Deconstructing job search behavior
利用在线招聘数据，通过共同申请模式构建网络来定义求职者的考虑集，分析人口统计特征和广告时机对申请的影响，发现失业者符合存量-流量匹配，而就业者更倾向于申请高于自身期望工资的职位。
Using online job board data, we study the key factors driving application decisions for unemployed and employed job seekers. We identify relevant job consideration sets using a network approach based on co-application patterns. We document how demographics and ad timing affect applications, finding evidence consistent with stock-flow matching for the unemployed. Furthermore, we show seekers respond strongly to misalignment in education, experience, wages, and location, generally applying where observable alignment is good, although employed seekers seem more ambitious, showing greater tolerance for underqualification in education and a tendency to apply for jobs above their declared wage expectations. Methodologically, we propose this network approach for defining consideration sets, helping address potential biases in standard market definitions. This evidence contributes to understanding search behavior and differences between seeker types.