It is well known that the college enrollment rates of blacks have historically trailed those of whites, although in recent decades the actual size of the racial gap has fluctuated. Prior research has shown that blacks are more likely than whites to attend college after high school graduation, net of socioeconomic background and academic performance. It has been suggested that this "net black advantage" may be spurious--due to blacks' relatively high enrollment rates in historically black colleges and universities. With data from the National Education Longitudinal Study of 1988-1994, this hypothesis is tested by examining black-white differences in enrollment in different types of colleges: any college, four-year colleges, non-black four-year colleges, and academically selective four-year colleges. Overall, results confirm the existence of a net black advantage at low levels of family socioeconomic background. The implications of these findings for racial equality in access to higher education are explored.


In this article we explore methods for using mothers’ interviews to gather data on their children’s family formation experiences. These methods constitute a cost-efficient means of gathering data for models of family background that include both intergenerational and sibling influences. To judge the utility of these methods, we examine the quality of
mothers’ reports across a range of their children’s family formation behaviors. The dimensions of reporting quality we analyze include completeness, precision, and accuracy of mothers’ reports. We use unique data from personal interviews with mother– child pairs to test the accuracy of these mothers’ reports. The results demonstrate that, with some
behaviors, a flexible data collection approach can gather complete, precise, and accurate information on an entire sibling set by interviewing mothers. Our examination of data quality also suggests important limits on the use of this approach. The quality of mothers’ reports depends on the subject matter, with mothers providing lower quality reports of
their children’s cohabitation behavior compared to their children’s marital, childbearing, and divorce behavior.


Demographic models can have two meanings, one broad and one narrow. In its broad meaning, demographic models refer to all mathematical, statistical, forecast, and microsimulation models that are applied to studies of demographic phenomena. In its narrow meaning, demographic models refer to empirical regularities in age patterns of demographic events. This article is concerned with demographic models in the narrow definition. Demographic models are widely used (a) to improve data quality and (b) to compare demographic outcomes and processes across populations or subpopulations. Both parametric and semiparametric specifications have been proposed for modeling age patterns of demographic events, giving rise to parametric models and semiparametric models. Successful applications of both types of models are found in research on mortality, nuptiality, and fertility. As an integral part of formal demography, demographic models have been linked closely to mathematical demography. In recent decades, however, statistical demography has played an increasingly important role in demographic models.






This article studies the regional variation in earnings inequality
in contemporary urban China, focusing on the relationship between
the pace of economic reforms and earnings determination.
Through a multilevel analysis, it shows that economic growth depresses
the returns to education and work experience and does
not affect the net differences between party members and nonmembers
and between men and women. Overall earnings inequality
remains low and only slightly correlated with economic
growth because, in faster-growing cities, the tendency toward
higher levels of inequality is somewhat offset by the lower returns
to human capital. A plausible interpretation is that these
results are largely due to the lack of a true labor market in urban


In this paper, the author develops a new class of discrete-time, discrete-covariate models for modeling nonproportionality in event-history data within the log-multiplicative framework. The models specify nonproportionality in hazards to be a log-multiplicative product of two components: a nonproportionality pattern over time and a nonproportionality level per group. Illustrated with data from the U.S. National Longitudinal Mortality Study (Rogot et al. 1988) and from the the 1988 June Current Population Survey (Wu and Tuma 1990), the log-multiplicative models are shown to be natural generalizations of proportional hazards models and should be applicable to a wide range of research areas.