Absolute income is a better predictor of coverage by skilled birth attendance than relative wealth quintiles in a multicountry analysis: comparison of 100 low- and middle-income countries

Table 1 Linear regression analyses to investigate how well relative quintiles, actual mean wealth index scores and absolute income (per quintile) predict SBA coverage (N = 1465 observations)

	SBA prevalence (coefficients expressed as percent point)
Analysis level	Cross-country analysis			Within country analysis
	Model 1	Model2	Model 3	Model 4	Model 5	Model 6	Model 7
Asset quintile 1	0 (reference) p < 0.001			0 (reference) p < 0.001			0 (reference) p = 0.139
Asset quintile 2	10.19 (1.05)			10.19 (1.17)			2.18 (3.93)
Asset quintile 3	18.66 (1.80)			18.66 (2.02)			5.36 (6.53)
Asset quintile 4	28.60 (2.23)			28.60 (2.49)			9.98 (9.23)
Asset quintile 5	40.04 (2.56)			40.04 (2.86)			11.79 (14.02)
Mean wealth scores		6.97 (1.82) P < 0.001			6.85 (3.29) P < 0.001
Log income^a			19.13 (1.24) p < 0,001			18.38 (1.31) p < 0.001	12.78 (6.13) p = 0.04
Survey specific intercepts	NO	NO	NO	YES	YES	YES	YES
R-squared	0.220	0.128	0.516	0.877	0.777	0,879	0,881

Robust standard errors in parentheses are clustered at the country level
^aIncome is expressed in 2011 purchasing power parity-adjusted international dollars. Model 1 and model 4: cross-country and within-country prediction of SBA coverage according to wealth quintiles. Model 2 and model 5: cross-country and within-country prediction of SBA coverage according to actual mean wealth scores. Model 3 and model 6: cross-country and within-country prediction of SBA coverage according to household income. Model 7: within-country prediction of SBA coverage according to wealth quintiles and household income

ISSN: 1471-2393