#306 #NoFilter Analyses
Debating the Value of an Engineering Degree in India, China-US Magnet Truce, and the Long Overdue Air Safety Regulatory Reforms
India Policy Watch #1: An Elaborate Charade
Insights on current policy issues in India
—RSJ
One of the underreported stories of Indian industry gained some media coverage a couple of weeks back when education entrepreneur, Peri Maheshwer, founder of Career360, wrote on his Facebook page about the falling value of an engineering degree in India in the past 15 years (there’s this YouTube episode too). Now a bit about Peri Maheshwer. He was once an editor of Outlook, the current affairs weekly that gave India Today a run for its money in the late 90s and early 2000s. Over the past two decades, he has built a reputation as a credible voice in the student counselling space that’s rife with charlatans. Talking of charlatans brings me to, arguably, Maheshwer’s biggest claim to fame for those of us unfamiliar with his work. He single-handedly brought down Arindam Chaudhuri and the IIPM group of institutions, which, through much of the 2000s, had defrauded thousands of students with their false claims of foreign degrees, placements and academic excellence. Almost everyone then could see IIPM was a scam but no one took them on. Chaudhuri lied about almost everything related to IIPM, flooded traditional media with ads to buy their silence and when challenged by Maheshwer, harassed him by filing frivolous cases against him in courts in remote corners of India (one in a Silchar district court, if I remember correctly). Maheshwer won that four-year battle with odds stacked against him and at a high personal cost. I bring this up to highlight the public-spirited nature of Maheshwer’s enterprise and his courage in speaking truth to power. The stagnation of entry-level salaries among traditional IT services players in India (both of Indian and foreign origin) isn’t exactly a scam. But very few speak about it or question it. Maheshwer has made a start with this.
The data here is quite compelling. In 2010, the entry-level salaries at the Top 5 IT services companies (including Accenture) was between ₹3-3.3 Lakhs. In 2025, that has moved to ₹3.6-3.8 Lakhs, an increase of about 20 percent in absolute value over a 15-year period. The cost of living over the 15 years has almost doubled if one were to take an average inflation of about 4.9 per cent. Simply put, the entry-level salary in IT services today is about 40-50 per cent lower than what it used to be in 2010.
Maheshwer points this out while contrasting it with two striking data points. Profits among these five companies have risen between 3.4-5 times during this period while the cost of a four-year engineering degree has gone up by 3.2-4 times during this period. From this data, Maheshwer arrives at a few conclusions or (to be fair to him) questions. Isn’t this profiteering on the part of IT services firms? They seem to have reached an ‘informal’ understanding of keeping entry-level salaries low while they continue to raise the price of their services, which is evident in their growth in profits. How has the cost of education continued to rise if the value of a degree has only fallen? Education loans (among the priciest loans offered by Banks) and their repayment might now take about a decade for students to pay off. With that level of indebtedness early on in the career, should it surprise anyone that their consumption remains low, demand for unsecured loans to tide them over short-term cash flow shortages continues to be high, and there’s almost a zero sense of loyalty to the employer. Maheshwer, who I suspect has a good, old-fashioned socialist bone in him, urges IT companies to share more of their profits with their employees and calls for policymakers to look at this asymmetry of cost of education outstripping the rise in salaries. He has received the standard rebuttals from the industry and others about this being a demand and supply issue, and there’s nothing more to see here.
Is this an issue for policymakers to step in? Is there a market failure here with strong evidence of cartelisation among players? I am not too sure if one were to just go by what seems to be Maheshwer’s understanding of the problem which is IT services companies are profiteering by suppressing wages of their employees. I suspect there’s more here than mere demand and supply at play or abuse of industry power by a few players to exploit employees. The state doesn’t need to intervene here. It will only make things worse. But the data that Maheshwer has put out is useful to reach other, possibly more nuanced, conclusions which might be useful for analysts and policymakers to consider.
Firstly, a good additional data to help with comparisons here would be the trend in operating margins (OM) among this set of players over the 15 years. From the highs of 28-34 percent, OM has seen a secular decline over this period by between 3-8 percent. Now a large determinant of operating expense increase in IT services is wage cost. They keep it in control by continuing to have a fairly robust growth in headcount by hiring in bulk at entry levels and keeping a tight lid on entry-level salary. Despite the anaemic growth in entry-level salary and a 2-3 times growth in headcount, the OM has declined consistently. Why? Is it because these companies have invested heavily in innovation, research and future-ready technology, thus increasing their expenses during this period? No publicly available information suggests this is the case. Then the only other conclusion to reach is that their revenue productivity isn’t going up. In simple terms, their offerings aren’t differentiated or value-adding enough to sustain pricing power. Any talk of Indian IT services moving up the value chain and doing higher-quality work should be seen in this context. Now, in my view, if there is a demand for a service and you can fulfil it profitably, it is all that we should expect a firm to do. There’s no high or low-end work; only high or low-quality businesses can or cannot sustain a business model for long. In that same breath, we should wonder if this is the nature of IT work that Indian firms are doing, what would the impact of coding LLMs be on this work in the very near future? How good quality a business model are they running, and how sustainable is this business model?
Secondly, maybe we should ask a different question, such as whether an engineering degree has fallen in value in the past 15 years. Perhaps, the right question is whether an engineering degree was overvalued in the past and is now reaching its right value? If the likes of TCS and Infosys aren’t able to innovate or demand a premium for their services (based on evidence of margins), is it because the engineers they have hired over the last two decades just don’t have the quality to create such differentiation? After all, what exactly are these firms selling except for the enterprise and skills of their staff? The other data to consider here is how the median salaries of the graduating batch of college engineers have moved over this time. The data from Tier 3 and below colleges is difficult to verify, but broadly speaking, for Tier 1 and 2 colleges, the median salary has moved from ₹4-7 Lakhs (in 2010) to about ₹8-12 Lakhs, keeping with the pace of inflation. The median Indian engineer is indeed getting a higher salary over this time. But the Indian IT services companies are making do with lower and lower decile of engineering graduates because either the quality of work coming their way doesn’t need higher quality talent or they have re-engineered the software development lifecycle to an assembly line model where you really don’t need too many super skilled employees or all round experts. Perhaps back in 2010, you still needed engineers who could apply themselves a bit more because the work was of better quality, or your internal processes (reusable code, standard platforms) were still underdeveloped. The engineer who joined the IT services then was therefore overvalued.
Thirdly, we come to the question of the cost of engineering studies going up by almost 3-4 times during this period. How should we see this? An obvious response to this is to highlight the doubling of the median salary of the graduating class, the emergence of significantly higher paying jobs at the top quartile of graduating class (product companies, startups, consulting and VC jobs) and the need for engineering colleges to keep up to the demand. But what about the lower deciles? What’s happening there? A couple of things here that go beyond the surface.
One, the Indian IT services companies have made virtue of the fact that they hire engineers for the jobs that they outsource to India from their global clients. An engineer is a very different and specialised resource in developed economies than what India produces on average. So when global clients hear a software engineer being staffed on their project, their comparison is to the image of the engineer they have in mind. It would surprise many in India to know that a global firm like Accenture barely had engineers in their ranks about two decades back when they hadn’t set up large India offices. The work they did for clients, arguably higher-end than Indian IT firms, didn’t require engineers. Strangely, this asymmetry of perception about Indian engineers persists even after three decades of outsourcing to India. This allows the India IT services sector to sustain the myth of hiring engineers and charging some kind of premium even when the work doesn’t demand an engineer. This chimera has kept the demand for engineers reasonably buoyant in volume terms, but not quality. The engineering colleges have merely catered to this demand by increasing seats, lowering minimum cut-offs for entry and letting everyone clear the course. A tech sector job is fairly cushy, the work locations are in tier 1 cities, the mobility factor within the sector is high, and there’s a legacy ‘halo’ effect of working in these firms that still works for middle-class families. So, the demand for engineering seats continues to remain high despite the real value of the degree falling quite sharply over this period.
Two, the lower entry-level salary of the likes of TCS and Infosys forms the benchmark salary for a whole lot of other companies, including those in manufacturing, that come for placements and eventually for the entire ecosystem of jobs that support a fresh software engineer in big cities. Also, the students from outside of Computer Science or Electronics engineering streams continue to join the IT services sector rather than working in their core industry. This is a complete inversion of the global trend where more engineers (as a proportion) are found working in the manufacturing sector than in IT services. This has meant global manufacturing companies find Indian engineering talent on the shop floor weaker than what they have, compounding the difficulties in setting up an outsourced manufacturing unit in India. I sense this is somewhat sweeping an assertion, but there’s more than a mere strain of truth here.
Lastly, I do have some sympathy for a particular line of argument that the IT services sector makes on this topic. The average graduating Indian engineer joining the likes of TCS and Infosys is almost unemployable. These companies spend anywhere between 6-9 months of training effort while paying these engineering graduates to teach them everything from writing, speaking and coding. Quite clearly, they could have hired any graduate and trained them, except they like the idea of hiring engineers for their clients. In their view, although they won’t say this openly, these aren’t entry-level employees working for them and earning their salaries. They should rather be viewed as apprentices who are being paid nominally to learn on the job. It is a huge drain for the IT sector to do this every year and yet lose more than a quarter of their people to attrition. Anyway, once the employees finish their training and get into the decidedly commoditised nature of work on offer, they do well, earning competitive and comfortable salaries in five years or so. There’s also the additional lure of going abroad and working at a client site that makes the initial years of the grind worth it. This isn’t a bad thing, except to admit it will mean dropping the charade that these companies hire the best talent from campuses to serve the needs of their global customers. And once you drop it, you will struggle to justify your pricing.
I might be repeating a point I have made a few times earlier. India's IT services sector has been agile and sharp in keeping with the broader trends in technology and riding those waves while simultaneously doing more and more commoditised work. This is an interesting paradox that they have been able to manage over the years, and this is a tribute more to their financial acumen than to their technological strengths. If you look through this, it is a house of cards that’s stayed freakishly resilient over the years. Will it survive the AI wave? I remain sceptical about the prospects of the biggest employer in the formal sector. Peri Maheshwer raises pertinent questions with his data. He thinks IT services firms are perpetrating a crime. A closer look would suggest they might be the eventual victims of this.
Global Policy Watch: No Time for Complacency on Rare Earth Magnets
Global policy issues relevant to India
—Tannmay Kumarr Baid & Pranay Kotasthane
(An edited version of this article appeared first on India’s World website on June 23)
Donald Trump recently announced a new Rare Earth Element (REE) deal with China. As per the deal, China is set to lift its export controls on any REEs exported to the United States in exchange for exemptions granted to Chinese students amidst the president’s crackdown on immigrant student visas.
Trump portrays this agreement as a significant win for his administration. The EV industry, which is dependent on rare-earth magnets, is also jubilant. However, the reality is more nuanced. The deal only provides a six-month license to US companies, after which China could reinstitute the controls if US-China relations do not improve. With a newly institutionalised export control regime, China is prepared to use its rare earth dominance for strategic gains again in the near future. There is, for instance, no such deal with India on the horizon, and Indian automotive companies continue to rely entirely on rare earth magnet supplies from China, which remain blocked. We thus argue that policymakers—particularly in the US, EU, and India—should look beyond short-term deals and accelerate efforts towards diversification.
Anatomy of China’s dominance
China’s primary advantage does not stem from exclusive dominance over natural reserves or possession of superior mining capabilities; rather, it lies in refining and processing. China controls approximately 90% of the global rare earth element (REE) refining industry. This dominance did not arise from advanced technological prowess but through government subsidies paired with predatory pricing practices. Take the Bayan Obo mine in Inner Mongolia, which is responsible for roughly 45% of global REE output on its own. In the early 2000s, subsidies on electricity, rail freight, and land-use fees at Bayan Obo shaved operating costs by up to $15 per kilogram compared with Australian peers. Beijing’s export quota regime, introduced in 1999, then ensured domestic refiners had first claim on cheap concentrate, which resulted in a fully integrated on-shore supply chain that no foreign competitor could match. Environmental compliance costs, meanwhile, remained minimal for Chinese firms because wastewater-treatment standards were loosely enforced. Estimates are that Chinese refiners spent only a fraction of what their international counterparts paid on effluent control. The result was that by the time overseas miners ramped up production in the late 2000s, China could undercut them by 30–40% on finished oxides.
Many countries attempted to diversify away from China; however, their efforts failed because whenever international suppliers sought to expand their operations, Chinese firms would flood the market with subsidised REEs at artificially low prices, driving foreign competitors out of business or forcing them into acquisitions by Chinese firms. This reversion to dumping rare earths was possible because China’s previous export restrictions lacked official status, were country-specific, and tended to be short-lived.
The recent controls, however, differ significantly. China officially declared these export controls, structuring them explicitly as retaliatory measures, thereby giving them a degree of permanence. The establishment of a formal bureaucratic structure to decide permits further signals China’s recognition of REEs as a tool it intends to use in the future.
The new agreement
The semi-permanent nature of the recently established export controls creates a promising window for global diversification initiatives. Countries will be forced to develop alternatives, encouraged by the clarity of China’s stance and a resultant increase in prices, which would make these alternatives economically viable.
However, Trump’s recent deal with China could disrupt this momentum to some extent. The scenario to draw attention to is where companies and governments become complacent because of the temporary relief. This kind of short-term period of stability could lull businesses into a false sense of security, which would deter necessary investments in furthering alternative technology and diversification strategies. To put the six-month licence in perspective, permitting a green-field REE refinery typically takes 18–24 months, while qualifying a new magnet supplier in an automotive supply chain can add another 12–18 months. In other words, by the time the current waiver lapses, even the fastest-moving projects will still be at the blueprint stage. Treating the reprieve as anything more than breathing space, therefore, risks freezing capital expenditure precisely when momentum is most needed.
Adding to this is the almost inevitable predatory pricing that China is likely to employ when any of these companies attempt to establish refining operations. In this scenario, if competition against China’s dumping doesn’t materialise, China’s dominance over the REE market would persist and likely strengthen.
Policy Recommendations
However, these situations can be avoided by active policy-making. Governments worldwide must actively support emerging private-sector efforts to diversify refining and processing. Here are a few actionable recommendations:
First, simplifying auctioning and subsequent licensing processes for new blocks of REEs as and when they are found. Australia’s 2023 Critical Minerals Strategy, for instance, appointed a dedicated “Critical Minerals Facilitator” that cut average approval times for exploration licences by nearly half; India and the United States could replicate that single-window model. Additionally, the Mineral Security Partnership [SA9] [PK10] (MSP) is another multilateral avenue for cross-country synchronisation on licensing and regulation for international projects, with the intent of cutting down approval time.
Second, when refining is set up in its fledgling state domestically, impose coordinated countervailing duties against Chinese REEs to protect against price dumping. The WTO-approved safeguard tariffs that the United States applied to Chinese solar panels in 2012 provide a template for such actions.
Third is to promote refining-technology innovation. This will happen anyway, as private companies establish refining processes; government research institutions can further accelerate the pace of research. One promising line of inquiry, for example, is ionic-liquid separation for low-grade clays, which reduces acid waste and energy use compared to traditional solvent-extraction cascades.
Fourth, work on initiatives to promote recycling and material and process substitution to reduce reliance on REEs as a whole. Urban Mining Company’s pilot plant in Germany, which harvests NdFeB magnets from end-of-life wind turbines and electric motors, is an example of how industrial recycling could look. Other examples include Hitachi in Japan and Conifer in the US, a company founded by two Indian Americans.
Fifth, work on stockpiling and resource sharing with international partners. Shared inventories in the next couple of years are essential to keep the supply chain running in the meantime. Japan’s JOGMEC programme, for example, already maintains a 60-day rare-earth stockpile for domestic manufacturers. Multilateral arrangements would provide similar insurance on a larger scale.
Overall, an increase in usage efficiency and global diversification within the supply chain must be the goals of policymakers, rather than treating the deal as an end in itself.
India’s Opportunity
Like in the United States and elsewhere, it’s reasonable for Indian automotive companies to ask for a quick resumption of rare-earth magnet supplies from China. It also makes sense for the Indian government to facilitate a quick resumption. However, the government needs to go beyond this band-aid solution because the next India-China confrontation is likely to see China using this tool again.
The good thing is that India isn’t in a particularly bad position either. According to the Atomic Minerals Directorate for Exploration and Research, India’s coastline hosts an estimated 5.1 million tonnes of monazite, translating into roughly 1.6 million tonnes of contained rare-earth oxides which forms the world’s sixth-largest light-REE reserve. For one, its monazite sands have many light rare earths, including neodymium, the rare earth most widely used in magnets. Second, IREL (India) Ltd is setting up a Rare Earth Permanent Magnet plant in Visakhapatnam, one of multiple upcoming projects. Third, several companies have plans to increase capacity in the rare-earth magnet recycling space. A few examples include companies like Attero (₹100 crore investment), JBM Group, and Exigo Battery Solutions (to process 2000 tons/annum of feedstock into high-purity oxides). Parallel overseas moves are also underway: KABIL, India’s joint venture with NALCO, HCL and MECL, has shortlisted deposits in Chile and Argentina, signalling that New Delhi is willing to take equity stakes abroad to secure heavy REE feedstock.
Within India, the stumbling block is government regulations and control. Deregulating all segments of this supply chain, fast-tracking environmental approvals, and funding exploration projects to reduce information asymmetry will go a long way toward lowering China’s control over rare earths.
In essence, countries need to guard against complacency that may arise as a byproduct of temporary deals signed with China to resume exports. Instead, it is time to accelerate diversification initiatives through various policies. In the midst of this, India, in particular, has an opportune window due to domestic reserves and new government initiatives, such as the National Critical Minerals Mission (NCMM) and KABIL. Diversification, in the form of refining improvements and substitutability research, will enable the world to reduce its reliance on China for REEs, and there is no better time than now to initiate this process.
India Policy Watch #2: Getting Things Right
Insights on current policy issues in India
—Pranay Kotasthane
In the previous edition, I wrote: “Following the AI 171 crash, the DGCA will be hyperactive for a few weeks, conducting spot checks and issuing notices at a frenzied pace, after which we will be back to where we were before the crash.” That’s pretty much how the last week went. The DGCA sent two teams to survey airports, and unsurprisingly found many cases of “recurring aircraft defects, indicating ineffective monitoring, inadequate rectification, and non-compliance with maintenance work orders”.
Nevertheless, my post was about the need to re-engineer the air safety apparatus. In the past week, I managed to gather some more details about India’s current system. It turns out that India has a dedicated Aircraft Accident Investigation Bureau (AAIB), which was constituted in 2012. This is the investigative agency tasked solely with doing a root-cause analysis of aircraft accidents in India. For example, it is the lead agency analysing the data of the AI 171 black boxes.
AAIB was established to comply with the practices of the International Civil Aviation Organisation (ICAO). And its reporting, staffing, and financing structure suggests it was a half-hearted attempt at barely complying with international standards rather than a serious regulatory reform exercise. Here’s why.
One, AAIB is merely an attached office, not an autonomous body. Under the civil aviation ministry, there are three types of organisations. First are the PSUs, such as the Airports Authority of India and Air India Asset Holding Limited. They have their own balance sheets and management boards, and are relatively the most autonomous of the three. The second type of organisation is autonomous bodies like the Airports Economic Regulatory Authority (AERA), Indira Gandhi Rashtriya Uran Akademi (IGRUA), and Rajiv Gandhi National Aviation University (RGNAU). These organisations have some autonomy in their day-to-day operations and financial management compared to attached offices, and have their own set of rules and regulations for internal administration. While they receive grants from the government, they have some flexibility in managing their funds. The third type of organisation is an “attached office”, of which the DGCA, AAIB, and Bureau of Civil Aviation Security (BCAS) are examples. These are the least autonomous organisations because they are under the direct administrative and financial control of the parent ministry. Their budget and expenditure are closely managed by the ministry. How will the regulators pull up other players and PSUs if they are the least autonomous bodies in the ministry?
Two, AAIB is severely underfunded. This week, you would have seen the news snippet that the government is now providing VIP-level security to the AAIB head. That’s good, but it doesn’t take away from the fact that this attached office has remained underfunded for several years. The latest budget doesn't even have a dedicated line item; it is clubbed under the Ministry Secretariat instead. This was also flagged by the latest parliamentary standing committee report, which said:
“the Bureau of Civil Aviation Security (BCAS) and the Aircraft Accident Investigation Bureau (AAIB) receive ₹15 crore and ₹20 crore, respectively. While regulatory compliance remains essential, the rapid expansion of aviation infrastructure - with airports increasing from 74 in 2014 to 147 in 2022 and a target of 220 by 2024-255 - necessitates proportional growth in security capabilities and accident investigation resources. In view of the growing complexity of aviation security threats and the critical nature of accident investigations, the Committee finds these allocations relatively modest.”
Not only is it underfunded, but the report also finds significant underutilisation of funds, with just seven per cent of the allocated funds having been deployed by the Secretariat and AAIB put together.
Three, this organisation only has deputed personnel. As of 2021, AAIB had not recruited directly from the market, which means it has limited expertise, even though it requires specialised human power. Incidentally, we know about its recruitment policies because questions have been raised on the floor of the parliament, not about the shortage of overall capacity, but about whether all the reserved quota seats have been filled in the ministry.
Thus, the least the government can do to honour the many people who lost their lives would be to undertake a regulatory reform that empowers all transport safety and accident investigation authorities. Just a few spot checks and shakedowns won’t do.
PS: I’m sure you've encountered those who argue that Air India’s privatisation is to blame for its current troubles. Remind them that when it was a PSU, it was consistently rated as one of the least safe airlines in the world. As far back as 2013, it was ranked the third least safe airline amongst 60 major players. Back then, the PSU released a statement trashing the report. It chose to raise questions about the report’s methodology rather than to introspect. If anything, keeping it as a zombie firm for this long is the primary reason for India’s airline troubles.
HomeWork
Reading and listening recommendations on public policy matters
[Article] Why is China, a closed politco-economic system, investing heavily in open source AI? This post by Kevin Xu provides some answers.
[Podcast] The next Puliyabaazi is an ode to Tokyo, a magnificently dense yet beautiful urban agglomeration of over 4 crore people.
[Article] The Economist argues increasing military production doesn’t require old-style industrialisation.
