When "Voice" Gets the Opportunity to "Exit": Evidence from Hong Kong

• In Event: Repression, Mobilization and Protest Under Autocracy

Abstract

Does having the outside option of emigration make people more or less likely to engage in pro-democratic protest? We utilize an age-based discontinuity in eligibility for the British National Overseas (BNO) passport in Hong Kong to obtain quasi-random variation in exit opportunities. Only those Hong Kong natives born before the territories’ handover on July 1st, 1997, were eligible for the BNO passport. Possession of the document constitutes a significant increase in exit opportunities: UK Prime Minister Boris Johnson announced in June 2020 that holders would obtain the right to take residence in the UK for multiple renewable 12-months periods with an immediate right to work and a clear trajectory to citizenship.

We employ a pre-registered regression discontinuity design and surveyed 763 young Hong Kongers in June 2020. We targeted exclusively native Hong Kongers who were born within an 18-month interval around the handover date (January 1996 - December 1998). Our identification assumption of the continuity of potential outcomes around the threshold is supported by the balance on all 19 political and socio-economic covariates we included in the survey. Crucially, the local average treatment effect of BNO status on protest intent refers to a subpopulation of considerable substantive interest – those who age-wise should be most likely to protest. We find that BNO status does not reduce intent to protest, the expected size of overall turnout in the annual July 1st protests, or the perceived efficacy of protest.

Further, we embed a survey experiment in which we prime half of our respondents with Boris Johnson’s announcement about (1) the exit opportunities attached to BNO status and (2) that about 40% of Hong Kong’s current population is eligible. Conceptually, we understand this treatment as priming the extent to which the population has exit opportunities. We find that this prime does not affect protest intent – both unconditionally and conditional on BNO status of the participant.

Overall, our study contributes to a distinguished, if observational, literature that examines how opportunities to emigrate (“exit”) influence individual’s willingness to engage in pro-democratic domestic protest (“voice”) (e.g., Hirschman 1970, 1978; Gehlbach 2006). By leveraging quasi-random access to exit opportunities, we demonstrate that (1) one’s exit opportunities and (2) perceiving signals of increased exit opportunities in the population do not reduce protest intent and perceived protest efficacy. We believe these findings to be highly policy relevant: Western policy makers deciding about the restrictiveness of immigration policies want to factor into their decision how much facilitating emigration would hurt domestic protest against (impending) authoritarianism in sending countries. Our findings suggest that announcing more liberal immigration policies does not reduce pro-democratic protest.

Authors

• Julian Michel, UCLA

• Margaret E. Peters, University of California, Los Angeles