Austin Clemens

I deleted my blog but I am doing some projects again so here's a landing page

Hi, I'm a social scientist, coder, data visualizer, and etc. You can find me on LinkedIn and BlueSky . I took my blog down a long time ago, but I'm occasionally doing little data projects these days. This landing page is a hub for them, and that's pretty much it!

• Media coverage of pandemic inflation was significantly higher than inflation alone would predict – Using a corpus of NYT, Washington Post, and USA Today articles, I looked at the level and tone of inflation coverage in the pandemic. The volume coverage was high, and stayed elevated through 2024 despite falling levels of inflation. This appears to be due at least in part to election narratives.
• The U.S. Inequality Tracker - I developed this tracker in collaboration with the Washington Center for Equitable Growth in 2025. It's an automatically updating look at income and wealth inequality in the U.S.
• Which political events does media focus on most? - This is an exploration of 1.5 years of Washington Post and New York Times articles in print, centered around 8 high-coverage events. I used machine learning and AI to identify these events and graphed total coverage of them by each paper per week.
• An Algorithm for Creating and Selecting Graph Axes - this is a post from my defunct data blog, first published in 2016. I've always been real proud of this algorithm, so I resurrected it here.
• A replication of Ben Harris and Aaron Sojourner's work on media tone – this is a replication, in Python, of Harris and Sojourner's Stata code, with a data cron that fetches new data and updates data visualizations. It is updated quarterly to show how the sentiment of media coverage of the economy differs from sentiment predicted by economic fundamentals.
• Stock market tracker for Trump's 2nd term - just a goofy little thing that tracks the three major stock indices in Trump's second term and compares it to Biden's term. Updated nightly.
• Exploring gendered language in mass culture, the media, and Victorian literature - this is a really old natural language project I did where I created 6 large corpuses and analyzed how men and women are described in them. Parts of it don't really work anymore and it's a bit archaic generally, but I always liked it.

Here's a pic of me working on my guitar chords.