Professional Documents
Culture Documents
ABSTRACT
INTRODUCTION
20,18%
18,08%
20,00%
15,00%
10,00%
8,87% 9,64%
17,34%
12,56% 13,48%
5,00%
0,00%
2006
2007
2008
2009
2010 2011
Years
2012
2013
2014
2015
40,00%
35,00%
30,00%
28,00%
27,00%
29,00%
28,50%
30,00%
29,00%
24,00%
25,00%
20,00%
15,00%
10,00%
5,00%
0,00%
Source: Keberpihakan.org (Center for Economics and Development Unpad)
2006
2007
2008
2009
2010
2011
Years
2012
Through this paper, authors would see the socioeconomics factors and measure the probability of
individuals to obtain higher education. This
motivation invites a question from authors, what are
actually the factors that affecting the rate of
participation in higher education in Indonesia?
Authors hypothesize that household income,
household asset, parents education background,
urban or rural school, and gender have impacts to
individuals to acquire higher education.
This paper proceeds in the following direction.
In the next section will be about previous findings
about factors that may affect individual to have
higher education. In the third section sets a model,
kind of data, and data analyze tools. The fourth
section presents result and discussions of author
estimates, while the final section summarizes and
concludes the discussion
LITERATURE REVIEW
1.
Household income
2.
Household Assets
RESEARCH METHODOLOGY
1. Methodology
Present article explore further and try to see the
factors of probability of being college student by
exploring the household and individual determinant.
The specification of econometric model can be
written as:
= ( , )
Where, is the latent dependent variable
that has the binary code. Which is 1 means that
individual i are the college student in 2014. While 0
means that individual i are not attending college in
2014. are the vector variables that included
household and individual characteristics such as
household size, age, sex, expenditure, household
asset, parent education, urban, and school type.
While is the error term. This research use probit
model to analyze the probability of being college
student in the particular age. Because the coefficient
from probit model can not be interpreted, so that
later we calculate the marginal effect to get the
number of probability.
2.
Data
Marginal
Effect
Coefficient
Household size
0.1781*
Household size2
-0.0146
0.5166***
0.1232
0.1101***
0.0263
Urban
0.2171**
0.0496
Poor status
0.0930
0.0232
Age
-0.1188***
Female
0.1235
0.0293
1.3045***
0.3358
0.7601*
0.1910
1.4048***
0.4705
0.8328**
0.1930
0.5721
0.1397
0.2186
0.0547
0.4336
0.1183
0.3785
0.0937
1.3101***
0.4338
0.1444
0.0369
_cons
-10.5247***
Number of obseravation
* p<0.1, ** p<0.05, *** p<0.001
1624
0.0425
-0.0035
-0.0283
APPENDIX
1. Summary of Statistics
Variable
Obs
Mean
Std. Dev.
Min
Max
Household size
Household size2
Log of per capita expenditure
Log of household asset
Per capita expenditure
Household asset
Urban
Poor status
Age
Female
Senior high school: SMA
Senior high school: SMK
Senior high school: MA
Senior high school type: Public
Senior high school type: Private
Parent education: Elementary
Parent education: Junior high
school
Parent education: Senior high
school
Parent education: College
degree
Parent education: Other school
1764
1764
1626
1761
1626
1764
1764
1764
1764
1764
1764
1764
1764
1764
1764
1764
4.37
23.62
13.83
18.01
1314585.00
183000000.00
0.70
0.02
20.70
0.53
0.43
0.43
0.11
0.54
0.46
0.27
2.12
22.58
0.68
1.67
1322959.00
291000000.00
0.46
0.13
2.07
0.50
0.50
0.50
0.32
0.50
0.50
0.45
1
1
12
5
158604
0
0
0
17
0
0
0
0
0
0
0
13
169
17
22
22800000
3950000000
1
1
24
1
1
1
1
1
1
1
1764
0.17
0.37
1764
0.40
0.49
1764
1764
0.13
0.01
0.33
0.07
0
0
1
1