Wednesday, August 13, 2014

Up to What Point?

This post is about the abortion debate, an atypical subject for me but I'll be writing about it in relation to my customary concerns about quantitative thinking. All I want to say about it is that whenever I try to follow the debate, I become frustrated by the vague propositions people use to frame the question. Here's an example, a graph from Pew:

The poll asks, "Do you think abortion should be legal or illegal in all or most cases?" The trouble with this is that there's no time axis in the question. I think for a large fraction of people---not everyone, certainly, but for a significant number of people---their actual views about this question depend strongly on how far along the pregnancy in question happens to be.

For a significant number of people the answer is, "Legal up to a point"---which makes it important to ask, "Up to what point?" But none of the polls seem to ask this question.

For each n = 1, 2, 3, ..., what percentage of Americans would outlaw abortion-on-demand in week if they could do so? Is there any research on this?

No comments: