Week One:
The poverty line is a critical threshold of income, consumption or access to goods and services
below which individuals are considered to be poor. The idea is that an individual needs to be
able to acceptipally participate in a society, which means that they need to be able to fulfill some
absolute notion of the ability to function in a society.
1) How to measure the PL?
a) Define a minimum consumption basket and calculate the income needed to
obtain it. The big question is what do we add to the basket? Food? Shelter?
Clothes? Education? Healthcare? Etc.
2) Are we talking about relative or absolute poverty?
a) Sometimes we can talk about absolute, and politicians like this as people don’t
see the relativity when they do. Hence, we always need to make sure we are
clear about whether it is relative or absolute poverty.
3) Temporary or Chronic?
a) Both are important, so it’s important to make a distinction between the two!
4) Households or individuals?
a) Typically we talk about households, you ask how many kids/adults are in the
household and then you see how much the household needs. Then look at
household income and you can divide by how many people live in it!
In conclusion: PLs are always approximations to a threshold that is fuzzy, more because the
effects of sustained deprivation are often felt at a later point in time. Poverty is a deeper and
less quantifiable concept.
Our final definition of the PL is: an expenditure threshold that is regarded as minimally
necessary for adequate participation in economic life. The real question is though, how can we
measure this?
Head Count (HC) You find the number of individuals so that 𝑦𝑖 < 𝑝
Where 𝑦𝑖 is the expenditure of individual i
And p is the poverty line.
Head Count Ratio 𝐻𝐶
where n is the total population
𝑛
(HCR)
Problems with HCR:
(1) It does not show the degrees of poverty (Insensitivity), perhaps
all people are very close, or very far from the poverty line.
Depending on which one it is, the good policy to pursue would
be very different!
(2) Bias policy, this is when the policy favors those close to the
poverty line in order to decrease the poverty stats, as those
people are now just above it, but the extreme cases remain.
, Poverty Gap Ratio ∑𝑦𝑖<𝑝(𝑝−𝑦𝑖)
(PGR) ⇒ The numerator is the income needed to bring all people
𝑛·𝑚
to the poverty line and the denominator is the income of society,
where m is the mean income.
We divide by nm because it gives an idea of how large the gap is
relative to the resources available that can potentially close the gap.
PGR is the measure of resources required to eradicate poverty.
Problems with PGR:
(1) Misleading in ‘rich’ but highly unequal societies, where PGR
looks small because nm is so large due to the few who have a
lot/almost all the income.
Income Gap Ratio ∑𝑦𝑖<𝑝(𝑝−𝑦𝑖)
(IGR) ⇒ where the numerator is the same but now the
𝑝·𝐻𝐶
denominator is the total income needed to bring all the poor people to
the poverty line. IGR reduces the denominator of PGR
It captures the acuteness of poverty as it measures it relative to the
total income needed to make that poverty go away. Both IGR and
PGR are not susceptible to policy distortion as HC is!
Sadly, no measure for relative deprivation (inequality) amounts to the poor. This is becoming
more and more important!
In the slides he also goes over the flier problem (same as in the book) and we see that a
personalized problem and small one we feel like we can help, so we are willing to donate more,
a huge problem we feel small though, so donate less.
Two Views on Aid:
Jeff Sachs (Columbia) Sachs believes that countries are poor because of endemic
problems, and hence they are unproductive and have no money to
deal with these endemic problems (weather, geography, malaria
infested, etc.) Thus, the solution is to break this trap and we can do
this with foreign aid!
Bill Easterly (NYU) Aid didn’t prove to work is what he said. He says it prevents people
from searching for their own solutions and undermines local
institutions. He says the poverty trap doesn’t exist, when markets
are free and incentives are right, people can find the way to solve
the problems.
Now, who do we believe?? We need evidence!
The issue is, if we compare countries, we see that those who received aid did not grow faster
than the rest, but it’s also possible that aid stopped these countries from growing even slower
than they are now! What we are missing is the contrafactual! And because of that, we can only
make big speculations. In this course we will focus on concrete examples, rather than the
massive problem. To solve these, we will use Randomized control Trials (which we will see next
week).
, Clips Week Two:
As always, it is important to remember that correlation does not mean causation. The correlation
fallacy is the logical mistake of believing that because two events occurred together, there is a
cause and effect relationship. We have a causal impact if we have the counterfactual, an
example would be:
How much money an individual would have earned if they did not go to school.
Comparing individuals over time will not in most cases give a reliable estimate of a program’s
effectiveness because we will most likely have unobserved factors.
Examples of textbooks and grades:
When doing this experiment you need to compare those in the program to people who
are not in it but exogenous to whatever characteristics that can affect the outcome of the
problem. A.K.A we need a comparison group. These people, with some conditions would/should
get similar results as these in the treatment. Hence, random sampling=good. See clip 2 if you
want to understand the mathematical formula for it. But basically you just compare the groups
and see the difference between them.
Here is the formula for the initial method of looking at the differences.
Then here we add and subtract the counterfactual, and rearrange it so we can see the
treatment effect and the counterfactual. Now, this shows that it is possible to have this selection
bias.
Clips Week Three:
Key questions about RCTS:
What does the timeline (1) Listing: short census of the population
look like? (2) Baseline survey: typically a 60-90 minute survey (not
