An Article by Alexis M. Arthur and Kimberly J. Vallejo

Abstract

This paper examines the impact of government education spending on two types of inequality within sub-Saharan Africa and Latin America. It develops a model to test the assumption that increased spending on education decreases inequality in general. While the results indicate a correlation between education spending and income inequality, quantitative analysis proves insufficient to explain the impact of education on broader social disparities, like  gender inequality. The researchers recommend that policymakers take a more nuanced approach to measuring the impact of education spending on social inequality, in particular through in-depth qualitative analysis. The paper also highlights the need for more dynamic measures of inequality, particularly regarding gender, and explores some of the challenges associated with gender inequality indicators. Moreover, the paper calls for improved data collection methods, particularly in regards to gender inequality statistics, to better inform social policy development.

Acknowledgments

The authors would like to acknowledge the work of fellow Cornell students Hannah Amsili (’10) and Tom Archibald (PhD Candidate, Dept. of Education) who contributed to the original version of this paper, written for the course Empirics of Development. The authors would also like to thank Associate Professor Parfait M. Eloundou-Enyegue, in the Department of Development Sociology at Cornell University for his ongoing support throughout the revision process.

Introduction

Recent international education initiatives, like Education for All and the Millennium Development Goals, advocate that developing nations increase their commitment to education as a means to achieve national development and increased social equality. This commitment to education is often held to be synonymous with an increase in national spending on education in proportion to overall public spending.  This belief, that education provides a foundation for development, draws on various economic assumptions of human capital, in addition to the expectation that increasing access to education strengthens democratic participation and poverty reduction. Through the study of two regions in the Global South, this paper analyzes the relationship between national investment in education and inequality. Understanding that inequality cannot be understood as a singular, static term, this paper argues that at least two forms of inequality, both income and gender, are correlated with education and that changes in the level of government spending on national public education will influence relative levels of both types of inequality. The relationship between education spending and income inequality in Latin America and sub-SaharanAfrica is examined first to test the hypothesis that a negative correlation between the two variables exists. The argument and the empirical model are subsequently expanded to test the relationship between government spending on education and gender inequality.

While the study found evidence linking increased educational spending to improved socio-economic wellbeing, results were inconclusive with regards to gender inequality. To further test the claim that educational spending is negatively correlated with overall social inequality, this paper calls for more nuanced analyses of the social returns on investment in education; particularly concerning the impact on gender inequality.

Overview of Literature and Rationale for Study

National and international policy-makers have increasingly relied on education as an engine for social and economic progress and development (UNESCO-UIS/ OECD 2003). In contrast to the strict recommendations and funding restrictions associated with the structural adjustment of the 1970’s and 1980’s, the World Bank now cites education as the “single most important key to development and to poverty alleviation” worldwide (UNESCO-UIS/ OECD 2003, 21). In light of policy initiatives like Education for All and the Millennium Development Goals, many developing nations have placed increased importance on education as a means to improve the potential economic capacity of their citizenry and thus fueling the overall development of the nation.

Studies which have followed the economic performance of OECD countries over a span of almost three decades have found that investments in education that  increase the average years of study of the population by just one year have seen an overall increase in GDP of about 6 percent (UNESCO-UIS/ OECD 2003, 22). Increased funding for education is seen to correlate with economic development in part because citizens are more prepared to participate in the formal economy if they have received more schooling. Developing higher literacy rates and more complex math and science skills are seen to directly improve society’s labor preparedness. The impact of increases in education spending (either to improve quality of schooling, access to schooling, or both) are thought to be exponential according to human capital theory. This economic theory understands the benefit of education as social rather than individual. The impact of improved education permeates the entire labor force, thus supporting links between increasing commitment to education and rising productivity and income. In short, individuals who acquire more years of learning influence the relative opportunities of others in the labor market. In doing so, rising incomes for the educated promote the accumulation of years of schooling for the nation at large (UNESCO-UIS/ OECD 2003).

Education  also boosts democratization and functions as an engine for general social cohesion and a means to fight poverty. Better educated citizens are more likely to participate in the democratic process and have a tendency to support democratic regimes over other political systems (Ganimian and Solano Rocha, 2011). Given the positive economic and democratic externalities typically associated with increases in national education levels, the assumption follows that increased spending on education (to improve access to schooling and the quality of national schools) will contribute to decreased income inequality within a nation. Those externalities, which benefit all of society according to their relative theories, thus suggest that we can expect income inequality to decrease over time as investment in schooling increases.

Schooling, however, is complex. Beyond being a tool for economic advancement and democratic development, education is a social institution. In order to fully comprehend the societal effect of schooling, education should be understood as a legitimizing institution, one that both constructs and alters “roles in society and authoritatively allocate[s] personnel to these roles” (Meyer, 1977, p.56). Educational institutions have the power to influence relationships between social groups and assign societal roles within the economic sphere”). Similarly to human capital theory, legitimation theory claims that education, as a social institution, has the power to influence and “transform the behavior of people in society quite independent of their own educational experience” (Meyer, 1977, p.56). In this regard, education is responsible for affecting power relations between social groups and hence impacting social inequality at large, not just for those who proceed through formal schools. If schools are thus legitimizing roles that promote social equality, it follows that society will exhibit less disparity between social groups. Critical analysis of the social structures being reproduced or transformed through the schools should be as central to the educational policy debate as is the correlation between income inequality and schooling.

Gender inequality emerged as an international concern in the 1970’s and has played an important role in international development policy. The elimination of gender disparity in education and employment are both integral to the achievement of the Millennium Development Goals and other large-scale development initiatives. The United Nations has recognized that “education is one of the most important means of empowering women with the knowledge, skills and self-confidence necessary to participate fully in the development process” (UNFPA, 2010, UNDP, 2010). Targeting the education of girls has been shown to yield much higher dividends than general increases in education spending. Educated mothers are better household negotiators and secure greater resources for their children, they participate more in the labor force and thus have higher family incomes, have healthier families, fewer children, and place greater value on the education of their children (UNFPA, 2010). Policy-makers should approach the connection between gender inequality and education spending in the same way as the relationship between income inequality and education. A better understanding of this gender relationship allows us to gauge how the funding of social institutions, like schools, is reproducing social structures and shaping gender roles. Former UN Secretary-General Kofi Annan, stated “[w]ithout achieving gender equality for girls in education, the world has no chance of achieving many of the ambitious health, social and development targets it has set for itself” (UN Secretary-General Kofi Annan, March 2005, quoted in Global Campaign for Education, 2005, p. 19).

Building the Models to Test Correlation of Education and Inequality

 

Explanatory Variable

Each model used in our analysis is based upon the independent variable of education spending as a percentage of overall government spending. Simple monetary measures of education spending, such as total US dollars, would have been flawed indicators due to their dependence on the relative income of each country. Taken not in absolute terms, but rather as a measure relative to the total government funding for social programs, education spending as a percentage of overall government spending proves to be a better indicator of a government’s commitment to providing education to its  citizens and is easily compared across countries and regions (Thomas, Wang & Fan, 2001, p.4). Education spending is also of primary concern to policymakers; understanding the relationship (if any) between government spending and broader social equality means we can then make better founded policy suggestions for governments regarding the allocation of funds to education. In setting the terms in percentages rather than absolute dollar amounts, the findings are more applicable to policy-makers across a variety of contexts. The conceptual model in Figure 1 represents the theoretical framework for the two empirical models outlined in the following sections:

Figure 1. Conceptual model of central theoretical framework.

 

Model 1: Dependent Variable

The income distribution inequality measure, the Income GINI Coefficient, helps us to better understand the relationship between education spending and overall income inequality. Findings  that support human capital theory would show that increased investment in “becoming more educated” makes citizens more prepared to participate in the formal economy and earn greater income[L16] . In increasing public spending, one should see that income disparity decreases over time as more citizens gain equal access to education and are thus able to participate on similar terms in the formal economy. The Income GINI is a better measure of socio-economic inequality than per capita income, which is simply an average and hence distorted by population size and outliers. While the Income GINI does have limitations, such as measuring income rather than wealth or wellbeing, it is both widely accepted and data is readily available across a broader range of countries and years. These factors make it useful for this type of regression analysis. Other studies have also pointed to a correlation between government spending on education and income inequality inAfrica (Global Campaign for Education, 2005). A similar association is expected in this regression.

Alternative variables, such the Education GINI (an indicator of the distribution of education across a society) could have been used as the dependent variable in this study. The Education GINI would have resulted in a slightly different analysis; providing insight into the equality/inequality inherent in the distribution of formal education in the selected regions. While one could argue that unjust social structures of a nation are directly related to the inequality inherent in the distribution of access to education, the measure is not widely used and thus insufficient data is available in this case (Thomas, Wang & Fan, 2001). Yet the Education GINI, due to its direct link to social inequality, would have proven extremely beneficial for social policy-makers if more widely available.

In the absence of an adequate Education GINI, there are few indicators that attempt to measure and compare educational outcomes within and across nations (Thomas, Wang & Fan, 2001). These indicators, which measure attainment levels, enrollment ratios, etc., tell us very little about the quality of schooling around the world and hence make this data virtually impossible to compare from country to country.
Mediating and Control Variables: Employment and Urbanization

In the 1950’s, Simon Kuznets developed the parameter that has shaped studies of inequality for the last 50 years. His “inverted u-curve” hypothesis stated that as countries entered the early stages of development and economic growth, income inequality would increase. However, in the intermediate stages it would slowly level out, and then as countries increased their wealth, inequality would decrease (Korzeniewicz and Moran, 2005). This theory was accompanied by the assumption that, as developing countries industrialized; the dual sector economy (agricultural and non-agricultural) would begin to change (Nielson and Alderson, 1997). As industrialization expanded, populations would shift from rural to urban areas, resulting in a rise in incomes (Simpson, 1990). This theory is based on the idea that non-agricultural employment yields higher wages, and that increased employment in urban areas would result in a greater proportion of the poor enjoying a greater share of overall income.

Urbanization and employment thus become obvious variables to include in a multivariate regression analysis of the relationship between education and income inequality. In this case, the mediating variable is employment, given that the direct means by which education yields greater access to income is through formal employment or wage labor. Urbanization is included as a control variable as it bears a direct impact on changes in the dependent variable, income inequality, as referenced in Kuznet’s theory. The extensive criticism of Kuznets’ work should also be taken into account here. The calculations are based on data in industrialized countries during the late nineteenth and twentieth centuries (Nielson and Alderson, 1997). Moreover, the argument that increased urbanization leads to decreased inequality ignores the contradictory view that income inequality is higher in urban areas (Nielson and Alderson, 1997).

Rising employment and economic growth does not necessarily translate into a reduction in income inequality. Despite rapid economic growth at the global level from 1995 – 2005, the 2008 International Labour Organization survey found that two thirds of countries for which data exists experienced increases in income inequality (International Labour Organization, 2008). The global situation is intensified inLatin America, where income inequality is pervasive despite rapid economic growth (Korzeniewicz and Moran, 2005). This is partly due to structural reforms implemented during this period, which emphasized a reduction in the role of the State, in particular, the State’s provision of public goods and services such as welfare (Grilli, 2005).

 

Model 2: Dependent Variable

The second model in this study applies the same structure of the first hypothesis, though poses a more nuanced question to explore the correlation between government spending on education and measures of gender inequality. Finding a useful indicator of gender inequality is of paramount importance, yet given the statistics generally available, this relationship proves to be far more difficult to understand. The gender equality indicators that are most readily available, including the GEM, GDI and to some extent the CPIA (all explained below), reveal very little about the relationship between education spending and overall gender equality. Data for gender empowerment is limited both in geographic coverage and time span. There is also limited information on the association between governmental spending on education and trends in gender equality measures, which makes the prediction of the association between the variables particularly difficult (Mahadevia and Hirway, 1996).

The Gender Empowerment Measure (GEM) is a measure of inequality between men and women’s opportunities in a country. It combines inequalities in three areas: political participation and decision making (parliamentary seats held), economic participation and decision making (legislators, senior officials, and management, professional and technical positions held), and power over economic resources (estimated earned income at purchasing power parity [PPP]).  The GEM is measured on a scale of 0 – 1, with 1 representing the highest level of gender equality. Indicators of economic and political participation are also included within the gender empowerment targets under the Millennium Development Goals.

The Gender-Related Development Indicator (GDI, where 0=low equality, 100= full equality) is an indication of the standard of living in a country. It considers the inequalities between men and women in the following areas: long and healthy life (life expectancy), knowledge (education, measured as the adult literacy rate and the combined primary to tertiary gross enrollment ratio), and a decent standard of living (estimated earned income at PPP).

The Country Policy and Institutional Assessment (CPIA) gender equality rating (where 1= low and 6=high) is another resource. In theory, this indicator of gender equality assesses the extent to which the country has installed institutions and programs to enforce laws and policies that promote equal access for men and women in education, health, the economy, and protection under law.

While this study explored the possibility of using all three common gender indicators mentioned above, the results proved inadequate in identifying a correlation. The GEM was selected as the most appropriate measure to use in the regression, however, due to insufficient reporting of the GEM indicator, the results of the model ultimately proved inconclusive.

 

Mediating and Control Variables: Female Employment and Fertility

Female employment acts as a mediating variable in the second model based on the theory that when girls and young women are more educated, they are subsequently more able to work and earn good wages, and thus have a greater likelihood of being “empowered”.[1]Yet while women’s entry into the labor force has boosted GDP growth in many industrialized nations, cultural norms surrounding women’s place in the workforce have limited gains in the Global South. In both Latin America and sub-SaharanAfrica women constitute over 70 percent of the non-employed population, highlighting the gap between male and female employment. Continued bias against women in the workplace also results in uneven employment and income growth between genders both across and within these regions (International Labour Organisation, 2008).

The fertility rate (the average number of children a woman will bear over her lifetime) is used as a control variable, based on the notion that higher birth rates impose a burden on women that prevents them from taking advantage of other opportunities, including education. It also increases health risks for women, particularly in areas with limited access and poor quality health care, such as sub-Saharan Africa. Yet issues of reverse and overlapping causation abound. It is plausible that education causes a decline in fertility, yet it is unclear how individual and societal factors interact in the intersection of empowerment and fertility. In any case, there are well-established—however complex—relationships between education and fertility as described briefly below.

There is “a growing empirical literature that suggests that gender equity in education promotes economic growth, reduced fertility, child mortality, and under-nutrition” (Abu-Ghaida and Klasen, 2004, p. 1075). Some interpretations also explore the role of empowerment of women and girls as a related factor. In a review of the literature, Basu (2002) explores some of the potential mechanisms by which these variables are related. She problematizes otherwise facile assumptions of how these relationships work, focusing on a number of complicating cases. For instance, in most cases the content of schooling in developing countries hardly seems conducive “to producing the relatively self-confident and effective mother and family planner that the educated woman appears to be” (p. 1780). This analysis corroborates Meyer’s (1977) work on the societal outcomes of education. Basu also examines ways in which issues of selection complicate understandings of these relationships: perhaps women who seek access to education are predisposed to having lower fertility rates for a variety of reasons. Also, because education may not affect gendered power relations around culturally normative decisions such as family size, an educated woman who is empowered economically may not be empowered reproductively. McDonald (2000) highlights another complicating factor: fertility change in society, according to most current theories of fertility transition, must be capable of being explained in individual terms. Yet gender equality, broadly understood, is a characteristic of society’s institutions.

It remains unclear what mechanisms link fertility and empowerment. In part, these shortcomings in gender-related indicators are summarized by Beteta (2006) as “[t]he selection of indicators of the GEM is generally biased towards measuring the empowerment of the better-off. In terms of political representation, the GEM uses female presence on national parliaments, institutions which normally concentrate national elites who already have access to education, as well as political and economic networks” (p. 222). She continues, “[t]his is also the case for the female share of economic decision-making positions, especially when it comes to education (as it specifically includes the percentage of women in professional and technical occupations). Another problem is that this indicator, as well as the income component, is measured only in the formal sector, which introduces a class-bias”, as generally a greater proportion of middle and upper class populations participate in the formal sector (Beteta, 2006, p. 222).

 

Statistical Methods & Results

This research uses a combination of univariate, bivariate and multivariate analyses.  The primary concern is with the effects of public education spending on income inequality, and one would expect to find a negative relationship based on the theoretical framework. The overall significance of the model is reflected by theAdjusted R Square, i.e., the percentage of inequality that can be explained by education spending considering the other co-variants included in the analysis.

Both models use a lag of ten years between education expenditure and the dependent variables to address concerns of simultaneity. The education spending data is taken from 1985-1995, while the GINI Index, urbanization and population data is taken from 1995-2005. In the second model, women’s empowerment and the other variables are measured in 2005 while education spending is measured ten years earlier (1995) due to the lack of GEM data for other years. This 10 year lag is hypothesized to be a sufficient amount of time to see the results of education spending trickle down into society. The rationale is that children entering school at age eight will have entered the work force by the time they are 18 years old.

 

Model 1:

GINI income = Education Spending as % of Government Spending + (Ed.Spending Sq.d) + % of Population in Urban Areas + % Employment to Population Ratio

GINI = β0 + β1Ed + β2Ed2 + β3URB + β4EMP + εi

Model 2:

Gender Empowerment Measure = Education Spending as % of Government Spending + (Ed.Spending Sq.d) + Fertility Rate (total births) + % Women’s Employment to Population Ratio

GEMP = β0 + β1Ed + β2Ed2 + β3FER+ β4WEMP + εi

 

For each equation, bivariate and multivariate regressions were run. The first bivariate scatter plot (see Figure 2), shows the relationship between education spending and the GINI Index, however there is no clear linear trend. For this reason, the possibility of a non-linear relationship between education and the two measures was considered. This was done by squaring the education spending variable to better represent the relationship between education spending and income inequality. Table 1 shows the findings the multivariate regression.

 

Table 1. Descriptive statistics.[2]

Variable

All Countries

Sub-SaharanAfrica

Latin America

Mean

Std. Dev.

N

Mean

Std. Dev.

N

Mean

Std. Dev.

N

Gini*

50.04

8.25

129

48.70

9.81

75

51.89

4.91

54

Education spending (%)

15.58

5.58

130

15.99

6.25

76

15.00

4.45

54

Urbanization (%)

39.06

20.25

201

29.91

13.56

141

60.55

16.85

60

Employment(%)

62.45

10.96

195

65.44

11.50

135

55.72

5.29

60

Gender Empowerment**

0.56

0.07

29

0.54

0.09

10

0.57

0.06

19

Fertility*

4.76

1.52

201

5.43

1.14

141

3.18

1.05

60

Women’s Employment (%)

50.65

16.09

186

55.69

15.72

135

38.05

8.23

54

* The GINI coefficient is measured on a scale of 1  – 100, with 0 indicating complete income equality

** Gender Empowerment is measured on a scale of 0 – 1, with 1 representing the highest level of gender equality.

*** This measures the average number of children born to women in their childbearing years.


Model 1 Results: Education Spending & GINI Index

Figure 2. Scatter plot of the Education Spending compared with the GINI Index for all years.

Table 2 demonstrates the results from the Model 1 multivariate regression. In the progression through the models, the Adjusted R square becomes more statistically significant, thus illustrating that the model itself provides a better explanation of the correlation between education spending and the GINI Coefficient.

 

Table 2. Presentation of findings from Gini and Education Spending Model

Model 1

Model 2

Model 3

Model 4

Education Spending(Independent variable)

0.136

1.248*

1.222*

1.314*

Education2

-0.031*

-0.030*

-0.033*

Urbanization(Control)

.073*

-.066*

Employment(Mediator)

-.526*

Intercept

47.871

38.886

35.693

73.442

R Square

.008

.042

.085

.445

AdjustedR Square

-.003

.021

.055

.42

 

 

The model derived here illustrates the same curvilinear relationship between education spending and inequality that Kuznet’s hypothesis supports in more general studies on inequality and development. A visual representation of this curvilinear relationship can be seen below in Figure 3. When education spending reaches 19.90%, the maximum point on the curve, the slope equals zero.[3]After education spending reaches its peak, there is a decline in the GINI Coefficient.

 

Figure 3. Education Spending and Inequality.

 

For a more in-depth analysis, the results of the first regression can be split and compared between the two regions under study. By adding a dummy variable, the difference between the income GINI in Sub-Saharan Africa and inLatin Americacan be separated and viewed in the light of their specific correlation to education spending, per region. The new model thus becomes:

D= dummy variable:Latin America= 1; Sub-Saharan Africa = 0

Gini = β0 + β1Ed + β2Ed2 + β3URB + β4EMP + β5D + εi

Table 3. Dummy variable results.

Coefficientsa

Model

Unstandardized Coefficients

Standardized Coefficients

t

Sig.

B

Std. Error

Beta

1 (Constant)

73.504

6.721

10.936

.000

Education Spending

1.302

.491

.867

2.650

.010

Education2

-.032

.013

-.797

-2.421

.018

Urbanization

-.094

.047

-.267

-2.007

.048

Employment

-.525

.069

-.715

-7.552

.000

Dummy Variable

1.760

2.071

.108

.850

.398

a. Dependent Variable: GINI

 

The results illustrate that the Income GINI inLatin Americais 1.76 times (or 76%) greater than the GINI in Sub-Saharan Africa. The 76 percent difference in the GINIs is explained by the variables included in the model above; education spending, urbanization and employment. This data is consistent with accepted assumptions about inequality inLatin America. Latin America is the world’s most unequal region, with five of the ten most unequal countries in the world located inLatin America(Inter-American Dialogue 2009).

 

Model 2 Findings: Education Spending & Gender Empowerment

Table 4 below shows the results of the second multivariate regression. While there is an increase in the adjustedR Squarevalue through the progression of the models, suggesting that female employment and fertility help to explain gender empowerment, none of the figures are statistically significant at the 5 percent level. These results do not contradict the theoretical framework; rather they make the case for greater data availability to enable better empirical analysis of the impact of education spending on gender equality.

Table 4. Presentation of findings from Women’s Empowerment and Education Spending Model

Model 1 Model 2 Model 3 Model 4
Education Spending(Independent variable)

.006

-.002

-.010

-.003

Education2

.000

.000

.000

Fertility(Control)

-.026

-.018

Women’s Employment(Mediator)

.000

Intercept

.470

.524

.681

.641

 

Further Analysis

These findings support the claims made by the Global Campaign for Education. The Campaign argued that in order to achieve universal primary education, countries had to first abolish school fees and make education accessible for all. This meant that more education had to be publicly funded and accessible for  girls. For most girls the major barrier to education was poverty, therefore abolishing school fees was the first step. However, the report also found that girls needed more female teachers, gender-sensitive learning materials, small class sizes and adequate instruction hours, and an environment that was free from sexual harassment and abuse. Countries that had made education accessible in these ways were spending at least 20 percent of their public budget on education (Global Campaign for Education, 2005). While the initial regression found the figure of approximately 20 percent of government spending to also be significant in promoting social equality, the data has struggled to reveal the extent to which any of these gender sensitive measures may also be taking place in the schools and having an empowering effect on women in society.

Quantitative research alone is insufficient to explain an issue as complex as inequality. Equality and gender empowerment may simply be improving overtime regardless of how much is being spent on education. Similarly, there is the argument that multiple factors are affecting the change in inequality and gender empowerment. Globalization, political climate, legal reform, and socio-cultural factors also affect income and gender inequality. A full analysis of competing theoretical frameworks for the explanation of diminishing inequality is beyond the scope of this paper. However, these considerations inform recommendations for policy-making and future research.

 

A Brief Consideration of Policy Implications

The above analysis confirms the assumption that education spending decreases income inequality over time, while supporting the international benchmark that 20 percent of government spending should be directed towards education in order to have an impact in reducing income inequality. However, the data also supports the assumption that while increased spending generally leads to improved access to education and to improvements in the physical aspects of the schooling (buildings and materials, hiring of more teachers), it does not necessarily bear an impact on the culture and pedagogy of schooling, including the (trans)formation or legitimation of gender power relations. The qualitative impact is impossible to measure through current statistical analysis alone.

The data also challenges the hypothesis regarding education spending and gender inequality. While it was suggested that education spending alone would be insufficient to promote equality in gender relations, inadequate data was available to show a real correlation between the two variables. As a result, the evidence is unable to support a closer examination of specific educational policies with the intent to improve gender equality. This underscores the need to improve quantifiable indicators of gender inequality at the national level. Without a quantitative analysis pointing to the importance of the correlation, qualitative research into the gendered impact of education spending policies may appear unfounded.

In both cases, policy-makers will need to expand qualitative and mixed methods research approaches to examine specific educational investments. Further research will enable policy-makers to make informed decisions not only on what percentage of government spending should be allocated to education, but how those funds should be allocated, particularly in the case of programs targeting young women and girls. Furthermore, policy-makers must propose and test new theoretical models regarding the ways education, empowerment, and gender inequality are interconnected.

 

Final Policy Recommendations:

1.     A more nuanced approach should be taken to analyze the impact of government education spending on income inequality. This approach would include qualitative and mixed methods research to measure the impact of education spending on income disparity within a nation.

2.     Policy-makers should improve methods of monitoring gender inequality within a nation. Including more precise indicators and data collection methods, would eliminate data gaps and enhance understandings of the relationship between education spending and gender inequality. This would bolster policy-making efforts to draft and evaluate school funding initiatives which specifically target girls’ empowerment.

3.     International institutions such as the UN, World Bank and Inter-American Development Bank should consider investing resources to support new measurements of inequality in education. Indicators, such as the Education GINI, could be used on a national scale or perhaps even at departmental/state levels within countries, to provide greater insight into the inequality inherent in the distribution of formal education.

In closing, policy-makers should consider that currently, intensive, qualitative studies of individual schools is perhaps the best way to  reveal the ways in which educational institutions are influencing social inequality beyond income disparity. Large scale studies using macro-level data may be useful in identifying signs that human capital is increasing and perhaps some positive gender status legitimation is taking place within the schools. However, close examination of pedagogy and the critical consideration of teacher training methods and curriculum should also be considered in policy-making efforts to create dynamic social change via a nation’s school system.

 

Bibliography

Abu-Ghaida, D., and S. Klasen. “The Costs of Missing the Millenium Development Goalon Gender Equity.” World Development 32, no. 7 (2004): 1075-1107.

Basu, A.M. “Why Does Education Lead to Lower Fertility? A Critical Review of Some of the Possibilities.” World Development 30, no. 10 (2002): 1779-1790.

Beteta, H.C. “What is Missing in Measures of Women’s Empowerment.” Journal of Human Development 7, no. 2 (2006): 221-241.

Bing, Wu Kin. “Education and Inequality in the Developing World.” In Inequality in Education: Comparative and International Perspectives, edited by D.B. Holsinger & W.J. Jacob, 86-127.Hong Kong: Comparative Education Research Centre, 2009.

Collins, R. “Functional and Conflict Theories of Educational Stratification.” American Sociological Review 36 (1971): 1002-1118.

Ganimian, Alejandro J., and Solano Rocha, Alexandra. “¿Cómo se desempeñaron América Latina y el Caribe en el Programa para la Evaluacion Internacional de Alumnos del 2009 (PISA)?.” Programa de Promoción de la Reforma Educativa en América Latina y el Caribe (PREAL). (2011).

Global Campaign for Education. “Girls Can’t Wait: Why Girls Education Matters and How to Make it Happen Now: Briefing Paper for the UN Beijing+10 Review and Appraisal.” Reproductive Health Matters 13, no. 25 (2005): 19-22.

Grilli, Enzo. “Political Economy and Economic Development in Lain Americain the Second Half of the 20th Century.” Journal of Policy Modeling 27 (2005): 1 – 31.

Inter-American Dialogue. How Poor and Unequal is Latin America and the Caribbean?Washington,D.C.: Inter-American Dialogue, 2009.

International Labour Organisation. World of Work Report 2008: Income Inequalities in the Age of Financial Globalization. International Labour Organisation, 2008.

Korzeniewicz, Roberto Patricio, and Timothy Patrick Moran. “Theorizing the Relationship between Inequality and Economic Growth.” Theory and Society 34 (2005): 277 – 316.

Lopez-Acevedo, Gladys. “Mexico: Evolution of Education and Inequality in the Last Two Decades.” In Inequality in Education: Coparative and International Perspectives, edited by D.B. Holsinger & W.J. Jacob, 417-443.Hong Kong: Comparative Education Research Centre, 2009.

Mahadevia, Indira Hirway and Darshini. “Critique of Gender Development Index: Towards and Alternative.” Economic and Political Weekly, 1996: WS87-WS96.

McDonald, P. “Gender Equity in Theories of Fertility Transaction.” Population and Development Review, 2000, 3 ed.: 427-439.

Meyer, John W. “The Effects of Education as an Institution.” The American Journal of Sociology, 1977: 55-77.

Nielson, Francois, and Arthur S Alderson. “The Kuzents Curve and the Great U-Turn: Income Inequality in U.S.Counties 1970 – 1990.” American Sociological Review 62, no. 1 (1997).

Rudra, Nita. “Openness, Welfare Spending, and Inequality in the Developing World.” International Studies Quarterly, 2004: 683-709.

Simpson, Miles. “Political Rights and Income Inequality: a cross national test.” American Sociological Review 55 (1990): 682 – 693.

UNESCO-UIS/ OECD. Financing Education: Investments and Returns; Analysis of the World Education Indicators, 2002 Edition. UNESCO-UIS/ OECD, 2003, 19-51.

V. Thomas, Y. Wang & X. Fan. Measuring Education Inequality. Policy Research Working Paper, The World Bank,Washington,D.C.: The World Bank Institute Office of the Vice President & Economic Policy and Poverty Reduction Division, 2001.


[1] We recognize that ‘empowerment’ is an ambiguous and contested construct, which renders clear-cut analysis difficult, however, it provides a useful assumption for our analysis.

[2] The majority of the data used in this research comes from the World Development Index World Bank, World Development Indicators & Global Development Finance, September 28, 2010, http://databank.worldbank.org/ddp/home.do (accessed November 27, 2010). The employment data was sourced from  Millennium Development Goals IndicatorsUnited Nations, Millennium Development Goals Indicators, 2010, http://mdgs.un.org/unsd/mdg/Data.aspx (accessed November 27, 2010). and the Gender Empowerment Measure is taken from the United Nations Development Program UNDP, 2010 Report, 2010, http://hdr.undp.org/en/statistics/data/ (accessed November 27, 2010).

[3] The following calculation illustrates this point: Y = β0 + β1X1i + … βkX+ εi ; Gini = β0 + β1Ed + β2Ed2 + β3URB + β4EMP + εi; Gini = 73.442 + 1.314(Ed) – 0.033(Ed2) – 0.066(URB) – 0.526(EMP)

First derivative of this equation: dGini/dEd = 1.314 – 0.066(Ed) ; 0 = 1.314 – 0.066(Ed); 0.066(Ed) = 1.314; Ed = 1.314/0.066; Ed = 19.90


Cornell Institute for Public Affairs

Written by Cornell Institute for Public Affairs

Leave a Reply

Your email address will not be published. Required fields are marked *

Time limit is exhausted. Please reload CAPTCHA.