Tuesday, August 30, 2005

I suppose I shouldn't be surprised...

... at writers of op-ed articles being less than 100% accurate. But something like this always makes me see red. Georgia has a new voting law which requires voters to show valid photo identification at the polls. Critics of the law contend that it discriminates against minorities, who are much less likely to have an acceptable form of ID. Defenders say that it is not discriminatory, but simply intended to reduce fraud. I have no strong opinion either way; I think requiring photo ID is desirable, but efforts should be made to ensure that every citizen can easily obtain a valid form of ID (which apparently is not the case in Georgia now, though the new law will go some way towards addressing this problem).

So what am I worked up about? These sentences in the linked article:
Of Georgia's voting-age population, 2,260,437 more people hold such identification than are registered to vote. Thus the number of voting-age citizens who lack photo identification cannot, as a matter of math, be large.
The phrase "as a matter of math" obviously implies that mathematics shows that the truth of their statement cannot be denied. Unfortunately, mathematics shows no such thing: If the only thing we know is that a set X is larger than another set Y, there is nothing obvious that can be said about the number of elements in Y that are not in X.

Granted that the writers are non-mathematicians, and may know nothing at all about set theory. We could be charitable and assume that they thought it was simply common sense, and would be supported by mathematics. But that doesn't support their assertion. Consider this statement, which is entirely equivalent (with the numbers reduced, but remaining roughly in proportion according to the U.S. Census Bureau):
"Of the 150 students in a high school, 50 more played basketball than baseball. Thus the number of students who play baseball but not basketball cannot be large."
Perhaps we have 30 who play only baseball, 80 who play only basketball, 30 who do both, and 10 who do neither. That sounds plausible, but one-fifth of the students play baseball but not basketball. Certainly any law which disqualifies a fifth of the voting population would be unacceptable. So common sense doesn't really help them.

Ok, so I'm overreacting; this isn't even a particularly egregious example of using 'mathematics' to mislead. But I wish there were a penalty for regular offenders... maybe insist that they take high-school math again?

No comments: