Trying to Make Sense of Texas Civil Rights Accounting
EEO Report: Working Note One
Not sure how we ended up looking at Table One of the Feb. 2005 report from the Texas Workforce Commission Civil Rights Division, but we were puzzled from the start.
At first, it didn’t seem strange that among an alleged 32.89 percent Hispanic American workers in the total workforce, only 15.2 percent were to be found in Administrative positions. Having studied Texas history, we could easily make sense of that.
For these same reasons, neither did it seem strange that among an alleged 11.24 percent African American workers in the total workforce, only 7.1 percent were classified in Administration.
But it was surely curious to find that among an alleged 84.30 percent Caucasian American workers in the total workforce, only 77.8 percent were in Administration. That was a shock on three counts.
First of all, it meant that all categories of workers by race-ethnicity were under-represented in the Administrative classification, which is quite a statistical achievement to think about in Civil Rights history.
Second, it meant that when you added up the total percentage of workers in the workforce, you got 84.3 Caucasian plus 11.24 African American plus 32.89 Hispanic American equals a workforce of 128.43 percent!
Third, it meant that in a mandatory report to the state, the Texas Workforce Commission on Civil Rights couldn’t even get the first row of numbers right.
And fourth, as we look at the twirling graphics at the TWC website announcing the new report posted Mar. 15 — we wonder, did anybody notice?
It is not difficult to figure out what went wrong here. In the second number of the chart — percent Caucasian American in total workforce — somebody plugged in the percent you would get for White workers if you didn’t subtract out Whites of Hispanic origin. While this is the number reported for White by the bureau of labor statistics, it is a deceptive number to use for Civil Rights purposes under the heading of Caucasian American ethnicity.
Note: on second reading, analysis in the next few paragraphs looks strange to us. We’d rather say, after further consideration, that better numbers were available from EEO reports, so why weren’t EEO reports used?
If you derived percentages from the overlapping numbers of race-ethnicities actually reported in row one, you’d start with a total Caucasian workforce of 65.64 percent. And if you started this way, you would notice in the second row of Chart One that Caucasians who hold 77.8 percent of positions in Administration are over-represented by 12 percentage points.
Furthermore, if you derive your first row percentages from the first row numbers provided for race-ethnicity, you’d find that the total workforce is not 32.89 percent Hispanic American as reported but 25.6.
Misdirection on the percent of total Hispanic Americans in row one is instructive, because the percent reported there (32.89) matches pretty closely with the percentage that the US Census Bureau reports for the Hispanic population as a whole in Texas (32.0). Just reading along with the Texas Civil Rights Commission’s first chart, you’d think the Texas workforce had achieved parity in Hispanic employment.
Ditto with the misdirection on African Americans. The 11.24 percent workforce shown matches up nicely with the 2000 census number of 12.0 percent Black Texas, which sure looks better than the 8.75 percent you’d have to publish in row one if you worked with the actual numbers in that same row.
As for percentages Caucasian, the impression that Caucasians with 77.8 percent of Administrative jobs, share some kind of under-representation in Texas Administration is dispelled by knocking down the 83 percent total workforce figure to 65, as we have seen. But now look at the 2000 census percentage of Whites who are not Hispanic and the dramatic heft of white power weighs in with a 25 percentage point differential between total population (52 percent) and total Administration (77.8).
Note: the preceding attempt to make sense of civil rights categories with non-civil rights numbers, although yielding results closer to the truth of the civil rights situation, employs methodologies that we do not recommend. Better to get the civil rights numbers from a proper civil rights source, rather than try to make bad numbers work for purposes they were never intended to serve.
Enough with the misdirection, already. The whole stupid report should be tossed back to the alleged Commission on Civil Rights with an angry, loud, and resolute demand: give Texas citizens numbers that dignify the importance of Civil Rights in this state.
The fact is that BLS numbers do not reflect either the categories of Caucasian or African American reported in table one, and the numbers are meaningless for a civil rights report.
Note: this was our first whack at the Civil Rights Report and it has some zest to it. But the analysis has been superseded by another day’s work.