Who goes to college in India, and who does not.

The top bar is the level of study, the bottom bar the social group. Higher education in India means, to a first approximation, the under-graduate degree: eight seats in ten sit at that level. Post-graduate is a distant second. The research track is thin, Ph.D.s are under one per cent of the room, and the certificate and M.Phil. rungs are shrinking as the system consolidates around degrees.

The social-group bar carries a residual. The survey names Scheduled Caste, Scheduled Tribe and Other Backward Classes enrolment directly; everyone else, the general category, minorities, the unclassified, falls into "Others", computed here as the remainder. Three named groups plus that residual sum to the whole.

Two lines, five points each: one per survey round. Both rise: the system expanded by roughly a fifth in four years. The female line (blue) rose slightly faster in percentage terms, and the Gender Parity Index held at or above 1.0: women enrol at the same rate as men. But in absolute numbers men still outnumber women, because the male college-age cohort is itself larger. Equal rates, unequal counts.

Read the slopes, not the levels. The 2020-21 step is the largest single-year jump: partly real expansion, partly a wider institutional net catching more colleges in the DCF (Data Capture Format) response. AISHE's coverage of institutions improved year by year, so part of any rise is the survey seeing more, not the system growing more.

Three GER lines, all rising, all ordered the same way: All Categories on top, Scheduled Caste below it, Scheduled Tribe at the bottom. The gaps narrowed a little over five years, but the ordering held. A Scheduled Tribe student of college age was markedly less likely to be enrolled than the All-India average throughout the window; the latest gap is in the note beside this one.

The lines move together because they share a denominator: the same age projection, sliced by group. When the projection is revised, all three shift. Read the gaps between the lines, which are more stable than any single level.

This is the chart behind the Gender Parity Index. GPI is a single number, female GER divided by male GER, and a single number hides how close the two rates are. Plotted as two lines, the story is visible: women's GER sat at or above men's across the whole window, peaking around 1.06 in the middle years and settling back to parity by 2021-22.

The lines nearly touch because the rates nearly match: that is what parity looks like. It does not mean equal numbers of men and women are enrolled: men still outnumber women in absolute enrolment (Plate 1), because the male college-age cohort is larger. Equal rates, unequal counts.

Two strips, one question each. The top strip asks does college reach here?: GER by state, against the All-India rule. It is a long downward slope: the city-territories where students migrate to sit at the top, the large poor states at the bottom. The bottom strip asks where are the students?, pure enrolment volume, and the answer is a cluster of big states. Uttar Pradesh and Maharashtra alone hold more than a quarter of all enrolment.

The two rankings tell different stories. A state can hold many students and still enrol a small share of its college-age population (Bihar); a city can enrol a high share of a cohort it did not raise (Chandigarh). Read the two strips together: access and volume are different stories.

Eight small panels, one per level of study, each on its own scale (Under Graduate would swamp the rest on a shared axis). Read the shape of each line, not the level: a rising line grew, a falling line shrank. The system consolidated around the degree: under-graduate and post-graduate rose; the diploma rungs held; and the two tail rungs moved hard in opposite directions.

M.Phil. fell by nearly three-quarters: the National Education Policy 2020 discontinued it, and the line shows the policy landing in the data. Integrated programmes (a combined bachelor's-master's) more than doubled, from a small base. Certificate enrolment halved. These are real shifts in what "higher education" means, visible only when the levels are apart.

Five faculty ranks, ordered senior to junior, each with a male dot (grey) and a female dot (blue). The line between the dots is the gender gap at that rank. The pattern is monotone: the higher the chair, the fewer the women. Professors are the most male; demonstrators and tutors, the bottom of the academic hierarchy, are the only majority-female rank.

This is the classic academic pyramid, sharpened by gender. Lecturers and assistant professors are the largest rank by far (seven in ten teachers), and women hold roughly four in ten of those posts. The thinning at the top is real, but the survey says nothing about why: promotion gaps, retirement cohorts, and historic hiring all leave the same trace in a stock count.

Four groups the survey counts separately from the SC, ST and OBC figures above: Persons With Disability, Muslims, Other Minority Communities, and the newer EWS category. Each group has two bars: men (grey) above, women (blue) below. These are counts: how many enrolled, not how large the group is. A small bar may mean a small community or a community that college reaches poorly: the survey serves no denominator to tell them apart.

The gender split within each group is its own signal. Muslims reach near-parity (women are half of Muslim enrolment); disability enrolment skews male, as disability identification itself does. EWS is the youngest category, its line in the data begins in 2020-21, after the reservation came in, so no trend is drawn for it yet.

One dashed line, five points. The pupil-teacher ratio is the system seen whole: every enrolled student divided by every teacher, across every institution. It fell from the high twenties to the mid-twenties over five years: a small improvement, because India added teachers slightly faster than students. The line is deliberately flat: this is a resource ratio, not a dramatic series.

Do not read it as a class size. A PTR of 26 means one teacher for roughly every 26 enrolled students across the whole system: a research supervisor with eight students and a lecture hall with 200 both feed the same average. It measures how thin the teaching staff is spread, not how big a classroom is.

Each bar is the number of regular students for every teacher in the state's colleges and universities. It is not a class size, but it is a fair measure of how thinly teaching is spread: in a state at the top of this chart, each teacher is shared among far more students.

The ratio is low in small states and union territories partly because their colleges are small. Read the large states against each other.

Where the bar falls short of the dot, women are a smaller share of those who teach than of those who learn. A young woman in college in such a state is less likely to be taught by a woman, and less likely to see one at the head of a department (Plate 6 shows how women thin out as the posts rise).

A count of colleges says how many doors there are, not how big they are: many colleges in Uttar Pradesh, Rajasthan and Karnataka are small private colleges with a few hundred students each. Read it with Plate 4, which shows how much of each state's young population is enrolled.