Exercises Part III

Question 1

What does cross-sectional dependence mean? See this question

Answer 1

Question 2

Discuss the two main approaches to model cross-sectional dependence.

Answer 2

Question 3

Answer 3

Question 4

Answer 4

Question 5

What is a random coefficients model?

Answer 5

Question 6

Answer 6

Question 7

Answer 7

Question 8

Answer 8

Mention: Unobserved Common Factors!

Question 9

(just b)

Answer 9

Question 10

(just c)

Answer 10

Question 11

Answer 11

Question 12

Answer 12

The assumptions mentioned in the screenshot are as follows:

1. Heterogeneity of (\beta_i): (\beta_i = \beta + \nu_i), with specific conditions on (\nu_i).
2. Unobserved common effects ((f_t)): Stationary with absolute summable autocovariances.
3. Individual-specific errors ((\epsilon_{it}) and (\nu_{it})): Independently distributed, following stationary processes.
4. Factor loadings ((\gamma_i) and (\Gamma_i)): Independently and identically distributed across units, with specific constraints.
5. Rank condition: The number of unobserved factors ((m)) must be less than the number of observations.

Question 13

Answer 13

Question 14

just b

Answer 14

Question 15

Answer 15

Question 16

Answer 16

Question 17

just b

Answer 17

Question 18

just c

Answer 18

Question 19

just d

Answer 19

Question 20

What is the common correlated effects estimation?

Answer 20

Question 21

What is pooled OLS?

Answer 21

Question 22

What is the fixed effects estimator?

Answer 22

Question 23

What is the Anderson-Hsiao estimator?

Answer 23

Question 24

What is the Arrelano-Bond estimator?

Answer 24

Question 25

What is the Nickell bias?

Answer 25

Question 26

“Cross-sectional dependence” in panel in panel data occurs when the error term of the panel data model exhibits serial correlation across time.

Answer 26

Question 27

Pesaran (2006) suggests using time series average to account for cross-sectional dependence.

Answer 27

Question 28

Spatial approach to model cross-sectional dependence can only be used for geo- graphical data as it requires a geographical distance.

Answer 28

Question 29

“Fixed effects estimator” is an estimator that is used to estimate panel data mod- els where there is an unobserved individual specific time-fixed component in the error term.

Answer 29

Question 30

“Nickell bias” is the bias that is caused by the presence of cross-sectional depen- dence in the error term of a panel data model.

Answer 30

Question 31

Anderson-Hsiao proposes a solution to the Nickell bias problem that arises in es- timating dynamic micro panels. They simply propose to use a bigger data set with a longer time series dimension to eliminate the bias.

Answer 31

Question 32

Arrelano and Bond estimator is an efficient estimator that eliminates the Nickell bias.

Answer 32

Question 34

What is the hetrogeneity of beta assumption? (From CCE)

Answer 34

Question 35

What is the unobserved common effects assumption? (From CCE)

Answer 35

Question 36

What is the individual specific error assumption? (From CCE)

Answer 36

Question 37

What is the factor loadings assumption? (From CCE)

Answer 37

Question 38

What is the rank condition assumption? (From CCE)

Answer 38