Why Programs Fail: What India's Monitored Projects Show
Programs fail by running late and over budget, and India's monitored central projects show many delays trace to land, clearances, financing and scope decisions. Eight pass-or-fail questions test a live programme.
By Sanjeev Kumar, Founder · Updated 19 Nov 2026

Key takeaways
- The delay causes MoSPI records are mostly decisions and dependencies outside the delivery team (our inference; MoSPI does not rank them).
- Delay vanishes from headlines when schedules are reset. In December 2023, 210 of 848 delayed projects stopped counting as delayed once measured against the latest schedule.
- The Project Management Institute's (PMI) 2016 survey found 54% more projects met original goals where benefits were identified before the start.
- Eight pass-or-fail questions test your own programme. Start with whether the original finish date is still on your status report.
In this article
- 1Why do programmes fail, judging by India's own project data?
- 2Which causes sit behind the delay in a failing programme?
- 3How does re-baselining hide programme delay?
- 4How can you test whether your own programme is heading the same way?
- 51. Does every status report show the original finish date?
- 62. Are benefits written down with an owner and a date?
- 73. Is every outside dependency closed before the work that needs it?
- 84. Does every pending approval have a named decision-maker and a due date?
- 95. Is every scope change approved and priced?
- 106. Does every component have a start and a finish date?
- 117. Is the sponsor accountable for benefits, not only delivery dates?
- 128. Is there a scheduled gate where the programme can be stopped?
- 13What do we teach about stopping programme failure early?
- 14Frequently asked questions
A delayed project can stop counting as delayed without moving any faster. In December 2023, 210 of 848 delayed central projects dropped off the list once their schedules were reset (our subtraction from the flash report of India's Ministry of Statistics and Programme Implementation, MoSPI).
That matters to anyone who runs or joins a programme: a set of related projects managed together to deliver benefits that none of them would deliver alone. This page covers programme management, not Andreas Zeller's debugging book of the same name. MoSPI monitors central-government projects costing ₹150 crore or more, and in March 2024 779 of 1,873 were behind their original schedule.
The pattern is twofold: schedule resets hide delay, and the causes agencies record mostly sit in decisions outside the delivery team (our reading). Below are the data, the check that exposes the reset, and eight pass-or-fail questions for your programme.
Why do programmes fail, judging by India's own project data?
Programmes fail by running late and over budget, and India's central-government record shows it at scale. In March 2024, 779 of 1,873 monitored projects were behind their original schedule, and anticipated cost exceeded original cost by 18.65% in aggregate.
MoSPI's Infrastructure and Project Monitoring Division publishes a monthly flash report on central-sector projects (those run or funded by India's central government) costing ₹150 crore (₹1.5 billion) or more. Six reports, side by side:
| Flash report | Projects monitored | Delayed against original schedule | Delayed share (our calculation) | Average delay of delayed projects, months | Cost overrun on original cost |
|---|---|---|---|---|---|
| December 2020 | 1,687 | 558 | 33.1% | 45 | 19.96% |
| December 2021 | 1,673 | 557 | 33.3% | 45.69 | 19.82% |
| December 2022 | 1,438 | 835 | 58.1% | 40.5 | 22.12% |
| December 2023 | 1,820 | 848 | 46.6% | 36.59 | 18.65% |
| March 2024 | 1,873 | 779 | 41.6% | 36.04 | 18.65% |
| 1,817 | 831 | 45.7% | 35.1 | 20.70% |
The delayed share is our calculation: delayed projects divided by projects monitored in each report. Each report covers a different set of projects, as projects are added and completed every month, so read the share as a range of 33% to 58%, not a trend. The December 2023 overrun was ₹4,82,530 crore on an original cost of ₹25,87,066 crore.
Delay was deep as well as common. Of the 848 delayed projects in December 2023, 323 (38%) were 25 to 60 months late and 123 (15%) were more than 60 months late. It also varied by sector: power had 59 of 98 projects delayed (60%), petroleum 91 of 157 (58%), roads 450 of 1,021 (44%) and railways 105 of 250 (42%).
These are infrastructure projects rather than programmes in the formal sense. We use them because they are the largest public record of how multi-party delivery slips. Whether business-change or IT programmes behave the same way is our inference, not something this data shows.
Which causes sit behind the delay in a failing programme?
Most of the delay causes MoSPI records are decisions or dependencies that someone outside the team must close (our reading; MoSPI does not rank them). The December 2023 flash report lists these reasons for delay, among others: land acquisition, forest and environment clearances, tie-up of project financing, finalisation of detailed engineering and change in scope, tendering, ordering and equipment supply, technical approval, contractual issues, litigation before the National Green Tribunal, geological surprises and COVID-19 lockdowns.
Our reading is that most of that list consists of items a delivery team cannot recover by working faster. Land, clearances, financing, engineering sign-off, scope and tendering each need a decision from someone outside the team. Geological surprises and lockdowns are the exceptions. The reasons are reported by the implementing agencies and are not ranked, so the list shows which causes occur, not which is largest.
A separate study points the same way. In a 2016 survey of 1,189 practitioners, the Project Management Institute (PMI) found that 54% more projects met their original goals and business intent where benefits were always or often identified before the project started. That is a relative difference, not a success rate. The same report found that 83% of organisations described their benefits realisation as immature, and that fewer than 40% of respondents said project managers were accountable for identifying benefits. It is one survey and ten years old, so we treat it as supporting evidence, not proof.

How does re-baselining hide programme delay?
Re-baselining makes delay disappear from the headline count. A baseline is the approved schedule and cost a programme is measured against, and re-baselining resets it to a later date. The delay still exists, but a report measured against the latest schedule no longer shows it.
MoSPI's reports give both counts, so the effect can be calculated:
| Flash report | Delayed against original schedule | Delayed against latest schedule | Projects that stop counting as delayed | Share of delayed projects (our calculation) |
|---|---|---|---|---|
| December 2021 | 557 | 420 | 137 | 24.6% |
| December 2022 | 835 | 673 | 162 | 19.4% |
| December 2023 | 848 | 638 | 210 | 24.8% |
| March 2024 | 779 | 567 | 212 | 27.2% |
| May 2024 | 831 | 554 | 277 | 33.3% |
The last two columns are our subtraction and division from the published counts. In these five reports, between 19% and 33% of delayed projects disappear once the schedule is reset. Some projects also cannot be measured at all: in December 2023, 298 of 1,820 projects reported neither a commissioning year nor a gestation period (the expected time from start to commissioning).
The latest MoSPI flash report in this comparison is May 2024, and later reporting may not let you repeat this sum. ThePrint reported in a 13 August 2026 column that since MoSPI moved to PAIMANA, its project monitoring platform, in September 2025, late projects are labelled "ongoing". ETV Bharat reported the April 2026 figures: 1,981 projects, original cost ₹37,12,662 crore, revised cost ₹42,78,402 crore. That is a 15.24% overrun by our calculation ((42,78,402 − 37,12,662) ÷ 37,12,662). This is a different project set from the 2020 to 2024 reports, so do not compare the two.
Our recommendation: keep the approved original finish date in every status report, next to the current forecast, and report how many dates have moved.
How can you test whether your own programme is heading the same way?
Run eight pass-or-fail checks on your programme. Questions 1 and 2 test for hidden delay and undefined benefits; questions 3 to 5 test the dependency, approval and scope causes in MoSPI's list; questions 6 to 8 test planning and governance (our recommendation). Fix failures in the order you find them, starting with any that cover a funding release.
1. Does every status report show the original finish date?
Pass: each component's original approved finish date sits beside its current forecast, and the report counts how many dates have moved. Fail: only current dates appear. To measure hidden delay, count components late against the original date, then subtract those late against the current date.
2. Are benefits written down with an owner and a date?
Pass: every benefit in the approved programme business case has a named owner, a measure and a target date, agreed before the next funding release. Fail: any benefit lacks an owner, a measure or a target date, or was not agreed before the next funding release.
3. Is every outside dependency closed before the work that needs it?
Pass: each land, clearance, permit, financing or supplier dependency is closed, or has an owner and a dated plan to close before the dependent component starts. Fail: any open dependency has no owner or no dated plan to close before the dependent component starts. Track them under programme risk management.
4. Does every pending approval have a named decision-maker and a due date?
Pass: both exist for every technical, engineering or tender approval. Fail: any approval has no named owner, or its due date has passed without a decision.
5. Is every scope change approved and priced?
Pass: every change since baseline has a recorded approver and a stated effect on cost and schedule. Fail: any change was absorbed without both. Scope management covers how to trace deliverables to benefits.
6. Does every component have a start and a finish date?
Pass: all of them do. Fail: any component lacks either.
7. Is the sponsor accountable for benefits, not only delivery dates?
Pass: the sponsor's name is on each benefit target, and the sponsor reviews them at every gate. Fail: the sponsor signs only schedule and budget.
8. Is there a scheduled gate where the programme can be stopped?
Pass: a dated decision point exists, with written stop-or-continue criteria. Fail: there is no gate, or the criteria are unwritten. See our stage gate process page for who decides at each gate.

What do we teach about stopping programme failure early?
We teach it in our Program Management Essentials course: 12 weeks, 10 modules and more than 15 hours of guided learning. The modules cover benefits planning and value realisation, governance structures, stakeholder engagement, risk management and strategic alignment, with templates, case studies and monthly live Q&A. It ends with a completion certificate, and session recordings stay available for three months.
The eight checks above draw on the same ground as those modules: benefits, governance, risk and alignment. The course is a foundation, not PgMP® (Program Management Professional) exam preparation. Sanjeev Kumar, our founder, teaches it with more than 20 years of experience, and we have mentored or trained more than 6,000 professionals. If you are unsure whether it fits your role, ask us for a free discovery consultation at support@augmentconsultancy.com or on +91 9816 360 116 (phone or WhatsApp).
Enrol in Program Management Essentials (US$99 per enrolment as of November 2026).
Frequently asked questions
Yes. Andreas Zeller's Why Programs Fail: A Guide to Systematic Debugging teaches how to reproduce, find and fix bugs in software. It is not about programme management. Its site is whyprogramsfail.com.
Sources
- 1.Ministry of Statistics and Programme Implementation (MoSPI), Flash Report on Central Sector Projects, December 2020, MoSPI, Government of India, Infrastructure and Project Monitoring Division
- 2.MoSPI Flash Report on Central Sector Projects, December 2021, MoSPI, Government of India, Infrastructure and Project Monitoring Division
- 3.MoSPI Flash Report on Central Sector Projects, December 2022, MoSPI, Government of India, Infrastructure and Project Monitoring Division
- 4.MoSPI Flash Report on Central Sector Projects, December 2023, MoSPI, Government of India, Infrastructure and Project Monitoring Division
- 5.MoSPI Flash Report on Central Sector Projects, March 2024, MoSPI, Government of India, Infrastructure and Project Monitoring Division
- 6.MoSPI Flash Report on Central Sector Projects, May 2024, MoSPI, Government of India, Infrastructure and Project Monitoring Division
- 7.The Strategic Impact of Projects: Identify benefits to drive business results, Project Management Institute (PMI), Pulse of the Profession in-depth report, 2016
- 8.When "delayed" becomes "ongoing", the growing transparency gap in India's infrastructure, ThePrint, opinion column, 13 August 2026
- 9.Infra projects see cost overrun of Rs 5.65 lakh crore in April, ETV Bharat, reporting MoSPI's April 2026 flash report, 25 May 2026
- 10.Why Programs Fail: A Guide to Systematic Debugging, Andreas Zeller, book website
- 11.Program Management Essentials, Augment Consultancy, course page, as of November 2026



