Q&A  | 

Digital tools to mitigate poverty with Anna Schrimpf, with the Nobel Award winning J-PAL

"M-PESA helped lift 194,000 mostly female-headed Kenyan households out of extreme poverty.”

Tags: 'Abdul Latif Jameel Poverty Action Lab' 'AI' 'Ana Schrimpf' 'ayuda al desarrollo' 'J-PAL' 'Pobreza'


Reading Time: 5 minutes

Anna Schrimpf is the Executive Director of J-Pal Europe, the European branch of the Abdul Latif Jameel Poverty Action Lab whose co-founder and co-director Esther Duflo received the Nobel Prize of Economics in 2019.
J-PAL is a global research center anchored by a network of over 190 affiliated professors working to reduce poverty by ensuring that policy is informed by scientific evidence. “The research in our network includes a large number of evaluations of interventions that seek to combat poverty by introducing technological innovations, across a variety of sectors”, explains Schrimpf.

The use of satellite imagery and a novel computer algorithm to evaluate the effectiveness of a payments for ecosystem services (PES) program in curbing deforestation is one of the very innovative approaches of J-Pal to reduce poverty.

How do you work with both governments and individuals to drive progress towards development goals?

It is vital for a functioning society to have a strong link between its citizens and its government. At J-PAL, through our Innovation in Government Initiative, we work with governments to support them in their use of evidence in policy making and to improve the effectiveness of their social programs, particularly for essential services, such as education, health, and social assistance.

International aid can be an effective tool to reduce poverty and complement governmental action, but for funds to be efficiently allocated, we must first learn which programs work, which do not, and why. Evidence from randomized evaluations helps us understand that, and can change the way we address problems related to poverty.

One example is support packages for the very poor. One approach, called the Ultra-Poor Graduation developed by the NGO BRAC, is a multifaceted livelihood program that provides low-income households with a productive asset, training, regular coaching, access to savings, and consumption support.

Are new approaches necessary in policymaking? What are some of the projects that J-PAL Europe is working on right now?

At J-PAL, we believe that innovating and testing with promising new approaches is key to finding solutions to the world’s complex poverty issues. To assist governments, international organizations, NGOs, and private actors with their innovation processes, we work with them to rigorously test and improve the effectiveness of social programs, broadly share evidence on what programs work best and why, and build their capacity to use evidence. 

For example, in Europe we have recently launched an initiative on social inclusion to understand what programs and policies can help foster the social inclusion of migrants and refugees across the continent.

Through this initiative, we have come across a number of innovative approaches to inclusion, from working with professional athletes and artists of migrant origin to prevent social exclusion in schools in Finland, to buddy matching programs between locals and migrants in Sweden, and to encouraging native and refugee students to think from one another’s perspectives to facilitate integration in Turkey.

Partnering with European governments and civil society to conduct rigorous evaluations in these topics will help us answer urgent policy questions and generate insights applicable to other questions. 

Can you provide some examples of the use of technological solutions to reduce poverty?

Turning to financial inclusion, mobile money and digital financial services hold the potential to help reduce poverty. Mobile money makes it easier and faster for households to send and receive money, which can improve households’ welfare.

For example, evidence shows that the growth of the mobile money service M-PESA in Kenya helped lift 194,000 mostly female-headed households out of extreme poverty. Mobile money can also give women more control over the use of funds. Building on this existing research, J-PAL recently launched a Inclusive Financial Innovation Initiative to better understand how digital finance can be used to accelerate financial inclusion and promote development in Indonesia and beyond.

Further, digital ID systems may make it easier to implement and target government welfare programs. Evidence shows that biometric IDs have increased the efficiency of welfare payments  by simplifying the system and reducing leakages. In addition, when linked with monitoring systems, digital IDs have reduced absenteeism among health workers, reduced data forgery, and improved patient adherence to treatment. However, while digital ID systems show promise, there is limited research in this area, a gap that J-PAL is looking to fill through our Digital Identification and Finance Research Initiative (DigiFI Africa) initiative.

In the field of education technology, evidence from one review summarizing over 120 studies suggests that educational software—or computer assisted learning—designed to help students develop particular skills at their own rate of progress has shown enormous promise in improving academic achievement, particularly in math.

In one intervention in India, a technology-based program called Mindspark, which uses games, videos, and activities and can provide personalized instruction adapted to the level of every student, improved students’ Hindi and math test scores.

Another promising approach is the use of technology-based nudges to encourage or remind students or their parents to do certain actions. Nudges have proven effective at facilitating school-parent communication and at constantly improving students’ educational outcomes, often at very low costs.

How can AI can contribute to poverty reduction?

On the topic of climate change, Seema Jayachdaran, a J-PAL affiliated researcher, and co-authors utilized satellite imagery and a novel computer algorithm to evaluate the effectiveness of a payments for ecosystem services (PES) program in curbing deforestation. In response to the COVID-19 pandemic, J-PAL affiliated researchers Joshua Blumenstock, Dean Karlan, and Christopher Udry are using mobile phone and satellite data to assist the Government of Togo to better target its emergency cash transfer program to the most vulnerable households. 

There are many exciting ways in which AI can contribute to poverty reduction. In Africa, there is suggestive evidence that digital technologies have a large potential to help improve public service delivery and citizen welfare, increase fiscal capacity, and reduce corruption. 

In other fields, such as education, programs have been able to adapt instruction to students’ learning levels and needs by leveraging AI and machine learning or to improve college matriculation rates by sending automated responses to common student questions.

Can you give us some examples of potential pitfalls of using technology in international development?

There has been recent excitement around the potential for technology to help reduce poverty, particularly when it comes to education. However, the hype around education technology has outpaced the evaluation of such programs. In a context of rising inequalities, it is important to step back and understand how and when technology can help—or hinder—student learning. 

For example, research has shown that, despite their popularity, programs seeking to expand access to technology alone, such as by supplying computers or access to the internet, did not in general improve grades and test scores. 

Further, while digital ID systems hold the potential to positively transform people’s lives, recent evidence from India highlights how attempts to reduce corruption in welfare programs by making ID requirements more stringent can also generate non-trivial costs in terms of exclusion and inconvenience to genuine beneficiaries. Implementation challenges of data privacy and misuse could also have significant negative effects on people. This is reflected in the pushback from civil society on digital ID systems in countries such as Kenya and India.

How big is technology based international development a business opportunity for tech companies and how can we make sure social interest comes first?

Technological innovations can have a transformative role in communities and the lives of citizens. For example, lack of identification in developing countries remains a major obstacle for accessing formal finance or formalizing businesses. Evidence from Malawi shows that fingerprinting for credit can lead to a better functioning credit market. Similarly, other Digital IDs and payment systems platforms provide an opportunity for businesses to develop and implement innovative digital tools and platforms as well as harness the benefits of such technology. Through DigiFI Africa, J-PAL aims to learn more about how rapid large-scale digital identification and payment systems, which increasingly attract the attention of African policymakers, can best be structured and implemented so that they have a positive impact on people’s lives.