Friday, December 15, 2017

Run-off system more democratic


Nigel AR Henry is a data scientist and pollster. He is former member of the Obama for America analytics team, and is currently the Chief Analyst and founder of Solution by Simulation.

Mark Fraser

The majority of the public discourse surrounding the controversial proposed run-off election system holds that it is undemocratic and distorts the intended will of the population.  

The run-off system calls for a second election if no candidate receives at least 50 per cent of the vote.  Most pointedly, commentators hypothesise that the run-off system will do little to loosen the stranglehold of our democracy by the two dominant parties - which is perhaps of particular significance given our population composed of two major ethnic groups in near-equal proportions.  Scrutiny of the numbers and empirical data elsewhere in the world argue against this view.


Effect on third parties

It is well established in theory and practice that the FPTP system entrenches a two-party system.  This is the reality in our local context, where voters complain of having to choose between the lesser of two evils in order to prevent the less preferred option from winning.  In cases when a significant share of the electorate did select a third-party -  22 per cent for the ONR in 1981, and 23 per cent for the COP in 2007 - the third party was left without a seat; very undemocratic results indeed.

Rather than concentrating power in two parties, the run-off system gives significant third parties a chance to win.

France, one of the few nations using the run-off system to elect representatives to single-member constituencies, has a multi-party system that is sustained through the run-off system.  Voters on the left of the political spectrum have the option to vote for any one of a number of left-wing parties, all of which field candidates in the first round.  In the second round, these voters then typically vote for whichever of the left-wing parties received the most votes, providing a consolidated vote against the right-wing in the second round.  

The level of understanding of this system is so strong that when two candidates of left-wing parties advance to the second round, the party who received less votes withdraws.

If the run-off system were in place in 2007, then-COP political leader Winston Dookeran would have likely been re-elected, and the COP would have likely picked up a second seat as well.  Mr Dookeran lost St Augustine by 617 votes, and would have qualified for a run-off against UNC candidate Vasant Bharath.  Of the 4,309 voters (24 per cent) in St Augustine who voted for the PNM, many would no doubt stay home in a subsequent run-off, but many would have returned to vote either for the PNM or the UNC.  

Although we lack polling to know for certain whether more PNM supporters would have preferred the COP to the UNC, many data points including analysis of 2010 polling strongly suggest that Mr Dookeran would have had no trouble making up the difference (especially since many UNC voters would also have stayed at home or switched their vote after being voted out of power).  The other seat in a similar situation was Tabaquite.

Most would agree that COP’s 23 per cent translating to two seats is a more democratic result than the zero seats obtained in the current system.  Furthermore, if voters knew of this possibility, perhaps even more disaffected UNC or PNM voters would have cast their votes for the COP to make a statement through the ballot box, without fear that their votes would unwittingly put their least-preferred option (the PNM or the UNC respectively) into power.

Smooth transfer of power

The PNM made a valid point by highlighting the possibility of losing candidates retaining office and a losing party continuing to govern pending runoff elections.  For a potential resolution to this dilemma we can look to the precedent of the United States.

The US allows each of its 50 states to decide how they would elect representatives to its national legislature (the “Congress”).  While the majority have retained FPTP, ten states use run-off voting.  

Depending on the state, these can be as late as the ninth week after the election.  Congress does not wait for these run-offs to be concluded before it sits and elects a new Speaker of the House.  Losers immediately demit office, while winners form their caucuses and take up committee assignments.  Seats subject to a run-off merely remain vacant until the results are certified.

The analogue in our situation would be for the rules regarding the appointment of the Prime Minister to remain unchanged.  Following the election, members of Parliament elected outright (ie, those who gained a majority of ballots cast) would write to the President informing him/her of which member (elected outright) they support as leader.  

If any member commands the majority of total seats (in other words at least 21 members), then that member would be invited to form the Government.  

If no member yet commands this level of support, only then a new Prime Minister would await the conclusion of the run-offs.

This period, while somewhat uncertain, would be no more perilous than the two-week period immediately following the infamous 18-18 result of the 2001 general election.

I would argue that with this modification, the unlikely chance of “uncertainty” is a small price to guarantee the certainty that comes with a government democratically elected by popular mandate.

Nigel AR Henry is a data scientist and pollster.  He is former member of the Obama for America analytics team, and is currently the Chief Analyst and founder of Solution by Simulation.