Identifying the Multiple Skills in Skill-Biased Technical Change

US median wages have stagnated for 30 years, yet this masks a surprising amount of variation among different groups. A large literature documents that skill-biased technical change (SBTC) has led to substantial real wage increases on average for those with at least a college degree, and declines on average for those with less education. However, SBTC is not a single dimensional construct. We identify seven orthogonal dimensions of skill across 514 occupations and examine how the earning power of each skill changed over the period 2006-2014. We find that (1) different skills accrue different levels of wage, (2) these wage effects vary on both the actual use of IT in the occupation as well as the general level of IT intensity in the industry, and (3) the technology-skill effects on wage are nonlinear. We discuss implications for workers, educators and policymakers.