A new research study released by the Indian Institute of Management Bangalore (IIMB) has found that teachers in Bihar’s public schools often misjudge how well their students are performing, and these misjudgments vary sharply by caste.The research shows that even when students from different caste groups score equally well on standardised tests, the perceptions of the teachers do not match the actual performance, with backward-caste students consistently rated lower than they deserve.The study, ‘Caste identity and teachers’ biased expectations: Evidence from Bihar, India’, conducted by Soham Sahoo, chairperson of Public Policy, and Ritwik Banerjee, chairperson of Economics, analysed data from 105 government schools across Bihar, gathered through a detailed survey of students, teachers, households, and schools.The students took standardised subject-specific tests in Maths, Hindi, and English, which determined their true academic rank within the class. Their teachers were separately asked to place them in the top, middle, or bottom group based purely on perception. The difference between test-based rank and teacher-assigned rank forms what the researchers call the ‘Evaluation Bias’.The study finds that the teachers’ perceptions are often inaccurate in general, but the inaccuracy becomes systematic when caste is involved. Teachers, especially those from forward castes, tend to underrate backward caste students, even when their test scores match those of forward-caste classmates taught in the same classroom.Backward-caste students, particularly those from SC and ST groups, are ranked 0.22 to 0.43 positions lower than their actual test-based rank when assessed by a teacher belonging to a forward caste. The study shows that backward caste students are also 17% to 27% more likely to be underestimated across all three subjects. When the researchers disaggregated backward caste into OBC and SC/ST categories, the gap was found to be driven mainly by the under-rating of the SC/ST students relative to the general category students.The researchers emphasise that this caste-based evaluation gap is not linked to differences in actual academic performance. Students across caste groups score similarly on standardised tests. Instead, the gap arises from teachers’ perceptions and inaccurate beliefs, rather than from concrete differences in ability. The study also examines other possible explanations such as class size, lower attendance of backward-caste students in classrooms led by forward-caste teachers, or the age and experience of the teacher and finds that these do not account for the gap. The authors argue that such misjudgments can have long-term consequences. If teachers consistently expect backward caste students to perform poorly despite evidence to the contrary, it may influence classroom interactions, affect the confidence and motivation of these students, and eventually widen educational gaps, an effect consistent with the Pygmalion effect, where teacher expectations shape student outcomes.The study adds to earlier research on discrimination in grading, but shifts the focus to the beliefs that precede grading, showing that biased expectations exist even before teachers assign marks. The researchers say this makes it crucial to correct misperceptions through caste-sensitisation programmes and simple feedback mechanisms that alert teachers when their perception does not match actual student performance.Published - November 19, 2025 07:27 pm IST