A masterpiece of electoral manipulation

One must admit that the Spanish electoral system is a masterpiece of electoral manipulation. This can be said with sincere admiration. Without needing to resort to the bonuses of 50 seats like Greece, or Italy's bonus for an outright majority, the Spanish system, designed so that the UCD (Union of the Democratic Center --a party founded in 1977 during the post-dictatorial transition) could win an absolute majority with the minimum number of votes possible, is less coarse but equally efficient. Thanks to subtle mechanisms, it has managed to incorporate three biases simultaneously.

First, it is relatively well-known that the lower the number of MPs per province, the more disproportionate the result. In Spain, the average number of representatives per provincial constituency is 6.7, ranging from Ceuta and Melilla's single seat to Madrid's 36. This means that three electoral systems coexist in Spain. One is of majority rule, which in practical terms means that only the two main parties gain seats in all the provinces from one to five seats (Zamora, Soria, Teruel, and 25 other constituencies); the second is a proportional system, where there are more than ten seats in play (as in Valencia, Madrid, and Barcelona); and the third is an intermediate system with few seats —between six and nine—, in which the conspiracy of the d'Hondt formula skews representation in favor of the majorities [see map].

The second bias is proportional, and it is based on the assignment of at least two initial representatives to each constituency, while in the others they are distributed following population-based criteria. The result of this is that some sparsely populated provinces have a much higher specific weight. For example, in Soria --with two deputies in 2011-- 28,000 voters are needed to elect one representative, while in Madrid --with 36 seats-- 100,000 voters are needed. In other words, the vote of a citizen in Soria is worth close to five times more than one in Madrid, in terms of parliamentary representation.

Lastly, the variation in the number of seats between districts generates a conservative bias that means that the PP’s absolute majorities have been approximately four points cheaper than one of the PSOE. As the PP is especially successful in Castile --where there are smaller districts--, the left --with a vote in larger constituencies-- needs more total votes to obtain an equivalent representation.

These three biases have important implications in the upcoming general elections on December 20th. Because of these differences in the size of constituencies, the distribution of votes for the new parties will be what makes the difference. As a result, what is most relevant for the distribution of seats is not only the percentage of support for each party but, more importantly, where it is obtained. After all, while electing a representative in Ávila may require more than 16% of the popular vote, one can be had in Madrid --although there are more of them-- for a mere 3% of the vote.

The geography of the vote, decisive

In this sense, the new parties are expected to find it much more difficult to gain representation in the smaller constituencies, where the bi-partisan system will withstand the challenge better. In any case, both Ciudadanos and Podemos are approaching a threshold that could allow them to achieve a good result. If they can concentrate their votes well, if they surpass the 16% mark in small provinces and invest their capital in the medium-sized ones, it could shift the balance. With a small vote differential in the right place, the smaller parties could go from 30 to 65 seats. Probably for this reason, more than ever, the geography of the vote will be decisive.

The author is a political scientist and Editor of Politikon