Governments worldwide are increasingly using biometrics to identify their own citizens. To Stanford Institute for Human-Centered AI Fellow Michelle Spektor, this raises interesting questions about how biometric identification (fingerprints and/or facial recognition) impacts the relationship between a state and its citizens, as well as how different states might reach different conclusions about designing and deploying such technologies.
To explore these issues, Spektor’s research follows the history of biometrics in two countries: the United Kingdom, which rejected a national fingerprint ID system in 2010; and Israel, which in 2017 adopted a national ID system that uses both fingerprints and facial recognition. Her research is still in progress but suggests that the increased use of biometrics to identify citizens is not inevitable: States don’t just create biometric infrastructure because it’s the most accurate or the most efficient option. Decisions to use biometrics are driven by social and political concerns. Also, social movements in opposition to these systems matter — especially when they appeal to broadly held societal concerns that go beyond civil liberties.
What inspired you to study the history of national governments’ use of biometrics to identify their citizens?
I have a small collection of family IDs that belonged to my dad, grandparents, and other family members. They immigrated to the U.S. as refugees from the former Soviet Union in the 1970s, and during their early years in the U.S., their documents listed them as “stateless,” which is a pretty jarring thing to see. So I believe there’s something important to be said about the ways states identify their people, and it can have significant social consequences. The details on an identification card don’t necessarily reflect who we are as individuals, but they do say something about what information the state issuing the ID considers important, which itself can tell us something about social and political notions of national belonging.
For over a hundred years, states have been using biometrics — in particular, fingerprints — to verify individual identity. For the most part, they’ve used them for monitoring marginalized communities, criminals, or colonial subjects. But over the last two decades, many states have taken up biometrics for purposes of citizen identification, and incorporated them into passports, national ID cards, or more rarely, national biometric databases. That makes this a really interesting time to think about why that might be happening.
And so the questions that motivate my research are: Why are some states doing this while others are not? What are the societal and political effects of using biometrics to identify a state’s citizens? And how do social and political concerns shape the design of technological biometric systems and decisions to implement them or not?
How can using a historical lens help us understand what is driving the current trend toward expanded use of advanced biometric technology for identifying citizens?
Taking a historical perspective helps foreground the social and political drivers behind states’ decisions to use or not use these technologies to identify their own citizens. There’s a tendency to attribute this development to innovation or to the belief that biometrics are the most accurate way to verify identity. But taking a historical perspective shows us that biometrics were just as cutting edge one hundred years ago as they are today. So biometric technologies being new doesn’t fully explain why governments choose to use them or not.
One common thread through the history of biometrics is the belief that biometric measurements say something about individual identity. Today, the implementation of a biometric system is often guided by the belief that biometrics are simply accurate tools for verifying identity. But if you look at 19th-century eugenics research, where biometrics originated, researchers measured the body with the aim of interpreting and classifying identity. And this foregrounds something that gets obscured by high-tech systems today: What’s actually being dealt with here are measurements of the body. So, for example, there has recently been a lot of important research on how biometric data contain biases and assumptions about the body in relation to race, gender, age, and disability. And by taking a historical perspective of biometrics, we see that eugenicists of the 19th and 20th century originally designed biometrics to do exactly that.
You are specifically studying the history of biometrics in the UK and Israel. Why these two countries?
In the last 20 years, the governments in the UK and Israel considered similar proposals for national biometric identification systems. The two governments’ stated aims for the systems were also similar: security, preventing identity theft and fraud, government efficiency, and technological progress, among other reasons. And there were also robust opposition movements in both countries that pursued similar strategies and arguments that relied on a mix of academic experts challenging government claims about how well biometric technologies worked and whether they would achieve their stated goals, and grassroots activists focusing on concerns about civil liberties and privacy.
But then in the end, the UK system was rejected in 2010, while the Israeli system became fully implemented and mandatory in 2017.
These two countries also have a shared history of biometrics. Many biometric techniques were developed by British anthropologists and statisticians in the mid-to-late 19th century as part of eugenics research and were also adopted by police departments. Colonial police departments also used these techniques across the British Empire, including in Palestine in the 1930s. And that marked the beginning of fingerprinting in Israel, because the Israel Police inherited that fingerprint system when Israel was established in 1948. Today, several countries that exist in former British colonial territories, including Israel, have national biometric systems for their own citizens, but the UK itself doesn’t.
Why do you think Israel adopted its proposed national biometric identification system while the UK did not?
There are a lot of reasons why some systems are implemented and others aren’t. They can relate to bureaucracy, private sector contracts, social movements, decisions about non-biometric alternatives, or even elections. But I’m most interested in how notions of national belonging, and politics of inclusion and exclusion, are built into biometric systems. My hunch is that this influences their design, how they’re perceived, and whether or not they’re implemented. To answer these questions, I use archives and oral history interviews to study past British and Israeli biometric systems. This allows me to see the political and cultural meanings different people have attached to biometrics over time, and how these influenced contemporary systems.
I think the histories of several prior systems in these two countries can shed light on this. The UK already rejected a national biometric system once before, when a National Anthropometric Survey was proposed in 1904. This system was meant to measure population health, but was also steeped in eugenics research, and the concerns that led to its rejection may lend insights into contemporary attitudes toward biometrics in the UK.
I also study the British colonial police service’s introduction of a fingerprint registry in Palestine in the 1930s, and how the Israel Police acquired the system in 1948. Since then, biometrics in Israel have taken on new uses and meanings. In 1974, the Israeli military began collecting the fingerprints of all soldiers enlisted through mandatory conscription, which serves a mix of state, military, and religious concerns. This database is managed by the Military Rabbinate, which uses it to identify soldiers who die during service. Biometrics in Israel are also connected with Israel’s security-technology complex, and technologies of surveillance and control. For example, in 1999 Israel began developing a biometric system to identify Palestinians from the Occupied Palestinian Territories who have Israeli work permits. Today this system is used at West Bank checkpoints.
In what ways can social concerns and opposition movements influence the use of biometrics?
I found that arguments about privacy or civil liberties didn’t gain a lot of traction in the debates over national biometric systems in the UK in 2004 and in Israel in 2008. Opponents of the systems had to adapt to political reality and think of ways to express concerns about privacy and civil liberties that would connect with issues that actually concern people. In Israel, for example, activists ran a campaign that focused on concerns that the government wouldn’t be able to adequately protect the data from being leaked. This data security argument had some success in Israel — a place where people care a lot about security implications of policies and technological projects.
In the UK, the groups opposing a national biometric identification system addressed privacy and civil liberties, and emphasized that a “papers please” mentality was an approach taken by oppressive governments. But some of their most compelling arguments to the public and to Parliament concerned costs and the government’s ability to carry out such a project.
It’s perhaps a bit disheartening to find that concerns about privacy, equity, discrimination, and one’s rights to control one’s own information or data about one’s body are not necessarily the ones that succeed when you’re trying to convince politicians or the public to reject biometric surveillance. But I would still say that these British and Israeli movements were influential. British Parliament canceled the national system following elections in 2010, and the system ratified by Israel’s parliament in 2017 included modifications that partially addressed activists’ concerns.
Can you talk a little bit more about how biometric systems relate to the politics of inclusion and exclusion?
The UK’s proposed biometric system from 1904 illustrates this well. At that time, there was widespread concern that the health of the people was declining. A government committee wanted to conduct a national anthropometric survey, which would measure British citizens’ bodies to determine the occurrence of what they called “physical deterioration.” They also hoped that the findings would inform social reforms to improve the living and working conditions of a marginalized population — the urban working class. But anthropologists who were consulted as experts on designing the survey had a different agenda. They wanted to use a national anthropometric survey to pursue their eugenics research, develop racial classifications of people in the UK, and advocate for anti-immigration policies that would exclude Jewish immigrants in particular.
This system was not implemented, but it’s an interesting one to consider because it involved two different groups hoping to use measurements of the human body to pursue really different projects of inclusion and exclusion. And some people might be quick to dismiss this example as irrelevant, but I suggest that it holds an important lesson for all biometric systems today. A single biometric system can have a variety of purposes and meanings at different times and for different people. For any biometric system, there’s always the question of who is to be included in it or excluded from it. And there’s also the question of what being included and excluded means — whether it’s to delineate citizens and non-citizens, or who is supposed to receive government services or certain benefits, or who is singled out for surveillance and control. This example, even though it’s a system from over 100 years ago that ultimately never happened, reminds us that biometric systems are just as much political projects as they are technological projects.
Ultimately, how do you hope your work will be useful?
I think opponents and proponents of biometric identification systems today could all learn from my work because it sheds light on the historical context behind these systems, the motivations behind them, and their effects on society.
These insights could inform efforts to make these technological systems more just and equitable. Maybe the answer is to not have them at all. But overall I hope understanding how political and social concerns shape the development of biometric systems will help us be better equipped to evaluate the significance and consequences of biometric systems today and in the future.
Stanford HAI's mission is to advance AI research, education, policy and practice to improve the human condition. Learn more.