The United Nations Development Programme (UNDP) launched a policy brief and tool to leverage Big Data and artificial intelligence (AI) for mapping vulnerability to help policymakers and community members combat poverty in Cambodia.
The web-based Cambodian poverty mapping tool can be utilized by policymakers and community members regardless of technical know-how. The tool is viewed as critical in addressing Cambodia’s rapid economic and climate change as it efficiently looks at the needs of the most at-risk communities.
The information in the tool allows users to follow up with in-depth inquiries, such as a lack of infrastructure like paved roads and inadequate health services, and direct new services to have the most impact and support targeted aid efforts.
This tool will provide policymakers with comprehensive, multidimensional information on poverty and interactions of overlapping deprivations often missed by traditional analyses.
“The integration of Big Data and AI in poverty mapping marks a leap forward in our ability to understand and combat poverty in Cambodia,” said Shakeel Ahmad, deputy resident representative of UNDP Cambodia and officer-in-charge.
“The innovative tool will equip policymakers with crucial insights, enabling them to make informed decisions and implement targeted interventions. Our goal is to ensure that everyone benefits from Cambodia’s progress toward sustainable development, leaving no one behind,” he added in an August 8 statement.
According to the Ministry of Planning, Cambodia reduced its poverty rate from 26.3 percent in 2014 to 17.8 percent in 2019. The UNDP’s 2023 Global Multidimensional Poverty Index report showed that multidimensional poverty in Cambodia decreased from 36.7 percent in 2014 to 16.6 percent in 2021, with one in five Cambodians moving out of poverty in just seven years.
However, various segments of the population, especially those living in remote rural areas remain vulnerable to economic shocks. About 47 percent of the population is vulnerable as they hover around the poverty line.
The Covid-19 pandemic could reverse many development gains, particularly those affecting poor and vulnerable communities but the lack of updated, granular and accessible data hampers efforts to track and evaluate the full extent of this impact.
UNDP developed the tool to address that challenge and launched a policy brief on August 8, highlighting key insights and reflections on the use of the tool for policy making, planning and budgeting, and its complementarity with traditional data sources.
UNDP Cambodia communications analyst Ath Chhunmuoy told CamboJA News via email that the project started in July 2021 and was set to be completed by the end of 2024, with the tool being transferred to the Ministry of Planning’s National Institute of Statistics.
“After technical training is provided to ministry officials, we expect them to take full ownership with updating the tool more frequently on a periodical basis to make it stay up to date,” she said.
The key objective is to enable users including policymakers, researchers and civil society organizations to explore poverty and vulnerability alongside specific indicators such as education, health, living standards and monetary factors from provincial to district and commune/sangkat levels, Chhunmuoy added.
The policy brief stated that in the future, poverty mapping will rely less on expensive, infrequent and inconsistent survey-derived data and more on remote sensing and big earth data that is collected frequently and at a regular rate.
Political analyst Em Sovannara said the poverty mapping tool is very useful because UNDP took many years to gather the data and report on the result. The tool provides government data which can be used to formulate policies or develop necessary infrastructure such as water, electricity, and roads.
“[ The government ] can develop policies easier to make the people in the poorer areas in the UNDP map data without having to spend a lot of time researching or hiring NGOs to access the data,” he told CamboJA News.
However, he questioned whether the government recognized or accepted the data in UNDP’s poverty tool.
He said the government may have known about the figures previously, although not as detailed as UNDP’s, and possibly in relation to the intention to restore or intervene in areas that face difficulty.
“Because in villages and communes where people don’t support the ruling party, is the ruling party really willing to develop or does it still discriminate against people with [opposing] political inclinations?” Sovannara said.
In the past, the government seemed to “discriminate against villages and communes” which did not support the ruling party, with development “slower than elsewhere”, and expenditure budget inadequate to meet demand, he alleged.
Ministry of Rural Development spokesperson Pit Karuna told CamboJA News that the launch of the poverty mapping tool would help the ministry access more specific data for rural development plans.
To achieve the first phase of the Pentagonal Strategy, the ministry identified four priorities, which are improving physical and transport infrastructure, and the living conditions of rural communities, diversification of rural economic activities, and strengthening institutions and general affairs.
Meanwhile, Ministry of Education, Youth and Sport spokesperson Khoun Vichheka said they have been assisting schools in remote and challenging areas since 2014.
“We focus on schools in regions with difficult access, low density [fewer than 10 people per square kilometer], frequent flooding or proximity to the border,” she said.
Previously, the disadvantaged areas in remote regions were identified through a collaboration between the ministry and other ministries following detailed discussions. Her ministry is currently reviewing this process and plans to hold similar in-depth discussions at an inter-ministerial level.
Ministry of Planning spokesperson Srey Da declined to comment, noting that it was not his area of expertise and would revert back with answers from relevant people. However, he had not replied at the time of publication.