By S. Sivasegaram –

Electoral reforms will be a bigger bone of contention in the drafting of the constitution than key issues facing the country. Each political party will be interested in securing its strength in future parliaments. Equally there will be personal ambitions at play. What is unfortunate about electoral reforms proposals thus far is that it they merely concern implications for the parliamentary strength of certain political parties and not democratic representation.

First Past the Post Scheme (1947-1977)

Until the passage of the constitution of 1978, the electoral scheme of the country was based on the “first past the post” (FPP) scheme for all elected bodies. It offers the benefit of stable parliamentary government in the context of a few serious rivals for power, but at the price of representation not being in fair proportion to the votes received by the parties and a severe bias against representation of all small political parties except those with regional power bases. The data below illustrates some of the worst disparities between the votes and seats secured by parties. For simplicity I have combined the data for the left parties contesting in their party names and that for the two main Tamil political parties of the time. The percentage of votes is shown after the party name and percentage of seats in parentheses.

(Note: SLFP refers to the MEP alliance in 1956; Left refers to all left parties contesting in their names; Tamil refers to combined TC & FP from 1947 to 1970 and the TNA in 1970. In 1947, independent candidates polled 29.2% of the vote and won 22.1% of the seats.)

Disparity existed all along between votes received and seats secured. The UNP suffered injustice in 1956 and 1970 and the SLFP even worse in 1977. This disparity is also distorted by strategic contests, electoral pacts and regional patterns of voting, mostly based on ethnicity. Since 1952 (results not shown), the FPP scheme has been disadvantageous to the loser among the main parties (and alliances). All small parties except Tamil nationalist parties— with a strong regional base in the North and East —suffered.

The data shows the Left as a net gainer— except in 1977 when it lost representation altogether. It was a marginal loser in 1960 March and June, and 1965 (results not shown). The performance of the Left between 1956 and 1970 is distorted by no-contest pacts with the SLFP. The results of 1956, 1970 and 1977 make a strong case against the FPP scheme where the winning party secured between 20 and 63% more seats than rightfully due, based on the proportion of votes gathered, while its main rival had between 70 and 80% less seats than due.

While the share of seats secured by the Tamil nationalist parties is more than their share of vote, the proportion of seats is less than what the percentage Tamil population would justify. Muslim representation was possible only as candidates of a Tamil or Sinhala nationalist party or as independents with the backing of a Tamil or Sinhala nationalist party, and there was no parliamentary political party with Muslim identity. The Hill Country Tamils who had reasonable— although less than proportional —representation in 1947 were since denied elected representation until 1977 when the granting of citizenship to a section of the population enabled the election of a single MP in 1977.

Transition to the Current Electoral Scheme

The method of election was changed under the constitution of 1978, for rather sinister considerations. The executive presidency was established in 1978 using the 83% strength of the UNP in the House. The method of election which enabled that majority in 1977 could under different conditions grant a similar mandate to another party or alliance which could reverse the changes introduced by the constitution of 1978 as well as add features unfavourable to the UNP. JR Jayewardene was aware of the risk, and designed the district-wise proportional representation (DPR) scheme— with a small national proportional representation (NPR) component —so that no party was likely to secure a sufficient majority to undo his constitution. But it also meant that, in years to come, no single party could secure an absolute majority in parliament. Subsequent elections showed that JRJ was correct in his calculations, so that the UNP never faced a debacle of the kind it did in 1956 or 1970. Even in 2010, when its share of votes was the poorest in party history, the UNP-led alliance secured 60 of 225 (or 26.7%) of the seats with 29.3% of the vote.

JRJ took no chances and introduced constitutional clauses requiring a referendum to amend parts of the constitution and a ruling of the Supreme Court on some others, as made clear by the Supreme Court ruling on the draft of the 19th Amendment in April 2015.

DPR, however, had its merits. It enabled the JVP to win seats in several districts and parties of the Muslim and Hill Country Tamil minorities to secure seats even in districts where their presence was of the order of 10%. The Sri Lanka Muslim Congress and the CWC benefitted most. Such electoral success also meant splits and a proliferation of parties reflecting personal rivalries rather than political differences. Also, vested interests bought into political parties by dubious means such as sponsorship of candidates and persuading a major party to nominate individuals named by them as candidates or to elect them on the “national list”.

Problems arose with DPR owing to the preferential vote to choose MPs from the electoral lists of political parties. The preferential vote heightened rivalries between competing members of a party so that, at times, intra-party rivalry and violence were far worse than inter-party rivalry and violence. Personal rivalry wrecked the unity of parties— especially the large parties —as each potential MP sought to outperform party colleagues.

Battles over Reforms

The clamour to revert to the “first past the post” (FPP) was strong even before the ouster of the UPFA regime in 2015. “Electoral reforms” were considered by the UPFA regime for a while, not in the interest of democracy but to soften intra party conflict and, rather cynically, to weaken smaller political groups which gained ascendency under DPR. Many a formula was suggested to placate allies of the two major parties representing minorities. There was no consensus as the Muslim and Hill Country Tamil parties would be big losers even with the best compromise between the DPR and the FPP. Allocation of a large share for the number elected based on DPR could have been a consolation. An arrangement was arrived at in mid June 2015 to increase the number of MPs from 225 to 237, with 145 elected on FPP basis at, 55 on DPR basis, and 37 on NPR basis. But disagreements persisted and there were fresh proposals for a yet-to-be-defined dual voting scheme.

DPR benefits the leading parties in a district but generally not the smaller parties. Increasing the number of seats allocated for NPR may help some of the smaller parties, as the national vote is the total of the votes gathered in all electorates by a party. Whatever the mix of DPR and FPP, it will hurt political parties of the Muslim and Hill Country Tamil minorities, mostly dispersed across the country. Tamil political parties will also be losers owing to the moving out of a sizeable section of Tamils from the North and East since the escalation of conflict. Lack of resources will prevent Tamil political parties of the North and East from fielding candidates outside, except perhaps in Colombo with a good number of Tamils. But they have so far avoided contesting outside the North and East.

DPR gave Muslim nationalist parties the chance to secure seats not only in the three districts of the Eastern Province with sizeable Muslim populations but also in the Districts of Mannar, Puttalam, Kandy, Colombo and Kalutara. The Hill Country Tamils boosted their representation in the Nuwara Eliya District and had the chance to secure seats in the Badulla District and, to a less extent, Kandy. The number of representatives of Hill Country Tamils will dwindle from the present seven to one or two, while that of Muslims will fall by a third or so unless Muslims are nominated under the NPR allocation of the major parties. Hill Country Tamil and Muslim nationalist political parties— which despite serious political flaws comprise the only political defence of these minorities against chauvinism —will suffer badly. A serious victim of the FPP scheme will be the JVP which, with its present share of votes, cannot secure even one seat on FPP basis or for that matter the DPR, with a 10% share of the vote in any district other than Colombo, Gampaha and Kurunegala, whereas the existing scheme would have let them secure around five seats by DPR even with around 5% of the votes going their way in each district. If all PR seats were allocated under NPR, the JVP will secure a significantly larger number of seats.

A worse victim of the FPP is the JHU, which lost its shine since it peak in 2004 with nine seats and needed the mercy of a bigger party to win three seats in 2010. Under a scheme combining FPP and DPR it cannot win any seats on its own. Yet it wants future parliamentary elections to be under the reformed scheme, as the JVP and the Muslim and Hill Country Tamil parties will be big losers and the JHU could hopefully secure a seat or two at the mercy of the UNP— or even the SLFP.

Even a significant increase in the share of DPR seats in parliament will not help a small party without adequate support in a district. (With 80 seats allocated to DPR, a party will typically require close to 10% of the vote in most districts to secure a DPR seat whereas, earlier, typically around 5%— even 3% in districts like Colombo and Gampaha with a large number of seats —could assure it a seat. If the NPR seats are set at 80, a small party with around 0.6% of the national vote has opportunity to secure an NPR seat and with 1% of the national vote could be almost certain of a seat.

Implications of Proposed Reforms

The analysis below is based on the hypothetical district wise voting pattern in Schedule 2 for two major alliances, three weaker parties and three minority parties with potential to secure parliamentary seats under FPP, DPR or NPR. Performance of other parties is considered insignificant for purpose of analysis. The pattern is based on voting patterns in General Elections and is not intended to reflect a true distribution of votes as the purpose here is to illustrate the implications of the reforms under consideration for parliamentary seats won by strong, weak and minority parties. The share of vote is considered only among parties likely to secure a seat under a PR scheme.

The FPP seats for a party was estimated based on a guesstimated strength of the party in the constituencies (which remain to be determined); DPR seats on its total vote and seats allocated in each district (which too is to be determined); and NPR seats on the national total votes received. Schedule 3 shows the number of seats won under FPP in a 240 seat parliament with 160 FPP seats and 80 PR seats. The accuracy of FPP seat estimates is unimportant for the present study of the impact of any likely PR scheme.

Schedule 4 shows the number of seats won by each party for two extreme cases of a total of 80 PR seats. In one, all PR seats are allocated district wise (DPR) and in the other all are allocated nationally (NPR). The DPR seats are analyzed district wise, based on votes as in Table 2 and PR seat allocation as listed in Table 3.

For purpose of comparison the FPP seat data— normalized to a total of 80 (half the seats estimated in Schedule 3 for a total of 160) —is presented on the last row of the schedule.

NPR is to the disadvantage of the leading national parties N1 & N2 and the minority party M3, and the benefit of the weaker national parties N3, N4 &N5. It has no effect on the minority parties M1 & M2. Thus, parties N3, N4 &N5 will be significant gainers while N1, N2 and M3 will be worse off as weightage is increased towards NPR so that N1, N2 and M3 will prefer a greater weight to DPR in any mix of PR seats comprising DPR and NPR. The difference in benefit to parties M1 and M2 cannot be very much for them to care about weightage. Only the weaker national parties N3, N4 and N5 benefit from assigning NPR a greater weight.

Shift of weightage from DPR to NPR will thus hurt parties with strong reliance on regional support while offering some benefit to all island parties which fall outside the first two in most electorates and thus secure only a few seats under DPR. However, for a party to benefit from NPR, it needs to contest in as many electorates as possible to garner votes to qualify for seats. Thus weaker parties will prefer a shift of weightage from DPR to NPR, and will be inclined to contest as many electorates as possible across the island to increase their national vote. Even then, voters could shun candidates of a relatively weak party despite sympathy for it, as stronger rivals will present it as a spoiler in a battle between leading candidates. Thus the smaller parties could, in reality, be bigger losers than the analysis shows, especially because they cannot always field credibly strong candidates in all districts; and, even if they do, potential supporters are often tempted to vote for a party that is least unacceptable to them, among likely winners.

The Case for Dual Ballot Paper

It is clear from the foregoing that the FPP scheme is heavily loaded against any political identity, be it ethno-religious or ideological, and only proportional representation based entirely on the national vote can do justice to minority political groups of any kind. The downside of the DPR scheme was seen over the past 35 years in various elections. While an island-wide PR scheme will amplify some of the problems, superposing a PR scheme— either on district basis or on national basis —on an FPP scheme as the existing proposals indicate, will, given the mechanics of FPP elections, deny representation of supporters of a political minority party.

Thus for the PR component of the votes cast to ensure that the smaller political parties have a fair deal but without denying some of the benefits of the FPP scheme, two important features need to be ensured in PR voting. Firstly, all PR is based on the national vote and, secondly, the PR vote is independent of the FPP vote in any constituency.

In short, one should be able to vote for a candidate to represent one’s electorate independently of one’s political preference, which need not coincide with any candidate’s political identity. In other words, one should be able to vote for a party even where the party has not fielded a candidate. This is possible with a dual ballot paper: a constituency ballot to elect the constituency MP; and a national ballot to elect MPs nationally. I will use the terms Electorate Vote and Party Vote, as they do in Germany and in New Zealand, to refer to the two parts of the ballot paper.

A hypothetical vote distribution under a dual ballot scheme is used in Schedule 5. It is assumed that around 25% of Tamils (casting around 250,000 votes) and that 60% of the Muslims and Hill Country Tamils (casting around 480,000 and 240,000 votes, respectively) live outside regions where their respective ethnic parties normally contest parliamentary elections. It is also assumed that around half of them opt for an ethnic party on the national ballot while the rest vote for all island parties. The share of the national minority vote received by the major parties in regions where an ethnic party does not contest is assumed to decrease by a large fraction for Party N1 and by smaller fractions for Parties N2 & N3, while N4 and N5 have no significant attraction for ethnic voters. A smaller national party N6 is introduced since it is possible for a smaller party to gather sufficient Party Votes to enter the fray, based on frustration with leading national as well as minority parties. An erosion of Sinhalese votes is also assumed for N1, N2 and N3 in favour of N5 and N6, as several voters who opt for a likely winner under FPP in the electorate will use more political and less personal criteria in casting the Party Vote.

Schedule 6 shows the number of seats secured by the parties under the dual ballot scheme.

Representation of ethnic minorities will, despite the improvement offered by the dual vote system, be less than what corresponds to their population. For example, Muslims comprising 9% of the population will improve their representation from 7 to 9, to reach just under 4% of the 240 seats. The Hill Country Tamils who form 4.5% of the population will improve representation from 3 to 5 seats, to reach 2% of the seats. The Tamils of the North and East lose out a little with around 9.5% of the seats for a population of over 11%. The picture will not be substantially different even if an ethnic minority chooses to cast its party ballot only for an ethnic political party, because Parliament is dominated by FPP MPs. One national party or another may nominate a few ethnic minority MPs from their Party List but that cannot offset the loss due to FPP domination of the election process.

Ethnic considerations, although seemingly undesirable for national harmony, are inevitable in an ethnically dominated political climate and amid electoral reforms that are certain to minimize the role of national minorities on policy making. The dual ballot paper scheme is a fair, although inadequate, way to compensate loss of representation in a predominantly FPP scheme.

Without separation of the election processes for Electorate and Party, not only any mix of DPR and NPR, even a boost in the number of PR seats to match the FPP seats will deny small political parties, especially national political parties. It is important to note that the dual ballot scheme assures representation even for small political parties with a modest support base by giving them opportunity to gather votes nationally without fielding candidates in all or even most electorates.

The dual ballot scheme is more democratic than PR schemes based on a single ballot, and its benefits will improve with growing social awareness of the voters and appreciation of how the scheme works. Besides being more democratic it offers other benefits for the electoral process.

1. It lets a voter express support for a party in a constituency while rejecting its candidate with whom the voter is unhappy and vice versa. Thus a political party is compensated for the damage done by a personally unpopular candidate and, on the other hand, a good candidate could get elected despite affiliation to a party that is unpopular.

2. It allows small parties to gather votes in electorates where their support base is weak but significant. Unlike with the single ballot, the dual ballot enables a small party to avoid fielding candidates for the sake of building its national vote, and thereby waste party and public resources and distort the outcome of an election, for only a marginal benefit.

Issues and Political Benefits

The dual ballot scheme has its problems. While the Electorate Ballot is straightforward, the Party Ballot may need to accommodate too many political parties to the point of being awkward.

The number of parties on the Party Ballot could be minimized by confining it to (a) parties already represented in Parliament; (b) political parties or groups fielding candidates in no fewer than a quarter of the constituencies so that parties that are unlikely to secure adequate votes nationally will not appear on the Electorate Ballot. To help clarity, the parties could be listed on the ballot paper in the order of their strength in the last parliament, followed by unrepresented parties listed in alphabetic order.

An unfortunate aspect of increased representation of minority nationalities has been that parliamentary representation has failed to bring political or social gain for the nationalities since the interests of parliamentary parties have little to do with the aspirations of communities. Parliamentary membership has been abused for personal gain and political favours including cabinet posts when governments badly need a majority or have only a fragile majority. This problem can only be cured by the voters publicly questioning their leaders.

A PR scheme is more democratic and politically sensible than electorate based schemes in matters of legislation and national policy and will accommodate a wide spectrum of political views in parliament. The FPP scheme is, however, effective in addressing issues at grassroots level and ensuring that an MP is attentive to the concerns of the people. If meaningful devolution of power is achieved, it may be possible to use only a PR scheme for elections to the central parliament and a mix of PR and FPP for Provincial Councils. The FPP alone could be used for local government. But no scheme can be democratic if the electoral process remains corrupt as it has been since 1978.

A likely beneficial outcome of the dual ballot scheme is that minority parties, to strengthen their parliamentary presence, will need to appeal to voters outside their regions. This means that they will need to adopt a less parochial approach that will bring them to terms with the reality that their respective communities are part of a multi-ethnic polity.

However, no electoral scheme in an unequal society can assure social justice. The purpose of pressing for proportional representation here is to secure a voice for sections of population in fora where there is conscious attempt to deny political minority representation.