David Pion-Berlin, "Informal Civil-Military Relations in Latin America: Why Politicians and Soldiers Choose Informal Venues." Armed Forces & Society 36, 3 (2010): 526-544.
This study examines the phenomenon of informal civil–military relations. Informal behaviors are those that normally do not occur within the chain of command, are not mandated by law, and do not conform to official procedures. Politicians and soldiers discover that formal, institutional routines are sometimes too constraining and that they can advance their interests more effectively by amending, circumventing, or violating those routines. The party most aggrieved by the rules of the game initiates an informal solution. Whether the other side goes along depends on how divergent its preferences are with the aggrieved party. Greatly divergent preferences result in unilateral informalities, less divergent but still negotiable positions yield bilateralconflictive
encounters, and convergent preferences result in cooperative ventures. Case studies on Chile, Argentina, and Bolivia exemplify three different kinds of informal encounters and their impacts on civilian policy choices and military interests.
I studied this issue in considerable detail for Chile in the late 1990s and early 2000s (and he kindly cites me in several places). Although I've moved in different research directions, there are still many unanswered questions and too few people tying to answer them. This is a good article because it makes an effort both to consider different types of informal encounters and the different outcomes they produce, all in a comparative context.
His examination of Chile, Argentina, and Bolivia shows how either civilian governments or the military leadership chose informal approaches because they felt that formal channels would not serve their purposes. I would quibble a bit with the case study choices, because the levels of conflict in each instance were quite different (Kirchner firing officers versus simply have a Mesa de Diálogo in Chile, for example) which might push us in an apples and oranges situation. To be fair, though, it is hard to find identical cases across different countries.
He acknowledges that much more work is required, but he does a nice job of pushing us forward with directed questions. I would suggest that we need to identify more causal mechanisms so that we can suggest concrete hypotheses. As I know from experience, that is much harder than it sounds.