Sunday, September 13, 2009

Electoral reform and the left

I'm in the process of writing a short article on electoral reform and left politics, as a contribution to the debate on PR kicking off in the British Labour party at the moment. Here it is:

The 2005 election was a turning point in contemporary British politics. It represented a historic moment for the centre-left, as Labour won a third consecutive election for the first time in its history, ensuring more than a decade of majority government for the party. But it also marked a point of no return for the British electoral system. Although Tony Blair’s Labour party won a comfortable majority of 66 seats, it won little more than a third of the popular vote on a very low turnout. Indeed the Conservatives were only a couple of percentage points behind, and won more votes than Labour in England. With the support of less than a quarter of the British total electorate, the current Labour government lacks legitimacy to govern as a majority administration. The ‘First Past the Post’ electoral model (FPTP) – which forces citizens into tactical voting and marginalizes the supporters of minor parties – is an increasingly inappropriate way of electing the representatives of a diverse and demanding civil society.

The argument for electoral reform – and specifically for adopting a form of proportional representation (PR) for Westminster elections – has long enjoyed support on the left. This support waned during New Labour’s triumphant early period, in which Tony Blair enjoyed such a colossal majority in the House of Commons that the Jenkins Commission was ignored and a manifesto commitment for a referendum on electoral reform conveniently forgotten. Now, with Labour facing defeat next year, many of us are rediscovering our interest in fair voting. But above and beyond the obvious short-term attractiveness of PR, which would limit the scale of a likely Tory victory, there are powerful longer-term considerations which the left should take seriously. To put it succinctly, there is abundant evidence from around the democratic world that PR gives the policies of the left a head-start, whilst First Past the Post benefits the right.

It is little known amongst non-political scientists that there is a good deal of research about the relationship between types of electoral systems and redistributive, progressive policies. This literature has identified a pretty clear negative correlation between 'majoritarian' or 'plurality' electoral systems - like First Past the Post in Britain - and redistributive welfare states. So, in countries like the UK, US, Canada and (formerly) New Zealand, majoritarian electoral rules have been associated with the kinds of hardline neoliberal policies of Thatcher, Reagan and others, whereas in continental Europe, home of the 'social market economy', forms of proportional representation are the norm. Although there are exceptions - Ireland has PR and little redistribution, France has a two-round majoritarian system and an exensive welfare state - the pattern over the post-war period is fairly clear.

Political scientists offer two broad interpretations of this trend.

Arend Lijphart has written extensively about the difference between 'majoritarian' and 'consensus' democracy. Majoritarian democracies tend to elect representatives in single member districts, so that the ‘winner takes all’ and large parties are over-represented at the expense of small ones. In consensus democracies, elections are governed by the principles of PR: representatives are elected in multi-member districts, so that the seats can be allocated to the various parties in proportion to their vote share. Small parties are able to win seats in parliament, and large parties are denied inflated majorities, and forced to seek alliances in order to form government coalitions. Electoral systems are part of a broader collection of institutions which push political systems in a particular direction: majoritarian institutions concentrate power around the representatives of the most powerful groups, while consensus institutions disperse it, allowing minorities the chance to influence, or even veto, policy decisions. As a result, Liphart argues, consensus democracy produces 'kinder, gentler' policy outcomes, including greater redistribution from the wealthy to the poor. Although this means that all groups, not just the poor, are able to push for policies favourable to their interests, a more inclusive system is obviously to the benefit of the most vulnerable social groups since they are least able to defend their interests in more competitive institutional environments.

Could Britain function as a ‘consensus democracy’? In a sense, thanks to reforms implemented by the Labour government, it already does. Devolution to Scotland, Wales and Northern Ireland represents a big step in the direction of a more consensus-oriented political system. PR has been adopted for elections in those three countries, and in election to the European Parliament and Greater London Assembly. The evidence from this recent UK experience is that although Labour has sacrificed the chance of dominating the Welsh and Scottish institutions – under FPTP Labour would have won comfortable majorities – this has not prevented the formation of centre-left coalitions, which have adopted consistently more progressive policies than in the UK as a whole. Extending consensus democracy principles to England and the UK as a whole would not automatically have the same effect – after the 2005 election a Conservative-Liberal coalition would have been just as feasible as a Labour-Liberal one – but what it certainly would do is ensure that a long period of Conservative dominance, such as the destructive 18 years under Thatcher and Major, could never happen again.

Of course, the flip side of this is that the three terms of comfortable majorities for New Labour since 1997 would also have been impossible, given the vote shares enjoyed by Labour fell well below the 50% threshold. Didn’t the huge majorities of 1997-2005 give Tony Blair the chance to reverse the damage of the Tory years, rebalancing public policy in favour of the dispossessed? This is true to some extent, but most on the left would concede that the New Labour project did not achieve as much as was hoped. Level of income inequality in Britain, which exploded after 1979, have been moderated under Blair and Brown, but remain far higher than in 1979, and indeed higher than in almost any Western European democracy. There are many interpretations of the New Labour project, but one perspective largely missing from the debate is the effect the FPTP electoral system had on the policy stances adopted by the party in the 1990s. Labour’s push to the centre was a direct result of FPTP, because FPTP makes it difficult for left parties to advocate extensive redistribution without losing support amongst middle class, centrist voters.

The reasons for this have been examined by Torben Iversen of Harvard University and David Soskice of Oxford, who have studied the effects of electoral systems on the incentives facing different political parties. In their analysis they assume three broad social groups - the High income, Middle income, and Low income groups - of equal size. The Low and Middle income groups together form a majority, and in an unequal society have a joint interest in confiscating part of the wealth of the High income group, and sharing the proceeds amongst themselves. This would mean for example, progressive taxation and redistribution through the welfare state. But although the Low and Middle groups have this incentive in every society, redistribution does not always happen, particularly in countries which have FPTP electoral rules. For Iversen and Soskice, FPTP makes redistributive coalitions more difficult to establish.

The logic is as follows: In a PR system, each group can form a political party which will enjoy a share of parliamentary representation roughly equivalent to the size of the group - here, 33% each. Low and Middle together have 66% of the votes in parliament, and can establish a government which would redistribute resources from High. They can bargain about how to distribute these resources among themselves, in the knowledge that if one group seeks an unfair advantage, the coalition would break down and both sides would lose out. As a result, redistribution is the likely outcome.

In a majoritarian system, the electoral rules tend to favour two large parties (as is evident in the US and UK), not three. So, in order to act jointly to achieve redistribution, Low and Middle must form a political party jointly. But, although Low and Middle have a joint interest in redistribution, they have divergent interests when it comes to distributing these resources amongst themselves. The Middle income group, in particular, may be concerned that the Low group could take control of the party and redistribute not only from the High income group, but also from the Middle. So in order to head off this possibility, Middle may choose instead to ally with High, and keep its income to itself.

This abstract theory provides a neat explanation of what has happened in the Labour party over the last couple of decades. In the 1980s, the hard left (representing, purportedly, the Low income group) made an attempt to take control of the Labour party, offering policies which were almost equally unattractive to the middle class as to the wealthy. Although this attempt was ultimately unsuccessful, it did alienate many middle class voters. In the UK, in the absence of PR, the only realistic prospect for redistribution from the rich to the middle and poor, is the election of a Labour government. But the Labour party in the 1980s threatened to redistribute from the middle to the poor as well, and therefore a sizeable chunk of the middle classes opted for an anti-redistribution alliance with the High income group, voting Conservative. To win back power in the 1990s Labour had to make strenuous attempts to convince Middle income voters that it would not redistribute to their disadvantage. The result: New Labour. However, New Labour's redistributive strategy was ultimately unstable, since it had to redistribute away from the middle classes to some extent, and part of this group is clearly shifting back to the Conservatives - in other words, an Middle-High alliance.

How would PR change this? By allowing a more stable alliance of Low and Middle income groups. Under majoritarian rules, negotiations between these groups must take place within the Labour party, and leaderships of political parties tend to be 'sticky' - if a party heads in the direction of one set of interests rather than another, it is difficult to get it to change direction. This means that trust between the groups represented within the party can easily break down – as they clearly did in the LAbour party in the early 1980s. When this happens, the most likely outcome is that the middle classes will desert the redistributive coalition, and opt instead for an alliance with the wealthy. It is not optimal - under redistribution they would be better off - but at least the risk of being 'exploited' by the Lower income group is minimized. This is probably what is happening to much of Labour's middle class support at present - they feel that they are paying high taxes for the benefit of Labour's working class and public sector base. Because it is so difficult to renegotiate the terms of the Low-Middle coalition inside a closed and centralized party like Labour, they opt instead for a low tax policy under the Conservatives which leaves them worse off then their ideal redistributive strategy, but worse off than policies that redistribute away from them.

This argument implies that Lower and Middle income groups would always benefit from a coalition against the rich (who wouldn't want to redistribute some of Fred Goodwin's wealth?), but that they find it harder to do so under majoritarian electoral systems than under PR. Under majoritarianism, parties need to do a very good job of representing and mediating between the diverse social interests voting for them. Present-day parties are less and less able to do this, for a variety of familiar reasons (declining membership, excessive reliance on corporate donations, tendency of politicians to leave in a protected bubble). In the absence of transparent and representative parties, PR offers the best hope of allowing social interests to be accurately represented in the institutions, and enabling them to develop a stable and effective relationship to bargain over the distribution of resources. The better redistributive record of countries with PR in Western Europe suggests this a plausible argument.

So is PR an open goal for the left? Not quite – there are two notes of caution to be made. The first is that the Iversen/Soskice theory offers a neat explanation for the more generous welfare states in the ‘consensus’ democracies, but we cannot be sure that adopting ‘consensus’ institutions would necessarily have the same effects in the UK. There are other factors at play which may work against progressive policies, such as the absence of strong and representative trade unions, which are a crucial underpinning for redistributive politics in consensus democracies, and even the Anglo-Saxon tradition of liberal individualism, which may lead even low income voters to resist increases in the size of the welfare state. If this is so, changing the electoral system won't necessarily lead to greater redistribution, if the broader institutions of the pro-redistribution coalition are not present. However, the evidence from Scotland and Wales so far is that governing coalitions have generally adopted progressive positions.

The second note of caution is that PR also has the potential to unleash unpleasant extremist forces, as the European elections showed us. There are ways of dealing with this - particularly high thresholds for representation, so that very small groups have trouble getting a foothold in the institutions - but it is a serious risk. But if PR allows the representatives of Lower and Middle income groups to do their job properly, there will be less incentive for marginalized groups to support extremist parties. The success of the British National Party in impoverished communities in Northern England is in large part the result of the inability of New Labour to speak to the most vulnerable parts of the electorate, which in turn is a consequence of First Past the Post.

PR is not a panacea for British democracy, nor for progressive policies in Britain. However it is hard to see how the UK could have become such a harsh and unequal society as it has in the last quarter century without the help of an electoral system which gave hard-line Conservatives almost two decades of untrammeled power, and forced the Labour party into relatively conservative positions in order to defeat them. The hope that Labour could use FPTP to transform British society has been disappointed by the experience of the past decade. Achieving progressive policies, and averting reactionary policies, will be easier under proportional representation.