More misestimations of Orleans population, registration

Sometimes, a blind pig finds an acorn. We observe a good example of this aphorism and its opposite (they often don’t find one) in a couple of announced studies of New Orleans’ population and voter registration totals which have in common that they both take rather debatable approaches to their investigations.

Greg Rigamer of GCR and Associates makes, to be frank, a mistake when he assumes just because registered voters in large part have not changed registration addresses to outside of Orleans Parish that they will be in Orleans to participate in the Apr. 22 and May 20 elections. As anybody familiar with the electoral behavior literature knows, it is hazardous at best to assume that just because displaced individuals have not changed their registrations that they faithfully will troop to Orleans Parish on election day – especially when the best estimates have roughly 275,000 Orleanians living out of state at the present, (and no more than a small fraction will avail themselves of early, mail-in, or satellite voting).

Yet the GCR Associates study the state paid for makes precisely the bad assumption that most on the registration rolls in Orleans will be present and ready to vote. The report’s author does not seem to understand that changing a registration in order to reflect the fact that they will not show up in April and May to vote, when faced with much bigger problems and entirely uncertain about when and if they will return to New Orleans, is probably the last thing on displaced Orleanians’ minds.

As I noted in a previous posting, my paper presented at March’s meeting of the Louisiana Political Science Association demonstrates, comparing historical rates of changes in registration totals to the present to show the actual “retention” rate of voters we can expect to be on the ground in Orleans on election day, these estimates are too optimistic on both the numbers that will be present (projected to be 130,000 in April, 138,000 in May) and in the proportion of blacks present (about 53 percent). (And my offer still is good: if anybody wants a copy of the paper, write me.) Data are no good unless you’ve got good theory behind it.

At least Rigamer actually does not overestimate too badly in both numbers and percentage of black population. Still, it’s disturbing that the state paid taxpayer dollars for an analysis that appears to be so unsophisticated and off the cuff (although neither is that inefficiency atypical of Louisiana government). Totally missing the mark is an academician who seems to offer little more than historical speculation and anecdotal evidence to say New Orleans won’t have a majority black population for some time (although, not having seen it, I’m having to go on the description afforded of it).

News flash: New Orleans did by the end of 2005 (about 80,000 blacks, 76,000 non-blacks, according to my estimates). Lesson (drawn from both examples): it takes good data used reasonably to draw truly valid conclusions about New Orleans’ future political geography.