Table 2a.
Robustness Checks for the Cross-Country Analysis
| (1) | (2) | (3) | (4) | |
|---|---|---|---|---|
|
| ||||
| VARIABLES | Negative Binomial | OLS | OLS | OLS |
| Number of Languages | Log Number of Languages | |||
| Variation in Elevation | 1.313*** (0.395) | 1.045** (0.473) | 1.309*** (0.445) | |
| Variation in Land Quality | 3.252*** (0.952) | |||
| Dispersion of Elevation | 0.429*** (0.130) | |||
| Dispersion of Land Quality | 1.315*** (0.399) | |||
| Variation in Climatic Suitability | 2.505*** (0.749) | |||
| Mean Climatic Suitability | 0.661* (0.349) | |||
| Variation in Soil Suitability | 3.785*** (1.324) | |||
| Mean Soil Suitability | 0.653 (0.474) | |||
|
| ||||
| Continental FE | Y | Y | Y | Y |
| Observations | 142 | 142 | 142 | 142 |
| R2 | . | 0.73 | 0.73 | 0.73 |
| Log pseudolikelihood | −536.365 | . | . | . |
Nonstandardized coefficients reported, with robust standard errors in parentheses
***
p < 0.01,
**
p < 0.05,
*
p < 0.1
All specifications include the same controls as those of column 5 in Table 1. Column (1) is estimated using a negative binomial and the dependent variable is the number of languages. Columns (2)–(4) use as dependent variable the ln Number of Languages. Column (2) uses the dispersion in elevation and land quality to capture geographic variability. (3) and (4) use the climatic and soil components of land quality, respectively.