The determinants of success in Baccalaureate in Morocco :
An analysis using spatial panel data
Jabrane amaghouss1 (GRES, Cadi Ayyad university)
Ibourk (GRES, Cadi Ayyad university)
1. Abstract:
Morocco has worked since independence to develop its education and training system to face
double challenge of upgrading and extension. Despite the reforms, there is a persistent
weakness of the Moroccan educational system and a process of catching up is still far.
Thus, we observe the contrasts between the developed regions (Rabat and Casablanca
respectively administrative and economic capital) and other regions especially the
mountainous one. Spatial analysis is widely discussed by reference to the dynamics of
monetary variables (income, earnings, GDP ...). Those based on socioeconomic variables is
rare. The objective of this research is to explore the extent of spatial disparities in education,
test the convergence performance of the Moroccan educational system (approached by the
success rate in the Baccalaureate) in a spatial perspective and to analyze the determinants of
these inequalities ((especially those related to the school context.) we rely on data from the
Ministry of Education (2009 to 2013). we are mobilizing the techniques of multivariate
analysis, Exploratory Analysis of spatial Data and techniques of spatial econometrics
especially recent developments in spatial panel data econometrics. results to be obtained
allow to draw some implications of economic policies.
Key words : Baccalaureate, Morocco, spatial, panel data, inequality
2. Introduction
Education is a key factor in the development of any country, and differences in living
standards are largely attributed to him. It contributes to the reduction of poverty and
inequality, but at the same time, it can lead to exclusion and marginalization (World Bank,
1999). Whatever the level of accumulation of human capital contributes to economic growth
1
The corresponding author. Email : [email protected]
of countries ( Mankiw et al 1992 . Aghion and Howitt, 1998, Benhabib and Spiegle, 2005), its
accumulation at the local level could be a good way to measure the changing economic
performance of regions (Cardenas and Pontoon, 1995). For example, wage differences
between regions are largely due to the level of education in Brazil (Azzoni and Servo, 2002).
In general, regions that invest more in human capital accumulation are developing rapidly.
The importance of education in the development process of developing countries has
given rise to several studies (World Bank, 2004; Bonal, 2004, Unesco , 2004; Amaghouss and
Ibourk, 2013). All these studies analyze the indicators of education in a classical perspective
(educational performance between sexes, between rural and urban areas). However, very few
studies proceed to the analysis of educational variables in a spatial perspective. During the last
decade, there has been a growing recognition of the importance of geography and space in the
analysis of economic convergence (Janikas and Rey, 2005; Mossi et al , 2003. Dietrich
Brauch and Monasterio , 2009). This convergence (between countries or within the same
country) is widely discussed with reference to the dynamics of monetary variables (income,
wages, GDP...). The analysis of the spatial convergence (within the country) based on
socioeconomic variables is rare. These indicators can be a complement to understand the
spatial dynamics of the regions. Baumol et al. (1994) argue that the analysis of social
variables could be the best way to study the differences in performance between regions
especially in the developing countries. Educational indicators can provide a more
comprehensive understanding of the asymmetries and imbalances in a given space and
between
different
regions.
In Morocco, the approximation of the levels of economic development is a major concern of
regional policies. Despite various attempts by various governments, Morocco is still suffering,
even before the independence, from a "regional imbalance" . This problem manifests itself in
particular by the existence of a large number of poor regions with low school performance. It
should also be noted that the Moroccan provinces are characterized by the persistence of a
dual structure between developed area ( the axis Kenitra – Settat) and a periphery in trouble.
Our goal in this paper is to show that taking into account the spatial dimension of the data,
that is to say, the location of the observations with respect to each other, allows to incorporate
some lessons and try to assess how extent the results usually found may be modified. Models
of spatial econometrics are required to justify their teachings in analyzing spatial disparities,
why geographic should be considered in the context of convergence models. The objective
here is to identify the relationship between spatial location and disparities in education in
Morocco. More specifically, it is shown that the inclusion of spatial disparities renews the
debate on the effectiveness of regional development policies. These inequalities cannot be
neglected and they allow a better understanding of the process of convergence between
provinces. It seeks to show that educational performances are related to distribution patterns
of educational services in the territory. It is to describe and visualize spatial distributions,
identify atypical locations and outliers, detect spatial association patterns and suggest spatial
regimes or other forms of spatial heterogeneity. More specifically, it seeks to answer the
following questions :
What are the characteristics of spatial disparities in Morocco? It comes down to analyze
basically the evolution of geographical disparities and the depth of interdependence amongst
the Moroccan provinces by constructing a Synthetic Spatial Index for Education (SSIE) based
on the method of the Principal Component Analysis.
How should we take into consideration these spatial disparities in our analysis of the
determinants of success in the Baccalaureate in Morocco?
This work allows us to offer geo-targeting policies aimed at reducing disparities between
the provinces. Specific regional development policies aimed at reducing the spatial
divergence, ie to attend the convergence between the best performing provinces and backward
provinces. Morocco was part of a broader process of regionalization; space groups obtained
could be a proposal for the next territorial division under Morocco.
3. Emergence of the theory of spatial justice: an overview of the
literature
Historically, Bowman (1942) was the first to rethink the geography as a determinant of
justice. If one believes Sorre (1957), classical geography does ignoring ethical considerations.
In France, the work of Jean Gottmann makes emerge explicitly the link between geography
and justice2.
In the 1970s, we have stated a progression of a new geographical welfare flow. Returning
to Marx's theory from a spatial perspective, Peet (1971) argue that inequality and poverty are
the functional components of the capitalist mode of production. Capitalism necessarily
produces unequal social structures. They are transmitted from one generation to another.
Assuming that the social geography of the city consists of a hierarchy of community settings,
it says that the capitalist mode of production reproduces the hierarchical structure of classes.
2
2
In his theory of justice, Rawls (1971) never mentioned the concept of space.
David Harvey does not use the term spatial justice but he uses the term "territorial social justice"
The persistence of poverty in American cities is the result of breeding an army of American
reservation. It concludes that inequality and poverty cannot be eradicated without
fundamental changes in the mode of production.
In an original way, David Harvey has the merit to transpose the concept of justice to
questions of geography. His conception of “territorial social justice” is an adaptation of Rawls
(1971) prepositions. In 1973, he published a work entitled Social Justice and the City. His
work marks a break with the traditional spatial analysis and new quantitative geography
(Brennetot, 2010).
Therefore, injustice is defined in relation to non-compliance with one of these two
principles. This author substitute "primary goods" by the concept of "needs" that includes
"food, housing, medical care, education, social and environmental services, consumer goods,
recreational Opportunities, neighborhood amenities, transportation facilities" (Harvey, 1973,
p. 102). The author joined the Marxist interpretation of urban development developed by Peet
(1971), that is to say, he denounces the distributive criterion of Rawls. Harvey is also
committed to prove the limits of liberal reformist model of territorial planning and operate a
critical study of spatial structures of capitalism.
The link between geography and justice also questioned Coates et al. (1977). They argue
that government strategies of welfare state neglect equity in the spatial distribution of welfare.
In another study, Smith (1979) takes the words of Rawls from a spatial perspective. Spatial
justice is not translated in the elimination of regional inequalities but in building equity. In
Chapter 3 of the Geography of welfare, Bailly (1981)3 provides a critical analysis of the
principles often used to assess the level of well-being. He advocates the consideration of
cultural and spatial criteria.
Starting from the center-periphery model of Amine (1973), Reynaud (1981) extends this
model to questions of spatial justice. It develops a theory of territorial injustice in which he
identifies divergences between spaces as disparities that develop between conflicted "sociospatial" groups.
Smith (1994) develops an evaluation model of welfare in which he proves that the overall
standard of living in the United States can hide spatial disparities. According to him, the
criterion of Lorentz and Rawls' principle of difference can substitute the principle of Pareto
optimum.
3
"The ideal society is no more the one in which a more optimal allocation of production factors and resources exist, but the one that
promotes the development of local cultural values”(Bailly 1981, p.99).
After the remarkable neglect of the field of spatial justice in the 1980s4, 1990s and 2000s
marked a renewed interest manifested by the multiplication of work on spatial justice. The
theory of spatial justice now covers many dimensions of globalization (Bret, 2009), urban
cohesion (Connolly et al, 2009; Soja, 2010); natural areas (Proctor and Smith, 1999; Lee,
2004); borders (Nussbaum, 2006), education (Mills, 2008; Habibi et al 2003.) scales (Fraser,
2008) and the environment (Flipo, 2007).
If we believe Bret (2009), the center-periphery model5 is the correct transposition of
spatial inequalities. Indeed, the center holds decision-making powers. In addition, the resident
population enjoys better living conditions at multiple scales (level of income, access to the
labor market, education, health…) These spaces are erected in the center because they attract
capitals, products and especially the flow of men (Bret, 2009). The periphery is defined by
opposition to the center. In general, it ensures unfavorable development's prospects. Some
research works wonder if this inequality between center-periphery constitutes an injustice
which arise therefore the inquiry of Bret (2009). He wondered whether the spatial equality is
consistent with the principle of maximum? The answer to this problem depends on the
potential role of the center relative to the periphery. In addition to its ability to attract, produce
and distribute wealth, the center can be a lever for development. In this context, territorial
inequalities are not an injustice because the periphery takes advantage of its position and is in
a less favorable position than if there was no central state. This is materialized by the transfer
to the periphery of the wealth produced in the center. Contrariwise, spatial inequalities can be
unfair if the center develops to the detriment of the periphery.
Having conducted this overview of the main trends in the theory of spatial justice, can we
then define these trends at the sectoral level especially within that of education? This
hypothesis is a subject of empirical verification in the following section.
4. Determinants of school performance : Literature Review
4.1 Quantity Vs. Quality of education
In a theoretical framework, education yields to two kinds of returns. The first one is
private while the second is social. Put simply, private return belongs to individuals so that
every additional year of schooling yields an increase in future income. Furthermore, social
return ensures a social cohesion due, among others, to the reduction of criminality, the
4
5
The concept of spatial justice appears despite all within a few summary work (Johnson et al, 1983.Brunet,1990).
Recall here that the center-periphery model can be applied to all geographical scales, from local to global
strengthening of citizenship (improving vote quality, creation of NGOs…). Ideally, social
return should exceed private return.
Some of the existing literature argues that a link exists between education and growth as
depicted by human capital theory (Mincer, 1958; Shultz, 1960, 1997 ; Becker, 1964 ; Denison
1962, 1979). In order to test this theory, education can be defined empirically in terms of
quantitative variables as well as qualitative variables.
Quantity of Education and Economic Performance
In a microeconomic framework, the Mincer earnings function (Mincer, 1958, 1974)
estimates educational returns, whereas in macroeconomics the link between education and
growth has never been proved, at least unanimously. Barro (1991), for instance, found a
significant positive relation between initial human capital ( rate of enrollment in 1960) and
growth. In 1992, Mankiw, Romer and Weil conclude that their results confirm the extended
Solow model. Benhabib and Spiegel (1994) estimate a Cobb-Douglas production function in
which physical capital as well as human capital are considered as production factors. Their
results revealed that human capital contributes to economic growth throughout two
mechanisms. First, the level of human capital influences directly the rate of technological
innovation produced locally as found also by Romer (1990). Second, the stock of human
capital affects the speed of the adoption of foreign technology, this assertion comes along
with the findings of Nelson and Phelps (1966). In a whole, according to the authors, all
countries would converge to the same level.
- Quality of Education and Economic Performance
The development of the qualitative approach of education has been initiated mainly in the
US. Indeed, the Coleman report (Coleman et al., 1960) is one of the earliest works that
established the relationship between educational outcomes and education resources.
Moreover, Mulligan (1990) and Lazear (2003) adopted a direct estimation of the impact of
achievements obtained in standardized tests. Actually, they used a representative database of
individuals, entering to the labor market, who finished their education. Their results suggest
that an increase of the standard-error of mathematical achievements increases earnings by
12%. Based on the higher education national survey of 1972, Murnane, Willet, Duhaldeborde
and Tyler (2000) investigate empirically the link between quality of education and individual
earnings. Results reveal that earnings for men and for women respectively increase by 15%
and 10% when the standard-error of the test score increases by 1%. However, the quality of
education in developing countries is overlooked. Hanushek and Wößmann (2007) suggest that
educational returns observed in developing countries are higher than those observed in
developed countries.
Since the 1990’s, several empirical works used scores of standardized test in order to test
the relation between quality of education and growth. Hanushek and Kimko (1995, 2000)
constructed an index of education quality, based on standardized tests. Their findings are in
favor of the positive relation between education and growth in the 1960-1990 period.
Furthermore, Barro (2001) used several standardized tests and combined it with quantitative
indicators of education. He found that both quality and quantity of education have an impact
on growth, however quality has a greater impact on economic growth. Wößmann (2002,
2003a) suggests that the proportion of variation in the level of economic development due to
human capital increases if the quality of education is considered. Bosworth and Collins
(2003); Ciccone and Papaioannou (2005) extend the works of Hanushek and Kimko, their
results confirms those of Barro (2001) regarding the importance of the role of education
quality on growth.
4.2 Determinants of schooling achievement
4.2.1 The intrinsic characteristics to the student
Some theorists explain differences in academic performance by individual differences among
students. They consider that school performance is positively correlated with intelligence
quotient (IQ). Genetic current explains school failure by disorders and deficiencies that are
intrinsic to the individual which may be detected by testing. Theproponents of this current
(Debray-Ritzen, 1978; Jencks, 1973; Le Gall, 1954; Terman, 1917); cited by Akoué (2007)
argue that academic success is a function of intelligence included in the genetic heritage.
Maehr, Pintrich & Linnenbrink (2002) show that the motivation and self-perception has links
closely with academic success.
4.2.2 The home environment
Several studies have developed the impact of the family environment, with a particular focus
on the cultural heritage. The academic performance of the student depend on the cultural and
linguistic bases held by the family environment. The children from higher socio-economic
level of families develop habits and tastes that are directly transferable in schools (Bourdieu,
1966, p.329). Mingat (1991) observed that the level of student achievement is strongly linked
to social class students. Children whose father exercises unskilled employment have the
lowest level acquisition, while children whose father is a technician have an average superior
results. Thus, the school "transforms cultural differences in inequality success "(Duru-Bellat,
2003, p.33).
4.2.3 The school context
While most individual and social factors are not controllable by agents school, those related to
the school context are controllable. Heyneman and Loxley (1983); cited by Duru-Bellat
(2003), showed that school factors were more influential regarding the success children, that
family factors. Among the factors related to the school context, we distinguished class size
(Fuller (1986); Akoué (2007)), Pedagogical Practices teachers (Lockheed and Verspoor
(1990); NLEP (2001)) School equipment (Ouellet (1987); Psacharapoulos and Woodhall
(1988)).
5. Methodology
As a first step, we analyze the spatial disparities of the rate of successful bachelor (and by
branch: scientific, technical and literary).
In order to properly account for spatial interactions, these methods take into account the
relative positions of the data through the inclusion of spatial weights matrix. Thus, the
comparison of a space observation with its neighbors is taken into account directly. In
addition, these methods provide measures of global and local spatial autocorrelation (Le Gallo
and Lame-Orain, 2004).
In a second time to confirm the results to be obtained by the exploratory analysis, we proceed
to the confirmation by the test of the hypothesis of absolute convergence of the Moroccan
provinces over the period 2013-2009. The estimated cross-sectional model of β-convergence
is our starting point. Formally, the model applied to our data sample is defined by the
following expression:
TX bac = α + βLog (bac) 2009 + ε
With TXbac is the growth rate of success in baccalaureate. Log (bac) is the logarithm of the
rate of success in baccalaureate and ε is an error term independent and identically distributed.
In the presence of spatial autocorrelation, the estimated equation by OLS (ordinary least
square) gives biased coefficients. So the first test to be performed is on the presence of spatial
effects.
Third, we analyze the determinants of success in baccalaureat. We focus our attention (due to
data availability) on the determinants related to the school context. The spatial regression of
the model using panel data can be written as follows:
bacit=Xitβ+ai+γt+vit
bac is the success rate of bachelor with X is the set of explanatory variables. In this study, X
includes the number of rooms, the rate of fellows, the number of teachers, the number of
administrative staff, the number of service personnel, the number of beneficiaries from
transport
services,
the
rate
of
beneficiaries
from
the
boarding.
Taking into account the spatial autocorrelation, the model can be done in several ways: By
spatial lag ((Case et al, 1993; Brueckner, 1998), a spatial error ((Rey and Montouri, 1999
Baumont et al, 2001) or by both. These models have different interpretations depending on
the nature of the convergence process. The objective of this research is to analyze what is the
model that best represents the problem of spatial dependence. We distinguish five general
types:
1) SAR model :
(bac)it = ρW(bac)it+Xitβ+ai+γt+uit
Where W is the weight matrix, a s a single effect, γ is a time effect and u is an error term.
2) The SDM model :
bacit=ρW(bac)it+Xitβ+WZitθ+ai+γt+uit ,
where Z is the matrix of spatial lagged regressors
3) The SAC model :
bacit=ρWbacit+Xitβ+ai+γt+vit ,
vit=λEvit+uit ,
4) The SEM model :
bacit=Xitβ+ai+γt+vit ,
vit=λEvit+uit .
5) The GSPRE model:
bacit=Xitβ+ai+vit ,
ai=ϕWai+µi ,
vit=λEvit+uit ,
All data come from the Ministry of Education during the period 2009-2013.
6. Results
6.1 Descriptive Statistics
Table: Rate of success in Baccalaureate per branch
From the table, we observe that the highest success rates are observed for scientific courses
followed by literary branches and finally by the original teaching.
Table : Evolution of numbers of bachelors in Morocco
The table clearly shows that the evolution of the graduates of the workforce
experiencing an upward trend.
Table : Evolution of numbers of bachelors in Morocco by province
6. The implications of this research
First, it is clearly obvious, from the global analysis, that there is a high spatial disparity in the
field of education among provinces as well as a poor performance of the Moroccan
educational system. Firstly, the concentration of provinces within a region having the same
characteristics is confirmed. Indeed, the significance of the quadrants EF (haven of wealth)
and FE (black sheep) is low throughout the period of study.
Secondly, two poles have appeared. The first consists of Rabat-Casablanca axis provinces and
some Sahara provinces. We call this group the center. The second pole is made of some High
Atlas and Rif (mountains) provinces
This work allows us to offer geo-targeting policies aimed at reducing disparities between the
provinces. Specific regional development policies aimed at reducing the spatial divergence, ie
to attend the convergence between the best performing provinces and backward provinces.
In general, the results of this study will allow policy makers to make better decisions on
regional development.
References :
Anselin L. (1988). Spatial Dependence and Spatial Structural Instability in Applied
Regression Analysis, Journal of Regional Science, 30, 185-207.
Anselin L. (1992). Spatial data analysis with GIS : An introduction to application in the social
sciences, Technical Report, 92-10.
Anselin L. (1995). Local Indicators of Spatial Association-LISA, Geographical Analysis, 27,
93-115.
Anselin L. (1998). Interactive techniques and exploratory spatial data analysis, Techniques,
Management and Applications, Wiley, New York.
Anselin L., Bera A. (1998). Spatial dependence in linear regression models with an
introduction to spatial econometrics. Chapter 7 (pp. 237-289) in Aman Ullah and David Giles
(eds.). Handbook of Applied Economic Statistics, NewYork: Marcel Dekker.
Azzoni, C. R. and Servo, L. M., 2002, “Education, cost of living and regional wage inequality
in Brazil”, Papers in Regional Science, 81(2):157-175.
Baumol, W. J., Nelson, R. R., and Wolff, E. N., 1994, Convergence of Produc-tivity. Crossnational studies and historical evidence, Oxford University Press, New York, NY.
Benaabdelaali, W. and A. Kamal. (2010). "The dynamics of educational inequality in
Morocco (1950-2010)" Paper presented at the second international conference of GDRI
DREEM on "Innovation and Economic Development In the Mediterranean Countries".
Bonal, X., 2004, “Is the world bank education policy adequate for fighting pov-erty? Some
evidence from Latin America”, International Journal of Educa-tional Development,
24(6):649-666.
Cardenas, M. and Pontoon, A., 1995, “Growth and convergence in Colombia: 1950-1990”,
Journal of Development Economics, 47:5-37.
El Horst, J.P. (2003). Specification and Estimation of Spatial Panel Data Models,
International Regional Science Review, 26: 244-268.
Ertur C., Koch W. (2004). Analyse spatiale des disparités régionales dans l'Europe élargie.
Pôle d'Economie et de Gestion, Université de Bourgogne.
Ertur C., Thiaw K. (2005). Croissance, capital humain et interactions spatiales : une étude
économétrique, Pôle d'Economie et de Gestion, Université de Bourgogne.
Haining R. (1990). Spatial Data Analysis in the Social and Environmental Sciences,
Cambridge University Press, Cambridge.
Janikas, M. V. and Rey, S. J., 2005, “Spatial clustering, inequality and income convergence”,
Région et Développement, 21:45-64.
KAPOOR, M., KELEJIAN, H.H. ET PRUCHA, I.R. (2007). Panel Data Models with
Spatially Correlated Error Components, Journal of Econometrics, 140: 97-130.
LEE, L.-F. ET YU, J. (2010). Estimation of Spatial Autoregressive Panel Data Models with
Fixed Effects, Journal of Econometrics, 154 : 165-185.
Lim, U., 2003, “A spatial analysis of regional income convergence”, Planning Forum, (9):6680.
Le Gallo J. (2002). Econométrie spatiale : L'autocorrélation spatiale dans les modèles
régression linéaire, Economie et prévision, 155, 139-157.
Le Gallo J., Dall'erba S. (2005). Croissance, convergence et interactions régionales : les outils
récents de l'analyse spatiale quantitative. Région et Développement, n⁰ 21, 5-11.
Morales, M., and C. Paz Terán. (2010) "Educational Inequality in Argentina: The best and
worst" Instituto de Estudios Laborales y del Desarrollo Económico, Documentos de Trabajo
No. 5.
Mossi, M. B., Aroca, P., Fernandez, I. J., and Azzoni, C. R., 2003, “Growth dynamics and
space in Brazil”, International Regional Science Review, 26(3):393-418.
Qian, X. and R. Smyth. (2008). "Measuring Regional Inequality of Education in China:
Widening Coast-Inland Gap or Widening Rural-Urban Gap?." Journal of International
Development 20(1): 132–
144.
Rey, S. J. and Janikas, M. V., 2005, “Regional convergence, inequality, and space”, Journal
of Economic Geography, 5(2):155-176.
Rey, S. J. and Montouri, B. D., 1999, “US regional income convergence: A spatial
econometric perspective”, Regional Studies, 33(2):143-156.
Thomas V., Wang Y., FanX., 2003, "Measuring education inequality: Gini Coefficients of
Education for 140 countries (1960–2000)",Journal of Education Planning and Administration,
17(1): 5–33.
UNESCO, 2004, “Global Education Digest 2004: Comparing Education Statistics Across the
World”, UNESCO Institute for Statistics, Montreal, Canada.
Wooldridge J. M. (2010). Econometric Analysis of Cross Section and Panel Data. MIT press
World Bank, 1999, Educational Change in Latin America and the Caribbean, World Bank,
Washington, DC.
YU, J., DE JONG, R. ET LEE, J.-F. (2008). Quasi-maximum Lekelihood Estimators for
Spatial Dynamic Panel Data with Fixed Effects When Both n and T are Large, Journal of
Econometrics, 146: 118-134.