Claude Steele's Whistling Vivaldi is an excellent new work of social science for the lay-reader about how stereotypes impact the performance and even the long-term health of stereotyped group via a phenomenon called stereotype threat. There is an ever-growing mountain of experimental evidence for the existence of such a phenomenon, which Steele contends has less to do with a stereotyped person internalizing their society's racism as part of their personality, although that is one influence, and more to do with the way these social realities are ubiquitous, resulting in the stereotyped party receiving continual cues about the poor expectations of his performance from his environment. Steele presents research that debunks the myth that the simple solution to this problem is always to work harder. In fact, having a lot at stake and working at the most challenging level are two important contributors to the threat effect.
Steel tracks the evidence collected to-date about the mechanism of stereotype threat at three levels - society, identity, and brain (though he offers the least support for the last, sadly, as there are some very good studies out there). Although Steele's history explores newsworthy domains such as as stereotypes about women being poorer at math than men, and about those with dark skin being less academically strong than those with light, but he makes the point repeatedly that identity threat can and does effect most everyone, offering good examples. He chooses not only the obvious settings of the classroom and the running track to make his point, but some unlikely ones as well - such as the Supreme Court. Steele's writing is attractively free of self-serving hyperbole, his anecdotes illustrate his points clearly, he doesn't oversell them, and the book is both compact and fast-moving.