Tracking individuals and households: longitudinal features of Danseong household register data

Abstract

Household registers, which were compiled every three years and contained a variety of demographic information, are invaluable sources with which to investigate family and demographic behaviors of historical populations in Korea, especially during the period of the eighteenth and nineteenth centuries. Linking three sets of household registers that covered the area of Danseong from 1780 to 1786, in this study we examine longitudinal features of Korean household registers. We assess the degree of attrition over six years at both individual and household level, and identify social and demographic characteristics associated with attrition. Our investigation shows a considerable degree of attrition of individuals and households over six years which is significantly associated with the social status of individuals and the household heads. We supplement the analysis of exits with the analysis of new entrances, which shows that in each year of register a substantial number of individuals, comparable to the size of individuals and households lost, were newly added. Those who newly entered the registers are very similar in their characteristics to those who exited the registers. We discuss the implications of our findings for studies that rely on Korean household registers to examine family and demographic behaviors of historical populations in Korea and East Asia.

Keywords::

• Danseong household register
• longitudinal features
• demographic behavior
• historical population
• social status

Notes

^1. For translation of Korean language terms, we follow the convention specified by the Korean government which can be found in the “Glossary of Korean Studies” provided by the Academy of Korean Studies (available at http://www.aks.ac.kr/glossary/default.asp).

^2. However, note that there are several missing registers during the period of 1606 to 1888, which makes it difficult to follow up the same individuals and households for long times. Moreover, registers for the nineteenth century do not contain all villages in Danseong. See Park and Lee (2008) for more information on Danseong household registers.

^3. There is disagreement among Korean historians on exactly how a specific social status was attached to a woman. In particular, scholars disagree to what extent a woman's social status was related to her husband's or her own father's status (see K. Kim, 2003). Whatever the specific mechanism was, the relevant point for the current study is that women as well as men were clearly distinguished into different social status groups in the household register.

^4. We thank an anonymous reviewer and the guest editor, Cameron Campbell for their suggestion of this analysis of new entrances supplementing the analysis of exits.

^5. We excluded 40 individuals whose information on gender was missing.

^6. Among 3,513 individuals who were recorded in the 1780 register but disappeared in the 1783 register, only 166 individuals showed up again in the 1786 register.

^7. We treated 166 individuals who disappeared in the 1783 register but somehow reappeared in the 1786 register as if they disappeared entirely in the 1783 register. We also conducted a supplementary analysis by treating them as if they stayed in both 1783 and 1786. The result was hardly different.

^8. For multivariate analysis, a very small number of individuals whose age was missing were included into the age category of 61+.

^9. J. Jung's (2003) study showed this potential line of research by comparing records contained in households registers and genealogies (and other sources), although it covered only very small areas and samples.