F12 Panel data II (dynamic models) Flashcards
What is a dynamic panel data model?
The dynamic contains a lagged term of the dependent variable
What is a lagged variabel?
The outcome is measured later in time (t+1). Could be yearly, quarterly, monthly etc.
There is a lagged effect vs. an immediate effect
When does it make sense to lag a variable?
It depends on the theory. If an effect of an intervention is expected to take time to manifest, then the variable should be lagged.
Military conflict –> causalities (immediate)
Causalities –> trust in government (lagged)
What is Jan’s strategy with lagged variables?
Try different lags. If the results a robust it’s more convincing.
What is the difference between a panel data model and a dynamic panel data model?
The dynamic model includes a lagged term (t-1) that is significant (different from zero).
In dynamic models there is a degree of autocorrelation in the dependent variable. The value in t-1 affect the value in t.
There is no autocorrelation in the error term per design. The linear regression assume that error terms are independent from each other.
What happens if you incorrectly apply pooled regression to a dynamic panel data format?
The lagged term (one of our predictor variables) is correlated with the error term and unit-specific fixed effects.
OLS becomes biased and inconsistent because we assume E(ε|x)=E(ε).
What are the two approaches to deal with dynamic panel data models?
Anderson Hsiao-estimator (first-difference estimator) or GMM (generalized methods of moments)
Explain the Anderson Hsiao-estimator
The estimator eliminates unit-specific FE by using a first-difference transformation (eliminate time-invariant factors).
The lagged term (t-1) is still correlated with the error term. This is fix by using another lagged term (t-2) as an instrument.
What does it mean to first-difference transform data?
We simply difference the variables We subtract the value from the previous period. A change variable capital Delta.
As time invariant factors does not vary over time their influence is ruled out.
Why is y_i,t-2 a valid instrument in dynamic panel data models?
y_i,t-2 only affect y_i,t through y_i,t-1, so it meets the conditions for a valid instrument (exclusion criterion).
Relies on the assumption of only one significant time period. If there are several significant lagged periods use GMM.
What are two other problems with the Anderson Hsiao-estimator
(1) If the effect of the lagged term is close to one it’s a sign of an extremely high degree of autocorrelation. Only marginal changes over time e.g. institutional settings are sticky over time (GINI). Especially a problem with binary measures (democracy or not).
(2) Small sample size. Difficult to estimate the correct value of lags.
What is GMM and when is it used?
GMM stands for generalized methods of moments and is used when there are more than one significant lagged period.
It’s possible to include further lags and additional instruments.
Improves efficiency and reduce bias of the estimator.
What is important for GMM?
All instruments must meet the exclusion criterion
How can you test whether all instruments meet the exclusion criterion in GMM?
Via the Sargan test. Tests whether the dependent instrument only affects the dependent variable through the autocorrelated term.
Stars/significance is bad.
What package is used in R?
PLM. Can among other things introduce multiple lags and see how many is significant.