Issue, No.9 (March 2019)
Extending educational attainment variables at LIS: On the importance of analyzing returns to education based on detailed education categories and years of education
Comparative studies of educational inequalities usually face a unique dilemma regarding the measurement of educational levels: whether to use a more precise, but less comparable, measure of educational levels or a rather crude, but more cross-nationally comparable measure. Other factors shaping inequality, such as income, wealth and even occupations are relatively comparable across periods and over countries using conventional standardization techniques (PPP, ISCO, ISEI, etc.). However, concerning education, especially outside the scope of the Bologna process, standardization may also represent a problem: the same nominal educational level might mean different things over time and between countries – in terms of duration required to earn this particular level of education, its prestige and its relevance for the labor market.
In comparative research, mostly three strategies are used for measuring and comparing the highest level of education: 1) years of schooling (in full-time equivalents), 2) consensual measurements of levels of education and 3) highly standardized measures such as the International Standard Classification of Education (ISCED) (UNESCO 2012, Schneider 2013). Years of schooling is easily measureable, but suffers from considerable measurement error. Problematic is, for instance, the reporting of years spent in full-time education, where the repetition of classes is included (e.g., ESS), or exclusion of vocational training in the count (e.g., ISSP). Moreover, especially in tracked educational systems, the duration and the level of education is not necessarily associated – or in other words, does not tell much about the stratification of a society. The forthcoming release of the updated LIS Database will introduce a standardized measure of years of education. This measure avoids some of these imprecisions by converting the country-specific measure of the highest educational achievement into years normally required to obtain these educational levels.
Despite several sophisticated solutions which try to harmonize educational levels, scholars often opt for a crude but consensual measurement of education, based on an ad-hoc harmonization of country-specific measures or a reduction of an international standardized scheme such as CASMIN or ISCED. Likewise, a three-category measurement was introduced in the LIS Database1 allowing comparisons of low/medium/high education over the LIS countries, often over several decades (England, Gornick & Shafer 2012, Bar-Haim et al. 2018, Bar-Haim, Chauvel & Hartung 2019).
Unfortunately, basic measurements do not allow to differentiate between crucially different educational certificates as well as tracks within educational levels, a fact that harms the ability to produce insightful research on education and inequality. A third measure, especially more recent studies have therefore used, is the ISCED, which was explicitly created by the UNESCO for the provision of harmonized international statistics on educational levels, i.e. school leaving certificates. The ISCED takes into consideration comparability issues in both highest (graduate) and lowest (preschool) levels of education. ISCED accounts not only for educational systems in high-income countries but also for middle and low-income countries. Several studies suggest that it performs better than other common educational categorization systems, at least in the European context. Most importantly, even in its reduced (1-digit) form, it distinguishes between nine levels and allows thus a much more refined investigation of educational inequalities.
The differentiation of educational degrees within crude educational levels, particularly within the tertiary level, has become much more important in times of the global educational expansion (Schofer & Meyer 2005, Bar-Haim & Shavit 2013). Since secondary education became almost universal in many countries, we are facing saturation of education or “educational inflation”. As a result, the differentiation between undergraduate vs. graduate degrees and even within the undergraduate degrees, e.g. BA (or equivalent), MA and PhD degrees, has stronger sociological and economic implications. For example, Posselt and Grodsky (2017) presented data for the U.S. that suggest that the wage gap between persons with a BA degree and persons with a high school diploma increased by 6% between 2000 and 2013, while the wage gap between persons with a graduate degree and those with a high school diploma increased by more than 17%. They also found that the importance of parental education in the U.S. remained stable for achieving undergraduate degrees between the 1970s and the 1990s, but increased significantly for obtaining PhD degrees.
Also for the U.S., Torche (2011) shows that the association between parental background and individual socio-economic outcomes (income and occupational standing) is strongly significant only among graduate degree owners, contrary to other educational levels, where this association does not exist. Despite the vast evidence of the importance of more detailed categories of educational levels, the difficulties to compare these levels prevented full-scale comparative studies to incorporate such a detailed scheme. The lack of a detailed comparative classification of education in many of the cross national and time-series data sets is a major setback. Therefore, the incorporation of the more detailed 1-digit ISCED 11 (UNESCO 2012) to most of the LIS datasets (variable educlev) is a major contribution for the comparative study of education-driven inequality2.
The potential contribution of incorporating a more detailed educational measure based on the ISCED in the LIS Database can be demonstrated by the example of trends in returns to education. In the past, studies of economic returns to education who employed LIS data had to focus on income differences between less than secondary, non-tertiary and tertiary education. However, as noted above, much of the change over time in returns to education can be found in the differences between Bachelor and Master degrees. In order to demonstrate this, we analyzed the income (unadjusted) returns to educational levels for the U.S. in 1991 and 2016. First, we used the less detailed educational variable (educ) and then we compared the results to those obtained using the forthcoming, more detailed ISCED variable (educlev). The results are shown in Fig. 1., where the grey (blue) bars represent the returns to education in 1991 (2016). The most important difference are the returns to higher levels of education. The dramatically increasing returns to BA and higher degrees based on the ISCED categories (right side of the figure), is masked to large extent when using the less detailed, three-category variable (left side of the table). This is due to the very small contribution of short-cycle tertiary education, which in the U.S. mostly refer to community colleges, to the increase in the returns to education in this broad category. These findings are in line with the literature that found substantial differences between short- and full-cycle tertiary education in the U.S. .
The descriptive results presented here, albeit preliminary, demonstrate the possible contribution of the new variable to the LIS data. Using a more refined, yet comparable measurement of educational levels, the LIS Database will allow to increase our knowledge on the role of education in inequality and stratification systems.
1 In addition to the more detailed country-specific measurements (LIS variable educ_c).
2 The forthcoming LIS variable educlev also incorporates the category ‘no education’ within the less than primary education category.
References
| Bar-Haim, E., & Shavit, Y. (2013). Expansion and inequality of educational opportunity: A comparative study. Research in Social Stratification and Mobility, 31, 22-31. |
| Bar-Haim, E., Chauvel, L., & Hartung, A. (2019). More necessary and less sufficient: an age-period-cohort approach to overeducation from a comparative perspective. Higher Education, https://doi.org/10.1007/s10734-018-0353-z. |
| Bar-Haim, Eyal, Chauvel, Louis, Gornick, Janet & Hartung, Anne (2018) Closing or persisting gender wage gap? A cohort analysis of education and earnings in the US and Europe. LIS Working Paper Series, No. 737. |
| England, P., Gornick, J., & Shafer, E. F. (2012). Women’s employment, education, and the gender gap in 17 countries. Monthly Lab. Rev., 135, 3. |
| Posselt, J. R., & Grodsky, E. (2017). Graduate education and social stratification. Annual review of sociology, 43, 353-378. |
| Schneider, S. L. (2013). The International Standard Classification of Education 2011, in Gunn Elisabeth Birkelund (ed.) Class and Stratification Analysis (Comparative Social Research, Volume 30) Emerald Group Publishing, 365 – 379. |
| Schofer, E., & Meyer, J. W. (2005). The worldwide expansion of higher education in the twentieth century. American sociological review, 70(6), 898-920. |
| Torche, F. (2011). Is a college degree still the great equalizer? Intergenerational mobility across levels of schooling in the United States. American Journal of Sociology, 117(3), 763-807. |
| UNESCO Institute for Statistics. (2012). International standard classification of education: ISCED 2011. Montreal: UNESCO Institute for Statistics. |