Forecasting Labor Markets with LSTNet: A Multi-Scale Deep Learning Approach

TL;DR

This paper introduces LSTNet, a deep learning model that forecasts short-term employment changes and evaluates industry health using multivariate labor market data, outperforming baselines and providing interpretable insights.

Contribution

The paper presents LSTNet, a novel multi-scale deep learning approach for labor market forecasting and industry health assessment, with improved accuracy and interpretability.

Findings

LSTNet outperforms baseline models in most sectors.

The Industry Employment Health Index correlates well with actual employment volatility.

The model provides accurate 7-day employment forecasts.

Abstract

We present a deep learning approach for forecasting short-term employment changes and assessing long-term industry health using labor market data from the U.S. Bureau of Labor Statistics. Our system leverages a Long- and Short-Term Time-series Network (LSTNet) to process multivariate time series data, including employment levels, wages, turnover rates, and job openings. The model outputs both 7-day employment forecasts and an interpretable Industry Employment Health Index (IEHI). Our approach outperforms baseline models across most sectors, particularly in stable industries, and demonstrates strong alignment between IEHI rankings and actual employment volatility. We discuss error patterns, sector-specific performance, and future directions for improving interpretability and generalization.