Akash Dubey

Akash Dubey

I've been working in software for over 10 years 😉

Economic Drivers of Puerto Rican Migration

Most historical analyses rely purely on text, but numbers often tell the most compelling stories of demographic shifts. This project is a comprehensive, interactive data visualization website that tracks and analyzes the profound economic drivers behind Puerto Rican migration to the mainland United States from 1950 to 2024.

View the Interactive Website | View on GitHub

Data-Driven Storytelling

The project bridges the gap between historical literature (specifically Piri Thomas’s Down These Mean Streets) and quantitative demographic realities. By fusing literary analysis with deep datasets from the U.S. Census Bureau, IPUMS USA, and Federal Reserve Economic Data (FRED), the platform visualizes the structural forces that drove mass displacement and resettlement.

The Three Eras of Migration

The interactive experience guides users through three distinct historical phases:

  1. The Great Migration (1950s–1970s): Driven by Operation Bootstrap’s shift from agriculture to manufacturing, displacing hundreds of thousands of workers who hyper-concentrated in New York City (representing 88% of mainland Puerto Ricans in 1940).
  2. The Dispersion (1980s–2000s): The shift away from New York and the rise of a U.S.-born mainland majority—the “Nuyorican” shift.
  3. The Modern Exodus (2010s–Present): Driven by modern debt crises and natural disasters, visualizing the massive geographic reorientation toward Florida and the American South.

Technical Implementation

The website is explicitly designed for smooth, scroll-driven storytelling.

  • Frontend: Built with vanilla HTML/CSS, utilizing CSS animations (IntersectionObserver) for staggered, fade-in loading of text and graphs as the user scrolls.
  • Data Visualization: Employs Python (Pandas) and Plotly to generate rich, interactive HTML charts. These standalone interactive iframe graphs allow users to hover over data points, zoom into specific timelines, and toggle demographic variables natively in the browser without a backend server.
  • Data Engineering: Cleaned, processed, and merged multiple decades of raw IPUMS microdata and historical census records to create continuous time-series analyses.