The difference between digital identity, identification, and ID

Caribou Digital’s style guide for talking about identity in a digital age

Thanks to my Caribou Digital colleagues, Savita Bailur, Nicki McGoh, Bryan Pon, and Emrys Shoemaker, for their inputs to this essay

Image by Nick Youngston Alpha Stock Images CC By-SA 3.0

How does a person prove who she is? How does she describe who she is? How does a state count and distinguish between the people under its care? How does a company tie together disparate information about a user to personalize a compelling new product while maintaining that user’s trust? And how are the answers to these questions changing as digital systems replace analog ones? Caribou Digital is privileged to be a part of several ongoing conversations and projects in this broad space.

Which leads us to the question of what we call this space. Are we working on “Digital identity”? “Digital Identification”? “Digital ID”? What terms do we use, day in, day out? Among the broader community, and even our own projects, different terms appear at different times.

For example, the World Bank promotes the term “Identification for Development” (ID4D) and has drawn together a diverse group of institutions to issue the “Principles on Identification.” The ID2020 alliance focuses on many of the same objectives through its “global partnership committed to improving lives through digital identity.” And USAID has recently written in this area under the banner of “Digital ID” (DID for short). The driving force behind many of these initiatives is Sustainable Development Goal 16.9: “By 2030, provide legal identity for all, including birth registration.”

This note is too short for us to propose a universal terminology. However, there is more at stake here than a simple “potato/potato, tomato/tomato” agreement to disagree. Instead, we believe that, as the communities of practice grow, intermingle, and intentionally modify “identity,” “identification,” and “ID” with the powerful word digital, the distinctions between these terms become both more salient and more important.

In this note we share the three reasons that have led us to draw distinctions between identity, identification, and ID, and share our own, still evolving, “style guide.” We hope others in the identity, identification, and ID communities will benefit from this exploration of our terminology, even if it doesn’t make sense to ask anybody to change their own usage of these terms.

Three reasons to distinguish identity, identification, and ID

One’s a relative social coordinate. One’s a process. One’s a thing. The distinctions aren’t always clear, and there’s some overlap between the terms. But,

  1. Identity often implies a kind of multidimensional social location of an individual relative to other people and institutions around him or her. Someone is uniquely, X, Y, and Z not only by being X, Y, and Z simultaneously, but also, in part, by being not A, B, or C.” Identity is an intangible, always contested something an individual creates, or perhaps has, as a result of their interactions with other human beings and systems.
  2. By contrast, identification often implies a process — it’s a better term to describe a proof, a system, or a transaction involving a subject and an evaluator, centered around verifying a claim that a person is one person and not any other. It also works well when referring to the recording of certain attributes — biodata, biometrics, claims — in a formal record, a “credential,” that grants specific rights or permissions to the individual. Identification is a concept we care about because it is that process which that grants access and rights; it is the representation of the individual within/to an administrative system.
  3. Different still is the idea of an ID. An ID often implies a tangible artifact — a document or element that supports a claim or signals that identification might be possible. Often, we use it as the manifestation of a credential, typically in physical form. But we don’t suggest that the ID means much without the identification systems behind it.

These aren’t absolute rules, but rather overarching impressions and shades of meaning. In making this distinction we draw guidance from Whitley, Gal, and Kjaergaard, 2014, as well as Gelb and Metz, 2018. Notably, Gelb and Metz’s book explores an “Identification Revolution,” not an “Identity Revolution.”

Identity is also part of a broad and distinct conceptual debate in the social sciences and humanities. Discussions of identity, today, are just as frequently about nationality, ethnicity, religion, gender, sexuality, etc. (and their intersections) as they are about formal documentation. Beneath this broader use of the word identity is a blurring of official and unofficial processes. It’s also a blurring of administrative and critical perspectives on a concept, which creates a tangle of still-contested sociological terms about things near to the core of what it means to be human (see Brubaker and Cooper, 2000). We are not saying that any one field has rights to this word identity . Rather we want to articulate that there are times when, to work in the international development sphere and promote systems that better facilitate institutional functions like legibility and service provisioning, one can avoid much of this contestation and confusion by using the narrower, more administrative idea of identification.

Finally, digitization complicates and expands almost every discussion of these terms: identity, identification, or ID. The global era of networked communication and computing has ushered in a new enthusiasm for the potential of identification systems to efficiently and comprehensively reach and serve the world’s poor and vulnerable. The achievement of SDG 16.9, a legal identity for all, will only be possible through the power of digital technologies. But, with digital, comes all the fluidity of information that can be moved, combined, and connected in ways we are only beginning to understand. Once several digital identification systems are connected, their interrelationships with the broader social ideas of identity become simultaneously stronger and less controllable. Digital traces, footprints from cell towers, or logs of transaction behaviors, can be used as inputs in identification systems or even “social credit scores,” with or without the consent or knowledge of the individuals in question. People’s own contributions to social networks, made in the service of creating and presenting elements of a social, cultural digital identity, can be used to segment, to divide, or to serve. The interplay between identity, identification, and ID is amplified in the digital era. That is, the lines between who we are, where we go, what we say, and what the state knows about us have never been more blurred. That blurring demands more clarity in our terms.

IDs in Jordan. Photo: © Emrys Schoemaker

A style guide for talking about identity

But if “legal identity” is good enough for the SDGs, isn’t it good enough for us? Perhaps not. In writing in this space, we have found that we need to use different terms at different times. When we’re careful, we will use these terms as follows:

  • We reserve Identity to refer to broad, dynamic, ongoing negotiations between individuals and societies. Indeed we avoid talking about “an identity” as a singular concept. As we detailed in our work in India, people can choose to present different parts of themselves at different times, to different people, and to different institutions. These processes are dynamic, relatively uncontrolled, and, especially in the digital age, a mishmash of signals, inputs, and artifacts. Identity is really, really complicated!
  • We use Identification,to refer to a transaction or process. Accordingly, identification demands at least two actors in that X is evaluating a claim by Y that Y is Y. This is no small feat — society works because hundreds of millions of identification processes occur each day, most without issue, with the support of the infrastructures we seek to improve and expand.
  • Finally, although a term like ID may roll off the tongue as a convenient shorthand, in light of identity (a negotiation between individuals and society) and identification (a process), the idea of ID is perhaps best reserved as a synonym for an artifact:a document, passport, badge, number, electronic token, that facilitates the process of identification.

Based on these core distinctions, some further useful terms can be derived

  • Identification systems are the interfaces between individuals and institutions, the systems issue credentials (IDs) and facilitate identification. The term systems is key in that it points to the fact that IDs (artifacts) don’t drop from the sky, and identification moments don’t just happen. They depend on assemblages of technology, laws, business models, bureaucracies, and conventions that wield power by validating some claims and not others, by counting some people and not others. Often the biggest challenge is that people fall between systems, as we found in our India research. Digital systems are providing new opportunities to use inputs and manage information in different ways. So digital identification systems are more dynamic, more flexible, and often less expensive to implement than analog identification systems.
  • At the risk of getting into the weeds, it’s worth noting that identification systems issue credentials — passports, employee badges, voter cards — based on attributes or claims submitted by the individual that are frequently accepted only in the form of some other credential. For example, when applying for a passport, the individual may have to present a birth certificate. This interdependency seems circular, but is actually accretive in nature: new identification credentials are built on top of previous credentials (commonly called “breeder documents”). Recently, digital identification systems have become more adept at transforming personally identifiable information, such as call logs, transaction records, biometrics, etc. as claims or attributes — sometimes without the consent or knowledge of the individual. But one’s gait or fingerprint is not an identity, it is a new kind of input, recognized by an identification system that facilitates identification.
  • To account for the interplay of multiple identification systems, we use the term national identification ecosystem to capture several systems at once. Echoing earlier ideas like Kim Cameron’s identity metasystem, an ecosystem lens helps us see connections between formal identification systems and other digital products operating in the background, between what is seen in terms of tangible, material artifacts (IDs), and what is unseen, in terms of traces left by our digital lives. We use identification to foreground the administrative systems and development processes of interest to our community, but it’s important to stress that the ecosystem includes many of identity elements, like social media, transaction behaviors, etc. Identification might be the most important outcome vis-à-vis the SDGs, but the ecosystem is broader than the concept of identification alone.

And finally, we can make an important, carefully considered exception to the way we use identity, when we modify it.

  • What’s a legal identity? Let’s return to SDG 16.9: “By 2030, provide legal identity for all, including birth registration.” The corresponding indicator, 16.9.1, is “proportion of children under five years of age whose births have been registered with the civil authority.” SDG 16.9 modifies identity with the critical adjective legal. This modification converts an amorphous, contested, constructed term like identity into an administrative binary. To modify identity with “legal” “formal,” or even “functional” is to invite categorization into those that do have that class of identity, and those that don’t. From a development perspective, each refers to a threshold. We may cover the distinctions between legal, formal, and functional identity in a later post. But for now, we acknowledge that much as the idea of a binary “digital divide” collapses as one explores it in more detail, the idea of a binary “identity divide” — between those who have, and those who don’t have, “an identity” — is both seductive and in alignment with how a lot of development practice works. Aspiring to ensure that everyone has a “legal identity” (i.e., is visible to and uniquely distinguishable by the state and other actors), is an important mission that will require the efforts of a broad array of actors.

Nevertheless, our careful use of identity, identification, and ID helps us remain cognizant of the fact that the challenge doesn’t end with the crossing of any binary, once people “have a legal/formal/official identity.” There are ongoing processes of negotiation and cultivation of identity that require maintenance. In the digital age, administrative identification and broader social, political, economic identity can never be detached. For instance, recently, a number of Rohingya, staying in camps in Cox’s Bazar in Bangladesh, went on strike demanding, among other things, that their smart card IDs (artifacts) issued by the UNHCR as part of an identification system use the word “Rohingya,” rather than “forcibly displaced Myanmar national.” In another part of the world, through our ongoing research in Brazil for UNICEF, we spoke to transgender youth frustrated by their inability to change the name or gender on their administrative birth certificate (and ID artifact) until age 18, which left them with an administrative artifact out of line with their evolving social identity. Moments such as these encapsulate the interplay of several of the concepts we discussed in this note; importantly, such moments occur in ecosystems where identities, identification, and ID artifacts are interconnected.

One of our concerns is that an unwitting or unreflective detachment or conflation of identity, identification, and ID may lead to the design of systems with sub-optimal outcomes. Technologies don’t really create or bestow identity, instead socio-technical systems facilitate identification. Indeed, a focus on good outcomes for individuals means that we must take the socio-cultural implications of administrative identity into account. James C. Scott makes this point in Seeing Like a State: the administrative gaze constructs elements of identities that individuals have to negotiate and manage, with all the complexities that rigid administrative identification processes may bring.

To conclude, we can’t ask everyone to share our language, but if you hear us switching from one term to another, even in the same meeting, now you’ll know why. Thanks for reading. Please let us know if you have any questions or comments.