Despite the worldwide prestige of America's doctoral programs in the humanities, all is not well in this area of higher education and hasn't been for some time. The content of graduate programs has undergone major changes, while high rates of student attrition, long times to degree, and financial burdens prevail. In response, the Andrew W. Mellon Foundation in 1991 launched the Graduate Education Initiative (GEI), the largest effort ever undertaken to improve doctoral programs in the humanities and related social sciences. The only book to focus exclusively on the current state of doctoral education in the humanities, "Educating Scholars" reports on the GEI's success in reducing attrition and times to degree, the positive changes implemented by specific graduate programs, and the many challenges still to be addressed.

Over a ten-year period, the Foundation devoted almost eighty-five million dollars through the GEI to provide support for doctoral programs and student aid in fifty-four departments at ten leading universities. The authors examine data that tracked the students in these departments and in control departments, as well as information gathered from a retrospective survey of students. They reveal that completion and attrition rates depend upon financial support, the quality of advising, clarity of program requirements, and each department's expectations regarding the dissertation. The authors consider who earns doctoral degrees, what affects students' chances of finishing their programs, and how successful they are at finding academic jobs.

Answering some of the most important questions being raised about American doctoral programs today, "Educating Scholars" will interest all those concerned about our nation's intellectual future.

Most skills acquired through on-the-job training may be specific to an occupation and therefore transferable to some but not all firms. This paper explores the relationship between the size of the local market for an occupation-specific skill and job-training outcomes. The Stevens (1994) model of training predicts that as market size increases, job turnover increases and training becomes more general. I test these predictions using data on blue-collar workers and variation in market size across U.S. metropolitan areas. The empirical results support the theoretical predictions and the impacts are most relevant at low levels of market size.