

Decomposing Youth Poverty in 22 Countries
315
is the proportion of young adults living in household type
i
in
Taiwan, and
P
i
is the net disposable income poverty rate for young
adults in household type
i
in Taiwan.
To test the BTST income poverty and welfare effectiveness, I
use some of the equations provided by Heuveline and Weinshenker
(2008). The rewritten equation is a function of three vector
factors:
P =
Σ
(
Hi * Mi * [Pi/Mi])
(2)
where
M
i
is the BTST income poverty rate for young adults in
household type
i
in Taiwan and
P
i
/M
i
is the ratio of the net
disposable income poverty rate to the BTST poverty rate for young
adults in household type
i
in Taiwan (i.e., welfare effectiveness). I
use
W
i
to represent
P
i
/M
i
.
Therefore, the equation is rewritten:
P =
Σ
(Hi * Mi * Wi)
(3)
I include 21 countries in addition to Taiwan. For these 21
countries, the same terms are presented in lowercase letters,
producing the following equation:
p =
Σ
(hi * mi * wi)
(4)
I perform decomposition analyses to determine the
contributions of living arrangements, welfare effectiveness, and
BTST income poverty to the difference in poverty rates between
Taiwan and each of other selected countries. The decomposition
equations are:
p-P =
α
-effect +
β
-effect +
γ
-effect
(5)
α
-effect = Q
(
h
i
)
– Q
(
H
i
)
(6)
β
-effect = Q
(
m
i
)
– Q
(
M
i
)
(7)
γ
-effect = Q
(
w
i
)
– Q
(
W
i
)
(8)
More detailed discussion on the analytic method and equations are