Overview
This self-initiated project was born out of my curiosity about how data can reveal deeper insights into economic mobility and social capital. Inspired by groundbreaking studies published in Nature by Raj Chetty and colleagues, I set out to explore how economic connectedness—the extent to which individuals from different socioeconomic backgrounds form relationships—shapes opportunities for upward mobility.
Leveraging the Social Capital Atlas, a dataset developed by Harvard, NYU, and Stanford, I conducted an independent data-driven analysis of the top 50 ZIP codes, counties, high schools, and colleges with the highest levels of economic connectedness across the United States. The dataset, sourced from 21 billion Facebook friendships provided by Meta’s Data for Good program, provided a unique lens into how social networks influence life outcomes.
This project required a multi-faceted approach, involving:
SQL-driven data exploration to extract, process, and rank economic connectedness across different geographic regions.
Schema analysis of a large-scale relational database, consisting of 92 attributes across four key tables (ZIP codes, counties, high schools, and colleges).
Query execution in Google Colab, using structured data analysis to derive meaningful patterns.
Findings compilation in a structured report, drawing insights on which communities foster economic mobility and why.
By merging data engineering with socioeconomic analysis, this project served as a practical exploration of how AI, big data, and analytics can be applied to social sciences and policy research.
Motivation Statement
This project was driven by my deep-seated interest in data-driven social research—specifically, how quantitative insights can help us understand and address economic inequality. The idea that who you know can impact your financial future as much as what you know is compelling, and I wanted to examine this phenomenon through rigorous data exploration and structured analysis.
As someone passionate about AI ethics, data science, and social impact, this project allowed me to bridge technical expertise with a meaningful societal question: how can data help us design more equitable economic systems? The dataset’s scale and complexity posed an exciting challenge, requiring me to think critically about query design, data extraction, and how to translate findings into real-world implications.
Beyond the technical aspects, this project resonated with my long-term vision of using AI and data for social good. It pushed me to think not just about how to process data, but how to interpret it in ways that inform policy, economic strategies, and structural reforms.
Reflections and Learnings
Taking on this project independently was a deeply rewarding learning experience, sharpening both my technical skills in data engineering and my ability to interpret data in a broader socioeconomic context. Some of my key takeaways include:
The Power of Large-Scale Data for Social Insights – Working with a dataset of 21 billion connections reinforced my belief in data-driven policymaking and the ability of big data to uncover systemic economic patterns.
Advanced Data Engineering and Query Optimization – Navigating a complex relational database with 92 attributes required schema fluency, structured query design, and efficiency in data extraction. This project honed my ability to work with large datasets methodically and effectively.
Interpreting Data Beyond Numbers – Extracting insights wasn’t just about running SQL queries; it required translating raw data into actionable conclusions that could inform real-world economic discussions. This reinforced the importance of communicating data-driven insights for both technical and non-technical audiences.
AI and Data Science for Societal Change – This project deepened my passion for using machine learning and big data to tackle systemic challenges. Understanding how economic connectedness influences upward mobility inspired me to further explore AI applications in public policy, urban planning, and economic development.
By taking an independent, research-driven approach, this project not only strengthened my technical foundation but also deepened my conviction in data’s role as a tool for meaningful social impact. Moving forward, I plan to build on this experience by continuing to explore how AI and data science can drive evidence-based change in economic research, policy-making, and AI ethics.
