In the two decades since the terrorist attacks of Sept. 11, 2001, Americans increasingly have been under surveillance. While much of the focus of concern is on the government, data supplied by technology companies from the private sector provides the crucial materials for intelligence gathering. Texas’s recent antiabortion law that empowers private citizens to spy on their neighbors to find “illegal” abortions is just one reminder that governments can offload surveillance to the private sector and avoid constitutional scrutiny.
Over the two decades, advances in computing and algorithmic technologies as well as the growth of Internet use have plunged us into a surveillance age dominated by some of the largest, most profitable corporations in the world. Yet none of this is exactly new. Although largely understood as a post-9/11 phenomenon, business surveillance systems have been operating in the background of American life for nearly 200 years.
Before computers, the guardians of capitalism disciplined American workers and consumers through the use of ink pens and ledger books, filing cabinets and typewriters, statistical tables and customer lists, telephones and their own alert ears and eyes. Indeed, modern capitalism is inseparable from surveillance. And for this reason, business and government surveillance have always been intertwined.
As far back as the 19th century, two important industries — credit reporting and insurance — devised sophisticated, information-intensive ways to manage risks and boost profits in the modern marketplace. In the 1840s, newly formed commercial credit reporting agencies began to collect information about American retailers and business owners. Since creditworthiness was linked to morality, agencies dug into the private lives of citizens, looking for evidence of known or suspected vices.
Dun and Bradstreet, the leading credit reporting firms at the time (separate companies until a merger in 1933), published credit rating books for subscribers. Within these volumes, tens of thousands of individuals were classified as trustworthy or disreputable according to alphanumeric codes. Those with the best reputations were rated A1, while those with dubious reputations were rated lower.
The U.S. government’s foray into private-sector surveillance began during the Civil War, not the war on terror. In May 1861, federal agents descended on Northern telegraph offices and seized transcribed messages in bulk. Nineteenth-century telegraph companies — a private industry — maintained vast archives of personal communications, which the U.S. government was able to use in wartime. Analysis of the messages, it was hoped, would reveal the activities of Southern conspirators behind Union lines.
As buying on credit became widespread in the 20th century, retailers and financial institutions used these tried techniques to rate consumers. Consumer credit bureaus took root in cities across the nation. By the 1930s, hundreds monitored the lives of ordinary Americans. Some of the largest bureaus in New York and Boston kept files on more than 1 million individuals, which included information collected from stores, employers, landlords, court records and police logs. Similarly, insurance companies delved into the lives of their customers, compiling demographic data and hiring investigators to examine the health and morality of applicants. Together, credit bureaus and insurance companies tracked the lives of millions of Americans and sorted them into categories of risk and value.
Outsourcing surveillance to the private sector relieved governments of the cost and burden, and it shielded state agents from public scrutiny. In the 1930s, the New York City Police Department relied on bartenders and tavern owners to monitor and control sexual behavior. LGBTQ patrons, who did not conform to the sexual and gender norms of the time, were removed from public spaces. The enforcement mechanism was the liquor control board, which threatened to revoke the liquor license of any bar or nightclub that refused to comply.
During this time, government agents from the FBI and Internal Revenue Service regularly visited credit bureaus for assistance with criminal or national security investigations. Credit bureau operators were all too happy to comply, with or without court orders. In fact, during World War II, Dun & Bradstreet bragged in its annual reports about how effectively it vetted defense industry workers, rooting out the “enemy agent, the Axis sympathizer, the disgruntled individual or the man or woman of weak character.”
The flow of surveillance worked in both directions. After World War II, businesses benefited from municipal records as well as growing government bureaucracies and data-gathering systems. By the 1970s, credit agencies and marketing firms had devised much more detailed portraits of people by linking personal information to census tract data, Zip codes, standard industrial codes and Social Security numbers.
Businesses also served as extensions of the government’s own screening and compliance programs. For example, as credit reporting become more extensive and, presumably, more reliable, federal home loan programs began to mandate that mortgage applications include private-sector credit reports and, later, credit scores for borrowers. As with algorithms and artificial intelligence today, the result was hidden racial biases. Invariably, the credit reports and scores worked against borrowers of color, since the data itself reflected the legacies of racial discrimination in hiring and housing.
Growing concern about the files available on American citizens led to congressional investigations on privacy in the 1960s and 1970s. Ironically, the first of these was in response to a government proposal to link the databanks of a range of federal agencies. But the investigations soon revealed that the private databanks of credit reporting, insurance and financial institutions were far more extensive than previously known, completely unregulated and largely hidden from view.
Out of these early investigations came legislation such as the Fair Credit Reporting Act of 1970. This new law aimed to mitigate the sort of built-in biases that excluded people of color and women from credit access, while also restricting who could access credit files and for what purposes. Ultimately the law did neither very well. The problem was that the act focused on privacy, a concept that proved ineffective in limiting the torrent of data that flowed into corporate databases and fed algorithmic decision systems.
In the aftermath of the 9/11 attacks, U.S. intelligence officials were criticized for failing to “connect the dots,” missing or misinterpreting vital information that could have averted the tragedy. To prevent another catastrophe, they contended that the obvious solution was to collect more data.
Indeed, the fall of the twin towers in 2001 served as a justification for more expansive government surveillance, but it also provided cover for more intrusive commercial surveillance. By that time, private surveillance infrastructures had already been built and were growing exponentially, a product of capitalism’s long-running drive to master people and markets through surveillance.
Today, our relationships, interests, affiliations and movements are captured by a combination of search engines, websites, social media platforms, voice assistants, home cameras, payment systems and app-laden smartphones. While government agencies struggle to connect the dots, corporate behemoths like Google, Facebook, Amazon and Acxiom produce pointillist portraits of each of us.
Privacy is of little use against surveillance if it means protecting individual identity. Data-mining programs and algorithms now sort us according to statistical correlations that are completely mysterious to the public, or even to the authors of the programs. Efforts to assure privacy often end up focusing on getting the information “right” — making it more accurate, more complete, less biased, less discriminatory — which only makes it more useful for this sort of surveillance practice.
While U.S. government surveillance is bound by constitutional protections, commercial surveillance systems operate with very few constraints. This is not a new problem, but it is increasingly consequential in the age of ubiquitous capitalist surveillance.