When English reached the firm names in your town

Two waves: IMPEX everywhere at once, English in turn#

The first wave is the 1990s one. IMPEX, IMEX and EXIM (import-export), anywhere in a name, peak in 1994 in county towns, other towns and communes alike. In 1997, 14.4 in 100 new firms in the county towns had such a name, 17.4 in the other towns and 13.7 in the communes; in Bucharest, 31.9. By 1999 the share had roughly halved everywhere: 6.7, 7.7, 7.4 and 13.0.

TRADING is an English word, but it behaves like IMPEX: it peaks in 1996–1997 in every kind of place and lags nowhere. It belongs to the import-export fashion, so it is not counted here. With it, Bucharest’s year would be 1992, not 2000.

Size goes with the year most closely, distance from Bucharest a little#

For 284 towns and cities (leaving out Bucharest, Ilfov and the five big cities: Cluj-Napoca, Iași, Timișoara, Constanța and Craiova), the year was set against three things: the 2002 population, the distance from Bucharest and the distance from the nearest of the six big cities.

Size goes with the year most closely. Of these 284 towns, half of those over 30,000 people passed the line by 2006, those of 10,000–30,000 by 2011, and those under 10,000 only by 2016 (with 23 still to get there, counted as after 2024). Another 21 towns have too few new firms for a year and are not in the test. Distance from Bucharest adds a smaller but real delay. Distance from the nearest big city adds nothing that stands out from chance.

The county towns, in order#

The 40 county towns that are cities, and Bucharest, by the year one new firm in 20 took an English name (the middle of the first window to pass the line). Ilfov’s county town, Buftea, is a small town and is counted with Ilfov. The five big cities were left out of the test above, because they were the very points the distance was measured from.

What the page does not show#

This page is about firm names, not the language people speak. It does not say that people in a place speak or learned English, and it does not say why names changed: no cause was tested.

Names and addresses are today’s, or the last before a firm was struck off. A firm that moved its office is counted in the new place from the year it was founded, and one that changed its name shows only the new name. This matters most next to big cities: Florești, next to Cluj-Napoca, passes the line in 1998, before Cluj (2003). But of the 28 English-named firms founded there before 2005 whose status is known in the register of October 2025, 14 were still active, against 168 of 547 of its other firms. It may suggest that some moved there later. So the apparent lead of the places around big cities needs further checking; it is not a finding.

Size alone does not decide the year: distance from Bucharest counts too. Differences between regions were not tested, so the page claims none.

Where both counting rules (the three-year window and the window widened to 100 firms) give a year and the two are more than three years apart, the card shows both: this happens in 143 places, 116 of them communes. Where the three-year rule gives no year, the card shows the widened window’s year.

How they were counted#

Firms. The trade register (ONRC, the open file of 19 October 2026, CC BY 4.0): every SRL, SRL-D (a starter limited company) and joint-stock company founded from 1991 to 2025, including those struck off. Each office was placed in one of the 3,181 administrative units by county and locality; Bucharest is one. About 9,000 firms could not be placed and are missing.

Words. A name counts as English when it has at least one of SOLUTION(S), SERVICES, SYSTEM(S), EVENT(S), PROJECT(S), LOGISTICS, CONSULTING, GROUP, HOME, SHOP, TECH: words Romanian does not spell that way. SERVICE (as in “service auto”, a garage), CONCEPT, MANAGEMENT, BUSINESS, SMART and DIGITAL were left out because they are also used as Romanian words. TRADING was taken out because it behaves like IMPEX.

A place’s year. For each year, new firms are taken in a window centred on it: three years, widened to five, seven or nine until it holds at least 100 firms. The place’s year is the first in which at least 5 in 100 firms in the window have an English name. The widened window keeps small towns from looking later just because they have few firms. Where the earlier test’s rule (a fixed three-year window with at least 30 firms and 5 English names) gives a year more than three years away, both years are shown.

What came first. The hypothesis, the word list, the line and the test were written down before any town was scored (6 November 2026). On that rule the order also followed size (−0.60 for population, +0.15 for distance from Bucharest). The widened window, taking TRADING out of each place’s year, and the spans were decided afterwards, before the cards were drawn.

Sectors. The mix of firms does not account for the gap between county towns and communes: in the earlier test (on the list with TRADING, firms of 2005–2025 with authorised activity codes), holding each sector at its national weight keeps 96% of the gap, and communes stay below county towns in construction, trade, transport and professional services.

What is not published. No firm is named, and the year of a place’s first English-named firm is not shown, because it would identify a single firm. The year of the fifth firm is not given exactly: the card says only by which round year (2000, 2005, 2010…) at least five had been founded. The 2021–2025 share appears only where there are at least five English-named firms.

Sources#

1. National Trade Register Office (ONRC): registered firms, 19 Oct 2026 (data.gov.ro, CC BY 4.0)
2. ONRC: authorised activity codes (OD_CAEN_AUTORIZAT), 19 Oct 2026
3. National Institute of Statistics: 2002 and 2021 censuses, population by administrative unit
4. INS: SIRUTA register of localities (kind of unit, county seats)
5. geo-spatial.org: administrative units of Romania (LAU), CC BY-SA 4.0

Related pages

• From IMPEX to SOLUTIONS: the words in Bucharest’s firm names, year by year
• Firms in a flat
• Company towns of Romania