The Subcontracted Republic
In the years the survey can be read on one definition, profit has never taken more of the value India's factories add. The real wage per worker has stood still since 2019-20, and two in every five workers on the floor came to be hired through a contractor, on four-fifths of a direct worker's yearly pay.
Take a thousand rupees of the value India's factories add: what they sell, less what they buy and the wear on their machines. In 2019-20, 403 of those thousand rupees were paid to the people who work in them, and 407 were kept as profit. Two years later, in 2021-22, the employees' part had fallen to 320 and profit's had risen to 556: the most in the 11 years the survey can be read on one definition.
A thousand rupees of the value factories add. 302 to their employees, 495 kept as profit, 203 to interest, rent and provident funds.
The employees' share creeps up: 340 of the thousand.
403 to employees, 407 to profit: as close as the two have come.
Then the coins move. In 2021-22, 556 of every thousand rupees are kept as profit, 149 more than two years before. The employees' part falls to 320.
And they stay there: 530 to profit, 339 to employees in 2023-24.
MoSPI, Annual Survey of Industries 2013-14 to 2023-24 20 of 20 cells MoSPI printed match
- Source
- MoSPI, Annual Survey of Industries 2013-14 to 2023-24: the primary unit-level records, weighted as MoSPI weights them
- Checked
- 20 of 20 cells MoSPI printed match at the printed precision (ASI Summary Results, Statement 3A (shares of NVA, 2013-14 to 2022-23)): each year's shares of net value added, from Statement 3A as printed, within 0.1 points
- Methodological notes
- all industries; every year rebuilt from the unit-level returns on one definition of input, output and rent, because the published series changes definition at 2015-16 and 2018-19
- 11 years: the long-run series reach back to 1974-75 for wages and employment, not for value added
Behind the shares are two different decades. Net value added in India's factories grew from Rs 9,02,055 crore in 2013-14 to Rs 21,13,386 crore in 2023-24, in current prices. Up to 2019-20, the wage bill grew faster than profit, and the employees' share rose from 30.2% to 40.3%. After it, profit took the recovery: 55.6% of value added in 2021-22, 53.0% in 2023-24.
Chapter 1Fifty years of the factory wage
Over the long run the factory wage rose. From 1979-80 to 2023-24 the real wage per worker changed by +2.42% a year deflated by wholesale prices and by +1.09% by workers' consumer prices, while the number of workers grew +2.38% a year. The recent years break the pattern. From 2019-20 to 2023-24 the number of factory workers grew 4.45% a year, while the real product wage moved −0.20% a year and the consumption wage +0.12%. More workers, and the wage per worker flat in real terms.
Fifty years of the factory wage
Wages to workers per worker in manufacturing, a year, and the number of workers below, 1974-75 to 2023-24. Choose the deflator. Years with no series are left as gaps. Hover a year.
MoSPI, Annual Survey of Industries 1974-75 to 2023-24 15 of 15 cells MoSPI printed match
- Source
- MoSPI, Annual Survey of Industries 1974-75 to 2023-24: the primary unit-level records, weighted as MoSPI weights them
- Checked
- workers, all industries: within 0.1% of the published series in 15 of 15 years checked, 1999-00 to 2023-24 (CSO Table 1 to 2008-09 and the RBI Handbook Table 31 from 2017-18)
- wages to workers, all industries: within 0.1% of the published series in 15 of 15 years checked, 1999-00 to 2023-24 (CSO Table 1 to 2008-09 and the RBI Handbook Table 31 from 2017-18)
- workers, all industries: within 0.1% of the published series in 7 of 8 years checked, 1981-82 to 1992-93 (CSO's Table 1, Annual Series (ASI 2008-09))
- total emoluments, all industries: within 0.1% of the published series in 5 of 8 years checked, 1981-82 to 1992-93 (CSO's Table 1, Annual Series (ASI 2008-09))
- Methodological notes
- Years without a series (not interpolated): 1975-76, 1978-79, 1983-84, 1984-85, 1987-88, 1989-90, 1993-94, 1994-95, 1995-96, 1996-97, 1997-98, 1998-99, 2005-06, 2007-08.
Chapter 2The contract floor
Of every hundred manufacturing workers in 2017-18, 36.5 were hired through a contractor; by 2023-24, 41.9. Their wages are a smaller part of the wage bill, which also pays supervisors, managers and other employees: 14.5% in 2017-18 and 17.8% in 2023-24. In 2023-24 a contract worker earned Rs 1,88,715 a year against a direct worker's Rs 2,36,621.
The gap between them narrowed. A contract worker's annual wage rose 46% in nominal terms, a direct worker's 36%, and the ratio went from 0.746 to 0.798. These are annual wages per person: they do not say whether a contract worker doing the same job is paid less.
A thousand workers on India's factory floors. 635 are on the factory's own rolls; 365 were hired through a contractor.
For each rupee a direct worker earned in the year, a contract worker earned 75 paise.
386 through a contractor.
405: the year profit took the most of what the factories added.
419 of the thousand, two in every five, now come through a contractor: 54 more than six years before.
They earn 80 paise for each rupee of a direct worker's year. The survey cannot say whether they do the same job; it can say they work under the same roof.
MoSPI, Annual Survey of Industries 2017-18 to 2023-24 Computed from the unit-level records
- Source
- MoSPI, Annual Survey of Industries 2017-18 to 2023-24: the primary unit-level records, weighted as MoSPI weights them
- Computed
- the panel's mutually exclusive staff groups add to Block E's total persons engaged in 7 of 7 years, exactly
- MoSPI prints no table of contract workers by division; the shares are computed from the factories' own returns
- Methodological notes
- the contract/direct ratio is the annual wage per person, not a worker-level causal wage premium
More contract workers, a narrower gap
Left: workers hired through contractors as a share of all workers, and their wages as a share of the wage bill. Right: the annual wage per person of direct and contract workers, and the ratio between them. Manufacturing, 2017-18 to 2023-24.
MoSPI, Annual Survey of Industries 2017-18 to 2023-24 Computed from the unit-level records
- Source
- MoSPI, Annual Survey of Industries 2017-18 to 2023-24: the primary unit-level records, weighted as MoSPI weights them
- Computed
- the panel's mutually exclusive staff groups add to Block E's total persons engaged in 7 of 7 years, exactly
- MoSPI prints no table of contract workers by division; the shares are computed from the factories' own returns
- Methodological notes
- the contract/direct ratio is the annual wage per person, not a worker-level causal wage premium
Chapter 3Where contract work is
The contract share is highest in coke & refined petroleum (76.1% of workers, where a contract worker earns 0.31 of a direct worker's wage). In motor vehicles it is 56.9% (0.60), in basic metals 53.9% (0.65). The largest employer, food products, has 37.9% (0.88); in wearing apparel it is 12.2%, and a contract worker earns almost what a direct one does (0.996).
Where contract work is, and what it pays
Manufacturing divisions (NIC 2008, two digits), 2023-24: contract workers as a share of workers, and a contract worker's annual wage per rupee of a direct worker's. The dashed line is parity. Hover a division.
MoSPI, Annual Survey of Industries 2017-18 to 2023-24 Computed from the unit-level records
- Source
- MoSPI, Annual Survey of Industries 2017-18 to 2023-24: the primary unit-level records, weighted as MoSPI weights them
- Computed
- the panel's mutually exclusive staff groups add to Block E's total persons engaged in 7 of 7 years, exactly
- MoSPI prints no table of contract workers by division; the shares are computed from the factories' own returns
- Methodological notes
- the contract/direct ratio is the annual wage per person, not a worker-level causal wage premium
The subcontracted republic
The measurements end here; this is the observatory's reading of them. The factory floor is where India was meant to build a working class with secure jobs and rising wages. What the survey records instead, since 2019-20, is a floor where the number of workers grows and the real wage per worker does not, where two in every five are hired through someone else, at a lower annual wage and outside the factory's own rolls, and where the share of the value added that stays with the owners reached its highest in the 11 years the survey can be read on one definition, just as the economy recovered from the pandemic.
Contract hiring is how a factory takes labour without the obligations of employing it: the contractor stands between the worker and the firm, and the firm's books show a service bought rather than a worker kept. The data cannot follow a single worker or name a cause. What they show is the shape of the arrangement, year after year: more value, more of it kept as profit, and more of the people who make it held at arm's length.
How this was measured
Sources. The Annual Survey of Industries, unit-level returns from 1974-75 to 2023-24, analysed by the India Labour Observatory, weighted by the multiplier. Manufacturing throughout.
Who gets the value. Shares of net value added (output less input less depreciation), all industries, every year from 2013-14 to 2023-24 rebuilt from the unit-level returns on one definition; checked against the same shares computed from MoSPI's Statement 3A as printed: 20 of 20 cells MoSPI printed match at the printed precision (ASI Summary Results, Statement 3A (shares of NVA, 2013-14 to 2022-23)): each year's shares of net value added, from Statement 3A as printed, within 0.1 points.
Checked against MoSPI's print. workers, all industries: within 0.1% of the published series in 15 of 15 years checked, 1999-00 to 2023-24 (CSO Table 1 to 2008-09 and the RBI Handbook Table 31 from 2017-18); wages to workers, all industries: within 0.1% of the published series in 15 of 15 years checked, 1999-00 to 2023-24 (CSO Table 1 to 2008-09 and the RBI Handbook Table 31 from 2017-18); workers, all industries: within 0.1% of the published series in 7 of 8 years checked, 1981-82 to 1992-93 (CSO's Table 1, Annual Series (ASI 2008-09)); total emoluments, all industries: within 0.1% of the published series in 5 of 8 years checked, 1981-82 to 1992-93 (CSO's Table 1, Annual Series (ASI 2008-09)). The checks are of all-industries totals, the level MoSPI publishes; the story's series are for manufacturing.
Cautions. ASI years are accounting years. Years with no series are not interpolated. In 2023-24 every non-responding unit has a multiplier of zero, and Block E, the employment block, gains a row for transgender workers.
Data and code. The India Labour Observatory computed every number on this page from the unit-level records MoSPI releases for these surveys, and checks the results against the tables MoSPI prints each time they are recomputed.
✦