Government Debt - Should we worry about it?
Sovereign debt is in the news again, as the US negotiates raising its debt ceiling. Is sovereign debt something we should worry about?
Writer note: in the coming articles, I will trial a short summary feature in the beginning of the article that will go over the key points of the article. Let me know in the comments if this is helpful.
Summary
Governments can fund their projects via taxes or debt. Which method they choose is driven by the desire to balance outcomes across generations.
Government debt in developed economies is currently most likely to be ‘free’ due to borrowing rates being below GDP growth rates.
However, as the debt amount continues to increase, this may cease to be the case. Furthermore, certain generations may become worse-off due to what is known as the crowding-out effect causing private investment to drop.
Fundamentally, debt should be seen as a useful tool for governments when considering inter-generational trade-offs.
Although the current fiscal (financial) costs of debt might be overstated, the question on what the debt is spent on is probably more important than the amount of debt.
Introduction
The topic of sovereign, or government, debt is in the news again due to a unique legal and historical quirk in the US. The US, unlike most countries, has a legal debt limit (“debt ceiling”) that needs approval from Congress to be increased. The US actually used to have an even more restrictive spending law, whereby every debt and bond issuance had to be signed off by Congress. This was changed to a debt ceiling in 1917.
As there are significant political negotiations occurring around the debt ceiling, we can ask about the risks (and benefits) of sovereign debt. When countries and governments spend more than they earn (mainly in tax revenues), they must run a deficit and turn to debt financing. This debt financing is done by the issuance of bonds. Who actually holds these bonds can be varied – government debt can be held by central banks, financial institutions, citizens of the country issuing the debt, and foreign citizens and institutions. Who holds government debt is on its own an important question in economics, but in this post we will focus on the impacts of debt on the economy.
Financing through Debt
Any government spending can be broadly financed in one of two ways: either collecting revenue from its citizens, usually via taxes, or issuing debt. This debt needs to be repaid at some point in the future by the citizens, which means that tax revenues will have to be raised. This leads to a choice of “tax now or tax later”. One theory about this choice, that acts as a helpful thought experiment, is the Ricardian Equivalence theory. Named after economist David Ricardo1, the hypothesis argues that the choice of debt spending or raising taxes is irrelevant. This is because citizens would assume that in order to pay off the issued debt, their taxes will be increased in the future. To prepare for that situation, they would set aside this money today, which would basically mimic being charged a tax. However, Ricardian Equivalence has been repeatedly shown not to be true in the data. From this, the discussion on government debt spending grew into one on whether debt spending can benefit everyone, including future generations.
The main issue that arises with debt financing is the inter-generational question – funding via debt today could result in higher taxes for future generations, while at the same time increasing consumption today. This leads to a trade-off of today’s generations having higher consumption at the expense of future generations. However, whether this trade-off is necessarily true is also hotly debated in economics. This leads to two main distinct questions of what is the impact of debt spending:
The fiscal impact – is issuing debt a fiscally costly or beneficial act in purely financial terms? Basically this question focuses on whether issuing debt costs us money, today or in the future, or has no negative financial impact (colloquially known as a ‘free lunch’). For example, if the interest rate on the debt is lower than the return gained by spending this debt, the debt would have no negative fiscal impact.
The welfare impact – unlike the fiscal impact, this issue focuses on whether individuals are better off in terms of actual consumption. From the perspective of individuals, money on its own is not important, but rather the goods and services we consume (economists usually refer to this as “utility from consumption”). Thus, the welfare impact focuses on: 1) how the benefits are spread out (for example, an extra $1 of consumption for a poorer person gives them more value than a richer person); 2) how are they spread out in time (for example, are all current and future generations better off, or just the current ones); and 3) if the benefits are certain or uncertain (people are risk averse, which means they prefer guaranteed outcomes over expected outcomes).
Benefits of Debt Spending
In 2019, Oliver Blanchard delivered a speech (text) to the American Economic Association (video) on the issue of sovereign debt in a low interest environment. This speech has been considered to be one of the most important speeches in recent times, as it reinvigorated the debate around debt and deficit spending. Blanchard points out that for nearly all developed economies, the nominal interest rate for debt is below nominal GDP growth rates (which still holds true today). If we were to think of it in terms of an investment, we could borrow money to fund the investment at a lower interest rate than the return from the investment. In this circumstance, borrowing can be seen nearly as free. (In the Appendix of this post, I go over the mechanism explaining why such debt is not only sustainable, but also welfare-improving for all individuals in the economy.) This situation creates a ‘free lunch’, meaning it has no fiscal, or financial, impact. Research by Mian, Straub and Sufi (2021) has shown this situation to probably be true in the US (and in Japan). To illustrate this effect numerically, today, the nominal borrowing rate in the US (a 30 Year Treasury is at 3.65%) is below the GDP nominal growth rate (around 9-10%), which would imply that borrowing would be cost-less for the US. This pattern currently holds for many developed economies. However, although the debt might be fiscally ‘free’, it still might not have a positive welfare impact (issue 2 listed above). To resolve this issue, government debt acts as a tool to balance the welfare outcomes across generations.
To achieve this welfare balance, government debt (and any government spending for that matter) can be used broadly for two things – investments or consumption (for example, if a government gives a tax cut today, people can go and consume this money on entertainment). The distinction between the two might not always be clear-cut. Generally, investment can be seen as interventions that could result in higher economic output, usually in endeavors that are under-funded by the private market.2 Examples of this include infrastructure spending on roads or mass transit or funding for university research. These interventions predominantly benefit future generations. Consumption type spending are programs such as social security or stimulus checks, as these are direct transfer to individuals. These interventions benefit current generations. An example of a mixed program can be the Child Tax Credit, as it can be seen as a human capital investment, as well as a consumption program, which benefits both generations.
In either case, the main purpose of government debt is to balance the welfare outcomes between generations. If long-run investments are funded via tax increases, the generation (“Generation 1”) that pays the tax might not survive until the benefits of these investments materialize. However, if the government funds these long-run investments with debt, this problem is resolved. Generation 1 now will hold government debt, and when they retire and have no sources of income, they will be able to use their debt claim to receive income. Through debt financing, the government is able to both fund long run investment and increase consumption for both generations. Note that in an optimal world, we would fund the investments via tax increases (preferably lump sum taxes) on Generation 1 and then the newly incoming generation (“Generation 2” or the children of Generation 1) would pay Generation 1 via transfer when they retire, offsetting the tax Generation 1 initially paid. However, the problem with this solution is mainly an issue of political commitment – there is no guarantee that Generation 2 will want to transfer money to Generation 1 when the time comes. In his speech, Blanchard argues that debt financing also assists with this problem and that some levels of debt are welfare improving for everyone – in this case for both Generation 1 and Generation 2 (see Appendix for more details).
Macroeconomic Costs of Debt Spending
Based on the above, there appears to be a growing consensus that the fiscal impact of debt in advanced economies is not that large, if any. However, on the issue of welfare impact, there is far less consensus with many economists arguing about the adverse impact of debt. Typically, the key costs of debt and deficit spending, as articulated by Boskin (2020) in a response to Blanchard (2019) are the following: 1) crowding-out private investment3; 2) inflationary pressures if the debt is held by the central bank4; and, 3) panic over high levels of sovereign debt. The first issue that is currently most pertinent is the impact of the crowding-out effect. By reducing private investment, the costs to the economy can be significant. Kumar and Woo (2010) argue that a 10 percentage point increase in the debt-to-GDP ratio, reduces GDP growth by 0.2 percentage points for advanced economies, and by 0.3-0.4 percentage points for developing economies.
However, regarding advanced economies, the crowding out argument has been challenged because of the recent extremely low interest rate environment prevalent in these advanced economies. During the last decade nominal interest rates were zero (and therefore real rates were negative as inflation was positive). This meant that the price of money or investing was ‘free’ (more precisely, negative). The main way this could have occurred is because there were insufficient investment opportunities in the market – there was no demand for investment capital, while the supply of capital was large. With high supply and low demand, the price of money (and theoretically of anything) can become negative. One way to correct the problem of negative real interest rates would have been through additional government investments via debt, which would increase demand for capital. As the government would demand more capital, the price would go up. Today, however, real interest rates may potentially turn positive soon (when nominal interest rates will be higher than inflation), the issue of crowding-out private investment by issuing government debt could re-emerge because the demand for capital will be high.
Another problem of government debt, acknowledged by Blanchard (2019), is if the debt interest rate return becomes larger than the growth rate of the economy. In this situation, the overall debt burden could start growing exponentially as interest payments become larger. Greenlaw et al. (2013) estimate that each 1 percentage point increase in debt-to-GDP increases borrowing costs by 0.045 percentage points. A similar effect was found by Laubach (2009). In instances where the debt would become fiscally unsustainable, there would be two ways to resolve the issue: either an increase in taxes to pay off the debt or a default (i.e. the government reneges on its debt obligation). Increasing taxes, although unpopular, is generally not too problematic, especially as it would be a one-off increase to bring the debt amount to sustainable levels. However, defaulting on debt could have potentially far wider implications. The risk of default was very prominent during the European debt crisis in 2009-2010, especially for Spain and Italy. As this is an important topic on its own, I will go over it in a separate post.
Conclusion
The issue of deficit and debt funding is a topic that stirs significant political debates. It is important, however, to note that debt funding is a tool for allocating economic resources across generations. Investments today generally pay off well into the future and thus, current generations, who are the voters, have low incentives to fund these projects through taxes. Moreover, these voters also do not believe in the altruism of future generations, as these future generations would need to compensate their predecessors for having undertaken the long-term investments. Debt financing resolves some of this intergenerational conflict.
The more important question, however, is how the funding is allocated. Public investments, such as infrastructure spending, are generally seen to be welfare improving for future generations, as the fiscal multiplier (i.e. how much return we get for a dollar of spend) is greater than 1 resulting in more resources available for future generations than the debt that needs to be repaid. In such a situation, debt financing allows us to distribute some of these benefits from future generations to current generations that hold the debt, making everyone better off. Even critics of large debt burdens believe that this type of public spending is beneficial (Boskin, 2020). Furthermore, this is somewhat confirmed by research by Peppel-Srebrny (2021) which suggests that government bonds issued for investment projects have lower interest rates than bonds for consumption purposes, as the bond holders expect the economy to benefit in the long run making repayment more likely, and thus they require a lower return.
Finally, this all leads to the question of what is the optimal or maximal amount of debt a government can take, which, unfortunately, is also an unanswered question. The European Union has what is called a 60 - 3 rule, whereby debt-to-GDP shouldn’t exceed 60%, while annual deficits should never be greater than 3% of GDP. However, whether this is a good approach has not been determined. Blanchard, in an IMF post said about this question the following:
“So my answer to the question is, I do not know what level of debt, in general, is safe. Give me a specific country and a specific time, and I will use the approach above to give you my answer. Then we can discuss whether my assumptions are reasonable. But don’t ask me for a simple rule. Any simple rule will be too simple.”
He elaborated that EU fiscal rule was probably too restrictive during the 2008 crisis and maintaining the 60 - 3 rule probably lowered consumption for all generations, making everyone else worse off. Regarding the US, recent research by Mehrotra and Sergeyev (2020) suggested that the US sustainable fiscal limit could be around 150% to 220% of GDP, while it currently sits at around 120%. Mian, Straub and Sufi (2021), on the other hand, believe that the fiscally ‘free’ limit might already have been attained.
Post Scriptum
We often hear the analogy that government budgets must be balanced (revenues must equal spending) just like our own household budgets. This is not a good analogy for many reasons but the two key ones are that: 1) A significant amount of government borrowing is held by its own citizens – so it is as if we were borrowing from ourselves, and 2) government spending is an inter-generational consumption smoothing tool. We do not make our own household spending and investment decisions based on what we think is a fair consumption level of our great-grandkids, because we do not get any benefit from it. It is precisely the situation of not internalizing the externality of our own actions, which is why we, within our households, personally heavily under-invest in the welfare of our descendants. Government debt allows us to internalize this externality.
Appendix – Fiscally Neutral Debt
To elaborate more on why debt might have no fiscal cost, Blanchard proposed a model where individuals live for two time periods – one when they are young and work, and one when they are old and retire (this is called an overlapping generations model or OLG model). When young people earn money via work, they consume some of this money and save the rest. Whatever they saved when they were young, they will consume when they are old. The money that is saved is invested in capital (we can think of this capital as factories, machines and buildings, but also intangible capital such as software) and earns some return. Note that the more people save, the lower the return to capital is as capital becomes abundant.
Next, Blanchard assumes a government issues debt. As this debt must be held by someone, for simplicity, he assumes that the young ‘buy’ this debt (i.e. invest in it). For holding this debt, they will receive some risk-free return (a risk-free return, as the name suggests, can be thought of as a guaranteed return). Note that the young can still save some of their money in the riskier ‘capital’ (by riskier, we mean its return is uncertain), but because they purchase some of the risk-free debt, they save less in the risky ‘capital’. From the issued debt, the government can transfer the entire issued debt amount (net off the interest needed to pay the debt holders) to the old for them to consume. As long as the risk-free interest rate of the government debt is below the growth rate of the economy (the growth rate of the economy is driven by factors such as population growth and technological improvement), the debt will be financially sustainable. Moreover, this world with debt issuance will make everyone better off! The main reason for this is that the young need some ‘instrument’ to save for retirement. In a world without debt, they would need to save using risky capital and because people generally are risk-averse, the uncertainty of these returns harms individuals' welfare (i.e. you prefer to know how much you’ll receive when you retire rather than have a probability of outcomes). With the certainty of debt returns, individuals will be better off.
Lastly, as a small aside, it is worth noting that a direct transfer from young to old would be even better from a welfare perspective. The mechanism of young saving through government debt is inefficient because its returns are below the growth rate of the economy – a direct transfer, bypassing any saving instruments, would have a return equal to the growth rate of the economy. Thus the ‘return’ from the direct transfer is greater than using any saving instrument – whether government debt or risky capital.
The concept of Ricardian Equivalence was formalized by Robert Barro, who is considered to be one of the founders of new classical economics.
Underfunding of certain projects occurs due to the problem that is called not internalizing externalities. Externalities are indirect costs or benefits that impact uninvolved parties caused by activities of another party (pollution impacting people’s health caused by factories and using your car leading to traffic congestion are examples of negative externalities; vaccinating oneself is a positive externality as it reduces the likelihood of others getting sick). Since these costs and benefits are not taken into account by private individuals when making decisions (i.e. they don’t ‘internalize’ them), some activities happen too much (pollution), while others not enough from the perspective of society.
Crowding out occurs when government spending distorts the market economy. In this case, the government could impact the market economy adversely in two ways. First, if citizens buy government debt, there is less capital available for other market participants (for example, to invest in setting up a new company), who are generally better at allocating money than the government. Secondly, by spending, the government might make it too costly for other market participants to purchase for their needs. For example, once the government starts significant infrastructure projects, demand for construction will increase, pushing construction costs up, which will make it much more costly for private investors to undertake their own construction projects.
The central bank, in this case, injects new money into the economy to purchase the debt, potentially fueling inflation. If debt is held by individuals or firms, debt issuance is simply a different allocation of the available money.