The Moorish Wanderer

Elections, Vote Spread and District Representation

Posted in Intikhabates-Elections, Moroccan History & Sociology, Morocco, Read & Heard by Zouhair ABH on October 19, 2012

The post was going to be about yet another (boring) summary of parliamentary politics, supplemented perhaps with equally boring statistics about why opposition parties find it so though to get a majority, regardless of institutional motives. My assumption about the whole thing in Moroccan politics is that institutional barriers – corruption, gerrymandering the historical tampering of the Ministry of Interior in each and every election up to, say 2002, are secondary to statistical realities.

First off, consider the number of invalidated ballots between 1963 and 2011, both in absolute terms and relative to valid ballots. It went from 3.67% in 1963, to 24.4% in 2011. The last Tangier By-election produced a little over 26% of invalidated ballots. Nonetheless, I for one would buy into the ‘nihilist’ viewpoint on turnout and invalidated ballots: by every measure of vote dispersion, these have proven to be potential vote winners, ‘King-makers’ as a matter of fact.

Multipartism hurts strong majorities: between 1997 and 2007, spread votes (the percentage of invalidated ballots relative to votes’ dispersion normalised to the number of competing parties) increased significantly, and could have been higher if more political parties participated in the 2011 Elections.

Let me explain: in all the ballot systems the Moroccan elections tried over the past half a century, the common element was for a candidate (or a list of candidates) to get most votes, but also to make their margin is large enough. Roughly speaking, Block Vote, Party Block Vote and Proportional Representation are conditioned on four factors: the winning majority, the margin with respect to the second candidate, the minor’s performance, i.e. the closest contender to the electoral threshold, and finally the number of competing candidates.

The two last factors seem contradictory, but in most districts, the cumulative share of ultra-minor parties (those with less than 6% of the total vote) is close to or larger than the candidate list with a plurality of votes. To sum up, the politics of legislative elections between large and small parties is that of a zero-sum game. Every vote cast in favour of small candidate lists tend to increase voter dispersion, and deprive larger, national parties from getting a seat on the margin, especially so when the competing lists have some ideological affinities.

Consider the first general election of 1963: Invalidated votes could have bridged about a third of aggregate dispersion of electoral votes.

In 5 provinces, these votes could have decided between the contenders, meaning the election was close enough for the invalidated ballots to matter: these provinces accounted for 31 seats out of 144 in the new Parliament, and could have delivered an absolute majority to a UNFP-Istiqlal coalition, or indeed consolidate FDIC’s own plurality. And there goes my argument: aside from gerrymandering, voter intimidation and disenfranchisement, the surest way to ‘control’ the outcome of an election at the local level remains more effective, more discrete and less controversial than any other strategy to manipulate elections. After all, representatives of Interior are registrars in all polling stations, and depending on the district, they have the final say as to how a ballot should be examined. If anything, there goes hard proof of election meddling: exceptionally high percentages of invalidated votes can be reasonably considered to be an indicator.

There is an ever more insidious way to manipulate elections, and it has shown its effectiveness in 2002: there were about 26 parties competing for 295 seats on local ballot, and as a result, the dispersion of votes increased, rendering invalidated votes even more important in determining the winner of a closely (and bitterly, like the Rabat 2007 Election) contested election. The 2002 Election outcome was perhaps the best argument for election reforms. Speaking of which, I have found some interesting results as to how the number of seats should be allocated, given some nation-wide indicators.

It is pretty much a given to allocate the most populated districts with the largest number of seats. Yet there are some instances where small districts get about the same number of seats than those in metropolitan areas. Consider Azilal and El-Jadida districts: both have been allocated 6 seats each, yet El Jadida has twice as many registered voters. Same goes for Oued-Dahab (registered population: 19,000) with two seats, same as Tata (registered population: 60,000). So the proposed idea is to get rid of arbitrary distribution of seats, and instead use the nation-wide distribution to get a standard figure per seat: 43,000 votes per seat (plus or minus 139 votes). Assuming a fixed minimum number of seats per district (there are about 16 districts out of 92 with a registered population of less than 43,000) the number of representatives elected on local ballot can be brought down to 282 or 289 (depending on the minimum number of seats)

 Local  |Votes/|Min |Max |Spread|Seats|
 Ballot |Seat  |Seat|Seat|Votes |Total|
--------+------+----+----+------+-----|
Existing|43,847|  2 |  9 |18,626| 305 |
--------+------+----+----+------+-----|
Option 1|43,391|  1 | 10 | 139  | 282 |
--------+------+----+----+------+-----|
Option 2|43,391|  2 | 10 | 139  | 289 |
--------------------------------------

In electoral terms, this means 34 seats are going to be deflated to 16, so as to get near-commensurate representation.

As it happens, these changes have a net neutral effect on the largest parties’ electoral advantages – this is particularly true given the fact that most gains are centered around the Souss-Massa, Tansift-Haouz and Doukkala regions, all of which have a relatively homogeneous representation in Parliament (most of the big political parties have at least one representative from their districts) yet the changes operated prior to the general election last year were modest.

As one looks upon the districts with most losses, those in the South for instance, their over-representation is tightly linked to their turnout: Taounate had a 45% turnout and Azilal a little under 70%. As for the Southern districts, their average turnout was around 62%. So this is perhaps the internal logic to the distribution of seats per district: those regions with higher turnouts tend to be allocated disproportionately larger number of seats, even as their populations (registered or otherwise) are smaller than nationwide means. A rational politician (and Istiqlal, USFP and RNI have been quite clever at this) do their best to control these districts – because they get seats with a minimal number of votes.

There are 55 seats eliminated from 38 districts, and 32 seats added to 22 districts.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: