Nicholas Bloom of Stanford and Josh Lerner of Harvard co-direct the National Bureau of Economic Research’s Productivity, Innovation, and Entrepreneurship initiative. On its program page, the NBER says the group studies “firm-level and economy-wide forces that determine productivity,” with emphasis on innovation, patenting, entrepreneurship, and the formation of new firms. The same page now highlights featured work on automation’s labor effects and the economics of large language models, a clear signal about where the questions are heading. That evolution is the story: the NBER productivity program is moving AI and new-work dynamics to center stage in mainstream economics.
What the NBER productivity program says it studies
According to the NBER program page, the group examines how management practices, investment spending, the labor force mix, and market structure shape productivity. It bridges firm-level detail with economy-wide outcomes. That breadth matters because the same forces that drive a single plant’s output per hour roll up into national productivity and wage growth.
The co-directors bring distinct lenses. Bloom’s work measures management quality across countries and ties it to performance differences, while Lerner’s research links venture capital, private equity, and incentives to innovation outcomes. The pairing points to a running theme on the page: productivity isn’t only technology; it’s how firms organize, finance, and staff that technology. The U.S. Bureau of Labor Statistics shows how productivity trends anchor growth and living standards. The program’s design suggests it wants the microfoundations behind those aggregates.
AI, automation, and the skill premium move up the agenda
One featured item is titled “New Work, New Technologies, and the Skill Premium,” with work by David Autor and coauthors. The NBER summary states the papers examine how new technologies erode demand for long-standing tasks via automation while also creating fresh demand for “new work.” That pairing fits decades of research on the skill premium and skill-biased technological change, but with a present-day twist: the scope of task change is widening. The thrust is simple and sharp—who gains depends on whether workers and firms pivot to emerging tasks faster than old ones are automated away.
The page also flags featured content on large language models, reflecting a live debate about whether generative AI expands or replaces routine cognitive work. Within this mix, the NBER productivity program places AI alongside more classic drivers like patenting and firm formation, which suggests the research community is treating generative AI as a first-order productivity question, not a side topic.
Why this matters for founders and policy, not just scholars
Program pages rarely read like headlines. But this one points to a near-term playbook for leaders outside academia. If the agenda ties productivity to how firms adopt technology, structure work, and finance innovation, then the operational levers matter as much as the tech itself. Founders can read it as a checklist: where are we creating “new work” that rivals what we’re automating; how clear is our process discipline; and are we investing in complementary skills, not just tools?
On IP, the emphasis on patenting syncs with evidence that protected knowledge can help firms scale innovations. The USPTO’s patent statistics offer a window into where invention is tilting by field, which feeds back into sector-level productivity. On entrepreneurship, Lerner’s long view on venture ecosystems dovetails with how early-stage finance shapes the innovation pipeline. Annual reports from groups like the National Venture Capital Association show funding cycles, but the NBER lens asks a deeper question: which funding structures and governance choices actually push the productivity frontier?
For policymakers, the agenda hints at the trade-offs that matter more than hype cycles. If technology shifts the task mix, training dollars should follow that mix, not only job titles. If investment drives productivity, then tax or credit policies that change the cost of capital can have real effects on firm upgrading. And if the composition of the labor force shapes outcomes, then immigration, education, and childcare policy are productivity policy, not just social policy. The way the NBER productivity program stitches these threads makes that linkage explicit.
How the program’s scope reframes “tech progress”
There’s a quiet but important framing here: innovation shows up in output when organizations absorb it. The page’s spotlight on management practices means diffusion, not just invention, is central. That syncs with a long-standing puzzle—why frontier firms pull away while others lag. Better-managed firms adopt faster, experiment more, and reallocate capital and talent with fewer frictions. In the AI context, this suggests variance across firms could widen before it narrows.
Putting AI research next to patenting and entrepreneurship also nudges analysts to track complements, not only substitutions. If automation depresses demand for certain tasks, complementary “new work” has to be designed, paid for, and measured. Productivity accounting then needs to catch intangibles that standard investment data miss, like software, data pipelines, and training time. The program’s own emphasis on investment and labor composition invites that accounting shift.
What to watch next from the NBER productivity program
Expect more work that measures where generative AI actually moves the needle, where it stalls, and why. Studies that marry firm-level adoption patterns with wage and output data will be especially valuable. Clear links between investment choices and productivity outcomes can guide both CFOs and city-level development plans. If featured work on large language models expands, watch for task-level taxonomies that separate augmentation from substitution in ways managers can act on.
Also watch diffusion. The gap between leaders and laggards can widen when new general-purpose technologies arrive. If management quality and financing conditions explain that gap, then policy and practice should target those bottlenecks. The NBER productivity program has the pieces—management, innovation finance, IP, labor composition—to study that system end to end.
The big takeaway is not a single paper result but the agenda itself. By putting AI, “new work,” investment, and entrepreneurship on the same page, the NBER is telling firms where to look for real productivity gains. For readers outside academia, that map is the value: it shows which margins are likely to matter over the next few planning cycles, and where evidence will arrive first.
In short, the NBER productivity program is signaling that the next wave of productivity research will live at the intersection of AI-driven task change, managerial execution, and the financing of innovation. That’s where founders, investors, and policymakers should meet it. For more on this, see bloomberg.com and nytimes.com.