In any political debate there are two currents underpinning each participant’s position: the preference-based one and the factual one. These aren’t typically distinguished very carefully, but they exist anyway. For instance, one could hold that more immigration from Latin America is desirable from a cultural point of view but that the economic absorptive powers of US society are not sufficient to support it. Or one could hold that it is both desirable and possible for the US law enforcement system to prevent large numbers of people from consuming some illegal drug. Etc, etc.
Not only is it important to separate these admittedly related currents, but it’s important to remember that different people and different debates are guided by different proportions in underlying justifications. It’s probably true that somebody, somewhere is guided by one exclusively without regard to the other, but in most cases these two exist side by side.
Especially since getting to graduate school I tend to rely less on (my-own-)preference-based justifications for my positions and more on factual ones. Obviously, there is still room for debate no matter which part you choose to rely on more; it could be that I have incorrect beliefs about facts, however much I try not to, and I could debate with somebody having the same preferences I have about alternative means of realizing them. The Affordable Care Act is a popular subject of these kinds of debates—I think most people probably want affordable, quality health care to be available to large numbers of people, but there is considerable room for disagreement about whether the ACA is an efficient way to accomplish this.
I don’t mean to slight preferences by saying they are all beyond rational investigation. Philosophy informs (creates?) our preferences, and to the extent that philosophy is about discovering truth, there is room for debate on correct positions. But this is not what I’m expert in, so I don’t make a habit out of it.
It shouldn’t automatically be cause for suspicion when preferences and factual suppositions point a person in the same direction. In fact, it’s probably the exceptional case in which they don’t. But in the interests of intellectual honesty it’s important to make the distinction so we know what we’re talking about and why. And in the interests of strategy, of “dialogue”, it’s important to remember that other people may not use the two threads in the same proportion that you do. Not only will this help you get to the root of the problem quicker, but it will help you from seeing the world like a medieval morality play in which every conflict is good vs. evil. Related: Which Side Are You On? Robert Wolff, Murray Rothbard, and Me by David D. Friedman.
Randall McElroy III 
Two philosophical objections:
a) First, a mostly terminological point: if moral realism is true (and it is, sez I), then values are a subset of facts, not an alternative to them.
But b) more importantly for present purposes, and leaving the terminological issue aside, I don’t think it makes sense to talk about relative strengths of normative vs. positive considerations, because neither one generates any action recommendation in the absence of the other. The positive belief “policy X will have result Y” is motivationally inert unless accompanied by a normative judgment as to whether result Y would be good or bad. Likewise, the normative belief “result Y would be good” is motivationally inert unless accompanied by a positive belief as to what I can do to bring about result Y. Since neither one has any motivational upshot absent the other, I don’t know what it could mean to weigh one more than the other –or how it could possibly be true that “somebody, somewhere is guided by one exclusively without regard to the other.” Nobody can be guided by one more than the other, since without the interaction of both there’s no guidance.
And your examples don’t seem to help: your immigration case, for example, seems to involve the judgments “immigration policy X would have results Y and Z”; “result Y would be good”; and “result Z would be bad.” In that case you have to weigh the values of Y and Z against each other, but you’re not weighing values against non-evaluative facts.
My argument was more restricted in scope, I was thinking about policy debates. Also, I said preferences instead of values because it seems to me most people don’t think about preferences too deeply, and it helps to keep me in line talking about things I know more rather than things I know less about. It’s anecdotal, but I think some people weigh the relative contributions differently: Tea Partiers and living wage advocates more on the preference side and think tank wonks more on the factual side.
Aside from that, I agree that in the “graduate” version of this “undergraduate” discussion there would be much more to include.
But I don’t know what “more on the preference side” and “more on the factual side” mean. Suppose I favour X because I believe (“factually”) that it will lead to Y, and I (“preferentially”) value Y. Does my position lean more to the factual side or the preferential side? If it does one of those things, how would you have to change it to make it do the other?
Another unstated assumption is the relatively uncontroversial idea (among economists!) that people don’t vary THAT much in preferences. If this is the case, we should debate a lot more about means, i.e. strictly about facts.
However, we could take an outlier, a KKK member for example, who may read the literature and realize that allowing more immigration would lead to more economic growth (which he prefers, ceteris paribus), but who thinks that whatever the other likely consequences there should be less immigration because more foreigners being here is ipso facto against his preferences. This is somebody motivated purely by preferences. Most of us are not like this, of course.
But I still don’t understand what you mean by someone “motivated purely by preferences.” Suppose Stan believes a) that laxer immigration laws would have two results: b) more economic growth, and c) more foreigners being here. Stan likes (a), but dislikes (b) even more than he likes (a). So Stan opposes laxer immigration laws. How is he more motivated by preferences than:
someone who disbelieves (a)?
someone who believes (a) but likes both (b) and (c)?
someone who dislikes (b), but likes (a) more than he dislikes (b)?
I just don’t understand what difference you’re talking about.
Sorry, I screwed up the letters. Ignore the above, here’s the right version:
But I still don’t understand what you mean by someone “motivated purely by preferences.” Suppose Stan believes a) that laxer immigration laws would have two results: b) more economic growth, and c) more foreigners being here. Stan likes (b), but dislikes (c) even more than he likes (b). So Stan opposes laxer immigration laws. How is he more motivated by preferences than:
someone who disbelieves (a)?
someone who believes (a) but likes both (b) and (c)?
someone who dislikes (c), but likes (b) more than he dislikes (c)?
I just don’t understand what difference you’re talking about.
At one point I thought maybe you were talking about the difference between valuing something as a means and valuing something as an end. But that doesn’t seem to line up too well with your examples either.
You’re looking at it from the opposite direction. We don’t know Stan’s preferences when we encounter him in a friend’s comment thread on Facebook saying the legislature should pass a living wage bill. We could jump to conclusions, as is common, and assume that Stan hates the poor and wants fewer of them to be employed. If that is the case we will use one line of reasoning with him. If instead he wants the poor to have jobs but has incorrect factual beliefs about what will promote that, we will use a different line of reasoning that doesn’t include the supposition that he is a bad person. We also don’t know ex ante how motivated his support for a policy position is by either his preferences or his factual beliefs, so we will have to dig a little.
I suggest that we will find more often than not that people have similar preferences (about end states) to ours, so we should usually talk about means more than desired ends. Anecdotally, when I hear people supporting living wage legislation it’s based on preferences and not at all on factual beliefs about what the real-world consequences will be, whereas when economists write policy papers some broad social goal is usually implicit while 99.9% of the paper is just about facts. So there’s not a stock answer to your questions above.
As a philosopher you could unpack the concept of preferences a lot more than economists do, or than we have to in order to get people not to automatically assign “good guy” and “bad guy” labels to supporters or opponents of a policy.
We could jump to conclusions, as is common, and assume that Stan hates the poor and wants fewer of them to be employed. If that is the case we will use one line of reasoning with him. If instead he wants the poor to have jobs but has incorrect factual beliefs about what will promote that, we will use a different line of reasoning that doesn’t include the supposition that he is a bad person.
Fine; and those two lines of reasoning do differ according to whether we take his disagreement with us to be primarily positive or primarily normative. But those interpretations don’t differ in treating his position itself as primarily positive or primarily normative (whatever that would mean); nor do the interpretations themselves differ in being primarily positive or primarily normative (again, whatever that would mean).