Yay system designing! The sentiment is good, correct incentives will change people's actions so they can maximise their return.
Indeed, this is why senior management is sometimes paid in shares of the company and by get bonuses according to the share price: What people want is the business to make profits. Senior management is making the big decisions which will impact the business's profits. If profits are good, then the company's shares go up and senior management get money. Incentives all align and it's great.
This also happens at the lowest level, where it's easy to track people's performance and easy to hold them accountable. In electronics stores people are paid by commission: What the business wants is more sales. Salespeople can generally try harder to secure more sales. If they sell more, they get more money. Yay incentives.
Issues can happen where you use the wrong metric, where you want people to innovate, where it can be costly or not-time-appropriate to measure or where people don't have the power. Considering each individually:
If you use the wrong metric, then you're creating this perverse incentive for people to do the wrong thing.
What if you're a banker and you need to approve home loans according to whether the person will be able to pay: What is the appropriate metric? You could say number of loans approved, but that creates a perverse incentive for people to approve all the dodgy loans.
The reason there is such a big backlash against using standardised testing in education is because it means teachers will:
- be incentivised to cheat for their students;
- be incentivised to only want to take the best students and not teach "unprofitable" classes with worse students; and
- be incentivised to not focus on a child's development, their social skills, their psychological well-being, etc. because it's not measured.
There's a similar issue in the healthcare industry: Tying doctor's pay to hospital profits may incentivise them to only accept the wealthy patients who will pay large bills.
If you want people to innovate, then it's hard to demand because innovation comes from lots of places. If people are being told to focus on a specific task because that's how they get paid, then the rational thing is for that person to purely focus on that and not try to improve things or solve bigger problems. In some cases, having a metric of "how many new ideas" might work, but that doesn't ensure quality. In some cases, having a metric of "how much money did your profits gain" might work, but often it runs into our next issue. If what was needed to be done is obvious that it can be metric-afied, it probably doesn't cover innovation or problem-solving.
If it is costly or not-time-appropriate to measure.
Firstly, if it's not time appropriate: as I was saying about innovation, the pipeline for innovation can take ages because of the trial and error that comes along with it. Also applicable if you were, say, the manager who decides where and when to build new factories. It is hard to create a metric around this because it would imply needing to hold off on paying the manager until the factory has been built, which can take ages.
Secondly. it can also be costly to measure. This is why the NFP is sometimes slammed for being inefficient. It is costly to measure the impact someone working for a non-for-profit has had on a community. If you were leading a project to bring education to a community, the pay-off not come for a long time (not-time-appropriate) it would also be hard to measure. So far many NFPs use process-measures (e.g. how many schools did we put up), but that creates a perverse incentive where NFPs put up a lot of schools... but that might not be what the community needs.
Finally, the big one is that it's often not equitable. Often people don't actually have much power to influence things, and sometimes that can be a huge issue. My friends who work in retail have quit when paid by commission because the sheer number of customers coming into the store is too low for it to be profitable. Bankers can't decide how many credit-worthy borrowers come to see them. Teachers and doctors aren't miracle workers, so if someone is deathly stupid or sick there's only so much they can do. Innovators can spend a lot of time working on something, and then it just doesn't work out. It's often unfair to tie people's pay to outcomes, because the outcomes aren't always a direct function of the worker's inputs.
Two more thoughts:
1. People also lie. A great example where people were held accountable (and fired if they didn't make standards) was at Enron. Everyone just lied about their results so they would get more pay, and in the end management was being fed false information.
2. System's not perfect but an entire area of HR is dedicated to correct incentives so I'm sure there's something behind it. Have you read about
Google's rule to Pay Unfairly and they've also written a new book about it (and other stuff)
their HR book?