over 9 years ago by Manthri.lk - Web Team under in ஆய்வறிக்கை


By Rohan Samarajiva & Sujata Gamage


Software is complex.  One set of instructions can contradict another, and cause the whole thing to freeze, or work differently from what was intended.  These bad outcomes are said to be caused by "bugs."  Debugging is a major part of software development.  The open source approach to software was pioneered by Linus Torvalds.  He wrote the first version of Linux and put it out to the community along with the source code.  They looked for the bugs and fixed them.  He converted users into co-developers.  It is low-cost, fast, and results in better software. 


For the success of the 100-day Program, we believe there is no alternative to the open-source approach.   Without this approach electoral reform will be still born.


Electoral reforms

The manifesto put forward to the people by the Common Candidate at the Presidential Election was an ambitious one.  But it was what brought the multiple parties and organizations together and what the people voted for.  Without its full implementation, trust in the political system and in the words of politicians cannot be restored.  Without the rebuilding of trust, there is little hope for the new political culture that President Sirisena is committed to build.


Back in January, we concluded that electoral system reform was the 100-day promise that was most likely to fall by the wayside.  It was complex and needed the consent of all the parties, large and small.  We were surprised that a detailed commitment to electoral reform had made it into the 100-day program and that it had been agreed to by so many parties and organizations.  Seeing this as a unique opportunity and challenge, we chose electoral reform as the focus of our contribution to the President’s mission of building a new political culture.


On the 2nd of March when the All-Party Committee was due to have submitted its report electoral reforms, we presented our results to a gathering of opinion leaders at the Nagarodaya Center in Colombo.  By this time, the political parties were under pressure from multiple sources.  The most prominent was the Ven. Maduluwawe Sobitha thero, the moral leader of the coalition that toppled President Rajapaksa.  Another factor was the SLFP’s stance that changes to the Presidential powers had to be done in tandem with electoral reforms.  Our original intention was not to invite politicians, but close to the date, we received requests to which we acceded.  The report on electoral reforms was thus presented on March 2nd as promised, though not by the all-party committee.


Rationale for reform

The Common Candidate’s manifesto stated that “the existing electoral system is a mainspring of corruption and violence. Candidates have to spend a colossal sum of money due to the preferential system.”  Former Minister Kumara Welgama saying on the record that he spent LKR 800 lakh (80 million) for his last campaign (2010) in the Kalutara District (1,576 sq. km; 897,349 voters) illustrates the problem. 


The quantum of funds required to mount an effective campaign constitutes a barrier to entry into politics by those who are not in a position to spend their own wealth or are unwilling to make the necessary moral compromises.  Corruption is a necessary outcome of the current system.


The present system creates incentives for internecine violence among candidates from the same party.  The introduction of preferential votes improved the previous system of proportional representation (PR) which assigned extraordinary power to political party bosses unconstrained by inner-party democracy, but drove up costs and violence.



We took the specific language used by the Common Candidate as our guide:


I guarantee the abolition of the preferential system and will ensure that every electorate will have a Member of Parliament of its own.  The new electoral system will be a combination of the first-past-the post system and the proportional representation of defeated candidates. Since the total composition of Parliament would not change by this proposal, I would be able to get the agreement of all political parties represented in Parliament for the change.  Further, wastage and clashes could be minimized since electoral campaigns would be limited to single electorates.


For comparison, we analyzed the New Zealand Mixed Member Representation (MMR) system (NZ Model) that has been in effect since 1996.  A simplified and improved version of the extremely complicated German system, it is among the best PR systems and does not rely on preferential votes.  It yields a seat allocation in the 120-seat Parliament that is reflective of percentage of votes garnered.  In certain circumstances a few additional seats may have to be given, making the size of Parliament variable.  Fifty out of 120 MPs come from the “other persons’ lists” given by political parties. 


The other persons’ list, introduced to Sri Lanka by the Local Authorities Elections (Amendment) Act (22 of 2012), is unpopular with politicians who face the hustings.  They do not like 30 percent of the seats in local authorities being given to “other persons” anointed by distant political bosses.  Voters are short-changed because inner party democracy is weak to non-existent.  Interestingly, the recommendations of the Select Committee (SC) chaired by Hon. Dinesh Gunawardene provided an alternative whereby all votes are given value, even those not cast for the candidate winning the electorate.


Analysis of options

The SC proposed that 140 seats should be decided on the basis of first-past-the-post (FPP) voting; and that 70 should be decided on the basis of allocating the unused votes or the remainder, to parties that had gained more than five percent of the vote.  Fifteen seats would be distributed based on PR to parties for “National List” appointments. 


We applied the NZ model and three variations of the SC model to the 2000, 2001, 2004 and 2010 general-election results.


·       The NZ model has never yielded a single-party government for New Zealand.  It is likely to yield a similar result in Sri Lanka, requiring the major parties to enter into alliances with small parties who will not be permanently exiled into the opposition as a result.  Our simulation saw the winning party gaining 102 seats in 2000, 103 in 2001, 103 in 2004 and 136 in 2010, an unusual year.  The NZ model gives considerable power to political parties.  It does not accommodate defeated candidates or “best losers.”  A fixed number of seats in Parliament is not guaranteed.    


·       The first variation of the SC model (SC1) is for 125 seats to be determined by FPP voting; 75 on basis of remainder votes and 25 from the National List based on PR.  The delimitation of 125 electorates will yield large units.  This may require an additional rule to accommodate multi-ethnic representation in areas such as Nuwara Eliya-Maskeliya and Colombo Central (multi-member constituencies before 1977).  This is feasible within the allocation of the 75 seats to “best losers” based on remainders.  Here all the criteria specified by the Common Candidate are met.  The small parties will continue to be represented in Parliament, more or less in line with the proportion of votes they garner.  Single-party governments are unlikely except in unusual circumstances such as the 2010 post-war election.  The National List allocation is not very different from the present and can accommodate a rule such as “every party appointing more than one National-List MP must appoint at least one woman.”


·       SC2 is the Select Committee’s recommendation:  140 seats determined by FPP voting; 70 on basis of remainder votes and 15 from National List based on PR.  The delimitation of 140 electorates will yield large units of around 1 lakh voters, though smaller than in SC1.  This may require an additional rule to accommodate multi-ethnic representation in places such as Harispattuva and Pottuvil.  This is feasible within the allocation of 70 seats based on remainder votes.  Here too all the criteria specified by the Common Candidate are met.  Small parties will continue to be represented in Parliament, more or less in line with their proportion of votes.  There is a greater likelihood of single-party governments, though the winner will enjoy smaller margins (one more than 113 seats in 2000; two more in 2001; no majority in 2004; and 20 more in 2010).  The National-List allocation is half of what exists now and may not be large enough to ensure adequate representation of women and professionals.  


·       The third variation of the SC Model (SC3) requires the conversion of the current number of polling divisions (160) to the same number of electorates by a delimitation commission.  Sixty seats will be allocated on the basis of remainder votes and five will be allocated to small parties.  The National List has to go.  The delimitation will yield smaller units.  The Common Candidate’s criteria are satisfied.  The increase in the FPP component is disadvantageous to the small parties.  The problem can be somewhat alleviated by implementing the Hakeem-Muthu Sivalingam proposal to the SC for an additional five-seat allocation.  Single-party governments will result (majority of six in 2000; seven in 2001; three even in the tight 2004 election; and 27 in 2010).  The loss of the National List makes adequate representation of women and professionals a challenge.  


The data sheets and methodology have been made available at Manthri.lk (click here) so that interested persons can replicate the results or develop their own scenarios and solutions.  We are sure the contributions of committed citizens will remove whatever weaknesses our solution contains and yield an electoral system worthy of the aspirations of our people and best suited to our conditions.


You can tweet the authors Prof. Rohan Samarajiva @samarajiva & Dr. Sujatha Gamage @sujatagamage


Please leave your comments below, alternatively email us [email protected] or tweet us @manthrilk or find us on facebook.com/manthrilk


This article is reproduced by Manthri.lk to encourage the debate on electoral reform. 

Manthri.lk has neither contributed or endorsed any of the views provided. 

(Check out MaithriMeter.lk - Scoring 100 days out of 100)