3 - Inequality and Development Flashcards
4 principles for being a good measure of inequality:
- Anonymity principle - it doesn’t matter who is earning the income.
- Population principle - how many people living in the country is irrelevant, it is the percentage of the pop earning the different levels of income that matters.
- Relative income principle - don’t measure the absolute amount of money but the relative. Inequality in a poor country can be same as in a rich, measure differences within country.
- Dalton principle - A regressive transfer lead to an increase in ineq. The rich takes from the poor.
Is Dalton principle valid the other way around?
No, Dalton isn’t valid for progressive transfers, i.e when rich gives to the poor, the inequality will not decrease.
5 measures of inequality:
- The Lorenz Curve
- Coefficient of Variation
- The GIni Coefficient
- The Theil Entropy Index
- Income shares and Kuznets ratios
Lorenz curve
Graphs the relationship between cumulative percentage of HH (x) to the cumulative percentage of HH income (y). This is typically bent away from the 45 degree line of perfect equality. The further away, the more inequalities.
How to do Lorenz step by step (7):
- Arrange HH income from lowest to highest.
- Calculate total income
- Calculate the fraction of income earned by each population decile.
- Calculate the cumulative fraction earned by the deciles.
- Set out the perfect equality line.
- Plot the Lorenz with the values you got. The more bowed the more unequal.
- The Gini coefficient is the area between the Lorenz and perfect line divided by the total area under the perfect line. (Easier to compare across time and space).
Coefficient of Variation
CV=sd/my
The standard deviation of expenditure divided by the average expenditure in the population.
Gini Coefficient
G= A/(A+B) = (2/nu)*cov(expenditure, rank)
Sooo either we can measure from the Lorenz where we take the area of A divided by A+B. Or we divide 2 by n*mean and them multiply by the covariance between a person’s expenditure and income rank.
If G=1 we have max inequality and with G=0 there is perfect equality.
Does Gini satisfy our four principles?
- Anonymity - yes
- Population - yes (if you double the population and replicate the same expenditures, you see that G remains the same)
- Relative - yes. If multiplying with some constant it should give the same result, same relative relation.
- Dalton - yes. Can also be tested by transfer money from low to high.
Theil Index
A measure that runs from 0 that is perfect equality to ln(n)= max inequality.
This measurement is additively decomposable and therefore we can divide into subgroups and compare inequality between groups.
Income shares
Another very common measure where we for ex calculate how much of the income that is held by the top 10% or the bottom 10%. Proportions.
Decomposability
Means that you can divide the pop into subgroups and calculate inequalities within and between these groups, when adding the within and between inequalities,w e get the total of the population.
–> help us focus policies on the right type. State-level or more local interventions.
What is Piketty’s main point about why inequalities have increased?
That r>g always. The return is always greater than the growth, or have almost always been at least. So the ones already owning a lot of assets, or are born rich will continue becoming even richer while the poorer have hard to reach the same speed since they have hard financing assets and earn the r.
Piketty’s solution to inequalities:
Must have global taxes on capital, otherwise you can place your assets in countries with low taxes….
Can growth create inequalities?
No, evidences show that on average will growth not increase/reduce inequalities. But initial inequality matters - low from beginning is good bc high ineq will grow larger with growth.
Channels through which inequalities impact growth (8):
- Aggregate saving
- Market failures
- Social and political instability
- Political economy and Fiscal policy
- Access to financial capital
- Education
- Incentives
- Domestic market size