CGAP recently commissioned our design firm, Continuum, to use applied product innovation techniques to find ways to help government-to-person (G2P) beneficiaries in Pakistan. We conducted research with beneficiaries to better understand the constraints on linking G2P payments to financial inclusion. The insights from this research provide a nuanced understanding of the difficulties of providing financial services to the very poorest in a society, specifically the barriers that come from extreme illiteracy. This research is applicable not only to other G2P payment providers and government ministries, but also to any financial provider that seeks to design a product that meets the needs and the capabilities of low-income clients.
Literacy is often a hidden hurdle to bringing financial inclusion to the unbanked. Systems that should work in theory break down when poor people are unable to learn how to use them or are unable to assume the accountability of consumers who know their rights and how to obtain recourse to maintain transparency and honesty in the system. If you cannot read or understand your receipt, how can you be confident that you have received the correct amount of money?
Literacy is such a basic part of how we communicate, interact, and think in the developed world and among the educated in developing countries that, at the start of our work, we did not fully grasp the barriers posed to extremely illiterate BISP recipients. The standard for literacy in Pakistan is to be able to write your own name. Unfortunately, most women in Pakistan are unable to do even this. Our fieldwork revealed that the women in our study could not read in their own spoken language, and many of them could not read anything at all. Many of them do not understand symbols, icons, illustrations, and instructions.
Photo Credit: Continuum Innovation
Because systems should be designed to work for the user, the onus is on the designer of the system to make sure users can access the system. Needless to say, challenge of communicating with BISP recipients is extreme.
Being illiterate is not limited to not being able to read. Research has shown that learning to read helps to develop a person’s ability to use language in general and to deal with abstraction. To determine the dimensions of illiteracy among BISP recipients, we developed some simple tests to find out what they could and could not do.
Writing and reading Urdu script. The most obvious manifestation of illiteracy is not being able to read or write. None of the women we spoke with could read or write Urdu script. Few could read or write their own name. Ironically, when BISP recipients get their BISP debit card it comes with written instructions for how to use it. None of the women we spoke with could understand any of the instructions.
Numbers. Most, but not all, BISP recipients could read English numbers (e.g., 1, 2, 3, 4, etc.) and knew what they represented. But being able to read a number does not mean that you can interact correctly with systems using numbers.
Direction. In English we write from left to write, in Urdu, from right to left. People who speak Urdu, Punjabi, or Saraiki, even if they cannot write, have learned to look at things from right to left. This is a problem when English numbers encroach into a mostly Punjabi world. We frequently saw women enter in their PIN code beginning from the right. Using an ATM is especially difficult because when the user punches in her PIN code, only asterisks appear on the screen so she has no feedback that what she is inputting does not match what is on her PIN card.
Currency. Every BISP recipient could identify the different notes in her currency. The denominations are written in the English number system, 20, 50, 100, 500, 1000, etc., so that reinforces their comprehension of numbers. But it does not mean they can recognize numbers in other contexts.
Parsing. To understand a receipt you have to pick out which characters are relevant and recognize what they represent. We found that the women could not recognize an amount on a receipt if it was close to other extraneous text or numbers of the same size.
Two obvious solutions to the literacy problem are to use icons to simplify communication or to use verbal instructions. We tested these ideas with BISP recipients and learned that what was simple to us, did not work for them.
Icons. Some have proposed using icons as a way to communication to illiterate people. The BISP instructions that came with the card included diagrams and icons that were intended to help the overwhelming majority of women who could not read the text. But icons are essentially shorthand for something that is already understood. Like acronyms, they make things simpler for those in the know, but are incomprehensible to the uninitiated. For example, to us the icon taken from an instruction sheet looked like a fanfold of five 500 Rupee notes, but to BISP recipients who were not used to icons it looked like a poorly drawn hand.
Verbal instructions. Another obvious option is to provide verbal communication. However, in testing instructions for how to use an ATM, BISP recipients struggled with making the translation from verbal instruction to manual action. Abstractions such as “top, right” or “the square” were confusing. Many literate people are used to dealing with abstractions and translating from one domain to another, but illiterate people are not practiced in this skill.
So, having determined what does not work, we looked for a way to communicate without words, diagrams or icons, or verbal instructions. The answer was simple: we used photographs. Photographs are literal ways to communicate—there is no abstraction. We found that BISP women were confident and eager to use an ATM after they were shown a series of photographs showing each step of the process. An example of photographic instruction is shown below.
Photo Credit: Continuum Innovation
The photographs should show exactly what the person has to do on exactly the same device that they will use. We prepared a short design guide to help banks in Pakistan and BISP to communicate with their customers on how to use the system, increase transparency, and improve trust.
BISP recipients play an important role in reducing corruption. But to do so, they need to understand what is going on and be able monitor the checks and balances in the system. By making the numbers on receipts large and with no other distracting characters or symbols around them, for example, it is easier for BISP recipients to monitor their payments and more difficult for them to be victims of fraud.
Photo Credit: Continuum Innovation
If banks and the government use the right approach, BISP recipients will learn how to use ATM and agent systems, make sure that they get the payments owed them, and protect themselves and the system from fraud. In time, as the BISP recipients gain confidence in the system, they will also gain the confidence to begin to use other simple financial products and take further steps toward financial inclusion. Designing systems to match the capabilities of the people they are meant to serve is essential to growing financial inclusion.
For our detailed recommendations about design elements for a financial system used by people without the ability to read, please see our design guide.
Harry West and Rachel Lehrer were part of the Continuum Innovation research team in Pakistan.
Very interesting project Rachel and Harry. Sometimes we design systems that are obvious for us but don't work in the target market and then wonder what went wrong.
It reminds me of the story of Karim Kohja, CEO of Roshan (Afghanistan) when he had 250 angry police officers asking for his payroll. Since they where illiterate, they couldn't realize their salaries where already paid.
I have several questions:
1. Do BISP recipients always do the same transaction? If not, how do they deal with the complexities of knowing which set of instructions to use?
2. The solution implies that there should be hardware standardization (or different set of instructions for each type of hardware) that could probably limit scalability. Is this a challenge in Pakistan?
3. How are the banks planning to educate recipients. Are they able to use available resources or will they need to create an education field team?
your questions are very valid. when you say same transaction, please do note that they receive the same amount but they may withdraw different denominations. so the eventual challenge may not be just that they have received the money, but how much of a balance do they have in their account (something an unbanked person may not note is written in the receipt and might be maintaining on physical ledgers or diaries) and your observation therefore is also correct, they may learn to withdraw 1000 rupees but if they need more, they'd eventually be stumped if not instructed. if you look at the ATM menus across various banks in pakistan, they are standardized within the boundaries of the bank itself but not across: so if you go to another banks atm you will see a differently designed screen. although the overall flow is the same but the prompts and layouts change - something that can be significantly detrimental for the target audience. and yes, this is a real challenge in pakistan where people are even being defrauded using these electronic channels which doesnt really help in increasing their trust in the system. for branchless and e banking to really reach the uneducated masses, there needs to be a mass scale overhaul of ADC channels keeping the unbanked in mind (like an accessibility feature in computers for the disabled, an option to switch from the conventional to a very scaled down interactive voice/graphics rich interface) whether banks/the govt is planning on educating the people is a hard question to answer becuase it seems there is more activity on launching these initiatives and lesser still on the follow through. it comes down to very relevant researches like these and steps taken by the private sector (participating banks, atm switch owners, branchless banking partners) to try to make these things easier for the customer. or, perhaps the state bank, as ultimately all of this results in the ultimate outcome of financial inclusion of the population.
Fascinating research on the challenges of communication across language and literacy boundaries. One can imagine that the lessons learned could have wider application, perhaps through the creation of instructional photo books or pamphlets. I wonder how effective a hybrid of photos with graphics would be, and whether that approach might begin to teach people who do not interpret graphics the skills to do so?
I think you mean to say Arabic numbers (or Hindu-Arabic), not English.