1405 words - 6 pages

Estimating the Elasticity of Demand for Gasoline

Professor Pushan Dutt

The graph below shows the evolution of the price of oil (adjusted for inflation) since

1957. Note a couple of sharp jumps and collapses in the price of oil.

1.

2.

3.

1973: : 2.75 % of global production was withheld; Prices in nominal terms

jumped from $3.5 a barrel to $10 a barrel

1979: 5.68 % of global production withheld; Prices in nominal terms jumped

from $15 to $32 a barrel.

2007: Oil prices increase from $60 to reach a peak of $128, followed by a rapid

collapse

Why do we observe such sharp swings in oil prices? The answer lies in the elasticity of

demand and supply. For this exercise we will focus ...view middle of the document...

We will also assume

that the demand function is log-linear. The advantage of this specification is that all the

estimated coefficients are elasticities. For example, in the specification below β is the

price elasticity and γ is the income elasticity of gasoline. Please note that we add a

negative sign in the price elasticity formula so the price elasticity will be – β.

log Gt log PGt log Yt log PNCt log PUCt

Step 2: Collect Data

Variable

G

PG

Y

PNC

PUC

Inflation

Population

Source

U.S. Energy

Information

Administration

U.S. Bureau of

Labor Statistics

Federal

Reserve

Economic Data

Website

http://tonto.eia.doe.gov/dnav/pet/his

t/mgfupus1m.htm

http://data.bls.gov/cgibin/surveymost?ap

http://research.stlouisfed.org/fred2/s

eries/A229RX0#

U.S. Bureau of http://www.bls.gov/cpi/data.htm

Labor Statistics

U.S. Bureau of http://www.bls.gov/cpi/data.htm

Labor Statistics

U.S. Bureau of http://www.bls.gov/cpi/data.htm

Labor Statistics

http://research.stlouisfed.org/fred2/s

Federal

Reserve

eries/POP

Economic Data

Notes

Thousands of barrels;

multiply by 42 to

convert to gallons

Monthly retail price

of a gallon of gas

Real per capita

disposable income;

Monthly, seasonallyadjusted data

Monthly, seasonallyadjusted data

Monthly, seasonallyadjusted data

We need to adjust

prices and income for

inflation

We need population

to get per capita

gasoline consumption

Is that too much work? Since I am feeling generous and since duplication of effort should

be avoided, go to the class website and retrieve Excel file: gasolinedata.xls

The raw data sits in the “raw data” spread sheet. The data you will use sits in the

“adjusted data” sheet. Here the gasoline consumption is in gallons and the price of

gasoline is adjusted for inflation. Price of used cars and new cars are price indices so we

don’t need to adjust this.

Step 3: Seasonal Adjustment of Gasoline Consumption

The sheet Chart 1 plots the evolution of per capita gasoline consumption over time. Do

you discern any pattern in the data that may require additional variables in the regression?

Which months are the peaks and which months are the troughs? Does this seem

reasonable?

Given that there is a clear seasonal pattern in the data we need to adjust gasoline

consumption in the spreadsheet “adjusted data.” A simple way to do is to calculate a 12month moving average

Click Data Analysis, and choose Moving Average

Highlight column labeled G in the spreadsheet “adjusted data”

Input 12 for 12-months in the Interval box.

Select cell I2 as the place to output the moving average. This will create a variable

for seasonally adjusted gasoline consumption. Call it SG.

Again, this is done in spreadsheet

Step 4: Basic Demand Function

Take the natural log of all the variables (base e not base 10; ln(x) in Excel). Put

them in a third spreadsheet. Call it “data for regression”. Name the...

Tap into the world’s largest open writing community