In my Facebook note, which have somewhat become the substitute for blogging, we have a productive exchange on, again, market. Specifically, on why most economists believe in the market, what are the limits of the market, and how the economics as a discipline has evolved and integrated the so-called 'non-mainstream' approaches. One of the 'non-mainstream' approaches is the field of behavioral and experimental economics. Basically, they show how the rationality assumption is often violated due to cognitive, emotional, bounded rationality etc.
I just recalled some readings by Harvard's Sendhil Mulianathan that addressed how psychology and behavioral perspectives can help us understand more why rationality assumption often fails, particularly in the context of poverty: this one, this one (with Richard Thaler), and this one (with Marianne Bertand and Eldar Shafir). Those three are basically emphasizing each other. He discussed some cases in which the rational maximization model may not be a very good approximation of human behavior, especially when we talk about poverty: underinvestment in education, undersaving, loss aversion in property rights assignment, misaligned teacher's motivation or low take-out rate of social programs.
I admit that, yes, we have to keep rethinking our epistemological position on rationality and how the market works and doesn't. On the other hand, we as economists do know that market often fails, hence it results in suboptimal outcome. But what we doesn't always know why it fails, let alone what solution should we prescribe. The reason is because "all working markets are alike, every failed markets fails in their own way." Meaning, we need to see things case by case and come up with specific - take a deep breath - policy implication, if any.
So why do we still stick to our mainstream or traditional economic tools? Because it is still a good tool. It enables us to: 1) compare the outcomes when the market works (called the benchmark condition) with the one under market failure, 2) analyze which assumptions are violated, 3) think about what - take a deep breath - policy implication, if any.