Sean Kay recently wrote an interesting article for the Duck of Minerva, arguing that academics can contribute to the policymaking process, especially through their ability to inform policy. As an academic with some exposure to decisionmaking processes in Indonesia, let me make several points. From my viewpoint, there are four major problems for academics who are hoping to be influential in domestic or foreign policy.
First, to put it bluntly, most academics lack access to the relevant decision-makers. While Kay did state the need for senior policy makers "to provide a climate that incentivizes policy staff time to gain fresh perspectives engaging with scholars," the real question is whether the ones that really matter, the ones who actually make decisions, will listen to them.
While a scholar might be a household name in academia, in the government, s/he is just one voice in the wilderness, especially with so much information to sift, data to analyze, and various competing interpretations swirling -- unless s/he knows someone important. At the same time, someone can be a mediocre and unimportant academic, and yet, thanks to his or her connections, they can call the president at 3 AM to offer thoughts. In essence, connection matters.
Second, as Jeffrey Stacey and Robert L. Gallucci noted, scholars need to drop the jargon of the academy and speak in the terms that policymakers understand. In short, keep it simple, stupid. Stacey is correct, that proper academic training does help immensely in dealing with policy-making, but not in the sense that he suggests, that "when the topic of democratization came up, no one else in the room knew that democracies do not fight wars with one another."
Honestly, that's a really dumb example for two simple reasons: (1) there is a voluminous literature in the IR debating whether that assertion is accurate, and more importantly (2) policymakers really don't care whether democracies fight wars with one another. What they care about is whether the nuggets that scholars bring to the table fit nicely with their goal and agenda, and that's to rise to the top of food chain.
Academic training helps give people a strategic-theoretical framework that they can employ to create a coherent "analysis" out of pile of the God-knows-what is on their desk. This analysis functions as a story, or a narrative, that's then told to relevant policymakers. This may or may not be a good thing, but in the end, analysts typically need to figure out what kind of actions they should recommend and whether those suggestions will bite them in the ass in the future. That's the best contribution graduate school can give people to prepare for a policy job, not informing us whether democracies will fight with each other.
Policy analysts read and digest articles and researches in order to improve the theoretical models inside their head, so when they are asked whether the proposed policy is good or not, they can confidently give input. However, the dirty secret is that analysts might not know what the hell is going on, because important matters frequently arise suddenly, out of the blue, such as during crises like the Arab Spring, where policymakers were all running like chickens with their heads chopped off and needed something to make sense of everything. Remember, policymaking is not Jeopardy. Trivia can help bolstering one's credibility, that he or she actually knows something, but at the same time, they have strong incentives to care about only the things that those higher in the food chain care about.
Third, and I learned it the hard way, is that academics often don't even understand the culture inside bureaucracy. This is what's going on inside bureaucracy. (Picture stolen from here)
Put simply, there are things that make sense from academics' perspective, and yet, thanks to the ingrained bureaucratic culture, it simply cannot be done, because that's how things work. For instance, I wrote a policy paper for my superior, stating that Indonesia needs to do "X." It came back with one reply, "that policy you suggested is dead on arrival because the top brass will never ever want to entertain that policy." So my paper was shot down, not because my policy recommendations weren't sound and sensible, mind you, but because they worked against every grain of the existing bureaucratic culture. It's clearly evident that analysts know what can or cannot be done only after he or she immerses themselves for quite some time within the relevant bureaucratic machinery.
Fourth and finally, if academics are really interested in influencing policy making, then they ought to make themselves either (1) a total expert on one important subject that has relevance to policy-making, e.g. military and politics in Latin America or (2) someone who knows a little bit of everything: economics, politics, psychology, sociology, public opinion, math, etc. No offense to any constructivists or feminists out there, but being an expert in a very obscure subject such as, lets say, "feminist movements in the Tibetan plateau" won't help cultivate a career in policy. It might be a very important subject in academia, but unfortunately, there's almost no relevance to real world policy making.