In a country such as Japan, where personal loyalties are strong and university pedigrees are often akin to a family bond, objective evaluation of grants can be a challenge.

The government should make sure that the country's basic foundations for science are in good repair — and that it is seeking out the most creative projects.

That is perhaps one reason why Japan has historically distributed comparatively little of its science funding through competitive grants, preferring instead to spread it more evenly by funnelling the money into the basic operating budgets of universities and research institutions. In 2009, only 13.8% of the country's reseach funding was allocated through competitive grant schemes — up from just 8.9% in 2002 — a considerably lower proportion than in the United States.

Although this approach might seem democratic and fair, Japanese officials have increasingly come to realize that it often fails in the larger test of picking the best science. During the past decade, as Japan has struggled with a stagnant economy, international patent disputes and competition from an emergent China, officials have repeatedly sought to make Japanese researchers stronger on the international scene by upping the competition at the domestic level.

Those efforts have not always paid off — for example, plans to double competitive funding between 2001 and 2005 fell short because the economy couldn't keep up and the science budget plateaued. But change may now be afoot with the massive ¥100-billion (US$1.1 billion) Funding Program for World-Leading Innovative R&D on Science and Technology (FIRST) — a plan that was slashed by the new government last year, but that is nevertheless handing out some of the biggest competitive grants Japanese researchers have ever seen (see page 966). The 30 winning scientists will now have four years to spend an average of ¥3.3 billion each, and will enjoy much greater autonomy under this 'researcher-focused' plan than that offered by previous grant programmes.

FIRST represents an ambitious and worthwhile step for Japan. However, the programme could have been better planned. Consider, for example, that most of the 30 FIRST winners are already world-leading innovators — they include a Nobel laureate and some of the top-cited scientists in their fields. The majority of them were already the recipients of significant grants, and four have been leaders or co-leaders at one of the five World Premier International Research Centers — a 2007 initiative intended to attract top-quality scientists from around the world to work with their Japanese counterparts. These outstanding researchers have accomplished much, and will undoubtedly accomplish more. But it would have been encouraging to see some new faces. Because Japan's strategy for world-class innovation rewards researchers solely on the basis of their past success, it runs the risk of missing some excellent ideas from younger, less-established researchers.

It is also worth asking whether all of these projects require such huge investment. Some clearly do — one example being the Hitachi corporation's effort to develop a holographic electron microscope. This could open a new window onto atomic structures, thereby benefiting hundreds of researchers in various fields. But it is less clear whether all of the FIRST projects will be able to make such productive use of their massive budgets.

A proportion of the FIRST money might arguably have been better spent on increasing the pot of money available through the already-existing mechanism for doling out small competitive grants: the 'Grants-in-Aid' programme administered by one of the country's leading research funding agencies, the Japan Society for the Promotion of Science. Such funding is crucial for small groups hoping to look into risky but potentially groundbreaking ideas — the way that many of the FIRST project leaders made the major discoveries that led them to fame. Moreover, the Grants-in-Aid programme has been forced to cut the average size of its awards in recent years, from ¥3.34 million in 2003 to ¥2.89 million in 2008, because Japan's overall research budgets have remained flat even as the country tries to support an increasing number of researchers. That surge was created, in part, by a programme launched in 1995 that successfully produced thousands of new postdocs but didn't find them jobs (see Nature 449, 1084–1085; 2007).

These concerns aside, however, Japan's government and research community must now follow through on FIRST, giving these projects the support and participation they need to realize their full value. But the next time the government invests large sums in competitive funding, it should first make sure that the country's basic foundations for science are in good repair — and that it is truly seeking out the most creative projects. They are not always in the obvious places.