We analyze the effect of ICT and R&D on total factor productivity (TFP) growth across different industries in Sweden. R&D alone is significantly associated with contemporaneous TFP growth, thus exhibiting spillover effects. Although there is no significant short-run association between ICT and...

We document a significant increase in the sorting of workers by cognitive and non-cognitive skills across Swedish firms between 1986 and 2008. The weight of the evidence suggests that the increase in sorting is due to stronger complementarities between worker skills and technology. In...

This review considers the evolution of economic thinking on the relationship between digital technology and inequality across four decades, encompassing four related but intellectually distinct paradigms, which I refer to as the education race, the task polarization model, the...

Several scholars as well as industry professionals have claimed that there is a "software-biased shift" in the nature and direction of innovation in that software development is a core part of innovation activities in firms across a wide array of industries. Empirical firm-level evidence of such...

We analyze investment decisions when information is costly, with and without delegation to an agent. We use a rational-inattention model and compare it with a canonical signal-extraction model. We identify three "investment conditions". In "sour" conditions, no information is acquired and no...

We exploit exogenous variation arising from the historical rollout of the Swedish railroad network across municipalities to identify the impacts of lowered interaction costs on innovative activity. A network connection led to a surge in local innovation due to an increased entry, productivity,...

Does capital accumulation increase labor demand and wages? Neoclassical production functions, where capital and labor are q-complements, ensure that the answer is yes, so long as labor markets are competitive. This result critically depends on the assumption that capital accumulation does not...

We analyze how output and wages behave under different scenarios for technological progress that may culminate in Artificial General Intelligence (AGI), defined as the ability of AI systems to perform all tasks that humans can perform. We assume that human work can be decomposed into atomistic...

This paper uses a large language model to develop an ex-ante measure of the commercial potential of scientific findings. In addition to validating the measure against the typical holdout sample, we validate it externally against 1.) the progression of scientific findings through a major...

We construct a grid that covers the key business functions of an establishment and the main technologies used in each of them. We populate this grid with data from over 20,000 establishments in 15 countries. We use this dataset to document novel "facts" about how establishments use technology,...