idea was undisputedly established by Sicherman and Galor in 1990 with their study “A Theory of Career Mobility”. In the previous decade, a growth of wage inequality across many western countries, especially in the United States and the United Kingdom, could be observed and constituted their starting point. In an intuitive way they explain this directly within the scope of the supply and demand model (2.1). Because the growing wage gap could in particular be observed across skill or education categories, Sicherman and Galor concluded in short that “the joint occurrence of rising returns to education and an increase in skill supply can only be explained by an even faster growth in skill demand” (Tåhlin 2006: 16 et seqq.). This means in respect to overeducation that its appearance is very unlikely and actually should not be real or at least not consistent in line with this theory. The excess demand has to lead to a full utilization of workers abilities. Hence, real overeducation can only exist voluntarily, which is only rational under career aspects in the short run.

[...]