That's why companies give you stock.

I'd say this is why companies give bonuses. Stock is fungible for cash unless the company is private, in which cases a stock bonus can be meaningfully different from a cash bonus.

I (mostly) disagree. If I receive an RSU grant that vests over 4 years, it's not just because the company hopes I'll stay that long but primarily because the company hopes that will align my incentives with the shareholders' incentives. If I work to make the company better over a multi-year period, we all win together. If they instead just give me the equivalent value in cash salary, my incentive is to extend my term of employment, not the value of the company.

That works for small companies. In my large group, whether I do a stellar job or slack on HN all day won't influence the stock price by a cent.

Do you have an idea of why the shareholders permit you to stay employed then?

I believe that they, via management, believe that the aggregate of what you and your colleagues do is contributing to the overall enterprise value, even if your individual contribution is much smaller than a penny per share.

He specifically said what he does won’t change the stock price by a cent, so he is presumably talking about his individual contribution, not the aggregate one. Which is his point —- giving him RSU’s won’t allow him to capture the value of going above and beyond, because even as you acknowledge yourself, his going above and beyond won’t move the share price by even a penny.

If every software engineer can lift the share price by a hundredth of a penny every year, should the company give shares as part of comp to their 5,000 engineers or not? Should they even employ engineers at all?

The fact that some engineers can shirk and ride on the efforts of their peers does create a tragedy of the commons possibility, but the most appropriate solution there involves strong local leadership, rather than the conclusion that RSU grants are ineffective.

The point is that they are ineffective for the stated purpose of incentivizing the engineers to work harder. To that end, bonuses are much more effective.

The difference is that anyone can buy stock in their publicly held employer (with a bonus or otherwise) and have the same incentive. RSUs force the incentive to exist (instead of directly rewarding performance), but the balance between RSU growth as reward and ability to influence RSU growth may be negligible except at VP levels and above. If I am reasonably certain that keeping a backend service running within SLA is unlikely to meaningfully move the stock price I certainly won't be putting in extra hours or weekends to achieve it. Additionally, it forces a minimum risk on the risk-reward calculation made by an employee.

Whereas banks and sales teams have long given bonuses because it's tractable to assign value to individual contributions, I think bonuses are the generally preferred incentive when tractable. Making partner is roughly the VP-level equivalent of RSU grants where company performance starts being tied to individual performance.

Where this becomes a bit more evident is that RSU refreshes are almost always based on performance review, like bonuses.

I see RSUs as a cheaper/more flexible compensation for employers and a tax reduction strategy for employees (both worthwhile in their own right).