Measures


Intergenerational persistence

Intergenerational persistence is the coefficient from a regression of respondents’ years of schooling on the highest years of schooling of their parents. It measures the estimated impact of one additional year of schooling of parents on the years of schooling of respondents. A high intergenerational persistence is equivalent to a low level of relative mobility. When no years of schooling variable is available in the data, but educational categories are available, these categories have been transformed into a years of schooling variable using UNESCO guidelines and data.


Absolute mobility (parents not having tertiary)

Absolute mobility (conditional on parents not having tertiary)  measures the share of a cohort which obtains more education than the maximum level of education of their parents. Education is measured in five categories based on the International Standard Classification of Education (ISCED): less than primary (ISCED 0), primary (ISCED 1), lower secondary (ISCED 2), upper secondary or post-secondary non-tertiary (ISCED 3–4), and tertiary (ISCED 5–8). Respondents whose parents have attained tertiary education are excluded since the respondents are unable to improve upon this level.


Intergenerational persistence (non-linear coeff 1)

NL1 (b1) and NL2 (b2) are the coefficients from an expanded model of IGP that includes parents’ years of schooling squared, such that child_educ=b0+b1*parent_educ+b2*parent_educ^2. Higher values of NL1 (b1) indicate greater intergenerational persistence and, hence, lower relative mobility in education. NL2 (b2) determines the non-linearity of respondents’ years of schooling on parents’ years of schooling.


Intergenerational persistence (non-linear coeff 2)

NL1 (b1) and NL2 (b2) are the coefficients from an expanded model of IGP that includes parents’ years of schooling squared, such that child_educ=b0+b1*parent_educ+b2*parent_educ^2. Higher values of NL1 (b1) indicate greater intergenerational persistence and, hence, lower relative mobility in education. NL2 (b2) determines the non-linearity of respondents’ years of schooling on parents’ years of schooling.


Correlation

Pearson's correlation coefficient between parents’ and respondents’ years of education. Higher values of the Pearson's correlation coefficient indicate greater intergenerational persistence and, hence, lower relative mobility in education.


Absolute mobility (weakly)

Absolute mobility (weakly) measures the share of a cohort which has more education than the maximum level of education of his/her parents, or at least the same level as his/her parent if they have completed a tertiary degree. Education is measured in five categories based on the International Standard Classification of Education (ISCED): less than primary (ISCED 0), primary (ISCED 1), lower secondary (ISCED 2), upper secondary or post-secondary non-tertiary (ISCED 3–4), and tertiary (ISCED 5–8).


Intergenerational privilege

Intergenerational privilege is the share of respondents who reaches the top quartile in educational attainment of their generation among all individuals who are born to parents in the top quartile of educational attainment of their generation.


Movement from bottom to 1st quartile

The movement from the bottom variables measure the share of individuals who reaches different quartiles of educational attainment of their generation among all individuals who are born to parents in the bottom half of their respective generations. In particular, BHQ1, BHQ2, BHQ3 and BHQ4 indicate the probability that an individual from the bottom half moves to the lowest (first), second, third or top (fourth) quartile, respectively.


Movement from bottom to 2nd quartile

The movement from the bottom variables measure the share of individuals who reaches different quartiles of educational attainment of their generation among all individuals who are born to parents in the bottom half of their respective generations. In particular, BHQ1, BHQ2, BHQ3 and BHQ4 indicate the probability that an individual from the bottom half moves to the lowest (first), second, third or top (fourth) quartile, respectively.


Movement from bottom to 3rd quartile

The movement from the bottom variables measure the share of individuals who reaches different quartiles of educational attainment of their generation among all individuals who are born to parents in the bottom half of their respective generations. In particular, BHQ1, BHQ2, BHQ3 and BHQ4 indicate the probability that an individual from the bottom half moves to the lowest (first), second, third or top (fourth) quartile, respectively.


Movement from bottom to 4th quartile

The movement from the bottom variables measure the share of individuals who reaches different quartiles of educational attainment of their generation among all individuals who are born to parents in the bottom half of their respective generations. In particular, BHQ1, BHQ2, BHQ3 and BHQ4 indicate the probability that an individual from the bottom half moves to the lowest (first), second, third or top (fourth) quartile, respectively.


Difference in years of schooling

The average difference in years of schooling between parents and their children (respondents). To account for the ceiling effect in this measure, only respondents whose parents are in the bottom half of the national distribution are used.


Intergenerational privilege (Asher, et al. 2017)

Intergenerational privilege (Asher, Novosad, and Rafkin (2017)) is a variant of the intergenerational privilege measure (IGP), where the transition matrix accounts for ties in the educational distribution by assuming an underlying continuous distribution of human capital.


Three-generational persistence (parental effect)

Three generations IGP are the regression coefficients from a regression of respondents’ years of schooling on both parents’ and grandparents’ years of schooling (child_educ=b0+b1*parent_educ+b2*grandparent_educ). ThreeGen_IGEp1 (b1) captures the effect from parents, and ThreeGen_IGEg1 (b2) captures the effect from grandparents. This is available only for a subset of countries where grandparental educational information can be extracted.


Three-generational persistence (grandpar. effect)

Three generations IGP are the regression coefficients from a regression of respondents’ years of schooling on both parents’ and grandparents’ years of schooling (child_educ=b0+b1*parent_educ+b2*grandparent_educ). ThreeGen_IGEp1 (b1) captures the effect from parents, and ThreeGen_IGEg1 (b2) captures the effect from grandparents. This is available only for a subset of countries where grandparental educational information can be extracted.


Relative intergenerational mobility in income

Measures the intergenerational income elasticity. Some estimates are obtained from the Equalchances (2018) database. Other estimates are obtained from the surveys in the GDIM using the following three-step procedure: (1) estimate an income equation from an older sample that is representative of the current population of parents (when they were younger, that is pseudo-parents), (2) use the estimated model coefficients (that is, return to education and experience) to predict parental earnings at the reference age using the retrospective data on parental age and education as predictors (explanatory variables), and (3) regress child earnings at the reference age on predicted parental earnings at the reference age. The resulting regression coefficient from step 3 is the estimate used.


Parental characteristics’ effect on persistence S1

S1, S2 and S3 are derived from decomposing the intergenerational elasticity of income. S1 is the share of income persistence that is the effect of parental education on offspring’s income via offspring’s education; S2 is the share of income persistence that is the effect of parental education on the determinants of offspring’s income that are independent of education; and S3 is the share of income persistence that is the effect of parental characteristics (other than education) that are related to income, on offspring’s income.


Parental characteristics’ effect on persistence S2

S1, S2 and S3 are derived from decomposing the intergenerational elasticity of income. S1 is the share of income persistence that is the effect of parental education on offspring’s income via offspring’s education; S2 is the share of income persistence that is the effect of parental education on the determinants of offspring’s income that are independent of education; and S3 is the share of income persistence that is the effect of parental characteristics (other than education) that are related to income, on offspring’s income.


Parental characteristics’ effect on persistence S3

S1, S2 and S3 are derived from decomposing the intergenerational elasticity of income. S1 is the share of income persistence that is the effect of parental education on offspring’s income via offspring’s education; S2 is the share of income persistence that is the effect of parental education on the determinants of offspring’s income that are independent of education; and S3 is the share of income persistence that is the effect of parental characteristics (other than education) that are related to income, on offspring’s income.