Skip to main content

Advertisement

Log in

Measuring educational inequalities: a method and an application to Albania

  • Original Paper
  • Published:
Journal of Population Economics Aims and scope Submit manuscript

An Erratum to this article was published on 10 July 2010

Abstract

In this paper, we investigate whether educational inequalities stem from differences between families or within families. In a poor economy, schooling is costly for parents, and education is likely to be unequally distributed among siblings. Drawing on discrete ordered choice models, we present a simple method to estimate the between and within components of both the explained and unexplained variance in education. For our empirical analysis, we use the Living Standard Measurement Study survey conducted in 2002 in Albania. We explain about 40% of the total variance and find that inequalities in education are mainly due to differences between families. Differences within families are smaller and are far less easily explained.

This is a preview of subscription content, log in via an institution to check access.

Access this article

Price excludes VAT (USA)
Tax calculation will be finalised during checkout.

Instant access to the full article PDF.

Institutional subscriptions

Fig. 1
Fig. 2
Fig. 3
Fig. 4

Similar content being viewed by others

Notes

  1. Behrman et al. (1995) further consider how parents allocate human capital among their children. In a pioneering piece of work, Sheshinski and Weiss (1982) provide a theoretical analysis of inequality within and between families.

  2. It is also possible that years of schooling are measured with error. A simple way to check this is to see whether the degree awarded is consistent with the level of schooling attained. For evidence on measurement error in years of schooling, see Ashenfelter and Krueger (1994).

  3. A more equal distribution of children’s education is expected over time, which is found in the context of Taiwan (Yang and Zhu 2003).

  4. One limitation of this work is that only school attendance is taken into account, whilst birth order may have a greater impact when considering years of education.

  5. However, no significant effects are found for the older and younger siblings of the opposite gender.

  6. For further information, see http://www.worldbank.org/lsms/. The data are available online.

  7. Community information was collected from interviews with individuals considered to have superior information about each module within a community, usually an elected or appointed community leader.

  8. We assume that children aged over 16 have either already finished primary school or have dropped out and will never finish. Given that primary school lasts 8 years, it may be that some children over 16 are still in primary school, for instance in the case of multiple grade repetitions. For children living at home and enrolled at school, we find that only nine (resp. 1) children aged 17 (resp. 18) are still in primary school.

  9. In our analysis, we consider a sample of children with an average age of 30 years. We use the term “children” in this paper because our focus is on sibships, and we merge the characteristics of these different children with those of their parents. Clearly, we do not account for the schooling attainment of young children in 2002.

  10. Children born after 1975 were 15 years old during the 1990s, so they were probably at primary school during the chaos in the educational system.

  11. We will see, however, that our estimates are not biased by this second effect.

  12. This proportion is less than half as large when the parents live in a rural area.

  13. More educated parents certainly have higher incomes, and it will be easier for richer parents to finance the cost of education. Unfortunately, we only have information on the current level of income in the data. We choose to exclude this covariate, as household income may be caused by children’s education, and the income measure is not related to the period when investment in education takes place.

  14. Ideally, the regression should include a measure of the parent’s socioeconomic position. Unfortunately, this information is only partially available in the LSMS questionnaire. In the labour module, we know whether the respondent has worked during the past 12 months and on occasion occupation in the last job, but no information is provided for those who are no longer working.

  15. See in particular the detailed discussion in Butcher and Case (1994) and Garg and Morduch (1998).

  16. A simple likelihood ratio test shows that the specification with interaction effects is preferred to the model with no interaction terms.

  17. As shown in Table 6, the estimates on the other variables are little changed from those discussed previously.

  18. Nevertheless, with additional interaction effects, we do not find that educational differences in the allocation of resources vary with the gender of the child.

  19. Specifically, the computation of the random effects ordered probit model is carried out via a program discussed in Frechette (2001), which makes use of analytical first derivatives.

  20. As we account for younger children with the censored ordered probit model with random effects, we also estimate a model with more detailed information on birth cohorts for the children (with annual dummies for the more recent year). This specification does not influence the values of the other estimates.

References

  • Altonji JG, Dunn TJ (1996) The effects of family characteristics on the return to education. Rev Econ Stat 77:692–704

    Article  Google Scholar 

  • Ashenfelter O, Krueger AB (1994) Estimates of the economic return to schooling from a new sample of twins. Am Econ Rev 84:1157–1174

    Google Scholar 

  • Basu K (1999) Child labor: cause, consequence and cure. J Econ Lit 37:1083–1119

    Google Scholar 

  • Becker GS, Tomes N (1986) Human capital and the rise and fall of families. J Labor Econ 4:S1–S39

    Article  Google Scholar 

  • Berhman J (1997) Intrahousehold distribution and the family. In: Rosenzweig MR, Stark O (eds) Handbook of population and family economics. Elsevier, North-Holland, pp 125–187

    Google Scholar 

  • Behrman J, Taubman P (1986) Birth order, schooling, and earnings. J Labor Econ 4:S121–S145

    Article  Google Scholar 

  • Behrman J, Pollak RA, Taubman P (1995) From parent to child. Intrahousehold allocations and intergenerational relations in the United States. University of Chicago Press, Chicago

    Google Scholar 

  • Bommier A, Lambert S (2004) Human capital investments and family composition. Appl Econ Lett 11:193–196

    Article  Google Scholar 

  • Butcher KF, Case A (1994) The effect of sibling sex composition on women’s education and earnings. Q J Econ 109:531–563

    Article  Google Scholar 

  • Butler LS, Moffitt R (1982) A computationally efficient quadrature procedure for the one factor multinomial probit model. Econometrica 50:761–764

    Article  Google Scholar 

  • Ejrnaes M, Pörtner C (2004) Birth order and the intrahousehold allocation of time and education. Rev Econ Stat 86:1008–1019

    Article  Google Scholar 

  • Emerson PM, Portela Souza A (2008) Birth order, child labor and school attendance in Brazil. World Dev (forthcoming)

  • Frechette GR (2001) Random effects ordered Probit. Stata Tech Bull 159:23–27

    Google Scholar 

  • Garg A, Morduch J (1998) Sibling rivalry and the gender gap: evidence from child health outcomes in Ghana. J Popul Econ 11:471–493

    Article  Google Scholar 

  • Gavaria A (2002) Intergenerational mobility, sibling inequality and borrowing constraints. Econ Educ Rev 21:331–340

    Article  Google Scholar 

  • Glewwe P (2002) Schools and skills in developing countries: education policies and socioeconomic outcomes. J Econ Lit 40:436–482

    Article  Google Scholar 

  • Greenhalgh S (1985) Sexual stratification: the other side of growth with equity in East Asia. Popul Dev Rev 11:265–314

    Article  Google Scholar 

  • Gross M, Glewwe P (2000) Designing household survey questionnaires for developing countries: lessons from 15 years of the living standards development study. Oxford University Press, Oxford

    Google Scholar 

  • Hauser RM, Daphne Kuo HH (1998) Does the gender composition of sibships affect women’s educational attainment? J Hum Resour 33:644–657

    Article  Google Scholar 

  • INSTAT (2003) Albania living standard measurement survey 2002. Basic information document. Mimeo, World Bank

  • Jamison DT, Lookheed ME (1987) Participation in schooling: determinants and learning outcomes in Nepal. Econ Dev Cult Change 35:279–306

    Article  Google Scholar 

  • Kaestner R (1997) Are brothers really better? Sibling sex composition and educational achievement revisited. J Hum Resour 32:250–284

    Article  Google Scholar 

  • Lillard L, Willis RJ (1994) Intergenerational educational mobility. Effects of family and state in Malaysia. J Hum Resour 29:1126–1166

    Article  Google Scholar 

  • Morduch J (2000) Sibling rivalry in Africa. Am Econ Rev 90:405–409

    Google Scholar 

  • Ono H (2004) Are sons and daughters substitutable? Allocation of family resources in contemporary Japan. J Jpn Int Econ 18:143–160

    Article  Google Scholar 

  • Parish WL, Willis RJ (1993) Daughters, education, and family budgets: Taiwan experiences. J Hum Resour 28:863–898

    Article  Google Scholar 

  • Sheshinski E, Weiss Y (1982) Inequality within and between families. J Polit Econ 90:105–127

    Article  Google Scholar 

  • Sudha S (1997) Family size, sex composition and children’ education: ethnic differentials over development in Peninsular Malaysia. Popul Stud 51:139–151

    Article  Google Scholar 

  • Thomas D (1994) Like father, like son: like mother, like daughter. Parental resources and child height. J Hum Resour 29:950–988

    Article  Google Scholar 

  • Thomas D, Schoeni RF, Strauss J (1996) Parental investments in schooling: the roles of gender and resources in urban Brazil. Mimeo, Rand Labor and Population Program

  • Yang DT, Zhu X (2003) Economic structural change and family investments in children. Mimeo, Virginia Polytechnic Institute

Download references

Acknowledgements

We are indebted to two anonymous referees, the editor Christian Dustmann and Paul Glewwe, for very helpful comments and suggestions on a previous draft. We express our gratitude to Andrew Clark for his careful reading of the paper. We also would like to thank Marc Gurgand, Francis Kramarz and seminar participants at the INED Education conference, the 1st Minnesota International Economic Development Conference, the 22nd Journées de Microéconomie Appliquée, the 19th Annual Conference of the European Society for Population Economics and the 1st Meeting of the Society for the Study of Economic Inequality. Any remaining errors are ours.

Author information

Authors and Affiliations

Authors

Corresponding author

Correspondence to François-Charles Wolff.

Additional information

Responsible editor: Christian Dustmann

An erratum to this article can be found at http://dx.doi.org/10.1007/s00148-010-0327-7

Rights and permissions

Reprints and permissions

About this article

Cite this article

Picard, N., Wolff, FC. Measuring educational inequalities: a method and an application to Albania. J Popul Econ 23, 989–1023 (2010). https://doi.org/10.1007/s00148-008-0201-z

Download citation

  • Received:

  • Accepted:

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1007/s00148-008-0201-z

Keywords

JEL Classification

Navigation