Documente Academic
Documente Profesional
Documente Cultură
ThisworkislicensedunderaCreativeCommonsAttribution-ShareAlike3.0License.Thislicenseallows
theusertoremix,tweak,andbuildupontheworkevenforcommercialpurposes,aslongastheuser
creditstheauthorandlicensestheirnewcreationsundertheidenticalterms.
http://creativecommons.org/licenses/by-sa/3.0/
http://learningresources.lse.ac.uk/
Yit 1 j X jit t i it
j 2
j2
i 1
Yit j X jit t i Ai it
In the third version of the fixed effects approach, known as the least squares dummy
variable (LSDV) method, the unobserved effect is brought explicitly into the model.
1
Yit 1 j X jit t i it
j 2
j2
i 1
Yit j X jit t i Ai it
If we define a set of dummy variables Ai, where Ai is equal to 1 in the case of an observation
relating to individual i and 0 otherwise, the model can be rewritten as shown.
2
Yit 1 j X jit t i it
j 2
j2
i 1
Yit j X jit t i Ai it
Formally, the unobserved effect is now being treated as the coefficient of the individualspecific dummy variable, the iAi term representing a fixed effect on the dependent variable
Yi for individual i (this accounts for the name given to the fixed effects approach).
3
Yit 1 j X jit t i it
j 2
j2
i 1
Yit j X jit t i Ai it
Having re-specified the model in this way, it can be fitted using OLS.
Yit 1 j X jit t i it
j 2
j2
i 1
Yit j X jit t i Ai it
Note that if we include a dummy variable for every individual in the sample as well as an
intercept, we will fall into the dummy variable trap.
5
Yit 1 j X jit t i it
j 2
j2
i 1
Yit j X jit t i Ai it
To avoid this, we can define one individual to be the reference category, so that 1 is its
intercept, and then treat the i as the shifts in the intercept for the other individuals.
6
Yit 1 j X jit t i it
j 2
j2
i 1
Yit j X jit t i Ai it
However, the choice of reference category is often arbitrary and accordingly the
interpretation of the i not particularly illuminating.
7
Yit 1 j X jit t i it
j 2
j2
i 1
Yit j X jit t i Ai it
Alternatively, we can drop the 1 intercept and define dummy variables for all of the
individuals, as has been done here. The i now become the intercepts for each of the
individuals.
8
Yit 1 j X jit t i it
j 2
j2
i 1
Yit j X jit t i Ai it
Note that, in common with the first two versions of the fixed effects approach, the LSDV
method requires panel data.
9
Yit 1 j X jit t i it
j 2
j2
i 1
Yit j X jit t i Ai it
With cross-sectional data, one would be defining a dummy variable for every observation,
exhausting the degrees of freedom. The dummy variables on their own would give a perfect
but meaningless fit.
10
Yit 1 j X jit t i it
j 2
j2
i 1
Yit j X jit t i Ai it
If there are a large number of individuals, using the LSDV method directly is not a practical
proposition, given the need for a large number of dummy variables.
11
Yit 1 j X jit t i it
j 2
j2
i 1
Yit j X jit t i Ai it
Equivalent to within-groups method:
k
Yit Yi j ( X jit X ji ) ( t t ) it i
j2
However, it can be shown mathematically that the approach is equivalent to the withingroups method and therefore yields precisely the same estimates.
12
Yit 1 j X jit t i it
j 2
j2
i 1
Yit j X jit t i Ai it
Equivalent to within-groups method:
k
Yit Yi j ( X jit X ji ) ( t t ) it i
j2
Thus in practice we always use the within-groups method rather than the LSDV method.
But it may be useful to know that the within-groups method is equivalent to modelling the
fixed effects with dummy variables.
13
Yit 1 j X jit t i it
j 2
j2
i 1
Yit j X jit t i Ai it
Equivalent to within-groups method:
k
Yit Yi j ( X jit X ji ) ( t t ) it i
j2
The only apparent difference between the LSDV and within-groups methods is in the
number of degrees of freedom. It is easy to see from the LSDV specification that there are
nT k n degrees of freedom if the panel is balanced.
14
Yit 1 j X jit t i it
j 2
j2
i 1
Yit j X jit t i Ai it
Equivalent to within-groups method:
k
Yit Yi j ( X jit X ji ) ( t t ) it i
j2
In the within-groups approach, it seemed at first that there were nT k. However n degrees
of freedom are consumed in the manipulation that eliminate the i, so the number of
degrees of freedom is really nT k n.
15
NLSY 19801996
Dependent variable logarithm of hourly earnings
OLS
Fixed effects
Married
0.184
(0.007)
0.106
(0.012)
Soon-to-bemarried
0.096
(0.009)
0.045
(0.010)
0.061
(0.008)
Single
0.106
(0.012)
R2
0.358
0.268
0.268
20,343
20,343
20,343
To illustrate the use of a fixed effects model, we return to the example in Section 1 and use
all the available data from 1980 to 1996, 20,343 observations in all.
16
NLSY 19801996
Dependent variable logarithm of hourly earnings
OLS
Fixed effects
Married
0.184
(0.007)
0.106
(0.012)
Soon-to-bemarried
0.096
(0.009)
0.045
(0.010)
0.061
(0.008)
Single
0.106
(0.012)
R2
0.358
0.268
0.268
20,343
20,343
20,343
The table shows the extra hourly earnings of married men and of men who are single but
married within the next four years. The omitted category in the first two columns is single
men who are still single four years later.
17
NLSY 19801996
Dependent variable logarithm of hourly earnings
OLS
Fixed effects
Married
0.184
(0.007)
0.106
(0.012)
Soon-to-bemarried
0.096
(0.009)
0.045
(0.010)
0.061
(0.008)
Single
0.106
(0.012)
R2
0.358
0.268
0.268
20,343
20,343
20,343
The controls (not shown) are the same as in the example in the first slideshow on panel
data.
18
NLSY 19801996
Dependent variable logarithm of hourly earnings
OLS
Fixed effects
Married
0.184
(0.007)
0.106
(0.012)
Soon-to-bemarried
0.096
(0.009)
0.045
(0.010)
0.061
(0.008)
Single
0.106
(0.012)
R2
0.358
0.268
0.268
20,343
20,343
20,343
The first column gives the estimates obtained by simply pooling the observations and using
OLS with robust standard errors. The estimates are very similar to those in the wage
equation for 1988 in the example in the first slideshow on panel data.
19
NLSY 19801996
Dependent variable logarithm of hourly earnings
OLS
Fixed effects
Married
0.184
(0.007)
0.106
(0.012)
Soon-to-bemarried
0.096
(0.009)
0.045
(0.010)
0.061
(0.008)
Single
0.106
(0.012)
R2
0.358
0.268
0.268
20,343
20,343
20,343
The second column gives the fixed effects estimates, using the within-groups method, with
single men as the reference category. The third gives the fixed effects estimates with
married men as the reference category.
20
NLSY 19801996
Dependent variable logarithm of hourly earnings
OLS
Fixed effects
Married
0.184
(0.007)
0.106
(0.012)
Soon-to-bemarried
0.096
(0.009)
0.045
(0.010)
0.061
(0.008)
Single
0.106
(0.012)
R2
0.358
0.268
0.268
20,343
20,343
20,343
The fixed effects estimates are considerably lower than the OLS estimates, suggesting that
the OLS estimates were inflated by unobserved heterogeneity. Nevertheless the pattern is the
same.
21
NLSY 19801996
Dependent variable logarithm of hourly earnings
OLS
Fixed effects
Married
0.184
(0.007)
0.106
(0.012)
Soon-to-bemarried
0.096
(0.009)
0.045
(0.010)
0.061
(0.008)
Single
0.106
(0.012)
R2
0.358
0.268
0.268
20,343
20,343
20,343
Our findings confirm that married men earn more than single men. Part of the differential
appears to be attributable to the characteristics of married men, since men who are soonto-marry but still single also enjoy a significant earnings premium.
22
NLSY 19801996
Dependent variable logarithm of hourly earnings
OLS
Fixed effects
Married
0.184
(0.007)
0.106
(0.012)
Soon-to-bemarried
0.096
(0.009)
0.045
(0.010)
0.061
(0.008)
Single
0.106
(0.012)
R2
0.358
0.268
0.268
20,343
20,343
20,343
However if we make married men the omitted category, as in the third column, we find that
soon-to-be-married men earn significantly less than married men. Thus part of the
marriage premium appears to be attributable to the effect of marriage itself.
23
NLSY 19801996
Dependent variable logarithm of hourly earnings
OLS
Fixed effects
Married
0.184
(0.007)
0.106
(0.012)
Soon-to-bemarried
0.096
(0.009)
0.045
(0.010)
0.061
(0.008)
Single
0.106
(0.012)
R2
0.358
0.268
0.268
20,343
20,343
20,343
Hence both hypotheses relating to the marriage premium appear to be partly true.
24
11.07.25