Where it actually went
Between 2012-13 and 2026-27 the Union budget more than tripled in rupees, so almost every line went up and a chart of rupees says only that things got bigger. A share is the one view where something has to fall for something else to rise. Education’s slice went from 4.68% to 2.60%. Road transport went from 1.43% to 5.79%.
What was chosen over what
Share of total Union expenditure. 2012-13 figures are actuals; 2026-27 is a Budget Estimate, so the last bar is what the government intends to spend rather than what it has spent. Shares are computed from nominal rupees on purpose — a share is a ratio inside one year, so inflation cancels out of both halves. 2014-15 is excluded throughout: scheme money was reclassified out of ministry demands that year, so its figures measure something narrower. Source: ministry-wise expenditure.
The spending is not the outcome
These two columns are not a subtraction. The posts counted on the right are employed by states and courts, not by the Union, so the money on the left was never what was supposed to fill them — health and education are largely State subjects, and the Union share is a fraction of the combined total. Read together they say something narrower and harder to dismiss: sustained spending at one level of government does not by itself staff the level that delivers the service. Cumulative figures exclude 2014-15, whose basis differs, and years after 2025-26, which have no published price index. Spending source: ministry-wise expenditure.
The parts of the answer that live elsewhere
A budget line can also fall short without any of this showing up in a share. It can be announced and then revised away inside the same year — 1,31,114src crore was, across 12 schemes, in 2025-26 alone. It can be voted in full and then not staffed, which is what the empty chair counts. And it can simply never reach the target the government set for itself, which is what the board opens with.