Time series approach on Philippines’ three economic participation using ARIMA Model

Abstract

The main objective of this study was to predict the three economic participation’s (unemployment, underemployment, employment) in the Philippines for the year 2020 progressively and respectively on a quarterly scale. With a time series approach, the researchers were able to produce ARIMA models that contribute to determining the future values using the quarterly data from the year 2005 to 2019, a total of 60 observations for each economic participation, with the help of the software R Programming. The ARIMA (2,1,0) and ARIMA (0,1,1) were the identified models that are the most adequate and appropriate used to forecast the future values of the three economic participation. These models have undergone series of diagnostics like the seasonally adjusted plot to remove the seasonality of the data, and the Augmented Dickey-Fuller test to check the stationarity which starts with differencing the data. The Augmented Dickey-Fuller test generated the p-values for each economic participation non-stationary which means that the models should undergo a first-order differencing. After differencing, these results were obtained. For the unemployment rate, the ARIMA (2,1,0) forecast the quarterly rate for the year 2020 which are 5.34, 5.29, 5.14, and 4.66. While for the underemployment rate, these values were produced, 15.90, 15.52, 16.15, and 14.90, respectively, by ARIMA (0,1,1). And for employment, ARIMA (2,1,0) was able to generate these values,94.60, 94.64, 94.89, and 95.46. The predicted quarterly values for the year 2020 show a declining trend for unemployment which consequently indicates an inclining path for employment. While the underemployment rate follows a trend from high to low for the first and second quarter, rises for the third quarter and decreases for the last quarter. The obtained results that a low percentage of unemployment and underemployment, subsequently gives employment a higher rate, and vice versa.