Atkinson

A detailed Summary of Atkinson

Many measures of inequality are used today to show how income/resources are unevenly distributed. The problem arises, however, because each measure of inequality is structured in such a way that does not give equal consideration to all parties being measured or contains other biases. One example would be the Gini coefficient, and how its measure is weighted heavily regarding the middle income distribution class1. Other measures of inequality also have unequal weighting favoring different members of the income hierarchy. By using a certain measure of inequality, our focus is biased by the measure's underlying weightings. Thus, it can be argued that we are making value judgements regarding who is considered more significant when measuring inequality.

Aware of these "hidden" value judgements, Atkinson devised a measure that tried to be more value-free. This paper takes a look at Atkinson's measure of inequality and whether or not it is truly value-free, or perhaps even more value laden than other measures.

Atkinson noted that the problem with current measures of inequality was that they each had a weighting scheme that favored different distributions of the income ladder. When someone uses a certain measure, they are-knowingly or

In my opinion, Atkinson's index is much more value-laden than other inequality measurements because there is an implied goal. This goal would correspond to making poorer people more equal with other distributions. Other indices may apply weight to certain distributions, but there is not a definite path for us to ascertain. For example, even though the Gini gives middle distributions more weighting, the Gini does not state that distribution inherent of middle distributions is what we should strive for. The increased weighting is only a result of the fact that more people incorporate the middle distributions. The Atkinson, on the other hand, increases emphasis transfers to the poor for all å other than zero.

unknowingly-giving more consideration to a certain income distribution. By doing so, they are, perhaps, stating that they think a certain income distribution is more important. Atkinson criticized this, and his solution was to make value judgements explicitly, rather than implicitly. But before a closer look at the Atkinson index, let's discuss the value judgement components of a few other inequality measurements.

The Gini coefficient, based on the Lorenz curve, is probably one of the most widely used indices for measuring inequality. It has many advantages over the above measurements such as avoiding emphasis on mean income. Instead it considers differences between every pair of incomes. A major problem when using the Gini measurement arises when comparing different Gini coefficients corresponding to different Lorenz curves. If the Lorenz curves do not cross, then we can conclude higher inequality with higher Gini coefficients4. However, if the Lorenz curves cross, then it is ambiguous as to which Lorenz curve represents higher inequality. Additional value judgements must be made to determine which Lorenz curve has more inequality. Thus, value judgements are unavoidable when using the Gini because Lorenz curves will cross from time to time. Furthermore, the Gini is disadvantaged because weighting given to certain distributions is arbitrary. Gini coefficients change when the number of distributions changes5. So if everyone's allocation remains the same, and we define income distributions differently, the Gini will change also. Value judgement must be made when considering how to define different distributions. Also, individual Gini measurements for each distribution do not add up to

Some common words found in the essay are:
, Furthermore Gini, Gini Gini, value judgement, measure inequality, mean income, atkinson index, value judgements, poorest person, Gini Lorenz, income distribution, income distributions, lorenz curves, richest person, lorenz curves cross, person poorest person, middle income distributions, social welfare function, current welfare welfare,

Approximate Word count = 1624
Approximate Pages = 6 (250 words per page double spaced)

Category: Miscellaneous