The Quiet Incentive Problem Baked Into Bitcoin's Design

You're a miner. A block just confirmed with 3.2 BTC in transaction fees sitting inside it, and you watched someone else claim it. You could build on top of that block, extend the honest chain, collect whatever fees trickle into the next one. Or you could stop, run the arithmetic, and start mining that same block height again from scratch, trying to take those fees for yourself.

That second choice is fee sniping. Right now it's almost always irrational. But the word "almost" is doing a lot of work in that sentence, and the calculus shifts dramatically as the block subsidy continues its scheduled decline toward zero.

What Fee Sniping Actually Is

Fee sniping is when a miner, instead of extending the longest chain, attempts to re-mine a recently confirmed block to claim its transaction fees. The attacker broadcasts an alternative block at the same height, hoping to build a longer chain from there and orphan the original.

The mechanism is straightforward. Say block 850,000 just landed with 2.8 BTC in fees alone. A miner with 15% of total network hashrate could spend their next several blocks' worth of work trying to re-mine block 850,000 rather than building block 850,001. Succeed, then extend that fork by one more block, and they've replaced the original miner's block and pocketed those 2.8 BTC.

The reason this rarely happens today isn't moral restraint.

It's math.

The Subsidy Acts as a Disincentive

Every block a miner produces earns two things: the block subsidy (newly minted bitcoin, fixed per epoch) and the transaction fees from included transactions. Currently the subsidy dwarfs typical fees, and that gap is doing most of the security work in this system, like water pressure keeping a pipe from running backward.

Walk through the arithmetic. Suppose the block subsidy is 3.125 BTC and fees in a target block are 0.3 BTC. A miner with 15% hashrate who spends one block's worth of effort on a sniping attempt has an expected payout of roughly 0.15 times 0.3 BTC, about 0.045 BTC. Honest mining during the same window yields an expected 0.15 times 3.425 BTC (subsidy plus fees), around 0.51 BTC. The honest strategy pays more than ten times as much. Sniping is a bad trade.

Now run the same numbers in a post-subsidy world. Subsidy is zero. That same block carries 3.2 BTC in fees (higher fees, because that's the only miner revenue remaining). The honest expected value from building the next block is 0.15 times whatever fees accumulate there. The sniping expected value is 0.15 times 3.2 BTC, or 0.48 BTC, plus the chance to collect fees from the block you build on top of your stolen one. Suddenly the incentive flips. Sniping a fee-rich block becomes the rational move, at least probabilistically.

This is the proportionality people mean when they say the risk scales with subsidy decline. Not a sudden cliff. A gradient that tilts a little more with every halving.

The Game Theory Gets Uglier With Coordination

One miner sniping is a nuisance. A coalition is a structural threat.

Imagine three mining pools controlling 40% of hashrate quietly agreeing to target any block with fees above a certain threshold. The expected value calculation above gets considerably more favorable at 40% than at 15%. Worse, if two of them succeed in re-mining a block, the third can immediately build on their fork, extending it faster than the honest chain and making the orphan permanent.

This kind of coordinated sniping doesn't require a 51% attack on the whole network. It just requires that the fee-sniping fork grows faster than the honest chain for a few blocks, which becomes increasingly plausible as the attacking coalition grows.

Bitcoin Core developers acknowledged this dynamic explicitly when they introduced the nLockTime-based fee sniping deterrent. Wallets implementing BIP 115 or the nLockTime convention set their transactions' locktime to the current block height, making those transactions invalid if included in a re-mined earlier block. It's not a complete fix. It's a friction mechanism that raises the cost of a sniping attempt by reducing the pool of transactions an attacker could stuff into their forged block, and it relies on widespread wallet adoption that has, in practice, remained incomplete, with uptake uneven across major wallet implementations years after the convention was introduced.

The Deeper Tension: Security Budget

Fee sniping is really a symptom of the broader security budget question, which is among the most genuinely unresolved issues in Bitcoin's long-term design. I'll say that plainly: unresolved, not "actively being addressed" or "a topic of ongoing discussion." The community has been circling it for well over a decade without landing anywhere.

Bitcoin's security rests on the assumption that honest mining is more profitable than dishonest mining. The subsidy enforces that assumption mechanically right now. As the subsidy declines across successive halvings (every 210,000 blocks), transaction fees must grow to compensate, or the incentive structure gradually weakens.

Fee sniping is one specific failure mode that emerges from that weakening. But notice it doesn't require fees to be low. It requires fees to be lumpy. A block with 10 BTC in fees sitting next to a block with 0.2 BTC in fees is a sniping magnet, even in a high-fee environment. Uniform fee pressure across blocks would reduce the temptation; volatile, spiky fee markets increase it.

Layer 2 systems like the Lightning Network complicate this further. If a large portion of transactions settle off-chain, on-chain fee revenue per block could remain structurally low even as Bitcoin's usage grows. That's a scenario where the security budget concern becomes acute, not because Bitcoin failed, but because it succeeded at scaling. The irony is not lost on anyone paying attention.

Researchers have proposed various mitigations beyond the nLockTime approach: demurrage-style mechanisms, fee smoothing across blocks, tail emission (a permanent small subsidy), changes to how fees are distributed. None have achieved anything close to consensus. The community remains genuinely divided on whether the problem is serious, addressable, or both.

What Honest Timelines Look Like

Vague warnings about "the future" don't help anyone think clearly, so here's the actual pace.

The block subsidy halves every 210,000 blocks, roughly every four years at a ten-minute average block time. The current subsidy is 3.125 BTC. By the time it reaches approximately 0.1 BTC, fees would need to constitute more than 96% of miner revenue for the security budget to match today's in bitcoin terms. Whether on-chain fee markets will deliver that is genuinely unknown, and anyone who tells you otherwise is selling something.

The full schedule runs out around the year 2140, when the last fraction of a bitcoin is mined. The sniping risk doesn't wait until 2140, though. It intensifies incrementally with each halving. At 50% subsidy contribution, sniping a fee-rich block already looks tempting for large miners with enough hashrate to make the gamble worthwhile.

Are you skeptical that fees alone can sustain security at today's levels? You should be. Hal Finney raised this concern in 2010. It remains open.

Friction, Not a Fix

The tools currently deployed against fee sniping are deterrents, not solutions. The nLockTime convention raises the cost of a successful re-mine. Mining pool reputation effects discourage overt collusion. The sheer current size of the honest chain makes short-range forks expensive.

Deterrents work until they don't. When the expected value of sniping a block exceeds the expected value of honest mining, and when a miner has enough hashrate to make the probability favorable, the economic logic points in one direction regardless of norms.

The subsidy is the thing keeping that calculation honest right now. It's scheduled to disappear. Build your mental model accordingly.