How unusual is the Netherlands 2023 gap between a leading party with less than a quarter of seats and the second party?

This may be my longest post title ever, but it is a question I needed to know the answer to. When the result of the recent Dutch election came in, the gap between the plurality party and the second one was surprisingly large. The PVV of Geert Wilders did not quite win a quarter of the seats (37/150), which is a pretty small share for a largest seat-winner. However, the second largest party is well behind, with only 25 seats for the alliance list of the Green Left and Labour. It seemed to me as if it might be unusual for such a small largest party to have such a large lead.

As the graph below shows, indeed it is unusual. The graph shows on the x-axis the seat share of the largest party (s₁), with the share of the second party (s₂) on the y-axis. Plotted are about 200 elections in parliamentary democracies. The solid line marks s₂=s₁, which is obviously the upper limit. If the top two parties tie in seats, the election will be plotted on this line. Cases in which the leading party has less than a quarter of seats are labelled, as are cases where the leading party has over two thirds or the second party is at less than about an eighth, as well as a few others that I’ll note below.

The dashed line represents s₂=s₁–0.05. As we can see, across the entire graph there are many cases with gaps bigger than 0.05 (i.e., a 10-seat difference if the parliament had 200 seats), but a gap at least that large is very rare if the largest party has less than 25% of the seats. In fact, aside from the Netherlands 2023, marked in magenta, there are only three such elections: Israel 2006 and Belgium in 1961 and 1995. We could give an honorable mention to Denmark 1973, where the leading party had 25.7% and the second party had 15.6%.

Of course, there is a crucial difference between the recent Dutch election and the other cases in which a leading party with only around a quarter of seats or less nonetheless had a five percentage-point lead in seats. In the Dutch case the leading party is a radical party, whereas in the other cases the leading party was one of the mainstream parties. The largest party in Denmark in 1973 was the Social Democrats, the main party of the center-left, while in the two Belgian elections it was one of the main center-right forces (Christian Social in 1961, Flemish Christian Democrats in 1995). The largest party in Israel in 2006 was Kadima, the centrist party formed by Ariel Sharon with the main purpose of building support for his unilateral withdrawal from Gaza. (Centrists do not always have good ideas, but I digress…)

Thus the outcome of such a weak leading party yet such a substantial lead over the next party is indeed unusual. It is even more unusual in that it is a party with extremist positions rather than a party of the center, around which various smaller parties would vie as potential coalition partners. Given the size of the lead, Wilders is in a relatively strong position, while given his extreme policy stances he is by no means assured of parlaying his lead into a prime ministership. And, of course, having less than half of a parliamentary majority is by definition rather weak, even before the ideological positioning is considered. The broad right won big, but it is too early to say that Wilders and his PVV have won in any meaningful sense of what it means to win an election in a parliamentary democracy.