Professor Arend Lijphart presented a lecture, “Consensus Decision Making”, at the Linen Hall Library. The event was hosted by Peter Emerson of the de Borda Institute.
Lijphart said that in his early years, he thought that for divided societies, majoritarian (first-past-the-post/FPTP) democracy was desirable but an ill-affordable luxury. He has now come to believe that FPTP is not even desirable, explaining his view in his recent book, Patterns of Democracy
He conducted research comparing FPTP versus coalition-based democracies. He found no large difference between the two types. The conclusion is that while the evidence does not particularly favour one type over another, one cannot dismiss coalition-based democracies as inferior to majoritarian governments.
Lijphart discussed the struggle to have proportionality accepted within electoral systems. Indeed, instead of introducing PR for its merits, it has been implemented defensively. One example was using PR in the UK to assure Conservatives during the expansion of the franchise to women. Another example is the use of two-seat constituencies in Chile, which assures the minority Conservative Party of seats (but which Lijphart argues contorts proportionality towards a false two-party system).
Lijphart cited a debate on PR that was published by the journal Representation. One of Lijphart’s objections to FPTP is that the majority Government has not usually represented the majority (popular will) of the people, specifically citing the case of the UK. But to me this has more to do with the greater-than-two-party system, which causes a split Opposition vote. While using FPTP in a greater-than-two-party system may create problems of fair representation, introducing PR in an established two-party system (e.g. USA) could also introduce its own problems.
He also argued that FPTP-based democracies suffer more from the lack of turnover in incumbent representatives, citing the USA as the worst case. For me, however, in the USA this probably has more to do with the fact that incumbents control the constituency boundaries, which was highlighted by Lijphart himself during the lecture.
He said that if PR was so unpopular by a simple majority of its elected representatives, then the larger parties would conspire to squeeze out the smaller parties, and that this has not been the case. Yet this made me think of the Good Friday Agreement and exclusion of non-unionist/non-nationalist parties in the decision making of the Northern Ireland Assembly. I queried Lijphart on this point in the Q&A session (see below).
More academically, Lijphart cited the work of Anthony Downs, who defended the merits of a two-party system: the consensual fight for the centre ground, median voter, encouraging moderation. This has been countered by Bing Powell, in his work, “Elections as a Test of Democracy”. Powell’s research shows that consensus-based democracies have a better link between the largest party and the median voter.
Donald Horowitz does not believe in PR but in the Alternative Vote. Horowitz argues that PR does not produce moderate outcomes; in fact, PR can exacerbate them. Lijphart counters this by saying that you can use types of list systems [e.g. of Belgium “Alliances”?]. Lijphart does not see why any party would not use a list system to appeal to a wider audience. He cited examples of parties using “coalition tickets”, with a list of candidates representing a diversity of society. In the Q&A session, I remarked that in a divided society such as Northern Ireland, parties could adapt to an imposed list system, for all societal segments except that which divides them. I said that I did not see either the UUP or SDLP moving towards a representative sample of society. The exception was the Alliance Party, which at every election actually presents a list of candidates from Protestant and Catholic backgrounds.
Lijphart argued that FPTP is more dangerous than PR for the task of compromise. Under FPTP, parties that go too far in conciliation may be punished by losing all of their constituency seats, whereas under PR the party should be able to retain some of them. He acknowledged the fear by parties of being punished by the voters. In the Q&A session, I argued that since 1998, the voters have indeed punished the UUP for having going too far. While the UUP has retained more seats under PR than FPTP (wait for the 2005 Westminster election!), PR itself has not lead to greater moderation (i.e. in Northern Ireland, PR appears to be no better than FPTP to encourage moderation or contest elections towards a median voter; this is important in the discussion on voluntary versus involuntary executive formation).
He cited India as an example of achieving proportionality without PR, but that it had been made more difficult under FPTP.
Lijphart’s last topic was the formation of executives in power-sharing governments. He appeared to be removing any substantive significance as to whether this was achieved by consociational methods or more general porportionality (i.e. voluntary agreements). In the Q&A, I asked him to elaborate on this, and in particular, his views on how executive formation has been applied under the GFA. He replied that Northern Ireland is unique in its rigid application of a consensus formula (citing d’Hondt). He then made the comparison with the interim government of South Africa, arguing that differences between the two were less than people thought. I was unsatisfied with the comparison, and I asked if he could make a one with Lebanon. Here, Lijphart argued that its method of appointing high office positions is the least favourable way, as it entrenches group identities. He also said that a result of an election simply reflects the ethnic makeup of society (i.e. why not employ census rather than bother with elections). In regards to consensual executive formation, Lijphart summarised that “neutral” proportionality is better, and he favours “steamlined” power sharing. By this I understood him to mean the compromise by elected “elites”, which is consistent with his consociational theories; he was not advocating d’Hondt or any other involuntary executive formation method.
One questioner asked about the requirement of overlapping consensus in order for consociationalism to work, and put it that there is a lack of overlapping memberships in Northern Ireland. Lijphart replied by saying that there are few overlapping consensus in divided societies anyway. His argument for consociationalism is to provide security within the largest segments in society, who will then have greater willingness to compromise. He cited the case of the Netherlands as an ideal example, but this was assisted by his remarks of the prevalence of mixed neighbourhoods. He argued against the voluntary (laissez-faire) nature of (monolingual/extreme) radio station provision in the Netherlands, and made the association of an improved shared environment there with the more universal nature of television. Put another way, he found the segregated radio stations running counter to the existence of mixed neighbourhoods.
In essence, for me, while Horowitz concentrates on producing compromise as an electoral outcome, Lijphart places a faith upon political elites (even if extreme opposites), working within a mixed/shared environment. If Horowitz is too determinist, then perhaps Lijphart is too hopeful. Lijphart himself said at the lecture that consociationalism relies upon passive followers and benevolent political elites making productive deals.
In regards to the role of a Prime Minister, Lijphart said that PMs in consensus democracies typically have less raw power than those in FPTP systems, because a PM in the former act a head of a coalition. Lijphart gave the example of German Chancellor Kohl replacing a Free Democrat minister with one of his own picking, with the Free Democrat Party saying, “hold on, we get to say who the Free Democrat representative in Cabinet will be”.
Lijphart also said that it was important the institutions of separation are not “engraved in stone”. He gave the example of the Dutch education system, where in 1917 the Protestant and Catholic churches wanted government-subsidised separate schools, which they achieved. However, Lijphart made the point that the government didn’t name the churches per se in the legislation, but agreed to subsidise any private school that proved viability, etc, with his argument being that this left scope for changing demographics, where another ethnic group could achieve its own schools. After the lecture, I had an opportunity to have a brief conversation with Lijphart, and I told him how shortly after the establishment of Northern Ireland, both Catholic and Protestant churches complained about the secular nature of the education system, with the result of ultimately full government subsidisation for Catholic schools and the retention of Protestant church nominations to the school boards of state schools.
Lijphart’s last comment was a declaration of his ideal model of proportional democracy: 5-6 members constituencies, elected by PR; and a 2-tier system (he cited Germany; I thought of Scotland). He favours the Danish system (naturally!): small geographical districts (encourages local accountability); multi-member representation (so a voter can affirm with at least one representative); and a 2-tier national system, with a 2% threshold for executive/cabinet membership.