Socio-Environmental Development Index
for the State of Bahía
Leonardo Araújo, Elaine Fernandes and Patrícia Rosado

Preparation of the idsa index includes the values estimated for the eci. This was also undertaken by factor analysis principal component method7 for the indicators of Gross School Attendance Rate; Income per capita; Life Expectancy at birth and eci .

The analysis undertaken resulted in two factors which succeeded in explaining the total variance of the data in the order of 66%.

According to Table 4, Factor 1 (f1) is directly related to the eci, Income Per Capita and the Gross School Attendance Rate. This precisely demonstrates the relationship previously touched upon that localities with more purchasing power exert greater pressure on the environment, reflected as a consequence in a low eci.

Factor 2 has a strong, direct relationship with the Life Expectancy at Birth indicator, a measure of health conditions in general. The similarities of this factor show that 88% of the variations of this indicator are explained by Factor 2.

The municipalities with the highest idsa were Salvador, Lauro de Freitas, Itabuna, Madre de Deus and Paulo Afonso. Table 5 shows the values of the ß referring to Equation 2, for each of the variables. It can be seen that the indicators with the greatest weighting in this idsa calculation are precisely the socioeconomic indicators (life expectancy and income per capita) and the municipalities that rank among the first five in the idsa are the most economically developed in the State. In addition, all of these municipalities show an eci greater than 0.5.

The municipalities with the worst eci on the other hand (América, Dourada, Santa Brígida, Quijingue, Nordestina, Cansanção) with the exception of the School Rate, show other indicators inferior to those observed for the State average, highlighting once again the socioeconomic indicators, which in the case of these municipalities were 40% less on average than those of the State.

7 The Bartlett and kmo tests were carried out. For this model, the Bartlett test obtained a value equal to 139,906, the probability significant at 1%, rejecting the nule hypothesis that there is no correlation between the variables. The value obtained for the Kaiser-Meyer-Olkin test (kmo) was 0.626 indicating that the sample used for factor analysis is suitable.