In my last post, I shared my belief that what we saw this Tuesday was a long-term strategy of the US right-wing coming to fruition. This “southern strategy” is dependent upon the right’s ability to mobilize white southerners through race pandering and appeals to their reactionary Christianity. This time around the members of the right’s strategic alliance were considered safe bets for Bush and Co. Preliminary returns confirm that these segments of the population were all solidly behind Bush: the South (58%), white men (62%), the rural voters (57%), and Protestants who attend Church weekly (70%).
On the flip side, people of color have long been considered a reliable base of the Democratic Party. The question of whether Kerry took these voters for granted or not aside, non-white folks certainly voted for Kerry (67% of men, 75% of women). The Black electorate led the way in this regard, with a whopping 88% going to Kerry nationally, with even higher percentages in the South (90%). However, these generalizations run into major problems when we look at the Latino vote. In response to a post by John Lacny which asked what is to be done about all these stupid white folks, Dennis O’Neil commented that we must wait and see who’s next to be reclassified as “white”. Additionally, Dennis shared an insight offered to him by Elly Leary that she thinks “non-Indio Latinos seem to be being offered the next pass in whiteness.” Not only does her analysis hit the nail squarely on its head, I think that this trend has been a long time coming.
The past methodology used by the US Census Bureau (in which people self-identify) hints at a way this shift in the comparative whiteness of Latinos may (has been) happen(ing). A little bit of history on this one. The first mention of Latinos in the US can be found in the foreign born tables (where states of origin would have identified where these "immigrants" came from). However, these numbers only counted "white persons" from any given country. Starting in the 1950 Census, Spanish Origin/Surname was used in some of the more detailed tables and Special Reports but figures for "white persons from X" are still the only ones available. Starting in 1960 the "white persons" is dropped from the general tables, but the wording of its replacement, "persons of foreign stock reporting X as country of origin," does not exactly strike me as poetic.
As many know, even the most current 2000 Census did not include “Hispanic or Latino” in its list of racial distinctions; instead it is treated as an ethnic identity. So first they ask if you are “Hispanic or Latino”, then follow up by asking you to identify your race. In the 2000 Census 37,659,799 people identified as "Hispanic or Latino." Half of these respondents (nearly 19mil) then identified as "White alone or in combination with one or more other races." Don't let the "or in combination..." get you, because this same question is asked with the other 5 race categories substituted for white as well. The Census even discounts this wording, omitting it completely from several tables. Regionally this breakdown gets even more interesting. The Census breaks the country down into NE, MW, S, and W regions, and in those regions with mostly “red states” (S then MW) you find more people who identify as "white or some..." rather than "some other race..."
However, on the state by state level its evens out. Here’s breakdown combining exit poll data for many “red states” (from CNN.com) with figures from the 2000 Census:
| STATE | % Population that is Latino | % Latinos Who Voted For Bush | % Latinos Who Identify as White | % Latinos Who Identify as Other |
| Arizona | 25.25 % | 43 % | 46.26 % | 45.59 % |
| Colorado | 17.10 % | 30 % | 48.55 % | 41.38 % |
| Florida | 16.79 % | 56 % | 74.79 % | 16.70 % |
| Georgia | 5.32 % | 56 % | 45.64 % | 42.51 % |
| Nevada | 19.72 % | 39 % | 50.48 % | 39.74 % |
| Oklahoma | 5.20 % | 74 % | 40.19 % | 44.94 % |
| Texas | 31.99 % | 59 % | 57.97 % | 36.25 % |
| New Mexico | 42.08 % | 44 % | 52.36 % | 40.09 % |
| Nationally | 12.55 % | 44 % | 50.12 % | 37.35 % |
When I first started writing this post I expected to see a sharp jump in the percentage of “Hispanic or Latino” and “White or some combination…” While this hypothesis does hold in states where there are big discrepancies between those who identify as white and those who identify as “other” (see: Texas and Florida), more generally these figures show many Latinos, regardless of state, already see themselves as white, and historically self-categorization has been step one in the “whitening process”.
This becomes all the more alarming when current population projections are brought into frame. The Census is done every 10 years; but estimates are made during the interim period, with major work going into the midpoint projections. When looking at the current population estimates available on the Census.gov I couldn't help but notice, what was to me, a striking absence in the Race and Hispanic Origin tables. The ability to self identify as "Other" had been removed from all "Hispanic or Latino" tabulations. It is not as if the Census lacks these figures, they collected and presented data for the 2000 year; instead, these projections automatically identify Latinos and persons of Hispanic Origin as white unless some other racial category (Black, Asian, Pacific Islander, Native American) mitigates against their whiteness. This is nothing less than a complete break with previous methods of American coloring. Instead of people having to prove themselves white (the one drop rule), here we are seeing people being forced to prove that they are not white.
Sticking with the states above, it is clear how this shift plays out. In only 3 years Arizona's "Hispanic or Latino" population ballooned from 46.26% white to 94.78% white. This same pattern holds true for the other "red states" examined above: 93.35% in Colorado, 94% in Florida, 91.47% in Georgia, 93.93% in Nevada, 89.53% in Oklahoma, 96.87% in New Mexico and 96.87% in Texas. Even in California, the heart of the Chicano power movement, the percentage of "white" Latinos is now at 94.13%.
While these changes are currently imposed by the Census Bureau, I think these trends will hold even once people are able to self-identify, which will (hopefully) happen with the 2010 Census. As with polling, questions that force people to choose between options that they feel have pre-assigned value judgments attached to them (Are you going to vote) are fundamentally flawed questions. In a racist, one-generation post-apartheid society what is more value-laden than race? In the absence of the safe "Other" or wording that more positively reflects the heritage of many Latinos (mestizo, Afro-Caribbean, etc) it shouldn't come as a surprise that folks would be less than willing to give up this newly endowed whiteness, even if it is a whiteness not enjoyed in one's day to day life (yet).
My facts and figures come from American FactFinder, the Historical Census Browser, Population Information in the Twentieth Century (General Tables Index) and Demographic Trends in the Twentieth Century (pages 71-111, especially pages 78, 86-7, 96-8) both put out by the Census Bureau (Demographic Trends in the Twentieth Century actually replaced Population Information in the Twentieth Century).
