About the project and methodology

Basepoint Doodles brings together the official economic statistics on Bulgaria and presents them so that the data speak for themselves. There is no commentary on the numbers: the selection, the calculations and the arrangement of the charts are the editorial choices.

• Now: what is happening in the economy right now.
• What drives growth: where growth comes from and what we sell.
• Catching up with the EU: convergence with the EU in five dimensions.
• Bulgaria and its peers: Bulgaria and six comparable economies.
• The economy and everyday life: wages, prices, housing and credit.
• Indicators: every chart, with search and tables.

Methodology

Where the data come from

Below are the sources and tables the site uses. Each name links to the official table or file.

• Eurostat — bop_c6_m, bop_c6_q, ei_bssi_m_r2, ei_cphi_m, jvs_q_r21, lc_lci_lev, lc_lci_r2_q, lfsi_emp_a, lfsi_emp_q, nama_10_a10, nama_10_a10_e, nama_10_lp_ulc, nama_10_pc, namq_10_a10, namq_10_gdp, namq_10_lp_ulc, nrg_cb_em, nrg_cb_pem, prc_hpi_q, prc_ppp_ind, prc_ppp_ind_1, sts_cobp_q, sts_copr_m, sts_inpp_m, sts_inppd_m, sts_inppnd_m, sts_inpr_m, sts_sepr_m, tipsii10, une_rt_m
• NSI — Export and import price indices, Employees under labour contract, Average wages and salaries, Retail trade
• BNB — Gross external debt, Foreign trade (FTRD), Interest rate statistics, Monetary statistics: loans and deposits, Foreign direct investment
• Employment Agency — Monthly bulletins
• Registry Agency — Property Register · Statistics by type of registration

At every update the site downloads the full available history, so that revisions to older data are picked up too.

Four different dates

• Observation period: the month, quarter or year the value refers to. It appears next to every chart.
• Source publication date: when the source last updated the table (for Eurostat, taken from the table itself).
• First detected: the date on which the site first found the new period in the source's data. The date is kept in a register that is never rewritten; its history starts with the launch of V2, so older periods have no such date. "Latest data" on the Now page is sorted by it.
• Site update: the date on which the current version was published. It says nothing about how fresh any individual chart is.

What the freshness labels mean

• current: no expected period is late.
• delayed: one expected period is missing. Only a label is shown.
• stale data: two or more are missing. The chart is dimmed.
• no data: the series is empty or broken and is not shown.

The status is calculated from the latest observation and the source's usual publication lag, not from the time of the download: a successful download cannot make a stale series current. If the site has not been updated in time, the page raises the status itself after the expected date of the next period, and never lowers it. That date follows the source's usual lag, not an official release calendar.

Bases and comparisons

• Base = the average for a year, not a single month: "2019 = 100" is the average for 2019. "The economy and everyday life" uses 2019 throughout, wages in Indicators use 2021, and house prices use 2015 (as Eurostat publishes them).
• "Start of the period shown = 100" is recalculated every time the period changes: the base is the first month for which every line has data, and it is stated on the chart.
• Comparisons in one row are read at one period: the newest one all the series share. That way a June wage is never divided by August prices.
• Country rankings use a common period. A country that has not yet published it, or already has a newer one, is listed separately with its own period.

How the derived figures are calculated

• Real wage: the index of the average gross monthly wage is divided by the HICP, both rebased to a common base; the annual change is taken from the resulting ratio. Percentages are never subtracted.
• Contributions to GDP growth: Eurostat's official ones (quarterly and annual). The gap to the published growth is shown as the statistical discrepancy.
• Contributions to GVA growth by industry: an approximation, the change in the industry's real volume divided by total GVA a year earlier. Chain-linked volumes are not additive, so the gap is shown separately.
• Contributions to inflation: by Eurostat's chain-linking method. The HICP is linked in December, so each contribution has a part from the current year (the year's weight times the change since December) and a part from the previous year. The four groups cover the whole basket; the remainder is rounding in the published indices (under 0.1 pp).
• Compensation per hour against productivity: compensation per hour worked is derived from the published ULC and productivity (ULC = compensation ÷ productivity), and productivity is then subtracted from it.
• Current account as % of GDP: the sum of the last 12 months divided by the sum of nominal GDP over the last four completed quarters. On "What drives growth" the monthly data are summed into quarters and divided by GDP for the same four quarters.
• 3-month rolling change: the average of the last three months against the same three months a year earlier; one comparison, not an average of three rates.
• Percentile: the share of all past observations of the same series that lie below the current one, with ties counted as half.
• Annual and monthly change: the annual change is calculated from the calendar adjusted index, the monthly change from the seasonally adjusted one.
• Sign: changes, contributions and balances are written with a sign (+/−); levels, shares, rates, indices and sums never are.

A missing period is never filled by interpolation, and a comparison never skips a gap in the series. If the input is missing, the output is empty.

Vintages and classification changes

• Purchasing power standards (PPS). The volume indices of GDP and of actual individual consumption (AIC) are in PPS; purchasing power parities (PPPs) are the conversion factors used to calculate them, and they allow a comparison after accounting for differences in price levels. The latest year is a preliminary estimate and Eurostat revises the whole history; the site downloads it again at every update.
• COICOP 2018. From 2026 the HICP follows ECOICOP 2 (13 groups), and price levels by category follow COICOP 2018 (from 2022). The old tables (prc_ppp_ind, prc_hicp_inw) have been discontinued; the category comparison with 2007 uses COICOP 1999 and is marked as such.
• Euro from 2026. The BNB's monetary and interest rate statistics are in euro from January 2026. The BNB notes a break in the comparability of interest rates on new business with earlier periods; the charts mark it.
• Delays by country. Countries publish on different days; comparisons use the common period, not each country's latest.

Registered sales of immovable property

• What is counted. The "Sales" column of the Registry Agency's quarterly statistics by type of registration in the Property Register: the number of sale deeds registered in the quarter.
• Not only homes. The column covers all types of immovable property — homes, land, garages, commercial property and others. A single sale deed may cover more than one property. The indicator measures activity in the property market, not the number of homes sold and not prices.
• By office. The statistics have one row for each of the 113 registration offices. The site shows the national total and the Sofia Registration Office; the other offices enter only the breadth indicator.
• National total. This is the "Total" row of the statistics. On every update the site checks that it equals the sum of the offices exactly; if it does not, the quarter is not published. In the statistics for the first three quarters of 2005 the "Total" row does not equal the sum of the offices and neither number can be taken as correct, so those quarters are empty.
• Year-on-year change. Compared with the same quarter of the previous year. If either of the two quarters is missing, there is no change.
• Net balance of offices. For each quarter every office is compared with the same quarter a year earlier: an increase (more sales), a decrease (fewer) or no change (exactly the same number). The net balance is 100 × (offices reporting an increase − offices reporting a decrease) ÷ offices with a valid comparison, in percentage points, between −100 and +100. Offices with no change stay in the denominator. An office without a figure in either of the two quarters has no comparison and enters neither the numerator nor the denominator. Each office has the same weight, regardless of its number of sales: the indicator says how widespread an increase or decrease is, not how large the change is.
• Missing quarters. The Agency's archive has no statistics for Q4 2020 and the first three quarters of 2021. They are empty, with no interpolation, and the year-on-year comparisons across them are empty too, up to and including Q3 2022.
• Office names. Over the years the Agency has written the same office in different ways (with or without the full "Registration Office" prefix, abbreviated, or wrapped over two lines). Every spelling is mapped to one canonical name, so each office keeps its whole history. A new, closed or renamed office stops the update until it is added by hand; Galabovo appears in the statistics from Q2 2007. In two releases (Q2 2009 and Q1 2012) the font has no text layer; the names there are recovered from the order of a neighbouring release and accepted only if one and the same key explains all 113 names.

Terms

• NSI: actual individual consumption in PPS and the 2024 release
• NSI: gross domestic product, national level
• Eurostat: HICP methodology and the methodological manual (2024)
• BNB: interest rate statistics

Every published series

The catalogue holds 528 series; each one lists its source, table, calculation and dates. The catalogue is also available as a file: catalogue.csv · catalogue.json.