The year 2016 saw a major rise of extreme movements in politics, especially in the major developed countries that use a majoritarian (first-past-the-post, then FPTP) electoral system: the US, with the election of Donald Trump as president, and the UK, with the Brexit. The two societies seemed extremely polarized in their opinions and inclined to unusually judge any statement more through their own confirmation bias than fact-checking it. Indeed, according to us, the UK seems to be drowning in tribalism.

This article intends to present a perspective on this kind of polarization and rise of irrational politics with anti-humanism traits, as rooted inherently into the first-past-the-post electoral system, and majoritarian systems in general. Then, we are going to address the fact that the newest technologies as big data and in general availability of detailed information make the parties more able to exploit this system to their advantage, inflating the problem.  Then, we seek to explain why FPTP is major actor into triggering and exaggerating the Brexit. Finally, a very brief and inexhaustive comparison with the proportional system, common in most of Europe, where instead extremists faced limited results in the latest years, the latest being Geert Wilders’s PVV performing so bad in the Netherlands a few days ago.

The basis for tribalism: tactical voting

Roughly speaking, the first-past-the-post system works as follows: the country is divided into portions, called constituencies in the UK, having each roughly the same population and each electing a single MP in the following way: a citizen may candidate for that constituency, the candidate who gathers more votes than the others is elected. Candidates are usually affiliated with parties. Unlike the proportional system, this ties the MPs to a geographical place, but in turn, each constituency suffers an effect of polarization we are going to explain.

A vote for a candidate that does not win the constituency is a vote effectively not represented because only the winner goes to the parliament and as such voting anyone but the winner is completely equivalent to not going to vote at all. Therefore tactical voting is encouraged and indeed the norm: if the voters of the constituency know in advance which candidates are the two likelier to win there (either because of polls or of previous elections), this will effectively polarize the society between those preferring one more than the other and voting the former, and those vice-versa, in order to make votes effective. Thus, this system forces two poles in the constituency out of a theoretically various and diverse political landscape.

In this way, the electoral law shapes the political debate greatly. Then, this system effectively turns away those who do not recognise themselves in either of the two most likely candidates in case there is a strong majority for a single candidate/party in a place it turns away everybody else, out of the reasoning “why should I vote if they win”. It shapes also the propaganda and the whole line of thought of the two parties of the constituency:  the forced bipolarism demands for the two poles to be able to attract anyone and self-contradicting to do so is just a mean to achieve the ends of victory, while establishing a geographical fidelity or creation of “red/blue strongholds” is most important so that  the party has not to bother with every constituency forever, even if that meant establishing a widespread cronyism within the constituency.

Hence, the fact that a party has to win places and not just people together with bipolarism, makes the system literally as “some interests with geographic ties vs other interests with geographic ties” that fosters an inherent tribalism and non-evidence-based political discourse. Once the tribalism is established and so evident in society, by means of this law, then it keeps exiting even in the referenda where the popular vote is counted and therefore there is bipolarism but no need of geography and, thus, tribes.


Win by distribution through technology?

Now we focus on the effects of the FPTP over all the country and not in a single constituency: for a party to win it has to get 50% of votes in 50% of constituencies, that theoretically are guaranteed to be more than 25% of the total votes, but can be far less than 50%.

Indeed, many elections in the UK resulted in winning parties having less votes than 50% of the population, and the last time a single party that governed won more than 50% of the popular vote in the UK was 1935. Non-representativeness of the house is particularly emphasised in the current UK parliament, where the ruling conservative party won the majority of the seats in the House of Commons yet having less than 37% of the popular vote.

The US presidential election system works with electoral colleges, that are a variant of FPTP and had elected for four times (the delegates for) a candidate that got less votes than another one: in 1876, 1888, 2000, 2016. The rational strategy for a party to win the parliament or for a candidate to win the US presidency is of course pursuing places and not the number of votes. Hence it becomes a sort of sociogeographic war where the successful parties aim at types of the electorate that are geographically tied and mostly ignore those who are homogeneously spread: tribalism again.

One may wonder then if winning by constituencies/states and ignore completely popular vote percentages is a feasible strategy for a party and not just mostly luck in some cases and a quirky coincidence that juts in the latest decades the FPTP has been more and more non-representative. First, a significant part of a party’s success is establishing a party on a basis that has clear geographic links, like, for instance, Labour is tied with industrial outskirts of cities or the SNP is clearly aimed at Scotland.

Yet, all these things are not relatively new. There is something new, though, because technology in the latest decade put a new actor in the play: big data, or the possibility to give everybody “the truth they want”. As Das Magazin first pointed out in a German article in December (followed then by Strade in Italian and in English), the Leave side as well as Trump supporters were helped by a data mining society, Cambridge Analytica that provided them extremely precise tools of psychological targeting that offered a prediction of the political opinions of the people the campaigners were to visit almost house by house, so that they went only where they could find receptive public and had instructions about the personality of the public and thereby what to say to convince them better. All this through analysis of social networks, indeed the list of the likes suffice.

Being able to have such a skill that determines with a high probability where and indeed who could be more likely your voter is the key to victory and makes winning by the sole distribution of votes even easier, thereby making a future with the next election results being less and less representative more likely. This is indeed a major problem not just for representing people’s opinion: what is the bias a political movement could have if they are constantly backed by 25% of the population?

Therefore, technology is making this system outdated and indeed risky and unwise if ever it were a democratic system at all.

The effect of the first past the post on Brexit

We may further argue that a spectacular example of the crisis of FPTP is Brexit. Wait, what? Brexit referendum did count the votes on the popular vote basis, not the constituency one, indeed.

The point is that FPTP surrounds Brexit and influenced it both before, during and after the referendum and if the MPs were approximable with rational players doing rational strategy, just because of FPTP, it would make sense for them:

  • to propose the referendum: so that the conservative party does not lose most constituencies out of UKIP competition, the conservatives agree to question the people on the EU;
  • to propose it with a vague question: other countries force their referenda to ask pertinent and stark questions like “would you abolish the first comma of art. n. 1234 of law no. 567?” even if that meant dozens of questions, whereas UK does not and indeed a vague question allows for more wishful thinking – and campaigning – about the problem, but most importantly gives more room for interpretation of the result to the parliament, that could on paper be sovereign, but is chosen through FPTP, and if FPTP were to dictate that an MP’s choice would mean loss of the seat, the MP would likely not do it, thereby giving priority to the distribution of the result through constituencies before the result itself in an actual implementation of it;
  • to prefer hard Brexit iinstead of no agreement: just by checking results and estimates by constituency, where available ( a sortable table in Wikipedia offers those pieces of data that are certain), one could see a strong majority for Leave bigger than in popular vote. In fact, Chris Hanretty estimated the results by constituency by means of Poisson regression (see the article on the Journal of Elections, Public Opinion and Parties and a divulgative article on ) and according to that study, more than 400 out of 574 English and Welsh constituencies likely voted to Leave. No wonder their MPs would not bet their seats on refusing Brexit and from a point of view of rational strategy, they would support Brexit even if they were to come to support the unthinkable ruin of the economy, because votes in a place, not economy, not fairness, not any kinship around Europe, not the Irish issue, get an English seat in Westminster;

  • for Labour to support hard Brexit: by the reason above, many Labour constituencies, including some “strongholds” voted to Leave, and no wonder they wish to keep their seats;
  • for people to indeed take decisions under fallacious arguments: as said above, the system puts the society into an unnecessary and exaggerated level of tribalism that proportional systems tend not to have. Tribalism in politics means more obfuscated choices, especially when the House of Commons is so non-representative of the election that generated it as it is now.

In a country governed by a proportional system, this would not happen to this scale because proportional takes into account only the share of popular vote and being backed by 48% of the population isn’t exactly no one to find voters among, nor there is a need to compete for the single voter in a swing constituency to win the all system, thereby curbing down the tribalism.


These points outline the 2016 crisis of inclusive institutions around the world as strengthened mainly by the electoral law. I believe indeed this is the case and that the FPTP is the major culprit, as indeed proportional countries seem to be able to cope better than majoritarian ones with xenophobe movements.

Proportional electoral laws work in a way that are multipolar, elections are participated, not won, your party usually represents a coherent position shared by your modest share of the electorate. Were the position no more coherent, the party is likely to split and then usually split the share of votes according to respective popular share. To pass a law you need to come to agreements with other parties so that in the parliament there is at least 50%, hence promoting consensus-building among parties, not division, not tribalism by default. The fact that you have to come to agreements largely works as a way to partly de-bias the effect of your party on the country because each party in the coalition collaboratively partakes to the formation of the law. This may lead to your dreams always partly fulfilled but never totally fulfilled, hence a lot of frustration, in return your own bias has been curbed down. On the other hand, the FPTP is just “choose a bias, mathematics says it is likely to be the most agreeable among the proposed” thereby lacking in the need of collaboration and sometimes even picking a very bad bias because of someone using more detailed information on the distribution of the voters than others.

For the wisdom of the crowd to work the best, you have to make them collaborate, not divide them.