A government goes bankrupt when it can no longer meet what it owes: it misses a payment, or it has to restructure debt on worse terms because no one will lend to it otherwise. In the real world this is usually the end of a slow squeeze, rising borrowing costs eating a growing share of revenue, rather than one dramatic moment. NationCraft turns that squeeze into a single checkable rule: its bankruptcy death condition fires immediately once a government's treasury balance falls below negative 500 million.
This page is published by the developer of NationCraft. It explains a real-world economic concept and uses NationCraft's simulation as a worked example of how that concept can be modeled in a game.
A government's spending in a given year rarely matches its tax revenue exactly. Wars, recessions, infrastructure, pensions, and welfare all create years where outgoings exceed income, and rather than raise taxes sharply or cut spending overnight, governments borrow to cover the gap. They do this by selling bonds: a bond is a promise to pay back a fixed amount later, plus interest along the way, in exchange for cash today. Investors, pension funds, other countries, and central banks buy these bonds because a government's promise to pay is normally considered one of the safest assets available. Borrowing itself is not the problem. Nearly every government on earth carries some level of debt at all times. The problem starts when the cost of servicing that debt grows faster than the revenue available to service it.
A credit rating is an assessment of how likely a borrower is to pay what it owes, on time and in full. For a government, agencies weigh things like the size of its debt relative to its revenue, whether it is currently running a deficit, and how stable its politics and institutions look. A high rating tells lenders the risk of not being repaid is low, so they will accept a low interest rate. A low rating tells lenders the risk is real, so they demand a higher rate to compensate for it, a risk premium.
This is where borrowing can turn dangerous. A downgrade does not just cost a government reputationally, it directly raises the interest rate on every bond it issues from that point forward. Higher interest payments eat further into revenue, which weakens the government's finances further, which can trigger another downgrade, which raises borrowing costs again. That feedback loop, worse credit leading to costlier debt leading to worse credit, is the mechanism behind most sovereign debt crises. It is not that a government spent too much in one bad year. It is that the market's price for lending to it kept rising faster than the government's ability to pay.
Economists draw a real distinction between two situations that can look identical from the outside.
A liquidity crisis is when a government's underlying finances are sound over the long run, it has enough future revenue to cover what it owes, but it cannot refinance debt that is coming due right now, usually because lenders have gotten nervous and interest rates have spiked. If it can bridge that gap, with a loan from another institution, a temporary rate cut, or simply time for confidence to return, the crisis passes without anyone losing money.
Insolvency is different: the government's economy fundamentally cannot generate enough future revenue to cover what it owes, on any realistic set of terms. No bridge loan fixes that. The only paths out are restructuring the debt (paying back less, or later, or both, by negotiated agreement) or an outright default.
The uncomfortable part is that these two situations are not always easy to tell apart in the moment, and a liquidity crisis can talk itself into becoming insolvency: if panic pushes borrowing costs high enough, refinancing at those new rates becomes unaffordable even for a government that was solvent at the old rates. The crisis becomes self-fulfilling.
One distinction matters more than almost any other: does a government borrow in its own currency, or in someone else's?
A country that borrows in its own currency, and controls its own central bank, can always produce the money to make a scheduled payment, in the narrow sense that the payment will technically be made. That is not free of consequence: printing money to cover obligations debases the currency and can ignite inflation, so the cost shows up as eroded purchasing power rather than as a missed payment. Some economists treat that inflation as a default by another name; strictly speaking, though, the bond gets paid.
A country that borrows in a foreign currency, or a shared currency it does not control, has no such option. If its reserves and revenue run short, it cannot create more dollars or euros the way it can create its own currency, so an actual missed payment or a negotiated haircut becomes the only way through. This is why sovereign defaults cluster so heavily among countries that borrowed outside their own currency: the escape route print-your-own-currency nations have is simply not available to them.
NationCraft is a mobile government and economy simulator, and its engine tracks a single treasury balance that every tick of the simulation updates: revenue comes in, expenses (including debt service) go out, and the difference lands in the treasury. The player can issue bonds across three durations to cover a shortfall, and each bond locks in an interest rate at the moment it is issued, based on the central bank policy rate, the government's current credit rating, and inflation at that time. Once issued, that rate becomes a fixed monthly payment for the life of the bond regardless of what happens afterward, so debt taken on to solve one month's problem keeps compounding pressure for years after the emergency has passed, the same mechanic that makes real sovereign borrowing dangerous: the bill outlives the crisis that created it.
Credit rating plays the same gating role it plays in the real mechanism above. A nation's rating worsens as its debt grows relative to its revenue and as its treasury goes negative, and a rating below a set floor blocks the government from issuing any further bonds at all: a nation already under strain finds it harder, or impossible, to borrow its way out, the liquidity-spiral mechanism above expressed as a hard rule rather than a market mood.
Bankruptcy itself is where NationCraft draws its clearest line. Most other ways to lose a government require a run of bad months in a row: a revolution needs satisfaction critically low for six consecutive months, a coup needs stability critically low for nine. Bankruptcy has no such grace period. The instant the treasury balance drops below negative 500 million, the run ends immediately, because that number represents obligations the government cannot currently meet, and the game does not model a slow-motion negotiation to work that out, only a binary outcome once it is crossed.
Any model of something this large has to compress it. NationCraft does not distinguish between a government borrowing in its own currency and borrowing in a foreign one, every bond in the game behaves the same way regardless of scenario, so the specific escape route available to a currency-issuing government in the real world is not represented as a separate mechanic. It also does not model debt restructuring or a partial default, a haircut negotiated with bondholders, only a single treasury threshold that ends the run outright. Those are real simplifications, made so a single-player simulation has one clear, checkable rule for a genuinely messy real-world phenomenon, and they are worth naming rather than leaving implied.
It means the government can no longer meet what it owes: it misses a scheduled debt payment, or it is forced to restructure its debt on worse terms because no one will lend to it on the old ones. In the real world this is usually the end result of a slow squeeze in which borrowing costs rise faster than revenue, not a single sudden event.
Spending rarely matches tax revenue exactly in any given year, so governments sell bonds, promises to repay with interest, to cover the gap rather than raising taxes sharply or cutting spending overnight. Borrowing itself is not the problem; nearly every government carries some debt. The danger starts when the cost of servicing that debt grows faster than the revenue available to pay it.
A high credit rating tells lenders the risk of not being repaid is low, so they accept a low interest rate. A low rating makes lenders demand a higher rate as compensation, which raises the government's interest payments, weakens its finances further, and can trigger another downgrade, a feedback loop that is the mechanism behind most sovereign debt crises.
A liquidity crisis is when a government's finances are sound over the long run but it cannot refinance debt coming due right now, usually because lenders have gotten nervous; a bridge loan or a return of confidence resolves it. Insolvency is when the economy fundamentally cannot generate enough future revenue to cover its debt on any realistic terms, so only restructuring or default resolves it. Panic can turn the first into the second if refinancing costs spike high enough.
Yes. A country borrowing in its own currency can always produce the money to make a scheduled payment, though doing so risks debasing the currency and igniting inflation instead. A country borrowing in a foreign or shared currency cannot create that currency out of nothing, so if its reserves and revenue run short, an actual missed payment or a negotiated haircut becomes the only way through, which is why sovereign defaults cluster among countries borrowing outside their own currency.
NationCraft tracks a single treasury balance every simulated month. Bonds the player issues lock in an interest rate that becomes a fixed monthly payment for the life of the bond, compounding pressure on the treasury over time, and a low credit rating blocks further borrowing once it falls under a set floor. Bankruptcy fires immediately, with no grace period, the instant the treasury balance drops below negative 500 million.
Related: how government budgets work, why inflation spirals, best economy simulation games.