Closing the Gap in Turkey: Evidence of Improved Quality and Reduced Inequality in an Expanding Education System


This page in:



Turkey’s remarkable economic growth over the last decade has been a much quoted success story. One often hears that the country trebled its per capita income, and has become the 16th largest economy in the world. One hears less often that this economic growth has been inclusive, accompanied by reduced poverty and expanded access to social services in health and education. And yet even these debates on expanded social services rarely move beyond quoting the headline numbers to look at the dynamics of change in the sector(s). This omission is unfortunate because the dynamics of change in the social sectors can be a harbinger for future progress. I want to draw the reader’s attention to the unheralded progress in the education sector.

Few issues are as controversial or hotly debated as education reform in Turkey. Most recently, controversy has surrounded the recent passage of the “4+4+4” law which has raised  the number of years of compulsory education from 8 to 12 years. The new law implies revisiting the rules governing transitions from lower to upper secondary schools, the choice of curricula and schools available to students at different levels of education. Some controversies are as old as the republic, such as the role of religious education in the education system of a secular state, and these continue to generate much emotion, now within the context of the changes implied by the new law.  Underpinning the latest incarnation of most of these controversies are differing views on the progress of the education sector in Turkey over the last decade.  
Interestingly, the debate on education quality is short on empirics. The World Bank with Turkey’s Ministry of National Education, analyzed the latest PISA data for the country and found several interesting and encouraging results. In short, the system has expanded impressively, with net enrolment rate in secondary education jumping from under 50% in 2001-02 to over 67% now (girls enrolment has gone from 43% to over 66% in the same period), adding over 1.5 million students to the secondary school system.  Data from PISA also tells us that this expansion has been accompanied by an overall improvement in PISA scores for 15 year olds of 20 points or more between 2003-2009 - equivalent to adding an additional half year of schooling to each student’s study period. Even more impressively, the improvement has come mostly from students from disadvantaged backgrounds. While a lot of work remains for the future, schools have become less segregated and inequality of access and educational performance had been reduced remarkably. 
Our analysis finds that Turkey’s performance in PISA is higher than would be predicted, taking into consideration its level of economic development and income. At around 450, Turkey’s PISA scores in reading, math, and science are better than average, when its level of GDP per capita is taken into account. The average PISA score of Turkish students in 2009 was approximately 10 points above that predicted by its income level. This performance looks even more impressive once the socioeconomic level of students, as well as the level of per capita GDP, is taken into account. By factoring this in, Turkey achieves almost 70 PISA points above average. In fact, in these terms, Turkey’s performance is nearly the highest among the 65 participating countries, on par with Korea or Singapore.

The increasing quality of education accompanied by reduced inequality in student performance has contributed to Turkey’s comparatively strong results in PISA. The country’s scores in the three PISA disciplines (reading, math, and science) increased sharply by 20 points or more between 2003 and 2009. At the same time, inequalities in student performance have decreased as these gains in PISA scores have come overwhelmingly from low and medium performers. For example, the performance of the bottom 1 percent of the cohort has gone up by 30 points in reading, 33 points in math, and 25 points in science; the performance of high-achievers has, on average, also improved, although by less than that of the lower performers. During the same period, the enrollment rate of 15-year-olds has grown by a strong 7.8 percent per year according to PISA, which makes these improvements even more remarkable.


While improvements in educational outcomes have, in part, resulted from the general increase in living standards (hence better overall opportunities for the new generation), increased effectiveness in the delivery of education has also played a significant role. Thus, the socioeconomic status of each student’s family mattered less for his or her PISA results in 2009 than it did in 2003. Our analysis shows that PISA score for students in the lower socio-economic quintiles improved much more markedly than for students in the top quintile. This is because the delivery of education services has become more effective in the period between these PISA tests - improving the schooling provided to almost all Turkish students and thereby reducing inequalities.


Despite these successes, some challenges still remain. The performance of Turkey’s average 15-year-old is still one full year (or 40 PISA points) behind the OECD average and a student’s socio-economic background remains a far more important determinant of educational outcomes in Turkey, than it does in other OECD countries. Around 25 percent of the Turkish 15-year-olds do not read well enough to be able to analyze and understand what they are reading and are therefore considered by the OECD to be “functionally illiterate.” However, this rate has been declining at a fast pace since 2003, when the equivalent proportion was 36 percent. Results and data from the 2012 round of PISA are expected in December 2013, and will allow for a more detailed analysis of progress made since 2009. Continued analysis of educational performance data should help shed light on Turkey’s ongoing educational debates and will hopefully contribute to greater consensus on what policies are needed to help the country  fully close the remaining gap to reach the standards of the world’s leading education performers like Korea, Singapore or Finland.

We discussed options for the future at a recent conference on education in Turkey that was jointly hosted by the World Bank and the Turkish Ministry of National Education. Will Turkey sustain or even accelerate recent gains and what measures might contribute most to this? I look forward to hearing your views online.

Also read this blog in Turkish.




Naveed Hassan Naqvi

Country Manager for Azerbaijan

Join the Conversation

Pelin Yenigun Dilek
July 18, 2013

Problem is that Turkey's social and governance indicators do not improve as much as its income indicators. There has to be a convergence by income indicators deteriorating or social and governance indicators improving even faster. Experiences of other countries Show that the gap almost always closes one way or the other.…

July 18, 2013

Glad to hear that the gap has closed over the years.

jagadish jagadishchandran
July 20, 2013

Continued analysis of educational performance data should help shed light on Turkey’s ongoing educational debates and will hopefully contribute to greater consensus on what policies are needed to help the country fully close the remaining gap to reach the standards of the world’s leading education performers like Korea, Singapore or Finland.

Semih Akçomak
July 30, 2013

When I read this piece I thought for a second that I live in another country! Regardless of status, education, job etc. almost all my acquaintances have been complaining how bad the recent changes in education system in Turkey. These are people who have children at schools. This piece argues exactly the opposite. My reservations: 1) The nice stata graph doesn't show much. It is a simple correlation, no causation; 2) The PISA scores is a nice source to compare different countries. However almost all professors at METU complains that the young generation has great difficulty in using the language; connecting related thoughts on a subject matter; and expressing themselves orally (even in Turkish!) most of which are analytical skills that PISA test can hardly measure. The current system is based on memorizing probable questions, certain style of solving problems, test technique etc. However we should educate our youngsters in such a way that they become skilled in obtaining new skills. This simple principle is applied in many developed countries but it lacks completely in Turkey 3) The second graph may tell you another thing if you do not look on the bright side: the current education system is destroying the top 15 percent performers! Don't countries depend on the bright brains?
The data to an outsider may tell you things that are quite different in reality. Here is my op-ed that I wrote some time ago on the current state of the Turkish education system I really wish that I am wrong and you are right!

Mehmet S.
October 07, 2013

I am not sure which Stata graph you are referring to, but I'm assuming it's average PISA scores and per capita GDP. The author is not claiming anything about causality, so I'm not sure why you find the graph to be inadequate. It is interesting that Turkey does pretty well controlling for SES, and should be investigated further -- I think this is the extent of the point that the author is making.
It's definitely certain, however, that a handful of professors at METU complaining about skills of their students doesn't necessarily mean that the students actually lack analytical skills! :)
As for your last point: it's the last 4 percent. This could be due to a number of reasons (the PISA test changes from year to year, the small sample of schools with such top-notch kids could've changed etc.). I'm not claiming it's not a problem, but it's nowhere near "destruction." These kids are already scoring way above OECD averages.

Bekir Gur
August 01, 2013

Great piece, Naveed. As you well know, there is a debate going on in Turkey about abolishing the entrance exams to high schools. I am happy to see that education is increasingly becoming a priority in public debates in Turkey. I think international experts have much to offer to these debates.
I think the current crisis in transition from middle school to high schools in Turkey is not related with the central exam per se. Rather, the current selection/placement system admits too much students to selective schools. Here is my reaction to the current debate on the transition:
"Quest for the reasonable in transition to high school"…

Yakup Ozsoy
August 19, 2013

World Bank supports a number of initiatives in health and education in Turkey. Recently, they have started a siginificant effort to prove how well Turkey did under the current government. Their method is to provide a rosy picture by emphasizing the good part of the story, accompanied by through analysis. At the end, after you are convinced that all is well in Turkey they provide a few problematic issues. In healthcare Lancet article of July 2013 is a very good example. In education this work presents the example. Leaving the fact that we are way behind OECD average in Pisa scores to the end and spending 90% of the piece to tell that we are 'bon pour l'Orient' does not appear to be a fair way of presenting current situation. Especially considering that Turkey lacks social scientists capable of analyzing the data and World Bank pieces usually find influential publication channels, it provides a very unbalanced description.

Naveed Naqvi
August 21, 2013

We take your point that Turkey has challenges ahead. The education system is still far from where it aims to be (at a par with the best schooling systems in the world), but data shows that the country has made remarkable progress towards its goal, even though it has not arrived at its intended destination yet. That progress is swift and has benefitted the poor deserves greater recognition, and is an impressive achievement. The blog is based on a more detailed report- "Promoting Excellence in Turkey's Schools"- which is available on the Bank website.

M Alper Dincer
September 27, 2013

PISA (2003-2009) by OECD, TIMSS (1999-2007-2011) by IEA and ÖBBS (2002-2005-2008) by MoNE are studies which contain detailed information on student achievement at varying grade levels and subjects in Turkey. All of these studies point out that student achievement in Turkey has been increasing and becoming more equitable since late 1990s/early 2000s. However, given that public education provision in Turkey experienced several significant interventions starting from late 1990s, the task of explaining the dynamics of this trend in student achievement is very challenging. First, education faculties was reformed in 1997. Thus, this reform may have caused a positive impact on teacher quality. Second, bylaw of norm-staffing was put in force in late 1999 or early 2000. This bylaw may have caused an improvement in the unequal distribution of teachers across provinces and cities, towns and villages. Third, recruitment of public servants (i.e. public school teachers) was tied to performance in centralized testing. Thus, nepotism was replaced by an objective measure for recruitment of teachers. This may have caused a positive impact on teacher quality, as well. Curriculum was reformed and weekly instructional times of math and science (measured by PISA, TIMSS & ÖBBS) was increased from 3 to 4 lesson-hours in 2005. This may have also caused a positive impact on learning. These hypotheses are waiting to be investigated. Increased effectiveness may be due to all of these factors or none them. We need lots of empirical studies and evidence on this front and I think, this blog-post and the WB report is an encouraging and timely start.

Emrah Arbak
November 28, 2013

I feel that the first diagram, which appears to be plotting the residuals of the PISA / GDP regression, could be misleading due to the financial crisis. For the basis year of 2009 used here, the income figures went down quite drastically in some countries but not in others. A high residual may simply be a reflection of the fact that the drop in GDP was relatively larger, which was definitely the case for Turkey. A better alternative would be to use the average incomes for a longer period (i.e. 2003-2012?) that contains 2009 as a focal point.
Moreover, the improvements plotted in the second diagram may be biased due to sample selection. The 2003 sample was comprised of 41 countries, most of which were advanced countries. The sample was enlargened to 65 countries in 2009 with the introduction of many emerging and less developed countries. This means that the (non-standardized) average scores for the 2009 sample were lower. The bias may occur since the analysis is based on standardized scores (mean = 500). I would assume that most -- if not all -- 2003 PISA countries would have experienced the same trend using those scores.