We can identify three social classes characterized by their posterior mean income (computed by averaging the transformation of the draws View the MathML source) given in Table 5. The first group of the mixture, those with the lowest mean income represents 30% of the sample on average. The second group represents the highest proportion of the population, 68% on average and has a much higher income, representing roughly twice of that of the poorest group. The last group, having the highest mean income, is much smaller, 2% on average of the whole sample. These posterior characteristics give arguments to interpret our mixture model as representing a heterogeneous population where each group corresponds to a particular income group. Of course, this is only an interpretation. The mean income of the first group increased very slowly over 1979–1992 and had a sharp increase in 1996. The income rise was more regular for the second group while the third group experienced a huge increase in income over 1979–1992, with a sharp decrease in 1996.