Science is science and policy is policy. In the end, if you want science to be useful, you have to recommend a policy, and scientists can do that, too. You just have to be clear on what is scientific results and what is policy recommendations based on said results.