# The Moorish Wanderer

## Who Will Get The Big Job?

Posted in Flash News, Morocco, Read & Heard by Zouhair ABH on December 10, 2012

5 candidates are competing to replace the dean of representatives and incumbent USFP premier, Abdelouahed Radi. 5 candidates with similar and different views, no doubt. And I thought this is the chance to assume, then prove political agents in Morocco are capable of rational strategies and decisions – and that will be put to the test, on the weekend 14-16th of december.

This is textbook game theory, where the game is as simple as it gets: during USFP’s convention, the delegates from all over Morocco need to vote for the candidate they believe can lead them to victory in 2016, or at least be in position to share some of the spoils a strong contender generates. I would like to think the centre-left party would do some introspective thinking about its political philosophy, its alliance strategy, but the would-be leaders would be hard-pressed to deliver results. So, setting aside the assumption of a selfless leader eady to sacrifice their (his) political future for the sake of the party, I would posit all 5 candidates have every incentive to go for -at least implicitely- an immediate positive return.

As mentioned before, the voting game is about declaring preferences. So it is quite possible the next USFP convention would end up in a deadlock – although it has a very low probability of happening:

$\mathbb{P}(deadlock)=1-\frac{\max {V}_{i=1}^5}{5!}$

(5! is the factorial, with a value 120, as there are 120 different combinations of listing preferences)

and even lower probability if USFP delegates are very adverse to rowdy convention outcomes (and recent history shows), ie. $u\left[\mathbb{P}(deadlock)\right]$ where u(.) is strictly concave to denote this risk aversion.

For each delegate thus, there is a ranking, I would like to add to a small constraint: the delegate has a strong preference for the first choice, their candidate, and then enunciate weak preferences for the remaining four:

$V_1 \succ V_2 \succeq V_3 \succeq V_4 \succeq V_5$

An equilibrium (the election outcome) does not necessarily mean a majority of delegates select the same first choice. To illustrate this, assume the rules of elections have been altered. Instead, the party convention will go through 5 ballots, each candidate is submitted to approval or rejection. The equilibrium here depends on the candidate order. This step is simply internalised by each and every delegate: they weigh in different outcomes, and eventually come up with a choice that maximizes a series of objectives (party victory, personal gain, idealistic aims, etc.) their respective preferences are solved using backward induction.

Since party convention rules for two ballots, preferences will be broken down in first-hand choices, and second-hand choices if the former fails.

Let us now consider more down-to-earth elements of this election: presumably, USFP needs a strong leader to measure up to Hamid Chabat, or Hakim Benchemas, or indeed the Head of Government, Abdelilah Benkirane. Meaning their next Secretary General has to adopt opposition-like strategies, including:

1/ Very active and visible on the media

2/ Established access to the same media

3/ A member of parliament (MP)

Point 2 and 3 are correlated – a member of parliament has a privileged access to mainstream media an outsider lacks. As it turns out, this rules out two candidates, Fathallah Oualalou and Mohamed Talbi (Zaidi, Lachguar and Malki being all three representatives) to be first choice for a whole lot of delegates. In fact, the weight party delegate allocate to this quality (being an MP) is roughly equal to the percentage of those delegates whose preferences are based upon that criterion: what is the probability of choosing a member of parliament as a first choice. answer: 60%. That is not to say 60% of the delegates will chose between Zaidi, Lachguar or Malki, but each delegate has a 60% chance of of choosing one of these candidates. But since it is expected some 1200 delegates will attend the convention, my assumption would be that 720 delegates will vote for one of the three ‘premium candidates’, and then scramble for the remaining 480 that votes primarily for the two other candidates, on the second ballot.

Let me make an assumption about the 480 ‘idealists’: presumably, a large percentage of those will vote for Fathallah Oualalou, and his votes will be crucial for the two parliamentary contenders. Notice I mention two, and not three. The proof is in three steps: first, assume all three candidates have equiprobability of getting the 720 votes – this means on average each one gets 240. A winner needs to get at least 361 additional votes, so the other 480 votes are not going to be split evenly across the first three contenders, which means only the first two of the parliamentary contenders really matter.

The crucial player in the two-ballots game is thus M. Oualalou: if he comes in second or first in the first ballot, he will be elected on the second,his support has every incentive to stick by him, in addition, support from his parliamentary competitor’s rivals will consolidate his lead. This is based on the assumption that delegates supporting a parliamentary candidate on the first ballot rank the other two behind M. Oualalou.

On the other hand, if M. Oualalou comes in as a third candidate, his support might make a difference. This is because one of the parliamentary contenders in the run-off could be too polarizing, and even support transferred from M. Oualalou might not be enough. This leads me to lay out some assumption about M. Driss Lachguar, whose own record shows he can be a serious contender, but his polarizing figure could produce a backlash and elevate another, ‘Dark Horse’ candidate to the Premier position.

In many respects, Mr. Lachguar is favoured to be USFP’s next boss: he is a member of parliament, has been involved in the decision to withdraw USFP from coalition talks and join in the Opposition, and rattled sabres over the appointment of M. Karim Guellab as Speaker of Parliament House. Yet many party delegate might not be interested to vote for him on the second ballot (that is, if he makes a first, or close second) and could vote for another, less illustrious candidate in an “Anti-Lachguar” stampede. And yet, there is a chance a Lachguar-Oualalou ticket might get a win, provided the following conditions:

1/ Lachguar supporters stick by him on the second ballot
2/ There is a common pool of Lachguar-Oualalou supporters
3/ Lachguar supporters are expected to cast slightly more votes than Oualalou’s

In fact, the minimum number of Oualalou supporters among the 480 delegates ready to switch on the second ballot, described earlier would be:

$240+480\times\alpha-\alpha^2\times (V_{FO},V_{DL})$

(this is assumed to differentiate between core Oualalou supporters and those likely to switch support to Lachguar

a simple FOC gives: $\max_\alpha V(FO,DL)= \dfrac{\partial V(FO,DL)}{\partial \alpha}=0$ yields $\hat{\alpha}=19.1\%$ so Lachguar only needs 20% of Oualalou’s supporters to throw their support behind him to win the ballot, should Lachguar come a close first or second.