Behavioral economics has transformed public policy by identifying how cognitive limitations and biases systematically distort individual decisions. Yet the field's dominant tools, nudges that target cognitive heuristics and choice architecture, leave a powerful behavioral mechanism largely unaddressed: affect (good-bad feelings). In the domains where behavioral public policy is most urgently needed, from climate change to effective charitable giving to long-term health investment, the primary failure is not cognitive but affective. People endorse high-impact options in the abstract but feel emotionally drawn toward low-impact alternatives. We introduce Affective Paternalism (AP) as a theoretical framework that positions affect as a policy resource rather than a bias to correct. AP's operative tool is the cudge: a liberty-preserving intervention that works by redesigning the emotional texture of choices rather than their cognitive structure. The CARE taxonomy (Create, Attenuate, Reinforce, Eliminate) provides a systematic framework for designing such interventions. Using scope insensitivity and impact neglect as the central case study, we show how CARE interventions can redirect emotional engagement from low-impact to high-impact behaviors without reducing individual welfare or constraining autonomy. We argue that in domains where felt reward is decoupled from actual impact, affective interventions offer a welfare-superior approach to behavior change: they achieve large societal gains at negligible affective cost to individuals. We further show that AP satisfies the criterion of asymmetric paternalism, creating large benefits for those whose decisions are affectively miscalibrated while imposing little or no cost on those whose felt responses are already well-aligned with welfare. We derive testable predictions and identify a research agenda for affect-informed behavioral public policy.