If you have paid attention to American politics for any amount of time during a discussion of spending bills, you will have heard some analogy that compares fiscal policy to a household.
This is a terrible comparison. Worse, it’s very misleading. Governments do not have to, and do not, behave like households. Governments have resources on a scale incomparable to any household. They can (unless they are in a currency union like the Euro) literally print money to pay their obligations. They can run deficits in many more years than surpluses without triggering any sort of panic. They can accumulate debts and never have to clear all of their debts.
That last point is probably the most crucial: governments do not have to clear all of their debts. As long as an investor who buys a ten-year bond can expect to get paid back, there is no problem. The government will be able to keep borrowing at reasonable interest rates, and unless an economic catastrophe produces a sharp drop in tax receipts, a reasonable rate of GDP growth will allow the government to take in enough money to pay back the bondholders.
The analogy is even more ridiculous when it is used to compare a household to the US government. Treasury bonds are not like a credit card or a mortgage. When the markets get scared, they don’t come running to American families, begging them to buy things with their credit card or acquire a new house. Investors move their money to secure investments, and US government debt is an extremely secure investment. When T-bill yields fall, it is because lots of investors want to buy government securities. This is good news for the government because it means the interest rates on public debt are lower. This will not happen if you have a variable-rate mortgage. In an economic crisis, banks will be scared to lend to individuals and charge a higher rate of interest, if they lend at all.
After twenty or thirty years, you are expected to pay a mortgage back. If a basic economic unit (individual or nuclear family) continually spends more than it takes in, the credit supply will stop. Lenders don’t think they will be paid back. At the front end of a mortgage, a household will be in far more debt than any country–several times its annual income. A family making $100,000 a year could have a mortgage for $800,000 for an income-to-debt ratio of 800%. Over the period of the mortgage, the family would eventually pay that debt off–all of it. That’s the expectation, and banks don’t lend if they don’t expect to be paid back in full. Government creditors (generally, and in the case of the United States, always) get paid back. The difference is that the family can’t permanently run a debt-to-income ratio of 80-100%–eventually lenders will doubt that they will get their money back. By contrast, people who buy public debt know that they will get paid back. American bondholders have always gotten back every penny that they have been owed, no exceptions. Because of this, and because an economy at a normal growth rate will allow the state to accumulate more in taxes when the debt must be paid, they can finance deficit spending on a year-to-year basis.
Here’s an example: US GDP growth has averaged around 2-3% for the last fifty years. Let’s take a conservative year-on-year growth rate of 2% over ten years. This means that the GDP will be about 20% larger at the end of the decade. In terms of debt, it means the country will have 20% more in tax revenue (ceteris paribus) at the end of the period. Family income is not assumed to grow consistently in this way, which is one of the reasons households have different debt-accumulation constraints. Therefore, an investor could expect that, even if the deficit in year 1 is 10% of GDP, and that deficit is financed entirely by issuing bonds, growth in tax revenue will allow the government to repay its debts with interest. (The US and UK have run deficits that are about 10% of GDP for the last few years).
To sum up: markets do not treat governments like banks treat your family. The US government debt is nothing like your mortgage.