Econometric Lec1

You might also like

Download as pdf or txt
Download as pdf or txt
You are on page 1of 72

THE NATURE OF

ECONOMETRICS

1
AND
REGRESSION ANALYSIS LECTURE

1
Chapter 0

Introduction

2
What is Econometrics

...Social Science in which the tools of


economic theory, mathematics and
statistical inference are applied to the
analysis of economic phenomena.

3
1. Economics Theory
“ Keynes postulated a positive relationship between consumption
and incomes”, i.e., people’s income

2. Mathematical Expression:
Consumption = f(Income) ==> C = f(Y)
MPC = dC/dY = f’(Y) > 0 ;assume 0 < MPC < 1
3. Statistics:
Year C Y Find the mean, variance,
standard deviation,
1980 2447.1 3776.3 correlation, etc.
1981 2476.9 3841.1
…. …. ….
4. Econometric - Regression model
C t = b1 + b2 Y t + u t => DC/DY = b2
=> estimating the relationship 4
The Role of Econometrics

Provide measurement and quantitative


analysis of actual economic
phenomena
or economic relationship based on

1. Economic theory
2. Economic data
3. Methods of model constructed
5
Economic Relationships:
Stock Market Index money supply
government
budget
Interest rate
Exchange
inflation trade Rate
deficit Properties Market
unemployment
Wage

capital gains tax


rent
control
crime rate laws
6
Economic Decisions

To use information effectively

}
economic theory
economic
decisions
economic data

*Econometrics* helps us combine


Economic theory and economic data.
7
How much ?

Listing the variables in an economic relationship is not enough.

For effective policy we must know the amount of change


needed for a policy instrument to bring about the desired
effect:

• By how much should the Federal Reserve


raise interest rates to prevent inflation?

• By how much can the price of football tickets


be increased and still fill the stadium?

8
Methodology of Econometrics
• Statement of theory or hypothesis
• Specification of the mathematical model of the
theory
• Specification of the econometric model of the
theory
• Obtaining data for the analysis.
• Estimation with statistical properties.
• Hypothesis testing
• Analyze and evaluate implications of the results
• Forecasting or prediction
• Using the model for control or policy purpose

9
10
Example: The Consumption Function

Consumption, C, is some function of income, inc :

C = f(inc)

For applied econometric analysis


this consumption function must be
specified more precisely.
11
Economic Empirical Study

Economic Theory; Past Experience, studies


C = f(Inc) ==>
Formulating a model: Ct = b1 + b2Inct + ut

Gathering data: Statistics monthly, quarterly, yearly data

Estimating the model: Simple OLS method or other advances


H0: b2>0,
Testing the hypothesis: positive relationship or not If not true

Interpreting the results:

Forecasting Policy implication and decisions


12
Chapter 1

The nature of
Regression
Analysis
13
Regression Analysis

...the study of the dependence of one variable


(called dependent variable) on one or more
other variables (called independent variables)
with a view to Estimating or Predicting
the mean or average of the former in terms of
the known or fixed values of the latter

14
Purpose of Regression Analysis

• 1. Estimate a relationship among


economic variables, such as Y = f(X).

• 2. Forecast or predict the value of


one variable, Y, based on the value of
another variable, X.

15
Terminology and Notation
Y = b1 + b2 X + u

Left hand-side Right hand-side


Variable: Variable:

Dependent Explanatory
Explained Independent
Predictand Predictor
Regressand Regressor
Response Stimulus or control
Endogenous Exogenous
16
Some Examples
• Consumption on Income
• GDP on Export, Money Supply
• Sales on Season, Advertising
Expenditures
• Income on Gender, Education, Years
of Traning

17
Some Comparisons

• Statistical vs. Deterministic Relationships


• Regression vs. Causation
• Regression vs. Correlation

18
Data

• Time-series Data
• Cross-Sectional Data
• Pooled or Panel Data

19
Time series data

20
Cross-section data and Pool (Panel) data

21
Chapter 2
Two-variable
Regression
Analysis : Some Basic
Ideas
22
Population Regression Function

23
Weekly Food Expenditures
Y = dollars spent each week on food items.
X = consumer’s family weekly income.

The relationship between X and the expected value


of Y , given X, might be linear:
E(Y|Xi) = f(Xi) = b1 + b2 Xi
Means that each conditional mean E(Y|Xi) is a
function of Xi, this equation is known as
the population regression function (PRF) 24
Y
Conditional means
E(Y|Xi)

Population regression line


149 (PRF)
101
65
Distribution of Y
given X=220
X
80 140 220
25
Average
Consumption Y
E(Y|x)
E(Y|x)=b1+b2x

DE(Y|X)
DE(Y|X) b2=
DX
Dx slope
b1{
Intercept X (income)

The Econometric Model: a linear relationship


between average consumption and income.
26
Linearity

• Linear in variables
• Linear in Parameters

27
Stochastic Specification of PRF

Given any income level of Xi, an family’s consumption is


clustered around the average of all families at that Xi, that
is, around its conditional expectation, E(Y|Xi).
The deviation of any individual Yi is:

ui = Yi - E(Y|Xi)

or Yi = E(Y|Xi) + u i
Shochastic error or
or Yi = b1 + b2 X + u i Stochastic disturbance
28
The Error Term
• Y is a random variable composed of two parts:

• I. Systematic component: E(Y) = b1 + b2x


This is the mean of Y.

II. Random component: u = Y - E(Y)


= Y - 1 - 2X
This is called the random or shochastic error.

Together E(Y) and u form the model:


Y = 1 + 2X + u
29
For examples:
given X = $80, the individual consumption are
Y1 = 55 = b1 + b2 (80) + u 1
Y2 = 60 = b1 + b2 (80) + u 2
Y3 = 65 = b1 + b2 (80) + u 3
Y4 = 70 = b1 + b2 (80) + u 4
Y5 = 75 = b1 + b2 (80) + u 5

Y^ = 65 = ^b + b^ (80)
1 1 2
Y^ = 65 = ^b + b^ (80)
Estimated average: 2 1 2
Y^3 = 65 = ^b1 + b^2 (80)
Y^4 = 65 = ^b1 + b^2 (80)
Y^5 = 65 = ^b1 + b^2 (80)
30
The reasons for stochastic disturbance

• Vagueness of theory
• Unavailability of data
• Direct effect vs indirect effect
• (Core variables vs peripheral variables)
• Intrinsic randomness in human behaviour
• Poor proxy variables
• Principle of parsimony
• wrong functional form
31
Unobservable Nature of Error Term

• Unspecified factors / explanatory variables,


not in the model, may be in the error term.
For example: Final examine score is not only depend on
class attended but also other unobserved factors such as
student ability, maths background, hard work effort, etc.

• Approximation error is in the error term if


relationship between y and x is not exactly a perfectly
linear relationship.

• Strictly unpredictable random behavior that


may be unique to that observation is in error. 32
Y (SRF) ^Y = ^b + ^b x
Y4 . 1 2

{ E(Y|x)=b1+b2x
Y3 (PRF)
}
Y2 ^
^ u2
Y 2
u2
E(Y|x2)

Y1 .} u 1
x
x1 x2 x3 x4
The relationship among Yi, ui and the true regression line.
33
The Sample Regression Function (SRF)
(SRF2)
Y ^ ^ ^
Y = b1 + b2x
Y4 ^u . ^Y = ^b + ^b x
4{ 1 2
(SRF1)
Y3 .}^
u3
Y2 ^u .
2{

^
Y1 } 1
.u
x
x1 x2 x3 x4
34
Different samples will have different SRFs)
SRF:
^ ^ ^
Yi = b1 + b2 Xi
^ ^
or Yi = b1 + b2Xi + u^ i Residual

or Yi = b1 + b2 Xi + ei

PRF:
E(Y|X) = b1 + b2 Xi
Yi = b1 + b2 Xi + u i Error term or
Disturbance
^
Yi = estimator of Yi (E(Y|xi)
^
bi or bi = estimator of bi 35
Chapter 3
Two-variable
Regression Model: The
Problem of Estimation

36
Ordinary Least Squares (OLS) Method

Yi = b1 + b2Xi + ui
u^ i = Y i - b^1 - b^2X i

Minimize error sum of squared deviations:


n n
å u^i2 = S(Y i - b^1 - b^2X i )2 = f(b^1,b^2)
i=1 i=1
37
Minimize w.r.t. b^1 and b^2:
n
f(b^1,b^2) = S(Y i - b^1 - b^2x i 2
) = f(.)
i =1

¶f(.)
= - 2 S (Y i - b^1 - b^2Xi )
¶ b^1

¶f(.)
= - 2 S Xi (Yi - b^1 - b^2Xi )
¶ b^2
Set each of these two derivatives equal to zero and
solve these two equations for the two unknowns: b^1 b^238
To minimize f(.), you set the two
derivatives equal to zero to get:

¶f(.)
= - 2 S (Y i – b1 – b2Xi ) = 0
¶ b^1

¶f(.)
= - 2 S xi (Yi - b1 – b2Xi ) = 0
¶ b^2

39
- 2 S (Y i - b1 – b2Xi ) = 0

-2S Xi (Y i – b1 – b2Xi ) = 0

S Yi - nb1 – b2 SXi = 0
2
S Xi Yi - b1 S X i - b2 S Xi = 0

nb1 + b2 S Xi = S Yi
S Xi + b2S Xi S Xi Yi
2
b1 =
40
n SXi b1
= S Yi
S i S i
X X 2
b2 = S Xi Yi
Solve the two unknowns
n SXi Yi - S Xi SYi
b2 =
n SX i - (SXi )
2 2

S(Xi - X )(Yi -Y) Sxy


= =
S( i
X - X ) 2
Sx 2

b1 = Y - b2 x 41
Y
Y4
. ^
Y = b1 + b2X

^
Y*
^*
Y3
.
{.
^u*
4
^*
^
Y* = b*1 + b* 2X
^
Y*
.
1 . 2
^
u*3{ Y4
u*2 {. Y2
^ .

{
^
u*1

.
Y1
Y3

x1 x2 x3 x4 x

Why the SRF is the best one?


^ is larger.
The sum of squared residuals from any other line Y* 42
The classical regression model:
Assumptions of LS Method
1. The linear regression Model:linear in
parameters
Y = b1+ b2X+ u
2. X values are fixed in repeated sampling, so that X is
not constant (X is nonstochastic).
3. Zero mean value of error terms (disturbance, ui),
E( ui | xi) = 0
4. Homoscedasticity or equal variance of ui, the
conditional variances of u i are identical, i.e.,
var(ui|xi) = s2 43
Homoscedasticity Case
f(Yi) Yi

re
d itu
en
p
.
ex .

x1=80 x2=100 income xi


The probability density function for Yi at two
levels of family income, X i , are identical. 44
Heteroscedasticity Case
f(Yi)

Y
i
r e
itu
n d
p e
ex
.
.
.
x1 x2 x3 income xt
The variance of Yi increases as family income,
xi, increases. 45
Assumptions (continue)

5. No autocorrelation between the disturbance.


cov(ui,uj|xi ,xj) = 0

6. Zero covariance between ui and xi, i.e.,


cov(ui,xi ) = E(ui,xi ) = 0

7. The # of observation (n) must be greater


than the # of parameters (k) to be estimated.
n>k
46
Assumptions (continue)

8. Variability in X values: The values in a


given sample must not all be the same,
at least two must different.

9. No specification bias or error: the regression


model is correctly specified.

10. There is no perfect multicollinearity. No


perfect linear relationship among the
independent variables. i.e.,
Xk ¹ l Xm
47
One more assumption that is often used in
practice but is not required for least squares:
• (Optional) The values of y are
normally distributed about
their mean for each
value of x:

Y ~ N [(b1+b2X), s2 ]

48
The Error Term Assumptions

• 1. The value of y, for each value of x, is


Y = b1 + b2X + u
• 2. The average value of the random error u is:
E(u) = 0
• 3. The variance of the random error u is:
var(u) = s2 = var(Y)
• 4. The covariance between any pair of u’s is:
cov(ui , uj) = cov(Yi ,Yj) = 0
• 5. u is normally distributed with mean 0, var(u)=s2
u ~ N(0,s2)

49
Simple Linear Regression Model

Yi = b1 + b2 X i + ui

Yi = family weekly expenditures


X i = family weekly income

For a given level of xi, the expected


level of food expenditures will be:
E(Yi|X i) = b1 + b2 X i 50
The population parameters b1 and b2
are unknown population constants.

The formulas that produce the


^ ^
sample estimates b1 (or b1)and b2(or b2) are
called the estimators of b1 and b2.
• When b1 and b2 are used to represent

the formulas rather than specific values,


they are called estimators of b1 and b2
which are random variables because
they are different from sample to sample
51
Estimators are Random Variables

• If the least squares estimators b1 and b2


are random variables, then what are their
means, variances, covariances and
probability distributions?

• Compare the properties of alternative


estimators to the properties of the
least squares estimators.
52
Precision or standard errors of LS
estimates
Given that both Yi and ui have variance s 2,
the variance of the estimator b2 is:

^2
s ^2
s
var(b2) = =
S(xi - x) 2
S 2

xi
b2 is a function of the Yi values but
var(b2) does not involve Yi directly. 53
Variance of b1

Given b1 = y - b2x
the variance of the estimator b1 is:

Sx i
2
Sx i2
var(b1) = s2 = s2
n S(x i - x)
2 2
nSxi

54
Covariance of b1 and b2

-x -x
cov(b1,b2) = s2 = s2
2
S(x t - x) S xi
2

55
Estimating the variance
of the error term, s 2

^u = yi - b1 - b2 xi
i

T
^
S
2
ui
^2
s = i =1

n- 2

^
s is an unbiased estimator of s 2
2

56
Properties of LS Estimators:
Gauss-Markov Theorem

Given the assumptions of classical linear


regression model, the ordinary least
squares (OLS) estimators b1 and b2
are the best linear unbiased estimators
(BLUE) of b1 and b2. This means that b1
and b2 have the smallest variance of all
linear unbiased estimator of b1 and b2.
Note: Gauss-Markov Theorem doesn’t
apply to non-linear estimators 57
The coefficient of Determination, R2 –
A Measure of “goodness of fit”

58
The coefficient of Determination, R2 –
A Measure of “goodness of fit”

59
The coefficient of Determination, R2 –
A Measure of “goodness of fit”

^ ^
(Yi - Y) = (Yi - Yi) + (Yi - Y)
u^
To measure variation:
^ 2 ^
S(Yi - Y)2 = S[(Yi - Y) i + (Yi - Y)]2
^ 2 ^
S(Yi - Y)2 = S(Yi - Y) i + S(Yi - Y)2

TSS RSS (Su^2) ESS


Total sum of squares Residual sum of squares Explained sum of squares
60
R2 - Measure of “goodness of fit”

2 ESS RSS
Define R = =1-
TSS TSS
Su
^ 2
i
=1-
S(Yi - Y)2
^
S(Yi - Y)2
R2 =
S(Yi - Y)2 1 ³ R2 ³ 0
61
Alternative R2 expression

^
ESS S(Yi - Y)2 Sy^ i2
2
R = = =
TSS S(Yi - Y) 2 Sy i
2

^ ^2
S(b2xi )2 b Sx 2
2 i ^ 2 Sxi2
= Sy 2 = = b
i Syi2 2
Syi2

^ 2
2
Sx Sxiyi 2
Sxi2 (Sxiyi)2
= b = =
2 Sy2 Sxi2 Syi2 Sxi2Syi2
62
Explanation of R2
• R2 tells us how well the sample regression
line fits the data
• The value of R2 lets us know how much
percent of variation in the dependent variable
can be explained by the independent
variable or by the regression model.
• Ex: R2=0.8: 80% of the variation in Y can be
explained by X.

63
Y As R2 = 0

SRF
Which SRF ?

Y
As R2 = 1
SRF

SRF go through all points


64
X
Coefficient of Correlation R

65
Numerical Example

• Manual Calculations
• Demonstration of using computer software
• Use the data in table 3.8

66
Chapter 4
Classical Normal Linear
Regression Model

67
Summary of BLUE estimators
Mean
E(b1)=b1 and E(b2)=b2

Variance

Var(b1 )=s 2
Sx i2
and s2
Var(b2)=
nSxi
2
Sxi 2

Standard error or standard deviation

Se(bk) = var(bk) 68
Estimated Error Variance
T
^
S
2
ui ^ 2) = s2
E(s
^2
s = i =1

n- 2

Standard Error of Regression (Estimate), SEE

T
^
S
2
^=
s ^
s =
2 ui
i =1 K
n- 2 # of independent
Variables plus
69 the
Constant term
Normality Assumption for ui
ui ~ N(0,σ2)

Y ~ N [(β1+
β2X), σ2 ]

70
Probability Distribution
of Least Squares Estimators

s2Sx i2
b1 ~ N b1 ,
nSx i2

s2
b2 ~ N b2 ,
Sx i2
71
THE END

72

You might also like