Kåre Vollan explores how Nepal’s electoral system was reformed following the Comprehensive Peace Accord and the introduction of the 2007 Interim Constitution. Whilst the resultant mixed system of proportional representation and first-past-the-post brought some progress for greater representation of historically marginalised groups, compromises have been brokered that have allowed privileged groups to maintain important and consequential advantages.
Electoral systems and political representation in post-war Nepal
The electoral system for the Constituent Assembly described in the Interim Constitution referred to the excluded groups that had been identified in the CPA. The mixed system it specified consisted of 240 members elected in single-member constituencies by FPTP and 335 elected through PR. Article 63(4) described how inclusiveness would be pursued for each system: in selecting FPTP candidates political parties would ‘take into account the principle of inclusiveness’, while in selecting PR candidates they would ‘ensure proportional representation of women, Dalit, oppressed communities/ indigenous peoples, backward regions, Madhesi and other groups, as provided in law’.
The electoral law of June 2007 laid out requirements for candidate lists and how the results of voting would be translated into seats. However, ‘other groups’ was interpreted to mean all groups not explicitly identified, which included the elite hill castes that had dominated Nepali political life for centuries, and not as ‘other excluded groups’, such as Muslims. In other words, the hill castes that would already be grossly over-represented through FPTP were also given a quota in the PR election.
Two elections were held for two different Constituent Assemblies in the interim period, in 2008 and 2013. Both these polls confirmed that hill (effectively ‘upper’) caste men have a large advantage in the FPTP election, whereas the quotas in the PR elections worked to some extent as intended. Dalits and women won larger representation in the PR elections, although the PR quota for hill castes and men effectively capped their representation. The hill castes increased their share of FPTP candidates from 41 to 55 per cent between 2008 and 2013. In PR, they got around their proportional share, as expected. Dalits won 2.9 per cent of FPTP seats in 2008 and just 0.8 per cent in 2013, showing that they need special arrangements to support their parliamentary representation.
While the analysis broadly shows hill castes are politically included, and Dalits and Muslims excluded, closer and longer-term study (see Table 2) also reveals significant variation in inclusion or exclusion within Janajati groups and Madhesi castes. The 2011 Census enumerated 125 castes or ethnic groups that can be classified within these larger categories. Detailed examination of these data by the author (see Vollan 2015) shows how these groups have fared in FPTP elections since 1991, and which groups need affirmative action in national elections and which do not.
The analysis shows that the included groups (ie hill castes as well as some Janajati and some Madhesi groups) that make up around 55 per cent of the population won between 87 and 91 per cent of seats in four of the five FPTP national elections between 1991 and 2013. The exception was in 2008, when their share dropped to 78 per cent, mainly due to the strong showing by the Maoists, who had a more diverse set of candidates even in the FPTP part of the election. The excluded groups who make up the remaining 45 per cent of the population, on the other hand, won between nine and 13 per cent of the seats, again except in the 2008 when they won 22 per cent – although this was still only around half of their proportional share.
This suggests that the proportional quotas would have been more effective if targeted towards the politically excluded groups, rather than to all groups, including the hill castes. Since the quotas in 2008 and 2013 were extremely complicated for parties to adhere to, more targeted quotas would also help simplify the system. One alternative would have been to simply require 45 per cent of those elected to come from politically excluded groups.
The 2015 Constitution was an opportunity to simplify the representation of groups in the PR elections by defining special representation only for otherwise excluded groups. There are various possible explanations as to why this did not happen. Madhesi castes and Janajatis might have been afraid that their total share would go down if only those subgroups in need of a quota were given one. There was also clearly an element of the ‘cream’ of each group protecting their existing privileges. For instance, the Yadavs of the Madhesi caste group would argue strongly for Madhesi representation in comparison to hill castes, but would be less willing to share their representation among all 28 Madhesi castes.
The new constitution has retained a mixed parallel system, but now with a majority (60 per cent) of FPTP seats, which reduces its inclusiveness. The principle of quotas for everybody was kept intact for PR while expanding the list of groups to include Muslims and Tharus – the latter being another politically excluded group. Without removing the privileged groups from the quota system, the new PR provision in fact makes the system even more complicated, while rendering the targeting of excluded groups ineffective.
The quota provisions introduced in the Interim Constitution made the Constituent Assemblies of 2008 and 2013 more diverse than any previous parliament in Nepal, particularly for Dalits and women. Group consciousness was high in the first Constituent Assembly, and women and Janajatis formed strong caucuses across parties. Elements of inclusion were prominent in the thematic committees drafting concept papers to inform assembly deliberations in 2008 and 2009. But, as time went by, and the final provisions were negotiated, inclusiveness and representation became less important. It remains to be seen if the complicated quota system will be implemented in good faith and will be sustainable over time.