Lecture 3 ............

continued from Lecture 2

Defining Econometrics, Economic Models, and Econometric Models

Instructor: Sunaina Dhingra

Email-id: [email protected]
Lecture Date: 11th August (12.20 pm)
Week 1-2. Review of statistical inference &
Nature of Econometrics and Economic Data:
• References: Lecture Notes & Chapter 1 of JW
 Econometrics and Its Applications

• Econometrics: based on the development of statistical methods for estimating

economic relationships, testing economic theories, and evaluating and
implementing government and business policy
• Analyzing micro and macro variables and estimating relationships between
different variables
• Forecasting macroeconomic variables
• Cobb-Douglas model of production
• Asset pricing model
• Effects of fertilizer use on crop yields
• Impact of school spending on student performance
 Economic Model
• Economic model consists of mathematical equations that describe various
relationships
• Consider an example of utility maximization model
𝑚𝑎𝑥𝑦1……,𝑦𝑛 u(𝑦1,……., 𝑦𝑛 ) subject to σ𝑁
𝑖=0 𝑦𝑖 ≤ 𝑚

𝑦𝑖 ≥0 for all i
• Max U() subject to budget constraints
Economic Model (contd.)

y=f (𝑥1 , 𝑥2 , 𝑥3 , 𝑥4 , 𝑥5 )

where ,
y = quantity of pizza
𝑥1 = price of pizza
𝑥2 = price of substitute good
𝑥3 = price of complimentary good
𝑥4 = income
𝑥5 = characteristics that affect taste
Economic Model (contd.)

Let’s consider another example of an equation that we can derive using informal
reasoning
wage = f(educ, exper, training)
Or y = f(𝑥1 , 𝑥2 , 𝑥3 )

where,
y = hourly wage
𝑥1 = years of formal education
𝑥2 = years of workforce experience
𝑥3 = did the employee receive any job training (yes-1; no-0)
 Econometric Model

Let’s consider a complete econometric model for example

wage = β0 + β1 educ + β2 exper + β3 training + μ

where,
β0 , β1 , β2 , and β3 = parameters of econometric model
μ = error term that contains unobserved factors
 Goals of Econometric Model

Two main goals

• Estimating parameters in the model
• Testing hypotheses
Types of Datasets
 Types of Economic Data

• The success of any analysis depends on the availability of

appropriate data.
• The three common types of economic data
• Cross-sectional data
• Time series data
• Pooled cross-sectional data
• Panel/longitudinal data
 Cross-Sectional Data
• For example, the table gives data of 526 working individuals for the year 1976.
The variables included in the table are:
• Observation number obsno wage educ exper female married
1 3.10 11 2 1 0
• Wage 2 3.24 12 22 1 1

• Number of years of education 3 3.00 11 2 0 0

4 6.00 8 44 0 1
• Experience earned 5 5.30 12 7 0 1
. . . . . .
• Indicator for gender . . . . . .
. . . . . .
• Marital status 525 11.56 16 5 0 1
526 3.50 14 5 1 0
 Time Series Data
• For example, the table gives data of Puerto Rico for 38 years. The variables
included in the table are:
• Observation number obsno year avgmin avgcov prunemp prgnp

• Year 1 1950 0.20 20.1 15.4 878.7

2 1951 0.21 20.7 16.0 925.0

• Minimum wage
3 1952 0.23 22.6 14.8 1015.9
• Coverage rate
. . . . . .
• Unemployment rate . . . . . .

• Gross national product 37 1986 3.35 58.1 18.9 4281.6

38 1987 3.35 58.2 16.8 4496.7

 Pooled Cross-Section Data

• For example, the table gives data of 520 houses. The variables included in the
table are:
• Observation number obsno year hprice proptax sqrft bdrms bthrms
1 1993 85,500 42 1600 3 2.0
• Year 2 1993 673,00 36 1440 3 2.5
3 1993 1,34,000 38 2000 4 2.5
• House price . . . . . . .
. . . . . . .
• Property tax . . . . . . .
250 1993 2,43,600 41 2600 4 3.0
• Square feet 251 1995 65,000 16 1250 2 1.0
252 1995 1,82,400 20 2200 4 2.0
• Bedrooms 253
.
1995
.
97,500
.
15
.
1540
.
3
.
2.0
.
. . . . . . .
• Bathrooms . . . . . . .
520 1995 57,200 16 1100 2 1.5
 Panel/Longitudinal Data

• Special type of pooled cross-section data

• Same cross-sectional units are surveyed over time
• Example: Census data
• Advantages of panel data
• Useful in establishing causality
• Ability to analyze behavioral lags or the outcome of decisions
Nature of Variables in
Empirical Analysis
 Types of Variables

• The different types of variables encountered in datasets used for

economic analysis are:
• Ratio scale
• Interval scale
• Ordinal scale
• Nominal scale
 Ratio Scale Variables

The three main properties of ratio scale variables are:

• Ratio of two variables (𝑦1 /𝑦2 )
• Distance between two variables (𝑦2 - 𝑦1 )
• Ordering of variables (𝑦2 ≥ 𝑦1 or 𝑦2 ≤ 𝑦1 )
 Interval Scale Variables

• Properties satisfied by interval scale variables

• Distance between two variables
• Ordering of variables
• Examples of interval scale variables
• Temperature scale
• Calendar years
 Ordinal Scale Variables

• The properties satisfied by ordinal scale variables

• Ordering of variables
• Examples of ordinal scale variables
• Grading systems in educational institutions: Grade A, Grade B,
Grade C, etc.
• Income classification: low income, middle income, and high
income
 Nominal Scale Variables

• They do not satisfy any of the properties of ratio variables

• Also referred to as categorical
• Special types are known as dummy/binary variables
• Examples of nominal scale variables
• Gender
• Marital status
• Religion
• Mode of transportation
 Discrete Variables

• Finite or countably infinite number of values

• Cannot take decimal or fractional values
• Example of discrete variable
• Number of books in the library
 Continuous Variables

• Infinite number of possible values

• Can take decimal or fractional values
• Examples of continuous variables
• Height of children in cm
• Test score in an exam
• Body temperature of those who have the flu
 Lecture 4
Relationship between variables

Instructor: Sunaina Dhingra

Email-id: [email protected]
Lecture Date: 14th August ( 9 am)
 Thinking About Two
Variables Simultaneously
 Probability Density Function of wage(y)
Histogram: f(y) or f(wage)
The STATA command used:

The STATA command used:

 Probability Density Function of education(x) Histogram:
f(x) or f(education)
The STATA command used:

The STATA command used:

Scatter Plot: wage(y) Against education(x)

The STATA command used:

 Conditional Distributions: PDF of y When
x=10,x=16 : f(y/x=10), f(y/x=16)

The STATA commands used:

 Joint Distributions: PDF of y and x,
f(x,y)

The STATA command used:

 Conditional Distribution: Using Expectation
Notation
• Expected value of a random variable x, E(x)=μ𝑥
• E(wage| education=10) = 3.8
• E(wage| education=16)=16
 Independence Between Variables
f(x,y)= f(x) f(y)

The STATA commands used: