Nova workboard

a blog from young economists at Nova SBE

Minimum Dropout Age: Pros and Cons

Education is one of the most important government expenditure components in most of the countries; this is justified by positive impacts on welfare. The arguments for educational investment are typically related with gains for the society as a whole (externalities) which makes plausible that people is willing to pay for the others’ education; then there are also equity concerns on these investments.

Generally, it is relatively easy to point several potential externalities such as crime reduction (more educated people tend to earn more money and they are less likely to commit crimes), political participation (more educated people tend to be more careful with political decisions and monitor them through voting) or even a more healthy society (higher levels of education seem to lead to more health awareness and preventive behavior).

At the first sight, the possibility of having negative side effects caused by education does not seem to be credible but they may exist. Increasing the school attendance and mandatory drop-out age is a good example: On one hand, it tends to decrease crime because students have less time available and it education may change individual’s behavior. On the other hand, individuals that are more likely to quit school and are forced by law to continue may cause negative externalities to the others. “Forced students” may reduce class quality with a “distracting/noisy” behavior. Indirectly, the previous policy can lead to larger classes[1] and an increase of demand for teachers can reduce the average teacher quality (necessary hire lower quality teachers to match the demand for education) which have an effect on educational outcomes.

Andersen, Hansen and Walker (2012)[2] analyze the impact of an increase of mandatory drop-out age on student victimization in the United States. The authors find that increasing minimum dropout ages increase the probability that females and younger students report missing days of school for fear of their safety and younger students are more likely to report being threatened or injured with a weapon on school property. They conclude that this policy replaces the place of crime and delinquency: from the streets to schools.

As in every policy, we should not be tempted to measure only the benefits and be careful on how to measure benefits. In the previous example, we should be careful when measuring the impact of schooling on criminality because there may be a crime reduction in the end but we cannot forget that there is a trade-off between street and school crimes and its consequences.

Guilherme Rodrigues 541


[1] There is some evidence on the negative impact of class size on educational outcomes but it is not consensual in the literature.

[2] Anderson, D. Mark, B. Hansen and M. B. Walker, 2012 ” The minimum dropout age and student victimization”, Economics of Education Review


Estimating an Educational Production Function: a challenge!

Until now there is no developed economic theory about why and how school governance affects student performance, unless one considers that schools are just like other enterprises – and teachers and school directors are like other workers. If so, literature offers relevant models as the one developed by Hoxby in 1995. Organizational decision-making models highlight costs associated with collecting and exchanging information as well as coordination, monitoring, transaction and agency costs and moral hazard issues.

       However a simple production model is behind most of the studies in Economics of Education, the empirical application of such basic economic models turns out to be frequently difficult. Moreover, Eric A. Hanushek[i] refers to a remaining considerable confusion about how analysis in this field should be conducted. Economics of Education constitutes a distinct field since results have almost direct effect into the policy process. 

       The most common inputs in educational production functions are school resources, teacher quality and family attributes[ii]; while the outcome is schooling or individual skills, historically, measured by student attainment/achievement or years of schooling completed. This measure assumes that a year of schooling produces the same amount of student skills over time and in every country as it only counts for time spent in school, without judging what happens there, therefore not providing an accurate representation of outcomes. Furthermore, achievement can be measured at discrete points in time, but the educational process is cumulative and inputs in past periods affect students’ current levels of achievement. The problem is that many times only current input measures are available.

       Misleading understandings of school mechanisms and neglected features of the data lead to analytical errors and contradictory results concerning the impact of such inputs. For instances, relatively to classroom resources, some studies claim that teachers’ education and teacher/pupil ratios have a positive and statistically significant impact on student performance but others documented a similarly negative correlation; some acknowledged a positive correlation between teachers’ experience and student performance while others show that this input is not significant.[iii]

       Solutions have been pointed out in different studies[iv]. Relatively to problems of omitting prior inputs of schools and families, one suggestion[v] has to do with the measurement of achievement relations at different points in time by concentrating on achievement growth, and so, focusing on education evolution between outcome measurement points. In this sense, prior inputs are included in the initial achievement levels that are measured.

       Additionally, panel data accounts for the causal relationship between resources and performance: suppose police makers provide more resources to students they think need them more, thus higher resources might just be signaling students with lower achievement.

       Finally, random field experiments can be an answer to deal with sample selection and other omitted variables of which Project STAR[vi] is an example. Even in this case, one may not neglect the fact that conclusions are being stressed out of a particular context with specific features.

       Then, choosing the method and the data to use – most empirical work relies on data constructed for other purposes – is crucial to get the most accurate results and infer about future educational reforms and policies.

Mariana (64)

[i] Conceptual an Empirical Issues in the estimation of Educational Production Function, Eric A. Hanushek;

[ii] In what concerns inputs, family background is usually measured using socio-demographic characteristics, parents’ education, income and family size; peer inputs as aggregates of students’ socio-demographic characteristics and achievement per school or classroom; and school inputs as teachers’ education, experience, sex, race, class size or teacher/pupil ratio, school facilities, administrative expenditures, and district or community features (i.e. average expenditure levels and location);

[iii] Revision by Eric A. Hanushek of ninety individual publications that appeared before 1995, containing 377 separate production function estimates;

[iv] compilation and deep analysis of them can be find in Hanushek (2003);

[v] Hanushek (1979);

[vi] Word et al.(1990).

viEducation Production Functions, Eric A. Hanushek (07).

African Education Crisis

One of the biggest challenges for Africa, the poorest continent in the world, is to reform all of its education system and change educational incentives. It is estimated that one third of African children are still out of school, and 61 million of them will reach adolescence without the most basic knowledge as reading and writing, undermining the development potential of the whole continent.

Countries like South Korea, Singapore based they economic successful path on their learning achievement.  Quite often African classrooms are crowded of students, have no textbooks, and teachers are not committed, being absent  the major part of the days in some countries, which means that even if children are in school probably they are not learning anything.

One survey made by the center for universal education at brookings/This is Africa estimates that 61 million children of primary school age in Africa, near 50%, will reach adolescence years without being able to read, write, or perform basic numeracy tasks. Though, the most puzzling fact is that half of these children have spent four years in school.

A countless number of resources is being lost in Africa, policy are unable to perform two essential things, first get children to go to school, and second, provide right conditions for them to learn effectively inside their schools. As some scholars use to say, the fundamental problem of Africans education is its incapacity to fill the gap built by the twin deficit in access and learning.

Anyhow, there is some good news. Something was done in the last years, particularly in what concerns primary school enrollment rates and in the percentage of income disposable to investments in schools infrastructures. Nevertheless, there are still about 30 million primary school-age children out of school. The Millennium Development Goals, a series of development targets settle in 2000 to be achieved by 2015, are more than threatened.

Furthermore, the inequality in the access to education constitutes a considerable challenge for the continent. Children from the richest 20 percent of households in Ghana remain on average six more years in school than those from the poorest households. Data tell us that being poor, rural and female is a triangle that hinders even more the already difficult education path in Africa. Beyond that, permanent conflict between ethnicities and regional forces obviously halt in some way every economy activity including educational services.

It seems quite evident that the potential of Africa, and is economic performance, is highly dependent on its labor force, and as consequence on its educational system. African nations and International authorities have to change their worries from school enrollments to learning. African children deserve it. And so does Africa.

Rui Rodrigues 556



Too Little Access, Not Enough Learning: Africa’s Twin Deficit in Education; Center for Universal Education, January 2013

Educational expenditure in Portugal

Everyone takes as given the goodness of expenditures with Education. However, we observe that different countries have quite discrepant levels of education expenditure, particularly if we look to the expenditure in education as a percentage of GDP. Even if we consider developed countries, we may face different kinds of preferences facing education, and also different degrees of efficiency, with some countries being capable of reaching the same outcome with fewer resources.

As we can see in the graph below, Portugal has increased substantially its percentage of GDP spent by the state since 1974, the year that the country found its own way towards democracy.



Despite this huge increase in the last years, Portugal is still around the OECD-15 average in terms of public expenditure in education, with about 5 percent of its GDP invested in education, still far from Denmark that invests almost 7.5 percent of its gross product in education.


 In relation to expenditure with teachers’ wages, Portuguese teachers earned more 25% in 2010 than in 2000 (USD converted PPP). Anyhow, Portuguese teachers still earn a smaller monthly salary than the average of the countries in OECD if we measure wages in converted US dollars using purchasing power parity.

By these figures, what we can say is that Portugal is not spending money in education beyond its possibilities.

Anyway, it’s important to consider the low levels of human capital presented by the Portuguese population. Only 35% of the Portuguese population aged 25 to 64 has completed at least upper secondary education. The European Union has an average of 73.4 percent of the population! In 2012 were still reported about 1 million illiterate Portuguese. This clearly states the long path that the Portuguese society and economy has to travel. Given these unbelievable low levels of qualification, a higher public expenditure in education from the Portuguese would not be odd.

Rui Rodrigues



Education at a Glance: OECD Indicators 2012

“Há quase um milhão de portugueses analfabetos”, Expresso, September 2012

Pordata, Base de Portugal Contemporâneo

Worldbank data



A comment on Sofial Amaral’s “We don’t need no education, teachers leave the kids alone?”

In a recent post, Sofia Amaral defends the generalized oppression of innocent children through compulsory schooling.

First, she argues that education can contribute to the reduction of criminality. Apparently, she thinks that it is perfectly fine to punish children—by forcing them to go to school—for crimes that they might commit in the future. Even leaving that aside, the evidence she cites (Lochner and Moretti, 2004) is that schooling decreases the probability of being arrested. Yet it seems plausible that more educated people face a lower probability of arrest conditional on crime. It’s not that they don’t commit crimes: they simply know how to get away with it. Lochner and Moretti know this, but their solution is to point at self-reported criminal activity: hardly the most reassuring evidence.

Second, she notes that education increases the probability of voting in elections. What she forgets to note is that level of education can be a good predictor of party affiliation. From my personal experience, schooling is often not so much education as socialist indoctrination: hardly a surprise, given that it is funded by a state that has every interest in self-preservation. From this point of view, compulsory schooling looks to be little more than glorified electoral violence.

Finally, she claims that a child’s educational achievement is positively correlated with his parents’. All well and good, but again, is this a reasonable motive for interfering with personal liberty, for the benefit of “society”? As Margaret Thatcher would say, “there is no such thing as society. There are individual men and women, and there are families.” My message to Sofia is simple: leave me and my family alone.



Lochner, Lance and Enrico Moretti, “The Effect of Education on Crime: Evidence from Prison Inmates, Arrests, and Self-Reports,” American Economic Review, September 2004, 94 (1), 155–189.

A perverse side of education in Mozambique

In my first comment I rose up Easterly’s voice and asked: “where has all the education gone”? I presented a model of Becker, aiming at explaining the disappointing results of the educational boom that took place for 4 decades. However, as I previously said, many factors might have undermined countries’ performances and many others might have committed the validity of some expectations. Stiglitz has once said, “What we measure affects what we do; and if our measurements are flawed, decisions may be distorted”. Following the nobel’s point, today my comment relies on an important feature of Mozambique’s society that is rooted in educational system and might have contributed to the misleading measurement and to the unsatisfactory results: Corruption.

Educational indicators for Mozambique point towards significant improvements. Indeed, as it can be read in some reports published by UNICEF, Mozambican children have now considerably better opportunities to be educated than before. This is observed at primary and secondary level, as gross enrollment ratios increased from 61 to 115 and from 7 to 25, from 1991 to 2010, respectively. Also, 61% of students completed the last grade of primary, in the same year, following and increasing trend. These results are equally preeminent for the adult population.

After this, it could be straightforwardly concluded that the educational system in Mozambique is achieving promising results. But, let me introduce you to some other facts that are not less real in Mozambique.

In its complex manner, Corruption arises all around the world. And, even though its definition is difficult to be set, Transparency International explain “When politicians put their own interests above those of the public, when officials demand money and favors from citizens for services that should be free; corruption is not just an envelope filled with money though – these people make decisions that affect our lives”. Figure 1 illustrates a world map colored according to the Corruption Perception Index, which measures the perceived levels of public sector corruption in 176 countries for the year of 2012. Two-thirds of the countries scored below 50 and Mozambique was shortly assigned with 31 scores, being the 123th more corrupt in the sample.

Figure 1: Corruption Perception Index Map, 2012


Source: Transparency International

A study from the same organization concluded that in 2011, 56% of the Mozambican interviewees perceived that in the previous 3 years, corruption level had increased. They were asked to assign a number from 0 – not corrupt, to 5 – highly corrupt to a set of 9 public institutions. Education system was ranked the second most perceived to be corrupt, scoring an average of 4 (Figure 2). Additionally, the study concluded that 68% had already paid a bribe to at least one of these services; and, from this share, no significant differences on grounds of gender, income level or age were noticed. Finally, upon the condition of having had contact with the services in the previous year, it was asked if they had paid a bribe to any of the institutions; behavioral indicators point towards police, medical and education system being the most paid (Figure 3).

Figure 2: Breakdown of institutions according to scored of Perceived Corruption in Mozambique, 2011


Source: Transparency International


Figure 3: Breakdown of institutions according to scored of Perceived Corruption in Mozambique, 2011Image

Source: Transparency International

After presenting such results, we have now another face of the education provided in Mozambique. If corruption refers to every kind of bribery that has the purpose of changing a policy outcome, it is difficult to rely on some indicators as a mirror of some situations, or at least in the same way as we do in many other countries. Education briberies take many perverse forms and affect all age students. Most of the times it is used to get a student passed, even if his/her progress doesn’t suggest so. Education in Mozambique is for many much more costly that one might think, is much more perverse in the way students perceive the society’s organization and is much less representative of the actual level of instruction that population has.

All in all, Education is not provided in the same way all over the globe, some write on computers, some on the floor. Some have high-qualified teachers, others have absent and low skilled ones. Many educational features vary across the globe; the meaning of education and its outcomes is one of them.

Joana Cruz Ferreira, 570 


–       Corruption in the Educator Sector (2007). Transparency International

–       UIS Statistics in Brief (2010), Education (all levels) profile – Mozambique. Unesco

–       World Data on Education. 7th edition (2010/11). International Bureau of Education, UNESCO

–       Corruption Perception Index (2012). Transparency International

–       Hardoon, Deborah, and Heinrich, Finn. Daily Lives and Corruption: Public Opinion in Mozambique (2001). Transparency International



Evidence on Socio-Economic Stratification

A permanent debate on economics of education regards the schools’ management and funding and how it relates to both efficiency and equity. If different schemes are responsible for differentiated school quality, it will lead to distinct learning opportunities and performances. Therefore, this creates, at the same time, a debate on the existence of stratification and its consequences on educational outcomes.

According to the OECD (2012), stratification “means creating classes of students according to their socio-economic backgrounds”. If we first relate this concept with the schools’ type of management (in almost every country, schools can be privately or publicly managed), evidence on OECD countries in figure 1 suggests that there is higher stratification in privately managed schools. The socio-economic background of students enrolled in such schools is higher than those attending publicly managed ones, with exception for Luxembourg and Chinese Taipei where the reserve happens. In countries like the Netherlands and Finland the difference is not statistically significant and so the socio-economic conditions are similar for students in both types of schools. Then, these seem to be the most equitable countries by providing access to private and public schools irrespectively of the student’s background.


Another important organizational aspect of schools is the funding source. It may be that some of the differences seen above are due to financial constraints of low-income families that prevent their children to attend privately managed schools. Therefore, some degree of State’s intervention in order to cover schooling costs could increase the proportion of students coming from a disadvantaged socio-economic background. For instance, in the Netherlands tuition costs are fully covered by the Government and schools are not allowed to ask for extra money to the parents.

In Figure 2, even though one cannot establish a causal relationship, it is clear that the level of public funding for privately managed schools is negatively correlated with socio-economic stratification. Consequently, countries where privately managed schools receive higher fractions of public funding face less stratification between public and private schools. According to the OECD (2012), “a 10 percentage-point increase in public funding for privately managed schools is associated with a 0.06 index-point reduction in stratification”.


Financial constraints can also have efficiency implications. Since privately managed schools will be highly demanded by those who can afford it while public schools will be attended mostly by children coming from a low-income family, this is likely to deliver very different learning experiences which will be reflected on the schools’ performance. This is a result not only of innate ability and school resources but also of a peer effect which is expected to yield better results in mixed contexts (that is, in classes with both low-income and high-income children).

Figure 3 shows that students in countries with less stratification tend to perform better in reading, while countries with more stratification tend to have lower results. Once again, this figure does not establish a causal relationship but adds important evidence for the debate by suggesting that it is possible to break the efficiency-equity trade-off.


I would like to conclude by stating that would be important to study the causal effects between these variables. Only then it is possible to derive policy implications. Some studies at the micro level have been conducted recently, namely through randomized experiments, but their external validity may not always be ensured, meaning that extrapolating its conclusions at a macro level is not straightforward.


Sofia Oliveira


Reference: OECD (2012), Public and Private Schools: How Management and Funding Relate to their Socio-economic Profile, OECD Publishing.

My family decides my future

It is a common fact that family background is a major determinant in the education production function for children. But how important is the childhood for an individual’s cognitive skills and future earnings? When looking at the impressive amount of empirical work done on the subject the most studies focus on the role of mothers. But the results are ambiguous.

Many American studies show that having a mother who stays at home compared to having a working mother has positive effects when it comes to school achievements. Ruhm (2004) finds negative effects of maternal employment in verbal ability of children aged 3 and 4 and on reading and mathematics for children ages 5 and 6. This negative effect is supported to have long run effects by Bernal (2008) that also uses American data and find that full-time maternal employment reduces children’s high school test scores. Furthermore, the negative effect of maternal employment is found to be more strongly negative among children of highly educated women (Gregg et al, 2005) and women with high socioeconomic status (Ruhm, 2008).

Looking at Northern European data positive or no effects are found. Dunifon et al. (2013, forthcoming) find using Danish register data a positive effect of mothers working full time in the first 3 and 15 years of their children’s life on the children’s performance in 9th grade. Furthermore, Rasmussen (2010) finds no effects of expanding the maternity leave by six weeks in Denmark in 1984 on high school enrollment, high school completion rates, and high school grades. Lastly, Rege et al (2011) find that children whose mothers lost their jobs in Norway due to a plant closing while the children were in eighth or ninth grade had a similar grade point average in 10th grade as other children.

The importance of the course of one’s childhood is emphasized by Heckman (2008). He concludes that 50 percent of inequality of the present value of lifetime earnings is due to factors determined by age 18. And the effect of family background as above discussed as maternal employment decisions does not only affect cognitive skill formation, but also many other skills, such as socioemotional skills, physical and mental health, perseverance, attention, motivation, and self confidence. All these skills are associated with higher procuctivity and lower inequality (Heckman, 2008).

Summarizing the above, family background plays an important role for a child’s educational attainment. Furthermore, when investing in reducing the inequalities of being born into a disadvantaged family, it is important to intervene at a early stage in the child’s life. Another challenge that prevail from the above results is that the effects of maternal employment is ambiguous. Recommendations of policy interventions cannot be generalized. Whether there are any significant cross-country differences needs to be investigated further. The political systems and the family life of USA and Northern Europe are very different, and could be possible drivers of the different results. But one main conclusion arises; your family decides a very big part of your future.   

By Anne


Bernal, R., 2008. “The Effect of Maternal Employment and Child Care on Children’s Cognitive Development,” International Economic Review 49(4): 1173-1209.

Dunifon, R, Hansen, A.T, Nicholson, S and Nielsen, L.P, 2013 –fourthcoming. ” The Effect of Maternal Employment on Children’s Academic Performance”.

Gregg, P., Washbrook, E., Propper, C., and Burgess, S., 2005. “The Effects of a Mother’s Return to Work Decision on Child Development in the UK,” The Economic Journal 115: F48-F80.

Heckman, J., 2008. “Schools, Skills, and Synapses”. Economic Inquiry, Vol 46(3): 289-324

Rasmussen, A.W., 2010. “Increasing the Length of Parents’ Birth-Related Leave: the Effect on Children’s Long-Term Educational Outcomes,” Labour Economics 17(1): 91-100.

Rege, M., Telle, K., and Votruba, M., 2011. “Parental Job Loss and Children’s School Performance,” Review of Economic Studies 78: 1462-1489.

Ruhm, C. J., 2004, “Parental Employment and Child Cognitive Development,” Journal of Human Resources 39(1): 156-192.

Ruhm, C. J., 2008, “Maternal Employment and Adolescent Development,” Labour Economics 15: 958-983.

The effect of maternal employment on children’s academic performance – across countries

From a child development perspective, mothers face a trade-off between time devoted to parenting and money when deciding on whether to work or not. Both money and time devoted to parenting are believed to have a positive effect on a child’s cognitive development. Therefore, it is relevant to estimate the effect of maternal employment on children’s development. Is it plausible to think that the direction of this effect changes across countries?

A research paper by Dunifon et al (2013) associates children’s well-being with their academic performance in high school, and estimates the causal effect of maternal employment on children’s educational outcomes. The study uses detailed Danish data on over 125,000 children born between 1987 and 1992. In two out of tree model specifications they find a positive significant correlation between maternal employment and children’s grade. The papers Instrumental Variables model specification suggests, that a child of a woman who work 30 or more hours per week while her child was under the age of four is predicted to have a GPA that is 5.6 percent higher, than an otherwise similar child whose mother worked between 10 and 19 hours per week. They find no evidence of a negative association between maternal employment and children’s grades. Nevertheless, this result is in contrast to a series of related studies that use data from U.S., Europe, and Canada.

Evidence based on longitudinal data from the United Kingdom and the United States generally suggests that full-time maternal employment during the first year of a child’s life is associated with poorer child outcomes, especially poorer cognitive outcomes.*     

The paper by Dunifon et al (2013) mentions the extensive paid leave time, relatively lower work hours and the generous early care and education programs as potential factors that will make the trade-off less constrained. Therefore, the implications of maternal employment for child well-being may differ from other countries.

The reason for the divergent result of maternal employment found by Dunifon et al (2013) compared to studies from other countries is not clear. It could be due to the unique situation faced by working mothers in Denmark, or due to sampling or other methodological reasons. Possibly it is plausible to think that country specific effects may influence the effect of maternal employment on child development, but more research is needed to conclude anything.





R. Dunifon, A. Hansen, S. Nicholson, L. Nielsen: “The Effect of Maternal Employment on Children’s Academic Performance”, January 2013 (WORK IN PROGRESS)


What explains the gender differences in enrollment rates in Denmark?

In Denmark the share of women taking a long university education (minimum of 5 years at University) has gotten bigger than the share of men. Looking at data from 2012 on highest attained education of the Danish population aged 25-29 the share of females with a long university education is 7.75 percent and the male share only 6.46. 

Share of male and female population aged 25-29 with a long educational degree in 2012





     Source: Statistics Denmark and own calculations.

By other means, a tendency has evovled in recent years where more young women than men are getting highly educated. But what are the possible explanations for this new development? The Danish job market is gender segregated with high share of women concentrated in the public sector and a high share of men in the private (European Commision, 2009). Also on average the men are getting higher wages than the women. Looking at the most recent data available (from 1992) on wages of Economists aged 30-34 the male wage is 13 percent higher than the female wage.

Average wage of economists in 1992 for men and women  (in DKK)





                                           Source: Statistics Denmark. The wage differences are similar for other higher educations and future years.


If we are looking at wages as the only gain from education according to the human capital theory (Becker, 1964), it is difficult to explain the gender differences of the educational choices of young people in Denmark. More men and fewer women should go into higher education since the benefits are on average higher for men than women. But could there also be some other explanations for the gender differences? The most obvious could be that men have a higher opportunity cost of choosing education instead of work because the wage level at all levels of education is higher for men than women. But that does not change the fact that the wage difference yields higher benefits of education for men.

The human capital theory is build on the theory that educational choices are based solely on the economic benefits and costs of education. This theory might not always be valid when investigating educational choices in highly developed societies. The data above suggests there are some unexplained reasons for choosing to go into higher education that is not associated with economic valuation. Could educational choices in countries that do not face strong liquidity constraints also depend on more sociological factors such as recognition, social and cultural capital? Many sociologists have developed theories on these subjects with the most famous being Pierre Bourdieu and Axel Honneth. Anderson and Honneth (2005) write about the importance of self-esteem and recognition of the society around you. Individuals need a level of self-esteem, trust and respect that can be gained through recognition of others. Being highly educated may lead to a higher level of recognition from others and a higher level of self-esteem. Can these non-economic aspect of power, respect and self-esteem gained from education be tested empirically? If yes, it might share some light to the new development in the Danish education system.


Anderson, J and Honneth, A. 2005, “Autonomy, Vulnerability, Recognition, and Justice “ Autonomy and the Challenges to Liberalism: New Essays, pp. 127-149.

Becker, GS, 1964, “Human Capital: A Theoretical and Empirical Analysis, with Reference to Education”. Chicago. University of Chicago Press.

European Comission, 2009. “Gender segregation in the labour market. Root causes, implications and policy responses in the EU.”



The Correlation Between Education and Fertility

The Nobel laureate William Shockley controversially argued in the late 19th century that the future of the world population was threatened as people with low IQs had more children than people with higher IQs. The first studies concerned with this issue found that there is an actual negative correlation between intelligence and fertility rates. The smarter people were, the lower was their fertility. On the other side those studies also found that the survival rate among more intelligent people was higher so that the overall net effect on population was unclear. Nevertheless, assuming that the survival rate in general increases due to better living conditions, the negative correlation between intelligence and fertility suggests that the average IQ should be decreasing.

In more recent studies, scholars extended their research from only national data towards international data and used educational attainment data as proxies for intelligence. The following chart shows data provided by the World Bank to picture the correlation between fertility rates and female education.

We can see that, throughout the different regions of the world, the completion rate of primary education among women not only increased but also that the fertility rate decreased. To see the change in both variables over time please click on the picture. It suggests a clear negative correlation between education and fertility throughout the world.

But the United Nations find five different explanations that cause a decline in fertility:

1)   The mortality decline in childbirth and the decline in child mortality.

2)   The decline in the female age at marriage.

3)   The increase in female literacy and education.

4)   The female economic participation.

5)   And the increased access to contraception.

The first cause implies that if more children survive, parents have less reason to have more children in order to be secured and taken care of when getting old. The second reason holds, as women getting married in a younger age will naturally be able to give birth to more children than a woman getting married later. This suggests that early childbearing can cause less education as women have less time to go to school when the first child is born. This again is directly related to the third cause. The fourth reason implies that if women have access to economic activities they will decide to have fewer children in order to gain income. And income is again related to intelligence and education.

This relationship between higher income and a decline in fertility is named the demographic-economic paradox. In the following, scholars found what is called the fertility-development controversy. We can see that there is not only a negative correlation between education and fertility but also in the Human Development Index and the fertility rate of a country.

This suggests that an increase in the standards of living reduces fertility and that education therefore seems to play an important role. But it is not clear whether or not education causes a decline in fertility or the other way round.

Written by Julia Seither


Graff, Harvey: “Literacy, Education and Fertility, Past and Present: A Critical Review”.

United Nations Report: “Completing the Fertility Transition”.

World Bank Data

A follow up on corruption and the educational system

Being corruption a pervasive and significant phenomenon that emerges everywhere and that is difficultly eradicated, it has recently attracted the attention of the development economic researchers. Though no study has been made on the grounds of corruption in Mozambican educational system, some have tackled the issue elsewhere.

First of all, it is important to acknowledge that corruption is a very broad concept, difficult to be defined and thus, to be measured. Researchers are still struggling to find the best way to approach it, which is why little work has been done.

A starting point to the literature was, for example, the contribution of Shleifer and Vishny (1993) who tried to define corruption. They stated that it is referred to “a sale of government property in order to satisfy personal desires rather than to maximize society’s welfare.” Some other theoretical papers can be found on this.

Nevertheless, considering a measurement of corruption in education, there are few works to be stressed; I choose the exceptional contribution of Ritva Reinikka and Jakob Svensson (2004) to present you today.  This consists on a natural experiment to measure the scope of corruption in educational system in Uganda. It represents a great stimulus to the literature, as authors designed a new tool to approach the matter- the public expenditure tracking survey. Such instrument is used to gather data on government flow and service delivery. Using panel data from one survey of primary schools, from 1991 to 1995, they measured the amount of payments from a World Bank fund aimed at covering the non-wages expenditures of public schools that actually reached the intended end-user.  This study states that, on average, 13% of the grants were actually received by schools. Authors could conclude that resource flows were endogenous to schools’ sociopolitical endowment. This is, some schools received nothing, and others could grab part of the grants due to their bargaining power. Results reveal that government officials were the ones capturing grants.

All in all, one might re-think what actually happened to the public investment on education that took place for 4 decades. Having a look at the government officials pockets could be one answer to Easterly’s work “where has all the education gone?”

                                                                                Joana Cruz Ferreira,  570


– Shleifer, Andrei & Vishny, Robert W, 1993, “Corruption,” The Quarterly Journal of Economics, MIT Press, vol. 108(3), pages 599-617, AugustReinikka,

– Ritva, and Svensson, Jakob. Local Capture: Evidence from a Central Government Transfer Program in Uganda  (2004)

Can Pre-School Education Serve as a Growth Engine?

In her last blog, Joana Cardim wrote about the return of pre-school education to future education and wages, and rose the question if Portugal should invest more in it so as to close the gap in educational attainment compared with other European countries. She refers to neurological and economic studies showing the impact of pre-school education in other nations, and argues that it positively affects the non-cognitive skills of pupils that are related to college attendance and outcomes.

The effects mentioned above can be described as being the private returns to education. Every additional year of schooling will increase an individual’s skills and is reflected in higher wages and educational attainments. But aside from the individual’s return to education, there is also an effect on the society if people obtain more education – referred to as the social returns to education. In September 2012 Timothy Bartik claimed at the TEDxMiamiUniversity conference that the effect of pre-school education on the local economy is even more important than the private returns. In his talk ‘Can pre-school save the economy?’ he argues that governments and policy makers have to change their viewpoint that pre-school education primarily affects the individual and recognize that their decision on the provision of pre-school education will directly influence the economic development and growth of the region.

His main argument is that a higher quality workforce – due to more education through pre-school programmes – will lead to higher wages, and thus economic development and growth. The enforcement of education will bring more and better jobs into a region or state and therefore promote higher earnings not only for the individuals having obtained this education but also for the remaining state’s residents. The provision of pre-school education will thereby have the same influence as business tax incentives programmes – but at a lower price. Timothy Bartik argues that according to his research an investment of 1$ of additional education expenditures in pre-school education will yield 2.78$ of per capita earnings of the state residents.

One of the explanations for this effect could be a peer effect inside firms. This means that people that work in an environment with better educated colleagues will be more productive, thereby making the firm more competitive, and ultimately fostering growth. Another reason for economic development can be found in the effects of education on crime and political participation. The negative correlation between education and crime suggests that pre-school education can have a significant impact on the reduction of crime, which again has a positive impact on economic development. Other social returns to education can be found in sustaining democratic attitudes and increasing political participation.

Nevertheless, Timothy Bartik received a lot of sceptics about his findings and their magnitude as for example a paper by Acemoglu cannot find significant social returns to education. Bartik’s policy recommendations should thus be analysed carefully in order to understand the real impact of pre-school education.

Written by Julia Seither

Education and Health- What kind of link?

In many countries and over several time periods, researchers have gone to great lengths to learn more about the association between education and health. Indeed, the rationale for focusing on this relation has been the fact that both education and health constitute two critical determinants for human capital accumulation. Essentially, there are three main potential reasons that support the view that there is a positive link between education and health. The first points out that individuals with a better health extract more benefits from education, while the second highlights the existence of common factors ( background) that have a similar impact on both domains. Finally, the third idea believes that education has a positive causal effect on health outcomes. However, as I will argue below, third-party effects can make these returns uncertain.

As illustrated in Fig.1, more educated people tend to have a higher life expectancy as well as better health outcomes. This is coherent with the view that education contributes to increasing health knowledge and healthier habits. In light with this view, Acemoglu, Johnson and Robinson[1] (2003) argue in favor of reverse causality between education and health as poor health conditions mean a lower life expectancy and, therefore, a decrease in the returns from investing in education. Additionally, an increase in educational attainment provides room for improved jobs that are typically associated with higher income, more financial stability, insurance coverage, better working conditions and employment benefits. All these effects together are very likely to boost self-esteem, reduce the level of stress and anxiety and hence lead to a better health. The adoption of risky-behaviours like the consumption of drugs, alcohol and smoking is shown to be lower the more educated individuals is. Field (2005)[2] found that parents whose children had more years of schooling were more likely to quit smoking. Moreover, in developing countries infant and child mortality, morbidity and epidemics are paramount issues that do not allow for proper human capital accumulation. In this context, Miguel and Kremer (2004)[3] showed that a deworming program in rural Kenya ended up being a very cost-effective measure as it accounted for fabulous within and cross-school externalities that consequently increased school attendance rates and learning. A large body of evidence has also proved that maternal education impacts infant and child health, especially birth weight babies. Nevertheless, Cutler and Muner (2006)[4] believe in the existence of third factors that explain the increase in both schooling and health, meaning that they are not necessarily correlated. For this reason, policy-makers when deciding in favor of subsidies for schooling -which could reduce costs of the health sector- should take into consideration that the observed correlation between education and health has not been strongly confirmed by sufficient evidence. In fact, unobserved factors that include family background or genetic traits could be explanations for the better health registered by more educated people. Funchs (1982)[5] also found that “individuals with lower discount rates are more likely to invest more heavily in both education and health”.

All in all, I firmly believe that more research in the field requires more powerful econometric tools to distinguish between the effects of education on health, and vice-versa, controlling for key-variables that might be capturing that relation.

Ana Correia

[1] Acemoglu, D., S. Johnson and J. Robinson (2003), “Disease and Development in Historical Perspective”, Journal of the European Economic Association 1,pp. 397-405.

[2] Field, Erica (2005), “Are there upward intergenerational education spillovers on health? The impact of children’s education on parents’ smoking cessation,” Harvard University

[3] Miguel, Edward, and Michael Kremer (2004), “Worms: Identifying Impacts on Education and Health the resence of Treatment Externalities”, Econometrica, 72(1), 159-217.

[4] David M. and Muney, Adriana Llleras, Education and Health: Evaluating Theories and Evidence, June 2006

[5] Fuchs, Victor R. (1982), “Time Preference and Health: An Exploratory Study”, in V. Fuchs (ed.)Economic Aspects of Health (Chicago: The University of Chicago Press).

Quality Teaching in Higher Education

Quality teaching in higher education is reputed to be very relevant for student learning outcomes. For this reason, improving quality teaching is essential and creates varied challenges for higher education institutions, taking into account that recently the educational system has received attention from different forces within it. One of the most common definitions applied to quality teaching is the use of pedagogical techniques to produce learning outcomes for students . Indeed, it comprehends dimensions such as the effective design of curriculum and course content, the learning contexts that should be put into practice, the feedback from the students as well as a proper assessment from students of the learning process.
In this context, Hopkins et al (1997) proposed three dimensions of effective teaching. First of all, we have teaching effects that include a combination of teaching skills and behaviours like the management of time and the promotion of independent working. Secondly, there is the acquisition of effective teaching models in the sense that quality teaching requires creating a suitable learning environment inside the classroom. The third factor puts emphasis on teacher artistry as a way of accounting for the teacher´s personal responsibility for generating all the conditions for effective learning.
Furthermore, quality teaching is nowadays subject to new paradigms, namely concerning the fact that when graduates enter the labour market they face higher uncertainty, risk, complexity and interdisciplinary work than before. As a result, the demand for soft-skills- mainly interpersonal skills- has risen over the past years thus putting pressure in teaching at the higher educational level. Additionally, students value highly equality of teaching and learning opportunities, fair assessments and expect to receive at universities the crucial tools for their career later on. Hence transmitting knowledge and expertise is no longer the only role of higher education teachers; they need to perceive their job as an important part of a dynamic learning community, for example through a more prominent bridge between teaching and research, innovative learning platforms, guiding and tutoring students and implementing assessment models aligned with student-centered learning. Central to this analysis is also the recognition that it is paramount that, at the same time, teaching practices and the degree of commitment to this new paradigm are adequately assessed and documented. Under a scenario of high competition at the high education level, the engagement in national and especially in international networks allows to share and expose the best practices of the quality teaching staff. Similarly, it is up to universities to instigate research-inspired teaching and to distinguish teaching excellence by divulging publicly their accomplishments and using them as role models for the rest. Also, a permanent upgrade of pedagogical skills based on professional development initiatives, training, peer-evaluation and constructive feedback are said to contribute to a “learning community” approach of quality teaching. A good example of how taking all of these aspects into account can lead the way to quality teaching is the Catholic University of Portugal (Porto) which was categorized as having a university-wide teaching-learning approach. Particularly, the University developed a Sistema de Garantia Interna de Qualidade (SGIQ) and a Skills Development Plan for Teachers that envisaged connecting the technological and pedagogical spheres.
All in all, it is expected from institutions to lead the way and instigate the match between the expectations of students and the requirements imposed by employers. Promoting institutions as effective learning communities where leadership and cooperation are regarded as key-determinants might also alleviate the tensions between proponents of innovation and those not as prone to change.

Ana Luísa Correia