Reference basket: 2021 structure
Romania · one dot = one thousandth of the basket share
The canvas data failed to load. All its figures are in the text below.
Each dot is one thousandth of the reference structure: one leu out of 1,000.
The 2021 INS weights describe the reference basket — 335 dots for food and 157 for housing, water and energy; the other eleven categories share the rest. Food’s published weight is 33.4%. Weights shown to one decimal place sum to 99.9%; the largest-remainder method normalises them to 1,000 dots, giving food 335.
Basket weights and price changes answer different questions.
We apply December 2021–December 2022 price changes to the 2021 basket structure. The weighted products sum to 15.65 percentage points, against the published HICP rate of 14.12%. In the modelled sum, food has 485 of 1,000 points, housing 241 and communications 3.
The same price changes, applied to different profile weights.
Choose a profile: the canvas estimates how the weighted sum would be distributed using that group’s 2024 INS survey structure. The farmer-profile model is 17.1%, +3.0 points above the published HICP rate of 14.12%. This is a model estimate, not a measure of inflation experienced by households.
Move the dial and the canvas resettles for each year.
For 2021–2025, the year-by-year view uses the previous year’s basket weights: 2020 weights for 2021 and the 2021 basket for the 2022 comparison. The modelled sum for 2022 is 15.65 percentage points, against the published HICP rate of 14.12%. For September 2026, the selector uses annual rates for that month and 2024 INS weights, the latest used here.
What moved
Two counts, one basket
A budget share describes how spending is divided. A weighted estimate shows how those shares combine with price changes. They are different statistics.
The first count uses the published 2021 reference-basket weights: 33.4% for food and 15.7% for housing, water and energy. Weights rounded to one decimal sum to 99.9%; dots are normalised to 1,000 by the largest-remainder method, so the 33.4% share appears as 335.
The second count is a simple model: each reference-basket weight multiplied by its annual category price change. For 2022, the 13 products sum to 15.65 percentage points, while the published HICP rate is 14.12%. In the modelled sum, food accounts for 48.5% and housing 24.0%; these are not shares of the official rate.
Categories with smaller price changes make up a smaller part of the modelled sum, even when their budget share is larger. Communications are one example: 5.4% of the reference basket and 0.3% of the estimate. The counts below compare the 2021 reference basket with the 2022 modelled sum; each view totals 1,000 dots.
Communications had a 5.4% share in the 2021 reference basket. Their 0.98% annual price change in 2022 gives them 3 out of 1,000 points in the modelled sum; this does not measure how much people spent on phone or internet services.
Transport had a 7.0% share in the reference basket and a 9.94% annual price change in 2022; it has 44 out of 1,000 points in the normalised model.
Category by category
Thirteen rows, two orders
The ranking can show either the basket share or each category’s share of the estimated sum of weight × annual price change.
The bar shows size relative to the top position. Weights are from the 2021 INS family budget survey; the second view multiplies each weight by its annual price change. Dots show each category’s share of the estimate, not an official HICP contribution.
Same rate, different basket
One HICP rate, estimates on different structures
The published HICP rate for 2022 was 14.1%. The chart compares it with estimates from applying 2022 price changes to each profile’s structure in the 2024 INS survey.
published HICP rate model estimate 2022 · 2024 INS profile weights × 2022 price changes
The model gives the highest value for the farmer profile, 17.1%, +3.0 points above HICP; for employees it estimates 15.3%. These are estimates using fixed 2024 profile weights, not measured costs for each group.
These estimates differ from HICP because they apply 2024 profile weights to common 2022 price changes; they do not show observed inflation rates for household groups.
How we counted
Method and limits
Basket weights come from the INS family budget survey, TEMPO-Online table BUF109I. Price changes for December 2021–December 2025 come from IAPC102. For September 2026, we use annual Eurostat HICP rates from prc_hicp_minr, RCH_A, updated 16 October 2026; all 13 divisions map to the model categories, with insurance and financial services separate from personal care and miscellaneous.
For 2021–2025, weights come from the previous year’s basket: 2020 for the 2021 point and 2021 for the 2022 comparison. The model multiplies each reference-basket weight by its annual category price change, then adds the 13 products. The 2022 comparison uses the 2021 basket and price changes from December 2021 to December 2022: the sum is 15.65 percentage points, against the published HICP rate of 14.12%, a difference of 1.53 points. This model does not decompose the difference into causes. Recounted dots show each category’s share of the modelled sum; they are not official contributions.
Dot allocations divide each category’s raw basket share by the sum of weights, then use the largest-remainder method so every canvas totals exactly 1,000. Food’s raw 33.4% share displays as 335 dots after rounding differences are distributed. Household profiles use the 2024 INS survey, the edition used here for social categories, applied to December 2021–December 2022 price changes; results are model estimates, not experienced inflation.
Three limits that matter
The current classification starts in 2020. The canvas compares 2021–2025 using INS data and adds September 2026 using Eurostat data.
Profile weights are fixed at 2024. Each group uses its 2024 INS spending structure for every year; only the applied price rates change. Results are estimates, not observed group costs.
2026 is still in progress. The latest point uses September 2026 annual rates, compared with September 2025, and 2024 INS survey weights. The published HICP rate is 6.5%; the modelled sum is 5.88 percentage points. This monthly observation is not a full-year result.
Values are recalculated from the cited sources; for 2026, Eurostat publishes the headline rate, while the weighted sum here is an estimate using 2024 INS weights.
Sources
What we read
- National Institute of Statistics (Romania), TEMPO-Online, BUF109I — Structure of total household consumption expenditure, by expenditure category, social category and area of residence, table, extracted 9 September 2026.
- National Institute of Statistics (Romania), TEMPO-Online, IAPC102 — Harmonised consumer price indices by group of goods and services, ECOICOP version 2 classification, table, extracted 9 September 2026.
- Eurostat, Romania HICP, September 2026, release of 16 October and prc_hicp_minr RCH_A series for the headline and division rates. HICP methodology separately describes official weights and contributions.