Discussions around precise definitions of poverty are often fraught affairs, not least because so much rests on just what side of the line you sit. New research from Aston University suggests we could do better by deploying machine learning to more accurately measure poverty in different countries.
The paper argues that current thinking around poverty fails because too much emphasis is placed on subjective notions of needs, with this often failing to capture the complexity of how people spend their money. By using AI to crunch through huge quantities of economic and spending data, the researchers believe their model provides a much richer perspective on the matter.
“No-one has ever used machine learning to decode multidimensional poverty before,” the researchers say. “This completely changes the way people should look at poverty.”
Measuring poverty
The researchers explain that most current measures of poverty aim to establish a monetary level, below which a person or household is defined as poor. These methods trace their origins back to early reformers from the 19th century.
At the moment, the World Bank sets this line at $1.90 per day. It’s a line that approximately 10% of the world’s population falls below. The figure is calculated using a subjective assessment of the level of income required to cover basic needs in the poorest countries.
The researchers analyzed data from the last 30 years in India to allow them to split expenditure into three distinct categories of basic food and spending on housing and transport. The team aims to recognize the interplay between each category, with additional spending in one place typically resulting in less spending elsewhere. They believe this provides a more holistic measure of poverty that can be more readily adjusted to the unique circumstances of individual countries.
This data, combined with data on asset and commodity markets, allowed them to develop a model that was able to accurately predict past poverty levels in India and the United States. The model also attempts to predict future poverty levels based on various economic assumptions.
“Current thinking on poverty is highly subjective, because ‘poverty’ will mean different things in different countries and regions,” the researchers say. “With this model, we finally have a multi-dimensional poverty index that reflects the real-world experience of people wherever they live and largely independent of the social class they are deemed to belong to.
“Importantly, it’s a model that takes account of the economic circumstances people find themselves in – and the factors that can make the biggest difference to their material wellbeing. As such, it can be an important tool for governments and policymakers globally in identifying poverty and putting in place interventions that really tackle it.”
