As Jonathan Edelstein noted a few days ago at The Head Heeb, two factions of Fatah managed to avoid running separately in the Palestinian elections scheduled for January 25, thanks to a late, creative, ruling by the electoral court to delay the registration deadline.
Younger Fatah activists had split off and formed a party called “Future” (an interesting, if presumably unintentional parallel to the not-so-young Ariel Sharon’s splinter of Likud: “Forward”).
The delay in the registration deadline permitted the two groups to reconcile and submit a joint list of candidates. Had they failed to do so, as Jonathan notes, an already weak party system would have been made even weaker. On the other hand, the fact that they will not be running separately denies voters a chance to choose clearly between the old guard and the new blood. Still more to the point, the Palestinian electoral system effectively gave the two Fatah wings no other choice (even if they almost blew it).
The Palestinian legislative elections will be held under a variation of a mixed-member system: Half the 132 seats will be elected by plurality in multi-seat districts* and the other half by closed party list PR in a single Palestinian territorywide district (Ste.-Laguë divisors, 2% threshold). The nominal tier of regional multi-seat plurality constituencies has an average magnitude of around four seats, and the voter may vote for as many candidates as there are seats in the district.
The system is thus mixed-member majoritarian (MMM, also known as parallel). That is, unlike in MMP systems, the share of votes won by each party will not determine the overall make-up of the parliament. Instead, each party’s share of votes determines only the share of seats it wins in the party-list tier, and then these seats are simply added to the totals won by each party’s candidates in the separate nominal-tier races.**
What that means is that a party that wins many nominal-tier seats will retain much of its seat-winning bonus–unlike in MMP. And given multi-seat plurality, the votes-to-seats disproportionality of this tier could be very large (unless many voters vote less-than-straight tickets, that is mixing and matching across parties).
While I do not know the regional breakdown of Fatah factional and Hamas support, with a recent territorywide polling breakdown that Jonathan gives, it is clear that two separate Fatah lists, also running separate full slates of candidates in the nominal tier, could lose badly to Hamas (especially given that Hamas voters are probably likely to be more disciplined, i.e. likely to vote the full slate of Hamas-nominated candidates in their districts). A unified Fatah list, on the other hand, should beat Hamas.
According to the poll, Hamas is at 31.4%, Future 26.8, and “official” Fatah at 17.7. If that breakdown held, the MMM system could give Hamas, running against separate Fatah factional slates, a seat share much greater than its vote share. But by rejoining forces, Fatah, if it can hold the combined 44.5% that the poll shows for its two wings, almost surely would win a majority of seats under MMM. It is worth noting, however, that this is a rather big “if” because, given the Fatah divisions, many voters may refuse to vote for the full Fatah slate in their nominal-tier district. That is, some voters may vote only for the candidates they recognize as “Future” or “official” Fatah members, and not for the candidates of both factions. To the extent that happens, and Hamas voters are more unified, the result could still favor Hamas relative to Fatah in the votes-to-seats conversion. The two Fatah factions are still better off running together–and trying to encourage full-slate voting–than they would be running separately.
The electoral system really gave them no other choice than to come back together, or else forfeit the leading position in parliament to Hamas. In that sense, then, the recent brinksmanship was sort of an unofficial “primary” by which the Future faction sought to get better nominations than the official Fatah had previously offered it. Apparently, it has worked, albeit only after armed skirmishes between the factions.
In another post a few days earlier, Jonathan commented on the benefits to a possible further delay in the Palestinian elections (which at one time had been set for last July):
If Israel votes first, it is likely that a stable center-left government will be established, and the new government will be able to facilitate the Palestinian election process without worrying about its standing in the polls. This, in turn, will increase the odds of a more moderate Palestinian parliament being elected. If the Palestinians vote first, the situation will be precisely the opposite: election-year politics will prevent Israel from making the concessions that might boost Fatah’s electoral chances, and a strong Hamas performance in January will give the right-wing parties a campaign issue for March.
*With one proviso: There are guarantees that a fixed number of seats must be won by Christians, so in districts where Christian seats are set aside, some non-Christian candidates might be skipped in favor of one or more Christians with fewer votes. (This mechanism is similar to the women’s-representation provision in Afghanistan’s SNTV system.)
**To my knowledge, MMP has never been used with a nominal tier of multi-seat districts. It could be done, in principle, but doing so would make it all the more important that the list-PR tier be really large (not less than 50%) in order to achieve proportionality.
The best gauge of Hamas’ regional strength is probably its performance in the municipal elections. This method of prediction has its flaws, not least that the local and national party systems don’t match, but it’s the only meaningful region-by-region data out there.
If voting patterns in the general election follow the municipals, Hamas will be strong throughout Gaza (except where clan loyalties favor Fatah) as well as Nablus, Jenin, Qalqilya and parts of the Hebron hills. Fatah will be strongest in Ramallah, Jericho, Tulkarem, Bethlehem and many of the smaller villages that are tied into the patronage system. Fatah should also do well in East Jerusalem, which is one reason why Israel’s threat to impede voting there is tactically stupid.
This probably shouldn’t be taken as gospel. The data for Nablus and Jenin are skewed by the fact that the municipal elections there occurred the day after the Fatah split on a day when many party activists were demoralized (although Hamas will probably be strong in these districts anyway). Some of the results may also be affected by the fact that municipal boundaries don’t correspond to parliamentary districts; for instance, Hamas might beneft from al-Bireh being included in the same electoral district as Ramallah. The popularity of particular candidates in the nominal tier, especially the independents, will also be a factor. The most that’s possible at this point is a back-of-the-envelope guess, with mine being that Fatah will win slightly less than half the seats.
One interesting thing about Hamas’ poll numbers, BTW, is that they haven’t fluctuated by more than four or five percent since it decided to participate in the elections. Despite all the angst about whether certain measures would weaken Fatah or strengthen Hamas, the only thing that really did affect its numbers was the Future faction’s breakaway. This might suggest that Hamas has a very disciplined core of voters but that the rest of the electorate, while willing to vote Hamas for city council, isn’t rushing to do so on the national level.
I’d also be interested in your thoughts on my first two posts about the Fatah split, and about dissident party factions’ use of general elections as a substitute for meaningful primaries.
Seed planted by Jonathan Edelstein — 28 December 2005 @ 21:39
One more thing: The six Christian seats are located in Bethlehem (2), East Jerusalem (2), Ramallah and Gaza City. Christians will be substantially overrepresented in the nominal tier, given that they form only 2 to 3 percent of the total Palestinian population. However, as in Lebanon, the Christian MPs will be elected by all voters in the district rather than just Christian voters, so their party alignment will probably reflect that of the district (the Christian representative from Gaza may well be a Hamas ally). The Palestinian Christians’ traditional preference for the PFLP probably won’t translate into control of the reserved seats, except in Bethlehem where the PFLP is strong anyway.
A few other Christian candidates - Hanan Ashrawi, for instance - are high enough on national lists that they’ll probably get elected. It doesn’t seem as if the Samaritan MP will be re-elected, though, now that there’s no longer a reserved seat for him.
And the list of newly registered candidates is now up on the Central Elections Commission’s web site. The intra-Fatah agreement appears to involve both the PR and nominal tiers; among other things, Jibril Rajoub and Mohammed Dahlan are both running from geographic districts rather than on the national list. That’s probably a good move; Dahlan should be able to break through the Hamas slate in Khan Yunis, and Rajoub stands a good chance of eating into Hamas’ support in Hebron. Strong local candidates are important for Fatah, especially in districts where the party as a whole is weak.
Seed planted by Jonathan Edelstein — 28 December 2005 @ 22:20
This is a reply to the footnote “To my knowledge, MMP has never been used with a nominal tier of multi-seat districts. It could be done, in principle, but doing so would make it all the more important that the list-PR tier be really large (not less than 50%) in order to achieve proportionality.”
I think you’re right, although it’s a shame it didn’t happen in the Netherlands. They recently had an MMP model almost ready to go forward, as demanded by a junior coalition partner, but then decided not to proceed. The model was never finalized, but the draft model had a lower tier of multi-seat districts.
I don’t follow the statement that the list-PR tier would then have to be really large. Assuming the lower-tier multi-seat districts are themselves run by list-PR, the nominal results are already partly proportional, and the compensatory tier could be smaller than usual. This is a possible model for Ontario, Canada, put forward by Wayne Smith in his personal capacity (he has since become president of Fair Vote Canada).
It has a nominal tier of 14 multi-seat districts with 67 Members of Provincial Parliament (MPPs) and 36 single-seat districts. The compensatory list tier has only 21 MPPs. I don’t advocate it for Ontario because it has province-wide lists for the compensatory tier, which I don’t believe the forthcoming Ontario Citizens’ Assembly would buy. Still, it’s an interesting hybrid alternative.
A similar model in actual use is the Austrian three-tier regional list system. One can consider it a mixed compensatory system, so you could call it multi-seat MMP if you want to. It must have helped inspire Hungary’s two-tier MMP system. Austria has 183 MPs in 43 small multi-seat districts, with compensation at the level of the nine provinces and further compensation at the national level. Perhaps 90 parliamentary seats come from the little regional voting districts, but remainders of less than 0.5 quotients are carried upwards. Perhaps 65 seats come from provincial voting districts, and perhaps 25 from the federal level. If Austria had more territory, they might have added some single-seat districts in rural areas and ended up with Wayne Smith’s hybrid model above.
And Austria has an interesting “flexible open” (or semi-open) list model with optional personal preference votes. All worth further study.
Seed planted by Wilfred Day — 29 December 2005 @ 14:06
I am glad to know someone reads my footnotes! Just to clarify, it is only a “nominal” tier if the seats in the lower tier are won by votes cast for individual candidates (i.e., by name, with no party-list allocation within that tier). So, nominal-tier rules would include plurality, majority, or SNTV/limited vote. They could also be AV (IRV)/STV. But not a list-PR (or list plurality) formula.
The Dutch mixed-member proposal that I saw at a symposium we had at UCSD with some Dutch government officials and academics was for SNTV in the lower tier (so multi-seat nominal voting as in Palestine, except only one vote per voter) and a compensatory PR tier. I would certainly count that as MMP (albeit a rather odd form). I have not followed what has happened with the proposal since, although there was recently a thread on it at Make My Vote Count.
Anyway, if the nominal tier were to be multi-seat plurality (with the voter having as many votes as there are seats in the district), you would need a pretty big list-PR tier to compensate and provide a high degree of proportionality.
While I agree that a system with two-tier PR, as in Austria, is in the family of compensatory systems (as is MMP), I would not put it in the category of mixed-member system. After all, what is ‘mixed’ about MMP (or MMM) is that members are elected under different principles (nominal and usually plurality in the lower tier; and list and usually PR in the upper), not simply that they are elected in tiers with different district magnitudes.
Also, I would not call the Hungarian system MMP. It is quite disproportional, because the seats in the two tiers are allocated in parallel, albeit after adjusting the list-tier votes by adding “unused” votes from the nominal tier, which is itself two-round majority/plurality, and hence quite disproportional. (In Hungary, the PR tier is actually itself two-tier, for three tiers in all; the national tier compensates for the regional PR districts, but fully not for the SSDs–what a complex system!) Both the Hungarian and Italian systems are in between MMM and MMP, but if I had to put them in one box, it would be the former (and that is what we did in the Mixed-Member Electoral Systems volume; see the 2×2 table on p. 15).
Seed planted by MShugart — 29 December 2005 @ 16:50
[...] On December 28, after surveying the electoral system, I said that this is the sort of electoral system that would manufacture a majority for a party of around 45% if that party’s voters tended to vote the full slate in the multi-seat districts. However, given divisions in Fatah, I doubted that would happen. Referring to the factions in Fatah that re-united only after the deadline for registering candidates was extended, I noted: That is, some voters may vote only for the candidates they recognize as “Future†or “official†Fatah members, and not for the candidates of both factions. To the extent that happens, and Hamas voters are more unified, the result could still favor Hamas relative to Fatah in the votes-to-seats conversion. [...]
Scion grafted by Fruits and Votes » Blog Archive » Palestinian exit polls suggest Fatah lead but Hamas advantage — 25 January 2006 @ 18:42
[...] Remember, before the election, in surveying the electoral rules, I noted that a parallel mixed-member system in which the nominal tier was multi-seat plurality (in the form of multiple non-transferable vote) would tend to create a highly disproportional result. [...]
Scion grafted by Fruits and Votes » Blog Archive » The magnitude of the Hamas sweep: The electoral system did it — 27 January 2006 @ 16:14