If you follow the news, you will have observed a recent coup in Myanmar. Myanmar had been a military dictatorship for 22 years until a semi-transition in 2011. Unlike the other autocracies of the region, the autocracy had elite support only in the military (no middle classes, communal or economic elites (Slater)). Myanmar’s regime was one of the most predatory in the world, relying on natural resources such as diamonds and brutally neglecting its people, restraining their collective action capacity to challenge the military (De Mesquita, political survival). The military regime once responded to a hurricane once by dispersing refugee camps and banning humanitarians to restrain collective action (Ibid).

Acemoglu and Robinson’s model provides a parsimonious description of the democratization game. De facto power fluctuates between the pre-democracy selectorate (PDS) and the pre-democracy ejectorate (PDE). Because power fluctuates, when the PDS is challenged by a revolution they cannot credibly commit to adopt pro ejectorate policies. The ejectorate knows that next year they may be weak again and the regime will change its mind. The military wanted to make concessions toward, but wanted a credible commitment that their monopoly on violence and access to rents from state industries would continue. But once democracy empowers the PDE, the democracy cannot credibly commit to maintain the military’s privileges. So the military “democratized” while reserving various veto powers and a, crucially, its monopoly on violence.

The military certainly retained a great deal of de jure power, appointing ¼ of the parliament, privatizing state-owned industries into military families, and influencing presidential nominations. But the military also retained great defacto power, including a majority on the National Defense Council, the vice presidency, and even the ability to compel public servants to attend rallies. The military always retained the ability to reassert control.

The military realized that once democracy consolidated under the opposition NLD, it would be too late to protect their interests. The timing of the coup supports this explanation; it was launched the day before the parliament accepted the NLD’s second landslide electoral victory. More electoral victories would enable NLD to consolidate and shift the balance of de facto power away from the military. Without a coup, the NLD could have waited until the military was weakest then broken them and dismantled their corruption machine. A faction of the military prefered a stunted national economy and international sanctions to this outcome, so they launched the coup.

The real puzzle is why were non-credible promises tolerated from Myanmar’s military, but not from the military in Spain, the monarchy in Britain or the UMNO in Malaysia in 2018. Surely most outgoing regimes would prefer to maintain a veto and their tanks. My tentative guess is that internal pressure on the military was weaker than most transitions (due to weak collective action and high state violence capacity). The transition relied more on pressure from foriegn sanctions. Foreign observers are less informed than challengers, and misread the credibility of the commitment. I am concerned about Sudan’s transition following the same path.

A second question is why the ANC’s promises to protect white South Africans through democratic institutions was credible, but the NLD’s promise to protect the Myanmar military was not. One theory is that the National Party was losing control in South Africa anyway, and trusting the ANC was the best remaining option. Anyone have another idea?

Citations: Slater \emph{Ordering Power} De Mesquita and Smith, \emph{The Logic of Political Survival} Acemoglu and Robinson \emph{The Economic Origins of Dictatorship and Democracy}