The Package Deal

Ask why a person’s view on capital gains predicts their view on pipeline permits, pandemic policy, and the interpretation of a novel, and you have asked the strangest question in public life. There is no philosophical thread connecting these positions. There is only a supplier.

An industry standard

Consider what a party platform, a denomination’s catechism, or a scene’s unwritten code actually is, commercially speaking: a bundle. Positions on trade, borders, energy, criminal sentencing, family law, and the meaning of history — issues with no logical dependency on one another — are packaged and priced as a unit, take-all-or-leave-all. The nearest business model is the old cable-television bundle, and the resemblance is not decorative; it is structural. Nobody wanted two hundred channels. The bundle existed because it was efficient for the seller: it moved weak products on the strength of strong ones, simplified billing, and made every household’s subscription identical and therefore cheap to service. Replace “channels” with “positions” and the sentence survives intact. A movement cannot mobilize millions around forty separate judgments; it can mobilize them around one membership. The bundle is not a conspiracy. It is a logistics solution — for the logistics of someone else.

The tell is that the bundles reshuffle. Positions migrate between packages from decade to decade — a stance on free trade, on war powers, on state surveillance will appear in one party’s bundle in one generation and the rival’s in the next, the way a channel jumps carriers when a contract lapses. The positions did not change their truth values in transit. They changed suppliers. Anyone whose convictions completed the same migration on the same schedule has learned something uncomfortable about who was doing their believing. If your opinions travel in a flock, it is worth checking whether they are your opinions, or the flock’s.

What bundling does to thought

The package deal has a second-order effect worse than its contents: it deactivates the faculty that would have evaluated the contents. Once the bundle is adopted, each new question arrives pre-answered — the subscriber needs only to check what the package includes this season. Simone Weil, watching the machinery up close, named the resulting condition “mendacity at the very centre of his soul,” and the punishment she named for it was “inner darkness” — hers is a claim about fidelity to the inner light, not about cognition, and it should not be quietly upgraded into one. The cognitive version is a separate claim with separate support: the member half-knows the answer was issued rather than earned, and spends real effort not knowing it more clearly. The laboratory evidence is consistent with it — partisan identity biases reasoning, and it does so at both ends of the ability range — though the stronger form of that finding, that skill actively makes it worse, is the one that failed to replicate. Skill, once enrolled, at minimum stops helping.

There is also an arithmetic problem, and it has to be stated with its assumption showing or it is a conjuring trick. If forty positions were independent of one another, the probability that any single supplier held the correct view on all forty would be, to a first approximation, zero. The word carrying the argument is independent, and it is precisely what a defender of bundles would dispute: positions cluster legitimately when they descend from a shared premise — a discount rate, a prior about state competence, a weighting of liberty against equality — and someone reasoning downward from one honestly held premise produces a bundle that looks issued and was not. The arithmetic does not touch that person. It lands on the reader who cannot name the premise that generates his forty answers, which is most readers — including, in fairness, whoever writes pages like this one and ships ten articles stapled together under a masthead. A reader who accepts the qualified version has conceded the sixth article; the rest is inventory management.

The unbundling, and its counterfeit

Industries built on bundles end the same way: unbundling, from below. Recorded music was sold as albums until it was sold as songs; television until it was sold as streams; newsprint until it was sold as articles. Belief is late to unbundle because its distributors do not invoice in money — they invoice in belonging, a currency with no competing mint. But the ledgers say the run on the bank has begun: the parties, pews, and lodges are emptying in every established democracy, and the leavers are not converting to a rival bundle. They are going à la carte.

Two honest warnings complete the picture. First, à la carte is work — forty questions, forty investigations — and no one completes it; the grown-up version is triage with open books: investigate what you can, suspend what you cannot, and never mistake the suspension for an answer. An honest I have not examined this is worth more than a bundle’s confident error, in exactly the way an empty ledger line is worth more than a forged one. Second, beware the counterfeit unbundling: the algorithmic feed also promises freedom from the package, then quietly assembles a new one from your clicks — a bundle with a sample size of you, sold back as taste. The test of genuine unbundling is the same as ever, and it is not tidiness: can you say, of each conclusion, how you got there and what you consulted that nobody handed you? A mind shopping à la carte can answer that question. A subscription cannot, whoever prints the invoice.