Understanding the Underlying Trends in US Construction Labor Wages: A Data-Driven Mixed-Method Computational Approach

Precisely forecasting construction costs is crucial for maintaining financial stability for contractors and the broader construction sector. Nonetheless, this task has long been acknowledged as challenging. Recent global events and inflationary pressures have notably driven up construction labor expenses. While existing research examined labor shortages, there remains a gap in understanding the diverse labor wage trends. This paper addresses this research gap following a multistep methodology, which included: (1) gathering construction labor data from 1999 to 2023; (2) conducting trend and statistical analyses to discern the underlying patterns in labor trends; (3) employing clustering analysis to categorize construction occupations based on their wage and employment changes; (4) assessing univariate time-series analysis to forecast median labor wages; and (5) utilizing bivariate vector autoregression models and Granger causality to assess the wage fluctuation transmission among various occupations. Trend analysis reveals wage correlations among most occupations, with consistent upward wage growths. Subsequent clustering analysis partitioned the occupations into four groups based on their differing wage and employment changes. Notably, lower-wage occupations, such as helpers for various construction trades, exhibited the highest wage increases and substantial workforce size reductions. Univariate models demonstrated adequate predictive performance for forecasting overall wage trends across occupations. Additionally, construction laborers and carpenters were identified as key occupations with high capacity to transmit wage fluctuations, while supervisory roles, electricians, and plumbing workers were found to be susceptible to receiving such fluctuations. This study provides valuable insights for contractors by (1) identifying trades with substantially increasing wages, guiding where additional contingencies could be allocated; (2) proposing a time-series approach as a useful tool for wage forecasting; and (3) identifying key occupations that transmit and receive wage fluctuations. Contractors can utilize these findings to proactively plan for labor wage changes, thereby enhancing financial robustness in the broader construction industry.