Wind project learning-by-doing is often cited as a reason that building more renewable energy capacity could make future projects less expensive. But a key question for developers, investors and policymakers is whose experience produces those savings: a company’s own project history, or knowledge that spreads across the industry? An analysis of U.S. wind project costs from 2001 to 2015 separates these possibilities and finds evidence consistent with cost reductions from a developer’s own experience, but not from other firms’ experience.
The distinction matters beyond renewable energy policy. For operations leaders and technical decision-makers, it illustrates why cost trends should not automatically be attributed to shared industry learning. Project scale, input prices and technical change can also affect costs, while practical knowledge may remain within the organization that gains it.
Separating firm learning from industry spillovers
Learning-by-doing describes productivity improvements that arise as an organization repeats a production or project delivery process. In wind development, experience can include learning to assess sites, manage permitting, coordinate transport to remote locations, and plan construction activities such as foundations, access roads and electrical connections.
Those capabilities could reduce the cost of a firm’s own later projects. Alternatively, information developed by one company might spread to competitors, creating industry-wide knowledge spillovers. These two mechanisms have different implications: a spillover benefits firms that did not bear the cost of gaining the experience, while firm-specific learning is captured by the organization doing the work.
To distinguish between them, the researchers built project-level measures of experience for individual developers and for other industry participants. They examined two ways of counting prior output: cumulative installed generating capacity, measured in megawatts, and the cumulative number of completed projects. The analysis also allowed the value of past projects to vary according to how long ago, and how far away, they were completed.
What the U.S. wind project cost analysis found
The study assembled information on U.S. wind projects completed through 2015 and estimated costs for 408 projects completed from 2002 to 2015. A project’s completion cost included development, equipment procurement and construction. The cost model accounted for factors that could otherwise be mistaken for learning, including project capacity, input prices, year, state, developer and turbine manufacturer.
Across model specifications, the results were consistent with a cost-reducing effect from a developer’s own accumulated experience. The estimated effect indicates that doubling a firm’s experience base was associated with a 1.3% to 1.6% reduction in the cost of installing a megawatt of capacity, all else being equal.
By contrast, the analysis found no evidence that experience accumulated by other wind developers reduced a firm’s project costs. This does not establish that information never moves between companies. Rather, within the study’s framework and data, the researchers could not identify a measurable cost benefit from competitors’ accumulated project experience.
The findings also caution against treating falling or rising average project costs as direct evidence of learning. For example, project costs rose during much of the 2000s even as investment in new wind capacity increased. Changes in major input prices and technology can influence costs at the same time as firms gain experience, making it important to account for these factors separately.
Experience can lose value across time and distance
Experience was not treated as a permanent, uniform asset. The analysis allowed older projects to contribute less to a developer’s current experience base than recent ones, and nearby projects to count more than those completed more than 100 miles away. The estimates suggest experience depreciates over time and distance.
There are practical reasons why project history might become less applicable. Technology and project conditions change, while knowledge of local contractors, infrastructure, regulations and physical conditions may be more useful in one region than another. A firm’s historical project count alone may therefore be a poor guide to how much relevant experience it can bring to a particular job.
The researchers also considered joint ventures and acquisitions. Their model tested how project experience might be divided among joint-venture partners and whether experience could transfer when a firm was acquired. The evidence on these mechanisms was inconclusive: the data could not reject the possibility that joint-venture experience was shared equally or that acquired experience was as valuable as experience developed internally.
Implications for project delivery and policy
For developers and project owners, the results support a careful distinction between organizational experience and general industry progress. A company’s accumulated capability may contribute to lower costs on its own future work, but the study does not show that competitors automatically gain the same benefit. When evaluating a project team, decision-makers may therefore want to consider the relevance of its previous work—not just the volume of past activity, but also its recency and geographic applicability.
For policymakers, the findings are relevant to arguments for supporting renewable investment specifically to accelerate industry-wide learning. The study finds no evidence of cost-reducing spillovers between developers in the period examined, so it does not substantiate that particular rationale for support. It does not assess other possible policy considerations, such as environmental effects.
Conclusion
The U.S. wind industry analysis points to a targeted form of learning: developers’ own experience is associated with lower costs on their subsequent projects, while measurable benefits from other firms’ experience were not identified. Because experience can lose relevance across time and distance, its value depends on more than accumulated output. For project planning and cost assessment, the practical lesson is to distinguish firm-specific capability from broader industry change and to account for the changing relevance of past work.