Welcome!

I am an economist at Federal Reserve Bank of St. Louis.

My research interests are in economic growth, innovation and technological change. I also have interests in the use of text as data. 


Email: aakash.kalyani@outlook.com. For my CV, click here.

Working Papers:

"New Technologies and Inequality" (Mar 2025) w/ Tarek Hassan and Pascual Restrepo.  -- New Paper

Slides Working paper     

This paper shows that the pace of technology creation is a key driver of wage inequality across space and time. College-educated workers learn how to use new technologies faster than others, and this advantage dissipates slowly over time as technologies become standardized and easier to use. 

"Economic Surveillance using Corporate Text" (June 2024) w/ Tarek Hassan, Stephan Hollander, Laurence Van Lent, Markus Schwedeler, and Ahmed Tahoun.  

Prepared for the Journal of Economic Perspectives

We describe application of simple methods from computational linguistics to analyze unstructured corporate texts for economic surveillance.

"The Creativity Decline: Evidence from US Patenting" (April 2024) 

Slides Working paper        SSRN        St Louis Fed Working Paper   Creativity by Patent ID (data)

The divergence between increasing patents and falling productivity growth is explained by a decline in patent creativity, as measured by the share of novel technical terminology in patents. A third of the decline in creativity is due to a lack of new inventors.

Non-academic discussion: Stl Fed, Forbes

"Theory Meets Textual Analysis: Measuring Firm-Level Labor Cost Pressures and Inflation Pass-Through" (Apr 2025) w/ Serdar Ozkan.  -- Draft coming soon


We use textual analysis of earnings calls and the theoretical principle that cost-minimizing firms equate marginal costs across variable inputs to develop a novel measure of firm-level marginal labor cost. This measure, which uses both theory and text, outperforms conventional slack variables in forecasting inflation.

Publications

"Diffusion of New Technologies" (2025) w/ Nicholas Bloom, Marcela Carvalho, Tarek Hassan, Josh Lerner, and Ahmed Tahoun. 

Quarterly Journal of Economics 

NBER WP SSRN Data   VoxEu   Slides    Slides (pdf)  

Using textual analysis of patents, job postings, and earnings conference calls, we track the development and diffusion of new technologies. The development of new technologies is highly concentrated geographically. Hiring spreads geographically and across skill levels as technologies mature, this process is slow, taking around 50 years.

Policy articles

Discussions