“But people aren’t rational the way economics assumes.”
Perhaps you’ve heard that levied as an objection to the use of economics. It’s a fair objection: I’ve often said that virtually any multidisciplinary conversation involving an economist is bound to devolve into a debate about the rationality assumption. Limits to the homo economicus version of rationality are also limits to the explanatory power of some aspects of introductory-level economics. They do not, however, compromise the case for laissez-faire. If anything, they strengthen it.
The crucial non-sequitur in arguments for government intervention based on individual irrationality is the one that assumes the same faults do not apply to the regulators, as well. Who, I asked for Forbes a few years ago, nudges the nudgers?
I worry, as Deirdre McCloskey did in her 2006 volume The Bourgeois Virtues, about the prerogatives of an immortal, all-seeing, all-encompassing, all-powerful state with a visible and presumably benevolent fist that reaches into our affairs and corrects that which is unwise or unjust or simply unlovely.
In a 2015 article in the West Virginia Law Review, legal scholars Gary M. Lucas and Slavisa Tasic explored the relationship between behavioral issues, law, and regulation. The cognitive limitations compromising the effectiveness and rationality of our commercial and social behavior are amplified by the political because we have even fewer incentives to behave rationally in politics. It is “rational” to remain ignorant in the run-up to an election as one’s vote is almost certainly not going to be decisive. It is also “rational” to cling to irrational beliefs—those that are demonstrably at odds with one’s stated goals—because these beliefs are costly to correct and because one might run afoul of friends and family indulging similarly irrational beliefs and earn their disapprobation.
Citing the work of Timur Kuran and Cass Sunstein, they note the importance of salience as a predictor of “availability cascades” (p. 223):

“In an availability cascade, people with imperfect information accept something as true because others believe it. Moreover, people feel social pressure to accept a particular belief and become reluctant to publicly express their skepticism. Special interest groups and other availability entrepreneurs attempt to create availability cascades to advance their agendas.””]
Think about that the next time a particular tragedy or issue dominates your social media feeds. The by-now-classic example is terrorism. It’s horrible, no doubt, but terrorism is so rare that it should barely register in any inventory of likely threats to our safety and security.
Lucas and Tasic also note the importance of salience for explaining why we regulate rather than subsidize. To use one example, it would be far more effective to write checks to low-income workers than to mandate that firms pay higher wages, but as they point out (p. 239) regulation conceals costs that would be obvious with subsidies. Even though the regulation is more wasteful than a subsidy, it is more politically palatable because the transfer isn’t as obvious.
A bit of introspection should convince all of us that we are bound by important and sometimes-depressing cognitive limitations. These limitations, I think, should make us less enthusiastic about using the state to solve our problems rather than more: externalities notwithstanding I bear almost all the costs of my impulsiveness or flawed beliefs about medicine or what have you. With access to the machinery of the state, I can inflict those costs on everyone.