Mumbai and Noida: India banks solely on police data for estimating deaths in road accidents, ignoring other critical sources, leading to a data gap that impairs its emergency response system, makes it difficult to design and implement accident prevention strategies, and leads to an underestimation of accidental deaths, experts say.

The World Health Organization (WHO) classified India's official data on road traffic fatalities as "unusable" or "unavailable" because of "quality issues" in its 2019 Global Health Estimates released in December 2020. This study is still the latest available source of global data on death and disability by region, country, age, sex and cause.

In its methods and country-level data sources handbook, the WHO pointed out that it has official figures from India for only 2001-2003 and 2010-2013. For the remaining years, it has had to collate numbers from the Global Burden of Diseases (2019) study produced by its partnering institute, the Institute of Health Metrics and Evaluation (IHME), an independent research center at the University of Washington, said a WHO statistician who did not wish to be named.

The Ministry of Road Transport and Highways (MoRTH)'s data on road crash deaths are not complete because the data reporting process of the police departments is flawed, there is significant underreporting especially in rural India, and not enough public trust in the police, experts told IndiaSpend.

The WHO classifies all countries into four categories on the basis of data availability on the cause of death: "high completeness", "low completeness and/or issues with cause-of-death assignment", "low completeness with severe quality issues" and "data are unavailable".

In the case of road deaths, for example, Japan is listed in the "high completeness" category and Thailand in "low completeness", while India's numbers are categorised as "data are unavailable". While Japan collates road deaths data from sources such as hospitals, ambulance services, forensic investigation of crash sites and insurance providers, Thailand banks on its ministries of public health, interior and transport and so on for estimates.

However, this gap is normal because the WHO and the Indian government use different methods to collect numbers and it is difficult to prove which method is more robust, Anshuman Mohanty, economic advisor, Transport Research Wing, MoRTH, told IndiaSpend.

A comprehensive crash investigation system is required to offer every stakeholder--the public, the police and policy makers--strong evidence on the contributing factors, wrote Benedicta Isaac-Kumar, a consultant with Chennai-based Citizen consumer and civic Action Group, a non-profit, voluntary and professional citizens group.

Why police data are incomplete

In India, there are two reports on data relating to road traffic accidents--the 'Road Accidents in India' report of the Ministry of Road Transport and Highways and the 'Crime in India' report of the National Crime Records Bureau (NCRB). As in many countries, the data are sourced entirely from the police departments of states and union territories.

Here is how the police collect crash information. Once they are informed about a road accident by someone involved in the mishap, a witness or a police officer at the site, a First Information Report (FIR) is recorded.Its content can only be changed with the permission of the High Court or the Supreme Court of India, explains a 2020 Indian Institute of Technology Delhi (IIT-Delhi) study. Once the investigation concludes, a case file is prepared. After this, the offending party is charged under provisions of the Indian Penal Code and the Motor Vehicles Act of India 1988.

In the case of serious injuries and instantaneous death, the survivor or the victim's family may file an FIR only if they want to pursue the case in a court or claim compensation, insurance after adjudication by the Motor Accidents Claims Tribunal.

The IIT study also reported that most cases are attributed to "human error" and this means that factors such as vehicular and infrastructural problems are not included. The cause of the crash has to be recorded as the "fault" of a "road user" as per the India Penal Code (IPC) and Motor Vehicle Act 1988, ignoring all other factors.

On the shortcomings of the data reporting mechanism in India, the MoRTH report itself pointed out that if a road accident victim dies in a hospital after 30 days of the accident, it is not reported as a road traffic fatality.

"Collecting data from different sources to build meaningful analysis of road crashes is vital," said Piyush Tewari, founder of the New Delhi-based SaveLIFE Foundation (SLF), an independent non-profit organisation focussed on improving road safety and emergency medical care. "For example, data can be collected from police officials, ambulance service providers, information from bystanders etc."

"We don't have any national trauma registry database where hospitals need to report road accident deaths or injuries, like now all hospitals need to report Covid-19 cases on a central portal," M.C. Misra, the former director of the All India Institute of Medical Sciences (AIIMS), told IndiaSpend.

Underreporting of deaths

Due to the inadequate police presence in India's rural areas, it is possible that about 50% of road crash deaths there go unreported, said the IIT study. For urban pockets, this figure is lower but still significant, 10-20%. In urban areas, most of the road accident deaths and injuries get reported which occur in a hospital. Due to the sparse density of hospitals in rural areas, the chances of missing accidents, deaths and injuries are higher.

Data on non-fatal crash injuries in India may also be unreliable because survivors may choose not to report them to the police or hospitals, the IIT Delhi study noted.

Lack of trust, bias

Road traffic accidents and injuries go unreported because of entrenched biases among both the police and civilians. Accident survivors in India are reluctant to report cases because they believe the police to be apathetic and also because they do not want to be caught up in lengthy police procedures, especially in rural areas, said G. Gururaj, former director of the National Institute of Mental Health and Neurosciences (NIMHANS). He pointed to the fact that in rural areas, an accident victim is first cremated and then the case reported to the police.

When accidents occur, police officials tend to assume that drivers of large vehicles such as truck drivers are at fault, ignoring the conditions of the vehicle, fatigue among drivers, road conditions, and so on, said Tewari of SLF, adding that this is why truck drivers tend not to report accidents. About 40% of respondents in the 2020 SLF survey on truck drivers said they were dissatisfied with their jobs because of harassment from the police, the transport authorities and local residents.

Skewed distribution of traffic resources

Police departments are also stretched for human and other resources. About 27% of the vacancies in traffic police departments across India have yet to be filled, according to the 2020 Bureau of Police Research & Development (BPR&D) report on 'Data on Police Organizations'. Eight states--Jharkhand, Rajasthan, Haryana, Gujarat, Telangana, West Bengal, Madhya Pradesh and Bihar--have more than 40% of posts waiting to be filled in their traffic police forces. Among them, Rajasthan, Madhya Pradesh, Telangana and Gujarat appear in the top 10 list of most number of road accidents in India in 2019, per MoRTH report.

The deployment of traffic constables and head constables in states does not align with the vehicular density of states. An average 26% of posts lie unfilled in the traffic police departments of 10 states with the most vehicular density, shows a combined analysis of the MoRTH's 2017-18 and 2018-19 Road Transport Year Book report and the BPR&D report.

Finding solutions

To allow the data on road accidents to be more robust, the government introduced the 2019 Motor Vehicles (Amendment) Act (MVAA). It aimed to enable officials to investigate road crashes for better and more comprehensive data.

The government also launched the Integrated Road Accident Database (iRAD) project. The iRAD is a central accident database management system launched in January 2020 that connects information from multiple sources--hospitals, ambulance services, blood banks, insurance companies, and so on. It was deployed for "live accident data entry" in six states--Maharashtra, Karnataka, Madhya Pradesh, Rajasthan, Uttar Pradesh and Tamil Nadu--in February 2021. Its data processing mechanism was still in progress as of March 18, 2021, Gadkari said responding to a question during the March 2021 parliament session.

The SLF has started training traffic police in India's 15 most accident-prone states, including Uttar Pradesh, Maharashtra, Odisha, Delhi, Telangana, Rajasthan, Jharkhand and Chattisgarh. The Uttar Pradesh police have started conducting forensic investigations for road crashes on the Yamuna Expressway in collaboration with the SaveLIFE Foundation.

However, data integration is challenging in India because it is "a poor country" with "a lot of diversity", Mohanty, economic advisor, Transport Research Wing, MoRTH, told IndiaSpend. "I don't have any update on the implementation of iRAD in India. I have to check," he added.

Case study: Why data is critical

In 2016, the SLF collaborated with the Maharashtra government on a Zero Fatality Corridor (ZFC) project on the 95.4-km Mumbai-Pune Expressway (MPEW), aiming to create a safety model that could be replicated across roads in India. They looked for solutions in four areas commonly stated as "4 E's of Road Safety": engineering, emergency care, enforcement and engagement.

The project undertook various initiatives--issued 46,563 challans using speed-traps, installed Vehicle Activated Speed Sign (VASS) to alert motorists about speeding and barricaded 10,300 parked or stalled vehicles to prevent rear-end collisions.

As a part of the initiative, around 150 km of Tactile Edge Lines--yellow road tiles to indicate hazards and the direction of travel--were installed on the highway to address fatigue-related crashes. The project also pinpointed 15 types of road engineering errors--poorly designed slopes and lack of siding space for trucks on uphill tracts--which had been repeated 2,145 times across the two corridors of the highway. Later, these engineering errors were rectified by the SLF in collaboration with the state government.

The SLF also developed a 'Chain of Survival' initiative consisting of on-ground, ambulance and in-hospital care services for road accident victims.

This model helped slash the number of road crash fatalities on the MPEW by 52% between 2016 and 2020--the numbers fell from 151 to 66. Replication of this model, the use of data science and forensic investigation--all these can make Indian roads safer, said Tewari of SLF.

We welcome feedback. Please write to We reserve the right to edit responses for language and grammar.