The economy front door: methods
Page: /economy/, “How is India’s economy doing?”. Code: apps/site/src/pages/economy/index.astro; the coming-releases list is step economy_calendar (compute/economy_calendar.py, publish/economy_calendar.py), bundle economy/calendar/upcoming.
The page writes no number by hand. It reads:
| Bundle | What it gives the page |
|---|---|
economy/models/scorecard | Per dimension: how many readings are stronger (prices: hotter) or weaker than usual; the weakest reading on the board |
economy/models/breadth_facts, breadth | Share of activity indicators improving, its median since 2008, a year ago, and the chart |
economy/models/composites_facts, activity, inflation | The two composites at the last month every theme reports (never the partial month), three months earlier, and the theme that moved most |
economy/models/liquidity_facts, fci_facts, fci | Banking-system liquidity and its run of surplus or deficit days; financial conditions |
economy/prices/facts | Headline and core CPI, the RBI’s target |
economy/bonds/facts, rbi/stats/repo | The ten-year benchmark yield; the repo rate |
economy/calendar/upcoming | MoSPI releases due, from the newest version of the Advance Release Calendar, and how many past releases slipped |
The verdict. From the two composites at their last complete month: activity above +0.5 and inflation above +0.5 is “running hot”; activity above +0.5 alone “growing, with prices in check”; activity below −0.5 with inflation above +0.5 “slowing, with prices still high”; activity below −0.5 alone “cooling”; otherwise “steady” (with “prices heating” when inflation alone is above +0.5). “Households are the soft spot” is added when fewer than half of three or more household indicators are improving.
What the page does not claim (from docs/methods/economy_models.md): composites are readings against the indicators’ own last three years, not growth or inflation rates; breadth, the composites and the financial conditions index failed or were null in the pre-registered tests C60-C63, so nothing here is said to predict GDP, industrial output or credit; breadth history is pseudo-real-time; stale indicators are left out, not carried forward.
Dimension links. Each of the ten questions links to its page where one exists (/economy/prices/, /economy/bonds/) and otherwise to where the data lives today (/macro/, /macro/pulse/, /rbi/), marked “Full page coming”.