In February, Google announced that it was committing to training 100,000 Black women in digital skills. This announcement arrived as a PR Hail Mary amid the ever-growing industry and academic outcry over Google’s firing of prominent, brilliant, respected A.I. researcher Timnit Gebru and recruiter April Christina Curley, both Black women and both exceptional contributors at the company. (Google claims that Gebru resigned; she says she did not.) The backlash occurred during a year of widespread protest against the centuries-old violence of racism and racialized capitalism in the United States.
This is not the first time that a prominent tech organization has attempted to “train up” Black Americans. From 1968 to 1972, at least 18 programs to provide computing skills training to Black and brown Americans were established in the United States. They were located in East Coast and California cities, with one in St. Louis, Missouri. In some cases, they were provided by government or social welfare organizations; others were provided by professional organizations, including the Association for Computing Machinery. The programs often targeted young adults; one offered at Johns Hopkins University enrolled and graduated five Baltimore high school students. Another in the Albany, New York, area enrolled 31 Black students, of whom a third were women. Some of them taught keypunch or mainframe operation, which included batch processing, while others actually aimed to teach computer programming in COBOL or FORTRAN. This era was dominated by mainframe computers, when computer programs were typically communicated to a machine via a series of carefully pre-punched cards, and those programs were run one after another in sequences known as batches. Trainees in a few programs received hands-on experience running the IBM 1401, the (much-heralded) IBM 360 system, and the IBM 7030. Though their operations (and quality) varied by site, most programs were connected to the ACM; for example, members organized and volunteered in programs, and others reported on these training efforts at ACM conferences. At the national level, the organization launched the Committee on Computing and the Disadvantaged.
A unifying theme motivating the programs appeared to be the notion that providing some basic digital skills training for Black and brown Americans, and thereby preparing them for entry-level jobs in the tech industry, would be a way to give them an “advantage.” In the words of many of the program organizers, and the ACM itself, these Black and brown Americans were “disadvantaged.” One scholar who has studied these programs, Arvid Nelsen, explained that most of the programs were established by white volunteers, and they—like many others at the time and since—conflated communities of color with people experiencing poverty, and both of those with urban neighborhoods. Indeed, those white 1960s-era volunteers, who were computing professionals themselves, exhibited a strange mix of motivations. On one hand, they thought that digital skills would provide a path out of poverty and toward stability, in the myopic way that those who are privileged by race and gender (in this case, white men) imagine that others can easily re-create their privilege and success. On the other hand, they feared the instability, protests, and uprisings for Black lives that continued in the U.S. during those years, and they imagined that preparing Black and Latino Americans for entry-level jobs would quell those crises. They lacked an understanding of what comprehensive economic and social justice looked like, to say the least.
In many cases, there was a disconnect between what the majority-white organizers imagined their programs would accomplish and how the programs themselves were carried out. The organizers typically demonstrated little awareness of the structural obstacles imposed by centuries of forced enslavement, unpaid labor, Jim Crow violence, and then separate and emphatically unequal education, compensation, health care, and housing. According to a report published in one 1969 ACM conference proceedings, the Philadelphia program admonished students that “any social problems that [they] might be experiencing cannot be an excuse for doing poorly. … The social aspects of a student’s life, regardless of how unpleasant, must be handled in an adult way and not be used as an excuse for under-accomplishment.” By contrast, the San Diego program arranged for Spanish-speaking instructors for Latino students, and even converted a 40-foot tractor-trailer truck into a mobile training facility so that students—who were spread across the sprawling city—would not have to spend upward of an hour commuting by bus to a central training location. Overall, however, the programs lacked both tech industry buy-in and continued funding, which is why many ran for only a few years. Those participants who completed the programs often could not secure jobs, even entry-level ones. In some cases, even outstanding program graduates were paid only $420 per month (about $3,100 today), which wasn’t nearly enough to support themselves and their families.
Furthermore, offering entry-level digital skills training to Black and Latino Americans did nothing to challenge the hierarchy of tech work already being established, in which white men were displacing Black and white women in computing jobs as the field become more socially, politically, and economically valuable (a process I describe in A People’s History of Computing in the United States, and one addressed by Mar Hicks in their book Programmed Inequality and by Nathan Ensmenger in his book The Computer Boys Take Over). In her book Recoding Gender, Janet Abbate describes a 1968 General Electric program designed to place Black Americans in the Albany-Schenectady, New York, area into programming jobs. The training was open to those without a high school diploma and to those with police records, and there was no admissions testing. Well over half the students completed the training. Yet as Abbate cogently observes: “Such experiments had little effect on the wider computing culture. Accepting black high-school dropouts as potentially talented programmers might have threatened the privileged status of technical skill. Instead, employers clung to the college degree as a promise of merit—although whether that meant specific knowledge, general intelligence, or social standing was often unclear” (emphasis mine).
In short, just because Black men and women receive digital skills training—whether in the 1960s or the 2020s—it provides no guarantee of employment, compensation, health insurance, job stability, job mobility, and overall power in the labor market. Perhaps some of the women trained by Google will get jobs. Some of their lives may improve. But it’s not a meaningful challenge to centuries of institutionalized segregation, separation, and subordination.
This criticism of such programs isn’t new, either. In a 1969 essay in ACM conference proceedings, Walter A. DeLegall, a Black computing professional at Columbia University, lamented with an ironic tone, “Too often, in well-meaning haste to set about the business of lifting up the unemployed and underemployed minorities through the magic of data processing training, certain serious problems are either overlooked or given only superficial attention.” DeLegall underscored the problems in these 1960s digital skills trainings, and the ways in which they paralleled the shortcomings of public education for Black and Latino students: “ineffective teachers, unsuitable textbooks, [and] inadequate facilities and equipment.” Throughout his essay, DeLegall pointed out the overarching structural and cultural policies that foreclosed opportunities for young Black and brown Americans; DeLegall contradicted the dominant white narrative that such individuals were deficient.
Google’s program perpetuates the fallacy—what I call the pipeline fallacy—that if we just get more Black women into STEM and get them to stick around (again, as if the individuals are the problem), we’ll magically have parity and equity in these areas. But we know that’s not true. Time and again, American culture values labor not by the labor performed, but by who is doing the work.
And who does the work can dictate how much the work is valued, rather than the other way around. For example, in 2017, scholar Miriam Posner analyzed the segregation of front-end versus back-end developers (history, telling once again). These are not hard and fast lines, but we can generally say that front-end coders work on what users see and experiences in their apps and online, whereas back-end coders “do the programming that works behind the scenes.” Posner observes: “Front-end dev work isn’t real engineering, the story goes. Real programmers work on the back-end, with ‘serious’ programming languages. Women are often typecast as front-end developers, specializing in the somehow more feminine work of design, user experience, and front-end coding.” And not surprisingly, front-end jobs pay about $30,000 per year less than back-end ones.
What would truly anti-racist and feminist work look like if Google executives took the (here I’m paraphrasing Abbate) perspective of Black Americans seeking to fix the system rather than whites seeking to “fix” Black Americans? Rather than just offering a flashy public relations ploy, Google could, for example, invest some of its millions into revitalizing Cradle to Kindergarten child care and education, creating a more promising pipeline for all Americans from birth. Likewise, Google could systematically overhaul its own cultural and hiring practices and set meaningful goals for recruitment, retention, and compensation to reflect the diversity of America’s population. Google’s digital training for Black women is emphatically not an effort to fix the system but rather an effort to “fix” Black women. And in the process, it will continue to reproduce the harmful status quo.