Are smaller School and Class size better for students?: A Review of Global Evidence

Introduction:

As many progressive reforms through the decades of the 60s and 70s allowed minorities and marginalized sections to access the existing school facilities, the average school size in the United States became large (Harvey, 1971). The rising demand for education, along with declining learning levels in schools, created the perception that the US was failing its students (A Nation At Risk, 1983). Policy-makers believed that it was because of large school-sizes, teachers were not able to teach, and students were not able to learn. The school-size reduction was the only way through which the American education system could be revamped. Since then, numerous studies have looked at the impact of school size reduction on the various aspects of schooling. Despite a large number of studies on the topic, the question of school size reduction has remained unsettled.

 Informed by these debates in the western countries where the school-size reduction was thought as a solution to many shortcomings, policy-makers in India decided to rationalize the school resources by fixing teacher-student ratio to 30 for primary classes and 35 in upper primary classes. Explicit norms and standards for the establishment of a school were promulgated (pg. 13, RTE-2009). By having these provisions in RTE, the policy-makers believed that this would not only improve the overall teacher shortages in government schools but would also result in the absolute learning gains among the students studying the government schooling system. As the elementary education has become the fundamental right, various civil society organizations mounted on state governments to fill up the positions of teachers to maintain the teacher-student ratio mandated in the right to education (RTE) Act, passed in 2009. With an intent to provide free and compulsory education to the children in the age group of 6-14 years, the policy mandated that the state (or any appropriate authority as defined by the law) should provide a school within the one-kilometre periphery of a habitation. Since the policy focused on input-based measures, most of the investment made under Sarva Shiksha Abhiyaan (SSA) went into providing infrastructure for the school. While the expectation was that the increased access to schooling would not only improve enrolment in these schools but also impact the learning levels among these students, the outcomes on students' achievement have been quite contradictory. The preference for private schools among the parents led to an absolute decrease in the number of students enrolled in government schools. Students' achievement in the same period also remained flat, where only a quarter of students enrolled in grade 3rd could perform grade-appropriate tasks (ASER, 2018). The falling enrolment in the government schools, accompanied by static students' outcomes led to a situation where many state governments started further rationalizing the resources (both the physical and human resources), manifested by various policy measures such as school consolidation, merger, etc. taken at the state level. Under these policies, state governments have sought to merge smaller schools (enrolment wise) to the nearest bigger schools.

 The review of literature:

There have been, over the years, numerous studies looking at school size and student achievements. During the debates over the report' Nation at the Risk', various researchers tried to look at the impact of school size on the students' achievement in a US context. The first such study, using the data from 1021 New York State elementary schools, found that the school size had a negative correlation with reading and mathematics achievements, controlling for the socio-economic status (SES) of students (Wendling & Cohen, 1981). In the very next year, a survey study concluded that the schools with 100 students or less had a significant impact on the satisfaction, attendance, and other non-cognitive outcomes of students (Lindsay, 1982). This result was again contradicted by a study conducted among the 287 elementary schools, which found that the school-size had little or no impact on the performances of students (Eberts, Keyhole, & Stone, 1984). However, this study did not consider the locations of the schools. The efforts were also made to analyze the impact of school-size on the students' drop-out rates (Pittman & Haughwout, 1987), though no direct link between these two variables could be established. However, results from a study by Toenjes (1989) showed school-size being a strong predictor in determining the drop-out risk.

 The black and Hispanic students seemed to drop-out more when school-size was large. This established also dispelled the notion that the drop-out among students a primarily a non-white phenomenon. Significant progress was made towards analyzing the impact of school-size on school performance when Friedkin and Necochea used the data from the 1983-84 California assessment to see whether school-size and SES had any effect on the school performance. While this could establish the link between the school-size and the socio-economic status of students, the authors also indicated a negative effect of school-size on school performance. (Friedkin & Necochea, 1988). Large school sizes negatively impacted the performance of students who came from a poor socio-economic background.

 With being cited as a significant predictor for students' success, the early decade of the 90s saw many studies contributing to understanding school-size linkages. In a meta-analysis of around 103 studies, Kathleen Cotton (1996) argued that there was no optimum school size which could be established from these studies. However, the small school size had a positive impact on attendance and drop-out rates among the students. The large school sizes seemed to impact the students belonging to low socio-economic status adversely. The evidence on students' achievement remained mixed in this meta-analysis. As several studies hinted towards the socio-economic status and the school-size being related, Craig Howley (1996) attempted out how exactly these two variables were linked. Using and SES as independent variables, this study analyzed the students in the 3rd, 6th, 9th, and 11th grade, and showed that the large school sizes were correlated with low achievement of students with poor socio-economic status, while the students from high socio-economic status seemed to be doing well. In a similar vein, a study by Craig Howley and Robert Bickel (1999) across states of Ohio, Georgia, Texas, and Montana showed that there was a correlation between SES status and students' achievement and smaller school sizes did improve the outcomes of students' belonging to poor socio-economic status.

However, the source of the interaction between these two factors remained confounded, and the study did not provide a definite answer in that case. Some other studies that have analyzed data from Project STAR (Student-Teacher Achievement Ratio) in Tennessee have also indicated achievement improvements for students in small-size schools compared to students in regular-size schools (Krueger, 1999). Although some researchers have contested the positive effects small schools can have on student achievement using data from Project STAR (e.g., Hanushek, 1999 Hanushek, E. A. (1999). Nye, Hedges, and Konstantopoulos (2000a), too found no evidence for the differential effects of small-size schools on students from poor socio-economic background. Adding to the literature on impact, Lindsay Page, Carolyn Layzer, Jennifer Schimmenti, Lawrence Bernstien, and Leslie Horst (2002) reviewed 30 studies to conclude that the small school sizes were positively correlated with the achievements of students from the poor socio-economic background.

 The review also showed that the schools with small-class size had lower drop-out rates, higher graduation rates, and higher attendance. Additionally, schools with small school sizes were as cost-effective as large schools. However, a review of the literature conducted by John Slate and Craig Jones (2005) on the effects of school size on students' achievement remained inconclusive. Similar results were repeated in a review conducted by The Dublin City University School (2007) where 60 studies were examined to determine the optimum size of class for both the costs and students' outcomes. Students' in small-class size schools seemed to be doing marginally better than large-size schools, though large-size schools might favour bright students as they had much better infrastructure and human resources to support them. A study by Jacob Werblow and Luke Duesbery (2009) supported the previous findings of small-school size positively impacting drop-out rates and attendance. However, the results on students' outcomes remained confounded. More recent studies such as Li and Konstantopoulos (2017), based on TIMSS (Trends in International Mathematics and Science Study) sample from 14 European countries, too have shown the school size-reduction having no impact on students' achievements. Even among the low socio-economic category, school-size did not seem to have an effect. Holmund and Bohlmark (2017) studied the reorganization of Swedish Middle Schools by exploited a policy change in brought out by the government and found no short and long-term difference in students' achievements. Thorsen's (2017) study of consolidation of 76 rural lower secondary schools in Norway too show no significant impact on the test scores. Large school sizes did not seem to have any effect on the learning outcomes of students.

Similarly, many developing countries have implemented school-size reduction programs, and results have been decisive in most of these cases. In a study on the school construction program in Indonesia, Duflo (2001) found that a new school per 1,000 children led to an increase of between 0.12 and 0.19 years of education. Furthermore, increasing the coverage of schools in rural Mozambique led to an improvement in primary school enrolment (Handa, 2002). In contrast, using a difference-in-difference approach with propensity scores, Liu et al. (2010) found that in rural China, the school merger policies of 2002 did not make any difference in student's performance evaluated in 2006. The effects of school merger policies are also mediated by what (Mo, D. et al., 2012) have called 'resource effect (an effect that appears to be associated with the better facilities and higher quality of teachers)". Students who moved to county schools benefitted no matter where they started (village or urban areas). In Argentina, adding pre-schooling to the existing schools and then raising the school size had a positive impact on the learning levels of students, and this resulted in an improvement of 0.23 standard deviation from the previous scores in the formal primary school exams (Berlinski, Galiani & Gertler, 2009). Increasing the size of schools in rural Afganistan not only increased the enrolment by 52 percentage but also improved the test scores by 0.65 standard deviations (Burde & Linden, 2013). In a study conducted in Burkina Faso, too, produced the similar results with enrolment increasing by 19 percentage points and test scores by 0.41 standard deviation due to the increase in the school size (Kazianga et al., 2013). School consolidation/reorganization policies might also have a negative effect on the access and participation of marginalized sections, especially in those cases where the size and nature of habitation are correlated with schooling outcomes (Kochar, 2008).

 In a different context, a survey of rural households in China in 2018 found that school closure reduced the number of years of schooling completed by girls above the age of 15, possibly due to the higher sensitivity of girls' enrolment towards distance and that of boys towards the quality of schools (Hannum et al., 2018). Despite the scores of studies on the impact of the students' achievements, the direction of evidence remains inconclusive. Overall, evidence suggests a marginal improvement in reading scores due to small school-size, but it is balanced by the negative, though non-significant impact on mathematics scores, due to school-size reduction. Similarly, the studies have been inconclusive while discussing any optimum school-size for schools. While many top-performing schooling systems (Finland, Singapore, Taiwan, Hong Kong, etc.) have mandated some limitation to their school-size, there is no one which might fit all schooling systems. There are scores of other factors whose interplay with school size impacts the students' achievement.

The relevance of these studies does not get reduced by the fact that they have been generally conducted in a western context (only a few in developing countries). We, too, have been feeling similar pressures in managing and administering our elementary education. While the dominant narrative has been to fix a limit for the school-size, oppositional voices have also been gaining momentum in the recent past. Since the publication of the Ministry of Human Resource Development (MHRD) guidelines, which stated one of its objectives as to "rationalize of small schools across the States for better efficiency", the school consolidation/reorganization processes have gained momentum. NITI Aayog, which works as an external agency, started signing a memorandum of understanding with state governments to restructure their education systems under a program called SATH (Sustainable Action for Transforming Human Capital). Niti Aayog engaged several external funding and grant-in-aid agencies such as BCG and Piramal Foundation for Education Leadership (PFEL) as knowledge partners to closely work with the states towards removing the inefficiencies from their schooling systems[1]. In their objective to control these cost-inefficiencies, state governments are in the process of merging or consolidating small schools into a large unit. The primary argument for this merger/consolidation process has been that it rationalizes both the physical and human resources in a high resource constraint environment. Whether one may like or dislike these policies, there is no denying the fact that school mergers and consolidations have become part of the dominant narrative. In such a scenario, analyzing and evaluating school merger/consolidation policies become obligatory on researchers, administrators, and policy-makers.

Is big the better: The politics of school mergers/consolidations in India

According to an estimate based on DISE data, there were around 1 lakh government schools which had 20 or fewer students enrolled in 2014-15. This number increases to 3.7 lakh if we consider the number of schools with 50 or fewer students enrolled. The cost of operating these schools is enormous, where the per-child expense comes around a staggering Rs 40,800 per year (as estimated by Kingdon in her paper). However, most of this expense goes into paying the salaries to the teachers (Kingdon, 2017). Also, those who support the policies of school mergers, consolidation, and closure, cite ASER surveys which have been publishing the disappointing figures for learning levels in rural India. The last decade has also seen the number of students going to government schools falling in an absolute sense, and at the same time, the number of students going to private schools has increased. The increasing cost to the exchequer has led many to argue that the governments should do away with the 'tiny' schools. Instead, composite schools can be created through integration and consolidation across levels. The supporters argue that the merger would bring efficiency to the system as composite schools would have better facilities, trained and qualified teachers, broader extracurricular activities, and diverse social experience. Despite school consolidation or rationalization policies getting the upper hand in the last few years, there have been no studies in India which have looked at the impact of these policies. Now it is the right time to evaluate the effect of these policies when the process of school rationalization has been formalized through a circular by the Ministry of Human Resource Development for better efficiency (MHRD, 2017). So, while a better academic environment might have been offered as a reason for school consolidations, but the primary drive inherent in these policies has been the economic logic of reducing inefficiencies. The draft National Education Policy (NEP) furthers this discourse by suggesting to merge small schools to form a school complex, which would provide pre-primary to secondary school education. While various media reports have covered the school merger and consolidation in detail, there has been no report which may put these numbers in perspective.

 The guideline issued by MHRD for school rationalization makes enrolment a criterion for school merger. According to the policy guidelines, states could have their criterion for school consolidation. For instance, Rajasthan decided to merger those schools first where there were zero enrolments shown in the DISE data. Soon, any school which had enrolment less than 30 was brought under the school rationalization process. Andhra Pradesh, on the other hand, prepared a list of schools where enrolment was less than 40. Also, governments do not generally provide information on consolidated schools. For instance, the Rajasthan government closed down around 20,000 schools, but no aggregated list of consolidated schools was provided.

 Additionally, these schools were not removed from U-DISE. MHRD report puts the number of closed schools to 7473 in 2013-14, 18819 in 2014-15, and 5517 in 2015-16. No official data on closed or merged schools have been released since.

 Evaluating the impact of school restructuring policies in India

Unlike the United States and other developed countries, where data sources are generally credible, and administrative data are made available regularly. In India, we have a severe lack of reliable administrative data sources. Self-reported data sources like DISE has a problem. If one wants to conduct a school consolidation study, she would require to have longitudinal data on the subjects, and also credible data on both the school and household-related factors. Lack of these datasets would make it difficult for individuals to run regression studies. The countries like India also lack experiments like the STAR experiment, where a database was created on the students who were enrolled in either small or large school sizes. As these experiments are costly, it puts individually-run trials out of the question. Even two recent studies, which were conducted in India, either used NSS survey data or collected their data by working with a local organization. Without denying the possibilities of quantitative studies on the topic, I would prefer to approach this phenomenon from a qualitative point of view.

 Before getting into the issues of 'learning outcomes', I would like first to study how these 'consolidation' or 'reorganization' policies have caused upheavals in the lives of children. How in states like Jharkhand, Chattisgarh these policies deprived children of one proper meal they used to get in their schools. Also, while these policies emphasized on 'consent' part and mandated that the consolidation/reorganization decisions would be taken only after consulting the school management committee, it would be interesting to study how exactly these processed were carried out. Studying children who have been left out during this process is as essential as studying any effect of them getting enrolled in a 'consolidated' school. Some preliminary questions such as how do these 'consolidated/merged' schools offer better governance, student wellbeing, and community satisfaction should first be answered before going into the issues of student outcomes. Any answer to these questions would require an in-depth investigation into this phenomenon. Interpretive, critical, and/or post-cultural approaches would be more suited to such investigations. Strategies such as interviews, focus group discussions (FGDs), and field notes, etc. would be deployed to understand the phenomena. An attempt would also be made to locate this phenomenon within local cultures in relation to national and international epistemologies and ontological shifts.

References

Berlinski, S., Galiani, S., & Gertler, P. (2009). The effect of pre-primary education on primary school performance. Journal of public Economics, 93(1-2), 219-234.

Burde, D., & Linden, L. L. (2013). Bringing education to Afghan girls: A randomized controlled trial of village-based schools. American Economic Journal: Applied Economics, 5(3), 27-40.

Cotton, K. (1996). School size, school climate, and student performance.

Duflo, E. (2001). Schooling and labor market consequences of school construction in Indonesia: Evidence from an unusual policy experiment. American economic review, 91(4), 795-813.

Friedkin, N. E., & Necochea, J. (1988). School system size and performance: A contingency perspective. Educational Evaluation and Policy Analysis, 10(3), 237-249.

Goodwin, D., Page, L., Layzer, C., Schimmenti, J., & Hadley, M. A. (2002). 55 Wheeler Street Cambridge, MA 02138.

Guidelines for Rationalization of Small Schools across States for Better Efficiency, MHRD, 2017 (https://mhrd.gov.in/sites/upload_files/mhrd/files/Guidelines%20for%20Rationalization.pdf)

Handa, S. (2002). Raising primary school enrolment in developing countries: The relative importance of supply and demand. Journal of development Economics, 69(1), 103-128.

Hannum, E., Liu, X., & Wang, F. (2018). Estimating the Effects of Educational System Contraction: The Case of China's Rural School Closure Initiative. SSRN. doi: 10.2139/ssrn.3140132

Hanushek, E. A., Mayer, S. E., & Peterson, P. (1999). The evidence on class size. Earning and learning: How schools matter, 131-168.

Holmlund, H., & B?hlmark, A. (2017). Does grade configuration matter for school performance? Short-and long-run effects of school reorganization (No. 2017: 6). Working Paper.

Howley, C. B., & Bickel, R. (1999). The Matthew Project: National Report.

Kazianga, H., Levy, D., Linden, L. L., & Sloan, M. (2013). The effects of" girl-friendly" schools: Evidence from the BRIGHT school construction program in Burkina Faso. American Economic Journal: Applied Economics, 5(3), 41-62.

Kingdon, G. G. (2017). The private schooling phenomenon in India: A review.

Kochar, A. (2008). Can schooling policies affect schooling inequality? An empirical evaluation of school location policies in India. Suman Bery Barry Bosworth Arvind Panagariya, 53.

Krueger, A. B. (2003). Economic considerations and class size. The Economic Journal, 113(485), F34-F63.

Lamdin, D. J. (1995). Testing for the effect of school size on student achievement within a school district. Education Economics, 3(1), 33-42.

Lindsay, P. (1984). High school size, participation in activities, and young adult social participation: Some enduring effects of schooling. Educational Evaluation and Policy Analysis, 6(1), 73-83.

Li, W., & Konstantopoulos, S. (2017). Does school size-reduction close the achievement gap? Evidence from TIMSS 2011. School Effectiveness and School Improvement, 28(2), 292-313.

Leufer, T. (2007). STUDENTS'PERCEPTIONS of the Learning Experience in a LARGE CLASS ENVIRONMENT. Nursing Education Perspectives, 28(6), 322-326.

Liu, C., Zhang, L., Luo, R., Rozelle, S., & Loyalka, P. (2010). The effect of primary school mergers on the academic performance of students in rural China. International Journal of Educational Development, 30(6), 570-585.

Mo, D., Yi, H., Zhang, L., Shi, Y., Rozelle, S., & Medina, A. (2012). Transfer paths and academic performance: The primary school merger program in China. International Journal of Educational Development, 32(3), 423-431.

National Commission on Excellence in Education. (1983). A nation at risk: The imperative for educational reform. The Elementary School Journal, 84(2), 113-130.

Nye, B. A., Hedges, L. V., & Konstantopoulos, S. (2000). Do the disadvantaged benefit more from small classes? Evidence from the Tennessee class-size experiment. American journal of education, 109(1), 1-26.

Overbay, A. (2003). School size: A review of the literature. Research Watch, 3, 1-12.

Pittman, R. B., & Haughwout, P. (1987). Influence of high school size on the drop-out rate. Educational evaluation and policy analysis, 9(4), 337-343.

Slate, J. R., & Jones, C. H. (2005). Effects of school size: A review of the literature with recommendations. Essays in Education, 13(1), 12.

Toenjes, L. A. (1989). Drop-out Rates in Texas School Districts: Influences of School Size and Ethnic Group.

Thorsen, H. (2017). The effect of school consolidation on student achievement.

Werblow, J., & Duesbery, L. (2009). The impact of high school size on math achievement and drop-out rate. The High School Journal, 92(3), 14-23.

[1] The project is being implemented in three phases over a period of 30 months, coming to an end in 2020. The two phases of the project have been completed. It is now in the third phase of implementation, which will last for 18 months. https://niti.gov.in/sustainable-action-transforming-human-capital-sath (accessed on 1st October)