The HCE test is complicated, but it basically amounts to ensuring that the disparity between 401k plan usage isn't too great by compensation (so like a Gini coe

The HCE test is complicated, but it basically amounts to ensuring that the disparity between 401k plan usage isn't too great by compensation (so like a Gini coefficient for 401ks). The way Google satisfies the test probably comes down to three factors: 1. All employees are opted into the plan at hire at 10% of salary, possibly even with a default 1% contribution increase per year (I'm unsure). 2. It's actually not exactly 50% match; it's 100% match on the first $3k and then no additional match on the next $3k, or in other words, it's 100% match that smoothly decreases down to 50% match by $6k and then stays at 50% all the way up to the IRS limit. This helps reduce the 401k "Gini coefficient"; it's progressive contributions. 3. A lot of the lowest rungs of jobs are TVCs, not employees, so they don't even count for this test because they're technically employees of another company (which itself would have a much worse 401k plan than Google's, if any). It's not great but it's relevant here :/