Protests in Macao Under Chinese Sovereignty: Against the State or for the State?

AbstractThe existing literature on contentious politics has long centred on “state-challenging protests,” referring to resistance that denies state legitimacy. However, the implications of “state-engaging protests,” referring to protests that seek state assistance in what are essentially private disputes rather than challenging state legitimacy, have yet to be thoroughly explored. By analysing 2,364 protests in Macao, this study compares the trends and characteristics of protests from 2000 to 2021. The results indicate that state-challenging protests that targeted the Macao government exceeded state-engaging ones but state-engaging protests that targeted the mainland Chinese authorities exceeded state-challenging ones. The conclusion strongly supports the thesis that identity affiliation affects the type of protests: citizens with a strong local Macao identity tend to be involved more in state-engaging protests, seeking state assistance. However, citizens with a strong Chinese national identity tend to engage more in state-challenging protests to criticise the local government. These findings provide a useful comparative perspective on the causes of state-challenging and state-engaging protests.Key Words: ChinaidentityMacaostate-challenging protestsstate-engaging protests AcknowledgementsThe authors are grateful to the Editor-in-Chief of the Journal of Contemporary Asia and the journal’s anonymous reviewers for their constructive feedback and criticisms on earlier drafts of this article. All mistakes are the responsibility of the authors.Notes1 Protests saw another peak in 2020 with demands for subsidies by different sectors (such as the casino sector and tourism) during the COVID-19 pandemic. Since Macao did not have gathering restrictions, there were no restrictions on protest.2 The dependent variables are measured as categorical dummy variables (0/1), and therefore the models have the same number of observations. This approach takes full advantage of the data over splitting the sample into groups, which will also make the estimation unreliable for the categories with a small N.