This is series of posts that identifies countries that almost comply with the definition of semi-presidentialism that is used in this blog, but which fail to do so on the basis of a certain, sometimes unusual provision, or where the date when semi-presidentialism started can be contested.

One of my colleagues was at APSA a few weeks ago and got talking to someone. ‘Who’s your professor?’, he was asked. My colleague gave him my name. ‘Oh’, the reply came, ‘he’s the one who thinks every country is semi-presidential’.

I’m not sure that this is how I want to go down in political science history (however small such a footnote might be). Anyway, it got me thinking. In what ways might I have been over-estimating the number of semi-presidential countries? Obviously, the wording of certain constitutions will always be open to interpretation (e.g. Art. 26 of Singapore’s constitution – does the fact that the president has discretion over the PM’s dismissal when the PM has lost the confidence of the legislature mean that the PM is not really responsible to the legislature in a standard way? In retrospect, it probably does mean that. The president is the one who decides, not the legislature). However, such ambiguities would only affect a handful of cases and I have already been careful to exclude various ‘on-the-cusp’ cases e.g. Guyana. However, is there anything that makes me systematically over-estimate the number of semi-presidential countries?

The one aspect that might make me do so is the inclusion of countries where the PM and cabinet are responsible to the legislature, but only as a function of a super-majority vote. For example, Art. 78 of the 1992 Constitution of Mali states: “The National Assembly can hold the government responsible by way of a motion of censure … Only the votes in favour of the motion are counted and the motion is adopted only if it wins the support of two-thirds of the total number of deputies in the Assembly”.

So, the government is collectively responsible to the legislature. Hence, Mali meets the definition of semi-presidentialism used in this blog. But, in practice, it is likely to be mighty difficult to bring the government down. So, the government is probably safe in office. If so, shouldn’t Mali be considered a presidential system?

I know that, for one, Matthew Shugart believes cases of super-majority responsibility should not be classed as semi-presidential (though I have seen Mali included in his list of semi-presidential countries).

Anyway, to the best of my knowledge, these are the cases of super-majority dismissal that are currently in operation:

Algeria: Art. 136
Egypt: Art. 127. (This is the situation following the constitutional amendments in 2007. Prior to this time, it was even more difficult to dismiss the government and, in effect, the government was not responsible to the legislature. Even now, though, a super-majority results in the government’s dismissal only after the second no-confidence vote. The first seems to act as a sort of warning and the president has the right to ignore it).
Mali: Art. 78
Rwanda: Art. 130 (but not under the previous constitution)
Togo: Art. 97

Note that in Kazakhstan, the 2007 amendment reduced the responsibility requirement to a simple majority from a two-thirds majority. The same is true in Tunisia since the 2002 amendment.

Also note that in Madagascar after the 1995 constitutional amendments Art. 94 stated that if the government requested a vote of confidence, then a simple majority was required, whereas Art. 97 stated that if the government is the subject of a vote of no-confidence, then a two-thirds majority was required. Obviously, though, Madagascar is no longer semi-presidential after the coup earlier this year.

So, the super-majority requirement would not get rid of many countries from my list. In any case, as I said earlier, the government is still collectively responsible under such a requirement. This is simply not the case under presidentialism. Moreover, in Mali, for example, the party system is very fragmented and the majority is usually a coalition of parties often with ethnic loyalties. It is not inconceivable that, for example, an anti-ADEMA coalition could form that might have the requisite super-majority support. By the same token, even in countries with no super-majority requirement, there can be dominant parties where the effect is to leave the government totally secure in office (e.g. Mozambique and Namibia).

On balance, therefore, I still think we need to define semi-presidentialism on the basis of constitutional rules and not on the basis of what, in practice, might be the case. Therefore, Mali etc should still be classed as semi-presidential. So, it looks like I have failed in my bid to reduce the number of semi-presidential countries (with the exception of Singapore, which was worth another look at).