> people who haven't studied economics enough to conceive of second- and third-order policy effects.

The problem is the effects don't stop at 3rd order, where economic models stop, yielding the consensus that trade liberalization is always good. The reality is different:

her work showed that trade liberalization had slowed the rate of poverty reduction in rural India. - the linked article

" ... none of the world's most successful trading regions, including Japan, Korea, Taiwan, and now mainland China, reached their current status by adopting neoliberal trading rules." - https://en.wikipedia.org/wiki/Comparative_advantage

It turns out that our evolved instincts can be better guides than simplified mathematical models.