Demographic processes in China and Vietnam: Three essays

This dissertation includes three separate papers that share an emphasis on the significance of socio-political and historical factors on the quality of demographic data in China and Vietnam. The first two papers investigate the interaction between data collection and family planning policy implementation in contemporary China. They consider how the quality of birth statistics may be compromised by the bureaucratic and political concerns of local officials whose job performance is evaluated on the basis of their ability to keep fertility low, and by individual couples, many of whom want more children than they are officially permitted. To search for evidence of underreporting of births, and unravel the complex mechanisms underlying the association between such phenomenon and family planning policies, the first paper conducts a microanalysis of birth data in a dozen villages in a Chinese county, as maintained by separate registration systems, the family planning and the health systems. The second paper draws deductions from application of statistical models to data collected in two successive surveys conducted in hundreds of villages in four counties. The results of these analyses indicate that the Chinese state's pervasive presence in people's reproductive lives, and the administrative interference with the collection and processing of fertility data, make China's existing family planning statistical system very sensitive to changes in policy and administrative procedures, at the expense of data quality. The third paper confronts the methodological problems posed by Vietnamese census data to estimate mortality during the 1979-1989 intercensal decade. To minimize bias arising from the political and social upheavals of the aftermath of war, e.g. massive intercensal outmigratory streams, internal population redistribution policies and a highly mobile population, this study performs an unprecedented adjustment for intercensal emigration. It also proposes steps to reduce errors in two successive census age distributions caused by differential census enumeration completeness and residual emigration that could not be accounted for by the adjustment procedure. The reconciliation of the results from various techniques allows one to identify a set of estimates that provide a more consistent, accurate measurement of mortality levels prevailing in Vietnam during the 1979-1989 intercensal period.