Unemployment remains a critical macroeconomic issue in Southeast Asia, significantly impacting the Philippines. This study aims to forecast the unemployment rate in the Philippines using ARIMA models to inform policy decisions and economic planning. The study utilized ARMA model for time series forecasting, leveraging past values and error terms to predict future data points. The methodology involved identifying suitable models through stationarity tests and autocorrelation analyses, estimating model parameters, and conducting diagnostic checks to ensure accuracy. Finally, the models were used to forecast future values and compute confidence intervals, providing insights into potential future trends. Monthly unemployment data from 2019 to 2023 was analyzed using the ARIMA (1,1,0) model, selected for its optimal fit based on statistical criteria. Initial findings indicated non-stationary unemployment data, necessitating first differencing to achieve stationarity. The ARIMA (1,1,0) model showed robustness, evidenced by diagnostic checks and Portmanteau test results confirming white noise residuals. Forecasts from 2024 to 2050 suggest fluctuating unemployment rates until 2034, followed by a gradual decline, indicating potential economic growth and improved labor market conditions. These forecasts align with the broader goal of sustainable economic development and highlight the importance of strategic policy interventions to foster job creation and economic stability in the Philippines. The study highlights the efficacy of ARIMA models in capturing unemployment trends and aiding policymakers in mitigating the socio-economic impacts of unemployment.
Keywords: unemployment rate, forecasting, ARMA model, Box-Jenkins Methodology, economic strategies, Sustainable Development Goals