Why Bulgaria's Narrow Roads Cost It 6% of GDP Author: Дністер Published: 2026-08-27T03:01:53.000Z Language: en URL: https://neurodrift.org/en/blog/ekonomika-dorih-1-1/ Original (Ukrainian): https://neurodrift.org/blog/ekonomika-dorih-1-1/ Tags: infrastructure, Bulgaria, economics, roads, logistics, EU A numbers-first breakdown: what an economy loses when there is no second lane between its cities. Bulgaria in figures (the Hemus motorway under construction since 1974, the EU's poorest region, 74 road deaths per million), the methodology for converting time into money, a micro-calculation for a single bridge, and an eight-country case table — from Poland to Norway. ----- .tbl-breakout{overflow-x:auto;margin:1.7em 0;border:1px solid var(--border);border-radius:10px;background:var(--bg-elevated);-webkit-overflow-scrolling:touch} .tbl-breakout table{width:100%;border-collapse:collapse;font-size:.88em;margin:0} .tbl-breakout th{font-size:.72em;text-transform:uppercase;letter-spacing:.08em;color:var(--accent);text-align:left;padding:.7em .85em;border-bottom:1px solid var(--border);vertical-align:bottom} .tbl-breakout td{padding:.6em .85em;border-bottom:1px solid var(--border);vertical-align:top;line-height:1.45} .tbl-breakout tr:last-child td{border-bottom:none} .tbl-breakout .nw{white-space:nowrap} @media (min-width:1000px){.tbl-breakout{margin-left:-28px;margin-right:-78px;width:calc(100% + 106px)}} @media (min-width:1400px){.tbl-breakout{margin-left:-44px;margin-right:-150px;width:calc(100% + 194px)}} @media (max-width:640px){.tbl-breakout table{font-size:.82em}} The OSRM routing engine, which computes travel time by road class with zero traffic, gives this estimate: Sofia → Ruse, 309 km — 4 hours 24 minutes. Average speed: about 70 km/h, and that is in an ideal world without trucks, tractors or roadworks. In the real world — slower. This is not a «bad road» in the pothole sense: the asphalt may be decent. It is a 1+1-class road — one lane there, one back, an overtaking window every few kilometers. That is what most of Bulgaria's intercity network looks like — and it is this configuration, not surface quality, that makes the country drive slowly. What it costs can be counted. The numbers are below. 51 yearsthe Hemus motorway Sofia–Varna has been under construction: started 4 October 1974, promised finish — 2029–2030 ~70 km/haverage speed Sofia→Ruse (309 km in 4:24, OSRM, zero traffic) up to €6.5bn/yrestimated losses from weak transport connectivity — ≈6% of 2024 GDP 74 / millionroad deaths per million residents in 2024 — 2nd place in the EU against an average of 44 01 · How many roads Bulgaria has — and how many are real Bulgaria's national road network at the end of 2025 runs to 20,028 km. Of that, motorways make up just over 900 km: 879 km in 2024, about 917 km by November 2025 after several Hemus sections opened. Expressways as a class barely exist — mostly on paper. In other words, over 95% of the network has no second lane: one each way, overtaking into oncoming traffic wherever markings and nerves allow. Geography is even more eloquent than arithmetic. The country's only completed motorway is Trakia (Sofia–Plovdiv–Burgas), the southern corridor. The North — half the country along the Danube — has no motorway at all: Hemus (Sofia–Varna, 420 km as designed) is roughly half-built after half a century, with about 220 km of the remainder not even started. There is no motorway to Ruse — the country's fifth city and its main land gateway to Romania. None to Vidin either. The result shows in Eurostat: Severozapaden (Bulgaria's North-West) is the poorest region of the entire European Union, at about 42% of the EU average GDP per capita in purchasing power terms (2023). Roads are not the only cause, and an honest analysis must say so: demography, geography and institutional quality all pull the same way. But Bulgaria's business press flatly calls the missing infrastructure «the root of the enormous economic imbalance» between North and South — and the correlation «where the motorway is, there the economy is» reads off the country's map without glasses. !Aerial view of a 1+1 road: a column of cars trails behind a truck while the oncoming lane sits empty 02 · How fast you actually go: two routes, one router To avoid arguing about feelings, we use one instrument — the OSRM router on OpenStreetMap data. It knows nothing of congestion and computes time by road class: motorway — fast, 1+1 through villages — slow. This is pure «speed the infrastructure allows». Route Distance Time (no traffic) Average speed Comment Sofia → Ruse309 km4 h 24 min~70 km/hno motorway; a classic 1+1 corridor Sofia → Varna (fastest route)524 km5 h 39 min~93 km/hthe router takes a longer arc via motorway sections Sofia → Varna (direct Hemus alignment, if completed)420 km≈4 h 00 min~105 km/hcalculated: 420 km at average motorway speed Note the second row: to go «fast» from Sofia to Varna you must drive a hundred kilometers farther than the unfinished Hemus's direct alignment. Speed is bought with distance — and it still takes over five hours for 420 km of straight road. Comparing rows three and two prices the unfinished corridor: roughly 1.5–1.7 hours per trip Sofia–Varna. Now multiply that by the law. EU Regulation 561/2006 caps a truck driver at 9 hours of driving per day (twice a week — 10), with a mandatory 45-minute break after 4.5 hours. A truck's operating radius is not kilometers — it is hours multiplied by speed: Average speed Legal daily range (9 h) What it means 50–60 km/h (the 1+1 network)450–540 kmSofia→Varna — nearly a full daily quota one way 80–90 km/h (a motorway network)720–810 kmSofia→Ruse→Sofia (624 km) — legal in a single day Same driver, same truck, same law — but on a 1+1 network the logistics company gets a third less radius per day. A Sofia–Ruse–Sofia run at 85 km/h is 7.3 hours at the wheel — legal. At 55 km/h it is 11.3 hours: illegal even with an extended day. Where a fast network allows a same-day turnaround, a slow one demands two days, an overnight stay and a second driver. That is not a metaphor — it is direct cost in wages, per diems and tied-up capital. 03 · The physics: why 1+1 breaks long before it «fills up» The Highway Capacity Manual — the standard engineering reference — gives a two-lane road a theoretical ceiling of about 3,200 vehicles per hour in both directions combined. But degradation starts much earlier, and not from «filling up» — from the disappearance of overtaking opportunities: one truck at 60 km/h plus an oncoming stream is enough for a column to grow behind the truck, and percent-time-following (the share of time you are locked behind someone slower) shoots up. Engineering practice holds that a two-lane road needs widening already at 10,000–15,000 vehicles per day — long before its formal capacity ceiling. That is why «adding overtaking windows» is a palliative: a window works only while oncoming flow is light. On a loaded corridor, a 1+1 road effectively travels at the speed of its slowest participant. Average 50–70 km/h on intercity corridors is not about drivers — it is about geometry. 04 · How time converts into money: the methodological rails Transport economics prices a slow road via three main line items. People's time. In European project-appraisal practice (the HEATCO methodology / European Commission guidance, in 2024 prices per the EUROCONTROL compilation), an hour of personal travel is valued at roughly €22, a business hour at about €55. Bulgaria's national values are lower (time value scales with incomes), but the order of magnitude stands: an hour in a column behind a truck is tens of euros of vanished value per car. Trucks' time and kilometers. A European truck costs its operator roughly €1.45–1.70 per kilometer in full cost terms, or €40–100 per hour depending on country and haul type. The World Bank, on global data, estimates that routing via a motorway instead of a lower-class road cuts transport cost by about 19%. Crashes. Road crashes cost the EU about 2% of GDP a year (≈€280 billion under the SafetyCube methodology; the cross-country range is 0.4–4.1% of GDP). The value of a statistical life used in European practice is about €4.7 million in 2024 prices. We return to this in section 08 — for Bulgaria it is no abstraction. Honesty frameAll three line items are methodological rails, not a verdict from a calculator. Time values are EU averages; a truck-hour's cost depends on the contract; but even at the most conservative values, the bill for a slow network runs to hundreds of millions a year for a country of Bulgaria's size. Below — two levels of calculation: micro and macro. 05 · The micro-calculation: one bridge and one corridor The Ruse–Giurgiu bridge. Along ~470 km of the shared Bulgarian-Romanian Danube border there are two road bridges: the Friendship Bridge at Ruse–Giurgiu (1954) and Vidin–Calafat (2013). A third is promised no earlier than 2035. About 1,300 trucks a day cross at Ruse leaving Bulgaria; the truck queue runs 3–6 hours in a normal period, 8–10 at peaks, with recorded cases over 12. !A truck queue before a Danube bridge at dusk: red taillights stretch to the horizon, a driver asleep in his cab Simple arithmetic (assumptions: 1,300 trucks/day × 3–6 hours of queue × €40–60 per truck-hour): €156–468 thousand of losses per day, i.e. €57–171 million a year — trucks only, the queue only, one bridge only. Excluding cars, wrecked delivery schedules, and the cost of a second driver. That is the price of two EU countries adding not a single new bridge across their shared river since 2013. The Sofia–Varna corridor. The unfinished Hemus costs ~1.5–1.7 hours per trip (section 02). For a truck that is €65–100 of extra cost per run; for a family car — over three person-hours per round trip. Multiply by decades and by everyone driving between the capital and the country's third city, and it becomes clear why the national-scale estimates land in the billions. 06 · The macro level: how much in GDP Bulgaria's most-cited estimate belongs to Prof. Christina Nikolova (University of National and World Economy, UNWE), who calculated at the Sofia Economic Forum, using the EU «Handbook on the external costs of transport» methodology: Bulgaria loses about €6.5 billion every year to weak transport connectivity with its neighbors and the state of its roads. Bulgaria's 2024 GDP was 204.9 billion leva ≈ €105 billion — so the figure means ≈6% of GDP annually. An important caveat: this is one researcher's estimate, it covers transport connectivity broadly (not only «the missing second lane»), and there is no independent government verification of the number. But even if the real figure is half or a third of it — that is still billions a year. The academic literature offers more cautious but stable rails. The Bom & Ligthart meta-analysis (578 estimates from 68 studies) puts the output elasticity of public capital at 0.08–0.12: a 10% increase in road capital is associated with roughly +1% of GDP in the long run. An independent line (Calderón, Moral-Benito, Servén) — 0.07–0.10. The IMF, in a classic World Economic Outlook chapter: +1 percentage point of GDP in public investment yields ≈+0.4% of output the same year and ≈+1.5% within four years in advanced economies. This is not «infrastructure as a magic wand» (see section 07 on the US and Norway), but it is a consistently positive, measurable effect. There is a third dimension — logistics competitiveness. In the World Bank's Logistics Performance Index (2023), Bulgaria and Romania share last place among EU countries (51st globally, scoring 3.2 against Germany's 4.1). For economies selling themselves as a nearshoring platform between Europe and the Middle East, this is not a decorative ranking: it is read by the same people who decide where to put the factory. 07 · Who has already done the math: eight countries Country / project What was measured Figure Source / status Romania (anti-case)losses from missing motorways~2% of GDP a year; 1,418 km of network against a ~3,000 km target by 2030Pro Infrastructura estimate (advocacy); Mercedes reportedly chose Kecskemét (Hungary) over Romania partly for infrastructure Poland (counter-example)the build-out effect after EU accession784 km (2004) → ~5,222 km (2025); 20.2% of all EU road-freight tonne-km; 32.7% of the EU's international haulageEurostat, fact; official estimate: ~16% of the gap to the EU closed thanks to cohesion policy India, Golden Quadrilateraleconometrics: firms along the routeraw-material inventories −6–12 days of production; new plants clustering within 0–10 km of the routeDatta (J. Dev. Economics 2012); Ghani–Goswami–Kerr (Economic Journal 2016) — the cleanest causal identification USA, Interstatesocial return on road capital~35%/yr in the 1950–60s → ~10% in the 1980s; contribution to productivity: 31% → 7–8%Nadiri–Mamuneas (FHWA); Fernald (AER 1999): a one-off effect — «a second Interstate would not pay off» Greece, Egnatia Odostravel time across the country's north670 km for €5.93bn; Alexandroupoli–Igoumenitsa: 11.5 h → 6 h 10 min (−46%)fact (project completed 2009) Ireland, the M networkintercity times after the 2005–2010 boomDublin–Cork: 4+ h → under 3; Dublin–Limerick: ~3 → ~2 h; one M7-corridor estimate: +€525m GDP/yrfact (times); the GDP estimate — a single study, handle with care Croatia, A1coastal accessibilityZagreb–Split: 5+ h → ~3.5 h; tourism ≈20% of GDPfact (time); the tourism link is directional, no clean causal estimate Norway, E39 (counterpoint)an honest megaproject CBA≈€36.5bn planned to halve a 21-hour drive; TØI's calculation for the Bokn–Klett section: minus NOK 56bnofficial appraisal — proof that not every road pays off Reading the table with an honest eye yields three conclusions. First: where the econometrics is careful (India, the US), the motorway effect is real and large — but largest where there was no network. The US earned 35% annual returns in the 1950s and 10% by the 1980s; Norway, which already has excellent roads, gets a negative appraisal for its new megaproject. Second: Bulgaria and Romania sit precisely in the «Indian» part of the curve, where returns are highest because the base network is missing. Third: Poland is the closest cultural-institutional proof that this works inside the EU within a single generation: in 20 years — from 784 km to 5,222 km, and to the status of Europe's logistics superpower with 43.7% of all EU cabotage. !A Polish motorway with a stream of freight trucks at sunset — the symbol of a logistics leap 08 · Blood as a balance-sheet item In 2024, 478 people died on Bulgarian roads — 74 per million residents, second place in the EU after Romania (78), against a European average of 44. Police recorded 7,173 crashes and 9,054 injured. A 1+1 road has a specific death profile: the head-on collision during overtaking — precisely the crash type that is structurally impossible on a road with separated directions. There is no official statistic on the share of head-on collisions in Bulgaria in open sources (telling in itself), but the Bulgarian media write the same thing after every high-profile overtaking crash. Apply the European methodology (crashes cost EU countries an average of 2% of GDP, up to 4.1% for the worst) and Bulgaria's bill for deaths alone runs to hundreds of millions of euros a year: 478 lives × €4.7 million per statistical life ≈ €2.2 billion of social losses, before the injured. Even scaling the VSL down two- or threefold for Bulgarian incomes, this is billion-scale — every year. 09 · The Swedish compromise: 2+1 as the cheap half of the answer Here the article must take an honest turn: a full motorway is not the only answer, and not always the right one. Since the late 1990s Sweden has been mass-converting dangerous two-lane roads into the 2+1 scheme with a cable barrier: three lanes, the middle one alternating direction every 1–2.5 km, and — the main thing — opposing flows physically separated. The research results: a 50–80% drop in fatalities (the flagship estimate for roads converted in 1998–2007 — minus 79%), at a conversion cost several times below building a full motorway — often just re-marking plus a barrier on an existing 13-meter roadbed. !A Swedish 2+1 road with a central cable barrier: the two flows physically separated For a network carrying 3,000–20,000 vehicles a day (which is most Bulgarian intercity corridors), 2+1 is the engineering-appropriate answer: it delivers no motorway speed, but it removes head-on collisions and guarantees regular overtaking. The right hierarchy for a poor country with a limited budget looks like this: motorways on the 3–4 main corridors with the heaviest traffic, 2+1 on everything else. The Bulgarian paradox is that the country does neither properly — even though a kilometer of motorway here is the cheapest in Europe (≈€3.1m/km against ≈€6.5m on average in Romania and up to €40m on difficult terrain). 10 · Why half a century of not building: Hemus as a diagnosis Construction of the Hemus motorway began on 4 October 1974. By Bulgaria's EU accession (2007), 33 years in, 170 km had been built. On the project's fiftieth anniversary, in October 2024, the state news agency wrote honestly: «still only half-built». The ministry's current promise is end-2029 to early 2030; of the remaining ~220 km, construction has not started on a single section, with permits issued for only ~36 km and works to begin in 2026. The money, meanwhile, was there. A 2021 audit found the Road Infrastructure Agency had handed the state company Avtomagistrali contracts worth nearly 4 billion leva (≈€2 billion) without tenders; then the classics: advances to contractors for works never started (343 million leva to one company), and 60 million leva, per investigative reporting, funneled through front firms and carried out in cash «in sacks and bags». The prosecutor who prepared a 400-page indictment in the «Hemusgate» case was removed from it in November 2023 — one step before trial. The indictment never saw a courtroom. In parallel runs chronic underfunding of the routine: over 35% of second- and third-class roads (7,000+ km) are in poor or unsatisfactory condition; the August 2026 vignette price hike of 30% will bring ≈€54 million a year — against needs the industry estimates an order of magnitude higher, and with the industry's own skepticism that the money will reach the asphalt at all. And where building finally does begin, projects can run into legitimate conflicts: the final section of the Struma motorway through the Kresna Gorge spent 28 years unresolved with environmentalists — and even the official cost-benefit analysis for the through-route variant came out negative (B/C 0.79). Not every kilometer one wants to build is worth building — but in Bulgaria, for decades, even the worthwhile ones don't get built. DiagnosisBulgaria's problem is not money (the kilometer here is Europe's cheapest, and EU funds exist), not technology, and not terrain. The problem is institutional capacity: procurement corruption, a broken «plan→tender→oversight→court» cycle, and a political horizon shorter than a construction cycle. A road under construction for 51 years is not an engineering object. It is a monument to a model of governance. 11 · Conclusions A 1+1 network is not «a bit slower». It is an average of ~70 km/h instead of ~105 on a motorway, a third off every truck's legal daily radius, queues at a single bridge costing €57–171 million a year, and second place in the EU for road deaths. The honest macro conversion looks like this: the most aggressive estimate — up to €6.5 billion/year (≈6% of GDP) for everything combined; the conservative academic rails (elasticity 0.08–0.12, returns highest exactly where the network is missing) still yield a billion-scale order. An exact number does not exist — but its order of magnitude is beyond doubt. The best evidence is not rhetoric but the neighbors: Poland turned 784 km into 5,222 km in 20 years and took a fifth of all EU road freight. Romania, which did not, pays an estimated ~2% of GDP a year and watches factories go to Hungary. The answer is not «pave everything with motorways»: the American and Norwegian data show returns fall with saturation. The answer is motorways on the main corridors plus Swedish 2+1 on the rest of the network: 50–80% fewer deaths at a fraction of the price. Bulgaria's bottleneck is neither budget nor terrain but institutions: Hemus has been under construction for 51 years not because it is hard, but because it is possible. Sources BTA: Bulgaria's Road Network Totalling 20,028 Km (2026) — national network length BTA: 50 Years after 1974 Groundbreak, Hemus Still Halfway Built (2024) — Hemus history, 170 km by 2007 BTA: Hemus Expected Complete by End-2029/Early-2030 (2025) — ministry promise; 220 km not started bne IntelliNews: How Millions Were Siphoned off Hemus Construction — BGN 4bn without tenders; advances Anti-Corruption Fund: Hemusgate (2024–2025) — cash «in sacks», removed prosecutor Eurostat: Regional GDP per capita (2023 data) — Severozapaden ≈42% of EU average Sofia Globe: Bulgaria EU's 2nd-Highest Road Death Rate 2024 — 74/million vs 44 EU; 478 deaths 24 Chasa: Prof. K. Nikolova's (UNWE) estimate — €6.5bn/yr; EC Handbook methodology NSI: Bulgaria GDP 2024 — BGN 204.9 bn — base for ≈6% of GDP friendshipbridge.eu: Danube Bridge Crossing 2025 — truck queues 3–6 h (peak 8–10); 2 bridges; third ≥2035 BTA: ~1,300 Trucks/Day at Ruse–Giurgiu — bridge traffic Regulation (EC) 561/2006 — 9 h of driving/day; breaks OSRM / OpenStreetMap routing (queries August 2026) — Sofia–Ruse 309 km/4:24; Sofia–Varna 524 km/5:39 — zero-traffic calculation EUROCONTROL Standard Inputs: Value of Travel Time — €22/h personal, €55/h business (HEATCO, 2024 prices) European Road Safety Observatory: Consequences of Crashes (2023) — crashes ≈2% of EU GDP; €280bn Bom & Ligthart, J. of Economic Surveys (2014) — meta-analysis of 578 estimates: elasticity 0.08–0.12 IMF WEO Oct 2014, Ch. 3: Is It Time for an Infrastructure Push? — public-investment multipliers Datta, J. of Development Economics (2012) — Golden Quadrilateral: inventories −6–12 days Ghani, Goswami & Kerr: Highway to Success (2016) — GQ: firm location and productivity FHWA: Productivity and the Highway Network (Nadiri–Mamuneas) — Interstate returns 35%→10% Fernald, AER (1999): Roads to Prosperity? — the one-off nature of the Interstate effect Notes from Poland: 20 Years in 20 Charts (2024) — 784 km (2004) → ~5,000+ km Eurostat: Road Freight Transport Statistics (2025–2026) — Poland: 20.2% of EU tonne-km; 32.7% international; 43.7% cabotage VTI: Traffic Safety Effects of Narrow 2+1 Roads in Sweden — −50–80% fatalities (−79% flagship estimate) Novinite: Bulgaria Builds Cheapest Highways in Europe — ≈€3.1m/km Romania-Insider: Highway Costs RO vs BG — RO ≈€6.5m/km, up to €40m Reuters: Romania's Roads to Nowhere — the ~2% of GDP estimate (Pro Infrastructura, advocacy) CEE Bankwatch: Struma Motorway CBA — ENPV −€239m, B/C 0.79 (through-route via Kresna) Novinite: Vignette +30% from Aug 1, 2026 — +€54m/yr; 35% of class 2–3 roads in poor condition World Bank: Why Transport Costs Are Stifling Growth — motorway instead of lower-class road: −19% transport cost Honesty and confidence limitsThe numbers carry different weight: network lengths, the Hemus timeline, mortality, Regulation 561/2006, Poland's Eurostat statistics — hard facts. The €6.5bn/yr estimate is one researcher's calculation (the broadest definition of losses); Romania's ~2% of GDP is an industry association's advocacy estimate; the bridge micro-calculation and truck radii are this article's own arithmetic with explicitly stated assumptions (€40–60/truck-hour, 3–6-hour queues). OSRM times are zero-traffic calculations: real times are worse. No exact official statistics exist for the share of head-on collisions or the share of the 1+1 network — both figures in the text are derived, not cited.