In a written question in the Rajya Sabha earlier this year, an MP highlighted that 27% of students in government schools are paying fees — including admission, tuition, examination and development charges — and sought a response from the Ministry of Education. The MP’s claim was based on the findings of the Comprehensive Modular Survey: Education (CMS-E), conducted by the National Statistics Office as part of the 80th round of the National Sample Survey (NSS) from April-June 2025 (below charts).
The finding that a sizeable share of students, even in government schools, paid fees surprised many, since it could suggest that those schools were illegally charging fees, raising concerns over the implementation of the Right to Education (RTE) Act, which mandates free education for all children up to age 14. The survey covered 52,085 households, including 28,401 in rural areas.
This article examines whether government schools are indeed charging such fees, or whether the findings reflect issues related to data collection, compilation or both.
To identify the source of the problem, the average course fees paid by students in rural and urban areas were compared using education expenditure data for students in Classes 1 to 8, who are covered under the RTE Act. For the analysis, the course fees reported in the survey — which include admission fees, tuition fees, development fees and other compulsory payments — were classified into four categories: no fees, ₹1-₹5,000, ₹5,001-₹10,000 and above ₹10,000.
The analysis showed that, among those reporting course fees above ₹10,000, the average course fee paid by students in government schools in rural areas was higher than that paid by their counterparts in private schools.
Second, the data across all four course-fee categories showed that the percentage difference in average total educational expenditure between students in government and private schools ranged from −50.4% to +4.0% in rural areas and from −57.2% to +34.3% in urban areas (below table). Both points challenge the widely held perception that students in government schools incur substantially lower educational expenses than those in private schools.
Third, several States reported unusually low proportions of students with no course fees in government schools. In rural areas, the share of such students is particularly low — in Ladakh (0%), Tripura (2.6%), West Bengal (43%) and Kerala (53%).
Fourth, the percentage of students with no course fees in government schools is substantially lower in rural areas than in urban areas. For instance, in Himachal Pradesh this was 55.6% and 78.3% respectively, with similar patterns observed in six other States, which is difficult to reconcile with any existing government policy.
Possible problems
A commonly cited explanation for such discrepancies is data-collection error.
First, even when survey investigators understand the intended meaning of the questions, they may struggle to communicate it to respondents in the same context in which they were designed. In many cases, terms clearly defined in academic or policy contexts are interpreted differently in everyday usage. For example, a key question on educational expenditure in this survey is Question 9(i) in Block 5. It asks about “course fees”, which include admission fees, tuition fees, examination fees, development fees and other compulsory payments. Interestingly, in common usage, the term “tuition fees” is often understood to mean fees paid for private coaching or supplementary education.
However, in school education, “tuition fees” refers to payments made to schools for teachers’ salaries and basic academic resources. While the survey intends to capture this, respondents may have reported spending on private coaching as tuition fees.
Second, information was collected using Computer-Assisted Personal Interviewing (CAPI), under which responses are recorded directly on tablets. While this approach may facilitate faster processing, it also introduces certain limitations. For example, in earlier paper-based NSS surveys, enumerators could leave questions blank when respondents gave no answer. With CAPI, non-response is generally not permitted, and enumerators must enter a value before proceeding. Previous surveys suggest that non-response, resulting in missing observations, was not uncommon. However, there were no missing observations in this survey, which used CAPI.
This raises the possibility that some course-fee values may have been entered merely to proceed to the next question. Moreover, even minor modifications may disrupt the process, and features such as autocorrect can sometimes distort recorded responses. Even the World Bank has noted that CAPI systems can be highly rigid.
If a large number of students do not have access to free education, the issue warrants serious scrutiny. Yet it is surprising that, despite the matter being raised in Parliament, neither the Central government nor any State government has come forward to investigate. Conversely, if the findings result from flaws in data collection and validation, as the evidence above suggests, those responsible must be held accountable. Such errors not only distort policy research but also tarnish the reputation of government schools and teachers.
Anish Gupta teaches economics at Delhi School of Economics.
Gaurav and Deepti Kushwaha are pursuing their PhD and internship, at the Delhi School of Economics and IEDS Noida, respectively.
Published – July 15, 2026 07:00 am IST