Access the full text.
Sign up today, get DeepDyve free for 14 days.
A Afonso (2010)
367J. Econ. Ineq., 8
It is well-established that human capital contributes to unequal levels of earnings mobility. Individuals with higher levels of human capital, typically measured through education, earn more on average and are privy to greater levels of upward change over time. Nevertheless, other factors may have an incremental effect over education, namely cog- nitive ability and the skill demands of employment. To deepen insight into whether these aspects contribute to earn- ings mobility over a four-year period, the present study examines positional change in Canada and Germany—two contexts typified as examples of liberal and coordinated market economies. A series of descriptive indices and relative change models assess how different measures of human capital are associated with earnings mobility. The results indicate that, while individuals with higher cognitive skills experience greater earnings stability and upward mobility in both countries, there is only an incremental effect of skills on mobility in Germany once we account for educational credentials. The results also provide evidence on the role of skill demands for earnings mobility; in both countries, advanced skills at work are associated with greater short-term mobility, even while controlling for cognitive ability and other factors. Together the results showcase how longitudinal data containing detailed measures of human capital allow for deeper insight into what facilitates earnings mobility. Keywords: Earnings mobility, Human capital, PIAAC , Relative income position, Cognitive skills JEL: I24, I26, J24, J31 1 Background and Goebel 2017; Bartels 2019). Earnings mobility is not Over the past decades, earnings inequality has increased only an important aspect of economic security and well- in many countries, including Canada and Germany. In being but also a factor that either increases or decreases Canada, inequality grew over the 1980s, especially at overall earnings inequality at the societal level. Because the top of the earnings distribution (Green et al. 2017). it may contribute to convergence and greater equaliza- Until the early 1990s, the increase in earnings inequal- tion (Friedman 1962) or generate inequality through ity in Germany was limited to the upper end of the earn- divergence and unequal change (Raferzeder and Winter- ings distribution but has continued steadily at both the Ebmer 2007), there is a need to understand which fac- upper and lower ends thereafter (Fitzenberger 1999; tors contribute to upward or downward mobility. Many Dustmann et al. 2009; Antonczyk et al. 2012; Grabka factors are at the societal, economic, or policy level, such as economic downturn, while others encompass employ- ment or individual characteristics (Choi 2016; Garnero *Correspondence: apullman@irpe-epri.ca et al. 2019). Education Policy Research Initiative, Graduate School of Public and International Affairs, Faculty of Social Sciences, University of Ottawa, With a focus on individual characteristics, human capi- 5004-120 University, Social Sciences Building, Ottawa, ON K1N 6N5, tal theory suggests that “the ability to deal successfully Canada with economic disequilibria is enhanced by education” Full list of author information is available at the end of the article © 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/. 10 Page 2 of 19 A. Pullman et al. (Schultz 1975, 843) and “steepens age-earnings profiles” two types of analysis: a series of descriptive mobility and (Becker 1962, 29). In this sense, human capital is deemed inequality indices; and multivariate linear regression to provide greater adaptability, ability, and/or opportu- models that measure change in the relative position of nity to favorably navigate changing economic circum- individuals between 2012 and 2016 in the within-country stances in ways that lead to positive earnings mobility. distribution of earnings. Rather than isolate the impact Yet, as signaling (Spence 1974) and cognitive psychol- of a particular feature of either country, our comparative ogy (Peng and Kievit 2020) theory and research suggests, case study approach aims to provide a rich picture of the there may be separate, confounding, or bidirectional associations among earnings mobility, cognitive skills, relationships among education credentials and cogni- job characteristics, and credentials in each context. tive skills. Although cognitive skills generally increase with higher education levels, they also vary at each level 2 Earnings mobility and human capital (OECD 2013a); as an example, not all individuals with a “Intraindividual” mobility measures change in earnings bachelor’s degree have higher cognitive skill levels com- among the same individuals over time (Shorrocks 1978). pared to those who only hold a high school diploma. In Also termed “intragenerational” mobility, research in this short, the study of cognitive skills may capture aspects area examines dynamic positional change (e.g., change of human capital not acquired within the school con- in an individual’s position in the earnings distribution), text and, thus, not reflected in educational credentials. individual growth (e.g., measures of trajectories of change In this sense, cognitive skills may become more relevant over time), long-term inequality (e.g., average earn- for earnings and earnings mobility over the course of ings and period-specific deviations), or risk (e.g., earn - an individual’s working life (Altonji and Pierret 2001). ings instability) (Jäntti and Jenkins 2015). A large body Higher credentials may also provide opportunities for of research on mobility also examines earnings growth earnings mobility through access to jobs with specific and variance over time using cross-sectional data, often characteristics, such as employment with higher skill use with the aim of decomposing permanent and transitory and demand, factors that are also associated with cogni- variance (e.g., Gottschalk and Moffitt 2009). The present tive skills (OECD 2013a). contribution, however, focuses specifically on short-term Using data from the Canadian and German longitudi- mobility over a four-year period among the same indi- nal components of the Programme for the International viduals as measured by upward or downward positional Assessment of Adult Competencies (PIAAC), the present change in earnings percentiles. study enhances our understanding of the associations A positional change signals increased or decreased between earnings mobility and cognitive skills, skill use earnings relative to others. There would be no positional and demands in employment, as well as educational cre- change if all earnings among a group of people increased dentials. A comparison between Germany and Canada or decreased to the same extent. Rather, positive change provides insight into if and how these associations vary takes place when only certain people experience an in two distinct contexts. As will be discussed further increase in earnings (e.g., upward movement from the in the literature review, Germany and Canada differ in 50th to the 60th percentile) and negative change takes many respects, especially in terms of their level of labour place when other people experience a decrease in earn- market regulation and educational systems. Germany ings (e.g., downward movement from the 60th to the 50th is typically identified as a coordinated market economy percentile). Some forms of positional change are associ- that promotes job-specific skills, while Canada is recog - ated with employment trajectories over the life course, nized as a liberal market economy that promotes gen- such as early-career increases or late-career decreases eral skill acquisition (Estevez-Abe et al. 2001). By adding in earnings (e.g., Raferzeder and Winter-Ebmer 2007). a comparative component, this study provides insight Employment transitions (i.e., change in hours or posi- into if and how the relationship between earnings mobil- tion) may also result in a positional change in earnings ity and human capital differs or is similar across varying (e.g., Kosteas 2009). Unlike an approach that measures contexts. change in real earnings over time, a positional change This study answers three primary research questions: approach provides greater insight into the “incidence, To what extent do individual differences in cognitive skills intensity and inequality of positional mobility” (Creedy contribute to differences in short-term earnings mobility? and Gemmell 2019, 753). How does accounting for skill use and demands at work, When it comes to the determinants of positional education, and additional individual characteristics change, human capital is seen as a key driver. Research change this association? And finally, how do the trends demonstrates that individuals with higher levels of edu- observed differ between Canada and Germany? Studying cation experience more upward mobility in their prime short-term mobility over a four-year period, we estimate age working years (Heckman et al. 1998; Connolly and Short‑term earnings mobility in the Canadian and German context: the role of cognitive skills Page 3 of 19 10 Gottschalk 2006) and less variation in earnings during Woessmann 2017). Education level remains constant for economic downturns (Rauscher and Elliott 2016). Using adults who do not return to school; yet, there is evidence Austrian data between 1994 and 2001, Raferzeder and that skill levels do continue to change over the life course Winter-Ebmer (2007) show that education is among the after finishing education (Cunha et al. 2006; Desjardins most important predictors for upward mobility: indi- and Warnke 2012). Therefore, education and cognitive viduals holding an academic qualification experienced, skill level represent related and complementary aspects on average, 6 percentiles greater relative growth com- of human capital that may have distinct associations with pared to individuals with a lower level of education, even earnings mobility. when controlling for a range of background, social, and Cognitive skill level also relates to the extent to which employment attributes. Likewise, Bachmann et al. (2016) an individual engages in skill-based activities (OECD and Aristei and Perugini (2015) analyze patterns of earn- 2013a). “Possessing” human capital through individual ings mobility across European countries and find that education and skill level may have a different relation - individuals with lower education levels have a reduced ship with earnings mobility compared to “using” or probability of upward earnings mobility. “applying” this capital. Re-framing human capital as Education is just one aspect of human capital that connected to everyday practices and the opportunity to may generate differences in earnings mobility and there employ capabilities is an extension made to neo-classi- may be different mechanisms through which educa - cal human capital theory (Klees 2016). Job-requirement tion influences individual earnings. Formal credentials analysis emphasizes the role of everyday activities on may produce “signaling” (Spence 1974) and “sheepskin” both cognitive skill level and earnings. Activities that are effects (Hungerford and Solon 1987) that provide access non-routine and require information-processing skills to labour market positions—and potentially earnings are associated with higher earnings (Green 2012; Ederer mobility over time—through status attainment rather et al. 2015; Mane and Miravet 2016; Mainert et al. 2018). than through the cognitive skill gains education can pro- Greater earnings for people who perform high-skill activ- vide. Because of this, education credentials are a proxy ities at work may be due to skill-biased technological for human capital and often reflective of learning earlier change that decreases earnings for workers performing in the life course and sociodemographic factors, such as routine tasks (Autor et al. 2003; Spitz-Oener 2006; Goos family background (Manzoni et al. 2014; Sakamoto et al. and Manning 2007). Few studies, however, consider the 2018). More direct measures of skill generate additional association between earnings mobility and the opportu- insight into how human capital promotes social and eco- nity for skill use and skill-demands of employment. Some nomic well-being in adulthood (Heckman and Corbin evidence comes from Coban (2017) who analyzes the 2016), which we discuss next, and are often more com- relationship between workplace activities and earnings parable measures across country contexts with different mobility in Germany between 1984 and 2014, and dem- education systems (Hanushek and Woessmann 2008). onstrates that people who perform mainly manual tasks u Th s, the present study contributes to existing research have lower mobility compared to those who perform pri- on human capital and earnings mobility by considering marily non-manual tasks. the contribution of cognitive skill level as a more direct Although prior research demonstrates that earn- but rarely measured aspect of human capital that may ings differentials relate to education (e.g., Connolly and have a distinct association with earnings mobility. Gottschalk 2006) and cognitive skills (e.g., Hanushek et al. 2015), as well as skill-based activities at work (e.g., 3 Extending human capital theory: distinguishing Liu and Grusky 2013), the majority of this research relies between acquired and utilized skills on cross-sectional data and does not generate insight into Multiple bidirectional pathways characterize the rela- how all three aspects relate to earnings mobility. While tionship between cognitive skill level and education, both it is unknown if all three aspects of human capital facili- of which also relate to personal resources and opportu- tate access to higher earnings through mobility, prior nities within social contexts. On the one hand, an indi- research indicates that the earnings returns to cognitive vidual’s skill level in adulthood is influenced by their skills increase with age (Lin et al. 2018) or time spend in educational background (OECD 2013a). On the other a job, for example, when employers learn more about the hand, early ability simultaneously contributes to that skills of their workers (Altonji and Pierret 2001). Educa- very level of educational attainment (Ou and Reynolds tion level is also associated with career progression, often 2014). Although average skill levels generally rise with through occupational sorting (Manzoni et al. 2014) and higher education levels and years of schooling (OECD job mobility (Becker and Blossfeld 2017). Therefore, the 2013a), they also vary in reference to other experiences, present study takes a complementary dynamic perspec- such as employment and training history (Hampf and tive that considers how earnings change over time among 10 Page 4 of 19 A. Pullman et al. Table 1 Canada and Germany country profiles Canada Germany OECD Job protection, union density, temporary employment Job protection against individual dismissal, permanent workers, 2013 0.92 2.53 2.03 Specific requirements for collective dismissal, 2013 2.97 3.63 2.89 Regulation on temporary forms of employment, 2013 0.21 1.75 2.07 Share of temporary employment (age 25–54), 2017 10.3 9.6 10.2 Trade union density, 2015 29.4 17.6 26.3 Collective bargaining coverage, 2015 28.4 56.8 32.7 Job tenure (% age 25–54), 2015 < 12 months 15.1 12.4 15.5 1–3 years 18.9 13.6 13.2 3–5 years 13.8 12.9 12.7 5–10 years 22.4 20.5 24.2 10 years + 29.9 40.6 31.6 Income inequality GINI index of disposable income (post tax/trans., age 18–65), 2015 0.322 0.301 0.315 Mismatch, 2016 Under-qualification 21.7 19.7 18.9 Over-qualification 16.2 17.2 16.8 Public expenditure on labour market programs, 2016 Total (as a percentage of GDP) 0.90 1.45 1.25 Total active measures 0.25 0.63 0.52 Training-related active measures 0.07 0.19 0.12 Passive measures 0.65 0.82 0.74 Source: The OECD indicators on Employment Protection Legislation. Scale from 0 (least restrictive) to 6 (most restrictive) Source: OECD.stat, Employment by Permanency (Dataset: Labour Force Statistics) Source: OECD.stat, Trade Union (Dataset: Trade Unions and Collective Bargaining) Source: OECD.stat, Collective Bargaining Coverage (Dataset: Trade Unions and Collective Bargaining) Source: OECD.stat, Employment by job tenure intervals (Dataset: Labour Force Statistics) Source: OECD.stat Income Distribution Database (Dataset: Social Protection and Well-being) Source: OECD.stat Mismatch (Dataset: Labour) Source: OECD.stat Public expenditure and participant stocks on LMP (Dataset: Labour) the same individuals in Canada and Germany. Compar- evidence that the association between cognitive skills ing these two contexts offers a way to understand if there and earnings is stronger in countries that have a greater is variation in the role of both acquired (i.e., cognitive dispersion in earnings (Jovicic 2016), while higher lev- skills and credentials) and utilized (i.e., skill-based activi- els overall are associated with a more equal distribution ties) human capital for earnings mobility and how it dif- (Afonso et al. 2010). Institutional and policy differences fers by context. may also matter, such as the level of unionization, the strength of employment-protection legislation, the size of 4 The German and Canadian contexts the public sector, and the amount of the minimum wage The association between human capital and earn - (Hanushek et al. 2015; Broecke et al. 2017). ings is known to vary by context. International com- Our research draws upon two contexts that are often parative research based on cross-sectional data shows typified as examples of liberal and coordinated market that although cognitive skill level typically has a posi- economies (Hall and Soskice 2001). In liberal market tive association on individual earnings both before and economies like Canada, competitive market arrange- after accounting for education level (e.g., Hanushek ments generate more flexible forms of employment com - et al. 2015), the association is often weaker in continen- pared to coordinated market economies like Germany tal European countries compared to the United States, where job protection is more stringent. As Table 1 indi- Canada, and the United Kingdom (Leuven et al. 2004; cates, Canada has comparably low levels of job protec- Blau and Kahn 2005). In addition, there is cross-sectional tion, regulation on temporary employment, and length of Short‑term earnings mobility in the Canadian and German context: the role of cognitive skills Page 5 of 19 10 job tenure compared to Germany and the OECD average; High rates of temporary employment in both Canada and however, Germany and Canada have a similar share of Germany may be a contributing factor as job change typi- temporary employees overall. Although Canada appears cally results in more volatile earnings over time (Aristei to deviate from a typical deregulated labour market in and Perugini 2015). terms of its high level of trade union density, collective Along with context differences, it is also important bargaining coverage across all employees is lower than to consider if cognitive skills, education, and the skill the OECD average and in Germany. use/demands at work have different relationships with Certain types of skills are also promoted within each earnings in Canada and Germany. Both countries have context, especially in relation to differences in education similar associations between cognitive skill level and systems, employment-based training, and vocational edu- cross-sectional earnings; for example, in a baseline cation (Estevez-Abe et al. 2001). Germany is often cited model, Hanushek et al. (2015) find that, for each standard as an example of a context that places high importance deviation increase in numeracy skills, earnings increase on standardized formal qualifications (Allmendinger and by 19.3% in Canada and 23.5% in Germany. In addition, Hinz 1997). It promotes high levels of vocational train- the earnings returns to literacy skills are largely the same ing to streamline the transition between school and work across groups with different educational credentials in and reduces education-employment mismatch (Hofacker both countries (OECD 2013a). To the best of our knowl- and Blossfeld 2011). This also means that individuals are edge, there is no research on the relationship between often closely tied to their occupational field and experi - skill-based activities and earnings that compare results ence fewer career changes. While Canada also offers for Canada and Germany. Although Pouliakas and Russo vocational education and training, a comparably smaller (2015) use data from both countries and find a positive proportion of post-secondary students pursue this path- association between various job tasks (e.g., abstract rea- way and instead gain credentials in college and univer- soning) and earnings using cross-sectional data, they sity programs that may offer some (but typically more only report pooled results and there are no conclusions limited) opportunities for on-the-job training, especially about the potential country differences. in fields that do not lead to specific forms of accredita - tion (Kirby 2007). As illustrated in Table 1, although Ger-5 Analytical approach many has a higher level of public expenditure on overall 5.1 Data and sub‑samples and training-based labour market programs compared Data for our analyses come from the Canadian and Ger- to Canada, the proportion of people in Canada and Ger- man samples of the PIAAC study. Initiated by the OECD, many considered as under- or over-qualified is similar. PIAAC aims to provide internationally comparable Prior research on earnings mobility in Canada and measures of skills among adults age 16 to 65 through a Germany highlights country similarities and differ - computer-assisted in-person survey that focuses on soci- ences. In terms of similarities, Chen (2009) shows that odemographic characteristics, employment, skill use at in the 1990s and early 2000s Canada and Germany had home and work, and assessments of cognitive skills in similar and high rates of two-year earnings stability (i.e., three domains: literacy, numeracy, and problem-solving approximately 45–50% of the sample remained in the in technology-rich environments. Both Canada and Ger- same earnings decile) and a lower rate of mobility com- many first participated in late 2011 and early 2012 and pared to other countries in the study. Yet, the similarity extended their PIAAC studies longitudinally, providing a in stability could be due to different reasons. Analyzing unique opportunity to examine how cognitive skill level is the role of institutions on earnings mobility, Pavlopou- related to short-term outcomes between 2012 and 2016. los (2007) and Pavlopoulos et al. (2010) argue that both The Canadian longitudinal data comes from the Lon - unionization (as in Canada) and employment protection gitudinal and International Study of Adults (LISA) sur- (as in Germany) may support greater stability and less vey that includes a sub-sample of PIAAC participants downward and upward mobility. Both forms of employ- (n = 8598 in 2012) who answered the complete PIAAC ment protection mean employers face barriers to firing background questionnaire and undertook the cognitive workers that results in fewer job changes and thus lowers skill assessments. Due to attrition, the overall sample mobility (Bachmann et al. 2016). Still, Chen (2009) dem- size of LISA-PIAAC respondents changed between 2012 onstrates that half of all workers do experience short- and 2016 and, by 2016, there were only 4796 respond- term upward or downward mobility in both contexts. ents. The German longitudinal data comes from the Different measures of mismatch may suggest different findings; for example, Pellizzari and Fichen (2017) suggest Germany has a higher proportion of over- Of note, these rates decrease to 12.9% for Canada and 14.8% for Germany and under-qualified workers compared to Canada. once a model accounts for years of schooling. 10 Page 6 of 19 A. Pullman et al. PIAAC-Longitudinal (PIAAC-L) study that followed up The first part of our analysis comprises three inequal - with PIAAC respondents over three additional waves. ity and mobility indices and uses the original continuous Like Canada, there is attrition over time and the German measure of adjusted monthly earnings. In addition, we sample diminished from 5465 in 2012 to 2967 in 2016. In transform earnings information into deciles to capture an all analyses, we apply longitudinal sampling weights that individual’s relative position in the distribution of earn- correct for attrition in both countries. ings in 2012 and 2016 and the experience of upward, Our analyses use survey responses and assessment downward, or no positional change over the four-year data from Canadian LISA-PIAAC and German PIAAC- period. As we will discuss further below, the final part L respondents who participated in the 2012 and 2016 of the analysis transforms 2012 and 2016 earnings into a surveys. Among these respondents, all analyses exclude measure of positional change between the two periods. individuals who were unemployed, self-employed, or in school (n = 2150 in Canada, n = 1410 in Germany) and 5.2.2 Independent variables did not report earnings (n = 626 in Canada, n = 174 in Our main independent variable of interest is numeracy Germany) in 2012 and 2016. To reduce the influence of skills, one of the three cross-nationally validated infor- outliers and atypical earnings, the top and bottom 1% mation processing skills measured in PIAAC 2012. This of the earnings distribution in 2012 and 2016 are also comprehensive assessment comprises of 56 items that excluded (n = 61 in Germany, n = 65 in Canada). Finally, test “the ability to access, use, interpret, and commu- there is a small amount of missing information at the nicate mathematical information and ideas in order to covariate level (n = 17 for Canada, n = 2 for Germany) to engage in and manage the mathematical demands of 4 8 which we apply listwise deletion. With these exclusions, a range of situations in adult life” (OECD 2013a, 59). our final sample sizes are 1,320 individuals in Germany According to the updated Cattell-Horn-Carroll theory and 1,938 individuals in Canada. of intelligence (McGrew 2009; Schneider and McGrew 2018), quantitative/math ability (i.e., Gq) is a broad skill 5.2 Variables domain at Stratum II and typically has the highest factor 5.2.1 Dependent variables loading on general mental ability (i.e., G). That is, it cor - Our main dependent variables are constructed from relates very highly with markers of general cognitive abil- adjusted self-reported before-tax earnings (not including ity. It is for these reasons that we consider the PIAAC bonuses) in 2012 and 2016. While the German PIAAC-L numeracy measure as a proxy of general cognitive ability. survey reports monthly earnings, the Canadian LISA sur- Our analysis uses numeracy skills, measured through vey reports weekly earnings that are multiplied by four 10 plausible values, in two ways. In the descriptive index- to ease comparability. As Fig. 1 illustrates, the earnings based analyses, scores are transformed into three coarse distribution is similar in 2012 across both contexts; how- categories representing assessed skill level, namely levels ever, in 2016, average earnings have changed to a much 0/1, 2/3, and 4/5, typical groupings that roughly meas- smaller degree in Canada compared to Germany. ure low, average, and high scores (for more information on level thresholds, see OECD 2013b). In the regres- sion-based analyses, numeracy scores are included as a continuous standardized measure that represents each It is necessary to detail the weighting strategy given the complex survey standard deviation increase in assessed score. Models design of PIAAC and survey design differences between Canada and Ger - involving cognitive ability are run separately for each many. First, all indices and models that measure skill level use the 10 plausible of the 10 plausible values and the results are aggregated values produced for the PIAAC assessment scores (OECD 2013b). Second, all longitudinal indices and models include the longitudinal sampling weights produced separately for Canada and Germany to account for survey design and attrition. For the Canadian analysis, the corresponding 1,000 bootstrap weights are also used to estimate the sampling variance. 4 7 In a very small number of cases, we were able to reconstruct missing val- An alternative approach to measuring mobility—one that is based on an ues using responses to other survey waves: five German respondents did individual growth approach rather than a positional change approach—is not report their firm size in 2012 but reported it in later waves without to measure the change in earnings between time one and time two as the experiencing a job change; three German respondents did not report their dependent variable. As reported in Appendix B, these models produce similar work hours in 2016 but did so in 2015; and three Canadian respondents did results to the positional change models in the main results section. not provide their highest education level in 2012 but did in 2014. 8 PIAAC includes assessments for literacy, numeracy, and technology-solv- To generate greater comparability between Canada and Germany and ing in technology-rich environments skills. As there is a strong correlation account for inflation, earnings are adjusted using purchasing power parities between all three skill domains, ranging from 0.740 to 0.868 in Canada and (PPP) information from the OECD (for more information, see: https:// data. 0.753 to 0.872 in Germany, we report results for numeracy skill only. Sensi- oecd. org/ conve rsion/ purch asing- power- parit ies- ppp. htm). tivity analyses show the results for literacy to be very similar and are avail- able in Appendix B. Given the construction of the dependent variables is based on an indi- vidual’s relative position in the earnings distribution, this small difference According to ongoing work (Engelhardt et al. n.d.), this correlation is should not impact the comparison of results between the two countries. r = .70 on the latent-variable level. Short‑term earnings mobility in the Canadian and German context: the role of cognitive skills Page 7 of 19 10 Fig. 1 Earnings Distribution in 2012 and 2016. Box-and-whiskers plots graphically portray the quartile distribution of adjusted monthly earnings in 2012 and 2016 for Canada and Germany. The top and bottom 1% of the earnings distribution are trimmed with corrected standard errors (for further details on vocational education and training (VET) post-secondary using plausible values, see Wu 2005). education (PSE) below the bachelor’s degree level but The other key independent variables of interest are above the high-school level (i.e., ISCED levels 4 & 5 with highest education level and skill use and demands at a VET specialization) ; (3) a non-VET PSE (or first stage work. As measured in 2012, four education levels distin- tertiary) credential below the bachelor’s degree level (i.e., guish among respondents who have: (1) a high-school ISCED levels 4 & 5 without a VET specialization); and 4) diploma or less (i.e., ISCED level 3 and under) ; (2) a 10 11 In Germany, this includes a large number of individuals with a credential In Germany, this category includes individuals with advanced (voca- below the high school level and a vocational qualification (i.e., “Berufsausbil - tional) qualifications not obtained at universities (i.e., “Meister” and dung”). “Berufs- / Fachakademie”). 10 Page 8 of 19 A. Pullman et al. a credential at the bachelor’s (BA) degree level or above level of earnings inequality in the original continuous dis- (ISCED level 5A/6 +). tribution of earnings by measuring the mean difference Two dummy variables measure if participants reported between all possible earnings in the sample, with higher engaging in advanced math and reading at work in 2012. values signaling greater levels of inequality. In Eq. (1), To measure advanced math at work, we use two PIAAC y (y ) represents the earnings of the individual and rep- i j background questionnaire items that ask how often par- resents the average earnings in the sample: ticipants use simple algebra, formulas, advanced math, or N N statistics at work and construct a dummy indicator that I = y − y Gini i j (1) captures people who engage in any of these activities at 2N y i=1 j=1 least once a month. To measure advanced reading at work, respondents who report reading professional journals, Second, to capture positional movement in the origi- publications, or books at least once a month are coded as nal continuous distribution of earnings, the Fields and using advanced reading skills. The analysis also includes Ok (1996, 1999) mobility index M estimates the aver- FO a dummy measure of workplace discretion in 2012 con- age overall level of change in monthly earnings between structed from a single question that asked respondents the 2012 and 2016 (i.e., earnings and earnings ), with i2016 i2012 extent to which they have flexibility in how they do their higher values signaling greater mobility overall: work and group people who answered to a “high” or “very high” extent as having discretion at work. M = earnings − earnings FO i2016 i2012 (2) i=1 5.2.3 Control variables Third, the Fields (2010) index M is a measure of the In the regression analyses, several additional independ- F extent to which mobility equalizes the distribution of ent variables capture possible individual, employment, earnings over time. A zero value indicates that mobility and geographical factors. We include dummy indica- does not change inequality in the distribution of earn- tors measuring gender and native-speaker status (i.e., ings between 2012 and 2016. A positive value indicates if a respondent is a native English/French or German lower inequality through greater upward mobility among speaker) and a categorical variable that captures three individuals located within the lower end of the earnings age groups in 2012 (age 34 or younger, 35–54, and 55 distribution. A negative value indicates higher inequal- or older). As changing jobs or working hours will likely ity over time through greater upward mobility among affect earnings mobility, a dummy variable measures if individuals located within the higher end of the earnings respondents changed jobs between 2012 and 2016 and distribution. The Fields index relies on the Gini index as a continuous variable measures the change in work- a measure of inequality in 2012 (y ) and a vector of earn- ing hours between the same period. Further job charac- ings change between 2012 and 2016 ( ): teristics that may influence the likelihood of receiving a raise comprise firm size, public/private sector, part-time I y Gini employment status, and more than one job in 2012. M = 1 − F (3) I y Gini 0 Given possible regional effects, binary variables measur - ing area of residence in 2012 capture provinces in Canada Fourth, descriptive statistics provide insight into and East/West Germany. A continuous measure captures earnings decile transitions; that is, the proportion of potential years of labour market experience by 2012 (i.e., respondents in Canada and Germany who changed earn- age minus six minus years of schooling). Two binary ings decile between 2012 and 2016. Upward mobility is measures also capture if a respondent was living with a measured as belonging to a higher decile in 2016 com- spouse or partner in 2012 and/or had a child under six pared to 2012, no change is measured as belonging to the years old in 2012. Appendix A provides summary statis- same decile in 2012 and in 2016, and downward mobility tics for all independent and control variables. is measured as belonging to a lower decile in 2016 com- pared to 2012. Given the distribution, the highest decile 5.3 Analysis cannot experience upward mobility, while the lowest As a descriptive overview of the nature of earnings mobil- decile cannot experience downward mobility. ity and its contribution to overall and group-based ine- In the second part of our analysis, we perform multi- quality in each country, descriptive analyses of earnings variate analysis and use linear regression models to gauge mobility examine three indices as well as decile transi- the relative contributions of cognitive skills, skill use and tions between 2012 and 2016. First, the Gini (1921) index demands at work, and education level to earnings mobil- I estimates the level of earnings inequality within each Gini ity. Our dependent variable is a mobility measure that country. Separately in 2012 and 2016, it summarizes the Short‑term earnings mobility in the Canadian and German context: the role of cognitive skills Page 9 of 19 10 Table 2 Earnings inequality and mobility indices for Canada and Germany Gini Fields and Ok Fields Can Ger Can Ger Can Ger 2012 2016 2012 2016 12–16 12–16 12–16 12–16 Overall 0.288 0.287 0.323 0.313 0.296 0.261 0.061 0.056 Numeracy Skills in 2012: Level 0/1 0.290 0.276 0.290 0.328 0.396 0.353 0.117 0.054 Numeracy Skills in 2012: Level 2/3 0.270 0.276 0.308 0.293 0.292 0.261 0.052 0.062 Numeracy Skills in 2012: Level 4/5 0.237 0.232 0.284 0.257 0.252 0.236 0.073 0.080 Advanced math at work in 2012 0.245 0.251 0.285 0.273 0.271 0.227 0.047 0.053 No advanced math in 2012 0.291 0.286 0.323 0.314 0.316 0.307 0.071 0.066 Advanced reading at work in 2012 0.265 0.258 0.289 0.274 0.268 0.239 0.071 0.061 No advanced reading in 2012 0.276 0.289 0.333 0.327 0.334 0.305 0.047 0.065 Low discretion at work in 2012 0.297 0.288 0.316 0.313 0.302 0.309 0.070 0.056 High discretion at work in 2012 0.267 0.277 0.317 0.302 0.289 0.240 0.047 0.056 Education: HS or less 0.265 0.278 0.299 0.299 0.325 0.293 0.059 0.061 Education: VET PSE (below BA) 0.263 0.282 0.305 0.272 0.292 0.255 0.036 0.930 Education: non-VET PSE (below BA) 0.293 0.265 0.281 0.249 0.297 0.223 0.109 0.860 Education: BA degree or above 0.248 0.237 0.265 0.250 0.278 0.241 0.082 0.074 captures the change in earnings percentiles between Overall, the Gini results are similar across both time 2012 and 2016. By definition, individual mobility can vary periods but are slightly lower in Canada. In both contexts between 99 (i.e., indicating an increase from the bottom and periods, the Gini index decreases among individu- to the top of the distribution) and -99 (i.e., indicating a als with the highest numeracy levels. In both Canada and decrease from the top to the bottom of the distribution). Germany, people who self-report performing advanced For example, an individual who moved from the 50th math and reading at work have earnings that are more percentile in 2012 to the 65th percentile in 2016 would equal compared to those who did not. The Gini index is have upward mobility of 15 percentiles. Models with this similar by level of workplace discretion in Canada and type of dependent variable are commonly termed rela- Germany in 2012 and 2016, although it is slightly higher tive change models (e.g., Raferzeder and Winter-Ebmer in 2012 among respondents in Canada who report low 2007). workplace discretion relative to those who report high We use a series of models to examine how the relation- discretion. In both Canada and Germany, the Gini index ship between mobility and skills changes when additional is typically lowest among individuals with a BA degree or independent variables are added to the model. Model 1 above in both 2012 and 2016; although in 2016, German only includes numeracy score and initial decile in 2012. respondents with a BA degree or above or non-VET PSE Model 2 adds indicators measuring skill use and demands below the BA degree level have markedly similar results. at work and Model 3 introduces the highest education The Field and Ok index provides insight into the overall level. As in the equation below, Models 4 adds a vector level of earnings mobility between 2012 and 2016, with βX of control variables. higher values signaling greater mobility in terms of either upward or downward change. The overall index score is Mobility =β + β Numeracy + β Initaldecile i 0 1 i 2 i slightly higher in Canada, suggesting greater earnings + β Mathatwork + β Readingatwork 3 i 4 i mobility compared to Germany. In both countries, indi- + β Discretionatwork + β VETPSE 5 i 6 i viduals with lower numeracy levels experience greater upward or downward mobility, as do people who self- + β non−VETPSE + β BAorabove 7 i 8 i report not performing advanced math or reading at work + βX + e i i and being in positions with lower levels of discretion in (4) 2012. Similar to the skill level results, respondents in Canada with higher credential levels typically experience 6 Results lower mobility compared to people with lower education. 6.1 Descriptiv e analyses of earnings mobility In Germany, the Field and Ok index is similar across all Table 2 presents the results of the Gini, Fields and Ok, and Fields indices for the Canadian and German samples. 10 Page 10 of 19 A. Pullman et al. Table 3 Proportion of downward, same, and upward decile change, 2012 to 2016 Canada Germany downward same upward downward same upward All 0.279 0.369 0.352 0.262 0.407 0.331 Numeracy Skills in 2012: Level 0/1 0.288 0.327 0.385 0.320 0.362 0.318 Numeracy Skills in 2012: Level 2/3 0.281 0.366 0.354 0.271 0.388 0.341 Numeracy Skills in 2012: Level 4/5 0.261 0.429 0.310 0.196 0.502 0.302 Advanced math at work in 2012 0.285 0.399 0.316 0.274 0.408 0.318 No advanced math in 2012 0.275 0.352 0.373 0.251 0.407 0.342 Advanced reading at work in 2012 0.287 0.356 0.373 0.268 0.406 0.326 No advanced reading in 2012 0.270 0.382 0.330 0.254 0.408 0.338 Low discretion at work in 2012 0.255 0.375 0.370 0.263 0.373 0.363 High discretion at work in 2012 0.309 0.363 0.328 0.261 0.426 0.313 Education: HS or less 0.304 0.354 0.342 0.295 0.376 0.329 Education: VET PSE (below BA) 0.328 0.412 0.260 0.238 0.438 0.324 Education: non-VET PSE (below BA) 0.253 0.353 0.393 0.227 0.442 0.330 Education: BA degree or above 0.238 0.371 0.390 0.216 0.445 0.339 three PSE education levels and lower when compared to downward decile change in Germany, yet they also have those with a HS diploma or less. slightly higher rates of upward change in Canada. For the Field index, positive and higher values signal In both contexts, individuals who did not perform that mobility has an equalizing effect on earnings ine - advanced math at work have somewhat higher rates quality. The overall Field indices suggest earnings mobil - of upward mobility, while those who did engage in ity equalizes earnings to a similar extent in Canada and advanced reading have higher rates of upward mobility Germany. However, mobility has a larger equalizing in Canada. Individuals who have low levels of workplace effect among individuals with the lowest numeracy levels discretion in both countries have higher rates of upward in Canada and the highest numeracy levels in Germany. mobility. In Germany, people with high discretion at In Canada, the Field index is higher among individuals work have greater stability and lower rates of upward who do not use advanced math and self-report low lev- mobility compared to those with lower levels of discre- els of workplace discretion, but it is lower among those tion. In Canada, those with high discretion have higher who do not have the opportunity to engage in advanced rates of downward mobility and similar rates of stability reading. In contrast, the Field index is similar by level of compared to individuals with low discretion jobs. In both workplace discretion and reading activities in Germany, countries, individuals with a BA degree or above have the but higher for individuals who do not use advanced math lowest rates of downward mobility. Although in Germany at work. By education level, earnings equalize to a greater upward mobility is similar across all education levels, in extent among individuals with a non-VET PSE creden- Canada, individuals with a non-VET PSE credential and a tial in Canada and those with a VET PSE credential in BA degree or above have higher rates of upward mobility Germany. compared to the other two levels of education. Table 3 assesses the proportion of individual upward or downward decile change and stability between 2012 and 6.2 Multivariate analysis 2016. The overall results are comparable between Can - Table 4 presents the results of relative change models that ada and Germany with only small differences. A slightly explain positional percentile change in monthly earnings higher proportion of individuals in Germany are in the between 2012 and 2016 for the Canadian sample. Model same earnings decile for both periods, while there are 1 demonstrates that, controlling for initial 2012 decile somewhat higher levels of upward and downward mobil- position, each standard deviation increase in numer- ity in Canada. Consistent with the findings in Table 2, acy scores relates to a 2.4 percentile upward change in people with higher numeracy skill levels experience the distribution of earnings over the four-year period. greater stability in Canada and Germany. Respondents Although this coefficient may seem small at first glance, who scored at skill levels 0 or 1 have the highest rate of it is important to consider it in reference to the distribu- tion of numeracy scores. For example, on average for the Short‑term earnings mobility in the Canadian and German context: the role of cognitive skills Page 11 of 19 10 Table 4 Relative change in Canadian monthly earnings percentiles, 2012 to 2016 (1) (2) (3) (4) ** ** Numeracy 2.433 (0.759) 2.212 (0.760) 1.315 (0.777) 0.449 (0.706) Initial position a *** *** *** *** Earnings decile in 2012 − 2.518 (0.224) − 2.696 (0.255) − 2.869 (0.250) − 2.999 (0.333) Work characteristics in 2012 Advanced math at work 1.102 (1.182) 0.999 (1.170) 0.832 (1.096) b * * Advanced reading at w ork 2.549 (1.264) 1.515 (1.193) 2.151 (1.045) High discretion − 0.113 (1.094) 0.149 (1.062) 0.696 (0.951) Education VET PSE (below BA) 1.249 (1.730) 1.625 (1.454) d *** ** Non-VET PSE (below BA) 5.393 (1.615) 4.588 (1.491) d *** *** BA degree or abo ve 7.329 (1.509) 5.760 (1.634) Controls e *** Female − 5.669 (1.202) Age: 34 and under − 0.400 (1.589) Age: 55 and older − 1.576 (1.715) Non-native language speak er 0.079 (1.438) h ** No job changes between 12–16 3.139 (1.214) *** Change in working hours 12–16 0.600 (0.068) i ** Firm size: 11–50 ppl 4.421 (1.612) i *** Firm size: 51–250 ppl 5.662 (1.668) i *** Firm size: 251 + ppl 6.705 (1.769) Public sector 0.514 (1.036) Part time status − 2.681 (1.602) ** Labour market experience − 0.247 (0.085) More than one job in 2012 − 0.011 (1.509) Living with partner in 2012 1.544 (1.127) Child under 6 in 2012 − 1.515 (1.282) Region included yes *** *** *** ** Constant 14.728 (1.330) 14.054 (1.304) 11.877 (1.476) 13.117 (4.035) R 0.12 0.13 0.15 0.35 Observations 1938 * ** *** Standard errors in parentheses; p < 0.05, p < 0.01, p < 0.001 2012 earnings decile as a continuous variable Reference group: little math/reading at work Reference group: low discretion Reference group: high-school diploma or less Reference group: men Reference group: aged 35–54 years Reference group: native speaker Reference group: job change between 2012 and 2016 Reference group: firm size of 1–10 ppl Reference group: private sector Reference group: full-time employed Reference group: only one job in 2012 Reference group: not living with partner in 2012 Reference group: no child under 6 in household in 2012 10 Page 12 of 19 A. Pullman et al. entire original PIAAC sample, the difference between distinct associations with earnings mobility over a four- each skill level (e.g., level 1 versus level 2) is approxi- year period in Germany once the model includes all con- mately one standard deviation. This means that in Model trol variables. 1 individuals with skills at level four (i.e., high skills) are Regarding the other variables of interest, the results estimated to experience a relative change in earnings indicate that advanced reading skills at work has a posi- of roughly 9 percentiles compared to individuals who tive relationship with earnings mobility across all models. scored at level zero (i.e., low skills). Controlling for work Like Canada, performing advanced math at work does characteristics reduces the numeracy coefficient slightly not have an association with later earnings mobility. Dif- to 2.2 in Model 2. Education level has a large effect on fering from the Canadian results, high discretion at work the model and reduces the size and significance of the is only associated with greater mobility over time in the numeracy coefficient to 1.3 in Model 3. Finally, once final model when controlling for all variables. Compared Model 4 introduces all control variables, the numeracy to individuals with a high school diploma or less, those coefficient reduces to a 0.4 percentile change. with a non-VET PSE credential below the BA level are In terms of the other independent variables of interest, privy to a 3.4-point change in their earnings position and advanced reading at work has an association with short- those with a BA degree or above have a 3.9-point change term positional change over four years and, in the final in their earnings position. Like Canada, there is no statis- model (i.e., Column 4), results in an average increase of tically significant difference in earnings position between 2.2 percentiles between 2012 and 2016 among individu- individuals who have a VET PSE credential below the BA als who self-report engaging in these activities compared level compared to those with a high school diploma or to people who did not. Although, the inclusion of edu- less. cation level in Model 3 reduces the size and the signifi - cance of the numeracy score coefficient, the final model 7 Discussion results show that individuals with a non-VET PSE cre- As earnings stability and the possibility of upward mobil- dential (below the BA level) have a 4.6 percentile change ity are key components of economic well-being, varying in earnings and those with a BA degree or above have a levels of mobility among only specific groups can imply 5.8 percentile change in earnings between 2012 and 2016 growing inequality (Oh and Choi 2018; Tansel et al. compared to people with a high-school diploma or less. 2019). Previous literature demonstrates that individuals Importantly, the results demonstrate that education and with higher education levels are more likely to experience skills do not have distinct correlations with positional upward earnings mobility and stability over short and earnings mobility in Canada when a regression model long periods of time (Heckman et al. 1998; Connolly and includes both variables. Gottschalk 2006; Raferzeder and Winter-Ebmer 2007; Table 5 provides the results of the relative change Rauscher and Elliott 2016). Nevertheless, the relation- model for Germany. In Model 1, each standard deviation ship between human capital and earnings mobility is less increase in numeracy scores results in a 3.0 percentile straightforward than commonly assumed when consider- increase in positional earnings between 2012 and 2016. ing cognitive skills and skill use/demands at work along- Similar to Canada, the skill coefficient becomes smaller side education level. when controlling for work, education, occupational, and Contributing to research on the influence of human other individual characteristics. Nevertheless, the size of capital on earnings mobility, the present article examines the association between skill and mobility for the Ger- how short-term earnings mobility over a four-year period man sample remains descriptively larger than the Cana- differs by skill level in Canada and Germany. By taking dian results across all models, and the coefficient remains into account skills, skill use and demands at work, and statistically significant in Model 3 when controlling for educational credentials, the aim is to identify how dif- education level. Like the results for Canada, the final ferent measures of human capital interact and are asso- model indicates that education and skills do not have ciated with earnings mobility. Our research draws upon two contexts that are often typified as examples of liberal 12 and coordinated market economies (Hall and Soskice To investigate how economic development is related to earnings mobil- ity, we also estimated the German results separately for West and East Ger- 2001). In a liberal market economy such as Canada, com- many. The results for the West German sample are broadly consistent with petitive market arrangements should result in greater the results in Table 5. The main differences are that the coefficient for numer - levels of mobility and a stronger association between gen- acy is slightly larger and also statistically significant in Model 4. For the East Germany sample, we find no statistically significant relationship between eral measures of human capital and positional change. In numeracy and earnings mobility, a result that is likely due to the small sample contrast, in coordinated market economies like Germany, size (n = 245). Using a larger dataset, future studies should further investigate where job protection is more stringent and high levels potential regional differences. of employment-based training and vocational education Short‑term earnings mobility in the Canadian and German context: the role of cognitive skills Page 13 of 19 10 Table 5 Relative change in German monthly earnings percentiles, 2012 to 2016 (1) (2) (3) (4) *** *** ** Numeracy 3.011 (0.630) 2.715 (0.635) 2.157 (0.675) 0.804 (0.539) Initial position a *** *** *** *** Earnings decile in 2012 − 1.924 (0.190) − 2.145 (0.232) − 2.330 (0.240) − 2.756 (0.289) Work characteristics in 2012 Advanced math at work 0.861 (1.096) 0.829 (1.086) 1.154 (0.906) b ** * *** Advanced reading at w ork 3.137 (1.011) 2.337 (1.015) 3.034 (0.873) c * High discretion 0.794 (0.907) 0.750 (0.912) 1.629 (0.817) Education VET PSE (below BA) 0.367 (1.731) 0.305 (1.390) d * * Non-VET PSE (below BA) 3.318 (1.471) 3.431 (1.404) d ** * BA degree or abo ve 4.761 (1.493) 3.910 (1.545) Controls e *** Female − 4.240 (0.865) Age: 34 and under − 1.554 (1.565) Age: 55 and older 0.886 (1.816) Non-native language speak er − 2.688 (1.808) h * No job changes between 12–16 2.911 (1.257) *** Change in working hours 12–16 0.760 (0.066) Firm size: 11–50 ppl 2.348 (1.243) i ** Firm size: 51–250 ppl 4.022 (1.257) i *** Firm size: 251 + ppl 5.592 (1.443) j * Public sector 1.882 (0.903) k * Part time status − 2.986 (1.380) *** Labour market experience − 0.305 (0.077) More than one job − 0.805 (1.533) m * Living with partner in 2012 − 2.022 (0.946) Child under 6 in 2012 − 0.193 (1.286) Region included yes *** *** *** *** Constant 11.516 (1.131) 9.962 (1.187) 9.836 (1.220) 17.014 (3.426) R 0.11 0.12 0.13 0.42 Observations 1320 * ** *** Notes: Standard errors in parentheses; p < 0.05, p < 0.01, p < 0.001 2012 earnings decile as a continuous variable Reference group: little math/reading at work Reference group: low discretion Reference group: high-school diploma or less Reference group: men Reference group: aged 35–54 years Reference group: native speaker Reference group: job change between 2012 and 2016 Reference group: firm size of 1–10 ppl Reference group: private sector Reference group: full-time employed Reference group: only one job in 2012 Reference group: not living with partner in 2012 Reference group: no child under 6 in household in 2012 10 Page 14 of 19 A. Pullman et al. emphasize industry-specific over general forms of human and portability of skills and how they impact the eco- capital (Estevez-Abe et al. 2001), we expect to find lower nomic behaviour of individuals in different contexts. mobility and a weaker association between general skills This study goes beyond this research and aims to pro - and mobility. However, our results demonstrate similari- vide a clearer understanding of how different measures of ties and differences in both countries, as well as patterns human capital—that is, education and cognitive skills, as that deviate from the liberal and coordinated market well as the opportunity for skill use and the skill-demands typology. of employment—are associated with earnings mobility. In terms of similarities between Canada and Germany, In this way, it also adds to the argumentation that there is the level of overall earnings inequality (as measured a need to re-situate research on the economic returns to by the Gini index) is slightly lower for individuals with human capital that is not based on individual skills alone, higher skill levels and educational credentials in both but on a combination of credentials, skills, and workplace countries. The Field and Ok index indicates that individu - opportunities that generate socioeconomic inequality als in Canada and Germany with higher skill levels expe- even over short periods of time (see e.g., Kleese 2016). rienced greater earnings stability between 2012 and 2016, Our research has implications for understanding the as did those who self-reported performing advanced relationship between human capital and earnings mobil- reading and numeracy workplace activities and holding ity. In line with theories of “signaling” and “sheepskin” a PSE credential. In the baseline multivariate models, effects, it confirms that people with higher education lev - there is a positive relationship between skills and upward els—especially non-VET PSE credentials in both Canada mobility for respondents in both Canada and Germany. and Germany—are privy to greater levels of earnings Advanced reading at work and holding a non-VET PSE mobility compared to those with lower levels of educa- credential below the BA level or a BA degree or above are tion even when controlling for observed ability. Policy also associated with short-term upward percentile mobil- that supports skill development often aims to improve ity, controlling for all other factors, in both Canada and economic well-being, with “the belief that better-edu- Germany. cated citizens yield a wealthier country […] a cornerstone Despite the overall cross-country similarities, the find - of public policy almost everywhere” (Hunter and Leiper ings are not universal and there are some notable differ - 1993, 22). Underlying these policies is belief in meritoc- ences between the Canadian and German results. Over racy that assumes people with higher education levels the four years under consideration, the Field index results have greater earnings due to their greater abilities. How- suggest that earnings became more equal for individuals ever, the results suggest that even among people with with a non-VET PSE credential below the BA level and the same ability (at least measured in terms of cognitive those with lower numeracy levels in Canada, while in skills), those with a non-VET PSE credential experienced Germany earnings became more equal among individuals greater mobility overall. Thus, it is not necessarily skill with VET PSE credential below the BA level and higher but also formal credentials that relate to earnings mobil- numeracy levels. The relative change model also demon - ity over time. strates that both numeracy skill and education levels have distinct associations with upward mobility in Germany 8 Conclusion prior to controlling for other variables in Model 3, while Although we provide several new insights into the the numeracy coefficient becomes non-significant once relationship between measures of human capital and the same Canadian model specification includes educa - earnings mobility, there are limitations that are neces- tion level. However, once Model 4 includes all control sary to discuss. Although the majority of our indica- variables, there is no distinct association between skills tors were measured prior to earnings mobility between and positional change in Germany. Finally, the relative 2012 and 2016, we do not establish causal mechanisms change models also illustrate that workplace discretion in this study as there may be unobserved confounders is associated with positive mobility in Germany but not that generate spurious relationships and reverse causa- Canada. tion is still possible. As the literature review discusses, Our findings offer important insight into the relation - cognitive skills and education have a reciprocal relation- ship between multiple measures of human capital and ship; that is, education credentials are strongly associ- earnings mobility. In part, the different measures connect ated with cognitive skills that, in turn, often increase to prior research that differentiates among what Estevez- at higher levels of education. Thus, it is likely that the Abe et al. (2001) term firm-specific skills (e.g., workplace reduction in the cognitive skill coefficient once a model tasks), industry-specific skills (e.g., VET education), and controls for education level (i.e., Models 3 and 4) is par- general skills (e.g., direct measures of cognitive skills). tially explained by variance in education that is associ- This and other typologies tend to focus on the specificity ated with cognitive skills earlier in the life course. Even Short‑term earnings mobility in the Canadian and German context: the role of cognitive skills Page 15 of 19 10 if the observed associations are causal, it is beyond the Table 6 Summary statistics scope of our study to fully unravel the mechanisms Canada (%) Germany (%) through which cognitive skills, education level, or other Numeracy: Level 0/1 12.38 9.88 measures of human capital affect earnings mobility. Numeracy: Level 2/3 72.14 70.85 A second limitation is that the analysis only contrasts Numeracy: Level 4/5 15.48 19.27 similarities and differences between the Canadian and Advanced math at work 36.53 46.59 German results and does not formally test differences Advanced reading at work 49.90 58.20 between contexts or sub-populations. We could for- High discretion 43.62 64.58 mally test governmental, policy, or educational differ - Education: High school diploma or less 30.47 53.37 ences cross-nationally through multi-level modeling if Education: VET PSE (below the BA level) 18.61 8.76 all countries participating in PIAAC (i.e., over 40 as of Education: non-VET PSE (below the BA 21.98 15.85 2020) included longitudinal earnings data. Because of level) the final sample size, the study does not assess how the Education: BA degree or above 28.94 22.01 findings differ by other sociodemographic characteris - Female 48.09 47.74 tics. Prior research demonstrates that returns to human Age: < 35 years 31.28 21.57 capital differ by gender and race (e.g., Hu et al. 2019) Age: 35–54 years 56.99 66.71 and thus future research must assess how the associa- Age: > 55 years 11.73 11.72 tion between mobility and human capital differs among Non-native language speaker 18.28 7.83 social groups. No job change between 12–16 68.53 77.84 Even with these limitations, our contribution has key Change working hours 12–16 (mean and 1.73 (11.46) 0.57 (9.65) strengths. It uses unique longitudinal data with direct SD) assessment tests that allow for an expansion of com- Firm size: < 10 ppl 17.7 22.49 monly used measures of human capital and furthers Firm size: 11–50 ppl 31.15 26.38 research on “the extent to which modern, knowledge- Firm size: 51–250 ppl 26.53 25.09 based labor markets reward skills” (Hanushek et al. Firm size: > 251 ppl 24.62 26.05 2015, 123). The study provides insight into how both Public sector employee 28.72 23.26 acquired and utilized human capital are associated with Part-time employee 9.53 20.49 earnings mobility, evidence that generates avenues for Labour market experience (mean and SD) 20.48 (11.62) 22.61 (10.05) future research and theoretical development. In par- More than one job in 2012 9.38 8.80 ticular, it is necessary to expand signaling theories that Living with partner in 2012 68.18 75.92 typically surround employer recognition of credentials Child under 6 in 2012 14.49 9.79 during the hiring process and consider the mechanisms Observations 1938 1320 behind why these same credentials are associated with Table shows percentages unless stated otherwise. Survey weights provided by earnings mobility, even when controlling for indi- the OECD are used vidual cognitive skills. While prior research indicates Reference group: little math/reading at work that perceived skills are related to work promotion and Reference group: low discretion retention (Furnham and Petrides 2006), it will also be Reference group: Native speaker necessary for future research to deepen insight into d Reference group: job change between 2012 and 2016 why cognitive skills and education level do and do not e Reference group: private sector have separate associations with earnings mobility in Reference group: full-time employee certain contexts. Our comparative case study approach Reference group: only one job in 2012 allows for the beginning of theory development as it Reference group: not living with partner in 2012 assesses the extent to which the findings are context Reference group: no child under 6 in household in 2012 dependent. Appendix A: Summary Statistics for Independent Variables See: Table 6. 10 Page 16 of 19 A. Pullman et al. Appendix B: Sensitivity Tests Relative change models with literacy instead of numeracy See Tables 7 and 8. Table 7 Relative change in Canadian monthly earnings percentiles, 2012 to 2016 (1) (2) (3) (4) ** ** Literacy 2.327 (0.710) 2.092 (0.703) 1.076 (0.738) 0.389 (0.683) Initial position a *** *** *** *** Earnings decile in 2012 − 2.450 (0.214) − 2.636 (0.249) − 2.822 (0.244) − 2.990 (0.332) Work characteristics in 2012 Advanced math at work 1.352 (1.170) 1.216 (1.155) 0.891 (1.084) b * Advanced reading at work 2.430 (1.265) 1.463 (1.195) 2.138 (1.043) High discretion − 0.183 (1.097) 0.123 (1.062) 0.677 (0.947) Education VET PSE (below the BA level) 1.433 (1.714) 1.678 (1.450) d *** ** non-VET PSE (below the BA level) 5.423 (1.593) 4.619 (1.469) d *** *** BA degree or abo ve 7.413 (1.517) 5.800 (1.628) Controls *** *** *** ** Constant 14.380 (1.288) 13.750 (1.259) 11.525 (1.409) 13.039 (4.039) R 0.12 0.13 0.15 0.35 Observations 1938 * ** *** Standard errors in parentheses; p < 0.05, p < 0.01, p < 0.001 2012 earnings decile as a continuous variable Reference group: little math/reading at work Reference group: low discretion Reference group: high-school diploma or less Table 8 Relative change in German monthly earnings percentiles, 2012 to 2016 (1) (2) (3) (4) *** *** ** Literacy 2.566 (0.574) 2.185 (0.575) 1.585 (0.594) 0.710 (0.556) Initial position a *** *** *** *** Earnings decile in 2012 − 1.802 (0.177) − 2.028 (0.221) − 2.246 (0.234) − 2.749 (0.286) Work characteristics in 2012 Advanced math at work 1.146 (1.115) 1.080 (1.103) 1.201 (0.927) b ** * ** Advanced reading at work 2.972 (1.013) 2.172 (1.014) 2.973 (0.874) c * High discretion 0.784 (0.904) 0.735 (0.913) 1.634 (0.814) Education VET PSE (below the BA level) 0.769 (1.733) 0.386 (1.405) d * * non-VET PSE (below the BA level) 3.673 (1.453) 3.518 (1.386) d *** ** BA degree or abo ve 5.190 (1.424) 4.026 (1.516) Controls x *** *** *** *** Constant 10.848 (1.059) 9.282 (1.113) 9.166 (1.132) 17.033 (3.404) R 0.10 0.11 0.12 0.42 Observations 1320 * ** *** Standard errors in parentheses; p < 0.05, p < 0.01, p < 0.001 2012 earnings decile as a continuous variable Reference group: little math/reading at work Reference group: low discretion Reference group: high-school diploma or less Short‑term earnings mobility in the Canadian and German context: the role of cognitive skills Page 17 of 19 10 Change in earnings models See Tables 9 and 10. Table 9 Change in monthly earnings between 2012 and 2016, Canada (1) (2) (3) (4) ** ** Numeracy 164.758 (51.020) 149.534 (51.434) 98.261 (52.779) 41.315 (49.442) Initial position a *** *** *** *** Earnings decile in 2012 − 70.573 (16.492) − 79.903 (18.362) − 90.075 (18.148) − 99.402 (24.029) Work characteristics in 2012 Advanced math at work 94.751 (85.715) 88.394 (86.093) 64.600 (83.474) Advanced reading at w ork 85.719 (84.010) 26.555 (81.798) 76.947 (72.825) High discretion 28.380 (75.708) 43.490 (74.343) 73.583 (67.010) Education VET PSE (below the BA level) 48.073 (127.489) 78.202 (108.759) non-VET PSE (below the BA level) d ** ** BA degree or abo ve 291.085 (96.151) 261.523 (94.451) Controls x *** *** *** *** Constant 962.432 (86.513) 922.661 (85.960) 806.987 (94.109) 1074.640 (278.466) R 0.02 0.03 0.04 0.25 Observations 1938 * ** *** Standard errors in parentheses; p < 0.05, p < 0.01, p < 0.001 2012 earnings decile as a continuous variable Reference group: little math/reading at work Reference group: low discretion Reference group: high-school diploma or less Table 10 Change in monthly earnings between 2012 and 2016, Germany (1) (2) (3) (4) *** *** ** Numeracy 197.478 (41.221) 176.658 (41.615) 146.669 (43.404) 47.131 (37.782) Initial position a *** *** *** *** Earnings decile in 2012 − 61.965 (14.195) − 75.886 (16.987) − 87.310 (17.531) − 109.993 (20.906) Work characteristics in 2012 Advanced math at work 75.909 (80.579) 73.791 (79.672) 83.195 (66.183) 75.909 (80.579) b ** * *** ** Advanced reading at work 192.579 (65.110) 145.322 (64.650) 179.622 (52.426) 192.579 (65.110) High discretion 24.618 (68.058) 25.302 (67.616) 81.918 (60.467) 24.618 (68.058) Education VET PSE (below the BA level) − 29.953 (107.275) − 44.950 (80.515) non-VET PSE (below the BA level) 132.630 (91.108) 129.414 (92.891) d * BA degree or above 298.565 (113.807) 231.408 (117.010) Controls x *** *** *** *** Constant 824.187 (76.815) 735.299 (75.775) 739.553 (76.598) 1115.776 (246.033) R 0.04 0.04 0.05 0.33 Observations 1320 * ** *** Standard errors in parentheses; p < 0.05, p < 0.01, p < 0.001 2012 earnings decile as a continuous variable Reference group: little math/reading at work Reference group: low discretion Reference group: high-school diploma or less 10 Page 18 of 19 A. Pullman et al. Abbreviations Institute for the Social Sciences, Survey Design and Methodology, P.O. Box 12 PIAAC : Programme for the International Assessment of Adult Competencies; 21 55, 68072 Mannheim, Germany. PIAAC-L: PIAAC-Longitudinal; LISA: Longitudinal and International Study of Adults; VET: Vocational education and training; PSE: Post-secondary education. Received: 5 May 2020 Accepted: 15 March 2021 Acknowledgements We gratefully acknowledge comments and helpful discussion with partici- pants of the research seminar “Scale development and evaluation” at GESIS. References Authors’ contributions Afonso, A., Schuknecht, L., Tanzi, V.: Income distribution determinants and pub- AP, BG, and CL contributed with the initial conceptualization of the study. lic spending efficiency. J. Econ. Ineq. 8(3), 367–389 (2010) AP prepared and analyzed the Canadian data. BG prepared and analyzed the Allmendinger, J., Hinz, T.: Mobilität und Lebensverlauf: Deutschland, Großbri- German data. AP conducted the literature research and wrote significant parts tannien und Schweden im Vergleich. In Hradil, S., Immerfall, S. (eds.) Die of the original manuscript. AP, BG, and CL contributed to the revision of the westeuropäischen Gesellschaften im Vergleich, pp. 247–285. VS Verlag für final manuscript. Sozialwissenschaften (1997) Altonji, J.G., Pierret, C.R.: Employer learning and statistical discrimination. Q. J. Authors’ information Econ. 116(1), 313–350 (2001) Ashley Pullman is a Senior Analyst at the Education Policy Research Initiative Antonczyk, D., DeLeire, T., Fitzenberger, B.: Polarization and rising wage in the Graduate School of Public and International Affairs at the University inequality: Comparing the U.S. and Germany. ZEW Discussion Paper No. of Ottawa. Her research focuses primarily on skill, education, and life course 10–015, Centre for European Economic Research, Mannheim (2012) transitions using Canadian microdata and administrative data. Aristei, D., Perugini, C.: The drivers of income mobility in Europe. Econ. Syst. 39(2), 197–224 (2015) Britta Gauly is a researcher at the Department of Survey Design and Methodol- Autor, D.H., Levy, F., Murnane, R.J.: The skill content of recent technological ogy at GESIS—Leibniz-Institute for the Social Sciences in Mannheim, Germany change: an empirical exploration. Q. J. Econ. 118(4), 1279–1333 (2003) and part of the German national project management for the Programme Bachmann, R., Bechara, P., Schaffner, S.: Wage inequality and wage mobility in of the International Assessment of Adult Competencies (PIAAC). Her main Europe. Rev. Income Wealth 62(1), 181–197 (2016) research interests comprise topics on adult education and the value of cogni- Bartels, C.: Top incomes in Germany, 1871–2014. J. Econ. Hist. 79(3), 669–707 tive skills in the labour market. (2019) Becker, G.S.: Investment in human capital: a theoretical analysis. J. Polit. Econ. Clemens Lechner is the head of the “Scale Development and Documenta- 70(5), 9–49 (1962) tion” team at the Department of Survey Design and Methodology at GESIS – Becker, R., Blossfeld, H.P.: Entry of men into the labour market in West Germany Leibniz-Institute for the Social Sciences in Mannheim, Germany. Since joining and their career mobility (1945–2008). J. Labour Market Res 50(1), GESIS in 2016, he has been involved as a principal and co-principal investiga- 113–130 (2017) tor in several major research projects on adults’ cognitive and socio-emotional Blau, F.D., Kahn, L.M.: Do cognitive test scores explain higher US wage inequal- skills development over the life span. ity? Rev. Econ. Stat. 87, 184–193 (2005) Broecke, S., Quintini, G., Vandeweyer, M.: Explaining international differences in Funding wage inequality: skills matter. Econ. Educ. Rev. 60, 112–124 (2017) This research has been financially supported by a travel grant offered by GESIS Chen, W.H.: Cross-national differences in income mobility: Evidence from – Leibniz Institute for the Social Sciences and The Social Sciences and Humani- Canada, The United States, Great Britain and Germany. Rev Income and ties Research Council of Canada Postdoctoral Scholarship. The opinions Wealth 55(1), 75–100 (2009) expressed in the article are those of the authors and do not necessarily reflect Choi, S.: How does job mobility affect inequality? Evidence from the South the views of the funding sources. Korean economic crisis. Soc. Forces 94(3), 1045–1073 (2016) Coban, M.: Wage mobility, wage inequality, and tasks: Empirical evidence from Availability of data and materials Germany, 1984–2014. Discussion Paper No.139, University of Würzburg, Both German and Canadian PIAAC microdata are not publicly available and Würzburg (2017) are accessible under restricted conditions. The authors accessed these data Connolly, H.C., Gottschalk, P.: Differences in wage growth by education level: through both the Research Data Centre PIAAC at GESIS– Leibniz Institute for Do less-educated workers gain less from work experience? IZA Discus- the Social Sciences and the Canadian Research Data Centre Network. As both sion Paper No. 2331, Institute for the Study of Labor, Bonn (2006) data sources are restricted for use, the authors do not have and cannot obtain Creedy, J., Gemmell, N.: Illustrating income mobility: New measures. Oxf. Econ. permission to share these data. Pap. 71(3), 733–755 (2019) Cunha, F., Heckman, J.J., Lochner, L., Masterov, D.V.: Interpreting the evidence on life cycle skill formation. In: Hanushek, E., Welch, F. (eds.) Handbook Declarations of the Economics of Education, pp. 697–812. North Holland, Amsterdam (2006) Ethics approval and consent to participate Desjardins, R., Warnke, A.J.: Ageing and skills: A review and analysis of skill gain All respondents gave informed consent prior to participation in the PIAAC and skill loss over the lifespan and over time. OECD Education Working study and its longitudinal components in Canada and Germany. As the follow- Papers No.72, OECD Publishing, Paris (2012) ing article is based on secondary data analysis, additional ethics approval is Dustmann, C., Ludsteck, J., Schönberg, U.: Revisiting the German wage struc- not necessary within the jurisdiction of this research. ture. Q. J. Econ. 124(2), 843–881 (2009) Ederer, P., Nedelkoska, L., Patt, A., Castellazzi, S.: What do employers pay for Consent for publication employees’ complex problem solving skills? Int. J. Lifelong Educ. 34(4), The authors consent to publication and are not aware of any reason why the 430–447 (2015) following study cannot be published. Engelhardt, L., Goldhammer, F., Baumert, J., Carstensen, C.H., Köller, O., Lüdtke, O., Schupp, J.: Separability of the PIAAC competence tests from general Competing interests cognitive ability (Manuscript in Preparation) The authors declare that they have no conflict of interest. Estevez-Abe, M., Iversen, T., Soskice, D.: Social protection and the formation of skills: A reinterpretation of the welfare state. In: Hall, PA, Soskice, D Author details (eds.) Varieties of capitalism: The institutional foundations of comparative Education Policy Research Initiative, Graduate School of Public and Interna- advantage, pp. 145–183. Oxford University Press, Oxford (2001) tional Affairs, Faculty of Social Sciences, University of Ottawa, 5004-120 Univer - Fields, G.S., Ok, E.A.: Measuring movement of incomes. Economica 66(264), sity, Social Sciences Building, Ottawa, ON K1N 6N5, Canada. GESIS—Leibniz 455–471 (1999) Short‑term earnings mobility in the Canadian and German context: the role of cognitive skills Page 19 of 19 10 Fields, G.S., Ok, E.A.: The meaning and measurement of income mobility. Liu, Y., Grusky, D.: The payoff to skill in the third industrial revolution. Am. J. Journal of Economic Theory 71(2), 349–377 (1996) Sociol. 118(5), 1330–1374 (2013) Fields, G.S.: Does income mobility equalize longer-term incomes? New meas- Mainert, J., Niepel, C., Murphy, K.R., Greiff, S.: The incremental contribution ures of an old concept. The Journal of Economic Inequality 8(4), 409–427 of complex problem-solving skills to the prediction of job level, job (2010) complexity, and salary. Journal of Business and Psychology, early release, Fitzenberger, B.: Wages and employment across skill groups: An analysis for 1–21 (2018) West Germany. Physica/Springer, Heidelberg (1999) Mane, F., Miravet, D.: Using the job requirements approach and matched Friedman, M.: Capitalism and freedom. The University of Chicago Press, employer-employee data to investigate the content of individuals’ Chicago (1962) human capital. J. Labour Market Res. 49(2), 133–155 (2016) Furnham, A., Petrides, K.V.: Deciding on promotions and redundancies: Pro- Manzoni, A., Härkönen, J., Mayer, K.A.: Moving on? A growth-curve analysis moting people by ability, experience, gender and motivation. J. Manag. of occupational attainment and career progression patterns in West Psychol. 21(1), 6–18 (2006) Germany. Soc. Forces 92(4), 1285–1312 (2014a) Garnero, A., Hijzen, A., Martin, S.: More unequal, but more mobile? Earnings McGrew, K.S.: CHC theory and the human cognitive abilities project: Standing inequality and mobility in OECD countries. Labour Econ. 56, 26–35 (2019) on the shoulders of the giants of psychometric intelligence research. Gini, C.: Measurement of inequality of income. Econ. J. 31(1), 124–126 (1921) Intelligence 37(1), 1–10 (2009) Goos, M., Manning, A.: Lousy and lovely jobs: The rising polarization of work in OECD: OECD Skill outlook 2013: First results from the survey of adults. OECD Britain. Rev. Econ. Stat. 89(1), 118–133 (2007) Publishing, Paris (2013a) Gottschalk, P., Moffitt, R.: The rising instability of US earnings. Journal of Eco - OECD: Technical report of the survey of adult skills (PIAAC). OECD Publishing, nomic Perspectives 23(4), 3–24 (2009) Paris (2013b) Grabka, M.M., Goebel, J.: Real income rose significantly between 1991 and Oh, H., Choi, Y.J.: Limited income mobility: empirical evidence from Korea. Soc. 2014 on average: First indication of return to increased income inequality. Indic. Res. 138(2), 665–687 (2018) DIW Economic Bulletin 7(5), 47–57 (2017) Ou, S.R., Reynolds, A.J.: Early determinants of postsecondary education Green, D.A., Riddell, C.W., St-Hilaire, F.: Income inequality: The Canadian story. participation and degree attainment: findings from an inner-city minority Institute for Research on Public Policy, Montreal (2017) cohort. Educ. Urban Soc. 46(4), 474–504 (2014) Green, F.: Employee involvement, technology and evolution in job skills: a task- Pavlopoulos, D., Muffels, R., Vermunt, J.K.: Wage mobility in Europe: a compara- based analysis. ILR Rev. 65(1), 36–67 (2012) tive analysis using restricted multinomial logit regression. Qual. Quant. Hall, P.A., Soskice, D.: Varieties of capitalism: The institutional foundations of 44(1), 115–129 (2010) comparative advantage. Oxford University Press, Oxford (2001) Pavlopoulos, D.: Wage mobility patterns in Europe. Dissertation, University of Hampf, F., Woessmann, L.: Vocational vs. general education and employment Tilburg (2007) over the life cycle: new evidence from PIAAC. CESifo Econ. Stud. 63(3), Pellizzari, M., Fichen, A.: A new measure of skill mismatch: theory and evidence 255–269 (2017) from PIAAC. IZA J. Labor Econ. 6(1), 1–30 (2017) Hanushek, E.A., Schwerdt, G., Wiederhold, S., Woessmann, L.: Returns to skills Peng, P., Kievit, R.A.: The development of academic achievement and cognitive around the world: evidence from PIAAC. Eur. Econ. Rev. 73, 103–130 abilities: a bidirectional perspective. Child Dev. Perspect. 14(1), 15–20 (2015) (2020) Hanushek, E.A., Woessmann, L.: The role of cognitive skills in economic devel- Pouliakas, K., Russo, G.: Heterogeneity of skill needs and job complexity: opment. J. Econ. Lit. 46(3), 607–668 (2008) Evidence from the OECD PIAAC survey. IZA Discussion Paper No. 9392, Heckman, J.J., Corbin, C.O.: Capabilities and skills. J. Hum. Dev. Capabilities Institute for the Study of Labor, Bonn (2015) 17(3), 342–359 (2016) Raferzeder, T., Winter-Ebmer, R.: Who is on the rise in Austria: Wage mobility Heckman, J.J., Lochner, L., Taber, C.: Explaining rising wage inequality: explora- and mobility risk. J. Econ. Ineq. 5(1), 39–51 (2007) tions with a dynamic general equilibrium model of labor earnings with Rauscher, E., Elliott, W.: Wealth as security: Growth curve analyses of household heterogeneous agents. Rev. Econ. Dyn. 1(1), 1–58 (1998) income and net worth during a recession. J. Fam. Econ. Issues 37(1), Hofacker, D., Blossfeld, H.P.: Globalization, uncertainty and its effects on early 29–41 (2016) family and employment lives. In: Blossfeld, H.P., Hofacker, D., Bertolini, Sakamoto, A., Tamborini, C.R., ChangHwan, K.: Long-term earnings differentials S. (eds.) Youth in globalised labour markets: Rising uncertainty and its between African American and white men by educational level. Popul. effects on early employment and family lives in Europe, pp. 9–38. Barbara Res. Policy Rev. 37(1), 91–116 (2018) Budrich, Opladen (2011) Schneider, J.W., McGrew, K.S.: The Cattell–Horn–Carroll theory of cognitive Hu, M., Daley, A., Warman, C.: Literacy, numeracy, technology skill, and labour abilities. In: Flanagan, D.P., McDonough, E.M. (eds.) Contemporary intel- market outcomes among Indigenous peoples in Canada. Can. Public lectual assessment: theories, tests, and issues, pp. 73–163. Guilford Press, Policy 45(1), 48–73 (2019) New York (2018) Hungerford, T., Solon, G.: Sheepskin effects in the returns to education. Rev. Schultz, T.W.: The value of the ability to deal with disequilibria. J. Econ. Lit. Econ. Stat. 69(1), 175–177 (1987) 13(3), 827–846 (1975) Hunter, A.A., Leiper, J.K.: On formal education, skills and earnings: the role of Shorrocks, A.: Income inequality and income mobility. J. Econ. Theor. 19(2), educational certificates in earnings determination. Can. J. Sociol. 18(1), 376–393 (1978) 21–42 (1993) Spence, M.: Job market signalling. Quart. J. Econ. 87(3), 355–374 (1974) Jäntti, M., Jenkins, S.P.: Income mobility. In: Atkinson, T., Bourguignon, F. (eds.) Spitz-Oener, A.: Technical change, job tasks, and rising educational demands: Handbook of income distribution, pp. 807–935. Elsevier, North Holland looking outside the wage structure. J. Law Econ. 24(2), 235–270 (2006) (2015) Tansel, A., Dalgıç, B., Güven, A.: Wage inequality and wage mobility in Turkey. Jovicic, S.: Wage inequality, skill inequality, and employment: evidence and Soc. Indic. Res. 142(1), 107–129 (2019) policy lessons from PIAAC. IZA J. Eur. Labor Stud. 5(21), 1–26 (2016) Wu, M.: The role of plausible values in large-scale surveys. Stud. Educ. Eval. Kirby, D.: Reviewing Canadian post-secondary education: Post-secondary 31(2–3), 114–128 (2005) education policy in post-industrial Canada. Can. J. Edu. Admin. Policy 65, 1–24 (2007) Publisher’s Note Klees, S.J.: Human capital and rates of return: Brilliant ideas or ideological dead Springer Nature remains neutral with regard to jurisdictional claims in pub- ends? Comp. Educ. Rev. 60(4), 644–672 (2016) lished maps and institutional affiliations. Kosteas, V.D.: Job level changes and wage growth. Int. J. Manpow. 30(3), 269–284 (2009) Leuven, E., Oosterbeek, H., Ophem, H.V.: Explaining international differences in male skill wage differentials by differences in demand and supply of skill. Econ. J. 114(495), 466–486 (2004) Lin, D., Lutter, R., Ruhm, C.J.: Cognitive performance and labour market out- comes. Labour Econ. 51, 121–135 (2018)
Journal for Labour Market Research – Springer Journals
Published: Apr 6, 2021
You can share this free article with as many people as you like with the url below! We hope you enjoy this feature!
Read and print from thousands of top scholarly journals.
Already have an account? Log in
Bookmark this article. You can see your Bookmarks on your DeepDyve Library.
To save an article, log in first, or sign up for a DeepDyve account if you don’t already have one.
Copy and paste the desired citation format or use the link below to download a file formatted for EndNote
Access the full text.
Sign up today, get DeepDyve free for 14 days.
All DeepDyve websites use cookies to improve your online experience. They were placed on your computer when you launched this website. You can change your cookie settings through your browser.