Even though an excerpt of each of these comments appears (as with all comments at F&V) over on the left sidebar, the following comments to an earlier post on Bolivia are so interesting and relevant to ongoing themes here that I wanted to “promote” them to the front page. Thanks, Wilfred and Miguel! Please note: None of the remaining text in this post is mine.*
FROM WILFRED DAY: If the results posted on the National Electoral Court’s website are official, Bolivia’s MMP system has just passed its first “sweep test” with flying colours.
Previous elections have seen no party win a majority, so a congressional coalition would form a government and elect a president. But this time one party, MAS, got 53.74% of the vote, making Evo Morales president outright. With a single ballot for president and congress, and a proportional system, MAS should have 70 of the 130 deputies. But Bolivia’s MMP system has nine separate calculations, one in each self-contained province. There were two risks:
1. Local sweeps by other parties could have deprived MAS of a majority. In the four provinces won by the more conservative PODEMOS, they might have swept the 26 local seats, 54% of the 48 seats from those provinces, with much less than 54% of the vote. With no German style-links between the provincial calculations — no national compensation — these sweeps could have deprived MAS of its majority. Early projections suggested this is just what would happen. In the result, however, PODEMOS won 17 of the 26 local seats in those four provinces, and in each of the four, it had enough votes left over to earn at least one list seat. No disproportionality. Take little Pando, their smallest province. PODEMOS almost won local seat #68, losing by 10 votes to the centrist UN. What happens if a recount changes that result? No change: PODEMOS earned three of the province’s five seats, and now has two local seats and one list seat. If UN lost seat #68, it would take that list seat from PODEMOS instead. The MMP universe is unfolding as it should.
2. The bigger risk might have been a “large-party bonus†for MAS. It swept all 15 local seats in Bolivia’s largest province, La Paz, and all but one in Cochabamba, Potosi, and Chuquisaca. (Those kind of sweeps happened in four of Scotland’s eight regions in both the last two elections. Last time, Labour got 38.8% of the seats with only 29.3% of the votes.) But in each province, MAS got enough votes to justify those sweeps, enough and to spare: at least one list seat in each of those provinces. In La Paz they got 66.63%, earning 5 list seats on top of their local sweep.
The final unoffical result seems to be 70 deputies for MAS, dead on target.
FROM A RESPONSE TO WILFRED BY MIGUEL CENTELLAS: Wilfred raises a good point. I was at first disappointed with MMP, because of the 2002 debacle. I still do believe (with my research evidence to support) that the 1997 & 2002 MMP elections polarized the electorate, which encouraged regionalized party alignments. Perhaps that breakdown was necessary for an eventual re-alignment into the current two-party system. Time will tell, of course. But it seems that the 2005 election was basically split into “blue†(MAS) & “red†(PODEMOS) regions, with some “battleground†departments to tip the balance. Either way, an effective number of parties of 2.2 is remarkable for a country that previously had an effective number of parties at 4-5.
* MSS: I should note that I did some very light editing of these entries.
With final results, we can see that the minor weaknesses of Bolivia’s MMP system are just what one would expect.
If it had a national compensation calculation, using d’Hondt, MAS would have had 73 seats, and actually got 72. PODEMOS would have had 39, but got 43, a bonus of 4. UN would have had 10, but got 8. MNR would have had 8, but got 7.
With nine self-contained states, small parties are going to have some wasted votes. But MAS being shorted a seat is counter-intuitive. I assume MAS is short one because PODEMOS won 3 of 4 small states. While LaPaz has a seat for each 33,167 voters, and Cochabamba a seat for each 27,227 voters, Tarija has one for each 15,149 voters, Beni one for each 11,409 voters, and Pando one for each 4,081 voters. This would be corrected with a German-style national compensation, but not in Bolivia.
UN has 34,717 wasted votes: 9,875 in Potosi, 9,793 in Tarija, 8,633 in Oruro, and 6,416 in Beni, where it won no seats. MNR has even more, 57,023: 24,482 in LaPaz, 12,773 in Cochabamba, 11,021 in Potosi, 6,705 in Chuquisaca, 2,042 in Pando. But MNR’s strength in small-states Beni and Tarija helped offset this.
UN and MNR were actually lucky: d’Hondt can disadvantage parties that win only one seat in a state where another calculation method would have given them two seats, but that didn’t happen this time.
In all, considering that Bolivia’s model could suffer the same large-party bonus as Scotland’s, it performed very well, this time.
Seed planted by Wilfred Day — 14 January 2006 @ 22:50