Google Money

If you google “money,” the search engine called Google, from which that verbification is derived, will tell you that there are 1.12 billion results. That sounds like a lot of anything, but consider this fact: Google Inc. makes $1.12 billion in profit every 33 days.

Last year approximately $60 billion in revenue flowed through Google, one of Silicon Valley’s tech giants. To put this kind of dollar churn in perspective consider that the state of California’s 2013 budget was $96 billion. Google’s own budget, it’s expenses on sales, marketing, salaries, and R&D was $45 billion, about half that of the largest state in the nation. At this point Google might as well be printing money. Google is sized like a mature blue chip industrial corporation, but it’s still growing like a startup. From 2012 to 2013 Google’s revenue rose 20 percent. It’s net income for 2013 was about $13 billion. That’s cash for the company and its shareholders. Google has so much money it has become something of a problem.

Where to put it all? The G-Men have tucked $58.6 billion into different pockets for safe keeping. Google has $10 billion in cash deposits. But why earn less than inflation when you can invest? Google’s financial engineers have bought $1.5 billion in foreign government bonds. Google owns your city’s debt, and the bonds of your local school district; the company owns $3 billion in municipal bonds. Google owns $7.3 billion in the shares of other corporations, perhaps the company you work for? Google might even own a slice of your mortgage; it holds $7.3 billion in federal agency-backed mortgage securities.

Part of the reason Google has so much cash is because it has come to dominate Silicon Valley’s key industry: advertising. More than 90 percent of Google’s earnings come through selling ads, either directly through its own web sites, or third party web sites.

But advertising is only hyper-profitable because Google has figured out how to dodge the tax man. Google’s effective tax rate was already well below the statutory 35 percent federal corporate rate, but last year it dipped even further, falling from 19.4 percent in 2012 to 15.7 percent in 2013. It’s hard to tell what Google actually forked over to the feds and to the foreign governments where it operates, but Google reports that its tax burden keeps dropping “primarily as a result of proportionately more earnings realized in countries that have lower statutory tax rates,” according to the company’s 2013 annual report. Many of the countries Google is referring to are considered tax havens.

But here too it’s hard to tell which countries specifically are hosting Google’s strategic shell companies, and how much Google is earning overseas. As Jeffrey Gramlich and Janie Whiteaker-Poe, researchers at the University of Maine pointed out in a recent study, Google began omitting information about its overseas subsidiaries in its SEC filings three years ago.

“Google’s 2009 SEC Form 10-K, filed in February 2010, disclosed 117 subsidiaries, 81 of which were located in 38 foreign countries,” note Gramlich and Whiteaker-Poe. “When Google issued its financial statements for the year ended December 31, 2010, its Exhibit 21 listed only two significant subsidiaries.”

So where did 98 percent of Google’s subsidiaries go? Did Google terminate them?

Gramlich and Whiteaker-Poe found that most of them still exist and are registered with authorities overseas, but that they have been purposefully omitted from Google’s SEC filings, likely in an effort to conceal tax planning strategies used to reduce taxes paid to the U.S. government and foreign nations. Google’s most recent filing with the SEC reveals only 3 subsidiaries. Two are in Ireland, and one is in Delaware, both jurisdictions that afford Google secrecy and basement corporate income tax rates.

Iron Cagebook: The Rational Ends of Facebook’s Patents

“No one knows who will live in this cage in the future, or whether at the end of this tremendous development, entirely new prophets will arise, or there will be a great rebirth of old ideas and ideals, or, if neither, mechanized petrification, embellished with a sort of convulsive self-importance. For of the fast stage of this cultural development, it might well be truly said: ‘Specialists without spirit, sensualists without heart; this nullity imagines that it has attained a level of civilization never before achieved.'”
—Max Weber, 1905

On November 12 Facebook, Inc. filed its 178th patent application for a consumer profiling technique the company calls “inferring household income for users of a social networking system.”

“The amount of information gathered from users,” explain Facebook programmers Justin Voskuhl and Ramesh Vyaghrapuri in their patent application, “is staggering — information describing recent moves to a new city, graduations, births, engagements, marriages, and the like.” Facebook and other so-called tech companies have been warehousing all of this information since their respective inceptions. In Facebook’s case, its data vault includes information posted as early as 2004, when the site first went live. Now in a single month the amount of information forever recorded by Facebook —dinner plans, vacation destinations, emotional states, sexual activity, political views, etc.— far surpasses what was recorded during the company’s first several years of operation. And while no one outside of the company knows for certain, it is believed that Facebook has amassed one of the widest and deepest databases in history. Facebook has over 1,189,000,000 “monthly active users” around the world as of October 2013, providing considerable width of data. And Facebook has stored away trillions and trillions of missives and images, and logged other data about the lives of this billion plus statistical sample of humanity. Adjusting for bogus or duplicate accounts it all adds up to about 1/7th of humanity from which some kind of data has been recorded.

According to Facebook’s programmers like Voskuhl and Vyaghrapuri, of all the clever uses they have already applied this pile of data toward, Facebook has so far “lacked tools to synthesize this information about users for targeting advertisements based on their perceived income.” Now they have such a tool thanks to the retention and analysis of variable the company’s positivist specialists believe are correlated with income levels.

They’ll have many more tools within the next year to run similar predictions. Indeed, Facebook, Google, Yahoo, Twitter, and the hundreds of smaller tech lesser-known tech firms that now control the main portals of social, economic, and political life on the web (which is now to say everywhere as all economic and much social activity is made cyber) are only getting started. The Big Data analytics revolutions has barely begun, and these firms are just beginning to tinker with rational-instrumental methods of predicting and manipulating human behavior.

There are few, if any, government regulations restricting their imaginations at this point. Indeed, the U.S. President himself is a true believer in Big Data; the brain of Obama’s election team was a now famous “cave” filled with young Ivy League men (and a few women) sucking up electioneering information and crunching demographic and consumer data to target individual voters with appeals timed to maximize the probability of a vote for the new Big Blue, not IBM, but the Democratic Party’s candidate of “Hope” and “Change.” The halls of power are enraptured by the potential of rational-instrumental methods paired with unprecedented access to data that describes the social lives of hundreds of millions.

Facebook’s intellectual property portfolio reads like cliff notes summarizing the aspirations of all corporations in capitalist modernity; to optimize efficiency in order to maximize profits and reduce or externalize risk. Unlike most other corporations, and unlike previous phases in the development of rational bureaucracies, Facebook and its tech peers have accumulated never before seen quantities of information about individuals and groups. Recent breakthroughs in networked computing make analysis of these gigantic data sets fast and cheap. Facebook’s patent holdings are just a taste of what’s arriving here and now.

The way you type, the rate, common mistakes, intervals between certain characters, is all unique, like your fingerprint, and there are already cyber robots that can identify you as you peck away at keys. Facebook has even patented methods of individual identification with obviously cybernetic overtones, where the machine becomes an appendage of the person. U.S. Patents 8,306,256, 8,472,662, and 8,503,718, all filed within the last year, allow Facebook’s web robots to identify a user based on the unique pixelation and other characteristics of their smartphone’s camera. Identification of the subject is the first step toward building a useful data set to file among the billion or so other user logs. Then comes analysis, then prediction, then efforts to influence a parting of money.

Many Facebook patents pertain to advertising techniques that are designed and targeted, and continuously redesigned with ever-finer calibrations by robot programs, to be absorbed by the gazes of individuals as they scroll and swipe across their Facebook feeds, or on third party web sites.

Speaking of feeds, U.S. Patent 8,352,859, Facebook’s system for “Dynamically providing a feed of stories about a user of a social networking system” is used by the company to organize the constantly updated posts and activities inputted by a user’s “friends.” Of course embedded in this system are means of inserting advertisements. According to Facebook’s programmers, a user’s feeds are frequently injected with “a depiction of a product, a depiction of a logo, a display of a trademark, an inducement to buy a product, an inducement to buy a service, an inducement to invest, an offer for sale, a product description, trade promotion, a survey, a political message, an opinion, a public service announcement, news, a religious message, educational information, a coupon, entertainment, a file of data, an article, a book, a picture, travel information, and the like.” That’s a long list for sure, but what gets injected is more often than not whatever will boost revenues for Facebook.

The advantage here, according to Facebook, is that “rather than having to initiate calls or emails to learn news of another user, a user of a social networking website may passively receive alerts to new postings by other users.” The web robot knows best. Sit back and relax and let sociality wash over you, passively. This is merely one of Facebook’s many “systems for tailoring connections between various users” so that these connections ripple with ads uncannily resonant with desires and needs revealed in the quietly observed flow of e-mails, texts, images, and clicks captured forever in dark inaccessible servers of Facebook, Google and the like. These communications services are free in order to control the freedom of data that might otherwise crash about randomly, generating few opportunities for sales.

Where this fails Facebook ratchets up the probability of influencing the user to behave as a predictable consumer. “Targeted advertisements often fail to earn a user’s trust in the advertised product,” explain Facebook’s programmers in U.S. Patent 8,527,344, filed in September of this year. “For example, the user may be skeptical of the claims made by the advertisement. Thus, targeted advertisements may not be very effective in selling an advertised product.” Facebook’s computer programmers who now profess mastery over sociological forces add that even celebrity endorsements are viewed with skepticism by the savvy citizen of the modulated Internet. They’re probably right.

Facebook’s solution is to mobilize its users as trusted advertisers in their own right. “Unlike advertisements, most users seek and read content generated by their friends within the social networking system; thus,” concludes Facebook’s mathematicians of human inducement, “advertisements generated by a friend of the user are more likely to catch the attention of the user, increasing the effectiveness of the advertisement.” That Facebook’s current So-And-So-likes-BrandX ads are often so clumsy and ineffective does not negate the qualitative shift in this model of advertising and the possibilities of un-freedom it evokes.

Forget iPhones and applications, the tech industry’s core consumer product is now advertising. Their essential practice is mass surveillance conducted in real time through continuous and multiple sensors that pass, for most people, entirely unnoticed. The autonomy and unpredictability of the individual —in Facebook’s language the individual is the “user”— is their fundamental business problem. Reducing autonomy via surveillance and predictive algorithms that can placate existing desires, and even stimulate and mold new desires is the tech industry’s reason for being. Selling their capacious surveillance and consumer stimulus capabilities to the highest bidder is the ultimate end.

Sounds too dystopian? Perhaps, and this is by no means the world we live in, not yet. It is, however, a tendency rooted in the tech economy. The advent of mobile, hand-held, wirelessly networked computers, called “smartphones,” is still so new that the technology, and its services feel like a parallel universe, a new layer of existence added upon our existing social relationships, business activities, and political affiliations. In many ways it feels liberating and often playful. Our devices can map geographic routes, identify places and things, provide information about almost anything in real time, respond to our voices, and replace our wallets. Who hasn’t consulted “Dr. Google” to answer a pressing question? Everyone and everything is seemingly within reach and there is a kind of freedom to this utility.

Most of Facebook’s “users” have only been registered on the web site since 2010, and so the quintessential social network feels new and fun, and although perhaps fraught with some privacy concerns, it does not altogether fell like a threat to the autonomy of the individual. To say it is, is a cliche sci-fi nightmare narrative of tech-bureaucracy, and we all tell one another that the reality is more complex.

Privacy continues, however, too be too narrowly conceptualized as a liberal right against incursions of government, and while the tech companies have certainly been involved in a good deal of old-fashioned mass surveillance for the sake of our federal Big Brother, there’s another means of dissolving privacy that is more fundamental to the goals of the tech companies and more threatening to social creativity and political freedom.

Georgetown University law professor Julie Cohen notes that pervasive surveillance is inimical to the spaces of privacy that are required for liberal democracy, but she adds importantly, that the surveillance and advertising strategies of the tech industry goes further.

“A society that permits the unchecked ascendancy of surveillance infrastructures, which dampen and modulate behavioral variability, cannot hope to maintain a vibrant tradition of cultural and technical innovation,” writes Cohen in a forthcoming Harvard Law Review article.

“Modulation” is Cohen’s term for the tech industry’s practice of using algorithms and other logical machine operations to mine an individual’s data so as to continuously personalize information streams. Facebook’s patents are largely techniques of modulation, as are Google’s and the rest of the industry leaders. Facebook conducts meticulous surveillance on users, collects their data, tracks their movements on the web, and feeds the individual specific content that is determined to best resonate with their desires, behaviors, and predicted future movements. The point is to perfect the form and fuction of the rational-instrumental bureaucracy as defined by Max Weber: to constantly ratchet up efficiency, calculability, predictability, and control. If they succeed in their own terms, the tech companies stand to create a feedback loop made perfectly to fit each an every one of us, an increasingly closed systems of personal development in which the great algorithms in the cloud endlessly tailor the psychological and social inputs of humans who lose the gift of randomness and irrationality.

“It is modulation, not privacy, that poses the greater threat to innovative practice,” explains Cohen. “Regimes of pervasively distributed surveillance and modulation seek to mold individual preferences and behavior in ways that reduce the serendipity and the freedom to tinker on which innovation thrives.” Cohen has pointed out the obvious irony here, not that it’s easy to miss; the tech industry is uncritically labeled America’s hothouse of innovation, but it may in fact be killing innovation by disenchanting the world and locking inspiration in an cage.

If there were limits to the reach of the tech industry’s surveillance and stimuli strategies it would indeed be less worrisome. Only parts of our lives would be subject to this modulation, and it could therefore benefit us. But the industry aspires to totalitarian visions in which universal data sets are constantly mobilized to transform an individual’s interface with society, family, the economy, and other institutions. The tech industry’s luminaries are clear in their desire to observe and log everything, and use every “data point” to establish optimum efficiency in life as the pursuit of consumer happiness. Consumer happiness is, in turn, a step toward the rational pursuit of maximum corporate profit. We are told that the “Internet of things” is arriving, that soon every object will have embedded within it a computer that is networked to the sublime cloud, and that the physical environment will be made “smart” through the same strategy of modulation so that we might be made free not just in cyberspace, but also in the meatspace.

Whereas the Internet of the late 1990s matured as an archipelago of innumerable disjointed and disconnected web sites and databases, today’s Internet is gripped by a handful of giant companies that observe much of the traffic and communications, and which deliver much of the information from an Android phone or laptop computer, to distant servers, and back. The future Internet being built by the tech giants —putting aside the Internet of things for the moment— is already well into its beta testing phase. It’s a seamlessly integrated quilt of web sites and apps that all absorb “user” data, everything from clicks and keywords to biometric voice identification and geolocation.

United States Patent 8,572,174, another of Facebook’s recent inventions, allows the company to personalize a web page outside of Facebook’s own system with content from Facebook’s databases. Facebook is selling what the company calls its “rich set of social information” to third party web sites in order to “provide personalized content for their users based on social information about those users that is maintained by, or otherwise accessible to, the social networking system.” Facebook’s users generated this rich social information, worth many billions of dollars as recent quarterly earnings of the company attest.

In this way the entire Internet becomes Facebook. The totalitarian ambition here is obvious, and it can be read in the securities filings, patent applications, and other non-sanitized business documents crafted by the tech industry for the financial analysts who supply the capital for further so-called innovation. Everywhere you go on the web, with your phone or tablet, you’re a “user,” and your social network data will be mined every second by every application, site, and service to “enhance your experience,” as Facebook and others say. The tech industry’s leaders aim to expand this into the physical world, creating modulated advertising and environmental experiences as cameras and sensors track our movements.

Facebook and the rest of the tech industry fear autonomy and unpredictability. The ultimate expression of these irrational variables that cannot be mined with algorithmic methods is absence from the networks of surveillance in which data is collected.

One of Facebook’s preventative measures is United States Patent 8,560,962, “promoting participation of low-activity users in social networking system.” This novel invention devised by programmers in Facebook’s Palo Alto and San Francisco offices involves a “process of inducing interactions,” that are meant to maximize the amount of “user-generated content” on Facebook by getting lapsed users to return, and stimulating all users to produce more and more data. User generated content is, after all, worth billions. Think twice before you hit “like” next time, or tap that conspicuously placed “share” button; a machine likely put that content and interaction before your eyes after a logical operation determined it to have the highest probability of tempting you to add to the data stream, thereby increasing corporate revenues.

Facebook’s patents on techniques of modulating “user” behavior are few compared to the real giants of the tech industry’s surveillance and influence agenda. Amazon, Microsoft, and of course Google hold some of the most fundamental patents using personal data to attempt to shape an individual’s behavior into predictable consumptive patterns. Smaller specialized firms like Choicepoint and Gist Communications have filed dozens more applications for modulation techniques. The rate of this so-called innovation is rapidly telescoping.

Perhaps we do know who will live in the iron cage. It might very well be a cage made of our own user generated content, paradoxically ushering in a new era of possibilities in shopping convenience and the delivery of satisfactory experiences even while it eradicates many degrees of chance, and pain, and struggle (the motive forces of human progress) in a robot-powered quest to have us construct identities and relationships that yield to prediction and computer-generated suggestion. Defense of individual privacy and autonomy today is rightly motivated by the reach of an Orwellian security state (the NSA, FBI, CIA). This surveillance changes our behavior by chilling us, by telling us we are always being watched by authority. Authority thereby represses in us whatever might happen to be defined as “crime,” or any anti-social behavior at the moment. But what about the surveillance that does not seek to repress us, the watching computer eyes and ears that instead hope to stimulate a particular set of monetized behaviors in us with the intimate knowledge gained from our every online utterance, even our facial expressions and finger movements?

Tech Boom = Declining Real Incomes for Most San Mateo Households

As the great California journalist Upton Sinclair once said, “it is difficult to get a man to understand something, when his salary depends upon his not understanding it.”

In my effort to show the Wall Street Journal’s tech reporter that real household incomes have declined in San Mateo County I dug up and sent a time series comparing 1999 to 2012. As the table below clearly shows, the median household income for Black families in San Mateo County declined by a staggering 27% over the last 13 years. For whites the decline was more modest, about 6% at the median.

What this means is that even the most privileged racial group which has the greatest educational and cultural capital is experiencing downward mobility in the epicenter of the tech boom. Blacks, Latinos, and other more marginalized groups have it a lot worse.

And of course these numbers don’t capture the families who moved out of San Mateo County because it has simply become too expensive, so they might even be conservative measures biased by displacement.

I showed the WSJ’s Manjoo the above table. He responded via Twitter that he thinks it merely shows the effects of two recessions (2001 and 2007-2010). So the tech boom is unrelated?

He then stated that what matters, in his opinion, are the income gains experienced since 2010. In 2010 the national economic recovery began, and it coincides with the heating up of the latest the tech boom in Silicon Valley and San Francisco.

So what happens if we look at household incomes in San Mateo County since 2010?

We find yet again that most households are losing out, in spite of, or perhaps because of the tech boom.

Tech Reaganomics: On Those “Rising Paychecks” of Workers in San Mateo County

A technology writer for the Wall Street Journal Farhad Manjoo has a defense of California’s tech industry in the current issue San Francisco Magazine. Manjoo’s core claim is that while northern California’s tech boom might be a source of problems like rising rental prices, and what he euphemistically calls a loss of “cultural diversity” (read: Black and Latino displacement), it’s still good for everyone in the Bay Area. It’s a trickle down economics argument, basically. To support his tech and wealth-friendly perspective Manjoo offers us what he says is a key economic stat, the “rising paychecks of workers in San Mateo County.”

Since Manjoo chose the words “paycheck” and “workers,” you’d think we can safely assume he’s trying to tell us something about the real incomes of the majority of the labor force in San Mateo County, the Bay Area’s tech epicenter. He’s not.

Perhaps because Manjoo is trying to portray the tech industry as a great economic engine creating jobs and wealth that trickle down to everyone he chose the Bureau of Labor Statistics’ County Employment and Wages Summary as the source for his “rising paychecks” stat. Here’s what he reported:

“At the end of 2011, according to the Bureau of Labor Statistics, people in the county just south of San Francisco earned about $81,000 a year on average. That’s a respectable figure—despite being a small, mainly suburban area, San Mateo had workers who were among the best paid in the nation. Then something extraordinary happened: Over the course of a single year, the county’s average pay shot up 107 percent. In the last quarter of 2012, San Mateo wage earners averaged about $168,480 a year. That made San Mateo by far the top-earning county in the nation[…]”

This is misleading and it undermines anything further Manjoo might have to say about economic inequality in Silicon Valley. Not that he tries to say anything substantively about it anyway. His article shrugs off growing inequality while offering up anecdotes and de-contextualized factoids he says show how tech wealth is being spread around.

To be fair, Manjoo does point out that the dramatic spike in wages in 2012 was due to the Facebook IPO which minted more than a few millionaires. That event skewed the county’s average upward.

But while the BLS wages statistic helps Manjoo talk about the princely incomes of the tech elite, it leaves his argument devoid of any accurate information about how income is actually distributed in San Mateo and the wider Bay Area. He ponders the lives of nomadic twenty-somethings who live out of their cars and spend their days trying to build startup companies in hackerspaces. He doesn’t spend any time thinking about half of the region’s workforce employed in low wage service sector jobs, amassing debt, afflicted by housing insecurity, with little promise of advancing. Without finding a stat to actually measure workers incomes he can say nothing of substance about the overall equity of the tech boom.

Given this fact you’d think Manjoo would ditch his unreliable “rising paycheck” statistic for something that actually measures the real earnings of the majority of workers, not wages as they’re defined by the BLS.

He doesn’t.

Instead Manjoo let’s these astonishingly high paycheck estimates stand with that little caveat about Facebook’s public offering. He then proceeds to claim that this skewed wealth creation over three months helped San Mateo rake in more tax dollars to fund local services, thereby lifting all boats. Then he rattles off that the county’s unemployment rate is at a seemingly healthy 5 percent. Finally he makes the following amazing claim for the entire Bay Area: “Every other local economic indicator—including per-capita income and employment in sectors outside the tech industry, as well as the aforementioned rental and real estate prices—is at or approaching an all-time high.”

That sentence is also very misleading and says nothing about human welfare. It’s like saying, because asset prices are high, and because lots of people are technically employed (forget their actual earnings or well-being) then the society is fine. It’s not.

But let’s just focus on income, the metric Manjoo never actually measured.

When he didn’t actually measure it he couldn’t say anything substantive about the situation of most workers in San Mateo, or the Bay Area today. Instead chose to make a Reaganomics argument about how enrichment at the top of society trickles down and benefits those at the bottom. Of course this isn’t true nationally, and every economic and social statistic, from real incomes to health outcomes demonstrates the suffering that growing inequality causes. The trickle down ideology holds no truer for the Bay Area if you actually look at people’s real incomes compared to the cost of living.

If Manjoo wanted to actually measure the paycheck of the average worker in San Mateo he couldn’t have picked a more misleading source. The BLS data he used to claim that the average worker earned $81,000 in 2011 is calculated as the mean average of all wages paid to every employee in San Mateo County covered by unemployment insurance. But in the BLS’s survey, wages are defined not just as cash in the form of wages or salary. The BLS includes “non-wage cash payments [including] employer contributions to certain deferred compensation plans such as 401(k) plans and stock options.” That’s the 2012 Facebook distortion right there. Manjoo recognizes this in his article, but he misses the significance of this technical note.

What Manjoo fails to see is that 2012 was only an anomaly in absolute scale. Otherwise every year in San Mateo the income figure reported by the BLS survey is greatly inflated by the over-sized compensation packages (that include yearly infusions of stock options) of the thousands of corporate executives who live there. Not only was the 2012 estimate of $168,480 skewed upward, the 2011 average of $81,000 which Manjoo calls a “respectable figure,” and which he assumes is normal, is also already skewed upward.

$81,000 would be respectable if it were a remotely accurate estimate of what workers actually earn in San Mateo on average, but it’s not. To understand what workers actually earn today in San Mateo a better source of information is the U.S. Census Bureau’s American Community Survey, 1-Year Estimate for 2012. For half of San Mateo’s labor force, wages fell below $41,274 in 2012.

What Manjoo avoids recognizing in his ode to the tech sector is that Silicon Valley is an extremely unequal place, and that this inequality is ripping apart the social fabric. Almost half of San Mateo’s households have incomes under $75,000 a year. Half of the county’s households —which average in size at 2.8 persons— subsist on an annual income that is well below Manjoo’s fictional average paycheck for a single worker.

Income distribution in the tech epicenter of San Mateo County is highly unequal. Black and Latino households are over-represented at the bottom the county’s income range.

There may very well be 40,000 households (15 percent) in San Mateo County whose income is above $200,000 a year and who are living very well. These are the tech executives, the private equity investors, and the lawyers and banker who work in San Jose, San Francisco, and the suburbs in between. Just the same there’s another 40,000 households that earn less than $30,000 a year. Many of these Silicon Valley denizens at the bottom of the hierarchy are Black and Latino families who live in the shadows of the region’s wealth and glory. Their struggles to survive are rarely reported. Instead “tech journalists” these days run around the Bay Area telling cute stories about killer apps and occasionally lamenting how the hyper-gentrification of San Francisco is paradoxically destroying opportunities for artistic and cultural consumption for the privileged techies. The real big story, however, is the massive redistribution of wealth and power to the top 5 or 1 percent, the shrinking of the middle class, and the immiseration of the bottom half of society.

Per capita incomes in San Mateo by race show us that the tech boom hasn’t been good for Black and Latino workers. While white workers make $63,000 per capita, Blacks make only $30,000 and Latinos even less with $20,000. Let that sink in. The per capita income for white San Mateo residents is double that of Blacks and triple that of Latinos. It’s well known that the tech sector is a white and Asian space, that the average software programmer or engineer is a young white or Asian male, and that the upper-most executive posts in tech and along Sand Hill Road (the finance capital for Silicon Valley) are filled with white men.

We also know that the tech boom is unequal in terms of gender. To jump right to the point, inequality in Silicon Valley follows a pretty typical pattern of racial and gender hierarchy in which white men rake in the biggest rewards by far, followed by white women and Asian men. At the very bottom of the income earnings distribution are Latinas and Black men. The median earnings for white men in San Mateo County are 200 percent higher than for Latinas. Half of all Black men in San Mateo earn less than $22,000 a year.

Of course only focusing on the socioeconomic statistics of San Mateo creates a skewed picture that itself doesn’t accurately reflect the changes being wrought on the Bay Area by the latest tech boom. The San Francisco Bay region is an integrated and inter-dependent economic unit of nine counties and dozens of cities. The tech sector is geographically concentrated in Santa Clara, San Mateo, and San Francisco Counties. These three counties also are home to the most affluent households in the Bay Area, so any income averages that don’t drill down to particular cities and neighborhoods will be biased upward and the growing poverty beneath the surface will be obscured. Any exclusion of Alameda and Contra Costa Counties —the East Bay where polluting industries are concentrated alongside hyper-segregated Black and immigrant communities— also creates a distorted picture of the Bay Area’s economic transformation.