New study: strong leadership practices in schools drive improved performance Dr Dumisani Hompashe, Research Associate at the Institute of Social and Economic Research at Rhodes University

Rhodes>Latest News

Research Associate at the Institute of Social and Economic Research at Rhodes University, Dr Dumisani Hompashe.
Research Associate at the Institute of Social and Economic Research at Rhodes University, Dr Dumisani Hompashe.

At the Basic Education Sector Lekgotla in 2019, President Cyril Ramaphosa said, “We need a functional leadership and accountability system that involves the school management teams, the teachers, the representative councils of learners, the school governing bodies and the community at large. With visionary leadership, we will succeed and our education system will achieve its intended objectives.” Implicit in the President’s words is the key role that should be played by school principals in leading their schools.

In the past, principals were regarded as managers and administrators who were preoccupied with issues of school discipline and financial management. However, according to the National Development Plan (NDP) school principals should also be involved in curriculum leadership, the core business of the school. Principals must lead their schools to ensure that quality teaching and learning takes place and instructional time is always protected.

Over the past two decades, the South African government has shown commitment to strengthen accountability of principals and schools. Over the same period there has been noteworthy increase in academic performance by South African students as shown by several international assessments, including the Trends in Mathematics and Science Study (TiMSS), the Progress in International Reading Literacy Study (PIRLS) and the Southern and Eastern Africa Consortium for Monitoring Educational Quality (SACMEQ), with most of the gains among the disadvantaged students who form the lower end of the distribution.  

Quality school leadership and management is an important determinant of quality teaching and learning, especially in low-resourced schools.   South Africa has continued to see improvements in mathematics scores in the TiMSS, with the strongest gains observed between 2015 and 2019 among the most disadvantaged students. To sustain these improvements, it is crucial to understand why this upward trajectory in learning outcomes is occurring.  The improvements in TiMMS scores are also observed in most of the participating countries.

The hypothesis driving a recent research report is that quality school leadership as implemented by a principal will lead to students’ improved learning outcomes. The study analyses the improvement between 2015 and 2019 Grade 9 Mathematics scores as reflected in the TiMSS results to explore the extent to which the performance gap is related to instructional leadership practices.  The study report addresses the association between instructional leadership and student learning outcomes in South Africa, and explores how the increase in South African Grade 9 Mathematics students’ learning outcomes is associated with changes in instructional leadership.

Providing a holistic and policy-focused South African perspective on the relationship between instructional leadership and educational outcomes, the study results support the ongoing implementation of key policies, shedding light on actions that could sustain these improvements.  Noteworthy policies include Section 16A of SASA, which makes provision for the principal to prepare an annual report in respect of the academic performance of the school, and for the principals of public schools that have been identified by the HOD to prepare an academic improvement plan at the beginning of each year. Section 58B of SASA provides for the HOD to annually identify any public school that is underperforming based on the report contemplated in Section 16A (1) (b) and from other relevant reports. The Whole School Evaluation policy provides for external moderation, on a sampling basis, of the results of self-evaluation carried out by the schools. The policy evaluates the effectiveness of a school in terms of the national goals.

 In 2015, about 19% of South African Grade 9 mathematics students achieved a score of between 400 and 475 points, a low level benchmark. In 2019, this increased to 25%, an increase of 6 percentage points. A similar pattern occurred in higher TIMSS benchmarks, as the proportion of students who achieved a score between 475 and 550 points, an intermediate level benchmark, increased from 11% in 2015 to 16% in 2019. There has also been a notable decline in the proportion of students who achieved a score below 400 points from 2015 to 2019. In 2015, 67% of students scored below 400 points, while in 2019, 54% of students scored below the lowest TIMSS benchmark.

Although SA performed poorly overall (relative to other countries) in the TIMSS assessment, the country showed noteworthy improvement at both the lower end and top end of achievement distribution. The ideal is a situation where there are minimal quality differences between schools (Reddy et al., 2016). Between school-inequality is calculated by using the intraclass correlation coefficient (ICC), which indicates the proportion of the total variance in the outcome variable that lies systematically between and within schools.

In the TIMSS 2015 and 2019 data, the share of overall inequality attributable to between-school differences in South Africa emerges as extremely high for a developing country, at least compared to Iran and Egypt, indicating that there is considerable inequality between schools in South Africa. South Africa’s between-school inequality with respect to Grade 9 mathematics, remained persistently high between 2015 and 2019 at 0.51. This means that school-level variables, and particularly instructional leadership variables describing what occurs in schools, deserve special attention. After excluding the top quintile of schools, South Africa’s proportion of between-school inequality emerges as typically low for a developing country.

The study is guided by this theoretical instructional leadership framework. Schools that practice instructional leadership, with an orderly and supportive environment, with principals who have adequate experience and training, will have effective teachers that are guided by competent instructional leaders. Such school settings will ultimately lead to improved learning in schools.

Variables of interest include the emphasis on academic success scales (principal-reported), the School safety and discipline scale (principal-reported) and the Principals’ educational qualifications. Explanatory variables include student gender, age, whether the language of instruction is spoken at home, the presence of internet, Socioeconomic status (SES), whether the school is located in an advantaged area and whether the school has a library.

To analyse the relationship between instructional leadership and the increase in mathematics students’ learning as reflected in TIMSS 2015 to 2019, the Oaxaca (Wahaka)-Blinder (OB) decomposition technique was used, separating the incremental gap in test scores into “explained” and “unexplained” components.  The “explained” component accounts for differences in observable characteristics of students, for example, some students are taught in schools with high emphasis in academic success, while other are taught in schools with low emphasis in academic success. The “unexplained” component means how such observable characteristics are better used.

The research reflects that school leadership variables have an especially large association with achievement in both years.  School emphasis on academic success and school discipline have a stronger association with achievement in 2015 than in 2019. Living in an urban area had a positive and significant association with better mathematics performance in both 2015 and 2019.

The purpose of decomposition was to investigate what changes may have occurred over time that would explain the increase in TIMSS mathematics scores between 2015 and 2019.  Differences in school location and student age each account for more than half the explained part of the performance gap, whereas differences in principals’ qualifications and socio-economic status do not seem to matter much. Differences in school emphasis on academic success contribute about a fifth of the explained part of the achievement gap.  However, the difference in discipline and safety decreases the explained part of the performance gap by more than 20%, whereas the difference in the presence of internet at home decreases the explained part of the gap by more than a third.

The impact on performance gap of studying in schools with high emphasis on academic success decreased from 2015 by more than a fifth, to a relatively lower impact in 2019 (-4.19), whereas the impact of studying in schools with discipline decreased by more than more than a third, to a lower impact in 2019 (-6.19). However, the impact on achievement gap of being a student of appropriate age and female student increased by a quarter and an eighth respectively from 2015, to a relatively large impact in 2019.

Inputs such as school emphasis on academic success, discipline and safety, school located in advantaged area, students of appropriate age and students that are female seem to be important contributors to the performance gap. This means that there are important returns to some of policy interventions that took place years ago.

Areas of delivery success highlighted in the Action Plan to 2024 report reflect improved implementation of management and leadership in schools. However, high between-school inequality means that the impact of these areas of delivery success are not equitably distributed in the entire schooling system. Policy should also focus on lower-quintile schools regarding instructional leadership.


Dr Dumisani Hompashe is a Research Associate at the Institute of Social and Economic Research at Rhodes University. A senior lecturer at the University of Fort Hare, he further serves as Research Associate at the Research for Socio-Economic Policy (ReSEP) and the Institute of Social and Economic Research in Stellenbosch University. His research interests include the quality of education and healthcare, inequality of education outcomes, and development economics. Initially qualifying as a secondary school teacher, Dumisani has since gained a BA in Economics and English (UNISA); BA (Hons) in Financial Economics and an MCom in Financial Markets (Rhodes University)and PhD (Economics) from Stellenbosch University.


Dr Dumisani Hompashe, Research Associate at the ISER, Rhodes University presented his new study on 25 August.  The session was chaired by ISER Researcher Dr Reesha Kara with respondent Dr Rethabile Mawela, Alan Macintosh Research Fellow - Institute for the Study of the Englishes in Africa (ISEA), Rhodes University. 
The study is the result of a collaborative research effort after the HSRC asked the ISER and the University of Fort Hare to analyse the results of the TIMMS survey.

We thank these institutions for their collaboration and support.

The full session can be watched here: