The Jobs Crisis Is Going To Get Worse: Nandan Nilekani

Raghavan Jagannathan. The Jobs Crisis In India. Pan Macmillan India. 2018. pp 384. Rs 469.

The biggest problem, or opportunity, for the current and many successive governments, is, and would be this - how to provide gainful employment to the millions of Indians entering the labour market every month?

That one question has several corollaries to it. What role would automation play in all of this? What kind of jobs would be the first victims of automation? Is it wrong to expect manufacturing sector to provide jobs at a large scale now? What are the problems specific to India? Can Indian education system read market signals quickly and skill the labour force according to that?

Here is an extract from the book The Jobs Crisis In India in which the author, R Jagannathan, the Editorial Director of Swarajya, spoke to Nandan Nilekani, co-founder and non-executive chairman of Infosys Limited and the man behind the Aadhaar unique identity project, for his perspectives on the themes flagged above and much else.

RJ: Technology is throwing up a jobs challenge. The problem seems to be global in nature, and not just restricted to India. Are governments anywhere near getting a handle on it?

NN: No, I think things are changing too fast, and the pace of change is driven by technology which moves very quickly, whereas the pace of job creation depends on societies and markets. Those don’t move as fast. The fundamental problem is speed mismatch, the nature of the problem changing, and the ability of the system to respond, the speed with which the system responds....

One thing, of course, is automation, which is affecting both services and manufacturing. The historical model of development in the emerging markets was to get onto the manufacturing treadmill, and then make more and more sophisticated things — go from shoes to planes or whatever. In that process, you create jobs, raise incomes, the kind of things the Chinese, the Japanese, and the South Koreans or South-East Asians did. That model is fundamentally broken. Manufacturing automation is going to dramatically eliminate a lot of hitherto low-skill jobs that could occupy people who may not have had the most advanced education.

That is on one side; on the other, there is globalisation itself. From the 1980s to 2012-13, globalisation went through a massive surge, and it is now in (partial) reverse gear. The Trump administration is trying to reverse NAFTA. There is thus an element of de-globalisation too.

The same thing is happening to service jobs, things like BPO, for example. One-time robotic process automation and even high-end jobs — legal assistants, for example — are being automated as ML is rising. So, there is an automation of manufacturing and service jobs, and this is happening at very high speeds. The historical fountains of job creation are under threat.

RJ: In the past, automation helped create lower-skilled jobs, so even people with high-school qualifications could get high-paying jobs. Is that different now? Also, is automation the new threat as globalisation was in the past?

NN: I think automation has been around for a long time, but historically the reason why globalisation was a bigger issue was that it could be personified emotionally. You can say I lost my job due to China or India. The idea that I lost my job because this machine was difficult to sell politically, but now it is becoming an issue, and there is anger against automators.

So, while public anger against globalisation could be directed at Mexico, or India, or China, the anger now is directed towards Silicon Valley, which is saying we will automate everything. Unfortunately, they have amplified this by saying we don’t need any more doctors, we don’t need any more truck drivers, no taxi drivers — and when you talk like that it upsets people. If you say ‘We will eliminate all truck drivers’, that’s 2 million jobs gone (in the US). Somebody will say we don’t need (so many) doctors as AI will remove that need... when you talk like that, Silicon Valley faces a backlash because they have become the personification of job destroyers. Global job destruction was always there. The backlash to Silicon Valley has been about use of indiscriminate language.

RJ: What are the problems specific to India?

NN: India has had an additional challenge. Most of the countries that got into manufacturing-led job creation and now face an automation threat had universal literacy. Japan had 100 years of literacy; the Chinese Communists invested in the iron bowl and teaching; all of South-East Asia is highly literate. Korea, the same. India is the only country facing the double whammy of this jobs crisis hitting us, and illiteracy. So, we are particularly unprepared for automation.

RJ: You have said in some presentation that manufacturing is essentially over as a job creator. Can you expand on this?

NN: Adidas is going to make shoes in Germany. They are only going to create about 600 jobs in the factory. If they were going to make 300 million shoes by hiring thousands of people in Asia, those jobs are gone. And that is happening sector by sector.

RJ: But this could work for Germany, which has an ageing population...

NN: The West always kept the advantage gained by changing the terms of trade. Today the competitive response of the West to ageing is automation. Why did they come to India 500 years ago — because they wanted to buy cloves and spices. Why did they go to China — to buy tea. But they changed the terms of the game. They commoditised the prices of those things and changed the terms of trade. Similarly, because the West is ageing, it is easier. If the choice is between migration and automation they will opt for the latter. Japan is facing the same things, they are automating like crazy....

The populous countries are facing this at a difficult time. Automation is happening, anti-globalisation is happening, there is massive overcapacity in most industries, and we are at the end of the commodity super-cycle.

With China’s overcapacity of the last twenty years, we don’t need more factories. There is enough steel production in the world. Plus, we have this massive domestic population that is not educated.

RJ: So time is not on our side. How will new manufacturing technologies like 3D printing impact jobs?

NN: I think we may need to evolve from the notion of large monolith factories hiring thousands of people. That’s not going to happen. When you have a very large factory making thousands of scooters or cars, you will automate that. Even Indians are doing that. You go to the Bajaj Auto factory, and they are automating in a big way. But you could have a new model of manufacturing, which is high-skilled, high-end cottage manufacturing, where, perhaps, every town, every nook and corner, has a 3D printer person. He just downloads the latest designs from the computer and makes the product. So, it’s about small-scale manufacturing with high-tech.

RJ: Will 3D manufacturing require high skills? And is 3D printing taking off?

NN: It goes back to education — that is our fundamental strategic problem. And yes, 3D is taking off. The Adidas factory uses 3D manufacturing. It basically makes layers and layers of shoes from a liquid. And there are different types of 3D, and it is evolving very fast with metals and composite materials. They [the advanced nations] will crack it.

RJ: So, one should not put great hope in manufacturing providing jobs?

NN: India expecting to become some kind of manufacturing back office of the world is not going to happen. China will be replaced by machines. China will not be replaced by Indians. China is one of the biggest adopters of robots. Foxconn’s factories are all robotic. The Chinese are ageing rapidly, so for them it makes sense.

RJ: Is the same problem true of financial services like banking, where too we have heard of banks like HDFC cutting jobs?

NN: Fundamentally financial services are delivered on the wire, on a phone. I can send money, receive money, make a deposit, sign a contract [digitally]. Why do I nee (sic) a branch? It’s all happening.

RJ: Will the job loss be made up in fin-tech?

NN: No, not enough jobs in fin-tech. My thesis is all this [digitisation] will reduce the cost of customer acquisition and the value of transactions. The reason why the mobile revolution happened in India was the genius of prepaid. Prepaid was a brilliant innovation, helping deal with Indians who had no credit records. The notion that you pay Rs 100 in advance, and buy talk time was a masterstroke. Every mobile became a bank as all the money, paid in advance, brought a lot of people into the system, and this was combined with sachet size. The mobile recharge cost came down to Rs 5. Combine prepaid with sachet and suddenly a billion people could buy talk time, and that is what will happen now with the new financial technologies. It will create zero friction, and make low-cost, small-size transactions possible. More people will join the formal economy, that is what will happen.

Jobs will not be created by Yes Bank starting more branches, but if Yes Bank delivers financial services to 100 million people and they get access to the formal economy, then they will somehow [make things happen]. The important thing is job creation is going to happen at millions of small units. Twenty thousand jobs are not going to be created by Maruti in Manesar; 20,000 jobs will happen when 20,000 small units hire 1 extra person. That is the way things will happen.

RJ: Will similar things happen in IT?

NN: You see, traditional services are getting automated. They will have to stay ahead of the curve and automate rather than resist it. But the flip side is the opportunity for the transformation of companies, which is also high because they are facing digital challenges. If I am Nordstrom, I worry about Amazon. If I am Ford, I worry about Tesla. The large incumbent players in every industry are facing digital challenges and they have realised that they have to transform themselves if they have to survive. So that is driving tech spending, that means people need to help them do it.

RJ: Can people do it inside their companies?

NN: Some big MNCs are saying a lot of this is strategic but may not have the talent pool to get it done, or the ability to re-skill their people to do this, whereas people like us, we can re-skill our people faster.

RJ: Autos, banking, and even IT may not employ too many directly. Where will the jobs happen?

NN: In autos, more jobs will happen in repair sales and services.

RJ: Real growth may happen in education and health and social services. But can’t they be automated too?

NN: They will be automated, but most of these will have a component of high touch. You can’t educate without teachers, you can’t give healthcare without nurses. Those jobs that have high empathy and people-oriented stuff, they will continue.

RJ: But in Japan old people learn to live with robot helpers.... What is to stop a hospital being run largely by robots?

NN: Your breakeven points are too high in India. At low-wage levels, replacing people with robots is a very expensive proposition. At low-wage levels robots are not economic. Our point of robotisation will come much later.

RJ: Do you see a change in the idea of ownership? Will people tend to hire more than own things in future? Will Uber and Ola reduce the need for people to buy cars? And how will this impact jobs?

NN: I think, on mobility, it will happen. A lot of Indians will just skip buying an automobile. It doesn’t make sense given our urban infrastructure. Shared assets are more utilised. In many areas you will have a sharing economy....

RJ: Why only mobility? Why would one want to buy a high-priced phone when it will be obsolete in eighteen months?

NN: Yes, you may have subscription services, you will sign up a five-year contract for phones, whenever a new version comes the operator takes it back and gives it to somebody else who is okay with a hand-me-down.

RJ: We are so behind in terms of skilling our people. How is that going to play out, and will ‘Skill India’ work?

NN: Again, if you use this as a supply-side problem — the government thinks of it as a supply-side problem — you are anticipating what supply is required, whereas it is a market issue. I can’t predict what demand will be there five years from now. If I say I will upgrade my ITI to provide some skill in welding, and, in five years, if there is not much need for welding, you are gone. So it must be demand led....

RJ: There is a gap between information about the skills in demand, the certifications that are required for them, and the supply of those skills....

NN: There is knowledge asymmetry. If you can reduce the time gap between job demand and signalling to the labour market, that’s where you get some action. And you need a flexible education system that can reorient itself for that system very fast. The agility is not there is the system.

RJ: Doesn’t the West have the same problems?

NN: The West has bigger problems. The bigger problem was that in the post-war era, between 1945 and the 1970s, high-school grads could get a high-paying job in a GM plant in Detroit and get lifelong healthcare and buy a house in the suburbs. That era is over, and a lot of the Trumpian issue is that the white middle-class has lost ground. The Trump constituency is the fifty-year-old jobless person in West Virginia. On the other side, in the 1945–65 period, a lot of the education was driven by massive schooling of ex-military people. Those who fought in the war got free education. Now, one of the biggest issues is the rising cost of education in America. You end up with people taking loans to go to college (student debt is over a trillion dollars) and there are no jobs for these people. So, both the high-school grads and regular grads have a problem.

RJ: Are MOOCs the answer to quick skilling?

NN: The first generation of MOOCs were static video. The second generation is more agile, gives smaller courses, but the long tail of people dropping off remains.

RJ: Is there a market signalling gap to MOOCs too?

NN: One is market signalling, but after you get that signal, you need access to that skill, you need credentials — no society has that. The closest is Singapore. They have this SkillsFuture, which gives people credits for relearning. The state does some scanning of market demands.

RJ: So, should not the state be doing market scanning and close the information gap?

NN: States can’t do that. The state is not good at anticipating markets. The state should create infrastructure that reduces the delays between market signals and acquisition of relevant skills. They should focus on reducing the time and not manage the signals.

RJ: Our education model is largely broken.... What needs to be done?

NN: It will require a leapfrog using highly tech-base access to infrastructure, access to infrastructure for anytime, anywhere learning, automatic assessment, and those kinds of things. We do a lot of that at EkStep, a philanthropy initiative, to prepare infrastructure for twenty-first century education. We need a signalling mechanism that reduces cycle time between demand for a service and the supply side getting up to it.

RJ: What should the government do?

NN: The government should create the enablers — not give money to an ITI to teach welding. It should create infrastructure to connect market and supply.

RJ: Will the huge supply-chain disruptions impact jobs? What if car demand shifts from internal combustion engines (with 20,000 movable parts) to electric vehicles (which have just 20)?

NN: Many things are happening. The impact on manufacturing will be huge, but supply chains will require a lot of people anyway. What is happening because of competition is the need for instant consumer gratification. Amazon offers same-day delivery in every major American city, and in some cases delivery (Amazon Now) within hours. That requires warehouses, delivery people, a new set of jobs. New supply chains based on end-user satisfaction.

RJ: What are the prospects for secure middle-class jobs? Is there a crisis ahead here?

NN: Yes, of course. More importantly, the life-time employment business is dead where you join a big bank as a probationary officer and retire as deputy MD or GM. And this is true for a lot of industries. [Given the pace of change] companies cannot also think long term.

RJ: Does the fact that India has a young population give us any benefits, since the young are easier to teach and skill?

NN: You have to get it right. A younger population is always an advantage, it is malleable, flexible, but if you haven’t got the supply-side right, and it is not aligned to market, that is a problem.

RJ: Where will new jobs come from?

NN: We should expect a larger number of smaller entities that create jobs and a lesser number of larger entities that create jobs.

RJ: Who should do the job of gathering information on job supply and demand and skill requirements?

NN: Ideally, governments should do it, but they don’t have the innovation skills for it. We need to think about interoperable infrastructure, which has many moving parts, from [skilling] content producers, to consumers, people who deploy this stuff, and then there are people who test you for that, give you certificates. So, a multi-stakeholder ecosystem is needed, and somebody has to orchestrate that. The orchestration is the problem. In India, this can ideally be done by states. If even a few states do it, and they succeed, it can be scaled up.

RJ: Can you give us three bullet points on what needs to be done?

NN: First, we have to encourage the creation of mass services — which is what IndiaStack tries to do: reducing friction, reducing interaction cost, reducing search costs, creating sort of Lego blocks that reduce friction and enable job creation. [IndiaStack is a set of application programming interfaces that enable governments, businesses, start-ups and developers to use the digital infrastructure to solve India’s hard problems towards ‘presence-less, paperless, and cashless service delivery,’ as noted on its website.]

Second, assume the future of jobs is the small enterprise. How do we make a million small enterprises, each with two employees? That’s two million jobs. Think small and at a scale. Reduce friction for a million small enterprises.

Third, leverage data. Individual data. My data is my concern — my payments, my tax payments — which I can leverage to get a loan. If I can send this data securely to someone on a contractual framework, I can monetise it to get a loan or a lower healthcare premium.

RJ: Are social tensions going to boil over due to this jobless crisis?

NN: It’s going to get worse. Your regional and caste movements are about jobs. Hardik Patel wants jobs. Jats, Gujjars, all have the same issues. The government can be only a limited source for jobs. Maybe we need more policemen, but that’s a fraction of the jobs society needs. You need entrepreneurs. You need this flexible and agile system where job creation is notified and people apply for jobs, see demand, and quickly get retrained.

RJ: Do you see that happening?

NN: Unlikely. Designing such a system is complicated. It requires systemic thinking.