Trade openness, GDP scale, and population size across 30 countries in a balanced 2010–2023 panel.
The GDP-only model shows a pronounced negative association between log GDP and trade openness: β = -24.49 with a 95% confidence interval of [-29.28, -19.70]. In other words, larger economies in this panel tend to trade less relative to the size of their own GDP.
When population is added, the GDP coefficient shrinks to -0.92 and its interval spans zero, while population remains materially negative at -20.69. Within these specifications, aggregate GDP does not dominate the trade-openness relationship once country size is represented another way.
The full model adds GDP per capita, where the positive β = +6.59 sits alongside a still-negative GDP coefficient (-7.51) and population coefficient (-14.10). That pattern is more consistent with openness being linked to economic intensity than to sheer scale alone.
| View | Catalog family | Commonality | Why it fits this task |
|---|---|---|---|
| Coefficient chart | Effect size / uncertainty comparison | 4 / 5 | Grouped estimates with confidence intervals make it easy to compare how GDP and population coefficients move across model specifications. |
| Trade vs log GDP scatter | Relationship / correlation | 5 / 5 | A scatter plot with fitted line is the canonical way to inspect whether the bivariate pattern visible in the panel matches the regression slope. |
| Country | Avg. trade openness | Avg. log GDP | Latest year |
|---|
| Model | Term | Estimate | 95% CI |
|---|