Poverty, Inequality and Development Flashcards
(41 cards)
Measurement Issues
Mobility:
Two societies, two groups, first earn $3000 or $2000
but stuck in jobs for life. Second, $1000 and $4000 but
move between jobs every month.
* First seems more equal but actually second is.
How income is earned:
Wages vs. rents vs. profits
* Charity vs. self-esteem
* Which factors owned?
Inequality Measuure
Assigns a level of inequality to each income distribution.
Principles of Measurement: Anonymity Principle
It doesn’t matter who receives the income (unimportant whether person A receives X and B receives Y, or A receives Y and B receives X)
Thus, can always rank income distributions such that Y1≤Y2 ≤ Y3 ≤ …Yn, ie. poorest to richest, wihtout losing valuable information.
Principles of Measurement: Population Principle
-> although its not intuitive, economic inequality calculations do not consider the size of the population
-> The size of the population is irrelevant
- > If double the population, relative inequality remains the
same
-> mathematically, regardless of the size of the population the distribution can be normalised to yield percentages/shares that can then be meaningfully compared across countries regardless of their size.
-> comparison purposes
Principles of Measurement: Relative Income Principle
Income levels in themselves are, NOT important
Not absolute but rather relative incomes that matter
* I(Y1, Y2 ,Y3…Yn) = I(aY1, aY2, aY3 …aYn)
* So following income distributions over 2 people equivalent:
– [$15,000, $45,000] and [$1000, $3000]
– Thus assumed that utilities are proportional to income
Shortfalls in absolute income rather addressed when studying poverty measurement
CRITICISM:
you could have a really equal society where everyone is below the poverty line
Principles of Measurement: Dalton Principle
->progressive transfer (rich to poor): makes it better for the income distrubution, inequality decreases
-> reggressive transfer (poor to rich): take from poor and give to rich, inequality increases
if you make a regressive transfer, what you end up with after the transfer must be less than what you started with (distribution is more unequal).
if one income distribution can be achieved from another by constructing a series of regressive transfers, then the former distribution MUST be more unequal that the latter…
For every income dist, I(Y1, Y2…Yn) and transfer t > 0,
I(Y1, …, Yi, …, Yj, …, Yn) < I(Y1, …, Yi + t, …, Yj + t, …, Yn) whenever Yi<Yj
IE.
For all of these measures of inequality, the gini coefficient, understand it does not take into account size pop etc FINISH ^^^^
Income Inequality PPF graph
Lorenz Criterion
“If Lorenz curve lies at every point to the RHS of the
Lorenz curve of some other distribution, the former is
more unequal that the latter”
I(Y1, Y2, Y3…Yn) >= I(Z1, Z2, .. Zn)
Not a new principle!
Ineqaulity measure consistent with the lorenz criterion ONLY if it simultaneously satisfies all 4 principles.
Incorporates anonymity, population and relative principles because curve drops all information on income and population in magnitudes and retains only population shares.
Does this satisfy the Dalton principle? YES. The Lorenz criterion’s foundation is that a transfer of income from a poorer to a richer person shifts the Lorenz curve upwards, indicating a higher level of inequality. Therefore, any measure that is Lorenz-consistent, like the Gini coefficient, will also satisfy the Dalton principle.
Lorenze Curve (Issues)
-> Good graphical illustration. but
-> cannot make a inequality statements when lorenz curves cross
I.e. not a complete ranking of income distributions
Ideally would like single number with which to compare countries/distributions
IQ measures: Range (R)
take difference of richest and poorest, and divide by the mean
R = 1/mu * (ymax - y1)
Good because:
-> easy to find the riches person
-> only need gdp, pop, richest person and pooerst
Bad because:
-> doesnt say anything about the people in the middle
Does this satisfy the dalton principle? NO because if you take two incomes from inbetween, and they make a regressive transfer, the range does not change, even if overall inequality does change.
IQ Measures: Mean Absolute Deviation
Notion; Inequality proportional to the distance from the mean income
M =
1/(mu*n) * sum [ (nj) * |yj - mu|
-> accounts for entire income distribution, thus better than range
-> calculate distance of each income from mean, sum them, then divide by total income
-> Often does not satisfy the Dalton principle.
To see this, take two incomes, both either above the mean or below the mean. A regressive transfer between the two (so that both incomes remain either above or below the mean, as before) will leave the mean absolute deviation unchanged
IQ Measures: Coefficient of Variation C
Give’s more weight to large deviations from mean: squaring means incomes further away from mean are given greater influence in the calculation
SD = sqrt [[ sum[nj / n] * (yj - mu)^2]
C = SD / mu
i.e. only relative incomes matter
Does this meet dalton principle? YES
Any regressive transfer increases the square of the larger number more than the square of the smaller.
IQ Measures: Gini Coefficient (G)
- incorporates sum difference of ALL pairs of incomes and total differences.
- normalised by dividing the populations squared and mean income
- as each income pair is counted twice (once for each Y variable perspective), you divide by 2
Gini = exact ratio of the area between the Lorenz curve and 45deg line, to the area below the 45deg line.
1: total inequality
2: total equality
In theory, the Gini coefficient can exceed 100% in extreme situations. For example, when handling negative wealth or income, the figure can theoretically be higher than 1; in that case, the Lorenz curve would dip below the horizontal axis.
Positives: allows for comparison across countries very easily
Criticism: requires almost perfect information - extremely difficult calculation.
Doesn’t tell us about the shape of this area - whether inequality is greater at higher income or lower income. You can have the same gini coefficient for both these scenarios, but without the graph cannot understand the nature of the distribution.
The Gini coefficient is an overall measure of a distribution that may mask extreme inequalities between certain groups of the population.
Dalton principle? YES! Gini coefficient or Lorenz curve are designed to be sensitive to transfers from rich to poor and thus satisfy the Pigou-Dalton principle.
Why should we care about extreme inequality?
- (+) Econ efficiency
few ppl accessing loans etc., typically lower savings - undermines social stability
institutions - generally ‘unfair’
Summary of Inequality measures
Inequality Trends:
Share of Total Income going to top 1% in English Speaking counries:
Share of Total Income going to top 1% in English Speaking counries:
1920, (-)ing until 1979, and then (+)ing to about the same as 1920 by 2014
Gini Coefficient
Wealth Inequality in Australia
-> ppl in highest 20% of wealth
Poverty Measures: Poverty Line
A minimum level of acceptable economic participation in a given society at a given point in time.
Poverty Line: expenditure threshold below which people are poor, such that they cannot adequately participate economically
Context of poverty line is important.
- expenditure vs consumption
- -> what ppl spend vs actually use
- -> in dev nat’s, consumption is usually preferred as it is more stable and is generally a better reflection of standard of living than expenditure
- relative v absolute poverty
- temporary vs permanent
- -> temporary due to shocks, or seasonal migration for employment
- level of observation
- -> individual vs houseshold, national v global
- fuzzy estimates
- -> in countries where data is hard to collect or verify, fuzzy estimates are used to make rough calculations for the poverty line
- -> poverty lines are constructed - not laws. The are often based on limited or incomplete data, arbitrary cutoffs, and normative judements (‘ what counts as basic needs, can that change from country to country, how does that influence comparison ‘)
Poverty Line Calculation
Poverty Line: expenditure threshold below which
people are poor, such that they cannot adequately
participate economically.
* y = income/expenditure
* i = refers to individual
* p = poverty line
* m = mean income
* n = number individuals in society
National Poverty Line
Children Living in Poverty
Poverty Measure: Head Count/Head Count Ratio
Count the number of those below poverty line (Head
Count, HC):
- HC: Number: yi<p
Relative incidence need to divide by total population
(Head Count Ratio) - 𝐇𝐂𝐑 = HC / n
Positives:
* Easy to calculate/interpret, minimal data requirements
* Widely used
Negatives:
* Fails to account for the extent of poverty
Poverty Gap Ratio
Poverty Gap Ratio: “Average of income needed to
get a ‘poor’ person to the poverty line divided by
mean income”
* A measure of resources which could close gap
it’s a deeper poverty measure than the headcount ratio because it accounts for how far below the poverty line people are, not just whether they’re below it.
𝐏𝐆𝐑 = sum(p - yi) / n*m
- ‘Total Poverty Gap’ = numerator
- Dividing by economy-wide income = misleading
-> Disguises inequality
–> small numbers of ppl with very large incomes drives up the total mean income
–> dividing by this number dilutes the PGR as it appears small relative to national income.
–> Disguises poverty close to the poverty line
–> An economy with large numbers of ppl just below the poverty line may still have a small PGR, despite large levels of absolute poverty.
+ Better than HCR as it accounts for depth of poverty
+ Helpful for budgeting anti-poverty programs
- assumes perfect redistribution, where (1) every dollar needed to close the poverty gap is collected without waste or leakage, (2) that money is transferred directly to the right individuals, in the exact amount needed, and (3) there are no administrative costs, political resistance, corruption, or targeting failures.
