The Debt That Doesn't Appear on the Balance Sheet
You open the government's published debt figures. The number looks manageable, perhaps 60 or 70 percent of GDP, the kind of ratio that earns a decent credit rating and keeps bond markets calm. Then an actuary quietly runs the numbers on public sector pension obligations and produces a figure two or three times larger. Same government. Same moment in time. Completely different picture.
This is not a rounding error. It is a structural feature of how governments account for what they owe their workers, and it has been quietly inflating the apparent fiscal health of states for decades.
Promises Written in Pencil, Recorded in Invisible Ink
Most public sector pensions are defined-benefit schemes. A teacher, a police officer, a civil servant is promised a specific income in retirement, usually calculated as a fraction of final or average salary multiplied by years of service. The government's obligation is real, contractual, and in many jurisdictions legally protected. What varies enormously is whether that obligation appears on the government's books at all, and if so, how it is valued.
Private companies, under accounting standards like IFRS and US GAAP, must recognize the present value of their pension obligations on the balance sheet. If a corporation's pension fund holds assets worth $8 billion but the discounted value of future payments to retirees is $10 billion, that $2 billion deficit sits there as a liability. Shareholders can see it. Analysts can price it. Governments operating on a cash basis often do nothing of the sort: they record a pension expenditure only when the cheque goes out, so the accumulated promise, the mountain of future payments owed to current and former employees, simply does not appear. It is rather like a mortgage borrower listing no debt on a personal balance sheet because the monthly payment hasn't cleared yet.
The UK's Office for National Statistics periodically publishes supplementary figures attempting to capture unfunded public pension liabilities. Those estimates have come in at several trillion pounds, dwarfing the headline national debt figures that appear in budget documents and press coverage. Similar analyses of US state and local government pensions, conducted by researchers at institutions including Stanford's pension tracker project, have produced aggregate shortfalls in the trillions of dollars once more realistic discount rates are applied.
The Discount Rate Is Doing Most of the Work
This is where the accounting becomes genuinely consequential, and where choices made by governments and actuaries can shift the apparent liability by enormous amounts. The discount rate, in short, is the most powerful lever in the room, and it is almost never discussed in budget speeches.
To calculate what a future pension payment is worth today, you discount it back at some assumed interest rate. A higher rate makes distant obligations look smaller. A lower one makes them look larger. The choice is therefore not merely a technical preference. It is a political one.
Consider a concrete scenario. A government owes a retired civil servant $50,000 a year for the next 20 years. Discounted at 7 percent, the present value of that stream of payments is roughly $530,000. Discounted at 3 percent, it climbs to around $745,000. Same obligation, same person, same payments: a difference of more than $200,000 depending on a single input assumption.
Many governments have historically discounted their pension liabilities using the expected return on invested assets, or using long-term bond yields at moments when those yields happened to be high. When interest rates fell to historic lows across much of the developed world, some schemes kept discount rates elevated by assuming strong equity returns going forward. This is the accounting equivalent of valuing your house at the price you hope to achieve rather than what a buyer would pay today, and it deserves to be treated with the same scepticism.
Financial economists, most notably Robert Novy-Marx and Joshua Rauh in their widely cited work on US public pensions, have argued that because pension promises are essentially risk-free obligations, they should be discounted at a risk-free rate, closer to government bond yields. When that adjustment is made, the reported funding gaps of public schemes tend to expand dramatically.
What People Get Wrong About the "Funded" Distinction
There is a common assumption that funded schemes, those where assets are set aside in a pension fund, are inherently transparent and safe, while unfunded or pay-as-you-go schemes are the dangerous ones. The reality is considerably less tidy.
A funded scheme with a reported 85 percent funding ratio can look healthy and still conceal a severe problem if that figure was calculated using an aggressive discount rate. Take two hypothetical public employees: Maria, who works for a state that marks its pension liabilities to market using a conservative discount rate and shows an honest 70 percent funding ratio, and David, who works for a state that uses an optimistic 7.5 percent discount rate and reports a flattering 95 percent funding ratio. David's pension may actually be the more precarious one. The number is just better dressed.
Unfunded schemes, for their part, are not automatically dishonest. Some governments explicitly disclose the scale of their pay-as-you-go commitments through supplementary reporting. The problem is not the funding model itself but the absence of consistent, mandatory disclosure that would allow citizens and investors to make meaningful comparisons across jurisdictions.
The Fiscal Illusion and Who Pays for It
When pension liabilities are hidden or understated, governments appear more solvent than they are. That apparent solvency allows politicians to resist tax increases, expand other spending, or borrow more cheaply than the true fiscal position would justify. The obligation doesn't disappear. It shifts forward, onto future taxpayers or onto employees who may find their promised benefits cut when the gap finally becomes impossible to paper over.
Detroit is the case that stays with you. The city entered bankruptcy proceedings in which pension obligations featured centrally, and retirees who had built careers around specific benefit promises received less than they were owed. The accounting treatment in the years before the crisis had not made those obligations visible in any way that triggered political action early enough to matter. History has a habit of making these lessons available only after the cost of ignoring them has already been paid.
Ask yourself this: have you ever found pension liabilities consolidated into your own country's headline debt figure, rather than buried in a supplementary annex that nobody reads? If the answer is no, you are looking at a fiscal statement that is, at minimum, incomplete.
The honest reckoning is not that every government pension system is a ticking catastrophe. Many are manageable, and some are genuinely well-funded. The problem is that the accounting frameworks in common use make it very difficult to tell the difference from the outside. A government that has made careful, conservative provision for its workers' retirements and a government that has quietly deferred an enormous liability can both produce budget documents that look, on the surface, broadly similar.
Debt that doesn't appear on the balance sheet is still debt. It is simply debt that has been made politically comfortable to carry, right up until the moment it isn't.