We agree that science policy can and should encourage research programmes that deliver public values — the core aspirations that motivate and justify public spending on science (Nature 473, 123–124; 2011).

The public expects climate research to generate useful, reliable knowledge. This is more likely when programmes engage with the policy and public context in which they are embedded. Such engagement is not easy, or popular. It demands creativity, resources and collaboration with diverse communities.

Scientists working on important societal problems are usually subject to the same metrics of success as theoretical physicists and mathematicians — peer-reviewed articles. Nevertheless, some scientists and organizations generate useful information for policy-makers and for public benefit. More must be done to recognize and reward their contribution, and encourage others to follow their example.

Thus, it is not a question of whether the United States has enough institutional capacity, as you ask, but how it can develop appropriate kinds of institutions. Pouring money into the same institutional framework is unlikely to bring climate research into line with society's expectations.