igmobil: A command for intergenerational mobility analysis in Stata

In this article, I describe a new command, igmobil, that computes up to 20 intergenerational mobility (IGM) indices for continuous (that is, income or years of education) or discrete (that is, educational or occupational level) variables. I consider three classes of IGM indices: 1) single-stage indices, 2) indices derived from a transition matrix between parents’ and children’s socioeconomic status, and 3) indices based on inequality measures. Users may add a fourth class to specify any possible IGM index not included in igmobil. Standard errors and confidence intervals are calculated using a bootstrap procedure. Users can customize many aspects of the program output, including the type and dimension of the transition matrix, the parameters for some IGM indices (like the ones involving generalized entropy measures and the Atkinson index), and how standard errors and confidence intervals are calculated.本文介绍了一款全新的计量命令igmobil，其可针对连续型变量（如收入、受教育年限）或离散型变量（如教育层级、职业层级）计算至多20种代际流动（intergenerational mobility, IGM）指数。本文涵盖三类代际流动指数：1）单阶段指数；2）基于父母与子女社会经济地位间转移矩阵推导得到的指数；3）基于不平等测度构建的指数。用户可自行新增第四类指数，以纳入igmobil未覆盖的任意代际流动指数。标准误与置信区间可通过自助法（bootstrap procedure）计算得到。用户可对程序输出的诸多维度进行自定义设置，包括转移矩阵的类型与维度、部分代际流动指数的参数（如涉及广义熵测度与阿特金森指数的相关指数），以及标准误与置信区间的计算方式。