Global ID Coverage, Barriers, and Use by the Numbers: An In-Depth Look at the 2017 ID4D-Findex Survey © 2019 International Bank for Reconstruction and Development/The World Bank 1818 H Street, NW, Washington, D.C., 20433 Telephone: 202-473-1000; Internet: www.worldbank.org Some Rights Reserved This work is a product of the staff of The World Bank with external contributions. The findings, interpretations, and conclusions expressed in this work do not necessarily reflect the views of The World Bank, its Board of Executive Directors, or the governments they represent. The World Bank does not guarantee the accuracy of the data included in this work. The boundaries, colors, denominations, and other information shown on any map in this work do not imply any judgment on the part of The World Bank concerning the legal status of any territory or the endorsement or acceptance of such boundaries. Nothing herein shall constitute or be considered to be a limitation upon or waiver of the privileges and immunities of The World Bank, or of any participating organization to which such privileges and immunities may apply, all of which are specifically reserved. Rights and Permission This work is available under the Creative Commons Attribution 3.0 IGO license (CC BY 3.0 IGO) http:// creativecommons.org/licenses/by/3.0/igo. Under the Creative Commons Attribution license, you are free to copy, distribute, transmit, and adapt this work, including for commercial purposes, under the following conditions: Attribution—Please cite the work as follows: World Bank. 2019. Global ID Coverage, Barriers, and Use by the Numbers: An In-Depth Look at the 2017 ID4D-Findex Survey, Washington, DC: World Bank License: Creative Commons Attribution 3.0 IGO (CC BY 3.0 IGO) Translations—If you create a translation of this work, please add the following disclaimer along with the attribution: This translation was not created by The World Bank and should not be considered an official World Bank translation. The World Bank shall not be liable for any content or error in this translation. Adaptations—If you create an adaptation of this work, please add the following disclaimer along with the attribution: This is an adaptation of an original work by The World Bank. Views and opinions expressed in the adaptation are the sole responsibility of the author or authors of the adaptation and are not endorsed by The World Bank. Third Party Content—The World Bank does not necessarily own each component of the content contained within the work. The World Bank therefore does not warrant that the use of any third-party-owned individual component or part contained in the work will not infringe on the rights of those third parties. The risk of claims resulting from such infringement rests solely with you. If you wish to re-use a component of the work, it is your responsibility to determine whether permission is needed for that re-use and to obtain permission from the copyright owner. Examples of components can include, but are not limited to, tables, figures, or images. All queries on rights and licenses should be addressed to World Bank Publications, The World Bank, 1818 H Street, NW, Washington, DC, 20433; USA; email: pubrights@worldbank.org. Image credit: Shutterstock/By Artistdesign29. About ID4D The World Bank Group’s Identification for Development (ID4D) initiative uses global knowledge and expertise across sectors to help countries realize the transformational potential of digital identification systems to achieve the Sustainable Development Goals. It operates across the World Bank Group with global practices and units working on digital development, social protection, health, financial inclusion, governance, gender, legal, among others. The mission of ID4D is for all people to be able to access services and exercise their rights, enabled by inclusive and trusted digital identification systems. ID4D makes this happen through its three pillars of work: ƒƒ Thought leadership and analytics to generate evidence and fill knowledge gaps; ƒƒ Global platforms and convening to amplify good practices, collaborate and raise awareness; and ƒƒ  ountry and regional engagement to provide financial and technical assistance for the C implementation of robust, inclusive and responsible digital identification systems that are integrated with civil registration. The work of ID4D is made possible through support from the World Bank Group, the Bill & Melinda Gates Foundation, UK Government, the French Government, the Australian Government and Omidyar Network. To find out more about ID4D, visit id4d.worldbank.org. To participate in the conversation on social media, use the hashtag #ID4D. GLOBAL ID COVERAGE, BARRIERS, AND USE BY TH E NUM B E R S: iii AN IN-DE PT H LOOK AT T HE 2017 ID4 D-FIND EX S URV EY Acknowledgments This paper was drafted by Anna Metz and Julia Clark with inputs from Leora Klapper, Saniya Ansar, and Jake Hess as part of the World Bank’s Identification for Development (ID4D) Initiative, under the supervision of Vyjayanti Desai. It benefited greatly from peer reviews by Lucia Hanmer and Kartik Srivastava (World Bank) and Alan Gelb (Center for Global Development) and continuous feedback from Jonathan Marskell and Robert Palacios. iv ID 4 D.WO R L D BA N K .ORG Contents About ID4D iii Acknowledgments iv Introduction 1 Data 3 Who does—and does not—have an ID? 4 Country-level variation 4 Individual predictors of ID ownership 6 Why don’t people have an ID? Barriers, needs, and alternatives 14 Reported reasons for not having an ID 15 Explaining coverage gaps for certain groups 16 How do IDs facilitate access to services? 18 Financial and mobile services 19 Public sector benefits and services 24 How non-existent or low-coverage ID systems limit opportunities 28 Conclusion 30 References 31 Annex 1. ID coverage by country and gender 33 Annex 2. ID module in the 2017 Global Findex survey 36 Annex 3. Regression results 38 GLOBAL ID COVERAGE, BARRIERS, AND USE BY TH E NUM B E R S: v AN IN-DE PT H LOOK AT T HE 2017 ID4 D-FIND EX S URV EY Tables and Figures Table 1. Likelihood of reporting different reasons for not having an ID in LICs by demographic group 17 Table 2. Surveyed economies with highest reported use of ID for financial and mobile services 21 Table 3. Likelihood of reporting use of IDs for mobile and financial services by demographic group 22 Table 4. Surveyed economies with highest reported use of ID to access government services and support 26 Table 5. Likelihood of reporting use of IDs for government benefits by demographic group 26 Table 6. ID Coverage: Demographic predictors (regressions for Figure 4) 40 Table 7. ID Coverage in LICs: Interaction models (regressions for Figure 7) 41 Table 8. ID Barriers in LICs: Demographic predictors 42 Table 9. ID Use in LICs: Correlations between having an ID and access to private-sector services 43 Table 10. ID Use in LICs: Correlations between having an ID and access to public-sector services 43 Table 11. ID Use: Correlations with reported use of ID for financial and mobile services 44 Table 12. ID Use: Correlations with reported use of ID for government services and support 45 Figure 1. Global ID coverage gaps by income group and region 4 Figure 2. County-level ID coverage by GDP per capita, excluding HICs 5 Figure 3. Comparing ID coverage and birth registration rates in LICs and LMICs 6 Figure 4. Individual-level predictors of the respondent having an ID (LICs only) 7 Figure 5. Share of men and women in low- and middle-income countries without an ID 8 Figure 6. Ten surveyed economies with the largest gender gaps in ID coverage 8 Figure 7. Probability of having an ID in LICs: differences for men and women 9 Figure 8. Mean ID coverage gap by household income quintile 10 Figure 9. Surveyed economies with the largest income gaps in ID coverage 11 Figure 10. Surveyed economies with the largest education gaps in ID coverage 12 Figure 11. ID coverage by age and education levels (LICs only) 13 Figure 12. Reasons for not having an ID cited by people without one 15 Figure 13. Country-level ID coverage and ownership of a financial account and mobile phone 19 Figure 14. ID ownership and use of private-sector services 20 Figure 15. Self-reported use of ID for financial and mobile services 21 Figure 16. ID ownership and barriers to account opening in LICs 23 Figure 17. Accessing services with and without a foundational ID in LICs 24 Figure 18. ID ownership and use of government-to-person (G2P) transfers 24 Figure 19. Self-reported use of ID for government benefits and services by gender 25 Figure 20. Surveyed economies with highest reported use of ID to access government financial support 27 Figure 21. Share of people unable to access services because they did not have the necessary identity documents 29 vi ID 4 D.WO R L D BA N K .ORG GLOBAL ID COVERAGE, BARRIERS, AND USE BY TH E NUM B E R S: vii AN IN-DE PT H LOOK AT T HE 2017 ID4 D-FIND EX S URV EY Introduction In an increasingly formalized, globalized, and digital world, people need secure and trusted ways to prove who they are for both in-person and online transactions. Without recognition of their existence and credentials that allow them to reliably assert their identity, people are likely to face barriers when attempting to open a bank account, obtain a mobile phone, receive government assistance, register to vote, and more.1 For this reason, United Nations (UN) Member States have adopted Sustainable Development Goal (SDG) Target 16.9: “to provide legal identity for all, including birth registration” by 2030. Yet despite widespread recognition of the importance of being able to prove who you are, the evidence around identification has faced a number of limitations, including precise measurements of who has—and does not have—access to proof of identity, what the main barriers to accessing identity documents are and how they affect different populations, and what the impact of increasing the coverage of identification (ID) systems is on individual and country-level development outcomes. To better understand the scale of the identification challenge, the World Bank developed the ID4D Global Dataset, which provided a first estimate of the number of people who lack official proof of identity—just under 1 billion as of 2018.2 This calculation relies on a number of data sources, including birth registration rates for children and administrative data from ID agencies for adults (or where this data are unavailable, voter registration rates). While the Dataset provides an estimate of the global identification gap, its reliability for measuring coverage at the national level varies by country due to gaps in administrative data.3 As a result, it is not ideal for country-level comparisons or providing more nuanced information such as disaggregation by gender, poverty, or level of education. Beyond measuring the coverage of ID systems, there is also a lack of systematic data on the barriers that people face to accessing IDs, how IDs are used, and the impact they have on people’s lives. While some case studies and other research—including the World Bank’s ID4D Diagnostics4, end-user research conducted by the World Bank and others (e.g., GSMA 2019a, Caribou Digital 2017), and multiple surveys in India (e.g., Abraham et al. 2018, Gelb et al. 2018)—have provided key insights regarding barriers and use, this information is often limited to a handful of countries and/or not representative. In addition, many of the efforts to measure coverage and understand the correlates of identification are focused on birth registration and birth certificates, rather than other types of foundational ID systems and credentials, such as national IDs, population registers, and unique ID numbers5 1 See, for example, Gelb and Diofasi Metz 2018 and World Bank 2017a. 2 ID4D Global Dataset. 2018, available at www.id4d.worldbank.org/dataset. 3 While data collected from ID authorities in some countries is very precise, many others lack up-to-date information on how many people living in their territory today are in possession of at least one government-recognized identity credential. In addition, the dataset uses voter registration as a proxy for ID possession by adults when administrative data is not available. While this data source is useful because of its near universal availability and the fact that most voter registration exercises require people to provide proof of identity, it is an imperfect measure. 4 See: http://id4d.worldbank.org/country-action/id4d-diagnostics 5 See, for example Corbacho et al. 2012; Apland et al. 2014, and Brito et al. 2017. GLOBAL ID COVERAGE, BARRIERS, AND USE BY TH E NUM B E R S: 1 AN IN-DE PT H LOOK AT T HE 2017 ID4 D-FIND EX S URV EY A new set of ID-related questions in the World Bank’s Global Findex survey6 help overcome some of these challenges by providing the first systematic, quantitative evidence on the correlates of ID ownership and use across 97 economies. Specifically, the ID4D-Findex data allow us to conduct an exploratory analysis that provides new insights on the following questions: 1. What individual and country-level factors are most associated with having an ID? 2. What barriers do different groups of people face to obtaining an ID? 3.  ow are IDs commonly used by individuals across demographic and socioeconomic groups H and in different country contexts? Overall, we find that the gap in ID coverage is heavily concentrated in low income countries (LICs), where an estimated 36 percent of adults7—more than 1 in 3—do not have an ID. Within countries, those least likely to have an ID are women, lower-educated adults and those out of the workforce, adults in the bottom of the income distribution, and people living in rural areas. These gaps in coverage are highly intersectional— for example, while women are less likely to have an ID than men overall, the gender difference is largely concentrated among low educated and rural populations. The data also reveal a number of patterns regarding how IDs are used in practice. In general, adults with IDs are more likely to have bank accounts, mobile phones, and receive government benefits and services, and to report that they have used their IDs for these purposes. However, ID use for private sector services (i.e., bank accounts and mobile phones) is much higher than for receiving government financial benefits or other services. We also find differences in terms of gender, education, income level, and rural/urban location in terms of how people report using their IDs. This note begins with a brief overview of the ID4D-Findex dataset and other data sources used in this analysis, and then presents the main results and discusses their implications. We conclude with the need for more evidence in particular areas and specific questions for future research. 6 See the Global Findex Database 2017: https://globalfindex.worldbank.org/ and https://id4d.worldbank.org/global-dataset 7 We use the term “adult” as shorthand for the population covered in this study; this includes those who are at least 15 years of age (the minimum age in the Findex survey) and who are above the required age for obtaining the national ID or equivalent (see following section on data for more information). 2 ID 4 D.WO R L D BA N K .ORG Data The ID4D-Findex data was collected as part of the 2017 round of the World Bank’s Global Findex survey, carried out by Gallup, Inc. as part of its Gallup World poll. The survey was conducted on representative samples of the non-institutionalized civilian population over age 15.8 The survey included three questions related to ID that were asked in 97 countries: (1) whether or not a person had their country’s national ID or equivalent foundational ID credential; (2) for those with the ID, whether they had used it for specific purposes; and (3) for those without the ID, what their reasons were for not having one. In five economies9 with no ID system or very limited coverage, people were asked a single question on whether or not they had been unable to access certain services due to lack of identity documents (see Annex 2 for full text of questions). Combined with other individual-level indicators from the 2017 Global Findex, these data provide new insights into who has an ID, how they are used, and persistent barriers to access. At the same time, the data have some limitations. First, while the ID4D-Findex questions cover countries that represent close the three-quarters of the world’s population, they do not cover all countries. For example, most high-income countries, some low- and middle-income countries, and some smaller economies are not included. In total, the countries included in the ID4D-Findex data account for a little over 80 percent of the population in low and lower-middle income countries. Second, the Findex survey is limited to people ages 15 and older, and therefore does not capture the identity gap or correlates of identification for the youngest people. Importantly, we have restricted the analysis in this paper to those respondents who are above the age when the ID is mandatory, using information from the ID4D Global Dataset.10 In countries where the mandatory ID age is greater than 15 years old, excluding these observations provides more conservative estimates of the gap in identification than if they were included, as young people may not have an ID simply because they are not yet eligible or required to have one.11 Finally, respondents were asked whether they had one specific identity credential, using local terminology to the extent possible (e.g. ‘Kartu Tanda Penduduk (KTP)’ for Indonesia or ‘Aadhaar’ for India). The specific credential referenced in each country was the document most likely to be used by people to prove who they are for a wide variety of purposes—i.e., a “foundational” ID12—which was typically the national ID card or another widespread form of identification in countries without a national ID.13 As a result, the ID coverage data may underestimate the number of people able to prove who they are by other means, including through functional IDs—e.g., voter registration cards, passports, or driver’s licenses—or through informal mechanisms. Despite these limitations, however, the ID4D-Findex data still provide the most systematic and cross-nationally comparable evidence to date on the adult coverage of foundational ID systems for a large majority of people in the developing world. 8 For more details on the Findex methodology, see: https://globalfindex.worldbank.org/sites/globalfindex/files/reports/2017%20Findex%20 full%20report_survey%20methodology.pdf 9 Central African Republic, Liberia, Nepal, Tanzania, Philippines 10 In countries where the ID is not mandatory, the minimum age for obtaining the ID is used. 11 By restricting the sample to respondents above the ID age, we drop 2888 observations across 80 countries (approximately 2.6% of the total sample). Overall, this has the effect of slightly increasing estimates of ID coverage across countries, as people below the ID age were less likely to have an ID than those above. For example, if these observations were included, the estimated percent of adults in LICs with an ID would be 59 percent, compared to 64 percent using the methodology in this paper. See Annex 1 for more detail. 12 Foundational ID systems are those designed to provide identification and credentials to the population for a wide range of purposes and public- and private-sector transactions. In most countries, this has typically included, for example, civil registers, national ID systems, and population registers. Functional ID systems are those creased to manage identification and authorization for specific sectors or use cases, such as voter registration (voter IDs), taxation (tax numbers), social protection (social security numbers), etc. See the ID4D Practitioner’s Guide (http://id4d.worldbank.org/guide) for more information on the different types of ID systems. 13 Note that when describing the results from the ID4D-Findex data, we use the term ‘ID’, ‘national ID’, and ‘foundational ID’ interchangeably. GLOBAL ID COVERAGE, BARRIERS, AND USE BY TH E NUM B E R S: 3 AN IN-DE PT H LOOK AT T HE 2017 ID4 D-FIND EX S URV EY Who does—and does not—have an ID? Like many development indicators, ID ownership varies significantly by region and income group, and access to identification within a given country is also unevenly distributed. Earlier findings from the ID4D Global Dataset, for example, suggest that most people without an ID live in the world’s poorest countries and in Africa.14 Furthermore, a growing body of research suggests that poor people, women and children, rural dwellers, persons with disabilities, refugees and stateless populations, religious and ethnic minorities, and other already vulnerable groups can face significant barriers to obtaining identity documents, and that these vulnerabilities are intersectional.15 For example, wealthy individuals in otherwise disadvantaged groups may be more easily able to overcome social and cultural barriers, while these issues may be compounded for poor people. The ID4D-Findex data allow us to examine the relationship between some of these national and individual-level predictors of ID ownership, providing new evidence on who is most likely to have an ID. Country-level variation The ID4D-Findex survey data confirm that the largest ID coverage gaps are in LICs and in Sub-Saharan Africa. Among LICs, 36 percent of adults do not have an ID, compared with only 5 percent in lower-middle income countries (LMICs). Due in part to its high concentration of lower-income economies, Sub-Saharan Africa lags behind other regions on ID coverage—an estimated 29 percent of adults in the region do not have their country’s foundational ID (Figure 1). Of the 21 countries included in the Findex with an estimated ID coverage of less than 80 percent, all but four (PDR Laos, Afghanistan, Pakistan, and Haiti) are in Sub- Saharan Africa (See Annex 1 for a list of coverage by country). Figure 1. Global ID coverage gaps by income group and region 14 For example, the 2018 ID4D Global Dataset estimates that approximately half of the one billion who lack proof of identity reside in Africa. See www.id4d.worldbank.org/dataset. 15 See, for example: FWLD 2015, IFES 2013, Oppenheim and Powell 2015, NRC 2016. 4 ID 4 D.WO R L D BA N K .ORG As shown in Figure 2 and Appendix, however, there is wide variation in coverage among LICs and countries in Sub-Saharan Africa and South Asia. Within Sub-Saharan Africa, ID ownership ranges from a low of around 21 percent in South Sudan16, to nearly universal access in Botswana and Rwanda. Although 95 percent of adults in South Asia are estimated to have an ID, this high coverage rate is heavily weighted by India’s large population and close-to-universal coverage of its Aadhaar ID system. Other countries in the region have larger ID coverage gaps; approximately 29 percent of adults in Afghanistan do not have an ID (Tazkera), and 20 percent of adults in Pakistan do not have the country’s Computerized National Identity Card (CNIC). In both countries, women are disproportionately represented among those without an ID (see below). Figure 2. County-level ID coverage by GDP per capita, excluding HICs In addition to income level and region, we might expect to find a strong relationship between birth registration rates—the official indicator for assessing progress toward SDG 16.9 and a crucial metric for ensuring proof of legal identity from birth—and the coverage of other foundational ID systems. However, the ID4D-Findex data show that high birth registration rates do not automatically translate into high ID coverage rates for adults, or vice-versa (see Figure 3). In economies such as Zambia or Bangladesh, for example, ID coverage is above 80 percent, while under-5 birth registration rates are below 25 percent. 16 At the time of survey completion in 2017, Malawi had the lowest ID coverage in Sub-Saharan Africa at 16 percent. However, the country has since made significant progress in coverage, registering over 9 million people for a new national ID (UNDP 2017). GLOBAL ID COVERAGE, BARRIERS, AND USE BY TH E NUM B E R S: 5 AN IN-DE PT H LOOK AT T HE 2017 ID4 D-FIND EX S URV EY In contrast, in economies such as Togo or Lao PDR, around 75 percent of births of children under age 5 are registered, while ID coverage is only approximately 40 percent. This finding points to the need for multiple measures to capture the nuances of access to proof of legal identity for different age groups and throughout a person’s lifetime. Figure 3. Comparing ID coverage and birth registration rates in LICs and LMICs Individual predictors of ID ownership Beyond country-level factors associated with ID coverage, the ID4D-Findex data also reveal some important individual-level predictors of who has an ID. All else equal, people are more likely to have an ID when they are in the workforce, married, and older,17 and less likely to have an ID when they have attained only a primary level of education, are female, live in a rural area, and are in the bottom 40 percent of the income distribution. As shown in the full regression results in Annex 3, Table 6, we find similar relationships across all developing economies; however, the size of the association between these demographic characteristics and the likelihood of someone having an ID is the greatest in LICs, and smaller and/or insignificant in middle-income economies. 17 However, as indicated by a small but statistically significant negative coefficient on the age-squared term, the marginal effect of age on the likelihood of having an ID decreases slightly over time. 6 ID 4 D.WO R L D BA N K .ORG As shown in Figure 4, these differences persist even when controlling for other factors. For example, adults living in a LIC with primary education or less are around 17 percentage points less likely to have an ID, compared with adults who have completed secondary school or above, all else equal. In addition, women in LICs are approximately 11 percentage points less likely to have an ID than men, adults living in a rural area are nearly 10 percentage points less likely to have an ID than a person living in urban areas, and those in the bottom of the income distribution are 7.4 percentage points less likely to have an ID than a person in the top of the income distribution. Each of these results is examined in more detail below. Figure 4. Individual-level predictors of the respondent having an ID (LICs only) Overall, an estimated 44 percent of women living in LICs do not have an ID, compared with only 28 percent of men living in LICs (see Figure 5). On average, the difference in ID coverage between men and women in LMICs is less than one percent (significant at the 99 percent level) and insignificant for UMICs. As Figure 4 demonstrates, this gap persists when controlling for other factors. GLOBAL ID COVERAGE, BARRIERS, AND USE BY TH E NUM B E R S: 7 AN IN-DE PT H LOOK AT T HE 2017 ID4 D-FIND EX S URV EY Figure 5. Share of men and women in low- and middle-income countries without an ID However, while men are more likely to have an ID than women in LICs on average, there is significant variation by country. As shown in Figure 6, Afghanistan has the greatest gender gap in ID ownership among the 97 economies surveyed: 94 percent of men have an ID compared with only 48 percent of women. Chad, Niger, Benin, South Sudan, and Ethiopia also have more than 20 percentage-point differences in ID coverage between men and women. In a handful of low- and middle-income economies—many in Europe and Central Asia—however, ID ownership is slightly higher for women than men (e.g., Bulgaria, Czech Republic, Greece, Haiti, Russia, Ukraine, Uzbekistan, Viet Nam). See Annex 1 for a gender breakdown of ID coverage for all countries. Figure 6. Ten surveyed economies with the largest gender gaps in ID coverage 8 ID 4 D.WO R L D BA N K .ORG In addition, the ID4D-Findex data also reveal that the gender gap is largely concentrated among low- educated women and women in rural areas. Figure 7 shows the predicted probability of having an ID based on gender, education level, geographic location, and marital status, controlling for other variables like household income, workforce status, and age. It shows that, all else equal, the probability that a woman living in an LIC with primary education or less has an ID is 65.4 percent, compared with 80.4 percent for a man with primary education or less (an approximate 15-percentage point gap).18 However, men and women with secondary education or higher have an essentially equal probability of having an ID. Similarly, the probability that a woman living in an LIC in a rural area has an ID is 68.8 percent, compared with 82.1 percent for a man (a 13.3-percentage point gap), all else equal, while the difference between urban men and women is negligible. Figure 7. Probability of having an ID in LICs: differences for men and women The reasons for the gender gap in ID coverage are complex and the ID4D-Findex data can only provide some clues. For example, married men in LICs are around 8.5 percentage points more likely to have an ID than unmarried men, all else equal; for women, however, marital status does not significantly change the likelihood of having an ID (see Figure 7). One speculative explanation for this trend is that after marriage, men often become the head of household, taking on more responsibility for accessing services for which an ID is often needed, such as mobile and financial services. Conversely, women who transition from their parent’s household to their husband’s household may not have a similar shift in responsibilities, leading to a relatively constant rate of possessing an ID before and after marriage. We also know that in many of the countries with large gender gaps, women face legal barriers to accessing identity documents. For example, in Afghanistan, Benin, and Pakistan, a married woman cannot apply for a national ID in the same way as a married man (World Bank 2018). Legal barriers to accessing IDs for women are often the result of prevailing social norms and tend to demonstrate deep rooted assumptions 18 Note, these predicted probabilities (also shown in Figure 7) are different than the weighted averages shown in Figure 5 (i.e., that 72 percent of men in LICs have an ID, vs. 56 percent of women). This is because they are based on country-level fixed effects logit regression models that control for a variety of factors at the individual and country level. See full regression results in Annex 3, Table 7. GLOBAL ID COVERAGE, BARRIERS, AND USE BY TH E NUM B E R S: 9 AN IN-DE PT H LOOK AT T HE 2017 ID4 D-FIND EX S URV EY about the appropriate role of women in society. Many of the countries with the largest gender gaps in ID coverage score badly on other indicators of gender inequality. For instance, Chad ranks 158th and Niger ranks 151st out of 160 countries on UNDP’s gender inequality index19. For both men and women, the ID4D-Findex data also demonstrate a significant association between household income, education, age, and geographic location and whether a person has an ID. Within countries, the poorest and least-educated segments of the population are less likely to have an ID than their wealthier and more educated peers. As with gender, these gaps are most prevalent in LICs and in Sub-Saharan Africa. Across LICs, 43 percent of people in the poorest income quintile lack an ID compared with a more modest 25 percent in the richest. For LMICs, the average difference in ID coverage between the top and bottom income quintiles is statistically significant at the 99 percent level, but much smaller: an average of 7 percent of adults in the bottom 20 percent do not have an ID, compared with only 4 percent in the top 20 percent. For the average UMIC, there is no statistically significant difference in ID ownership across the household income distribution. Figure 8. Mean ID coverage gap by household income quintile 19 See: http://hdr.undp.org/en/content/gender-inequality-index-gii. 10 ID 4 D.WO R L D BA N K .ORG Within-country ID coverage gaps based on income, education, and locality also vary significantly by country. Of surveyed economies, the “income gap” is greatest in Togo, Lao PDR, Mozambique, and Ethiopia, where there is a greater than a 30 percentage-point difference between the top and bottom quintiles of the income distribution (Figure 9). Niger, Haiti, and Benin, also all have income gaps greater than 20 percentage points. In addition, we see gaps between the richest and poorest in some upper-middle and high-income countries that also have relatively high coverage in their ID systems, such as Greece (16.5 percentage points). Figure 9. Surveyed economies with the largest income gaps in ID coverage Looking at the “education gap”, we also see a handful of middle-income countries with large differences in coverage between the most and least educated adults despite overall high ID ownership rates. This includes some surveyed economies in Europe and Central Asia, such as Ukraine (12.5 percentage point gap) and Georgia (11.9 percentage point gap), as well as Malaysia (12.6 percent gap). However, the largest gaps in ID ownership between lower- and higher-educated adults are concentrated in Sub-Saharan Africa (e.g., 43.6 percentage points in South Sudan, 34.6 in Niger, and 32.4 in Mozambique). GLOBAL ID COVERAGE, BARRIERS, AND USE BY TH E NUM B E R S: 11 AN IN-DE PT H LOOK AT T HE 2017 ID4 D-FIND EX S URV EY Figure 10. Surveyed economies with the largest education gaps in ID coverage In some cases, the lack of ID may reflect general marginalization of low-educated people. As shown in Figure 11, adults with only a primary education are much less likely to have an ID than those with a secondary education or higher at any age. However, there is also a clear jump in ID ownership around age 18, when people are likely to be taking exams to complete secondary school and enroll in university, both of which typically require proof of identity. 12 ID 4 D.WO R L D BA N K .ORG Figure 11. ID coverage by age and education levels (LICs only) As with gender, we also observe intersectionality between education and geographic location. All else equal, the predicted probability that a person has an ID is lower on average for adults with primary education or less than those with secondary education or more or in the top 60 percent of the income distribution. However, the size of the gap between lower and higher educated adults is greater for those living in rural areas than it is for those living in urban areas—see Annex 3 for regression results. Together with the interaction models summarized in Figure 7, this provides strong evidence that the ID coverage gap is most concentrated among low-income, low-educated women living in rural areas. GLOBAL ID COVERAGE, BARRIERS, AND USE BY TH E NUM B E R S: 13 AN IN-DE PT H LOOK AT T HE 2017 ID4 D-FIND EX S URV EY Why don’t people have an ID? Barriers, needs, and alternatives The gaps in coverage described above may be the result of overall supply constraints as well as multiple economic, social, and procedural barriers that affect people at the individual level.20 For example, people often face high direct and indirect costs to obtaining national IDs or other foundational documents. A synthesis of ID4D Diagnostics in 17 African countries shows that fees for ID cards can be as high as US$ 8-10, and applicants will often need to spend an additional US$ 10–25 on travel costs and supporting documentation (World Bank 2017b). Similarly, ID4D end-user research from one East Asian country shows that even if there is no direct charge for the ID, the combined travel costs, bribes, and lost wages incurred during the application process can come to US$ 15, the equivalent of over two to three days’ income. People living in the most remote and marginalized communities often experience the highest costs due to the large distances to the nearest registration office. These barriers may be multiplied if applicants need to present supporting documents that require additional fees and visits to government offices. Furthermore, people in marginalized groups may also be less likely have the supporting documentation required to obtain national IDs, such as birth certificates or certificates of nationality. Globally, for example, only 56 percent of children under the age of 5 born to families among the poorest 20 percent in their countries had their births registered, compared to 82 percent of those among the richest 20 percent (UNICEF 2017). Beyond these barriers, people may also not have an ID because it is not necessary for their daily lives. In particular, demand for a specific ID is influenced by its perceived usefulness and the availability of alternative methods for proving identity. For example, some people may see little need for an ID if they do not commonly use services or perform transactions that would require it. This might be because formal services are not available in their geographic area, because someone else in the household is interfacing with service providers on their behalf, and/or because services are accessible through channels that do not require government-recognized proof of identity. In addition, people may have multiple identity documents, such as passports, driver’s licenses, voter ID cards, beneficiary IDs for a government program, or student or employee IDs that may fill their identification needs. To unpack why people do not have an ID and to better understand the demand- and supply-side factors related to obtaining one, survey respondents who previously answered that they did NOT have the national ID or equivalent were asked whether any of the following four reasons explained why they did not have this document (with multiple answers possible per person): ƒƒ It is too difficult to apply ƒƒ You don’t have the necessary documents ƒƒ You have another form of identification ƒƒ You don’t need an ID for any purpose 20 For a more in-depth discussion of different types of barriers, see the ID4D Practitioner’s Guide, available at http://id4d.worldbank.org/ guide. 14 ID 4 D.WO R L D BA N K .ORG The first two options are more indicative of supply-side constraints to obtaining an ID, whereas the latter two relate more to demand. These four options, of course, are not exhaustive of the possible reasons why someone may not have an ID: 29 percent of respondents without an ID did not choose any of the four possible responses. However, they provide some initial insights that complement the above findings and serve as a useful starting point for deeper investigation. Reported reasons for not having an ID As shown in Figure 12, supply-side barriers are slightly more commonly cited than demand-side factors in countries with ID coverage of less than 80 percent. However, the precise challenges that people selected vary by country. In Gabon, for example—where difficulties with ID card printing and the scarcity of access points have been well documented21—nearly 70 percent of people without an ID indicated that this was because the process to apply was too difficult. Not having supporting documents—which might include a birth certificate, certificate of nationality, proof of address, ID cards of the applicant’s parents, etc.—is also a common barrier to obtaining an ID in many countries. In Lesotho, where under-5 birth registration is less than 50 percent and an even higher share of adults have no birth certificates, lack of necessary documents was cited by over half of those without an ID (UNICEF 2017). Figure 12. Reasons for not having an ID cited by people without one 21 See, for example, http://www.gaboneco.com/carte-nationale-d-identite-une-piece-d-etat-civil-devenue-rarissime.html and http://www. gabonactu.com/gouvernement-envisage-detablir-cartes-nationales-didentite-cedoc/. GLOBAL ID COVERAGE, BARRIERS, AND USE BY TH E NUM B E R S: 15 AN IN-DE PT H LOOK AT T HE 2017 ID4 D-FIND EX S URV EY Although supply-side barriers appear to be more common than demand-side factors in low-coverage countries, this is not always the case. In Afghanistan, for example, 70 percent of adults without an ID—the vast majority of whom are women—responded that they do not have one because they have no need for it, as did 40 percent of those in Ethiopia. In a few countries with low ID coverage, we also see high proportions of adults without an ID reporting the possession of other identity documents, including in Gabon (56 percent), Mozambique (55 percent), and Benin (49 percent). Without more detailed information, however, we cannot accurately assess how well these alternative identity documents empower their holders to access public services, open a financial account, obtain a SIM card, or access other services and rights. The scale of these challenges becomes evident when we look at the prevalence of demand- and supply- side ID barriers across the total adult population. In South Sudan, for example, an estimated 78 percent of adults do not have an ID, and 55 percent of those without an ID give the reason that it was too difficult to apply (see Figure 12). Together, this implies that approximately 43 percent of all adults in South Sudan do not have an ID due fully or in part to the arduous application process. Similarly, more than 1 in 4 adults in Togo and Niger do not have the necessary documents to apply for the ID. Here, we also see that although demand-side factors are still less common, they affect significant portions of the adult population. In Lao PDR, for example, multiplying the ID coverage gap (56.6 percent) by the response rate for not needing an ID, we find an estimated 27 percent of all adults in the country see no need for the ID. Explaining coverage gaps for certain groups The ID4D-Findex data also allow us to analyze whether certain groups are more or less likely to cite different reason(s) for not having an ID. All else equal, the regression results summarized in Table 1 (and shown in full in Annex 3, Table 8) indicate that women are less likely to report finding the application process too difficult, compared with men. While this could mean that men do in fact have a harder time meeting the application requirements, it could also be related to the fact that women are also more likely than men to cite not needing an ID as their reason for not having one. If women are simply making fewer attempts to apply for an ID, they may not have as much first-hand experience with difficult application processes. For example, approximately 23 percent of Ethiopian women do not have an ID because they do not see a need for it, compared with only 8 percent of Ethiopian men.22 As discussed in the previous section, women’s reduced demand for ID could partially be a function of the fact that—in male-headed households—men may be more likely than women to perform many of the formal transactions that require an ID (e.g., opening a bank account, registering property, obtaining a SIM card, etc.). This estimate is based on multiplying the percent of women and men without an ID in Ethiopia (49 for women and 30 for men) by the 22  percent of those without an ID that reported “no need” as a reason for not having one (47 percent for women and 26 percent for men). 16 ID 4 D.WO R L D BA N K .ORG Table 1. Likelihood of reporting different reasons for not having an ID in LICs by demographic group Too difficult No documents Other ID No Need One of Only One of Only One of Only One of Only many reason many reason many reason many reason reasons given reasons given reasons given reasons given Female – – – + + Low edu. + + + + – – Age – – + + Rural + – + Key: + = more likely to cite reason (positive correlation), – = less likely to cite reason (negative correlation), based on logit models for LICs with country-level fixed effects, controlling for age-squared, workforce status, household income, and marital status. Empty cells have results that are not statistically significant at the 90% level or above; see Annex 3, Table 8 for full regression tables. We also see that adults with primary education or less find it harder to apply and face greater difficulty providing documentation of their identity than those who have attained secondary education or higher. Less educated adults are also less likely to report not having an ID because they have other forms of identification. Adults living in rural areas are more likely than urban dwellers to cite the difficulty of the application process or not needing an ID as their only reason for not having an ID. These findings are consistent with the fact that people in rural areas often have to travel longer distances to apply for an ID and—due to less frequent contact with the State or other service providers—may have less need for an ID overall. Furthermore, it may demonstrate one reason why we see lower ID coverage among rural women: they are more likely than either urban women or rural men to say they do not need an ID. Consistent with the above findings that younger adults are less likely to have IDs, we also see that as a person’s age increases, they are less likely to attribute their lack of identification to not having documents, and more likely to say it is because they have another ID. Notably, we do not see a statistically significant relationship between a person’s household income and the likelihood that they identified certain barriers over others. This is counter-intuitive, because—as discussed above—the direct and indirect costs of obtaining an ID can be quite high and are likely to place a disproportionate burden on the poorest people. However, although we presume that for most people, financial issues with obtaining an ID would fall under the “too difficult to apply” response category, the lack of a separate category on costs may be adding noise to these results. GLOBAL ID COVERAGE, BARRIERS, AND USE BY TH E NUM B E R S: 17 AN IN-DE PT H LOOK AT T HE 2017 ID4 D-FIND EX S URV EY How do IDs facilitate access to services? From a development perspective, one central reason why access to government-recognized identity credentials matters, is because exercising one’s rights and accessing basic services and economic opportunities often require official proof of identity. In many countries, a government-issued ID is required to prove one’s identity and/or eligibility for government services and safety nets, enroll in school, apply for a driver’s license, to obtain a building permit, or to register a business or obtain a job in the formal sector. In turn, unique and verifiable IDs can facilitate more effective delivery of services and payments, helping to minimize fraud and leakages and improve targeting. Beyond basic services, having reliable proof of identity is also essential for exercising basic rights like voting, or obtaining passports and visas needed to travel. Similarly, having an ID can be critical for unlocking access to private-sector services. In the financial sector, providing reliable proof of identity is typically required for opening an account due to Know-Your-Customer (KYC) requirements and Customer Due Diligence (CDD) regulations intended to prevent money laundering and terrorist financing. According to the 2017 Global Financial Inclusion and Consumer Protection (FICP) survey based on responses from financial sector authorities in 124 jurisdictions, 90 percent of commercial banks require a government-issued ID to open a deposit account.23 In addition, similar KYC requirements are often enforced for buying a mobile phone or SIM card or opening a mobile money account. This is increasingly the case as more and more countries—including 79 out of the 97 of the ID4D-Findex countries surveyed and 150 countries globally—have introduced mandatory SIM registration. At the same time, having an ID does not automatically translate into access to services, and not having an ID does not necessarily mean that people will always be excluded from services. Whether a specific ID is required for a service—and whether this requirement is enforced—varies significantly by context. Furthermore, for an ID to enable access to services like government benefits or banking, such services must actually be available to people in their country and/or geographic area. And even if services that require an ID are locally available, having a government-recognized ID may be a necessary but not sufficient condition for access if the application or enrollment process requires multiple forms of identification, proof of address or employment, and is associated with high fees or other indirect costs. The ID4D-Findex data provide the first opportunity to analyze the relationship between having an ID and access to services cross-nationally. In addition to the ID-specific questions discussed above, the Findex survey includes a wide variety of indicators on whether people use specific financial services, have mobile phones, and receive different types of government transfers. Furthermore, it includes a series of questions that asked respondents who reported having an ID whether they had ever used it for one of the following purposes (with multiple answers possible):24 ƒƒ To apply for a government service ƒƒ To receive financial support from the government ƒƒ To use financial services ƒƒ To apply for a SIM card/mobile phone service 23 http://www.worldbank.org/en/topic/financialinclusion/brief/ficpsurvey. 24 While these options capture many of the main uses of ID, they are not exhaustive. Overall, 23% of respondents did not choose any of the four responses. 18 ID 4 D.WO R L D BA N K .ORG Knowing when and how IDs facilitate access to services and how people use them in their daily lives can help us better understand the barriers faced by those without one and identify opportunities to strengthen inclusive service provision. Financial and mobile services The ID4D-Findex data show that people with IDs use financial and mobile services more frequently than people without IDs. This trend is evident in Figure 13, which shows a positive correlation between countries’ adult ID coverage and rates of owning a financial account and mobile phone. In particular, we see a strong relationship between countries that have high ID coverage and those with high mobile phone ownership. The relationship between ID access and mobile access can be mutually reinforcing: the desire to access mobile services can serve as an incentive for obtaining an ID, while having an ID can unlock access to SIM cards and mobile money accounts, creating new opportunities. Countries with high ID coverage may also attract more mobile service providers, since it will be easier for providers to onboard customers if they already have the identity documents needed to satisfy SIM registration and/or KYC requirements. Figure 13. Country-level ID coverage and ownership of a financial account and mobile phone At the individual level, we also see that adults with IDs are more likely to use financial and mobile services than those without an ID. In LICs, an estimated 65 percent of adults with IDs have a mobile phone, compared with only 40 percent of those without an ID. Similarly, 33 percent of adults with an ID have an account at a financial institution or a mobile money account, compared with only 19 percent of those without and ID. GLOBAL ID COVERAGE, BARRIERS, AND USE BY TH E NUM B E R S: 19 AN IN-DE PT H LOOK AT T HE 2017 ID4 D-FIND EX S URV EY Figure 14. ID ownership and use of private-sector services While we cannot establish a causal relationship with the Findex data—i.e., that having and ID directly led to account or mobile ownership—the regressions summarized in Annex 3 (Table 9) show that the positive relationship between having an ID and using financial and mobile services persists after controlling for gender, age, location, education level, income, workforce status, marital status, and the country in which a person is living—all factors that are likely to be correlated both with having an ID and having access to services. All else equal, an adult in an LIC with an ID is 2.6 times more likely to own a mobile phone, 2.9 times more likely to have an account at a financial institution, 2.1 times more likely to have a mobile money account, and 2.8 times more likely to own a debit card than adults without an ID. In addition, adults who have a debit card and an ID are 5.2 times more likely to have the card in their own name than debit card owners without an ID. Given the global gender gap in access to financial services, these findings supports the case that access to an ID is important for women’s economic empowerment. In some countries, we have also seen a clear link between extending the coverage of IDs with access to financial services. In India, for example, the share of adults with an account has more than doubled in the last six years, from 25 percent in 2011 to 80 percent in 2017, with women and the poor seeing some of the biggest gains.25 An important factor driving this increase was a government policy launched in 2014 to boost account ownership among unbanked adults through providing all residents with a unique, verifiable identity through the Aadhaar program and linking account opening to this process. When asked directly whether they had used their IDs for specific purposes, respondents confirmed frequent use of these credentials to access financial services and mobile phones. Approximately 57 percent of adults with ID had used it to apply for a SIM card or mobile phone service, and 45 percent had used it to access financial services. As shown in Figure 15, reported use patterns differ slightly by region, though in all except for South Asia, reported use for mobile services is significantly higher than for financial services. The higher reported use of IDs for mobile services fits with the fact that mobile services are more available in most developing countries, and around 80 percent of the surveyed population owns a mobile phone. Furthermore, SIM card registration is mandatory in all LICs included in the ID4D-Findex sample (GSMA 2019b). 25 2011 and 2017 Findex data. 20 ID 4 D.WO R L D BA N K .ORG Figure 15. Self-reported use of ID for financial and mobile services When we look at the country level in Table 2, we also see that the ten economies with the highest share of adults who report using their IDs for financial services are mostly higher income economies, including many in Europe, where account ownership is also above 80 percent. The 10 economies with the highest share of people reporting using their IDs to access mobile phone is more diverse, spanning multiple regions and levels of economic development. SIM registration is mandatory in each of these countries, and mobile phone ownership is relatively high. In addition, mobile money accounts are relatively widespread in several of these countries, such as Kenya (75 percent), Gabon (45 percent), and Botswana (24 percent). Table 2. Surveyed economies with highest reported use of ID for financial and mobile services Economy Used ID for Account Economy Used ID for Mobile Phone financial services ownership SIM/mobile ownership Israel 81% 94% Kenya 92% 88% Latvia 78% 93% Botswana 84% 84% Greece 77% 85% Turkmenistan 81% 75% Belarus 76% 81% Singapore 81% 93% Poland 74% 90% Gabon 78% 82% Croatia 72% 92% Algeria 77% 86% Kenya 71% 84% Tunisia 76% 72% Estonia 71% 98% Belarus 75% 89% Mongolia 69% 94% Poland 74% 88% India 64% 80% Togo 74% 64% GLOBAL ID COVERAGE, BARRIERS, AND USE BY TH E NUM B E R S: 21 AN IN-DE PT H LOOK AT T HE 2017 ID4 D-FIND EX S URV EY As with the likelihood of owning an ID, the ID4D-Findex data allow us to examine correlations between how people report using their IDs and other demographic factors. All else equal, we find that women, low- income and low-educated adults, younger adults, and those living in rural areas are less likely to report using their ID for either SIM/mobile access or financial services, compared with men, higher-income and more educated adults, older adults, and those living in urban areas (see Table 3 and full regression results in Annex 3, Table 11). At the same time, we see that lower-income and lower-educated adults are more likely to reply that obtaining a SIM or mobile service was the only use for the ID among the options presented to them, compared with higher-income and more educated adults. Similarly, women and low-educated adults were more likely to only select the financial services response than were men or more educated adults. Table 3. Likelihood of reporting use of IDs for mobile and financial services by demographic group Used ID for SIM/mobile Used ID for financial services One of many uses Only use given One of many uses Only use given – – – + Female Bottom 40% inc. – + – Low education – + + – Age + + Rural – – – Key: + = more likely to report use (positive correlation), – = less likely to report use (negative correlation), based on logit models with country-level fixed effects that control for the variables listed above as well as whether the respondent has a mobile phone or financial account, age-squared, workforce status, and marital status. Empty cells have results that are not statistically significant at the 90% level or above; see Annex 3 for full regression tables. When we look at the reasons people give for not having a financial account, we find further evidence that having an ID is associated with reduced barriers to financial inclusion. As shown in Figure 16, only 17 percent of unbanked adults in LICs with an ID cited “lack of documents” as a reason for not having an account; compared with 33 percent of those without an ID. At the same time, it is clear that not having an ID is only one barrier to accessing financial services. First, an ID may only be part of the documentation required to open an account; in many economies, financial institutions also require proof of address, proof of employment, or proof of income. Second, other factors—such as a lack of money, being too far away from a financial service provider and finding the account opening process too expensive—are cited as or more frequently than missing documentation as core barriers to account opening. Therefore, although having proof of identity eliminates a hard constraint to access, many other barriers remain and must be addressed holistically to ensure full financial inclusion. 22 ID 4 D.WO R L D BA N K .ORG Figure 16. ID ownership and barriers to account opening in LICs Just as not having an ID is only one obstacle to achieving to inclusive development, the ID4D-Findex data also show that having a national ID is not always a necessary condition for access to services. The requirements for proving one’s identity vary by country and the service being provided, and a specific ID may be necessary in some cases but not in others. For example, Figure 17 shows mobile phone and financial account ownership rates in LICs by whether or not the person has the foundational ID referenced in the survey (“has ID”). Here, we see that 45 percent of adults with the ID have a financial account, and 65 percent of those with the ID have a mobile phone. However, we also see that while access to these services is lower for people without the foundational ID, it is not zero. Of the adults without the foundational ID living in LICs who reported “owning another ID” as a reason for not having it (“has other ID”), 28 percent have access to an account, and 47 percent have access to a mobile phone. Among adults without the ID who did not report having another form of ID (“no ID”), account access is estimated to be 15 percent, while mobile phone access is estimated to be 37 percent. While this measure is imperfect—it is likely that some adults in the “no ID” category do have other forms of identification but simply did not list this as a reason for not having the ID—it indicates that some people are able to access these services through other channels that do not require government-recognized forms of identification. In many cases, for example, women may have (or have access to) mobile phones or financial accounts that are not in their own name. Similarly, certain basic accounts or those provided through non-regulated entities may require less stringent CDD during account opening. In other cases, it may be that requirements for showing government-recognized proof of identity (e.g., for SIM registration) are unevenly enforced. GLOBAL ID COVERAGE, BARRIERS, AND USE BY TH E NUM B E R S: 23 AN IN-DE PT H LOOK AT T HE 2017 ID4 D-FIND EX S URV EY Figure 17. Accessing services with and without a foundational ID in LICs Public sector benefits and services The ID4D-Findex data show that people with an ID are more likely to receive public sector benefits than those without it, but to a lesser degree than private sector services. For example, Figure 18 shows that adults with IDs are slightly more likely to access public-sector benefits than those without IDs, including government transfers, pensions, and agricultural payments. In LICs, for example, an estimated 36 percent of adults without an ID received agricultural payments, compared to 40 percent of those with an ID. Of this 36 percent, only one quarter reported having another form of identification as a reason for not having the ID, suggesting that in many countries these payments may not be contingent on having government- recognized identity credentials. Thus, while having an ID may make it easier to access government benefits and services, it is not always a requirement. Figure 18. ID ownership and use of government-to-person (G2P) transfers 24 ID 4 D.WO R L D BA N K .ORG The relationship between having an ID and receiving some government transfers persists after controlling for other factors that are likely to influence both receiving these benefits and ID ownership, although it is smaller than for financial and mobile services as discussed above. All else equal, adults living in LICs with an ID are 1.4 times more likely to have received a government transfer and 1.5 times more likely to have received a government pension in the past 12 months than those without an ID (see Annex 3, Table 10 for full regression results). When we look at people’s self-reported use of IDs, we find that approximately 33 percent of adults in LICs have used it to receive financial support from the government, and 16 percent have used it to apply for government services. These rates are higher in LMICs, where an estimated 39 percent of adults report using the ID for financial support and 27 percent report using it to access services. While these usage rates are lower than for financial and mobile services, they still indicate that around a third of adults with IDs are using them for state-provided programs. When we consider a regional and gender breakdown, we also find some significant differences. In East and South Asia, for example, men are more likely to report using their ID to access government financial support and services, compared to women. Conversely, women in Europe and Central Asia and Latin America are more likely to report using the ID for government services, as equally as likely to report using it for government financial support. Looking at reported use at the country level also makes it clear that these global and regional averages mask significant variation in the use of ID for access to government benefits. Figure 19. Self-reported use of ID for government benefits and services by gender The economies with the highest share of adults who report using their IDs for government services and support tend to be ones where the public sector is engaged in extensive service provision (see Table 4). Eastern European and Central Asian countries stand out in this regard. For example, an estimated 89 percent of adults with an ID in Belarus, 79 percent in Kazakhstan, and about 70 percent in Estonia, Russian Federation, Turkmenistan, and Ukraine cite using their ID to access government services. Similarly, the share of adults who report using their IDs to receive government financial support is 53 percent in Latvia, 42 percent in Ukraine, and 40 percent in Belarus. These findings make sense given that the coverage of social protection programs is extensive in former Soviet republics in particular, and a relatively high share of adults in these countries also report receiving government transfers in the past 12 months.26 For instance, 44 percent of adults in Ukraine received a government transfer within the last 12 months, as did 41 percent in Latvia and 25 percent in Belarus. It is important to note that these three indicators—(1) whether the ID was used for government financial support (at any point in the past), 26  (2) whether the person received government transfers in the past year, and the (3) national coverage of social protection programs—are conceptually related but not equivalent. GLOBAL ID COVERAGE, BARRIERS, AND USE BY TH E NUM B E R S: 25 AN IN-DE PT H LOOK AT T HE 2017 ID4 D-FIND EX S URV EY Table 4. Surveyed economies with highest reported use of ID to access government services and support Economy Used ID for Economy Used ID for Received Coverage of government govt financial govt transfer social protection services1 support1 in last year1 programs2 Belarus 89% Latvia 53% 41% 80% Israel 80% Thailand 48% 35% 79% Kazakhstan 79% Ukraine 42% 44% 70% Morocco 76% Belarus 40% 25% 76% Turkmenistan 73% Namibia 38% 18% 16% Estonia 73% India 37% 8% 94% Russian Fed. 72% Kazakhstan 37% 16% 50% Greece 69% Israel 36% 32% - Ukraine 68% Moldova 34% 15% 57% Moldova 67% South Africa 34% 26% 80% Sources: 1) 2017 ID4D-Findex survey and, 2) World Bank ASPIRE database (https://data.worldbank.org/indicator/per_allsp. cov_pop_tot?view=chart). On the African continent, a handful of countries also have a high share of ID holders who report using these to access government services, including an estimated 76 percent in Morocco, 61 percent in Rwanda, and 60 percent in Namibia. In South Africa, where the government has a substantial social protection regime estimated to cover 80 percent of the population, some 34 percent of ID holders report using their ID to receive government financial support. In Namibia, where safety nets are smaller in scope, around 38 percent of adults with an ID still report using it to access government support. We also see people frequently using their IDs to receive financial benefits in a handful of South and East Asian countries that invest heavily in social protection, including Thailand (48 percent) and India (37 percent). Table 5. Likelihood of reporting use of IDs for government benefits by demographic group Used ID for govt services Used ID for govt financial support One of many uses Only use given One of many uses Only use given – + + – Female Bottom 40% inc. – + + Low education – + – + Age + + + Rural – + + + Key: + = more likely to report use (positive correlation), – = less likely to report use (negative correlation), based on logit models with country-level fixed effects that control for the variables listed above as well as whether the respondent reported receiving a government transfer, age-squared, workforce status, and marital status. Empty cells have results that are not statistically significant at the 90% level of above; see Annex 3 for full regression tables. 26 ID 4 D.WO R L D BA N K .ORG When we look at the demographic factors that predict people’s use of IDs for public benefits, we see similar relationships as those described for private sector use. All else equal, women, low-income and low-educated adults, younger adults, and those living in rural areas are less likely to report using their ID to access government services, compared with men, higher-income and more educated adults, older adults, and those living in urban areas (see Table 5, and full regression results in Annex 3, Table 12). Women and lower educated adults were also less likely than men and higher educated adults to report using the ID to access government financial support; while older adults and rural dwellers were more likely to cite this use, compared with younger adults and urban dwellers. At the same time, we see that women, lower-income and lower-educated adults, and those living in rural areas are more likely to state that obtaining government services or government financial support were the only purpose for which they used their ID among the options presented to them.27This would align with the fact that in many cases, poor people and rural dwellers are the beneficiaries of targeted financial support from the government, and may have obtained IDs specifically to access these benefits, while higher income individuals and urban dwellers may be accessing a larger variety of services that require an ID. Furthermore, although men in most low- and middle-income economies are more likely than women to use their ID to receive financial support overall, women are significantly more likely to report using their ID for this purpose in a number of wealthier economies and former Soviet republics such as Ukraine, Belarus and Kazakhstan (see Figure 20). This reflects the fact that in many of these economies—e.g., South Africa— women are often the beneficiaries of social protection programs such as child support grants. Figure 20. Surveyed economies with highest reported use of ID to access government financial support 27 Some of these results are only significant when looking at the full sample of countries, as opposed to LICs only. See Annex 3, Table 13 for full results. GLOBAL ID COVERAGE, BARRIERS, AND USE BY TH E NUM B E R S: 27 AN IN-DE PT H LOOK AT T HE 2017 ID4 D-FIND EX S URV EY Evidence from several economies, such as South Africa and Pakistan, suggests that government social protection initiatives can spur ID ownership among the targeted population if having an ID is set as a condition for accessing these programs (World Bank 2019, Cheema et al. 2015). Having an ID can then open new opportunities for people, with often particularly large benefits to members of vulnerable and marginalized groups, while reliable and high-coverage digital ID systems can help facilitate the more effective delivery of government transfers. Ensuring that all people have secure, easily verifiable identity credentials that can facilitate access to bank or mobile money accounts is also becoming more critical as government payments are moving from cash to digital channels. Among adults in Thailand who have an account, about 14 percent opened their first account to receive government transfers; in Argentina 11 percent did so. In both economies, governments and payment recipients could leverage a robust and widely accessible ID system with universal (>99%) coverage. The survey data also point to potential opportunities for financial inclusion through greater digitization of government payments: globally, 130 million unbanked adults who already have an ID still receive cash payments for government transfers and public pensions. In many ways, the relationship between ID coverage and government-to-person (G2P) transfers is likely to be mutually reinforcing. Without high levels of ID coverage, it may be difficult for countries to roll out transfer programs, as they may not have a reliable way to ensure that these benefits reach their intended recipients. At the same time, rolling out a new transfer program may actually be one of the primary drivers for improving ID coverage. How non-existent or low-coverage ID systems limit opportunities Not having trusted and widely accepted proof of identity can inhibit participation in social, political and economic life and can be a barrier to exiting poverty. This may be the case for people who live in a country where a particular ID is widely required but do not have one , as well as people living in countries that do not have foundational ID systems designed to provide proof of identity for a wide variety of purposes. How severely people will be impacted by the lack of IDs will depend on each country’s overall identification landscape (e.g. are there widely accessible and widely accepted alternatives to a specific foundational ID credential?) and policies related to documentation requirements for accessing services (e.g. to what extent are government-recognized IDs needed to access services?). In five economies without a foundational ID system or with very limited coverage—Central African Republic (CAR), Liberia, Nepal, the Philippines, and Tanzania28—respondents were asked whether the lack of necessary identity documents inhibited them from: ƒƒ applying for a government service ƒƒ accessing government financial support ƒƒ accessing financial services ƒƒ applying for a SIM card 28 As of 2017, no national ID or similar foundational ID system existed in Nepal, Liberia, or the Philippines (although projects to implement ID systems are currently underway in Philippines and Liberia). Tanzania has a national ID system, but coverage of the ID card was very low (~10 percent) in 2017. CAR previously had a limited coverage national ID system but the issuance of cards was halted during the civil conflict in 2013 and has not resumed. 28 ID 4 D.WO R L D BA N K .ORG As shown in Figure 21, not having an ID negatively affects particularly large segments of the adult population in CAR and Liberia across multiple areas of life. In CAR, nearly 30 percent of adults reported not being able to access a government service, government financial support, or financial services because they did not have the necessary identity documents, and 34 percent reported difficulties applying for a SIM card. In Liberia, about one in five adults report not being able to access basic services due to a lack of identification. In both countries, we see no evidence of significant variation in lack of access by gender. Figure 21. Share of people unable to access services because they did not have the necessary identity documents In Tanzania and in the Philippines, lack of ID limits access to services for fewer people: about 15 percent of adults report being unable to access a government service and unable to use financial services due to not having the required IDs. In each case, while foundational ID systems are non-existent or low coverage, there are a number of functional ID systems (e.g., for voting, tax, and social programs) that people frequently use to access services. In Nepal—where there is no foundational ID but citizenship certificates function as a de facto national ID—not having an ID seems to be a barrier for less than 10 percent of potential service users, and is most commonly cited as an obstacle for accessing mobile services. In Tanzania, men are slightly more likely than women to report that they could not access a mobile service or SIM (20 percent for men vs. 15 percent for women) or government financial support (13 percent for men vs. 8 percent for women) due to a lack of identity documents; we see no statistically significant gender differences in Nepal. GLOBAL ID COVERAGE, BARRIERS, AND USE BY TH E NUM B E R S: 29 AN IN-DE PT H LOOK AT T HE 2017 ID4 D-FIND EX S URV EY Conclusion The ID4D-Findex data show conclusively that the identification challenge is most concentrated in lower income countries and marginalized groups. In LICs, over one-third of adults does not have a national ID or similar foundational identity document. Within LICs in particular, whether a person has an ID or not is highly influenced by their demographic characteristics, including gender, income, education, and geographic location. Furthermore, we see that access to IDs is intersectional: lower-income, lower-educated women in rural areas are the least likely to have an ID, and thus are at the greatest risk of being left behind. These gaps are the result of a complex set of reasons, including both demand- and supply-side factors. Overall, about one in three people without an ID in LICs find it “too difficult” to obtain one and almost as many cite not being able to produce the necessary supporting documents. Adults with an ID are more likely to access public- and private- sector services, but how people use their IDs varies by context. Overall, having an ID is positively associated with owning a financial account and mobile phone, and with receiving government transfers, even after controlling for other factors. In many developing countries, where government services and cash transfer programs reach only a limited segment of the population, ID use is greater for private sector services. Of the four options included in the survey, applying for a SIM card/mobile phone service is the most frequently cited reason for using one’s ID, followed by accessing financial services. Overall, women, low- income and educated adults, and those living in rural areas are less likely to use their ID for accessing public and private sector services, pointing to broader inequities in participation in economic life. In countries without a foundational ID system, lack of identification can be a significant barrier for people accessing basic services. However, having an ID does not remove all the barriers to inclusive development, and lacking an ID does not always prevent access to services. Although adults with an ID are more likely to access services—and those without them are more likely to experience difficulties—obtaining an ID does not always translate into inclusion. For example, while people with an ID are less likely to say they do not have an account because they lack the necessary documents, some 17 percent still report having insufficient documentation. Furthermore, many people—particularly in developing countries—who do not have their country’s foundational ID are still able to open accounts at financial institutions, obtain mobile phones and SIM cards, and receive government benefits such as agricultural payments. Given that ID may improve access to services in some contexts—but is not always required or sufficient— it is therefore essential that governments and service providers to balance the benefits of improving identification against the risk of exclusion when a (specific type) of government-issued ID is required to access a service or program. Where a specific ID is made mandatory for a service or benefit, this may exclude people who were able to access it previously, particularly if those people are also in groups (e.g., low-income and low-educated women) that are less likely to have ID. Countries therefore need to simultaneously strengthen and improve the accessibility of ID systems and look for opportunities to reduce documentation and identification requirements where appropriate. These results also point to a number of areas for future research. Ultimately, ensuring universal access to identification is only the first step; in order for ID systems to be catalytic for individual welfare and development, they must be trusted, empowering, and applied appropriately. Additional work is therefore needed to better measure how—and when—people use their IDs, the barriers they do (and do not) face if they lack a particular ID, and the impact of making an ID mandatory for services that people used to access informally. In addition, more detailed quantitative and qualitative work is needed to better understand how the quality and type of various ID systems—e.g., digital, biometric, mobile, etc.—affect the accessibility and convenience of different services and benefits. 30 ID 4 D.WO R L D BA N K .ORG References Abraham, Ronald, Elizabeth S. Bennett, Rajesh Bhusal, Shreya Dubey, Qian (Sindy) Li, Akash Pattanayak, and Neil Buddy Shah. 2018. State of Aadhaar Report 2017-18. IDinsight. https://static1.squarespace. com/static/5b7cc54eec4eb7d25f7af2be/t/5bbd2874c8302561862f03d4/1539123330295/ State+of+Aadhaar+Report_2017-18.pdf. Apland, Kara, Carolyn Hamilton, Brad K. Blitz, Mary Lagaay, Dr. Rajith Lakshman, and Elizabeth Yarrow. 2014. Birth registration and children’s rights: a complex story. https://www.planusa.org/docs/ reports/2014-birth-registration-research-full-report.pdf. Brito, Steve, Ana Corbacho, and Rene Osorio. 2017. Does birth under-registration reduce childhood immunization? Evidence from the Dominican Republic. Health Economics Review, 7 (1):14. doi:10.1186/s13561-017-0149-3. Caribou Digital. 2017. Identities: New practices in a connected age. Farnham, Surrey, United Kingdom: Caribou Digital Publishing. https://www.identitiesproject.com. Cheema, Iftikhar, Maham Farhat, Simon Hunt, Sarah Javeed, Katharina Keck, and Sean O’Leary. 2015. “Benazir Income Support Programme: Second Impact Evaluation Report”. Oxford, UK: Oxford Policy Management. https://www.opml.co.uk/files/Publications/7328-evaluating-pakistans-flagship-so- cial-protection-programme-bisp/bisp-second-impact-evaluation-report.pdf?noredirect=1. Corbacho, Ana, Steve Brito, and Rene Osorio Rivas. 2012. Birth Registration and the Impact on Educational Attainment. Washington, DC: Inter-American Development Bank. Dahan, Mariana, and Alan Gelb. 2015. “The Role of Identification in the Post-2015 Development Agenda.” Working Paper. Washington, DC: World Bank. FWLD (Forum for Women, Law, and Development). 2015. Acquisition of Citizenship Certificate in Nepal. FWLD Publication 169. Kathmandu: FWLD. http://fwld.org/wp-content/uploads/2016/06/ Acquisition-of-Citizenship-Certificate-in-Nepal-Estimation-and-Projection.pdf. Gelb, Alan, and Anna Diofasi Metz. 2018. Identification Revolution: Can Digital ID Be Harnessed for Development? Washington, DC: Center for Global Development. Gelb, Alan, Anit Mukherjee, and Kyle Navis. 2018. Digital Governance in Developing Countries: Beneficiary Experience and Perceptions of System Reform in Rajasthan, India. Washing- ton, DC: Center for Global Development. https://www.cgdev.org/publication/digital-governance- developing-countries-beneficiary-experience-and-perceptions-system. GSMA. 2019a. Digital identity opportunities for women: Insights from Nigeria, Bangladesh, and Rwanda. https://www.gsma.com/mobilefordevelopment/wp-content/uploads/2019/05/Digital-identity- opportunities-for-women-Insights-from-Nigeria-Bangladesh-and-Rwanda-Web.pdf. GSMA. 2019b. Access to Mobile Services and Proof of Identity 2019: Assessing the impact on digital and financial inclusion. https://www.gsma.com/mobilefordevelopment/wp-content/uploads/2019/02/ ProofofIdentity2019_WebSpreads.pdf. GLOBAL ID COVERAGE, BARRIERS, AND USE BY TH E NUM B E R S: 31 AN IN-DE PT H LOOK AT T HE 2017 ID4 D-FIND EX S URV EY IFES (International Foundation for Electoral Systems). 2013. Survey Assessing Barriers to Women Obtaining Computerized National Identity Cards (CNICs). Washington, DC: IFES. http://aceproject.org/ electoral-advice/archive/questions/replies/277728362/962062828/IFES-PK-Survey-Assessing- Barriers-to-Women.pdf. NRC (Norwegian Refugee Council). 2016. Securing Status: Syrian refugees and the documentation of legal status, identity, and family relationships in Jordan. https://reliefweb.int/sites/reliefweb.int/files/ resources/securing-status%5B1%5D.pdf. Oppenheim, Ben and Brenda Marea Powell. 2015. Legal Identity in the 2030 Agenda for Sustainable Development: Lessons from Kibera, Kenya. New York, NY: Open Society Justice Initiative. https:// www.justiceinitiative.org/uploads/0a6472de-a975-4a3b-b3ad-2b979891d645/legal-identity- 2030-agenda-lessons-kibera-kenya-2051216.pdf. UNDP (United Nations Development Programme). 2017. “Over 9 million people register for National Identity Cards as Malawians make history”. UNDP Malawi website. December 5, 2017. http://www.mw.undp. org/content/malawi/en/home/presscenter/articles/2017/12/05/over-9-million-people-register-for- national-identity-cards-as-malawians-make-history-.html. UNICEF. 2017. Global databases: Birth registration. https://data.unicef.org/topic/child-protection/birth- registration/. World Bank. 2017a Principles on identification for Sustainable Development: Toward the Digital Age. Washington, DC: World Bank Group. World Bank. 2017b. The State of identification systems in Africa – a synthesis of country assessments. Washington, D.C. : World Bank Group. http://documents.worldbank.org/curated/en/156111493234231522/The- State-of-identification-systems-in-Africa-a-synthesis-of-country-assessments. World Bank. 2018. Women, Business, and the Law. Washington, DC: World Bank Group. https:// openknowledge.worldbank.org/bitstream/handle/10986/29498/9781464812521.pdf. World Bank. 2019. South Africa ID Case Study. Washington, D.C.: World Bank Group. http://documents. worldbank.org/curated/en/315081558706143827/South-Africa-ID-Case-Study. 32 ID 4 D.WO R L D BA N K .ORG GLOBAL ID COVERAGE, BARRIERS, AND USE BY TH E NUM B E R S: 33 AN IN-DE PT H LOOK AT T HE 2017 ID4 D-FIND EX S URV EY Annex 1. ID coverage by country and gender ID coverage for adults at or above ID coverage for all the mandatory/required age adults (no age cutoff) Country Sample Size All Women Men Gender Gap Sample Size All Afghanistan 1000 71.4% 48.4% 94.2% 45.8%*** 1000 71.4% Albania 985 92.8% 92.1% 93.5% 1.4% 1000 90.1% Algeria 1016 95.1% 94.5% 95.7% 1.1% 1016 95.1% Argentina 99.3% 999 99.4% 99.2% -0.2% 1000 99.2% Azerbaijan 987 99.5% 99.5% 99.5% 0% 1000 98.8% Bangladesh 947 88.7% 87.8% 89.5% 1.7% 1000 82.9% Belarus 1053 98.9% 99.2% 98.6% -0.7%* 1053 98.9% Benin 907 48.4% 38.0% 59.1% 21.1%*** 1000 46.0% Bolivia 915 97.8% 97.6% 98.0% 0.4% 1000 97.7% Bosnia and Herzegovina 972 98.0% 98.4% 97.6% -0.8% 1000 94.1% Botswana 985 96.8% 97.2% 96.4% -0.8% 992 96.0% Brazil 943 99.0% 99.4% 98.6% -0.8% 1000 98.4% Bulgaria 1000 99.3% 100.0% 98.5% -1.5%*** 1000 99.3% Burkina Faso 989 80.8% 74.9% 86.2% 11.3%*** 1000 80.7% Cambodia 1600 88.6% 88.9% 88.3% -0.5% 1600 88.6% Cameroon 992 76.2% 74.5% 77.9% 3.4% 1000 76.2% Chad 910 39.0% 21.0% 58.7% 37.7%*** 1000 37.3% Chile 1000 99.2% 99.2% 99.1% -0.1% 1040 98.8% China 3595 98.7% 98.5% 98.9% 0.4% 3627 98.5% Colombia 943 98.9% 99.2% 98.6% -0.6% 1000 91.6% Congo 921 63.7% 60.1% 67.4% 7.3%* 1000 59.3% Costa Rica 968 97.8% 97.4% 98.3% 1% 1000 97.6% Cote d’Ivoire 972 70.4% 71.0% 69.8% -1.2% 1000 68.1% Croatia 945 99.5% 99.4% 99.5% 0.1% 1000 99.4% Czech Republic 1000 98.5% 99.2% 97.8% -1.4%** 1000 98.5% Dominican Republic 979 89.9% 90.7% 89.2% -1.4% 1000 87.6% Ecuador 929 99.6% 99.6% 99.6% -0.1% 1000 99.5% Egypt 990 93.7% 90.2% 97.2% 7%*** 1000 92.2% El Salvador 954 96.3% 95.8% 96.9% 1% 1000 88.5% Estonia 1000 98.6% 98.2% 99.0% 0.9% 1000 98.6% Ethiopia 979 60.4% 50.8% 70.5% 19.7%*** 1000 59.0% Gabon 976 69.9% 67.2% 72.7% 5.4% 1000 68.4% Georgia 1000 94.1% 94.1% 94.1% 0% 1000 94.1% Greece 1000 92.8% 94.4% 91.1% -3.3%* 1000 92.8% Guatemala 904 93.0% 93.3% 92.7% -0.6% 1000 81.7% Guinea 994 44.6% 39.4% 49.6% 10.1%** 1000 44.6% Haiti 474 79.5% 83.0% 75.8% -7.2%* 504 72.5% Honduras 921 93.2% 91.6% 94.9% 3.3%* 1000 83.8% Hungary 999 99.7% 99.6% 99.8% 0.2%* 1000 99.7% GLOBAL ID COVERAGE, BARRIERS, AND USE BY TH E NUM B E R S: 35 AN IN-DE PT H LOOK AT T HE 2017 ID4 D-FIND EX S URV EY ID coverage for adults at or above ID coverage for all the mandatory/required age adults (no age cutoff) Country Sample Size All Women Men Gender Gap Sample Size All India 96.8% 3000 96.7% 96.8% 0.1% 3000 96.8% Indonesia 969 95.8% 95.7% 95.9% 0.2% 1000 90.4% Israel 995 99.1% 99.7% 98.5% -1.1%* 1000 97.8% Jordan 987 90.2% 90.9% 89.6% -1.3% 1012 87.7% Kazakhstan 990 93.8% 92.7% 95.0% 2.3%* 1000 91.2% Kenya 967 94.3% 93.0% 95.6% 2.6% 1000 90.9% Kosovo 981 92.7% 94.2% 91.2% -2.9% 1000 90.3% Kyrgyz Republic 983 93.2% 92.8% 93.6% 0.8% 1000 89.8% Lao PDR 946 43.4% 42.6% 44.3% 1.7% 1000 40.7% Latvia 1001 98.9% 98.8% 99.1% 0.3% 1002 98.9% Lebanon 1000 96.7% 96.8% 96.6% -0.2% 1000 96.7% Lesotho 976 70.9% 69.1% 72.7% 3.6% 1000 69.7% Lithuania 970 81.7% 81.0% 82.5% 1.5% 1000 81.5% North Macedonia 985 98.8% 98.4% 99.2% 0.8% 1008 94.0% Madagascar 938 85.4% 80.5% 90.9% 10.5%*** 1000 78.9% Malawi 956 16.0% 16.1% 16.0% -0.1% 1000 15.7% Malaysia 1004 94.2% 94.3% 94.1% -0.2% 1004 94.2% Mali 995 70.7% 63.1% 78.5% 15.5%*** 1000 70.7% Mauritania 966 89.6% 89.0% 90.2% 1.2% 1000 88.9% Mexico 960 90.1% 89.3% 91.0% 1.6% 1000 89.2% Moldova 985 98.4% 98.8% 98.0% -0.9% 1000 97.1% Mongolia 987 97.6% 98.3% 96.9% -1.3% 1000 95.7% Montenegro 971 98.5% 99.0% 98.0% -1% 1000 93.8% Morocco 4822 98.5% 98.3% 98.7% 0.4% 5110 92.7% Mozambique 986 58.2% 51.5% 65.3% 13.8%*** 1000 58.1% Myanmar 1600 88.8% 87.2% 90.6% 3.4%* 1600 88.8% Namibia 978 91.9% 93.4% 90.1% -3.3% 1000 90.6% Nicaragua 983 92.1% 91.2% 93.1% 1.9% 1000 89.8% Niger 854 48.4% 34.5% 61.6% 27.1%*** 1000 44.8% Pakistan 1491 86.5% 77.5% 95.1% 17.6%*** 1600 79.5% Panama 924 95.3% 96.9% 93.6% -3.3%* 1000 92.1% Paraguay 999 99.6% 99.2% 100.0% 0.8%*** 1000 99.6% Peru 1000 98.4% 97.9% 98.9% 1% 1000 98.4% Poland 956 98.8% 98.6% 99.0% 0.4% 1000 93.3% Romania 1001 99.0% 98.7% 99.4% 0.7%* 1001 99.0% Russian Federation 1974 98.0% 99.1% 96.6% -2.5%*** 2000 97.9% Rwanda 983 92.9% 91.5% 94.4% 2.9% 1000 90.6% Senegal 988 72.6% 72.1% 73.1% 1% 1000 72.4% Serbia 999 97.0% 98.0% 95.9% -2.1% 1000 96.7% Singapore 1000 94.7% 92.1% 97.5% 5.3%*** 1000 94.7% Slovak Republic 999 99.4% 99.5% 99.4% -0.1% 1000 99.4% South Africa 981 92.4% 91.0% 93.9% 2.9%* 1000 91.9% South Sudan 875 21.8% 11.5% 32.4% 20.9%*** 1000 21.4% 36 ID 4 D.WO R L D BA N K .ORG ID coverage for adults at or above ID coverage for all the mandatory/required age adults (no age cutoff) Country Sample Size All Women Men Gender Gap Sample Size All Sri Lanka 1093 93.5% 91.6% 95.8% 4.2%* 1104 92.1% Thailand 1000 99.5% 99.0% 100.0% 1%*** 1000 99.5% Togo 998 39.6% 31.6% 47.7% 16.1%*** 1000 39.8% Trinidad and Tobago 501 94.9% 93.9% 96.0% 2.1% 504 94.9% Tunisia 949 96.5% 95.2% 97.8% 2.6%** 1001 93.0% Turkey 1000 96.7% 95.7% 97.6% 1.9% 1000 96.7% Turkmenistan 983 98.4% 98.2% 98.6% 0.4% 1000 96.7% Uganda 980 83.5% 82.3% 84.9% 2.6% 1000 81.4% Ukraine 998 97.3% 98.4% 95.8% -2.7%** 1000 97.3% Uruguay 998 99.7% 100.0% 99.5% -0.5%*** 1000 99.7% Uzbekistan 992 95.3% 96.8% 93.6% -3.2%* 992 94.5% Venezuela 1000 98.3% 98.1% 98.6% 0.5% 1000 98.3% Vietnam 1000 94.1% 95.9% 92.1% -3.8%* 1002 94.1% Zambia 968 89.0% 87.2% 90.9% 3.8%* 1000 86.5% Zimbabwe 975 86.2% 84.9% 87.7% 2.8% 1000 84.1% Note: All analyses reported in this paper use a restricted sample that includes only those respondents at or above the age where obtaining an ID is either allowed or required (the first group of columns above). As shown here in the final two columns for illustration, using the entire sample provides slightly lower coverage estimates in most countries, which makes sense as we would expect fewer people below the ID cutoff age to have one. The Gender Gap column shows the percentage point difference between men and women and whether this is significant at the 99% (***), 95% (**), or 90%(*) confidence level based on design-based t-test. GLOBAL ID COVERAGE, BARRIERS, AND USE BY TH E NUM B E R S: 37 AN IN-DE PT H LOOK AT T HE 2017 ID4 D-FIND EX S URV EY Annex 2. ID module in the 2017 Global Findex survey FIN48. Do you personally have a [insert local terminology for national ID card]? CIRCLE ONE RESPONSE: ROUTE: Yes 1 (Continue) No 2 (Skip to FIN50) (DK) 3 (Refused) 4 FIN49. Have you ever used your [insert local terminology for national ID card] to do any of the following? (Randomly rotate FIN49A-FIN49D) YES NO (DK) (REFUSED) FIN49A. To apply for a government service 1 2 3 4 FIN49B. To receive financial support from the government 1 2 3 4 FIN49C. To use financial services 1 2 3 4 FIN49D. To apply for a SIM card/mobile phone service 1 2 3 4 38 ID 4 D.WO R L D BA N K .ORG FIN50. Please tell me whether each of the following is A REASON why you do not have a [insert local terminology for national ID card]? Is it because… (Display FIN50A first and then randomly rotate FIN50B-FIN50D)? YES NO (DK) (REFUSED) FIN50A. You have another form of identification 1 2 3 4 FIN50B. You don’t need an ID for any purpose 1 2 3 4 FIN50C. It is too difficult to apply 1 2 3 4 FIN50D. You don’t have the necessary documents 1 2 3 4 FIN48. [Asked only in countries with no national ID or equivalent] Have you ever NOT been able to do any of the following because you did NOT have the necessary identification documents? (Randomly rotate FIN48A-FIN48D) YES NO (DK) (REFUSED) FIN48A. Apply for a government service 1 2 3 4 FIN48B. Receive financial support from the government 1 2 3 4 FIN48C. Use financial services 1 2 3 4 FIN48D. Apply for a SIM card/mobile phone service 1 2 3 4 GLOBAL ID COVERAGE, BARRIERS, AND USE BY TH E NUM B E R S: 39 AN IN-DE PT H LOOK AT T HE 2017 ID4 D-FIND EX S URV EY Annex 3. Regression results Table 6. ID Coverage: Demographic predictors (regressions for Figure 4) Has an ID LICs only LICs w/FE LMICs w/FE LMICs w/FE All w/FE (1) (2) (3) (4) (5) Female 0.598*** 0.517*** 0.808 0.985 0.769*** (-0.04) (-0.04) (-0.09) (-0.16) (-0.05) Primary school education or less 0.352*** 0.364*** 0.432*** 0.468*** 0.438*** (-0.02) (-0.03) (-0.05) (-0.1) (-0.04) In workforce 1.699*** 1.773*** 1.366* 1.662* 1.505*** (-0.12) (-0.14) (-0.19) (-0.35) (-0.14) Bottom 40% of income distribution 0.714*** 0.644*** 0.886 0.753 0.780*** (-0.05) (-0.05) (-0.08) (-0.14) (-0.05) Age 1.107*** 1.119*** 1.239*** 1.105*** 1.175*** (-0.01) (-0.01) (-0.03) (-0.03) (-0.01) Age2 0.999*** 0.999*** 0.998*** 0.999*** 0.998*** (0) (0) (0) (0) (0) Rural 0.631*** 0.558*** 0.809 0.746 0.738*** (-0.05) (-0.05) (-0.09) (-0.15) (-0.06) Married 1.088 1.274** 1.474** 1.854** 1.527*** (-0.07) (-0.1) (-0.21) (-0.35) (-0.13) Constant 0.663* 0.978 0.129*** 1.709 21.315*** (-0.12) (-0.22) (-0.06) (-0.92) (-12.18) Country Fixed Effects N Y Y Y Y Income Group Fixed Effects N N N N Y N 16618 16618 37525 36537 102983 F Statistic 90.5 60.8 41.9 21.7 59.2 Prob > F 0 0 0 0 0 Note: * p<0.05, ** p<0.01, *** p<0.001 Odds ratios from logit models using survey weights and design-based standard errors. Excludes respondents below the required/ minimum age for obtaining the ID. Income groups based on World Bank’s 2017 classification. 40 ID 4 D.WO R L D BA N K .ORG Table 7. ID Coverage in LICs: Interaction models (regressions for Figure 7) Has an ID (1) (2) (3) (4) (5) (6) Female 0.833 0.576*** 0.524*** 0.884 0.708*** 0.516*** (-0.09) (-0.07) (-0.05) (-0.12) (-0.07) (-0.04) Primary school education or less 0.481*** 0.362*** 0.364*** 0.362*** 0.367*** 0.525*** (-0.05) (-0.03) (-0.03) (-0.03) (-0.03) (-0.07) In workforce 1.756*** 1.953*** 1.773*** 1.765*** 1.745*** 1.768*** (-0.14) (-0.25) (-0.14) (-0.14) (-0.14) (-0.14) Bottom 40% of income distribution 0.643*** 0.645*** 0.656*** 0.643*** 0.635*** 0.646*** (-0.05) (-0.05) (-0.08) (-0.05) (-0.05) (-0.05) Age 1.122*** 1.119*** 1.119*** 1.120*** 1.114*** 1.119*** (-0.01) (-0.01) (-0.01) (-0.01) (-0.01) (-0.01) Age2 0.999*** 0.999*** 0.999*** 0.999*** 0.999*** 0.999*** (0) (0) (0) (0) (0) (0) Rural 0.562*** 0.557*** 0.558*** 0.803 0.556*** 0.758* (-0.05) (-0.05) (-0.05) (-0.1) (-0.05) (-0.09) Married 1.264** 1.272** 1.273** 1.270** 1.802*** 1.277** (-0.1) (-0.1) (-0.1) (-0.1) (-0.22) (-0.1) Female x primary school or less -0.05 (0.553***) Female x workforce 0.859 (-0.13) Female x bottom income 0.656*** (-0.08) Female x rural 0.542*** (-0.08) Female x married 0.560*** (-0.07) Primary school or less x rural 0.632** (-0.09) Constant 0.781 0.904 0.97 0.715 0.939 0.77 (-0.18) (-0.21) (-0.22) (-0.17) (-0.21) (-0.18) Country Fixed Effects Y Y Y Y Y Y N 16618 16618 16618 16618 16618 16618 F Statistic 59.2 58.4 58.5 58.6 57.3 58.8 Prob > F 0 0 0 0 0 0 Note: *p<0.05, ** p<0.01, *** p<0.001 Odds ratios from logit models using survey weights and design-based standard errors. Excludes respondents below the required/ minimum age for obtaining the ID. Includes LICs only based on World Bank’s 2017 classification. GLOBAL ID COVERAGE, BARRIERS, AND USE BY TH E NUM B E R S: 41 AN IN-DE PT H LOOK AT T HE 2017 ID4 D-FIND EX S URV EY Table 8. ID Barriers in LICs: Demographic predictors Reason for not having an ID (multiple answers possible) Too difficult Too difficult only No docs No docs only Other ID Other ID only No need No need only (1) (2) (3) (4) (5) (6) (7) (8) Female 0.74** 0.72* 0.85 0.72 0.83+ 0.93 1.26+ 1.97** (-0.07) (-0.1) (-0.09) (-0.19) (-0.09) (-0.16) (-0.16) (-0.49) Primary school 1.39** 1.57** 1.48*** 1.57* 0.57*** 0.41*** 1.2 1.46 education or less (-0.16) (-0.25) (-0.18) (-0.35) (-0.07) (-0.07) (-0.17) (-0.43) In workforce 1.37** 0.96 1.24* 1.07 1.20+ 1.07 1.19 0.85 (-0.14) (-0.16) (-0.11) (-0.18) (-0.13) (-0.19) (-0.14) (-0.19) Bottom 40% of 1.03 1.09 0.92 0.74 0.99 1.03 0.99 0.92 income distribution (-0.1) (-0.15) (-0.1) (-0.17) (-0.1) (-0.18) (-0.11) (-0.16) Age 1 1 0.96*** 0.92** 1.04*** 1.07*** 1 1.05 (-0.01) (-0.02) (-0.01) (-0.02) (-0.01) (-0.02) (-0.01) (-0.03) Age2 1 1 1.00** 1.00** 1.00** 1.00** 1 1 (0) (0) (0) (0) (0) (0) (0) (0) Rural 0.96 1.43* 1.03 0.83 0.78+ 0.82 1.1 1.60* (-0.12) (-0.23) (-0.15) (-0.24) (-0.11) (-0.18) (-0.16) (-0.37) Married 1.09 0.96 1.1 1.14 1.06 1.12 1.07 0.85 (-0.12) (-0.16) (-0.13) (-0.36) (-0.12) (-0.17) (-0.13) (-0.19) Constant 0.86 0.04*** 0.42** 0.09** 0.05*** 0.01*** 1.28 0.03*** (-0.27) (-0.02) (-0.14) (-0.07) (-0.02) (-0.01) (-0.39) (-0.02) Y Country Fixed Effects Y Y Y Y Y Y Y N 5841 5841 5841 5841 5841 5841 5841 5841 F Statistic 8.4 4.1 7.4 4.2 14 7.3 10.9 11.3 Prob > F 0 0 0 0 0 0 0 0 Note: + p<0.10, * p<0.05, ** p<0.01, *** p<0.001 Odds ratios from logit models using survey weights and design-based standard errors. Excludes respondents below the required/ minimum age for obtaining the ID. Includes LICs only based on World Bank’s 2017 classification. For columns 1, 3, 5, and 7 the dependent variable is whether or not the respondent gave the following answers as one of their responses: ‘It is too difficult to apply’ (1), ‘Does not have the necessary documents (3)’, ‘Has another form of identification’ (5), and/or ‘Does not need and ID for any purpose’ (7). For columns 2, 4, 6, and 8, the dependent variable is whether the respondent gave the above as their only answer. 42 ID 4 D.WO R L D BA N K .ORG Table 9. ID Use in LICs: Correlations between having an ID and access to private-sector services Has … Mobile phone Account (FI) Account (mobile) Debit card Debit card Credit card in own name (1) (2) (3) (4) (5) (6) Has ID 2.55*** 2.85*** 2.07*** 2.83*** 5.26*** 1.82*** (-0.19) (-0.3) (-0.17) (-0.39) (-1.72) (-0.3) Constant 5.11*** 0.05*** 0.01*** 0.02*** 0.57 0.01*** (-1.17) (-0.01) (0) (-0.01) (-0.44) (0) Country FE Y Y Y Y Y Y Controls Y Y Y Y Y Y N 16618 16618 15748 16618 1924 16618 F Statistic 53.5 44.4 59.6 33.4 4.8 11.1 Prob > F 0 0 0 0 0 0 Note: + p<0.10, * p<0.05, ** p<0.01, *** p<0.001 Odds ratios from logit models using survey weights and design-based standard errors. Excludes respondents below the required/ minimum age for obtaining the ID. Includes LICs only based on World Bank’s 2017 classification. Controls include dummies for gender, education level, income level, workforce status, age and age-squared, rural location, and marital status. Table 10. ID Use in LICs: Correlations between having an ID and access to public-sector services In the past 12 months, has received … Government transfer Government pension Agriculture payment (1) (2) (3) Has ID 1.42* 1.54* 0.93 (-0.21) (-0.26) (-0.07) Constant 0.01*** 0.00*** 0.04*** (0) (0) (-0.01) Country Fixed Effects Y Y Y Controls Y Y Y N 16618 16618 16618 F Statistic 6.8 7.8 28.4 Prob > F 0 0 0 Note: + p<0.10, * p<0.05, ** p<0.01, *** p<0.001 Odds ratios from logit models using survey weights and design-based standard errors. Excludes respondents below the required/ minimum age for obtaining the ID. Includes LICs only based on World Bank’s 2017 classification. Controls include dummies for gender, education level, income level, workforce status, age and age-squared, rural location, and marital status. GLOBAL ID COVERAGE, BARRIERS, AND USE BY TH E NUM B E R S: 43 AN IN-DE PT H LOOK AT T HE 2017 ID4 D-FIND EX S URV EY Table 11. ID Use: Correlations with reported use of ID for financial and mobile services Used ID for … (multiple answers possible) SIM/mobile SIM/mobile ONLY Financial services Financial services ONLY (1) (2) (3) 4) (5) (6) (7) (8) SIM registration mandatory 1.96*** 4.48*** (-0.3) (-0.17) Owns mobile phone 5.10*** 3.80*** 2.38*** 2.17*** (-0.43) (-0.18) (-0.25) (-0.21) Owns account 10.16*** 10.02*** 4.67*** 4.43*** (-1.05) (-0.54) (-1.18) (-0.46) Female 0.55*** 0.56*** 0.87* 0.83*** 0.85+ 0.91* 1.38 1.46*** (-0.04) (-0.02) (-0.06) (-0.04) (-0.08) (-0.04) (-0.28) (-0.11) Primary school edu. or less 0.58*** 0.77*** 1.46*** 1.44*** 0.61*** 0.65*** 1.64* 1.29** (-0.05) (-0.03) (-0.12) (-0.09) (-0.05) (-0.04) (-0.32) (-0.12) In workforce 1.26** 1.37*** 0.82* 0.87* 1.24+ 1.41*** 1.34 0.88 (-0.11) (-0.07) (-0.07) (-0.05) (-0.14) (-0.07) (-0.28) -0.09) Bottom 40% of income dis. 0.72*** 0.70*** 1.15+ 1.14* 0.89 0.67*** 0.87 1.01 (-0.06) (-0.03) (-0.08) (-0.07) (-0.08) (-0.04) (-0.2) (-0.09) Age 1.02 1.04*** 0.99 1.00 1.04** 1.04*** 1.01 1.00 (-0.01) (-0.01) (-0.01 (-0.01) (-0.02) (-0.01) (-0.03) (-0.01) Age2 1.00* 1.00*** 1.00 1.00** 1.00** 1.00*** 1.00 1.00 (0) (0) (0) (0) (0) (0) (0) (0) Rural 0.61*** 0.83** 0.84* 0.92 0.80* 0.86* 0.94 1.17 (-0.07) (-0.05) (-0.08) (-0.06) (-0.07) (-0.06) (-0.18) (-0.13) Married 0.99 1.19*** 0.88 0.99 1.16 1.05 1.35 1.19* (-0.08) (-0.06) (-0.07) (-0.06) (-0.11) (-0.05) (-0.28) (-0.1) Constant 1.41 0.24*** 0.22*** 0.05*** 0.15*** 0.16*** 0.01*** 0.01*** (-0.51) (-0.04) (-0.07) (-0.01) (-0.05) (-0.03) (0) (0) Sample LICs only All LICs only All LICs only All LICs only All Country Fixed Effects Y N Y N Y Y Y Y Income Group Fixed Effects N Y N Y N N N N N 10777 92537 10777 92537 10777 92537 10777 92537 F Statistic 38.7 152.3 16.4 64.6 34.9 44.6 5.6 16.3 Prob > F 0 0 0 0 0 0 0 0 Note: + p<0.10, * p<0.05, ** p<0.01, *** p<0.001 Odds ratios from logit models using survey weights and design-based standard errors. Excludes respondents below the required/ minimum age for obtaining the ID. Includes LICs only based on World Bank’s 2017 classification. For columns 1, 2, 5 and 6, the dependent variable is whether or not the respondent gave the following answers as one of their responses: ‘Used ID for SIM/mobile services’ (1 and 2), ‘Used ID for financial services (5 and 6)’. For columns 3, 4, 7 and 8, the dependent variable is whether the respondent gave the above as their only answer. 44 ID 4 D.WO R L D BA N K .ORG Table 12. ID Use: Correlations with reported use of ID for government services and support Used ID for ... (multiple answers possible) Gov’t services Gov’t services ONLY Gov’t financial support Gov’t financial support ONLY (1) (2) (3) (4) (5) (6) (7) (8) Received gov’t transfer w/in 12 mo. 2.11*** 2.37*** 0.46+ 0.56*** 8.35*** 7.51*** 5.36*** 2.93*** (-0.29) (-0.14) (-0.2) (-0.08) (-1.11) (-0.54) (-1.9) (-0.33) Female 0.82* 0.79*** 1.74** 1.42*** 1.02 0.88* 1.73 1.34** (-0.07) (-0.03) (-0.3) (-0.11) (-0.13) (-0.04) (-0.62) (-0.15) Primary school education or less 0.51*** 0.59*** 1.07 1.50*** 0.72** 0.87* 1.54 1.75*** (-0.04) (-0.03) (-0.2) (-0.12) (-0.08) (-0.05) (-0.58) (-0.23) In workforce 1.45*** 1.31*** 0.77 0.77** 1.24 1.17** 0.46* 0.81+ (-0.15) (-0.07) (-0.16) (-0.07) (-0.17) (-0.06) (-0.17) (-0.1) Bottom 40% of income distribution 0.88+ 0.75*** 1.44* 1.20* 0.9 1.04 0.96 1.61*** (-0.07) (-0.03) (-0.25) (-0.09) (-0.09) (-0.06) (-0.32) (-0.17) Age 1.04** 1.04*** 1.04+ 0.98+ 1.04* 1.04*** 1.07+ 1.02 (-0.01) (-0.01) (-0.02) (-0.01) (-0.02) (-0.01) (-0.04) (-0.02) Age2 1.00** 1.00*** 1.00 1.00* 1.00+ 1.00*** 1.00 1.00 (0) (0) (0) (0) (0) (0) (0) (0) Rural 1.00 0.85* 2.50*** 1.32* 1.04 1.28** 1.56 1.57** (-0.09) (-0.05) (-0.6) (-0.15) (-0.14) (-0.1) (-0.87) (-0.25) Married 1.04 1.03 0.98 1.03 1.13 0.90+ 1.87+ 0.69** (-0.09) (-0.05) (-0.17) (-0.1) (-0.13) (-0.05) (-0.65) (-0.09) Constant 0.49* 0.56** 0.01*** 0.03*** 0.08*** 0.06*** 0.00*** 0.00*** (-0.15) (-0.12) (0) (-0.01) (-0.03) (-0.01) (0) (0) Sample LICs only All LICs only All LICs only All LICs only All Country Fixed Effects Y Y Y Y Y Y Y Y N 10777 92537 10777 92537 10777 92537 10392 92152 F Statistic 24.7 60.2 7.2 24 15.6 31.9 9.2 10.3 Prob > F 0 0 0 0 0 0 0 0 Note: + p<0.10, * p<0.05, ** p<0.01, *** p<0.001 Odds ratios from logit models using survey weights and design-based standard errors. Excludes respondents below the required/ minimum age for obtaining the ID. Includes LICs only based on World Bank’s 2017 classification. For columns 1, 2, 5 and 6, the dependent variable is whether or not the respondent gave the following answers as one of their responses: ‘Used ID to receive government services’ (1 and 2), ‘Used ID to receive government financial support (5 and 6)’. For columns 3, 4, 7 and 8, the dependent variable is whether the respondent gave the above as their only answer. GLOBAL ID COVERAGE, BARRIERS, AND USE BY TH E NUM B E R S: 45 AN IN-DE PT H LOOK AT T HE 2017 ID4 D-FIND EX S URV EY id4d.worldbank.org