One of the foundational principles of Sherlock Holmes’ Science of Deduction can be summarised in the quote: “It is a capital mistake to theorize before one has data”.

This statement is not just applicable to forensics or investigation but to almost every field of study, from science to social sciences and even policy making. Many governments across the world today accept the efficacy of using data in policy formulation.

Advertisement

Thus, NITI Aayog recently announced its decision to partner with institutions to promote evidence-based policy making. Similar endeavours have been made by the Delhi Government. These reforms are in line with the global thrust towards data-driven policy making.

Data-driven policy making is seen as the gold standard in governance as the use of evidence or data is crucial in identifying the problems plaguing the society and formulating appropriate measures.

Long before the advent of Computers and Information Technology in our lives, the large-scale sample surveys pioneered by Prof. P.C. Mahalanobis collected valuable information. This data was meant to act as an instrument in the formulation of sound policy decisions. However, the policies adopted in the decades following Independence, rather than being data driven were based on inherent biases and electoral calculations.

Take the case of Education in India. The education policy since independence has placed focus on improving Tertiary Education. However, this has been at the cost of Primary and Secondary education. Gopal Krishna Gokhale had highlighted the importance of primary education as early as 1911 by proposing free and compulsory primary education for all.

However, it wasn’t until almost a century later that Right to Education guaranteeing free and compulsory education until age of 14 was considered a fundamental right. On the other hand, China with a well-structured education policy, managed to achieve the goal of universal primary education as early as 1980s. This undue bias against primary and secondary education has adversely affected the manufacturing sector in India due to a severe lack of medium-skilled workers.

The stated goals of many of the policies on education are primarily based on a metric of funds allotted or on outcomes like enrolment ratios. Less than appropriate use of available data has skewed the impact of these policies away from their intended beneficiaries. Also, at times these ‘stated goals’ are intended to paint a brighter than real picture of the state of the country.

Take a look at the education sector, especially the estimates of Gross Enrolment Ratio (GER) and Net Enrolment Ratio (NER). According to DISE Statistics, India achieved close to full NER at the prjmary level back in 2011-12 itself. Further, estimates from 2015-16 suggest that GERs at secondary and higher secondary level are close to 80 per cent and 56 per cent respectively.

On the face of it, these numbers show some progress in the country’s education, however a deeper scrutiny would reveal that the inadequacy of measures like GER and NER fail to shed light on the true state of education.
The reasons are twofold. One, the figures of enrolment or even completion of a grade fail to measure the learning outcomes of a child.

The recently published Annual Survey of Education Report (ASER) 2017 iterates the poor quality of learning prevalent in India. The report mentions that about 25 per cent of the enrolled children in the age group of 14-18 are still unable to read basic texts in their mother tongue. A major proportion of students show an inability to apply basic literacy and numeracy skills in day to day use.

Two, the NER and GER only consider the total number of students registered or enrolled in a particular grade and ignore whether these students attend classes or even pass the grade. Due to the Right to Education (RTE) Act, the difference between students enrolling for a particular grade and completing, called the transition rate, is very small until Class 8. However, transition rate is very high in secondary school nearing 17 per cent due to the prevalence of high drop-out rates.

A better measure to look at educational attainment would be to observe the number of students who complete a grade as a percentage of people who are eligible (i.e. above the minimum age required to complete) to complete a particular grade. The data from the NSS 71st round would place this figure at 44 per cent and 29 per cent for secondary and higher secondary education respectively.

Even if the focus on the quality of learning is ignored due to difficulty in acquiring data on the quality of learning, the mere percentage of grade completion itself is extremely poor. These are a lot worse for a girl child, with the completion rates close to 42 and 26 per cent for Secondary and Higher Secondary levels.

Until recently there has been a lack of a systematic and targeted approach to correcting such biases. An ideal policy to address this problem must be designed to target the behavior of citizens and households. This is done through a systematic study of data and identification of factors and the manner in which they affect behaviour. For example, Econometric Analysis reveals that the educational attainment of a girl child is an outcome of a host of factors.

Enrolment levels of girls reduce with distance from school. This can be seen as lower willingness of the family to send a girl to a school if the distance increases. To address this the Rashtriya Madhyamik Siksha Abhiyan (RMSA), the flagship scheme of the Union Government to promote secondary education, plans to create hostel facilities for secondary-level girl students in educationally backward blocks to ensure that girls aren’t denied education due to a long commute.

In light of this, the ‘Mukhyamantri Balika Cycle Yojana’ by the Bihar government which provides cycles or an amount of Rs 2,000 to girls enrolled in Classes IX and X is also a step in the right direction.

Another factor which affects the enrolment levels of a girl is the lack of adequate toilets in schools. To address this the percentage of schools with a girls’ toilet under RMSA increased from 64 per cent in 2010-11 to 98 per cent in 2015-16 (CPR Report). These are well framed schemes which tackle the inconvenience faced by girls, especially of menstruating age, in having to make a long commute without having adequate sanitation facilities.

The impact of distance to school and lack of sanitation facilities on learning outcomes is well known. However, a more robust data gatherring and analysis would reveal a host of other (and even unexpected) factors which have a strong effect on behavior. Taking another example; further analysis of the 71st Round of NSS reveals that the presence of a male sibling increases the probability of a girl completing higher secondary school by about 30 per cent.

This result is not what would be expected given that the conventional literature in this field sees the relation between siblings as a form of competition for education due to scarcity of resources. Multiple explanations can be offered to understand this, with some being more likely than others. For example, a girl with a brother may feel safer going to school.

An appropriate policy intervention should target the issue of lack of security for girls. Similarly, every scheme and policy intervention must target the root causes of problems plaguing our society, rather than merely addressing apparent symptoms.

Similar studies must be conducted to ensure that people’s behaviour can be deduced to bring about greater development and empowerment of citizens. The framing of policies must reflect an understanding of these underpinnings.

This can only be done if data backed deductions take a front seat in policy formulation. Without such a data-driven process, policy makers tend to rely on anecdotal evidence or even their personal opinion (which could be clouded). This may eventually result in even a well-intended policy falling short of achieving its goals.

The writer is a research scholar at the Delhi School of Economics.