Abstract:
In a recent investigation of ways to enhance the role of construction industry in the economy of Kenya, methods of modeling the construction activity were explored. This is because there remains a problem in the construction industry failing to meet demand for constructed facilities in the country at the macro-level. In the exploration, it was noted that ARIMA (Autoregressive Integrated Moving Average) models have widely been used in modeling construction industry – either as a whole or the different sectors of the industry - around the world. This modeling approach which is also popularly known as the Box-Jenkins approach was adopted in this study. This paper is a report of the study. Construction output data collected and analyzed in this paper was obtained from the Kenya National Bureau of Statistics (KNBS) for the period between 1977 and 2019 – 43 years. The developed ARIMA model was evaluated on its predictive power on the basis of an out-of-sample forecast. The study adopted two measures of accuracy - MAPE and RMSE - which produced fairly good results. Therefore, the model can be used for forecasting construction output in Kenya. Finally, it is recommended that the model be applied in setting construction production targets at the macro-level and guiding policy formulation to stimulate higher performance of the construction industry in Kenya.Keywords:Box-Jenkins approach, Construction industry, Construction output, Forecasting.
