The advent of the global economy has brought about unprecedented scale and opportunities for the best to grab a much larger market, resulting in substantial growth in the wealth and power at the top.

This is the story used to explain the sharp rise in pay for CEOs and top artists. Every company wants the best CEO (for the company), and who doesn’t want to listen to the best music (for their tastes)? Yet this story doesn’t appear to be universal. How much more do the top quintile of doctors and engineers earn than the average?

The income difference between the best engineers and the average engineers is closer to 3x than 100x, yet I would much rather work with one excellent engineer than ten mediocre ones. The top engineer builds toward the future, efficiently and with a sense of quality, while the mediocre ones build toward a jumbled product that leaves them reacting to bugs for who knows how long. The top engineer leaves things in a much better state than what he inherits, while the mediocre ones leave a trail of debt. Yet the top engineer doesn’t see proportionate rewards.

What went wrong?

Executives and artists are disciplined and measured by the markets. The CEO is responsible for the P&L; the artist is responsible for her sales. It’s clear when one artist is better than another–just look at their sales.

For the most part, engineers and doctors aren’t subject to the discipline of the market because impact is indirect and hard to measure. The top engineer’s investment in quality (particularly those that don’t have immediate product impact) pay dividends over a long period of time and is difficult to attribute to him–if you’ve ever been thrown into a dungeon of legacy code, you’ll understand the cost of prolonged neglect. In the short run, working twice as hard to produce work that’s higher quality may result in another 5-10%, hardly proportional.

So as a result, you have engineers earning low six-figures (not small, of course) while the executives are taking home hundreds of times as much despite getting fired. In a standard tech company, somewhere between a top engineer and the COO is a discontinuity from the Zone of Diminishing Marginal Returns to the Winner-Takes-All Zone.

Paul Graham argued that measurement and leverage are necessary to enter the latter zone. He follows that with his typical “Join/found a startup” sell.

Yet there’s a tension between measurement and leverage. The easiest path to measurement is to follow Graham’s advice where you are subject to the discipline of the market–your work will determine whether you sink or swim. However, a position in an established company with broad reach will have much bigger leverage: A 5% revenue increase, 5% cost reduction and 5% increase in the productivity of engineers would be tremendously valuable. Unfortunately, few people work so close to the core metrics that those metrics can completely characterize their contributions. Even those who work directly with those metrics won’t see rewards proportionate with their impact, perhaps because the company wants to avoid the perverse incentive of everyone migrating to the more measurable teams.

So how do you position yourself in the latter zone when startups don’t have enough leverage (unless you get lucky), and large companies are unwilling or unable to reward proportionately with impact? Quantitative trading firms are clearly an option–they often try to poach engineers from tech firms by promising multiples the rewards. But if you like the technology sector, is there another way?

There is a currency that you acquire through working with people that’s a lot more valuable in the long term than the pay (Unless you’re Henrique)–the trust and respect of your colleagues. While it’s hard for companies to measure and reward good work, it’s recognized by your peers. This is the fuel behind upward momentum.

A reputable company with the trust of peers of exceptional people is a launchpad, whether to your own company (where you can have immediate leverage because you don’t have to start from step zero) or a more leveraged position somewhere else.

At the end of the day, it’s others’ expectations of you based on your reputation that determine your opportunities.