The Right to Education (RTE) Act has been a cornerstone in changing the education landscape in India. With the introduction of the Sarva Shiksha Abhiyan, the goal of ensuring universal primary education was aggressively pursued, and a significant quantitative impact in terms of the enrolment ratio has been seen. For the past six years now, enrolment in the country has been around 96%, which may seem a great feat. However, an assessment of the actual learning levels reveals the flip side of the coin. It is almost as if the common ‘volume versus quality’ trade-off has played its part in this scenario too, like any other.
The Annual Status of Education Report (ASER) 2014 indicates how this linear approach hasn’t reaped the benefits it should have; learning levels of students are still a huge concern. According to the survey, almost 50% of Class V students were not able to read basic sentences, and more than 70% were unable to perform simple division.
Thus, it is important for state administrations to realize that improving infrastructure and resources should be accompanied by commensurate learning levels of students. Thus, the need for a measure of efficiency emerges in order to assess education systems in their ability to convert educational inputs to outputs. This can help provide an objective way for states to get feedback on their education delivery process and do away with the practice of judging the performance of states based solely on their inputs, or outputs.
The objective of this work, therefore, is to develop a methodology to measure the relative efficiency of the education delivery process and provide insights on what states can learn from peer-to-peer exchanges. Since there are multiple inputs and outputs, the conventional notion of efficiency defined as the ratio of output to input would not work here. The field of Operations Research provides a suitable methodology in the form of Data Envelopment Analysis (DEA), which has been used extensively in several investigations and researches across countries and sectors for efficiency analyses. DEA compares each entity (states in this case) with its peers in the set, and assigns a relative efficiency to it. For states that are marked efficient, it does not imply that there isn’t room for improvement; it simply means that in ‘relative’ terms, there is no other state performing better than the given one.
The first step in efficiency measurement using DEA is to identify relevant inputs and outputs for the educational process. The learning outcomes reported by ASER are used as outputs, namely reading levels in local language, basic arithmetic ability and learning levels in English. Similarly, the resources and infrastructure provided by state authorities to facilitate education are the quantifiable inputs.
The RTE lays down certain minimum requirements, and the percentage of schools adhering to those norms serve as input values in this method. The seven factors, as mandated by RTE, considered in this analysis are pupil-teacher ratio, classroom-teacher ratio, availability of drinking water, availability of usable toilets, availability of buildings and playgrounds, availability of library with books and mid-day meals being served. Two additional inputs to represent the socioeconomic background of students as well as the local village infrastructure are also used. These account for the specific conditions within a state. The data is obtained from ASER reports of 2011, 2012, 2013 and 2014.
The findings presented here are based on research conducted by our team at IIT Delhi, and a detailed research paper elaborating a part of this research is under consideration for publication. Interesting insights about the standing of various states with respect to each other emerge from the analysis. While there are 12 inefficient states from the 2014 data, an extension of the same to previous years (2011-13) yields a few patterns. Gujarat, Jammu & Kashmir, Karnataka, Rajasthan and Tamil Nadu have consistently been performing poorly, and are inefficient across all four years. Punjab had been performing efficiently until 2014, when it slipped slightly. On the contrary, Uttar Pradesh used to be inefficient in 2011- 12, but has remained in the efficient group since 2013, indicating improvement in its education delivery.
Since DEA compares each state to all others while computing efficiency, some states act as the superior efficient peers, whose better performance results in inefficiency of others. Himachal Pradesh and Manipur are two states that have consistently been the efficient peers for the most number of inefficient states.
For each of the inefficient states, it is also possible to highlight the output attribute that needs particular attention, improvement in which will lead to the maximum rise in efficiency. The importance of comparing performance on grounds of efficiency as opposed to merely outcomes is reinforced by the fact that while outcome-oriented rankings would classify Punjab and Sikkim as high-performers, the analysis shows that they are not performing to their fullest potential. Similarly, from an input-oriented perspective, Gujarat, Karnataka, Maharashtra, Rajasthan and Tamil Nadu are seemingly providing good resources, but are not able to translate them to equally good learning achievements.
Hence, in order to not trade off quality against volume, as recently emphasized by the prime minister, a careful inclusion of inputs as well as outputs is needed in assessment of the status quo, and data-driven insights need to be drawn to identify the right focus areas for improvement. DEA fulfils all such requirements, and can aid in the policymaking process in other sectors too. A sound elementary education system is essential for our country to tap the potential of its vast human resource, and the importance of data-driven policy in this context can never be overemphasized.