The United States, as Winston Churchill once observed, can usually be relied upon to do the right thing — after exhausting all of the other options available. The aphorism applies neatly to the research component of President Bill Clinton's eighth and last federal budget, which was delivered to the Congress on Monday.

The budget proposal would invest an additional $2.8 billion in research and development, with most of the money being made available for university grants from the National Science Foundation (NSF) and the National Institutes of Health (NIH), the two agencies with the strongest track record of stringent peer review (see page 585). It also incorporates two major initiatives — in nanoscience and information technology — that will again place heavy emphasis on basic research conducted in universities.

The Clinton administration has acknowledged the need to shift the United States' massive research and development portfolio in the direction of peer-reviewed university research almost from the outset. Science in the National Interest, a report published by the White House Office of Science and Technology Policy in 1994, endorsed such a shift as the best way to extract maximum value from research dollars.

But the administration's economic advisers always held technology development efforts as the top priority in its research and development portfolio, and much political effort went into trying to expand such programmes, in the face of justified congressional scepticism. Less politically contentious, but equally insidious, was the administration's tendency to maintain support for intramural programmes, however unproductive, in government laboratories and research centres. Progress in strengthening non-biomedical university research has therefore been unnecessarily slow, and it is only now, with extra money available for government expenditure as a whole, thanks to the budget surplus, that decisive action can be taken to expand the NSF.

Not everyone agrees that this extra money is available, of course. Much of the $40 billion of extra spending for the fiscal year 2001 contained in Clinton's budget proposal will be fiercely contested in the Congress, where some have already dismissed it as a pre-election spending binge.

But the research component of that plan is in pretty good shape. The NIH will, as usual, do better than the president suggests. And support for the NSF has broadened over the past three or four years. It used to struggle in the Congress because it has no big laboratories, and thus no natural constituency. But fiscal conservatives appreciate the NSF's efficiency and, provided its director, Rita Colwell, can persuade Congress that the nanoscience and information technology initiatives are consistent with NSF standards, its budget prospects are good.

The same cannot be said for the Department of Energy's component of the budget proposal. Most of that is for energy-related research, which the Republicans loathe, and for the Spallation Neutron Source, which they have chosen to associate with the political fortunes of Al Gore, the likely Democratic nominee in November's presidential election. But whoever wins that election will inherit a strong research portfolio that, one way and another, has strong bipartisan support.