Responding to concerns over the uncertain career paths for postdocs in Japan, we recently carried out a survey at seven universities — Osaka, Tohoku, Hokkaido, Waseda, Nagoya, Yamaguchi and Kyushu — and the institute of physical and chemical research (RIKEN). Our fields included science, engineering, agriculture and health care. It was the first such survey at multiple institutions in Japan.

A total of 3,870 people responded just after the end of fiscal year 2005: about a quarter of the roughly 15,000 postdocs in Japan during that period. Two-thirds (2,592) stayed postdocs at the same institution they'd been in during the year; 8% (310) had become postdocs at other institutions; 19% (752) were doing other work, studying or unemployed; and the occupations of the remaining 6% (216) were unknown. Of the 752 who changed their type of work or role, 82% entered research and development (R&D) professions and 9% entered occupations requiring specialized knowledge (such as teachers, doctors, occupations related to intellectual property, coordinators for industry–university collaboration, or science and technology communicators).

The percentage of Japanese versus foreigners who became postdocs at other institutions — as well as who became non-postdoc R&D workers — were similar. Of Japanese postdocs, 72% stayed in Japan and 7% went to the United States. Among non-Japanese, 24% stayed in Japan and 20% moved to China.

More women were unemployed at the end of 2005 than men. More engineering postdocs than scientist postdocs became R&D professionals outside Japan; in Japan, more engineers than other postdocs became private-sector R&D personnel. The average age of becoming a lecturer was 34.2, associate professor 36.9, professor 44.4 years. The number of people becoming postdocs at other institutions decreased with age.

As this was the first large survey of its kind, it is not clear whether job prospects are getting better or worse. But it is worth noting that more than 80% of those who obtained non-postdoctoral positions were able to enter R&D professions. For each institution, the results suggest that the different forms of career development or support may be necessary, depending on the field, to diversify postdocs' career options.