Predistribution

Predistribution interventions build assets and income at the bottom and middle of the distribution rather than (or in addition to) extracting from the top. The label is Jacob Hacker’s (2011); the underlying idea is older, going back to Thomas Paine’s Agrarian Justice (1797) and the citizen-endowment proposals of Ackerman & Alstott (1999). The political appeal is that these interventions are framed as universal entitlements rather than as taxes on the rich, which can build broader coalitions and create durable constituencies.

The empirical case is uneven across instruments. Cash transfers (including conditional, unconditional, and universal designs) have a large and well-replicated evidence base on poverty reduction and a smaller but growing evidence base on long-run developmental effects (Hoynes et al. 2016 on EITC; Bastian & Michelmore 2018; Akee et al. 2010 on Cherokee dividends). Baby bonds — government-funded endowments paid at age 18 or 21 — are theoretically attractive and have been adopted at small scale in Connecticut and the U.K., but lack long-run outcome data. Sovereign wealth funds with citizen dividends (Alaska Permanent Fund, Norway’s GPFG) are the longest-running examples and show strong evidence on poverty reduction and political popularity.

A separate strand is employee ownership (ESOPs, worker cooperatives, employee ownership trusts). The U.S. ESOP system covers ~14 million workers; evidence (NCEO 2024; Kruse et al. 2010) shows higher wages, lower turnover, and modest wealth-building effects, though selection bias is hard to rule out. The U.K. Employee Ownership Trust regime (2014) has grown rapidly and provides cleaner identification.

Instrument comparison

Instrument	Funding source	Evidence base	Wealth-concentration effect
Universal Basic Income (UBI)	General taxation	Pilots (Finland, Kenya, Stockton)	Modest direct; large indirect (bargaining power)
Negative income tax / refundable credit	General taxation	Strong (EITC, CTC)	Moderate
Baby bonds	General taxation	Limited (early-stage programs)	Potentially large long-run
Sovereign wealth fund + dividend	Resource rents / SWF returns	Strong (Alaska, Norway)	Moderate; highly durable
Employee Stock Ownership Plans	Tax-advantaged corporate transactions	Moderate (NCEO, Kruse)	Moderate; concentrated in mid-firms
Employee Ownership Trusts (U.K.)	Tax-advantaged sales	Emerging (post-2014)	Moderate
Public banking / postal banking	State capital	Limited in U.S.; strong in EU/Asia	Modest

The Hacker/Ackerman framing and the politics of universalism

The term “predistribution” was introduced by political scientist Jacob Hacker in a 2011 Policy Network essay and developed in his subsequent work. Hacker’s argument is that the welfare-state debate of the 20th century focused too heavily on post-distribution — taxes-and-transfers that redistribute income after market processes have already produced unequal outcomes — and too little on pre-distribution, the institutional design of markets themselves (labor-market rules, financial regulation, antitrust, education, asset-building) that determines what the market generates before any tax or transfer is applied.

The intellectual foundation for the asset-building strand is older. Thomas Paine’s Agrarian Justice (1797) proposed a stake — £15 paid to every citizen at age 21 from a tax on landed inheritance — as compensation for the privatization of common land. Bruce Ackerman and Anne Alstott’s The Stakeholder Society (Yale, 1999) updated the proposal as an $80,000 grant to every American at age 21, financed by a 2% wealth tax. The contemporary baby-bonds proposals (covered below) are direct descendants of this lineage.

The political appeal of predistribution is universalism. Means-tested redistribution programs (food stamps, Medicaid, SNAP) build narrow constituencies of recipients who can be politically isolated and program-cut. Universal programs (Social Security, Medicare, the proposed UBI and baby bonds) build broad constituencies that include the politically powerful middle class and become much harder to dismantle once enacted. The empirical track record supports this: in the U.S., universal programs have consistently expanded over time while means-tested programs have been repeatedly contracted (Wilson 1987; contemporary work by Theda Skocpol, Suzanne Mettler).

The trade-off is that universal programs are more expensive per dollar of poverty-reduction effect, because they pay benefits to non-poor recipients along with poor ones. The political response to this critique is that the durability of universal programs — their capacity to persist across hostile administrations — more than offsets the per-dollar inefficiency, because the expected poverty reduction over time is what matters and means-tested programs face high probability of cuts.

The interventions on this page split between asset-building (baby bonds, ESOPs, sovereign-wealth dividends) and income-floor (UBI, negative income tax, refundable tax credits). Both work on the predistribution margin but address different dimensions of the wealth-and-income gap.

UBI pilot evidence

The empirical record on Universal Basic Income consists of several pilot programs of varying scale, design, and evaluation rigor. The strongest evidence comes from four sources.

Finland (2017–2018) — Kela randomized 2,000 unemployed adults to receive €560/month for two years vs. control. Headline results:

• Employment effect: +6 days of employment per recipient over the reference period — small, but the treatment group worked slightly more than control, contradicting the “UBI reduces work” prediction.
• Wellbeing: meaningfully positive on perceived financial security, mental health, trust in institutions, and confidence in the future.
• Interpretation complicated by a 2018 unemployment-benefits “activation model” change that affected the control group.

Stockton, CA SEED (2019–2021) — 125 randomly selected residents received $500/month for 24 months. Results from the first year:

• Full-time employment among recipients rose from 28% to 40% in the first year, vs. 5 percentage points in the control group.
• Increased financial stability, liquidity, and mental health.
• Spending was concentrated in essentials (food, home goods, utilities, gas); only 1% on tobacco and alcohol.
• Mayor Michael Tubbs subsequently launched Mayors for a Guaranteed Income (now Center for Guaranteed Income Research) to scale the model; ~25 cities now run their own pilots.

GiveDirectly Kenya (2017–present) — Long-running RCT in Kenya providing $22.50/month to ~6,000 households across hundreds of villages for 12 years (a true universal long-horizon design at sub-national scale). Interim findings show meaningful poverty reduction, asset accumulation, no employment-effort decline, and no measurable inflation in recipient communities. The program is the longest-running large-scale UBI evaluation in the world and is ongoing.

Mincome (Manitoba, Canada, 1974–1979) — Historical reference point. Provided guaranteed income to all residents of Dauphin, Manitoba for 4 years. Forget 2011 found reduced hospitalization rates, better mental health, more high-school completion, and modest labor-supply reductions concentrated in mothers (more time with infants) and teenagers (more time in school) — both arguably positive social effects.

The cross-pilot pattern is consistent: modest unconditional cash transfers do not measurably reduce work effort, do measurably improve health and financial stability, and produce essentials-focused spending rather than the “wasted on alcohol” pattern critics predict. The evidence is strongest for poverty reduction at the bottom and weakest for the broader wealth-concentration effect, because pilot scales have been too small to test general-equilibrium and political-economy mechanisms.

Baby bonds: from Hamilton proposal to state implementation

Baby bonds are the most concrete recent expression of the asset-building strand of predistribution. The contemporary U.S. design was developed by Darrick Hamilton (The New School) and William Darity Jr. (Duke) over a decade of policy work, with the seminal proposal calling for $50,000–$60,000 federal endowments at birth, scaled to family wealth at birth, with proceeds available at age 18 for asset-building purposes (education, business start, home purchase, retirement saving).

The federal proposal has not advanced (Senator Cory Booker’s American Opportunity Accounts Act of 2018, 2021, and 2023 reintroductions all stalled), but state-level implementation has begun:

Jurisdiction	Year established	Initial endowment	Eligibility
Connecticut	2023 (in force)	$3,200/child	All children covered by HUSKY Medicaid (poverty-eligible)
California	2022 (HOPE pilot)	Up to ~$8,000	Children orphaned by COVID + foster youth
Vermont (pilot)	2024 (in development)	TBD	Pilot scope
District of Columbia	2021 (Child Wealth Building Act)	Up to $25,000	Family-income-tested
Washington State	2022 (Future Fund proposal)	Pending	Pending
U.K. Child Trust Fund	2002 (closed to new entrants 2011)	£250–500	Universal cohort 2002–2011; matured 2020+

The Connecticut program, in force since July 2023, is the first U.S. state-level baby bonds program and provides $3,200 per child born to a HUSKY-covered (Medicaid) family, accumulating with investment returns until age 18 and then available for housing, education, business start-up, or retirement savings. As of 2025, more than 13,000 children had been enrolled. The first beneficiaries reach age 18 around 2041, when the first long-run effects will be measurable.

The U.K. Child Trust Fund (2002–2011) is the closest existing reference point. The Labour government provided £250–500 vouchers to every newborn, with additional contributions allowed up to a cap and tax-free growth until age 18. The program was closed to new entrants in 2011 by the Coalition government, but the original cohort has been maturing since 2020. Early evaluations show modest asset effects and substantial uptake heterogeneity by parental engagement.

The structural critique of baby bonds is that the asset accumulates over 18 years and produces benefits at age 18 — a long horizon that limits political payoff and depends on durability of the program design across multiple administrations. The empirical question of whether the asset-at-18 design actually produces the intended downstream outcomes (reduced racial wealth gap, expanded entrepreneurship, increased home ownership) will not be answerable for another 15–20 years.

Sovereign wealth funds: Alaska, Norway, and proposed U.S. designs

The longest-running and best-evaluated form of universal asset distribution is the sovereign wealth fund with citizen dividends. Two cases dominate the empirical record.

Alaska Permanent Fund Dividend (PFD, 1982–present) — Established by constitutional amendment in 1976; first dividends paid 1982. Fund principal is currently ~$80B; annual dividend has ranged from ~$300 to ~$3,200 per resident depending on returns and political adjustment. The empirical evidence on the PFD (Berman 2024, Poverty & Public Policy; Jones & Marinescu 2022, AEJ: Policy):

• Poverty effects: substantial and durable. A 2024 paper found the PFD reduced the number of Alaskans below the poverty threshold by 20–40%; rural Alaska Native poverty rates fell from ~28% to ~22%.
• Labor effects: essentially none. Jones & Marinescu document zero effect on aggregate employment and a 1.8 percentage point increase in part-time work — directly contradicting the prediction that universal cash transfers would reduce labor supply.
• Health effects: positive for children. Strong evidence of reduced obesity and reduced child abuse/neglect among recipients with young children.
• Political durability: extraordinarily high. The PFD has survived 40+ years across all party administrations; recent attempts to reduce or eliminate it have produced sharp political backlash. This is the cleanest evidence for the universalism-builds-durability argument.

Norway Government Pension Fund Global (GPFG, 1990–present) — World’s largest sovereign wealth fund (~$1.7T as of 2024); funded by oil revenues invested in global equities and bonds. Norway does not pay direct citizen dividends (proceeds fund the general budget under a 4% spending rule), but the fund effectively underwrites Norway’s universal welfare state and serves as intergenerational savings against the day oil revenues run out. The Norwegian model has been studied extensively as a template for resource-rich countries managing extractive wealth without producing oligarchic capture.

Proposed U.S. designs include various “American Stake” or “Public Wealth Fund” proposals from across the political spectrum:

• Bruenig (People’s Policy Project) “Social Wealth Fund” — Federal SWF capitalized by progressive taxation (estate tax, financial transaction tax) and corporate share contributions, paying universal dividends.
• Sherraden / Goldberg “American Opportunity Account” — Asset-building accounts model.
• Conservative variants (“Federal Reserve dividend” proposals; some Singapore-style designs) emphasize financial-stability uses of SWF proceeds.

None has substantively advanced, but the policy infrastructure (the proven Alaska model + the proven Norwegian management framework) means that the design questions are well-understood when political opportunity emerges.

Employee ownership: ESOPs, EOTs, and Mondragon

Employee ownership is the third major predistribution lever — broadening capital ownership rather than redistributing capital income. Three institutional designs dominate.

U.S. Employee Stock Ownership Plans (ESOPs) — Established by ERISA 1974; tax-favored mechanism for company sales to employee-owned trusts. The U.S. has ~6,500 ESOP companies covering ~14M workers, predominantly in middle-market manufacturing and services. NCEO 2024 statistics show:

• ESOP companies pay 5–12% higher wages on average than comparable non-ESOP firms.
• Lower turnover (~50% of comparable firms in many sectors).
• Average ESOP retirement account balance ~2.5× the average 401(k) balance.
• Bankruptcy rates ~50% of comparable firms.

The selection-bias caveat is real (firms that adopt ESOPs may differ from those that don’t on unobserved dimensions), but multiple matched-firm studies and the U.K. EOT evidence (cleaner identification) point in the same direction.

U.K. Employee Ownership Trusts (EOTs) — Established by Finance Act 2014. Owner sells controlling interest to trust held for benefit of employees; capital gains are tax-exempt for the seller, providing a clean tax incentive for employee-ownership conversions. EOT adoption has grown rapidly: ~1,800 EOTs in 2024, doubling roughly every 18 months. The U.K. design is cleaner than U.S. ESOPs because it doesn’t require the leveraged-trust structure that complicates ESOP financial engineering.

Mondragon Cooperatives (Spain, 1956–present) — World’s largest worker cooperative federation; ~80,000 worker-owners across 95+ cooperatives in manufacturing, retail (Eroski supermarkets), finance (Caja Laboral / Laboral Kutxa), and education. Mondragon survived the 2009–2012 Spanish crisis with substantially fewer layoffs than comparable conventional firms (cooperative members took pay cuts and inter-cooperative transfers to preserve employment). The model has been hard to replicate at scale outside the Basque Country, but informs the worker-cooperative federations in U.S., U.K., Italy (LegaCoop), and elsewhere.

The shared pattern across employee-ownership models is that they produce measurably better worker outcomes (wages, turnover, financial security) and equivalent or better firm performance but face high adoption costs — the conversion from conventional ownership to employee ownership requires either a willing seller, tax incentives, or both. The U.K. EOT example shows that the right tax incentive structure can produce rapid uptake; the U.S. ESOP regime has more political support across parties (rare in inequality policy) than is widely appreciated and could be substantially expanded with relatively narrow tax-code changes.

For the wealth-concentration agenda, employee ownership operates on a different margin than the cash-transfer interventions discussed above — it builds productive capital in worker hands rather than transferring consumption flows. The two are complementary, not substitutes, and a coherent predistribution agenda includes both.

Objections it must clear

Predistribution — building assets and bargaining power before the market distributes — is aimed squarely at the desert and entitlement objection and the mechanical-not-extractive objection. The entitlement story assumes a fair starting point and objects to redistributing the result; predistribution accepts the premise and attacks the assumption, equalizing starting endowments rather than re-slicing outcomes. Baby bonds and child asset accounts do not tax the winner after the race — they change who lines up where at the start, which is precisely the “justly acquired” premise the entitlement defense needs and the evidence does not supply. Against the cure-is-worse objection that transfers breed dependency, asset-based predistribution is the form most resistant to it: it funds ownership, education, and enterprise rather than consumption.

Distributional effects: gender and race/ethnicity

Predistribution is the one lever explicitly engineered around the racial wealth gap. The baby-bonds proposal of Hamilton and Darity (2010) sizes a publicly funded, progressively scaled trust to each child’s family wealth, so children born into low-wealth families — disproportionately Black and Hispanic — receive the largest endowments. Zewde’s (2020) simulation finds that a universal, wealth-progressive baby bond would cut the median Black-white young-adult wealth gap dramatically — from a roughly tenfold ratio toward near-parity — precisely because the program’s size is inversely keyed to existing wealth. This baby-bonds design is race-neutral in eligibility but race-equalizing in effect, because the targeting variable (family wealth) is itself so racially skewed. It contrasts with the explicitly lineage-targeted reparations case of Darity and Mullen’s From Here to Equality (2020), which conditions eligibility on documented descent from enslaved ancestors rather than on wealth.

The gender dimension is strongest in the cash-transfer wing of predistribution. Unconditional income and asset support raises women’s intra-household bargaining power and buffers the care-driven earnings penalty that depresses women’s lifetime wealth; the most robust evidence comes from multi-country transfer experiments showing durable gains in assets and autonomy for women recipients (Banerjee et al. 2015), though that evidence is global rather than U.S.-specific. Design choices — whether endowments are individual or household, and to whom transfers are paid — determine whether predistribution narrows the gender wealth gap or merely relocates it. The sharpest test is whether an endowment reaches the compounded bottom, the households where low family wealth, gender, and a marginalized race, caste, or indigenous status coincide, or stops at the margins that single-axis targeting already covers, the position the comprehensive strategy treats as the real target. Disability belongs in the same frame, since a household supporting a disabled member faces higher costs and thinner earnings at once and is easily missed by a wealth-only trigger.