Excerpted from David Judd and Zakiya Khabir:

“In the U.S., the stereotype of the “techie” is probably a young white or Asian man who earns a relatively high salary as a programmer in Silicon Valley or some large coastal city. But while this image has a basis in the exclusionary demographics of programming, it shouldn’t be taken as representative of workers in the tech industry as a whole.

This remains the case even if we define the industry in relatively narrow terms–as made up of information technology businesses, ranging from titans like Google, Microsoft and Apple to the myriad smaller software companies and Internet startups that share commonalities of interest, strategy, competition and dependence.

It might be defensible to exclude telecommunications and aerospace businesses that tend to sell hardware to very different markets, are more likely to be unionized and have “old economy” business models, although they employ thousands of engineers and other technically skilled workers.

But it would seem false to exclude companies like Amazon and Uber, which share software companies’ origins, culture, venture capital-backed funding model and focus on technical innovation and “scalability”–even though they sell traditional goods and services like books, groceries and cab rides. And it would be truly strange to exclude computer hardware businesses when these, in the form of companies like HP, Intel and Apple, not only have often grown or pivoted to compete in software, but also established Silicon Valley in the first place.

There are still a million people employed in computer and electronics manufacturing in the U.S., many of them performing classic assembly-line work. Production workers in Silicon Valley’s factories have historically skewed immigrant, female, and Latina or Southeast Asian. The 55 mainly Latina women who struck a Versatronex plant in 1992 to protest low wages and lack of benefits represent a side of Silicon Valley that is very real, despite being very unlike what we see on HBO’s series of the same name.

It’s true that the hardware industry in the U.S. has shrunk in recent years, although it’s unclear whether the trend will continue, and if it does, at what pace of decline. The semiconductor fabrication engaged in by Versatronex has largely moved overseas in the last two decades, cutting short attempts by workers to organize. But even if the offshoring of manufacturing continues, that will not mean the Internet has moved the U.S. economy beyond material goods.

Instead, Internet companies are increasingly integrated with a growing logistics industry that relies on work which is just as regimented.

Journalist Mac McClelland made this clear in a Mother Jones article describing her time in one of the warehouses that companies like Amazon and Apple rely on to store and ship their products:

– The place is immense. Cold, cavernous. Silent, despite thousands of people quietly doing their picking, or standing along the conveyors quietly packing or box-taping, nothing noisy but the occasional whir of a passing forklift. My scanner tells me in what exact section–there are nine merchandise sections, so sprawling that there’s a map attached to my ID badge–of vast shelving systems the item I’m supposed to find resides. It also tells me how many seconds it thinks I should take to get there.

These companies, like those which sell the services of drivers, cleaners or delivery people, can rely on the pressure of millions of unemployed people to keep workers stuck in jobs with low pay and unsafe, humiliating conditions.

But for good measure, they typically rely on “temporary” staffing agencies and “independent” contractors to maintain some distance from their worst-treated workers–which helps reduce the potential of either consumer backlash or joint organizing with those segments of their workforce that have more immediate leverage. In addition, these giant employers have placed most of their larger facilities in isolated areas with few potential allies or alternative employment opportunities for workers.

This doesn’t mean that logistics workers are powerless–far from it. In principle, in the era of same-day delivery and just-in-time production, the potential to disrupt chokepoints like warehouses offers real power to an organized workforce.

In practice, whether the leverage of those who work at these critical points in the supply chain correlates with the necessity of their labor or with the number of substitutes available on the market depends on whether these workers are negotiating collectively or as individuals. As labor historian Kim Moody points out, the longshore workers who today have some of the strongest and most militant unions were once among the most precarious of employees.

Still, with the exception of some strikes in Walmart’s supply chain, this potential is mostly unrealized, at least in the newest corporate giants. As a result, these workers are left with a miniscule slice of the tech industry’s revenues.

The era of “big data”–of companies built on collecting, centralizing and monetizing information about human activity in new ways–still requires humans to sift and act on that information at key points. We are talking about as many as 100,000 workers–often based overseas in low-wage economies like the Philippines, but also employed near corporate headquarters in the U.S. Facebook might have most of us working on our own time to produce its raw content, but that content still needs to be filtered and curated by paid labor into usable forms.

Even those tech companies which sell only software rely on a physical infrastructure of office buildings, fiber-optic cables and datacenters. They also have the same need for administrative work, receptionists and office managers, and HR and support staff as any similar type of organization.

Despite pay that tends to be modest at best, the people who fill these positions often have better conditions than workers performing sweated labor in a warehouse or small factory, and in some circumstances, they have been able to organize.

Silicon Valley janitors have been unionized for two decades–since a bitter 1992 campaign during which employers used workers’ immigration status to fire them in large numbers. Facebook’s bus drivers voted to unionize last fall. These drivers are now paid $18-20 an hour: above the minimum wage, but not above a living wage in a region where the cost of living for a family of four is $75,000 per year and up.

Many of the white-collar non-technical workers in the industry are isolated at smaller companies. They are disproportionately women, who are particularly likely to face gender discrimination working with better-paid engineers and managers who are predominantly male. It’s rare that a startup is willing to admit in a job ad that it’s seeking a “Girl Friday”–but it’s common for the industry to expect a woman who might be hired as user support to serve, in the words of Melissa Gira Grant, as “a kind of domestic worker, a nanny, housemaid and hostess, performing emotional labor that is at once essential and invisible.”

Google employs about as many men as women in non-technical roles, but more than four-fifths of its technical employees are male–and only 2 percent and 1 percent are Latina/o or Black, respectively. The numbers at other big tech companies that have released them, such as Apple and Amazon, are similar.

Economy-wide, according to the Bureau of Labor Statistics, Black and Latina/o employment among software developers and programmers rises to one in 20 each, still disproportionately low–while women remain at one in five. Women leave technical roles at more than twice the rate of men, while Blacks and Latinos are structurally excluded from entering these jobs in the first place.

One study of the tech industry found significant racial disparities around compensation, with white men paid best and Latinos making around $15,000 less, on average, than their non-Latino colleagues after controlling for education, occupation, age, geography, gender, citizenship status, marital status and children in the home.

A great deal of attention is being paid the industry’s lackluster numbers around gender and race. Less discussed is how sexual orientation, gender identity, family status and age can also count against tech workers in companies where the phase “culture fit” is a cover for hiring only those that fit into the “bro-grammer” stereotype.”