Basic assumptions: 1. Three relevant contenders: TB, MG, CA. 2. There is no first‐round absolute winner. 3. All contenders win enough votes in the first round to give them bargaining power in the second round i.e. either of the three can change the balance of power by transferring their votes in a potential agreement. Critical assumptions: 1. TB is not interested in the chief‐of‐government (PM) position. 2. Both MG and CA derive some utility from the PM position, albeit smaller than the President (PR) position. The rationale behind these assumptions goes as follows: TB once in a first‐rank position will certainly not be interested in a second‐rank position. Moreover, staying away from governance might bring him back in power in the next round of presidential elections (2014). The other two contenders will certainly be marginalized by their own parties if they fail to win. But securing the PM position might still preserve their leader status and provide them same bargaining power within their own political structures. Following these assumptions, one can draw up a sequential, two‐stage, non‐repetitive game. Starting with the final stage of the game the potential outcomes would look as follows:
1:TB,MG
2:TB,CA
3:MG,CA
TB:PR = 100 MG:PM = 50 CA = 0
TB:PM = 0
TB = 0
TB:PR = 100
TB:PM = 0
TB = 0
TB = 0
TB = 0
MG:PR= 100
MG:PR= 100
MG = 0
MG = 0
MG:PM = 50
MG:PR = 100
MG:PM = 50
CA = 0
CA: PM = 50
CA: PM = 50
CA:PR = 100
CA: PR = 100
CA: PM = 50
CA: PR = 100
The values 50 and 100 are chosen arbitrarily to reflect the level of utility achieved by a player in a particular outcome. A simple analysis shows that the outcome with TB president has purely statistical 2/8 = 25% odds, while the outcome with MG respectively CA president have 3/8 = 37,5% odds each. Evidently, the outcome with a president other than TB has 75% odds. These odds are statistical in the sense that this is a scenario when Nature would be allowed to choose the outcome i.e. when the players do not employ specific strategies that would eventually improve their chances of winning.
Let us turn to some of the potential strategies available each of the three players. Obviously, although the structure of the game is relatively simple (two‐stage, unrepeated) the range of strategies available makes it very difficult for a thorough analysis. If TB lures the other two into not cooperating with each other in the first round, the game will end up in the scenario 1 and 2. If they do cooperate, there is an option that they both might get elected, and not TB i.e. scenario 3. Strategies: S1 – cooperate; S2 – Do not cooperate Game 1: TB vs CA TB cooperate TB no‐coop CA coop (100, 50) (50, 0) CA nocoop (0, 50)) (50, 0) Game 2: TB vs MG: similar play‐offs with TB vs CA Game 3: CA vs MG MG coop MG no‐coop CA coop (100, 50) or (50,100) (50, 0) CA no‐coop (0, 50) (50,50) If CA cooperates he surely gets 50. If he does not coop, he will get either 100 or 0. Game 1 has a Nash equilibrium (i.e. coop‐coop) which is also Pareto optimal. Since the alternative presented by Game 3 might offer a similar outcome, but conditioned on some other factors, I would infer that there is still a higher chance that if TB passes in the second round, the other two would rather prefer a prime‐ministerial position with TB president, as opposed to the alternative of forging an alliance with the remaining runner‐up. This, of course, will alter the initial odds, i.e. when the game is played by Nature’s choice.