Presidential Term Limits and Regime Type

After my post on president age, it struck me that in just 2 years, three of the 15 presidents ran for a 3rd or 4th term. That seems like a lot, since presidential term limits (usually 2) are one of the continent's democratic successes. 

Presidential term limits are particularly important in "presidential regimes" in which the president is both head of state and head of government. In semi-presidential regimes, where the prime minister is head of government and answers to both the president and the legislature, presidential term limits might play a different role. The president is already constrained to a greater degree by the legislature, and prime ministers generally don't have term limits, so term limits on the presidency might not be as high-stakes. 

So I thought it would be interesting to look at the status of term limits in presidential and semi-presidential ECOWAS countries separately. (There are currently no parliamentary regimes in the region.) Shugart and Carey further differentiate semi-presidential systems between premier-presidential and president-parliamentary forms of semi-presidentialism. In premier-presidential systems, the prime minister is stronger relative to the president - the prime minister and cabinet, once in office, answer only to the parliament/legislature. In president-parliamentary, the prime minister and cabinet answer to the president and to the parliament. This means that even in the president-parliamentary system, the mere presence of a prime minister should not be enough to make the system semi-presidential, if the president alone can appoint and dismiss the prime minister. 

As a first step to identify the regime type in 2015, I looked at Matt Golder's data, as compiled with a lot of other data in the Quality of Government dataset. I'm looking at gol_inst. 

This list has some problems. First, non-democracies aren't included so we're missing cases such as Togo and Gambia. And some of the democracies are miscategorized. Benin should not be categorized as semi-presidential - the office of prime minister is not even mentioned in the Constitution, it's an office that presidents can choose to create and abolish. In the past 7 years there has only been one prime minister  that served for less than one year, when Boni Yayi was trying to tap a successor. So I would call Benin a presidential regime. 

I'll look at Robert Elgie's data instead. This isn't missing any countries (dictatorships are included), I don't see any errors, and it has the added benefit of separating president-parliamentary (which is really close to presidential, particularly in Africa where presidents have a lot of power) from premier-presidential (where it might make some sense, in some cases, to think of the prime minister rather than the president as the head of government). 

This is a good list. It isn't fooled by the presence of a prime minister. Guinea, for example, lists the prime minister office in its constitution, but the prime minister answers only to the president. 

Just categorizing the cases took longer than I expected, so I don't have much time now to look at how that relates to term limits. I'm just going to put the Posner and Young Table 11.1 data (ends in 2015) into this list and see what happens. I'll go past that another time. I'm pretty sure none of these regime types changed since 1990. 

I'm going to check this for correctness and then update since 2015. For example, it appears Cote d'Ivoire did have term limits. But just from this preliminary analysis, it looks to me like term limits were less prevalent, and more likely to be respected, in premier-presidential regimes than in presidential regimes. 

This looks like promising enough for me to try turning this into a research paper, or at least a research note, for publication. So I might not share more on this blog - contact me if you're interested.