Transition — From Scarcity Economics to Abundance

Most economies today are organized around the assumption that labor scarcity, capital scarcity, and resource scarcity are the binding constraints on output. If even one of those assumptions weakens — and the evidence in /abundance/economic-mechanism suggests at least labor scarcity is starting to — the institutional scaffolding built on it does not adapt automatically. The transition to abundance is not a switch; it is a multi-decade reconfiguration of property rights, fiscal architecture, governance, and political coalitions. This page is about what that reconfiguration looks like, what has historically gone wrong, and what concretely we would expect to see if it were going right.

Why transitions are the hard part

The direction of long-run economic change is often visible decades before its institutional consequences are absorbed. The English wool trade was reorganizing rural life in the sixteenth century; the laws governing rural employment, common land, and parish relief did not catch up until the nineteenth.¹ The microprocessor was a commercial product in 1971; competition policy treating platform dynamics as the central question of antitrust did not become mainstream in the OECD until the 2010s. Direction is the easy part of any structural transition. Speed and the distribution of costs are what determine whether the transition is endured, redirected, or reversed.

Karl Polanyi’s The Great Transformation (1944) is the canonical statement of this point.² Polanyi argued that the construction of a self-regulating market economy in nineteenth-century Britain was not a natural evolution but a political project, and that it was met by an equally political “counter-movement” of self-protection by society — factory legislation, municipal sanitation, trade unions, the gold standard’s eventual abandonment, social insurance. The market was made; the response to the market was made; the interaction of the two is what Polanyi called the “double movement.” Polanyi’s empirical claim is contested in detail, but the structural claim — that any disembedding of economic life provokes a reactive re-embedding, and that the speed and form of the reaction determines political outcomes — has held up well.³

Daron Acemoglu and James Robinson’s Why Nations Fail (2012) gives this a sharper institutional edge.⁴ Their distinction between extractive and inclusive institutions is, in essence, a typology of which transitions go well and which do not. Inclusive political and economic institutions diffuse the gains and costs of structural change widely enough that the political coalition for continued reform survives the disruption it causes. Extractive institutions concentrate the gains in a narrow elite that, even when nominally pro-modernization, becomes the primary obstacle to the next phase of it once their position is secured. The history of resource booms (see /abundance/precedents) is largely a history of extractive institutional capture defeating ostensibly favorable economic conditions.

The reason this matters for the abundance transition specifically is that the canonical historical illustration of the counter-movement winning is recent: the Reagan-Thatcher rollback of the post-war social-democratic settlement in the early 1980s.⁵ Margaret Thatcher took office in May 1979; Ronald Reagan in January 1981. Within a decade, top marginal tax rates in the United Kingdom had fallen from 83 percent to 40 percent and in the United States from 70 percent to 28 percent, union density had collapsed in both countries, capital controls had been dismantled across the OECD, and the institutional logic of the Beveridge-era welfare state had been put on the defensive in a way it has never recovered from. The rollback was not the inevitable working-out of impersonal economic forces; it was a political project, organized over two decades by an identifiable network of think tanks, donors, and politicians, and it succeeded because the institutional defenders of the prior settlement underestimated both the speed of the counter-movement and the breadth of its coalition.⁶

That is the empirical baseline for any discussion of the abundance transition. Direction is the easy part. Speed and distribution determine whether the institutional response arrives in time — and whether, having arrived, it survives the next political cycle.

Historical analogue 1: agricultural-to-industrial transition

The longest transition in modern economic history, and the one whose distributional pain is best documented, is the shift from agrarian to industrial employment in Britain and then in the rest of the North Atlantic. It is the right place to start because it tells us, in the most concrete terms, what happens when the productive capacity of an economy reorganizes faster than its institutions.

Robert Allen’s The British Industrial Revolution in Global Perspective (2009) is the most rigorous recent reconstruction of this period.⁷ Allen’s central empirical point is that the British industrial revolution was driven by a specific factor-price configuration — high wages relative to capital and especially to coal — that made labor-saving innovation rational in Britain in a way it was not elsewhere. The institutional precondition for that wage configuration was the prior reorganization of rural life: the enclosure movement, which between roughly 1750 and 1850 converted millions of acres of common and open-field land into individually owned, market-oriented estates, displacing a substantial fraction of the rural population into wage labor.⁸

The displacement was not gentle. Contemporary observers, including those broadly sympathetic to the industrial project, described the conditions of the new urban working class in terms — overcrowding, sanitation collapse, child mortality, occupational disease — that would today be classified as humanitarian emergencies. The “satanic mills” period that Polanyi documents in detail lasted, on any reasonable reading, two full generations: from the parliamentary enclosures of the late eighteenth century through the Factory Acts of the 1830s and 1840s, the public health reforms of the 1870s, and the early social insurance legislation of the 1900s. The institutional response — public education, child labor laws, sanitation infrastructure, eventually pensions and unemployment insurance — arrived after, not during, the displacement that made it necessary.⁹

The American version of the same transition is shorter but structurally similar. Agricultural employment in the United States was approximately fifty percent of the labor force in 1870; it is approximately one and a half percent today, depending on how one counts agricultural support services.¹⁰ That collapse — from one in two workers to roughly one in seventy-five over five generations — happened despite, not because of, planned institutional response. The Homestead Act, the land-grant universities, the Department of Agriculture’s extension service, and the New Deal farm programs are all in some sense responses to the agricultural transition, but they are responses oriented around preserving the political position of the farm sector, not around managing the transition of displaced agricultural labor into urban employment. The actual institutions that absorbed displaced agricultural labor were ad hoc: the migration networks of the Great Migration, the urban political machines of the late nineteenth and early twentieth centuries, the public education system, and eventually the Wagner Act and the post-war welfare state.

Two empirical lessons stand out from the agricultural-to-industrial analogue. The first is that institutional response lagged the productive transformation by roughly fifty years. The conditions that produced the Factory Acts of the 1830s were already visible to anyone willing to look in the 1780s; the conditions that produced the New Deal of the 1930s were structurally in place by the 1880s. The lag is not a quirk of specific countries; it appears across the comparative historical literature.

The second lesson is more subtle. The transition was endured rather than reversed only because the eventual institutional response was sufficient — sufficient in scale, sufficient in coverage, sufficient in the political legitimacy it conferred on the broader project of industrial modernization. There were moments in the nineteenth century when the political coalition for continued industrialization was genuinely fragile: 1848 across continental Europe, the British Chartist agitation of the 1840s, the American populist movements of the 1890s. The fact that industrial modernity ultimately won is contingent on the institutional response arriving when it did. Had the Factory Acts been delayed another generation, or had the welfare-state expansion of the early twentieth century been blocked by the kind of capital-mobility constraints that are now standard, the political settlement we now treat as the natural endpoint of industrialization might never have stabilized.

The lesson for the abundance transition is uncomfortable: a fifty-year institutional lag is the historical default, and the transition is endured rather than reversed only because the response, when it finally comes, is large and inclusive enough. We do not have a fifty-year window. The deployment timelines for the capital-stock changes that drive the abundance question — discussed in detail in /abundance/economic-mechanism — are closer to a decade than to a half-century. The structural mismatch between deployment timescale and institutional response timescale is the central problem this page is about.

Historical analogue 2: post-war welfare-state expansion

The second analogue is more encouraging, because it is the one case in modern history where the institutional response was not just adequate but anticipatory. Between roughly 1942 and 1975, the OECD economies built — in something like a single thirty-year window — the institutional floor that defines what we now mean by a developed-country social contract. It is the only example we have of a deliberate, planned, and largely successful institutional transition at scale, and it deserves close attention as a model for what an abundance transition would have to look like.

The intellectual anchor is the Beveridge Report of November 1942, formally titled Social Insurance and Allied Services.¹¹ Beveridge identified what he called the “five giants” — want, disease, ignorance, squalor, and idleness — and proposed a unified, contributory, universal scheme of social insurance covering unemployment, sickness, retirement, and survivor benefits, complemented by a national health service, family allowances, and a commitment to full employment. The report was published while the war was still being fought, and it was read with extraordinary breadth: roughly six hundred thousand copies sold in the United Kingdom alone within months, against a population of about forty-eight million. It became, almost immediately, the political horizon of post-war reconstruction not just in Britain but across Western Europe.

The American counterpart is the Servicemen’s Readjustment Act of 1944, the GI Bill, which provided returning veterans with tuition and living-expense support for higher education, low-interest mortgages with little or no down payment, and unemployment compensation during the transition to civilian work.¹² The Bill was an act of mass institutional construction: it sent roughly eight million veterans to college or vocational training, financed something on the order of two and a half million home purchases in its first decade, and effectively constructed the post-war American middle class as a political entity. The fact that its benefits were administered in racially discriminatory ways — particularly the housing component, through the cooperation of federal underwriting with local redlining — is part of the historical record and matters for any serious assessment of how to design universal programs in heterogeneous societies.¹³

The German Sozialstaat, rebuilt under the Adenauer governments of the 1950s on the prior Bismarckian foundation, completed the trio. By the mid-1970s, the OECD economies had converged on a recognizable institutional pattern: universal or near-universal pensions, unemployment insurance, public health systems (with significant cross-country variation in delivery), public education through at least secondary level, and a broadly progressive tax system funding the whole. Gøsta Esping-Andersen’s The Three Worlds of Welfare Capitalism (1990) is the classic comparative typology of how this institutional pattern decomposes into liberal, conservative-corporatist, and social-democratic variants.¹⁴ Esping-Andersen’s typology is contested in detail and has been extended in important ways since — particularly to account for the situation of women in the welfare state, and for the Mediterranean and East Asian variants that did not fit neatly into the original three — but the underlying empirical claim, that there are a small number of stable institutional configurations rather than a single welfare-state model, has held up.¹⁵

Three things about the post-war expansion are relevant to the abundance question. First, the timescale was approximately thirty years from intellectual articulation (Beveridge) to institutional maturity (the mid-1970s). That is fast by historical standards — much faster than the agricultural-to-industrial response — but it required exceptional conditions: the political opening created by total war, a generation of policy intellectuals who had spent the 1930s thinking about exactly these problems, and a sustained period of high productivity growth that made the fiscal arithmetic of expansion comparatively easy.

Second, the expansion was successful in part because it was framed not as compensation for a losing economic position but as the universal floor on which a productive, mobile, and politically stable society would be built. The Beveridge framing of want, disease, ignorance, squalor, and idleness as common adversaries — rather than as the fault of those experiencing them — was deliberately constructed to support a universalist political coalition. The link to /abundance/the-floor is direct: the political economy of universalism is qualitatively different from the political economy of targeted relief, and the difference is much of what makes a settlement durable.

Third, the post-war settlement was built before the productive transformation that would test it. The Beveridge Report preceded the post-war productivity boom; the American GI Bill preceded the suburban-industrial expansion of the 1950s; the German Sozialstaat preceded the Wirtschaftswunder. The institutional response was, for the first and arguably only time in modern economic history, ahead of the productive change rather than behind it.

The argument for the abundance transition follows almost mechanically: it needs a Beveridge-timescale institutional response, not an enclosure-timescale one. A thirty-year window of deliberate institutional construction, beginning before the full productive transformation arrives, is the demonstrated historical model for transitions that do not produce two generations of avoidable damage. The question is whether the political and intellectual conditions that made the post-war expansion possible can be reconstructed, or approximated, in a context that lacks the unifying force of total war.

Historical analogue 3: the digital transition

The third analogue is the one we have lived through and the one whose lessons are most directly transferable. From ARPANET’s first packet-switched communications in 1969 to ubiquitous mobile internet in the mid-2010s — about thirty years of serious diffusion, depending on where one places the start and end points — the digital transition reorganized work, communication, retail, finance, and political coordination in essentially every advanced economy. We have, in this case, both the productive transformation and a sufficient institutional record to evaluate the response.

The earliest empirical puzzle was that the productive transformation took longer than expected to register in measurable productivity. Robert Solow’s much-quoted line from 1987 — “you can see the computer age everywhere except in the productivity statistics” — captures the puzzle in its sharpest form, in a New York Review of Books review of Stephen Cohen and John Zysman’s Manufacturing Matters.¹⁶ The productivity paradox was eventually resolved, in part: by the late 1990s, large gains were appearing in the U.S. data, particularly in sectors that had heavily restructured around information technology, and the apparent earlier absence was reinterpreted as a measurement and lag problem rather than as evidence that the technology did not matter.¹⁷ But the resolution was partial. The productivity gains that did materialize were uneven across sectors, uneven across countries, and — crucially — uneven across workers within the sectors that did benefit.

David Autor, Frank Levy, and Richard Murnane’s 2003 paper in the Quarterly Journal of Economics on the skill content of recent technological change is the canonical statement of the second mechanism.¹⁸ They argued that information technology substitutes for routine cognitive and manual tasks while complementing non-routine analytical and interactive tasks, and that this skill-task complementarity explains much of the polarization of the U.S. wage distribution since the late 1970s. The empirical literature that followed has refined the argument in important ways, but the basic mechanism — that the digital transition raised the returns to a particular kind of cognitive skill and depressed the returns to a particular kind of routine work — is now standard.

The third mechanism is the one most relevant to abundance: winner-take-most platform dynamics. The economics of two-sided markets, network effects, and zero-marginal-cost reproduction systematically generate concentration in digital sectors in a way that the institutional toolkit of twentieth-century antitrust was not designed for and has only recently begun to address. The empirical result is visible in the market structures of search, social networking, mobile operating systems, online retail, and cloud computing. Each of these markets is, globally, dominated by between one and four firms, with margins and persistence that would have been considered prima facie evidence of insufficient competition under any pre-1990 antitrust framework.¹⁹

Three lessons from the digital transition matter for the abundance one. First, large productive gains do materialize, but with substantial lags and substantial measurement difficulty. We should expect the productivity statistics around the abundance transition to be ambiguous for some years even if the underlying productive transformation is real.

Second, the gains are heavily concentrated unless institutional design is explicitly directed against concentration. The digital transition was permitted to proceed for about three decades with essentially no institutional adjustment to the new market structure. The result is the level of platform concentration we now observe, and the political backlash it has produced. The abundance transition, which has stronger underlying tendencies toward concentration than the digital one (because the orchestration layer discussed in /research/economic-orchestration is itself a winner-take-most market structure), cannot afford a similar three-decade institutional vacation.

Third, the political response to concentrated digital gains has been slow, fragmented, and largely defensive. Antitrust enforcement against the dominant platforms has accelerated in the EU and is beginning in the U.S., but the substantive response is occurring twenty to thirty years after the relevant market structures consolidated. That is the institutional-response timeline we should expect to inherit by default, and the timeline the abundance transition has to beat.

The political economy of the transition

The barriers to a successful abundance transition are not primarily technical, fiscal, or analytical. They are political, in the specific sense that they involve the strategic behavior of organized groups whose position is favored by the existing institutional configuration. Three threads of political-economy literature are particularly useful for thinking about what those barriers look like and how, in principle, they have been overcome in past transitions.

Veto points and incumbent capture

Mancur Olson’s The Rise and Decline of Nations (1982) is the foundational reference.²⁰ Olson argued that long periods of political stability allow distributional coalitions — industry associations, professional licensing bodies, organized labor in protected sectors, regional political machines — to accumulate, each of which has strong incentives to lobby for transfers and protections that are individually small but collectively growth-suppressing. Olson’s empirical claim about the long-run consequences (that countries with more accumulated coalitions grow more slowly) is contested, but the underlying mechanism — that organized incumbents capture institutional veto points and that the transition costs of dislodging them are typically borne by diffuse and unorganized populations — is well-established in the political-science literature.

The contemporary illustrations are easy to enumerate. The political battles between ride-hailing platforms and taxi medallion holders in the early 2010s involved, in city after city, a small and well-organized incumbent group (medallion owners and the financial institutions that had financed them) defending a position threatened by a structurally cheaper service. The fact that the incumbent position eventually lost in most cities is less interesting than the fact that the political conflict took roughly five years to resolve in each jurisdiction and required coordinated political mobilization by the entrant. Multiply that pattern across the sectors in which automation is now plausibly competitive, and the cumulative deployment delay is structural.

A second illustration: the slow rollout of automation in agriculture in the United States is partly a function of the availability of low-wage seasonal labor under the H-2A visa program and the broader structure of immigration policy. The labor-supply margin makes the labor-saving capital margin relatively less attractive at the prices observed; immigration policy is in this sense an industrial policy, with consequences for the trajectory of capital deployment that are largely invisible in the public debate.²¹ The general point is that the political economy of incumbent protection operates through many channels, most of them not obviously labelled as such.

Geography of effects

The second political-economy thread concerns the spatial distribution of transition costs. The literature on the U.S. industrial heartland — the Rust Belt — and on the French banlieues (and the analogous deindustrialized regions of the UK Midlands, the Ruhr, northeast Italy, and parts of post-Soviet Central Europe) is extensive and increasingly well-quantified.

The single most influential paper on this question is David Autor, David Dorn, and Gordon Hanson’s “The China Syndrome,” published in the American Economic Review in 2013.²² Autor, Dorn, and Hanson exploited variation across U.S. commuting zones in their exposure to import competition from China after China’s WTO accession in 2001, and showed that the local labor-market effects of the trade shock were large, persistent, and concentrated in ways that previous trade theory had predicted would be diffused away by labor mobility. The follow-up paper by Acemoglu, Autor, Dorn, Hanson, and Price in 2016 estimated total manufacturing-employment losses attributable to the China shock at roughly two to two and a half million U.S. jobs through the early 2010s.²³

The theoretical significance of these papers is not the magnitude of the China shock specifically; it is that they overturned a generation of trade-theory presumption that adjustment costs are second-order and that displaced workers are quickly re-employed elsewhere. The empirical evidence is that adjustment is slow, geographically concentrated, and politically consequential. The 2016 U.S. presidential election, the 2016 Brexit referendum, the 2017 French presidential first-round geography, and the 2018 Italian election are all readable in part as the political downstream of these adjustment costs. Any abundance transition that produces displacement on a scale comparable to or larger than the China shock — and many credible scenarios do — will face a similar political downstream unless the geographic and temporal distribution of costs is actively managed.

Cross-border coordination

The third thread is the hardest. Even a domestic political coalition successfully constructed around an abundance settlement — a sovereign wealth fund, a dividend, a universal services basket — runs into the constraint that capital is mobile across jurisdictions while institutions are not. An automation tax, a robot tax, a profits-on-AI tax, or any analogous instrument imposed in one jurisdiction without coordination will see its base shrink as activity migrates to non-taxing jurisdictions. This is the same logic that produced the international tax-competition race-to-the-bottom of the 1990s and 2000s on corporate income taxation.

The closest precedent for the kind of coordination an abundance transition would require is the OECD/G20 process on a global minimum corporate tax, formally announced in October 2021 and progressively implemented through 2024.²⁴ Pillar Two establishes a fifteen percent minimum effective tax rate on the profits of large multinational enterprises, with a top-up mechanism that allows source countries to collect the difference if profits have been booked in lower-tax jurisdictions. Emmanuel Saez and Gabriel Zucman’s 2024 work on the rollout documents both the genuine progress (over one hundred and forty jurisdictions agreed to the framework; implementing legislation has passed in the EU, UK, Japan, South Korea, and several others) and the substantial limitations (incomplete U.S. implementation, carve-outs for substance-based activities, uncertainty about enforcement against firms in non-participating jurisdictions).²⁵

The lesson is mixed. On one hand, Pillar Two demonstrates that international tax coordination at scale is possible. On the other, the timescale matters: serious negotiations began in 2013 with the OECD’s BEPS process; the framework was agreed in 2021; implementation is still in progress in 2026. That is roughly a decade from start to substantive operation, against a capability-deployment timeline for the underlying productive change of three to five years. The mismatch between coordination timescale and deployment timescale is structural, and it is one of the central reasons for emphasizing — as we do throughout this content section — the importance of beginning institutional construction before the deployment has fully arrived rather than after.

Governance: who decides

A persistent open question, surfaced in the /manifesto and not resolved there, is how decisions about a system whose effects are global ought to be made when the political institutions with democratic legitimacy operate at the national scale. The abundance transition raises this question in particularly acute form, because the decisions at issue — about the orchestration layer, about the basis for any global dividend, about the coordination of automation taxation — are intrinsically transnational and intrinsically high-stakes.

There are working models, but they work within narrow domains. The Bank for International Settlements is the institutional home of central bank coordination, with a record of producing internationally adopted standards (the Basel Capital Accords being the most consequential) over multi-decade timescales.²⁶ The IETF and the broader internet standards community produces, through a rough-consensus and running-code process, the protocols on which global digital communication runs. ICANN governs the domain name system. The IMF and the World Bank have effective convening power on macroeconomic policy. The IPCC produces scientific consensus on climate. The WHO coordinates pandemic response, with mixed results.

The pattern across the working cases is that legitimacy is built around technical narrowness: the institution does one thing, its expertise is concentrated and specifiable, and the political stakes are kept (where possible) below the threshold at which national sovereignty arguments override the gains from coordination. The instruments of an abundance transition do not fit that pattern. A global automation tax, a global dividend, an orchestration layer for productive coordination — these are decisions with first-order distributive consequences, of the kind that under the Westphalian settlement only national governments are presumed to have the legitimacy to make.

Robert Dahl’s Democracy and Its Critics (1989) is the most careful treatment of the legitimacy question at scales above the nation-state.²⁷ Dahl argued, persuasively in our reading, that democratic legitimacy at the supranational level cannot be borrowed wholesale from the national level; it must be constructed, with explicit attention to the conditions — equal voting weight, effective contestation, access to information, equal opportunity to set the agenda — that confer legitimacy on national-level democratic institutions. The European Union is the running natural experiment. Its institutional progress on the technical side of integration has been substantial; its progress on what Dahl would call the conditions of democratic legitimacy has been slower, contested, and reversible.

The argument that follows is uncomfortable but, we think, structurally correct: the hardest part of the abundance transition is not designing the technical mechanisms — the analytical work in /research/economic-orchestration and /abundance/economic-mechanism is hard but tractable — but constructing the legitimacy of the institutions that will operate them. Legitimacy infrastructure has to be built as a precondition for institutional authority, not retrofitted as post-hoc cleanup once the technical institutions are already operating and unaccountable. The relevant link to /safety/principles and /safety/responsible-development-policy is direct: the same argument that applies to the legitimacy of operators of advanced AI systems applies, with the same force, to operators of an orchestration layer or a global dividend mechanism.

Risks specific to the abundance transition

Beyond the generic political-economy risks of any large institutional transition, the abundance trajectory has specific failure modes that are worth naming explicitly. Each of the following deserves serious analytical attention; none of them is hypothetical.

Risk 1 — Concentration of the orchestration layer

The orchestration layer described in /research/economic-orchestration and /abundance/economic-mechanism — the set of computational and informational systems that would coordinate productive activity at scale — is, by its technical structure, prone to concentration. The economics are similar to those of cloud infrastructure or platform markets: large fixed costs, near-zero marginal cost of additional users, network effects across users and complementors, and substantial returns to incumbency in training data and operational experience. Without explicit institutional design directed against concentration, the default outcome is that one or a small number of operators end up with effective control of the layer.

Concentration of the orchestration layer concentrates authority. Concentration of the dividend or services floor described in /abundance/the-floor concentrates dependency. Both are problems; the combination of the two — a population whose material standard depends on a floor mechanism, operating on top of an orchestration layer controlled by a small number of firms — is a configuration that the constitutional traditions of liberal democracies have no working precedent for handling. The institutional question is how to keep the orchestration layer plural, inspectable, and contestable, without sacrificing the coordination gains that justify its construction in the first place.

Risk 2 — Capture of sovereign-wealth-style funds

The institutional analysis in /abundance/precedents draws a sharp contrast between Norway’s Government Pension Fund Global and Mongolia’s Human Development Fund. The Norwegian fund has, since its first deposits in 1996, accumulated assets in excess of a trillion U.S. dollars, operates under a published fiscal rule (the spending limit is roughly the expected real return, currently set at three percent), and has maintained transparency, professional management, and political insulation across more than two decades and multiple changes of government.²⁸ The Mongolian fund, established in 2009 to distribute mining rents to the population, was structurally captured within three years: its dividend payments became electoral instruments, its borrowing-against-future-revenues mechanism produced a fiscal crisis, and the fund as originally conceived effectively ceased to exist by 2012.²⁹

The relevant variable is not the underlying resource base; it is the institutional design. Norway began with strong rule-of-law institutions, an established central bank with operational independence, and a political culture of consensus around resource management. Mongolia did not. The lesson for the abundance transition is that a sovereign-wealth-style fund as the asset base for a dividend is only as good as the institutional configuration in which it sits. Building the fund without first or simultaneously building the institutional configuration is, on the historical evidence, the failure mode.

Risk 3 — Incidence drift over decades

A third risk, less dramatic but potentially equally consequential over the relevant timescale, is that programs designed to be progressive in incidence drift toward regressivity as their constituencies and administrative practices evolve. The U.S. mortgage interest deduction is the canonical case: introduced essentially as an incidental feature of the 1913 income tax, it became, over the course of the twentieth century, one of the largest tax expenditures in the federal budget, with benefits concentrated overwhelmingly in the top quartile of the income distribution because the deduction is worth more to households with higher marginal tax rates and larger mortgages.³⁰ The political coalition defending the deduction is now sufficiently entrenched that even modest reform attempts have been politically costly.

Any abundance-transition instrument with a long expected life — a dividend, a sovereign fund, a universal services basket — is exposed to the same risk. Progressivity in design does not guarantee progressivity in incidence twenty or forty years later, and the institutional and constitutional protections required to maintain progressivity over the relevant timescale are not standard features of the policy toolkit. Building them in from the start is harder than it looks; retrofitting them once incidence has drifted is harder still.

Risk 4 — Workforce skill atrophy

The post-labor question is treated in detail in /abundance/post-labor; we will not duplicate the argument here. The transition-specific risk is that during the period in which automation is structurally displacing labor but a stable abundance settlement has not yet been institutionalized, the economy may pass through an extended phase in which large segments of the workforce are out of stable employment without yet having access to the institutional supports that would maintain skills, social participation, and the kinds of meaning-bearing activity that paid work has historically provided. The empirical literature on long-term unemployment is unambiguous about the human and economic costs of even a few years out of work; a transition that produces a decade or more of structural underemployment for any large cohort is unlikely to be politically endurable, regardless of how the eventual settlement is designed.

Risk 5 — Geopolitical asymmetry

Countries that adopt abundance institutions before their international peers may face a transitional period of strategic disadvantage. An automation tax, a slower deployment of labor-displacing capital under a managed-transition framework, or a heavy domestic investment in floor-provision institutions all carry, in the short term, real opportunity costs relative to a jurisdiction that does not adopt them. The historical analogy is the early-twentieth-century debate about labor standards: countries that imposed eight-hour days, child labor restrictions, and minimum wages on themselves did so against a real fear that they would lose ground to countries that did not. They did, on balance, lose some ground; they also produced more politically stable societies, and the institutional standards eventually diffused. The transitional period was nonetheless costly, and the political coalitions that bore the cost had to be sustained through it.

The abundance transition is exposed to the same logic, with an additional complication: capability for automation may itself be a strategic asset, in ways that ordinary industrial competition was not. A jurisdiction that constrains its own automation deployment to manage transition costs may find itself at a security disadvantage that is hard to recover from. This is one of the considerations that makes cross-border coordination, as discussed above, structurally important rather than optional.

Risk 6 — Backlash and reversal

The final transition-specific risk is the most general: Polanyi’s double movement applies to abundance settlements as much as to any other institutional configuration, and the historical record on the durability of welfare-state expansions is mixed. The Reagan-Thatcher rollback of the 1980s, discussed above, is the canonical post-war reversal; it succeeded in part because the political coalition defending the post-war settlement had become identified with specific incumbent interests (unionized industrial labor, public-sector employees, recipients of specific transfer programs) rather than with the universal logic of the original Beveridge framing.

An abundance settlement that becomes politically identified with specific beneficiary groups — rather than with the universal logic of a productive transformation that benefits the entire society — is exposed to the same kind of reversal. The argument for universalism in /abundance/the-floor is, in part, an argument about what kinds of settlement are politically durable across the kinds of political cycles that are likely to occur over a thirty-year transition.

Pathways that have worked

Not all of the historical record on resource-windfall and major-transition management is failure. There are working examples; they are instructive, and they share institutional features that are recognizable and, in principle, reproducible.

Norway’s resource governance

The Norwegian model, treated in more detail in /abundance/precedents, deserves a second mention here because of what it demonstrates about institutional design under transition. The Government Pension Fund Global is operationally independent, rule-bound (the spending rule is published and binding), broadly diversified (the fund holds equity, fixed income, and real estate across global markets), transparent (holdings are published; an ethics council operates publicly), and depoliticized (the management is delegated to Norges Bank Investment Management, with parliamentary oversight rather than parliamentary direction). Cappelen and Mjøset’s 2009 analysis of the Norwegian model documents in detail how each of these features was a deliberate institutional choice rather than a cultural inheritance.³¹

The model is portable in principle. The institutional preconditions — operational independence of the financial authority, a political culture that accepts rule-bound rather than discretionary management of large pools of public assets, a public commitment to transparency that makes deviation politically costly — are present in many but not all candidate jurisdictions. Where they are present, the Norwegian model is a working blueprint. Where they are absent, building them is a precondition for any analogous fund to function.

Singapore’s CPF and HDB housing

A second working model, structurally different from Norway’s, is the coupled system of the Singaporean Central Provident Fund and the Housing and Development Board.³² The CPF is a mandatory savings system with high contribution rates (employer plus employee contributions in the high tens of percent of wages, depending on age and income); the HDB is a public housing system that has, since the 1960s, built and sold housing on long leases to a substantial majority of Singaporean households. The two are coupled: CPF balances can be used for HDB purchases, and the resulting structure is one in which roughly eighty percent of Singaporeans live in HDB-built housing on a long-lease basis and accumulate housing equity through their working lives.

The Singaporean model is not a universalist welfare state in the European sense; it does many things that European welfare states do, but through a different institutional channel — savings and asset accumulation rather than current transfers. For our purposes, the lesson is that the institutional form of an abundance settlement is not unique; there are multiple institutional configurations that can deliver broadly similar outcomes (universal access to housing, retirement security, basic services), and the right configuration for a given jurisdiction depends on its prior institutional inheritance and political culture.

Mauritius and Botswana on resource curse avoidance

The third set of pathways comes from the comparative political economy of resource-rich developing countries. Most resource-rich developing countries underperform their non-resource-rich peers on long-run growth, the stylized fact that motivates the “resource curse” literature. A small number do not, and the question is what they did differently.

Acemoglu, Johnson, and Robinson’s 2003 analysis of Botswana is the most-cited single case study.³³ Botswana, at independence in 1966, was among the poorest countries in the world. The discovery of major diamond deposits shortly after independence could plausibly have produced a Mongolian-style outcome. Instead, Botswana over the following four decades produced one of the highest sustained growth rates in the developing world, on the basis of an institutional configuration that included strong property rights, a politically credible commitment to investing resource rents in infrastructure and human capital, a partnership with De Beers that — for all its complications — produced fiscal terms favorable to the state, and a political culture that maintained competitive elections and constitutional government across the entire post-independence period.

Mauritius is the parallel case in a different sector — sugar and then export-oriented manufacturing rather than minerals. The institutional features that account for the divergence from regional peers are similar: stable property rights, competitive politics, deliberate investment in education, and an effective state capacity to convert sectoral rents into broad development.

The lesson, again, is that institutional design matters more than the underlying productive base. A jurisdiction that builds the institutional preconditions can capture the gains of an abundance transition broadly; a jurisdiction that does not will see those gains captured narrowly even if the underlying productive transformation is identical.

A 30-year scenario

It is useful to sketch, explicitly, the range of possible outcomes over the thirty-year window in which most of the substantive abundance transition is plausibly going to occur. We see three broad branches.

Branch 1 — Optimistic

In the optimistic branch, the institutional response described above arrives roughly on the Beveridge timescale. An orchestration layer is built with explicit institutional design directed against concentration; sovereign wealth or equivalent dividend mechanisms proliferate, with rule-bound governance modeled on Norway; universal basic services frameworks scale across the OECD and into middle-income countries; cross-border coordination on automation taxation succeeds, on a Pillar Two model but operating faster. Abundance becomes a structural rather than a marginal feature of advanced economies between roughly 2030 and 2050; the global income distribution begins to compress meaningfully by the 2050s.

The leading indicators of the optimistic branch, observable well before its outcome is settled, would include: stabilizing or rising labor share of national income across major OECD economies; visible OECD-level coordination on AI and automation taxation, with implementation timelines in single-digit years rather than the decade-plus of Pillar Two; a wave of new sovereign wealth or analogous dividend mechanisms with rule-bound governance, in jurisdictions beyond the existing handful of resource-rich exceptions; falling rather than rising geographic concentration of displacement effects within major economies; and a measurable shift in the sectoral composition of public investment toward floor-provision institutions.

Branch 2 — Muddled

The muddled branch is the modal outcome on the historical base rate. Rich countries, individually, build domestic versions of an abundance settlement; cross-border coordination stalls; large parts of the global south are excluded from the institutional response. The intra-OECD income distribution stabilizes; the global income distribution widens as the orchestration and floor mechanisms remain effectively national. The political consequences are absorbed within rich-country institutions, which are robust enough to handle them; the political consequences in middle-income and low-income countries — particularly those whose development trajectories had been premised on labor-intensive export manufacturing — are more severe and produce migration, instability, and political pressure that propagates back into rich-country politics.

The leading indicators of the muddled branch include: domestic OECD adoption of floor mechanisms succeeds in three or four advanced economies by the early 2030s, but cross-border coordination on the corresponding revenue base does not; a global cross-jurisdictional gap between AI capability deployment and governance widens; sovereign wealth funds proliferate within OECD jurisdictions but not in middle-income ones; and the global Gini coefficient begins to widen after a multi-decade narrowing.

Branch 3 — Regressive

The regressive branch is the one in which concentration wins. The orchestration layer ends up controlled by a small number of operators, with capture across multiple jurisdictions. Sovereign wealth analogues, where they exist, are repurposed for non-dividend uses — strategic industrial policy, geopolitical positioning, security investment — rather than for broad transfers. Floor mechanisms are built in some places but politically rolled back in others. Abundance becomes a private-club good, accessible to those who hold the right assets in the right jurisdictions and unavailable to others. The political coalitions explicitly committed to rolling back universalist provisions win at scale, in a Reagan-Thatcher pattern but applied to a productive base that is structurally less able to deliver mass employment than the one those reforms operated on.

The leading indicators of the regressive branch include: market concentration in orchestration-relevant sectors continuing to accelerate rather than stabilize; explicit capture of sovereign wealth or analogous funds for non-dividend purposes; political coalitions running on rollback platforms winning major elections in major jurisdictions; visible erosion of antitrust capacity in the major regulatory authorities; and a divergence between productivity statistics and median wages widening rather than narrowing.

What we would know by 2032: by the early 2030s, most of these leading indicators will have resolved enough to make the branch identification reasonably confident. We would know whether OECD-level coordination on AI/automation taxation was on a single-digit-year track or had stalled; whether sovereign wealth or analogous dividend mechanisms were proliferating beyond the existing exceptions; whether labor share was stabilizing or continuing to decline; whether the geography of displacement effects was being managed or was producing the same political downstream the China shock did; whether antitrust capacity in the major jurisdictions was strengthening or eroding. None of these is hard to measure; the indicators are observable in standard public data. The question is whether the institutional response is built before, or after, the branch is settled.

Where to read further

For the historical-precedent baseline against which any abundance settlement should be evaluated, see /abundance/precedents. For the underlying analysis of how an orchestration layer and dividend mechanism could be constructed, see /abundance/economic-mechanism. For the institutional logic behind universalist floor provision, see /abundance/the-floor. The post-labor question is treated in /abundance/post-labor. For the broader case for why this is the appropriate frame for thinking about advanced AI development, see /manifesto. For the safety-side counterpart of the institutional argument made here, see /safety/responsible-development-policy.