Intersectional inequality in digital development: reflections of a “brown woman”

“All my life I’d heard people tell their black boys and black girls to be “twice as good,” which is to say “accept half as much”. This sentence resonated with me when I read Ta-Nehisi Coates’s Between the World and Me last year. When I first arrived in the UK with my family, my Indian parents expected me to excel and so I attended a briefing meeting about applying to Oxbridge. A teacher came up and asked why I was sitting there. Did I really think I was “Oxbridge material”, she asked? Maybe she is right, I thought at the time, maybe I’m just not that bright. But later, I wondered—that suppression of ambition—would it have been the same, said in the same way, in the same tone, to someone who was a person of colour? To accept half as much?

Perhaps it was that teacher’s comment together with other experiences that motivated me to work in international development (and to teach). Subhashish Bhadra from Omidyar Network put it beautifully when he said his motivation to work in development was about the “ability to be”: “about the women who exercise their ‘ability to be’ in corporate boardrooms and other positions of power. It is about the young boy who exercises his ‘ability to be’ himself in the privacy of his room, away from the prying eyes of family and society. It is about the wheelchair-bound woman’s ‘ability to be’ in spaces that are accessible to her. It is about a gay man’s ‘ability to be’ a customer of dating apps without worrying about being identified. What is life without this ‘ability to be’?”

While working with others to be the best they could, he too wants to be the best he can. In the same vein, what ability do we have “to be” as women of colour, especially as development is about empowering others? First, it’s this tension I’d like to speak about—what is it like to be “a brown woman” (will talk more about that term itself) working in development? Second, I’d like to move beyond binaries of white vs non-white—while these discussions are much needed, I’d also like to raise the intersectionality of inequality in development, and now in digital development.

I’ll be honest—it took me ages to write this. I was nervous of Twitter trolls (known to attack women more than men), of gaining a reputation of “biting the hand that feeds me”, and feeling that I haven’t really suffered like some people have (certainly, no colleague has called me a f*ing bitch as recently happened to Congresswoman Ocasio-Cortez in the USA).

But then I thought these might be all reasons that people don’t speak up. Surely if we work in technology and international development (sometimes called ICTD or digital development), where we discuss key questions of bias, discrimination and exclusion, we should also question our motivations, concerns, and even legitimacy in working in this space? It’s perhaps fitting that so much of the research I’ve been doing in the past few years has been on identification and identity in a digital age, and here these are thoughts on identity—of others and of mine.

It’s becoming clearer and clearer that the development industry suffers from both racism and gender inequality. In terms of race and bias, there have been recent articles questioning the white gaze and how this is built into structural inequalities and how aid flows replicate soft power relationships between former colonial powers and former colonies. USAID, after all, is “from the American people” and UKAID is “from the British people”. Over 1000 staff signed an internal statement on institutionalised racism at Médecins Sans Frontières, perpetuated by hiring practices, culture and a leadership from a “privileged white minority”. Rosebell Kagumire wrote a piece in The Guardian on how she witnessed racism within the UN system.

Add technology to the mix. Eliza Anyangwe makes the point that the white saviour has now morphed into the white startup entrepreneur interested in the African continent while Larry Madowo wrote recently on how Silicon Valley venture capitalists have deep pockets for African startups “as long as you’re not African”. More than ten years ago Gitau et al wrote on how the ICT for development field has so few authors based at African institutions, the latter being biased towards South Africa and Nigeria (it would be interesting to see if and how things have changed today). Earlier this week, there was an excellent discussion on racism in ICTs and development, held by the ICT4D Meetup group.

Gender inequality has been a major component of the development agenda (it has a whole Sustainable Development Goal!) although women’s empowerment has faced critique of its own: Batliwala, Rowlands, and Cornwall are a few of those who have identified the challenge of empowering women without considering the wider implications, and this is especially in the digital age when social norms are not addressed (for example in the increasingly crowded space of digital ID, women simply don’t have access to technology or still rely on male relatives as gatekeepers). In her must-read book Invisible Women, Caroline Criado Perez gives multiple examples of how design by men is for men. 

Combine race and gender though and we hear very few examples from women of colour working in development—what are the challenges they face? How much are they heard?  #AidToo emerged in the wake of #Me too as well as the Oxfam and Save the Children abuse cases—but how much are cultures changing? 

In setting up a blueprint for Black Lives Matter in the development sector, Hannah Ryder writes that at the root of racism lies power and prejudice. But power and prejudice emerge in all forms. I’ve certainly been at the receiving end of discrimination. My voice does not carry the same weight as a white man’s. I’m not saying this to sound like a victim. I’ve worked with same great colleagues but also faced micro-aggressions and then downright discrimination. In a simple case, a white male manager at the World Bank plagiarised my work. When I brought it up (which took a lot of courage), he asked whom people would believe—him or me. Several members of my team complained on my behalf. Everyone was aware of it, and as is typical in systemic discrimination, no one did anything about it. I left. 

A cartoon in the New Yorker which went viral

But inequality is—in Crenshaw’s now-famous term—intersectional. Race, class, gender, physical appearance, age and other individual characteristics intersect with one another and overlap (what about caste in development? Or disability? Or LGBTQ communities? And the intersections between any of these?). It wasn’t just the white manager at the World Bank—other examples of discrimination I faced were being awarded a grant from a tech company, only to be cut out of it by the white female academic who I applied with. Or an Indian male who started shouting at me because he misunderstood something I’d written in an email and told me “I was a lowly consultant who shouldn’t aspire to more”. Yes, there are always such individuals, but at the receiving end, one can’t help think about power and prejudice—would these individuals have reacted differently if it was a white man with the same skill set? Going back to Subashish Bhadra’s words, interactions such as these quell a person’s ability “to be”. They scream “you are not as good as me, and I have the power to do whatever I want”. 

This intersectionality of inequality in development needs to acknowledge all power imbalances. Is a “person of colour” a helpful term in banding everyone together? What about Latinx? And what about white? Gina Crosley-Corcoran wrote an interesting article deconstructing white privilege and how poverty and class illustrate that white is not a homogeneous category, just as gender, being Asian or being a person of colour is not. Indi Samarajiva wrote an honest but shameful (for me) article on how racist South Asians can be. Racism against African students in India, for example, has been appalling. “Fair” is still a common aspiration in India. Skin lightening creams continue to be sold, rebranded from “fair and lovely” to “glow and lovely”. Power and prejudice exist in different forms.

Another intersectional concern for me is that I might be considered more “empathetic” or understanding of women and/or under-represented minorities. But here too there is nuance—I grew up mostly in the UK after the age of 11—do I really understand issues in India as well as someone based in India, not to mention “all brown women”? (Afua Hirsch’s book Brit(ish): race, identity and belonging is excellent). And what happens when I’m on a panel for representation? Whom am I exactly representing? (See Kelan’s critique of inviting a female speaker purely to avoid manels).

On a panel for representation

And what about those we work with? As a researcher, the responsibility of communicating other voices is always challenging. Researchers perhaps have the luxury of asking questions, rather than proposing policies. We don’t talk about “beneficiaries” (horrible word) but of respondents. But it can still feel extractive, and we are still plagued with worries: how much do I understand after two weeks of “fieldwork” (another horrible word)? What right do I have to ask this mother to give up her time and come answer questions for me (compensation for time is also a contentious topic).

One of the biggest challenges is being able to interview time-strapped women for research, but without their voices, are they being represented? One book in development that left a deep impression was Cooke and Kothari’s “Participation: The New Tyranny” which deals with how participation can be oppressive in itself. In digital development, it can also be frustrating, because time and again, the message is, technology is not a silver bullet—think about the context, think about the social norms, but the technologists still seem to find this surprising.

Then again, we talk of the need for “local” researchers and capacity building—but what does local mean? During my PhD research in rural Karnataka, it was hard to find a research assistant from Bengaluru (or Bangalore) willing to go to the village of Budhikote for extended periods. A “local” researcher from the next town to Budhikote played on the urban/rural hierarchy and often laced his speech with words such as “uneducated villagers”. So what does local mean?

Like I said at the beginning, I feel as if I’m not saying anything new, but power and prejudice remain. There is so much left to deconstruct … unpaid internships, unequal pay, contracts of small pieces of work where intellectual capital is lost with the end of a project, no stability for the contractor, visas and right to work etc etc … I’ve worked across the private sector, academia and NGOs. Going only by what I have encountered, in terms of comparison and perhaps unexpectedly so, the tech sector was the most inclusive (although we know about Satya Nadella’s comments on women needing to be patient for a pay rise). Discrimination in academia is bad (in terms of gender as well as race—we know about academic review processes that commonly disregard women’s additional unpaid work, now exacerbated during COVID, or how male professors are frequently given higher reviews than female ones, even if they teach an identical course. And similar discrimination comes up in terms of race. But I have found the development space the most imbued with power and lack of accountability. 

Not all women and not all “women of colour” will feel the same of course. Plenty of women and women of colour are doing amazing work, and are great role models. And not all women, or women of colour, are angels—Sarah Drinkwater makes the point that one of the reasons she admired Elizabeth Holmes, the CEO of Theranos, was likely because there were so few female role models. She writes “if we had more women on stage, we would pick up the charlatans earlier, right? Because it doesn’t matter what demographic you’re from, you can still be kind of a charlatan”. 

At the end of this you might ask “what do you suggest we do?” On the one hand, like Browning and Kagumire, I feel a sense of fatigue around both race and gender, compounded by the intersectional element of age. People have talked about inequality in different forms for years. Is anything changing? Do I have any right to speak on behalf of others, even though I hope I am relaying their voices? As I get older, I want to be a role model for junior colleagues, students and my young nieces, but feel like I haven’t figured out the leaning in myself (see Michelle Obama’s response to lean in). The challenge of working twice as hard, to paraphrase Coates, is tiring. On the other hand, we know the efforts our mothers and grandmothers and other strong women made. We need to build on those and continue fighting for a more equitable future.

#Metoo, the COVID-19 virus and BLM are providing ample opportunities for the development sector to question power and prejudice, whether in terms of gender, race or other aspects of exclusion. Ryder provides a thorough blueprint for development organisations to move towards equality and diversity. Similarly, Engine Room has written on deconstructing knowledge and power in work. And Maria Faciolince, leading the Power Shifts initiative at Oxfam, has compiled a comprehensive list of resources on dismantling individual and institutional racism in development. 

So on a larger scale, it’s about constantly absorbing and educating oneself (such as by reading some of the references mentioned above). In action, it’s about humility, inclusivity, and diversity. Intersectionality shows that power is not binary: it morphs and shifts, and in many cases stems from insecurity (“I need to do this for my career, my ego etc”). We need to treat others with respect and kindness, as equals, and feel free enough to speak up when we don’t see it (I’ll admit it though—I could have mentioned other cases but didn’t feel comfortable doing so).

I wrote this as a personal piece but then shared it with the company I consult for, Caribou Digital. Together with other colleagues concerned about issues of diversity, we’re working on a discussion session at our upcoming virtual retreat, and what actionable change we can work towards. Many of these issues are deeper ideological and structural issues in development, but we can do our best as an organisation working in this space. Another colleague and I started a code of conduct for Caribou Digital which we’re working on improving—we need to run this through everything we do. I still believe most of us working in development have good intentions, but what does good mean, and what biases are embedded in it? I’d love to hear your stories—good and bad—and together I hope we can make a change both in the development space and beyond.

Thanks to all those who gave feedback on drafts, and to Sharon Rhodes for proof-reading!

Confusing biometric ID experiences at a young age: voices from Thailand (5/5 blogs on youth & ID for UNICEF)

Hélène Smertnik and Savita Bailur

This blog is one of five blogs from our work on children and the tension between identification and identity in a digital age. Other blogs include:

  • Identification and identity for children in a digital age
  • Identification, identity and sexuality for youth in Brazil
  • Refugee identification and identity for children in Lebanon
  • Financial inclusion, identification and identity for children in Kenya

At Caribou Digital, with the support of UNICEF, we undertook a multi-country research study to understand the interplay and challenges between identification, identity, and digital ID for children in their journey to adulthood, from 0 to 18. What tensions appear between the ID credentials (ID) one owns, the way one is identified (identification), and the way one identifies oneself (identity) during this formative period?

To do so, we conducted fieldwork in four countries: Kenya, Brazil, Lebanon, and Thailand. Each country surfaced a different theme and ID challenges: in Brazil, we saw tension between identification and gender/sexual identity; in Kenya we saw the consequences of early and informal financial inclusion due to the high penetration of mobile money; and in Lebanon, tensions in the identity-making of refugee children.

In Thailand we focused on the experience of acquiring identification credentials at a young age, especially when the credentials are biometric.

The graphic below maps a Thai child’s accumulation of ID credentials as well as personal data, from childhood to adulthood. ID credentials are represented by the solid grey blocks, divided by development age ranges. These credentials are both legal, such as a birth certificate, and non-legal, such a Facebook account. The floating dots illustrate the child’s growing amount of personally identifiable data (PII) over time (e.g. a date of birth, a password, or a Google search). By the time a child reaches 18, a huge amount of PII is available on them.

Identification, identity, and digital ID in Thailand for children between 0 and 18

Visual developed with UNICEF

In 2005, the government of Thailand decided to shift from a paper-based ID to a biometric smart ID; in 2011, this was followed by a change in the age to obtain the national ID from 15 to 7. These policies intend to improve access to essential services as well as curb identity theft by creating one unique smart card from a young age.

Note that in this blog we focus on the topic of children with Thai citizenship. Non-citizens have additional challenges. We spoke to a few Thai Yai (non-Thai) children and their families, who raised concerns of not being able to access services, particularly higher-level schooling resulting in potential unemployment because of a lack of ID. However, we did not explore this in-depth, and this is a topic that deserves much deeper discussion in Thailand. (We do address the challenges of refugees, identification, and identity in our blog on Lebanon [link].)

While there have been discussions in Thailand around the issues of seven-year-olds obtaining an ID (the article referenced is in Thai but discusses the importance of photo identification, something a birth certificate lacks), the larger discussion revolves around the biometric capture. Research on the topic of biometric ID enrollment for children globally often focuses on benefits for governments, technical aspects, (e.g. the recognition accuracy of children’s biometric data), or privacy issues (e.g. risk of hacking of stored data). A recent study by UNICEF (July 2019) provides comprehensive guidance for actors looking at implementing biometric authentication projects for children. While a large part of the study looks at the technical and legal aspects of implementation, it also warns against the risks specific to children as well as the question of user acceptance.

We conducted this qualitative study over two weeks, at the end of 2018 — interviewing over 80 children, youth, caregivers, and frontline workers (e.g. medical staff and school teachers) in and around Chiang Mai — to further explore user acceptance by looking at ID enrollment through the eyes of youth. It was preceded by a short, online U-Report survey conducted by UNICEF Thailand to collect preliminary information on youth’s experiences and opinions around identification and biometrics. During our fieldwork, we focused on Chiang Mai (Thailand’s second-biggest city) and its surrounding areas. We interviewed a total of 100 respondents through focus groups and interviews. These included accompanied children (11 to 14), youth (15 to 20), teachers, nurses, parents, and caregivers as well as a few experts and ID practitioners.

At seven years old, children are asked to obtain identification credentials without having the ability to fully understand the meaning of it

Lawan (pseudonym), turned seven a few months ago and had just recently gone with her mother to register for her National Identity card (NID). According to her mother (Lawan was too shy to tell us), Lawan saw the NID as a toy: “all her friends at school had it and so she wanted one as soon as she could to not feel left out.”

While Lawan’s mother was happy about her daughter’s enthusiasm, she noted the difficulty of getting an ID at seven: “I would’ve preferred for her to do it at 15, as in the past. The child is too young to understand and it is difficult to explain what the NID is.”

Sarah, a 16-year-old adolescent, also highlighted the issue of not understanding the value of the NID when so young. She remembers that it was just something that needed to be done. “It is only when I passed my first school exam (at 13), requiring my ID, that I realised that this is an important document. My parents still keep my ID today.”

Getting a biometric ID at a young age: intimidating and confusing

Several adolescents we spoke with recalled uncomfortable experiences when getting their biometric ID. Being so young and without a full understanding of what the NID is, or what the process entailed, meant that when errors occurred or authentication failed they would take it as a personal matter, as Waen’s quote illustrates:

“The first time the officials could not see [read] my fingerprint on the machine. I thought something was wrong with me and the agent didn’t help, he just told my parents we would need to come back. I was worried.” — Waen, 14 year old

In addition to the uncertainty it can create for young children, the process of collecting data itself was sometimes remembered as quite intimidating, as Vanida, now 13 years old, explained: “They told me not to smile and not to move. I remember I was a bit nervous when they took my photo…the place felt like a prison.”

The fact that children did not know the agents collecting their biometric data added to their discomfort: “We go do the biometric registration with our parents or our teachers, but still we don’t know the people who do the collection. It would be better if it was our teachers.” — Aroon 12-year-old male student

Among the different methods of biometric identification in Thailand, most of the children and adolescents interviewed preferred having fingerprints taken, followed by photographs, and lastly iris scans, which felt the most invasive. Respondents found the fingerprint scanning was the easiest method as they had already done it for school activities. Similarly, UNICEF’s evaluation of user acceptance of biometric data suggests that facial recognition and fingerprint methods were the most acceptable while iris scan was the least acceptable. In addition, a short online U-Report survey was conducted by UNICEF Thailand for this research. The results, out of a total of 71 responses indicated that, overall, adolescent girls (15 to 20 years old) felt less comfortable with biometric data collection than boys.

Therefore, looking at user acceptance before looking at efficiency is recommended. In the case of iris scans, UNICEF’s report suggests it is a highly accurate method of identification but considering its low acceptance by children, shouldn’t be encouraged at that age.

Youth expressed the desire for more control over the ID process

“To give your biometric information, which will stay with the government for the rest of your life, you should be the one who decides what information you give. With Facebook, I decide which picture goes up, I decide which information I give. I would like to be able to decide more when it comes to government information.” — Phueng, 18-year-old female student

Group of 18 to 20 year old students at Chiang Mai University. Credit: T. Rawanghet.

When asking youth to compare the government’s ID procedures and requirements with social media registration and use, especially the ability to choose what information they post online, youth brought up the issue of consent. Older teen respondents understood that the government collected their personal data for security purposes (both the U-report responses and the interviews highlighted this). Still, youth regularly raised “asking for consent” and “providing clear information on the reasons for biometric data collection” as ways to give youth a sense of ownership over their ID rather than just a document for the government to keep track of its citizens.

“When I use my fingerprints on my phone, I know they stay on my phone and I own it. But if I use the data at my workplace, to enter the building, I don’t own the data anymore. Same with the government, they put my information somewhere and I don’t have access.” — 19-year-old student

Takeaways

While acquiring the NID at seven enables more seamless procedures for the government (e.g. usage of one document from an early age instead of the accumulation over time of different documents providing access to different services as portrayed in the graph above), the experiences of children and youth highlights the nervousness some young recipients of ID feel; this should not be underestimated.

Accordingly, here are a few recommendations for key stakeholders to ensure adoption of ID by children is both efficient and makes sense to, rather than confuses, children:

  • The Government, when implementing ID enrollment at a young age, should ensure that the process is as easy as possible for parents and children, that both are adequately informed, and that the process takes into account the young age of the citizens. The government should also consider the most comfortable methods of capturing biometric information on children.
  • Frontline workers, such as schools or maternity clinics, have a key role to play in ensuring that the procedures are done in a way that is understood and adapted to children. Allowing adults to be close to their children, taking the time to explain what is being done and giving information at the end of the process are all important parts of the process which appear to be overlooked.
  • Parents and caregivers need to be well informed as the process relies heavily on them. It is key to involve them and ensure that they are able to inform their children adequately. This is undoubtedly the same as all other bureaucratic procedures, but there is an additional level of vulnerability for both parent and child when biometrics are being captured and children are being handled by strangers.

Acknowledgements

We were particularly aware of the complexity of talking about ID with children and youth and underwent a comprehensive UNICEF Ethics Board review prior to beginning fieldwork. We also made sure to create a friendly, non-intimidating atmosphere with the usual contract of anonymity. We were very grateful for the time and attention that children, youth, families and officials gave us.

We would like to sincerely thank UNICEF Thailand for their support during their field research, in particular Jakub Lambrych for his drive during the research as well as Ploycarat Nana’s support in conducting the U-Report Survey. Finally, our sincere thanks to our Research Assistant, Thanatporn Rawanghet, for her dedication, coordinating the field research, recruiting respondents, interpreting and leading interviews and focus groups, transcribing them over the two weeks and her great support with analysising the research findings.

In addition to the above, we are very grateful to Jonathan Donner (Caribou Digital) and the UNICEF team for feedback. These blogs represent our views and not those of UNICEF.

Social media as self-identity — opportunity and risk for young refugees in Lebanon (4/5 blog on youth & ID for UNICEF)

Hélène Smertnik and Savita Bailur

This blog is one of five blogs from our work on children and the tension between identification and identity in a digital age. Other blogs include:

  • Identification and identity for children in a digital age
  • Identification, identity and sexuality for youth in Brazil
  • Biometric identification and identity for children in Thailand
  • Financial inclusion, identification and identity for children in Kenya

At Caribou Digital, with the support of UNICEF, we undertook a multi-country research study to understand the interplay and challenges between identification, identity, and digital ID for children in their journey from 0 to 18. What tensions arise between the ID credentials (ID) one owns, the way one is identified (identification), and the way one identifies oneself (identity) during this formative period?

To answer these questions, we conducted fieldwork in four countries: Lebanon, Brazil, Kenya, and Thailand. Each country surfaced a different theme and ID challenges: in Brazil, we saw tension between identification and gender/sexual identity; in Kenya we saw the consequences of early and informal financial inclusion due to the high penetration of mobile money; and in Thailand, the confusing experience of acquiring biometric ID at a young age.

Here, we focus on the experiences of young Syrian and Palestine (the preferred use rather than Palestinian) refugees, navigating their identity on social media in Lebanon. We look at both the opportunities and risks that social media presents for refugee youth.

This qualitative research was conducted over two weeks during which we interviewed a total of 100 respondents through focus groups and interviews organised with the support of local NGOs in three refugee camps in the Bekaa Valley, Akkar, and Shatila. These included accompanied children (11 to 14), youth (15 to 20), parents, caregivers, frontline workers (school heads and teachers, health workers, and NGO representatives), and government experts.

The ID ecosystem for refugees is complex

Obtaining identification documents as a refugee (when they can) is complicated. In addition, some refugee parents may not want their children to be identified in their host country under the assumption that their children will not be able to return home. Legal procedures and requirements for acquiring ID are beyond the scope of this research; rather we focus on refugees’ experiences of identification and identity before adulthood.

The graphic below maps some of the ID credentials and personally identifiable data (PII) a child with refugee status may accumulate, from childhood to adulthood. ID credentials are represented by the solid grey blocks, divided by development age ranges. These credentials are both legal, such as a birth certificate, and non-legal, such a Facebook account. The floating dots illustrate the child’s personally identifiable information (PII) as it expands over time (e.g. a date of birth, a password, or a Google search). By the time a child reaches 18, a huge amount of PII is available on them. The key credentials for Syrian and Palestine refugees in Lebanon includes but is not limited to: a birth certificate or civil extract, a family booklet, a national ID card, a residency permit, and a UNHCR card for Syrian refugees or UNRWA card for Palestine refugees. We note that the status and history of Syrian and Palestine refugees in Lebanon are widely different. This diagram only aims to give an idea of the type of documents they both can obtain.

Identification, identity and digital ID in Lebanon for children between 0 and 18

Visual developed with UNICEF

Creating an identity vs. identification, experiences from childhood

The refugee respondents in our research had come to Lebanon as families. We noted that children start building their identity, as Syrian or Palestinian, within their household as well as at school, through storytelling. Parents and educators play an important role in this foundational step of understanding where refugee children come from as reflected in the respondent quotes below.

Here at the center (NISCVT), we made a tree during our activities. Each branch of it represented a town in Palestine. My family comes from Yaffa — a town known for citrus fruit — so we used lemons to represent it.” — Palestinian children in Shatila refugee camp

“At five or six our parents start teaching us that we live in a camp, start telling us about our family and the country we come from. This is how we start building our identity.” — Syrian adolescent

“We, mothers, play a big role in this (non-formal identity). I always taught my children to be proud of themselves, even though they are not Lebanese on paper. Even if they have to use a Syrian ID.” — Lebanese mother married to Syrian man

The first tension between identification and identity arises as early as 8 or 9 years old: a few respondents felt that their identity had nothing to do with the identification credentials they owned. The manager of a center for Palestinian children in Shatila refugee camp, in the outskirts of Beirut, explained:

“An ID document that says ‘refugee’…doesn’t tell you where you are from or who you are.”

Growing your identity through social media…

From a young age, refugee children may use social media to connect with family members and friends and to continue developing a sense of belonging to a country. A 9-year-old Palestine refugee girl in Shatila camps shared her experience:“I started a Facebook page when I was eight for my family in Palestine to connect and tell stories. Even some relatives in other camps joined the group to communicate and share. Now, we are using WhatsApp more as it’s easier.”

At adolescence, respondents we spoke with suggested that social media, notably Facebook, could be an outlet for expressing one’s identity and interests as well as for developing a sense of community. This was the case of this 18-year-old Syrian mother: “I like to paint. When I finish a painting, I like to post a picture of it on Facebook and get feedback from others. And I can also do so in person; through discussing it with others.”

Some adolescents mentioned they used social media to also connect to humanitarian organizations, like this young 15-year-old Palestinian girl and her group of friends: “We express our solidarity through some posts on NGOs’ Facebook pages when something bad is going on in Palestine such as the war in Gaza. I also have my own writing blog and I’ve connected with new people through it.”

… but with limits

Digital platforms and social media can be an empowering tool but also a possible nuisance for refugees — particularly as they become adolescents. After 15, young boys are careful to not share personal information on digital platforms for fear of being tracked by their government, in particular for fear of having to go back to complete the military service. A 20-year-old male Syrian medical student now living in Akkar refugee camp explained his cautious approach to using social media: “I do put some things on Facebook. Maybe an article occasionally. I hardly shared any photos and now I definitely don’t anymore. Especially after the crisis, after 2014. I am wanted by the army because I have not completed my military service in Syria. You normally have to complete it when you turn 18. If you don’t serve, you can go to prison for six months. This is the first stage, I don’t remember what comes after that.”

There are multiple pulls of identity-making, in addition to that of being a refugee. The adolescent girls we spoke with said that, after marriage, they do not use social media much and, when they do, they may use their husbands’ accounts. Several of those we spoke to said they were under the legal age of marriage but it was accepted. However, with this came both online and offline restrictions:

“I am not allowed to leave the house alone. Only with my mother or mother-in-law. Not even with my sisters or his. The mothers would always have to be with us. I just have my own personal hobbies. I like to do hair and makeup, but I just do this myself at home. I don’t share it with anyone or online.” — 16-year-old Syrian mother and wife

“I do not use social media. My husband doesn’t accept it. He has Facebook, and I can use his account sometimes. He has given me a phone, but he doesn’t like me using Facebook. I tried to convince him a few times but he refused. His reason is that he worries about other men that might be on there.” — 17-year-old Syrian refugee

For refugees, the risks of becoming too visible increase at adolescence and can have serious implications in the real world with government entities tracking young men or as cultural norms prohibit young women’s expression on these platforms.

Takeaways

It is key to acknowledge this tension around identification and identity for refugee children. Social media offers an incredible opportunity to connect with family, peers, and community when uprooted, but it also leads to the risk of too much visibility — particularly for adolescents. Key actors, notably at the community level but also within international organisations, can help ensure this tension is understood and managed adequately.

International organisations working with refugees, such as UNHCR, UNRWA and community-level organisations, should include materials on how to manage social media and its implications for both adults and children in their training and information sessions.

This recommendation is in line with the need for “data accountability” to be built into future policy and guidance documents for Accountability to Affected Populations (AAP) discussed in another piece of Caribou Digital research looking at identification systems for refugees.

Acknowledgements

We were particularly aware of the complexity of talking about ID with children and youth, furthermore in a refugee context, and underwent a comprehensive UNICEF Ethics Board review prior to beginning fieldwork. We also made sure to create a friendly, non-intimidating atmosphere with the usual contract of anonymity. We were very grateful for the time and attention that children, youth, families and officials gave us.

We would like to sincerely thank UNICEF Lebanon for their support during their field research as well as our Research Assistant Diala Hajal for all her help and efficient coordination. We would also like to thank Aimee Ghanem for her great help in speaking with experts in the field. Finally but not least we sincerely thank the UNRWA team for their time and review as well as the NGOs that enabled us to meet with respondents, creating a safe environment: the Collective for Research & Training on Development — Action in Beirut, The National Institute of Social Care and Vocational Training in Shatila refugee camp, Relief and Reconciliation in Akkar, the International Rescue Committee in the Bekaa Valley, and the Norwegian Refugee Council.

In addition to the above, we are very grateful to Jonathan Donner (Caribou Digital) and the UNICEF team for their feedback. These blogs represent our views and not those of UNICEF.

Early financial inclusion can lead to exclusion: experiences of Kenyan youth (3/5 blog on youth & ID for UNICEF)

Hélène Smertnik and Savita Bailur

This blog is one of five blogs from our work on children and the tension between identification and identity in a digital age. Other blogs include:

  • Identification and identity for children in a digital age
  • Identification, identity and sexuality for youth in Brazil
  • Refugee identification and identity for children in Lebanon
  • Biometric identification and identity for children in Thailand
An engaged group of students, ages 15 to 18, at their high-school in Kisumu, Kenya’s third biggest city. Photo credit: H. Smertnik

At Caribou Digital, with the support of UNICEF, we undertook a multi-country research study to understand the interplay and challenges between identification, identity, and digital ID for children in their journey from 0 to 18 years old. What tensions appear between the ID credentials (ID) one owns, the way one is identified (identification), and the way one identifies oneself (identity) during this formative time?

To do so, we conducted fieldwork in four countries: Kenya, Brazil, Lebanon, and Thailand. Each country surfaced a different theme and ID challenges: in Brazil we saw differences between identification and sexual identity; in Lebanon, tensions in the identity-making of refugee children; and in Thailand, the confusing experience of acquiring biometric ID at a young age.

In Kenya, we focus on youth’s early and, as a result, informal access to financial tools which are only legally accessible at 18; we found that an “inclusive” journey towards financial independence can potentially become exclusionary when started too early.

This qualitative research was conducted over two weeks during which we interviewed a total of 100 respondents through focus groups and interviews. These included accompanied children (11 to 14), youth (15 to 20), parents, caregivers, frontline workers (school heads and teachers, health workers, and NGO representatives), and government experts. We focused on Kisumu (Kenya’s third largest city) and its surrounding rural areas.

By middle childhood, children are accessing mobile apps and, often, mobile money

The graphic below maps a Kenyan child’s accumulation of ID credentials as well as personal data, from childhood to adulthood. ID credentials are represented by the solid grey blocks, divided by development age ranges. These credentials are both legal, such as a birth certificate, and non-legal, such a Facebook account. The floating dots illustrate the child’s personally identifiable information (PII) as it expands over time (e.g. a date of birth, a password, or a Google search). By the time a child reaches 18, a huge amount of PII is available on them.

Identification, identity, and digital ID in Kenya for children between 0 and 18

Visual developed with UNICEF

By middle childhood, awareness and use of ID credentials begins: around age 13, most children will pass their first school exams for which they will need to show an ID (i.e. their birth certificate). While youth have to wait until 18 to obtain their National ID card (NID) and obtain full and formal independence, particularly financial independence, some of the adolescents we spoke with didn’t wait until 18 to begin using their parents’ or older siblings’ mobile phones, notably to access financial apps.

Youth are cautious about their online presence but less cautious about mobile money use while underage

“Many African parents say that to perform, their children shouldn’t have a phone…The only digital platform I use is YouTube. I love fashion and watch plenty of shows. But any other trace of me on the internet I would not want.” — Paulette, 16 year old student in Kisumu

Many of the youth we spoke with were as cautious of their presence online as Paulette, not wanting to share too much of their personal information online (although they may also be unaware of how much data they generate). This attitude stemmed from peer warnings and parents’ own caution around having their children use the phone. Despite this caution, the same youth often started using their parents’ mobile money account as young as 12 or 13. While technically illegal (the owner of the phone and the M-PESA account should conduct the transaction and show an ID), interviewees stated it was common practice.

“I use my phone to learn and meet people but no dating!!! For M-Pesa, it’s my mother who is registered. She gives me her ID number and I go to the agent. No need for a physical ID. When I use my sister’s account, I use my sister’s ID.”Shamza, 15 year old student in Kisumu

Accessing mobile money from a young age is not the primary issue here: after all, some children receive “piggy bank” money from a young age. The problem arises from the wider array of mobile financial services, particularly loans, that become suddenly available to youth while they are still not fully aware of the implications of using these services.

Becoming financially included too early in life can lead to exclusion later

“I was in need of a bit of cash and downloaded Tala on my phone… actually I was seeing a girl and wanted to take her out. I read some bad reviews on Facebook, someone saying that they didn’t know why they were blacklisted when they had reimbursed, but I decided to try anyway. I needed an M-PESA account to register with Tala, and my brother was ok so I registered on M-PESA with his ID. The app says it collects some information on us — not sure what — to see if we can get a loan. For me it took two days and I got the loan.”Paul who started using apps at 16

Discussing with Paul about the implications of getting loans before being legally able to and using an ID that wasn’t his revealed that he had not considered these matters earlier. He did realise that his brother would be exposed, as the ID holder, if he ever did wrong. He also noted that it may mean that his brother could take his money and that seemed to be a motivation for him to change his mobile money account to his own ID.

Accessing financial tools and thus being “included” relies on identification, specifically the National ID (NID). Over the years M-PESA have made the NID requirement more and more stringent in order to avoid fraudulent usage of the service. However children and youth conducting transactions either on behalf of their parents or for themselves via their known and trusted neighbourhood agent is a grey zone. While this flexibility is most likely advantageous, in some situations it could make youth less conscious of the impact of using mobile money for services other than simply sending and receiving money.

Loan apps such as Tala, as Paul mentions, or Branch offer quasi-instant loans to youth who do not always know or understand the terms and conditions and sometimes find themselves blacklisted and unable to access financial services for failing to repay as little as 200 KES (2 USD), seriously impacting their credit history as they grow older.

Youth are particularly exposed and vulnerable as the main target of this easily accessible money (e.g. radio stations in Kenya make these loans sound almost like free money). If apps do not ask for too many formal credentials and make the background checks invisible, they appear extremely attractive. However, it’s likely that poor or no repayments will place debtors on blacklists that will affect them and possibly financially exclude them as adults.

Takeaways

The familiarity with using mobile phones from a young age — as well as less stringent KYC requirements compared to banks — may lead some youth to undervalue the formal nature of digital financial transactions and consider the phone as a gateway to easy money. As we have seen, this has implications for both the youth using these digital financial services without understanding the consequences (particularly loans) and the holder of the ID who will be held accountable for any misuse. Special attention therefore needs to be paid to youth’s usage of digital financial services:

  • For government and community level engagement to ensure education of both families and youth about digital financial literacy and privacy (not sharing IDs the way one doesn’t share an ATM pin). Parents should know to adopt the right attitude when asking their children to do transactions for them and convey the official nature of financial transactions which require official documentation.
  • For regulators to engage with digital financial service providers, such as Tala, on the need to ensure strict ID requirements and cross checks to ensure the age and solvency of the borrower.
  • Frontline workers, particularly teachers, can play a role in directly informing children of the need for caution and the implications of using financial services.

Acknowledgements

We were particularly aware of the complexity of talking about ID with children and youth and underwent a comprehensive UNICEF Ethics Board review prior to beginning fieldwork. We also made sure to create a friendly, non-intimidating atmosphere with the usual contract of anonymity. We were very grateful for the time and attention that children, youth, families, and officials gave us.

We would like to thank UNICEF Kenya for their support during this field research, particularly Minu Limbu and Moses Rono. Finally, our sincere thanks to our Research Assistant, Diana Mwaga, for her support, coordinating the field research, recruiting respondents, interpreting interviews and focus groups, and transcribing these over the two weeks.

In addition to the above, we are very grateful to Jonathan Donner (Caribou Digital) and the UNICEF team for feedback. These blogs represent our views and not those of UNICEF.

Identification, identity, and sexuality in Brazil (2/5 blog on youth & ID for UNICEF)

Savita Bailur, Cecília Peres, and Hélène Smertnik

This blog is one of five blogs from our work on children and the tension between identification and identity in a digital age. Other blogs include:

  • Identification and identity for children in a digital age
  • Financial inclusion, identification and identity for children in Kenya
  • Refugee identification and identity for children in Lebanon
  • Biometric identification and identity for children in Thailand
Bianca talking about her experience of identification and identity. Photo credit: S. Bailur.

“My national ID says my name is Rodrigo, but on my school card and Facebook, I have my social name — I am Bianca. I was told I couldn’t change my official name until I was 18. I had to live with this name and formal identity all through school. I went to the registrar as soon as I could after my 18th birthday, but they said I had to have a body examination or a letter from a psychologist to change my name. Then I can only change everything else after I change my birth certificate.”

At Caribou Digital, with the support of UNICEF, we undertook a multi-country research study to understand the interplay and challenges of identification, identity, and digital ID for children in their journey to adulthood, from 0 to 18 years old, especially for emerging economies in a digital age. To do so, we conducted fieldwork in four countries: Brazil (which we discuss here), Kenya, Lebanon, and Thailand. Each country surfaced a different theme — in Kenya we saw the consequences of early and informal financial inclusion due to the high penetration of M-PESA; in Thailand we saw how children and youth experience biometric identification; and in Lebanon we examined the tensions in the identity-making of refugee children.

In Brazil, one theme was the perceptible conflict between sexual identities in the making and the static identification given at birth. For some, there is a gap between the identification issued at birth and a child’s evolving sense of their own gender/sexual identity, and our research in the South American nation uncovered evidence of how children and adolescents struggle to bridge this gap.

The graphic below maps a Brazilian child’s accumulation of ID credentials as well as personal data, from childhood to adulthood. ID credentials are represented by the solid grey blocks, divided by development age ranges. These credentials are both legal, such as a birth certificate, and non-legal, such a Facebook account. The floating dots illustrate the child’s personally identifiable information (PII) as it expands over time (e.g. a date of birth, a password, or a Google search). By the time a child reaches 18, a huge amount of PII is available on them.

Identification, identity and digital ID for children between 0 and 18 in Brazil

Visual developed with UNICEF.

We began each focus group with older youth (14–18) in Brazil by asking if they could say what kind of identifiers would be acceptable on a digital national ID (a digital version of the “carteira de identidade”, or RG) and what wouldn’t. In one focus group, youth felt that a digital ID could have the following:

A focus group exercise on what data is acceptable to include on a national ID. Photo credit: S. Bailur.

In the photo above, youth were comfortable with identifiers of name, signature, birth place, birth date, blood type, passport number, profession, gender, race, photograph, tax number, and parents’ names. On the other hand, they were less comfortable with identifiers such as a landline or mobile phone number, bank account, political affiliations, allergies/vaccines, whether you were eligible for social security, criminal record, religion, purchase history, and absolutely any connections to social media. While we will not go into detail on this distinction here, there is a clear desire to keep social media (construction of one’s identity) separate from identification.

A focus group exercise on what is not acceptable on a national ID. Photo credit: S. Bailur.

Waiting until 18 for one’s identity

Bianca’s experience and the quotation that began this blog captures the tension between identification and identity. Bianca was an 18-year-old transgender woman we met in the town of Gravatá in northeastern Brazil. She told us “I knew who I was at 13. I came out at 13 … it was like a stab wound for my mother. My father never accepted it. But I saw myself and I see myself as a girl, then a woman. I am a woman. I left my home at 13 and left school at 15 and now I try to earn a living as a cleaner. I hate my deep voice and my facial hair.”

Bianca knew she was Bianca at 13, but on all her formal identification she had always been Rodrigo. In March 2018, Brazil’s Supreme Court (STF) ruled that transgender citizens above 18 years old could alter their names and gender on their birth certificates without previous obstacles such as reassignment surgery, court order, or medical clearance. All they needed to do from now on is simply take the required documents to the registrar and proceed with the desired changes. However, they have to change their birth certificates, a sometimes lengthy process, before they can change virtually any other document.

We met Bianca and “Frida” (below) in a poorer neighborhood of Gravatá in northeastern Brazil. Photo credit: S. Bailur.

Reconciling legal and social name (identification and identity) is challenging

Bianca’s chosen name is what Brazilians formally call the social name (nome social) — i.e. the one by which transgender individuals are socially identified, which can legally replace the name given at birth on the documents issued by the Brazilian government once an individual has turned 18 years old. The Supreme Court’s 2018 decision allowing for legal name changes was an important step towards expanding freedom of expression and mitigating the social exclusion experienced by one of Brazil’s most marginalized groups.

Nonetheless, when Bianca tried to legally change her name and gender, she was told that a binding medical examination was still applicable — a harsh blow to her identification journey. “Doctors define your identity,” she said. Bianca has a voter ID (título de eleitora), her national ID card (carteira de identidade, or RG), her birth certificate, her SUS (Sistema Único de Saúde, Brazil’s Unified [Public] Health System) card, her taxpayer ID card (Cadastro de Pessoa Física, or CPF) and work permit card (carteira de trabalho) — all but one of these names Bianca as “Rodrigo”. All Brazilian citizens are eligible for these credentials, though several interviewees from our higher income demographics were unaware of the SUS (public health card). Importantly, Bianca doesn’t identify with all of these identification credentials in the same way. Of all her identification credentials she identifies most strongly with her voter ID which lists her social name and has no picture.

“I hate situations in public establishments because I say my name which is different from the one public workers see on my ID documents. And the picture on ID documents is so different from how I look in person. They have a hard time accepting my social name and won’t type it down/write it on papers (for example, at the hospital, on a medical report). So every time I have to think — who am I? I found it wonderful because at 13 I could be whoever I wanted to be on social media — I can type any name.” As an illustration of how little Bianca identifies with her legal name and how she asserts her identity, she carries a copy of Presidential Decree No 8.727 or the “social name decree”, inside her purse wherever she goes in case she needs to prove that she has the right to be called Bianca.

Other individuals in focus groups with LGBTQ groups confirmed that Brazilian identity credentials issued at birth became either hated or just irrelevant because details could not be changed until the age of 18: “I hate looking at my birth certificate”,said one respondent. Another recounted how hard his mother fought to assert his right and change his credentials. A third teenager said they didn’t identify with either male or female genders and left forms blank when gender-neutral or “prefer not to say” options were unavailable.

Theo is proud of his social name on the SUS card. Photo credit: S. Bailur.

When your identity is taken out of context and used against you

The digital age exacerbates the tensions between identification and identity for LGBTQ groups, especially in more conserative cultures. Marwick and Boyd wrote famously on concerns of social media context collapse (“all my friends can see everything”) and strategies pursued by teenagers. Duguay writes about LGBTQ youth in the United Kingdom who perform carefully on social media to separate their worlds. Finally, Dhoest and Szulc also write about first and second generation LGBTQ immigrants as well as “sexual refugees” (who were previously persecuted on the basis of their sexuality) in Belgium, and develop ways of demarcating their sexual identities from their other identities when they want/need to keep their sexual identity separate.

In emerging contexts, the less savvy may be at risk if they don’t know how to navigate these worlds. One example in Gravatá was of “Frida”. Frida is a gay 18 year old. When he was around 14 he decided to come out. He took a picture of himself dressed as Frida Kahlo and posted it on Instagram (as shown in the photo). A few weeks later he applied for a job, advertised on Instagram, as a retail assistant and was offered the position. A few days after that, his boss’s wife called him and told him not to bother coming in. He was sure it was because she was a strict Pentecostal Christian, searched back through his profile, and found this photo. He knows the position was then filled by someone else. Frida had not expected his identity to be taken out of context or that the signals he was giving out in a social context would be used against him … as identification.

“Frida”’s Instagram photo which he feels was used against him. Photo credit: S. Bailur.

Takeaways

Obviously, not all LGBTQ youth are unaware of how to navigate the tensions and consequences of identification/identity. However, we have found that all children and youth are navigating the choppy waters of identity-making, and in some contexts, these social identities may be used against them as identification (see, for example, the case of financial identity in Kenya or refugee identity in Lebanon) because they don’t understand the full ramifications. Sexuality is one another way that social identities can serve as a vector for discrimination.

Discovering one’s sexuality in childhood is always fraught with challenges. However, in this digital age, where children grow up quickly and are also more vulnerable, we suggest a few things:

  • While respecting that a minimum legal age for changing identification credentials remains 18, anyone interacting with an LGBTQ youth should recognise and respect social names. In many countries, for example, a natural way to begin a conversation after an introduction might be “what would you like to be called?” Or “can I call you ‘x’”.
  • Children undoubtedly need digital literacy education. Those coming out may need to be made more aware of privacy and context collapse without compromising their exploration. Social media platforms should play a key role here in being more transparent to children and youth about how data is used, “friends” are suggested, and so on.
  • Parents, guardians, (older) family relatives, teachers, medical professionals, NGOs all need to help LGBTQ youth in identification/identity overlaps and challenges, especially from the age of 11 to 13, when many of the respondents said they were becoming aware of their sexuality.
  • Laws on identification must be observed at the local level. A Presidential Decree (№8.727), stating that the social name must be respected, already exists in Brazil, and the Supreme Court has ruled in favor of a simpler social name adoption process (no longer requiring medical examinations at 18). However, the fact that Bianca was unaware of this shows her own lack of knowledge, or at least that of the person(s) she spoke to. Although registrars across the country are legally required to respect the Presidential Decree, they still add obstacles, whether out of ignorance or prejudice. Municipalities (where street-level bureaucrats are most in contact with citizens) must raise awareness about the rights of transgender citizens, including with regard to identification and identity. This would mean cooperation between the mayor, the city representatives, the municipal secretaries of justice/human rights, and so on, to make these formal decisions known and demand they be respected.
  • Finally, we must push for legislation and accountability when PII data (identification) is taken out of context to punish or discriminate against those under 18 who are coming to terms with their identity.

Methodology

This qualitative research was conducted over two weeks in July 2018. During our field work, we focused on Recife, Brazil’s fourth largest city in the northern state of Pernambuco, and the smaller town of Gravatá, an hour from Recife. We interviewed a total of 80 respondents through focus groups and interviews. These included accompanied children (11 to 14), youth (15 to 20), parents, caregivers, youth, frontline workers (school heads and teachers, health workers, NGO representatives), and government experts.

We were particularly aware of the complexity of talking about ID with children and youth and underwent a comprehensive UNICEF Ethics Board review prior to beginning fieldwork. We also made sure to create a friendly, non-intimidating atmosphere with the usual contract of anonymity. We were very grateful for the time and attention that children, youth, families, and officials gave us.

We would like to thank UNICEF Brazil (both in Recife and Brasília) for all their help. Finally, our sincere thanks to our Research Assistant, Cecília Peres, for her support, coordinating the field research, recruiting respondents, interpreting interviews and focus groups, and transcribing these over the two weeks.

In addition to the above, we are very grateful to Jonathan Donner (Caribou Digital) and the UNICEF team for feedback. These blogs represent our views and not those of UNICEF.

Identification and identity for children in a digital age (1/5 blog on youth & ID for UNICEF)

Savita Bailur and Hélène Smertnik

This blog is the first of five blogs from our work on children and the tension between identification and identity in a digital age. Other blogs include:

  • Identification, identity and sexuality for youth in Brazil
  • Financial inclusion, identification and identity for children in Kenya
  • Refugee identification and identity for children in Lebanon
  • Biometric identification and identity for children in Thailand

Children: Not being seen enough and being seen too much

What child has not at some stage felt embarrassed or complained about their baby pictures being shared? What about now, when parents post photographs of children on social media without their permission? When these photographs can stay online indefinitely? Caribou Digital, with the support of UNICEF, undertook a multi-country research study to understand the interplay and challenges of identification, identity, and digital ID for children in their journey to adulthood, from 0 to 18 years old.

The crux of our argument was this: There is an intense (and much-needed) focus on the identification of children, particularly on birth registration. At the same time, while children are at risk of not being identified, there is equal concern that they are at risk of being too identified . . . through social media, the data being collected about them, and profiling from birth. Alongside the birth registration advocates, there are children and privacy advocates. What happens when these two seemingly different worlds interact? Our research showed that identification and identity are interlinked and each impacts the other with major consequences for children and the adults they become.

Photo credit: Piron Guillaume on Unsplash.

Defining identification, identity, and digital ID

Let’s elaborate on these three terms before exploring why they are critical in the context of children. As our colleague Jonathan Donner explored in “The difference between digital identity, Identification, and ID”:

  • Identification implies a process: a proof, system, or transaction involving a subject and an evaluator, centered around verifying a claim that a person is one person and not any other. It also refers to the recording of certain attributes — e.g. biodata and biometrics, which result in a “credential.” A child is identified by relevant authorities as being born by name, parents, date and place of birth. All this results in a birth certificate.
  • Identity, on the other hand, implies a kind of multidimensional social location of an individual relative to other people and institutions around him or her. According to the World Economic Forum, “identity is, literally, who we are: a combination of personal history, innate and learnt beliefs and behaviours, and a bundle of cultural, family, national, team, gender or other identities.” Brubaker defines identification as someone else doing the identifying, whereas an identity is typically self-owned. A child may decide they want a different name from that on their birth certificate, or become estranged from one or both parents and decide they do not want to be associated with them.
  • Finally the credential that results from identification is commonly an ID — and now, there is digital ID. One definition of a digital ID from the World Bank is that of “a collection of electronically captured and stored identity attributes, including biographic data (e.g. name, age, gender and address) and biometric data (e.g. fingerprints, iris scans and facial photographs) that uniquely describe a person within a given context and are used for electronic transactions.” At Caribou Digital, however, we find the term “digital ID” problematic; these days, nothing is either completely digital or non-digital. Instead, the term digital ID sows confusion: it can be used to describe a range of disparate things, from biometric passports to social media accounts. It can mean both identification (someone else doing the identifying, e.g. a biometric passport) and identity (your own identification, e.g. a social media profile). This tension between identification and identity creates a challenge for children as they grow from birth into adulthood.

Visualising a child’s ID journey through development stages

To explore the tensions and overlaps between identification, identity, and ID, we started by looking at all the different IDs (digital and non-digital) a child collects in four different countries: Brazil, Kenya, Lebanon, and Thailand. Through interviews and focus groups, we spoke to around 100 children, caregivers, and expert interviewees in each country, amounting to over 400 respondents in all. We worked with these countries because of UNICEF’s country office interest. Through our research, we noted that children collected ID at different stages, roughly corresponding to what UNICEF has defined as the four developmental stages from birth to adulthood: 0–6 years old, 7–11 years old, 12–14 years old, and 15–18 years old.

We mapped these cohorts in the graphic below:

Identification, identity, and digital ID for children between 0 and 18

Visual developed with UNICEF.

In the graphic above, the fan shapes represent the developmental stages from birth to adulthood while the dots correspond to the amount of PII (personally identifiable information) being collected about the child (name, date of birth, passport number, driving licence number, national ID number, etc.). As the graphic illustrates, the number of dots increases as the child gets older. By 18, a vast amount of data has been collected.

Potential outcomes

As soon as a child’s PII exists, it can be taken out of context. This is certainly not a new phenomenon, but with digitisation, time and space have been flattened and anybody can access, and potentially misinterpret, a person’s PII. For this reason, we talk about “identification and children in a digital age” rather than digital ID. Equally, we don’t separate the two extremes of identification and children in terms of birth certificates or the other end of children and social media — they are inextricably linked.

The consequence of time and space flattening is that people become identifiable long before they become adults. In our Kenya research we saw how this could have implications for financial inclusion later. In other cases, as children grow up, they may feel a disconnect between their identification (that they are refugees in Lebanon, for example) and how they would like to identify themselves (perhaps as a medical student).

The consequences of the conflict between identification and identity for children are as follows:

  • Data can be taken out of context. In Brazil, one respondent, “Frida”, a gay teenager, posted a photo on Instagram in which he was dressed as Frida Kahlo for a party at the age of 14. When he later applied for a job through Instagram, an initial job offer was retracted — he suspects this was because the employer scrolled through his profile and did not approve of his sexuality.
Frida’s Instagram photo which he feels was used against him. Photo credit: S. Bailur.
  • Those in fragile situations may “restrict their identities” through fear of being identified. In Lebanon, once Ali turned 15 he became more and more cautious of his usage of Facebook. As a Syrian refugee he felt at risk of being tracked down and geo-localised in Bekaa valley and either deported back to Syria to do his military service or jailed for fleeing and not having done his service. As reports of countries using smartphone data to track refugees have surfaced, we know that this is a legitimate risk.
  • Others, who may not understand the implications, may make rash decisions that impact them later. In Kenya, Paul, 16, had a date and needed some quick cash to pay for the dinner. He downloaded a loan app on his phone and, after sharing his Facebook details, was screened and received a small loan — enough to cover dinner. Paul felt the whole experience was almost like a game: they didn’t ask anything “serious” like banks do. More critically, Paul had used his brother’s ID to access the phone and app, which means his brother would be implicated if he can’t repay the loan (though he was more worried about his brother using the money!). He has heard of colleagues who weren’t able to open a “real” bank account because of not paying back loans from these “fun” apps. Following a slew of these apps (such as Tala), and the risks they hold, the Kenyan government has recently begun cracking down on small digital loans.
  • Younger children are too young to think about constructing their identity, but the ID process can be intimidating. In Thailand, Wren got her biometric National ID at 7 and her parents didn’t know how to properly explain the meaning of having or obtaining an ID. All she remembers is that a complete stranger asked her to stand still and not smile in an office space that felt a bit like a prison. In retrospect, she would have liked to have more agency over this process — for it to be either explained better or to have had the ID issued later.

Takeaways

While we go into each of these stories and more in our country blogs, overall, we believe we need to discuss this tension between identification and identity more, especially for children who may be struggling or who will not understand the implications until they’re older:

  • Children need greater protection from and awareness of their online behaviour and the data trails being created (see the stories above from Kenya and Lebanon).
  • We need to extend regulations such as GDPR and the right to be forgotten world wide so that data is not used out of context (for example, a Google search on something a child did when young).
  • We also need stronger regulations and messaging from country governments — see the Kenyan government’s Cyber Crimes Act making it illegal to post your naked child’s photo online.
  • Parents, caregivers, and peers need to share and reiterate knowledge of safe behaviour, both offline and online — for example, not allowing a child to use a parent’s identification to collect rations or transfer mobile money — or accept the consequences.
  • We need to constantly respect children and seek their consent when it comes to identification, as we found in our research on collecting biometrics in Thailand, or in Brazil on using social names where teenagers who were coming out felt their identity didn’t conform with their ID. This identification process will shape who they become.

When we asked children and families in our research what an ID meant to them, they talked about:

  • Belonging to a country
  • Being an adult
  • Not being a “bad person” or a terrorist (depending on how old they were)

As they age, children are very aware of the need for an ID, but less so of the intersection between identification and identity, especially in a digital age — this is where we feel more research is needed. For more details on each country study, please see our blogs on Brazil, Kenya, Lebanon, and Thailand. Each country surfaced a different theme: in Kenya we saw the consequences of early and informal financial inclusion due to the high penetration of digital financial services; in Thailand we learned how children and youth felt about biometric identification; in Lebanon we witnessed tensions in the identity-making of refugee children; and in Brazil we saw the conflict between sexual identities in the making and the static identification issued at birth.

We would be keen to hear your thoughts on each of these themes as identification and identity for children more generally.

Acknowledgements

In addition to our country research teams, UNICEF country offices, expert interviewees and respondents thanked in each blog, we are very grateful to Jonathan Donner (Caribou Digital) and the UNICEF team for feedback. These blogs represent our views and not those of UNICEF.

Which DFS features matter more to women than men (research for the Gates Foundation from Côte d’Ivoire and Kenya)

Written for International Women’s Day 2020

For the past six months, we’ve been working with the Bill & Melinda Gates Foundation on fascinating gender and DFS (digital financial services) research. The project originated as a gender-sensitive evaluation of the Level One Principles — a set of principles and features to ensure digital financial services are more inclusive.

In the process, our research unearthed women’s opinions and concerns about DFS more broadly. On International Women’s Day, and at a time when DFS are so key for women’s empowerment, we’d like to surface five emerging themes we observed in the fieldwork. Our full research will be published later this year, with data from Caribou Data.

99% school fees are paid by mobile money in Côte d’Ivoire. Photo credit: H. Smertnik

1. DFS providers and mobile network operators need to be clearer about costs

Both in Kenya and Côte d’Ivoire, women in our sample (80 end users in each country, 40 male and 40 female) stated concerns around DFS transaction costs more than men — not only that they were high but also opaque and they didn’t know what they would be charged for. Men were often so aware they had developed workarounds, such as sending multiple lower value transactions below the threshold where they would be charged. One young female student in Nairobi also used this method to avoid fees but on the whole, the women we spoke to did not. Instead, female respondents complained of “hidden costs” and “losing money” in DFS.

Lack of clarity over costs leads to suspicion about mobile money in Côte d’Ivoire. Photo credit: H. Smertnik

2. Money hanging between networks leads many women to multi-SIM or physically transfer money

In both countries, cross-network transactions initiated by women were rare. In Kenya, this is because M-Pesa is so dominant so transfers to/from Airtel are infrequent. In Côte d’Ivoire, women either relied on agents to conduct transactions or “multi-SIMed” with DFS accounts from both Orange and MTN, rather than sending from one network to another. Women also reported that mobile money to bank transactions had challenges, such as money left “hanging”.

“Last weekend, I did a transaction from my M-Pesa to Equity Bank. I found that my money didn’t reach the account I was sending it to, it was hanging for two days… after two days it was reversed to me, with no message. I had to call Safaricom and Equity customer care to understand what had happened.”

In our research, women did not always see mobile to bank transactions as an end-to-end process, and often withdrew money from a mobile money account and physically deposited it into the bank. This cost them time and money but felt more trustworthy.

3. Women value real-time transactions and notifications but want control over these

Previous research has found that women tend to be more conservative about experimenting with new technology or financial products, so consistent service reliability is crucial especially in building trust in the service. One way of mitigating fear is the need for more agent interaction for women than men, where an agent will verify the transaction.

Since there’s suspicion about money being “lost in the system”, real-time transactions and notifications provide a great deal of confidence and trust. Among those interviewed, women emphasized the need for notifications more than men, because it gave them security and saved them following up. In Côte d’Ivoire, one interviewee remembered taking her brother-in-law with her for support:

“I paid a bill and have not received a message for two days. My brother-in-law and I went to the agency to file a complaint. It was found that indeed I had paid the bill by mobile money and that I had not received a message.”

Transactions and notifications of transactions are also affected by poor network quality — for example in Côte d’Ivoire, respondents felt that to send money “to the interior” they had to have an MTN SIM as well as Orange because Orange had poorer network coverage outside of Abidjan.

Even agents in Côte d’Ivoire said consistent notifications would improve their work because their female customers would feel more confident. In Kenya, however, with more advanced mobile money use over Côte d’Ivoire, women wanted better control over notifications so the men in their lives didn’t see them.

Female agent in Lari, Kenya. Photo credit: H. Smertnik

4. Agents are critical for women but quality varies

Uniting “tech and touch” is critical for women. While M-Pesa agents proliferate in Kenya ( — “You might get only one T-cash agent, or one Airtel money agent, but M-Pesa agents … they’re everywhere, everywhere is green,” said a female student in Nairobi), in Côte d’Ivoire, women reported that agent availability and quality varied.

Respondents didn’t express preference for male or female agents, but just wanted a professional who could provide advice and clarity on costs. They also wanted more standardized support rather than varied levels of help though they acknowledged personal relationships mattered here.

Purely digital products were not fully trusted: “these Tala apps, they don’t even have agents so do they really exist? How can they help you if there is no one to go to?”.

On the other hand, men don’t express this need for agents, instead preferring to conduct transactions themselves or asking friends for help.

However, this also increases women’s dependence on agents, and from an agent perspective, seems to add some pressure: one agent in Côte d’Ivoire resented the consumer education he had to provide:

“It’s not my job to raise awareness. When I did my training, I was shown how to do transactions, how to welcome clients and how to avoid risks and mistakes. I do advise clients how to reduce costs though. I earn very little. It exhausts me and brings me nothing.” Agent in Marcory, Abidjan

5. Lack of ID creates DFS dependency for women

One of the benefits of mobile money is that it often has lower or tiered KYC (“know your customer”) requirements than banking. More often than men, women lack required identification documentation — in low-income countries, 44% of women do not have an ID, compared to 28% of men.

Female respondents in Côte d’Ivoire definitely appreciated lower KYC for mobile money — as one respondent explained: “In banks they ask you for evidence of your salary, bills, things like that. We don’t have all that.” However, it is also common for women to use a SIM registered in someone else’s name and ID because it was too time-consuming to get their own ID — which often caused further problems down the line, like direct access to funds.

Going forward

We’re still analyzing our interviews (in addition to quantitative data), but these findings build on earlier literature on gender and DFS, making the strong case that digital financial service providers and mobile network operators need to understand the needs of women as distinct to those of men and invest more in consumer education and customer service. On International Women’s Day we need to remember that while DFS provide great opportunities for women’s empowerment — the ability to be economically independent — technology shouldn’t create new challenges for women.


We would like to thank our research partners on the ground:

  • in Kenya, the AFROES research team, led by Gathoni Mwai and Sylvia Oloo
  • in Côte d’Ivoire, Empow’Her research team, in particular Soazig Barthelemy, Chloe Roncajolo and Serge Couadio.

And we would like to thank the Global Centre for Gender Equality at Stanford University for feedback, in particular Nicole Figot, Margaret Greene and Angela Hartley.

Videos from women, work and ID research in Bangladesh and Sri Lanka (12/12 blog)

Savita Bailur and Hélène Smertnik

If you’re more of a visual person, please take a look at these videos from our research in Bangladesh and Sri Lanka on the experiences and challenges of getting and using an ID for these seven women. All respondents gave their permission and approval (Tilakeshwari did not wish to be filmed and so with her permission we overlaid B-roll with her voice).

All the blogs from our research are on Caribou Digital’s Medium page.

Thanks to Big Blue Communications and Hélène Smertnik for producing these videos.

This research is funded by Australia’s Department for Foreign Affairs and Trade, via its support for the Commonwealth Digital ID Initiative. We would like to express our thanks to all our respondents, the DFAT iXc team, DFAT Sri Lanka, Sparkwinn Research, Big Blue Communication and our advisory panel for their support.

When ID works for women: summary findings from Bangladesh (11/12 blog on research for Ausaid from Bangladesh and Sri Lanka)

Hélène Smertnik and Savita Bailur

“Before we didn’t have any freedom and respect. With an ID, I can work and get the money and freedom to spend for the family. I feel much better about it now that I am earning and it is not just my husband supporting us.” — Ruma, garment factory worker

In 2019, we conducted fieldwork in Dhaka, Bangladesh to understand the role of ID for women and work, with the support of Australian Government through the Department of Foreign Affairs and Trade and as part of the Commonwealth Digital Identity Initiative. Our research sought to explore:

  1. The potential role of ID in enabling women to access a wider variety of income generating opportunities
  2. The potential role of ID in enabling women to retain and control their own income

We collectively spoke with more than 80 people (mostly women), including domestic workers, garment factory workers and online workers to answer these two questions. We also interviewed employers in these three sectors and development experts in Dhaka and its outskirts (see interview and focus group breakdown at the end of the blog).

Here we share our summary findings of that research.

More detailed blogs on the role of ID for women in Bangladesh are available on Caribou Digital’s Women, Work and ID page.

A focus group discussion with female garment factory workers. Photo credit: BizInsights

ID holds great value in Bangladesh … for the most part

The Bangladesh national ID was originally a paper-based laminated card issued as part of voter registration in 2006 to everyone over 18. As a result, many of those we interviewed used “voter ID” and “national ID” interchangeably — the voter ID has become the national ID. For the past three years — since 2016 — new smart IDs with chips and biometrics have been issued. This has led numerous people to exchange/upgrade their paper-based national ID for the “smart ID”.

The value of ID for respondents is very clear. It was cited as critical by our respondents:

  • For children’s education — to enroll children at school
  • To deliver a child at the hospital
  • To travel outside the home and not be bothered when you get home late at night
  • To bury your family members (you need your ID and their ID)
  • To rent a house, in particular “in better areas”
  • To buy land
  • To open a bank account or a formal savings account
  • To buy a SIM card
  • To be recognised as a citizen of Bangladesh and not be considered a foreigner/ refugee

At the same time, the importance of ID varied according to the job women were looking for or held, as explored next.

“Better”, safer jobs inevitably require ID

While ID was deemed essential to access numerous services, we were interested in diving deeper into exploring whether ID increases access to a wider variety of income generating opportunities and allows women to retain, control and grow their own income. On this, our findings were mixed.

We found there is a spectrum of ID requirements: the more formal the job (i.e. with a contract, better standards and benefits including paid leave), the more likely the need for ID. The experiences of domestic, factory and online workers illustrated that spectrum.

Domestic workers are less likely to view ID as essential to accessing work

Domestic workers were the largest demographic in our sample who did not hold national IDs. For example, Shilpi, a domestic helper who works in four homes has not yet found the time to obtain her ID. Shilpi has reasoned that as it is not currently required for her do her job, she cannot justify the effort and time off work to obtain her ID.

ID is required if women want to increase employment opportunities by working online

On the other hand, Jesmin, a beautician who works through the online platform Sheba xyz, would not have been able to work through the platform without an ID. To verify her identity, Jesmin had to upload her National ID and two photos to join the platform. Without this verification of identity, accessing the opportunities on the platform is not possible. She encouraged her colleague to join her online business but the colleague couldn’t justify going through the hassle of getting an ID, so she is only able to work in beauty salons — with lesser pay.

ID affords access to better benefits and wages in compliance factories

Factory workers were aware of the need to have an ID to work in compliance factories and have the chance to get better pay and more benefits than in non-compliant, non-formal factories. Moushumi, a garment factory employee we interviewed, said: “We won’t be able to find jobs anywhere without an ID card. Nobody will want to hire us because they can’t trust who we say we are. You need it to open a bank account, then you will need it to join a union and learn about your rights at work. You need it if something happens to you, and you want the benefits to be passed to your family.”

ID is an essential credential for financial inclusion, but not sufficient

The connection between ID and financial inclusion is less direct. Respondents were hesitant to open a bank account or mobile account due to a variety of reasons like wanting easy access to cash and feeling like they wouldn’t have access with an account. Some felt they didn’t have enough to save, or were not comfortable with their digital literacy skills to use mobile money. Additionally, the process of opening an account as a woman often means providing personal information to a male agent, which can be uncomfortable.

Because of these obstacles, women often either use their husbands’ accounts or register their SIMs under male members of the household. This brings up the question if these women would have true ownership over their finances if they used mobile money — on the contrary, they may become more dependent on their husbands. Having an ID is the critical first step to financial inclusion, but without facilitated procedures and education on the benefits of having their own accounts, just having an ID is not sufficient for women to feel confident opening accounts.


Follow Caribou Digital on Twitter for more updates on their research.

When ID works for women: summary findings from Sri Lanka (10/12 blog on research for Ausaid from Bangladesh and Sri Lanka)

Hélène Smertnik and Savita Bailur

By Hélène Smertnik and Savita Bailur, Caribou Digital

Following our research in Bangladesh, we conducted fieldwork in Sri Lanka to explore the role of ID in both enabling low-income women to access a wider variety of income generating opportunities and to retain and control their own income, as part of the Commonwealth Identity Initiative.

We spoke with more than 80 people (mostly women) in and around Colombo, including domestic workers, garment factory workers and online workers. We also interviewed employers in these three sectors and development experts.

This blog shares the summary of our research in Sri Lanka.

More detailed blogs on the role of ID for women in the domestic work, garment factories and online sectors in Sri Lanka are available on Caribou Digital’s Women, Work and ID page.

ID ownership, including the National ID card, is widespread among women in Sri Lanka

By and large, Sri Lanka is a country where ID does work for women to the extent that access and ownership of ID credentials is prevalent in Sri Lanka, with both high birth registration (97% — according to UNICEF’s 2017 data) and high National ID (NID) enrollment (95% of men and 90% of women according to the World Bank’s 2017 Findex survey). Almost everyone we interviewed in Colombo and the outskirts had their identification credentials, though we acknowledge this might be different in other parts, especially the post-war north and east of the country.

ID is widespread in Sri Lanka thanks to a robust registration system, supported by schools and the village councils (Gram Sevakas). The NID enrollment occurs at age 16 when students sit for “O-level” school exams, though obtaining the NID is not linked to passing the exam itself.

The government is also moving towards smart IDs, which have a machine readable barcode in addition to fingerprinting. Some of the women we spoke to had heard about these and a few had them, though the majority still had paper, laminated IDs (see photo above of Navodika and Rangala, two factory workers with both types of ID).

Having an ID gives more women more security and choice in the work they do

ID ownership does provide more security (for example, in domestic work, both for the employer and employee) but also provides room for women to consider choices in terms of what type of work to choose. Many domestic workers we spoke with defined domestic work as a “better work” option compared to factory work, because it gave them flexibility (e.g. in terms of work hours and leave). This was in contrast to Bangladesh, where many domestic workers we spoke to felt that was the only option as they did not have an ID and did not feel educated or skilled enough) for “better work”. This flexibility was considered a priority for those women, especially once married and after becoming a mother, even if there is no contract or benefits. The fact that they had the NID meant that that first critical requirement to access work was not an issue to overcome. Tilakeshwari speaks to that choice:

“I have a baby and need to look after her…she is only 2 ½ years old. I can send her to daycare until 1pm, so domestic work is the best option for me.”

… But the prevalence of NID does not protect women from working underage

Ownership of the NID (from the age of 16) does not protect under-age workers as some formal employers, such as factories, still allow under-18s to work, despite seeing their age on their NIDs. Though the factories of Sri Lanka appear to be more regulated and strict compared to those of Bangladesh when it comes to ID requirements to work, respondents such as Navodika, shared various workarounds. We should be particularly concerned about recruitment agencies as loopholes — some respondents said they keep fake IDs to get around ID issues when a girl is under age or does not have an ID:

“When the agency officials found out that I didn’t have one [National ID card], they used to give me someone else’s and send me for [daily] jobs”.

Cultural preference means informal lending continues — having an ID can increase access to formal financial services but alone isn’t enough

Formal banking is well established (we recognize this may be different outside of Colombo, especially in the north and east of the country), enabled by a strong ID penetration. However, many women we interviewed still use informal methods to save and take loans. These informal services do not always require ID, e.g. borrowing and saving through friends/neighbour groups (“seettu”) or pawning of their jewelry but are what women are used to and find most convenient.

Similarly, whether domestic, factory or online workers, most of the respondents still relied heavily on cash. Mobile money services were not commonly used for payments for a range of reasons both cultural and practical (though not because of the lack of ID) such as the need to go to an agent, the fear of hacking or making a mistake, the concern of using data or even that the fear that the husband may think they are using the phone too much. Even online workers, i.e. small entrepreneurs promoting their products and services online, used cash on delivery methods, instead of mobile money or bank transfers for payments.

Those women who use bank services tend to be those with higher incomes, such as Sarah who advertises her hair oils online. Though she still gets paid cash on delivery for her products, she then deposits money into a bank to save and gather interests. Her story highlights ID is a critical stepping stone to build a formal business and grow profits, but she does acknowledge the role of others who helped her access online work which is as important a factor as having an ID.

Takeaways and thoughts

  • Sri Lanka is a country where ID has worked for women: the current ID enrollment process through school is streamlined and has ensured high ID uptake across the country, including for women and girls. This means lack of ID is unlikely to be a reason for women to be denied work. ID enrolment in school (or in fact starting at birth) is a core learning for countries which don’t have high ID enrollment, especially for women (though it also means that girls need to stay in school until 16).
  • However, loopholes still persist in restricting the value of ID — one is that for women who want to work in factories but don’t have an ID for some reason, manpower/recruitment agencies can act as ID intermediaries and how they are regulated/monitored needs more attention. One expert interviewee, an NGO worker, recounted a case of a girl using someone else’s ID through an agency and when she was badly burned in a chemical spill, the factory refused to help, saying she had used a fake ID.
  • Similarly, while ID is required to access formal financial services like banking, ID doesn’t necessarily lead to formal financial inclusion for women as non-formal methods (e.g. seettu and pawning) are still preferred over formal banking services. There is a role for employers to play in encouraging women’s financial inclusion.
  • To conclude, having an ID has been essential for women in Sri Lanka to have access to work, and we certainly saw the value of it, as compared to Bangladesh, where women in our demographic reported it harder to obtain and therefore restricting their choices (although we acknowledge respondent bias). However, cultural aspects (e.g. the prevalence of informal lending despite good ID coverage) show that ID is a necessary but not sufficient factor for financial inclusion. ID is only one component of women getting access to better work and financial inclusion.

Follow Caribou Digital on Twitter for more updates on their research.