Rural-urban migration at high urbanization levels
用巴西449个城市数据检验哈里斯-托达罗模型在高城市化水平下的适用性，发现基础模型不成立，但加入非正规就业和住房市场后解释力增强，且对仅受过小学教育的工人更有效。
This study assesses the empirical relevance of the Harris-Todaro model at high levels of urbanization – a feature that characterizes an increasing number of developing countries, which were largely rural when the model was created 50 years ago. Using data from Brazil, the paper compares observed and model-based predictions of the equilibrium urban employment rate of 449 cities and the rural regions that are the historic sources of their migrant populations. Little support is found in the data for the most basic version of the model. However, extensions that incorporate labor informality and housing markets have much better empirical traction. Harris-Todaro equilibrium relationships are relatively stronger among workers with primary but no high school education, and those relationships are more frequently found under certain conditions: when cities are relatively larger; and when the associated rural areas are closer to the magnet city and populated to a greater degree by young adults, who are most likely to migrate.