China has recently undergone extensive reforms in its basic urban old-age pension policy, which intends to benefit all urban workers, including internal rural-to-urban migrant workers. In 2011, the Social Security Law, the first comprehensive law of China attempting to remedy social inequality, came into force. China has established pension schemes for both rural and urban workers.All workers employed in urban regions are entitled to the urban pension scheme, which offers pension benefits based on the cost of living in cities. However, implementation of the system and loopholes in the legislation have resulted in very low participation by migrant workers in the basic urban old-age pension scheme. As of the end of 2013, the migrant worker coverage rate of the oldage pension scheme remained less than 20% (NBS, 2014a). This gives rise to the following question: What implementation gaps hinder migrant workers from taking up this old-age scheme?Implementation gaps may occur throughout the entire policy implementation process, in which stakeholders at all levels (including policy makers, bureaucrats and recipients) play a role in the success of the policy. This thesis examined the implementation gap in the urban pension scheme at the bottom level, where rural-urban migrant workers constitute the potential recipients.Conceptualised using Mazmanian and Sabatier's (1983) implementation theory and the theory of bounded rationality, the study focused on factors that influenced migrant workers' intentions to join the urban old-age pension scheme and their perceptions of taking up the scheme.The research adopted a mixed-method approach with quantitative survey and qualitative interview data collected and analysed sequentially.In Phase 1, over 200 migrant workers were surveyed in Shenzhen, China, on their understanding of the urban pension scheme (including their intentions to join) and on factors that might influence their intentions to join (individual, land, settlement, and employment). Bivariate analysis and multiple regression models were used to statistically examine the main factors that influenced migrant workers' intentions to join the scheme. This analysis showed that workers' settlement plans (the likelihood to settle in cities and the decision to go back to their rural iii hometown) and land expropriation experience significantly influenced the intention to participate in the urban pension system. Furthermore, approximately 85% of survey participants did not understand the basics of the urban old-age pension scheme.Phase 2 consisted of semi-structured interviews with 12 migrant workers, which were designed to explore their perceptions of joining the urban old-age pension scheme. Based on preliminary results from the quantitative study, six migrant workers with land expropriation experience and six without were selected from the survey participants. Interview transcripts were subjected to thematic analysis. The rich elaborations in these interviews provided a detailed picture of the processes by which migra...