[ThrustVectoring]

It's not means testing that's the problem. The problem is that a linear combination of benefit programs with a below-unity benefit-to-income slope has an above-unity slope, which means that making an additional dollar disqualifies you from more money than the dollar gives you. The benefit programs as a whole aren't designed holistically to avoid this (and other) traps, so any welfare overhaul that consolidates multiple independent programs into a systematic whole should solve these problems.

[dllthomas]

And even before unity-in-dollars, it can be way below unity-in-utility for opportunities available.