Hyderabad's Pharma Cluster

On the western edge of Hyderabad there is a belt of factories that make the active ingredient in other people's medicines. From the road they look like any industrial estate: sheds, tank farms, flare stacks, a boundary wall. What determines whether each one is running at capacity is not visible from the road at all. It is a set of documents held in Maryland.

That is the unusual thing about this cluster, and it is the thing most descriptions miss. It was not built where the raw material was, or where the market was, or where the port was. It was built on a legal permission and then sustained by a regulatory one — and both of those are paperwork.

The decision that made copying lawful

The Patents Act of 1970 did something deliberate: for pharmaceuticals and agrochemicals it recognised process patents and not product patents. A company could patent a route to a molecule, but not the molecule. If you found a different route, you had a lawful product.

The effect on capability was immediate and on a scale that is hard to overstate. Indian firms learned to reverse-engineer, to design alternative synthetic routes, and to run the chemistry at a cost that the originators could not match. The industry that grew out of that — bulk drugs, fermentation, intermediates — was technically sophisticated and commercially cheap at the same time, which is a rarer combination than it sounds.

The regime ended when India brought its patent law into line with the WTO agreement on trade-related intellectual property in 2005. Product patents returned. The industry did not collapse; it moved up the chain, from copying molecules to manufacturing them under contract, from selling in markets that did not ask questions to selling in the ones that ask the most.

From a chemical to a dossier

What a modern plant sells is not a powder. It is a dossier: a document set that proves the product is the same as the reference drug, that the process is reproducible, that the impurities are known and bounded, and that the plant's air, water and records meet a standard. The powder is the by-product of the dossier being accepted.

This is why the cluster's economics look strange from outside. A generic manufacturer's largest fixed cost is not the reactor; it is the analytical laboratory, the documentation team and the regulatory affairs department. The chemistry is a commodity; the evidence is the asset. Firms in the belt compete for chemists and for review slots at the same time, and a plant with a good process and a weak file is worth less than a plant with an average process and a strong one.

The regulator is the map

Hyderabad became a pharma city partly by decision — industrial estates were laid out in the 1970s and 1980s, and the state built the infrastructure that fermentation needs: power, water, effluent lines, cheap land. The other half is an accident of timing. When the generics wave arrived in the 1990s and 2000s, the firms with the chemistry and the English-language documentation were already here.

What followed is a geography written by inspection reports. A facility that passes a US inspection can export to the largest market in the world. A facility that receives a warning letter, or an import alert, loses that market on a defined date, and no amount of commercial demand changes it. Plants open, expand, get reclassified and stop, and the local economy reads the sequence through hiring and land prices.

Genome Valley, laid out from the late 1990s as an organised cluster for biotech and contract research, is the same logic in physical form: a park built to attract companies whose main requirement is a compliant building and a regulatory file.

Manufacturing is not discovery

One thing the cluster has not done, in seventy years of chemistry, is discover new drugs at scale. The capability it built is the capability to make a known molecule cheaply, reliably and to a standard — which is an industrial skill, not a scientific one, and the two do not convert into each other automatically. A company that can run forty steps at commercial scale can still be unable to decide which molecule is worth running.

The reasons are ordinary: discovery is a decade-long gamble priced against a patent monopoly, it needs patient capital that does not exist in the local market, and it needs clinical trial infrastructure that took decades to build elsewhere. What the cluster did instead was insert itself into someone else's discovery pipeline as the manufacturer — lower margin, far lower risk, and a business that scales with other people's patents rather than with its own.

The bottom of the chain

Bulk drug manufacturing is chemistry at scale, and chemistry at scale produces waste. The solvent recovery is never complete; the aqueous stream carries whatever did not crystallise. For two decades the belt around Patancheru and Bollaram ran its effluent through common treatment plants and into the Manjira, and in 2007 a group of researchers published what they found downstream: concentrations of individual antibiotics in the effluent in the tens of milligrams per litre, which is to say at levels that no dilution could explain away.

The finding mattered for two reasons. It put a number on an externality that had been treated as a nuisance, and it introduced a category — a manufacturing effluent that is itself pharmacologically active — that regulators elsewhere had not designed for. Since then, zero liquid discharge rules, tighter consent conditions and the cost of compliance have pushed some production out of the belt. The waste did not disappear; some of it moved.

The dependence nobody chose

The cluster's weakest point is upstream and it is not in Hyderabad. India imports a large share of its active ingredients and key starting materials from China — commonly estimated at around two-thirds — because Chinese producers achieved scale and cost in fermentation and in the intermediate steps that feed it. The dependency was structural rather than strategic: each firm bought the cheapest input, and the sum of those decisions was a national vulnerability.

It became visible in 2020, when shipments stopped and several essential medicines went short. The response — incentives for domestic production of key starting materials and APIs — is an attempt to rebuild a link in the chain that was discarded because it was unprofitable. That is the hardest kind of link to rebuild, because the cost advantage that killed it has not gone away.

So the cluster sits between two documents: an approval that lets it sell, and a shipment that lets it make. It is a very sophisticated industrial base organised around paper, and paper is exactly what can be withdrawn.