I study how social welfare depends on strategic communication between a scientist and a policymaker and examine how the policymaker responds to electoral concerns using Crawford and Sobel’s classic model. Defining the social welfare to be the welfare of representative agents, whose preferences may be different from those of the policymaker and/or the scientist, I find that ex-ante social welfare is maximized, regardless of the preferences of the representative agent, if and only if the preferences of both the policymaker and the scientist are perfectly aligned. Communication is deliberately distorted by the policymaker when he has electoral concerns in a democracy.