The authors assess two forecasting methodologies’ abilities to contribute to predicting China’s efforts to indigenize lithography technology, an essential component of manufacturing advanced semiconductors. The authors posed two forecasting questions about China’s potential success or failure, first to a group of experts in a one-day Delphi-method workshop and second to a crowd of generalists on the RAND Forecasting Initiative platform.

How to Forecast China’s Lithography Leap

A Comparative Analysis of the Delphi Method and Crowdsourced Forecasting

Marie Jones, Fabian Villalobos

ResearchPublished May 7, 2026

Forecasting is an analytic tool used to mitigate the risk of surprise because of an uncertain future by collecting quantitative probabilities about a specific future event. In this report, the authors compare two forecasting methodologies—the Delphi method and crowdsourced forecasting—using the context of techno-economic competition and the People’s Republic of China’s efforts to indigenize lithography production capabilities.

Lithography is a crucial process in semiconductor manufacturing that uses light to print intricate patterns on silicon wafers and make computer chips powering modern electronic devices, including those that run artificial intelligence technologies.

In fall 2024, the authors posed two questions to (1) industry experts and academics using the Delphi method during a workshop and (2) RAND’s crowdsourced forecasting platform (known as the RAND Forecasting Initiative). The authors asked forecasters to provide probability estimates to evaluate whether China will be able to procure and operate deep ultraviolet and extreme ultraviolet lithography machines for making cutting-edge computer chips by 2026 and 2030, respectively.

In this report, the authors share their process for and results from comparing the two forecasting methodologies, offering recommendations for researchers’ future efforts and for policymakers making decisions under uncertainty.

Key Takeaways

  • Both forecasting groups identified similar factors influencing China’s likelihood of success or failure, and the groups offered complementary rationales to explain their estimates. The Delphi workshop group was 12 percent more accurate than the crowd of generalists on the first question after placing greater weight on the factor of the timeline being too short for China to achieve a breakthrough.
  • Both crowdsourced forecasting and expert elicitation are highly customizable methodologies that can be applied to a wide range of research budgets and accommodate various participant group sizes.
  • The authors found the active forecasting time frame to be a more important structural difference in the two methodologies than cost or other factors.

Recommendations

  • Researchers running future studies should keep collecting crowdsourced data until the study question or questions are resolved, when possible, to track change over time.
  • Researchers should consider enrolling experts and forecasters in forecaster training before they participate in a Delphi workshop because external studies show that exposure to proven forecasting methods can help mitigate biases.

Topics

Document Details

Citation

Chicago Manual of Style

Jones, Marie and Fabian Villalobos, How to Forecast China’s Lithography Leap: A Comparative Analysis of the Delphi Method and Crowdsourced Forecasting. Santa Monica, CA: RAND Corporation, 2026. https://www.rand.org/pubs/research_reports/RRA4482-1.html.
BibTeX RIS

Research conducted by

This research was independently initiated and conducted using gifts for research at RAND’s discretion from philanthropic supporter Coefficient Giving. This work was undertaken by the RAND Forecasting Initiative within RAND Global and Emerging Risks.

This publication is part of the RAND research report series. Research reports present research findings and objective analysis that address the challenges facing the public and private sectors. All RAND research reports undergo rigorous peer review to ensure high standards for research quality and objectivity.

This document and trademark(s) contained herein are protected by law. This representation of RAND intellectual property is provided for noncommercial use only. Unauthorized posting of this publication online is prohibited; linking directly to this product page is encouraged. Permission is required from RAND to reproduce, or reuse in another form, any of its research documents for commercial purposes. For information on reprint and reuse permissions, please visit www.rand.org/pubs/permissions.

RAND is a nonprofit institution that helps improve policy and decisionmaking through research and analysis. RAND's publications do not necessarily reflect the opinions of its research clients and sponsors.