That's the title of an interesting post by Julian Sanchez.
Before answering the question, he first provides some context.
One thing to bear in mind is that even informed and intelligent people do not typically arrive at their political views by an in-depth review of the evidence in each particular policy area. Most of us can only be really expert in one or two spheres, and in others must rely heavily on those who possess greater expertise and seem to share our basic values. In practice, most people select a “basket” of policy views in the form of an overarching political ideology—which often amounts to choosing a political community whose members seem like decent people who know what they’re talking about. So we needn’t assume the majority view of the intellectual class represents the outcome of a series of fully independent judgments: A relatively mild bias in one direction or another within the relevant community could easily result in an information cascade that generates much more disproportionate social adoption of the favored views. So any potential biasing factors we consider need not be as dramatic as the ultimate distribution of opinion: Whatever initial net bias may exist is likely to be magnified by bandwagon effects.
It's a worthwhile point. Most people just can't be experts in so many areas, but one thing the Internet has done is made it easier and easier to choose from a wide selection of baskets of viewpoints or opinions. In other words, it's easier now than ever to sound smart on a wide range of topics. That may sound more condescending than I mean it to: often a gateway to forming one's own opinions is trying others on for size, and as with innovation in other fields often it's easier to build off of or react to another person's ideas than to birth one from scratch.
Back to the question posed in the title of Sanchez's post.
Here, then, is an alternative (though perhaps related) source of potential bias. If the best solutions to social problems are generally governmental or political, then in a democratic society, doing the work of a wordsmith intellectual is a way of making an essential contribution to addressing those problems. If the best solutions are generally private, then this is true to a far lesser extent: The most important ways of doing one’s civic duty, in this case, are more likely to encompass more direct forms of participation, like donating money, volunteering, working on technological or medical innovations that improve quality of life, and various kinds of socially conscious entrepreneurial activity.
You might, therefore, expect a natural selection effect: Those who feel strongly morally motivated to contribute to the amelioration of social ills will naturally gravitate toward careers that reflect their view about how this is best achieved. The choice of a career as a wordsmith intellectual may, in itself, be the result of a prior belief that social problems are best addressed via mechanisms that are most dependent on public advocacy, argument and persuasion—which is to say, political mechanisms.
It's worth contrasting the situation with the technology industry where a great deal of quarterback coaching and navel-gazing occurs online. It's not all just people signaling how smart they are: if you have a legitimate issue with something, the densely interconnected network of folks online will likely hear you if you express your opinion cogently.