Marriage outcomes of displaced women

Frances Lu, Sameem Siddiqui, and Prashant Bharadwaj

Journal of Development Economics, Volume 152


This paper examines the marriage outcomes of displaced women. Early marriage has been shown in the literature to have detrimental effects on women by lowering educational attainment and increasing fertility rates.

The authors document marriage patterns among displaced and non-displaced women using 12 representative survey datasets from seven countries (Armenia, Cambodia, Colombia, India, Iraq, Kyrgyz Republic, and Nepal). The data shows that, across countries and over time, young displaced women are more likely to marry early compared with young women who are not displaced, while displaced and non-displaced men, regardless of age, appear to marry at similar rates.

The authors then focus their analysis on the particular case of Muslim refugees who were forcibly displaced from Indian Punjab to Pakistani Punjab following the partition of India and formation of Pakistan in 1947. Using a ‘difference-in-differences’ approach, the authors compare the age at which young displaced women married relative to two groups: (1) native-born women in the same age cohort; and (2) older displaced women (aged 30 to 32 at partition) who were married before partition. The analysis draws on data from Pakistan’s 1973 Housing, Economic, Demographic Characteristics (HED).

Main findings:

  • Compared to native-born women of the same age group, displaced women who were adolescents (aged 13 to 17) at the time of partition married 0.28 years earlier, were 3.8 percentage points more likely to marry before the age of 18, and were 5.3 percentage points more likely to be married during partition (1947–1949). For displaced women who were very young children (aged 1 to 5) at the time of partition, there were only small negative and statistically insignificant effects on age of marriage and likelihood of marriage before age 18, when compared with native born women of the same age cohort.
  • Displaced women who were adolescents (aged 13 to 17) at the time of partition were less likely than native-born women of the same age to continue their education and had higher fertility. However, displaced women who were much younger (aged 1 to 5) at partition were 3.5 percentage points more likely to be literate, 3.1 percentage points more likely to have completed primary school, and had more surviving children than native born women of the same age.
  • Displaced women don’t appear to have “lower spousal quality” than native-born women. In comparison to native-born women who were ages 30 to 32 at partition, young children at the time of partition appear to have married more educated and younger men, while for adolescent women at the time of partition there were no effects on the characteristics of spouses.

The authors conclude that displaced women tend to marry earlier than non-displaced women, but that these impacts depend on the timing of displacement during their lifetime. The authors suggest several potential mechanisms that might explain these results: (a) the threat of gender-based violence towards displaced women may have induced families to marry off daughters in order to afford them the protection of a male spouse; (b) the negative income and wealth shock precipitated by displacement may have induced families to marry off daughters to alleviate financial pressures; or (c) the displaced population in Pakistan had a higher share of males than the non-displaced population, which would have differentially increased the demand for women among the displaced population, leading to earlier marriage of displaced women.