Table 3. Robustness of terms (surrogate data) for democracy (D) and log GDP per capita (G).

The three columns furthest to the left shows the most eight most frequently recurring terms among the top 100 models for (D) and (G) for 1000 generated surrogate data sets. The columns to the right of show how often the terms appear in combination to each other. Red bars means a positive sign on the term and blue bars negative.

Democracy
Frequency per term			Frequency of combination
Term in model	Procent		DG	1/(1 + D)3	−D/(1 + G)	D	−1/(1 + D)
−D3	0.75		0.50	0.41	0.23	0.19	0.19
DG	0.62			0.39	0.05	0.11	0.19
1/(1 + D)3	0.56				0.15	0.11	0.16
−D/(1 + G)	0.35					0.14	0.08
D	0.30						0.07
−1/(1 + D)	0.28
Log GDP per capita
Frequency per term			Frequency of combination
Term in model	Procent		D2	D/(1 + G)	Const	1/(1 + D)3	1/(1 + D)2
D	0.29		0.01	0.01	0.02	0.08	0.07
D2	0.27			0.02	0.06	0.02	0.03
D/(1 + G)	0.24				0.01	0.07	0.05
Const	0.22					0.02	0.01
1/(1 + D)3	0.20						0.00
1/(1 + D)2	0.18