For years, we've been told that we don't produce enough math and science teachers. Increasing teacher retirements and increasing student enrollments, we're told, have forced many school systems to lower standards to fill teaching openings, leading to high levels of underqualified teachers and, in turn, to lower student performance. Numerous high-profile reports have directly tied mathematics and science teacher shortages to a host of education and social problems, including the inability to meet student achievement goals, low U.S. performance compared to other nations, the minority achievement gap, poor national economic competitiveness, and even threats to national security. The antidotes have been straightforward: Reformers, legislators, and school officials have implemented a wide range of initiatives designed to recruit able candidates into teaching and to increase the supply of qualified math and science teachers. But, after several decades, we seem to have made little progress. Why? Despite the long-standing prominence of this issue, there has been surprisingly little empirical research on math and science teacher supply, demand, and shortages. Although it has been widely assumed that student enrollments and teacher retirements have far outstripped the production of new math and science teachers, there do not appear to be any studies that test this claim with data. Several years ago, I set out to use the best national data available to answer these questions. Together with my research assistants, David Perda, Lisa Merrill, and Henry May, I analyzed two decades of data from several national surveys conducted by the U.S. Department of Education. I sought the answer to five questions: * What has happened to the demand and need for math and science teachers? * What are the main sources of new hires? * Has the new supply been sufficient to cover increases in student enrollments and increases in teacher retirements? * How many schools have trouble finding qualified candidates to fill their math and science job openings? and * Does the problem vary by state or locale? Many of our findings contradict conventional wisdom and, no doubt, will be viewed by some as heretical. It became clear that the reasons for these staffing problems are more complex and varied than simply an insufficient production of new teachers (Ingersoll and Perda 2010; Ingersoll and May 2010). THE DEMAND Beginning in the mid-1980s and continuing to the present, elementary and secondary student enrollments in the United States have grown steadily. Over the same period, high school graduation course requirements increased in the core academic subjects, especially in math and science. This led, in turn, to a dramatic rise in the number of students taking math and science courses over the past two decades. Math course enrollments grew by 69%, and science course enrollments grew by 60%. In addition, during this period, the number of teacher retirements increased by a striking 141%. All of these factors led to a large jump in the demand for new math and science teachers, and we would expect worsening shortages. THE SOURCES Unlike other nations, the United States produces new teachers through large numbers of accessible, widely dispersed teacher-preparation programs--an average of 25 per state--usually in departments or colleges of education. But, despite the large number of teacher-preparation programs, the traditional pipeline of college students with freshly completed education degrees is, surprisingly, a relatively minor source of new teachers. It constitutes less than a quarter of new math and science teacher hires in any given year. A larger source is those entering teaching with noneducation degrees in math or in one of the sciences, such as biology or chemistry. An even larger source of new hires is what is called the reserve pool, primarily former teachers who left teaching to return later (see Figure 1). …