Models of Fertility and Women's Work
回应Cramer对作者1978年论文的批评，论证非递归模型在均衡条件下适用于研究工作与生育率关系，并通过敏感性分析表明工作对生育率的影响几乎为零。
Comments on an article by Cramer in an earlier issue of the Journal (1980). The most basic question raised is the appropriateness of using nonrecursive models to study the work-fertility relationship. The authors argue that such models are appropriate under equilibrium conditions and that the assumption that these conditions were met in the authors 1978 data was not unreasonable. Other issues concern the inadequacy of the instruments in the authors 1978 paper. Cramer virtually dismissed the results based on his replication of a 2-stage least squares model a model which the authors note was flawed. The authors further demonstrate the robustness of their own results using Cramers recommended sensitivity analysis on their own model and data. The effect of work on fertility is shown to be effectively zero under a variety of model specifications. It is argued that Cramer was incorrect in concluding that the authors results were due to statistical problems. (authors modified)