The hypothetical most popular president (from the perspective of the entire population) would lose in the primaries. Their own party would never nominate them, because they would seem like a sell-out to them.
Imagine the following 3 candidates:
A—strongly against illegal immigrants, and quite racist against the legal ones
B—suggests to stop illegal immigration, but make the legal immigration much easier
C—wants to make immigration easier; refuses to debate illegal immigration because “no one is illegal”
A would win the Republican primaries, C would win Democratic primaries. B would lose both.
For most people, B would be preferable to either A or C. But they won’t get to make that choice. They will have to choose between A and C.
.
Let’s make it more complicated, and split the candidate B into two similar candidates: B-R and B-D. Both B-R and B-D have the same position on immigration, but on different topics, B-R leans slightly Republican, and B-D leans slightly Democrat.
A hypothetical rational Republican might say “let’s nominate B-R for our party, because they are most likely to get elected, and at least they agree with us on many other issues—the 100% chance of B-R winning is preferable to a 50% chance of A and a 50% chance of C”. A hypothetical rational Democrat might similarly prefer a 100% chance of B-D over a 50% chance of A and a 50% chance of C.
(And basically, this is what the median voter theorem suggests: that the election will ultimately be between B-R and B-D, rather than between A and C.)
But in a situation with primaries, B-R will lose to A, and B-D will lose to C.
.
I suspect that to a smaller degree this might be a problem with political parties in general, even if without primaries it is probably much smaller. Individuals need allies to win, and people like B-R and B-D won’t find many enthusiastic allies in their respective parties.
I don’t understand what you’re saying here, but I want to understand.
Can you explain it like I’m 5?
The hypothetical most popular president (from the perspective of the entire population) would lose in the primaries. Their own party would never nominate them, because they would seem like a sell-out to them.
Imagine the following 3 candidates:
A—strongly against illegal immigrants, and quite racist against the legal ones
B—suggests to stop illegal immigration, but make the legal immigration much easier
C—wants to make immigration easier; refuses to debate illegal immigration because “no one is illegal”
A would win the Republican primaries, C would win Democratic primaries. B would lose both.
For most people, B would be preferable to either A or C. But they won’t get to make that choice. They will have to choose between A and C.
.
Let’s make it more complicated, and split the candidate B into two similar candidates: B-R and B-D. Both B-R and B-D have the same position on immigration, but on different topics, B-R leans slightly Republican, and B-D leans slightly Democrat.
A hypothetical rational Republican might say “let’s nominate B-R for our party, because they are most likely to get elected, and at least they agree with us on many other issues—the 100% chance of B-R winning is preferable to a 50% chance of A and a 50% chance of C”. A hypothetical rational Democrat might similarly prefer a 100% chance of B-D over a 50% chance of A and a 50% chance of C.
(And basically, this is what the median voter theorem suggests: that the election will ultimately be between B-R and B-D, rather than between A and C.)
But in a situation with primaries, B-R will lose to A, and B-D will lose to C.
.
I suspect that to a smaller degree this might be a problem with political parties in general, even if without primaries it is probably much smaller. Individuals need allies to win, and people like B-R and B-D won’t find many enthusiastic allies in their respective parties.