# How much can a government borrow?

Few days ago, I read a write-up titled [No one knows how much the government can borrow](https://noahpinion.substack.com/p/no-one-knows-how-much-the-government)

The authors basically argues the following:
1. _How much can the government (US presumably) safely borrow?_ is an important question.
2. It seems that _no one knows the answer to this question_ as there seems to be an acute lack of research/literature on the topic. The author says there's quite some literature on historical cases of country defaults, as well as how countries ot out of hyperinflation, but little on what _actually triggers a default/hyperinflation_.
3. The author finds a paper which finally sheds some light on the topic and the main finding is that _hyperinflation episodes tend to be preceded by several years — maybe 2 to 12 years — in which inflation is high but not yet “hyper”._
4. One of the very interesting takeways was that Japan [In 2019, Japan's debt reached 223% of GDP](https://en.wikipedia.org/wiki/National_debt_of_Japan) and as of 2020 might as well be above [250% GDP](https://worlddebtclocks.com/japan), yet it somehow does not yield a state's default on its obligations.

Now I am not an economist (moreover, I am rather suspicious of that discipline in general), but this topic interests me. The other day, I had a conversation with a friend who is very worried about the public debt to GDP ration of Italy and how it can negatively affect EU, Eurozone and specifically northern-EU countries (because they will pay for the lazy Italian pensions, right?). So the question arises: **How much debt/GDP can Italy afford? Where is a breaking point? And what options different stakeholders in the game will have once a default of Italy is triggered?**

Since in the news everybody has an opinion on this, I asked around the Intertubes and among my smart friends, naively thinking that this must be a well researched topic in that high profile discipline. After all, maybe I am suspicious of economists, but I do treat most of their work with high respect. Well, my inquiries yielded nothing useful.

So here I am asking for pointers to useful literature on the topic. **_What do we know about how much a country like Italy can borrow before it ends up in trouble?_**

@FailForward I would argue that this is actually a much easier thing to reason than you might realize and that aside from a few quirks the same rules as a business borrowing for growth applies, or entity.

> How much debt/GDP can Italy afford?

The first point though is that the criterion for measurement is incorrect here. The ratio of debt to GDP is not an effective or useful measure, the reasoning is simple when you consider the case of entities in general.

If you have 0 income as either a company or a country, and more expenses than income, borrowing can, and often is, the best option for you, presuming that money is used properly. Many companies start out with no income or even product, make a loan from a bank, and use that money to grow and become profitable, eventually paying off the loan and being in the positive. Obviously this in no way guarantees default.

So the real question becomes, what is the critereon for evaluating appropriate debt. The answer to that would be that the interest rate on the total outstanding loans need to always be less than the ratio between the growth rate of the GDP and the discretionary income of the country in question applied to the amount by which the discretionary income would grow as a result of the loan.

A country, like any individual has some discretionary income, that is, the money it collected from taxes after all its fundemental and unavoidable expenses are met. So for example it is their tax income minus their foreign debt burden (minimum payments on foreign debt), the payout on government bonds that were issues that expire that year, as well as the virtual loss of money (or gain) through inflation that occured that reduced the nominal income.. what is left over is their discretionary income.

Next one would look at how well a government is spending that discretionary income in order to fuel future growth. If we see that 100B DI tends to result in a 5% GDP growth over time then we can safely say the way that country spends that 100B has a 5% GDP growth rate.

So the question becomes is the additional income that is gained by the increase in GDP (factored over the entire life of the loan) **more** than the total cost of the loan.

So say the overall effect on GDP throughout the life of the loan (once the loan is paid back) is 10% and this represents a 50B in the GDP, and presuming that without the loan their discretionary income would have resulted in only a 5% GDP increase, and thus represents a 25B overall gain over the same lifetime of that loan. If the total cost of the loan after interest is more than 25B then the loan cost you more than you gained from it, therefore it has pushed you farther to default.

As long as you are making more money from your loans than the loan is costing you then no matter how big the loan is, its a good idea.

@freemo Thanks for the lecture, I do understand these technicalities. Here I am interested in "applied macroeconomics" - if such a thing even exists.

Finances (and thus debt) of individuals lives in a fixed context (from their perspective). A country of certain size and standing is, however, well capable to change, or at least significantly influence the context: e.g, by printing money by mechanisms like QE, changing legal and political framework around its debt, use its trade or military gross tonnage to effectively blackmail its creditors, etc. Individuals have none of that.

At the core of any loan is in the end _trust_. To my understanding, a default happens when creditors no longer *believe* you can service your obligations. But that "feels" too weak. I would hope there's more human knowledge on this topic.

That is why I am interested in serious literature on the topic, rather than contrived technical arguments (I have myself a full bag of those). Do you perhaps have any pointers?