The Invisible Commodity

The Logistics, Labour, and Fortunes Behind Human Hair Extensions

1. The Origin: Ritual, Waste, and Survival

At the Kalyanakatta, the tonsuring hall of the Sri Venkateswara Temple at Tirumala in Andhra Pradesh, devotees offer their hair in fulfilment of a vow. In 2012 the Wall Street Journal reported that the official in charge of the temple’s hair collection, categorisation and sale dealt with about 900 kilograms of hair a day (Wall Street Journal, via Hinduism Today). The temple’s trust, the Tirumala Tirupati Devasthanams (TTD), does not keep the hair. It sells the hair by electronic auction to traders supplying wig and extension makers. What begins as a religious offering enters a commercial chain whose prices can be recorded more easily than its full provenance.

This article asks what happens to the hair’s value, and to what anyone knows about where it came from, between the shaving hall and the salon chair.

Temple hair is one of two broad collection systems examined here, and by one estimate it is not the larger one. The anthropologist Emma Tarlo puts daily tonnage across all Indian temples combined at about 5 to 6 tonnes. The rest, often called comb waste, comes from homes. Women collect fallen hair from combs and save it until a collector comes to the door.

The two systems differ in how the first sale is priced, how the hair is processed, who bears the physical toll, and what can be proved about where a finished product originated. Hair shaved in one piece at a temple keeps root and tip aligned, forming the raw material for cuticle-intact hair extensions, often sold as Remy. Hair picked from combs or swept from floors is tangled and mixed, and must be manually untangled, stripped of cuticle with acid and coated with silicone to be used at all.

2. The First Transaction: Who Sets the Price?

At the temple, price formation is institutional. TTD cancelled four conventional auctions in 2011 because it suspected bidders were colluding, and moved to electronic auctions, which that official said were more transparent and competitive (Wall Street Journal, via Hinduism Today). Deccan Herald reported that the 2011-12 auctions raised ₹200 crore, against ₹54 crore the year before. It attributed ₹130 crore to a first phase in September 2011 and ₹70 crore to a second phase held shortly before its 26 February 2012 report, and its standfirst puts the quantity sold at over 550 tonnes (Deccan Herald, 2012). The same paper reported at the time that the first auction alone raised about ₹133 crore for 465,000 kilograms. That report’s summary line, however, gives the weight as 4.65 tonnes, a hundredfold difference within a single source (Deccan Herald, 2011). A separate Business Standard report independently corroborated the maiden auction’s ₹133 crore revenue against a reserve price of ₹105 crore, though it did not specify the weight sold (Business Standard, 2011). The 465,000-kilogram figure is more plausible against the reported total of over 550 tonnes across both phases, whereas the 4.65-tonne summary figure appears to be a unit or transcription error. However, the reports do not explain the discrepancy, and neither figure should be treated as settled fact. The Wall Street Journal reported ₹200 million for the last financial year and ₹54 million for the year before. Each is one tenth of the corresponding Deccan Herald figure, since ₹1 crore is ₹10 million. The Journal’s reported US$3.6 million conversion is consistent with its own rupee figure, but it does not resolve the discrepancy. A first auction reported at about ₹133 crore cannot fit within a ₹20 crore annual total. No TTD or MSTC annual record was located, so ₹200 crore remains Deccan Herald’s reported total, not an independently reconciled or official figure.

Later figures are budget estimates, not realised sales: ₹125 crore for 2018-19 (Business Standard, 2018) and ₹106.7 crore for 2020-21 (Deccan Herald, 2020). Because the bases differ, they are not combined with the 2011-12 figure into a trend.

A single auction reported on 31 March 2018 shows how steeply length sets the price. As the report states them, TTD sold hair in five length categories: first type at 31 inches and above, second at 16 to 30, third at 10 to 15, fourth at 5 to 9, and fifth at 5 inches and below (the report’s fourth and fifth bands overlap at 5 inches). The first type sold at ₹23,071 per kilogram, the second at ₹17,223 and the third at ₹2,833. The fourth type was quoted at ₹1,194, but none of the 700 kilograms offered sold. The fifth type sold at ₹33. Only 500 of 7,800 kilograms of first-type hair were sold, and only 2,000 of 97,000 kilograms of fifth-type hair (Asianet Newsable, 2018). By listed price per kilogram, the first type fetched about eight times the third type, a comparison of prices and not of total proceeds. These are one auction’s results, not current prices, and the length cut-offs are not fixed: a 2022 report gives different thresholds, such as above 27 inches for the first category (Kerala Kaumudi, 2022).

Informal collection works by negotiation between individuals. Kashmir Life reported in 2017 that hair collected from roads in Srinagar was valued at ₹2,000 per kilogram and clean hair from door-to-door collection at ₹2,400. The report does not clearly say which step of the trade those rates describe, so they are not used here to calculate anyone’s margin. The same report describes collectors who are paid in goods such as hairpins, rubber bands and cosmetics (Kashmir Life, 2017). In Bangalore in 2023, a collector who sells to a dealer reported that black hair generally fetches ₹5,000 to ₹6,000 per kilogram, falling to ₹3,000 to ₹4,000 in the rainy season, and that rates are seasonal and give her no fixed income (PARI, 2023).

None of these prices can be set against the temple’s. They come from different years, cities and quality standards, and each informal figure is one trader’s account. What they establish is a difference in structure: a published auction by grade at one end, and private, seasonal, partly in-kind deals at the other.

3. Inside the Processing Economy

Once hair leaves the first buyer, the documented steps are mostly manual. Kashmir Life describes waste hair bundled into balls, put into bags, carried by bicycle to bus stands and trucked to processing hubs. Workers, mostly women in home workshops, wash the hair, untangle it with iron brushes or pins, grade it by length and tie it into small bundles called choti (Kashmir Life, 2017). A 2022 report describes hair gathered in South India, transported to West Bengal, and sorted and detangled before shipment to China (The National, 2022). Untangling takes time: Kashmir Life reported that an untangler produced 50 to 100 grams of choti a day, earning ₹150 to ₹250 (Kashmir Life, 2017). That is a piece-rate observation from one area, not a baseline for the whole sector.

The chemistry is less well documented. A hair-extension brand’s educational pages describe non-Remy hair soaked in acid baths to strip its cuticles, reducing tangling from cuticles that face in opposite directions (Bellami, cuticle alignment; Bellami, Remy vs non-Remy). The same pages state that the stripped hair is coated with silicone to mimic natural sheen. That coating is temporary: as silicone washes out, the hair tangles again. No scientific or industrial study detailing the acid concentrations, bath durations, silicone formulations or effluent treatment was found in the sources reviewed. The chemical process is described by industry participants, but its technical parameters remain largely proprietary.

At each handoff, the records needed to calculate value added and verify origin become harder to find. When hair is untangled at home, workers are paid in cash or by the piece, leaving few payroll records. When choti bundles are consolidated by local traders, hair from dozens of collectors is combined. Once hair is mixed and chemically stripped, its physical origins are indistinguishable. The processing economy transforms a traceable religious offering or an identifiable household scrap into an anonymous bulk material.

4. Across Borders: Mapping the Trade Flows Between India, China, and Beyond

Customs records reveal parts of the trade, but the product codes set the limits of what they can prove. Human hair moves primarily under three headings of the Harmonized System: HS 050100 (human hair, unworked, whether or not washed or scoured; waste of human hair), HS 670300 (human hair, dressed, thinned, bleached or otherwise worked; wool, other animal hair or other textile materials, prepared for use in making wigs or the like), and HS 670420 (wigs, false beards, eyebrows and eyelashes, switches and the like, of human hair). None of these codes separates temple hair from comb waste. HS 670300 lumps worked human hair with wool and animal hair, and HS 670420 includes false beards, eyebrows and eyelashes alongside wigs and extensions. Official trade totals show that goods in these categories crossed a border at declared values. They do not show which hair went into which product, or how much value was added at each stop.

According to UN Comtrade data on WITS, India reported $574.4 million in exports under HS 670300 in 2024, of which $468.35 million, or 81.5 percent, went to China (World Bank WITS, India exports of HS 670300, 2024). In the same year, India’s exports under HS 050100 were $27.02 million, with $13.79 million, or 51.0 percent, exported to China (World Bank WITS, India exports of HS 050100, 2024). Together, these figures confirm that China is the primary direct destination for Indian hair leaving under both headings. Unworked hair accounts for less than 5 percent of India’s combined export value under these two codes. Most declared trade leaves India as dressed, thinned or worked hair under HS 670300, where preliminary grading, sorting and washing have already taken place.

A China Daily report republished on the Henan provincial government’s website describes Juancheng County in Shandong and the city of Xuchang in Henan as centers of hair processing (China Daily, via Henan Provincial Government, 2026). It states that Juancheng firms handle early-stage sorting, cleaning and processing, and supply hair to manufacturers in Xuchang, which turn it into finished wigs and hairpieces for export. The report names one company, Henan Rebecca Hair Products Co. Ltd., and describes it as having an annual production capacity of millions of hairpieces, with research, development, production and sales operations. The account is an official or state-media overview, but it matches the division of labour described in industry sources: initial processing in specialized clusters, followed by manufacturing of finished goods in industrial centers with established access to export markets.

Africanews reported that Xuchang exported over $2 billion of hair products in 2024, that the city makes more than half of the world’s wigs, and that it has more than 4,000 hair-related enterprises (Africanews, 2025). It described Rebecca as a major producer whose products are sold in more than 40 countries and regions. The city-level total and enterprise counts are reported by state or trade bodies through the news service, and are not matched to individual transaction records or customs declarations. They give a sense of the scale claimed for Xuchang’s hair-products industry, but they should be treated as reported aggregate figures, not audited trade statistics.

On the consumer side, the United States reported $768.93 million of imports under HS 670420 in 2024 (World Bank WITS, US imports of HS 670420, 2024). The US import figure covers only goods entering under HS 670420, not the whole hair-extension market. Nor can it be directly compared with a retail-market estimate without a consistent product definition and methodology. The trade data establishes substantial cross-border movement of prepared hair and finished hair articles, but it cannot measure the retail market, the prices and costs along an individual supply chain, or the provenance of an individual batch. Trade statistics can map the movement of broad product categories, but they cannot show whether the value of a particular batch rises fairly, or where that value is retained.

5. The Human Cost: Labour Conditions, Health Risks, and the Xinjiang Enforcement Record

The clearest evidence of hardship concerns collectors. PARI’s Bangalore collector reported covering 12 to 15 kilometres on foot because, she said, bus conductors do not let collectors board state buses, and said the work leaves her with pain in her body and neck (PARI, 2023). A freelance journalist reported in 2025 that collectors in Bengaluru must separate hair from plastic wrappers, food waste or syringes, that handling discarded hair carries social stigma, and that many collectors live in makeshift colonies on the city’s edges (Vasudevan Sridharan, The New World). The anthropologist Dipti Bapat has worked with hair collectors from two denotified nomadic tribes, the Waddar and Vedhwa Waghri, who according to her trade goods for saved hair and earn about a quarter of the minimum wage of a central government construction worker or cleaner (SAPIENS).

Inside the temple’s institutional system, a 2022 report on barbers’ complaints said they are paid ₹11 per head and that employees allege their income is under ₹8,000, without stating the period that figure covers (Kerala Kaumudi, 2022).

The sources reviewed yielded no facility-level occupational-health study, inspection record or worker-interview series for processing workshops and wig factories, so no specific exposures or diseases are claimed for those workplaces. The conditions inside the middle of the chain are the least documented part of it.

Forced labour enters the story through one enforcement action. Effective 17 June 2020, US Customs and Border Protection (CBP) issued a withhold release order directing all US ports of entry to detain hair products produced by Lop County Meixin Hair Product Co. Ltd. of Xinjiang, and merchandise made wholly or partly from them. CBP said the order was based on information that reasonably indicated the use of prison labor, with further indicators of forced labor that included excessive overtime, withheld wages and restricted movement. The order rests on 19 U.S.C. 1307, which bars imports made wholly or partly with forced labor (CBP, withhold release order). On 1 July 2020, CBP announced that officers at the Port of New York/Newark had detained a shipment of almost 13 tons of hair products worth over $800,000 under that order. The importer of record could export the shipment or demonstrate that the goods were not made with forced labor (CBP, shipment detention). A withhold release order is an agency enforcement action, not a court finding. It applies to one named company, and it does not show conditions across China’s hair industry or across Xinjiang. The sources reviewed do not establish the shipment’s final disposition.

Global figures give context but are not hair-industry data. The International Labour Organization estimated that 27.6 million people were in forced labour on any given day in 2021. Its separate estimate of $236 billion in annual illegal profits covers the private economy only and excludes state-imposed forced labour (Associated Press report on the ILO’s 2024 study).

6. Who Captures the Fortune? What a Measured Value Ladder Needs

To show who earns what, a value ladder must follow one specified product from purchase to sale, in a common unit, with every price identified as observed or estimated. The sources reviewed support only the bottom rungs, and those rungs cannot be chained together.

*Reported prices at different collection and sale points, not a comparable price series*

Reported rate	Price	Where and when	Source
Road-collected hair	₹2,000 per kg	Srinagar, 2017	Kashmir Life
Clean, door-to-door hair	₹2,400 per kg	Srinagar, 2017	Kashmir Life
General rate a collector gives for black hair she sells to a dealer	₹5,000 to ₹6,000 per kg (₹3,000 to ₹4,000 in rains)	Bangalore, 2023	PARI
Temple e-auction, first to fifth type	₹23,071 down to ₹33 per kg	Tirumala, March 2018	Asianet Newsable

The Kashmir Life source does not clearly identify the transaction stage for its two rates, so they are not treated as an aggregator’s purchase price, and no collector or aggregator margin is calculated from them. The same report quotes traders’ own stated margins of ₹20 to ₹100, ₹50 to ₹100 and ₹100 per kilogram, but does not tie them to those two rates (Kashmir Life, 2017). The clearest observed pattern is within the temple auction, where grade matters enormously: by price per kilogram, the first type fetched about eight times the third. Nothing in these rows says what factories, distributors, brands or salons earn.

Three measurements are easily confused. Price is what a buyer paid for a defined lot. Value added is the additional value created at a stage, after accounting for inputs and processing. Profit margin is the revenue remaining after the relevant costs. Every figure above is a price, and a difference between two prices is neither value added nor profit. Nothing here shows that brands or stylists earn the largest margins. Prices would also need to be converted to a common unit. A price quoted for a 100-gram bundle, for example, must be multiplied by ten to compare it with a per-kilogram price.

No documented transaction was found for the later stages: factory export price, freight and duty, distributor price, retail price, or salon labour. A matched transaction would need to reveal the input mass and output mass, processing costs, export value, distribution price, and retail price. Those records would make it possible to calculate yield loss and estimate value added at each stage rather than guess. Invoices or customs entries for a single consignment, a processor’s records, and a retailer’s price and cost for the same product would supply the necessary evidence. Until someone publishes them, who captures the fortune remains an open empirical question.

7. The Traceability Test: Regulatory Mandates and the Limits of Origin Claims

Three legal frameworks make supply-chain evidence increasingly consequential, operating at different levels of the value chain. In the United States, the Uyghur Forced Labor Prevention Act (UFLPA) became law on 23 December 2021, and its rebuttable presumption took effect on 21 June 2022. Under the UFLPA, goods produced wholly or partly in Xinjiang, or by an entity on the UFLPA Entity List, are presumed prohibited from entering the United States. An importer seeking to overcome that presumption must satisfy the law’s clear-and-convincing-evidence standard (CBP; WilmerHale). In the European Union, Regulation (EU) 2024/3015 entered into force on 13 December 2024 and applies from 14 December 2027 (European Parliament Legislative Observatory; European Commission). Article 3 prohibits economic operators from placing or making available products made with forced labour on the Union market, and from exporting them. The regulation’s investigation and enforcement procedures determine whether a product falls within that prohibition (Regulation (EU) 2024/3015, EUR-Lex). Alongside these product-level market prohibitions, the European Union’s Corporate Sustainability Due Diligence Directive (CSDDD), originally adopted as Directive (EU) 2024/1760, was subsequently amended by Directive (EU) 2026/470. The amended framework changes the directive’s scope and implementation timetable, so its requirements should be assessed against the amended text rather than the original directive (European Commission; Directive (EU) 2026/470, EUR-Lex). None of these regimes makes a genetic test a substitute for evidence about how and where a product was made.

Whether origin can be shown has two separate parts. The first is scientific. Forensic laboratories routinely recover mitochondrial DNA from shed hair, but it has less discriminating power than a nuclear profile, and its quantity and quality fall along the length of the shaft. Sequencing work by FBI Laboratory scientists on rootless shed hairs found nuclear DNA, in fragmented form, to be abundant relative to mitochondrial DNA: at least 88 percent of human sequence reads in every sample examined, and generally more than 95 percent (Genes, 2018). The authors stress that the fragments are very short, so routine profiling methods still fail on most such hairs. The samples came from nine donors, and two of them had permanently coloured hair. The study reported no apparent link between hair treatment and mitochondrial DNA quantity or quality in the recent hairs, but two samples cannot show how the bleaching and dyeing used on extension hair affect these markers. Even a technically successful DNA test would point to a person or a maternal line, not establish who collected the hair, which trader bought it, or whether the tested sample remained linked to a documented batch.

The second part is documentary, and it is central to proving where a product came from. Trade accounts describe hair passing through several hands: collectors, local middlemen who send it by courier and air to processing factories in Bengal, and then processors in China (Kashmir Life, 2017). Whether lots are blended along the way, how often, and whether purchase records survive each handoff are not documented in the sources reviewed. Brands do make sourcing claims. One brand announced extension packs each paired with a QR code that lets stylists see the step-by-step process (MedEsthetics), and another says it has a protocol to track hair from donation to the finished product (Rapunzel of Sweden). Both are the companies’ own descriptions. No audit, survey of brands or sample of origin certificates was located that tests them.

The central question remains unanswered: can buyers and regulators verify a hair product’s origin through commercial records, even when processing and batch mixing complicate the chain of custody? That requires records connecting collection, purchase, processing and the finished product, backed by independent verification. Until those records can be matched to a specific product, the chain remains only partly visible. The sale may be recorded, the shipment counted and the finished product priced, while the route between them remains unproved.