India's roads carry 71% of the country's freight and 85% of its passenger traffic, and the government spent 1.22 lakh crore rupees building and maintaining them in the year just allocated. India also lost roughly 1.7 lakh people to road accidents in 2024, more than any other country on earth, on a network the government has spent the past decade racing to expand at record pace. Both facts describe the same infrastructure. Understanding whether India's road spending has actually paid off requires putting a number on every part of that ledger, not just the parts that make for a good ribbon cutting photo.
Chapter 1What does a kilometre of Indian highway actually cost to build?
The baseline figures NHAI itself uses are specific. A four lane highway costs roughly 8 to 9 crore rupees per kilometre, including land acquisition, and a six lane highway runs closer to 14 crore rupees per kilometre. Expressways cost considerably more once you move beyond a standard highway. The Delhi Mumbai Expressway, at 1,386 kilometres and a total budget exceeding 1 lakh crore rupees, works out to roughly 72 crore rupees per kilometre, nearly five times the basic four lane rate, reflecting the access controlled design, larger land footprint, and higher grade construction that comes with an expressway built for sustained high speed traffic.
A genuinely reliable apples to apples comparison against other countries' per kilometre costs is harder to pin down than it looks, since land acquisition costs, terrain, labour costs, and design standards vary enormously between projects even within India, let alone across countries, and most international comparisons circulating online do not disclose enough methodology to trust. What is verifiable is the scale of India's own ambition. NHAI's FY26 pipeline alone covers 2,188 kilometres at a combined capital cost of 1.15 lakh crore rupees, and the country has been adding highway at a pace of roughly 34 kilometres a day in FY24, a figure the government cites often as a marker of progress.
Chapter 2Is kilometres per day actually the right thing to be measuring?
This is where the story gets less flattering, because a headline construction pace and a durable road are not the same achievement, and India's own accident data shows the gap between the two widening rather than closing.
Pothole related road fatalities rose from 1,555 in 2020 to 2,385 in 2024, a 53% jump over five years, according to Lok Sabha data compiled by the Ministry of Road Transport and Highways. Pothole related accidents rose from 3,713 to 5,432 over the same period. Total pothole linked deaths between 2020 and 2024 reached 9,438, an average of nearly five deaths a day, on infrastructure the government has simultaneously been expanding at record pace. Reactive patchwork repairs, rather than proper scientific resurfacing, are the pattern MoRTH's own reporting attributes much of this to, with poor drainage and waterlogging accelerating pothole formation after heavy rainfall, a cycle that repeats every monsoon season on roads that were, in many cases, only recently built or resurfaced.
What do bad roads actually cost in lives, and where specifically?
India recorded approximately 1.7 lakh road deaths in 2024, the highest total of any country in the world. The distribution of where those deaths happen is the more damning detail. National highways make up only about 2% of India's total road network by length, yet accounted for roughly 30% of all road fatalities, with 29,018 deaths recorded on national highways alone in just the first half of 2025. The Ministry's own black spot classification, a 500 metre stretch recording at least five accidents and ten or more fatalities within three years, has identified 13,795 such locations across the national highway network, concentrated heavily in Punjab, West Bengal, Tamil Nadu, and Karnataka. Total road accidents climbed from 3.7 lakh in 2020 to 4.8 lakh in 2023, broadly tracking the pace of vehicle ownership growth over the same period.
The demographic pattern compounds the economic cost well beyond the immediate tragedy. Roughly 68% of male victims and 58% of female victims fall between 18 and 45 years old, the country's most economically productive age bracket, and the International Labour Organization's modelling used in recent South Asian road safety analysis estimates that each road fatality represents an average loss of 32 productive working years.
Chapter 4So what does this actually cost the economy, in a single number?
Government and World Bank estimates converge on a range of 3% to 5% of GDP lost to road accidents annually, a figure repeated consistently enough across independent assessments to treat as directionally solid even though the precise methodology varies between sources. Breaking that headline number down into its components matters more than the top line figure itself. Lost productivity, driven overwhelmingly by working age fatalities and the ILO's 32 year loss calculation, is by far the largest single component, estimated at roughly 58 billion dollars a year on its own, ahead of direct costs like medical treatment, emergency response, vehicle damage, and insurance claims combined.
That comparison is worth sitting with directly. The productivity lost to road accidents in a single year runs close to four times what the government spent building and maintaining the entire national highway network the same year. Even treating both figures as rough estimates rather than precise accounting, the scale gap between the investment and the ongoing loss is too large to be a rounding error.
Chapter 5Has all this construction actually lowered the cost of moving goods, the thing it was supposedly for?
This is the one place the data tells a genuinely more encouraging story than the accident numbers, though it comes with an honest caveat about how reliable the most commonly repeated version of this story actually is.
For years, India's logistics cost was cited almost universally as 13% to 14% of GDP, a figure that became the standard talking point for why India was uncompetitive against manufacturing rivals. That widely repeated number traces back largely to a 2015 AVALON Consulting study commissioned by the Confederation of Indian Industry, which used a narrower and less rigorous methodology than what has followed since. The first genuinely systematic government backed study, released in 2025, put the real figure at 7.97% of GDP, and NCAER's own detailed 2023 analysis, built from Supply and Use Tables rather than industry survey extrapolation, arrived at a comparable range of 7.8% to 8.9% for 2021-22. The most recent government assessment for 2026 places the number at roughly 8%.
That gap does not mean the 13% to 14% figure was invented out of nothing, or that logistics costs have not genuinely fallen with GST, FASTag, and highway expansion all playing a real role. It means the number most commonly cited in the media and in policy debate for the better part of a decade was built on weaker foundations than the confidence with which it was repeated, and the real improvement, from a properly measured baseline closer to 8.9% down toward 7.97%, is a smaller and more modest win than the popular "cut logistics costs nearly in half" framing suggests.
Chapter 6What's still broken, even with that genuine improvement?
Fuel accounts for 42.1% of road freight expenses in India, worsened by steep diesel taxation and time lost at toll plazas, and the deeper structural problem sits in how little use India gets out of the trucks it already owns. An Indian truck covers roughly 250 to 300 kilometres a day. An American truck covers 700 to 800 kilometres a day. A Chinese truck covers more than 500. That utilisation gap, not the headline logistics cost percentage, is where KPMG's own 2026 analysis argues the next real efficiency gain sits, since every 10 percentage point improvement in truck utilisation is estimated to cut long haul freight rates by 6% to 7%.
That inefficiency does not land evenly. Businesses with turnover below 5 crore rupees spend close to 17% of their output value on moving goods, more than double the 7.6% large firms with turnover above 250 crore rupees manage to keep logistics costs down to, meaning the same road network that has genuinely gotten more efficient at the aggregate level is still extracting a disproportionately larger toll from the smallest businesses least able to absorb it.
Chapter 7Who is actually paying for all of this, and how much?
Toll revenue is the most direct answer to who funds the newer, better maintained stretches of this network. NHAI collected over 41,000 crore rupees in toll revenue in FY 2024-25, a 12.5% year on year rise to an all time high, with projections reaching 1.3 lakh crore rupees by 2030 as FASTag coverage and road monetisation both expand. FASTag adoption now covers 97% of vehicles, collecting toll electronically across more than 1,150 fee plazas, and the Build Operate Transfer model that funds a large share of expressway construction means users are, in effect, paying twice, once through the fuel taxes and cess that fund general road budgets, and again directly through tolls on the specific stretches built to recover private capital investment faster.
A car travelling the full length of the Delhi Mumbai Expressway once it is fully operational will pay close to 2,950 rupees in toll for a single one way trip across 1,386 kilometres, a direct, itemised charge on top of whatever fuel and vehicle costs the same journey already carries, and one that a road user pays regardless of whether the specific stretch beneath their wheels happens to be one of the well maintained new sections or an older, pothole prone one nearby.
Chapter 8So, has India's road spending actually been worth it?
The honest answer is genuinely split down the middle of this piece's own numbers, not evenly, but split. The logistics cost story is a real, if more modest than advertised, success: costs have fallen from a properly measured baseline near 8.9% of GDP toward roughly 8%, a meaningful gain even after discounting the inflated 13% to 14% figure that dominated public debate for a decade. The construction pace, 34 kilometres a day, is a genuine engineering and execution achievement by any global standard. Against that sits an accident toll that costs the economy 3% to 5% of GDP every single year, with lost productivity alone running at four times the entire annual roads and bridges budget, and a pothole death count that rose 53% over exactly the five years the government was building highway at record speed. India has gotten significantly better at building roads fast. The unresolved question, visible in its own accident data rather than in any outside critique, is whether it has gotten anywhere near as good at keeping them safe once built, and on the numbers available, that second achievement is running well behind the first.