Econometric analysis of macroeconomic factors influencing construction labour productivity at industry level: evidence from Australia

Understanding the influence of macroeconomic factors on construction labour productivity is essential for developing strategies that mitigate the adverse effects of economic fluctuations. However, previous studies have rarely examined the dynamic lead-lag relationships between macroeconomic indicators and labour productivity in the construction industry using time-series forecasting models. This study addresses this gap by identifying key macroeconomic indicators, such as the Producer Price Index (PPI), Gross Domestic Product (GDP), and Consumer Price Index (CPI), that have a significant temporal influence on construction labour productivity (LP), and by developing a robust LP forecasting model. Time series data on labour productivity and economic indicators were sourced from the Australian Bureau of Statistics. Descriptive analysis, stationarity checks, breakpoint and Granger causality tests were performed to determine the leading indicators and select an appropriate multivariate model. Granger causality results identified PPIs for timber, plumbing, and appliances as significant predictors of construction labour productivity. Vector Autoregression (VAR) model was developed and validated using diagnostic tests, including Residual Serial Correlation and Heteroskedasticity tests, confirming model reliability. The results of impulse response functions showed that a one-standard-deviation shock to LP can lead to a substantial and immediate increase in LP. Furthermore, the variance decomposition test revealed that in the initial periods, nearly all the variance in LP is explained by its own shocks; however, over time, the influence of other variables, such as the PPI of appliances and timber, grows. Out-of-sample forecasting demonstrated high predictive accuracy, with RMSE of 1.36, MAE of 1.11, and MAPE of 1.07%. These findings demonstrate the model’s robustness and practical utility. This research contributes to the existing body of knowledge by identifying key macroeconomic factors that influence construction labour productivity at the industry level and by offering a predictive tool to assist contractors, project managers, and policymakers in anticipating productivity trends.