If you were to hear that 100% of adults in the Netherlands are literate and only 20% of adults in Chad are, would you be surprised? Would you assume that the numbers are based on adults being asked to read something to prove that they were literate? Seven hundred and seventy-three million individuals in the world reportedly cannot read. A majority of them live in just 10 countries and 2/3rds of them are women. Where do these numbers come from?
In my work at Curious Learning, I have written these figures and used them in presentations more times than I can remember. As demoralizing as these numbers are, it is even more disturbing that they might be a vast underestimation or, even worse, completely meaningless. What if dramatically more people are illiterate? If one assumes that most people who cannot read are in remote areas, unorganized settlements, or are refugees, how do countries count the number of people who cannot read? In this essay, I will clarify how literacy is measured, why this is important, and what could help solve the problems associated with determining literacy rates.
Literacy rates in international databases (like here and here) refer to the percentage of people aged 15 and above who can read and write a short sentence about their life. Most high-income countries report that 95-100% of adults can perform that task. For example, the US and Germany report that 99% of all adults can read at this level. Furthermore, many middle-income countries have literacy rates between 60-80% and many low-income countries report that 20-40% of adults can read. (For more information on historical growth in literacy rates I recommend this blog post.)
There is some interesting history behind this standard. The first literacy surveys were conducted in the 1600s in England to determine how many people in a district could sign their names on a document. In the early 20th century, more and more countries became curious about how many people in an area could do more than sign their name. Adults were asked to read and write their names as well as a short sentence about where they live and work. However, as more people around the world attended school and learned to read, the task of counting them became difficult. Therefore, surveys were developed to hopefully accomplish the same goals and reach more people. Current literacy rates are based on information gathered in the following ways:
Survey of the head of household
Survey of Individuals
Inference from School Completion Rates
Assessment
To clarify, the first two methods rely on asking either a head of household or an individual if they and their household members can read and write a short sentence about their life and home. Countries that report literacy rates based on inference assume that if a child has completed a particular grade in school they can read. Countries that report literacy rates based on literacy assessments derive these numbers from international assessments like the Program for the International Assessment of Adult Competencies (PIAAC) or the Literacy Assessment and Monitoring Programme (LAMP). The most recent data from 168 countries reported that 52 gathered literacy rates by means of a survey of the head of household, 81 with a survey of individuals, 5 using inferences from school completion rates, and 27 from literacy tests.
There are a number of problems with these methods of reporting literacy rates. Surveys are problematic because of how subjective the answers can be. If you are surveying the head of the household or an individual, how can you rely on the answers you receive? Why would someone answer no to the question, ‘can you or anyone in your household read and write’? Why would they say yes? In some areas of the world it could be dangerous for a husband to admit that his wife or daughters can read. On the other hand, some families might feel embarrassed to admit that no one in their household can read. In some low-income countries, women cannot be the head of the household, so single mothers with children are not even asked if they can read, which means they are not counted.
There are also problems associated with inferring literacy rates from school completion records. In some countries, it is a reliable means of reporting a basic level of literacy. Singapore, for instance, requires students to pass sets of exams before moving on to another grade or graduating so it is safe to assume that if a child completes secondary school they can read. However, in Subsaharan Africa 68% of children move on to secondary school even though only 12% of grade 4 children can read for meaning. Finishing primary school does not mean a child can read. Even in countries where a higher literacy rate might be assumed, there is no consistency in the level of completion that signifies that the child is ‘literate’. In Greece, you are assumed to be literate if you complete grade 6, but Paraguay defines a person as literate if they complete grade 2.
Deriving literacy rates from literacy assessments seems ideal. Two large scale assessments have been developed to determine literacy rates. The Program for the International Assessment of Adult Competencies has been administered to adults in 30, mostly wealthy, countries. The Literacy Assessment and Monitoring Program has been field tested in four countries. This development is very encouraging because an assessment represents the most accurate means of checking reading abilities.
However, administering these assessments is expensive, demands experienced personnel and has not historically included refugees or people who live in remote locations. The most recent data is from 2014 and most of the 28 countries that use either of the assessments submitted results from the early 2000s. If the literacy numbers are correct, these countries also represent a collection of countries sorely in need of investment in education. For example, Malawi (66%) and Sierra Leone (48%) each report comparatively low literacy rates, but since those numbers are from 2015 and 2013 respectively, they do not reflect the needs of children currently in school or the impact of programmatic interventions since then.
Which points to a general problem in all forms of reporting on literacy rates in that the data used in international counts is incredibly outdated. The most recent counts are from 2015-2016, but the oldest is a survey of heads of households collected by Poland in 1978. A nontrivial number of countries have not updated their literacy numbers in 20 years. Of course, the counts do not include the effect of school closings due to COVID as well as changes in education spending or increases in school attendance. For example, India has not reported literacy data since 2011, but the Ministry of Education has reported that the number of out of school primary aged children was cut in half between 2006 and 2014. If some of those children were in Grade 1 in 2014 they would now be in Grade 8 and in just a few more years would be considered an adult by literacy standards. Do we know how these children have fared? Has school access resulted in learning? Alas, we do not know.
In light of how unreliable the data is, why does the world even compile data on literacy rates? Why do we even care how many people can or cannot read? The fact that substantial effort is expended to gather any kind of literacy data reveals the importance the world places on learning to read and write. The United Nations recognizes that learning to read is a basic human right and that literacy is the foundation of economic growth and development. Because of this, the numbers are linked to development dollars, education spending, and international support for poverty reduction programs. Literacy is also considered a necessary base for social growth. More practically, we know that when an individual can’t read they have little to no economic freedom, they earn less money over a lifetime, are less likely to understand how diseases like COVID and monkeypox are spread, and are therefore less likely to protect themselves. Non-literate adults are more likely to be radicalized by violent groups or become a victim of them. They are less likely to understand or participate in efforts to combat climate change, but are more likely to face the extremely dire consequences. The success of every program or intervention aimed at eliminating poverty depends on people learning to read and write. These numbers have the potential to indicate how likely it is that a particular country or region will experience economic growth or radicalism, violence, and poverty.
Literacy skills also reflect the ability of a society to generate wealth. Being able to read and write a short sentence does not reflect the level of literacy needed to participate in the modern economy, which means that the standard itself used to evaluate literacy is problematic and is a questionable goal. An interesting way of thinking about this relationship is in terms of supply and demand. Literacy has become a skill required for the jobs that are currently available, as well as those available in the future. Therefore, the demand for literacy is high and growing. Educating children, by definition, means preparing them for an economy at least 20 years into the future. Literacy supply refers to the capacities of the present workers and the skills of the next generation. A dramatic mismatch between demand and supply is a key factor in high unemployment and economic stagnation. Furthermore, a key contributor to economic growth is investment in the education of adults (present workers) and children (next generation of workers).
Without reliable data, it is impossible to understand the scale of the problem. A low standard of literacy in high-income countries masks inequities among different groups. For example, a 2012 study comparing UNESCO literacy numbers to an assessment of adult literacy skills found that both Italy and Japan report almost 100% literacy rates. However, Italy achieved a lower average score on an assessment of adults than Japan. In the United States, 97% of adults reportedly can read and write a short sentence about their life, however, studies of adult literacy have found that at least 19% of adults only read at a Grade 4 level or below. There are many studies across low and middle-income countries where the reported literacy rates do not reflect actual literacy capacities. As reported in a Global Education Monitoring report from 2021 “Almost half of all 20- to 24-year-olds who had completed lower secondary school in 18 low- and lower-middle-income countries could not read a simple sentence.”
The relationship between investments in education and economic growth would seem to be obvious. But, studies are unclear about what investments make the biggest difference. Should a country with limited means prioritize early education or secondary school? Is it better to train teachers or build schools? To better understand the relationship between the development of human capital and economic growth it is necessary to have a reliable measure of literacy learning, one of many desired outcomes of education investments. Studies have shown that using a measure of schooling makes the relationship to economic growth unstable, but when the same relationship is analyzed using a measure of learning instead the relationship is strong. Just attending school does not mean that a country will receive economic benefits from investments in education. It is only when learning is measured that a strong correlation with economic growth is revealed.
There you have it, in a nutshell. The only way to reliably measure literacy and to have a number that can be used to elucidate the success of education investments is to measure actual learning. How can learning be measured reliably, frequently, and inexpensively? The proliferation of mobile devices offers a unique opportunity to do that at a fraction of the cost of traditional literacy assessments while maintaining reliability. Using apps that are designed to be engaging and game-like it is possible to repeatedly and reliably assess literacy skills and learning. There are various efforts in development to achieve a reliable, tech-based measure of learning (See here and here and here). At Curious Learning, we are developing games that are open-source and designed to be adapted to the languages needed by the most vulnerable populations. While implementation is limited to those with smartphones, those numbers are rising exponentially. Technology and good data offer the best hope for making substantive progress in education and truly eliminating illiteracy everywhere.
Written by Stephanie Gottwald, Co-Founder and VP of Content and Research, with help from Tinsley Jane Galyean.