Forecasting USA Unemployment Rate Base on ARIMA Model

Abstract

This paper presents a detailed analysis of unemployment rate forecasting, a critical subject for various stakeholders including policymakers, businesses, and individuals. Amid significant economic events such as the global financial crisis and COVID-19 pandemic, the need for precise unemployment forecasts has become crucial. The research utilizes an Autoregressive Integrated Moving Average (ARIMA) model to analyze US unemployment rate data from 2000 to 2023, sourced from the Federal Reserve Economic Data (FRED). The paper identifies seasonality patterns, executes appropriate data transformations, and incorporates the Box-Jenkins methodology for ARIMA model identification. The findings reveal the model's resilience, demonstrating accurate forecasts despite significant disruptions. These insights offer valuable contributions in understanding labor market dynamics, facilitating informed decision-making and strategic planning. The paper highlights the robustness of ARIMA models, and their potential to adapt to rapid changes in the economic landscape, thereby proving invaluable in forecasting unemployment rates.