TEST 1 + Final Flashcards
What does the derivative of a function represent?
slope of the tangent line to f(x) at specific point
Covariance is?
comovement between two variables
measures wheter they change is similar or opposite directions
p-value
describes the probability that in another similiar experiment you would observe the same outcome given H0 is true
descibe instrumental variable approach
when you have a seperate variable thats associated w/ the independent variables but not associated w/ dependent variable
Regression Dicontinuity approach
approach that uses a distinct cutoff to compare otherwise similar groups to identify causal effect of treatment/intervention
natural experiment difference in differences approach
approach studies differential effect of a treatment on a “treatment group” and a “control group”
what is omitted variable bias?
occurs when significant independent/explanatory variables of the dependent variable are not included in the model
2 different types of omitted variable bias
- independent variables that are uncorrelated with our already included independent variables (effects error term)
- independent variables that are correlated with our already included independent variables
(beta hat will be biased)
3 main categories of risk preference
- risk loving
- risk neutral
- risk adverse
breif overview of the planet money podcast, What Causes What
focuses on the distinction between correlation and causation
talks about real life examples life schooling, crime, and health
log-level model
estimates a constant percentage change relationship between y & x
log-level model might be good when we have a non-linear relationship between the dependent variable and the independent variables
Constant Relative Risk Aversion(CRRA)
what is value of w for each risk preference?
0 < w <= 1 (risk adverse)
w=0 (risk neutral)
-1 <= w < 0 (risk loving)
TR maximization
what is value of ED (elasticity of demand) when TR is maximized?
ED= -1
basically what is the approach for calculating comparitive statics ?
calculate EP*,x and EQ*,x
wrt some variable x that will be named, like the price of the other good
formula for Arrow Pratt measure of risk aversion?
r(I)=
r(I)= - U’’(I) / U’(I)
*the larger the number, the more risk adverse an individual or consumer group is relative to a group with a lower Arrow-Pratt risk aversion measure
What does the Arrow Pratt Measure of Risk Aversion tell us ?
consumers or companies risk preferences relative to eachother –> the larger the number, the more risk adverse an individual or consumer group is relative to a group with a lower Arrow-Pratt risk measure
what is the value based approach?
choose inputs that maximize outputs/revenues given a set of cost constraints
what is the cost based approach?
maximize profit by minimizing cost
do the value based and cost based approach yield different profit maximizing conditons?
no
what will HOD =
if there are constant returns to scale?
HOD=1
*if HOD=5, production will increase by 5 units for every unit of input
income elasticity of demand
what are the elasticity values for inferior,normal, and luxury goods
inferior goods: Eq,i < 0 as i increases,QD decreases
normal goods: Eq,i > 0 as i increases, QD increases
luxury goods: Eq,i > 1 as i increases, QD increases at a higher rate
cross price elasticity of demand
what are the elasticity values of goods that are complements, unrelated, and substitutes ?
Complements: EQc,Pk < 0 as Pk increases, QD decreases
unrelated: EQc,Pk = 0 as Pk increases, no change in Qc
substitutes: EQc,Pk > 0 as Pk increases, Qc increases
true or false
in certain portions of the classic production function, it is possible for quanity produced to decrease as additional inputs are needed
true
think… w/ too many people working, output could be decreased
market case
under market case, producers objective is to maximize profits, what is the condition at which this objective is satisfied?
slope of π = slope of PPF
or
MRPT= - P2 / P1
mathmatically:
max π = P1Q1 + P2Q2
s.t. Q1 = f(Q2) - PPF
elasticity curve scenarios
should you increase or decrease price to increase revenue while:
- operating on inelastic portion of demand curve
- operating on elastic portion of demand curve
inelastic portion of demand curve: increase price to increase revenue
elastic portion of demand curve: decrease price to increase revenue
elasticity values
define values for:
- perfect inelastic
- inelastic
- perfectly elastic
- elastic
perfect inelastic: |ED| = 0
inelastic: 0 < |ED| < 1
perfectly elastic: |ED| = 1
elastic: 1 < |ED|
exponential demand functions
for exponential demand functions, ED is __________
constant
factors the effect price ED
what are the four factors that affect price ED
- substitutibility of other goods –> greater substitutibility will make the demand curve more elastic
- product durability–> demand for more durable products is more elastic
- time period–> some demand curves are more inelastic in short run and more elastic in long run
supply curve shifters
what are examples of supply curve shifters:
- input prices
- technology and production efficiency
- size of market
- production conditions ( eg weather)
IV
key assumption of IV:
parrallel trends assumption, observe parallel trends between treatment/control group in the pre period
IV
describe approach:
have seperate variable thats associated w/ independent/explanatory variable but not associated w/ dep variable
RD
describe approach:
utilizeds distinct cut offs or thresholds that determine the assignment of treatment to determine causal effects of the intervention/treatment
covariance
Definition:
tells us the comovement of two variables
>0 positive relationship
<0 negative relationship
correlation
defn:
tells the magnitude and direction relationship of two variables
ranges between -1 and 1
economic signifigance
what does economic signifigance mean?
magnitude
eg: percent increase in ticket sales given a one unit increase in the specified independent variable
Poisson Model
what scenario is a Poisson model useful?
when dealing with count variables as our dependent variable, especially when count variable has a lot of zeros
log-level model
what scenario is a log-level model useful?
when the dependent and independent variables have a non-linear relationship
*it assumes constant percentage relationship
natural experiment DD model
defn:
strategy used when we observe control and treatment groups in real life, and then compare to determine causal relationships between dependent and independent variable
