Despite a plethora of normative discussions on gender equality as well as empirical studies on gender discrimination and gender effects in various settings, there is a paucity of large-scale empirical studies on son preference by ordinary people in asset distribution. Using an idiosyncratic data set on more than 1800 notarized or authenticated wills in Taiwan, this Article investigates whether testators show son preference in distributing estates in wills, and if so, what the driving factors are. It finds that son preferences exist in no more than 29% of the studied wills. Moreover, no matter whether son preference is broadly or narrowly defned, and no matter whether the sample is limited to land distribution or not, the pattern is consistent. Aboriginal people exhibit less son preference, as a few ethnic groups are matrilineal. Female testators do not tend to favor sons. In wills distributing more valuable estate and those distributing land, son preference is more pronounced. Notarized wills tend to contain son-preferring provisions, likely because notarized wills, due to their formal validity, have usually been upheld if disputes arise, despite sonpreferring provisions often violating the mandatory share law. Thus, when preparing a will that favors their sons, testators elect to notarize their wills. Finally, the strategic bequest theory explains the testator decisions in some wills, while the altruism theory has little explanatory power.
Son Preference: An Empirical Study of Estate Distribution in Wills, (Vol. 57.2 Spring 2025)
15 Jul 2025