I’ve long been skeptical of Botswana’s classification as democratic/free by standard datasets we political scientists rely on. Based on a pre-election report by Amy Poteete for The Monkey Cage, skepticism seems justified. Mysterious accidents and other odd events involving political rivals, increasing partisan use of state assets…
The Botswana Democratic Party (BDP), the only ruling party Botswana has ever known, is feeling the heat, and is experiencing internal tension, including cases of losers in candidate selection running against the party’s official choice.
In the 2009 election, the BDP won over three fourths of the seats* on just 53% of votes (FPTP). That’s some serious disproportionality, but a result like that reveals electoral precariousness.
Using the dataset of the Constituency Level Electoral Archive, I took a look at just how many seats the BDP won only narrowly in 2009. In that election, its median margin over the runner-up was .188 of the district’s valid vote. It would need to lose at least 15 seats to fail to retain its majority. The 15th most marginal seat was won by .116. (Five were under .05.) There were 17 seats that it won with vote shares below that of its nationwide share of .533, including seven won with under half the vote. Thus the party looks somewhat vulnerable if there is a modest swing against it or if the defecting candidates are in close districts and take some chunk of the BDP vote with them.
Nonetheless, the BDP in 2009 did not face a single opponent in many districts; of the 17 districts where its vote was below the nationwide share, it won with a median margin of .081. Thus unless its opposition is more coordinated in 2014, the BDP will probably hang on even in the face of an adverse swing.
The election is on 24 October.
* The data I am working with show 44 seats, which would be 77.2%, while Poteete says 79%.