random effects estimation random effects regressions when the observed variables of interest are...
TRANSCRIPT
Random effects estimation
RANDOM EFFECTS REGRESSIONS
When the observed variables of interest are constant for each individual, a fixed effects regression is not an effective tool because such variables cannot be included.
it
s
ppip
k
jjitjit tZXY
121
it
k
jjitj
iti
k
jjitjit
utX
tXY
21
21
1
Random effects estimation
In this section, we will consider an alternative approach, known as a random effects regression that may, subject to two conditions, provide a solution to this problem.
it
s
ppip
k
jjitjit tZXY
121
it
k
jjitj
iti
k
jjitjit
utX
tXY
21
21
RANDOM EFFECTS REGRESSIONS
2
Random effects estimation
The first condition is that it is possible to treat each of the unobserved Zp variables as being drawn randomly from a given distribution.
it
s
ppip
k
jjitjit tZXY
121
it
k
jjitj
iti
k
jjitjit
utX
tXY
21
21
RANDOM EFFECTS REGRESSIONS
3
Random effects estimation
If this is the case, the i may be treated as random variables (hence the name of this approach) drawn from a given distribution and we may rewrite the model as shown.
it
s
ppip
k
jjitjit tZXY
121
itiitu it
k
jjitj
iti
k
jjitjit
utX
tXY
21
21
RANDOM EFFECTS REGRESSIONS
4
Random effects estimation
We have dealt with the unobserved effect by subsuming it into a compound disturbance term uit.
it
s
ppip
k
jjitjit tZXY
121
itiitu it
k
jjitj
iti
k
jjitjit
utX
tXY
21
21
RANDOM EFFECTS REGRESSIONS
5
Random effects estimation
The second condition is that the Zp variables are distributed independently of all of the Xj variables.
it
s
ppip
k
jjitjit tZXY
121
itiitu it
k
jjitj
iti
k
jjitjit
utX
tXY
21
21
RANDOM EFFECTS REGRESSIONS
6
Random effects estimation
If this is not the case, , and hence u, will not be uncorrelated with the Xj variables and the random effects estimation will be biased and inconsistent. We would have to use fixed effects estimation instead, even if the first condition seems to be satisfied.
it
s
ppip
k
jjitjit tZXY
121
itiitu it
k
jjitj
iti
k
jjitjit
utX
tXY
21
21
RANDOM EFFECTS REGRESSIONS
7
Random effects estimation
If the two conditions are satisfied, we may use this as our regression specification, but there is a complication. uit will be subject to a special form of autocorrelation and we will have to use an estimation technique that takes account of it.
it
s
ppip
k
jjitjit tZXY
121
itiitu it
k
jjitj
iti
k
jjitjit
utX
tXY
21
21
RANDOM EFFECTS REGRESSIONS
8
Random effects estimation
First, we will check the other regression model assumptions. Given our assumption that it satisfies the assumptions, we can see that uit satisfies the assumption of zero expected value.
it
s
ppip
k
jjitjit tZXY
121
itiitu
0)()()()( itiitiit EEEuE
it
k
jjitj
iti
k
jjitjit
utX
tXY
21
21
RANDOM EFFECTS REGRESSIONS
9
Random effects estimation
Here we are assuming without loss of generality that E(i) = 0, any nonzero component being absorbed by the intercept, 1.
it
s
ppip
k
jjitjit tZXY
121
itiitu
0)()()()( itiitiit EEEuE
it
k
jjitj
iti
k
jjitjit
utX
tXY
21
21
RANDOM EFFECTS REGRESSIONS
10
Random effects estimation
uit will satisfy the condition that it should have constant variance. Its variance is equal to the sum of the variances of i and it. (The covariance between i and it is 0 on the assumption that i is distributed independently of it.)
it
s
ppip
k
jjitjit tZXY
121
itiitu
22,
2222 2 itiitiitiitu
0)()()()( itiitiit EEEuE
it
k
jjitj
iti
k
jjitjit
utX
tXY
21
21
RANDOM EFFECTS REGRESSIONS
11
Random effects estimation
uit will also be distributed independently of the values of Xj, since both i and it are assumed to satisfy this condition.
it
k
jjitj
iti
k
jjitjit
utX
tXY
21
21
itiitu
0)()()()( itiitiit EEEuE
22,
2222 2 itiitiitiitu
it
s
ppip
k
jjitjit tZXY
121
RANDOM EFFECTS REGRESSIONS
12
Random effects estimation
Individual Time u
1 1 1 + 11
1 2 1 + 12
1 3 1 + 13
2 1 2 + 21
2 2 2 + 22
2 3 2 + 23
it
k
jjitjit utXY
2
1 itiitu
However there is a problem with the assumption that its value in any observation be generated independently of its value in all other observations.
RANDOM EFFECTS REGRESSIONS
13
Random effects estimation
Individual Time u
1 1 1 + 11
1 2 1 + 12
1 3 1 + 13
2 1 2 + 21
2 2 2 + 22
2 3 2 + 23
it
k
jjitjit utXY
2
1 itiitu
For all the observations relating to a given individual, i will have the same value, reflecting the unchanging unobserved characteristics of the individual.
RANDOM EFFECTS REGRESSIONS
14
Random effects estimation
Individual Time u
1 1 1 + 11
1 2 1 + 12
1 3 1 + 13
2 1 2 + 21
2 2 2 + 22
2 3 2 + 23
it
k
jjitjit utXY
2
1 itiitu
This is illustrated in the table above, which shows the disturbance terms for the first two individuals in a data set, assuming that there are 3 time periods.
RANDOM EFFECTS REGRESSIONS
15
Random effects estimation
Individual Time u
1 1 1 + 11
1 2 1 + 12
1 3 1 + 13
2 1 2 + 21
2 2 2 + 22
2 3 2 + 23
it
k
jjitjit utXY
2
1 itiitu
The disturbance terms for individual 1 are independent of those for individual 2 because 1 amd 2 are generated independently. However the disturbance terms for the observations relating to individual 1 are correlated because they contain the common component 1.
RANDOM EFFECTS REGRESSIONS
16
Random effects estimation
Individual Time u
1 1 1 + 11
1 2 1 + 12
1 3 1 + 13
2 1 2 + 21
2 2 2 + 22
2 3 2 + 23
it
k
jjitjit utXY
2
1 itiitu
The same is true for the observations relating to individual 2, and for all other individuals in the sample.
RANDOM EFFECTS REGRESSIONS
17
Random effects estimation
Individual Time u
1 1 1 + 11
1 2 1 + 12
1 3 1 + 13
2 1 2 + 21
2 2 2 + 22
2 3 2 + 23
it
k
jjitjit utXY
2
1 itiitu
The covariance of the disturbance terms in periods t and t' for individual i is decomposed above. The terms involving are all 0 because is assumed to be generated completely randomly. However the first term is not 0. It is the population variance of .
2
))((
''
''
ititiititiii
itiitiitituu
RANDOM EFFECTS REGRESSIONS
18
Random effects estimation
Individual Time u
1 1 1 + 11
1 2 1 + 12
1 3 1 + 13
2 1 2 + 21
2 2 2 + 22
2 3 2 + 23
it
k
jjitjit utXY
2
1 itiitu
OLS remains unbiased and consistent, despite the violation of the regression model assumption, but it is inefficient because it is possible to derive estimators with smaller variances. In addition, the standard errors are computed wrongly.
2))(( '''' ititiititiiiitiitiitituu
RANDOM EFFECTS REGRESSIONS
19
Random effects estimation
Individual Time u
1 1 1 + 11
1 2 1 + 12
1 3 1 + 13
2 1 2 + 21
2 2 2 + 22
2 3 2 + 23
it
k
jjitjit utXY
2
1 itiitu
We have encountered a problem of the violation of this regression model assumption once before, in the case of autocorrelated disturbance terms in a time series model.
2))(( '''' ititiititiiiitiitiitituu
RANDOM EFFECTS REGRESSIONS
20
Random effects estimation
Individual Time u
1 1 1 + 11
1 2 1 + 12
1 3 1 + 13
2 1 2 + 21
2 2 2 + 22
2 3 2 + 23
it
k
jjitjit utXY
2
1 itiitu
The solution then was to transform the model so that the transformed disturbance term satisfied the regression model assumption, and a similar procedure is adopted in the present case.
2))(( '''' ititiititiiiitiitiitituu
RANDOM EFFECTS REGRESSIONS
21
Random effects estimation
Individual Time u
1 1 1 + 11
1 2 1 + 12
1 3 1 + 13
2 1 2 + 21
2 2 2 + 22
2 3 2 + 23
it
k
jjitjit utXY
2
1 itiitu
However, while the transformation in the case of autocorrelation was very straightforward, in the present case it is more complex and will not be discussed further here. Nevertheless it is easy to fit a random effects model using regression applications such as Stata.
2))(( '''' ititiititiiiitiitiitituu
RANDOM EFFECTS REGRESSIONS
22
Random effects estimation
Individual Time u
1 1 1 + 11
1 2 1 + 12
1 3 1 + 13
2 1 2 + 21
2 2 2 + 22
2 3 2 + 23
it
k
jjitjit utXY
2
1 itiitu
Random effects estimation uses a procedure known as feasible generalized least squares. It yields consistent estimates of the coefficients and therefore depends on n being sufficiently large. For small n its properties are unknown.
2))(( '''' ititiititiiiitiitiitituu
RANDOM EFFECTS REGRESSIONS
23
The table shows the results of performing random effects regressions as well as OLS and fixed effects regressions. In the next slideshow we consider how we should choose the appropriate estimation approach.
RANDOM EFFECTS REGRESSIONS
NLSY 1980–1996Dependent variable logarithm of hourly earnings
OLS Fixed effects Random effects
Married 0.184 0.106 – 0.134 –(0.007) (0.012) (0.010)
Soon-to-be- 0.096 0.045 –0.061 0.060 –0.075married (0.009) (0.010) (0.008) (0.009)(0.007)
Single – – –0.106 – –0.134(0.012)
(0.010)
R2 0.358 0.268 0.268 0.346 0.346
n 20,343 20,343 20,343 20,343 20,343
24
Copyright Christopher Dougherty 2011.
These slideshows may be downloaded by anyone, anywhere for personal use.
Subject to respect for copyright and, where appropriate, attribution, they may be
used as a resource for teaching an econometrics course. There is no need to
refer to the author.
The content of this slideshow comes from Section 14.3 of C. Dougherty,
Introduction to Econometrics, fourth edition 2011, Oxford University Press.
Additional (free) resources for both students and instructors may be
downloaded from the OUP Online Resource Centre
http://www.oup.com/uk/orc/bin/9780199567089/.
Individuals studying econometrics on their own and who feel that they might
benefit from participation in a formal course should consider the London School
of Economics summer school course
EC212 Introduction to Econometrics
http://www2.lse.ac.uk/study/summerSchools/summerSchool/Home.aspx
or the University of London International Programmes distance learning course
20 Elements of Econometrics
www.londoninternational.ac.uk/lse.
11.07.25