CC BY
287
Section 8. Economy Demography
DOI: http://dx.doi.org/10.20534/EJEMS-17-1-56-62
Sahibi Youness, University of Abdelmalek Essaadi PhD student, Faculty of Legal, Economic and Social Sciences
E-mail: [email protected] Moustapha Hamzaoui, University of Abdelmalek Essaadi PhD, Professor, Faculty of Legal, Economic and Social Sciences E-mail: [email protected]
Inequality of opportunity in education: the effect of circumstances on individuals education in Morocco
Abstract: This paper discusses the effect of inequality of opportunity on education in Morocco. Starting from an analysis of theoretical and empirical papers, we used the data from the National Survey on the levels of Living of Households 2006/07 to provide an estimation of the importance of these circumstances in education inequality in a country that has long suffered from inequality in education. We apply six circumstances variables which lie beyond the control of the individual — father's and mother's education; gender; milieu; region of birth and expenditure and make an inequality of opportunity index using two methods. Then we decompose this index using Shapley decomposition. Our result show that a significant part of inequality in education access is due to inequality of opportunity, and the most important source of this inequality of opportunity is area of residence.
Keywords: Inequality, circumstances, education.
Introduction
Many countries in the world have experienced a substantial increase in inequality over the last few years. In Morocco, although significant progress had been made at the national level, there are still differences between the treatment individuals. This inequality has a bad impact because it violates the principles of social justice [20] such as equal rights. It also has a negative effect on the efficiency of the economic system, children's education, employment, and so on.
For many years, inequality studies focused only in terms of income in relation to economic growth. However, this old debate introduced by authors such Kuznets, S. (1955) [16] and Lewis, W. A. (1954) [17], was replaced by a new one focusing on inequality ofopportunity. This new concept aims to assess the effects of inherited social factors such as family background, race, and place of birth on the outcome. Also, inequality of opportunity was extended to different social fields such as education, health and employment.
In this paper, we focus on inequality in education. We make this choice because, firstly, the education of children is in large part influenced by circumstances of birth. Indeed, children cannot improve their living standards without their parents. This situation makes child education very sensitive to family backgrounds. Secondly, because of the importance given to promoting education in Morocco. The country established an emergency program in 2009. This continued program aims to resolve the problem of education access. However, despite the progress made, the problem of inequality in education continues to slow down the process of human development in Morocco as in the case ofdeveloped and developing country (Tawil, S., Cerbelle, S., Alama, A., &Unesco, 2010) [22] Gosta Esping-Andersen. (2001) [14].
The main objective ofthis paper is to identify the share of total inequality in education due to circumstances beyond the control of the individual and to highlight the effect of these circumstances. To achieve this objective,
we estimated the effect of circumstances variables on education levels. We also studied the effect of inequality of opportunity on each level of education (primary, secondary, and higher). Then, we decompose inequality of opportunity into five groups and studied the contribution of each in total inequality.
The theory background: from primary goods to capability
Starting from the egalitarian philosophers such as Raw-ls, J. A. (1971) [19] and Dworkin, R. (1981a) [9], the principle of equality of opportunity received more attention.
Rawls, J. A. (1971) [21, 92] introduce the notion of primary goods "things which a rational man wants whatever else he wants". These goods include basic liberties and rights, freedom of movement, free choice of occupations, access to political and other offices, income, wealth, and the social bases of self-respect. For Rawls, J. A. (1971) [19], the distribution of these primary goods is essential to perform social justice. This justice is expressed in two principles according to this author:
"[...] each person is to have an equal right to the most extensive scheme of equal basic liberties compatible with a similar scheme of liberties for others" [19].
"Social and economic inequalities are to be arranged so that they are both (a) reasonably expected to be to everyone's advantage, and (b) attached to positions and offices open to all." [19].
The first principle is based on Jean-Jacques Rousseau theory of social contract, which claims legal rights. The second aims to resolve the problem of inequality and ensure a fair equality of opportunity. In order to achieve this principles, Rawls, J. A. (1971) [19] consider inequality in the distribution of these primary goods acceptable only if it benefits the poor members of society.
Dworkin, R. (1981b) [10] has expanded Rawls [19] approach by integrating individual physical characteristics like health, talents and capacities among the resources that people have. According to him, people with no physical capacities should be compensated because this situation is an unchosen circumstance.
However, this perception of resources equality (physical capacity and primary goods) was widely criticized by Sen, A. (1995) [21]. One of his criticisms was that people required different amounts of goods to meet the same needs. To handle this, he introduced a new concept called "functioning". This concept is related to what an individual can achieve, taking into account their capacities and resources. The different combinations of this functioning represent what Sen [21] called "capabilities". Thus capability is the expression of the freedom for an in-
dividual to choose between different living conditions or achievements.
Rawls, J. A. (1971) [19], Dworkin, R. (1981) [9] [10], and Sen, A. (1995) [21], have a different perception of inequality. However, all these authors believe that an equitable society provides equal chances to all individuals to attain achievements. Indeed, once the individuals have an equal opportunity they can choose their achievements according to their efforts. That's why the difference in the degree of achievement can be determined in large part by the inequality of opportunity. However, the question is how these inequalities can be measured?
Measuring the inequality ofopportunity in education
A large number of papers which attempted to analyze the question of equality of opportunity have faced up to the problem of measuring this kind of inequality. In order to cope with this problem, the authors distinguish between two parts of inequality of opportunity: inequalities that can be explained by the circumstances and those that are explained by the effort.
According to Checchi, D., Peragine, V., &Serlenga, L. (2010) [8], the first one is more ethically acceptable than the second one. Indeed, it's ethically offensive to think that circumstances which are beyond an individual's control have a large influence on individual outcomes. However, it's not offensive to think that a great effort should generate a great outcome.
One of the papers which the authors tried to measure equality of opportunity was Bourguignon, F., Ferreira, F. H. G., & Menendez, M. (2007) [5]. These authors try to quantify the role of inequality of opportunity associated with circumstances in current earnings in Brazil. They measure circumstances with five variables related to father's and mother's education, father's occupation, race, and region of birth. Using these circumstances variables, they analyzed how the proportion of income inequality explained by the circumstances has evolved. However, this type of study does not directly address the inequality of opportunities in education. Other authors like Binder, M., & Woodruff, C. (2002) [3], tried to analyses the education achievement as a cause of income inequality. Despite the first who studied this type of inequality in the case of education as an endogenous phenomenon was ROEMER, J. E. (1998) [20]. He distinguishes between the circumstances beyond a children's control which influence their abilities. These circumstances include family background, culture, and social milieu. For ROEMER, J. E. (1998) [20], if we want to handle the effect of the circumstance there need to redistribute educational resources like teachers, books, schools,
buildings, in order to compensate this effect. However, in order to ensure an equal redistribution it's necessary to target the children who have problems in their education by measuring the educational achievement. That's why different authors used three different approaches in order to measure individual educational achievement.
The first approach focused on academic mobility. This method has been used by Birdsall, N., Behrman, J. R., & Szekely, M. (1998) [4], for Latin America. The weakness of this approach is the use of prospective data instead retrospectives data. Indeed, the education achievement of an individual is linked to his academic life as a whole.
The second approach used by Chavez-Juarez, F., & Soloaga, I. (2015) [6], was the number ofyears ofschool-ing. This measure is considered imperfect because it does not reflect the educational performance of the student. Indeed, a high number ofyears of schooling cannot reflect the achievement because it does not take into account the repeated school years. In their estimation, Chavez-Juarez, F., & Soloaga, I. (2015) [6], used Mexican data and they found that all variables of parent's background influence the probability of education achievement. They also observed that the effects of circumstances are the highest for 9 and 12 years in their schooling.
The third approach used by Ferreira, F. H. G., & Gig-noux, J. (2011) [11] was on the basis of two types of measures: quantity and quality. The quantitative measure is enrollment age; the quality measure is a test score. This approach is considered to be the best because it reflects the academic achievement of the student. The result of this paper suggests that the most important circumstances are the family background variables.
In this paper, we use the level of education (primary, secondary and the second) in order to measure individual educational achievement. This approaches is quite similarly to that of Ferreira, F. H. G., & Gignoux, J. (2011) [11]. However, the level of education approach, unlike a number of years of schooling, reflect the educational performance.
Research methodology: Variables selection and estimation method
There are two main approaches to analyzing inequality of opportunity: the ex-ante approach focuses on the circumstances, and the ex-post approach focuses on the effort. In this paper, we are interested in the ex-ante approach because children's education depends on the circumstances in which they grew up.
Different empirical studies use the ex-ante approach like Bourguignon, F., Ferreira, F. H. G., & Menendez, M. (2007) [5], Gignoux, J., & Ferreira, F. H. G. (2010) [13],
Checchi, D., & Peragine, V. (2009) [7] and Chávez-Juárez, F., & Soloaga, I. (2015) [6]. In these studies, authors has proposed several methods to deal with ex-ante inequality of opportunity. However, the most popular method was the regression. With this method, we can know if circumstances variables affect the outcomes significantly.
In this paper, we estimate inequality of opportunity for a multi-way categorical dependent variable with four ordered values which measures the level of education, by using the method proposed by Ferreira, F. H. G., & Gignoux, J. (2011) [11].The idea behind is to estimate the outcome y of an individual with a set of circumstances C. Formally, this equation can be written as: y = E [ y\C ]
Then, the relative measure of inequality is obtained by applying common inequality measure I (.) to equation (1):
q = I (y )
This equation can be estimated by different methods depending on the outcome. In this papers, we used discrete outcome. That is why we used Barros, R. P. de, Ferreira, F. H. G., Vega, J. R. M., &Chanduvi, J. S. (2009) [2] method. These authors use a logit or probit model to estimate the equation of Ferreira and Gignoux (2011).
The idea behind the dissimilarity index is if we have equal opportunity, then the correspondence should be exact between the access rate of the groups defined by circumstances and the average access rate of the whole population. This correspondence is measured by the dissimilarity index ranges from 0 in the case of perfect equality to 1 in the perfect inequality.
1 N I —I
/ = D(y)= Ny #-y
Where y = E [y ].
However, according to Chávez-Juárez, F., & Soloaga, I. (2015) [6], this dissimilarity index has a limit related to his scale invariant. However, scale invariance is not a problem in the inequality index unless the inequality measure cannot satisfy the axiom of translation invariance. Thus, we can show that dissimilarity index does not satisfy this axiom:
1 N I _ I
D (y+V ) = ^—\ e\A+y - y - v\ =
2N (y + p )
—\ - y| = (y) -p)1 1 y + p
2N
y
To cope with this problem, Chavez-Juarez, F., & Soloaga, I. (2015) [6] proposed to simply drop in the denominator, which is the cause of the scale invariant. He modified the dissimilarity index:
This index can be decomposed into groups using Shapley method. This approach divides the outcome into shares that reflect the contributions of several groups in total inequality of opportunity. To compute this type of decomposition, we must calculate the average marginal effect of each circumstance variable on the global index.
In this paper, we compared the Chavez-Juarez, F., & Soloaga, I. (2015) [6] method (ws) and the Barros, R. P. de, Ferreira, F. H. G., Vega, J. R. M., &Chanduvi, J. S. (2009) [2] method (pdb). After that, we decompose our measure of total inequality of opportunity into groups, and we attribute a part of total inequality to each group of circumstances using Shapley decomposition. We used six circumstance variables. The selection of this variables is based on the theoretical and empirical contributions in literature. Among these variables, we have:
• Father's education: Much empirical evidence exists on the relationship between the father's and children's education. According to Pleck, E., H., &Pleck, J. H. (1997) [18], fathers are more likely to get involved actively in the care and education of children if they feel competent to do so. This competence can comes from his own education. In the same way, Altintas, E. (2016) [1] examined time US parents spent in activities vital for child development between 1965 and 2013. He found that highly educated fathers spend more time on developmental childcare than their less-educated peers, even after controlling for their spouses' education.
• Mother's education: this variable certainly has an impact on their children's education. Indeed, educated mothers can outsource activities such as cleaning. Thus she can spend her time with her child. In the same way, educated mothers can help more with homework which helps their children to have good marks in the school. This mother's effect was observed by Hernandez, D. J. & Napierala, J. S. (2014) [15], in this study about the effect of mother's education in the US. His results show that there are enormous disparities between children whose mothers had not graduated from high school, and those whose mothers had a university education.
• Milieu and region: these two variables represent the spatial factors. This spatial variables allows to assess the inequality of opportunity due to the environment which offers more or fewer opportunities. In order to analyze the effect of this spatial factor Ferreira, S. G., & Veloso, F. A. (2006) [12] studied intergenerational educational mobility in Brazil, they estimate Brazilian inter-
generational elasticities. These result show that mobility patterns differ across regions and races in Brazil.
• Gender: gender inequality is more pronounced in developing countries like Morocco. This inequality was largely studied in economic literature. Indeed, in many poor countries, society suggests that women should marry early, have children young and must devote to her family which causes high school abundant in women population.
• Expenditure: this variable measures the difference in wellbeing. A low expenditure can indicate malnutrition or lack of health care which affects children's education.
The Data
The main data source used in this work is the national survey on the levels of living of households 2006/07 (ENNVM). It was collected from a sample of 7062 households, representative at the national level. The main purpose of this survey is to gather informations about living standards in order to help define and evaluate social programs for better social inclusion of low-income households in the development process.
The success of this survey required the selection of a representative sample of households to provide reliable estimates of socio-economic indicators of the population. For conducting the survey on the living standards of households, a stratified three-stage was adopted. The first stage is the selection of 600 Primary Units from the 1848 Primary Units constituting the master sample. In the second stage, secondary units was selected from the each primary unit of the sample. In the third stage, 12 households are selected with a random with equal probability in each sample secondary unit.
In order to ensure to take people who finished their education, our sample includes 16053 adults aged 25 years and older.
Inequality of opportunity in education: Empirical results.
In this section, we present the result of estimating inequality of opportunity in education. In a first step we show, the result of ordered logistic regression, which links circumstances variables to the level of education (primary, secondary and higher). In the second step, we show the result of estimating the level of inequality of opportunity by comparing the Chavez-Juarez, F., & Soloaga, I. (2015) [6] method (ws) and the Barros, R. P. de, Ferreira, F. H. G., Vega, J. R. M., & Chanduvi, J. S. (2009) [2] method (pdb). Finally, we perform the Shapley decomposition to identify the main determinants of inequality of opportunity in education.
The Wald chi-square of 1025.99 with a p-value of which is higher than the coefficient of primary school.
0.000 tells us that our model as a whole is statistically significant. The estimated output indicates circumstance variables have the expected sign. The father's education level increases the chances of increasing the child's educational level. In fact, the coefficient of higher education is higher than the coefficient of secondary
The same result is observed in the mother's education, expect for the higher education, which had no significant effect. Also, the coefficients of the father's education are higher than those of the mother. This result means the father's education has a greater effect than the mother's education on the children's education in Morocco.
Table 1. - depicts the estimation of the ordered logit model
Student characteristics Coefficients
Father's education (omitted=no education)
Primary 0.65212361***
Secondary 1.3905079***
Higher 2.0588086***
Mother's education (omitted=no education)
Primary 0.30259131***
Secondary 0.93468989***
Higher 0.77145082
Area (omitted=rural)
Urban 1.6500741***
Region (omitted=center)
South -0.65875404**
North -0.13813977*
Gender (omitted=Male)
Female -0.75457395***
log expenditure 0.65113631***
cons -1.899032***
Pseudo R2 0.1595
Wald chi2 (11) 1025.99
Prob > chi2 0.0000
/cutl 5.377427
/cut2 8.35164
/cut3 9.431044
legend: * p<0.1; ** p<0.05; *** p<0.01
Moreover, place of residence seems to have a significant effect. Thus, living in an urban area compared to rural increases the probability to have a good education. Also, living in the South and in the North compared to the center decreases the probability of having a great education.
As also expected, the gender seems to have a significant effect on the level of education. Being a woman reduces the probability of the increase in the levels of education.
In the same way, consumption which is a measure of the standards of living affects positively the level of education. Thus, the poor have less probability of the increase their level of education in relation to the rich.
Regarding the effect of total inequality of opportunity, Table 3 show the part of the variability of each dependent variable categories attributed to inequality of opportunity. This table tells us variability in primary access is due to inequality of opportunity varies between 12.5% using Barros, R. P. de, Ferreira, F. H. G., Vega, J. R. M., &Chanduvi, J. S. (2009) [2] method (PdB) and 37.38 using Soloaga, I., and F. Chávez Juárez (2015) [6] method (WS). In addition, variability in higher education access due to inequality of opportunity varies between 35% using PdB method and 17.36% using WS method. However, both PdB and WS method shows that variability in secondary education access due to inequality of opportunity is around 33%.
Table 2. - Inequality of opportunity in the level of education
Threshold PdB ws
Primary 0.125477 0.373801
Secondary 0.330674 0.330129
Higher 0.350004 0.173621
This total inequality of opportunity was decomposed into five groups of variables in order to measure the contribution of each group of circumstances (Table 3).
Table 3. - Decomposition of Inequality of opportunity index (Shapley method)
Variable Value Percentage
Group1 0.075014 21.27%
Group2 0.040447 11.47%
Group3 0.118294 33.54%
Group4 0.01208 3.42%
Group5 0.104977 29.76%
TOTAL 0.352742 100.00%
Table 5 suggests the geographical variables (region and milieu) is associated with the largest share of inequality of education it represents 33.54% of inequality index. Followed by expenditure, which is a proxy of wellbeing. This variable accounts for 29.76% of total inequality of opportunity. Also, the father's education ranks third for about 21.27%, while mother's education do not account for much: only half the share of father's education (11.47%). On the other hand, Mother's and father's education combined account for 32.74%.
The lowest contribution is gender. This variable accounts for just 3.42%, which is almost ten times less than the group of region and milieu. Unlike a lot of study that considers the inequality between sexes are the most im-
portant, our results show that although it is important he come in last place after the other circumstances that have considered. However, the inequality between gender can be considered the most ethically offensive after expenditure because these inequalities can be a large inside the same community and even inside the same family where parents consider the education of their male child most important than their female child which can cause serious social problems.
Conclusion
We have asked to what extent inequality of opportunity affects individual education success in Morocco. We used two measures of inequality of opportunity: the Barros, R. P. de, Ferreira, F. H. G., Vega, J. R. M., &Chanduvi, J. S. (2009) [2] dissimilarity index and the Wendelspiess Chávez Juárez (2013) [6] modified dissimilarity index. Even the limit of our measure which is not integral to the part of the effort, this measure can help to build a policy which provides an equal opportunity in order to promote education in Morocco. Our results indicate the proportion of inequality in education explained by inequality of opportunity varies around 12% to 37% depending on the measure and level of education.
Unlike other studies which attribute inequality of opportunity first to gender, in a Morocco context the predominant effect is the place of residence. This result suggests a good policy based on spatial redistribution of educational resources like teachers, books, schools, buildings [20] can improve education effectively.
Also, the substantial effect of expenditure on inequality of opportunity suggests that conventional policies based only on education are not sufficient. A more global approach is needed, which also integrates the poverty reduction policies.
References:
1. Altintas E. (2016). The Widening Education Gap in Developmental Child Care Activities in the United States, -1965-2013. Journal of Marriage and Family, - 78 (1), 26-42. URL : http://doi.Org/10.1111/j omf.12254
2. Barros R. P. de, Ferreira F. H. G., Vega J. R. M., & Chanduvi J. S. (2009). Measuring Inequality of Opportunities in Latin America and the Caribbean (World Bank Publications). The World Bank. Retrieved from https://ideas. repec.org/b/wbk/wbpubs/2580.html
3. Binder M., & Woodruff C. Inequality and Intergenerational Mobility in Schooling: The Case of Mexico. Economic Development and Cultural Change, - 2002.- 50 (2), - 249-67.
4. Birdsall N., Behrman J. R., & Székely M. (1998). Intergenerational Schooling Mobility and Macro Conditions and Schooling Policies in Latin America (Research Department Publications No. 4144). Inter-American Development Bank, Research Department. Retrieved from https://ideas.repec.org/p/idb/wpaper/4144.html
5. Bourguignon F., Ferreira F. H. G., & Menéndez M. Inequality of Opportunity in Brazil. Review of Income and Wealth, - 2007. - 53 (4), 585-618. URL: http://doi.org/10.1111/j.1475-4991.2007.00247.x
6. Chavez-Juarez F., & Soloaga I. (2015). Scale vs. Translation Invariant Measures of Inequality of Opportunity When the Outcome is Binary (SSRN Scholarly Paper No. ID 2226822). Rochester, NY: Social Science Research Network. Retrieved from URL: http://papers.ssrn.com/abstract=2226822
7. Checchi D., & Peragine V. Inequality of opportunity in Italy. The Journal of Economic Inequality, - 2009. - 8 (4), - 429-450. URL: http://doi.org/10.1007/s10888-009-9118-3
8. Checchi D., Peragine V., & Serlenga L. Fair and unfair income inequalities in Europe - 2010. (Working Paper No. 174). ECINEQ, Society for the Study of Economic Inequality. Retrieved from URL: https://ideas.repec. org/p/inq/inqwps/ecineq2010-174.html
9. Dworkin R. What is Equality? Part 1: Equality of Welfare. Philosophy and Public Affairs, - 1981a. - 10 (3), 185-246.
10. Dworkin R. What is Equality? Part 2: Equality of Resources. Philosophy and Public Affairs, - 1981 b. - 10 (4), 283-345.
11. Ferreira F. H. G., & Gignoux J. The measurement of educational inequality: achievement and opportunity - 2011. (No. WPS5873) (P. 1-43). The World Bank. Retrieved from URL: http://documents.banquemondiale.org/cu-rated/fr/2011/11/15456651/measurement-educational-inequality-achievement-opportunity
12. Ferreira S. G., & Veloso F. A. Intergenerational Mobility of Wages in Brazil. Brazilian Review of Econometrics, -2006. - 26 (2), - 181-211. URL: http://doi.org/10.12660/bre.v26n22006.1576
13. Gignoux J., & Ferreira F. H. G. Inequality of opportunity for education: the case of Turkey - 2010. -No. 75447 - P. 1-28. The World Bank. Retrieved from URL: http://documents.worldbank.org/curat-ed/en/2010/03/17364383/inequality-opportunity-education-case-turkey
14. Gosta Esping-Andersen. A new challenge to social cohesion? Emerging risk profiles in OECD countries. In What Schools for the Future - 2001. - P. 205. OECD Publishing.
15. Hernandez D. J., Napierala J. S., & Foundation for Child Development. (2014). Mother's Education and Children's Outcomes: How Dual-Generation Programs Offer Increased Opportunities for America's Families. Disparities among America's Children. - No. 2. Foundation for Child Development. 295 Madison Avenue 40th Floor, New York, NY 10017. Tel: 212- 867-57-77; Fax: 212- 867-58-44; e-mail: [email protected]; Web site: URL: http://www.fcd-us.org.
16. Kuznets S. Economic Growth and Income Inequality. The American Economic Review. - 1955. - March.
17. Lewis W. A. Economic Development with Unlimited Supplies of Labour. The Manchester School, - 1954. - 22 (2), -139-191.
18. Pleck E. H., &Pleck J. H. Fatherhood ideals in the United States: Historical dimensions. In In M. E. Lamb (Ed.), The role of the father in child development - 1997. - P. 33-48, third edition. New York: John Wiley & Sons.
19. Rawls J. A Theory ofJustice. Harvard University Press. - 1971.
20. Roemer J. E. Equality of Opportunity. Harvard University Press. - 1998.
21. Sen A. Inequality Reexamined. Harvard University Press. - 1995.
22. Tawil S., Cerbelle S., Alama A., & Unesco. Education au Maroc analyse du secteur. Rabat: UNESCO, Bureau Multipays pou - 2010.
