AQRM part 2: causal analysis when endogeneity likely

?
  • Created by: charlie
  • Created on: 21-05-17 15:04
O N B S A M W V O R W T F I X E I S X E T
I O F L H A E S M U E D N V K V S H S P O
E I E K X U M L H V X M Y R R Y N A E A D
L T C Y V E G T A G H P I E T A O R R N W
D A G S O I P G N M I Q L L S I I P U W B
U S A E U V W O M T A E O E M H T R T B B
P I G Q B I Q D G F V M A V L O P D A D M
P M D M X A Q R K A C E X A R I M P E C E
P O K L R Q Y U N R Y I R N R G U R F T A
M D T F S R Y C J O G F O C J U S O F O J
E N Q L N C E L B V Q L K E B J S B F U K
K A E F S W Y H S U E U W S S T A L I C M
N R M S E H K F W L O V I T U N D E D N Q
C V A A O F X D A Q O Q T R M F R M N G I
C I K V T N X V O G D M H O U G Y S I Y U
T I B M U N D I T M H U K N Y N Z M F M H
V A B O W M R P F H D V L G V S Z P F E D
M M U H S V L E T V P L A I K P U N I F H
O V W O V N R J Y I P H Q V Q I F E D C I
C C T N L C C R X X A Y P P I S W U Q S P
G F T O Y T P H W F N T T G W G I H P O A

Clues

  • (1) trend captures all relevant differences before + after cut-off (2) IV exogenous (after adjusting for running variable will ply affect Y through X) + (3) No other factors affecting D (other policies) (5, 2, 11)
  • F-stat (2, 9, 4, 2)
  • F-stat>10, 2SLS estimator will be closer to true parameter value (stronger for larger sample sizes) (2, 9, 6, 2)
  • IV assigned randomly as easier to defend exogeneity assumption (2, 13)
  • observe 2 groups over time (treatment/comparison)/ comparison is different from treatment + untreated in both periods/ treatment experiences policy change (4, 2, 4, 8)
  • only ST effect (other factors begin to affect after cut off)/ need large sample size (introduce range of other factors attributing to Y) (5, 2, 8)

Comments

No comments have yet been made

Similar Economics resources:

See all Economics resources »See all research methods resources »