5 ms·

The Economics of Pay Incentives: a 116% Productivity increase

50 points by smanek 18y ago

ars 18y ago
I'd hate to be a worker there - you go from an easy job to a really hard one - without really getting more money. I wonder if they took that into account. If I knew about this, I'd find a different farm to work on. Also adding teams added a lot of stress to the workers, again that's not someplace I'd like to work. I think they should have rolled back the teams idea and stopped.
- paddy_m 18y ago
  Well if everyone followed your path, they would find themselves making minimum wage. You could chose to work at the minimum wage farm, or the productive farm where you would be making more than minimum wage. You would also find that the minimum wage farm would be likely to go out of business.
  - ars 18y ago
    No, the implication was that the whole point of doing this was to get as much output as possible while paying minimum wage. That was the whole premise - the whole thing about tuning the piece rate - it was to make sure each person earned minimum wage, and no more, but also not less, because otherwise they are paying extra (to make up the difference), without getting more fruit.
    - gojomo 18y ago
      I think you misread the article. The farmowner was happy to pay more than the minimum wage for more-than-minimum productivity.
      - ars 18y ago
        I'm getting down mods, and you're getting ups. But I read it _again_ and I don't see anywhere at all where it says that! Exactly the opposite: "their wages didn't fall below the legal hourly minimum. Farmer Smith tried to adjust the piece rate each day so that it was always adequate but never generous" Where are you seeing that he is paying more? All I see is him paying the same, but getting more output via competition. And again: "(If the managers made a mistake in their estimate, and the pickers didn't earn minimum wage, Farmer Smith would make up the shortfall"
        dkokelley 18y ago
        Exactly the opposite: "their wages didn't fall below the legal hourly minimum. Farmer Smith tried to adjust the piece rate each day so that it was always adequate but never generous" That was before the economists tried the new payment methods. The entire paragraph, for reference: The owner had been paying a piece rate - a rate per kilogram of fruit - but also needed to ensure that whether pickers spent the day on a bountiful field or a sparse one, their wages didn't fall below the legal hourly minimum. Farmer Smith tried to adjust the piece rate each day so that it was always adequate but never generous: The more the work force picked, the lower the piece rate. But his workers were outwitting him by keeping an eye on each other, making sure nobody picked too quickly, and thus collectively slowing down and cranking up the piece rate. To summarize, the farmer was paying a modified market rate for the productivity: using minimum wage as the mean to determine the rate to be paid out. The problem was that the workers were the only supply and they were collaborating to lower the output (even in good fields) so that they would work less for the same pay. A good analogy for this would be OPEC deciding to lower oil production. The oil supply is there, but by exporting less they do less work and increase the market value by limiting supply. After the new methods of payment were introduced, the more productive workers could earn more by being more productive.
        DougBTX 18y ago
        After the new methods of payment were introduced, the more productive workers could earn more by being more productive. The change was to do the calculation of how much to pay per kg before the day's work, rather than after. The most productive were always paid the most, and the least productive were still paid as close to minimum wage as possible. With the rate set at the start of the day, the people working in a particular field had no control over that day's pay per kg. As noted in footnote 5, the workers also had little control over the price for future days, because it was harder for them to correlate their productivity with the pay rate: First, ... it is difficult for workers to disentangle changes in the piece rate due to changing conditions and those due to management learning about workers' true ability ... Second, the effect of a worker's current performance on the unit wage she faces in the future is weak as the unit wage is field-day specific and workers are reallocated to different fields in different days So, instead of using the day's average, the farmer sets the rate based on an historical average, or a direct measurement of how hard a particular field is to work. In either case, it disconnects how hard someone works, and how much the they get paid that day. This makes collusion between workers harder, as they can no-longer game the system by forcing those directly working with them to work less. They would need co-operation between everyone working on the farm - a level of control that only the farm owner has. Looking at Table 1, you'll see that productivity rises from 4.92 kg/hr to 8.06 kg/hr. The difference between the hardest workers and the weakest workers can be approximated using the standard deviation, which went from 3.13 to 4.10. Putting that into a little table: weakest workers hardest workers kg/hr before 1.79 8.05 kg/hr after 4.93 12.16 Now, we know that the farmer will set the pay rate so that the weakest workers receive minimum wage. For simplicity, we can call the minimum wage £5 an hour. In the "before" case, he must pay £5 for 1.79 kg, so the pay rate = 5/1.79 = £2.79/kg. And "after", he must pay only 5/4.93 = £1.01/kg. This gives the following table: weakest workers hardest workers before £5.00/hr £22.45/hr after £5.00/hr £12.28/hr Now you can clearly see ars's worry. The hardest workers are picking 50% more, but are being paid almost 50% less, while the weakest workers pick 175% more, but are paid the same amount. No wonder the farmer was pleased, but it is hard to see how the workers will be happy with the "work harder and then you will get paid more" line.
  - DougBTX 18y ago
    ... or the productive farm where you would be making more than minimum wage. There is nothing to say that you will get paid more working in the "productive" farm, especially when "high productivity" is defined as "you doing more work for less money with the same tools". The farmer has all the data, he sets your pay, and he wants you to work for minimum wage. The trick used to boost the amount of work done was to reduce the influence that the workers had on their pay, not to increase their pay. Doing some quick calculations using the numbers in the paper linked to from the article, (and assuming that the farmer was successful in setting his pay structure so that those working the least were paid a minimum wage of £5/hr), those working the hardest worked harder (from 8.05 kg/hr to 12.16 kg/hr) but got paid less (from £22.45/hr to £12.28/hr). Assuming you don't care whether the farm goes out of business or not (you're only working there for one summer) which one would you like to work in again?
tomjen 18y ago
That story was interesting, but it is properly worth pointing out that the productivity was fruit picking, a line of work were the results are easy to measure and compare, the requirements of skill are modest and the difference between the highest and lowest performer small. Care should therefore be taken before this is applied to other areas where these assumptions no longer holds, such as software development.
- Alex3917 18y ago
  "a line of work were the results are easy to measure and compare, the requirements of skill are modest and the difference between the highest and lowest performer small." A 'routinized job' if you will.
- netcan 18y ago
  True. That is what makes it a great read (I thought it was) & a great experiment. You can easily understand the pieces. But hopefully, some rules & guidelines might emerge from this type of experimentation. These would (if they are good rules & guidelines) be applicable more widely. It'd be difficult to work out the laws of mechanics by examining an engine. But playing with a lead ball in a lab can eventually help you build a better engine.
- vitaminj 18y ago
  Definitely. And it should also be noted that the work in the story is seasonal. I'm not sure that the incessant competition in this performance-pay model would be entirely conducive to retaining staff long-term. However, this may not be a big problem with unskilled industries where staff turnover is high anyway.
  - netcan 18y ago
    Yes. The seasonal thing is central. Important assumptions here are basically: A - Workers have no incentive not to screw the employer. B - Employing/Retaining staff's not an issue. This is also part of the seasonal work. It's not worth looking for a different job when it's only a couple of weeks. If this were a more perfect market (infinite time, perfectly informed participants ...), the decreasing marginal rates (to avoid being over generous), would disappear. You'd be better off pinching your neighbour's picker that picks twice 2x fruit with 2 X minimum wage then keeping you 0.9x picker at 1 X minimum wage.
witten 18y ago
For another take on the subject: http://www.joelonsoftware.com/articles/fog0000000070.html http://www.joelonsoftware.com/articles/fog0000000070.html (Incentive Pay Considered Harmful)
- gaius 18y ago
  Joel would say that, since "incentive pay" comes out of his profit! The prospect of more money wouldn't convince me to do a job I actively disliked but at the end of the day I'm a professional. I do what I do for money. The company isn't shy about boosting its revenue and making profits, why should I be?
  - cdr 18y ago
    Joel would be one of the least concerned about how incentive pay affects his profit - he spends way more than average per developer, but in the areas where he thinks it has a positive effect, like above average salary, good desks/chair, etc. You may dislike Joel, but but that's a stupid argument.
    - gaius 18y ago
      I like Joel, and have subscribed to his blog for years, and what he say often seems to make a lot of sense. However, I know a lot of people in this industry, and I don't know a single one who uses a Fog Creek product. That, to me, speaks volumes.
      - cdr 18y ago
        You don't personally know anyone who uses Fog Creek software, and that means what? Fog Creek is a failure? Fog Creek looks to be doing quite well whether you know anyone who uses their products or not.
        gaius 18y ago
        It means that Joel's writing, while good writing, carries no more weight than any other tech blogger (of which there are many).
aristus 18y ago
Also be careful about side effects. In my town the city tried a neat rule for garbage men: finish your route and you can go home, but we'll pay you the full-time wage. This had two results: 1) garbage routes were completed more quickly 2) on-job injuries (pulled backs and hernias) went way up
- josefresco 18y ago
  Great little story, I wonder if they cut back health benefits for injuries related to "hauling ass" if it would re-balance.
boredguy8 18y ago
That works because the managers can do the job of their employees. Most tech fields to which I've been exposed, managers might have once been able to do the job of employees, but are either out of touch with the technology of today or their skills have ossified. Not to mention the managers that could never do the job in the first place...

50 points by smanek 18y ago

ars 18y ago

I'd hate to be a worker there - you go from an easy job to a really hard one - without really getting more money. I wonder if they took that into account. If I knew about this, I'd find a different farm to work on. Also adding teams added a lot of stress to the workers, again that's not someplace I'd like to work. I think they should have rolled back the teams idea and stopped.

paddy_m 18y ago
Well if everyone followed your path, they would find themselves making minimum wage. You could chose to work at the minimum wage farm, or the productive farm where you would be making more than minimum wage. You would also find that the minimum wage farm would be likely to go out of business.
- ars 18y ago
  No, the implication was that the whole point of doing this was to get as much output as possible while paying minimum wage. That was the whole premise - the whole thing about tuning the piece rate - it was to make sure each person earned minimum wage, and no more, but also not less, because otherwise they are paying extra (to make up the difference), without getting more fruit.
  - gojomo 18y ago
    I think you misread the article. The farmowner was happy to pay more than the minimum wage for more-than-minimum productivity.
    - ars 18y ago
      I'm getting down mods, and you're getting ups. But I read it _again_ and I don't see anywhere at all where it says that! Exactly the opposite: "their wages didn't fall below the legal hourly minimum. Farmer Smith tried to adjust the piece rate each day so that it was always adequate but never generous" Where are you seeing that he is paying more? All I see is him paying the same, but getting more output via competition. And again: "(If the managers made a mistake in their estimate, and the pickers didn't earn minimum wage, Farmer Smith would make up the shortfall"
      - dkokelley 18y ago
        Exactly the opposite: "their wages didn't fall below the legal hourly minimum. Farmer Smith tried to adjust the piece rate each day so that it was always adequate but never generous" That was before the economists tried the new payment methods. The entire paragraph, for reference: The owner had been paying a piece rate - a rate per kilogram of fruit - but also needed to ensure that whether pickers spent the day on a bountiful field or a sparse one, their wages didn't fall below the legal hourly minimum. Farmer Smith tried to adjust the piece rate each day so that it was always adequate but never generous: The more the work force picked, the lower the piece rate. But his workers were outwitting him by keeping an eye on each other, making sure nobody picked too quickly, and thus collectively slowing down and cranking up the piece rate. To summarize, the farmer was paying a modified market rate for the productivity: using minimum wage as the mean to determine the rate to be paid out. The problem was that the workers were the only supply and they were collaborating to lower the output (even in good fields) so that they would work less for the same pay. A good analogy for this would be OPEC deciding to lower oil production. The oil supply is there, but by exporting less they do less work and increase the market value by limiting supply. After the new methods of payment were introduced, the more productive workers could earn more by being more productive.
        DougBTX 18y ago
        After the new methods of payment were introduced, the more productive workers could earn more by being more productive. The change was to do the calculation of how much to pay per kg before the day's work, rather than after. The most productive were always paid the most, and the least productive were still paid as close to minimum wage as possible. With the rate set at the start of the day, the people working in a particular field had no control over that day's pay per kg. As noted in footnote 5, the workers also had little control over the price for future days, because it was harder for them to correlate their productivity with the pay rate: First, ... it is difficult for workers to disentangle changes in the piece rate due to changing conditions and those due to management learning about workers' true ability ... Second, the effect of a worker's current performance on the unit wage she faces in the future is weak as the unit wage is field-day specific and workers are reallocated to different fields in different days So, instead of using the day's average, the farmer sets the rate based on an historical average, or a direct measurement of how hard a particular field is to work. In either case, it disconnects how hard someone works, and how much the they get paid that day. This makes collusion between workers harder, as they can no-longer game the system by forcing those directly working with them to work less. They would need co-operation between everyone working on the farm - a level of control that only the farm owner has. Looking at Table 1, you'll see that productivity rises from 4.92 kg/hr to 8.06 kg/hr. The difference between the hardest workers and the weakest workers can be approximated using the standard deviation, which went from 3.13 to 4.10. Putting that into a little table: weakest workers hardest workers kg/hr before 1.79 8.05 kg/hr after 4.93 12.16 Now, we know that the farmer will set the pay rate so that the weakest workers receive minimum wage. For simplicity, we can call the minimum wage £5 an hour. In the "before" case, he must pay £5 for 1.79 kg, so the pay rate = 5/1.79 = £2.79/kg. And "after", he must pay only 5/4.93 = £1.01/kg. This gives the following table: weakest workers hardest workers before £5.00/hr £22.45/hr after £5.00/hr £12.28/hr Now you can clearly see ars's worry. The hardest workers are picking 50% more, but are being paid almost 50% less, while the weakest workers pick 175% more, but are paid the same amount. No wonder the farmer was pleased, but it is hard to see how the workers will be happy with the "work harder and then you will get paid more" line.
- DougBTX 18y ago
  ... or the productive farm where you would be making more than minimum wage. There is nothing to say that you will get paid more working in the "productive" farm, especially when "high productivity" is defined as "you doing more work for less money with the same tools". The farmer has all the data, he sets your pay, and he wants you to work for minimum wage. The trick used to boost the amount of work done was to reduce the influence that the workers had on their pay, not to increase their pay. Doing some quick calculations using the numbers in the paper linked to from the article, (and assuming that the farmer was successful in setting his pay structure so that those working the least were paid a minimum wage of £5/hr), those working the hardest worked harder (from 8.05 kg/hr to 12.16 kg/hr) but got paid less (from £22.45/hr to £12.28/hr). Assuming you don't care whether the farm goes out of business or not (you're only working there for one summer) which one would you like to work in again?

tomjen 18y ago

That story was interesting, but it is properly worth pointing out that the productivity was fruit picking, a line of work were the results are easy to measure and compare, the requirements of skill are modest and the difference between the highest and lowest performer small. Care should therefore be taken before this is applied to other areas where these assumptions no longer holds, such as software development.

Alex3917 18y ago
"a line of work were the results are easy to measure and compare, the requirements of skill are modest and the difference between the highest and lowest performer small." A 'routinized job' if you will.
netcan 18y ago
True. That is what makes it a great read (I thought it was) & a great experiment. You can easily understand the pieces. But hopefully, some rules & guidelines might emerge from this type of experimentation. These would (if they are good rules & guidelines) be applicable more widely. It'd be difficult to work out the laws of mechanics by examining an engine. But playing with a lead ball in a lab can eventually help you build a better engine.
vitaminj 18y ago
Definitely. And it should also be noted that the work in the story is seasonal. I'm not sure that the incessant competition in this performance-pay model would be entirely conducive to retaining staff long-term. However, this may not be a big problem with unskilled industries where staff turnover is high anyway.
- netcan 18y ago
  Yes. The seasonal thing is central. Important assumptions here are basically: A - Workers have no incentive not to screw the employer. B - Employing/Retaining staff's not an issue. This is also part of the seasonal work. It's not worth looking for a different job when it's only a couple of weeks. If this were a more perfect market (infinite time, perfectly informed participants ...), the decreasing marginal rates (to avoid being over generous), would disappear. You'd be better off pinching your neighbour's picker that picks twice 2x fruit with 2 X minimum wage then keeping you 0.9x picker at 1 X minimum wage.

witten 18y ago

For another take on the subject: http://www.joelonsoftware.com/articles/fog0000000070.html http://www.joelonsoftware.com/articles/fog0000000070.html (Incentive Pay Considered Harmful)

gaius 18y ago
Joel would say that, since "incentive pay" comes out of his profit! The prospect of more money wouldn't convince me to do a job I actively disliked but at the end of the day I'm a professional. I do what I do for money. The company isn't shy about boosting its revenue and making profits, why should I be?
- cdr 18y ago
  Joel would be one of the least concerned about how incentive pay affects his profit - he spends way more than average per developer, but in the areas where he thinks it has a positive effect, like above average salary, good desks/chair, etc. You may dislike Joel, but but that's a stupid argument.
  - gaius 18y ago
    I like Joel, and have subscribed to his blog for years, and what he say often seems to make a lot of sense. However, I know a lot of people in this industry, and I don't know a single one who uses a Fog Creek product. That, to me, speaks volumes.
    - cdr 18y ago
      You don't personally know anyone who uses Fog Creek software, and that means what? Fog Creek is a failure? Fog Creek looks to be doing quite well whether you know anyone who uses their products or not.
      - gaius 18y ago
        It means that Joel's writing, while good writing, carries no more weight than any other tech blogger (of which there are many).

aristus 18y ago

Also be careful about side effects. In my town the city tried a neat rule for garbage men: finish your route and you can go home, but we'll pay you the full-time wage. This had two results: 1) garbage routes were completed more quickly 2) on-job injuries (pulled backs and hernias) went way up

josefresco 18y ago
Great little story, I wonder if they cut back health benefits for injuries related to "hauling ass" if it would re-balance.

boredguy8 18y ago

That works because the managers can do the job of their employees. Most tech fields to which I've been exposed, managers might have once been able to do the job of employees, but are either out of touch with the technology of today or their skills have ossified. Not to mention the managers that could never do the job in the first place...