During the POLPAN Seminar on May 7, 2019, Dr Zbigniew Karpinski (Polish Academy of Sciences) and dr John Skvoretz (University of South Florida) presented a talk entitled “Status Allocation from Elementary Allocation Principles”.

Abstract:

Status allocation is a process by which individuals with varying credentials and background characteristics are allocated by complex labor market processes to occupational positions that vary systematically in their rewards such as prestige, autonomy, and compensation.  To simplify, the outcome of such a process is a two-way table of origins by destinations, the rows of which are ordered from highest to lowest category of credential – most often educational attainment – and the columns are occupational categories ordered from highest to lowest in terms of standing in the occupational hierarchy. Previous literature has focused on two principles by which allocation can be assessed: the meritocratic principle and the lottery or chance principles (Krauze and Slomczynski 1985). The concern of the work has been to assess how far an observed distribution is from the distributions that would obtain under pure meritocracy versus pure lottery. In our work, we treat these principles as elementary allocation principles that combine via estimated mixing parameters to produce an observed distribution, that is, an observed distribution is a compound of the two elementary distributions governed by a mixing parameter that expresses, in effect, the proportion of the table attributable to one principle versus another. We generalize the constant mixing model to a differential mixing model in which the mixing parameters can be origin specific. Our framework allows for systematically addressing a number of research questions: (1) What values does the mixing parameter typically take on? (2) How well does the constant mixing model or the differential mixing model fit a specific observed outcome?  (3) Is the rank of the origin category related to the value of the origin-specific mixing parameter? (4) Do the mixing parameters differ systematically for different populations or over time? We examine these questions with data on educational origins and occupational destinations from EU for over 30 countries from 2004 to 2017 for men and women separately. We find that the differential model fits better than the constant model, that mixing coefficient for top origin category is much larger than the mixing coefficient for lower categories, that the allocation of women is more meritocratic that the allocation of men, that allocation (based on the differential model) becomes less meritocratic over time, and that countries differ widely how much meritocracy versus lottery contributes to status allocation.