China's reform of the Hukou system and consequences for workers and firms

Abstract

This thesis contains three interconnected studies on China’s reform of the Household Registration System, or the Hukou System, which regulates internal migration. In Chapter 1, we describe the general institutional background of the Hukou System, look into a reform of the Hukou System adopted by local governments since 2002, and give the outline of each core chapter. In Chapter 2, we explain how the Hukou reform was gradually adopted by local governments in China. We propose a theoretical framework to describe local governments’ incentives to adopt the Hukou reform. By collecting information on the reform date at the prefecture level as well as a number of prefecture-level explanatory variables, we use two different approaches to empirically study local governments’ decisions to adopt the Hukou reform. First, we use a simple logit model to estimate how local economic factors in the pre-reform year affect the probability of early reform adoption. We group prefectures into “early reform” prefectures and the remaining prefectures based on a threshold point in time. Second, we use a complementary log-log model to model the duration of time until reform adoption. We transform our data into discrete-time duration data for the period 2002-2007 given that the reform was adopted by local governments at different points in time. Prefecture which did not adopt the reform until after 2007 are treated as censored. Our estimates indicate that prefectures with a higher degree of labour misallocation and higher demand for labour are adopting the reform earlier, and prefectures with a higher level of expenditures involving public services and social welfare are adopting the reform later. In Chapter 3, we examine whether the Hukou reform increases migration into the prefecture. Because information on migrant stocks and flows by prefecture is not consistently available to us, we construct three different indicators to capture prefecture-level migrations in China from 2000 to 2010. The first indicator refers to the number of non-agricultural Hukou holders registered at the prefecture level, which is available from 2000 to 2008. We use this indicator to capture whether the reform makes it easier for migrants to convert to a local Hukou. The second indicator refers to migration inflows from other prefectures, which is available in 2000 and 2005. We use this indicator to capture whether the reform induces new migration inflows from other prefectures. The third indicator refers to the size of total urban population, which is available in 2000, 2005 and 2010. We use this indicator to capture whether the reform increases the net migration inflows (migration inflows minus migration outflows). We use a difference-in-differences framework with prefecture-level fixed effects to estimate the reform effect on migration. Our estimates indicate that the Hukou reform increases the number of non-agricultural Hukou holders, increases migration inflows from other prefectures, and increases the total urban population in the long term, suggesting that the reform is effective in increasing migration. Also, we find heterogeneous effects of the reform for different skill groups. We find that the reform effect is larger for unskilled workers relative to skilled workers. In Chapter 4, we study how local manufacturing firms react to the Hukou reform. To be exact, we study whether manufacturing firms are growing faster in reform prefectures relative to non-reform prefectures. We focus on firm employment using the Chinese Annual Survey on Industrial Firms (CASIF) from 2000 to 2007. We use a difference-in-differences framework with firm-level fixed effects to estimate the reform effect on firm employment. We also use a propensity score matching approach to deal with the reform endogeneity suggested by the Chapter 2. Our estimates indicate that the reform increases local manufacturing firms’ employment on average. In addition, the reform impact on firm employment is larger for firms located in labour intensive industries and for firms located in industries that use migrant workers more intensively. Finally, we find that the reform also affects state-owned firms. In contrast to the non-state-owned firms that expand after the reform adoption, the state-owned firms downsize. Chapter 5 summarizes the main findings of the thesis and discusses our future research directions.

Date of Award	8 Jul 2018
Original language	English
Awarding Institution	Univerisity of Nottingham
Supervisor	Richard Upward (Supervisor), Paolo Epifani (Supervisor) & Juliane Scheffel (Supervisor)