Institutional distance and bilateral trade: does the rule of law substitute for trade agreements?

When two trading partners differ sharply in governance quality, can a regional trade agreement compensate for the resulting friction, or does it require a baseline of institutional compatibility to function? We construct a bilateral institutional distance measure fromWorld Governance Indicators (Rule of Law) and interact it with RTA status on a directed panel of 111 countries and 302,186 pair-year observations spanning 1996 to 2023. Estimation relies on Poisson pseudo-maximum likelihood with exporter-year, importer-year, and pair fixed effects, following the structural gravity specification of Yotov et al. (2016). The interaction coefficient is −0.072 (p = 0.004), indicating that each unit of institutional distance reduces the trade-creating effect of an agreement by roughly 7 percentage points. At zero distance the agreement raises trade by 20.3%; at a distance of 2.6 on the WGI scale the gain vanishes entirely. Quartile decomposition reveals a monotone gradient: the penalty concentrates in the third (−0.101, praw = 0.015, pHolm = 0.031) and fourth (−0.124, praw = 0.005, pHolm = 0.015) quartiles. The mechanism takes time. In the first five years after agreement entry the interaction is statistically indistinguishable from zero. It emerges at years five through nine (−0.079, p = 0.003) and peaks beyond fifteen years (−0.145, p < 0.001). Narrowing the treatment to FTAs eliminates the interaction, suggesting that the complementarity channel operates through non-tariff provisions—dispute resolution, customs cooperation, regulatory mutual recognition—rather than tariff reductions. A clean placebo lead on the interaction (p = 0.90), stability across six alternative specifications, and leave-one-out analysis confirm the finding. The design of trade agreements matters more than their existence when partners sit on different institutional ground. JEL Classification: F13 , F14 , F15 , C23 , P48