{"title":"From data to decision: Alleviating poverty and promoting development through measuring the unmeasurable economic numbers","authors":"Emmanuel A. Onsay , Jomar F. Rabajante","doi":"10.1016/j.socimp.2025.100138","DOIUrl":null,"url":null,"abstract":"<div><div>This work has been carried out and is currently being conducted in the poorest region of Luzon, Philippines. Since poverty is multifaceted and considered unmeasurable in social science, it is notoriously difficult to measure. The methods currently used to measure poverty require a significant amount of time, money, and labor. This poses challenges for policymakers in implementing poverty-reduction policies. Indigenous communities are among the most disadvantaged, vulnerable, and neglected populations in society, facing complex and diverse socioeconomic situations. Poverty stands as one of the oldest and most challenging social problems to have ever existed. Research on indigenous peoples typically takes a qualitative approach, whereas studies on poverty tend to be broad, making them susceptible to significant sampling errors and primarily intended for national policy-making. Using community-based monitoring system (CBMS) data, we achieved a prediction accuracy of 92.60–98.00 % using Random Forest classification and reduced traditional survey and data processing costs by up to 70 %. The proposed model incorporates 27 socioeconomic variables and enables localized policy targeting. Therefore, it is crucial to assess multifaceted poverty and simulate socioeconomic conditions for each tribe to foster economic development. By training and testing datasets, this work proposes new metrics and illustrates the effectiveness of machine learning in predicting poverty. Lastly, the results provide various localities with customized policy targeting tools for poverty alleviation. These techniques can be replicated, adapted, or repurposed by other researchers to assist impoverished populations in improving their well-being.</div></div>","PeriodicalId":101167,"journal":{"name":"Societal Impacts","volume":"6 ","pages":"Article 100138"},"PeriodicalIF":0.0000,"publicationDate":"2025-08-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Societal Impacts","FirstCategoryId":"1085","ListUrlMain":"https://www.sciencedirect.com/science/article/pii/S2949697725000372","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 0
Abstract
This work has been carried out and is currently being conducted in the poorest region of Luzon, Philippines. Since poverty is multifaceted and considered unmeasurable in social science, it is notoriously difficult to measure. The methods currently used to measure poverty require a significant amount of time, money, and labor. This poses challenges for policymakers in implementing poverty-reduction policies. Indigenous communities are among the most disadvantaged, vulnerable, and neglected populations in society, facing complex and diverse socioeconomic situations. Poverty stands as one of the oldest and most challenging social problems to have ever existed. Research on indigenous peoples typically takes a qualitative approach, whereas studies on poverty tend to be broad, making them susceptible to significant sampling errors and primarily intended for national policy-making. Using community-based monitoring system (CBMS) data, we achieved a prediction accuracy of 92.60–98.00 % using Random Forest classification and reduced traditional survey and data processing costs by up to 70 %. The proposed model incorporates 27 socioeconomic variables and enables localized policy targeting. Therefore, it is crucial to assess multifaceted poverty and simulate socioeconomic conditions for each tribe to foster economic development. By training and testing datasets, this work proposes new metrics and illustrates the effectiveness of machine learning in predicting poverty. Lastly, the results provide various localities with customized policy targeting tools for poverty alleviation. These techniques can be replicated, adapted, or repurposed by other researchers to assist impoverished populations in improving their well-being.