Our build-up to Tuesday’s hearing on the ACLU’s case against Anaheim continues! Part 1 offered the basics (and a couple of video interviews.) Part 2 covered eight slides from the Demographer’s Report that appears to be Anaheim’s primary line of defense. Welcome to Part 3 — more slides, hopefully less commentary. (We’ll see about that.) Continuing from where we left off …. which I believe would be Slide 9.
Slide 9, 2008 City Council Results Show Latino(ish) Power(ish): As you’ll recall from reading Part 2 of this series … which I’m sure you’ve read …
OK, go back and read it … we’ll wait … welcome back! … did you refer back to Part 1 as directed? … OK, go read Part 1 … no, you don’t have to if you don’t want to …
… the Demographer’s Report — the work of hired Demographer Dr. Peter Morrison that is supposed to save Anaheim from a fate worse than … worse than the suppression of the political power of Latino communities that exists now (at least from the perspective of the City Council’s governing Murrjority) …
see, if you’d read at least one of the previous parts, then you’d know what a “Murrjority” is!
… anyway, the Demographer’s Report was trying to prove that Latinos are actually really successful in Anaheim politics, despite any appearances to the contrary, by showing the John Leos finished third in City Council races of 2010 and 2012 — each of which filled two seats. (The fact that Leos ended up exactly as much “on the Council” as Rudy Gaona, who finished last in 2010 and next to last in 2012, is apparently not important. See Part 2 for more fun being made of this notion.)
We now move to the 2008 election results in which not just one name showing the power of Latino candidates (that being Leos) gets circled, but two names get circled: Harry Sidhu and Lorri Galloway!
It should be noted for the record that Harry Sidhu, while tan of skin, is of Asian Indian ancestry and is not Latino. (Using the vocabulary work of the week, “one does not operationalize Latino as including “people from India.”) The point being made, I suppose, is that whites don’t have a stranglehold over the City Council — the reverse of which which, as you might expect, is not an allegation that needs to be proven in the ACLU’s case.
But enough about Sidhu. Let’s operationalize Lorri Galloway. Does she count as Latino or not? (If not, would she count if she pronounced her name as “guy-yoh-ay”?) Her ancestry is half-Filipino and half-Spanish. That doesn’t mean that she’s not Latino. This, however, does: Galloway does not trace any of her ancestry back to Latin America. The Spanish half of ancestry is from Spain. She’s like an American who’s French or Italian, not like one who’s Mexican, Guatemalan, Salvadoran, or Colombian….
Can I pause for a moment to give voice to what you’re probably thinking: how utterly stupid this is?
The idea that Anaheim’s compliance, or lack of compliance, with the California Voting Rights Act should turn on whether Lorri Galloway is completely idiotic. The ACLU, I presume, would like to focus on more traditional measures such as whether Latinos vote differently from non-Latinos and whether non-Latinos get their way, in an at-large voting system (one without districts) more often than one would predict. THAT is not stupid. But the City of Anaheim wants to say that Galloway’s twice being elected to Council proves that Latinos have an equal shake — that is, they want to talk about the success of Latino candidates getting elected rather than the success of Latino voters electing the candidates of their choice. This is the wrong measure to use — the wrong operationalization of “success” — but hey, the ACLU can play that game if the city wants to.
So, the ACLU points out that under the law — nope, she doesn’t count as Latino. And that led to this post from the Anaheim business establishment’s paid yappy-dog, Matt Cunningham:
[Quoting a Voice of OC story]
The city challenges the validity of the lawsuit in court documents, arguing that members of minority groups have consistently been elected to the City Council. According to the city, 10 seats have been up for election since 2002, with seven of those seats filled by either “Asian” or “Hispanic” council members.
One of those council members is former Councilwoman Lorri Galloway, who is Spanish and Filipina, a mix that ACLU attorney Robert Ruben said doesn’t meet the criteria for Latina under the Voting Rights Act.
What a sad, infuriating spectacle. But not an unexpected one. When racial bean-counting is the coin of the realm for acquiring political power and “representation,” should we be surprised when such argument breaks out over whether or not someone is really Latino (or Asian or whichever ethnicity is deemed in need of increased “representation)? If single-member council districts become a reality in Anaheim, we can expect more of such demeaning spats.
Aside from giving Matt a round of very slow applause for that “racial bean-counting” remark in the context of discrimination against Latinos, he doesn’t seem to get that the ACLU is simply talking about what criteria are established by the law. It’s not a matter of what should be, but a matter of what is. Matt is like the guy pulled over doing 90 on the 5 near Camp Pendleton who yells at the officer because he doesn’t think that there should be speed limits in undeveloped areas. But that’s the law, Matt!
So, if Dr. Peter Morrison, the Demographer of the Demographer’s Report, wants to count Galloway as a Latina for the purpose of proving that Anaheim doesn’t discriminate, you would think that he would either use the legal definition of Latino or at least put an asterisk next to her name and run the results with and without her. But, as you’ll see, he didn’t.
This affected his results. But before we get to that …
Slide 10, Look, Richard Chavez also finished third in 2006! The only actually interesting thing in this slide is that it reminds us that Chavez — along with Bob Hernandez — both won election to the City Council in 2002. Both were Latinos; both were firefighters. That’s … unusual, isn’t it? Yes, it is! In fact, one might call it an outlier, one of those data points that is so unusual that taking it into account actually messes up one’s predictions! You see, this was the first election that occurred after 9/11 — and did I mention that they were firefighters? (As I understand it, they sure did!) I asked Dr. Morrison about this in my first interview with him, which you can read and watch at the bottom of Part 1:
GD: Did you have any concern that 2002 was an outlier?
PM: I had heard that it was the wake of 9/11 and that some regard it as an outlier.
GD: No, an outlier in terms of the results of the elections. Two Latinos are elected.
PM: Well, I don’t know what you mean by an “outlier.” I mean, it’s a situation where two got elected, and somebody said it was because they were firemen because of 9/11.
This gives me a good chance what I — and I think most social scientists — mean by “outlier”: it’s a data point that is so discrepant from others that including it in the description of the present results or especially in the prediction of future results would leave one making worse predictions than one would otherwise, because the special and unusual circumstances that led to the result are unlikely to recur. When you come upon an outlier that seems to be driving your results, you don’t say “yippee, I lucked into an outlier that makes my hypothesis seem to be more likely than it probably is!” You probably exclude it from your dataset and deal with it separately, as in: “this analysis excludes data from 2002’s election — which seems to show that, in the wake of a national catastrophe after which firefighters are venerated, firefighters will tend to win the next election even if they are Latino.”
This is about to become important.
Slide 11, Council Election Winners, 2002-2012: Here’s what the Demographer did next. He added up all of the 64 candidates for Anaheim City Council who ran in the six elections from 2002 to 2012– not including the Mayors, even though the Mayors were also chosen at-large in the exact same elections. (I feel obligated to note that including the Mayors an unbroken string of six “non-H white terms,” would have severely undermined the city’s defense.) And then he said: “OK, now how likely were you to win the election if you were part of a given race or ethnicity?”
NOW YOU ARE ABOUT TO FIND OUT WHY I HAVE BEEN GOING ON AND ON ABOUT OPERATIONALIZATION!
Let’s not worry about the three people for whom a race could not be determined. (If it turns out that they were white, of course, then including them in the analysis would have again damaged the city’s defense.) Let’s just look at the other three categories: 29 (non-Hispanic) whites, 18 Latinos, and 14 Others. (“Other” includes Filipinos, Blacks, Asians (both East and South — and maybe West), and other Others, such as Native Americans.) How many of the 12 winners of these six elections fall into each category?
(I’d like to point out, by the way, that one generally doesn’t try to make broad generalizations from a sample size of 12. It really doesn’t lead to stable or reliable results. But tossing that aside with all of the other problems….)
Well, who were the winners?
Bob Hernandez (2002, 2006) — LATINO
Richard Chavez (2002) — LATINO
Lorri Galloway (2004, 2008) — CLASSIFIED AS LATINO
Harry Sidhu (2004, 2008) — OTHER
Lucille Kring (2006, 2012) — WHITE
Gail Eastman (2010) — WHITE
Kris Murray (2010) — WHITE
Jordan Brandman (2012) — WHITE
And from that — and from ignoring 2 wins for Mayor by Tom Tait (WHITE) and 4 wins by Curt Pringle (GLEAMING WHITE) — Dr. Morrison concludes that Latinos do better in Anaheim than Whites or Asians because they win more often.
As I said, there are all sorts of problems here. The small sample size really doesn’t allow you to produce valid results. Candidate success isn’t really the right measure of voter empowerment to examine. Going back (intentionally or not) to include the only election in which two Latino candidates were elected and then no further is odd, unless there is a strong reason to suspect that there was something different about 2000 and before than is present now. (My guess is that we’ll find that something major did change in city politics — but more like in around 2006-2010.) The researcher knew that 2002 was a likely anomaly — and for reasons that don’t actually favor the election of Latinos in 2014 and beyond.
The main justification for doing the research this way is, essentially, that this is all we can do. This is the easiest approach. And it happens to be exactly what the client wanted to see. (And, one might cynically suppose, the check is going to clear.) But let’s put that aside and ask: does it make sense on its own terms?
No, not really — although it’s better than the slides that come next.
The first thing we can note is that re-election in Anaheim has been almost automatic, with only Chavez losing (by a few hundred votes) in 2006. So: Hernandez in 2006, and Sidhu and Galloway in 2008 don’t actually tell us much about the effect of race or ethnicity so much as about the power of incumbency — leaving us with (taking out Chavez 2006 in as well) only eight races to consider. But let’s be gracious and put them back in.
Let’s take another look at those five Latino winners. Hernandez and Chavez are (so far as I know) clearly Latino, but 2002 was an outlier — an anomalous year (firefighters after 9/11) that doesn’t tell us much about the future — and 2006 was a re-election bid. Lorri Galloway is not of Latin American ancestry, but European and Asian/Pacific Islander. (Maybe she should count as “half a Latino and half an Asian” for these purposes, but no more. That other famous political biracial guy, Barack Obama, gets counted as Black because he’s seen as Black. If a survey showed that the Filipino-Iberian with the Irish surname Galloway is really seen by voters as Latino, then I might change my tune — but for now I don’t see that her success should lead us to predict that future Latinos should do well (although the Irish surnamed Filipino-Iberians may take heart.)
So, among the five “Latino” victories, I’d say that one — Hernandez in 2006 — may actually have much merit in predicting what will happen in Anaheim at-large elections from 2014 and beyond, and that one was a re-election. Counting Galloway as an “Other,” and stopping before 2002 (five races isn’t much worse than six for predictive purposes — they’re both lousy) and adding back in the at-large Mayor’s races, we get these figures:
LATINO — 14 or fewer candidates, 1 victory
WHITE — 29 or fewer candidates, 5 victories
OTHER — 16 or fewer candidates, 4 victories
That gives a very different picture of where Anaheim stands (even though it’s still asking the wrong question.) Luckily things can’t get any worse than this … or can they?
Slide 12, Had Top 3 Candidates Won, 2002-2012: Oh dear. I’m not going to repeat the section I wrote previously about how, in a race with three candidates, the people pushing the top two candidates — who swept every race in 2010-2012 (except maybe last year’s Mayor’s race) — would simply run three candidates rather than two. Instead, I’ll tell you a story about Barack Obama’s 2012 campaign.
At one point (if memory serves, and if not just go with the story), Obama’s strategy people were told a reporter that Romney’s people claimed that, based on the polls state races, if we had a national popular vote instead of the electoral college, Romney would be in an even better position to win. And the Obama people said: “yeah — of course if there were no electoral college we’d be taking most of our money out of Iowa and Colorado and Virginia and focusing on rolling up the vote totals in California and New York and Illinois.” The point being: talking about what would happen under a different election system based on the behavior of candidates until the current election system is … let’s go with “misguided” and “pointless.” It is surely not the sort of thing that a Judge will accept, not so long as the opposing attorney has the good sense to point it out. It just doesn’t matter.
So what does the Demographer — who, I remind you, is not a Political Scientist — do? He tests how well Latinos would do with a six person City Council based on who would have finished in third place under the current system. This is worthless. This is like saying that if I were a woman I’d bleed all over my thighs every month because a study of my hygiene habits show that I’ve never once used a tampon or sanitary pad in my life. Of course, IF I WERE A WOMAN I’D ACT DIFFERENTLY IN THAT RESPECT, get it? And if we had a 6 or 8 person City Council, candidates and their supporters would act differently too.
Slide 13, Had Top 4 Candidates Won, 2002-2012: This slide is even more worthless, except that looking at it may make that menstruation analogy recede more quickly into your memory. (Seriously, I’m just trying to keep you awake at this point.)
Slide 14, Where Anaheim’s Latino Voters are Concentrated: I actually like this slide. This maps housing tracts onto precincts to show where the Latino voting population is most concentrated. Good Demography! The darker the brown, the more Latino. Well, OK, I guess! He’s the demographer!
Slide 15, Where Leos Placed First in 2012: Oh, yeah — a demographer, but not a political scientist. A political scientist would know that operationalizing the extent to which a district is inclined to vote for a Latino based on how likely its voters were to prefer a Republican with a possibly-Greek-sounding name is not very fruitful. If he’s followed the campaign at all, he’d probably hypothesize that there are the areas with the highest membership in public services unions, especially OCEA, who would have been targeted to receive Leos’s literature and where he’d have sent more canvassers and put up a disproportionately larger number of signs. But let’s double-down and put up a slide of where Leos finished second, and then where he finished either first or second — because we’ve never before encountered the phrase “garbage in, garbage out.”
Slide 16, Where Leos Placed Second in 2012: So this is where the Greek-sounding-named Republican OCEA-supported candidate finished second…
Slide 17, Where Leos Placed 1st OR 2nd in 2012: … and this shows where Leos placed either first OR second in 2012. I wonder how that matches up with where Anaheim’s Latinos live? Maybe if we overlay slide 17 over slide 14 ….
Slide 18, Where Leos Placed 1st Overlapping SMD: … and, here’s what the Demographer does instead. He chooses a unit of analysis called an SMD — which can be but need not be used in drawing districts, and he shows that it’s not really a good fit — with Slide 15, which we should have already established is worthless compared to Slide 17, because we would not necessarily expect Latinos to vote for a Republican over a Democratic-party endorsed candidate … and because of so many Latino candidates diluting the vote … and … and … that’s it for me. The takeaway was, somehow, that Latinos would do worse with eight districts than with six. It’s possible, but I don’t believe for a moment that this presentation does much to prove it. I’d be happy to see voters choose between six and eight.
Slide 1, The Key Take-Away Points, Revisited: This is what Anaheim’s hired demographer, Dr. Peter Morrison, told the Council and would presumably like to tell the Judge.
(1) “Latinos will gain electoral strength as Anaheim’s population evolves.” Probably. It’s not clear how fast, though — and there’s no reason to dilute their support in the meantime. (Also, how is Anaheim Hills going to feel without districts if Anaheim eventually grows to 60-70% Latino and its Latinos will choose who represents Anaheim Hills on the Council? If this is true, they’d be smart to lock in reform now!)
(2) “Many adults aren’t now eligible to vote — but they will be!” I think that the line is supposed to be something like many people in Anaheim who can’t vote now, because of age and citizenship, will eventually be able to vote. This seems pretty much like the first point — and also doesn’t justify delay. Besides, if Anaheim is economically poorly off — say because the current City Council seems intent on looting it — then young people, disproportionately Latino, may end up leaving to look for jobs elsewhere.
(3) “Past Council elections suggest how future Latino candidates could get elected.” The best paths at this point seem to be: “be a firefighter running just after 9/11” and “be an Iberian-Filipina with an Irish surname.” Many, many Latinos — including most of those who have been candidates, can’t follow this advice. And that’s fine, because it’s actually pretty bad advice.
None of this would matter if the Council had simply rejected it. It would have been unfortunate, but not really damaging to their court case. Building a house of cards isn’t much of a problem unless someone tries to stand on it. The only way someone could really mess up would be to rest the bulk of the argument against voter-based district explicitly on this flawed report, so that its flaws would taint the vote of the City Council itself and justify the Judge’s tossing out their vote.
Guess what? In our next installment, it will be time to go back and talk some more about Councilmember Kris Murray, answer to the ACLU’s prayers!