Heteroskedasticity Robust Standard Errors
When the assumption of homoskedasticity is violated, OLS remains unbiased and
consistent, but the standard errors are biased and we cannot rely on the usual t statistics, F
statistics and LM statistics for drawing
Day 14: 05/07/14
Class Notes
How does microcredit work?
2) Use of regular and frequent payments
o Rigid, frequent repayment schedule provides fiscal discipline to borrowers
o Weekly meetings with bank officers may increase trust, willingness to stay on
t
Day 12: 05/01/2014
Think Again: Brain Drain
Brain Drain into the United States
However, brain drain actually benefits both countries
For every emigrant stolen there are others who take advantage of the situation
o It can cause an increase in skill acqui
05/15/2014
Class Notes
TopDown View
Variation in public good access cannot entirely be explained by proxies for important
factors in the bottomup model
Bottomup models predict very slow changes in public goods in most situations
Do Leaders Matter? Ma
4/23/2014
Day 9: Land Markets
DE Ch. 11: Markets in Agriculture: Overview
Imperfect Market Cases
o Information: Unobserved Actions
Contracts must be suitably limited: incomplete market here stems from lack of
observability
Problem of unobserved actions
4/29/2014 Land Markets
So far
Inverse farm sizeproductivity relationship
Seems attributable to incentive issues
o Contracting issues in hired labor
o Sharecropping inefficient
Why not fixed effects
Implications
Larger farms less productive
This woul
Day 13: 05/06/14
Class Notes
Credit Markets
Why does credit matter?
Finance investment
Finance working capital
o Shortterm loans for buying foods on wholesale for retail businesses
Smooth consumption
Neoclassical model of the capital market
Everyone
4/17/2014 Child Labor, Education, & Human Capital
Economics 24
What is child labor?
Making your kid fold blanket a form of child labor?
Politicized definition: work that is bad for children
o bad depends on what children would do if they were not workin
Econ 24 Stats continued
Special Issues with log Dependent Variables
When the dependent variable is in logs, there are a few extra things to remember besides
interpreting coefficients in percentage terms. First, taking the exponent of the predicted
ln(y) w
Econ 24
Stat concepts
Linear Probability Models
When the binary variable is the dependent variable, we call that a linear probability
model. The linear probability model will violate the assumption of homoskedasticity, but
is often a good place to start.
Econ 24
Using Dummy Variables
Dummy variables are derived from categorical variables you need one less dummy
variable than there are categories. A dummy independent variable can be thought of as
shifting the intercept, that is as estimating a separate int
Econ 24
Statistical Testing Multiple Linear Restrictions
In order to test multiple linear restrictions, or exclusion restrictions, we have to perform
an F test. The idea is to see if the change in SSR from restricting the model is big
enough to reject the
Econ 24
Notes for stat
Classical statistical testing chooses a significance level, , first say 5% meaning there
will only be a 5% chance of rejecting the null if it is really true. That means we will
reject the null if our estimate is beyond 5% of the dis
Econ 24
Stat concepts we need to know
Statistical Testing ttests
Now that we have a standard error for our OLS estimate, we can use it for statistical
testing. Unless we have large samples, and can rely on asymptotic normality, we must
add yet another as
One of Nepals top developmental priorities in the next five years must be the
full realization of its hydro capabilities.
CURRENTLY

Our countrys hydropotential is 83,000 MW.
 Which would make us one of the leading producers of hydropower in the
world.
EFFECTS

Only 15% of our country currently has access to electrical power supply
 Through realization of our Hydropotential this figure would vastly increase and
increase the standard of living for our people

Revenue for such projects could be raised