Canada and the UK have first-past-the-post too, yet lack the hard two party system the US has.

One factor is both Canada and the UK have much stronger regional identities than the US - some Americans talk up how strong American regional identities are, but the US has nothing comparable to Scotland or Quebec.

Another factor is the parliamentary system encourages stronger party discipline, which leads to more parties having narrower definitions, compared to the two big tent parties in the US who have such weak party discipline, the party leadership has very little control over its elected officials.

The greater significance of primaries probably also contributes - although in recent decades Canada and the UK have started copying that institution, but still to nowhere near the same degree. A strong primary system weakens a party’s self-control and internal cohesion, and hence promotes it becoming a bigger tent that helps further further entrench the two party system.

Look at the current split of the UK right between Conservatives and Reform - such a split would be far less likely in the US, because the Republican brand is so vague and generic it is much harder for a renegade right wing party to succeed (and the same applies on the left)

> We're stuck with it for now.

One advantage the US has, is electoral systems is mostly a state-level decision, even for federal elections. In most other federations, the voting system used in federal elections is a federal competency. And with 50 states, it only takes one to introduce reforms. But, while there has been some experimentation with alternatives, by and large American states haven’t used their competency to carry out electoral reform. I suspect that is because Americans are fundamentally conservative-in the sense of resistance to change, even self-identified progressives in the US tend to focus their energies on changing certain key issues, and on other issues will let the status quo be.

And yet the two main parties here do feel the need to collude to exclude third other parties. Which demonstrates that they see third parties as a legitimate threat.

And third parties can get enough vote share to tip the outcome ("if only all those people hadn't thrown their votes away, my side would have won!"). Which means they're not the non-entities that theory suggests they are.

And parties aren't static, but have to adjust to match the electorate. There isn't a static steady-state to eventually reach.

And if you've seen discussions about the Democrat party in the US being a "big tent" party that's hamstrung by needing to appease moderates or the Republican party needing to kick out various extremists to gain legitimacy (why yes, most of the discussions I see do come from people on what we call the "left" here, how could you tell?), they sound like there's something similar going on to what I see in discussions about countries with proportional / parliamentary systems having to form coalitions post-election. Ie there's the same sort of coalition-building going on, it just happens before rather than after and isn't explicitly made legible in the labels candidates use for themselves.

The spoiler effect means you don’t need collusion to explain the results. Both parties are enormously incentivised to stamp out and subsume ideologically-near competitors. If they don’t, the other party wins.