Improving India’s School System: A Bottom-up Approach

Vijaya Sherry Chand and Kathan Shukla

In a recent article (Opportunity to Rethink Educational Policymaking), we discussed the problems of ‘top-down policymaking’ as it happens in our context: failure to build a body of policy-relevant knowledge based on empirical, context-sensitive studies, and the tendency to rely on short-sighted political compulsions. To counter these, we need to (a) develop an educational ecosystem that does not rely only on teachers and teacher-trainers, but draws on other kinds of human resource such as curriculum developers, administrators, educational data analysts, school psychologists, and test developers and psychometricians and (b) make this ecosystem a base for ‘bottom-up’ decentralized policymaking and evaluation. In this article, we contend that for these two responses to take shape, schools should learn to demand and expect service of high quality; along with that, the higher levels of our system need to reorient themselves towards greater autonomy and decentralized functioning, and a networked structure at the district-level with a range of skills and professional service to schools as its mission, should evolve.

The development of a decentralized ecosystem depends on two crucial assumptions: the willingness and capability of the current state and district levels for educational research and training (state councils and their district institutes of education and training) to foster such an ecosystem, and second, a reorientation of the school management system represented by Samagra Shiksha towards supporting bottom-up policies. The former is important because, at least in the near future, there is no alternative to this system and one can only hope for a reorientation of attitudes and capabilities. The latter is crucial since a huge one-way data-collection system has become entrenched—there is hardly any evidence available on how the data (both administrative and academic) that all states collect from the grassroots flows back to the schools in the form of actionable points for improving quality. If the centralization implied by a uni-directional flow of data and information is to be countered, ensuring data analytical capabilities at decentralized levels becomes necessary. 

What should be done about the first assumption? Two changes are needed. First, at the state-level, the Councils should, either through appropriate recruitment or networking, develop interdisciplinary expertise on designing relevant interventions, evaluating programs, developing sharp policy reports, and supporting multi-disciplinary teams at the district-level. Such expertise would require a diverse mix of curriculum experts, data analysts and psychometricians, educational economists, sociologists, psychologists, and so on. The Councils would likely argue they already have departments with such experts. However, the test of the system is in the relevance and applicability of its outcomes—research that feeds into policy. Here the deficiencies of our systems become apparent. The Councils should open themselves up to an assessment of their knowledge-production performance; policy consulting with state-level administrators; media outreach to convey educational issues and policies to the average citizen; and developing the next generation of researchers for district-level work. As they do this, the Councils will obviously have to factor in political compulsions and populism that will dominate in any democratic set-up. So, a fifth test of a good Council is in how well it matches strategic advice on informed policy-making with tactical ‘schemes’.

The second change needed is at the district level—the same honest assessment of the quality and relevance of the outputs generated by the district-level counterparts has to be undertaken.  Such reflective assessments should open the door to diversity of expertise needed and to a willingness to examine more localized alternatives to supporting educational achievement.

The second assumption has to do with the school management system that operates in parallel with the research function through Samagra Shiksha and its constituents. Nobody nowadays questions the need for data or evidence-driven policy adjustments.  The problem is with the interpretation of “data-driven” decision-making. What we have seen is a one-way flow of data from the grassroots to some mysterious centralized location; no doubt some use of the data is made to generate macro-level representations of what is happening to education, and decisions may indeed be based on it. But we need to move beyond this—and back to a fundamental premise underpinning any useful information system: data should lead to insights for action by the actors at the cutting edge of education. Repeatedly telling these actors that achievement levels are miserable, but not generating solutions to tackle these, serves no purpose. The answer lies in ensuring that data processing capabilities are developed at the local level. The ability to generate insights locally from the data collected locally is liberating and motivates action, since there is a greater sense of autonomy and control over the data that ‘I am generating at my level’.

An example will help. Recently we did a small exercise with the block resource centre coordinators of one district. A lot of data on children’s test results were being sent “upward”—it seemed to disappear into an abyss, with nothing coming back in the form of processed advice on action. We asked a simple question: With the same data, can you identify which teachers and schools in your block need more attention, in terms of problematic learning issues (by subject, by class)? Comparing the school’s performance with the block-level mean and standard deviation, it was possible to identify specific classrooms that could be earmarked for personal visits and academic follow-up—classrooms with low achievement and low variation (stagnating classrooms) and low achievement and high variation (vulnerable classrooms, where some children do well in an overall environment of poor performance). The control over simple analytical tools (using spreadsheets) and the facilitation of further action through qualitative exploration of the good teachers/ classrooms in order to learn from them, serve to promote autonomous functioning at a decentralized level. It needs no orders or approval from the top.  

With the two assumptions out of the way, it should be possible to develop a decentralized educational ecosystem that can undertake a range of ‘bottom-up’ policy studies: efficacy studies; replication studies to develop a deeper understanding of contextual factors; scaling up interventions that work; evaluation of field practices; and so on. Now comes the crux of the model: the school should be at the centre of the ecosystem.  Its support should come from the networked decentralized ecosystem which would include the research and training structures, Samagra Shiksha constituents, and other human resources relevant to the district’s needs. But who will drive this network? The answer is obvious, but is likely to be unpalatable to well-entrenched administrative and political interests: the school. Have our schools ever demanded they be served by the educational system instead of being blamed all the time? No. Have they realized that a decentralized educational ecosystem that is responsive to their needs is possible? The answer again is no. Our schools will have to learn to demand and expect good service. A cynical response to this prescription will be derision: our schools just do not have the capability. But that is precisely the challenge for our educational leadership: educating schools about their right to demand high quality support.

This article was first published in Business Line on Campus on 29 June 2020.

(Vijaya Sherry Chand and Kathan Dushyant Shukla are faculty members at the Ravi J. Matthai Centre for Educational Innovation, IIM Ahmedabad.)