Every policy relies on explicit or implicit assumptions about how people make choices. Those assumptions typically rest on an idealised model of how people think, rather than an understanding of how everyday thinking actually works. This year's World Development Report argues that a more realistic account of decision-making and behaviour will make development policy more effective. The Report emphasizes what it calls 'the three marks of everyday thinking.' In everyday thinking, people use intuition much more than careful analysis. They employ concepts and tools that prior experience in their cultural world has made familiar. And social emotions and social norms motivate much of what they do. These insights together explain the extraordinary persistence of some social practices, and rapid change in others. They also offer new targets for development policy. A richer understanding of why people save, use preventive health care, work hard, learn, and conserve energy provides a basis for innovative and inexpensive interventions. The insights reveal that poverty not only deprives people of resources but is an environment that shapes decision making, a fact that development projects across the board need to recognise. The insights show that the psychological foundations of decision making emerge at a young age and require social support. The Report applies insights from modern behavioural and social sciences to development policies for addressing poverty, finance, productivity, health, children, and climate change. It demonstrates that new policy ideas based on a richer view of decision-making can yield high economic returns. These new policy targets include: the choice architecture (for example, the default option); the scope for social rewards; frames that influence whether or not a norm is activated; information in the form of rules of thumb; opportunities for experiences that change mental models or social norms. Finally, the Report shows that small changes in context have large effects on behaviour. As a result, discovering which interventions are most effective, and with which contexts and populations, inherently requires an experimental approach. Rigor is needed for testing the processes for delivering interventions, not just the products that are delivered.