Outline Review of Econometrics I Panel Data Fixed Effects Next Time References

Regression with Panel Data: Part I

Osman DOGAN

1 / 37
Outline Review of Econometrics I Panel Data Fixed Effects Next Time References

Regression with Panel Data: Part I

Today outline:
1 Review of Econometrics I
2 Panel data
3 Regression with entity fixed effects
4 Regression with time fixed effects
Readings:
1 Stock and Watson (2020, Chapter 10).
2 Hanck et al. (2021, Chapters 10).

2 / 37
Outline Review of Econometrics I Panel Data Fixed Effects Next Time References

What is econometrics?

Stock and Watson (2020) give the following definition.

Definition 1
Econometrics is the science and art of using economic theory and statistical
techniques to analyze economic data.

All econometric work relies on some untestable assumptions.

Therefore, the economist Keane (2010, p. 48) writes that
... there is no way to escape the role of assumptions in statistical work,
so our conclusions will always be contingent. Hence, we should be cir-
cumspect about our degree of knowledge. In the words of Maimonides:
“Teach thy tongue to say ‘I do not know,’ and thou shalt progress.”

3 / 37
Outline Review of Econometrics I Panel Data Fixed Effects Next Time References

Econometric models
We covered the single and multiple linear regression models. The multiple
linear regression model is stated as
Yi = β0 + β1 X1i + β2 X2i + . . . + βk Xki + ui , i = 1, 2, . . . , n. (1)
We considered this model under the following assumptions:
Assumption 1 (Zero-conditional mean assumption)
The conditional distribution of ui given X1i , . . . , Xik has mean zero, that is,
E (ui |X1i , Xi2 , . . . , Xik ) = 0 for i = 1, 2, . . . , n.

Assumption 2
(X1i , Xi2 , . . . , Xik , Yi ), i = 1, 2, . . . , n, are i.i.d. draws from their joint
distribution.

Assumption 3
Large outliers are unlikely: X1 , . . . , Xk , and Y have finite fourth moments.

Assumption 4
There is no perfect multicollinearity.
4 / 37
Outline Review of Econometrics I Panel Data Fixed Effects Next Time References

Econometric models

The most important assumption is the zero-conditional mean assumption.

Assume that there is only one regressor. Then, how should we think about
E(ui |Xi ) = 0?

E(ui |Xi ) = 0 =⇒ corr(X, ui ) = 0 (2)

The contrapositive statement of (2) is

corr(X, ui ) ̸= 0 =⇒ E(ui |Xi ) ̸= 0 (3)

Thus, if we think that the error term includes a variable that is correlated with
the regressor X, then we can claim that E(ui |Xi ) ̸= 0.

5 / 37
Outline Review of Econometrics I Panel Data Fixed Effects Next Time References

Econometric models

The control variable approach is one of the methods that can be used to
achieve the zero conditional mean assumption.

Definition 2 (Control Variable)

A control variable W is a variable that is correlated with, and controls for, an
omitted causal factor in the regression of Y on X, but which itself does not
necessarily have a causal effect on Y .

An effective control variable satisfies the conditional mean independence

assumption.

Definition 3 (Conditional mean independence)

Let Xi denote the variable of interest and Wi denote the control variable(s).
W is an effective control variable if conditional mean independence holds:

E (ui |Xi , Wi ) = E(ui |Wi ) (4)

6 / 37
Outline Review of Econometrics I Panel Data Fixed Effects Next Time References

Econometric models

The conditional mean independence means that given the control variable, the
mean of ui doesn’t depend on the variable of interest.

Key Concept 1 (Control Variables)

Consider Y = β0 + β1 X + β2 W + u, where
1. X is the variable of interest
2. W is an effective control variable so that E (ui |Xi , Wi ) = E(ui |Wi ).
Then, we have the following results:
(a) β1 has a causal effect, i.e., it has a causal interpretation.
(b) The OLS estimator β̂1 is unbiased.
(c) The OLS estimator β̂2 is in general biased.

7 / 37
Outline Review of Econometrics I Panel Data Fixed Effects Next Time References

Econometric models

We extended the multiple linear regression model by considering the non-linear

relationship between the dependent variable and regressors.
We used the following approaches to incorporate the non-linear relationships:
1 The polynomial regression model,
2 The natural logarithm transformation,
3 Interactions between independent variables.
When we have binary dependent variables, we can use one of the following
models:
1 Linear probability model:
P (Yi = 1|X1i , X2i , . . . , Xki ) = β0 + β1 X1i + β2 X2i + . . . + βk Xki ,
2 Probit model:
P (Yi = 1|X1i , X2i , . . . , Xki ) = Φ (β0 + β1 X1i + β2 X2i + . . . + βk Xki ),
3 Logit model:
P (Yi = 1|X1i , X2i , . . . , Xki ) = F (β0 + β1 X1i + β2 X2i + . . . + βk Xki ).

8 / 37
Outline Review of Econometrics I Panel Data Fixed Effects Next Time References

Causality

We saw that a study gives a causal relationship if it has internal validity.

Definition 4 (Internal Validity)

A statistical analysis is said to have internal validity if the statistical inferences
about causal effects are valid for the population being studied.

Internal validity has two components:

1 The estimator of the causal effect should be unbiased and consistent.
2 The standard errors are computed in a way that makes confidence intervals have
the desired confidence level. That is, t-tests, p-values and F-tests should be valid.

9 / 37
Outline Review of Econometrics I Panel Data Fixed Effects Next Time References

Causality

Five threats to the internal validity of regression studies:

1 Omitted variable bias,
2 Wrong functional form (misspecification of functional form),
3 Errors-in-variables bias,
4 Sample selection bias,
5 Simultaneous causality bias.

All of these imply that

E (ui |X1i , . . . , Xki ) ̸= 0. (5)

10 / 37
Outline Review of Econometrics I Panel Data Fixed Effects Next Time References

Randomized control trials

Another way to achieve the zero-conditional mean assumption is to use the

randomized controlled experiments.

Definition 5 (Randomized control trial)

A randomized controlled trial is a type of scientific experiment that aims to
measure the effectiveness of some intervention (treatment or policy) through
randomly allocating subjects to two or more groups, treating them differently,
and then comparing them with respect to a measured response.

In the test score example, one can imagine randomly assigning “treatments” of
different class sizes (STR) to different groups of students.
If the experiment is designed and executed so that the only systematic
difference between the groups of students is their class size, then in theory this
experiment would estimate the effect on test scores of reducing class size,
holding all else constant.

11 / 37
Outline Review of Econometrics I Panel Data Fixed Effects Next Time References

Prediction and forecasting

We use the estimated regression model to get predicted values on the

dependent variable:

Ŷi = β̂0 + β̂1 X1i + β̂2 X2i + . . . + β̂k Xki , i = 1, 2, . . . , n, (6)

where Ŷi denotes the predicted value.

Thus, prediction is the process of using information on some variables
(regressors) to make statement about the value of the dependent variable.
A forecast is a prediction about the value of a variable in the future. We will
see more on this topic in this semester.
You do not need to know a causal relationship to make a good prediction.
A good prediction requires a good fit: High R̄2 and small SER.

12 / 37
Outline Review of Econometrics I Panel Data Fixed Effects Next Time References

Data

In econometrics, data come from one of two sources: experiments or

nonexperimental observations of the world.

Definition 6 (Experimental data)

Experimental data come from experiments designed to evaluate a treatment or
policy or to investigate a causal effect.

Definition 7 (Observational data)

Data obtained by observing actual behavior outside an experimental setting are
called observational data.
Whether the data are experimental or observational, data sets come in three
main types:
1 Cross-sectional data
2 Time series data
3 Panel data

13 / 37
Outline Review of Econometrics I Panel Data Fixed Effects Next Time References

Data

Panel data, also called longitudinal data, are data for multiple entities in which
each entity is observed at two or more time periods.

14 / 37
Outline Review of Econometrics I Panel Data Fixed Effects Next Time References

Data

Panel data are usually observed at regular time intervals (monthly, quarterly,
yearly, etc.), and are balanced (all units are observed at all periods).
Panel data can be
□ a short panel: many units and few time periods,
□ a long panel: many time periods and few units,
□ a large panel: many units and many time periods.

Why are panel data useful?

1 With panel data, we can control for factors that could cause omitted variable
bias if they are omitted.
2 With panel data, we can control for factors that vary across entities but do not
vary over time.
3 With panel data, we can control for factors that vary across time but do not vary
over entities.
4 Panels give more sources of variation, less collinearity among the variables, and
yield more efficient estimators.
5 With panel data, it is easier to study dynamic models.

15 / 37
Outline Review of Econometrics I Panel Data Fixed Effects Next Time References

Notation for panel data

We will use a double subscript it to distinguish entities (states) and time

periods (years).
1 We will use i to denote entities, and assume that there are n entities:
i = 1, 2, . . . , n.
2 We will use t to denote time periods, and assume that there are T time periods:
t = 1, 2, . . . , T.
Suppose we have one dependent variable and only one regressor. Then, the
data are:

(Xit , Yit ), i = 1, 2, . . . , n, t = 1, 2, . . . , T (7)

If we have k regressors. Then, the data will be denoted in the following way:

(X1it , X2it , . . . , Xkit , Yit ), i = 1, 2, . . . , n, t = 1, 2, . . . , T (8)

16 / 37
Outline Review of Econometrics I Panel Data Fixed Effects Next Time References

Data

We will work with traffic fatality data set (fatality.csv).

The data are for the lower 48 U.S. states (excluding Alaska and Hawaii),
annually for 1982 through 1988.
Observational units: n = 48 states and T = 7 years, n × T = 336 observations.
The main variables are described in the following table:
variable description
state state id (FIPS) code
year year
mrall per capita vehicle fatality (number of traffic deaths per 10.000
people)
beertax the tax on a case of beer
mlda minimum legal drinking age
jaild = 1 if state requires mandatory jail sentence for an initial drunk
driving conviction
comserd = 1 if state requires mandatory community service for an initial
drunk driving conviction
vmiles total vehicle miles traveled annually
unrate unemployment rate
perinc per capita personal income

17 / 37
Outline Review of Econometrics I Panel Data Fixed Effects Next Time References

Data
Consider simple linear regressions between fatality rate and the real tax on a
case of beer in 1982 and 1988. The estimated models are:
\
FatalityRate = 2.01 + 0.15BeerTax (1982 data),
\
FatalityRate = 1.86 + 0.44BeerTax (1988 data).
The regression results indicate a positive relationship between the beer tax and
the fatality rate for both years.

(a) 1982 Data (b) 1988 Data

18 / 37
Outline Review of Econometrics I Panel Data Fixed Effects Next Time References

Data

These results are contrary to our expectations: alcohol taxes are supposed to
lower the rate of traffic fatalities.
As we known from Econometrics I, this is possibly due to omitted variable bias,
since both models do not include any covariates:
1 Quality (age) of automobiles
2 Quality of roads
3 Culture around drinking and driving
4 Density of cars on the road

Because of these omitted variables, our results suffer from the omitted variable
bias.
Panel data lets us eliminate omitted variable bias when the omitted variables
are constant over time within a given state.

19 / 37
Outline Review of Econometrics I Panel Data Fixed Effects Next Time References

State fixed effects

Consider the following panel data model:
FatalityRateit = β0 + β1 BeerTaxit + β2 Zi + uit , (9)
where, Zi is a factor that does not change over time.
Suppose Zi is not observed, so its omission could result in omitted variable
bias. However, the effect of Zi can be eliminated using T = 2 years.
The key idea: Any change in the fatality rate from 1982 to 1988 cannot be
caused by Zi , because Zi (by assumption) does not change between 1982 and
1988.
Consider (9) when t = 1982 and t = 1988:
FatalityRatei1982 = β0 + β1 BeerTaxi1982 + β2 Zi + ui1982 , (10)
FatalityRatei1988 = β0 + β1 BeerTaxi1988 + β2 Zi + ui1988 . (11)
Assume that E (uit |BeerTaxit , Zi ) = 0. Subtracting (10) from (11) gives:

FatalityRatei1988 − FatalityRatei1982 = β1 (BeerTaxi1988 − BeerTaxi1982 )

+ (ui1988 − ui1982 ) (12)

20 / 37
Outline Review of Econometrics I Panel Data Fixed Effects Next Time References

State fixed effects

Specifying the regression in changes in (12) eliminates the effect of the

unobserved variables Zi that are constant over time.
In (12), the new error term (ui1988 − ui1982 ) is uncorrelated with both
BeerTaxi1988 and BeerTaxi1982 .
This “difference equation” in (12) can be estimated by OLS, even though Zi is
not observed.
Note that this difference regression does not have an intercept because it was
eliminated by the subtraction step.
The estimated model is

− FatalityRatei1982
FatalityRatei1988\
= −0.072 − 1.04 (BeerTaxi1988 − BeerTaxi1982 ) .

The estimated effect of a change in the real beer tax is negative, as predicted
by economic theory.
An increase in the real beer tax by $1 per case reduces the traffic fatality rate
by 1.04 deaths per 10.000 people.
21 / 37
Outline Review of Econometrics I Panel Data Fixed Effects Next Time References

State fixed effects

What if you have more than 2 time periods (T > 2)? Consider:

Yit = β0 + β1 Xit + β2 Zi + uit , i = 1, 2, . . . , n, t = 1, 2, . . . , T. (13)

Here, Zi is an unobserved variable that varies from one state to the next but
does not change over time.
Because Zi varies from one state to the next but is constant over time, the
model in (13) can be interpreted as having n intercepts, one for each state.
Specifically, let αi = β0 + β2 Zi . Then, (13) can be written as

Yit = β1 Xit + αi + uit , i = 1, 2, . . . , n, t = 1, 2, . . . , T. (14)

Definition 8
The model in (14) is the fixed effects regression model, in which
α1 , α2 , . . . , αn are treated as unknown intercepts to be estimated. These terms
α1 , α2 , . . . , αn are also known as entity fixed effects.

22 / 37
Outline Review of Econometrics I Panel Data Fixed Effects Next Time References

State fixed effects

Consider the model in (14). The state-specific intercepts also can be expressed
using binary variables to denote the individual states. Define
(
1 if i = 2,
D2i = (15)
0, otherwise.

Then, the fixed effects regression model in (14) can be written equivalently as

Yit = β0 + β1 Xit + γ2 D2i + γ3 D3i + . . . + γn Dni + uit , (16)

where
1 β0 , β1 , γ2 , γ3 , . . . , γn are unknown coefficients to be estimated,
2 the dummy variables D3i , . . . , Dni are defined as in (15).

23 / 37
Outline Review of Econometrics I Panel Data Fixed Effects Next Time References

State fixed effects

We claim that the models in (14) and (16) are equivalent. How can we see this
equivalence?
We can determine the the relationship between the coefficients of (14) and
(16). When i = 1, we will get
1 Y1t = β1 X1t + α1 + u1t from (14) and
2 Y1t = β0 + β1 X1t + u1t from (16).
Thus, we have α1 = β0 . Similarly, when i ≥ 2, we have
1 Yit = β1 Xit + αi + uit from (14), and
2 Yit = β0 + β1 Xit + γi + u1t from (16).

Thus, we have αi = β0 + γi for i ≥ 2.

24 / 37
Outline Review of Econometrics I Panel Data Fixed Effects Next Time References

Estimation

Summary: Two ways to write the fixed effects model

1 Fixed effects form:
Yit = β1 Xit + αi + uit . (17)
2 Dummy variable form:
Yit = β0 + β1 Xit + γ2 D2i + γ3 D3i + . . . + γn Dni + uit . (18)
When there are multiple regressors, X1 , X2 , . . . , Xk , then the fixed effects
models can be written in the following ways:
1 Fixed effects form:
Yit = β1 X1it + β2 X2it + . . . + βk Xkit + αi + uit . (19)
2 Dummy variable form:
Yit = β0 + β1 X1it + β2 X2it + . . . + βk Xkit
+ γ2 D2i + γ3 D3i + . . . + γn Dni + uit . (20)

25 / 37
Outline Review of Econometrics I Panel Data Fixed Effects Next Time References

Estimation

There are three estimation methods that we can use:

1 Use“n − 1 dummy regressors” OLS regression,
2 Use “entity-demeaned” OLS regression,
3 “Changes” specification, without an intercept (only works for T = 2). We
already saw this method.

These three methods produce identical estimates of the regression coefficients,

and identical standard errors.
Methods (1) and (2) work for general T . Method (1) is only practical when n
is not too big.
In the case of Method (1), we need to first create the binary variables
D2i , . . . , Dni . Then, we can estimate (16) by OLS.
Inference (hypothesis tests, confidence intervals) is as usual (using
heteroskedasticity-robust standard errors).
The first method is impractical when n is very large (for example if n = 1000
workers).

26 / 37
Outline Review of Econometrics I Panel Data Fixed Effects Next Time References

Estimation

For the entity-demeaned OLS regression, consider the “time average” of (14):
T T T
1 X 1 X 1 X
Yit = β1 Xit + αi + uit . (21)
T t=1 T t=1 T t=1

Subtracting (21) from (14) gives

T T
! T
!
1 X 1 X 1 X
Yit − Yit = β1 Xit − Xit + uit − uit
T t=1 T t=1 T t=1
=⇒ Ỹit = β1 X̃it + ũit , (22)
 
where Ỹit = Yit − T1 Tt=1 Yit , X̃it = Xit −
P 1
PT
T t=1 Xit , and
 
ũit = uit − T1 Tt=1 uit .
P

In (22), Ỹit and X̃it are “entity-demeaned” terms.

Thus, in order to estimate (14), we first need to construct the entity-demeaned
terms Ỹit and X̃it . Then, we can regress Ỹit on X̃it by using the OLS method.

27 / 37
Outline Review of Econometrics I Panel Data Fixed Effects Next Time References

Estimation

Consider the following regression model:

FatalityRateit = β0 + β1 BeerTaxit + αi + uit . (23)

We will use the plm package to estimate the fixed effects models.
As for the lm function, we have to specify the regression formula and the data
to be used in our call of plm function.
Additionally, it is required to pass a vector of names of entity and time
variables to the argument index.

# Application to Traffic Deaths

r1 = plm ( mrall∼beertax , data = mydata , index = c ( " state " ," year " ) ,
model = " within " )
r1 _ vcov = vcovHC ( r1 , type = " HC1 " )
r1 _ se = sqrt ( diag ( r1 _ vcov ) )
stargazer ( r1 , se = list ( r1 _ se ) , type = " text " , title = " Estimaiton Results " )

The estimation results are given in Table 1. The estimated model is

FatalityRateit = −0.66BeerTaxit + α (24)

\ bi

28 / 37
Outline Review of Econometrics I Panel Data Fixed Effects Next Time References

Estimation

The coefficient on BeerTaxit is negative and significant. The interpretation is

that the estimated reduction in traffic fatalities due to an increase in the real
beer tax by $1 is 0.66 per 10.000 people.

Table 1: Estimation Results

Dependent variable:
mrall
beertax −0.656∗∗
(0.289)
Observations 336
R2 0.041
Adjusted R2 −0.120
F Statistic 12.190∗∗∗ (df = 1; 287)
Note: ∗ p<0.1; ∗∗ p<0.05; ∗∗∗ p<0.01

Conclusion: By including state fixed effects in the fatality rate regression, we

can avoid omitted variables bias arising from factors like cultural attitudes
toward drinking and driving, which vary across states but remain constant over
time within a state.

29 / 37
Outline Review of Econometrics I Panel Data Fixed Effects Next Time References

Time fixed effects

There can be omitted variables that might vary over time but not across states:

□ Safer cars (air bags, etc.) and changes in national laws.

Let St denote the combined effect of variables which changes over time but not
states (“safer cars”). Then,

Yit = β0 + β1 Xit + β2 Zi + β3 St + uit , i = 1, 2, . . . , n, t = 1, 2, . . . , T.

(25)

This model can be reformulated to have an intercept that varies from one year
to the next. Considering t = 1982 and ignoring Zi from model yields:

Yi1982 = β0 + β1 Xi1982 + β3 S1982 + ui1982

= (β0 + β3 S1982 ) + β1 Xi1982 + ui1982
= λ1982 + β1 Xi1982 + ui1982 ,

where λ1982 = (β0 + β3 S1982 ).

Similarly, if we set t = 1983, we will get Yi1983 = λ1983 + β1 Xi1983 + ui1983 ,
where λ1983 = (β0 + β3 S1983 ).
30 / 37
Outline Review of Econometrics I Panel Data Fixed Effects Next Time References

Time fixed effects

Thus, in general, we can write the model with the fixed effects in the following
way:
Yit = β0 + β1 Xit + λt + uit , i = 1, 2, . . . , n, t = 1, 2, . . . , T. (26)

Definition 9
The time fixed effects regression model with a single X regressor is
Yit = β0 + β1 Xit + λt + uit , where λ1 , λ2 , . . . , λT are known as the time fixed
effects.
Just as the entity fixed effects regression model can be represented using n − 1
binary indicators, the time fixed effects regression model can also be
represented using T − 1 binary indicators:
Yit = β0 + β1 Xit + δ2 B2t + δ3 B3t + . . . + δT BTt + uit , (27)
where δ2 , δ3 , . . . , δT are unknown coefficients and
(
1 if t = 2,
B2t = (28)
0, otherwise,
and other dummy variables B3t , . . . , BTt are defined similarly.
31 / 37
Outline Review of Econometrics I Panel Data Fixed Effects Next Time References

Time fixed effects

Summary: Two ways to write the time fixed effects model

1 Time fixed effects form:

Yit = β1 Xit + λt + uit . (29)

2 Dummy variable form:

Yit = β0 + β1 Xit + δ2 B2t + δ3 B3t + . . . + δT BTt + uit , (30)

The model in (29) can be estimated “year-demeaned” OLS regression:
1 Deviate Yit , Xit from year (not state) averages,
2 Estimate by OLS using “year-demeaned” data.
The model in (30) can be estimated “T − 1 binary regressor” OLS regression:
1 Create binary variables B2, . . . , BT ,
2 Regress Y on X, B2, . . . , BT using OLS.

32 / 37
Outline Review of Econometrics I Panel Data Fixed Effects Next Time References

Entity and time fixed effects

We can also have a regression model that has both entity and time fixed effects.

Definition 10
The combined entity and time fixed effects regression model is
Yit = β1 Xit + αi + λt + uit , where αi is the entity fixed effect and λt is the
time fixed effect.

This model can equivalently be represented using n − 1 entity binary indicators

and T − 1 time binary indicators, along with an intercept:

Yit = β0 + β1 Xit + γ2 D2i + γ3 D3i + . . . + γn Dni

+ δ2 B2t + δ3 B3t + . . . + δT BTt + uit , (31)

where β0 , β1 , γ2 , γ3 , . . . , γn , δ2 , δ3 , . . . , δT are the unknown coefficients.

33 / 37
Outline Review of Econometrics I Panel Data Fixed Effects Next Time References

Estimation

Consider the following regression model:

FatalityRateit = β0 + β1 × BeerTaxit + αi + λt + uit (32)

# Application to Traffic Deaths : Both state and time fixed effects

r2 = plm ( mrall∼beertax , data = mydata , index = c ( " state " ," year " ) ,
model = " within " , effect = " twoways " )
r2 _ vcov = vcovHC ( r2 , type = " HC1 " )
r2 _ se = sqrt ( diag ( r2 _ vcov ) )
stargazer ( r2 , se = list ( r2 _ se ) , type = " text " , title = " Estimaiton Results " )

The estimation results are given in Table 1. The estimated model is

FatalityRateit = −0.66BeerTaxit + α
bi + λ (33)
\ bt

The result −0.66 is close to the estimated coefficient for the regression model
including only entity fixed effects.
The coefficient is less precisely estimated but significantly different from zero at
10% significance level.

34 / 37
Outline Review of Econometrics I Panel Data Fixed Effects Next Time References

Estimation

In view of (24) and (33), we conclude that the estimated relationship between
traffic fatalities and the real beer tax is not affected by omitted variable bias
due to factors that are constant over time.
Table 2: Estimation Results
Dependent variable:
mrall
beertax −0.640∗
(0.350)
Observations 336
R2 0.036
Adjusted R2 −0.149
F Statistic 10.513∗∗∗ (df = 1; 281)
Note: ∗ p<0.1; ∗∗ p<0.05; ∗∗∗ p<0.01

35 / 37
Outline Review of Econometrics I Panel Data Fixed Effects Next Time References

Next Time

We will revisit fixed effects regression models and make a summary.

We will see the fixed effects regression assumptions.
We will see an application on drunk driving laws and traffic deaths.

36 / 37
Outline Review of Econometrics I Panel Data Fixed Effects Next Time References

Bibliography I

Hanck, Christoph et al. (2021). Introduction to Econometrics with R. url:

https://www.econometrics-with-r.org/index.html.
Keane, Michael P. (2010). “A Structural Perspective on the Experimentalist
School”. In: The Journal of Economic Perspectives 24.2, pp. 47–58. (Visited on
01/01/2024).
Stock, James H. and Mark W. Watson (2020). Introduction to Econometrics.
Fourth. Pearson.

37 / 37