Alliance Portfolio Similarity
Also known as: Alliance Portfolio Similarity Scores, S-Score of Alliance Similarity, Tau-b Alliance Similarity, Alliance Profile Similarity
Alliance portfolio similarity measures how alike two states' overall patterns of alliance commitments are. Each state has a 'portfolio' — the profile of defense pacts, neutrality agreements, ententes, or no tie it holds with every other state — and the similarity of two portfolios is summarized in a single dyadic score. Signorino and Ritter (1999) showed that the long-dominant Kendall's tau-b measure is flawed for this purpose and introduced the S-score as a better-behaved alternative. These scores are a standard proxy for shared interests and have been used to operationalize utilities in expected-utility models of war.
Key highlights
- Provides a single, comparable dyadic summary of two states' security alignment.
- The S-score corrects documented defects of the older tau-b measure.
- Available over long historical periods from standard alliance datasets.
- Operationalizes shared interests and the utilities used in expected-utility models of war.
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 alliance portfolio similarity when you need a dyadic, behavioral proxy for shared strategic interests or revealed foreign-policy alignment, especially for operationalizing utilities in formal models or controlling for affinity in conflict analyses. It is well suited to long historical spans where alliance data exist. It is less appropriate when alliances poorly reflect the interests you care about, when more direct measures (UN voting, issue positions) are available and preferable, or when the tau-b/S choice and weighting decisions would materially confound the analysis without justification.
Strengths & limitations
- Provides a single, comparable dyadic summary of two states' security alignment.
- The S-score corrects documented defects of the older tau-b measure.
- Available over long historical periods from standard alliance datasets.
- Operationalizes shared interests and the utilities used in expected-utility models of war.
- Alliances are an imperfect and partial signal of underlying interests.
- Results can be sensitive to coding of alliance types and to capability weighting.
- Portfolios with few alliances yield unstable or uninformative similarity scores.
- The measure is correlational and does not establish that similarity causes cooperation.
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
Why is the S-score preferred over Kendall's tau-b?
Signorino and Ritter showed that tau-b, a rank-correlation statistic, conflates the similarity of two portfolios with the marginal distribution of alliance categories and can return counterintuitive values — for instance, sensitive to how many ties of each type exist regardless of whether the same partners are involved. The S-score measures the actual distance between two portfolios across partners on a clean −1 to +1 scale and avoids these artifacts, making it a more valid measure of alignment.
Should the S-score be weighted by capabilities?
It depends on the theory. Unweighted S treats every state's alliance the same; capability-weighted S gives more importance to agreement (or disagreement) over ties with powerful states, reflecting the idea that aligning with a great power matters more than aligning with a minor one. The choice should follow from whether your argument is about formal alignment per se or about strategically consequential alignment.
How does alliance portfolio similarity relate to foreign-policy similarity scores?
Alliance portfolios are one input to the broader concept of foreign-policy similarity. The same S-score machinery is applied to other revealed positions — most prominently UN General Assembly voting — to measure overall foreign-policy affinity. Alliance portfolio similarity is the security-specific version; foreign-policy similarity scores generalize the approach across issue domains and data sources.
Sources
- 1.Signorino, C. S., & Ritter, J. M. (1999). Tau-b or not Tau-b: Measuring the similarity of foreign policy positions. International Studies Quarterly, 43(1), 115–144.
You have read it. What now?
Cite this page
ScholarGate. (2026, June 22). Alliance Portfolio Similarity. ScholarGate. https://scholargate.app/international-relations/alliance-portfolio-similarity