Will Wilkinson has a solid takedown (and an interesting followup) of Jon Chait's self-satisfied New Republic piece, which proposed that liberals are distinguished from conservatives by their relentless empiricism and flexibility.
Two brief additions. While I generally agree with Will, I don't think Chait's wholly off base either: I do think liberals tend to focus more on empirical details, though that may just be my own sample bias. But I'm not convinced that's always better. Assume for the moment that we're dealing with an issue where, per impossible, there isn't actually a deeper normative divide turning on, say, the intrinsic value of choice as opposed to efficient promotion of health or widget production or whatever. One of Hayek's great insights was that the point of adhering to principles is to prevent you from having to make case-by-case decisions. This, Hayek suggests, is because the great virtue of markets isn't the efficient allocation of "given" resources to satisfy "given" human needs or wants, but rather the discovery process they facilitate. The best case for allowing markets greater leeway was, he thought, that regulators and planners couldn't anticipate what millions of entrepreneurs and consumers might invent or want. And since these unknowns are precisely the market's strength, a case-by-case examination would always tend to err too far in favor of regulation, because planners would incorporate the gains from past learning and innovation, perhaps even figure out a more efficient way to satisfy old needs by old means, and the gains from future innovation erased by intervention would remain invisible. If Hayek was right, then, a certain indifference about the empirical details in a particular debate about the wisdom of innovation could constitute a kind of rational dogmatism.
The second thing I want to add is more an illustration by way of anecdote of Will's general point. Last summer, I did a debate on the Kojo Namdi show with a guy from SmokeFree DC about the wisdom of a proposed smoking ban. One of the callers was a doctor who said something to the effect of: "Well, the evidence for how harmful cigarettes are is so overwhelming that only some kind of ideology could inspire you to oppose such a clearly benificial measure."
Now, I'm not wholly sold on the case against secondhand smoke, but that's beside the point, since I hadn't really been disputing that cigarette smoke is unhealthy. The real difference between us was that my ideology was that people should be free to make choices that are unhealthy, and his ideology was that a healthy population is the summum bonum, and if people don't know what's good for them, there's nothing intrinsically good about letting them make "bad" choices.
We moved on to other topics, so I didn't have a chance to bring this point out, but what was interesting was that he seemed to think that I was the only one in the argument working from ideological premises. In a sense, his expertise prevented him from recognizing the real bone of contention, because the empirical data swamped and obscured the normative component. Hyperfocus on the means can, it seems, serve as a distraction from reflection on the end.