Abstract:          Based on the dynamic monitoring data collected by the National Health and Family Planning Commission on floating population in 2015， this paper applies logistic regression model to analyze migrant long-term residence tendency in cities． The results show that migrant long-term residence tendency is influenced by their demographic， human capital， and economic characteristics， through which this migrant group has demonstrated a new feature of differentiation．Improvement of education enhances individual＇s long-term residence tendency． Junior college education being the dividing point， the higher educated migrants present a stronger tendency of residence． The increase of income level has a significantly effect on the residential preference of the entire group． Income of 6000 yuan being the dividing point， the high-income group of migrants have a stronger long-term residence tendency than the low-income group． When added into the model the housing expenditure， the increased housing expenditure elevates migrant residential preference． However， it also has a negative effect on the residential preference of the high-income group．

Keywords: Migrants , Long-term Residence Tendency , Duration of Stay

摘要：       文章根据 2015 年中国流动人口动态监测调查数据， 运用二元 Logistic 模型对流动人口在城市的长期居留意愿进行分析， 发现个体特征、 人力资本特征、 经济特征共同影响流动人口在城市的长期居留意愿，并使得这一群体内部出现了分化的新特征。研究表明: 教育程度的提高在普遍增强个体留城意愿的同时，以大学专科学历为界限，高学历者呈现出更加强烈的居留意愿。收入水平的增加显著影响了整个群体的居留意愿， 以 6000 元收入水平为分界点， 表现为高收入流动人口的城市居留意愿远大于中低收入者。在追加了住房支出变量后， 发现伴随着住房支出的增加提高了流动人口的城市居留意愿，但同时又对高收入流动人口的长期居留意愿表现出抑制作用。

关键词: 流动人口 , 长期居留意愿 , 流动时间