Two recent elections in Germany and New Zealand have reminded us in what way some electoral systems can create difficulties and can accentuate the well-studied problems associated with voting. That the choice of electoral systems has an impact on the final outcome is hardly news. Condorcet in the 18th century showed how different electoral systems could lead to different results with the same voters’ preferences. What is interesting in these current cases is that both Germany and New Zealand use the same electoral system of proportional representation.
The September election in Germany ended in an impasse for more than three weeks. Gerhard Shröder and the SPD had to admit defeat for Angela Merkel to become Chancellor and form a coalition government. Shröder and the SPD have already denounced the new government as “Anglo-Saxon” and argued that it will not work (see The Economist).
At the same time in New Zealand, an election held on September 17th also led to an impasse for almost four weeks. While Germany got the spotlight, hardly anyone paid attention to NZ. The incumbent NZ Prime Minister, Helen Clark, and the Labour Party obtained 50 seats in Parliament (more than 60 would have been needed to have the majority) and the main opposition party – the National Party led by Don Brash – obtained 48 seats (see results). There were 19 parties standing for the election and eight of them obtained seats in Parliament (there is a 5 percent threshold).
Clark announced her new government earlier this week. While the Green Party was left out (thank God), three minor parties are involved: New Zealand First, United Future, and Jim Anderton’s Progressive. In other words, Clark’s third term is more than ever dependent on the goodwill of other parties, especially New Zealand First and its leader, Winston Peters, who is now foreign minister.
The Electoral Act 1993 introduced proportionality (called mixed member proportional – MMP) in the New Zealand electoral system. The first election under MMP was held in October 1996. It has now been more than 10 years since the change and four elections (including September ‘05) have taken place.
For many observers in New Zealand, there is enough evidence to step back and take a good look at the effects of MMP. Roger Kerr has called the MMP system: Much More Paralysis. He argues that government spending tends to be higher under MMP because of the need to buy smaller parties into coalition governments (as the presence of Winston Peters in a previous government showed).
Moreover, there has been a clear change in the pace of the reform process after MMP was first used in 1996. In fact, as I argue in a forthcoming paper in The Independent Review, backsliding has taken place in NZ over the last ten years – a period during which coalition governments have become common. This was a concern expressed by Evans, Grimes and Wilkinson in their excellent JEL paper in 1996 (Economic Reform in New Zealand 1984-95: The Pursuit of Efficiency). As they put it:
The political sustainability of the reforms awaits the test of the first general election to be held under the new mixed-member proportional voting system. The process of balancing demands for economic efficiency versus rent-seeking and redistribution will be complicated by the advent of the new electoral system. Individual political parties seem likely to focus on narrower constituencies than in the past. (Evans et al. p. 1893)
The reforms in New Zealand during the 1984-95 period were made possible because of the first-past-the-post (FPP) system. Under FPP, each Member of Parliament is elected because he/she gained more votes than any other single candidate in his/her particular electorate. FPP was replaced because it tends to foster a two-party system and delivers majority governments (and thus ignores third parties even when they achieve a significant level of support).
MMP gave birth to a host of new and interesting parties in New Zealand, such as ACT, Libertarianz, and Atearoa Legalise Cannabis (close to the Green Party), although it gave birth to a lot of bad ones as well. ACT has played a very important role in the last ten years in proposing economically sound and rational policies and denouncing bad political behavior. However, it is not clear to me that the price to pay to have more representation is worth it.
True, countries with FPP can be stuck in sub-optimal equilibria for a long-time. However, it is also clear that deep policy reforms have taken place under FPP systems in a way that is perhaps impossible with MMP. Germany and New Zealand are in difficult situations: they both have similar electoral systems and they both have coalition governments. Whether Merkel will be able to implement social change à la Ludwig Erhard is difficult to say. While this is what Germany badly needs, my guess is that it won’t happen. As for Clark, she will be in the hands of her coalition partners and more backsliding is to be expected for the next three years in New Zealand.
As Condorcet showed, all electoral systems have problems. I believe the criterion to use for selecting a voting system is whether it allows for social change to take place as opposed to giving representation to small parties, which at best ends up reinforcing the status-quo and at worst leads to backsliding.
I don’t know whether much research has been done on the subject, but my casual observation, as continental Europe demonstrates, is that proportionality is generally bad for positive social change. Canada, Britain and the US all have FPP systems, which produce clear winners and in which the balance of power cannot be held by small parties. This perhaps explains in part the superiority of Anglo-Saxon jurisdictions over others.