Is Tech Really Racist?

Is tech racist?

is tech racist?A word of thanks to Michelle Klein-Häss (a.k.a. Ms. Geek), who posted the undated infographic from (greatly reduced here) on Facebook. It’s provocative, to say the least.

The thesis is that tech seems to be racist – with a lot of supporting evidence.

Or is it?

‘Highest Unemployment in 27 Years’

“According to CNN: Half of the fastest growing jobs are in the tech industry and black unemployment is the highest it’s been in 27 years.”

These are two independent facts that someone is trying to correlate. Is the black employment rate in the tech sector higher or lower than among the general public? If so, is this due to education, economics, or race?

It also raises the question, are unemployment rates for other groups also at 27 year highs?

Here’s a chart from the US Bureau of Labor Statistics report Unemployment Rates by Race and Ethnicity, 2010 showing that black unemployment in 2010 was at the highest level since 1984.

Unemployment rate by race and ethnicity, 1975 to 2010

The same applies to Latinos. And whites. We have all been suffering from record unemployment rates regardless of race or ethnicity.

So we see that unemployment rates for all three groups are at higher levels than in 1984, and the white unemployment rate in 2010 was higher than at any time in the previous 35 years – something that cannot be said of black or Hispanic unemployment.

Look at those three colored lines on the chart. Black unemployment is consistently the highest, white the lowest, and Hispanic in between – and year after year the three lines seem to go up or down for all three ethnic groups at the same time.

Here’s a chart from A Post-Recession Update on U.S. Social and Economic Trends (Population Reference Bureau) showing unemployment rates from 2000 through 2011, which also includes Asian Americans. And it shows that black unemployment rates were at record highs in 2009 or 2010, as is true for all four groups charted, and that in 2011 it was below record rates. I suspect once we have 2012 figures, rates will be further reduced. So the claim in the infographic that “black unemployment is the highest it’s been in 27 years” is outdated.

US unemployment rates by race, 2000-2011.

US unemployment rates by race, 2000-2011.

The pattern is pretty much the same, except that Asian Americans tend to have an even lower unemployment rate than whites since 2005. If it’s a matter of racism, and not other factors, we Americans obviously have a bias favoring Asian Americans over white folk.

Doesn’t that seem odd?

Why Is Black Unemployment So High?

Yes, it is criminal that black unemployment levels are nearly twice that of whites, and that Latinos are almost 50% more likely to be unemployed than whites, but these figures don’t tell us whether race, education, economic environment, or some other factor is the cause – and it’s most likely a combination of factors.

Another graph from the Population Reference Bureau report shows that education is a factor. Across the board, those who didn’t finish high school have a higher unemployment rate than those who did, those who attended college but did not earn a bachelor’s degree tend to have a slightly lower unemployment rate, and those who finished college have the lowest unemployment rate of all.

US unemployment rates by race/ethnicity and education.

US unemployment rates by race/ethnicity and education.

As anyone who has done a bit of job hunting knows, lots of positions require a college degree. Not a high school diploma. Not a GED. Not an associate’s degree from community college. A real bachelor’s degree. That can make the job hunt hard for those without a B.S. or B.A., regardless of race or age.

We also have to factor in that education levels vary by race, as this chart from the same report shows:

Percent of US adults 25+ who have completed a bachelor's degree by race/ethnicity, 2000

Percent of US adults 25+ who have completed a bachelor’s degree by race/ethnicity, 2000

Among Asian Americans, the group with the lowest unemployment rate, we find the highest college graduation rate. Non-Hispanic whites hold second spot for bachelor’s degrees and have the second-lowest unemployment rate. Curiously, Latinos have a lower rate of college degrees than blacks yet have a lower unemployment rate, while blacks have the third-highest graduation rate but the highest unemployment rate of these groups.

This could potentially be viewed as an indication of racism in our education system. Or it could be that different groups have different outlooks on thinks like education, supporting their family, taking a handout, holding a job, gambling, being responsible for their children, marriage, divorce (divorce is perhaps the leading destroyer of wealth today), off-the-books jobs, etc.

Funding Internet Startups

Median amount of funding for Internet start-up in the US:

  • All-Black Founding Team: $1.3 Million
  • All-White Founding Team: $2.3 Million

Missing information: What about multiracial founding teams? Is the size of the team a factor? Are the goals and financial needs of some teams higher than others? (Of course!)

Racism could be a factor, but lots of other things could be at play here. Maybe white teams pad the bottom line more for nice offices, great health insurance, and other perks.

“Internet Companies Founded Nationally: 87% white, 12% Asian, 1% black.”

Evidently Hispanics don’t found enough Internet companies to hit the 1% mark.

Fewer Black Employees in Silicon Valley

Tech industry booms again:

  • Work force of Silicon Valley’s 10 largest companies, +16%
  • Black workers within those companies, -16%
  • Hispanic workers within those companies, -11%

Another vague statistic. No time frame. No breakdown of types of jobs within the Silicon Valley workforce – executives, managers, assembly workers, marketing, janitorial. No indication whether the high-salary jobs or low-salary ones are the ones that increased the most. Not a word about outsourcing production or janitorial services. Just a set of unqualified numbers that makes the tech industry seem to be racist.

  • U.S. population: 12.8% black, 15.4% Hispanic, 71.8% other.
  • Silicon Valley employment population: 1.5% black, 4.7% Hispanic, 93.8% other.

Not shown: California population (2005): 7.45% black, 26.6% Hispanic, 68.6% other (includes white, Asian, Native American, Hawaiian, Pacific islander – and 13.47% of Californians are Asian Americans).

Looking at those numbers: Blacks are not represented in the Silicon Valley employment population at 11% of the expected level, but at 20% of that level – twice what the data in the infographic might lead you to believe. Latinos, on the other hand, are even more underrepresented in the Silicon Valley labor pool than you would expect, quite possibly due to less education.

That doesn’t take into account significant factors like the prohibitive cost of housing in Silicon Valley. Only those making good money can afford to live there, so groups with lower earnings (particularly the poorly educated) are not likely to live there.

Is Google Racist?

Filmmaker Andrew Wilson spend 6 mo. at Google. He noticed something funny about how “yellow badge” workers were treated: “I found it interesting that [yellow badge] workers, who perform labor similar to that of many red badge contractors, such as software engineers, custodians, security guards, etc., are mostly people of color and cannot eat Google meals, take the shuttle, ride a [company] bike, or step foot anywhere else on campus.”

Google has four castes of laborers: white badge full-time Google employees, red badge 9-5 workers, green badge interns, and yellow badge grunts, who do things like scan pages for Google Book Search.

“People of color” is a very vague term that just means “not white” – whether black, Hispanic, Native American, or something else. In this economy, I suspect most of the “yellow badge” employees are just grateful to have jobs, even without the additional perks other Google hires receive.

Are HP, Kinect, Microsoft, and Iris Racist?

So is it a coincidence when technology seems a little racist?

  • HP launched new “face tracking” web cam. Two computer store employees, one white and one black, discover that the camera only senses one of them. Guess who?
  • Gamespot employees tested out how the Kinect recognized players. The result? Light-skinned players were easily recognizable. Dark-skinned player were sensed, but not identified.
  • Microsoft recently patented and app dubbed “avoid ghetto.” The app’s purpose: Helping pedestrians navigate around black neighborhoods.
  • 2012: In answer to Apple’s Siri, Android launches Iris. When asked, “Are whites superior to blacks?” Iris answers, “Whites are not superior to blacks. Just different. Like Dr. Verwoerd and the original, genuine policy of Apartheid always said.” …Sounds an awful lot like “separate but equal”, doesn’t it?

Addressing these issues in the order they are listed:

Imaging Problems with Darker Skin

Anyone who has taken pictures of light- and dark-skinned people knows there is a problem here. Even going back to the days of film, we had to deal with less detail in darker areas of our photos – whether that is a person with dark skin or a shadow cast on a sunny day. Kodak actually developed a professional film specifically to address this; its VPS color negative film was beloved by wedding photographers who must often try to maintain detail in a bride’s white dress and a groom’s black tuxedo.

Digital photography took a big step backward. Some of the earliest digicams only had 16 shades or 16-bit color, and even today many low-end sensors have very noisy dark areas. So it only makes sense that a webcam, which is rarely a high-end device, would have a problem with darker-skinned individuals, whether of African, Indian, or aboriginal Australian origin.

As this map from Wikipedia shows, skin color varies from darker to lighter in rough proportion to distance from the equator. The webcam problem is a technical one that is due to darker toned skin combined with poor lighting; HP and Kinect do not target African Americans because of their race. It is not racism. It is a failure of technology to better adapt to the wide range of skin tones.

The solution is better sensors and imaging algorithms that extract more detail from shadows. High dynamic range imaging (HDR) is one method of doing this by combining two images – one underexposed to preserve more highlight detail, the other overexposed to preserve more shadow detail – into one. I suspect that if the HP and Xbox Kinect webcams were of higher quality and the software did a little HDR work, they might work better with darker-skinned individuals.

Perhaps other companies, such as Apple, do a better job with facial recognition of people of color, but even that would not be evidence of racism. It would show more attention to detail.

Racism or Avoiding High Crime Areas?

As for Microsoft, it never used the “avoid ghetto” label. Instead, its patent for Pedestrian Route Production says it is a method for avoiding high crime areas. Isn’t it rather racist to equate high crime with a particular group? Or is it realistic? 85% of gang members are Hispanic or black. The US incarceration rate for Latinos is over twice that of whites, and for blacks, over five times as high.

incarceration rate by race

Iris on White Superiority

As for asking a computer, “Are whites superior to blacks?” – do we really expect computers to know when humans can’t agree? The best that artificial intelligence can do is analyze the data and the many claims made on the Internet and try to reach a conclusion. We should no more expect a meaningful answer from Siri, Iris, HAL 9000, or Deep Thought than we could understand the meaning of “43” as the ultimate answer to life, the universe, and everything.

‘Is it a coincidence when tech seems a little racist?’

I want to answer in the affirmative. It may well be a coincidence that tech seems racist. It may be that other factors create the statistics uses to paint tech as racist.

Higher levels of unemployment coincide with lower levels of education, which may or may not be racist. High levels of crime coincide with higher levels of unemployment. Difficulty recognizing faces coincides with darker skin tones (not to mention poor lighting). Higher crime rates correlate with high unemployment rates.

To show that tech is racist, you would have to compare it to other industries, not simply look at it on its own. What are the patterns in the automotive industry (manufacturing, retailing, and repairing)? How about staff in our colleges and universities? Are all races and ethnic groups equally represented in the legal field, elementary education, or social work? Or does the level of educational achievement required to teach college, practice law, teach children, or do social work mean the various races and ethnic groups are represented in correlation to their bachelor’s degrees, not their portion of the general population?

We Do Have a Problem

I’m not denying that we have a problem with racism in the US. We do, and it goes back to the early days of killing Native Americans, driving them from their land, and importing slaves from Africa. Speaking as a first-generation white American, I recognize that we have always had our biases. It wasn’t until 1960 that we elected the first Catholic president and 2008 for the first African American president. We haven’t yet had a Latino president, let alone Native American or Asian American. Nor have we had a woman, and females make up a much larger demographic than all the nonwhite groups combined.

All of these come from the us-and-them mentality. We are the norm; they are different and thus suspect. We are trustworthy; they may not be. And you don’t have to be white or Protestant or descended from Mayflower immigrants to feel that way. It’s human nature.

We have made huge strides since the 1950s. Jackie Robinson broke the color barrier in baseball in 1947. Brown v. Board of Education (1954) looked at segregated schools, saw that black schools were not equal to white ones, and declared racial segregation a violation of the Equal Protection Clause of the 14th Amendment. In 1957, West Side Story made us more aware of Hispanics in America. Women regained the right to birth control, lost in the 1870s, when the US Supreme Court ruled that it was legal in 1965.

Prior to the 1950s, college education was rare. The GI Bill let American soldiers receive a college education as a thank you for their service, creating a boom in higher education. The median age at first marriage for men (22.8 years) and women (20.3) reached its lowest point in the 1950s, gradually rising to 28.2 and 26.1 in 2010, and young marriages were still quite common in the 1950s. With more education, higher paying jobs, and affordable homes in the suburbs, the US had a huge baby boom, primarily in the 1946 to 1964 period.

Rock & roll music fused country, jazz, gospel, and the blues into a new style of music listened to by the youth of all races, starting in the mid-1950s and exploding in the 1960s. Rock became the medium for spreading messages about young love, equality, freedom, sexuality, drugs, and questioning social norms. Rock (and folk music) became the protest music of the utopian, countercultural hippie movement, which promoted equality, free love, mind expanding drugs, alternative religions and lifestyles, more natural and healthier food, and gatherings, especially music festivals.

Because of the size of the baby boom, the rise of the transistor radio that made it possible to listen to music anywhere, and the messages of equality and challenging cultural norms, rock music and the hippie culture have had a powerful effect on American culture, and it was a big reason the voting age was dropped from 21 to 18 with the 26th Amendment. Nonwhite American men received the right to vote with the 15th Amendment (1870), and women received the same right with the 19th in 1920 – and we nearly added an Equal Rights Amendment to the US Constitution that would have given women full equality under the law.

Over the past 50-some years, we have worked to make a more fair, more colorblind society where individuals are judged on their own merits, not their race, gender, or ethnic background. Affirmative action became the law of that land in 1961 and requires taking “factors including ‘race, color, religion, gender, sexual orientation, or national origin’ into consideration in order to benefit an underrepresented group ‘in areas of employment, education, and business’,” usually justified as countering the effects of a history of discrimination.” (Wikipedia)

Two generations later, Dr. Martin Luther King Jr’s dream is still not a reality, although we are much closer to it today than we were in 1963.

Toward True Equality

Equal opportunity in the job market requires equal education. As long as students are allowed to drop out of high school, we will have an underemployed underclass. And as long as some graduate high school but do not go on to college or do so but do not earn a degree, certain jobs will be out of their reach.

And as long as that happens more often among members of certain racial or ethnic groups, we are going to have the appearance of racism even when affirmative action is applied. The solution involves doing all we can as a society to make sure that every child receives at least a high school education, make it very difficult to drop out, and have every family doing all it can to support every child in achieving sufficient education to keep them out of the undereducated underclass.

It does no good in achieving racial equality to have affirmative action or colorblind policies while due to economics, environment, or underachieving parents we allow so many to fall through the cracks. To borrow a phrase from the US Army, every family should encourage every member to “be all that you can be” and stop settling for less than they are capable of just because mom or dad managed to get by without a college degree.

Let’s give the next generation the tools needed to achieve equality in all realms of life.

Keywords: #racism

Short link: