Conversable Economist

The United States has had an industrial policy aimed at boosting its domestic shipbuilding industry since the passage of the Merchant Marine Act of 1920, commonly known as the Jones Act. Whatever the arguments for the passage of the bill a century ago, it has over time been a disaster for the US maritime industry, and continues to impose significant costs on other parts of the US economy. Colin Grabow goes through the arguments in “Protectionism on Steroids: The Scandal of the Jones Act” (Milken Institute Review, Second Quarter 2024, pp. 44-53).

The Jones Act “requires that vessels engaged in domestic transportation be registered and built in the United States as well as crewed and at least 75 percent owned by U.S. citizens.” However, the underlying rule goes back to an 1817 law “prohibiting foreign vessels from transporting goods within the U.S.”

The political problem back in 1920 was that as US shipbuilding and shipping costs were protected from foreign competition, they were no longer cost-competitive. in terms of production costs, and shipping by US-owned firms was not cost-competitive, either. Grabow gives an example of one 19th-century firm that shipped from New York to Belgium to California–because it was cheaper to pay for two “foreign” trips with non-US firms than to pay a US shipping firm to go direct from New York to California.

The gap in US ship-building costs has only widened. Current estimates are that “large cargo ships constructed in U.S. shipyards today cost at least 300 percent more than the competitive world price.” For operating costs, Grabow cites a 2018 report from the Government Accountability Office which finds: “According to U.S. Maritime Administration (MARAD) officials, the additional cost of operating a U.S. flag vessel compared to a foreign-flag vessel has increased—from about $4.8 million annually in 2009 and 2010 to about $6.2 to $6.5 million currently—making it harder for such vessels to remain financially viable.”

The consequences of this US attempt at a pro-shipbuilding and pro-US-shipping industrial policy have been awful. Here are some of them.

1) The US shipbuilding industry, with no need to respond to international competition, has become irrelevant in global markets. Here’s a table on large ocean-going ships under construction from the Congressional Research Service (“U.S. Commercial Shipbuilding in a Global Context,” November 15, 2023).

The CRS reports:

During World Wars I and II, the United States built thousands of cargo ships. These were sold to merchant carriers after the wars, including foreign buyers, but were soon replaced by more efficient ships built in foreign yards. In the 1970s, U.S. shipyards were building about 5% of the world’s tonnage, equating to 15-25 new ships per year. In the 1980s, this fell to around five ships per year, which is the current rate of U.S. shipbuilding. … The Jones Act’s domestic construction requirement likely underpins the entirety of U.S. commercial ship construction. None of the U.S.-flag international trading fleet is domestically built, though shipbuilders could take advantage of both the loan guarantee and tax shelter programs discussed above. No overseas purchase of large U.S.-built ships has occurred in decades because U.S.-built ships can be four or more times the world price.

Indeed, the US military relies on Chinese-built ships to support its military vessels: “Three of the ten commercial oil tankers selected to ship fuel for DOD as part of the newly enacted Tanker Security Fleet are Chinese-built. As for dry cargo supplies for DOD, 7 of the 12 most recently built ships in the Maritime Security Fleet are Chinese-built.

2) The higher costs of Jones-Act-compliant US shipping naturally impose heavy costs on places like Hawaii, Alaska, and Puerto Rico. Weird consequences result, and Grabow provides a number of examples. Puerto Rico gets its liquified natural gas from Nigeria, because there are no Jones-Act-compliant US ships to transport natural gas within the United States. US lumber producers complain that they have a disadvantage vs. Canadian firms, because the US lumber producers must use higher-cost Jones Act ships to send their products to US destinations, while Canadian lumber producers can use cheaper international shipping companies.

3) One might think that a natural transportation advantage for the United States would be to take advantage of maritime shipping via oceans on both sides. But the high cost of Jones-Act-compliant US shipping means more trucks and freight trains, with costs including traffic congestion, highway repair, and greater pollution.

4) Various specialized uses of ships become more costly. For example, if you want offshore wind-power to be an important part of future US electricity generation, you should know that it is considerably more costly to build with Jones-Act-compliant ships. Even basic tasks like dredging US ports and rivers are slower and more costly because the Jones Act (along with some other legislation of that time) shuts off the supplier of higher-quality and lower-cost dredging ships made elsewhere.

Supporters of industrial policy have a tendency to brush aside examples like the Jones Act: “Sure, that’s a foolish way to implement industrial policy, but my plan is a smart way to do so.” “Yes, the Jones Act is a problem, but the way to fix it is with much bigger government subsidies to expand US shipbuilding.” But the Jones Act is a classic example of a special interest law that benefits a small and very vocal group, while imposing large but diffuse costs. The problems of the Jones Act have been well-known for decades, and nothing has changed. Every proposal for industrial policy faces similar political economy dynamics.

Thus, it seems to me that the challenge for supporters of industry policy is not just to pick some alluring industries and then to hand out government favors like Halloween candy, but to specify in advance how they intend to measure success or failure of these subsidies–perhaps with a series of goals that must be met over time or else the subsidies get turned off. In South Korea, for example, which is often cited as an example of successful industry policy, the government subsidies for certain industries were often made contingent on the industries expanding their export sales at prevailing prices in international markets. When industrial policy goes poorly, as in the Jones Act, the costs are broadly felt across an array of related industries.

Extremely high levels of wealth are were not typically generated by people who were saving out of the income that they earned. Instead, high levels of wealth are typically about assets that rose considerably in value–sometimes land or real estate, often stock in a company. Billionaires like Elon Musk or Kim Kardashian don’t have a basement full of dollar bills, like Scrooge McDuck. Instead the bulk of their wealth is held in shares in corporations, where those shares have risen in value over time.

Thus, if you want to impose taxes that will affect the wealth distribution, raising the top-level income tax rates is not the most useful answer. “How to Tax Wealth,” a group of economists from the IMF (Shafik Hebous, Alexander Klemm, Geerten Michielse, and Carolina Osorio-Buitron, IMF How to Note 2024/001, March 2024). They write:

This note discusses three approaches of wealth taxation, based on (1) returns with a capital income tax, (2) stocks with a wealth tax, and (3) transfers of wealth through an inheritance (or estate) tax. Taxing actual returns is generally less distortive and more equitable than a wealth tax. Hence, rather than introducing wealth taxes,
reform priorities should focus on strengthening the design of capital income taxes (notably capital gains) and closing existing loopholes, while harnessing technological advances in tax administration—including cross-border information sharing—to foster tax compliance. The inheritance tax is important to address the
buildup of dynastic wealth.

I’ll add a few more words about the three options.

Income from capital can arrive in various ways, including interest payments, rent payments, dividends, share repurchases, or a pass-through firm that distributes profits to owners. But here, I want to focus on the problem of taxing capital gains: again, when you look at billionaire-level wealth, the wealth is commonly built on how the value of an asset, like stock ownership, has risen over time.

As the authors explain, the common approach is to tax capital gains when they are “realized”–that is, when the asset is sold. But this approach raises two issues. One is that if the asset has been held for a substantial period of time, the capital gains during that time have gone untaxed until they are realized–and deferring taxes for years is a substantial benefit.

The other issue is that it is often possible to roll one capital gain into a new asset without being taxed on the gain. In a US context, individuals can roll the capital gain from one house into the purchase of another house. If someone dies while holding stock, there is a “step-up” where the heir can value the stock at the price at the time of death, so the gains during the lifetime of the previous owner are not taxed.

The IMF authors describe the resulting problems in this way:

• Tax avoidance is encouraged, as there is an incentive to turn income into capital gains to benefit from lower taxation. For example, investment funds can reinvest rather than distribute earnings, and bonds can be designed to increase in value rather than pay interest.
• Tax legislation and administration increase in complexity as there is a need to address loopholes. For example, zero-coupon bonds are often taxed on their implied interest.
• Horizontal equity is diminished, because similarly profitable investments are taxed differently depending on the form in which they generate income.
• Vertical equity is diminished, because the share of income earned as capital gains rises with wealth and income. In the United States, the top 0.001 percent of taxpayers earned 60 percent of their income as capital gains (IRS 2022). In the United Kingdom, among the top 0.01 by income, almost 60 percent receive at least 90 percent of their remuneration in capital gains (Advani and Summers 2020).
• There is a lock-in effect as investors prefer to hold on to an asset even if the expected future returns are lower than those of alternative investments, as long as the tax saving from not realizing a capital gain outweighs the difference in returns.21 This leads to inefficient capital allocation. Some countries tax capital gains at lower rates (especially for long-term gains) to reduce this effect but thereby exacerbate the relative undertaxation of capital gains.
• In an international context, tax avoidance and evasion occur even on realized capital gains. For example, instead of trading a security directly, investors can trade a depository receipt in an offshore market that does not tax capital gains. Similarly, rather than directly selling a real asset, stocks or entire companies (registered in a different, conduit, country) that derive their values from that underling asset can be traded. The revenue loss can be significant in the case of high-value assets such as natural resources.

A final issue with taxation of capital gains involves inflation. If the increase in the value of my asset (say, my house) over time just matches inflation, then should this gain be treated as “income” to me when I sell the house?

There are ways to address all of these issues, but they aren’t simple.

With regard to a wealth tax, one immediate concern with a wealth tax is that a number of countries with wealth taxes decided to repeal them: basically, they were too much trouble to administer for too little revenue gain. The authors note:

[A]mong OECD members, those levying an explicit wealth tax declined from 12 in 1990 to only 3, while the Netherlands de facto also levies a wealth tax as part of its personal income tax (as does, outside the OECD, Liechtenstein). And where employed, the wealth tax is not a significant source of revenue, because of high
exemption thresholds and widespread evasion, amid severe enforcement challenges (Kopczuk 2019; Advani and Tarrant 2021). At 1.4 percent of GDP over the 2018–20 period, Switzerland has the highest revenue yield globally, but the country does not levy a capital gains tax (and its wealth concentration is high by international standards [Föllmi and Martínez 2017]). With the existing wealth taxes mostly modest and limited, studying them will not necessarily be indicative about the effect of more comprehensive or higher wealth taxes.

I’ve written earlier posts about countries dropping their wealth taxes here and here, and the case of equality-minded Sweden dropping its inheritance tax here.

One can make a case on paper for a tax on the super-wealthy, like those with more than a billion dollars in wealth. But the reality that even lower wealth taxes were too difficult to collect should raise some doubts. And even a substantially more aggressive wealth tax on the super-wealthy would have limited effects on revenue: “The EU Tax Observatory (2023) estimates that a wealth tax of 2 percent on the world’s top billionaires in 2023 (about 2,800 billionaires, 30 percent of whom are in the United States according to the report) can raise about $250 billion (or 0.2 percent of world GDP).”

In terms of incentives, a wealth tax applies whether or not there is income. Imagine a risky investment. With a tax on capita income, the tax revenue goes up if the investment is a success–say, if it doubles in value–but the tax rate goes down and even becomes negative if the investment fails–say, falls to half its value. With a wealth tax, the investor still owes the wealth tax on whatever remains even if the investment has failed: in this way, a wealth tax increases risk. In addition, because wealth is typically held not in rolls of dollar bills, but in assets, paying a wealth tax may require selling off some of the asset itself.

With regard to an inheritance tax, the primary goal is to limit the intergenerational transfer of extreme wealth. The authors write:

Empirical evidence shows that the share of inherited wealth in overall wealth is large, though precise figures are hard to come by. One difficulty is that estimates differ much depending on whether capital income earned on inherited wealth is counted as part of the inherited share or not. Davies and Shorrocks (2000) argue that a share of 35–45 percent is a reasonable estimate, based on balancing different assumptions made in papers yielding much higher or lower estimates. With more detailed and recent data, which are available for a few European countries, Piketty and Zucman (2015) report results for France, Germany, and the United Kingdom, finding that in 2010, the share of inherited wealth ranges from just over 50 percent in Germany to close to 60 percent in the United Kingdom. Moreover, as shown by Acciari and Morelli (2020) using Italian data, inheritances appear to become larger (from 8.4 to 15.1 percent of GDP between 1995 and 2016) and more concentrated over time. According to a UBS (2023) report, new billionaires acquired greater wealth through inheritance than entrepreneurship.

The creativity of tax attorneys will pose challenges for an estate tax. What if a wealthy person leaves their money to a trust? What the wealthy person leaves the money to a nonprofit, but also establishes their children with extremely well-paid jobs at that nonprofit? In what ways can wealth be transformed into untaxed forms? How does an inheritance tax after death compare with tax treatment of large gifts that are given during life? Is it right if an inheritance tax forces a family to sell off, say, a family home or a family farm?

In most countries, the inheritance or estate tax raises a relatively small amount.

For a follow-up to this post, see “The Super-rich and How to Tax Them” (November 17, 2020).

inheritance taxes, with a big

Empirical evidence shows that the share of inherited wealth in overall wealth is large, though precise figures are hard to come by. One difficulty is that estimates differ much depending on whether capital

The IMF publishes a Fiscal Monitor report twice a year about levels of spending and taxes around the world. The April 2024 report, subtitles “Fiscal Policy in the Great Election Year,” contains some warnings about the size of US budget deficits.

For context, here a table with fiscal balances for high-income countries, with actual data for 2019-23, and projected data from 2024-29. You can see that before the pandemic in 2019, the US already had a higher-than-average budget deficit. When the pandemic hit, deficits go up everywhere, but among high-income countries are largest in the US. The US deficits are projected to be much higher than those of other high-income countries moving forward: for example, in 2025 the US fiscal deficit is 7.1% of GDP, while other advanced economies excluding the US have an overall deficit of 2% of GDP.

I’ll note in passing that this is not a party-line issue. US deficits were already high under President Trump. The increase US government spending in response to COVID was bipartisan. The high deficits are now persisting under President Biden.

The IMF describes the US budget situation this way (mentions of figures omitted):

In 2023, the United States experienced remarkably large fiscal slippages, with the general government fiscal deficit rising to 8.8 percent of GDP from 4.1 percent of GDP in 2022, despite strong growth. Income tax revenues fell sharply, by 3.1 percentage points of GDP, owing to lower capital gains taxes in 2023 and delayed tax payment deadlines. Spending, in turn, increased by 1.3 percentage point of GDP.

The overall fiscal deficit is projected to persist at more than 6 percent of GDP over the medium term. Financing costs have increased substantially in recent years. Nominal yields on 10-year US Treasury bonds surged from below 1 percent in 2020 to 5 percent in October 2023, the highest level in 16 years, before receding to about 4 percent more recently amid a rapid pickup in inflation and inflation expectations. …

The rise in nominal term premiums also contributed to the surge in nominal Treasury yields in mid-2023. This rise reflects several factors, including the perceived risk of sustained inflation and uncertainty about the future path of monetary policy (US Congressional Budget Office 2023). Further, the Treasury’s plans to issue more debt, coinciding with quantitative tightening, likely contributed to heightened volatility in bond markets and a rise in term premiums … An empirical analysis to quantify the spillovers of US long-term nominal interest rates to nominal rates in other economies suggests that a 1 percentage point spike in US rates is associated with a rise in long-term nominal interest rates that peaks at 90 basis points in other advanced economies, with a persistent impact over many months. For emerging market economies, the same spike in US rates is associated with a peak increase in long-term interest rates of about 100 basis points. Moreover, it is possible that uncertainty about US fiscal policy and long-term rates could adversely affect financial conditions elsewhere. …

In sum, the previous analysis points to risks from loose fiscal policy in the United States along several dimensions. Loose US fiscal policy could make the last mile of disinflation harder to achieve while exacerbating the debt burden. Further, global interest rate spillovers could contribute to tighter financial conditions, increasing risks elsewhere.

The IMF also traces most of the surge in US core inflation rates to the very high budget deficits:

It is remarkable and discouraging that in the run-up to a US presidential election this fall, a central economic issue like the federal budget is not seriously discussed–indeed, it is barely mentioned.

The Dunning-Kruger effect can be paraphrased in this way: “On any particular topic, people who are not experts lack the very expertise they need in order to know just how much expertise they lack.” Corey S. Powell interviews David Dunning on how the underlying idea has been developed since the original paper published in 2000 (“David Dunning: Overcoming Overconfidence,” Open Mind, April 5, 2024).

For those who have only seen “Dunning-Krueger effect” deployed as an insult, it’s perhaps useful to briefly review the original paper from 25 years ago: “Unskilled and Unaware of It: How Difficulties in Recognizing One’s Own Incompetence Lead to Inflated Self-Assessments,” in the December 1999 issue of the Journal of Personality and Social Psychology (77:6, 1121-34). The paper opens with a lovely opening anecdote:

In 1995, McArthur Wheeler walked into two Pittsburgh banks and robbed them in broad daylight, with no visible attempt at disguise. He was arrested later that night, less than an hour after videotapes of him taken .from surveillance cameras were broadcast on the 11 o’clock news. When police later showed him the surveillance tapes, Mr. Wheeler stared in incredulity. “But I wore the juice,” he mumbled. Apparently, Mr. Wheeler was under the impression that rubbing one’s face with lemon juice rendered it invisible to videotape cameras (Fuocco, 1996). …

We argue that when people are incompetent in the strategies they adopt to
achieve success and satisfaction, they suffer a dual burden: Not only do they reach erroneous conclusions and make unfortunate choices, but their incompetence robs them of the ability to realize it. Instead, like Mr. Wheeler, they are left with the mistaken impression that they are doing just fine. … [A]s Charles Darwin (1871) sagely noted over a century ago, “ignorance more frequently begets confidence than does knowledge” (p. 3).

The actual study involved surveys of dozens of Cornell undergraduates. From the abstract:

People tend to hold overly favorable views of their abilities in many social and intellectual domains. The authors suggest that this overestimation occurs, in part, because people who are unskilled in these domains suffer a dual burden: Not only do these people reach erroneous conclusions and make unfortunate choices, but their incompetence robs them of the metacognitive ability to realize it. Across 4 studies, the authors found that participants scoring in the bottom quartile on tests of humor, grammar, and logic grossly overestimated their test performance and ability. Although their test scores put them in the 12th percentile, they estimated themselves to be in the 62nd. Several analyses linked this miscalibration to deficits in metacognitive skill, or the capacity to distinguish accuracy from error. Paradoxically, improving the skills of the participants, and thus increasing their metacognitive competence, helped them recognize the limitations of their abilities. 

Obvious questions arise. Do the results from these tests and from college undergraduates generalize to other settings and populations? This is where Powell’s interview with Dunning comes in. Dunning describes how he sees the main insight in this way:

The Dunning-Kruger result is a little complicated because it’s actually many results. The one that is a meme is this idea: On any particular topic, people who are not experts lack the very expertise they need in order to know just how much expertise they lack. The Dunning-Kruger effect visits all of us sooner or later in our pockets of incompetence. They’re invisible to us because to know that you don’t know something, you need to know something. It’s not about general stupidity. It’s about each and every one of us, sooner or later.

You can be incredibly intelligent in one area and completely not have expertise in another area. We all know very smart people who don’t recognize deficits in their sense of humor or their social skills, or people who know a lot about art but may not know much about medicine. We each have an array of expertise, and we each have an array of places we shouldn’t be stepping into, thinking we know just as much as the experts. My philosopher friend and I call that “epistemic trespassing,” because you’re trespassing into the area of an expert. We saw this a lot during the pandemic. … I think it was Vernon Law, the baseball pitcher, who said that life is the cruelest teacher because it gives you the test before it provides the lesson.

Is the Dunning-Kruger effect just a statistical artifact?

The critique is that the Dunning-Kruger effect is a statistical artifact known as regression to the mean. People who are poor performers on a test can only overestimate themselves. Those who are high performers can only underestimate themselves, so it’s a measurement error, an artifact. We talk about that issue in the original article. We did a nine-study series investigating regression to the mean. Other people have done studies that call the artifact into question. The critique tends to focus on the first two studies of a four-study paper in 1999. I can’t dismiss the irony of people not taking into account the 25 years of research that have happened since.

Dunning discusses social norms like “do not insult other people” and, at least as a first approximation, “if someone tells us something, we’ve been taught to assume it’s true.” These rules function fairly well in person-to-person interactions, but not on social media.

 think what’s interesting about the internet and social media is that it takes us out of the setting where we learned all these politeness rules. Right here, you and I are having a conversation. We’re in a relationship. Twitter is not that. On Twitter, I proclaim something by posting, and you come along a few hours later and you proclaim. We’re not interacting, we’re proclaiming asynchronously. The kindness rules and the politeness rules are not in play.

My anthropologist friends remind me that every time a new communication technology comes around, such as the telegraph or telephone, there is a breakdown in social norms. Whatever politeness rules have been built up don’t yet apply to the new platform. We’re in the middle of that right now. I think what’s happening with social media is that we haven’t developed the politeness rules that we have for face-to-face interaction.

David A. Price interviews Ulrike Malmendier, “On law versus economics, the long-term effects of inflation, and the remembrance of crises past” (Econ Focus: Federal Reserve Bank of Richmond, First/Second Quarter 2024, pp. 22-26). One theme of the interview is Malmendier’s recent work which emphasizes that living through a salient event can leave a lasting mark.

I mentioned how my early life path was influenced by my dad experiencing World War II and how everything can get destroyed — the house gets destroyed, you lose all your possessions and savings, and maybe your country’s currency isn’t worth anything anymore. One way of looking at the effects of this is simply in terms of information: After such an experience, you have new data about what can happen. That’s the traditional economic view. But I’d argue that there’s an element beyond the intellectual. When it’s your own life, you tend to put a lot of weight on what has happened to you. You’re pushed toward overweighing outcomes that have happened to you. I first worked on that in the context of the stock market, with a paper Stefan Nagel and I wrote on Depression babies in the U.S. We showed that people who experience big crashes of the stock market tend to shy away for years and decades from investing anything in the stock markets.

How might this experience of overweighting the past be playing out since the pandemic? One possibility is in how younger people, who had not previously experienced inflation, think about inflation.

For starters, look at inflation, which started creeping up since 2021, and then in 2022 you were getting close to the double digits. There was such a sharp contrast between the long period of the Great Moderation and all of a sudden that price shock kicking in. For older people, who have seen high inflation before in the ’80s or even the ’70s, I’m predicting they’re just taking that into the average of the long period of low inflation since the early 1980s and of their experience of high inflation in the 1970s and early 1980s. Given their long history of experiences, the new spike does not get too much weight. It just goes up a bit.

But for young people in the United States who basically had seen no inflation at all outside of textbooks, it’s a different story. All of their life before they had experienced very low inflation, and then all of a sudden there’s the spike. Initially, then, they might be a little slow to react. But if the spike in inflation lasts long enough — it isn’t just a two-month blip — they realize, whoa, the world I live in is different than the world I thought I was living in, where high inflation happens only in textbooks. So the weight they put on that experience increases and can in fact end up being much higher than for older generations because the new experience makes up a much larger part of their lives after it has happened for two years or so.

Another possibility relates to the question of why worker who were willing to come into the office five days every week have now become unwilling to do so.

One area where I do expect big experience effects from recent years is living through the COVID-19 crisis and many of us being relegated to working from home. I do expect there to be a lasting change in how we view the value of social interaction, the value of working from home versus working at your workplace. The leadership here at the Haas School of Business, where I am right now, is encountering exactly this issue. They wonder why the same people who were happily coming in five days a week before COVID absolutely refuse to do so now. It’s clearly an experience that has changed people. In the classical economic model, you would just talk about the information obtained from that experience and maybe the setup cost of learning Zoom. But that can’t explain everything. We knew the length of our commutes before COVID.

And yet, personally experiencing what remote work and cutting out your commute means for your personal life makes an enormous difference. You have to experience it first, not because of lack of information, not because you cannot add and subtract hours spent in the car versus not, but because it just enters your decision-making differently if you have physically experienced it.

In current work, Malmendier is looking at how the experience of being a CEO through a period of corporate success or failure leaves a mark on that person–and finds that being CEO during a period of corporate turmoil can literally take years off a person’s life. In a past essay back in 2015 for the Journal of Economic Perspectives (where I work as Managing Editor), Malmendier (with Geoffrey Tate) “Behavioral CEOs: The Role of Managerial Overconfidence.” Although she was not emphasizing the theme of how past experiences affect current judgments at the time, it seems obvious to wonder if those who have the past record to end up as CEOs may develop an inflated opinion of their own judgment and skills as they move forward.

Over the last half-century or so, the share of corporate stock that is owned by investors with taxable mutual funds or brokerage accounts has fallen dramatically. Steven M. Rosenthal and Livia Mucciolo tell the story in “Who’s Left to Tax? Grappling With a Dwindling Shareholder Tax Base” (Tax Notes, April 1, 2024).

Here’s their figure showing a breakdown of who owns stock in US publicly traded corporations. Back in the 1960s, 80% of this ownership was in the form of taxable accounts. But the share of US corporate stock held by foreign investors and retirement accounts has risen substantially, and nonprofits own a chunk of US corporate stocks as well. So in the last two decades, only 20-30% of US corporate stock is in taxable accounts.

Rosenthal and Mucciolo offer some additional discussion of how these groups are taxed. For example, dividends paid by US firms are taxable, even when paid to foreign investors, but these payments are governed by international treaties. They explain: “However, the rate is often reduced by tax treaties between the United States and the home country of the foreign investor: from 30 percent to 15 percent on portfolio investment dividends, for example, and 5 percent or even 0 percent on dividends from direct investments.” Foreign investors do not pay capital gains on stocks to the US government–instead, such gains are taxable in their home country. If US firms use the increasingly common practice of distributing funds to their investors by repurchasing their shares, then such payments are treated as capital gains, not dividends.

For retirement accounts, the common practice is that the money is not taxed when it goes into the account, and the returns are not taxed as they occur over time. Instead, retirement money is taxed as income to the taxpayer when it is received after retirement. Nonprofit, of course, are not subject to income taxes.

With these patterns in mind, proposals for taxing owners of corporate stock as a group–not just the minority who hold their investments in taxable brokerage and mutual fund accounts–are going to run into complexities. Dramatic changes in retirement accounts or international tax treaties are not a simple matter, in politics or economics. Jacking up taxes on the 20-30% of shareholders who are taxable would created incentives to push their share even lower. One can make an argument that a reason for an explicit tax on corporate income is that it has become so difficult to tax the gains to shareholders of those firms.

The authors describe the challenges without trying to spell out policy recommendations. They note: “The transformation over the past 60 years in the nature of U.S. stock ownership from overwhelmingly domestic taxable accounts to overwhelmingly foreign and tax-exempt investors has many important policy implications, including how we can most effectively tax corporate profits; who is affected by changes in corporate taxation; and the form of corporate payouts to shareholders. Policymakers must continue the process, only now beginning, of grappling with the dwindling shareholder tax base.”

Statistics are not reality, but they are a map to reality, and that map is central to the basic knowledge needed for modern government. Or at least so argues Michel Foucault in Security, Territory, and Population: Lectures at the College du France, 1977-1978 (edited by Michel Senellart, translated by Graham Burchell, originally published 2004, translation into English published in 2007). For example, he argues: “[T]his knowledge of things that comprise the very reality of the state is precisely what at the time was called `statistics.’ Etymologically, statistics is knowledge of the state, of the forces and resources that characterize the state at a given moment.”

Here’s a passage from Foucault’s lecture of March 15, 1978. Hat tip: I was introduced to this essay by a LinkedIn post from Noah Williams at the University of Miami. My previous readings of Foucault did not take me this deep into his writings!)

[A]t the level of content, what must be known in order to be able to govern? I think we see an important phenomenon here, an essential transformation. In the images, the representation, and the art of government as it was defined up to the start of the seventeenth century, the sovereign essentially had to be wise and prudent. What did it mean to be wise? Being wise meant knowing the laws: knowing the positive laws of the country, the natural laws imposed on all men, and, of course, the commandments of God himself. Being wise meant knowing the historical examples, the models of virtue, and making them rules of behavior. On the other hand, the sovereign had to be prudent, that is to say, to know in what measure, when, and in what circumstances it was actually necessary to apply this wisdom. When, for example, should the laws of justice be rigorously applied, and when, rather, should the principles of equity prevail over the formal rules of justice? Wisdom and prudence, that is to say, in the end an ability to handle laws.

At the start of the seventeenth century I think we see the appearance of a completely different description of the knowledge required by someone who governs. What the sovereign or someone who governs, the sovereign inasmuch as he governs, must know is not just the laws, and it is not even primarily or fundamentally the laws (although one always refers to them, of course, and it is necessary to know them). What I think is new, crucial, and determinant is that the sovereign must know those elements that constitute the state … That is to say, someone who governs must know the elements that enable the state to be preserved in its strength, or in the necessary development of its strength, so that it is not dominated by others or loses its existence by losing its strength or relative strength. That is to say, the sovereign’s necessary knowledge (savoir) will be a knowledge (connaissance) of things rather than knowledge of the law, and this knowledge of things that comprise the very reality of the state is precisely what at the time was called “statistics.” Etymologically, statistics is knowledge of the state, of the forces and resources that characterize the state at a given moment. For example, knowledge of the population, the measure of its quantity, mortality, natality; reckoning of the different categories of individuals in a state and of their wealth; assessment of the potential wealth available to the state, mines and forests, etcetera; assessment of the wealth in circulation, of the balance of trade, and measure of the effects of taxes and duties, all this data, and more besides, now constitute the essential content of the sovereign’s knowledge. So, it is no longer the corpus of laws or skill in applying them when necessary, but a set of technical knowledge that describes the reality of the state itself.

Walter Walter O’Leary, a Managing Partner at South Pointe Capital, and a colleague of Noah Williams at the University of Miami, pointed out the origins of the terminology of “statistics from the Online Etymology Dictionary:

1770, “science dealing with data about the condition of a state or community” [Barnhart], from German Statistik, popularized and perhaps coined by German political scientist Gottfried Achenwall (1719-1772) in his “Vorbereitung zur Staatswissenschaft” (1748), from Modern Latin statisticum (collegium) “(lecture course on) state affairs,” from Italian statista “one skilled in statecraft,” from Latin status “a station, position, place; order, arrangement, condition,” figuratively “public order, community organization,” noun of action from past-participle stem of stare “to stand” (from PIE root *sta- “to stand, make or be firm”).

OED points out that “the context shows that [Achenwall] did not regard the term as novel,” but current use of it seems to trace to him. Sir John Sinclair is credited with introducing it in English use.

The broader meaning “numerical data of any sort collected and classified systematically” is from 1829; hence the study of any subject by means of extensive enumeration. Abbreviated form stats is recorded by 1961.

This early notion of statistics as statecraft is (of course) appealing to me. Indeed, it helped me to crystalize one form of my discontent with how modern politics is often practiced. I would probably be uncomfortable with a government should be run by economists and other technocrats. But I would like to feel that a larger share of politicians have more than a passing and outdated acquaintance with the statistics that provide a map of “the forces and resources that characterize the state in a given moment.”

For more than a half-century, a UCLA-based research group has been carrying out surveys of incoming first-year college students. There are lots of questions about the decision process the students went through in applying, and about their expectations and priorities. The data tables from the 2022 survey, from the Higher Education Research Institute at UCLA, are available here.

I’ll focus here on a single question, which asks about what life objectives are important. The first column shows the overall answer: the other two columns are divided into answers from males and females. (These figures are taken from several different tables: for those who want to dig deeper, the underlying tables offer a number of other breakdowns.)

It’s worth remembering that survey responses are always a mixture of what the person actually believes and what they feel is a desired or appropriate answer. With that noted, it’s interesting to consider some of the gender gaps here. For example, incoming first-year college students who are male are notably more likely to list “raising a family” and “becoming successful in a business of my own” as essential or very important. Females are notably more likely to emphasize “working to achieve greater gender equity,” along with a variety of other social goals like “working to correct economic inequalities,” “working to correct social inequalities,” “improving my understanding of other cultures and countries,” “heling to promote racial understanding,” participating in a community action program,” “keeping up to date in political affairs,” “becoming involved in programs to clean up the environment,” and “helping others who are in difficulty.”

With that difference in mind, it’s interesting “being very well off financially” is by far the highest value for both male and female incoming first-year students (although I’m not sure how the overall average can be lower than it is for males and females taken separately). And it’s interesting that despite the emphasis on social goals in female responses, males are slightly more likely to emphasize the goal of “developing a meaningful philosophy of life.”

One substantial swing over the decades has been in these two answers concerning “being well-off financially” and “developing a meaningful philosophy of life. Back in the 1985 survey, the reporting of the survey includes this figure. In the 1966 survey, develop a meaningful philosophy of life was a high priority for a much larger share of first-year college students. But priorities shift, the lines cross in the late 1970s, and by the 1980s “be financially very well off” is well in the lead.

What’s going on here? Here are some thoughts and reactions

1) One reason is a dramatic shift in the gender mix. Back in the 1966 survey, only 31.6% of females listed “be financially very well off” as a top priority, compared with 54.1% of men.

2) The US economy suffered “stagflation” of high inflation and unemployment in the 1970s, which probably made financial concerns more salient. In recent decades, the Great Recession and the pandemic recession have kept financial concerns salient.

3) The cost of higher education has risen dramatically. As I’ve pointed out in the past, when I was thinking about college in the late 1970s, I had a lot of friends attending the big local state university–in my case, the University of Minnesota. At that time, it was possible to cover all of U of M tuition and a share of living expenses by working at a minimum-wage job full-time over the summer and 10 hours/week during school. That’s no longer even close to true at the University of Minnesota, much less at the pricier private colleges and universities. When college and universities have a high price, students are going to become more focused on financial goals.

4) The share of high school students who go immediately to a post-secondary program in the next year was around 45% in the 1960s. Before the pandemic, it had reached nearly 70%, before dropping off. My guess is that a substantial share of this expansion of enrollment was from people who were more interested in economic goals than in a “meaningful philosophy of life.”

5) The priorities of students will shape the intellectual climate of a college or university.

6) It’s interesting to me that the survey question doesn’t just ask about being “well-off,” but about being “very well-off.” There will be a distribution of economic outcomes. Perhaps being in the middle of that distribution–say, from the 40th to the 60th percentile–can be counted as “well-off.” But when people talk about being “very well-off,” it seems to me that they are thinking about being in the upper reaches of economic outcomes. It is statistically not possible for all college students to end up in the upper reaches of economic outcomes. Developing a philosophy of life that you find meaningful is possible for everyone; in contrast, making 85% of college and university students very well-off is statistically impossible.

The widespread current belief about globalization, I would say, is that it has been in decline since the Trump presidency, as a result of increased tariffs, a sharpening of global conflicts with China and Russia, and disruptions from the pandemic. But even with this perceived decline, a common belief is that globalization is an overwhelming force in both the US and global economy. Both of these beliefs are likely incorrect. Steven A. Altman and Caroline R. Bastian provide evidence and discussion in the DHL Global Connectedness Report 2024.

Altman and Bastian argue that globalization, fully understood, reached an all-time high in 2022 and remained near that level in 2023. They look at a variety of globalization trends. For example, this figure shows patterns of globalization n goods, investment, migration, and travel.

As you might expect, international growth in data flows has been remarkable: “The amount of data crossing national borders over the internet has nearly tripled since 2019, fueling dramatic increases in international information flows.” However, as the report points out, growth in data flows has also been very rapid within countries, so it’s not clear if the international share of internet traffic is rising.

Some of this growth in internet traffic is being driven by global commerce in service industries, where jobs are performed in one country and the work produce is transmitted digitally.

Information about science, intellectual property, and patents is globalizing as well.

An underlying theme here is that people sometimes talk about “globalization” as if it was purely a policy choice, and fully under the control of political actors. The rules about globalization make a difference, of course, but globalization is also driven by economic forces. In particular, the digital revolution has made it much easier to hear about distant products, to manage far-flung supply chains, to have services provided elsewhere, and to cooperate in matters of science and innovation. In some ways, the physical disruptions of the pandemic supercharged the importance of these digital connections.

But despite the growth of globalization to at or near all-time highs, it is paradoxically true that the extent of globalization is often overstated. Especially for the world’s largest economies, like the United States, what happens in the economy is overwhelmingly influenced by domestic factors. Even for the world economy as a whole, most flows are primarily domestic.

One might rejoice or lament in the extent of globalization. That’s a topic for another day. But either way, it’s useful to see the extent and trends of globalization more clearly.

Certain basic investment models are based on just two investment options: a safe asset like US Treasury bonds and a risky asset like a stock-market index fund. The underlying idea is that you can choose the riskiness of your preferred portfolio by having a larger or smaller share of the safe asset. For example, a person who reaches retirement can decide to take less risk by reducing what share of their portfolio is invested in stocks.

But where corporate bonds fit in this scenario? There is something like $66 trillion outstanding in corporate debt around the world. Some of that debt is highly rated “investment grade” bonds, which in some ways resemble government debt–that is, safe but with lower interest rates. Another part is high-yield debt, sometimes called “junk bonds,” which seems closer to equities in having more risk and higher returns. (A professor of mine used to refer to “junk bonds” as “equity in drag.”) How do corporate bonds fit in the eco-system of finance?

Elroy Dimson, Paul Marsh, Mike Staunton offer a long-run view in “Corporate bonds and the credit premium.” It appears in the publicly available part of the UBS Global Investment Returns Yearbook 2024, which is subtitled: “Leveraging deep history to navigate the future.” They begin with a wry comment: “Traditionally, bonds have been seen as boring, relative to stocks. In choosing the name James
Bond, Ian Fleming said, `I wanted the simplest, dullest, plainest-sounding name I could find.’”

However, there’s big money involved in bonds: “[D]ebt securities worldwide have a value of some USD 136 trillion compared with around USD 100 trillion for global
equities. The debt total comprises some USD 70 trillion in government debt and USD 66 trillion of debt securities issued by corporations. Of this amount, corporate bonds account for around USD 45 trillion, the remainder being other corporate issues.”

Here’s a figure showing the distribution of debt securities around the world. The US dominates corporate bond markets in part because of the sheer size of its economy, but also because of the depth of its financial markets. Also, the role of corporate bonds is different in the US economy than in many other countries. In “bank-centered” economies, corporations often have a very close ties (including cross-ownership) to one or several banks, and thus can raise money through those connections. In contrast, US financial markets are more likely to push companies that want to raise money to go “to the market,” and raise funds from outside investors.

What do returns on corporate bond investments look like over time? The figure shows returns on US Treasury bonds as the light-gray line, and returns US corporate bonds as the dark-gray line. The red line shows the gap between the two: that is, what is the extra return or “premium” that an investor gets for taking on the extra risk of corporate bonds: “”Over the 158 years spanned by Figure 77, the credit spread has averaged 1.58% with a standard deviation of 0.73%. The lowest spread was 0.42% in 1965, while the highest was 4.53% in 1931.”

Corporate bonds have higher return, because of their higher risk. As a result, they will outperform safe US Treasuries over the long run

One obvious risk for corporate bonds, as compared to US Treasury bonds, is that corporations are more likely to default than the US government. Of course, a default doesn’t mean that investors lose 100% of the value, but they can often lose half or ever more. However, the share of the total value of US nonfinancial corporate bonds ending up in default has been declining over time.

To return to the question at the top, what role should corporate bonds be playing for investors? In a situation with a safe asset like Treasury bonds and a risky asset like stock market index funds, are corporate bonds more-0r-less superfluous? The UBS authors emphasize that they are not offering advice about future investing. Like the advertisements for investing are required to say: “Past performance is no guarantee of future results.”

However, they do cite several studies that looked back in time and calculated what would have been an optimal portfolio if you could choose between government debt, corporate debt, and stocks. Looking back at the past, at least, the optimal long-run portfolio would have includes corporate bonds. Indeed, one study found that over the long run, the optimal portfolio from 1936-2014 would have included a larger share of corporate bonds than either government debt or stocks.

What about strategies to make money in bonds short-term? Again, the authors are careful to point out that it’s easy to point to the investments that would have worked in the past. But especially after such strategies have been publicized, there’s no guarantee that the strategy will continue to work in the future. But for completeness, the authors also mention a few limited and intriguing exceptions. One involves “fallen angels,” which is corporate debt that was once highly-rated as investment-grade safe, but now its credit rating has declined to the point that it has become a high-yield “junk bond.” Here’s the dynamic that can unfold:

Most mandates for IG [investment-grade]corporate bond managers require them to sell bonds within a relatively short time-span if they get downgraded from IG to HY [high yield] status (typically bonds being downgraded to BB). These bonds are commonly referred to as “fallen angels”. The need for a substantial number of
investors to divest within a limited window appears to have created price pressures that temporarily reduce prices below their fair values. Historically, it has been profitable to buy these fallen angels. Ben Dor et. al. (2021) analyzed the fallen angels effect and report an extensive pattern of strong price reversals. Their results suggest that investors start selling in anticipation of the downgrade before it happens and continue to sell until around three months after. The price falls are then reversed and fallen angels outperform by a total of 6.6% in the two-years post downgrade. The greater the initial under-performance, the higher the subsequent returns. They conclude that this is due to price pressure rather than an overreaction
to the information implied by the downgrade.

This “fallen angel” price dynamic has been noted for some years now even after being published in the literature, but at last check, it does not yet seem to have disappeared. Of course, it could easily disappear in the future.