Supporting Material is available for this work. For more information, follow the link from
the Table of Contents to "Accessing Supporting Material".
Table of Contents
A System Dynamics model of primary and secondary
education in Nicaragua
M.A. Altamirano, C.E. van Daalen
Delft University of Technology, Faculty of Technology, Policy and Management
P.O. Box 5015, 2600 GA Delft, The Netherlands
Phone: +31 15 2781143
Email : m.a.altamirano@student.tbm.tudelft.nl ; elsd@tbm.tudelft.nl
Abstract
A System Dynamics model of primary and secondary education in Nicaragua has been
developed to assist the National Ministry of Education in analysing the system of
schooling and investigating the impacts of different possible policy decisions. The model
shows that without a change in policy there will only be a small increase in the
percentage of primary and high school coverage and that the number of illiterate people
will more than double over a period of twenty years. The consequences of various
policies have been investigated using the model. These policies include implementing
literacy programs and introducing a program in which families in extreme poverty
receive a subsidy. Preliminary results show that by combining these two policies, an
effect on school coverage as well as on the number of illiterate people can be achieved.
Keywords
national educational system, System Dynamics model, primary education, secondary
education, Nicaragua
Introduction
This paper presents a System Dynamics model of the national system of primary and
secondary education in Nicaragua. The model has been developed to support the National
Ministry of Education of Nicaragua (MECD) in the preparatory phase of the process to
take part in the first group of countries for the World Bank “Education for All” (EFA)
Fast Track initiative’. The aim of the model is to serve as a tool, within this process of
intensive planning, for the analysis of the costs and impacts of different possible policy
measures and their combinations on the overall performance of the national system of
education.
Background
The MECD of Nicaragua is the administrator of the national system of education, which
includes preschool, primary school and high school. Given that nearly the 80% of the
students at these three levels attend public schools, the MECD is the main provider of
education for the country. At present, the educational system of Nicaragua has a coverage
" http://www 1 .worldbank.org/education/adultoutreach/index.efa.asp
of nearly 30% for preschool, 70% for primary school and 45% for high school. The
average number of schooling years for the country is approximately five’.
The main issue addressed in this study was to identify the policy and capacity gaps that
will need to be resolved in order to meet the Millennium Development Goal of providing
every girl and boy with quality primary education by 2015. The investigation should help
to define which goals are reasonable (illiteracy rate, coverage of the programs, average
schooling years), given the financial assistance that will be received, and to determine the
best way to achieve these goals. A System Dynamics model was developed to support the
Direction of Policy Research of the MECD in investigating these issues.
National educational system
The national educational system of Nicaragua is schematised in Figure 1. The two most
important subsystems are demand and supply of education. The demand subsystem
represents the demand for preschool, primary and high school, by the national population
in all the different age ranges. The supply subsystem represents the installed capacity in
terms of teachers and facilities. The top of the diagram presents the instruments or policy
measures, e.g. subsidies for students, literacy programs and an increase of teachers’
salaries, which can be implemented by the MECD to influence the system. The left hand
side of the diagram shows external factors, which affect the performance of the system,
but are not under the control of the MECD. Finally, the main performance indicators are
shown on the right hand side of Figure 1. The performance indicators have been based on
MECD plans (MECD, 2001a).
? Unless otherwise noted, the statistical data cited in the text was provided by the General Direction of
Policy Research of the MECD.
MECD
(Policy Measures)
S
e.g. literacy e.g. subsidies to e.g. increase of
programs cover costs teachers’ salaries,
OUTCOMES OF
vy INTEREST
EXTERNAL average number of
2: schooling years
birth rate
emand supply illiteracy rate
mortality rate Subsystem Subsystem
coverage of the
different programs
national economy
etc.
total budget needed
Figure 1. National educational system.
Related models on basic education
The national educational system of Nicaragua will be represented at a relatively high
level of aggregation, simulating the flow of students throughout the whole educational
system of the country from preschool to high school for both private and public
education. The model aims to help the Ministry of Education to determine the costs of
achieving certain goals and policies (policy and capacity gaps), which is key information
in the process of negotiation within the international cooperation.
Various System Dynamics models that also analyse issues in basic education have been
described in the literature (e.g. Karadeli et a/., 2001 and Terlou ef al., 1991). In models of
educational systems, students move through a chain of classes. A certain percentage of
children from the previous class enters a class each year, a certain percentage progresses
from the current class to the next class and a certain percentage of children drops out of
school each year. The way in which the percentages are determined, the level of detail
contained in the model and the types of factors taken into account in each of these models
depends on the questions to be answered by the model. The two most relevant other
models will be discussed below and will be compared to the model of the situation in
Nicaragua. Following this, the model of the national educational system of Nicaragua
will be described in more detail.
Karadeli et al. (2001) have developed a model to analyse the future quality of the Turkish
basic educational system depending on the budget of the Ministry of National Education.
In the model, students proceed through primary education and secondary education. This
model distinguishes urban and rural school, and within the urban schools a distinction is
made between public and private schools. For each of these different school types, the
model looks at groups of girls and boys. The quality of education is determined by the
student to teacher ratio and the student to class ratio. These ratios are also used to
calculate the passage rates between classes. The number of teachers and the number of
classrooms is calculated using the amount of money available for teachers and
classrooms.
The main differences between the model by Karadeli et al. (2001) and the current model
is that Karadeli et al. distinguish six different groups of students, whereas the current
model is more aggregated with respect to these groups. A single group of students, which
represents the national total, is considered, because the main question is how to reach the
goal that every child should finish primary education, independently of the distribution.
On the other hand Karadeli et al. (2001) have aggregated all classes in primary school
into one stock, because in primary school in Turkey repetition is negligible. In the model
described here, it is important to look at each of the classes in primary school because the
characteristics with regard to repetition and progression are different (repetition rates are
high 6 to 10%). Moreover, dropout rates are considerably higher in the first years of
primary than in the following years, i.e. 11% in first grade vis-a-vis 7% in fifth grade. In
addition, for the situation in Nicaragua it is also important to take into account literacy
classes.
The model described by Karadeli et al. (2001) contains influences of the student to
teacher ratios and the student to class ratios on the progression of students. In the model
described here such a relationship has not been modelled. The students to teacher ratio
has no influence in the dropout or promotion rates, because in principle the model is used
to calculate the number of teachers required given a certain actual ratio, which has been
assumed to be fixed at 40 for primary and high school. Thus measured by this indicator,
the “quality” is assumed to remain the same.
Terlou et al. (1991) have developed a System Dynamics model to investigate the low
efficiency of primary education in Latin America. The model looks at the progression
through primary school and includes causal chains leading to the progression, dropout
and repetition of students. A difference between the model by Terlou et a/. (1991) and the
model described here, is that the current model is geared specifically towards the national
schooling system of Nicaragua, whereas the model by Terlou et al. shows one average
chain of classes in Latin America. The current model is more aggregated than the model
described by Terlou et a/. (1991) in that the model by Terlou ef a/. represents the causes
leading to dropout and repetition in some detail and in the current model these causal
factors were not quantified. The dropout and repetition rates have fixed values (based on
historical data) that remain the same unless a new policy is implemented. Financial
aspects and literacy classes are not included in the model by Terlou ef al. (1991), but are
important for the current model as they are part of the national schooling system.
In summary, the model described in this paper can be compared to the model by Karadeli
et al. (2001) with respect to the level of aggregation i.e. of a national educational system.
It can be compared to the model by Terlou ef a/. (1991) with respect to the types of
problems it addresses, which are related to dropout and repetition rates.
System Dynamics model of the national educational system of Nicaragua
The System Dynamics model which been developed is based on the considerations that
were mentioned above and consists of four main sectors: demand for education, supply of
education, cost calculations, and calculation of performance indicators. These four
sectors will be explained below.
A. EDUCATION DEMAND
This subsystem generates the demand for education for all the age ranges and programs.
The subsystem consists of an aging chain in which children move from one level to the
next. Figure 2 shows that children stay at home for the first two years of their life, after
which they can go to preschool at the age of three. Approximately 30% of children from
3 to 6 years attend preschool education, however only 10% of the children of 3 years old
join the first level of preschool.
From the first level of preschool 80% of the children move on to the second level, and
20% drop out of school. At the second level, the children are joined by another group of
children of four years old who did not attend the first level. Figure 2 also shows children
proceeding through to the third level, from which they can then go onto primary school.
raya) PRESCHOOL 6
First Level Second Level Third Level
3 years : years 5 years
mation
rd Level
Children growing
Oto 2 years
Dropout 2nd level
g
in
Dropout 3rd level
uf of tip System of the Syfftem
4 years
Out of the fy stom
syed
CChildrenfo y ears
Children 3 year
without Pre Schoo!
Children 5 yeres. ———«Chlidren 6 years
going OS os
Figure 2. Children prior to preschool and in preschool.
Figure 3 shows the part of the model in which the children move through the first years
of primary school. The percentage of children who drop out in the first year of primary
school is higher than in the following years (MECD, 2001b). A difference between
preschool and primary school is that some children have to redo a year, which has been
included in the model. The children who drop out of primary school usually do not return
and become functional illiterates after five or more years.
A By = Enrollment
ee esse
Pre Schoo! Dlofowt Dropput =
bat oy a me
Tae cen Sain smcuaa
entaje of Chuldeen
enteljng Primary Schoo!
Children of6 year
iterates
2 years of 15 years
Iiiterafes
1 year of 46 years
iterates
16 or more years
Initerates without any education
of 15 years
Figure 3. Part of the primary school section.
With this same logic flow all the children from six to twelve years pass through the
educational system. A group, defined by the promotion rate which is approximately 90%
of the children enrolled in 6" grade of primary school, moves on to the first year of high
school, another group joins the labour force and the rest remains in the same class
redoing the year. For high school, the flow of students has the same structure as the flow
of students in primary school. Finally, a percentage of the students who finish high school
continues studying and joins university.
B. EDUCATION SUPPLY: TEACHERS AND INFRASTRUCTURE
The education supply sector represents the installed capacity the MECD has to offer in
response to the demand for free or public education. The body of teachers grows annually
through hiring of new teachers and is assumed to decrease only with the retirement,
which has been assumed to occur after approximately 30 years of service. In the case of
primary and high school a distinction has been made between non-graduated and
graduated teachers.
me) TEACHERS OF HIGH SCHOOL
Graduated teacher__
@
Graduated Teachers
%O
GT retired
hired Teachers
receiving training
Teachers HS graduating
Non graduated
teachers on to Training
Teachers non g rat
— ©
graduated hired
Teachers non graduated
retired or fired
Figure 4. High school teachers.
The infrastructure component of this sector consists of the number of classrooms. The
stock of classrooms increases annually with the construction of new ones and decreases
when they reach the end of their useful life (see Figure 5). The average lifetime of a
classroom is 10 years, but improved maintenance can increase the average useful life of a
classroom up to 20 years *,
jas
CONSTRUCTION AND MAINTENANCE OF CLASSROOMS 8
Classrooms in good Classrooms
Conditions Classrooms in damage processwaiting for
5 vant renovation
rs =
Classroms J<—#
constructed 4 ee i Oe
cidserooms | rs)
rejaced
K tional useful life
ie to maintenance
Classrooms Useful life\clasgroom
I towartis end 4
usefl life
\ Maintenance
YES\NO
Classrooms
waiting for
replacement
Towards total damage
Figure 5. Infrastructure component of the model.
3 Source: General Division of Investments and Cooperation, MECD.
C. COST CALCULATION SECTOR
This sector calculates the total budget, including operational and investment costs,
needed to cover the demand for free or public education. Operational costs are calculated
per program and for each program the most important factors consist of the number of
students and number of teachers. Minimum standard costs per students are used, which
includes books, tables, chairs and backpacks with notebooks (Franco, 2001).
It is important to highlight that the model is representative for the whole national
educational system. Therefore, the demand subsystem includes the enrollment in private
and public schools. However, within the cost calculation sector a distinction is made
between the enrollment in public and private schools, given that it is the budget for only
the public education sector that is calculated.
D. CALCULATION OF PERFORMANCE INDICATORS
This sector also represents a group of calculation processes. However, the indicators
resulting from these calculations are representative of the performance of the whole
national educational system and not only of public education.
The main performance indicators are the following:
e Coverage per program: is calculated taking into account the population inside and
outside the program within the right age range.
e Illiteracy rate: reflects the percentage of people aged 15 and over who cannot, with
understanding, both read and write a short, simple statement about their everyday life.
e Average number of schooling years: average number of years of schooling completed
among the population aged 15 and older.
Model testing
Consistency checks and extreme conditions tests were applied to the model. Following
these, a quantitative comparison was carried out for the demand sector and for the
calculation of the performance indicators. For the other sectors (i.e. costs and supply of
education), the data used in the model represent ideal parameters such as minimum
standard costs and hiring of the teachers required. Therefore, a proper comparison with
actual data cannot be carried out for these sectors. The main results of the comparison
between the actual data and the values generated by the model, are shown in Appendix A.
In general it can be observed that there is a high degree of similarity between the actual
values of the national education system and the values generated by the model. The
differences in the enrollment percentages for preschool, primary and high school are of
the order of 2 to 4%. Therefore, it can be assumed that the model shows a general
behaviour representative of the behaviour of the real system it intends to simulate.
Model behaviour
The behaviour of the most important performance indicators (output criteria) are
reviewed below for the medium (1998-2008) and the long-term (2008-2018). This
systematic review aims to increase the understanding of the overall performance of the
system for the next 20 years when no additional policies are implemented.
Preschool coverage
The model shows an increase from 24% in 1998 to 30% in 2002, which is most probably
due to the implementation and expansion of a program of community preschools. This
program is more flexible and accessible to rural areas, than the formal way of working
and is mainly based on teachers working voluntarily with no need for building
classrooms. The actual enrollment of these community preschools increased from 24,000
children in 1995 to 100,000 in 2001 (UNICEF, 2001). From 2002 until 2018 the model
shows that coverage remains stable at 30%. However, the absolute number of children
enrolled increases in the model from approximately 160,000 in 2003 to 196,000 in 2018.
Primary school coverage
This performance indicator represents the percentage of children from 6 to 12 years of
age who are receiving primary education. In the first two years there is an increase of 3%.
From 2000 until 2008 a minor increase of 1% occurs and afterwards this remains stable
with a coverage of 77%.
It is important to mention that the improvement of coverage for the first years could be
overestimated due to the fact that there were no accurate statistics on the number of
children who are not in school and thus the initial values had to be extrapolated. However
this does not present a problem because from 2001 onwards the model does keep accurate
accounting of the children out of the system.
Primary school relative coverage
This performance indicator differs from the previous one in that it represents the
percentage of the total number of children from 6 to 15 years who cannot participate in
any other education program than primary school. This new group includes the children
who, although they are older than 12 cannot attend high school because they have not
finished primary school and cannot yet take part in literacy programs due to their age.
Therefore this indicator represents the magnitude of the challenge for the primary
education subsystem more realistically. After 2000 this indicator remains constant at
68%, which is 9% less than the coverage estimated by the previous indicator.
In terms of the absolute number of children enrolled in primary school, there is in
increase of 80,000 children in the first 10 years and of 89,000 in the last 20 years.
High school coverage
This indicator represents the percentage of the total number of children between 13 and
17 years of age within the high school educational system. As can be seen in Figure 6, the
improvement of primary coverage of 3% that occurred in the first 2 years has an effect on
the coverage of high school in 2005. Afterwards the coverage only improves by 1%,
remaining stable at 49% from 2010 onwards.
High school relative coverage
As was the case for the primary school relative coverage, this indicator aims to calculate
the real gap faced by the high school subsystem. Therefore it represents the percentage of
the children between 13 and 17 who have completed their primary education. Figure 6
shows that after an improvement of 6% from 1998 to 2004, coverage remains at 73%
until 2018. Thus, measured in this way, high school shows a performance similar to
primary school which showed a coverage of 77%.
@ 1: Coverage High School 2: Relative Coverage High School
(MMM pf 0.720 edeeeeedeetn ee tercsraneeseenrereeretes
4]
a errsrienverstsierrerers oss oesece pasion eset Nera umm
—— nt
4 rit basses eneeruonermance frmopenatooen peel 4
es
1]
| 00 7
1998.00 2003.00 2008.00 2013.00 2018.00
lPage 1 2003.00 03:19 p.m. Mé, 18 de Feb de 2004)
s 3 BF 2? Performance Indicators: High School
Figure 6. Performance indicators: high school.
Illiteracy rate
If no literacy program is implemented the illiteracy rate tends to increase, as could be
expected. From 1998 to 2000 the illiteracy rate increases by 2% and it increases by 1%
from 2003 onwards and reaches a value of 30% (Figure 7).
Total number of illiterates
Even though an increase of only 3% in the illiteracy rate does not seem dramatic, when
looking at the actual number of illiterate people the increase does show a serious
problem. The number of illiterate people increases from approximately 640 thousand in
1998 to 1.11 million in 2008 and accounts for 1.54 million in 2018 (Figure 7). This
means that the initial number more than doubles, and there is an annual increase of
approximately 45 thousand illiterate people.
300000
Figure 7. Performance indicators: illiteracy.
Average number of schooling years
As shown in Figure 8 the average number of schooling years increases from 4.6 in 1998
to 5.8 in 2018, which is still under the average of 6 years that Costa Rica reached in 2000,
and under the average of 6.2 years that Chile had already reached in 1960 (UIS/OECD,
2002). However, the curve does not show exactly the same slope throughout the years. It
is somewhat steeper from 1998 until around 2012, which could be due to the
improvements in the first years 5 years in the coverage of the different regular programs.
Afterwards it shows the tendency to flatten out, which means that significant increases of
this indicator cannot be expected later on.
Figure 8. Performance indicators: average number of schooling years.
11
From the review of the behaviour of the performance indicators of the model some
important observations can be made:
e Primary school: even if 100% of the children of 6 years of age enrolls in primary
school, with the current dropout rates, the overall coverage of primary school will
never reach 100%. Therefore, the dropout rates should be considered important
leverage points for further improvement of system performance.
e High school: the new indicator high school relative coverage, which is not a
commonly used indicator but is included in the present model, allows us to observe
that this subsystem has a real performance very similar to the primary school
subsystem. However, a significant improvement in the effectiveness and coverage of
primary school, which now represents a bottleneck, would require an enormous
expansion of high school capacity in order to maintain the actual performance.
e Illiteracy: the illiteracy rate in itself is not sufficient as an indicator of the illiteracy
problem in a country. The value of 2.4 times the initial number of illiterate people is
evidence of the chronic problem of the educational system, especially in primary
schools where the dropout rates are particularly high.
The performance of the system discussed above will be considered to be the zero option
for the comparison of alternative policy measures that will be presented in the following
section.
Investigating the impact of different policies
The impact of two different policy options will be studied below. The first policy consists
of intensifying literacy programs, and the second consists of providing a subsidy to cover
the opportunity costs of parents when their children go to school.
Policy option 1: Literacy programs
Literacy programs aim to alleviate the illiteracy problem by implementing a three year
course for functionally illiterate people who are over the age of 15 years. Literacy
programs in Nicaragua started only recently (in 1998). Present enrollment is quite low,
but efforts in this direction could be stepped up.
The MECD is considering proposing a target of 50 to 60 thousand people a year to
graduate from the literacy program to the international cooperation, with the aim to solve
the illiteracy problem by the year 2015. The graduation target corresponds to an
enrollment number of around 60 thousand people each year in the first level. This means
that the enrollment of the three levels should make up a total of around 180,000 people
receiving education within the literacy program. The implementation cost of this policy is
estimated at around 8 to 9 million US dollars per year.
In the current model, a literacy program does not show any impact in the performance of
preschool, primary and high school education. It is important to mention that in reality
there does seem to be an influence of the level of education of the parents on the
assistance of children and their performance in school, which has been widely mentioned
in previous studies. A review of 67 research studies on literacy programs (Padak and
Rasinski, 1997) found that as parents spend time in adult literacy programs, their attitudes
towards education change: the more literate they become, the more value they perceive in
education, the more they support their childrens’ learning and the more they become
involved in their childrens’ schools. The result is that their childrens’ school achievement
jumps (Padak and Rasinski, 1997). The same report found that as parents continue to
spend time in literacy programs, their children attend school more regularly, achieve
higher IQ scores, and are more likely to complete their education. Moreover, the results
of a set of comparative tests applied around 1990 to measure the academic achievement
of Latin American students showed that children whose parents have a low educational
level score poorly on all these tests (UNESCO/OREALC, 2000). The scores obtained by
children who took the UNESCO/OREALC test increased by 1.1 points for every
additional year of their parents’ education. However, this feedback effect was not
included in the model, given that the statistical data available did not allow for a proper
quantification of this relationship in the model. This should be taken into account when
interpreting the model results.
Illiteracy
When introducing the literacy policy, the model shows a significant improvement of the
illiteracy rate from 30% in 1998 to 21% in 2008 and 15% from 2017 onwards (Figure 9).
When compared to the base situation the illiteracy rate is 16% less when the literacy
program is implemented. This improvement is significant compared to the literacy
indicators for the other Latin American countries from 1980 to 2000. Most of the
countries show an improvement of 2 to 3% per year. The only two countries that show a
comparable improvement in 10 years, are Brazil and Peru, which show a decrease of
around 7% in the illiteracy rate from 1980 to 1990 (UNESCO, 1995).
As expected, the number of illiterate people also decreases vis-a-vis the situation when no
policy is implemented. Under this policy the number of illiterate people reaches a value
of approximately 610 thousand in 2018, which represents 930 thousands less than for the
zero option.
However, it should be mentioned that when compared to the number of illiterate people
in 1998, which accounts for around 640,000, the situation does not seem to improve as
much as expected. The actual reduction is only around 30,000 persons in 20 years, even
though as shown in Figure 9, approximately 790,000 persons finish the literacy program
in the same period of time.
This policy is effective in decreasing the illiteracy rate and shows a significant
improvement compared to the base situation, but only this measure is not enough to
achieve the improvements aimed with the EFA plan. As can be seen from the figures
presented above, this program can ameliorate the illiteracy problem but will not eradicate
it unless the actual efficiency of the basic education system improves.
@ 1: Witerstes total 2: Adults from 17 to 60 years with an... 3: literacy Rate
yy tate ne
2
3: 1.00
ul 950000 |
3 0.60
d= + 1
’
y |
2 00000 ‘
: 1998.00 2003.00 2008.00 2013.00 2018.00
Page 1 . 2006 00 14:20 don 19 feb 2004
Veet 2? Performance Indicators: literacy
Figure 9. Illiteracy indicators when the literacy policy is implemented.
Average number of schooling years
The literacy policy causes an increase in the average number of schooling years from 4.6
in 1998 to 6.3 in 2018. Compared to the base case this means an improvement of half a
year in the overall average of the country.
Policy option 2: Subsidy to cover the opportunity costs
This policy intends to prevent children from dropping out of school and in this way to
ensure that a larger percentage of school-age children attend school. As stated by the
USAID Office of Women in Development (USAID, 1997), an important barrier for a
total coverage of the primary education is the cost of schooling, which is often
prohibitively expensive for families in poverty. In addition to paying for their daughters’
books, clothing, and other costs, parents must factor in the foregone income from the
child earning a wage or helping out with household tasks.
In Brazil a program, called Bolsa-Escola, which provides financial assistance to parents
has been in place since 1995. This program consists of granting a monthly allowance to
families with children aged 7 to 14 and an average family income below that of the
average state income, as long as the parents keep their children in school. With
approximately 8 million children being beneficiated each year, Brazil has come very
close to fulfilling the basic principle of making education universal. In 1999, 96% of all
children aged 7 to 14 were attending school, compared to 89% in 1994",
A similar program has been implemented in Guatemala. This country has been a leader in
offering scholarships to keep indigenous girls in school (USAID, 1997). USAID's Basic
Education Strengthening (BEST) Project built on earlier efforts to provide scholarships
4 Providing Universal Education, Embassy of Brazil in London, 2004. Website:
http://www. brazil.org.uk/page.php?cid=504
and to evaluate the effect of these scholarships on girls’ school dropout rates. When local
community workers assisted parents’ organizations in providing scholarships for girls,
the first grade promotion rates for scholarship recipients were at least 20 percentage
points higher than for non-scholarship recipients over a three-year period. In 1996, for
example, 87.5 percent of the girls in the scholarship program went on to second grade, as
compared to 61.9 percent of a control group of girls who did not receive scholarships
(USAID, 1997).
The policy which was investigated using the model, consists of giving a monetary
subsidy equivalent of 100 US dollars to the parents of children in risk of leaving primary
school due to poverty reasons. This is equal to the subsidy received by the Brazilian
families under the Bolsa-Escola program mentioned above. The national Nicaraguan
census of 1998 shows that 19.6% of the children who are actually in the educational
system live in extreme poverty (EMNV, 1998). This group of children is considered to be
the target group for the present policy. The results to be presented correspond to the
impact of the policy when 30% of these children receive the subsidy. An important
assumption in the model is that when assigning the subsidy, students from first to sixth
grade are prioritized in ascending order. The program is assumed to be successful in 30%
of the cases, given that in some cases it does not succeed in retaining the students at
school or the subsidy is assigned to children who would continue studying anyway. This
percentage of success of 30% is considered comparable to the success of the Brazilian
program.
The resulting costs of the implementation of the program are also estimated to amount to
approximately 8 to 9 million US dollars per year. The direct costs of the subsidy program
amount to four to five million dollars and the remainder goes towards the additional
operational costs of the primary program due to the increase in enrollment. Moreover, the
capacity needed to accommodate this growth in enrollment means around three thousand
more teachers and classrooms compared to the base case, and around ten thousand more
than the actual capacity. This seems a major challenge when analysing the causes of the
actual deficit of teachers the educational system faces presently. Until now an increase of
800 teachers per year has been necessary, but has not been realized because the budget is
insufficient. Moreover, covering the teacher deficit, even when enough money would be
available, would mean a higher proportion of non-graduated teachers. The teacher
training colleges are not educating/graduating enough teachers per year. The colleges are
being used to only 50% of their total capacity because there is not enough budget to cover
scholarships and there is not enough demand. Increasingly fewer high school students
choose a teaching profession due to the low socio-economical status of teachers and the
low salaries. The monthly salary amounts to only 70 American dollars for primary school
teachers and 150 for high school teachers.
Primary school coverage
Unlike the literacy policy, the subsidy does have a direct impact on the coverage of
primary and high school, which can be seen in Figures 10 and 11. The coverage of
primary school increases from 73% in 1998 to a maximum of 83% from 2009 onwards
(Figure 10). Compared to the base case this represents an improvement of 6% which also
holds for the indicator of relative coverage.
@ 1: Coverage Preschool 2: Cowerage Primary 3: Relative Coverage Primary
1 0.30 ee)
jit boconconnonesacenn
ene
2
3
00° + "
1998.00 2003.00 2008.00 2013.00 2018.00
ear 2008.00 12:11 don 11 maa 2004
Seer 2? Performance Indicators PreSchool and Primary
Figure 10. Coverage of preschool and primary school when the subsidy policy is
implemented.
High school coverage
The coverage of high school, as shown in Figure 11, reaches a maximum of 53%. This
corresponds to an improvement of approximately 4% with respect to the maximum of
49% coverage in the base case. However, the indicator of relative coverage does not
show any improvement when compared to the zero option. This shows that the policy
indeed did not affect the internal efficiency of the high school subsystem, in which
dropout rates remain the same.
@ 1: Coverage High School 2: Relative Coverage High School
i} 0.53 73 %
a 0.504
1
Hi 0.00 .
1998.00 2003.00 2008.00 2013.00 2018.00
Page 1 2013.00 03:35 pm. ue, 19 de Feb de 21
Seger ? Performance Indicators: High School
Figure 11. Coverage of high school when the subsidy policy is implemented.
Illiteracy rate
The illiteracy rate shows an improvement of 2% when compared to the base case, which
means an increase of 1% compared to the initial value of 27% (Figure 12). Similarly the
magnitude of the illiterate population shows an improvement of only approximately 120
thousand people when compared to the base case.
@ 1: iliterates total 2: Adults from 17 to 60 years wit... 3: Iliteracy Rate
i) Ee 401954 freeones leerio2 «© pemnentererenny ah
3: 1.00
1]
i
3
I 2.
| 300000 pa
"1998.00 2003.00 2008.00 2013.00 2018.00
. 2017.00 12:11 don 11 maa 2004
BaF ? Performance Indicators: lliteracy
Figure 12. Illiteracy rate when the subsidy is implemented.
Average number of schooling years
The implementation of the subsidy to cover the opportunity costs has an impact on the
average number of schooling years of just one fifth of a year more than when no policy is
implemented.
Table 1 shows a summary of results for different variations of the subsidy policy. The
number of students to receive a subsidy was altered as well as the effect of the subsidy on
the children remaining in school.
17
Table 1. Performance indicators for a number of alternatives in the implementation of the
subsidy policy.
No Policy 30% 69% 49% 73%
Subsidy (30%+30%effect) 30% 15% 53% 73%
Subsidy (60%+30%effect) 30% 81% 57% 73%
Subsidy (60%+40%effect) 30% 86% 60% 23%
Subsidy (100%+40%effect) 30% 90% 64% 23%
Subsidy (100%+100%
effect)
No Policy 31% 15 5.8
Subsidy (30%+30%effect) 28% 1.4 6.0 9 million
Subsidy (60%+30%effect) 25% 1,3 6.3 18 million
Subsidy (60%+40%effect) 23% 1.1 6.5 22 million
Subsidy (100%+40%effect) 20% 1 6.8 37 million
Subsidy (100%+100%
effect) 18% 0.9 7.3 40 million
Combining literacy programs and subsidies to cover the opportunity costs
As the results of the two different types of policies (i.e. literacy programs and providing
financial assistance to poor families) show, it is not possible to judge one of the policies
as performing better than the other. It seems that each has a positive effect on different
factors of the education problem and neither of these alone improves the overall
performance of the educational system.
For this reason, the model was run to simulate the impact of both policies at the same
time. Table 2 presents the performance of each of the two policies, as well as the
performance of both policies when implemented at the same time. The numbers in the
table are shown relative to the base case. As can be seen from the results, the
implementation of both policies does achieve an improvement in all the performance
indicators.
18
Table 2. Relative performance of policies compared to base case for the year 2018.
Literacy 0% 0% 0% 0% 0%
Subsidy (30%+30%
effect) 0% 6% 6% 4% 0%
Literacy & Subsidy 0% 6% 6% 4% 0%
Subsidy
(60%+30%effect) 12% 8% 0%
Literacy j 0.5 9 million
Subsidy
(30%+30%effect) -3% -0.1 0.2 9 million
Literacy & Subsidy -19% -l 0.7 18 million
Subsidy
(60%+30%effect) -6% -0.3 0.5 18 million
Conclusions and discussion
An aggregated model of the educational system of Nicaragua was developed in order to
support the National Ministry of Education (MECD) in identifying and analysing the
consequences of policies that are aimed at improving the coverage of the different
educational programs, reducing illiteracy and increasing the average number of schooling
years of the population. The model shows that without a change in policy there will only
be a small increase in the percentage of primary and high school coverage and that the
number of illiterate people will more than double over a period of twenty years. The
consequences of various policies have been investigated using the model. The two
policies that were studied in the current analysis consisted of intensifying literacy
programs and introducing a program in which families in extreme poverty receive a
subsidy to cover the opportunity costs of their children when they go to school.
The model shows that the education problem in the country has many components and
therefore a combination of different policy measures seem to be necessary in order to
solve the problem of different target groups. In addition, the effectiveness of each of the
educational programs influences the performance of the other subsystems. In this respect,
the effectiveness of the basic education system (primary education) showed to be key for
a sustainable improvement of the illiteracy problem as well as for a higher performance
of the high school program. Dropout rates in primary school showed to be important
leverage points for a significant improvement of the education system as a whole.
Therefore, further research of policies, such as the subsidy to parents, that aim to directly
influence these rates, is needed.
More information and understanding of the retention and promotion functions in primary
school is required in order to obtain a more accurate evaluation of the policy alternatives.
A better understanding and quantification of the factors affecting these rates would allow
the incorporation of feedback effects of the different policies in the model. One feedback
effect that has been widely mentioned in the literature and deserves special attention is
the relationship between the education of the parents and the probability of children
dropping out of school. The question is how the positive effect of literacy programs on
the drop out rates of primary school and vice versa can be measured. This effect has not
been included in the present model, but it could have a significant effect on the behaviour
of the model with respect to retention rates in school. It would mean that the results of the
literacy programs would not only influence the number of illiterate people but also have
an influence on the drop out rates in schools.
To conclude, it is important to mention that this paper aims to call for suggestions for
further enrichment of the model and for ideas on the most accurate ways to model policy
initiatives and measure their feedback in the educational system.
References
EMNV (1998). Encuesta Nacional de Hogares sobre Medicion del Nivel de Vida, 1998.
Instituto Nacional de Estadisticas y Censos (INEC).
Franco, S. De (2001). Informe Final Sobre el Marco Operativo del Plan Nacional de
Educacion: Costos Estandares de la Educacion General y Proyecciones 2002-2010.
Universidad Thomas More para MECD.
Karadeli, N.; O. Kaya and B.B. Keskin (2001). Dynamic Modeling of Basic Education in
Turkey. Senior graduation project, Bogazici University, Turkey.
MECD (2001a). Plan Nacional de Educacion (2001-2005), Ministerio de Educacion,
Cultura y Deportes, Managua.
MECD (2001b). Tasas de Escolarizacion por Nivel “TNE y TBE” (Preescolar, Primaria
y Secundaria). Ministerio de Educacion, Cultura y Deportes, Managua.
Padak N. and T. Rasinski (1997). Family Literacy Programs: Who Benefits? Ohio
Literacy Resource Center, Kent State University, OH.
Terlou, B.; E. van Kuijk and J.A.M. Vennix (1991). A System Dynamics model of the
efficiency of primary education in Latin America. Proceedings of the International
Conference of the System Dynamics Society, pp. 578-587.
UIS/OECD(2002). Financing Education - Investments and Returns. Analysis of the
World Education Indicators, 2002 Edition, UNESCO/OECD, Paris.
UNESCO (1995). Compendium of Statistics on Illiteracy, UNESCO, United Nations,
New York, NY.
UNESCO/OREALC (2000). Informe Subregional de América Latina de EFA.
Santiago,Chile: OREALC.
UNICEF (2001). The State of the World’s Children 2001. ISBN 9280636332, 116 p.
USAID (1997). Educating indigenous girls in Latin America: Closing the gap.
Information Bulletin, October 1997. A Publication of USAID’s Office of Women in
Development.
20
Appendix A. Quantitative comparison
[ DEMAND FOR EDUCATION |
Unless otherwise noted, the statistical data cited was
Source: provided by the General Direction of Policy Research
of the MECD
Variable Year Actual Value calculated by the
value model
(*1000) (*1000)
Total Pre-School 2000 167 159
enrollment 2003 179 160
Coverage of Pre-School 2003 30% 30%
. 2000 838 870
Total Primary enrolment 2001 867 393
Total High School 2000 315 307
enrolment 2001 335 328
. 2000 5,072 5,105
Total Population 3002 5.342 3.391
Illiterates 2001 715 729
CALCULATION OF PERFORMANCE INDICATORS OF THE
EDUCATION SYSTEM
Human Development Report 2003. UNDP.
Source:
www.undp.org
Indicator Year Value Value calculated by the
model
Adult literacy rate (% 2001 69.1% 11%
age 15 and above)
Combined primary, 2001 65% 68%
secondary and tertiary
gross enrollment ratio
Official Presentation, MECD:
"Estado de la formaci6n de la ciudadania y Recursos
Humanos”, MECD
Souree: II. La Educacion en Nicaragua
Diagnéstico a lo Interno del Sistema Educativo
http://www.mecd.gob.ni/educ.as
Indicator Year Actual value Value calculated by the
model
2003 5 years 4.9 years
Average schooling years
Population without a
complete High School
Education (from 18 to
45 years)
2003
1,435
thousand
(71.5%)
1,488 thousand
Education Indicators 2000, Nicaragua, Ministry of
Source: Education, Culture and Sports (MECD)
http://www.mecd.gob.ni/
Indicator Year Actual Value | Value calculated by the
Model
Coverage of Pre-School | 2000 27% 29%
Coverage of Primary 2000 80% 76%
Coverage of High 2000 43% 41%
School
Official Presentation, INATEC (National Institute for
Technical Education):
Sores: System of Technical Agricultural Education and
. Training (SETAC)
http://www.senar.org.br/seminario/ppt/Luz%20Marina/d
)%201UZ%20MARINA-EPT-2-2003.ppt
Indicator Year Actual value | Value calculated by the
model
Average Schooling 2003 4.8 years for | 4.9 years
years
men
5 years for
women
Back to the
22
