(Response composed while reading, so it’s a little disorganized, and most of it wasn’t written after I’d read the whole post. If something seems like a tangent to or misunderstanding of what you wrote, please assume that it is, but if it’s interesting feel free to remark on it anyway.)
I find the interaction between
Generally, if a politician supports one “Republican” position, he’ll support most of the other “Republican” positions.
and
Suddenly, it’s not one-dimensional any more. There are people who like low taxes and also like gay marriage.
noteworthy. Is the contrast that when making policy, people follow the party line, but when polled about their real opinions, they stray more? Or is it that politicians whose job security depends on pleasing a certain group of people will play it safe, and private individuals are freer to express their varied views? Or are those the same thing, in practice? It seems like you’re comparing apple politicians to orange citizens, but I might be missing something.
People who describe themselves as Democrats, for example, span a much wider range of views than people who describe themselves as “Republicans.”
Interesting! I didn’t know that. It doesn’t surprise me, but I wouldn’t have trusted my intuition about it—it’s easy for me to see more variety among Democrats because I live in an area that has lots of them. (I actually have unusually specific data about that, because I’ve been a poll worker and thus have skimmed through the entire list of party affiliations for a nearby district. IIRC there were more Green Party members than Republicans.)
If you support spending more on X, you must either support less spending on some other thing Y, or support raising taxes, or support running a higher deficit.
That depends on whether you’re polling people about practical policy or about what they wish could/would occur. These are two very different questions. I may very much want to increase the budget for education, but simultaneously understand that it’s not feasible to do so at the moment.
You cannot simultaneously be in favor of reducing carbon emissions, and be in favor of a set of policies that, on net, increase carbon emissions. Or, you can, but you’d have to be either ignorant or confused.
Similarly, if poll questions took into account the link between these policies, we’d rob ourselves of data about how many respondants are ignorant or confused. I think you underestimate the significance of the fact that people DO hold conflicting positions, sometimes strongly, and this puts the people who actually have to make feasible policy in an awkward position.
Or is that the point? That popularizing the second model encourages people to hold incompatible positions, whereas encouraging the use of the third encourages more practicality? I’m not sure I would try to push people to hold beliefs about politics or policies that are internally consistent. If I tell you that I want two things of which I cannot plausibly have both, I’ve told you more about my preferences than I would have if I’d just told you what compromise I would settle for. As the situation changes in the future, it might become possible to reconcile my two previously incompatible preferences, and it’s useful to know that I’d support that.
That said, I really like your description of a graph of policy preferences. I think it’s very useful for the purpose you describe—identifying root differences—and not just in politics. Two people who disagree strongly about which of two anime shows is better are more similar than two people who disagree about whether animated television is inherently childish and a waste of time. Seen in a certain light, this is also very much like the ancestral/derived trait analogy I used in the OKC post a while back: “ancestral” preferences would be the influential nodes, and “derived” ones the leaf nodes. I wonder where else interesting one can go with that pattern.
“Hi, I actually support monarchy,”
This made me laugh because I imagine it being used as a greeting. (“Hello, I’m Jesse.” “Hi, I actually support monarchy.”)
I think both your conclusion and your example of it are excellent, but I feel like the final point deserves more words than it got. I’m left wondering about specific major instances and ramifications of making the error you describe. (Is there already a name for it? If not, can we call it the space squid fallacy? Is there some bureau that decides these things?)
I think that there’s such a thing as party discipline among politicians, such that breaking with one’s party is a career risk, but ordinary folks can believe whatever they want to believe with little cost.
I don’t want to eliminate polling and replace it with something else. Polling describes people’s opinions; my “graph” describes the logical and causal relationships between policies. They describe different things. Also, making a graph for all of politics is impossible—you’d have to be Laplace’s Demon to do that. I could imagine making a very small toy example on a very specific project.
I agree that the graph setup is useful for analyzing disagreements and value differences in general, not just political ones.
The “fallacy” isn’t exactly a fallacy in reasoning, but it’s instrumentally irrational. One example would be infighting between very similar groups: “People’s Front of Judaea” versus “Judaean People’s Front.” It’s tempting to spend all your time fighting your friends—but if you focus on fighting over which anime shows are best, you won’t get across the message to people outside your subculture that anime in general is a good medium.
It’s tempting to spend all your time fighting your friends—but if you focus on fighting over which anime shows are best, you won’t get across the message to people outside your subculture that anime in general is a good medium.
Indeed. The value of which depends on why you’re arguing, I suppose—you might be artists trying to gain acceptance, but you might just be trying to decide what to watch tonight.
(Response composed while reading, so it’s a little disorganized, and most of it wasn’t written after I’d read the whole post. If something seems like a tangent to or misunderstanding of what you wrote, please assume that it is, but if it’s interesting feel free to remark on it anyway.)
I find the interaction between
and
noteworthy. Is the contrast that when making policy, people follow the party line, but when polled about their real opinions, they stray more? Or is it that politicians whose job security depends on pleasing a certain group of people will play it safe, and private individuals are freer to express their varied views? Or are those the same thing, in practice? It seems like you’re comparing apple politicians to orange citizens, but I might be missing something.
Interesting! I didn’t know that. It doesn’t surprise me, but I wouldn’t have trusted my intuition about it—it’s easy for me to see more variety among Democrats because I live in an area that has lots of them. (I actually have unusually specific data about that, because I’ve been a poll worker and thus have skimmed through the entire list of party affiliations for a nearby district. IIRC there were more Green Party members than Republicans.)
That depends on whether you’re polling people about practical policy or about what they wish could/would occur. These are two very different questions. I may very much want to increase the budget for education, but simultaneously understand that it’s not feasible to do so at the moment.
Similarly, if poll questions took into account the link between these policies, we’d rob ourselves of data about how many respondants are ignorant or confused. I think you underestimate the significance of the fact that people DO hold conflicting positions, sometimes strongly, and this puts the people who actually have to make feasible policy in an awkward position.
Or is that the point? That popularizing the second model encourages people to hold incompatible positions, whereas encouraging the use of the third encourages more practicality? I’m not sure I would try to push people to hold beliefs about politics or policies that are internally consistent. If I tell you that I want two things of which I cannot plausibly have both, I’ve told you more about my preferences than I would have if I’d just told you what compromise I would settle for. As the situation changes in the future, it might become possible to reconcile my two previously incompatible preferences, and it’s useful to know that I’d support that.
That said, I really like your description of a graph of policy preferences. I think it’s very useful for the purpose you describe—identifying root differences—and not just in politics. Two people who disagree strongly about which of two anime shows is better are more similar than two people who disagree about whether animated television is inherently childish and a waste of time. Seen in a certain light, this is also very much like the ancestral/derived trait analogy I used in the OKC post a while back: “ancestral” preferences would be the influential nodes, and “derived” ones the leaf nodes. I wonder where else interesting one can go with that pattern.
This made me laugh because I imagine it being used as a greeting. (“Hello, I’m Jesse.” “Hi, I actually support monarchy.”)
I think both your conclusion and your example of it are excellent, but I feel like the final point deserves more words than it got. I’m left wondering about specific major instances and ramifications of making the error you describe. (Is there already a name for it? If not, can we call it the space squid fallacy? Is there some bureau that decides these things?)
I’ll reply to a few of your points.
I think that there’s such a thing as party discipline among politicians, such that breaking with one’s party is a career risk, but ordinary folks can believe whatever they want to believe with little cost.
I don’t want to eliminate polling and replace it with something else. Polling describes people’s opinions; my “graph” describes the logical and causal relationships between policies. They describe different things. Also, making a graph for all of politics is impossible—you’d have to be Laplace’s Demon to do that. I could imagine making a very small toy example on a very specific project.
I agree that the graph setup is useful for analyzing disagreements and value differences in general, not just political ones.
The “fallacy” isn’t exactly a fallacy in reasoning, but it’s instrumentally irrational. One example would be infighting between very similar groups: “People’s Front of Judaea” versus “Judaean People’s Front.” It’s tempting to spend all your time fighting your friends—but if you focus on fighting over which anime shows are best, you won’t get across the message to people outside your subculture that anime in general is a good medium.
Indeed. The value of which depends on why you’re arguing, I suppose—you might be artists trying to gain acceptance, but you might just be trying to decide what to watch tonight.