Here is a question we have encountered often in client meetings.
How many independent mechanics are there in India?
The answers come quickly and they do not agree. Half a million. Two million. Five. Ten. What is odd is that nobody in the room seems bothered by the spread. Somebody says the sector is informal and there is no real way to know, everyone nods, and the meeting moves on.
We have been in such conversations for years now. We have also, on occasion, been the ones giving a number we could not fully defend. This is not a complaint about other people. But we no longer think a shrug is the right response, because we realise now that...
The number is not missing.
It was just never collected. The usual places you would look are all built to find registered businesses. GST picks up firms above a turnover threshold, and most repair shops sit below it. Udyam is voluntary, so it captures the ones who chose to sign up, which is a particular kind of owner and not a representative one. The Economic Census does better, but it goes looking for establishments, and a good mechanic working out of a fixed spot on a main road with a toolbox and no shutter is not an establishment in that sense. Labour force surveys ask people what they do for a living, and a man who repairs motorcycles most of the year and works his land during sowing will answer that question differently depending on when you catch him.
None of these instruments is badly made. They are just pointed at a formal economy, and this trade is not in it. Which means the absence of a number tells you something about our methods of counting, not about whether the thing can be counted.
Start from the vehicles instead
The vehicles are documented, even where the workshops are not. We know roughly how many two-wheelers, cars, tractors and trucks have been sold in India each year going back decades. We can make reasonable estimates of how long each type stays on the road, which varies a great deal between a scooter in a metro suburb and a commuter motorcycle in rural Bihar. That gives us a working estimate of the fleet.
From there it is a series of questions...
Each of which has a defensible answer. How far does a vehicle of each type travel in a year, and for tractors, how many hours does it run? What servicing does that distance or those hours call for, given the schedules manufacturers publish? How much of a workshop's day is actually scheduled servicing, as against straightening bent handlebars, tightening and lubricating chains, broken cable repairs which is where a lot of the real work sits? How many hours of labour does all of that add up to across the country in a year? And how many of those hours can one mechanic realistically deliver, allowing for the fact that nobody is productive for every hour they are at work?
Answer those and you have an estimate of how many mechanics the country's vehicles require. That is worth having. But it tells you what the work demands, not what exists.
Then start from the streets
So we run a second estimate from the other end. How many workshops are there per thousand vehicles, or per thousand people, in a village, a district town, a mid-sized city? How many people work in each? What proportion are one man on his own?
Multiply that out across the settlement pattern and you get a second number, built on entirely different foundations from the first.
Two independent methods are the point of the exercise
If they land close to each other, the answer is worth something. If they diverge, that gap is not necessarily a failure. In a trade like this, a shortfall between the labour the country needs and the labour it has does not appear as unemployment. It appears as a three-day wait for a repair, or as maintenance postponed until something breaks, or worse as work handed to a person who is not yet ready to do it.
The parts of this piece a desk search cannot supply
There are three numbers in this that we cannot look up anywhere, and they will materially shift the answer. What share of a workshop's day is scheduled servicing rather than running repairs. How many people work in a typical shop, and how that changes between a highway town and a metro. What proportion of the trade is a single man working alone, and what proportion has no premises at all, only a fixed place he can be found.
Those come from visiting workshops and asking. So we are going to go and ask, and publish what we find alongside the working behind it.
What we will publish
Over the next few months we will work through this on our LinkedIn page and here, one step at a time. Two-wheelers first, since they are the largest and the most widespread. Then cars, light commercial vehicles, trucks and tractors together.
We will publish a range rather than a single figure. Partly because of uncertainty, but mostly for a more interesting reason, which is that the answer depends on what you agree to count. The man who repairs punctures and trues wheels, the one who services cabin air conditioners, the denter, the painter, the auto electrician — each of them is either a mechanic too or he is not. One man under a tarpaulin sheet repairing two-wheelers is either a workshop or he is not. These decisions change the number substantially. We will make our choice explicit and show what the number looks like under the other definitions too.
Some of what we put out could be wrong, too. We would rather be corrected while the working is still open than after we have published a conclusion.
The next piece deals with a question that sounds trivial and is not: what counts as a workshop, and who counts as a mechanic.