← All posts
Policy & Infrastructure

GatiShakti at Five: India Has Learned to Plan Together. Now It Has to Move Freight Off the Road

Five years of planning numbers

PM GatiShakti National Master Plan turned five on October 13. In 2021 it began with a simple idea: build one digital map that every infrastructure ministry has to plan on, so that a new highway doesn't get built across a planned rail line, or a port expansion doesn't wait years for a road that no one scheduled.

The anniversary figures from PIB show how far the platform has come:

  • 23,493 data layers on one GIS platform.
  • 58 central ministries and departments, plus all 36 states and union territories, onboarded.
  • 396 projects worth Rs 18.66 lakh crore reviewed by the Network Planning Group, which every logistics and connectivity project above Rs 500 crore now passes through.
  • 256 of those projects formally sanctioned, and 198 under construction.

A week before the anniversary, on October 6, the Union Cabinet approved an Integrated Transport and Logistics Authority (ITLA), set up under DPIIT. It will prepare a National Transport Master Plan looking ten or more years ahead across roads, railways, ports, civil aviation, inland waterways, coastal shipping, urban mobility and logistics. It will check whether each transport ministry's five-year and annual plans fit that master plan. It will carry out technical appraisal of government projects above Rs 500 crore and assess their impact after completion. Financial appraisal stays with the existing process.

Together, these show the planning side of the system has matured. India now has a shared map, a review step for large projects and, soon, a body that holds every ministry's plan up against a single long-term one.

But every number in that list measures planning. None of them measures whether it has become cheaper to move goods across India.

What the logistics-cost figure does and doesn't show

The best available figure comes from the Economic Survey 2025-26. Citing a DPIIT-NCAER study, it puts India's logistics costs at 7.97% of GDP in FY24, down from 8.84% in FY23 and 8.79% in FY22. The Survey credits GatiShakti, the Dedicated Freight Corridors, Bharatmala and Sagarmala for the improvement.

That figure needs careful reading. For most of the last decade, the number in official speeches and policy documents was 13-14% of GDP. The NCAER estimate uses a different method. So the fall from 13-14% to under 8% is not a like-for-like gain. Much of it comes from measuring differently, not from moving goods more cheaply.

The year-on-year fall within the NCAER series, from 8.84% to 7.97%, is a more honest signal. Even that rests on a thin base. India has no annual official series on logistics costs, and the LEADS index, published since 2018, relies mostly on how shippers and operators perceive conditions. On the external measure, India ranked 38th of 139 in the World Bank's 2023 Logistics Performance Index, up from 44th in 2018. The National Logistics Policy targets a place in the top 25 by 2030.

So five years in, India can show a better planning process and a falling cost estimate. It can't yet show, year by year and on a fixed method, that the planning is what lowered the cost.

Where freight travels decides the cost

The clearest way to see what drives logistics costs is to look at what each mode charges. KPMG's estimates of the cost of moving one tonne of goods one kilometre:

  • Road: Rs 2.5-3.0
  • Rail: Rs 1.5-1.8
  • Waterway: Rs 1.0-1.2

Road costs roughly twice as much as water and about 60-70% more than rail. Most Indian freight still moves by road. Every percentage point of freight that shifts to rail or water lowers the national cost without anyone needing a new subsidy.

The difficulty is that freight demand is growing 7-8% a year. If most of that growth stays on trucks, logistics costs can rise again even with better planning, because the cheaper modes don't gain share. Better coordination stops ministries building things that clash. It doesn't by itself move a container from a truck to a train.

That depends on things the master plan maps but doesn't deliver on its own: rail sidings that reach industrial parks, terminals at the right places on the freight corridors, inland waterway berths with road and rail links at both ends, and tariffs and turnaround times that make rail and water a sensible choice for shippers who are used to trucks.

A different scorecard for the next five years

GatiShakti's first five years were judged, reasonably, by adoption: how many ministries joined, how many data layers were added, how many projects were reviewed. Those were the right measures for building a shared system. They are the wrong measures for whether the system works.

The next five years need a scorecard built on results:

  • Modal share. The share of freight tonne-kilometres moving by rail, coastal shipping and inland waterways, reported every year. This is the figure most closely tied to cost.
  • An annual logistics-cost series. A yearly estimate on a fixed method, published by the government, so that FY25, FY26 and later years can be compared with FY24 without arguments over methodology.
  • Project outcomes. ITLA's post-completion impact assessments, made public, showing whether reviewed projects delivered the traffic, travel times and costs promised when they were approved.
  • Last-mile links. How many industrial parks, ports and freight terminals have working rail or waterway connections, not just planned ones.

ITLA's mandate already includes impact assessment and a ten-year master plan. Whether India gets this scorecard depends largely on whether those assessments are published and whether the master plan sets modal-share targets that can be checked.

What five years have shown

GatiShakti answered a question that held Indian infrastructure back for decades: whether ministries that each own a piece of the transport system can plan on the same map. Five years in, the answer is yes.

The question for the next five years is narrower and harder. India needs to show, on numbers that are published every year and measured the same way each time, that the planning is moving cargo off the road and making it cheaper to move goods. Until that series exists, GatiShakti's results will keep being measured in data layers, and that is the wrong unit.