You know how some households have a chore chart on the fridge that everyone actually follows, while other households have the same chart but it's decorative — nobody checks, nobody enforces, the dishes pile up anyway? The difference isn't the chart. It's whether the people in the house believe they personally have the standing to call out a violation. That's the core mechanism here: constitutions are chore charts. The question is whether the culture produces people who enforce them. The committed claim: individualistic national culture causes higher constitutional compliance, and this compliance channel is a previously unidentified transmission mechanism from culture to long-term economic development. The authors study 115 countries cross-sectionally, using novel indicators of constitutional compliance — not just whether a constitution exists, but whether its provisions are actually respected in practice. They find that Hofstede-style individualism scores predict compliance levels, and the result survives instrumental variables estimation. This sits inside a long-running fight in institutional economics. One camp (North, Acemoglu, Robinson) says formal institutions — constitutions, property rights, courts — drive development. Another camp (Greif, Tabellini, Alesina) says informal institutions — culture, norms, trust — are what make formal institutions work or fail. This paper lands firmly on the informal side, but with a specific mechanism: individualistic cultures produce citizens who treat constitutional constraints as personal entitlements worth enforcing, rather than aspirational text written by elites. The paper explicitly frames this as supporting the 'embeddedness' thesis — that formal rules only bind when the informal environment makes them credible. The architecture is standard cross-country regression with instrumental variables. The IV strategy is the load-bearing wall here. Culture is obviously endogenous to institutions, so the authors need instruments that predict individualism but don't directly cause constitutional compliance. The abstract doesn't name the specific instruments, which is a gap for evaluation — but IV estimation at least signals awareness of the endogeneity problem. The sample of 115 countries is respectable for this type of work, covering most of the world's constitutional regimes. The integrity picture is mixed. Cross-country regressions are the workhorse of comparative political economy, but they're also where omitted variable bias goes to hide. 115 observations is a decent N for cross-country work but tiny by ML standards — every additional control variable eats statistical power fast. The religion finding — that Muslim-majority countries show lower compliance, consistent with prior work attributing development gaps to institutional quality — is the kind of result that could reflect genuine institutional dynamics or could reflect omitted variables correlated with both religion and colonial history. The paper's acceptance at the European Economic Review (conditionally) is a meaningful quality signal; this is a top-field journal with serious refereeing. The milestone question is tricky because this is a social science contribution, not an engineering paper with a next-number target. The real unlock would be panel data — tracking countries over time as cultural values shift (and they do shift, slowly) to see if constitutional compliance follows. Several countries are running natural experiments right now: Hungary, Poland, Turkey, and India have all seen constitutional norms erode under populist governments. If individualism scores in those countries predict the speed or depth of erosion, that would be strong out-of-sample confirmation. The obvious experiment the authors didn't run: within-country variation. If individualism drives compliance, regions within federal systems (U.S. states, German Länder, Indian states) with varying individualism levels should show varying compliance with sub-national constitutions. My read on why they didn't run it: (a) sub-national constitutional compliance indicators don't exist in standardized form, and building them is a separate multi-year project. This is a data availability constraint, not an evasion.
