Wooldridge Control Function Approach
Wooldridge Control Function Approach
Lecture 6
Control Functions and Related
Methods
Jeff Wooldridge
NBER Summer Institute, 2007
1. Linear-in-Parameters Models: IV versus Control
Functions
2. Correlated Random Coefficient Models
3. Some Common Nonlinear Models and
Limitations of the CF Approach
4. Semiparametric and Nonparametric Approaches
5. Methods for Panel Data
(1)
Ez u 1 0.
(2)
(3)
(4)
(5)
(6)
estimates.
(7)
(8)
(9)
(10)
(11)
If y 2 1z 2 e 2 0, u 1 , e 2 is independent of
z, Eu 1 |e 2 1 e 2 , and e 2 ~Normal0, 1, then
Eu 1 |z, y 2 1 y 2 z 2 1 y 2 z 2 ,
where / is the inverse Mills ratio
(IMR). This leads to the Heckman two-step
(12)
(13)
(14)
(15)
(16)
(17)
10
11
(18)
(19)
(20)
If we also assume
u 1 , v 2 ~Bivariate Normal
(21)
12
(22)
ASFz 1 , y 2 N 1 x 1 1 1 v i2 ,
(23)
i1
13
(24)
14
15
(25)
16
(26)
18
(27)
2.
just replace y 2 with
19
(28)
20
(29)
1 y 2 1 1 z 2 /1 z 2
21
(30)
(31)
(32)
22
(33)
ASFz 1 , y 2 N 1 x 1 1 , v i2 ;
i1
23
(34)
g 1 z 1 , y 2 , u 1 dF 1 u 1 ,
(35)
24
(36)
25
(37)
26
(38)
Chamberlain approach:
c i1 1 z i 1 a i1 , a i1 |z i ~ Normal0, 2a 1 .
(39)
27
(40)
(41)
(42)
Then
Ey it1 |z i , y it2 , v it2 e1 y it2 z it1 e1
e1 z i e1 e1 v it2
where the e subscript denotes division by
1 2e 1 1/2 . This equation is the basis for CF
estimation.
28
(43)
29
(44)
e1 z i e1 e1 v it2 .
N 1 e1 y t2 z t1 e1
(45)
i1
30
(46)
31