Democratization, autocratization, and coups: Assessing the impact of regime changes on forced migration
: (Alternative Format Thesis)

  • Sung Won Lee

Student thesis: Doctoral ThesisPhD


Thesis abstract

Human displacement occurs when people are motivated to leave their home country and when those who are eager to escape are given the opportunity to do so. In this regard, regime transition in the country of origin can be a key to understanding the phenomenon of forced migration: people’s motivations and opportunities for fleeing can be increased or limited by such a process. However, relatively few studies have systematically addressed how regime transition in the country of origin of forced migration affects the exodus of people, although studies on the causes of forced migration are extensive. This thesis aims to investigate how various forms of regime transition affect forced migration. The types of regime transition on which this study focuses are democratization, autocratization, and a radical replacement of the regime by a coup d’état.

The first paper analyzes the impact of regime change through a coup d’état on the volume of refugee outflow generated. Much research points to the relevance of political instability and violent armed conflict in explaining refugee flows, yet we still know little about whether and how other forms of political instability such as coups d’état create incentives for forced migration. We, therefore, develop a theory to describe the influence of coups on individuals’ decisions to flee, highlighting the fact that coup events exacerbate individuals’ physical and economic insecurity and thus increase incentives to flee. We consider successful and failed coups and assess our claims using data for all countries between 1980 and 2015. Our most conservative estimates suggest that successful coups have a substantive effect on the number of refugees: forced migration from countries that have experienced a successful coup is 40 percent larger than in countries that have not experienced a coup. We illustrate the theoretical mechanisms by analyzing four coups in two counties: Uruguay in 1973 and 1976 and Egypt in 2011 and 2013.

The second paper determines whether regime shifts play a role in creating or constraining willingness and opportunities to flee and if they cause an increase in the number of people in the country of origin with the intention of seeking asylum elsewhere. Specifically, regime shifts refer to changes in political regimes: such regimes can vary qualitatively, as some regimes move toward democracy (democratization), whereas others shift along the spectrum in the opposite direction toward autocracy (autocratization). Using panel data covering the period from 2000 to 2016 across 118 countries, my results show that the number of asylum seekers tends to increase when a country experiences either sudden democratization or autocratization. Nevertheless, the level of increase is dependent on the political regime of the time. Specifically, for democratized countries, the increase in asylum seekers tends to be smaller if the countries experience a further democratic regime shift. Such a diminishing marginal effect is explained by the decrease in willingness to escape gradually outweighing the effects of greater opportunity to escape, which occurs in more democratic countries. In other words, as democratic countries become more democratic, the outward flow of the population becomes less prominent because people are less motivated to escape, even if the opportunity to escape is greater. On the other hand, during autocratization, the increase in asylum seekers is larger if the countries were more democratic (or less autocratic) prior to the regime shift. There is an increasingly marginal effect because in more democratic milieus, opportunities to escape still exist that accommodate the increase in willingness to escape as a result of autocratization. In other words, if an already autocratic country further autocratizes, the population outflow is mitigated, as the citizen’s opportunities to escape the country are marginal to start with and limited further, even if their motivation to escape increases.

This study not only analyzes countries that have already experienced refugee outflows but also focuses on an area that has not experienced a mass outflow of forced migrants. The third paper examines different paths of regime development that the North Korean regime can follow, including 1) regime collapse, 2) status quo, and 3) democratic transition, and analyzes the possible impacts of each regime development path on forced migrations from North Korea. A statistical model is developed to examine the effects of the three regime variants on population outflow across autocratic milieus. The study employs a panel dataset covering 138 countries from 2000 to 2016. My findings show that the collapse of the regime and democratic transition in autocratic countries can increase the magnitude of the outflow of forced migrants, while the maintenance of the regime’s status quo has no significant impact on it. In other words, the findings imply, on the one hand, that if North Korea follows a path of regime collapse or democratic transition, the volume of forced migration generated may increase. On the other hand, if the current regime continues to maintain power in a manner resembling its current state, population outflow from North Korea will be deterred. The key conceptual framework of this study explains the relationship between a regime transition and population outflows, as well as distinguishing the motivation and opportunities for people to flee by controlling for the presence in a country of an autocratic military regime. In particular, the motivation and opportunities for North Koreans to flee may be increased or restricted depending on the path the regime transition follows.
Date of Award20 Jan 2021
Original languageEnglish
Awarding Institution
  • University of Bath
SupervisorTimo Kivimaki (Supervisor), Katerina Vrablikova (Supervisor) & Mauricio Rivera (Supervisor)


  • Regime transition
  • Forced migration

Cite this