Education
Benedikt Goderis
Education

Which countries have the most highly skilled populations?

Education provides people with the skills they need for economic and social participation in society. How do countries in Europe, Northern America, Oceania, Japan and Korea compare when it comes to the average level of skills of their populations? And to what extent do they offer equal opportunities for all in the classroom? In Chapter 2 of Public sector achievement in 36 countries, we use various indicators to address these questions. Here we briefly present those which cover the largest number of countries and the longest periods of time.

Average performance in mathematics and reading

The report first looks at cognitive skills, analysing the average performance of 15-year-olds in mathematics and reading. Figure 1 reveals large differences between countries. In 2012, Korea and Japan had the highest scores and Southern, Central and Eastern Europe the lowest. Western and Northern Europe, Oceania and Northern America fell in between, with each region having comparable results. Intraregional differences were the largest in Central and Eastern Europe, with scores ranging from low in Bulgaria and Romania to quite high in Poland and Estonia.

In 2012, average performance in maths and reading was by far the strongest in Korea and Japan

However, a comparison with data from earlier years - 2000 for reading skills and 2003 for maths skills - shows that the differences between countries used to be even bigger. This is because most countries with low initial scores improved their performance over the period until 2012, while most countries with good scores at the outset fell back. Poland (maths and reading), Romania (maths only), and Latvia (reading only) achieved the biggest improvements in test scores, while Sweden and Finland witnessed the largest drops.

Figure 1
Average maths and reading scores in 2000/2003 and 2012
Mean PISA maths scores
Country 2003 2012
Country 2003 2012
Austria 505.54 505.61
Belgium 514.53 529.29
France 494.98 510.80
Germany 513.53 502.99
Ireland 501.50 502.84
Luxembourg 489.85 493.21
Netherlands 522.97 537.82
Switzerland 530.93 526.56
United Kingdom 493.93 495.44
Denmark 500.03 514.29
Finland 518.76 544.29
Norway 489.37 495.19
Sweden 478.26 509.05
Greece 452.97 444.91
Italy 485.32 465.66
Portugal 487.06 466.02
Spain 484.32 485.11
Bulgaria 438.74 413.45
Croatia 471.13 467.25
Czech Republic 498.96 516.46
Estonia 520.55 514.58
Hungary 477.04 490.01
Latvia 490.57 483.37
Lithuania 478.82 486.42
Poland 517.50 490.24
Romania 444.55 414.80
Slovak Republic 481.64 498.18
Slovenia 501.13 504.46
Australia 504.15 524.27
New Zealand 499.75 523.49
Canada 518.07 532.49
United States 481.37 482.88
Japan 536.40 534.14
Korea 553.77 542.23
Mean PISA reading scores
Country 2000 2012
Country 2000 2012
Austria 489.61 492.06
Belgium 508.62 507.13
France 505.48 504.74
Germany 507.68 483.99
Ireland 523.17 526.67
Luxembourg 487.81 479.42
Netherlands 511.23 513.12
Switzerland 509.04 494.37
Denmark 496.13 496.87
Finland 524.02 546.47
Norway 503.94 505.28
Sweden 483.34 516.33
Greece 477.20 473.80
Italy 489.75 487.47
Portugal 487.76 470.15
Spain 487.94 492.55
Bulgaria 436.13 430.40
Czech Republic 492.89 491.58
Hungary 488.46 479.97
Latvia 488.69 458.07
Poland 518.19 479.12
Romania 437.60 427.93
Slovak Republic 462.77 469.16
Australia 511.80 528.28
New Zealand 512.19 528.80
Canada 523.12 534.31
United States 497.58 504.42
Japan 538.05 522.23
Korea 535.79 524.75

Notes: PISA conducts internationally standardized and nationally representative tests to measure the performance of 15-year-olds in reading, maths and science. Performance is mapped on a scale with a mean of 500 test-score points and a standard deviation of 100 points across the OECD countries. The figure above shows the average PISA maths scores in 34 countries (Cyprus and Malta not shown due to missing data) in 2003 and 2012. The diagonal line separates the countries that improved their score (located above the line) from the ones that saw their score deteriorate (located below the line). A difference of 41 score points corresponds to approximately one year of schooling. Due to missing data for 2003, we use 2006 instead of 2003 data for the United Kingdom, Romania, Bulgaria, Estonia, Croatia, Lithuania, and Slovenia.
Source: OECD (2014a).

Notes: PISA conducts internationally standardized and nationally representative tests to measure the performance of 15-year-olds in reading, maths and science. Performance is mapped on a scale with a mean of 500 test-score points and a standard deviation of 100 points across the OECD countries. The figure above shows the average PISA reading scores in 29 countries (Croatia, Cyprus, Estonia, Lithuania, Malta, Slovenia and the United Kingdom not shown due to missing data) in 2000 and 2012. The diagonal line separates the countries that improved their score (located above the line) from the ones that saw their score deteriorate (located below the line). A difference of 41 score points corresponds to approximately one year of schooling. Due to missing data for 2000, we use 2003 instead of 2000 data for Luxembourg, the Netherlands and Slovak Republic.
Source: OECD (2014a).

Equity in education opportunities

In 2012, inequality in educational opportunity was low in the Netherlands, the UK and Switzerland, but high in France and Belgium.

Education does not only provide the skills needed to participate in society, it also aims to promote equal opportunity (Van de Werfhorst and Mijs, 2010). It is not easy to measure the extent to which education offers equal opportunity in the studied countries. According to the OECD (2014b), equal opportunity does not suggest that all students should perform equally, but rather that students’ socioeconomic status should have little or no impact on their performance. In line with this definition, the OECD has created an indicator of inequality of opportunity by assessing the extent to which differences in maths performance are explained by differences in the socioeconomic status of students’ parents. Figure 2 shows that in 2012, average inequality figures were the lowest for Korea and Japan, followed by Northern Europe and Northern America. By contrast, inequality was higher in the rest of Europe and in Oceania. However, these regional averages mask considerable differences within regions. This occurs most notably in Central and Eastern Europe, where inequality was very low in Estonia (not shown due to missing data for 2003) and very high in Slovakia and Hungary. Western European countries also showed considerable differences, with inequality being low in the Netherlands, the United Kingdom (not shown due to missing data for 2003) and Switzerland, but high in France and Belgium. Germany, the Netherlands and Switzerland achieved the largest declines in inequality between 2003 and 2012, while Spain, Latvia and France witnessed the largest increases.

Figure 2
Inequality based on parental socioeconomic status in 2003 and 2012
Country 2003 2012
Country 2003 2012
Austria 15.06 15.82
Belgium 23.00 23.00
France 20.24 22.47
Germany 23.83 16.91
Ireland 15.68 14.61
Luxembourg 16.55 18.28
Netherlands 18.31 11.51
Switzerland 18.01 12.85
Denmark 17.33 16.49
Finland 10.50 9.35
Norway 12.07 7.37
Sweden 14.28 10.62
Greece 15.96 15.48
Italy 12.30 10.11
Portugal 18.53 19.65
Spain 12.64 15.80
Czech Republic 18.50 16.16
Hungary 25.71 23.06
Latvia 11.87 14.71
Poland 16.46 16.62
Slovak Republic 23.62 24.62
Australia 14.00 12.35
New Zealand 16.55 18.37
Canada 10.20 9.42
United States 19.04 14.81
Japan 11.78 9.81
Korea 14.52 10.09

Notes: The figure above shows inequality in educational opportunity in 27 countries (Bulgaria, Croatia, Cyprus, Estonia, Lithuania, Malta, Romania, Slovenia, and the United Kingdom are not shown due to missing data) in 2003 and 2012. The diagonal line separates the countries that improved their score (located below the line) from the ones that saw their score deteriorate (located above the line). Inequality is measured as the R-squared of country-specific bivariate regressions of a student’s maths score on his/her parental socioeconomic status (PISA).
Source: OECD (2014b).

Relating average performance to equal opportunity

To assess how equal opportunity in education is related to average performance, Figure 3 plots inequality against the average maths performance of countries in 2012. It seems that countries with better average performance also do better in providing equal opportunity. However, the relationship is fairly weak and some countries, such as France, the United Kingdom, Belgium and Finland, show distinctly different levels of inequality, while their average performance levels are similar. These findings are consistent with the notion that policies aimed at improving educational performance only go some way towards achieving equal opportunity for students from disadvantaged social backgrounds.

Figure 3
Inequality based on socioeconomic status plotted against average maths performance – 2012 data
Country % of variation in performance explained by socioeconomic status mean PISA maths score
Country % of variation in performance explained by socioeconomic status Mean PISA maths score
Austria 15.82 505.54
Belgium 19.63 514.53
France 22.47 494.98
Germany 16.91 513.53
Ireland 14.61 501.50
Luxembourg 18.28 489.85
Netherlands 11.51 522.97
Switzerland 12.85 530.93
United Kingdom 12.49 493.93
Denmark 16.49 500.03
Finland 9.35 518.75
Norway 7.37 489.37
Sweden 10.62 478.26
Greece 15.48 452.97
Italy 10.11 485.32
Portugal 19.65 487.06
Spain 15.80 484.32
Bulgaria 22.30 438.74
Croatia 11.99 471.13
Czech Republic 16.16 498.96
Estonia 8.65 520.55
Hungary 23.06 477.04
Latvia 14.71 490.57
Lithuania 13.77 478.82
Poland 16.62 517.50
Romania 19.28 444.55
Slovak Republic 24.62 481.64
Slovenia 15.56 501.13
Australia 12.35 504.15
New Zealand 18.37 499.75
Canada 9.42 518.07
United States 14.81 481.37
Japan 9.81 536.41
Korea 10.09 553.77

Notes: The figure above shows average performance and inequality in educational opportunity in 34 countries (Cyprus and Malta are not shown due to missing data) in 2012. The dashed line represents the regression line of a bivariate regression of the percentage of variation explained by socioeconomic status on the mean maths score.
Source: OECD (2014b).

Performance in maths and reading is explained by …

The previous section dealt with the large differences we have observed between countries in the maths and reading performance of their 15-year-old children. Which factors explain these differences in cognitive skills? Is it mostly differences in the individual characteristics of children or their family background that matter? Are differences in school inputs, such as the quality of teachers or instructional resources, important? What is the role of institutional factors, such as the degree of school autonomy? The body of academic literature includes a large number of studies that address these questions.[1]

… student and family background, …

The results of our literature review indicate that student and family background are important determinants of the performance of 15-year-old children.

… the quality of teachers and instructional material …

Evidence on the effects of school inputs is mixed. While the quality of teachers and instructional resources really seems to matter, quantitative measures such as expenditure per student and class size are less relevant. This suggests that the education authorities should invest in talented teachers and first-rate instructional resources rather than in smaller classes.

… and institutional features of schools and the education system.

Three institutional features are important in explaining performance: accountability, school autonomy and the degree to which state schools face competition from private schools.

Institutional features of schools and education systems also seem to be important in explaining performance. Three distinct characteristics are positively associated with the test scores of 15-year-olds: (a) accountability; (b) the autonomy of schools and (c) the degree to which state schools face competition from private schools.

For policymakers wishing to boost the cognitive skills of 15-year-olds, we provide a brief summary of the recommendations found in the literature:

Ensure that an authority other than the school itself is responsible for the content of examinations. This makes it possible to compare the performance of children across classes and schools. Consequently, it facilitates the monitoring of the performance, not only of children, but also of teachers and schools.

Ensure that teaching activity is monitored by school staff or external inspectors.

Ensure that schools use assessments of children’s achievements to compare their performance to other institutions at the district or national level.

Ensure that schools use assessments of children’s achievement to determine a child’s promotion to the next grade.

Grant schools autonomy in decisions concerning processes and teaching staff.[2] This can improve pupils’ performance because schools have easy access to local information. The measure is particularly important if accountability mechanisms are in place, such as the external responsibility for exam content mentioned above. However, greater school autonomy is not recommended for budget setting and establishing educational programmes since it could also prompt opportunistic behaviour (Hanushek and Woessmann 2014).

Ensure that a high proportion of schools are privately operated while a large share of school funding is public. According to Hanushek and Woessmann (2014), public funding may be an important precondition for the competitive effects of privately operated schools to take effect. This is because low-income families may otherwise not have the financial means to send their children to a private school.

The following tables document for each country: the presence of the accountability mechanisms mentioned above; the level of school autonomy with regard to setting the budget, establishing the starting salaries of teachers, determining the content of courses and hiring teachers; and the percentage of children in privately operated schools and the percentage of school funding provided by national governments.

Figure 4
Institutional features of school systems (2012)
Country External exams Assessments used for retention/promotion Assessments used to compare schools Monitoring of lessons by principal Monitoring of lessons by external inspectors Formulating budget Establishing starting salaries Determining course content Hiring teachers Privately operated schools Government funding
External exams Assessments used for retention/promotion Assessments used to compare schools Monitoring of lessons by principal Monitoring of lessons by external inspectors Formulating budget Establishing starting salaries Determining course content Hiring teachers Privately operated schools Government funding
Austria 0 94 28 74 29 30 7 74 54 9
Belgium 0 96 23 65 48 79 4 75 90 87
France 96 62 12 73 80 3 83 30 17
Germany 35 96 43 67 22 15 2 72 64 5 95
Ireland 100 62 77 13 48 76 4 72 87 56 90
Luxembourg 100 94 74 48 6 82 21 70 70 15 97
Netherlands 100 98 70 87 42 100 88 98 100 66 96
Switzerland 0 86 41 83 29 72 23 63 98 6 94
United Kingdom 100 69 96 97 68 91 80 97 100 44 92
Denmark 100 10 55 64 17 92 30 92 100 23 91
Finland 100 93 46 31 2 70 15 76 86 3 100
Norway 100 1 68 48 11 82 12 66 96 2 100
Sweden 0 43 90 80 27 89 64 80 100 14 100
Cyprus 16 78
Greece 0 98 17 8 21 79 5 5 6 2 86
Italy 100 87 65 17 1 23 7 88 14 5 52
Malta
Portugal 0 98 85 60 4 82 9 34 76 10 82
Spain 0 95 44 10 15 85 6 57 34 32 88
Bulgaria 100 65 86 97 49 67 82 40 99 1 97
Croatia 100 88 66 93 34 75 2 55 99 2 94
Czech Republic 100 79 58 98 33 91 92 100 100 8 94
Estonia 100 82 65 90 8 89 26 98 100 2 97
Hungary 100 69 78 97 13 77 48 86 100 16 93
Latvia 100 97 92 100 41 95 56 62 100 2 95
Lithuania 100 85 61 98 38 79 78 90 100 1 97
Poland 100 98 58 94 16 48 19 100 98 3 96
Romania 78 70 68 73 58 52 34 69 67 1 93
Slovak Republic 100 93 64 98 27 77 58 97 100 9 96
Slovenia 100 93 59 94 5 75 22 88 100 2 95
Australia 81 63 56 70 11 85 20 84 83 39 73
New Zealand 100 77 93 97 32 100 18 99 100 5 81
Canada 51 95 82 82 21 63 18 58 86 8 88
United States 7 57 94 100 42 76 54 74 98 5 92
Japan 100 90 17 81 26 46 30 96 33 30 74
Korea 100 56 70 96 68 48 9 97 51 47 53

Notes: The variable ‘External exams’ indicates the extent to which standards-based external examinations for students in secondary education exist in the system.
Sources: OECD (2010), OECD (2012), OECD (2013).

Notes: The variable ‘Assessments used for retention/promotion’ indicates the percentage of students in schools whose principal reported that assessments of 15-year-old students are used to decide on students’ retention or promotion.
Sources: OECD (2010), OECD (2012), OECD (2013).

Notes: The variable ‘Assessments used to compare schools’ indicates the percentage of students in schools whose principal reported that assessments of 15-year-old students are used to compare schools to district or national performance.
Sources: OECD (2010), OECD (2012), OECD (2013).

Notes: The variable ‘Monitoring of lessons by principal’ indicates the percentage of students in schools whose principal reported that the principal or senior staff have monitored maths teachers by observing lessons.
Sources: OECD (2010), OECD (2012), OECD (2013).

Notes: The variable ‘Monitoring of lessons by external inspectors’ indicates the percentage of students in schools whose principal reported that inspectors or other externals have monitored maths teachers by observing lessons.
Sources: OECD (2010), OECD (2012), OECD (2013).

Notes: The variable ‘Formulating budget’ indicates the percentage of students in schools where the principal reports that schools have a main/substantial responsibility for formulating the school budget.
Sources: OECD (2010), OECD (2012), OECD (2013).

Notes: The variable ‘Establishing starting salaries’ indicates the percentage of students in schools where the principal reports that schools have a main/substantial responsibility for establishing teachers’ starting salaries.
Sources: OECD (2010), OECD (2012), OECD (2013).

Notes: The variable ‘Determining course content’ indicates the percentage of students in schools where the principal reports that schools have a main/substantial responsibility for determining course content.
Sources: OECD (2010), OECD (2012), OECD (2013).

Notes: The variable ‘Hiring teachers’ indicates the percentage of students in schools where the principal reports that schools have a main/substantial responsibility for appointing teachers.
Sources: OECD (2010), OECD (2012), OECD (2013).

Notes: The variable ‘Privately operated schools’ indicates the percentage of students in privately operated schools (as opposed to publicly operated schools).
Sources: OECD (2010), OECD (2012), OECD (2013).

Notes: The variable ‘Government funding’ indicates the percentage of the total funding of students’ schools that comes from government sources.
Sources: OECD (2010), OECD (2012), OECD (2013).

References

Our review of this literature draws heavily on Hanushek and Woessmann (2010), Hanushek and Woessmann (2014), and Hanushek, Piopiunik and Wiederhold (2014).
Examples of process and staff decisions are those concerning purchase of supplies, budget allocation and hiring and paying teachers within a given budget.