Foreign Policy Alignment in Asymmetric Alliances

Alignment scores over time, by alliance type

Average alignment score by year, grouped by each country's alliance relationship with the United States that year: NATO, the inter-American treaty system (Rio Pact / OAS), other bilateral U.S. allies (e.g. Japan, South Korea, Australia, the Philippines), or no alliance.

View underlying yearly averages as a table

Source. Alignment scores are posterior means from the paper's latent measurement model; see the published paper (linked above) for full data and methodological details. Alliance status is coded year-by-year from ATOP v5.0 (Leeds et al. 2002), classifying each country-year by its most encompassing active pact with the United States: NATO (ATOP 3180) > Rio Pact / OAS (ATOP 3075, 3150) > any other bilateral or multilateral U.S. defense pact > no alliance with the U.S.

Who aligned most closely with the U.S., year by year?

Countries ranked by posterior alignment score in a single year. Step through 1950–2010 to see how the ranking changes.

1950

#	Country	Alliance status	Alignment score

Alignment trajectories: rank and score over time

Pick countries from the dropdown below to compare two views of the same trajectories, 1950–2010. Posterior mean alignment score with 95% credible interval (upper chart) shows the underlying score itself. Rank over time (lower chart) places each country in that year's alignment order — rank 1 is most closely aligned with the U.S. that year. Each chart also carries its own alliance-membership strip along its bottom axis, one row per selected country. Both charts share one country selection below them — add or remove a country once and both update together. Hover a point for exact values; click a chip to isolate that country's line (and its strip row) in both charts at once, click again to release; use × to remove it from both.

Reading this. Each shaded band is the 95% credible interval from the posterior distribution of the latent alignment measure for that country; the solid line inside it is the posterior mean. As in the rank chart below, each country's line gets a shade of its alliance-category color plus its own dash pattern, reassigned by position among the countries currently selected, so two lines in the same category stay distinguishable; each band is tinted to match its own line. Bands are drawn at a reduced, count-scaled opacity so a handful of overlapping intervals stay legible together — as with the rank chart below, this works best with a handful of countries at once (roughly 8–10 before color/dash combinations start repeating). Each selected country gets its own strip along the bottom axis, marked at its left edge with a swatch in that country's line color and dash pattern; each strip marks years that country held NATO (blue), Rio Pact/OAS (orange), or another U.S. defense-pact membership (aqua), with no color meaning no U.S. alliance that year. As with the rank chart below, Germany's line merges West Germany (1955–1990) and unified Germany (1990–2010) into one continuous trajectory.

Reading this. Rank is computed year-by-year from the same posterior mean alignment scores as Figure 1 and Table 1 (lower rank number = more aligned that year); it is not smoothed. The axis always spans the full range of the countries currently plotted, so no line is ever clipped. Each line's color is a shade of its alliance category (the small dot on each chip above shows the plain category color), and each line also gets its own dash pattern — both are reassigned by position among the countries currently selected, so two lines in the same category (say, two NATO members) never look identical. This works best with a handful of countries at once; past roughly 8–10 simultaneous lines, color/dash combinations start repeating and hovering becomes the reliable way to tell lines apart. Germany's line merges West Germany (COW 260, 1955–1990) and unified Germany (COW 255, 1990–2010) into one continuous trajectory. All 200 countries in the dataset are available in the dropdown above. As in the score chart above, each selected country also gets its own strip along the bottom axis, marked at its left edge with a swatch in that country's line color and dash pattern; each strip marks years that country held NATO (blue), Rio Pact/OAS (orange), or another U.S. defense-pact membership (aqua), with no color meaning no U.S. alliance that year.

How input variables contribute to a country's alignment score

Ten input variables underlying the alignment score, grouped into economic, military presence, non-proliferation, and coalition/diplomatic categories. Pick a country and a year to see each variable's value and its percentile among other countries that year.

1980

Reading the bars. Each track spans the 0th–100th percentile of that variable among all countries with data in the selected year; the marker and fill show where this country-year falls. Binary variables (treaty in force, weapons pursuit, coalition participation) collapse to two clusters — a country with the variable "on" always outranks every country with it "off" that year. Some variables (joint military exercises, sanctions coalitions, chemical/biological weapons pursuit) have gaps in coverage; "no data" means that variable isn't observed for this country-year, not that the true value is zero. Source. These are the input variables underlying this paper's latent alignment measure; see the published paper (linked above) for full data and methodological details.

Alignment scores before and after NATO accession

Pooled alignment scores in the years around NATO accession, for every country that joined after 1949 with data on both sides (Greece/Turkey 1951–52, West Germany 1954, Spain 1981, post-1997 enlargements). Year 0 is the accession year.

Pooled event study. Each point averages the posterior mean alignment score across all accession cases at that many years before/after joining NATO (line thickness of evidence varies; see n in tooltip).

Average alignment score in the 10 years before (hollow) vs. after (filled) accession, per country.

Major-power support signals alongside U.S. alignment scores

McManus & Nieman (2019) estimate how strongly a major power (the U.S., Russia, or China) signals security support to a country in a given year. Each point is one country: the x-axis is that power's signaled support, the y-axis is the country's own alignment score with the U.S.

1980

Source. Major-power support signals: McManus & Nieman (2019) (Major Protégé Dataset v1.1; see References). Values are posterior means on that paper's own latent scale — comparable across countries within a year, but not in the same units as this paper's alignment score, so read the chart for pattern (correlation), not for matching magnitudes. r is the Pearson correlation between the two measures among countries with both values in the selected year.