Green contamination

Political scientists who study mixed-member systems often refer to a concept of “contamination”.* One can find somewhat different meanings of the term in different works, but the most common conceptualization is the entry of candidates in nominal-tier districts where they can’t realistically win, but where their presence may help boost the list. The notion of “contamination” assumes that we should see a reduction towards two significant parties in the single-seat district, but this reduction may be retarded because of smaller parties’ vote-seeking for their party lists. The reasoning goes: the party may fail to mobilize supporters for its list if it is seen to have abandoned the district by not running a candidate.

Is the concept realistic? Scholarly literature is mixed (so to speak). But sometimes politicians act as though they believed political scientists. The Green Party in New Zealand seems to believe in contamination. (That was a fun sentence to write!)

First-term Green MP Holly Walker, elected form the list (like all Green MPs), is withdrawing her candidacy, according to the NZ Herald. Her list candidacy, that is. The article further notes,

Ms Walker, who will still contest the Hutt South seat in order to help the Greens’ party vote, said she was announcing the decision with “real sadness”.

So, even though she won’t be returning to parliament, her electorate campaign remains active, in order to attract votes to the list.

The NZ Green Party is an especially tough test for the contamination thesis, because Green voters in recent elections have shown a very strong tendency to split their tickets, strategically voting for the local Labour candidate. Ticket-splitting runs against the grain of the contamination thesis, which several authors have said is valid only to the extent that voters are generally reluctant to split their tickets.

However, the basic notion of contamination, as I articulated it above, is that the party can’t be seen to have abandoned the district. To do so would reduce their ability to get list votes. The logic does not actually require that voters vote for the district candidate. It only requires that the party’s putting a human face on the list by nominating a candidate in the district helps the party’s list vote. That seems to be precisely what the Greens and Walker are counting on.

* Other authors call it “spillover”. A short bibliography on the topic (and I should note that not all these authors claim to find such an effect, and some specifically argue against it):

Cox, Karen E., & Schoppa, Leonard J. (2002). “Interaction effects in mixed-member electoral systems: Theory and evidence from Germany, Japan, and Italy.” Comparative Political Studies, 35, 1027-1053.

Crisp, Brian F., Joshua D. Potter, and John J. W. Lee. 2012. “Entry and Coordination in Mixed-Member Systems: A Controlled Comparison Testing the Contamination Hypothesis.” The Journal of Politics 74 (02): 571–583.

Ferrara, Federico, & Herron, Erik S. (2005). “Going it alone? Strategic entry under mixed electoral rules.” American Journal of Political Science, 49, 16-31.

Herron, Erik S., and Misa Nishikawa. 2001. “Contamination Effects and the Number of Parties in Mixed-Superposition Electoral Systems.” Electoral Studies 20(1): 63– 86.

Karp, Jeffrey A. 2009. “Candidate effects and spill-over in mixed systems: Evidence from New Zealand.” Electoral Studies 28(1): 41–50.

Ellis S. Krauss, Kuniaki Nemoto and Robert Pekkanen, “Reverse Contamination: Burning and Building Bridges in Mixed Member Systems,” Comparative Political Studies, Vol. 45, No. 6 (June 2012), pp. 747-77.

Moser, Robert G., and Ethan Scheiner. Electoral Systems and Political Context. Cambridge, 2012.

19 thoughts on “Green contamination

  1. I should note that I have a couple of works in progress that express skepticism about “contamination” in New Zealand. One is a chapter in a forthcoming book (accepted by University of Michigan Press), and the other is a conference paper under revision based on survey data and exploiting the effects of redistribution (redistricting) on voter knowledge of candidates.

    I think I might have been the first to use the term, “contamination”, in a similar context: I use it in my dissertation (1988) to refer to the entry of small parties in plurality presidential races concurrent with PR legislative contests. It is also used this way in Shugart and Carey (1992). I don’t think my prior use of this term in the context of presidentialism has ever been cited in the literature on mixed-member systems. Which is just as well, as I really do not like it (anymore).

    One big problem I have with the very concept, other than its being a rather ugly term, is that it assumes that the natural order of things under plurality rules is TWO candidates. But, as is a frequent theme at F&V and in some of my academic writing, this notion is seriously outdated. Most FPTP contests are highly “contaminated” regardless of whether there is a PR vote at the same time!


    • > “assumes that the natural order of things under plurality rules is TWO candidates”

      Duverger’s Law is often mistaken for a universal effect of SMP, but it really applies only in one type of party configuration; where the Big Two are far ahead of everyone else in votes, they will stay far ahead, and indeed likely end up even further ahead. Whether this occurs because SMP lets the majors warn minor party supporters not to waste their votes on loony utopians (which tends to be the British explanation), or whether because it occurs SMP encourages the majors to consider, address and incorporate the very valid concerns, complaints and issues raised by the minor parties (which seems to be the explanation more popular in the US), SMP supporters don’t seem to have quite figured out yet, but hey, eppure si muove. Whether Perot or the LibDems lost votes 1992-96 or 1987-97 because voters realised they were locked in third place, or because Clinton and Blair co-opted the issues they raised, either way, their vote still went down.
      As I might have noted here at F&V before, though, SMP’s barrier to new entrants is very much a “glass ceiling” in one sense: it will repel intruders below a certain velocity, but once it starts to crack, it quickly shatters into pieces. Once enough viable candidates are in the race, the candidate with 15% support suddenly has a real chance of winning – maybe not with 15%, but if you can “squeeze” the candidates on 5% and 10%, you might get up to 30% and have a plurality. Examples are found in Papua New Guinea and the last couple of GOP presidential primaries.


      • Tom, note that the GOP primaries are not by M=1 plurality. Depending on the state, they are either M>1 plurality, or are PR (with a high threshold).

        I suppose that does not have much to do with Greens, MMP, or contamination. But just for the record.


      • Mss, true, delegate seats at the party convention are not SMP. But Ron Paul isn’t contesting the GOP primaries because his end goal is to serve one of two or three thousand delegates meeting for two weeks every quadrennium. He’s eying the single Presidential seat, for the whole four years. Winning delegate seats is but a means to that end. And the Presidency (or the party’s presidential nomination therefor), however, is a single office so the threshold is a simple majority – ie, if not 50%+1 outright, then at least a large enough plurality to be the last one standing after repeated convention ballots.


        • Tom, correct. But that it is a sequential process–both in terms of its being indirect and in terms of different states voting at different times–makes the incentives for consolidation rather different from a one-shot plurality race. And, to be clear, as you noted in another comment, one-shot plurality has “Duvergerian” tendencies. Except when it doesn’t. I like the “glass ceiling” analogy.

          (It looks like you tried to post this first under a different handle. I think you have used it before, so I am surprised the spam filter caught it. I would have cleared it when I saw it, but I realized you had already re-posted.)


      • Aha, I see. Yes, delegate strategy and the rolling campaign does change things, as both supporters and opponents of a single-shot, simultaneous national primary have noted. The result changes the incentives and strategies for tactical voting and “squeezing” rival candidates. (As one example, vote-swaps will only work if there are separate electoral districts). To quote (from memory) a Readers Digest article from 1984 explaining “how America’s elections work”: the small states, voting earlier, can advance an unknown new candidate but the big states, voting later and with more delegates, can crush the upstart if he [sic] turns out unsuitable.
        Since I smacked down a WaPo journalist in the “shutdowns elsewhere?” thread yesterday for failing to note that the US presidency is not a nationwide plurality election, karma has boomeranged here.
        (WordPress has a long memory! I last used that other nom de guerre half a decade ago, once or twice. WordPress shut me out from commenting for months until I managed to recover my password).


        • Tom, one might expect Australians to be more adept at avoiding boomerangs.

          And, while I was distressed to see the problems you were having commenting, I was cheered to know that Word Press maintains heightened security.


  2. In Germany, Baden-Wurttemberg state elections would be an obvious counter-example. Since the Green deputy or deputies from the region (one of the four regions) will be the local candidate(s) who got the most votes, of course voters have a double reason to vote for their local Green: it counts as a party vote, and boosts the regional ranking of their local candidate.

    But even nationally, do voters think the same way? Is a candidate’s local success a factor at future state conventions to rank the candidates in that state?

    Germany’s rising “split ballots” reached 22% in 2002. (More since then?) But the overwhelming majority of members have run in constituencies.The main path to Parliament is to be first chosen by the rank and file of a constituency, in the two main parties, and this norm is probably copied by the smaller parties. “Even among small parties that have very little chance of winning direct seats, the most advantageous positions on lists are occupied by people who also run in constituencies.” (Massicotte) The norm is established by the main parties whose deputies switch categories. “For example, the 1998 federal election saw a major constituency shift. Victorious in 221 constituencies in 1994, the CDU/CSU won only 112 in 1998. Meanwhile, the SPD went from 103 to 212 direct seats. No fewer than 124 members changed category: 73 incumbent list members (all from the SPD, except 2) became constituency members, whereas 51 incumbent constituency members (all from the CDU/CSU) held their seats thanks to party lists. The change was especially pronounced in the Land of Schleswig-Holstein, where a high number of constituencies have slim winning margins: 13 of the 24 federal members elected in 1998 went from one category to the other. In complete electoral upsets, there is even more category shifting. The 2003 election in Lower Saxony provides a good example. Constituencies won by the SPD fell from 83 to 9. Those won by the CDU rose from 17 to 91. As a result, nearly a third of the members elected in 2003 held a different type of mandate from the one they held previously.” (Massicotte.) So a list deputy must compete to serve the constituency where he or she will run next time, since the list seat may evaporate if the party does better. This norm must inspire even the Greens to act as good local deputies, and to demonstrate their worth by getting lots of “first votes.” Does this help their party? No doubt, since Germans generally do not make a difference between the two “classes” of deputies. A popular Green deputy from Stuttgart will surely help the Greens there, even though he was elected on the list.


  3. “Is a candidate’s local success a factor at future state conventions to rank the candidates in that state?” Probably, although I am not sure this has been shown.

    And, yes, in New Zealand and Germany, overwhelmingly candidates are nominated on both the list and in a district–small parties and large ones alike.

    There is much more shuttling between list and district in Germany than in New Zealand, where the normal trend is if you lose in your district but remain via the list, you don’t get renominated (or you bow out “voluntarily”) for the next election.


  4. Wilf, I am not sure what you mean Baden-Wurttemberg to be a counterexample of. You refer to incentives to vote for the party’s candidate in the district and for its list. That is the behavior that the typical “contamination” arguments expect.


    • I simply meant that, if “contamination” is the entry of candidates in nominal-tier districts where they can’t realistically win, but where their presence may help boost the list, the concept does not apply in Baden-Wurttemberg’s no-list model. I then suggested that, even at the federal level, some Germans most likely think the same way: since they generally do not distinguish between an MP from their constituency elected directly or elected on the list, they expect candidates to be on both ballots.


      • Right, Wilf, regarding the B-W state system, which I’d forgotten.

        Of course, “they expect candidates to be on both ballots” is precisely the point made by the “contamination” thesis.


  5. In a situation such as in Germany, where voters get a ballot for the election which determines the overall numbers of representatives for each party, and a completely separate ballot for who will represent their constituency, the local candidate for a particular party shouldn’t affect its votes in the proportional election, and vice versa. They are two separate things.

    In parliaments elected entirely from single member districts, small parties will tend to concentrate their resouces on one or two seats, just to get a foothold in parliament, and often voters in the targetted seat will vote for the small party just to ensure it has representation. The two larger parties constantly urge voters to vote strategically for them, or not “waste their vote” by voting for a preferred smaller party. All this “contaminates” the vote in single member district elections, in that it leads to a different result than if all voters just voted their preferences.


  6. If everybody in an MMP election split their vote, What would happen to the MMP system? Would it become a parallel system and no parties are running decoy lists?

    Why was a two vote MMP system embraced in Germany? I don’t think there is anything wrong with a one vote MMP system. It is just different.

    Why aren’t there two vote systems for a two tier list system? What makes MMP so special to have two votes with the list vote being more important?

    How can a two vote MMP system make electorate MPs more accountable if there already have a place on the list as well?

    Wikipedia has an article about Leveling Seats;
    Is the list component in an MMP system; leveling seats?


    • “Is the list component in an MMP system; leveling seats?”
      Effectively, yes. The difference is only the degree to which they increase proportionality. In PR systems with levelling seats, this is usually only minor, something like 3-5 points lower in the Gallagher index. In MMP it is obviously quite a bit more,


    • “Is the list component in an MMP system; leveling seats?”
      Effectively, yes. The difference is only the degree to which they increase proportionality. In PR systems with levelling seats, this is usually only minor, something like 3-5 points lower in the Gallagher index. In MMP it is obviously quite a bit more.


  7. I don’t think split votes would be an issue in Germany. I believe that any ballots for independents are excluded from the proportional count. Even if that is no longer true, seats in the Bundestag are distributed proportionally. If a party ran s decoy list and got X percent of the vote, it would only get X percent of the seats in a much larger Bundestag.


  8. There’s another factor at work in this specific case: in NZ, a party’s maximum permitted expenses are limited by the number of electorates it contests. If the Greens withdrew their candidate, their expense cap would be ~$25k lower.


  9. Pingback: MMP weekend: Germany and New Zealand 2017 | Fruits and Votes

Leave a Reply

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

You are commenting using your 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 )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.