#587 1981 · Haiyan Shirt Factory (海盐衬衫总厂, Bu Xinsheng) · Manufacturing / textiles
A near-bankrupt state shirt factory paid workers by the piece with no ceiling, but the real fix was on defects: one ruined shirt cost two good shirts' worth of pay, a penalty too steep for speed to ever be worth the quality risk.
the problem
piece-rate pay alone rewards speed and gives workers every incentive to cut quality corners
background
By the late 1970s, China's state-run collective factories, including the Haiyan County shirt factory in Zhejiang, ran on flat wages regardless of output — the 'big pot' (大锅饭) system that paid the same whether a worker produced ten shirts or a hundred, flawless or defective. The factory was losing money and had no mechanism to reward the workers actually capable of producing more, faster or better.
The obvious first fix, switching to pure piece-rate pay with no floor or ceiling, solves the speed problem but creates a new one: a worker paid strictly by volume has every incentive to rush and let quality slip, since a merely proportional penalty for a defective unit, losing the pay for that one bad shirt, still leaves speed as a winning strategy if it lets a worker produce enough extra units to cover the loss.
what everyone would do
The obvious first fix for a flat-wage 'big pot' system with no output incentive was switching straight to pure piece-rate pay with no ceiling, rewarding volume directly, an approach that solves the speed problem but leaves a proportional penalty for defects, losing pay just for the one bad shirt, that still lets a rational worker treat quality as a tradeable cost against enough extra speed.
what they saw
Bu Xinsheng saw that a merely proportional defect penalty didn't actually close off the quality-for-speed trade, it just priced it: a worker losing only the value of one ruined shirt could still come out ahead by producing enough extra units fast enough to cover that loss. The fix wasn't proportional punishment, it was making the penalty for a defect steeper than the value of the corner cut, costing a worker two good shirts' worth of pay for one ruined shirt, so no plausible amount of extra speed could ever make cutting quality corners a winning strategy.
the move
In 1981, under director Bu Xinsheng, Haiyan switched workers to pure piece-rate pay with no ceiling, 'produce more, earn more' (多劳多得), but paired it with a deliberately asymmetric defect penalty: ruining one shirt cost a worker the value of two good shirts, not one. The penalty was calibrated so that no amount of extra speed could make cutting quality corners a winning trade, summarized in the factory's own blunt slogan, 'you smash my brand, I smash your rice bowl' (你砸我的牌子，我砸你的饭碗).
why it works
Pricing a ruined shirt at double the value of a good one meant the math no longer favored rushing under any realistic production speed, since the penalty exceeded whatever gain a worker could recoup by producing faster, closing the exact loophole a merely proportional penalty leaves open. Because the piece-rate pay itself still had no ceiling, workers retained the full incentive to genuinely produce more, the fix specifically targeted the quality-corner-cutting behavior without dampening the volume incentive that was solving the factory's original output problem. This precisely calibrated asymmetry is what let Haiyan capture both benefits simultaneously, total industrial output value rising 4.3 times between 1978 and 1983 while profit rose 4.2 times, evidence the factory gained genuine volume without the quality collapse a pure, unpenalized piece-rate system would have risked.
the payoff
By 1983, the factory's total industrial output value reached ¥10.29 million, up 4.3 times from 1978; profit rose 4.2 times and tax contribution 2.6 times over the same period. The reforms drew national attention when People's Daily profiled Bu Xinsheng in November 1983 and Communist Party General Secretary Hu Yaobang personally endorsed the experiment, followed by state television coverage in February 1984 — People's Daily later noted Bu was, by coverage volume, the second-most-featured 'model figure' in the newspaper's history after Lei Feng.
where it breaks
The mechanism depends on the penalty actually being enforceable and consistently applied, since an asymmetric defect penalty that management fails to levy reliably, whether from favoritism, oversight gaps, or workers finding ways to hide defects before inspection, would collapse back into the same corner-cutting incentive a merely nominal penalty creates. It also depends on workers actually being able to detect and control the quality of their own output through effort and care, since penalizing defects heavily only changes behavior if the defect rate genuinely responds to how carefully or quickly a worker works, rather than being driven by factors like faulty equipment or poor materials outside the worker's control. And setting the penalty too steep relative to a worker's typical earnings risks discouraging workers from taking the piece-rate work at all, or pushing them toward excessive caution that undermines the volume gains the pay structure was meant to unlock in the first place, meaning the exact ratio, here two good shirts for one ruined one, has to be calibrated to the actual economics of the specific product and workforce rather than applied as a fixed universal multiple.
what came after
Bu Xinsheng was named one of China's 100 official 'Reform Pioneers' (改革先锋) at the 40th-anniversary celebration of reform and opening-up in December 2018, specifically honored as a pioneer of urban collective-enterprise reform, and the Haiyan shirt factory's asymmetric quality penalty is still cited in Chinese management writing as an early, working example of designing an incentive's downside steep enough that cutting corners is never rational, rather than merely proportional to the harm done.
references
- [1]改革先锋步鑫生:撕掉递来的条子,"谁也不许左右企业"新浪财经 (Sina Finance), 2019finance.sina.com.cn
- [2]【改革先锋国企人②】城市集体企业改革的先行者步鑫生央广网 (China National Radio), 2018news.cnr.cn