What If It Weren't D'Hondt? November 29 Under the Systems of the UK, France, Italy and Greece

With the vote we forecast today, the electoral system does not change who adds up to a majority: the conservative People's Party (PP) and the far-right Vox would have one under almost any system. What changes is everything else. Under a British-style first-past-the-post system, the PP would have an absolute majority on its own in nine out of ten simulations and Vox would be left with nine seats; under a two-round system, as in France, the Socialists (PSOE) would reach 154 seats, 46 more than under the current system, and the right would reach a majority in only six out of ten cases; under national proportional representation, Vox and Frente Amplio, the left alliance succeeding Sumar, would win 28 more seats between them. D'Hondt, so often criticised, sits in the middle; and switching only the formula to Sainte-Laguë would cost the PP about 11 seats without changing the right's majority.

24 October 2026 · Miguel · Leer en español

After every election night the same complaint is heard: the D’Hondt system favours the big parties and punishes the small ones. That is partly true, but the question is: compared with what? To answer it, we have applied to the vote we forecast today for November 29 five electoral systems used in other democracies, in addition to Spain’s own. These are the same 2,000 simulations as the forecast, with the same votes; only the way of turning them into seats changes. (D’Hondt is the highest-averages method used to allocate seats in each province of the Congress of Deputies, which has 350 seats; a majority is 176.)

Six ways to distribute 350 seats

• D’Hondt by province (the current system). Each province allocates its seats by D’Hondt, with a 3% threshold in the province.
• National proportional. A single 350-seat district for the whole of Spain, allocated strictly proportionally (largest remainders), with a 3% threshold of the national vote. This is closest to the systems of the Netherlands or Israel.
• Single-round majoritarian, as in the UK or the United States. Spain divided into 350 single-member districts; the most-voted candidate in the district wins the seat, even without a majority.
• Two-round majoritarian, as in France. The same 350 districts. A candidate who exceeds 50% wins in the first round; if nobody does, the candidates with at least 12.5% of the electorate (at minimum, the top two) go through to the second round, and the most-voted wins there.
• Mixed, as in Italy. 129 deputies (37%, as in Italy) are elected in single-member districts in a single round, and the other 221 by national proportional representation, with no compensation between the two parts.
• Winner’s bonus, as in Greece. National proportional allocation, but the most-voted party receives a bonus of seats if it exceeds 25% of the vote: 20 extra seats plus one for every half point above 25%, up to a maximum of 50, in the Greek law for 300 seats; here, scaled up to 350.

The results

Absolute majority: 176 seats (the dashed line marks the middle of the chamber).

The first conclusion is that a PP and Vox majority does not depend on the system: under five of the six, they add up to one in more than 90% of simulations. What the system decides is whether the PP governs alone, how much Vox weighs and whether Frente Amplio exists in Congress.

National proportional: Vox and Frente Amplio, the big winners

If the whole of Spain were a single district, the PP would lose about 3 seats compared with the current system and the PSOE would gain 3. The big winners would be Vox, going from 66 to 77, and above all Frente Amplio, going from 12 to 29: it is the party most punished by allocation by province, because in many small provinces its vote does not reach a seat.

There is an important caveat. With a 3% threshold of the national vote, as we have used, almost all nationalist and regionalist parties would disappear: ERC and Junts (Catalan pro-independence parties), EH Bildu (the Basque left), the PNV (Basque nationalists), the BNG (Galician nationalists) and Coalición Canaria (Canary Islands) each have less than 3% of the votes of the whole of Spain, even though they are big in their own territory. With a lower threshold, such as the Netherlands’, they would keep roughly the seats their votes entitle them to.

Single-round majoritarian: a PP with an absolute majority and a Vox almost wiped out

This is the system that changes the result most. With 31% of the vote, the PP would win on average 221 of the 350 districts (between 153 and 268 in 90% of simulations) and would have an absolute majority on its own in 90% of cases. The PSOE would be left with 105. Vox, with almost 19% of the vote, would win only 9 districts: it finishes second or third in almost all of them. Frente Amplio would disappear.

The only parties that hold out are those that concentrate their vote in one territory: EH Bildu (7 seats) and the PNV (5) would win their districts in the Basque Country and Navarre. This is what happens in the United Kingdom, where in 2024 the Labour Party won 63% of the seats with 34% of the votes, and where the Scottish and Welsh nationalists survive thanks to their concentration.

This system is also the most uncertain: the PP could win between 153 and 268 seats, because in hundreds of districts the gap between first and second is small and one national point changes dozens of seats.

Two rounds, as in France: the big winner is the PSOE

With two rounds, almost no district is decided in the first: only 2%. In the rest, the second round pits mainly the PP against the PSOE (63% of cases), and in six out of ten there are three or more candidates in the second round, like the French “triangulaires”.

What is decisive is whom voters whose party has been knocked out vote for in the second round. We estimated it with the second preference voters declare in the CIS (the state polling institute): 66% of Sumar voters would put the PSOE as their second choice, 58% of Vox voters the PP, and only 28% of PP voters Vox. The result is that the PSOE, which collects most of Frente Amplio’s vote, gains almost 50 seats compared with a single-round system: 154. The PP is left at 181, with an absolute majority in 57% of simulations, and the PP and Vox total falls to 63%. Vox, once again, barely wins seats.

It is the only system in which a majority for the right stops being almost certain: the left (PSOE, Frente Amplio and Podemos) would have a majority in 28% of simulations, and with the nationalists, in 36%.

Mixed, as in Italy: the PP close to a majority

With 129 single-member districts and 221 proportional seats, the PP would win most of the districts and total 169 seats, seven short of an absolute majority, which it would reach in only 41% of simulations. Vox would drop to 51 and Frente Amplio to 18. A PP and Vox majority would still be almost certain (98%).

Winner’s bonus, as in Greece

With 31% of the vote, the PP would receive a bonus of about 37 seats and reach 152. It would not be enough for an absolute majority (reached in only 2% of simulations): the Greek bonus is designed to give a majority to whoever exceeds 40%, as New Democracy did in 2023 with 158 of 300 seats. Vox and Frente Amplio, with the rest allocated proportionally, would end up on 68 and 25.

Same provinces, a different formula: Sainte-Laguë and Imperiali

Part of the complaint about D’Hondt is not really about the formula but about the small provinces. To separate the two, we repeated the allocation with the same 52 constituencies and the same 3% threshold, changing only the numbers each party’s votes are divided by:

• Sainte-Laguë, named after the French mathematician André Sainte-Laguë (1910): divides by 1, 3, 5, 7… instead of 1, 2, 3, 4…. It is used in Germany and New Zealand, and is one of the alternatives to D’Hondt most often proposed in Spain.
• Modified Sainte-Laguë, with a first divisor of 1.4 instead of 1, as in Norway, to hold back the smallest parties somewhat.
• Imperiali, which divides by 2, 3, 4… and favours the big parties even more than D’Hondt. It has been used for decades in Belgian municipal elections.

Under Sainte-Laguë, the PP would lose about 11 seats and the PSOE about 5. They would go mainly to Frente Amplio (7 more) and Vox (6 more), and to a lesser extent to mid-sized parties such as Adelante Andalucía, the BNG or Podemos (one more each). With the modified version, the effect is roughly halved. Under Imperiali, the opposite: the PP would rise to 148 and Vox would fall to 56.

Compared with national proportional representation, the formula explains half of what Vox loses under the current system (6 of 11 seats) and two fifths of what Frente Amplio loses (7 of 17); the rest is down to the size of the provinces. And, once again, the PP–Vox majority does not change under any of the three formulas.

What this says about D’Hondt

The Spanish system is not as disproportionate as is often said. Compared with national proportional representation, it slightly favours the PP and punishes mainly Frente Amplio and, to a lesser extent, Vox, because of the effect of small provinces that we cover in this series. But compared with majoritarian systems, which give absolute majorities with a third of the vote, D’Hondt by province is much closer to proportionality.

And a lesson for November 29: the right’s majority is not an effect of the system. With the vote we forecast, the PP and Vox would have it under almost any way of counting.

What the calculation does not capture

• Parties and voters would change their behaviour. Under a majoritarian system, small parties would ally or withdraw, and many voters would vote tactically from the first round. We apply the systems to today’s votes, which are cast with D’Hondt in mind.
• The districts are notional. We have divided each province into districts of similar population by joining consecutive census sections in the order of their code, which groups municipalities and neighbourhoods, although it does not guarantee that each district is a continuous territory. Districts drawn differently would give different results, especially in the majoritarian systems.
• The second rounds use the CIS second preference from the 2023 campaign (study 3416), and assume that anyone who does not have a finalist as a second choice abstains.
• The 3% national threshold in the proportional, mixed and Greek systems leaves out the nationalist parties; it is a decision of the exercise, not a law of nature.
• The overseas vote is not included in the district-based systems, which are calculated with the vote of the census sections.

How we calculated it

• Votes. The first 2,000 simulations of today’s forecast, with each party’s vote in each province. For the districts, the 2023 vote of each census section is shifted (on a logistic scale) until it matches the provincial vote of each simulation.
• Districts. 350 (and 129 for the mixed system) distributed among provinces according to the 2023 census, with a minimum of one per province.
• Check. With the current system, the simulation gives the PP 133.4 seats on average, practically the same as the published forecast (133.5).
• Code. articulo_sistemas.py, in the altnalysis repository.