Getting teachers who make a difference

Teachers matter …

One point of broad agreement in education is that teachers matter greatly. Students of certain teachers simply do better in a way that has a marked effect on social and economic outcomes. For example, a recent study drawing on data covering about 2.5 million US children found that, after correcting for other factors, pupils assigned to teachers identified as delivering better educational results “are more likely to attend college, attend higher-ranked colleges, earn higher salaries, live in higher [socio-economic status] neighbourhoods, and save more for retirement. They are also less likely to have children as teenagers.”[6] Professor Schwartz believes that “the single most important input variable [in education] is the quality of teaching.” However, teacher quality, notes William Ratteree – until recently, education sector specialist at the International Labour Organisation – "is a mix of factors which are difficult to pin down.”

Much of the research in this area has focused on what education systems can do to ensure that they find teachers who add value. Even here, though, says Professor Hanushek, “the rules tend to be country-specific.” McKinsey’s 2010 report, How the World’s Most Improved School Systems Keep Getting Better, argues that the best interventions even depend on the current state of the school system. In McKinsey’s view, systems currently marked by “fair” levels of performance should focus on teacher accountability, while “good” systems are likely to benefit more from enhancing the status of the teaching profession.

… But what matters for getting good teachers?

Despite such variation, a number of insights are broadly applicable. The first is that teacher pay has surprisingly little relevance to education results. In LCDB data, minimum and maximum instructor salaries at all education levels – measured as a percentage of average national income – have no long-term link to the inculcation of cognitive skills, as measured by standard international tests. Indeed, the only statistical correlation between pay and educational outcomes is a tendency of higher maximum salaries – as a percentage of the national average – at a number of teaching levels to lead to lower secondary school graduation results. A closer look at this counter-intuitive result reveals that – within the data set available – higher GDP countries do not pay teachers as high a percentage of the average wage as lower GDP ones. In other words, as economies grow, teacher salaries do so at a slower rate. Thus, the implicit correlation actually reveals again the link between higher GDP and certain better educational results.

Chart 4: Ratio of average teacher salary at primary, lower and upper secondary levels over the average gross wage, selected countries, 2010

Chart 4: Ratio of average teacher salary at primary, lower and upper secondary levels over the average gross wage, selected countries, 2010


Sources: Economist Intelligence Unit and OECD.

The lack of correlations in this area is consistent with much detailed research on the link between pay and results, which is typically found to be weak or non-existent.[7]  Performance-based pay is an exception: it does tend to lead toward better outcomes.[8]  On the other hand, in some cases high salaries without quality differentiation create problems. Mamadou Ndoye recalls that, when he was Minister of Basic Education in Senegal, the level of pay made it impossible to hire more teachers, so he had to engage in difficult negotiations to be allowed to bring in volunteers to help. Overall, in the words of Mr Cappon, “Teachers must be reasonably well paid, but this pales in comparison with other factors.”

Experts interviewed for this study repeatedly point to several of these other factors which are essential in promoting teacher quality:

  • Attracting the best people to the profession: Getting good teachers begins with recruiting talented individuals. Finland and South Korea – two perennially cited examples of education success and the top countries in our Index – obtain their annual teacher intake from the top 10% and 5% of graduating students respectively. The key to such success is the status in which teaching is held culturally. Here money can have some effect, not just as a simple inducement but as a signal of status. The South Korean government uses high levels of teacher pay in this way both to compensate for large class sizes and to indicate the importance it accords to the profession.
  • Providing the right training: The training of these new recruits has to be appropriate to the conditions in which they will work. This varies by country. The Finnish system, for example, benefits from teachers having graduate degrees. On the other hand, Nahas Angula, Prime Minister of Namibia, points out that his country’s policy of requiring all teachers to have an undergraduate degree may be driving up the cost of education when other training would suffice for primary grades. Teacher training also needs to be ongoing. This has a very practical reason – that no teacher’s college course will maintain complete relevance across decades of work – but also a demonstrative one. As Mr Cappon notes, “teachers need to be lifelong learners themselves. You can’t inculcate a love of learning unless you live it.” Effective professional development needs to address not just upgrading the knowledge of teachers – providing, for example, a better understanding of new technology and teaching strategies – but also allow them to advance along their career path into more senior positions where relevant.
  • Treating teachers like professionals: Consistent with the need to promote the status of teaching is its treatment as a profession. Mr Ratteree notes that “things like continual professional development and professional autonomy can be powerful incentives for better learning outcomes.” Mr Cappon agrees: “Teachers must be seen as professionals who exercise judgement, not just technicians.”
  • Implementing clear goals and effective oversight, and then letting teachers get on with it: Professors Hanushek and Woessmann both point to this combination of accountability and independence as consistently correlated with improved outcomes. Says the latter: “Education economists emphasise the need to think about incentives for people in the system to use resources efficiently. These are mostly framed by the surroundings of the education system, the accountability system and whether schools can act autonomously. There is clear evidence of strong relations between these and improved outputs.” Professor Schleicher agrees. High-performing school systems, he says, combine demanding standards, low tolerance of failure, and clear articulation of expectations with “a lot of professional responsibility within a collaborative work organization at the front line,” for both teachers and schools.

None of these on their own is enough. Instead, they form an overlapping, and mutually supporting, set of strategies to provide the high-quality teachers that are so important for education and to use them in the most effective ways.[9]

[6]Raj Chetty, John N. Friedman, Jonah E. Rockoff, The Long-term Impacts of Teachers: Teacher Value-added and Student Outcomes in Adulthood, National Bureau of Economic Research Working Paper 17699, December 2011,

[7]One notable exception is P. Dolton and O. D. Marcenaro-Gutierrez, “If you pay peanuts do you get monkeys? A cross-country analysis of teacher pay and pupil performance”, Economic Policy (2011) 26: 5–55.

[8]David N. Figlio and Lawrence Kenny, Individual Teacher Incentives and Student Performance, National Bureau of Economic Research Working Paper 12627, October 2006,; Ludger Woessmann, “Cross-Country Evidence on Teacher Performance Pay”, Forschungsinstitut zur Zukunft der Arbeit Discussion Paper 5101, July 2010.

[9]For a similar discussion of the key success factors in teacher development see Michael Barber and Mona Mourshed, How the World’s Best Performing School Systems Come Out on Top, McKinsey and Co., 2007, pp 15-23.


[previous page]  [next page]

Download the report

Download PDF

The Learning Curve videos

View videos

The Learning Curve case studies

Read case studies

Suggestion box

Please contribute any comments, thoughts, feedback or suggestions

Make a suggestion