Forecasting the 2022-23 tech layoffs using epidemiological models

TL;DR

This paper applies epidemiological SIR models to predict the end of the 2022-23 tech layoffs, estimating a return to normal levels by late 2023 without relying on economic forecasts.

Contribution

It introduces a novel approach of using SIR epidemiological models to forecast tech layoffs, providing a data-driven prediction method.

Findings

Layoffs follow an SIR model pattern

Layoffs expected to normalize by late 2023

Model fits observed data well

Abstract

Many large and small companies in the tech and startup sector have been laying off an unusually high number of workers in 2022 and 2023. We are interested in predicting when this period of layoffs might end, without resorting to economic forecasts. We observe that a sample of layoffs up to March 31, 2023 follow the pattern of noisy observations from an SIR (Susceptible-Infectious-Removed) model. A model is fitted to the data using an analytical solution to the SIR model obtained by Kr\"{o}ger and Schlickeiser. From the fitted model we estimate that the number of weekly layoffs will return to normal levels around the end of 2023.