Hedonic Regressions

1
Hedonic Regressions

• Price indices, quality control, and valuation of
  public services

2
Some Theory

• Multiple characteristics of goods
• Ala carte or combination platter
• From Industrial Organization, think of bundling
• Contribution/value of specific attributes
• Value of crash safety in car price
• value of school quality in home price
• Equilibrium of Supply and Demand

3
Early Hedonics

• Frederick Waugh - Quality Factors Influencing
  Vegetable Prices (1927)
• Andrew Court - Hedonic Price Indexes with
  Automotive Examples (1939)
• Gregory Chow - Technological Change and the
  Demand for Computers (1967)

4
Formulate a Model

• What determines the Price of Asparagus?
• Anything that shifts supply or demand
• Characteristics of the vegetable
• Time of year
• Good year - weather, crop disease, insect
  infestation
• Tastes and preferences
• Income
• Population

5
Formulate a Model

• What relationships do you expect?
• Does price rise or fall with Green?
• Does it rise or fall with Nostalks?
• Does it rise or fall with Disperse?
• Are marginal effects of characteristics constant?
• Do marginal effects of some characteristics
  depend on other characteristics?

6
Waughs Data

• http//research.umbc.edu/coates/ec611/Waugh.txt
• Open Excel or Quattro
• Copy the data from the web browser
• Paste it into Excel and parse it into separate
  columns
• Save as a text file (tab delimited)
• Read into Stata using the insheet command

7
Waughs Data

• Determined price, size and color variation, and
  the number of stalks in bundles of asparagus
• Price is price of a bundle relative to daily
  average price
• controls for seasonal and day-to-day changes
• Size measured as the variation in the diameter of
  the stalks in the bundle

8
Waughs Data

• Color measured as the number of inches of green
  on the stalks
• Number of stalks in a bundle

9
Check the Data

• 4 variables, 200 observations
• Find Means and Standard deviations
• Find correlations

10
Issues of estimation method

• Heteroskedasticity?
• Autocorrelation?
• Simultaneity?
• Omitted variables?
• Endogenous regressors?
• Correct functional form?

11
Regression

• Linear model
• Interpretation
• Statistical significance
• Economic significance
• Semi-log model
• What happens to observations?
• Interpretation

12
Regression

• Log-log model
• Interpretation
• Significance

13
Summary

• Formulate a Model
• Literature
• Theory
• Gather Data
• Check data
• Descriptive statistics

14
Summary

• Estimate the Model
• Test specification
• Test hypotheses
• Write it up