Plus ca change, and all that ...

I've made my submission to the Electoral Commission's MMP Review. Can you say the same?

The deadline for those wanting to make oral submissions to the Electoral Commission's review of MMP is today. Of course, I'm not one to miss a chance to open my big mouth, so I snuck my submission in last night - as did others equally enamoured with the intracacies of electoral system design and regulation. God, we're a sad and pathetic bunch.

If you haven't had a say yet, don't worry. You can still do it here, up until May 31 ... it's just you won't be able to speak directly to the Electoral Commission about what you think. 

Anyway, I've posted my full submission below (all 5000 words of it - sorry!). I've already covered the first issue (the threshold for entering Parliament) in a previous post, so forgive the repeat performance. But on the other issues, I surprised myself by how conservative I am. Basically, I think MMP is working pretty well and should be left pretty much as is.

Now - tell me why I'm wrong.


1. The thresholds for the allocation of list seats

1.1          I note that there are two alternate thresholds that a party may surmount in order to share in the allocation of list seats. The first I shall term the “Party Vote threshold”, whereby a party that gains 5% or more of the Party Vote is entitled to share in the allocation of list seats. The second I shall term the “Electorate Seat exception”, whereby a party that fails to meet the Party Vote threshold nevertheless may share in the allocation of list seats if one or more of its candidates win an electorate seat. I shall discuss the Electorate Seat exception first, then the Party Vote threshold.

1.2          The original reason for adopting the Electorate Seat exception is not entirely clear to me—the Royal Commission’s recommendation that it be included in New Zealand’s MMP model was not supported by any argument. I suspect it simply was carried over wholesale from the (then) West German example (where it exists because of the potential for regional parties that have support in particular areas of that federal country but are not popular enough nationally to break the 5% threshold), without really considering its relevance for a New Zealand environment. 

1.3          Thus, the only possible justification for the Electorate Seat exception that I can discern is that it helps to rectify some of the disproportionate outcomes that would result from simply having a 5% Party Vote threshold. That is to say, the Electorate Seat exception serves as a de facto way of easing what otherwise is an overly restrictive limit on parliamentary representation. However, I think that it does not do this job particularly well, and that using it for this purpose creates problems of its own.

1.4          First of all, the Electorate Seat exception has been hit-and-miss in its maintenance of proportionality. Most notoriously, Rodney Hide's victory in Epsom in 2008 meant ACT's 3.65% of the Party Vote brought 4 other MPs into Parliament, while Winston Peter's loss in Tauranga meant NZ First's 4.07% of the Party Vote was wasted. Equally, in 2002 Jim Anderton brought Matt Robson into Parliament with him by winning Wigram and getting 1.7% of the Party Vote. But in 1999, the Christian Heritage Party's 2.38% of the Party Vote was wasted—not to mention the Christian Coalition's wasted 4.33% in 1996. So at best, the Electorate Seat exception retains proportionality for only some lucky players. This seems inherently unfair and contrary to the underlying principles of MMP.

1.5          Second, the Electorate Seat exception only helps retain proportionality by (ironically) giving the voters in particular electorates a disproportionate say regarding the overall composition of Parliament via their Electorate Vote. So, voters in electorates like Epsom, Ohariu, Wigram, etc are able to "double dip", in that their Electorate Vote helps ensure one party is represented in Parliament whilst their Party Vote helps elect list MPs for another party. And voters in such electorates do consciously double dip, as the split Electorate/Party Vote statistics demonstrate. In contrast, it does not really matter whom the voters in seats not being contested as “electorate lifeboats” cast their Electorate Votes for—this will not alter the overall shape of Parliament one iota.

1.6          Third, smaller party reliance on the Electorate Seat exception opens the way for inter-party deals over the outcome of the vote. Not that there is anything illegal about one party giving a message (explicit or tacit) to their supporters that they should cast their votes for a candidate from another party. However, it seems fair to say that such spectacles have not gone down well with the voting public; indeed, they have been one of the main grounds for dissatisfaction about MMP.

1.7          Consequently, I recommend that the Electorate Seat exception should be abolished.

1.8          I further recommend addressing the only possible justification for the Electorate Seat exception’s existence—to rectify some of the otherwise disproportionate outcomes that would result from simply having a 5% Party Vote threshold—by lowering the Party Vote threshold. A few preliminaries on this issue, before I say what I think should be done about it.

1.9          First of all, for me the only good reasons for having any Party Vote threshold at all must stem from claims about the functioning of the legislature or government. In other words, I do not put much stock in arguments that without a Party Vote threshold you might get "silly" or "extremist" people elected to Parliament. Using a threshold to save the voters from their own (allegedly) poor choices seems overly paternalistic in the New Zealand environment. (I note that in other jurisdictions—such as Germany—there may be historical reasons for wishing to ensure “extremist” parties are not able to gain representation in the legislature; I do not think those reasons apply here.)

1.10      That said, there still are a couple of matters that I think need taken seriously. The first is the risk pointed to by the Royal Commission back in 1986: without a threshold, Parliament could become so fragmented amongst small political parties that the difficulties of putting together (and maintaining) a governing coalition will impair effective government. And effective government is a good thing that we ought to seek to preserve.

1.11      The second has to do with the functioning of the legislature. MMP is designed to enable parties to place representatives in Parliament to engage in lawmaking/governing. But to be able to effectively engage in lawmaking/governing, you have to be a part of a team of like-minded people and not a lone gunman. Of course, a single MP can act as an occasional flag-waver for her or his supporters on the floor of the House, or may even be able to trade her or his vote into a measure of executive power. But their parties are, for all intents and purposes, non-contributors to the day-to-day functioning of Parliament.

1.12      I admit that there is no objective metric for deciding how important these matters are vis-à-vis the issue of proportionality (which is, after all, what MMP is meant to deliver). And the relative importance of each issue also depends on what the individual observer thinks is likely to happen if the rules on the Party Vote threshold are changed. Will parliamentary representation really fragment to a significant degree once voters realise that a vote for (say) the Aotearoa Legalise Cannabis Party likely will not be wasted? Or are politics in New Zealand now so stable that we have relatively fixed party loyalties that will not crumble into micro-party splinters?

1.13    Having said all that, my own view is that a Party Vote threshold of 2.5% is warranted. I think 5% has proven too high a threshold in practice, in that it has inhibited the emergence and establishment of new political movements. However, I think that (in the interests of creating a legislature that consists of actual parties (in the sense of teams of MPs) and is likely to continue to provide a stable basis for government) requiring parties to achieve a level of Party Vote support that will return at least 3 MPs to Parliament is still a desirable thing to do.

1.14    Consequently, I recommend that the Party Vote threshold should be lowered to 2.5%.


2. List members contesting by-elections.

2.1          There seems to me to be only one real reason for concern about this issue. The election of a list MP at a by-election almost certainly will cause a change in the numbers in the House, in that a party can gain an extra parliamentary seat through the list MP vacating their seat before being declared the MP for the electorate, and thereby get “replaced” in the House from the party’s list. In other words, a win in a by-election may effectively result in two MPs from a party being elected, thereby distorting the proportionality of Parliament as established at the last election. I do not regard this “problem” to be sufficiently pressing to justify changing the law to bar list MPs from standing as candidates in by-elections.

2.2          It is important to recognise that MMP does not claim to guarantee a proportionately balanced Parliament for the entire parliamentary term; rather, it produces a (relatively) proportionately balanced Parliament based on the polling day results. The balance of numbers in Parliament may then change over the next three years for a whole host of reasons: parties may split asunder (or merge together); MPs may declare themselves to be independents or members of a different party; or an electorate MP may vacate her or his seat and be replaced after a by-election by a representative from a different party.

2.3          Consequently, any by-election always has the potential to create a “distorting” outcome, irrespective of whether list MPs are candidates or not. And there is no difference, in terms of overall proportionality, between a party winning an “extra” seat by taking an electorate off a competitor and it winning an “extra” seat by retaining the electorate and gaining a “replacement” list MP. So if the concern really is to prevent a change to the proportionality of Parliament, the answer would seem to be to get rid of by-elections altogether and just allow the party that holds the electorate seat to nominate the replacement MP.

2.4          Of course, this can never happen as voters hold a residual belief that they ought to be permitted to choose who is the best individual to represent their electorate as an MP. Consequently, if the voters in an electorate believe they ought to be the ones to choose a replacement electorate MP, and that choice always has the potential to “distort proportionality”, then why restrict the voters’ potential choices through barring list MPs from contesting the by-election? Insofar as the voters of an electorate see a difference between an MP sitting in the House as a list MP and sitting as the representative of their region, then it seems to me that they should be free to endorse a list MP as “theirs”. Any argument that the list MP is already acting as a representative and so does not need to be chosen as the local electorate representative then can be deployed as a campaign message by the other candidates in the by-election, with the political marketplace left to sort out the rights and wrongs of the issue.

2.5          Indeed, barring list MPs from contesting by-elections could itself result in some quite distorting forms of political behaviour. Many list MPs already function as de-facto electorate MPs, in that they have a strong local profile that has been established over a number of elections. Preventing such list MPs from contesting a by-election in “their” electorates would deprive the voters of the opportunity of choosing between them and the new candidate put up by the party that currently holds the seat. Consequently, the list MP’s party would have to select a new candidate for the by-election—who almost certainly will not then contest the next general election. This seems a somewhat odd state of affairs. 

2.6          There then is a final, subsidiary argument that voters might not actually understand the consequences of electing a list MP as their electorate MP, in that they fail to realise it also will bring in a new list MP for the party concerned. So, for instance, voters may cast a ballot for Candidate A (who happens to be a list MP) without fully realising that if Candidate A wins then Individual X also will become an MP (as they are next on the party’s list of candidates from the last election). However, barring individuals from standing as candidates on the assumption that the voters cannot understand the consequences of electing them seems a bad approach to take. For one thing, it is not clear why voters should be deemed unable to understand this particular consequence as opposed to any other consequence of their vote at the by-election (such as, for instance, depriving a Government of its majority in the House). For another, the potential “two-for-one” consequence of electing a list MP is something you would expect other candidates to educate the voting public on. If having Individual X enter Parliament really is a bad outcome, then that is a message Candidate A’s opponents repeatedly will hammer. So once again, the political marketplace seems able to resolve this issue.

2.7          Consequently, I recommend no change to the law on this issue.


3. The rules allowing candidates to both contest an electorate and be on a party list

3.1          As far as I can see, there are two reasons why it may be thought desirable to prohibit candidates from running both in individual electorates and on party lists. First, it is alleged that allowing candidates to enter Parliament by both avenues allows individuals who have been rejected by an electorate to become MPs in any case. In particular, this is said to blunt the voters’ ability to remove individual electorate MPs that they believe have failed to carry out their job properly. Second, the desire to retain a high list place as insurance against losing their electorates is said to make electorate MPs less interested in listening and responding to their electorates and more concerned with cultivating the favour of their parties. I believe both of these reasons are flawed, and furthermore that prohibiting individuals from running both as electorate and list candidates would have negative consequences.

3.2          At the outset, I simply cannot see any reason to generally prohibit candidates who stand in individual electorates—at least, those electorate candidates who are not sitting electorate MPs—from entering Parliament through the party list. Claiming that an individual candidate’s failure to gain a plurality of the Electorate Vote in a particular geographical area somehow makes that individual unfit to represent the party based on the nationwide vote for it as a party completely misunderstands the way MMP works. The decision of individual electorates as to who will represent them as their constituency MP is no longer the fundamental basis upon which Parliament is selected. It is the Party Vote and the subsequent allocation of party list seats that accomplishes this task. Consequently, trying to create a rigid divide between electorate and list candidates seems to me to be an attempt to reassert a First Past the Post approach to representation that fundamentally is at odds with how MMP actually works.

3.3          There may, however, be an additional reason to seek to prevent “failed” electorate MPs—that is, MPs who previously have been elected via the Electorate Vote but subsequently lost the electorate contest to another candidate—from reentering Parliament through the party list. It is at least arguable that where an electorate has decided it no longer wants an individual to represent it as their local MP, a decision by the MP’s party to featherbed the candidate on its party list should not be permitted to trump that judgment.

3.4          Before turning to the merits of this argument, I note that the problem (if, indeed, it is a problem) is not that great in terms of the numbers of individuals involved. After each MMP election, at most 5% of the MPs returned to the House are former electorate MPs who have lost their electorate seats. Furthermore, most such individuals either retire from Parliament before the next general election or are moved far enough down their party’s list so as not to get returned. Consequently, the combined effects of bruised personal political ego and a party’s desire for “fresh faces” means that the number of “failed” electorate MPs who remain in Parliament beyond the subsequent parliamentary term is very small indeed.

3.5          Whether even this small number of “failed” electorate MPs returning through the party list is a problem at all then depends upon an assumption that the Electorate Vote a candidate receives is a direct judgment on their qualities as an individual. This assumption is highly dubious. Even though voters have the ability to split their votes under MMP, the vast majority continues to “two tick” for the party they prefer. (And a non-trivial number of those who do split their vote will do so on the basis of the party that the candidate they are voting for represents, not who they are as an individual). It is thus extremely problematic to equate the decision of a plurality of an electorate’s voters to choose an alternative candidate over a sitting MP with a judgment on the fitness of that MP to serve in Parliament. More likely, it represents the voters’ changing assessment of the parties that the various candidates represent, with those assessments then flowing on to the decision who to choose as a local representative.

3.6          Furthermore, there are problems with what constitutes “the judgment of an electorate”. Given that the boundaries of electoral districts are subject to regular revision, the fact that an electorate appears to have rejected a sitting MP in favour of another candidate may be attributable to a change in the identity of the voters making the judgment, rather than any assessment that the MP has failed to perform their role satisfactorily.

3.7          I do not think that the argument for preventing even so-called “failed” electorate MPs from returning to Parliament via the party list holds up. It simply fails to recognise the fact that the Electoral Vote in many (even if not in all) cases is not a judgment on the qualities of the individual candidate contesting it.

3.8          What, then, of the alternative argument that allowing candidates to stand both in an electorate and on the party list undermines their relationship with the electorate and makes them too dependant on their party’s favour? Again, I think this is oversold.

3.9          First of all, there is no real evidence that I am aware of that electorate MPs under MMP are more beholden to their parties than was the case under First Past the Post (where all MPs were notionally directly accountable to their electorates). So the claim that MMP and the desire to obtain a good position on the party list somehow has made electorate MPs more likely to put their parties ahead of the voters in their electorates appears purely speculative. Furthermore, even if electorate candidates were prohibited from standing on the party list, there are reasons why they still would continue to owe a considerable duty of loyalty to their party. As already noted, party identification plays a big part in the voters’ decision as to who to support with their Electoral Vote. And political parties not only choose which candidates to include on the party list, but also who to endorse in individual electorates. Furthermore, the desire for advancement within the party, in particular promotion to cabinet (if and when the party wins a place in Government), encourages MPs to be seen to be good team players. These incentives will remain in place even if an MP is forced to rely solely on the Electoral Vote to return to Parliament.

3.10      Equally, under MMP there are incentives for electorate MPs (as well as list MPs) to challenge party policies that they believe are out of synch with the wishes of their electorate. The same voter dissatisfaction that places an electorate-only MP at risk of losing her or his seat also will impact on the overall share of the Party Vote that the party receives. The lower the share of the Party Vote, the fewer party list seats that will be won—as well as the less chance the party will have of forming part of the next government. This basic dynamic of political competition works to keep individual MP’s minds focused on the voters’ wishes irrespective of how they came to be elected. Simply put, whether candidates enter Parliament through the Electorate Vote or the Party Vote, it’s the same voters with the same view of their party’s actions and policies that put them there.

3.11      Additionally, any move to prohibit candidates from standing both in an electorate and on the party list would have a range of foreseeable negative consequences. It would effectively stop parties running senior MPs in electorates where they are not guaranteed victory. Instead, senior MPs without a “safe” electorate base would become list-only candidates, thereby actually further distancing them from day-to-day contact with the voters. By the same token, electorates not considered “safe” by one or another party would likely see a reduction in the quality of candidates offered to them. All parties would have to seek out more candidates than currently is the case, which again can be predicted to result in lower quality candidates contesting the election. So, to give a concrete example from my own electorate, Dunedin North likely would go from having 3 sitting MPs contesting the Electorate Vote (and turning up to meet-the-candidates meetings, etc) to having 1 sitting MP doing so, with a range of “cannon fodder” opponents put up against him by the opposing parties.

3.12      I would make one last point here. I suspect the Commission will be encouraged to adopt a range of apparently clever ways to square the circle on this issue. So, for instance, I have seen suggestions that only those electorate candidates that receive more Electorate Votes than the Party Votes cast for their party in an electorate should be able to be elected via the party list. (The assumption is that this indicates the candidate is personally more popular than her or his party.) I would caution against any such rule, or other attempt to allow some electorate candidates to be able to be elected from the party list, while others may not. Such rules open up incentives to engage in gaming behavior, such as vote splitting by a candidate’s opponents in order to try and disqualify her or him from entering Parliament. Equally, they can have unforeseen and quite unintended consequences. Instead, I would urge the Commission to retain a simple approach here, whether that be no rule at all (i.e. candidates may run both in electorates and on a party list) or a rule that candidates may not run in both capacities.

3.13      My recommendation is that the law in this area not be changed, and that candidates remain free to stand both as electorate candidates and on a party list.


4. The rules for ordering candidates on party lists.

4.1          I understand the concern behind this issue to be that while list MPs can claim a form of democratic mandate by virtue of their entering Parliament based on their party’s share of the Party Vote, the closed-list nature of New Zealand’s version of MMP means that that mandate is of a particularly attenuated nature. In particular, the way in which candidates come to occupy their position on a party’s list may be opaque to those outside the party, whilst voters may have little idea which precise individuals their Party Vote is helping to deliver into Parliament (or, even worse, may think that their Party Vote is helping to deliver the wrong individual(s) into Parliament).

4.2          There are two (not necessarily mutually exclusive) ways in which this issue might be addressed. First, the selection and placement of candidates on a party’s list could be made “more democratic” by tightly prescribing the ways in which parties must create their candidate lists. Second, the voters could carry out the actual selection of individuals to fill the party list seats directly—that is, an “open list” approach could be adopted. While I am less sure on this issue than I am on the others discussed, I think that on balance both of these approaches ought to be rejected in New Zealand.

4.3          Section 71 of the Electoral Act 1993 already imposes a requirement that political parties follow “democratic procedures” of a sort when selecting their candidates (be these electorate or list). However, the actual content of this requirement is not great; a candidate selection process that allows individual party members to “participate” in the selection by way of party officials elected by delegates at the annual party conference meets its terms (see the judgments in Payne v National Party of New Zealand and Payne v Adams). As such, it provides only the most minimal guarantee that the candidates endorsed by the party are those that the individual party members really want.

4.4          However, requiring parties to adopt some “more democratic” process of selecting and ranking their candidates—demanding, for example, that candidates are chosen on a “one member/one vote” basis—runs into the problem of party autonomy. The internal structures and rules of each party reflects the particular ideological views that the party exists to promote. So, for example, the involvement of unions in the selection of Labour Party candidates is a reflection of that party’s emergence from (and ongoing commitment to) the organised labour movement. Equally, the Green Party’s requirement for gender balance on its party list reflects that party’s broad commitment to issues of social equity. The risk is that however attractive it may be to demand that political parties be “properly democratic”, adopting prescriptive rules about how parties may (and may not) select candidates may effectively stop some parties from internally organising in ways that best meet their goals and beliefs.

4.5          By the same token, the claim that voters ought to be able to directly select which individuals they want to see entering Parliament from a party’s list has an immediate appeal. After all, if those list MPs will be representing the voters, then surely the voters ought to get to say who they will be. However, there are both principled and practical problems with this claim.

4.6          The principled objection is that as the Party Vote is a vote for the party qua party, then it should be for the party as an organisation to decide (according to its own internal rules) who will represent it in Parliament. This is again an assertion of party autonomy; the party’s determination as to who is best suited to promote the particular policies and principles it represents ought to be respected. (As an aside, I note that this principle has long been a part of New Zealand’s electoral law; even under First Past the Post, voters were simply called on to endorse or reject those candidates the parties placed before them. There never has been a right for voters to “write in” a different candidate’s name on the basis that they would rather a party had nominated her or him instead of the actual person endorsed.)

4.7          In terms of practical problems, there is the question of how an open list system would operate. One option would be for voters to cast their Party Vote directly for an individual candidate on the party list, with the candidates receiving the most such votes then filling the party’s list seats. Thus, if 30% of the voters cast a vote for the various candidates on a party’s list, then the party would gain 36-38 seats in Parliament—with the list seats the party receives being filled by the list candidates that received the most individual votes. However, while this may ensure the “most popular” list candidates enter Parliament, whether it really will solve the “attenuated mandate” problem is debatable. It can be predicted that the great bulk of votes will be given to the highest profile party candidates, the party leaders in particular. This may leave very few votes for other candidates, especially those chosen to fill the last list seats for the party. Simply put, if a list MP enters Parliament because (say) a thousand (at most) voters indicated they wanted to see her or him elected, I doubt this satisfactorily would answer complaints that list MPs have an insufficient mandate.

4.8          A second option may therefore be to allow individual voters to rank the party list according to their own preferred order, with the selection of individuals to fill list seats then being made according to the ordered preferences of the voters (however this is then discerned—presumably through some form of STV mechanism). This approach would have the benefit of being more nuanced than a crude “choose your preferred list MP” one. Again, however, I foresee problems with it. First of all, it would markedly complicate the process of counting votes and declaring who has been elected to Parliament (as well as making this a much more expensive exercise). Second, I think there are real questions about whether voters will have adequate information to conduct this ranking exercise—especially when party lists may have up to 60 names on them. Finally, I wonder whether voters really want to do so, or whether the Australian experience (where the vast majority of voters simply tick “above the line”, thus endorsing the party’s own ranking of candidates) will be replicated here.

4.9          On balance, therefore, I am of the opinion that whatever disquiet may be expressed about list MPs “not really representing anyone”, the potential fixes to this problem trench too deeply on the right of parties to select who they think can best advance their programme, as well as raise some quite challenging practical issues of implementation.

4.10       I recommend no change be made to the law on this issue.


5. The effect of a party winning more electorate seats than its party vote share entitles it to.

5.1          I must admit to being largely agnostic on this particular issue. There seems to me to be no good answer to it, outside of allowing the House of Representatives to grow large enough to remove any effect on proportionality. However, as this falls outside the permitted grounds of this review, there is no point pursuing it.

5.2          I recommend no change be made to the law on this issue.


6. The effects of the ratio of electorate seats to list seats on proportionality in certain circumstances.

7.1 I do not see this as a particular issue of current concern. I note that in 2000 the Department of Statistics projected that in 2051 there will be 77 electorates and 43 list seats. Those numbers do not lead me to believe that there will be any real problems of proportionality caused before that date.

7.2 I recommend no change be made to the law on this issue.