The UK Borders Agency has recently come under fire for looking into the use of DNA tests and isotope analysis to determine the true nationality of asylum seekers. It is not just refugee support groups who are outraged, scientists are equally upset (perhaps more). The problems are many: there is no reason to think ancestry and ethnicity fits with nationality, the relevant genetics and isotope data is noisy, the research may not have been vetted for reliability, and it is not inconceivable that noise in the tests could be used as excuses for dismissing people who actually have valid asylum reasons (like linguistic tests occasionally do).
The project is unfortunately just the latest example that governments may be too eager to buy snake oil: on this blog I have previously criticized the use of voice-based lie detectors, the legal use of fMRI to determine guilt, ethics for military robots, pre-emptive DNA testing and electronic voting machines. The problem here is not that these technologies can't work, but that they are deployed far earlier than any careful demonstration that they actually work well enough to fulfil their purpose. It is a "science fact" problem: it is hard these days to tell what has proven to work, what is being developed and what remains a theoretical possibility. Especially when it is being pushed by enthusiastic researchers and salesmen.
I believe individuals should be allowed to buy untested products as long as they are aware of their lack of evidence (this is why I have some problems with how much alternative medicine is sold, since it downplays the lack of testing and plays up anecdotal evidence and appeals to authority). But I think we should hold government branches to higher standards. When somebody puts crystals in their gas tank to improve mileage or drinks herbal tea against their cancer, the only one who might be hurt is the person. But if government institutions (or for that matter other powerful institutions like major corporations or NGOs) fall for it, their actions will have wide consequences. Putting crystals in gas tanks wastes money (and gives it to people prepared to sell badly investigated tools). Promoting tea rather than proven treatments for cancer will seriously harm many people. With greater power comes greater responsibility.
The problem with snake oil technology in government functions is that it 1) quickly gains a veneer of respectability by association, 2) significant amounts of public money get invested, and 3) this creates incentives for people involved to claim everything is fine since they now have a vested interest. The debate about the efficiency of CCTV in the UK is a good example, where people affiliated with the police and surveillance community claim great effect and outside groups find few if any (in particular, crime reduction might be overstated).
Is this a moral problem? I think so. If an action affects other's epistemic reasoning, especially in a morally sensitive area (e.g. genetics, deception, surveillance), then it is wrong to act without considering the risk that the act spreads misinformation. This corresponds to issue 1 and 3 above. There is also a problem of political responsibility corresponding to issue 2; wasting taxpayer money unduly is a breach of the trust citizens put in their elected officials.
What ought to be done? Scientists likely have a professional duty to speak out when they see this kind of mistakes, and also hold a moral responsibility for not overstating the case for their pet project. There is a need for more rational scepticism among government officials – given how many previous projects have failed and been a waste of effort, any new project should be scrutinized heavily. Ideally governments should act in a more evidence based manner. While this might be hard in many areas where cause and effect are complex, many of the above technologies could be tested far better. But testing means allowing dissenting voices and critical experiments to be done, and following even uncomfortable advice. It would imply that a suggested policy could be vetoed by science, something few politicians would like. So to achieve the overall goal the public must be clear in their demands that they want real results, not just impressive posturing.
As Richard Feynman said, "For a successful technology, reality must take precedence over public relations, for Nature cannot be fooled."