31474179
Description
Flashcards by Linnéa Ekberg, updated more than 1 year ago
|
Created by Linnéa Ekberg
over 4 years ago
|
|
| Question | Answer |
| Why is demand estimation useful? | Estimate mark-ups (market power) Merger analysis (degree of substitution) Define relevant market |
| Define the relevant market | All products that are interchangeable or substitutable in terms of price, characteristics, or regular use |
| Which relevant markets need to be defined? | Product market Geographical market |
| What kind of substitutability is there? | Demand and supply |
| What factors affect elasticity? | Existence of substitutes Substitutability Transparency Quality perception |
| What is the uniproduct monopoly markup? | (p-mc/p) = -1/elasticity |
| What is the cournot markup? | (p-mci/p) = -si/elasticity si: market share |
| What is the average markup? | sum(-si^2/elasticity) = -HHI/elasticity |
| What is the formula for concentration in the market? | sum(si^2) |
| Why is there a problem with endogeneity in demand estimation? | Price and quantity are interdependent |
| How can the endogeneity problem be solved? | Instruments in form of demand and supply shifters |
| Give examples of demand and supply shifters | Demand: income change supply: entry, capacity expansion |
| What makes a good instrument? | It is correlated with the price but not with the error term |
| How can preference parameters be estimated for the demand side? | Multinomial logit using random utility model by McFadden |
| What does a simple logit model give? | The probability of success of failure |
| How is a coefficient from a logit model interpreted? | The change in the log odds of the outcome for a one unit increase in the predictor variable |
| What are the benefits of using a simple logit model? | Easy to implement No assumption about distributions |
| What are the disadvantages of using a simple logit model? | Independent irrelevant alternatives |
| How can the problem with independent irrelevant alternatives be sovled? | Nested logit model Random coefficient model |
| What is the formula for a logit model? | F(z) = exp(z) / (1+exp(z)) |
Want to create your own Flashcards for free with GoConqr? Learn more.