I was just going to say something like this. In The Logic of Life, Tim Harford writes about how CEO pay reformers thought that making CEO pay public would cause companies to reduce pay because high pay was so outrageous. The opposite happened.
Why? Because who wants to be in the bottom 50%? Practically no one. If you open up salary information, everyone is going to want to be "above average," even in companies that pride themselves on hiring exceptional people--like Google.
This year, the median is $90K, so everyone will want $95K. Next year, it'll be 95K, so everyone will want $100K, and so on.
Fog Creek has a "ladder" system based on some fairly objective criteria. Based on years of experience and certain assessments of project scope and skill level, each employee gets a number between 8 and 16 that determines pay. This means that compensation is based on performance.
Now, the only point of controversy is whether the employee and employer agree on performance. If there's a discrepancy there, at least there's something meaningful to discuss and, if the difference is irresolvable, then they should separate.
This isn't perfect, but the gossip-driven salary-discovery system is even more imperfect and has the same problems.
In Finland, tax data is public -- you can go look up what your coworkers are making without the consent of either them or their employers. I don't think it has caused a race to the top in any meaningful way.
I think the "race to the top" people are talking about is the psychological tendency of most people to think they are better than average.
So, if you know the average salary is $X and the top salary is $2X, and you're making $1.1X then you ask for a payrise to $1.6X because "you know you're better than most of the people here". Then everyone does. And so on.
I think the way around that is to match transparent salaries with a transparent ladder of responsibilities/expectations/experience, like you were talking about with Fog Creek.
Then there's a clear standard for people to evaluate what they are worth to the company.
I can see the point you're trying to make, but humanistically and morally speaking, this isn't true.
A person with $10 million who puts $3 million into a business isn't taking that much of a risk. There's some risk there and, yes, it should be rewarded. But a person with no net worth is taking a huge risk every time he takes a new job: career risk.
This dynamic exists in startups as well. The people who are taking the real risks are the founders, not the venture capitalists.
This sort of misses the point. Maybe it's true that the employee is taking relatively more risk based on his life situation, but that doesn't matter. What matters is the business needs $3 million. The only way it can get that is by promising a sufficient return to whoever is willing to risk that much. What the amount means to the investor personally is irrelevant. The reality is there are a lot more people willing to risk their time (take a job) than $3 million, therefore the $3 million has a higher price attached. From the perspective of someone investing $3 million, seeking the best possible return the market will provide is no different than a programmer seeking the best possible wage the market will provide.
A race to the highest price the market can sustain. Companies will never pay so much they make a loss on a person, so the marginal revenue from having an additional employee is a firm upper bound on how much they will pay that employee.
This is good from an employee's perspective, but it reduces profits for companies. As such, it is unlikely that many companies would voluntarily support transparency.
It could, but it could also save companies major headaches because people would at least have accurate information instead of gossip. Average salaries would go up a little bit (10-15%?) but there'd be a lot less volatility, and managers wouldn't have to worry about people coming in pissed off that they've been underpaid for three years.
