Piece rate is good

Tyler Cowen of Marginal Revolution mentions the work of the economist Edward Lazear yesterday. Lazear argues that piece rate may boost productivity which brings me to another of my pet peeves: most people assume that piece rate is a nasty exploitive practice. I’d never be accused of being an apologist for the apparel industry but this myth is one that needed to die a long time ago. The truth is -at least in sewing factories- piece rate is popular -very popular! Why hasn’t anyone ever bothered to ask sewing operators what they think of piece rate? Sewing line operators prefer piece rate, they make more money -not less- and productivity increases. Everybody wins. What’s not to like?

From the Stanford Graduate School of Business magazine comes this example:

In his paper “Performance Pay and Productivity” Lazear writes: “When workers receive an hourly wage, the more able receive no reward for putting forth additional effort. As a result, they choose to reduce effort to the minimum required. When a piece rate is paid, workers who produce more, earn more. This provides an incentive to workers to put forth more effort.”

Workers at Safelite did put forth more effort: Their everyday output increased. Meanwhile, absenteeism dropped off sharply. The level of paid sick hours fell by 61 percent, and the company lost fewer of its most efficient workers. “People who like a piece rate are the most able; those who like an hourly rate are the least able,” says Lazear. “So when companies institute a piece rate plan, turnover among the most able goes down.”

The net result was that the cost of production for the company fell. While pay for the average employee rose 9 percent, productivity rose 24 percent. As Lazear writes, “The piece rate plan…made both capital and labor better off.”

There’s very little resistance from the labor force, including unions, says Lazear, because workers are guaranteed a minimum hourly rate. And while there’s more incentive to get work done quickly, quality doesn’t seem to suffer. In the case of Safelite, each windshield could be traced back to the employee who installed it, so there was the possibility of censure for doing slipshod work. This seems to have kept standards high.

Now, if you’re just starting an in-house sewing operation, I know it’s hard to figure out a fair compensation structure. For that reason, I don’t recommend you do that right away. At the outset you should pay hourly. Once your sewing operators have learned the job and you’ve seen a rough correlation of productivity (time per product completed), you can develop a piece rate structure based on completion time. Now, you must structure piece work so that nobody loses. By that I mean you have to make sure that your least productive worker can make the previously established hourly wage.

If you’re gravitating towards a team sewing structure (Toyota Sewing System), you may see some fall-out. By team sewing, I mean that a unit of stitchers will produce products jointly and the team’s pay is based on per unit completion. In these cases, you may be very surprised to learn that your most productive team workers will quickly become more intolerant of their under performing co-workers than you ever could. In these cases, it’s not unusual that teams will attempt to push co-workers out of their unit. In these cases, I’d suggest the first strategy would be to encourage the more productive workers to help the least productive workers to become more efficient. If under performers continue to lag, you may need to put them into another team. If the worker is pushed out of the second unit, they may well leave on their own. Otherwise, it’s likely you’ll find this person is in a job that’s not well-suited for their abilities. In this case, ask them what job they’d like or what job they feel they could be successful doing. You may be very surprised to find that what they’d really want to be doing could be a better fit for both of you and their subsequent productivity could soar.

There are two downsides to piece work that will require management and tact. The first downside is when you first implement the piece rate structure; you may find that you’ve mis-timed an operation and are in fact, paying double what you had been paying. In this case you have two choices. One, you can decide that the increase in production (greater capacity) is worth it and the price you’re paying is fair compensation. Second, you can adjust the pay rate. Exercising the second option will most likely result in some short term resentment. I’d recommend practicing an open-book strategy if you find that you’ve miscalculated the appropriate pay per operation. In this case, I’d think you could openly discuss this and negotiate to split the difference.

The second downside to piece rate is the barriers it can potentially create toward impeding incremental improvement (Kaizen). In becoming a lean manufacturer, I mentioned that an overly rigid piece rate structure will prevent workers from reporting problems if they lose money by speaking up. You must design your feedback system in such a way that employees don’t lose machine time (income). Perhaps one solution would be to pay a bonus for problem reporting.

I’d be interested in hearing about how all of you have used pay and incentive systems, the problems with each as well as any evolutionary pay strategies you’ve developed.

Get New Posts by Email


  1. christy fisher says:

    I have done piece rate for years (and as a former line sewer, I loved it). My workers can average $15-$20 an hour after they get the hang of a new pattern.
    Something we did on the line, that I have implemented into my system is that I start a new worker(or a new pattern) with hourly and pay piece rate VS hourly (whichever is greater) for a set period of training…the workers soon jump past the hourly with the piece rate.
    We set piece rates by timing 10-20 of a new style.
    Often I do the first batch myself to make sure the rates are fair.. then I pass a batch to my prime asssemby person.
    This is where KNOWING HOW TO SEW AND ASSEMBLE THINGS YOURSELF comes in very handy.
    I never give workers something that I wouldn’t be able to do myself at a fair rate.
    You also do not want to time your rates with a “home sewer”..they will be slow and your rates will be way too high.

  2. Eric H says:

    Productivity is normally defined as “output per unit of input”. If workers are producing more pieces per hour (increased labor productivity), and they are paid by the piece, then they get paid more per hour, not per finished piece. The employer’s costs per unit remain the same, but you get more units per hour (which reduces fixed price costs per unit, like electricity to light the factory, rent, interest on loans to buy equipment, management costs, etc.).

    I think Kathleen has identified a huge problem with piece rate pay with respect to jidoka (line stops initiated by operators), but we should look at that as an opportunity and challenge rather than fatal flaw. A bonus for problem reporting is an interesting start, but figuring out the correct compensation for suggestions is just as difficult for that as for figuring out the piece rate.

    In Kaikaku, I think it was, Norman Bodek relates the story of a Japanese worker who was paid roughly $0.20 for each suggestion. He and his wife sat at the table every night and brainstormed all the suggestions they could think of. Were they all worth $0.20? Probably not. Were a few worth more? Probably. Were one or two worth 10s or 100s of thousands? Possibly. Was the cost worth it to the employer? Almost certainly. Imagine your entire workforce staying up every night to come up with 10-20 or more suggestions – what are the odds that a few might result in radical savings?

    On the other hand, there are likely to be lots of silly, vague, impossible, unfocused, useless, and redundant suggestions, so should a person be compensated as a percentage of the potential savings? That would certainly get suggestions that are focused, and employees eager to see them through to completion. Potential savings are notoriously hard to calculate, but I do remember reading recently about an IBM program where a suggestion program resulted in $Millions in savings and $Thousands in cash rewards. IBM decided the program was too expensive (!), rescinded it, and the suggestions dried up. Apparently, their bean counters couldn’t subtract.

  3. Teijo says:

    I feel compelled to comment on Japanese terminology, since I read and write the language.

    Unless the author uses unusual ideograms, the term “jidoka” means “automation.”

    Japanese terms used in “kaizen” (improvement) manuals usually have exact English language equivalents. Consultants like to use unusual words because they can make the customer think about what is being said and tend to raise the (perceived & monetary) value of advice given. However, Japanese terms are not magic, and often using the English is more precise, and more comprehensible to the reader.

  4. In Sri Lanka we have monthly minimum salary, irrespective of number of work days in a month. So the Salary in January and the February are same, though the numbers of days in each are 28 and 31 respectively.

    To motivate better workers to achieve higher productivity, we use individual efficiency based incentive.

    One of the simple systems in use is following:

    It is measured on a days out put. If you achieve over 60%, 70%, up to about 120%, the sewing factories payment range is $0.25 to $2.5, extra per day for Machine operators.

    There are many variations to the above, from time to time based on management thinking prevailing at that point in time.

  5. SB says:

    Keerthi, Is there any system in place to encourage feedback from workers, as to how to improve the process or to indentify problems?

  6. Eric H says:

    Thanks, Teijo. I have been skeptical of the translations given to such words in the consultant-driven press, and you have provided an illustration of the reason for my skepticism. In the canonical “Lean” books, jidoka is usually translated as a “line stoppage” or a mechanism by which that is accomplished. I think I will accept your advice on this.

  7. Hugh E. says:

    My colleague and I recently began to manage a sewing factory in which all seamstresses are paid on piece rates. We have observed some notable facts regarding piece rate that are not immediately obvious to those unfamiliar with the system.

    First, I will agree wholeheartedly with Kathleen that sewing line operators love piece rates. We try to set rates such that an average seamstress exerting herself somewhat is able to exceed the state-mandated minimum wage. Setting rates too low eliminates all motivation, as no matter how fast they work, the operators will never break through minimum wage. The result of our set point is that our average operators do well, and our superstars do very well indeed, earning very high hourly rates.

    I concur with Eric H.’s comment above that increased productivity on piece rate increases operator earnings while maintaining unit costs. Also, in the case of a production push that requires overtime, the only additional payment nominally due to the operators is 1/2 of their guaranteed hourly rate – if they’re good enough to be allowed to work overtime, they’re already earning their hourly rate in piecework. The high producers often create enough value in their overtime to eliminate the need for any ‘unearned’ overtime payment, which means that their extra time is essentially free. As long as they’re earning more at piece rate than their guaranteed rate x 1.5, the additional cost is obviated by their productivity.

    My argument against piece rates is that while they do motivate operators very well to push as many pieces through their machines as possible in any given time, they are a direct barrier to other initiatives I’m trying to introduce to the floor, such as recording of quality statistics, new ways of organizing production, etc., etc. Operators see any new initiative as getting in the way of pushing parts through the machines, and thus become resistant to the idea of any change at all. We are going to try team sewing here shortly – I’ll post the results.

    Finally, on the use of Japanese terminology in Lean efforts: I’ve lived in Japan for five years and speak the language almost fluently. Reading Lean articles for me sometimes takes on the feeling of reading religious tracts – the true believers roll out their jargon in Japanese, and the uninitiated are expected to be awestruck. I once read an article entitled “The Gemba Walk” in which somebody waxed eloquent on the critical importance of “Walking the Gemba” and all of the ritualistic hokum that surrounded it. The manager described in the article apparently spent so little time on his factory floor that his appearance there on a “Gemba Walk” was “an event” worthy of awe and deference and special procedures. “Gemba” in Japanese literally means “The Actual Place” or “The Site” – phrases not too awe-inspiring. I’ve got a factory floor, too, and I generally walk down there about a dozen times a day. It’s no big deal. I say hi to my ladies, and we get a bunch of stuff done. I’m a great supporter of Lean efforts, and am working to move our operations in that direction. But I don’t need to do it in Japanese.

  8. Eric H says:

    Ironically, I scrupulously avoid the use of any reference to Japanese phrases or personalities in my own workplace. If I say, “We’re going to work on process improvement,” I can almost hear them thinking, “Oh good, as long as we don’t have to do Lean, kaizen, quality circles, TQM, or any of those Japanese fads!”

    The only exception is that I refer people to our copy of Hirano’s 5 Pillars, which struck me as an amazing book.

  9. Mike C says:

    I concur with Eric H.’s comment above that increased productivity on piece rate increases operator earnings while maintaining unit costs. Also, in the case of a production push that requires overtime, the only additional payment nominally due to the operators is 1/2 of their guaranteed hourly rate – if they’re good enough to be allowed to work overtime, they’re already earning their hourly rate in piecework. The high producers often create enough value in their overtime to eliminate the need for any ‘unearned’ overtime payment, which means that their extra time is essentially free. As long as they’re earning more at piece rate than their guaranteed rate x 1.5, the additional cost is obviated by their productivity.

    This is not true in the United States.

    You cannot set an arbitrary “artificial regular rate” to avoid paying overtime. A quick Google on “piece rate overtime” will yield loads of information on the topic.

    There is much about the FLSA that I don’t care for (especially all the parts about not allowing adults to enter into mutually beneficial contracts) but its still the law and attempts to circumvent it can yield *very* costly results.

  10. Alison says:

    I’d be interested in knowing what operators are getting for piece rate. Kathleen, in //fashion-incubator.com/mt/archives/problems_in_problem_prevention.html you said that “Even in 1995, the average stitcher earned $9.73-$11.27 on piece rate.” If I understand Hugh E’s post correctly, his superstars earn up to 150% of his state’s hourly minimum wage, and even a little more. Depending on the state he works out of, that means that his slow operators earn betweeen $5.15 and $8.50 per hour and his superstars earn between about $8.00 and $13.00 per hour.

    I don’t know what the change in the cost of living has been since 1995 – but if it went up an average of 2.5% per year, then the comparable 2005 stitchers should be making an average of $12.50 to $14.50 per hour.

    What are operators actually getting these days? (And does anybody know what the change in cost of living has been since 2005?)

  11. Alison says:

    Ok, found the US CPI. The CPI for 2005 was 128.4% of the CPI for 1995. So the 1995 “average stitcher” earning $9.73 to $11.27 per hour would need to be making $12.49 to $14.47 per hour in 2005. What I said.

  12. Bill Waddell says:

    My two cents:

    First, lean is hard enough without having to learn to speak Japanee at the same time. The definition of “jidoka”, as far as I am concerned is “irrelevant”. I have noticed that there are a lot of Lean ‘experts’ who seem to know a lot more about Japanese vocabulary than they do about manufacturing.

    Second, piece work is an alluring siren song that manufacturers find hard to resist, but there are huge problems with it. Quality often suffers, for one. The bigger problem is that most companies have much higher overhead costs than direct labor costs. Piece work seeks to optimize direct labor, but does not contribute to overhead cost reduction, and often creates incentives for workers to do things that increase overhead costs. There are very good reasons why the best performing lean companies pay for skills, rather than production.

  13. don says:

    The problem I see with piecework is does it match Takt time. If not then overproduction follows. This results in increased cost thru inventory, movement, space, and the possibility of obsolesence and scrap.

  14. Kathleen says:

    This results in increased cost thru inventory, movement, space, and the possibility of obsolesence and scrap.

    Before I forget, read The genius of superfactory readers on evolving excellence blog. It’s a post that Bill Wadell wrote about me and the particular idea I’m about to explain to you.

    This goes to the heart of my continual assertion in lean circles that not all industries are created equal. Due to inherent differences between classes of industries, they won’t be able to apply lean principles in the model most currently espoused and discussed. For example, agriculture. You cannot pull corn like cars. Does this mean agriculture can’t get leaner? Of course not.

    I believe that industries that are “closer to the dirt” like agriculture, apparel and housing (echoing the primary needs as described by Maslov as food, clothing and shelter) will have facets of their businesses that cannot follow the model most currently depicted by Toyota. Lean looks different in different industries. Making single count units to order in apparel isn’t tenable anymore than producing one single serving of corn is. I argue that some industries can never totally eliminate batch production due to fundamental differences inherent to their products.

    I’ve been trying to direct the discussion of lean -in this corner of manufacturing- toward exploring what concepts apply to us and which don’t. Some facets are already used by many firms, they just don’t know that’s what it is. We’ve done some of these things well before Lean became known. For example, before 1980, it was largely unheard of for a manufacturer to cut and sew a line without having secured the orders beforehand. That was “lean” and that’s the way manufacturers used to produce.

    Rather than using the protocols established by Toyota -such as the disparity you mention with takt time- I think we need to explore what concepts apply to us, what concepts are a force in our industry but not in others, and what it is that we can reasonably accomplish, all while reducing waste. The fact remains, for most cut and sew businesses, at least some batch processing is unavoidable. Ask any farmer. He’ll tell you the same :)

  15. Andy Chang says:

    Piece Rate is good for motivation and quality can be easily tied into the system using a tier structure.

    In my factory I think the pain is calculating the over time pc rate, it is very labor intensive and with over 2000 workers, it requires a team of people to calculate.

    A second problem is as productivity is improved, it is more difficult for factory to re-adjust pc rate once it has been published. As pricing pressure becomes difficult to abosorb, it is difficult for sewing line workers to accept the fact that their piece rate need to chage as well

  16. Alison Cummins says:

    “As pricing pressure becomes difficult to abosorb, it is difficult for sewing line workers to accept the fact that their piece rate need to chage as well.”

    Yes, but this goes for anybody. I work in telecom, and we’ve been subjected to huge pricing pressures over the last seven years. We’ve had layoffs, restructurings, and simply more work piled on our plates with the casual instruction to “work smarter, not harder.” But nobody has come to me and said “Oh, by the way, your salary is going to be cut 10%.” Because I would quit if they did.

  17. Jim G says:

    Was doing a search for piece rate and this site came up. I am the Production Manager for military supplier. We have pyrotechnic work, CNC machine work, welding, assembly and plating work. I’ve been with the Co. for 18 years, managing for 12. The biggest problem I’ve had is getting new supervisors (that come from other buisnesses) or assistants to ‘buy into’ the rate system.
    We have always used the rate system and as I started out on the assembly floor, I love it. As long as the checks and balances are in place to assure quality, everybody does win. The problems begin when you lose those checks and balances.
    Our Engineering dept sets our preliminary rates and then we adjust up or down as necessary. Our plant wide average is 140% give or take. We adjust rates anytime our average for the given job is either below 90% or above 200%, (we have around 250 different jobs). Generally the open book format for discussing rate changes has worked for me. I talk about the pay structure and how line works aren’t suposed to make more than their bosses.
    We pay our overtime rated work by using 150% of the base wage and then use the ‘rate formula’ as normal.
    I shy away from much of the new era lingo, whether it’s english or not. I find most of the time we over-complicate things. It’s true saying, “just get it done” or “what ever it takes” are the poorest examples of leadership, but I resist change for the sake of change and just trying something because it’s a new fad.
    I disagree with the continuous improvement plans to the degree that does that mean perfection is the goal? As long as we have human workers (forever I hope) perfection can’t be achievied.

    Nice blog to see others working the same things I do!

  18. Laura says:

    I’ve worked as a sewing machine operator at several companies and I loved working piece-rate. I also was chosen to sew samples for an important client and set the rate on their products. In that situation, the machine operator can manipulate the system by setting the piece-rate lower so as to make it easier to overshoot that rate later, but it’s completely unnecessary (and it would eventually be reset to a higher rate). I would set the piece-rate for products by sewing at a comfortably quick pace, not trying to slow down and not trying to sew too quickly. When I started work as a machine operator, I was a rising senior in HS and minimum wage was about $4.25, IIRC. At 17, I was very pleased to be making $6 or more per hour. Much later in life, I was an electronics tech for a large, well-known corporation that makes very expensive music systems. Although the pay was much higher and the benefits were better, I faced much more discrimination (being one of the few women in that job) and worse working conditions than in the sewing jobs I’d had. It was a soulless and boring job.

Leave a Reply

Your email address will not be published.

This site uses Akismet to reduce spam. Learn how your comment data is processed.