The Package Deal

An opinion is supposed to be a conclusion - the thing you are left holding after you have looked. Most political opinions are not conclusions. They arrive as a set: one membership, many outputs, and the tell is that a stranger who learns your position on a single issue can predict your positions on ten others that share no premise with it.

The pattern is easy to see once named. Views on net zero predict views on the Middle East. Views on lockdowns predicted views on statues. A person's judgement of the monarchy is a decent guide to their judgement of electric cars, and neither follows from the other by any chain of reasoning a philosopher would recognise. Issues with nothing in common move together, the way starlings do - thousands of birds turning at once, in formation, with nothing you could call a decision anywhere in the flock.

The case for holding opinions in bulk deserves to be made properly, because it is stronger than it sounds. Nobody can master epidemiology, energy grids, asylum law and the constitution in one lifetime. Deferring to a coalition that broadly shares your values is the division of labour applied to knowledge - the same sensible outsourcing that lets you cross a bridge without checking the engineer's arithmetic. On that account the bundle is not a failure of thinking but a shortcut through it, and a defensible one.

If that were all it was, the borrowing would show. A borrowed opinion should be held the way borrowed facts are held - lightly, at second hand, with the confidence of the source rather than the certainty of the witness. That is not what we observe. Three things give the game away.

First, the certainty runs backwards. People are often at their most ferocious on precisely the questions they have outsourced. The person who has worked a question out can usually tell you what would change their mind. The person wearing the position usually cannot.

Second, the flip. When a coalition changes its position, its members change theirs - quickly, together, and without any new evidence arriving. Causes migrate across the spectrum within a generation - free speech has changed hands inside living memory - and each tribe's certainty survives the journey intact. Reasoning does not behave like this. Membership does.

The gradient is the giveaway

Third, and decisive: if bundling were a symptom of inattention, it would be strongest among people who barely follow politics and weakest among those who follow it closely. The measured truth is the reverse. Philip Converse found in 1964 that ordinary voters' positions barely constrained one another - knowing one told you little about the rest - while the political class held theirs in tight formation. Pew Research Center measured the same pattern half a century on: by 2014 the share of Americans holding uniformly liberal or uniformly conservative views had doubled in two decades, from 10 per cent to 21 - and the rise was steepest among the most politically engaged. Those are American datasets, but the mechanism does not check passports.

That finding deserves a moment of discomfort. The more political information a person consumes, the more predictable their opinions become. Consumption is not teaching people how to weigh evidence. It is teaching them the playlist - which opinions go together, which combinations mark you as sound, which single deviation marks you as suspect.

There is a respectable objection here: perhaps the bundle has a hidden logic after all - deep values, a general disposition toward change or toward caution, expressing itself across issues. That explains a lean. It cannot explain a lockstep, and it cannot explain why tribes align on questions of plain fact - what a drug trial showed, what a grid can carry - where a disposition has no purchase. Values tilt the field. They do not call every play.

So the honest description of the package deal is this: it is a uniform. The bundle exists because it is worn, and it is worn because the price of one wrong opinion is not intellectual but social. Break formation on a single issue and no counter-argument arrives; a temperature change does. The bundle is enforced at the dinner table and in the group chat, never in the seminar.

There is a cheap test, and it stings. Name the issue where you disagree with your own side. Not a hedge, not a triangulation - a position you would say out loud at a table full of your team. Anyone reasoning issue by issue accumulates these deviations naturally, the way anyone walking their own route ends up off the marked path somewhere. If you have none, the conclusion is uncomfortable but hard to dodge: the opinions are not yours. You are wearing them.

The masthead exists because of this. Judge each argument on its own merits and you will land, sooner or later, off every party's map - which is not a failure of coherence but the only sign of it. An opinion you could never imagine surrendering was never a conclusion. It was a badge.