Real-world policymakers face pressure to take action, to legislate, and to attempt to solve problems even in imperfect ways. What kind of paternalistic policies can we reasonably expect policymakers to create? We argue that public-choice pressures will tend to produce suboptimal paternalistic policies, even if we assume behavioral paternalists’ conclusions about human behavior are generally correct. Rational ignorance, bureaucratic self-interest, concentrated benefits and diffuse costs, the influence of rent-seekers and moralists, and other factors will tend to shape policy in undesirable ways. If policymakers are susceptible to biases such as those attributed to regular people, the results could be even worse. Biases with the potential to adversely affect policymaking include action bias, overconfidence, confirmation bias, availability and salience effects, affect and prototype heuristics, and present bias. Because the political sphere offers weak incentives for the self-correction of biases, we expect such biases to be more significant in the public than in the private sphere.