The Public Pension System Is a Ponzi

The PUNZI Papers Essay Six · The Public Pension System Is a Ponzi

This essay is part of The PUNZI Papers, a series of essays explaining the reason for our project, PUNZI WARS, and the potential we think it points to.

Visit us: https://punzi.xyz Follow us on X: @PUNZIwars

Advisory: PUNZI WARS is a game and an on-chain research experiment. It is not an investment. Participate at your own risk.

A public pension pays today’s retirees out of today’s workers’ contributions, holds almost nothing in reserve, and produces none of what it pays out. It is among the largest Ponzi-like structures ever built, it is entirely legal, and it worked for a century only because each generation of contributors was outnumbered by the next. That stopped being true fifty years ago, in plain view, and the structure underneath was never rebuilt for what comes after.

A Retirement Age of Seventy

The first state pension to ever exist was, by design, a promise most of the people paying for it would not live to collect. In 1889 the German Reichstag passed the Old Age and Disability Insurance Law, the legislative core of what historians now call the Bismarckian welfare state. The bill set a retirement age of seventy. Otto von Bismarck made no secret of why: an old-age scheme run by the imperial state would, he hoped, undercut the rising Social Democratic Party by showing that the empire, not the socialists, could deliver security to working people. Modest pensions would be paid out of compulsory contributions from current workers and their employers, with a subsidy from the imperial treasury on top.

It is the retirement age that should arrest a modern reader. Seventy. At the time, male life expectancy at birth in Germany was approximately forty. That figure is dragged down by infant and childhood mortality, and a German man who reached twenty in 1880 could, on average, expect to live into his late fifties - still well below the age at which the pension began to pay. The scheme was built to pay benefits to a small, statistically improbable minority of long-lived workers, financed by a much larger mass who would, in expectation, contribute their whole working lives and never collect a pfennig.

That is what made the original arithmetic work, and it is the same arithmetic that makes every later pension work: a great many contributors standing behind each beneficiary. Bismarck achieved the ratio by drawing the beneficiary class deliberately narrow. His successors achieved it, for a while, by luck. Germany’s retirement age, under the post-2007 reforms, is rising towards sixty-seven, while male life expectancy at birth is near seventy-nine; the expected stretch of retirement has gone from a fragile possibility for a handful of survivors to a near-certainty of a decade or two for almost everyone.

The OECD-area ratio of contributors to beneficiaries has fallen from roughly six or seven workers per retiree in the middle of the last century to a little above three today, with the trajectory pointing towards two by mid-century. The promise that paid a few survivors out of the contributions of the many now has to pay almost everyone, out of the contributions of barely twice their number. The luck has been running out for decades, in full view of everyone whose job is to watch it, and the institutions built on top of it were not rebuilt. They were left to drift towards the cohort least able to survive the failure.

The Structure Is a Ponzi

Strip the politics away from a public pay-as-you-go pension and the structure that remains is the one this series has been tracking. It has the three features the opening essay used to define a Ponzi-like structure, and it has all three at once.

1. Earlier participants are paid from later ones. Current retirees receive benefits funded by the payroll taxes of current workers, not from any accumulated stock of their own saved earnings.

2. Operation depends on continued inflow. The system can meet its obligations only as long as workers keep contributing at a sufficient ratio to retirees. Slow the inflow, demographically or fiscally, and the structure cannot meet what it owes.

3. The system does not, on its own, produce the goods and services retirees consume. It allocates claims on present-day production; what retirees actually consume comes out of the output of present-day workers. The pension system is the transfer mechanism, nothing more.

That is the Ponzi-like structure in the analytical sense developed in the opening essay. It is not a Ponzi scheme: it is built deliberately, run capably, and defended in good faith. It becomes Ponzi-like the moment it funds old promises with new contributions and calls the demographic tailwind solvency. The observation carries no moral charge, and it was made explicitly, and approvingly, by Paul Samuelson in his 1958 paper An Exact Consumption-Loan Model of Interest with or without the Social Contrivance of Money, which characterised pay-as-you-go social insurance as actuarially unsound by the standards of private finance and socially beneficial all the same, so long as population and productivity were both growing. Henry Aaron formalised the point in 1966 as the social insurance paradox: a pay-as-you-go system can offer a return equal to the sum of the population-growth rate and the productivity-growth rate. In the high-growth decades after the Second World War, that beat any honest accumulation alternative. The favourable arithmetic was always a function of the growth rates. Slow either one, and the paradox flips.

The participant classes are wider than the contributor–beneficiary pairing suggests: current retirees who receive benefits, current workers who contribute and accrue future entitlements, future workers who will be both on terms set by present political decisions, employers who remit the employer share, the state that administers the system and stands as its ultimate backstop, and the bondholders of pension-issuing entities. That last class is worth pausing on. Where a trust fund exists at all, it typically holds claims on the issuing government itself. The fiscal capacity the bondholders are counting on is the fiscal capacity the pensioners are counting on. The two claims are, in the end, on the same balance sheet.

What the Structure Meets

A pension system is not optional in any plausible modern political economy, and the honest critique begins only after that is granted. A society without organised old-age provision tends to produce material elderly poverty at scale, and intergenerational stress on the younger members who absorb the cost informally. The pre-Social Security United States is the canonical reference: old-age poverty rates in 1935, measured against contemporaneous standards of subsistence, are generally estimated at somewhere between forty and fifty per cent. The 1935 Act, the post-war expansion of the programme, and the broader safety net drove that figure down decisively, to roughly a quarter of those sixty-five and over by 1970, and below ten per cent by the 2010s. The exact numbers depend on the measure used; the direction and the magnitude are not seriously contested.

What the system does is not easily replicated by private markets. It insures against the risk of outliving one’s savings, pooling longevity risk across millions so that an uncertain individual loss becomes a manageable aggregate cost; private annuity markets, where they exist at all, have historically reached only a small fraction of workers, defeated by adverse selection and cost. It puts a floor under old-age consumption that private savings did not produce before 1935 and would not produce now. It steadies the macroeconomy, since retirees’ incomes do not collapse when the labour market does. And it lets older workers leave in orderly fashion, opening positions for younger entrants.

So the structure cannot be dismissed merely because it is Ponzi-like; the pattern is doing genuine work. The question is whether the structure that meets these needs stays efficient, transparent, and durable once the demographic luck runs out, or whether it has hardened into a promise kept in place less by what it can still deliver than by how much now depends on no one asking what backs it.

A Brief Institutional History

Bismarck’s 1889 statute set the pattern most later systems would follow in outline: a compulsory contribution by workers and employers, administered by the state, with benefits payable at a defined age against a formula linking benefit to contribution history. It was a state instrument with political objectives from the start, the insurance function notwithstanding. The British Old Age Pensions Act of 1908 took a leaner shape (means-tested, non-contributory, funded out of general revenue, at five shillings per week for persons over seventy), more redistributive than the Bismarckian model but sharing the central logic all the same: present pensioners paid from present taxpayers. The United States Social Security Act of 1935 split the difference, establishing a contributory, payroll-tax-funded old-age insurance programme with benefits indexed to contribution history. Replacement rates were low, and coverage was incomplete. The programme then expanded through four decades of amendments.

The years from roughly 1945 to 1975 were the high tide of pension construction. Most OECD states built or substantially expanded public systems against three favourable winds at once: a sustained baby boom, rising labour-force participation among women, and rising productivity. The combination produced a contributor-to-beneficiary ratio and a wage-growth profile under which pay-as-you-go arithmetic worked beautifully. Replacement rates rose, eligibility widened, and in many systems the retirement age came down. The reckoning began in the 1970s, when total fertility rates in most developed countries fell below replacement, in some cases sharply, while longevity kept rising. The ratio underneath pay-as-you-go began to move in a direction it has not since reversed.

The first major American response was the National Commission on Social Security Reform, known after its chair as the Greenspan Commission, which reported in January 1983. The Old-Age and Survivors Insurance trust fund was, by then, within months of being unable to pay full benefits on schedule. The 1983 amendments that followed accelerated scheduled payroll-tax increases and introduced partial taxation of benefits for higher-income recipients, both of which the Commission had recommended. They also included a gradual increase in the full retirement age from sixty-five to sixty-seven, which the Commission had not recommended at all; that one was added on the House floor during debate. The package was designed to restore the trust fund to seventy-five-year actuarial balance, out to about 2058. The reform demonstrated something the institutions rarely state plainly: the promises are a political-economic instrument, adjustable by congressional action, not a closed actuarial commitment. What is owed can be rewritten by the body that owes it, and the people relying on the promise are the last to be asked.

The private landscape was shifting at the same time. The Employee Retirement Income Security Act of 1974 created the legal framework for private plans, with fiduciary duties for sponsors and federal insurance, through the Pension Benefit Guaranty Corporation, for defined-benefit plans that failed. Section 401(k) of the Internal Revenue Code, added in 1978 and operationalised in the early 1980s, launched what would become the dominant private retirement-savings instrument.

Through the 1990s and 2000s, corporate sponsors moved decisively away from defined-benefit plans, which placed investment and longevity risk on the employer, towards defined-contribution plans, which moved those risks off the employer’s balance sheet and onto the individual employee, a transfer of risk presented as a feature of choice. The shift had a second effect, less often remarked. Automatic enrolment and target-date defaults route a continuous, price-insensitive bid into the equity market every payday: tens of millions of contributions invested on schedule, with no reference to whether prices are high or low, because the worker has been defaulted into buying. A system sold as individual retirement saving doubles as a permanent forced inflow into the asset markets, exactly the kind of captive demand on which any structure that depends on continued inflow can quietly rely.

Public-sector pensions, by contrast, stayed largely defined-benefit and entered the 2020s in widely varying conditions. The Pew Charitable Trusts’ annual State Pension Funding Gap report has consistently named Illinois, New Jersey, Kentucky, and Connecticut among the most severely underfunded by the plans’ own actuarial assumptions, with funded ratios in some cases well below fifty per cent. The Detroit municipal bankruptcy of 2013 and the Puerto Rican fiscal restructuring of 2017 each forced reductions in promised public-pension benefits, which is to say that the legal protections around those benefits are not absolute when the issuing entity runs out of money.

How the Money Actually Moves

The mechanics differ across the three principal forms of provision, and the differences decide who absorbs the loss when one arrives.

Pay-as-you-go public systems. The mechanism is direct: this year’s payroll-tax revenue funds this year’s retiree benefits. A trust fund usually sits between inflow and outflow as an accounting buffer, but in most national systems that fund holds claims on the same government that pays the benefits. In the United States, the OASI trust fund holds non-marketable Treasury securities issued by the federal government; the asset-side claim and the liability-side obligation are owed by the same entity. The fund is, structurally, a transfer mechanism rather than an independent reserve of value. A few systems differ in substance and not merely in detail (Norway’s Government Pension Fund Global and Canada’s CPP Investment Board hold genuinely external assets at scale), but most do not.

Defined-benefit private plans. The employer commits to a defined benefit, typically calculated against years of service and final or career-average pay, and contributes to a separately held plan trust that invests in a portfolio of assets. At retirement the participant receives a defined payment whose continuity depends on the plan’s ongoing solvency and, if the plan fails, on the PBGC’s federally backed guarantee up to statutory maximums. Investment, longevity, and sponsor-bankruptcy risk all sit at the plan and sponsor level, not on the individual.

Defined-contribution plans. Here the risks shift entirely onto the individual. The worker contributes from wages, the employer may match, the balance accumulates in an account whose value at retirement determines retirement income, and the participant chooses, within a menu, how it is invested. Three distinct risks consequently land on one person: whether the chosen funds perform, whether market returns happen to be favourable in the years immediately before and after retirement (which has an outsized effect on the lifetime outcome), and whether the participant outlives the balance. In an annuitised defined-benefit plan the last two are pooled across the plan’s members; in a typical defined-contribution plan they are not. The structural difference is not who bears the risk in good times. In good times everyone is fine. The difference is who bears it in bad times. Defined-benefit plans pool the loss across sponsor, beneficiary, and, for public plans, taxpayers. Defined-contribution plans concentrate it on the individual retiree, who absorbs it alone, and at the worst possible moment.

Efficiency, Transparency, Longevity

Efficiency

Public pay-as-you-go systems are administratively cheap. The Social Security Administration reports administrative costs of well under one per cent of benefits paid, which makes it one of the most efficient large-scale income-transfer operations in modern government: a payroll tax collected by the IRS, a benefit calculated from a formula, a cheque sent monthly, with no fund manager, advisor, or broker-dealer standing between the contribution and the benefit.

Defined-contribution plans cost considerably more (mutual-fund expense ratios, plan-administration fees, advisor fees, and further layers of revenue-sharing and recordkeeping), and the compounded effect of even a modest differential over a forty-year horizon is large. Edward Wolff’s 2011 study The Transformation of the American Pension System documents the scale of the intermediary extraction in the defined-contribution era. Defined-benefit efficiency turns on funding discipline and investment governance: a well-funded, well-governed plan can deliver retirement income at a cost competitive with anything, while a poorly funded one enters a familiar dynamic. The sponsor, facing an unfunded liability and a contribution schedule it cannot or will not meet, reaches for higher-risk returns to close the gap, pulling in alternative managers, hedge-fund vehicles, and private-equity allocations whose fees are large fractions of the gross return. The federal guarantee behind private defined-benefit plans completes the pattern: a sponsor whose broken promises will be partly caught by a public insurer has less reason to fund them fully while it still can. The guarantee meant to protect beneficiaries quietly subsidises the underfunding that endangers them.

Transparency

This is the axis on which the pension system performs worst, and it does so by design rather than by accident. Contributors to public pay-as-you-go systems do not, as a general rule, understand the pay-as-you-go nature of the thing they pay into. The ‘I paid into it’ framing, ubiquitous in any discussion of Social Security, implies a vested claim on an accumulated personal fund. That fund does not exist. Today’s contributions fund today’s retirees, who in their working years funded an earlier generation, in a chain reaching back to the beginning of the system. None of this is hidden from anyone who looks (the Trustees Report explains the structure plainly every year), and yet the political language around the system, and the language participants use about it, almost universally implies an accumulated-fund model the system does not implement.

The institutions that administer it have no reason to correct a misunderstanding their solvency rests on. The 1983 episode showed the mechanics with unusual clarity. The task before Congress was, in effect, to negotiate which generation would carry which share of a shortfall already implicit in the programme’s design: the retirement-age increase shifting cost forward to workers born after 1937, the payroll-tax acceleration shifting it back onto workers then in the labour force, the taxation of benefits shifting it onto higher-income retirees. The package was politically durable in large part because most contributors and beneficiaries did not perceive these as the cost-allocation decisions they were.

The structure is renegotiated over the heads of the people who pay for it, and the contributor base is informed of the result, never consulted on the mechanics. Defined-contribution plans are formally transparent where public pensions are not (the participant gets a statement, a list of holdings, expense disclosures, a current balance), and that transparency is real at the account level and functionally incomplete on the only question that matters: whether the accumulation will be enough to support consumption through retirement. The projection involves assumptions about returns, longevity, and expenses that most participants are not equipped to make, and that most sponsors do not present in a way that would force a worker to confront an inadequate trajectory in time to correct it.

Longevity

Longevity is the axis on which pension systems are most directly under pressure, and the one on which the framework predicts most confidently. The OECD-area old-age dependency ratio (population aged sixty-five and over per hundred of working age) has risen from approximately fourteen in 1960 to over thirty in 2020, and the OECD’s own projections show it exceeding fifty in most developed countries by 2050, with Japan, Italy, Spain, and South Korea substantially higher. The mechanical implication for pay-as-you-go is direct: the same benefit promise, paid to a larger relative beneficiary class out of contributions from a smaller relative contributor class, requires higher contributions, lower benefits, later retirement, or a backfill from general fiscal resources.

It is tempting to assume that asset-funded systems escape this, since they do not depend on a contributor-to-beneficiary ratio. But the assumption is only half right. Asset-funded systems do reduce the direct dependency of benefit payments on current contributions; they do not escape the demographic problem, because the goods and services retirees consume are produced by workers, and the prices of the assets retirees rely on are partly a function of demand from saving workers. When the cohort drawing assets down in retirement exceeds the cohort accumulating them, asset prices face a structural headwind. The empirical magnitude is contested; the direction is not.

When demographic pressure forces a pay-as-you-go system into adjustment, the reform options narrow to four, and they have not changed in decades.

1. Raise the retirement age. Shifts cost onto workers approaching retirement.

2. Raise contribution rates. Shifts cost onto current workers.

3. Reduce benefits. Shifts cost onto current and future retirees.

4. Close the gap with general fiscal resources, i.e. transfer the problem onto the sovereign balance sheet, which transfers it again to whoever bears the cost of sovereign borrowing or taxation.

There is no fifth option, and no authority can conjure one. The arithmetic of pay-as-you-go forces a choice across these four, and every one of them is a decision about whose savings, whose working years, and whose old age absorb a shortfall the demographics have already fixed. The choice is not symmetric, and it is rarely made by those who will bear it.

Who Bears the Cost

The pattern of cost allocation across restructuring episodes is consistent enough across countries and decades to state directly. Current retirees are generally protected: cuts that take effect on people already drawing payments are the rarest reform in nearly every episode, because the constituency is concentrated, organised, and reliable at the ballot box. Workers near retirement bear the late-career adjustments and the retirement-age extensions, with less remaining working life across which to adjust. Workers mid-career bear the contribution-rate increases and the slower accrual, spread over a longer horizon and rarely the visible flashpoint of the debate. Young workers bear the structural cost of a system whose generous past commitments now extract a larger contribution share than they will, on present trajectories, get back, and they bear the implicit fiscal cost when general taxation backfills shortfalls.

Public-sector workers in underfunded plans face a particular risk of benefit reduction in restructuring: the Detroit bankruptcy of 2013 cut pensions for retired city workers, police, and firefighters; the Puerto Rican restructuring of 2017 cut them for retired public employees and teachers. The protections around public pensions are strong in normal times and limited in fiscal emergencies. Defined-contribution participants bear sequence-of-returns risk alone: a worker who reaches retirement age in 2009 and one who reaches it in 2007 face different lifetime outcomes from materially similar saving histories, because the drawdown of late 2008 fell on different stages of their accumulation, and there is no pooling mechanism to smooth it.

The externalities do not stay inside the pension system. Underfunding crowds out other state spending; by the mid-2020s, the share of Illinois’s general fund consumed by required pension contributions had risen to around a fifth or more, with corresponding pressure on every other category of expenditure. And the political economy of reform delay has a final feature worth naming: the cost of waiting falls disproportionately on cohorts who are not yet voting. A reform deferred from today until ten years from today reallocates a share of the structural cost onto people who are, today, children or not yet born.

There is no representational mechanism, in any democracy, by which the absent cohort registers an objection. The result is not an oversight but a structural bias (every actor has a reason to defer, and no actor speaks for the cohort the deferral is loaded onto), and it compounds over decades into the configuration most developed-country pension systems are now in: a debt to the unborn, accrued by the institutions sworn to protect them.

The cost does not stay inside the pension system because the pension system does not stand on its own. Its trust fund, where one exists, holds claims on the same sovereign whose debt the sovereign-debt essay watched being rolled forward; its funded variants hold the same equities the Federal Reserve essay in this series questions, at the same asset prices the central bank lifts and cannot always defend. The pension is downstream of every structure this series of essays has examined. It is the household’s claim on a future in which the sovereign stays solvent, the assets hold their prices, and the companies keep earning. When those assumptions hold, the promise is kept; when they do not, it fails for the one cohort with no working years left in which to absorb the failure.

The Authority Behind the Bargain

The three axes are three readings of one underlying condition: a functional opacity that hides what the bargain actually carries, laced with a trust or authority overlay that licenses the opacity to operate. In public pension systems the condition has institutional names. At the base sits the state, full faith and credit behind every promise as the ultimate backstop. Above it sits the trust fund, where one exists, and the annual Trustees Report, which together perform the appearance of an accumulated reserve and present projections under stated assumptions. Above that sits the pension formula, decades of regulatory codification converting a contribution stream into a stated future benefit by mechanical operation, which performs the appearance of a calculated entitlement. ERISA and the PBGC sit alongside, federally insuring private defined-benefit promises up to statutory caps.

The political constituency is itself a layer, the millions who say ‘I paid into it’ and the millions who say ‘I earned it’. State constitutional protections for accrued public-pension benefits, where jurisdictions have them, are another. And over all of it sits the Greenspan template, the settled understanding that restructuring is always available and is always a renegotiation among interest groups rather than a default. Each layer signals to the contributor and the retiree that someone responsible has already done the analytical work. Usually no one has. The credibility is the product, and the analysis is the thing it is sold in place of.

Functional opacity supplies the material the credibility waves through. The pay-as-you-go nature of the system is disclosed in every Trustees Report and is not the model contributors carry in their heads. The trust fund’s assets, where they consist of claims on the same government that pays the benefits, are an accounting reflection of historical inflow–outflow gaps rather than an independent reserve, and the distinction is not part of the standard public discussion. The actuarial assumptions (the population-growth rate, the productivity-growth rate, the discount rate at which liabilities are valued) are disclosed in technical appendices and determine whether a plan is reported as ‘fully funded’ or ‘underfunded’. The four-way reform menu is present in every actuarial report and absent from the political language. The combination is what carries the bargain.

The mechanics alone would force a public reckoning, every demographic cycle, with which cohort will carry which share of the structural cost; the authority overlay defers it by certifying that the system is funded, supervised, and backstopped; and the opacity hides how much of the stability depends on a demographic ratio that has been turning against it for fifty years. The contributor and the retiree do not learn they are inside it until the next restructuring episode, in which the cost allocation arrives as a fait accompli rather than the choice it actually is. By then the moment of discovery is already the moment of cost.

The Pension Ponzi

Pension systems are not frauds. They are widely beneficial public institutions that have measurably reduced old-age poverty across the developed world, and the populations they serve have, on balance, materially better lives than the elderly of the pre-pension era. The achievement is real and should not be minimised in the course of criticising the structure that produced it. But the structure is Ponzi-like at the level of mechanics, and the pattern’s vulnerabilities apply. Present payments come from present contributions; the arithmetic depends on a demographic dynamic that has been turning against it for fifty years and will keep turning for at least the next twenty-five; the reform options are arithmetically constrained; and the political economy of delay shifts cost forward to cohorts whose vote does not yet count.

On the three axes the picture is mixed but legible: public pay-as-you-go is highly efficient and well-funded defined-benefit competitively so, while underfunded defined-benefit and retail defined-contribution are not; transparency is poor in nearly every implementation, since contributors do not understand the pay-as-you-go nature of what they pay into and defined-contribution participants do not understand whether their accumulation will be enough; and longevity is the central problem and the axis on which the structure is most directly under pressure.

Demography is the one variable that does not negotiate, and it has moved against the structure for fifty years with the steadiness of a tide. The ratio of workers to retirees has fallen from roughly six or seven to one towards two to one, and no authority overlay can vote it back up. The reform menu has four items and has not changed in decades (later retirement, higher contributions, lower benefits, or a transfer onto a sovereign balance sheet that is already carrying the rest of the edifice), and each merely selects who absorbs a shortfall the arithmetic has already fixed. Detroit and Puerto Rico have shown what the resolution looks like in practice: the benefits are cut, the legal protections hold in normal times and yield in fiscal emergencies, and the loss falls on retired city workers, police, firefighters, and teachers who have no working years left in which to recover it. This is the promise made to the people least able to return to work once it is broken, resting on a foundation that is itself being rolled forward, unpriced, beneath the savings, the sovereign credit, and the asset prices it silently depends on.

When the demographic tide finishes turning, it will not be a bad quarter. It will be handed, whole, to the one cohort with nowhere left to stand. Whether that combination is removable, in this institution or in any Ponzi-like structure of any kind, is the question this essay does not answer. It leaves it with the reader, who will be standing inside the structure on the day the inflow no longer covers what was promised.

Conclusion · The PUNZI Papers Essay Six

This completes the present essay of The PUNZI Papers, a series of essays explaining the position behind our project, PUNZI WARS, and the potential it points to.

In the next essay we discuss The $715 Trillion Derivatives Ponzi.

Visit us: https://punzi.xyz Follow us on X: @PUNZIwars

Advisory: PUNZI WARS is a game and an on-chain research experiment. It is not an investment. Participate at your own risk.

Sources

1. Paul A. Samuelson, ‘An Exact Consumption-Loan Model of Interest with or without the Social Contrivance of Money’, Journal of Political Economy 66, no. 6 (December 1958): 467–482.

2. Henry J. Aaron, ‘The Social Insurance Paradox’, Canadian Journal of Economics and Political Science 32, no. 3 (August 1966): 371–374.

3. Report of the National Commission on Social Security Reform (Greenspan Commission), January 1983.

4. Organisation for Economic Co-operation and Development, Pensions at a Glance, annual editions.

5. Pew Charitable Trusts, The State Pension Funding Gap, annual editions.

6. Board of Trustees, Federal Old-Age and Survivors Insurance and Federal Disability Insurance Trust Funds, Annual Report (OASDI Trustees Report), annual editions.

7. Edward N. Wolff, The Transformation of the American Pension System: Was It Beneficial for Workers? (W. E. Upjohn Institute, 2011).

8. Otto von Bismarck, Reichstag Speech on Old Age and Disability Insurance, March 1889, in Reichstag Verhandlungen, vol. 105.