I'm Thomas Blanchet 🧔🏼♂️. I'm a data scientist and economist:
- 👨🏼💻 I'm currently data scientist at Untapped Global. Before, I was the statistical coordinator for the World Inequality Lab at the Paris School of Economics, and in 2021-2022, I worked for the University of California, Berkeley.
- 💻 I mostly work with R, Python, and Stata. And sometimes with Java and C++.
- 🌐 I also have experience coding for the web, using HTML/CSS/JavaScript and React.
- 🎓 I have a Master's in Data Science from ENSAE ParisTech and a Ph.D. in Economics from the Paris School of Economics. Over the years, I've done work in many areas of data science, including descriptive analysis, machine learning, predictive modeling, causal inference, bayesian methods, time series analysis, stochastic calculus, extreme value theory, etc.
- 🏛 You may also want to look at my academic website. My work has been published in international peer-reviewed journals such as the American Economic Journal: Applied Economics and the Journal of Monetary Economics. It has been featured in numerous outlets, including the New York Times, the Financial Times, Time Magazine, Bloomberg, Le Monde, and the White House's Economic Blueprint.
- 📫 You can reach me at [email protected].
What can you find here? (Click below to expand)
🛠 Code for the many tools that I have developed over the years to help people use and produce new economic data.
- gpinter — R package with an online interface, used by researchers worldwide to make use of the historical data on income and wealth published by tax authorities.
- enforce — a Stata command to adjust values and fill in missing data based on an arbitrary set of accounting identities between variables. The tool relies on methods from quadratic programming and singular value decomposition to intelligently perform the task.
- bfmcorr — a Stata command to correct representativity in surveys using tax data, using tools from calibration theory and a custom-made methodology to automatically determine the optimal way of combining the data.
- WID.world download tools — R and Stata commands that use the WID.world JSON API to easily download wealth and income data in statistical software.
📖 Replication packages for some of my academic work.
- Uncovering the Dynamics of the Wealth Distribution — a new methodology that uses stochastic calculus to decompose the dynamics of wealth inequality in the United States. See also the companion website, which applies the paper's tools to study wealth taxation, and its GitHub repository.
- Real-Time Inequality — a first-of-its-kind tool that combines many economic datasets to nowcast the distribution of income and wealth in the United States. See also the companion website that publishes monthly and quarterly data updates.
React
Typescript
Django
Laravel
Facebook
Microsoft
Google
Alibaba
D3
Tencent