Common linked UN votes covered only 2002–2010 in the tested 1992–2019 comparison of five permanent Security Council members

We tested whether linked UN resolutions could support a direct comparison of voting agreement across 1992–2005 and 2006–2019. We used all ten pairs among the five permanent Security Council members. The inputs were UN voting data version 30.0 and one fixed author file of 955 resolution links. After flag exclusions, 2,102 roll calls remained. Only 19–20 linked families had valid votes in both periods. Their 93–95 roll calls all occurred in 2002–2010. The common-family sample therefore did not preserve the time coverage of the full sample. Russia–China agreement rose by 7.72 percentage points in all included votes, by 1.83 points when all votes were restricted to 2002–2010, and by 1.32 points in 19 common families. One family supplied the last gain. Other country pairs also changed in size or direction under the sample rules. The score gaps cannot be assigned to a full-period agenda effect. This finding concerns a direct common-family comparison, not dynamic models that connect years through different items. A separate exact calculation reproduced all 20,696 valid pair scores and all 50 comparison settings. The package contains input votes, links, date coverage, code, figures, results, source hashes, and reproduction instructions.

United Nations · voting agreement · resolution agenda · temporal coverage · repeated resolutions · measurement audit