The Volokh Conspiracy

Andrew Sullivan, InstaPundit, and Ann Althouse link to this quiz that purports to measure where you stand on the liberal-conservative spectrum. I tend to agree with InstaPundit: "it was a pretty dumb test."

"Which do you trust more: The Pentagon or The U.S. Postal Service?," the test asks. Well, trust to do what? The USPS is generally pretty good at delivering the mail. Its success rate at its tasks (delivering each letter entrusted to it) is almost certainly higher than the Pentagon's success rate at its tasks (accomplishing every military objective that it's asked to accomplish). On the other hand, the degree of difficulty is a bit higher for the Pentagon's tasks, no?

Likewise, which do I trust more: trial lawyers or doctors? What does that mean? Which is more honest? I expect both are on balance roughly equally honest, just like most professional groups are on balance roughly equally honest. Which is better at accomplishing the tasks that are entrusted to it? Hard to tell. Which better serves the country? Also hard to tell, given how large a group trial lawyers are (if you interpret trial lawyers literally, they'd include defense-side lawyers as well as plaintiff-side lawyers, plus of course criminal lawyers on both sides); but that's pretty clearly not a matter of "trust."

Or how about Q 10: "Which would curb violent crime most? [A] Stricter controls on the sale of guns [B] Mandatory sentences for those who use guns in the commission of a crime [C] Both." What about "neither," if you think that mandatory sentences are a bad idea (because you believe in leaving judges with considerable discretion) but you also think gun control is a bad idea — or for that matter if you think that both A and B are just unlikely to have any real effect?

Similarly, consider Q 11: "In the long run, do you think we can reduce crime more by building more prisons or providing more financial assistance to rebuilding our inner cities? [A] Build prisons [B] Rebuild cities [C] Both." What answer should be given by a libertarian who thinks that neither "building more prisons" nor "providing more financial assitance to rebuilding our inner cities" will reduce crime?

Or consider Q 17: "The religious right is a threat to our political system [Agree / Disagree]" — what about people who strongly disagree with the religious right but see it as just a wrongheaded part of our political system, and not a threat to it?

I realize that all these tests, especially ones that try to collapse your views onto one dimension, are flawed; but this one strikes me as especially flawed.

	[Eugene Volokh, January 5, 2007 at 3:41pm] Trackbacks Trust Test: Andrew Sullivan, InstaPundit, and Ann Althouse link to this quiz that purports to measure where you stand on the liberal-conservative spectrum. I tend to agree with InstaPundit: "it was a pretty dumb test." "Which do you trust more: The Pentagon or The U.S. Postal Service?," the test asks. Well, trust to do what? The USPS is generally pretty good at delivering the mail. Its success rate at its tasks (delivering each letter entrusted to it) is almost certainly higher than the Pentagon's success rate at its tasks (accomplishing every military objective that it's asked to accomplish). On the other hand, the degree of difficulty is a bit higher for the Pentagon's tasks, no? Likewise, which do I trust more: trial lawyers or doctors? What does that mean? Which is more honest? I expect both are on balance roughly equally honest, just like most professional groups are on balance roughly equally honest. Which is better at accomplishing the tasks that are entrusted to it? Hard to tell. Which better serves the country? Also hard to tell, given how large a group trial lawyers are (if you interpret trial lawyers literally, they'd include defense-side lawyers as well as plaintiff-side lawyers, plus of course criminal lawyers on both sides); but that's pretty clearly not a matter of "trust." Or how about Q 10: "Which would curb violent crime most? [A] Stricter controls on the sale of guns [B] Mandatory sentences for those who use guns in the commission of a crime [C] Both." What about "neither," if you think that mandatory sentences are a bad idea (because you believe in leaving judges with considerable discretion) but you also think gun control is a bad idea — or for that matter if you think that both A and B are just unlikely to have any real effect? Similarly, consider Q 11: "In the long run, do you think we can reduce crime more by building more prisons or providing more financial assistance to rebuilding our inner cities? [A] Build prisons [B] Rebuild cities [C] Both." What answer should be given by a libertarian who thinks that neither "building more prisons" nor "providing more financial assitance to rebuilding our inner cities" will reduce crime? Or consider Q 17: "The religious right is a threat to our political system [Agree / Disagree]" — what about people who strongly disagree with the religious right but see it as just a wrongheaded part of our political system, and not a threat to it? I realize that all these tests, especially ones that try to collapse your views onto one dimension, are flawed; but this one strikes me as especially flawed.