NOA-POVERTY

A Norm Optimisation Approach to tackle poverty

The use of Artificial Intelligence to mitigate discrimination against the poor

An Agent-Based social model towards the UN Sustainable Development Goals

Would lower levels of discrimination against the poor contribute to the reduction of poverty levels?

We use Artificial Intelligence simulations to foresee the impact that discrimination against the poor has on poverty.

Our aim is to guide a new generation of poverty reduction policies that mitigate poverty not only by the redistribution of wealth, but also by mitigating the discrimination against the poor.

The state of the art in Artificial Intelligence and social sciences working hand-in-hand with NGOs to find alternative ways to mitigate poverty

NOA-Poverty is a multidisciplinary project being developed by computer scientists and social science researchers from the IIIA-CSIC (Barcelona) and the Technology Ethics Center at the University of Notre Dame (United States) with the collaboration of NGOs (Fundació Arrels and Cáritas). The project has just won the AI for Good Best Project Award at the International Joint Conference on Artificial Intelligence (IJCAI-2022) and has already produced a published paper in the IJCAI proceedings.

Contact

gcurtore@nd.edu

Motivation

Traditional poverty reduction policies based on the redistribution of wealth have proved ineffective in the recent decades: the United Nations informs that currently 700 M people (10% of the total world population) still live in extreme poverty. Only within in the EU-27, 96.5 M people are presently at risk of poverty or social exclusion (21.9% of EU-27 population) according to Eurostat.

New and alternative solutions are needed to work towards the #1 UN Sustainable Development Goal (poverty eradication) and the “Leaving No One Behind” principle.

Rationale: Aporophobia

The phenomenon of discrimination against the poor has not received the attention it deserves in the literature. It was not until 2017 that philosopher Adela Cortina coined the term aporophobia to describe the rejection of the poor (1). In 2021 the Spanish legal framework was pioneer to include this discriminatory phenomenon as an aggravating factor for hate crimes (2) and this year NOA-Poverty team members provided the first dataset informing about bias against the poor in Word2Vec and GloVe embeddings by using word vector representations (3).

To date, no evidence has been provided whether aporophobia hinders the success of poverty reduction policies. This study aims to fill in that gap.

(1). Cortina, A. (2017). Aporophobia el rechazo al pobre. Paidós.
(2). Boletín Oficial del Estado. Circular 7/2019, de 14 de mayo, de la fiscalía general del estado, sobre pautas para interpretar los delitos de odio tipificados en el artículo 510 del código penal. BOE – A – 2019 – 7771, 2021. Ministerio de la Presidencia, Relaciones con las Cortes y Memoria Democrática (Gobierno de España).
(3). Curto, G., Jojoa Acosta, M.F., Comim, F., García-Zapirain, B., (2022). Are the poor being discriminated against on the Internet? A machine learning analysis using Word2Vec and GloVe embeddings to identify aporophobia. AI & Society.1

Our outputs

Guidelines

To suggest alternative ways to mitigate poverty by acting on the discrimination against the poor in context-specific regulatory systems

Aporophobia-meter

To track discrimination against the poor and see its correlation with economic indicators and historical events

Graphical User Interface

To aid policy makers see the impact of potential regulations related to poverty and discrimination before these are actually implemented