Federal Reserve Board research finds Ideal Point Distance specifications materially change measured trade and financial fragmentation

In a FEDS Note, Federal Reserve Board staff revisit the Ideal Point Distance (IPD) measure of geopolitical distance derived from United Nations General Assembly voting and show that methodological choices can materially affect conclusions about geoeconomic fragmentation. Across alternative IPD definitions, the analysis finds evidence of post-2022 fragmentation in goods trade along geopolitical lines, while the portfolio-investment results are weaker and more sensitive to how geopolitical alignment is measured. The note compares a baseline IPD estimated on all votes from 1946–2023 with distances measured in 2021 against alternatives using 2023 distances, economic votes only from 1971 onward, and an estimation window starting after 1990. When countries are grouped into U.S.-leaning, China-leaning, and nonaligned blocs, 48 percent of countries change blocs under the 2023 IPD, 31 percent under the economic-votes IPD, and 2 percent under the post-1990 IPD. In gravity regressions with a post-invasion indicator for 2022–2023 using UN Comtrade trade data for 2001–2023, trade between countries in different blocs is estimated to be 11.8 percent lower than within-bloc trade under the baseline IPD, 24.1 percent lower using the 2023 IPD, and 9 percent lower using economic-votes IPDs, while pairs involving nonaligned economies are not statistically different from within-bloc trade. Using IMF Coordinated Portfolio Investment Survey data for 2015s1–2023s2, portfolio holdings show only limited evidence of fragmentation, including an estimated post-2022 decline between opposing blocs of about 0.3 percentage points in the baseline and post-1990 specifications, with less robust results under the 2023 and economic-votes IPDs; a complementary specification using bilateral IPD directly also indicates negative post-2022 effects for both trade and portfolio holdings.