IMAGE: SHUTTERSTOCK
BALENDRA SHAH’S government has a vision. Nepal will become a digital dynamo, exporting $30bn worth of software, services and solutions within a decade. That is the promise of the Rastriya Swatantra Party (RSP), which swept to power in March with ambitious plans to turn the country into an IT powerhouse. The target is inspiring, the kind of round number that sounds good in press releases and plays well in Kathmandu’s cafés and events. It is also fantastical.
Consider the starting point. Nepal’s annual exports of IT services—software development, business process outsourcing, cloud-based support and the like—now stand at roughly $1bn, according to the Nepal Association for Software and IT Services Companies (NAS-IT), an industry association, though there is no independent verification of that figure. “As an industry practitioner, I am very confident we have reached that billion-dollar number,” Gaurav Pandey, president of NAS-IT, told the press early this year. That is an achievement in itself: exports have more than doubled in three years, lifted by a young, English-speaking workforce and a favourable time zone for Western clients.
Yet to strike $30bn within a decade, the industry would need to grow at an annual rate of nearly 40%. For context its recent blistering pace has been 11% a year, estimates the World Bank. More than tripling the growth rate would require not only better policies but also a transformation of Nepal’s economic geography; its education system; and its place in global supply chains. No country has ever pulled off such a leap in ICT services without a corresponding explosion in physical infrastructure as well as capital investment and skills: things Nepal embarrassingly lacks.
The RSP’s target becomes even more delusional when you break down what $30bn actually means. That sum would be more than ten times Nepal’s total annual exports of goods, which stand at roughly $2.3bn, and would exceed the country’s entire GDP of about $45bn by a wide margin. India, the undisputed heavyweight of South Asian IT, exported around $205bn in software and IT-enabled services in the most recent fiscal year. For Nepal to match that intensity, its IT sector would need to be nearly three times more dominant relative to the size of its economy than India’s is. That is impossible.
None of this is to say Nepal cannot grow its IT sector. It can, and probably will. The ingredients are there: a median age of 25; widespread English proficiency; and a diaspora that sends back remittances, as well as technical know-how and client connections. The government’s declared “Decade of Information Technology” has already produced sensible steps, including a 75% tax exemption on export earnings for IT firms. That has spurred a wave of new registrations, from 2,000 or so companies a few years ago to over 4,000 today. Walk through any business district in Kathmandu and you will find cramped offices full of young developers writing code for Australian health-tech startups or American fintech firms.
But the constraints are just as pervasive. Nepal’s broadband infrastructure, though improved, remains patchy. Fixed fibre-optic coverage leaves nearly half the population more than ten kilometres from a backbone node. Bandwidth is expensive, partly because the country must import it via India and China, adding foreign-exchange risk to an already fragile economy: in 2023 alone bandwidth imports cost Rs4.7bn. Electricity outages, once a daily nuisance, have become less frequent but still disrupt work in ways that competitors in Bangladesh or Vietnam do not tolerate.
And then there is the brain drain. Nepal produces roughly 10,000 engineering and IT graduates a year, but a staggering proportion leave for Australia, Europe or the Gulf within 24 months of graduation. The ones who stay tend to work remotely for foreign firms, which is fine for individual incomes but does little to build the institutional capacity or managerial depth that a $30bn export industry would need.
There is, however, a more immediate threat to the RSP’s dreams: artificial intelligence. The same technology that promises to boost productivity is eating the entry-level jobs on which Nepal’s IT outsourcing model depends. A World Bank report released in October found that about 7% of jobs in South Asia are highly exposed to AI with low human-AI complementarity, meaning they are ripe for replacement.
The most vulnerable workers are young, moderately educated and concentrated in IT and business services: a perfect description of Nepal’s tech workforce. Since the launch of ChatGPT, monthly job listings for the most automatable white-collar roles have fallen by around 20% the world over. Entry-level coders, bug fixers and customer-support agents are finding that the work they trained for is being done faster and cheaper by machines. As one industry observer puts it, “Basic coding, bug fixing and support jobs are being automated. AI is changing the job market very fast”.
The trouble is that Nepal’s comparative advantage—a young, English-speaking workforce willing to perform routine digital tasks for low wages—is exactly what generative AI is best at replacing. Advanced voice agents and conversational AI are making traditional call-centre jobs obsolete. A detailed analysis of Nepal’s BPO sector published this year warned that generative AI “is poised to eliminate thousands of traditional voice-based BPO jobs in Nepal by automating customer service”. The low-value, script-driven BPO job, the report concluded, “is dying”.
None of this is to say that AI offers no opportunities. The same report points to a potential $500m annual market for “AI Tutors” specialising in Reinforcement Learning from Human Feedback (RLHF): the painstaking work of teaching large language models human-like judgement.
But seizing that opportunity means a workforce trained in new skills not the old ones. As one tech analyst notes, “New jobs are being created, just not the old kind. There is growing demand for AI trainers, prompt writers and people who can test AI models”.
Whether Nepal’s education system can pivot quickly enough to supply those workers is an open question. The World Bank has warned that without stronger digital foundations and skills training, Nepal risks seeing its youth and graduates “sidelined as AI reshapes professional work across borders”.
The RSP might argue that its target is not a forecast but a goal, something to stretch towards. That is a familiar political dodge. But goals that bear no relation to reality do not stretch: they mislead. They encourage the wrong kind of policy, such as throwing money at vanity projects like hardware manufacturing, which Nepal has no comparative advantage in, rather than fixing the basics: reliable power, cheaper bandwidth, faster visa processing for foreign clients and a regulatory regime that does not treat internet service providers like tax evaders. The effective tax burden on smartphones in Nepal can surpass 70%, making them unaffordable for the very rural populations that might otherwise become the next generation of programmers and support staff. Even the corporate tax rate for telecom and internet service providers stands at 30%, the same rate levied on alcohol and tobacco: a peculiar disincentive for a sector the government claims to champion.
The big problem is one of sequencing. IT services tend to flourish only after other foundations are in place, such as predictable regulation and a financial system capable of handling cross-border payments without friction. Nepal offers neither: its policy framework is dated and erratic, and its central bank is cautious to the point of paralysis. If it is serious the RSP should push through the harder reforms—privatising telecoms; liberalising capital controls; and overhauling university curricula—that would be necessary even to approach a $10bn export target, let alone $30bn.
A more plausible path looks like this. If Nepal maintains its current growth rate of 11% a year, IT exports would reach roughly $3bn by 2035. If it accelerates to 20%—a heroic assumption given the infrastructure and skills bottlenecks—the figure might touch $6bn. To hit $30bn, the industry would need to add the equivalent of India’s entire IT services growth over a decade, but starting from a base one-fiftieth the size.
The RSP deserves credit for raising the profile of an industry that has been overlooked. But a serious IT policy would set realistic milestones; invest in the tedious work of upgrading vocational training and spectrum allocation; and accept that $30bn is not even in the realm of plausibility. For now domestic software writers will keep doing what they do best: solving other people’s problems. The country’s own problem—how to turn digital promise into national prosperity—will take longer to debug. ■
