Education and Globalization in Southeast Asia Issues and Challenges

118 Mohamad Fahmi

Labour market incentives or marginal benefits of extra education are
denoted by Eij and represent the extra lifetime earnings expected if students
attend higher education. Cij is the forgone earnings or opportunity cost of
individuals aged 17–24 years old if they attend higher education.
To obtain lifetime earnings, I estimate the predicted earnings of the
three age cohorts in the upper secondary and higher education group.
The age cohorts are 25–34, 35–44 and 45–55 years old. I obtain the lifetime
earnings predictions based on some of the regressors above. From these
three separate estimations in each group, I obtain the average for each
individual to estimate the predicted life cycle earnings.
To estimate forgone earnings, I follow Flanerry (2009) and use the
gap between predicted earnings of those who do not continue to higher
education and enter the labour market and predicted of expected income
from any work while study in higher education. The earnings forgone
maybe written as

(5) Ci = (Y^w – Y^s)

where Y^w is the predicted earnings of individuals who decide not to continue
to higher education and Y^s is the predicted income of individuals from
any kind of work when they study in higher education.
I follow Fairlie (2005) and use non-linear Blinder-Oaxaca Decomposition
to estimate the gap from the binary choice model. The non-linear
decomposition between public and private schools can be written as:

(6) H

• j − H
  
  
  • k^ =
  
  
  
  

where Xij and Xik are independent variables for school type j and k. bj and
bk are vectors of coefficient estimates for school types j and k respectively.
Nj and j are the sample sizes for school types j and k. Technically, I use
fairlie a STATA module created by Jann (2006) to estimate the gap.

[Σ

Nj

i=1

F(Xijb^j)

Nj –Σ

Nk

i=1

F(Xikbj)

Nk ]+

[Σ

Nk

i=1

F(Xikb^j)

Nk –Σ

Nk

i=1

F(Xikbk)

Nk ]