Parliamentary majority in Poland?

Poland uses a proportional system, and has a generally very fragmented party system. Yet the Law and Justice Party (PiS) may have won a majority of seats today.

The headline in says “Polish right sweeps parliamentary elections”.

Somehow I don’t think of 39% of the vote as a “sweep”. However, if the exit polls are accurate, the wasted votes (below the threshold) were so high that PiS could have around 242 of the 460 seats.

However, if you read far enough down in the Politico piece, you see, “In previous elections, Polish exit polls have not always been accurate. There is a chance that some of the smaller parties balancing on the edge of the 5 percent threshold needed to enter parliament, could still squeeze in.” In which case, it would be rather less sweepy.

Also: “A coalition of left-wing parties failed to make it past the 8 percent threshold to get seats in parliament”.

WAIT. Is the threshold 5% or 8%? Yes; 5% for a single party, 8% for a pre-electoral coalition. And apparently there could be a lot of votes that were cast for parties (or coalitions) that did not clear it.

The outcome is not really a complete shock, at least in terms of vote percentage. The largest party has had 39-41% of the vote in a few elections since 1991. In fact, as recently as… 2011, when it was Civic Platform that won 39.2% of the vote (and 45% of the seats).

Moreover, the presidential candidate of the PiS was just elected in May; no candidate of Civic Platform even entered. (The runner-up was an independent backed by Civic Platform.) Thus this election was held within the “honeymoon” of the president opposed to the incumbent government, and so a surge in the PiS’s vote is what I would have predicted even knowing nothing about Polish public opinion.

However, an absolute majority of seats would be a first for Poland since the fall of the communist government in 1989. It would not be, however, the first time a lot of votes were wasted below the threshold (see 1993, 1997, and 2005). It would just be the first time that a leading party on around 40% of the votes and a big wasted vote happened in the same election. That combo is a recipe for a majority, even under “proportional” representation.

14 thoughts on “Parliamentary majority in Poland?

  1. Maybe just a quibble, but I wouldn’t call Poland “very” fragmented. To me, a very fragmented system is one where the biggest party typically gets around 20-25% of the vote (or less), like Finland, The Netherlands or Brazil. In the past 20 years, Poland has only had one such result (in 2005).

    • Yes, a quibble. The effective number of parties has been trending mostly downward since 1991, so my statement is a bit outdated. But Nv was never lower than 4.50 till 2005.

      If my posts were vetted by an editor first, probably the line in question would have been changed to something like how remarkable it is that there could be a parliamentary majority because it would have been hard to predict such a thing given how fragmented the party system was in the early years of Polish democracy. And then it would have proceeded to note that we have had Polish elections with a leading party around 40% of the votes before, and we have had Polish elections before with a lot of below-threshold votes. If the two conditions ever coincided, a parliamentary majority would be likely, and so here we are in 2015…

      By the way, in 1991, Nv=13.82, Ns=10.86. OK, that’s fragmented! There was not yet a threshold at the time. In 1993, with the threshold, the figures dropped to 9.81 and 3.88 (!!), respectively.

  2. Poland’s National Election Commission is now publishing election results (in Polish only) here for the Senate (elected by FPTP). To be certain, PiS has secured a large overall majority of upper house seats (58 of 91 declared so far), but at the same time it hasn’t been a complete wipeout for PO, which has held on to 29, mostly on or around northwestern Poland, as shown on the official results site’s election map (and rather unlike the map shown on’s live election blog, which had PO ahead in just two regions). That said, I don’t work much with Polish Senate elections, so I don’t know how much (or how little) their results vary from the far more important Sejm polls, although I would expect both to be more or less in line in terms of votes cast for competing parties.

    On the subject of polls but on another country and continent, Argentina’s pollsters had an awful night yesterday: down to the exit polls they forecast an easy victory for the ruling party in yesterday’s elections there, but instead the race turned out to be unexpectedly tight and there will be a runoff vote for the presidency next month. But back to Poland, I suppose we’ll know soon how well (or how badly) did pollsters fare over there.

  3. Final result:
    PiS 37.6% 235 seats
    PO 24.1% 138
    K’15 8.8% 42
    .Modern 7.6% 28
    PSL 5.1% 16
    (German minority – 1 seat)

    Below the threshold:
    UL 7.6% none
    Korwin 4.8% none
    Together 3.6%

    So PiS won a majority for a lower share of the vote than the Liberals in Canada last week. However, I think the majority is so narrow that it’s likely they will lose it through defections by the end of the term.

    • Would Poland had been better of having district threshold instead of a national 5% as to avoid so many wasted votes? Would this have changed the result?

      • Based on my fairly rough calculations, the results below are what would happen if the threshold was on a regional level.

        Law and Justice 225 (-10)
        Civic Platform 130 (-8)
        Kukiz 41 (-1)
        Modern Poland 23 (-5)
        Peasants 17 (+1)
        United Left 18 (+18)
        Korwin 4 (+4)
        Together 1 (+1)
        German Minority 1 (-)

  4. Grofman and Lijphart (1984) noted that the most “reinforced” PR systems, eg Spain, can produce results less proportional than the most proportional WTA systems. The USA’s at that time was an example because many congressional districts were very safe for one or another party and there were only two main contenders. Thus winning 1/435 of the House of Representatives required something approximating 1/435 of the votes nationwide (in party terms – of course a primary victor might win the party’s nomination with only 15 or 20% of the total votes). Gerrymandering, especially if computer-aided, changes the picture, though.

    • Like Zimbabwe, where the 1980 elections were held under party-list proportional representation with an average magnitude of 10, and produced disproportionality of 6.91 on the Gallagher index. At the next elections, held under single-member plurality, the index was 2.58.

    • Disproportionality, as typically measured through nationwide aggregate data, remains very low in the US even in recent elections. It also remains low in some African countries (e.g. Ghana and Zambia) where parties have strong regional basis and there are not very many parties that win many votes but consistently come in only second or third (like UK Liberals and LibDems, or now UKIP). Under those conditions, aggregate disproportionality is low. Of course, district-by-district proportionality is quite another matter.

      I am fairly certain, by the way, that Grofman and Lijphart did not use the term “reinforced” PR for Spain. They probably do not use it even for Greece, except perhaps in the context of indicating that is what it is commonly called in Greece. Spain’s PR produces relatively high disproportionality due to its low magnitude, of course, unlike Greece, where the current law does so via a bonus for the plurality party. In Spain, much of the disproportionality comes from the existence of a significant nationwide third party that wins seats in few districts–somewhat analogous to the British example, albeit less extreme because those few Spanish districts have high magnitude.

      • Sorry, “reinforced” was my paraphrase, not G & L’s original wording. More accurately, from memory, it was something like “the most proportional non-PR systems” vs “the most disproportional PR systems”, with the USA and Spain as the respective flagships.

      • … of course, a result where (eg) Republicans win 51% of seats with 49% of votes and Democrats win 49% of seats with 51% of votes (eg, Bush v Gore), (or vice versa), will register as fairly “proportional”, but of course the crucial factor is that the “wrong” party won. Whereas Canadian provincial-type sweeps where a popular vote just over 50% for one party wipes out the opposition in the legislature will show up as “disproportionate”, but it doesn’t change the basic question of who should have won.

  5. Pingback: Trudeau promise #1: Electoral reform | Voting in Canada

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 )

Google+ photo

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

Connecting to %s