When flaws in staking systems are pointed out, the most common response is deceptively simple:
“We’ll just increase slashing penalties.”
This is an understandable instinct—and usually the wrong one.
Slashing is a tool, not a foundation. Increasing it without addressing structural issues often produces perverse effects: validator centralization, risk-averse behavior, and an overreliance on trusted infrastructure providers.
The core issue is that slashing only matters while funds are slashable. Once assets become withdrawable, the threat evaporates. You can increase the percentage slashed, extend the unbonding period, or add more offense categories—but if the validator knows that their exposure eventually drops to zero, the system still fails the permanence test.
There’s also a deeper problem: aggressive slashing incentivizes validators to externalize risk.
Large validators can absorb slashes. Small validators cannot. Over time, this selects for professionalized operators, custody services, and opaque delegation structures. Ironically, a mechanism intended to enforce honesty often accelerates centralization.
Worse, slashing is reactive. It assumes misbehavior can be detected, proven, and agreed upon in time. Many of the most dangerous behaviors—soft censorship, delayed finality, subtle reorg assistance—live in gray zones where attribution is murky and enforcement is politically charged.
A system that relies too heavily on slashing is admitting something quietly: it doesn’t trust its own incentive structure.
Slashing should exist, but it should be the last line of defense, not the primary guarantee. If honest behavior only persists because punishment is terrifying, the system is brittle. Remove the enforcers, and everything collapses.
A healthier model treats slashing as friction, not fear. The real work is done by making misbehavior economically irrational before punishment is even considered. That usually means permanence, asymmetry, or commitments that outlive participation—not louder punishments.