Skip to main content
Log in

Children’s Living Arrangements and On-time Progression Through School in Latin America and the Caribbean

  • Original Paper
  • Published:
Journal of Family and Economic Issues Aims and scope Submit manuscript

Abstract

We used data from Demographic and Health Surveys and the Mexican Family Life Survey to test how children’s living arrangements were related to their progress through school in countries comprising three-quarters of the population in Latin America and the Caribbean. Our results indicated that family instability presents a challenge for educational progress: Large proportions (23–60 %) lived apart from at least one biological parent, and children in stepfamilies did not show better educational progress than children living with single parents. In some countries, living with other men was associated with worse educational outcomes, and the occasional advantage associated with living with other women was modest. Other adults in the household did not appear to buffer negative effects associated with parental absence.

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.

Fig. 1
Fig. 2

Similar content being viewed by others

Notes

  1. Children living in female-headed households in sub-Saharan Africa had better educational outcomes (Lloyd and Blanc 1996; Townsend et al. 2002), but the Townsend study nonetheless found that coresidence with both parents was superior to living with one or none. This is part of a growing body of work showing that female headship is a poor proxy for single parenthood (see also Clark and Hamplová 2013).

  2. In the pooled sample, the language test score differences were not significant and the math test score differences were modest but significant. The lower rates of grade repetition among children from two-parent homes were statistically significant and of more important magnitude.

  3. Stacer and Perruci’s (2013) analysis of family structure and school involvement did not distinguish between two-biological-parent families and other two-parent families, and they did not find greater involvement at school among two-parent families as a whole.

  4. Retrieved from http://dhsprogram.com/data/available-datasets.cfm.

  5. Retrieved from http://www.ennvih-mxfls.org/english/ennhiv-3.html.

  6. The 1996 Brazil DHS was the oldest data in our sample and was of particular concern because the economy has changed so rapidly in recent years. Although most censuses do not distinguish between children and stepchildren of the household head, the 2010 Brazilian census was one of the few censuses that did. Thus a similar analysis to what we conducted here was possible for the subsample of children/stepchildren of the household head. The results were very similar to an analysis of the 1996 Brazil DHS limited to children/stepchildren of the household head. This eased concern that using older data may have been problematic.

  7. The DHS included questions in the individual woman’s interviews that identified whether the woman was in union as well as whether her partner was in the household. We did not, however, want to utilize that information since it would have required limiting the sample to children living with interviewed mothers, consequently eliminating children living with only their biological father as well as children living with neither parent. Our constructed stepparent variable was of higher quality than the “probable stepparent” variable used in harmonized census data from the International Public Use Microdata Series (IPUMS) because we knew whether both biological parents were in the household: the IPUMS methodology only captures stepparents whose age makes them improbable biological parents, while our methodology included biological parents’ partners when the other biological parent was absent.

  8. 1 = poor floor, poor drinking water, and poor toilet; 2 = 2 of the following (poor floor, poor drinking water, and poor toilet); 3 = 1 of the following (poor floor, poor drinking water, and poor toilet); 4 = 0 or 1 of the following (poor floor, poor drinking water, and poor toilet) and a radio; 5 = 0 or 1 of the following (poor floor, poor drinking water, and poor toilet) and electricity; 6 = 0 or 1 of the following (poor floor, poor drinking water, and poor toilet) and a television; 7 = 0 or 1 of the following (poor floor, poor drinking water, and poor toilet) and a refrigerator; and 8 = 0 or 1 of the following (poor floor, poor drinking water, and poor toilet) and a car. Some countries have a DHS-provided wealth index that divides households into wealth quintiles within the country (relative wealth). In no case where the Giroux wealth index could be compared to the relative wealth measure was the statistical significance of living arrangements variables affected by the choice of wealth control. Thus we used the Giroux index rather than the DHS-provided relative wealth index to retain the maximum number of countries and comparability across diverse settings in the pooled sample.

  9. 2015 population data were from the World Population data sheet (PRB 2015). We retained all of the observations for Mexico rather than Brail (the largest country in the region) because there would not have been enough observations for a proportionate sample from Mexico. Brazil had 7034 observations, Mexico 4368, Colombia 1685, Peru 1073, Guatemala 557, Haiti 375, Dominican Republic 361, Bolivia 361, Honduras 285, Nicaragua 217, and Guyana 24. Some regions, particularly in Guyana, were dropped due to lack of variation in on-time progression within some regions in the limited sample.

  10. The reference category in Table 4 is two biological parents, but we also formally tested whether the apparent differences between absent-partner and single-parent homes were statistically significant.

  11. The effects of parental education and household assets were significant in every country; the urban advantage was significant in six countries.

  12. This female advantage in education in Latin America and the Caribbean has been documented before (e.g., Creighton and Park 2010; Grant and Behrman 2010; Knodel and Jones 1996; Willms and Somer 2001); having found it here indicated that although the DHS does not specialize in collecting education data, the data were nonetheless of high quality.

References

Download references

Author information

Authors and Affiliations

Authors

Corresponding author

Correspondence to Laurie F. DeRose.

Ethics declarations

Ethical Approval

All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards. All of our data come from secondary public data sources.

Rights and permissions

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

DeRose, L.F., Huarcaya, G., Salazar-Arango, A. et al. Children’s Living Arrangements and On-time Progression Through School in Latin America and the Caribbean. J Fam Econ Iss 38, 184–203 (2017). https://doi.org/10.1007/s10834-016-9502-7

Download citation

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1007/s10834-016-9502-7

Keywords

Navigation