Where do you really stand in Spanish politics?
The usual left-to-right grid has stopped explaining what happens. Nine axes at once, anchored in the canonical debates of political theory and calibrated against what the parties actually vote in Congress.
A Spanish political test that places you on nine axes at once, anchored in the canonical debates of political theory and calibrated against how parties actually vote in Congress.
The site is in Spanish, and it runs entirely in your browser: no account, no server, no telemetry.
Why nine axes and not one
The left-to-right line has stopped explaining what happens. Two people can agree completely on economics and split on territory, or think alike on culture and diverge on authority. Flattening all of that onto a single line hides exactly the information you need to understand alliances, friction and where consensus actually exists.
So the test measures nine debates separately.
| Axis | What it measures |
|---|---|
| Economic | The State's role in the economy: regulation, taxation, ownership of strategic sectors, labour law, price controls |
| Tradition | Whether inherited cultural institutions (family, customs, national identity, the 1978 Constitution) are presumed valid and preserved, or put up for examination and reform |
| Religion | The institutional role of religion in the State: tax and property privileges, funding of religious schooling, religious symbols in public life |
| Civil liberty | How far the State may coerce the individual outside economics: bodily autonomy, surveillance, imprisonment, the right to protest, conscription. Fiscal coercion belongs on the economic axis |
| Free speech | Whether opinions can be crimes. Measured symmetrically: blasphemy laws, offences against State symbols, hate speech by protected category, criminalised historical memory, State regulation of media |
| Identity | Which group the State should actively recognise and protect. One pole is minority and difference-based identity; the other is majority or national-community identity |
| Territory | Spain as a unitary entity with concentrated power, or a plurality of nations and communities with recognised self-government |
| International | The relationship with supranational frameworks (EU, international courts) and with the movement of people: immigration, borders, regularisations |
| Regime | Defending the current institutional framework (the 1978 Constitution, parliamentary monarchy, established party system) against wanting to replace it with a different one |
The free speech axis is the one I would point at first. Measuring it symmetrically separates the actual liberal (Mill) both from a right that punishes offences against symbols and from a left that punishes offences against categories. Most tests only measure one of those two directions, and so they score half the illiberalism in the room.
Anchored in the canon, not in the news cycle
Every axis names the classical debate it measures and cites the canonical thinkers at both poles. That is a deliberate design decision: if you land far from a pole, you know which intellectual tradition you are disagreeing with, rather than just seeing a number.
| Axis | The traditions it spans |
|---|---|
| Economic | Smith, Hayek, Friedman, Nozick / Marx, Keynes, Polanyi, Piketty |
| Tradition | Burke, de Maistre, Oakeshott, Scruton / Voltaire, Paine, Condorcet, Mill |
| Religion | De Maistre, Donoso Cortés / Locke, Jefferson, French laicism |
| Civil liberty | Mill, Locke, Tocqueville, Constant, Berlin / Hobbes, Schmitt |
| Identity | Taylor, Young, Honneth, Butler / Schmitt, Herder |
| Territory | Althusius, Tocqueville, cooperative federalism / Bodin, Jacobin and Bourbon centralism |
| International | Kant, Habermas, Nussbaum / Schmitt, Herder, political realism |
| Regime | Madison, Constant / Schmitt, rupturist traditions |
A test tied to whatever is trending ages in a year. One tied to the arguments these people were already having ages considerably slower.
How it scores
Fifty-four statements, answered from "strongly disagree" to "strongly agree" on a scale of −2 to +2. Each axis gets a balanced set of statements in both directions, so agreeing with everything does not push you to a pole.
Each answer pushes in a defined direction, and the normalised sum is rescaled to −100 to +100 per axis. Transparent arithmetic, not a black box.
Affinity with each of the seventeen parties is Euclidean distance in the nine-dimensional space, expressed as a percentage. On top of that it counts, statement by statement, how many you actually share with each party, which is where the hidden agreements show up: you often share more specific positions with a party than your overall score suggests.
Calibrated on votes, not speeches
Party positions come from what they vote in Congress and what they pass when governing, not from what they say at rallies.
That is the decision the whole thing stands or falls on. Coding parties from their manifestos is easier and produces a map of how parties would like to be seen. Coding them from roll-call votes produces a map of what they do. The positions are editable, published and arguable, and no party has signed off on them.
Nothing leaves your browser
No server, no cookies, no telemetry. The shareable URL encodes the answers in itself, so whoever opens it sees your result without anything of yours being stored or tracked.
Free, with no ads, no sponsors and no data for sale.
The maps
Three ways to read the same nine numbers:
- A 2D map where you choose which two axes to project. Each party is a point.
- A 3D cube with three axes at once, rotatable by dragging. Parties regroup as you change the projection.
- A parallel coordinates view where each party is a polyline crossing all nine axes. Where lines cross, parties diverge; where they run parallel, they agree. It is the view that shows blocs and fractures fastest.
What each party actually did
The test scores you against seventeen parties. To make that claim honestly, each party has a documented history: how it was founded, who split from whom, which refoundations it went through and who has led it. Sixty-eight thousand words across the seventeen, and every figure in them is backed by a literal quotation from a cited source. A verifier runs before publishing and fails the build if any four-digit year or figure in the prose does not appear verbatim in one of that section's sources, which is the only way to stop dates from being invented.
Where a court is involved, the wording is exact. Being declared a partícipe a título lucrativo is civil liability, and under article 122 of the Spanish penal code it presupposes that the beneficiary neither took part in the offence nor knew of it. Calling that "a corruption conviction" is false, and so is pretending the accounting findings were not made. The pages say which is which.
The nominal voting record
Congress publishes how every one of the 350 deputies voted, but it makes you click a calendar, pick a day and open one file per vote. There is no endpoint.
I crawled the whole legislature: 2,053 votes across 138 plenary days, which is 719,600 individual votes. Every deputy has a page with their complete record grouped by initiative, and every significant bill has a page listing who voted for it and who against, by name.
Out of that comes a figure nobody publishes, because computing it means counting all 350 votes on every single division to find each group's majority position: party discipline in the Spanish Congress runs at 99 to 100 per cent. The deputy who breaks ranks most did so 26 times out of 1,900.
The trap in that calculation is worth recording. Nine deputies changed parliamentary group mid-term. Measuring them against the group they had left produced a beautiful, completely false result: one of them appeared to have defied her own group 424 times, when what she had actually done was walk out of it. Group membership had to be stored as a timeline, not as a field.
And the record starts in 2011 because that is when Congress started publishing it, not because I stopped there. The Diario de Sesiones goes back to 1977, but for an ordinary division it prints only the aggregate: "votes cast, 326; in favour, 176; against, eight". Not one name, not even a breakdown by group. I checked two entire legislatures: of 3,013 divisions in the ninth, exactly two carry a roll call, and both are investiture votes.
The uncomfortable question
The site carries a section that is not really about the test. While the political conversation is taken up by identity, symbols and headlines, generative AI is exposed to displace a large share of current jobs in developed economies according to the IMF and the OECD, and Spain's birth rate hit a historic low in 2024, well under the 2.1 needed for generational replacement.
No party across the spectrum has a serious plan for either. The test will not fix that, but it can ask you something awkward: are you voting for whoever is at least looking in that direction, or for whoever shouts loudest about the conflict of the week?
No todo lo que importa hace ruido. No todo lo que hace ruido importa.
An orientation test, and an independent project not affiliated with any party.