Colin Macilwain argues that the United States would not need to spend US$3 billion annually on a programme to encourage young people to pursue careers in science, technology, engineering and mathematics (STEM) if market forces were right (Nature 497, 289; 2013). But conditions and opportunities in science have not visibly improved for students in the past three decades.

Twenty-four years ago, the US government was also trying to attract young people into STEM (T. Packard Eos 70, 709; 1989). Then, 66% of the ocean-science community was living hand-to-mouth on short-term government grants. A university professor was expected to do 40 hours of teaching and administration a week, and 40 hours of research. Researchers who did not receive funding from their universities dared not spend time away from their work, lest their publication record should drop.

It was clear to me at the time that if research centres, universities, governments and societies wanted more people to work in science and technology, then salaries, job stability and job security would have to improve. They still have not.

Young people continue to shun research and instead opt to use their mathematical skills in accounting, their analytical skills in investment banking and their love of science in medicine. Macilwain blames business for the woeful range of scientific opportunities available to graduates. Whether the fault lies with business, government or universities, the educational pipeline in science and engineering does not work because graduates are scared off by what they see as a meat grinder at the other end.