Disruption Index (CD-Index)
Also known as: CD Index, Consolidation-Disruption Index, CD5 Index, Disruptiveness Measure
The disruption index, or CD index, classifies a scientific paper or patent by how the work that cites it treats the work it built on. Introduced by Russell Funk and Jason Owen-Smith in 2017 as a dynamic network measure of technological change, and popularized for science by Lingfei Wu, Dashun Wang, and James Evans in 2019, it asks a simple structural question: when later researchers cite a focal work, do they also keep citing that work's own references, or do they cite the focal work instead of its predecessors? If subsequent work cites the focal item but largely ignores its references, the item has disrupted its field, eclipsing what came before; if subsequent work cites both the item and its references together, the item has consolidated existing knowledge. The index runs from -1 (purely consolidating) to +1 (purely disrupting) and has become a standard tool for measuring whether contributions push science in new directions or deepen established lines.
Key highlights
- Captures the qualitative character of a contribution, disruptive versus consolidating, that raw citation counts cannot distinguish.
- Relies only on citation-network structure, so it scales to tens of millions of papers and patents without reading text.
- Produces a bounded, interpretable score in [-1, 1] that is comparable across fields once windows and norms are held fixed.
- Validated against canonical paradigm-shifting and incremental works, and linked to substantive correlates such as team size.
Intuition
This section is available to Pro members. Upgrade to Pro
How it works
This section is available to Pro members. Upgrade to Pro
When to use it
Use the CD index when you want to characterize the nature of a contribution rather than merely its volume of citations, distinguishing work that opens new directions from work that deepens established ones. It is well suited to large-scale comparative studies of teams, fields, funding regimes, or eras where you have clean, reasonably complete citation data with both forward and backward links and a stable observation window. It is appropriate for papers, patents, or other citing objects as long as the citation graph is dense enough to populate the three citer types. It is less appropriate for very recent work whose forward window has not matured, for sparsely cited items where the index is dominated by noise, or for disciplines with idiosyncratic citation norms that violate the assumption that omitting references signals replacement. Because the measure is purely structural, it should be paired with content-aware reading when classifying any individual paper.
Strengths & limitations
- Captures the qualitative character of a contribution, disruptive versus consolidating, that raw citation counts cannot distinguish.
- Relies only on citation-network structure, so it scales to tens of millions of papers and patents without reading text.
- Produces a bounded, interpretable score in [-1, 1] that is comparable across fields once windows and norms are held fixed.
- Validated against canonical paradigm-shifting and incremental works, and linked to substantive correlates such as team size.
- Highly sensitive to the citation window and to the completeness of backward references, so data quality strongly affects scores.
- Sparsely cited works yield unstable, noisy indices dominated by a handful of citers.
- Field differences in citation and referencing norms can bias cross-disciplinary comparisons unless carefully normalized.
- A purely structural definition can mislabel individual papers whose citing patterns reflect convention rather than genuine replacement of prior art.
Common pitfalls
This section is available to Pro members. Upgrade to Pro
Applications
This section is available to Pro members. Upgrade to Pro
Frequently asked
How is the CD index different from a citation count?
A citation count measures how much attention a work received; the CD index measures the character of that attention. Two papers can have identical citation counts yet opposite CD scores: one disruptive, because later work cites it instead of its predecessors, and one consolidating, because later work cites it alongside its predecessors. The index reads the structure of the citing community's behavior rather than its volume, so it distinguishes works that redefined a field from works that extended it, a distinction citation counts are blind to.
What does a negative CD value mean?
A negative value means the work is consolidating. Later papers that cite it tend also to cite the references it was built on, so the community treats the focal work as continuous with prior art rather than as a replacement for it. Consolidating is not a pejorative: much essential science deepens and validates existing lines of inquiry. Funk and Owen-Smith and later Wu, Wang, and Evans designed the index precisely so that the two ends of the scale, disruption near +1 and consolidation near -1, name two legitimate and complementary modes of scientific contribution.
Why is the choice of time window so important?
Disruption is revealed only as the literature reacts to a work, so the index depends on how long you wait. A short window may catch only early adopters and understate eventual disruption, while comparing works observed over windows of different lengths conflates timing with character. The CD5 convention fixes a five-year forward window so that every work is judged over the same maturation period. Without a fixed window, CD scores are not comparable across works of different ages, which is one of the most common sources of error in applying the measure.
Sources
- 1.Funk, R. J., & Owen-Smith, J. (2017). A Dynamic Network Measure of Technological Change. Management Science, 63(3), 791-817.
- 2.Wu, L., Wang, D., & Evans, J. A. (2019). Large teams develop and small teams disrupt science and technology. Nature, 566, 378-382.
You have read it. What now?
Cite this page
ScholarGate. (2026, June 23). Disruption Index (CD-Index). ScholarGate. https://scholargate.app/bibliometrics/cd-index-disruption