Macro Roundup

State of play of autonomous driving technology"...In contrast to investors and CEOs, academics who study artificial intelligence, systems engineering and autonomous technologies have long said that creating a fully self-driving automobile would take many years, perhaps decades. Now some are going further, saying thatdespite investments already topping $80 billion, we may never get the self-driving cars we were promised. At least not without major breakthroughs in AI, which almost no one is predicting will arrive anytime soon—or a complete redesign of our cities. A recently published paper called “Why AI is Harder Than We Think” sums up the situation nicely. In it, Melanie Mitchell, a computer scientist and professor of complexity at the Santa Fe Institute, notes that as deadlines for the arrival of autonomous vehicles have slipped, people within the industry are redefining the term. To gauge today’s machine-learning systems, she developed a four-level scale of AI sophistication. The simplest kind of thinking starts with skill-based “bottom-up” reasoning. Today’s AIs are quite good at things like teaching themselves to stay within lines on a highway. The next step up is rule-based learning and reasoning (i.e., what to do at a stop sign). After that, there’s knowledge-based reasoning. (Is it still a stop sign if half of it is covered by a tree branch?) And at the top is expert reasoning: the uniquely human skill of being dropped into a completely novel scenario and applying our knowledge, experience and skills to get out in one piece. Problems with driverless cars really materialize at that third level. Today’s deep-learning algorithms, the elite of the machine-learning variety, aren’t able to achieve knowledge-based representation of the world, says Dr. Cummings. And human engineers’ attempts to make up for this shortcoming—such as creating ultra-detailed maps to fill in blanks in sensor data—tend not to be updated frequently enough to guide a vehicle in every possible situation, such as encountering an unmapped construction site. Machine-learning systems, which are excellent at pattern-matching, are terrible at extrapolation—transferring what they have learned from one domain into another. For example, they can identify a snowman on the side of the road as a potential pedestrian, but can’t tell that it’s actually an inanimate object that’s highly unlikely to cross the road. Small, low-speed shuttles working in well-mapped areas, bristling with sensors such as lidar, could allow engineers to get the amount of uncertainty down to a level that regulators and the public would find acceptable. (Picture shuttles to and from the airport, driving along specially constructed lanes, for example.)

Christopher Mims, "Self-Driving Cars Could Be Decades Away, No Matter What Elon Musk Said,"Wall Street Journal, June 5, 2021, https://www.wsj.com/articles/self-driving-cars-could-be-decades-away-no-matter-what-elon-musk-said-11622865615

Self-Driving Cars Could Be Decades Away, No Matter What Elon Musk Said

In 2015, Elon Musk said self-driving cars that could drive “anywhere” would be here within two or three years.

In 2016, Lyft CEO John Zimmer predicted they would “all but end” car ownership by 2025.

In 2018, Waymo CEO John Krafcik warned autonomous robocars would take longer than expected.

In 2021, some experts aren’t sure when, if ever, individuals will be able to purchase steering-wheel-free cars that drive themselves off the lot.

In contrast to investors and CEOs, academics who study artificial intelligence, systems engineering and autonomous technologies have long said that creating a fully self-driving automobile would takemany years, perhaps decades. Now some are going further, saying that despiteinvestments already topping $80 billion, we may never get the self-driving cars we were promised. At least not without major breakthroughs in AI, which almost no one is predicting will arrive anytime soon—or a complete redesign of our cities.

Even those who have hyped this technology most—in 2019 Mr. Musk doubled down on previous predictions, and said that autonomous Tesla robotaxiswould debut by 2020—are beginning to admit publicly that naysaying experts may have a point.

“A major part of real-world AI has to be solved to make unsupervised, generalized full self-driving work,” Mr. Musk himselfrecently tweeted. Translation: For a car to drive like a human, researchers have to create AI on par with one. Researchers and academics in the field will tell you that’s something we haven’t got a clue how to do. Mr. Musk, on the other hand, seems to believe that’s exactly what Tesla will accomplish. Hecontinually hypesthe next generation of the company’s “Full Self Driving” technology—actually a driver-assist system with a misleading name—which is currently in beta testing.

A recently published paper called “Why AI is Harder Than We Think” sums up the situation nicely. In it, Melanie Mitchell, a computer scientist and professor of complexity at the Santa Fe Institute, notes that as deadlines for the arrival of autonomous vehicles have slipped, people within the industry are redefining the term. Since these vehicles require a geographically constrained test area and ideal weather conditions—not to mention safety drivers or at least remote monitors—makers and supporters of these vehicles have incorporated all of those caveats into their definition of autonomy.

Even with all those asterisks, Dr. Mitchell writes, “none of these predictions has come true.”

In vehicles you can actually buy, autonomous driving has failed to manifest as anything more than enhanced cruise control, like GM’s Super Cruise or the optimistically named Tesla Autopilot. In San Francisco, GM subsidiary Cruise is testing autonomous vehicles with no driver behind the wheel but a human monitoring the vehicle’s performance from the back seat. And there’s only one commercial robotaxi service operating in the U.S. with no human drivers at all, a small-scale operation limited to low-density parts of the Phoenix metro area, from Alphabet subsidiary Waymo.

Even so, Waymo vehicles have been involved in minor accidents in which they were rear-ended, and their confusing (to humans) behavior was cited as a possible cause. Recently, one was confused by traffic cones at a construction site.

“I am not aware we are struck or rear-ended any more than a human driver,” says Nathaniel Fairfield, a software engineer and head of the “behavior” team at Waymo. The company’s self-driving vehicles have been programmed to be cautious—“the opposite of the canonical teenage driver,” he adds.

Chris Urmson is head of autonomous trucking startup Aurora, which recently acquired Uber’s self-driving division. (Uber also invested $400 million in Aurora.) “We’re going to see self-driving vehicles on the road doing useful things in the next couple of years, but for it to become ubiquitous will take time,” he says.

Key to Aurora’s initial rollout will be that it will only operate on highways where the company has already created a high-resolution, three-dimensional map, says Mr. Urmson. Aurora’s eventual goal is for both trucks and cars using its systems to travel farther from the highways where it will at first be rolled out, but Mr. Urmson declined to say when that might happen.

The slow rollout of limited andconstantly human-monitored“autonomous” vehicles was predictable, andeven predicted, years ago. But some CEOs and engineers argued that new self-driving capabilities would emerge if these systems could just log enough miles on roads. Now, some are taking the position that all the test data in the world can’t make up for AI’s fundamental shortcomings.

Decades of breakthroughs in the part of artificial intelligence known as machine learning have yielded only the most primitive forms of “intelligence,” says Mary Cummings, a professor of computer science and director of the Humans and Autonomy Lab at Duke University, who has advised the Department of Defense on AI.

To gauge today’s machine-learning systems, she developed afour-level scaleof AI sophistication. The simplest kind of thinking starts with skill-based “bottom-up” reasoning. Today’s AIs are quite good at things like teaching themselves to stay within lines on a highway. The next step up is rule-based learning and reasoning (i.e., what to do at a stop sign). After that, there’s knowledge-based reasoning. (Is it still a stop sign if half of it is covered by a tree branch?) And at the top is expert reasoning: the uniquely human skill of being dropped into a completely novel scenario and applying our knowledge, experience and skills to get out in one piece.

Problems with driverless cars really materialize at that third level. Today’s deep-learning algorithms, the elite of the machine-learning variety, aren’t able to achieveknowledge-based representationof the world, says Dr. Cummings. And human engineers’ attempts to make up for this shortcoming—such as creatingultra-detailed mapsto fill in blanks in sensor data—tend not to be updated frequently enough to guide a vehicle in every possible situation, such as encountering an unmapped construction site.

Machine-learning systems, which are excellent at pattern-matching, areterrible at extrapolation—transferring what they have learned from one domain into another. For example, they can identify a snowman on the side of the road as a potential pedestrian, but can’t tell that it’s actually an inanimate object that’shighly unlikely to cross the road.

“When you’re a toddler, you’re taught the hot stove is hot,” says Dr. Cummings. But AI isn’t great at transferring the knowledge of one stove to another stove, she adds. “You have to teach that for every single stove that’s in existence.”

Some researchers at MIT are trying to fill this gap by going back to basics. They have launched ahuge effortto understand how babies learn, in engineering terms, in order to translate that back to future AI systems.

“Billions of dollars have been spent in the self-driving industry and they are not going to get what they thought they were going to get,” says Dr. Cummings. This doesn’t mean we won’t eventually get some form of “self-driving” car, she says. It just “won’t be what everybody promised.”

But, she adds, small, low-speed shuttlesworking in well-mapped areas, bristling with sensors such as lidar, could allow engineers to get the amount of uncertainty down to a level that regulators and the public would find acceptable. (Picture shuttles to and from the airport, driving along specially constructed lanes, for example.)

Mr. Fairfield of Waymo says his team sees no fundamental technological barriers to making self-driving robotaxi services like his company’s widespread. “If you’re overly conservative and you ignore reality, you say it’s going to take 30 years—but it’s just not,” he adds.

A growing number of experts suggest that the path to full autonomy isn’t primarily AI-based after all. Engineers have solved countless other complicated problems—including landing spacecraft on Mars—by dividing the problem into small chunks, so that clever humans can craft systems to handle each part. Raj Rajkumar, a professor of engineering at Carnegie Mellon University with a long history of working on self-driving cars, is optimistic about this path. “It’s not going to happen overnight, but I can see the light at the end of the tunnel,” he says.

This is the primary strategy Waymo has pursued to get its autonomous shuttles on the road, and as a result, “we don’t think that you need full AI to solve the driving problem,” says Mr. Fairfield.

Mr. Urmson of Aurora says that his company combines AI with other technologies to come up with systems that can apply general rules to novel situations, as a human would.

Getting to autonomous vehicles the old-fashioned way, withtried-and-true“systems engineering,” would still mean spending huge sums outfitting our roads with transponders and sensors to guide and correct the robot cars, says Dr. Mitchell. And they would remain limited to certain areas, and certain weather conditions—with human teleoperators on standby should things go wrong, she adds.

This Disney animatronic version of our self-driving future would be a far cry from creating artificial intelligence that could simply be dropped into any vehicle, immediately replacing a human driver. It could meansafer human-driven cars, and fully autonomous vehicles in a handful of carefully monitored areas. But it would not be the end of car ownership—not anytime soon.

Ian Goldin, Pantelis Koutroumpis, François Lafond and Julian Winkler, "Why is Productivity slowing down?" Oxford University, May 9, 2021, https://www.oxfordmartin.ox.ac.uk/downloads/academic/ProductivitySlowdown.pdf

Their takeaways, “….Comparing the period after 2005 to the 1995-2005 decade, there is a generalised slowdown in productivity in advanced economies, with labor productivity growth falling from around 2% to less than 1%. While part of this decline may be due the end of convergence and a return to “normal” rates of growth at the frontier, it remains puzzling in view of its scale, and considering the context of continued deployment of new digital and other technologies. The literature points to wide and varied reasons, which our review has sought to systematically evaluate. We identified a small set of forces and mechanisms that taken together explain most of the slowdown. Table 11 summarizes the results, although it is limited to the explanations that we have evaluated quantitatively Accelerated mismeasurement, which is mostly due the inherent difficulty of measuring aggregate price changes when creative destruction and quality change are important, contributed less than 15% of the slowdown in the US, although there are large uncertainties associated with this estimate. The reduction in the contribution of capital deepening contributed almost 45% of the slow down in the US.We did not evaluate the relative contributions of specific underlying causes quantitatively, but, considering the literature on the slowdown of investment, we make a rough estimate that the two categories of factors contributed about equally. The first set of factors, which we call Financial Crisis, includes weak demand and credit constraints during, and to some extent in and after the financial crisis. The second set of factors, which we group under the name “secular decline of investment”, recognizes that investment may have been weak due to more structural changes, such as the increasing share of intangibles and globalization, with additional hypotheses including changes to corporate governance, and the weakening of competition. There is also a recognition that intangible investment, due to its higher potential for spillovers, may affect labor productivity through TFP, that is, above and beyond its impact through capital deepening. Reusing published elasticities, we find that the slowdown in intangible assets accumulation may have had a substantial effect on the TFP slowdown. Conventional growth accounting finds almost no role for a decline of human capital accumulation in the US, and a weak role at best elsewhere. But several labor-related mechanisms may have affected TFP, including aging and labor market institutions A key feature of the last two decades is the fast growth of global trade after 1995, and its collapse during the financial crisis. Any positive effect of trade would therefore translate into a productivity slowdown, and we do indeed find substantial effects. A large part of the current discussion on the productivity slowdown centers on business dynamism and competition, documenting trends such as declining entry-exit, increasing markups and concentration, an increasing divergence of the most productive firms. There is no consensus in the literature as to whether these trends are intrinsically good or bad for productivity and welfare, as concentration can reflect output-restricting dominant positions or a better allocation of resources to highly productive firms in an economy where firms are increasingly operating under increasing returns to scale, due in particular to the increasing prevalence of intangible investment. We do not attempt to settle this debate, but to reflect the potential importance of these trends we use recently developed estimates of the contribution of allocative efficiency to TFP growth that are calibrated on firm-level markup time series. We find that increasing allocative efficiency was a stronger contributor to TFP growth in the decade prior to 2006, implying that the lower contribution of increasing allocative efficiency to TFP growth is a substantial contribution to the TFP slowdown…”

6) Tech change

5) Dynamism, rise in market power

4) Impact of trade, this is basically the China Shock as supply chains have shifted away from higher income countries

3) Human capital, changes in populations demographics, included educational and skill levels

2) Lower rate of capital deepening per workers, they look at two effects a cyclical driver related to the crisis (so lower AD, credit constraints) and secular factors that are independent of business cycle (so the rise of intangibles, and a continued shift to foreign investments over domestic investment

1) Output is increasingly mismeasured. “…Mismeasurement of labor productivity can be due to three main sources: a mismeasurement of nominal output (perhaps due to changing boundaries of GDP), a mismeasurement of deflators, which has received the most attention in the literature, and a mismeasurement of labor inputs….potential biases in deflators. The main sources of bias include: issues with sampling and measuring prices and relative weights of items in consumption (or other final demand) baskets; issues with aggregating low-level price changes into indices, in view of the difficulty of assessing whether the shares of each item in the baskets are changing because of substitution induced by changes to relative prices; issues with the addition of new products and removal of disappearing products; issues with assessing quality change; more broadly, issues with new forms of commercialization (i.e. new retail outlets)….”

They offer a useful list of theories for the slowdown, though none that we are unfamiliar with:

In terms of TFP they largely cite some of Baqaee & Farhi work that decompose TFP into an “allocative efficiency” and a technology component, where allocative efficiency is ~ markups“…They find that allocative efficiency contributed around half of TFP growth between 1997 and 2014;we compute that it also contributed to roughly half of its slowdown between 1997-2005 and 2006-2014….”

Impact of capital deepening, “…We find that a decline in the rate of capital deepening has contributed to the slowdown, mostly driven by non-ICT physical capital, but also ICT capital and Intangibles. We distinguish two core arguments to explain this phenomenon. The first relates to the financial crisis, and suggests that the decline in investment is a cyclical phenomenon driven by financial constraints and weak aggregate demand. A second candidate explanation recognizes that the slowdown started before the crisis, so that structural factors may have been more important, including primarily a change in the composition of capital towards intangibles (which are riskier), but also lower competition, increasing short-termism, and the off-shoring of physical investment in the context of increasingly global value chains. We are unable to derive relative contributions for all of these factors, but as a whole we estimate that capital deepening explains 44% of the labor productivity slowdown in the US. In addition to this, we report evidence of the slowdown in intangible capital services, which it is argued to contribute to lower TFP growth…”

Trends in US and comps, “.. First, there was no convergence in the 1995-2005 decade, so slower convergence cannot explain the slower rates in Europe after 2005 compared to 1995-2005. Second, while it is true that the frontier (US) may be returning to more “normal” rates of growth after the 1995-2005 IT boom, the rates of labor productivity growth in all countries are still broadly speaking lower than at any time in the 20th century, and are low in all the 5 countries we study at the same time. Thus, while we see part of slowdown as simply reflecting the fact that all 5 advanced economies are now more or less frontier economies progressing at a “normal” rate, the rates observed are very low by historical standards and appear surprising in a context of technological transformations. That said, throughout the paper, we examine the role of both secular trends, such as aging and structural change, and cyclical or market phenomena, such as declines in investment…”

Note the paper is largely a lit review

Though you would want to see a more detailed writeup of the Oxford paper that was included in the morning roundup, on the slowdown of labor productivity and makes an effort to break down drivers for the US and some comps (France, Germany, Japan, UK), “….Comparing the post-2005 period with the preceding decade for 5 advanced economies, we seek to explain a slowdown of 0.8 to 1.8pp. We trace most of this to lower contributions of TFP and capital deepening, with manufacturing accounting for the biggest sectoral share of the slowdown. No single explanation accounts for the slowdown, but we have identified a combination of factors which taken together account for much of what has been observed….”

Jesse Eisinger, Jeff Ernsthausen and Paul Kiel, "The Secret IRS Files: Trove of Never-Before-Seen Records Reveal How the Wealthiest Avoid Income Tax,"Propublica, June 8, 2021, https://www.propublica.org/article/the-secret-irs-files-trove-of-never-before-seen-records-reveal-how-the-wealthiest-avoid-income-tax

The Secret IRS Files: Trove of Never-Before-Seen Records Reveal How the Wealthiest Avoid Income Tax

In 2007, Jeff Bezos, then a multibillionaire and now the world’s richest man, did not pay a penny in federal income taxes. He achieved the feat again in 2011. In 2018, Tesla founder Elon Musk, the second-richest person in the world, also paid no federal income taxes.

Michael Bloomberg managed to do the same in recent years. Billionaire investor Carl Icahn did it twice. George Soros paid no federal income tax three years in a row.

ProPublica has obtained a vast trove of Internal Revenue Service data on the tax returns of thousands of the nation’s wealthiest people, covering more than 15 years. The data provides an unprecedented look inside the financial lives of America’s titans, including Warren Buffett, Bill Gates, Rupert Murdoch and Mark Zuckerberg. It shows not just their income and taxes, but also their investments, stock trades, gambling winnings and even the results of audits.

Taken together, it demolishes the cornerstone myth of the American tax system: that everyone pays their fair share and the richest Americans pay the most. The IRS records show that the wealthiest can — perfectly legally — pay income taxes that are only a tiny fraction of the hundreds of millions, if not billions, their fortunes grow each year.

Many Americans live paycheck to paycheck, amassing little wealth and paying the federal government a percentage of their income that rises if they earn more. In recent years, the median American household earned about $70,000 annually and paid 14% in federal taxes. The highest income tax rate, 37%, kicked in this year, for couples, on earnings above $628,300.

The confidential tax records obtained by ProPublica show that the ultrarich effectively sidestep this system.

America’s billionaires avail themselves of tax-avoidance strategies beyond the reach of ordinary people. Their wealth derives from the skyrocketing value of their assets, like stock and property. Those gains are not defined by U.S. laws as taxable income unless and until the billionaires sell.

To capture the financial reality of the richest Americans, ProPublica undertook an analysis that has never been done before. We compared how much in taxes the 25 richest Americans paid each year to how much Forbes estimated their wealth grew in that same time period.

We’re going to call this their true tax rate.

The results are stark. According to Forbes, those 25 people saw their worth rise a collective $401 billion from 2014 to 2018. They paid a total of $13.6 billion in federal income taxes in those five years, the IRS data shows. That’s a staggering sum, but it amounts to a true tax rate of only 3.4%.

It’s a completely different picture for middle-class Americans, for example, wage earners in their early 40s who have amassed a typical amount of wealth for people their age. From 2014 to 2018, such households saw their net worth expand by about $65,000 after taxes on average, mostly due to the rise in value of their homes. But because the vast bulk of their earnings were salaries, their tax bills were almost as much, nearly $62,000, over that five-year period.

That works out to a true tax rate of 0.1%, or less than 10 cents for every $100 he added to his wealth.

In the coming months, ProPublica will use the IRS data we have obtained to explore in detail how the ultrawealthy avoid taxes, exploit loopholes and escape scrutiny from federal auditors.

But few specifics about individuals ever emerge in public. Tax information is among the most zealously guarded secrets in the federal government. ProPublica has decided to reveal individual tax information of some of the wealthiest Americans because it is only by seeing specifics that the public can understand the realities of the country’s tax system.

Consider Bezos’ 2007, one of the years he paid zero in federal income taxes. Amazon’s stock more than doubled. Bezos’ fortune leapt $3.8 billion, according to Forbes, whose wealth estimates are widely cited. How did a person enjoying that sort of wealth explosion end up paying no income tax?

In 2011, a year in which his wealth held roughly steady at $18 billion, Bezos filed a tax return reporting he lost money — his income that year was more than offset by investment losses. What’s more, because, according to the tax law, he made so little, he even claimed and received a $4,000 tax credit for his children.

His tax avoidance is even more striking if you examine 2006 to 2018, a period for which ProPublica has complete data. Bezos’ wealth increased by $127 billion, according to Forbes, but he reported a total of $6.5 billion in income. The $1.4 billion he paid in personal federal taxes is a massive number — yet it amounts to a 1.1% true tax rate on the rise in his fortune.

ProPublica’s data shows that while some wealthy Americans, such as hedge fund managers, would pay more taxes under the current Biden administration proposals, the vast majority of the top 25 would see little change.

The tax data was provided to ProPublica after we published a series of articles scrutinizing the IRS. The articles exposed how years of budget cuts have hobbled the agency’s ability to enforce the law and how the largest corporations and the rich have benefited from the IRS’ weakness. They also showed how people in poor regions are now more likely to be audited than those in affluent areas.

ProPublica is not disclosing how it obtained the data, which was given to us in raw form, with no conditions or conclusions. ProPublica reporters spent months processing and analyzing the material to transform it into a usable database.

We then verified the information by comparing elements of it with dozens of already public tax details (in court documents, politicians’ financial disclosures and news stories) as well as by vetting it with individuals whose tax information is contained in the trove. Every person whose tax information is described in this story was asked to comment. Those who responded, including Buffett, Bloomberg and Icahn, all said they had paid the taxes they owed.

A spokesman for Soros said in a statement: “Between 2016 and 2018 George Soros lost money on his investments, therefore he did not owe federal income taxes in those years. Mr. Soros has long supported higher taxes for wealthy Americans.” Personal and corporate representatives of Bezos declined to receive detailed questions about the matter. ProPublica attempted to reach Scott through her divorce attorney, a personal representative and family members; she did not respond. Musk responded to an initial query with a lone punctuation mark: “?” After we sent detailed questions to him, he did not reply.

One of the billionaires mentioned in this article objected, arguing that publishing personal tax information is a violation of privacy. We have concluded that the public interest in knowing this information at this pivotal moment outweighs that legitimate concern.

The consequences of allowing the most prosperous to game the tax system have been profound. Federal budgets, apart from military spending, have been constrained for decades. Roads and bridges have crumbled, social services have withered and the solvency of Social Security and Medicare is perpetually in question.

There is an even more fundamental issue than which programs get funded or not: Taxes are a kind of collective sacrifice. No one loves giving their hard-earned money to the government. But the system works only as long as it’s perceived to be fair.

Our analysis of tax data for the 25 richest Americans quantifies just how unfair the system has become.

By the end of 2018, the 25 were worth $1.1 trillion.

For comparison, it would take 14.3 million ordinary American wage earners put together to equal that same amount of wealth.

The personal federal tax bill for the top 25 in 2018: $1.9 billion.

The bill for the wage earners: $143 billion.

The idea of a regular tax on income, much less on wealth, does not appear in the country’s founding documents. In fact, Article 1 of the U.S. Constitution explicitly prohibits “direct” taxes on citizens under most circumstances. This meant that for decades, the U.S. government mainly funded itself through “indirect” taxes: tariffs and levies on consumer goods like tobacco and alcohol.

With the costs of the Civil War looming, Congress imposed a national income tax in 1861. The wealthy helped force its repeal soon after the war ended. (Their pique could only have been exacerbated by the fact that the law required public disclosure. The annual income of the moguls of the day — $1.3 million for William Astor; $576,000 for Cornelius Vanderbilt — was listed in the pages of The New York Times in 1865.)

By the late 19th and early 20th century, wealth inequality was acute and the political climate was changing. The federal government began expanding, creating agencies to protect food, workers and more. It needed funding, but tariffs were pinching regular Americans more than the rich. The Supreme Court had rejected an 1894 law that would have created an income tax. So Congress moved to amend the Constitution. The 16th Amendment was ratified in 1913 and gave the government power “to lay and collect taxes on incomes, from whatever source derived.”

In the early years, the personal income tax worked as Congress intended, falling squarely on the richest. In 1918, only 15% of American families owed any tax. The top 1% paid 80% of the revenue raised, according to historian W. Elliot Brownlee.

But a question remained: What would count as income and what wouldn’t? In 1916, a woman named Myrtle Macomber received a dividend for her Standard Oil of California shares. She owed taxes, thanks to the new law. The dividend had not come in cash, however. It came in the form of an additional share for every two shares she already held. She paid the taxes and then brought a court challenge: Yes, she’d gotten a bit richer, but she hadn’t received any money. Therefore, she argued, she’d received no “income.”

Four years later, the Supreme Court agreed. In Eisner v. Macomber, the high court ruled that income derived only from proceeds. A person needed to sell an asset — stock, bond or building — and reap some money before it could be taxed.

Since then, the concept that income comes only from proceeds — when gains are “realized” — has been the bedrock of the U.S. tax system. Wages are taxed. Cash dividends are taxed. Gains from selling assets are taxed. But if a taxpayer hasn’t sold anything, there is no income and therefore no tax.

Contemporary critics of Macomber were plentiful and prescient. Cordell Hull, the congressman known as the “father” of the income tax, assailed the decision, according to scholar Marjorie Kornhauser. Hull predicted that tax avoidance would become common. The ruling opened a gaping loophole, Hull warned, allowing industrialists to build a company and borrow against the stock to pay living expenses. Anyone could “live upon the value” of their company stock “without selling it, and of course, without ever paying” tax, he said.

Hull’s prediction would reach full flower only decades later, spurred by a series of epochal economic, legal and cultural changes that began to gather momentum in the 1970s. Antitrust enforcers increasingly accepted mergers and stopped trying to break up huge corporations. For their part, companies came to obsess over the value of their stock to the exclusion of nearly everything else. That helped give rise in the last 40 years to a series of corporate monoliths — beginning with Microsoft and Oracle in the 1980s and 1990s and continuing to Amazon, Google, Facebook and Apple today — that often have concentrated ownership, high profit margins and rich share prices. The winner-take-all economy has created modern fortunes that by some measures eclipse those of John D. Rockefeller, J.P. Morgan and Andrew Carnegie.

In the here and now, the ultrawealthy use an array of techniques that aren’t available to those of lesser means to get around the tax system.

Certainly, there are illegal tax evaders among them, but it turns out billionaires don’t have to evade taxes exotically and illicitly — they can avoid them routinely and legally.

Most Americans have to work to live. When they do, they get paid — and they get taxed. The federal government considers almost every dollar workers earn to be “income,” and employers take taxes directly out of their paychecks.

The Bezoses of the world have no need to be paid a salary. Bezos’ Amazon wages have long been set at the middle-class level of around $80,000 a year.

For years, there’s been something of a competition among elite founder-CEOs to go even lower. Steve Jobs took $1 in salary when he returned to Apple in the 1990s. Facebook’s Zuckerberg, Oracle’s Larry Ellison and Google’s Larry Page have all done the same.

Yet this is not the self-effacing gesture it appears to be: Wages are taxed at a high rate. The top 25 wealthiest Americans reported $158 million in wages in 2018, according to the IRS data. That’s a mere 1.1% of what they listed on their tax forms as their total reported income. The rest mostly came from dividends and the sale of stock, bonds or other investments, which are taxed at lower rates than wages.

As Congressman Hull envisioned long ago, the ultrawealthy typically hold fast to shares in the companies they’ve founded. Many titans of the 21st century sit on mountains of what are known as unrealized gains, the total size of which fluctuates each day as stock prices rise and fall. Of the $4.25 trillion in wealth held by U.S. billionaires, some $2.7 trillion is unrealized, according to Emmanuel Saez and Gabriel Zucman, economists at the University of California, Berkeley.

Buffett has famously held onto his stock in the company he founded, Berkshire Hathaway, the conglomerate that owns Geico, Duracell and significant stakes in American Express and Coca-Cola. That has allowed Buffett to largely avoid transforming his wealth into income. From 2015 through 2018, he reported annual income ranging from $11.6 million to $25 million. That may seem like a lot, but Buffett ranks as roughly the world’s sixth-richest person — he’s worth $110 billion as of Forbes’ estimate in May 2021. At least 14,000 U.S. taxpayers in 2015 reported higher income than him, according to IRS data.

There’s also a second strategy Buffett relies on that minimizes income, and therefore, taxes. Berkshire does not pay a dividend, the sum (a piece of the profits, in theory) that many companies pay each quarter to those who own their stock. Buffett has always argued that it is better to use that money to find investments for Berkshire that will further boost the value of shares held by him and other investors. If Berkshire had offered anywhere close to the average dividend in recent years, Buffett would have received over $1 billion in dividend income and owed hundreds of millions in taxes each year.

Many Silicon Valley and infotech companies have emulated Buffett’s model, eschewing stock dividends, at least for a time. In the 1980s and 1990s, companies like Microsoft and Oracle offered shareholders rocketing growth and profits but did not pay dividends. Google, Facebook, Amazon and Tesla do not pay dividends.

In a detailed written response, Buffett defended his practices but did not directly address ProPublica’s true tax rate calculation. “I continue to believe that the tax code should be changed substantially,” he wrote, adding that he thought “huge dynastic wealth is not desirable for our society.”

The decision not to have Berkshire pay dividends has been supported by the vast majority of his shareholders. “I can’t think of any large public company with shareholders so united in their reinvestment beliefs,” he wrote. And he pointed out that Berkshire Hathaway pays significant corporate taxes, accounting for 1.5% of total U.S. corporate taxes in 2019 and 2020.

Buffett reiterated that he has begun giving his enormous fortune away and ultimately plans to donate 99.5% of it to charity. “I believe the money will be of more use to society if disbursed philanthropically than if it is used to slightly reduce an ever-increasing U.S. debt,” he wrote.

So how do megabillionaires pay their megabills while opting for $1 salaries and hanging onto their stock? According to public documents and experts, the answer for some is borrowing money — lots of it.

For regular people, borrowing money is often something done out of necessity, say for a car or a home. But for the ultrawealthy, it can be a way to access billions without producing income, and thus, income tax.

The tax math provides a clear incentive for this. If you own a company and take a huge salary, you’ll pay 37% in income tax on the bulk of it. Sell stock and you’ll pay 20% in capital gains tax — and lose some control over your company. But take out a loan, and these days you’ll pay a single-digit interest rate and no tax; since loans must be paid back, the IRS doesn’t consider them income. Banks typically require collateral, but the wealthy have plenty of that.

The vast majority of the ultrawealthy’s loans do not appear in the tax records obtained by ProPublica since they are generally not disclosed to the IRS. But occasionally, the loans are disclosed in securities filings. In 2014, for example, Oracle revealed that its CEO, Ellison, had a credit line secured by about $10 billion of his shares.

Last year Tesla reported that Musk had pledged some 92 million shares, which were worth about $57.7 billion as of May 29, 2021, as collateral for personal loans.

The IRS records provide glimpses of other massive loans. In both 2016 and 2017, investor Carl Icahn, who ranks as the 40th-wealthiest American on the Forbes list, paid no federal income taxes despite reporting a total of $544 million in adjusted gross income (which the IRS defines as earnings minus items like student loan interest payments or alimony). Icahn had an outstanding loan of $1.2 billion with Bank of America among other loans, according to the IRS data. It was technically a mortgage because it was secured, at least in part, by Manhattan penthouse apartments and other properties.

Borrowing offers multiple benefits to Icahn: He gets huge tranches of cash to turbocharge his investment returns. Then he gets to deduct the interest from his taxes. In an interview, Icahn explained that he reports the profits and losses of his business empire on his personal taxes.

Icahn acknowledged that he is a “big borrower. I do borrow a lot of money.” Asked if he takes out loans also to lower his tax bill, Icahn said: “No, not at all. My borrowing is to win. I enjoy the competition. I enjoy winning.”

He said adjusted gross income was a misleading figure for him. After taking hundreds of millions in deductions for the interest on his loans, he registered tax losses for both years, he said. “I didn’t make money because, unfortunately for me, my interest was higher than my whole adjusted income.”

Asked whether it was appropriate that he had paid no income tax in certain years, Icahn said he was perplexed by the question. “There’s a reason it’s called income tax,” he said. “The reason is if, if you’re a poor person, a rich person, if you are Apple — if you have no income, you don’t pay taxes.” He added: “Do you think a rich person should pay taxes no matter what? I don’t think it’s germane. How can you ask me that question?”

Skeptics might question our analysis of how little the superrich pay in taxes. For one, they might argue that owners of companies get hit by corporate taxes. They also might counter that some billionaires cannot avoid income — and therefore taxes. And after death, the common understanding goes, there’s a final no-escape clause: the estate tax, which imposes a steep tax rate on sums over $11.7 million.

ProPublica found that none of these factors alter the fundamental picture.

Take corporate taxes. When companies pay them, economists say, these costs are passed on to the companies’ owners, workers or even consumers. Models differ, but they generally assume big stockholders shoulder the lion’s share.

Corporate taxes, however, have plummeted in recent decades in what has become a golden age of corporate tax avoidance. By sending profits abroad, companies like Google, Facebook, Microsoft and Apple have often paid little or no U.S. corporate tax.

For some of the nation’s wealthiest people, particularly Bezos and Musk, adding corporate taxes to the equation would hardly change anything at all. Other companies like Berkshire Hathaway and Walmart do pay more, which means that for people like Buffett and the Waltons, corporate tax could add significantly to their burden.

It is also true that some billionaires don’t avoid taxes by avoiding incomes. In 2018, nine of the 25 wealthiest Americans reported more than $500 million in income and three more than $1 billion.

In such cases, though, the data obtained by ProPublica shows billionaires have a palette of tax-avoidance options to offset their gains using credits, deductions (which can include charitable donations) or losses to lower or even zero out their tax bills. Some own sports teams that offer such lucrative write-offs that owners often end up paying far lower tax rates than their millionaire players. Others own commercial buildings that steadily rise in value but nevertheless can be used to throw off paper losses that offset income.

Michael Bloomberg, the 13th-richest American on the Forbes list, often reports high income because the profits of the private company he controls flow mainly to him.

In a statement, a spokesman for Bloomberg noted that as a candidate, Bloomberg had advocated for a variety of tax hikes on the wealthy. “Mike Bloomberg pays the maximum tax rate on all federal, state, local and international taxable income as prescribed by law,” the spokesman wrote. And he cited Bloomberg’s philanthropic giving, offering the calculation that “taken together, what Mike gives to charity and pays in taxes amounts to approximately 75% of his annual income.”

The statement also noted: “The release of a private citizen’s tax returns should raise real privacy concerns regardless of political affiliation or views on tax policy. In the United States no private citizen should fear the illegal release of their taxes. We intend to use all legal means at our disposal to determine which individual or government entity leaked these and ensure that they are held responsible.”

Ultimately, after decades of wealth accumulation, the estate tax is supposed to serve as a backstop, allowing authorities an opportunity to finally take a piece of giant fortunes before they pass to a new generation. But in reality, preparing for death is more like the last stage of tax avoidance for the ultrawealthy.

University of Southern California tax law professor Edward McCaffery has summarized the entire arc with the catchphrase “buy, borrow, die.”

The notion of dying as a tax benefit seems paradoxical. Normally when someone sells an asset, even a minute before they die, they owe 20% capital gains tax. But at death, that changes. Any capital gains till that moment are not taxed. This allows the ultrarich and their heirs to avoid paying billions in taxes. The “step-up in basis” is widely recognized by experts across the political spectrum as a flaw in the code.

Then comes the estate tax, which, at 40%, is among the highest in the federal code. This tax is supposed to give the government one last chance to get a piece of all those unrealized gains and other assets the wealthiest Americans accumulate over their lifetimes.

It’s clear, though, from aggregate IRS data, tax research and what little trickles into the public arena about estate planning of the wealthy that they can readily escape turning over almost half of the value of their estates. Many of the richest create foundations for philanthropic giving, which provide large charitable tax deductions during their lifetimes and bypass the estate tax when they die.

Wealth managers offer clients a range of opaque and complicated trusts that allow the wealthiest Americans to give large sums to their heirs without paying estate taxes. The IRS data obtained by ProPublica gives some insight into the ultrawealthy’s estate planning, showing hundreds of these trusts.

The result is that large fortunes can pass largely intact from one generation to the next. Of the 25 richest people in America today, about a quarter are heirs: three are Waltons, two are scions of the Mars candy fortune and one is the son of Estée Lauder.

In the past year and a half, hundreds of thousands of Americans have died from COVID-19, while millions were thrown out of work. But one of the bleakest periods in American history turned out to be one of the most lucrative for billionaires. They added $1.2 trillion to their fortunes from January 2020 to the end of April of this year, according to Forbes.

That windfall is among the many factors that have led the country to an inflection point, one that traces back to a half-century of growing wealth inequality and the financial crisis of 2008, which left many with lasting economic damage. American history is rich with such turns. There have been famous acts of tax resistance, like the Boston Tea Party, countered by less well-known efforts to have the rich pay more.

One such incident, over half a century ago, appeared as if it might spark great change. President Lyndon Johnson’s outgoing treasury secretary, Joseph Barr, shocked the nation when he revealed that 155 Americans making over $200,000 (about $1.6 million today) had paid no taxes. That group, he told the Senate, included 21 millionaires.

“We face now the possibility of a taxpayer revolt if we do not soon make major reforms in our income taxes,” Barr said. Members of Congress received more furious letters about the tax scofflaws that year than they did about the Vietnam War.

Congress did pass some reforms, but the long-term trend was a revolt in the opposite direction, which then accelerated with the election of Ronald Reagan in 1980. Since then, through a combination of political donations, lobbying, charitable giving and even direct bids for political office, the ultrawealthy have helped shape the debate about taxation in their favor.

One apparent exception: Buffett, who broke ranks with his billionaire cohort to call for higher taxes on the rich. In a famous New York Times op-ed in 2011, Buffett wrote, “My friends and I have been coddled long enough by a billionaire-friendly Congress. It’s time for our government to get serious about shared sacrifice.”

Buffett did something in that article that few Americans do: He publicly revealed how much he had paid in personal federal taxes the previous year ($6.9 million). Separately, Forbes estimated his fortune had risen $3 billion that year. Using that information, an observer could have calculated his true tax rate; it was 0.2%. But then, as now, the discussion that ensued on taxes was centered on the traditional income tax rate.

In 2011, President Barack Obama proposed legislation, known as the Buffett Rule. It would have raised income tax rates on people reporting over a million dollars a year. It didn’t pass. Even if it had, however, the Buffett Rule wouldn’t have raised Buffett’s taxes significantly. If you can avoid income, you can avoid taxes.

Today, just a few years after Republicans passed a massive tax cut that disproportionately benefited the wealthy, the country may be facing another swing of the pendulum, back toward a popular demand to raise taxes on the wealthy. In the face of growing inequality and with spending ambitions that rival those of Franklin D. Roosevelt or Johnson, the Biden administration has proposed a slate of changes. These include raising the tax rates on people making over $400,000 and bumping the top income tax rate from 37% to 39.6%, with a top rate for long-term capital gains to match that. The administration also wants to up the corporate tax rate and to increase the IRS’ budget.

Some Democrats have gone further, floating ideas that challenge the tax structure as it’s existed for the last century. Oregon Sen. Ron Wyden, the chairman of the Senate Finance Committee, has proposed taxing unrealized capital gains, a shot through the heart of Macomber. Sens. Elizabeth Warren and Bernie Sanders have proposed wealth taxes.

Aggressive new laws would likely inspire new, sophisticated avoidance techniques. A few countries, including Switzerland and Spain, have wealth taxes on a small scale. Several, most recently France, have abandoned them as unworkable. Opponents contend that they are complicated to administer, as it is hard to value assets, particularly of private companies and property.

What it would take for a fundamental overhaul of the U.S. tax system is not clear. But the IRS data obtained by ProPublica illuminates that all of these conversations have been taking place in a vacuum. Neither political leaders nor the public have ever had an accurate picture of how comprehensively the wealthiest Americans avoid paying taxes.

Buffett and his fellow billionaires have known this secret for a long time. As Buffett put it in 2011: “There’s been class warfare going on for the last 20 years, and my class has won.”

Robert Hall and Marianna Kudlyak, "Why Has the US Economy Recovered So Consistently from Every Recession in the Past 70 Years?" Federal Reserve Bank Of San Francisco, June 2021, https://www.frbsf.org/economic-research/files/wp2020-20.pdf

In terms of the LFP issue Ben noted, “…Variations in labor-force growth. The DMP model of Mortensen and Pissarides (1994) has a constant labor force. Extensions to endogenous participation may involve positive or negative co-movements of participation and unemployment. Figure 31 shows that participation rate grew during the years up to 1990 when the rising rate for women was a key factor for overall participation (to achieve a basic adjustment for demographic influences, the data refer to ages 25 through 54).In the recovery from the 2020 recession, participation was essentially unchanged. In the recovery from the 2007-09 recession, participation declined…”

To them this implies there are structural headwinds to increasing employment rapidly, they hypothesis (but don’t quantify) that it is a combination of negative feedback from unemployment to tightness, vacancy costs, recruiting process and externalities, composition effects, scarring effects, the separation rate act to bend employment recoveries into parallel paths, “…If employers find it potentially profitable, they will exert the same effort to lay on new workers if the unemployment rate is 10 percent or 4 percent—vacancy creation is infinitely elastic. Contagion may arise from the congestion that occurs in the recruiting process when employers are flooded with applicants. Or contagion may involve changes in the equilibrium search and recruiting strategies of job-seekers and employers that impede matching and lower the efficiency of the matching process. These modifications of the DMP model lower the elasticity of vacancy creation and slow down the rate of recovery of unemployment…We noted that part of the high level of unemployment soon after a crisis reflects a change in the composition of unemployment toward individuals with naturally lower job-finding rates. This is a source of lower matching efficiency, a decline in one of the DMP model’s driving forces. A related phenomenon is the lower incidence of on-the-job search when unemployment is high. Again, this is a source of lower matching efficiency….”

They establish that"...A remarkable fact about the historical US business cycle is that, after unemployment reached its peak in a recession, and a recovery begins, the annual reduction in the unemployment rate is stable at around one tenth of the current level of unemployment... Hall and Kudlyak (2020a), we study US business-cycle recoveries over the past 70 years. We focus on the unemployment rate. Our key results are (1) the recovery process takes place reliably, regardless of the nature of the shock that causes the preceding economic contraction, and (2) the recovery process is similar in all of the ten past recoveries—unemployment falls by about 0.1 log points per year. Figure 1 displays the log of the unemployment rate during the 10 recoveries since 1948, with the recession spells of sharply rising unemployment left blank. Throughout the paper we exclude the recovery from the pandemic recession that started in 2020. The key fact about recoveries is apparent in the figure: Unemployment declines smoothly but slowly throughout most recoveries most of the time, at close to the same proportional rate. In the log plot, the recoveries appear as impressively close to straight lines…”

Their takeaway, “…Why has the US economy recovered so consistently from every recession in the past 70 years? Our answer:Recoveries are endogenous—there is a natural force causing job-seekers to match with available jobs and to lower unemployment. The bulge of unemployment created by a crisis at the beginning of a recovery creates a negative feedback to labor market tightness, endogenously slowing the recovery…”

Share of Business Output, “..Figure 6 shows the percentage of sectors with low, medium, and high C4 in business output in 2002 and 2017.43 Sectors with a low concentration ratio, below 50 percent, had a much higher share in business output than medium (50-80 percent) and high (over 80 percent). Furthermore, low concentration industries became a larger share of the economy. While the low concentration sectors constituted 62 percent of the economy in 2002, by 2017, their share had grown to more than 80 percent. Inverse patterns occurred in medium and high levels of concentration. The share of economic output from highly concentrated industries fell from 10.6 percent in 2002 to 4.7 percent in 2017. In short, the U.S. economy is not becoming monopolistic, filled with giant rapacious firms gobbling up market share….”

Concentration and Profitability, "... There is essentially no relationship between industry profitability and the concentration ratio in 2017 (a correlation coefficient of 0.04)..."

The technology frontier isn’t concentrate, “…Anti-corporate populists have taken particular aim at technology sectors, claiming that “Big Tech,” particularly the Internet industry, is concentrated. However, the number of advanced technology industries with high levels of concentration is modest (see figure 5). Using the Brookings Institution list of 4-digit NAICS code of advanced technology industries, ITIF constructed a list of advanced technology industries at the 6-digit NAICS code level.34 Of 135 industries, only 8 have C4 ratios above 80 and 10 times more have C4 ratios below 50. Moreover, there were more unconcentrated tech sectors in 2017 than in 2002. The argument that tech sectors are becoming concentrated does not hold…”

Concentration and Prices, "...Surely concentration must enable firms to charge higher prices. In fact, of the 36 industries with a C4 ratio of over 70 percent in 2017, and for which there was data available on price changes from the BLS producer price index (PPI), 22 (61 percent) saw price increases from 2002 to 2017 that were lower than the economy-wideppI. In fact, the correlation between the C4 ratio and the change inppI was actually negative (-0.31), meaning the more concentrated the industry, the lower the price increase..."

The evidence, “…In 2017, 643 industries (76 percent) were unconcentrated with less than a 50 percent C4 ratio (see figure 1). A total of 173 (20 percent) were moderately concentrated with a C4 ratio between 50 percent and 80 percent. And just 35 industries (4 percent) were highly concentrated with a C4 ratio of 80 percent or more. Even at 80 percent, this means the top four firms had only 20 percent of the market share if they split it into equal shares….”

Change btw 2002-2017, “…On average, concentration increased only 1 percentage point between 2002 and 2017 after taking the simple average across all industries of the differences between C4 from both years (34.3 percent in C4 from 2002 and 35.3 percent in C4 from 2017). Given that industries with C4 ratios below 50 percent are considered unconcentrated, this is a very low number. The concentration of the eight largest firms (C8) increased even less, from 44.1 to 44.7 percent. Even considering the eight largest firms, the concentration ratio remained lower than 50 percent. Overall, 467 sectors (55 percent) increased in concentration, while 384 (45 percent) decreased (see figure 2). Again, this is hardly evidence of widespread growth of monopoly. Moreover, among the sectors that saw an increase, only 152 (18 percent of the total) increased by more than 10 percentage points..

Industry by concentration rate, “…Some industries clearly got less concentrated, while others got more. Table 1 shows the 20 industries with the greatest increases in C4 ratio. Only 30 percent had C4 ratios above 80 percent in 2017. And even for some of them, there was little risk of firms exerting much market power. For example, industries such as other performing arts companies, luggage and leather goods stores, geothermal power generation, and paint and wallpaper stores all face significant competition from firms in other industries such as movie theaters, department stores, and natural gas power generation. For some other industries, the U.S. trade balance deteriorated, meaning that imports took a larger share and provided more competition.30 This includes newsprint and electrical lighting manufacturing, in which the United States is the largest importer (importing 20 percent of the total international trade for both industries).31 For other industries, such as taxi service and travel agencies, the Internet enabled significant economies of scale and cost reductions, such as with the rise of Uber and Lift for taxis and Travelocity and Expedia for travel….”

Key factoids, “… Just 35 of 851 industries (4 percent) were highly concentrated, with the top-4 firms (the C4 concentration ratio) holding more than 80 percent of the market….In 2017,80 percent of U.S. business output was from industries with low levels of concentration, with that share increasing from 62 percent in 2002…..Fifty-five percent of industries increased concentration between 2002 and 2017; 45 percent decreased…..The average C4 ratio increased by just 1 percentage point between 2002 and 2017, from 34.3 percent to 35.3 percent, while the average C8 ratio increased even less, from 44.1 percent to 44.7 percent…..There was a slight negative correlation between the C4 level in 2002 and the percentage point change in C4 between 2002 and 2017….Among the industries with increases in concentration, only one-third increased by greater than 10 percentage points…..Of the 20 industries showing the greatest increase in the C4 ratio from 2002 to 2017, only 30 percent had C4 ratios above 80 percent in 2017…..Of the 115 industries with a C4 ratio of 60 percent or more in 2002, the majority got less concentrated, with the average C4 declining 4 percentage points…..For every advanced technology industry with a C4 ratio over 80 percent, there were 10 with a C4 ratio below 50 percent….Producer prices rose less from 2002 to 2017 in industries with higher levels of concentration than overall prices…”

Robert Atkinson and Filipe Lage de Sousa use Census data to argue that there hasn’t been an increase in concentration, asBen noted they aren’t using HHI series that the vast majority of people looking at this space do. Their bottom line, “…And the facts (data from the Economic Census) do not support assertions of monopolization. More than 80 percent of business output is in sectors with low concentration ratios, including many advanced technology industries. Moreover, from 2002 to 2017, concentration mostly stayed low and increased very little. In addition, more-concentrated industries did not on average show greater increases in prices or boosted profits than others…”