Code and data used in the paper: Changing Educational Homogamy: Shifting preferences or evolving educational distribution?

Naszodi and Mendonca (2022) studies educational homogamy in the US and four European countries (France, Hungary, Portugal, Romania). The authors apply two methods for identifying the directly unobservable shifts in marital preferences from one generation to another. Both methods identify changes in preferences through their effects on the share of educationally homogamous couples with a counterfactual decomposition. The first method was developed by Naszodi and Mendonca (2021). It uses only aggregate census data on the equilibrium matching outcome. It is implemented in R, Excel and Visual Basic and can be downloaded from doi:10.17632/x2ry7bcm95.2 (currently available in version 2). The aggregate census data used in the benchmark analysis of the paper by Naszodi and Mendonca (2022) can be downloaded from this site (see: Changing Educational Preferences - ContingencyTables.xls). These data are aggregate census data from IPUMS on the joint educational distributions of young male and female partners/spouses in the studied five countries covering two census waves. The second method was developed in Naszodi and Mendonca (2022) and implemented in R. This method uses dating data. First, it estimates the education level-specific distributions of the reservation points of the marriageable young women and men in the population from data on the distributions of the search criteria (on the education level of the preferred partners) of a sample of dating site users. Then, the method uses the estimated reservation points distributions together with the aggregate Gale-Shapley algorithm and the census data on the joint educational distributions to construct various counterfactuals for the decompositions. The uploaded files allow one to replicate the calculations in the paper by Naszodi and Mendonca (2022) using the reservation points-based second method. Also, it allows one to apply this method in other empirical settings as well. References: Naszodi, A. and Mendonca, F. (2021) "A new method for identifying the role of marital preferences at shaping marriage patterns." Journal of Demographic Economics, 1-27. doi: https://doi.org/10.1017/dem.2021.1 Naszodi A, Mendonca F (2022). Changing educational homogamy: shifting preferences or evolving educational distribution? Journal of Demographic Economics 1–31. https://doi.org/10.1017/ dem.2022.21纳佐迪（Naszodi）与门东萨（Mendonca）（2022）针对美国及四个欧洲国家（法国、匈牙利、葡萄牙、罗马尼亚）的教育同配（educational homogamy）现象展开研究。两位作者采用两种方法，用以识别代际间婚姻偏好中直接不可观测的变化。两种方法均通过反事实分解（counterfactual decomposition），基于其对教育同配夫妻占比的影响，捕捉偏好的变动。 第一种方法由纳佐迪与门东萨（2021）提出，仅依托均衡匹配结果（equilibrium matching outcome）的总体普查数据实现。该方法可通过R、Excel及Visual Basic运行，可从doi:10.17632/x2ry7bcm95.2下载（当前版本为2.0）。纳佐迪和门东萨（2022）的基准分析中使用的总体普查数据，可从此站点下载（文件名为：Changing Educational Preferences - ContingencyTables.xls）。此类数据源自IPUMS的总体普查资料，涵盖研究涉及的5个国家的年轻男女配偶/伴侣的联合教育分布情况，包含两轮普查数据。 第二种方法由纳佐迪与门东萨（2022）提出，通过R语言实现。该方法使用约会数据：首先基于约会平台用户样本中对伴侣教育水平的择偶标准分布数据，估算人口中适婚青年男女的教育水平特定保留点（reservation points）分布；随后结合估算得到的保留点分布、总体Gale-Shapley（盖尔-沙普利）算法，以及联合教育分布的普查数据，构建各类反事实场景以完成分解分析。 本次上传的文件可复现纳佐迪和门东萨（2022）论文中基于保留点的第二种方法的计算过程，同时也支持将该方法应用于其他实证场景。 参考文献： 纳佐迪A，门东萨F（2021）《识别婚姻偏好对婚姻模式塑造作用的新方法》，《人口经济学杂志》，1-27。doi: https://doi.org/10.1017/dem.2021.1 纳佐迪A，门东萨F（2022）《变化的教育同配：偏好转移还是教育分布演变？》，《人口经济学杂志》，1–31。https://doi.org/10.1017/dem.2022.21