Full Text
THE LAST ACCOUNTANT AND THE END OF MANUAL TAX COMPLIANCE
Small enterprises employ labor whose only function is preparing tax records for the government; compliance costs approach two percent of gross domestic product and fall hardest on the smallest firms. Our paper asks whether a state can collect tax from small enterprises with zero manual accounting work, answering through design analysis of payment-native taxation, where assessment, collection and record-keeping complete at settlement. Electronic-invoicing mandates in Italy and Peru digitized documents yet left small firms dependent on accountants, while audit evidence finds evasion near zero wherever third parties report transactions. Our central proposal is a statutory exchange: any small enterprise routing every payment through an optional central-bank payment instrument receives exemption from bookkeeping, filing and invoicing duties, with invoices generated automatically, tax collected at settlement under any rate regime, and cash staying lawful. Current tax-compliance accountants hold a voluntary employment guarantee to retirement with advisory redeployment, and citizen trust rests on Fiscal Secularity, the constitutional separation of money, taxation and governance: deductions stay visible at a capped rate, and sealed records open only through a judicial gate whose accesses are logged for the citizen concerned. Our paper specifies a piloted evaluation with parameters named in advance, adoption, compliance hours against standard-cost baselines, firm entry, access-log complaints, anticipating near-complete compliance-hour elimination for adopters at near-zero transition cost as most certified accountants approach retirement. A design recommendation follows: automating state-facing paperwork frees labor without destroying consumer demand, and relief aimed at the smallest firms removes a penalty on growth rather than protecting smallness.
INTRODUCTION: THE PROBLEM OF COMPLIANCE LABOR
Every economy contains a class of labor that produces nothing a customer can buy. A worker in a bakery who bakes creates goods; a worker in the same bakery who prepares tax records creates, for the benefit of the state, a copy of information the bakery already generated by transacting. The classical tradition condemned the excess directly: the fourth maxim of taxation in Smith (1776) demands that every tax take and keep out of the pockets of the people as little as possible beyond what enters the public treasury. The modern literature gives the same idea an empirical body. Americans will spend more than 7.9 billion hours complying with tax filing and reporting requirements in 2024, equal to 3.8 million full-time workers doing nothing but tax return paperwork, and total compliance costs reach approximately 546 billion dollars, or nearly two percent of gross domestic product (Tax Foundation, 2024, computed from Internal Revenue Service estimates). The compliance literature adds a distributional verdict: tax compliance costs are regressive in nature, imposing a heavier burden on small businesses compared to larger ones, and small businesses function, in the standard phrase of the field, as "unpaid tax collectors" (Pope and Rametse, 2001).
Our paper asks one question and answers one question: can a state collect tax from small enterprises while requiring zero manual accounting work from those enterprises? Our answer proceeds by design analysis rather than by advocacy of any particular tax rate or tax base. Our paper defines payment-native taxation, distinguishes payment-native taxation from the document digitization currently marketed as automation, grounds the enforcement mechanism in the strongest experimental evidence available in public economics, proposes a five-provision legislative package centered on a voluntary exchange of payment routing for compliance exemption, designs the transition of the existing accounting workforce as a first-class component, and specifies a piloted evaluation with parameters named in advance. Foundational to the whole argument stands one axiom our research program applies across many papers: an economy contains a productive layer, where goods and services people need come into existence, and an instrumental layer of money, records and administration that exists only to serve the productive layer; good design keeps the instrumental layer thin, cheap and unable to mistake itself for the point.
DEFINITIONS, THE AUTOMATION LADDER, AND PRIOR ART
Payment-Native Taxation Defined. Our paper uses the term automated taxation strictly. Automated taxation means that the assessment of tax, the collection of tax, and the record-keeping surrounding tax are all completed by the payment infrastructure itself at the moment a payment settles, with no separate human bookkeeping anywhere in the chain. Any arrangement demanding less should be called digitization of paperwork, because such an arrangement digitizes an artifact while preserving the human workflow around the artifact.
The Ladder. Existing regimes occupy four rungs. Rung zero holds manual paper filing. Rung one holds digitized documents: clearance-model electronic invoicing, in which the state receives and validates every invoice in real time. Rung two holds state-computed assessment: pre-filled returns built from third-party data, which the taxpayer merely confirms. Rung three holds payment-native taxation, in which the payment event is the tax event and no filing exists at all. Every celebrated digital tax success of the past decade sits on rungs one and two. The contribution of our paper consists in naming the ladder, demonstrating with evidence that rungs one and two do not remove the accountant, and specifying rung three as an implementable, voluntary design for small enterprises.
The Italian Lesson. Italy provides the best-in-class case precisely because Italy leads the field. Italy is widely regarded as a pioneer, and since January 2019 all domestic business-to-business and business-to-consumer invoices must be issued electronically and transmitted through the Sistema di Interscambio, the central exchange platform of the tax authority, and by January 2024 all remaining exemptions were removed (European Commission, 2025). The platform validates format, mandatory fields and identification numbers; a human being, or software a human must configure and pay for, still composes every invoice. The burden consequently relocated instead of disappearing: the initial burden was significant, particularly for micro-enterprises adapting to XML-based invoicing, and implementation into existing accounting and enterprise systems proved very costly, specifically for smaller entities, sparking a new market for e-invoicing consolidators (Grant Thornton, 2024). The decisive fact concerns the accountant: after universal adoption, survey evidence shows that 26 percent of Italian founders fully delegate accounting activities to their accountant, and 69 percent rely entirely on their accountants for value-added-tax and social-contribution deadlines, while 52 percent of companies still manage most financial processes manually or with disconnected tools (wamo and FidoCommercialista, 2026). The accountant survives every digitization of the document because the tax code's complexity, not the paper, created the accountant. Honesty requires the counterweight: the Italian system matured, now supplies pre-filled value-added-tax returns, and demonstrably serves the state, since the value-added-tax compliance gap fell in Italy by 10.7 percent, 12.7 billion euro, in 2021 compared to 2020, the largest drop among the EU-27 (European Commission VAT gap reporting). Italy therefore proves two claims at once: national real-time transaction infrastructure is feasible, and collecting documents about payments, rather than collecting tax inside payments, leaves the accountant standing.
The Peruvian Evidence. Causal evidence confirms the Italian reading. Using quasi-experimental variation in the roll-out of value-added-tax electronic invoicing in Peru, adoption increased reported firm sales, purchases and tax liabilities by over five percent in the first year, concentrated among small firms and less compliant sectors (Bellon, Dabla-Norris, Khalid and Lima, 2022). Two further findings serve our thesis: some firms exited when the requirements were announced, revealing real adoption costs, and existing stocks of tax credits offset the reform's revenue gains, so digital invoicing requires complementary reforms. Document digitization helps the state, imposes costs on firms, and cannot deliver the full dividend, because the credit-and-refund architecture claws gains back.
Prior Art. Payment-native taxation carries prior art in four families. First, the settlement-collected transaction levy: in simplest form, the Automated Payment Transaction tax consists of a flat tax levied on all transactions, automatically assessed and collected when transactions settle through the electronic technology of the banking and payments system (Feige, 2000), a proposal serious enough to have been presented to the United States President's Advisory Panel on Federal Tax Reform in 2005, with the elimination of filing stated as an explicit property, since automated recording of payments eliminates the need to file tax and information returns. Second, withholding at source: the Organisation for Economic Co-operation and Development names the automation of taxation processes within payroll software in pay-as-you-earn systems as a current and widespread example of taxation performed inside taxpayers' natural systems. Third, state-computed assessment: Estonian taxpayers review pre-filled forms and approve with a digital signature in three to five minutes, with one-click returns available in under a minute (European Commission, 2019). Fourth, the official strategic vision: the same organisation sets out a vision under which taxation becomes a more seamless and frictionless process over time, with the payoff stated plainly, because moving taxation into taxpayers' natural systems aligns tax with taxable events, producing fewer errors and substantially reduced burdens (OECD, 2020a). Our proposal is the terminal station of a line already drawn by the field's most conservative institution.
The Enforcement Mechanism. The mechanism behind the entire design carries the strongest causal evidence in tax economics, developed as an extension of the canonical evasion model of Allingham and Sandmo (1972), since the empirical results are explained by extending the standard model of rational tax evasion to distinguish self-reported from third-party reported income. In a randomized audit experiment covering over 40,000 Danish taxpayers, the evasion rate was 0.3 percent for income subject to third-party reporting and 37 percent for self-reported income, and since 95 percent of income was third-party reported, overall evasion stayed very modest; furthermore, audits and audit threats affected only self-reported income, with no effect on third-party reported income (Kleven, Knudsen, Kreiner, Pedersen and Saez, 2011). Complementary field experiments in Chile show that announcing additional monitoring has less impact on transactions already covered by a paper trail, indicating the trail's preventive deterrence effect, while information without credible deterrence does not self-enforce (Pomeranz, 2015). Payment-native taxation is the limiting case of the Danish result: when every taxable event is a third-party-verified settlement, the 0.3 percent regime extends to the whole participating economy, and the Chilean boundary, namely a trail that historically stopped at the final consumer, closes, because a payment rail records the retail transaction as completely as the wholesale transaction.
WHY THE STATE SHOULD FUND THE AUTOMATION
The Category Error and the Cost-Shifting Asymmetry. Compliance labor belongs to the instrumental layer, and the instrumental layer has quietly transferred most of the tax system's true operating cost onto private firms (for the survey of compliance and enforcement economics, see Slemrod, 2019). Private compliance cost for the United States income tax runs near ten cents per dollar collected, while the tax authority's own administrative cost runs near 0.6 cents per dollar, so private compliance costs dwarf public administrative costs in almost all cases (Slemrod, 1996; Sandford, 1995). Roughly ninety-four percent of the collection machinery's cost sits off the state's books, carried by firms and, through firms, by the smallest firms most heavily. The state authored the obligation; the state is the cheapest cost avoider; the state should fund the redesign.
Regressivity as Structure. Compliance cost behaves as a fixed head tax per firm. Econometric work finds compliance costs growing only 0.34 to 0.42 percent for each one percent of firm size, with a high constant indicating fixed cost elements (Eichfelder and Schorn, 2012), producing measured burden ratios across firm sizes of twenty-one to one, and up to one hundred and two to one under variable monetization (Internal Revenue Service Research, 2012), while in lower-income countries some small businesses incur compliance costs exceeding twelve percent of turnover (International Centre for Tax and Development, 2023). A four-country comparative study concluded that compliance costs remain high and regressive and do not appear to diminish over time, while the managerial benefits of compliance bookkeeping exist yet few small businesses can value them (Evans, Hansford, Hasseldine, Lignier, Smulders and Vaillancourt, 2014); our design absorbs the managerial-benefits counterpoint, because the transaction stream preserves the information while deleting the labor. The growth mechanism appears in the same literature: high compliance costs divert resources from productive activities and increase input costs without creating additional output (Dabla-Norris, Misch, Cleary and Khwaja, 2017).
Firms That Refuse to Grow. Burdens with thresholds teach firms to stay small. United Kingdom administrative records document bunching in annual turnover just below the value-added-tax registration threshold (Liu and Lockwood, 2021), and follow-up work finds annual turnover growth slowing by about one percentage point as firms approach the threshold, with no offsetting growth after crossing (Liu, Lockwood and Tam, 2024); the mechanism is explicit, since firms keep reported turnover below the threshold either legally, by restricting the scale of operations, or illegally, by misreporting sales. The same pathology appears at the French fifty-employee wall: size-contingent regulations cost the equivalent of a 2.3 percent variable labor tax, with welfare losses up to 3.4 percent of gross domestic product under wage inflexibility (Garicano, Lelarge and Van Reenen, 2016), and twelve percent of firms stayed at exactly forty-nine employees for two consecutive years, against two percent at fifty-two. Threshold burdens even corrupt the statistical record, since a large share of French employers manipulate declared workforce size, and the manipulation itself shapes the observed size distribution (Askenazy, Breda, Moreau and Pecheu, 2022). Our design answers with smooth tapering rather than cliffs, as the proposal section specifies.
The Return on Public Investment, and Regime Independence. Public digitalization pays for the state directly: the use of digital signatures in Estonia is estimated to save two percent of gross domestic product every year (OECD, 2020b, reporting a government estimate), and in Estonia total business tax compliance takes roughly fifty hours a year (World Bank Paying Taxes research). The growth channel is quantified across 118 economies: firm entry responds to the ease of paying taxes regardless of the corporate tax rate, and a ten percent reduction in tax administrative burden associates with a three percent increase in annual business entry (Braunerhjelm and Eklund, 2014, as reported in the World Bank Paying Taxes research), while complexity associates with evasion, and high compliance costs associate with larger informal sectors and less investment. The final clause carries the design consequence our paper builds on: the burden lives in process, meaning filings, payments and documentation frequency, not in rates. China cut compliance from 832 hours and thirty-seven payments per year to 138 hours and seven payments without abolishing taxation. The automation investment therefore pays under a value-added tax, under a turnover levy, under any regime a legislature chooses, and our paper accordingly presupposes no particular tax.
FROM FREED LABOR TO NATIONAL CAPABILITY
Misallocation, and a Reallocation with No Demand Cost. Growth economics names compliance labor precisely: misallocated resources. When capital and labor are hypothetically reallocated to equalize marginal products to the degree observed in the United States, manufacturing total-factor-productivity gains reach 30 to 50 percent in China and 40 to 60 percent in India (Hsieh and Klenow, 2009). Labor whose entire function is reproducing records for the state holds a marginal product of production near zero, so reallocation of such labor is a first-order productivity gain. One asymmetry makes the gain unusually clean, and one sentence of our research program states the asymmetry: automate the state ruthlessly, automate the market carefully. Automating a factory risks the demand spiral of fewer workers and fewer consumers; automating state-facing paperwork destroys no consumer demand, because the paperwork was never a product anyone consumed. Freed compliance labor is the purest reallocatable resource an economy holds.
Security Capability and Cultural Influence. A larger productive base underwrites national capability in two documented directions, and both are held here as instruments for citizens, never as ends. On security: economic strength and military power have been highly correlated in the rise and fall of major nations since 1500 (Kennedy, 1987), summarized in Kennedy's own words: "wealth is usually needed to underpin military power, and military power is usually needed to acquire and protect wealth". The policy relevance runs through the constant triple tension between investment, defense and consumption: recovering one to two percent of gross domestic product of pure waste relaxes the triple tension without touching either investment or consumption. On culture: economic resources feed attraction as well as coercion, since economic resources can produce soft as well as hard power (Nye, 1990; 2011), and South Korea proves the conversion is designable, because the 1998 Hallyu Industry Support Development Plan raised cultural spending from 14 million dollars to 84 million dollars by 2001 (Sellars and Leasure, 2025), the 1999 framework statute for promoting cultural industries institutionalized support (Republic of Korea, 1999), and by 2021 the content industry generated 137.5 trillion won in sales and 12.45 billion dollars in exports, with measured spillovers of 180 million dollars of consumer-goods exports and nearly three thousand jobs per additional 100 million dollars of content exports (Invest Korea, 2023, reporting Korea Creative Content Agency data). Cultural influence is downstream of fiscal surplus plus deliberate design; compliance automation supplies the surplus.
Why Replacing the Value-Added Tax Helps, Stated Within Strict Limits. Our paper does not claim that the value-added tax burdens production in any special economic sense; the critique runs through administration and visibility only, and rests on three charges. Charge one, compliance intensity: complying with value-added tax takes a case-study company longer on average than complying with corporate income tax, because the credit-invoice mechanism runs on matched document chains, which is exactly why the accountant survives digitization. Charge two, fraud architecture: carousel fraud exploits the fact that the tax is not immediately applied to cross-border business transactions, letting fraudsters buy free of tax, sell tax-inclusive, and vanish; the magnitudes are large, since around 90 billion euro of value-added-tax revenue was lost in the European Union in 2022, of which at least 13 billion euro links conservatively to carousel fraud (European Commission, 2024; Poniatowski, Smietanka and Skowronek, 2024), with carousel fraud making up roughly one quarter of the total gap, and the enforcement contest is permanent, even though national digital reporting requirements did add an estimated 19 to 28 billion euro of revenue between 2014 and 2019 (European Parliamentary Research Service, 2025). A settlement-collected levy holds no credit chain, no refund claim, and no missing-trader position to construct. Charge three, salience: consumers underreact to taxes that are not salient, demonstrated by an eight percent demand fall when tax-inclusive prices were posted (Chetty, Looney and Kroft, 2009); a tax living permanently inside posted prices is the maximally non-salient design, and a democracy cannot audit what citizens cannot perceive. The named theoretical objection to any transaction levy must appear here, and our paper cites the objection first: the Diamond-Mirrlees production-efficiency theorem dictates that taxes should not distort production, implying no taxation of intermediate inputs or turnover (Diamond and Mirrlees, 1971). The modern answer is empirical: the theorem relies on perfect enforcement, meaning zero evasion at zero administrative cost, and where enforcement is imperfect, turnover taxes reduced evasion by up to 60 to 70 percent of corporate income in Pakistani administrative data, and switching from profit to turnover taxation raised revenue by 74 percent without reducing aggregate profits (Best, Brockmeyer, Kleven, Spinnewijn and Waseem, 2015), with the distortion bounded because the turnover rate was 0.5 percent against a 35 percent profit rate. Cascading through supply chains remains a real cost, favors vertical integration, and must stay openly acknowledged; low capped rates, the small-producer exemption of the proposal, and compliance savings that dominate for small firms are the counterweights, and the rate itself remains an open parameter of the design rather than a claim of our paper.
THE ACCOUNTANT TRANSITION: GRANDFATHER THE PERSON, SUNSET THE FUNCTION
The Political Economy of Concentrated Losers. Reforms with diffuse gains and concentrated losses fail routinely, because organized losers mobilize while diffuse winners sleep. The compensation principle says winners could compensate losers; our design actually does so, converting the accounting profession from veto player into implementation partner: the people who know where every buried complexity lies become the people paid to help retire the system that buried the complexities.
The Cohort, the Exposure, and the Honest Reading of Automation Forecasts. Accountants and auditors held about 1.6 million jobs in the United States in 2024 at a median wage of 81,680 dollars (Bureau of Labor Statistics, 2025). The canonical exposure study placed the occupation near the top of the automation distribution, with a 94 percent estimated probability that automation systems replace accountants and auditors (Frey and Osborne, 2017), and clerks and tax examiners near 0.93. Honesty demands the critique: the same predictions add no forecasting power for realized occupation-level employment changes between 2013 and 2018 (Coelli and Borland, 2019), and task-based reanalysis finds lower exposure once social-interaction bottlenecks are counted (Arntz, Gregory and Zierahn, 2016). Realized data nonetheless show the split our design expects: bookkeeping and auditing clerks decline six percent over 2024 to 2034 while accountants and auditors grow, with the profession moving toward analytical and advisory work (Bureau of Labor Statistics projections). The clerical automation parable carries the same two-sided lesson: automated teller machines cut tellers per urban branch from twenty to thirteen, banks opened 43 percent more branches, and teller jobs did not disappear (Bessen, 2015), yet once mobile banking automated the whole task rather than a slice, teller employment entered a prolonged decline. Payment-native taxation is whole-task automation of compliance bookkeeping; the guarantee below exists for exactly that reason.
The Redeployment Thesis. Redeploying accountants as small-enterprise productivity advisors is not hope; the underlying intervention carries randomized evidence. Free consulting on management practices, randomly assigned across large Indian textile plants, raised productivity by 17 percent in the first year through quality, efficiency and inventory improvements, worth roughly 325,000 dollars of additional annual profit per firm, and informational barriers explained prior non-adoption (Bloom, Eifert, Mahajan, McKenzie and Roberts, 2013). The consulting content, namely measurement routines, standardization and feedback loops, is the accountant's native skill set aimed at production instead of at the tax code. External validity stays flagged: one industry, large firms, professional consultancy; hence pilot and measure.
The German Template and the Demographic Tailwind. A real state ended an entire occupation with zero forced dismissals, and every mechanism transfers. German hard coal declined across sixty years from more than 600,000 employees (Oei, Brauers and Herpich, 2020); a 2007 agreement among company, federal government, regions and union ended mining by 2018 in a socially just manner, with employment reorientation, transfer options and early retirement rights (Agora Energiewende, 2025); the toolkit included early-retirement adjustment money, retraining, transitions for younger workers, and no layoffs for operational reasons (union-negotiated measures documented in Environmental Research: Energy, 2025), plus three years of decommissioning work followed by five years of bridge payments to pension, and statutory pension top-ups for workers aged fifty-eight and above (Library of Congress, 2020). The demographic parallel is stronger for accountants than for miners: approximately 75 percent of certified public accountants are near retirement age, while accounting graduates fell twenty percent since 2010, about 300,000 accountants and auditors left their positions within three years, and roughly 190,000 to 200,000 positions stand open (industry data: AICPA Trends, 2025, and press reporting), so natural attrition performs most of the transition, current market shortage absorbs much of the rest, and the state guarantee functions as a backstop option rather than a payroll. The anticipated reviewer objection, namely why retire a profession in shortage, answers itself: the shortage is a shortage of compliance labor for a compliance system the proposal deletes; deleting the system deletes the demand. The institutional funding pattern has European lineage reaching back to the 1951 European Coal and Steel Community fund for training and redeployment of workers (Robert Schuman Foundation, 2020), with a corrective lesson from the auditors: place-based transition funds showed limited focus and impact on job creation (European Court of Auditors, 2022). Our design therefore attaches guarantees to persons, not places, and deploys advisors demand-led through vouchers rather than through grants to regions. The governing sentence of the whole section: grandfather the person, sunset the function.
SMALL FIRMS FIRST
The Arithmetic of the Smallest Firm. In a four-person firm where one person handles compliance, one quarter of the firm's human capital produces nothing sellable; in a three-person firm, one third. Reality approaches the arithmetic: the average surveyed entrepreneur spends 36 percent of the work week on administrative tasks, with three in ten spending between one quarter and one half of working hours on such tasks (Time etc survey, reported in Forbes, 2023); small firms average two days per month on financial administration, thirteen months of work for twelve months of pay, with half of small-business chief executives spending four hours weekly on payment issues (Sage, 2025); and owners spend nearly five hours per pay period calculating, filing and remitting payroll taxes (Intuit QuickBooks survey, reported by Bloomberg Tax, 2020), while 41 percent of small business owners report three to ten hours monthly on payroll taxes, with another ten percent above ten hours (National Small Business Association survey), under real penalty exposure, since United States civil penalties exceeded 65.5 billion dollars in fiscal 2023, including 8.5 billion dollars on employment-tax problems (Internal Revenue Service Data Book, 2023). Survey figures are graded as surveys; the regressivity evidence presented earlier carries the load.
Who Creates Jobs, and Who Keeps the Surplus. The famous claim that small firms create the jobs needs an academic correction, and the correction sharpens the targeting: once firm age is controlled for, no systematic relationship between firm size and growth remains, and startups and young businesses drive gross and net job creation (Haltiwanger, Jarmin and Miranda, 2013). Young firms are almost always small and meet the fixed compliance cost at the moment of entry, exactly where the entry elasticity documented earlier operates. A second targeting reason concerns where surplus goes: affiliates of foreign multinationals are an order of magnitude more profitable than local firms in low-tax countries, and an estimated 36 percent of multinational profits shift to tax havens globally (Torslov, Wier and Zucman, 2023), with over 600 billion dollars shifted annually and the European Union losing an estimated 18 percent of corporate tax revenue, and with the authors noting that profit shifting lowers effective rates for multinationals relative to local firms, affecting competition. Domestically owned small firms cannot shift, and so carry both full tax and full compliance burden, penalized twice. The interpretive debate around such estimates stays flagged (see Tax Foundation, 2018, for the tax-competition reading); the framing is fiscal integrity and structure, never nationality.
The Layering Axiom and Compounding in Both Directions. The administrative-reform literature states our axiom in its own words: "Reducing administrative compliance costs means eliminating non-productive expenditures for business" (den Butter and Hudson, 2009, as quoted in the Standard Cost Model literature). One sentence of analogy and no more: an administrative apparatus that consumes the productive capacity the apparatus exists to serve behaves like a misdirected immune system, while the plural, redundant allocation of capital across many banks, a design our research program endorses elsewhere, behaves like a healthy one. Accumulation compounds: cumulated regulation slowed United States growth by an average of 0.8 percent per year since 1980, leaving the 2012 economy nearly 25 percent smaller than the constant-regulation counterfactual (Coffey, McLaughlin and Peretto, 2020), and because growth is exponential, a seemingly small annual figure grows into a dramatic difference in levels. Removal compounds identically in reverse: the Dutch programme cut regulatory costs on business by 25 percent cumulatively in four years, eliminating four billion euro of administrative burdens by 2007 and lowering burdens from 3.7 to 2.8 percent of gross domestic product (World Bank, 2007, reporting Kox, 2005), using a numerical target, an independent measurement body and budget-cycle linkage, the governance template our proposal adopts, since a public percentage target concentrates pressure while a separate measuring organization removes the bias to underestimate costs. Two steelmen receive direct answers. First, small-firm favoritism can freeze economies into low-productivity dualism, and group-affiliated firms are indeed more productive on average; our instrument therefore removes a distortion rather than adding one: no subsidy for staying small exists anywhere in the design, thresholds taper smoothly, and the stated goal is growth through the size distribution, because our paper protects no firm's smallness and instead removes a penalty on becoming bigger. Second, regulation protects people; our target is the information obligation, never the substantive protection, a line the Standard Cost Model tradition itself draws by excluding administrative actions that correspond to what an entity would normally do absent any legal obligation: keep the regulation's goal, delete the regulation's paperwork.
THE PROPOSAL: FIVE PROVISIONS
Provision One, the Compliance-Exemption Rule. Any enterprise at or below the smallness threshold that routes all business payments through the payment instrument issued by the national central bank owes no bookkeeping, no filing, and no invoicing obligations of any kind; the transaction stream is the books. Invoices generate automatically as a by-product of payment rather than as a precondition of payment, the exact inversion of the Italian clearance model described earlier. Assessment and collection complete at settlement under whatever rate schedule the legislature has chosen, because rate schedules are arithmetic once the stream is the ledger; a single flat rate maximizes what the automation reaches, and the choice of regime remains outside the claims of our paper. The rule is opt-in: an enterprise preferring conventional books keeps conventional obligations, and cash remains fully lawful. Participation is purchased with freedom from paperwork, never compelled.
Provision Two, Sovereign Rails. The instrument is issued by the central bank rather than procured from a multinational processor, for three reasons. Payment rails are the deepest marketplace of a nation and belong to the instrumental layer that sound design keeps thin, cheap and universal; population-scale public instant-payment systems in several large economies demonstrate feasibility, and settlement-collected bank-transaction levies have operated at national scale with material yields, since such levies in Latin America yielded between 0.3 and 1.9 percent of gross domestic product at rates of 15 to 150 basis points (Matheson, 2011; Coelho, Ebrill and Perry, 2001). Public rails run at cost with guaranteed access, where private rails extract per-transaction rents and can exclude. The complete transaction graph of a nation is strategic information whose custody must not sit, by architecture, in foreign private hands, a claim about custody and architecture, never about nationality. The rails give the state a sovereign settlement-and-record infrastructure, sealed by default.
Provision Three, Fiscal Secularity. The historical failure of settlement-collected taxation was political, not technical: total transaction visibility implied total surveillance, and the institutional literature names trust as the binding constraint, since real-time seamless taxation requires legal and privacy frameworks without which taxpayers would not trust the system (OECD, 2022). Provision Three supplies the missing precondition: Fiscal Secularity, the constitutional and technical separation of money, taxation and governance. Records are immutable, sealed by default, and conditionally accessible only through a judicial gate; every access is itself logged, and the log entry, meaning who looked, when, and on what stated ground, becomes permanently visible to the citizen concerned, on the Estonian precedent where any transaction or information access is recorded in several places and citizens can monitor the time and access point of their data files through the government portal (OECD, 2020b). State action upon records is confined to theft and fraud; evidence obtained outside the gate is inadmissible; freezes are judicial, logged, appealable and deliberately slow; the ledger is append-only, with corrections performed as new court-ordered compensating transactions rather than administrative edits; the system's design is fully public, because security must never depend on secrecy of design: sealed data, transparent design. Two further features answer the strongest empirical warning against automated collection. Electronic toll collection, by lowering the salience of tolls, left rates twenty to forty percent higher than counterfactual, with charge-setting less sensitive to the electoral calendar (Finkelstein, 2009): silent rate creep is the documented failure mode of frictionless collection. The answers are architectural: every deduction and the deduction's destination appear on every transaction record, restoring the salience that automation would otherwise destroy, and the rate is capped constitutionally, with the cap deliberately difficult to amend. The exemption buys participation; Fiscal Secularity buys trust; without the second, the first would be surveillance bait.
Provision Four, the Accountant Guarantee. Current tax-compliance accountants hold a voluntary, closed-cohort menu: bridge to pension on the German mechanics described earlier; redeployment into a small-enterprise productivity advisory corps, funded through vouchers that small firms redeem with certified advisors, public or private, so demand rather than administration steers deployment; system-auditor and onboarding roles inside the new payment-rail administration; or ordinary market absorption, which the present shortage makes likely. Training pipelines redirect away from the sunsetting function; a statutory sunset date closes the cohort. No person is conscripted; the guarantee binds the state.
Provision Five, Rail-Guards. Free state tools cover every remaining obligation during transition, on the precedent that free government tools already reduce compliance costs for the smallest businesses to near zero (practitioner documentation of the Italian state tooling). Disputes resolve algorithmic-first: consensual refunds trigger automatic tax clawback, duplicate detection is trivial on a single rail, and human arbitration exists only on appeal, so the dispute system cannot grow into a new bureaucracy. Credit on top of the universal rail comes from plural competing banks, never from the central bank, and transaction-stream underwriting widens small-firm credit access. The smallness threshold tapers smoothly, with full exemption at ten or fewer employees declining linearly to zero at twenty-five, so splitting a company into fragments gains almost nothing; structured fragmentation, detectable as a graph pattern of shared owners, shared addresses and self-circulating flows, is classified as fraud and sits above the judicial threshold: money is inviolable, lying to the system about identity is not. A standing measurement office with a public numerical burden target operates on the Dutch pattern. Investor-facing reporting of listed enterprises stays untouched, and voluntary management accounting remains private practice.
Sequencing and the Political Deal. Rights first, meaning the Fiscal Secularity statute and entrenchment; rails second; offer third; sunset fourth. Rights precede rails so that no interval exists in which the state holds the data without the constraints. Every constituency receives a named benefit: the smallest firms receive zero paperwork; accountants receive a dignified guarantee; the treasury receives the third-party-reporting compliance regime and the deletion of the credit-chain fraud surface; citizens receive visible taxation and constitutionally inviolable money; large firms lose nothing; the privacy-minded keep cash, optionality, and the access log. The historical objection that bank debit taxes failed in Latin America receives a design answer rather than a denial: the record shows short-term revenue success with significant adverse allocational impacts, and better performance where intermediation was deep and rates modest (Coelho, Ebrill and Perry, 2001), including base contraction of 28 to 47 percent with deadweight losses of 30 to 45 percent of revenue and cascading pressure toward vertical integration (Kirilenko and Summers, 2004, as analyzed in Matheson, 2011). The failures ran through easy exit into cash and checks in a pre-digital payments world, through mandatory surcharges stacked on top of existing taxes, and through zero trust architecture. The present design inverts all three conditions, offering digitized payment economies, an opt-in exemption that replaces rather than stacks, and Fiscal Secularity, while adopting the record's one positive lesson verbatim: the rate must stay modest, hence the cap.
EVALUATION DESIGN, OPEN PARAMETERS, ANTICIPATED FINDINGS
Our paper specifies a pilot in one region or sector: an opt-in cohort against matched controls, with outcomes named in advance, namely adoption rate among eligible firms; compliance hours measured against Standard-Cost-Model baselines; firm entry and formalization rates; tax-gap movement; dispute volumes and resolution times; and access-log complaint rates as the trust metric, a measure our paper believes to be novel as a headline indicator of state-citizen data relations. Open parameters, named for reviewers and legislatures alike: the threshold and taper values; the rate regime and cap level; voucher size and advisor certification; the sunset date; and the entrenchment procedure for the cap. Anticipated findings, stated falsifiably: majority adoption among eligible small firms within the pilot horizon, driven by the exemption's cash value; near-complete elimination of measured compliance hours for adopters, against the fifty-hour Estonian floor as the benchmark to beat; entry effects consistent with the published entry elasticity documented earlier; fraud displacement toward identity misrepresentation rather than transaction concealment, detectable as graph anomalies; and net transition cost near zero, given attrition demographics and advisory output valued against the Bloom benchmark. A null result on adoption would itself be informative, indicating that trust rather than paperwork binds, and would redirect design effort toward the Fiscal Secularity layer.
LIMITATIONS AND HONEST UNCERTAINTIES
Six limits stay open. First, cascading: any turnover-style instantiation taxes gross flows at every stage, favoring consolidation; the counterweights, meaning capped low rates, the small-producer exemption, and compliance savings that dominate for small firms, are arguments rather than proofs, and the Pakistani evidence arrives from a 0.5 percent rate, so extrapolation upward is untested. Second, the financial sector: securities, interbank and currency flows dwarf production flows, and full-rate treatment would suppress them; scenario analysis rather than assertion belongs in the companion work of our research program. Third, external validity: the Danish, Chilean, Peruvian and Indian results emerge from specific institutions; the pilot exists because transplantation is a hypothesis. Fourth, capture: a constitutional cap and visible logs are designs against political drift, and designs can fail; the Finkelstein result is a warning about exactly such drift, answered but not abolished. Fifth, the informal margin: opt-in design converts informality into a standing offer rather than a dragnet, and the conversion rate is unknown. Sixth, survey-grade inputs: owner-time figures and profession-pipeline numbers derive from industry sources and are marked as such wherever used; no load-bearing conclusion rests on them alone.
CONCLUSION
The tax systems of the industrial era hired, in effect, millions of private workers to copy information the economy already produced, paid them out of the pockets of the smallest firms, and called the arrangement compliance. The digital settlement rail makes the arrangement optional for the first time: where every payment is a third-party-verified record, the record-keeping profession's compulsory layer loses its object, evasion approaches the measured floor of third-party-reported income, and the state can afford to buy, with an exemption, a guarantee, and a constitutional promise, what audits never achieved. Our paper has argued that the exchange is implementable now, regime-agnostic, politically survivable because every veto player is compensated, and honest about the objections of theory, history and salience, because each objection is cited first and absorbed by design. The title names the intended ending: the last compliance accountant does not lose a job; the last compliance accountant finishes a career with a public guarantee, trains a small firm to measure and improve real production, and retires from a function the country has outgrown, with honor, on schedule, and unreplaced.
Small enterprises employ labor whose only function is preparing tax records for the government; compliance costs approach two percent of gross domestic product and fall hardest on the smallest firms. Our paper asks whether a state can collect tax from small enterprises with zero manual accounting work, answering through design analysis of payment-native taxation, where assessment, collection and record-keeping complete at settlement. Electronic-invoicing mandates in Italy and Peru digitized documents yet left small firms dependent on accountants, while audit evidence finds evasion near zero wherever third parties report transactions. Our central proposal is a statutory exchange: any small enterprise routing every payment through an optional central-bank payment instrument receives exemption from bookkeeping, filing and invoicing duties, with invoices generated automatically, tax collected at settlement under any rate regime, and cash staying lawful. Current tax-compliance accountants hold a voluntary employment guarantee to retirement with advisory redeployment, and citizen trust rests on Fiscal Secularity, the constitutional separation of money, taxation and governance: deductions stay visible at a capped rate, and sealed records open only through a judicial gate whose accesses are logged for the citizen concerned. Our paper specifies a piloted evaluation with parameters named in advance, adoption, compliance hours against standard-cost baselines, firm entry, access-log complaints, anticipating near-complete compliance-hour elimination for adopters at near-zero transition cost as most certified accountants approach retirement. A design recommendation follows: automating state-facing paperwork frees labor without destroying consumer demand, and relief aimed at the smallest firms removes a penalty on growth rather than protecting smallness.
INTRODUCTION: THE PROBLEM OF COMPLIANCE LABOR
Every economy contains a class of labor that produces nothing a customer can buy. A worker in a bakery who bakes creates goods; a worker in the same bakery who prepares tax records creates, for the benefit of the state, a copy of information the bakery already generated by transacting. The classical tradition condemned the excess directly: the fourth maxim of taxation in Smith (1776) demands that every tax take and keep out of the pockets of the people as little as possible beyond what enters the public treasury. The modern literature gives the same idea an empirical body. Americans will spend more than 7.9 billion hours complying with tax filing and reporting requirements in 2024, equal to 3.8 million full-time workers doing nothing but tax return paperwork, and total compliance costs reach approximately 546 billion dollars, or nearly two percent of gross domestic product (Tax Foundation, 2024, computed from Internal Revenue Service estimates). The compliance literature adds a distributional verdict: tax compliance costs are regressive in nature, imposing a heavier burden on small businesses compared to larger ones, and small businesses function, in the standard phrase of the field, as "unpaid tax collectors" (Pope and Rametse, 2001).
Our paper asks one question and answers one question: can a state collect tax from small enterprises while requiring zero manual accounting work from those enterprises? Our answer proceeds by design analysis rather than by advocacy of any particular tax rate or tax base. Our paper defines payment-native taxation, distinguishes payment-native taxation from the document digitization currently marketed as automation, grounds the enforcement mechanism in the strongest experimental evidence available in public economics, proposes a five-provision legislative package centered on a voluntary exchange of payment routing for compliance exemption, designs the transition of the existing accounting workforce as a first-class component, and specifies a piloted evaluation with parameters named in advance. Foundational to the whole argument stands one axiom our research program applies across many papers: an economy contains a productive layer, where goods and services people need come into existence, and an instrumental layer of money, records and administration that exists only to serve the productive layer; good design keeps the instrumental layer thin, cheap and unable to mistake itself for the point.
DEFINITIONS, THE AUTOMATION LADDER, AND PRIOR ART
Payment-Native Taxation Defined. Our paper uses the term automated taxation strictly. Automated taxation means that the assessment of tax, the collection of tax, and the record-keeping surrounding tax are all completed by the payment infrastructure itself at the moment a payment settles, with no separate human bookkeeping anywhere in the chain. Any arrangement demanding less should be called digitization of paperwork, because such an arrangement digitizes an artifact while preserving the human workflow around the artifact.
The Ladder. Existing regimes occupy four rungs. Rung zero holds manual paper filing. Rung one holds digitized documents: clearance-model electronic invoicing, in which the state receives and validates every invoice in real time. Rung two holds state-computed assessment: pre-filled returns built from third-party data, which the taxpayer merely confirms. Rung three holds payment-native taxation, in which the payment event is the tax event and no filing exists at all. Every celebrated digital tax success of the past decade sits on rungs one and two. The contribution of our paper consists in naming the ladder, demonstrating with evidence that rungs one and two do not remove the accountant, and specifying rung three as an implementable, voluntary design for small enterprises.
The Italian Lesson. Italy provides the best-in-class case precisely because Italy leads the field. Italy is widely regarded as a pioneer, and since January 2019 all domestic business-to-business and business-to-consumer invoices must be issued electronically and transmitted through the Sistema di Interscambio, the central exchange platform of the tax authority, and by January 2024 all remaining exemptions were removed (European Commission, 2025). The platform validates format, mandatory fields and identification numbers; a human being, or software a human must configure and pay for, still composes every invoice. The burden consequently relocated instead of disappearing: the initial burden was significant, particularly for micro-enterprises adapting to XML-based invoicing, and implementation into existing accounting and enterprise systems proved very costly, specifically for smaller entities, sparking a new market for e-invoicing consolidators (Grant Thornton, 2024). The decisive fact concerns the accountant: after universal adoption, survey evidence shows that 26 percent of Italian founders fully delegate accounting activities to their accountant, and 69 percent rely entirely on their accountants for value-added-tax and social-contribution deadlines, while 52 percent of companies still manage most financial processes manually or with disconnected tools (wamo and FidoCommercialista, 2026). The accountant survives every digitization of the document because the tax code's complexity, not the paper, created the accountant. Honesty requires the counterweight: the Italian system matured, now supplies pre-filled value-added-tax returns, and demonstrably serves the state, since the value-added-tax compliance gap fell in Italy by 10.7 percent, 12.7 billion euro, in 2021 compared to 2020, the largest drop among the EU-27 (European Commission VAT gap reporting). Italy therefore proves two claims at once: national real-time transaction infrastructure is feasible, and collecting documents about payments, rather than collecting tax inside payments, leaves the accountant standing.
The Peruvian Evidence. Causal evidence confirms the Italian reading. Using quasi-experimental variation in the roll-out of value-added-tax electronic invoicing in Peru, adoption increased reported firm sales, purchases and tax liabilities by over five percent in the first year, concentrated among small firms and less compliant sectors (Bellon, Dabla-Norris, Khalid and Lima, 2022). Two further findings serve our thesis: some firms exited when the requirements were announced, revealing real adoption costs, and existing stocks of tax credits offset the reform's revenue gains, so digital invoicing requires complementary reforms. Document digitization helps the state, imposes costs on firms, and cannot deliver the full dividend, because the credit-and-refund architecture claws gains back.
Prior Art. Payment-native taxation carries prior art in four families. First, the settlement-collected transaction levy: in simplest form, the Automated Payment Transaction tax consists of a flat tax levied on all transactions, automatically assessed and collected when transactions settle through the electronic technology of the banking and payments system (Feige, 2000), a proposal serious enough to have been presented to the United States President's Advisory Panel on Federal Tax Reform in 2005, with the elimination of filing stated as an explicit property, since automated recording of payments eliminates the need to file tax and information returns. Second, withholding at source: the Organisation for Economic Co-operation and Development names the automation of taxation processes within payroll software in pay-as-you-earn systems as a current and widespread example of taxation performed inside taxpayers' natural systems. Third, state-computed assessment: Estonian taxpayers review pre-filled forms and approve with a digital signature in three to five minutes, with one-click returns available in under a minute (European Commission, 2019). Fourth, the official strategic vision: the same organisation sets out a vision under which taxation becomes a more seamless and frictionless process over time, with the payoff stated plainly, because moving taxation into taxpayers' natural systems aligns tax with taxable events, producing fewer errors and substantially reduced burdens (OECD, 2020a). Our proposal is the terminal station of a line already drawn by the field's most conservative institution.
The Enforcement Mechanism. The mechanism behind the entire design carries the strongest causal evidence in tax economics, developed as an extension of the canonical evasion model of Allingham and Sandmo (1972), since the empirical results are explained by extending the standard model of rational tax evasion to distinguish self-reported from third-party reported income. In a randomized audit experiment covering over 40,000 Danish taxpayers, the evasion rate was 0.3 percent for income subject to third-party reporting and 37 percent for self-reported income, and since 95 percent of income was third-party reported, overall evasion stayed very modest; furthermore, audits and audit threats affected only self-reported income, with no effect on third-party reported income (Kleven, Knudsen, Kreiner, Pedersen and Saez, 2011). Complementary field experiments in Chile show that announcing additional monitoring has less impact on transactions already covered by a paper trail, indicating the trail's preventive deterrence effect, while information without credible deterrence does not self-enforce (Pomeranz, 2015). Payment-native taxation is the limiting case of the Danish result: when every taxable event is a third-party-verified settlement, the 0.3 percent regime extends to the whole participating economy, and the Chilean boundary, namely a trail that historically stopped at the final consumer, closes, because a payment rail records the retail transaction as completely as the wholesale transaction.
WHY THE STATE SHOULD FUND THE AUTOMATION
The Category Error and the Cost-Shifting Asymmetry. Compliance labor belongs to the instrumental layer, and the instrumental layer has quietly transferred most of the tax system's true operating cost onto private firms (for the survey of compliance and enforcement economics, see Slemrod, 2019). Private compliance cost for the United States income tax runs near ten cents per dollar collected, while the tax authority's own administrative cost runs near 0.6 cents per dollar, so private compliance costs dwarf public administrative costs in almost all cases (Slemrod, 1996; Sandford, 1995). Roughly ninety-four percent of the collection machinery's cost sits off the state's books, carried by firms and, through firms, by the smallest firms most heavily. The state authored the obligation; the state is the cheapest cost avoider; the state should fund the redesign.
Regressivity as Structure. Compliance cost behaves as a fixed head tax per firm. Econometric work finds compliance costs growing only 0.34 to 0.42 percent for each one percent of firm size, with a high constant indicating fixed cost elements (Eichfelder and Schorn, 2012), producing measured burden ratios across firm sizes of twenty-one to one, and up to one hundred and two to one under variable monetization (Internal Revenue Service Research, 2012), while in lower-income countries some small businesses incur compliance costs exceeding twelve percent of turnover (International Centre for Tax and Development, 2023). A four-country comparative study concluded that compliance costs remain high and regressive and do not appear to diminish over time, while the managerial benefits of compliance bookkeeping exist yet few small businesses can value them (Evans, Hansford, Hasseldine, Lignier, Smulders and Vaillancourt, 2014); our design absorbs the managerial-benefits counterpoint, because the transaction stream preserves the information while deleting the labor. The growth mechanism appears in the same literature: high compliance costs divert resources from productive activities and increase input costs without creating additional output (Dabla-Norris, Misch, Cleary and Khwaja, 2017).
Firms That Refuse to Grow. Burdens with thresholds teach firms to stay small. United Kingdom administrative records document bunching in annual turnover just below the value-added-tax registration threshold (Liu and Lockwood, 2021), and follow-up work finds annual turnover growth slowing by about one percentage point as firms approach the threshold, with no offsetting growth after crossing (Liu, Lockwood and Tam, 2024); the mechanism is explicit, since firms keep reported turnover below the threshold either legally, by restricting the scale of operations, or illegally, by misreporting sales. The same pathology appears at the French fifty-employee wall: size-contingent regulations cost the equivalent of a 2.3 percent variable labor tax, with welfare losses up to 3.4 percent of gross domestic product under wage inflexibility (Garicano, Lelarge and Van Reenen, 2016), and twelve percent of firms stayed at exactly forty-nine employees for two consecutive years, against two percent at fifty-two. Threshold burdens even corrupt the statistical record, since a large share of French employers manipulate declared workforce size, and the manipulation itself shapes the observed size distribution (Askenazy, Breda, Moreau and Pecheu, 2022). Our design answers with smooth tapering rather than cliffs, as the proposal section specifies.
The Return on Public Investment, and Regime Independence. Public digitalization pays for the state directly: the use of digital signatures in Estonia is estimated to save two percent of gross domestic product every year (OECD, 2020b, reporting a government estimate), and in Estonia total business tax compliance takes roughly fifty hours a year (World Bank Paying Taxes research). The growth channel is quantified across 118 economies: firm entry responds to the ease of paying taxes regardless of the corporate tax rate, and a ten percent reduction in tax administrative burden associates with a three percent increase in annual business entry (Braunerhjelm and Eklund, 2014, as reported in the World Bank Paying Taxes research), while complexity associates with evasion, and high compliance costs associate with larger informal sectors and less investment. The final clause carries the design consequence our paper builds on: the burden lives in process, meaning filings, payments and documentation frequency, not in rates. China cut compliance from 832 hours and thirty-seven payments per year to 138 hours and seven payments without abolishing taxation. The automation investment therefore pays under a value-added tax, under a turnover levy, under any regime a legislature chooses, and our paper accordingly presupposes no particular tax.
FROM FREED LABOR TO NATIONAL CAPABILITY
Misallocation, and a Reallocation with No Demand Cost. Growth economics names compliance labor precisely: misallocated resources. When capital and labor are hypothetically reallocated to equalize marginal products to the degree observed in the United States, manufacturing total-factor-productivity gains reach 30 to 50 percent in China and 40 to 60 percent in India (Hsieh and Klenow, 2009). Labor whose entire function is reproducing records for the state holds a marginal product of production near zero, so reallocation of such labor is a first-order productivity gain. One asymmetry makes the gain unusually clean, and one sentence of our research program states the asymmetry: automate the state ruthlessly, automate the market carefully. Automating a factory risks the demand spiral of fewer workers and fewer consumers; automating state-facing paperwork destroys no consumer demand, because the paperwork was never a product anyone consumed. Freed compliance labor is the purest reallocatable resource an economy holds.
Security Capability and Cultural Influence. A larger productive base underwrites national capability in two documented directions, and both are held here as instruments for citizens, never as ends. On security: economic strength and military power have been highly correlated in the rise and fall of major nations since 1500 (Kennedy, 1987), summarized in Kennedy's own words: "wealth is usually needed to underpin military power, and military power is usually needed to acquire and protect wealth". The policy relevance runs through the constant triple tension between investment, defense and consumption: recovering one to two percent of gross domestic product of pure waste relaxes the triple tension without touching either investment or consumption. On culture: economic resources feed attraction as well as coercion, since economic resources can produce soft as well as hard power (Nye, 1990; 2011), and South Korea proves the conversion is designable, because the 1998 Hallyu Industry Support Development Plan raised cultural spending from 14 million dollars to 84 million dollars by 2001 (Sellars and Leasure, 2025), the 1999 framework statute for promoting cultural industries institutionalized support (Republic of Korea, 1999), and by 2021 the content industry generated 137.5 trillion won in sales and 12.45 billion dollars in exports, with measured spillovers of 180 million dollars of consumer-goods exports and nearly three thousand jobs per additional 100 million dollars of content exports (Invest Korea, 2023, reporting Korea Creative Content Agency data). Cultural influence is downstream of fiscal surplus plus deliberate design; compliance automation supplies the surplus.
Why Replacing the Value-Added Tax Helps, Stated Within Strict Limits. Our paper does not claim that the value-added tax burdens production in any special economic sense; the critique runs through administration and visibility only, and rests on three charges. Charge one, compliance intensity: complying with value-added tax takes a case-study company longer on average than complying with corporate income tax, because the credit-invoice mechanism runs on matched document chains, which is exactly why the accountant survives digitization. Charge two, fraud architecture: carousel fraud exploits the fact that the tax is not immediately applied to cross-border business transactions, letting fraudsters buy free of tax, sell tax-inclusive, and vanish; the magnitudes are large, since around 90 billion euro of value-added-tax revenue was lost in the European Union in 2022, of which at least 13 billion euro links conservatively to carousel fraud (European Commission, 2024; Poniatowski, Smietanka and Skowronek, 2024), with carousel fraud making up roughly one quarter of the total gap, and the enforcement contest is permanent, even though national digital reporting requirements did add an estimated 19 to 28 billion euro of revenue between 2014 and 2019 (European Parliamentary Research Service, 2025). A settlement-collected levy holds no credit chain, no refund claim, and no missing-trader position to construct. Charge three, salience: consumers underreact to taxes that are not salient, demonstrated by an eight percent demand fall when tax-inclusive prices were posted (Chetty, Looney and Kroft, 2009); a tax living permanently inside posted prices is the maximally non-salient design, and a democracy cannot audit what citizens cannot perceive. The named theoretical objection to any transaction levy must appear here, and our paper cites the objection first: the Diamond-Mirrlees production-efficiency theorem dictates that taxes should not distort production, implying no taxation of intermediate inputs or turnover (Diamond and Mirrlees, 1971). The modern answer is empirical: the theorem relies on perfect enforcement, meaning zero evasion at zero administrative cost, and where enforcement is imperfect, turnover taxes reduced evasion by up to 60 to 70 percent of corporate income in Pakistani administrative data, and switching from profit to turnover taxation raised revenue by 74 percent without reducing aggregate profits (Best, Brockmeyer, Kleven, Spinnewijn and Waseem, 2015), with the distortion bounded because the turnover rate was 0.5 percent against a 35 percent profit rate. Cascading through supply chains remains a real cost, favors vertical integration, and must stay openly acknowledged; low capped rates, the small-producer exemption of the proposal, and compliance savings that dominate for small firms are the counterweights, and the rate itself remains an open parameter of the design rather than a claim of our paper.
THE ACCOUNTANT TRANSITION: GRANDFATHER THE PERSON, SUNSET THE FUNCTION
The Political Economy of Concentrated Losers. Reforms with diffuse gains and concentrated losses fail routinely, because organized losers mobilize while diffuse winners sleep. The compensation principle says winners could compensate losers; our design actually does so, converting the accounting profession from veto player into implementation partner: the people who know where every buried complexity lies become the people paid to help retire the system that buried the complexities.
The Cohort, the Exposure, and the Honest Reading of Automation Forecasts. Accountants and auditors held about 1.6 million jobs in the United States in 2024 at a median wage of 81,680 dollars (Bureau of Labor Statistics, 2025). The canonical exposure study placed the occupation near the top of the automation distribution, with a 94 percent estimated probability that automation systems replace accountants and auditors (Frey and Osborne, 2017), and clerks and tax examiners near 0.93. Honesty demands the critique: the same predictions add no forecasting power for realized occupation-level employment changes between 2013 and 2018 (Coelli and Borland, 2019), and task-based reanalysis finds lower exposure once social-interaction bottlenecks are counted (Arntz, Gregory and Zierahn, 2016). Realized data nonetheless show the split our design expects: bookkeeping and auditing clerks decline six percent over 2024 to 2034 while accountants and auditors grow, with the profession moving toward analytical and advisory work (Bureau of Labor Statistics projections). The clerical automation parable carries the same two-sided lesson: automated teller machines cut tellers per urban branch from twenty to thirteen, banks opened 43 percent more branches, and teller jobs did not disappear (Bessen, 2015), yet once mobile banking automated the whole task rather than a slice, teller employment entered a prolonged decline. Payment-native taxation is whole-task automation of compliance bookkeeping; the guarantee below exists for exactly that reason.
The Redeployment Thesis. Redeploying accountants as small-enterprise productivity advisors is not hope; the underlying intervention carries randomized evidence. Free consulting on management practices, randomly assigned across large Indian textile plants, raised productivity by 17 percent in the first year through quality, efficiency and inventory improvements, worth roughly 325,000 dollars of additional annual profit per firm, and informational barriers explained prior non-adoption (Bloom, Eifert, Mahajan, McKenzie and Roberts, 2013). The consulting content, namely measurement routines, standardization and feedback loops, is the accountant's native skill set aimed at production instead of at the tax code. External validity stays flagged: one industry, large firms, professional consultancy; hence pilot and measure.
The German Template and the Demographic Tailwind. A real state ended an entire occupation with zero forced dismissals, and every mechanism transfers. German hard coal declined across sixty years from more than 600,000 employees (Oei, Brauers and Herpich, 2020); a 2007 agreement among company, federal government, regions and union ended mining by 2018 in a socially just manner, with employment reorientation, transfer options and early retirement rights (Agora Energiewende, 2025); the toolkit included early-retirement adjustment money, retraining, transitions for younger workers, and no layoffs for operational reasons (union-negotiated measures documented in Environmental Research: Energy, 2025), plus three years of decommissioning work followed by five years of bridge payments to pension, and statutory pension top-ups for workers aged fifty-eight and above (Library of Congress, 2020). The demographic parallel is stronger for accountants than for miners: approximately 75 percent of certified public accountants are near retirement age, while accounting graduates fell twenty percent since 2010, about 300,000 accountants and auditors left their positions within three years, and roughly 190,000 to 200,000 positions stand open (industry data: AICPA Trends, 2025, and press reporting), so natural attrition performs most of the transition, current market shortage absorbs much of the rest, and the state guarantee functions as a backstop option rather than a payroll. The anticipated reviewer objection, namely why retire a profession in shortage, answers itself: the shortage is a shortage of compliance labor for a compliance system the proposal deletes; deleting the system deletes the demand. The institutional funding pattern has European lineage reaching back to the 1951 European Coal and Steel Community fund for training and redeployment of workers (Robert Schuman Foundation, 2020), with a corrective lesson from the auditors: place-based transition funds showed limited focus and impact on job creation (European Court of Auditors, 2022). Our design therefore attaches guarantees to persons, not places, and deploys advisors demand-led through vouchers rather than through grants to regions. The governing sentence of the whole section: grandfather the person, sunset the function.
SMALL FIRMS FIRST
The Arithmetic of the Smallest Firm. In a four-person firm where one person handles compliance, one quarter of the firm's human capital produces nothing sellable; in a three-person firm, one third. Reality approaches the arithmetic: the average surveyed entrepreneur spends 36 percent of the work week on administrative tasks, with three in ten spending between one quarter and one half of working hours on such tasks (Time etc survey, reported in Forbes, 2023); small firms average two days per month on financial administration, thirteen months of work for twelve months of pay, with half of small-business chief executives spending four hours weekly on payment issues (Sage, 2025); and owners spend nearly five hours per pay period calculating, filing and remitting payroll taxes (Intuit QuickBooks survey, reported by Bloomberg Tax, 2020), while 41 percent of small business owners report three to ten hours monthly on payroll taxes, with another ten percent above ten hours (National Small Business Association survey), under real penalty exposure, since United States civil penalties exceeded 65.5 billion dollars in fiscal 2023, including 8.5 billion dollars on employment-tax problems (Internal Revenue Service Data Book, 2023). Survey figures are graded as surveys; the regressivity evidence presented earlier carries the load.
Who Creates Jobs, and Who Keeps the Surplus. The famous claim that small firms create the jobs needs an academic correction, and the correction sharpens the targeting: once firm age is controlled for, no systematic relationship between firm size and growth remains, and startups and young businesses drive gross and net job creation (Haltiwanger, Jarmin and Miranda, 2013). Young firms are almost always small and meet the fixed compliance cost at the moment of entry, exactly where the entry elasticity documented earlier operates. A second targeting reason concerns where surplus goes: affiliates of foreign multinationals are an order of magnitude more profitable than local firms in low-tax countries, and an estimated 36 percent of multinational profits shift to tax havens globally (Torslov, Wier and Zucman, 2023), with over 600 billion dollars shifted annually and the European Union losing an estimated 18 percent of corporate tax revenue, and with the authors noting that profit shifting lowers effective rates for multinationals relative to local firms, affecting competition. Domestically owned small firms cannot shift, and so carry both full tax and full compliance burden, penalized twice. The interpretive debate around such estimates stays flagged (see Tax Foundation, 2018, for the tax-competition reading); the framing is fiscal integrity and structure, never nationality.
The Layering Axiom and Compounding in Both Directions. The administrative-reform literature states our axiom in its own words: "Reducing administrative compliance costs means eliminating non-productive expenditures for business" (den Butter and Hudson, 2009, as quoted in the Standard Cost Model literature). One sentence of analogy and no more: an administrative apparatus that consumes the productive capacity the apparatus exists to serve behaves like a misdirected immune system, while the plural, redundant allocation of capital across many banks, a design our research program endorses elsewhere, behaves like a healthy one. Accumulation compounds: cumulated regulation slowed United States growth by an average of 0.8 percent per year since 1980, leaving the 2012 economy nearly 25 percent smaller than the constant-regulation counterfactual (Coffey, McLaughlin and Peretto, 2020), and because growth is exponential, a seemingly small annual figure grows into a dramatic difference in levels. Removal compounds identically in reverse: the Dutch programme cut regulatory costs on business by 25 percent cumulatively in four years, eliminating four billion euro of administrative burdens by 2007 and lowering burdens from 3.7 to 2.8 percent of gross domestic product (World Bank, 2007, reporting Kox, 2005), using a numerical target, an independent measurement body and budget-cycle linkage, the governance template our proposal adopts, since a public percentage target concentrates pressure while a separate measuring organization removes the bias to underestimate costs. Two steelmen receive direct answers. First, small-firm favoritism can freeze economies into low-productivity dualism, and group-affiliated firms are indeed more productive on average; our instrument therefore removes a distortion rather than adding one: no subsidy for staying small exists anywhere in the design, thresholds taper smoothly, and the stated goal is growth through the size distribution, because our paper protects no firm's smallness and instead removes a penalty on becoming bigger. Second, regulation protects people; our target is the information obligation, never the substantive protection, a line the Standard Cost Model tradition itself draws by excluding administrative actions that correspond to what an entity would normally do absent any legal obligation: keep the regulation's goal, delete the regulation's paperwork.
THE PROPOSAL: FIVE PROVISIONS
Provision One, the Compliance-Exemption Rule. Any enterprise at or below the smallness threshold that routes all business payments through the payment instrument issued by the national central bank owes no bookkeeping, no filing, and no invoicing obligations of any kind; the transaction stream is the books. Invoices generate automatically as a by-product of payment rather than as a precondition of payment, the exact inversion of the Italian clearance model described earlier. Assessment and collection complete at settlement under whatever rate schedule the legislature has chosen, because rate schedules are arithmetic once the stream is the ledger; a single flat rate maximizes what the automation reaches, and the choice of regime remains outside the claims of our paper. The rule is opt-in: an enterprise preferring conventional books keeps conventional obligations, and cash remains fully lawful. Participation is purchased with freedom from paperwork, never compelled.
Provision Two, Sovereign Rails. The instrument is issued by the central bank rather than procured from a multinational processor, for three reasons. Payment rails are the deepest marketplace of a nation and belong to the instrumental layer that sound design keeps thin, cheap and universal; population-scale public instant-payment systems in several large economies demonstrate feasibility, and settlement-collected bank-transaction levies have operated at national scale with material yields, since such levies in Latin America yielded between 0.3 and 1.9 percent of gross domestic product at rates of 15 to 150 basis points (Matheson, 2011; Coelho, Ebrill and Perry, 2001). Public rails run at cost with guaranteed access, where private rails extract per-transaction rents and can exclude. The complete transaction graph of a nation is strategic information whose custody must not sit, by architecture, in foreign private hands, a claim about custody and architecture, never about nationality. The rails give the state a sovereign settlement-and-record infrastructure, sealed by default.
Provision Three, Fiscal Secularity. The historical failure of settlement-collected taxation was political, not technical: total transaction visibility implied total surveillance, and the institutional literature names trust as the binding constraint, since real-time seamless taxation requires legal and privacy frameworks without which taxpayers would not trust the system (OECD, 2022). Provision Three supplies the missing precondition: Fiscal Secularity, the constitutional and technical separation of money, taxation and governance. Records are immutable, sealed by default, and conditionally accessible only through a judicial gate; every access is itself logged, and the log entry, meaning who looked, when, and on what stated ground, becomes permanently visible to the citizen concerned, on the Estonian precedent where any transaction or information access is recorded in several places and citizens can monitor the time and access point of their data files through the government portal (OECD, 2020b). State action upon records is confined to theft and fraud; evidence obtained outside the gate is inadmissible; freezes are judicial, logged, appealable and deliberately slow; the ledger is append-only, with corrections performed as new court-ordered compensating transactions rather than administrative edits; the system's design is fully public, because security must never depend on secrecy of design: sealed data, transparent design. Two further features answer the strongest empirical warning against automated collection. Electronic toll collection, by lowering the salience of tolls, left rates twenty to forty percent higher than counterfactual, with charge-setting less sensitive to the electoral calendar (Finkelstein, 2009): silent rate creep is the documented failure mode of frictionless collection. The answers are architectural: every deduction and the deduction's destination appear on every transaction record, restoring the salience that automation would otherwise destroy, and the rate is capped constitutionally, with the cap deliberately difficult to amend. The exemption buys participation; Fiscal Secularity buys trust; without the second, the first would be surveillance bait.
Provision Four, the Accountant Guarantee. Current tax-compliance accountants hold a voluntary, closed-cohort menu: bridge to pension on the German mechanics described earlier; redeployment into a small-enterprise productivity advisory corps, funded through vouchers that small firms redeem with certified advisors, public or private, so demand rather than administration steers deployment; system-auditor and onboarding roles inside the new payment-rail administration; or ordinary market absorption, which the present shortage makes likely. Training pipelines redirect away from the sunsetting function; a statutory sunset date closes the cohort. No person is conscripted; the guarantee binds the state.
Provision Five, Rail-Guards. Free state tools cover every remaining obligation during transition, on the precedent that free government tools already reduce compliance costs for the smallest businesses to near zero (practitioner documentation of the Italian state tooling). Disputes resolve algorithmic-first: consensual refunds trigger automatic tax clawback, duplicate detection is trivial on a single rail, and human arbitration exists only on appeal, so the dispute system cannot grow into a new bureaucracy. Credit on top of the universal rail comes from plural competing banks, never from the central bank, and transaction-stream underwriting widens small-firm credit access. The smallness threshold tapers smoothly, with full exemption at ten or fewer employees declining linearly to zero at twenty-five, so splitting a company into fragments gains almost nothing; structured fragmentation, detectable as a graph pattern of shared owners, shared addresses and self-circulating flows, is classified as fraud and sits above the judicial threshold: money is inviolable, lying to the system about identity is not. A standing measurement office with a public numerical burden target operates on the Dutch pattern. Investor-facing reporting of listed enterprises stays untouched, and voluntary management accounting remains private practice.
Sequencing and the Political Deal. Rights first, meaning the Fiscal Secularity statute and entrenchment; rails second; offer third; sunset fourth. Rights precede rails so that no interval exists in which the state holds the data without the constraints. Every constituency receives a named benefit: the smallest firms receive zero paperwork; accountants receive a dignified guarantee; the treasury receives the third-party-reporting compliance regime and the deletion of the credit-chain fraud surface; citizens receive visible taxation and constitutionally inviolable money; large firms lose nothing; the privacy-minded keep cash, optionality, and the access log. The historical objection that bank debit taxes failed in Latin America receives a design answer rather than a denial: the record shows short-term revenue success with significant adverse allocational impacts, and better performance where intermediation was deep and rates modest (Coelho, Ebrill and Perry, 2001), including base contraction of 28 to 47 percent with deadweight losses of 30 to 45 percent of revenue and cascading pressure toward vertical integration (Kirilenko and Summers, 2004, as analyzed in Matheson, 2011). The failures ran through easy exit into cash and checks in a pre-digital payments world, through mandatory surcharges stacked on top of existing taxes, and through zero trust architecture. The present design inverts all three conditions, offering digitized payment economies, an opt-in exemption that replaces rather than stacks, and Fiscal Secularity, while adopting the record's one positive lesson verbatim: the rate must stay modest, hence the cap.
EVALUATION DESIGN, OPEN PARAMETERS, ANTICIPATED FINDINGS
Our paper specifies a pilot in one region or sector: an opt-in cohort against matched controls, with outcomes named in advance, namely adoption rate among eligible firms; compliance hours measured against Standard-Cost-Model baselines; firm entry and formalization rates; tax-gap movement; dispute volumes and resolution times; and access-log complaint rates as the trust metric, a measure our paper believes to be novel as a headline indicator of state-citizen data relations. Open parameters, named for reviewers and legislatures alike: the threshold and taper values; the rate regime and cap level; voucher size and advisor certification; the sunset date; and the entrenchment procedure for the cap. Anticipated findings, stated falsifiably: majority adoption among eligible small firms within the pilot horizon, driven by the exemption's cash value; near-complete elimination of measured compliance hours for adopters, against the fifty-hour Estonian floor as the benchmark to beat; entry effects consistent with the published entry elasticity documented earlier; fraud displacement toward identity misrepresentation rather than transaction concealment, detectable as graph anomalies; and net transition cost near zero, given attrition demographics and advisory output valued against the Bloom benchmark. A null result on adoption would itself be informative, indicating that trust rather than paperwork binds, and would redirect design effort toward the Fiscal Secularity layer.
LIMITATIONS AND HONEST UNCERTAINTIES
Six limits stay open. First, cascading: any turnover-style instantiation taxes gross flows at every stage, favoring consolidation; the counterweights, meaning capped low rates, the small-producer exemption, and compliance savings that dominate for small firms, are arguments rather than proofs, and the Pakistani evidence arrives from a 0.5 percent rate, so extrapolation upward is untested. Second, the financial sector: securities, interbank and currency flows dwarf production flows, and full-rate treatment would suppress them; scenario analysis rather than assertion belongs in the companion work of our research program. Third, external validity: the Danish, Chilean, Peruvian and Indian results emerge from specific institutions; the pilot exists because transplantation is a hypothesis. Fourth, capture: a constitutional cap and visible logs are designs against political drift, and designs can fail; the Finkelstein result is a warning about exactly such drift, answered but not abolished. Fifth, the informal margin: opt-in design converts informality into a standing offer rather than a dragnet, and the conversion rate is unknown. Sixth, survey-grade inputs: owner-time figures and profession-pipeline numbers derive from industry sources and are marked as such wherever used; no load-bearing conclusion rests on them alone.
CONCLUSION
The tax systems of the industrial era hired, in effect, millions of private workers to copy information the economy already produced, paid them out of the pockets of the smallest firms, and called the arrangement compliance. The digital settlement rail makes the arrangement optional for the first time: where every payment is a third-party-verified record, the record-keeping profession's compulsory layer loses its object, evasion approaches the measured floor of third-party-reported income, and the state can afford to buy, with an exemption, a guarantee, and a constitutional promise, what audits never achieved. Our paper has argued that the exchange is implementable now, regime-agnostic, politically survivable because every veto player is compensated, and honest about the objections of theory, history and salience, because each objection is cited first and absorbed by design. The title names the intended ending: the last compliance accountant does not lose a job; the last compliance accountant finishes a career with a public guarantee, trains a small firm to measure and improve real production, and retires from a function the country has outgrown, with honor, on schedule, and unreplaced.