Minorities and other marginalized groups have not always enjoyed the best relationship with science. In the 1930s, researchers from the US government started a series of experiments that recruited hundreds of African American men infected with syphilis, then left their disease untreated to study its natural progression. (The government did, however, provide free burial insurance.) More recently, American Indians from the Havasupai tribe sued Arizona State University in Tempe over claims that geneticists had collected and analysed blood samples from tribe members without obtaining proper consent. The two parties settled that suit last year. Indigenous peoples in other countries such as Australia also have historical reasons to be suspicious of mainstream scientists.

For more than a decade, US leaders have been trying to move beyond that troubled past and recruit minorities into science and engineering. There are strong moral arguments for doing so. But in times of massive budgetary shortfalls, morals do not guarantee funds. Congress and the public should recognize the powerful practical reasons to support programmes that aim to raise the numbers of minorities in science.

Minorities represent a relatively untapped pool from which to draw the next generation of scientists.

A key issue is that of numbers. There is concern in the United States about the shrinking proportion of home-grown scientists. Foreign-born students, particularly from China and India, account for almost all of the growth in the number of science doctoral degrees granted in America. And many then take their skills back home. Minorities make up a growing share of the US population and represent a relatively untapped pool from which to draw the next generation of scientists.

They also bring fresh ideas to research. This sometimes results in the pursual of topics that can help specific communities but have not managed to capture the attention of mainstream researchers. An example of which is Katie McDonald, who embarked on a research project as an American Indian student at a tribal college in Montana. She found higher-than-expected levels of mercury in local fish and has helped her own tribe to avoid health problems (see page 25).

Bringing more diversity into the ranks of researchers will help to overcome the lingering suspicion of science that persists in some minority communities. In doing so, it will encourage members of the public to accept the products of research, whether they are government health recommendations or reports about the changing climate. Without that kind of trust, researchers could see their work ignored by segments of the population.

For these and other reasons, the US government has poured substantial funds into pulling more underrepresented minorities into science. The National Science Foundation spent more than $110 million on this in 2010, and other agencies, such as the National Institutes of Health, NASA and the US Department of Education, also have programmes to boost minority participation in science.

These initiatives still have a long way to go. The National Research Council (NRC) reported last year that underrepresented minorities made up 28% of the US population in 2006 but accounted for only 9% of college-educated Americans in the science and engineering workforce.

And in some cases, the numbers are proving hard to move. In 2008, American Indians comprised just 0.7% of the bachelor's degrees awarded in science and engineering — a proportion that is unchanged since 2000. Science bachelor's degrees earned by black students has also stayed constant at 8.3%. For doctoral degrees, the figures are even starker. American Indians, who represent 1% of the population, earn only 0.3% of the PhDs in science and engineering. Black people make up 13% of the US population but accounted for just 3% of the doctoral degrees awarded in 2008 in these fields.

The problem creates a vicious cycle. Similar proportions of minority and white students enter university intending to study science. However, the completion rate for minorities is lower. Many factors contribute to this gap, according to the NRC, but one remains the poor diversity of university faculty members and the scarcity of role models in science for students from minority groups.