Electoral reform is not a panacea, but it will have effects

Posted by Dan O’Brien

Electoral systems are like exchange rate regimes – none is perfect and all have downsides. I have not suggested that changing Ireland’s system is a panacea, but rather that it, and the making more normal the way the executive is formed, would be the two biggest steps in improving the effectiveness of government in Ireland.

Mixed member proportional systems, such as those in Germany and New Zealand, have single seat constituencies (elected on a FPTP basis) and (closed) national lists which ensure proportionality. I am not suggesting the first part.

I propose keeping multimember constituencies and keeping STV for them (which would not mean all these seats going to FF and FG or a number of other consequences you suggest). The only change in this aspect would be that the constituencies would be roughly twice the size, as they would return, in total, only 83 TDs. This would have the additional benefit of further reducing excessive localism as each TD would need a higher quota and would therefore need to appeal to a larger number of voters (let me be clear here: it is all about trying to get a balance between local and national focus, and not about being hostile to local focus).

As for the national list, I would suggest a closed list. I don’t see the need for this to be linked to the other half, to ensure perfect proportionality, although it could be.

Parties would have an incentive to make the list as attractive as possible and to avoid putting cronies on it. Neither Germany nor New Zealand has any real problem with parliaments packed with cronies. In addition, Ireland’s political parties appear to be trusted by voters – their pecking order in seat numbers hasn’t changed since the1930s, the longest such period of stability in any democracy – so why should there be such fear of trusting them with lists?

There is the wider issue of the relationship between political institutions and government effectiveness. As readers of this site know, there is a vast body of literature on this (as it happens UCD’s Sebastian Dellepiane-Avellaneda, has a nice review article of the literature in the January issue of the British Journal of Political Science). But the literature has focused on differences between developed and developing countries rather than the smaller differences among developed countries. There is an absence of comparative work on the relationship between electoral systems on the one hand and, on the other, parliamentarians’ calibre, how they use their time and legislative output. Despite this, it seems as if political scientists in Ireland mostly support retaining the STV because they believe it to be at least as good as other systems. If this is the case why have the dozens of countries that have considered its merits over decades not opted for it?

Advertisements

11 thoughts on “Electoral reform is not a panacea, but it will have effects

  1. “Despite this, it seems as if political scientists in Ireland mostly support retaining the STV because they believe it to be at least as good as other systems. If this is the case why have the dozens of countries that have considered its merits over decades not opted for it?”

    Because if it is equally good or bad, why would one change to another system that’s equally good/ bad? This would be a costly exercise with no obvious impact. I think one point many agree on is that too much of the behaviour of TDs is attributed to the electoral system and not enough to the parliamentary and governmental system. As David Stanton TD has pointed out forcefully, changing the selection method won’t make a difference if the new TDs have to work under the same restrictive rules.

    The arguments for change don’t really stand up empirically or theoretically. Let’s assume the purpose is to get a different calibre of TD who makes better policy, holds the government to account and spends less time on pot-hole filling and passport delivery. Few would disagree with that as a goal. We know that many legislators working under other systems have full constituency post bags. They may have different types of issues, I don’t know. But the passport delivery problem might in fact be more easily solved by making the public services more responsive to the needs of citizens and by taking away the fast-track facility that government departments give to TDs. So I might spend a day trying to get on the the Department of Social Protection, but my TD has a hotline to it. Of course I’ll contact my TD. TDs have a collective action problem. None of them may want to do this stuff, but they all feel they have to. They can solve this by making it virtually impossible for them to deliver public services more efficiently than citizens. Then there might be more pressure to improve the services offered the public. Theoretically, there is an assumption that the only way that candidates can compete for each other is providing such services, but there is nothing to stop TDs competing on performance in national policy debates. The problem is that TDs have little incentive to do this when the Dáil rules make it very difficult to shine.

    I suspect that some other rule changes could deliver the outcomes that most agree on, and they do not require electoral system changes. And electoral system changes on their own would not deliver those outcomes.

  2. “The only change in this aspect would be that the constituencies would be roughly twice the size, as they would return, in total, only 83 TDs. This would have the additional benefit of further reducing excessive localism”
    Rather, constituencies would be half their current size, so even if there was a clear relationship between geographical size and localism, it is most unlikely the excessive localism referred to would be reduced.
    On the issue of why others have not adopted STV, the key point is that there is generally very little electoral reform in liberal democracies. Richard Katz has an excellent chapter on this Michael Gallagher and Paul Mitchell’s ‘The Politics of Electoral Systems’.
    He equates electoral rules to the rules of a game that the parties in government are winning. Since it is usually the government that has the power to initiate electoral reform, why would those who are winning change the rules of the game? They would only do if they expected defeat and thought that victory could be secured under another system (a la Berlusconi in Italy in recent years).
    To wonder ‘why have the dozens of countries that have considered its merits over decades not opted for it’ misses the realpolitik of electoral reform. It is not too often that countries (in the form of national electorates) really get a chance to opt for it. Rather it is the political elite who decide such matters. Parties operating in non-STV systems fear PR-STV because of the power it gives to the voters. This is one of the reasons why many parties prefer a mixed or pure-list system where they can exercise some control over who gets to sit in parliament.
    If it was the case that many political parties across the world were advocating for PR-STV wouldn’t many of us be suspicious? It’s therefore perhaps a testament to STV that many have feared opting for it. In some rare moments where electors (in the form of a citizens’ assembly) were given the opportunity of opting for a system, they have favoured STV, such as in British Columbia, Canada (as did a majority of the electorate in a 2005 referendum, but not the required 60%).
    To re-iterate the danger in attributing too much causal effect to the electoral system, look at the example of independent (non-party) TDs. Many have gone on record claiming that the likes of Jackie Healy-Rae, Finian McGrath et al are a product of PR-STV. However in 3 other island states (the Australian Commonwealth Senate, Malta and the Tasmanian House of Assembly), with similar political cultures to our own, where PR-STV is also practiced, very, very few Independents are elected to parliament. At the last parliamentary election in Malta (the only other national lower house of parliament using STV), as few as 22 people voted for independent candidates.
    Indeed, having just returned from a fieldtrip to Australia, they laugh at the causal effects we attribute to PR-STV. They also see little problem in opting for this system, as a number of state parliaments adopted it for their upper houses in recent years: Victoria in 2003, Australian Capital Territory in 1992, South Australia in the mid-1970s and Western Australia in 1987 (David Farrell co-wrote a great book on this: ‘The Australian electoral system’).

  3. @ Liam – how would constituencies be half of their current (geographic)size? Surely there would be half as many constituencies, on average with each one being twice the physical size?

  4. @Dan
    “There is the wider issue of the relationship between political institutions and government effectiveness. ”
    Why not separate the members of the Government completely from the Dáil/Senate, if government effectiveness is what concerns you, along with the method of forming Cabinets only from the TDs (with an option for two Cabinet members to be drawn the Senate, but who must not be the Taoiseach, Tánaiste or Minister for Finance, under the current constitution) ?

    Dr. Niamh Hardiman (http://www.ssisi.ie/Hardiman26-11-09.pdf) pointed out that “The main concern expressed about the electoral system is that it does not supply us with
    people who are skilled in specific policy areas.”

    If this is the issue you want to address by changing the electoral system, why not advocate a complete separation of the Dáil/Senate from the executive side of government?

    For better or worse, we draw a lot of our thinking from the UK.
    So it was interesting to see a well informed and experienced public servant call for a separation of powers in the UK. Andrew Turnbull (a former Cabinet Secretary and Head of the home Civil Service) pointed out in a Financial Times article in June 2009
    “…. two principles come to mind – the separation of powers and parliamentary sovereignty. The first has been adopted by many countries but the separation of executive, legislature and judiciary has been espoused most zealously by the US. Each branch of government has its own sphere, creating a system of checks and balances. The striking thing about the UK is not how much we have adopted this principle but how little. Before the Constitutional Reform Act 2005 we tolerated a lord chancellor who sought to be a minister, a judge and Speaker of the House of Lords. But the key area where the separation of powers is not observed is in the overlap of executive and legislature……..The “nearly complete fusion” of executive and legislature is also harmful to the legislature….. Democracy would be better served by ministers who are less political and more expert but held to account by an independent, self-governing and self-confident parliament.”
    http://www.ft.com/cms/s/0/73f524ca-4faa-11de-a692-00144feabdc0.html

    The US is the best known example of this form of government, based on strong federalism and a written constitution.
    The EU institutions (Commission as the initiator, Council as the main legislator with Parliament getting more powers ) also reflect “separation of powers” influences.
    The County Management system is this republic was also clearly influenced by a separation of powers model. Note that it was introduced initially as a response to corruption and then to improve effectiveness, at a time when County Councils raised most of their funding from property taxes and were responsible for the provision of many services eg. health.

    In considering the 1980s crisis here, two friends and I advocated a complete separation of powers here, having considered the electoral system, public service reform etc. In 1987, I summarised our argument as follows
    “Our present structure is like a see-saw, with the elected representative function at one end and the Minister/executive role at the other. Any rise in the effectiveness of one implies a drop in the other. A new structure is needed which would cut the tie so that each can be improved without weakening the other equally necessary activity. There are very few useful changes that can be made without constitutional amendment to those articles which specify the form of government. It would be a pity to waste energy by attempting to fine-tune the 1920’s-based system by, for example, changing the electoral system or restructuring the Senate. Without much more effort, we could have a completely new model that will bring us to the year 2000 and beyond, by giving our government system the means to be successful while increasing democratic accountability. Only thus can our skills and energies be mobilised to open the paths to better standards of living and greater justice for all who wish to live and work here.” (Need Government Fail? Business & Finance May 1987)
    http://irishpoliticalreform.wordpress.com/resources/submitted-papers/donal-obrolchain-paper/
    In 1986, IPA’s Administration published our more argument for the separation of powers
    http://193.120.95.144/politics/design-for-democracy.pdf
    In 1987, Donal de Buitléar examined the electoral system by simulating the possible effects of a MMP system
    http://193.120.95.144/politics/Donal%20O%27Brolchain/Electoral%20Reform%20-%20A%20red%20herring%20Administration%2035%202%20Sept%201987.pdf
    As an attempt to capture the imagination, I wrote Ireland’s Second Republic which the (now-defunct) Department of Public Service published
    http://193.120.95.144/politics/Donal%20O%27Brolchain/Ireland%27s%20Second%20Republic%20Seirbh%C3%ADs%20Phoibl%C3%AD%201987%20Vol%208%202.pdf
    (It still rankles that the editor of Seirbhís Phoiblí removed references I made to Citizens’ Inititiative from that piece, without asking me or letting me know!)
    In 1998, we looked at the report of the Constitution Review Group (chaired by Dr. T. K. Whitaker) and the first two reports of the All Party Oireachtas Committee on the Constitution (then chaired by Brian Lenihan, now Minister for Finance)
    Re. The Whitaker report, we wrote that “The elegance of its writing and the clarity of it presentation cannot disguise the fact that the Whitaker Report is a deeply conservative document, at least insofar as it deals with the institutions of Government”
    http://193.120.95.144/politics/Donal%20O%27Brolchain/Institutional%20Reform%20%20Social%20Policy%20in%20Ireland%201998.pdf
    Recent efforts at Dáil and public service reform shows that the late Professor John Kelly’s comment still holds ie. “Ireland’s political and official rulers have largely behaved like a crew of maintenance engineers, just keeping a lot of old British structures and plant ticking over”
    The well-documented public lack of trust in public institutions suggests that it will take a lot more than changing the electoral system to rebuild this Republic.

  5. Some brief responses.

    “Parties operating in non-STV systems fear PR-STV because of the power it gives to the voters.”
    The democratic wave which began in the 1980s allowed new democracies a blank slate for many things, including choice of electoral system. The vast majority of decisions on which system to go for took place in that context, not in already established democracies changing their systems. Therefore, the natural anti change bias of already established parties, and individuals therein, was not a factor in the big majority of cases.

    “Because if it is equally good or bad, why would one change to another system that’s equally good/ bad?”
    I don’t see how one can arrive at the conclusion (from no country opting for STV over decades) that systems are equally good/bad.

    “As David Stanton TD has pointed out forcefully, changing the selection method won’t make a difference if the new TDs have to work under the same restrictive rules.”

    Imagine 83 individuals from a national list. They would have for more time on their hands as they would do far less constituency work and spend more time in the Dail. With this extra time, is it conceivable that they would not use it to devote greater attention to parliamentary business. For instance, with 83 people focused on being full time parliamentarians, can anyone really doubt there would be more private member’s bills?

    “Theoretically, there is an assumption that the only way that candidates can compete for each other is providing such services, but there is nothing to stop TDs competing on performance in national policy debates.”

    Given the “choice architecture” facing Irish voters, I would imagine that scoring highly on delivering to the constituency reaps far more votes than good speeches in the Dail. (Surely there is some empirical work on what influences voters’ choices.)

    More generally, the relationship between institutions and culture in complicated (it works both ways and there are feedback loops). But changing culture is far more difficult than changing institutions. “the passport delivery problem might in fact be more easily solved by making the public services more responsive to the needs of citizens”. I disagree. Making public services more responsive is notoriously difficult everywhere.

    @Donal
    Seperating the two branches was the second plank of my proposal on the aftershock programme. Your drawing attention to it as far back as the 1980s and the near complete silence since is depressing.

  6. Dan: I guess the political science community and you are simply going to have to agree to disagree on much of all of this. Let me have one more go to try and address some of your points.

    1.) There is a large and growing literature on the politics of electoral reform and design (I address much of this in the forthcoming new edition of my textbook on electoral systems) by authors like: Benoit, Boix, Shugart, Rahat, Renwick, etc., etc. The consensus in pretty much all cases is that when it comes to electoral system design (i.e. the design of new electoral systems for new democracies and the reform of electoral systems in existing democracies) the interests of the party political elite trumps all else. The fact that party interests predominate is one good reason why systems like STV or fully open list barely got a look in: party elites don’t like electoral systems that give voters close direct contact with politicians.

    2.) Mixed-member proportional systems vary in all sorts of weird and wonderful ways (see Shugart and Wattenberg, Blais and Massicotte, etc.), but what they ALL share in common is the fact that they mix two types (tiers) of TDs. As I tried to explain in my previous postings, this can result in a number of issues — not all of them desirable — including:
    – different types of TDs, elected under different systemic logics, operating the same territories with not always pleasant conesquences
    – depending on details such as electoral thresholds and electoral dynamics the distinct possibility that larger parties (esp FF) will win the lion’s share of the lower tier (STV) seats, with the smaller and micro parties picking up most of the latter.
    – Zombie politicians (explained in a previous posting) unless these have been specifically outlawed (in which case, for instance, the leadership of larger parties would likely have to hedge their bets and stay within the STV race — surely not a desideratum in your model)
    – You’ve still not answered the question of closed or open lists. If the latter, then constituency activity will continue apace for so long as the DEMAND is there for it; if the former, then these upper tier TDs become creatures of the party selectorates, beholden on them to remain in office (because they need their votes to be ranked high on the lists). In either version, for so long as the DEMAND from reference groups (be they the electorate or the selectorate) continues to privilege local issues, then the upper tier TDs will continue to supply this service.
    – Your electoral system risks high levels of voter confusion and invalid votes. Voters will be presented with two sets of instructions: to rank order candidates on one side of the paper, and to tick a party box on the other. As Scotland shows, this sows confusion.

    3.) For details on what voters look for in Irish elections, I refer you to the Irish National Election Study, led by Michael Marsh at TCD. This shows categorically that the number one thing that matters for Irish voters is contact with TDs. This is why so many of us in the political science community suggest that the basis for a solution to all of this is a demand-side one rather than the supply-side one you think it is. We need to change the mind set of citizens who expect constituency service from their politicians. Ways to change this include strengthening local government (so that local politicians can take up the slack) and — yes, because it actually can work! — improving the interface between relevant public service departments and citizens.

    Electoral reform may well be worth considering as part of a wider package of reform, as part of a process of clearing the air and starting anew. Personally, I certainly have no problem in considering an alternative electoral system for Ireland — but this would have to be part of a wider pattern of reform (because on its own it simply will not work), and it must be informed by what we know about those alternatives. I’m sorry, but in my book, your proposals fail on both grounds.

  7. Matt: re-constituency size, maybe I wasn’t wholly clear. There are 43 constituencies now, so by a rough approximation, if there are 83 in a hypothetical Dáil, these would have to be half the current geographical size? (of course, this premise ignores the existence of bailiwicks, but deputies, as I’m sure they do, are supposed to represent the entire constituency)

  8. @ Liam: I could be wrong here but, from my reading, the proposal was 83 TDs to be returned from multimember constituencies.

    “I propose keeping multimember constituencies and keeping STV for them (which would not mean all these seats going to FF and FG or a number of other consequences you suggest).”

    So there would be about 20-25 constituencies for the lower tier returning 3-5 TDs in each one. As I understand it, the country would be divided into fewer, larger chunks for the lower tier under this system.

  9. Matt: i find it an unusual suggestion: two levels of PR? I wonder are there are any mixed member systems where the constituency TDs are picked by PR? These would also be quite large constituencies for TDs to manage, and outside of Dublin, Cork, Limerick and Galway, almost all the rest would include more than one county per constituency. This doesn’t end localism as the examples of Laois/Offaly and Carlow/Kilkenny indicate.

  10. Yes – I’ve heard the Green Party floating a very similar system – where the lower tier will be PR-STV for fewer seats and the upper tier will be ‘open’ in that candidate ordering on the upper tier list goes by a sort of ‘best loser’ system.

    I don’t think that there’s much evidence that it would make politics less personalised than it currently is. In the Committee on the Constitution’s report on the proportion of time spnt by members on constituency versus legislative work, the geographic area of each constitunecy was not a significant predictor of time spent on constituency work (the regression tables are in the appendix of the report). Indeed, the distance between Dublin and the constituency was the only significant ‘geographic’ factor that turned up in the analysis (larger distances from Dublin were assoicatied with a greater proportion of time spent on constituency work).

  11. “If PR-STV is such a great system why isn’t it more widely used?”

    It’s a plausible point, though one that could be levelled against almost every system, given that most have their own individual features at the level of detail. What can be said is that PR-STV embodies the features that characterise the electoral systems of most EU member states, especially the smaller ones: PR in multi-member constituencies, all MPs elected from geographical constituencies, voter power to choose among candidates of the same party. That doesn’t prove that systems based on these principles are ‘best’, but it does rather suggest that the basic design of the system is not inherently outlandish or eccentric. It also suggests that we should pause before ascribing causal responsibility for any supposedly unusual feature of Irish politics to the electoral system, or before assuming that if we move away from a Denmark–Finland–Sweden type of electoral system we are somehow more likely to have Denmark–Finland–Sweden type politics.

    Anyway, as David F and Liam W have pointed out, the extent to which an electoral system is employed is not necessarily an indicator of ‘merit’, given that many factors determine electoral system choice. The current debate in the UK reminds us of this: electoral reform has got onto the agenda only because of the balance of political forces after the 2010 election, and if the UK moves to the alternative vote as a result of the process in train, this will not be because anyone thinks AV is the best system. Rather, it will be because it is the furthest the Lib Dems could get on the road towards their desired objective of PR, and the most the Conservatives were prepared to concede away from their preferred model of single-member plurality (‘first past the post’).

    Sad to say, in the imperfect world in which we live, a country’s use or non-use of a particular electoral system is not necessarily based on a solemn judgement by its people, politicians and academics as to what the ‘best’ system is. Decisions tend to be based on a combination of factors, including inertia, familiarity, imperfect information, some sense of the national interest, and consideration of partisan advantage.

    Thus, the reason PR-STV was adopted in Ireland was not because there was a thorough examination of the merits of all known electoral systems and PR-STV was rated highest; rather, it was chosen because pretty much every political actor favoured PR, and STV was the only form of PR that was at all familiar to people in Ireland in the early 1920s. Likewise, a major reason why it is not widely used around Europe is not because countries have thoroughly investigated its qualities and rejected it but because it is simply not well known in most European countries.

    The 1959 and 1968 referendum debates in this country were characterised by wildly exaggerated claims on both sides as to the likely consequences of acceptance or rejection of the proposal to change to single-member plurality. If there is another referendum, let’s hope that the arguments take account of the mass of evidence now available as to how electoral systems do in fact function: what aspects of politics we can expect to change, and what aspects we should not expect to change, if the electoral system is altered.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s