Sometime before the COVID-19 pandemic crimped global travel, I found myself in Hong Kong, in the elegant coffee shop of the city’s flagship Mandarin Oriental hotel. The venue’s aging elegance was more than made up for by its clientele—energetic, wealthy Mainlanders. I was being interviewed there by a young Chinese journalist. Pleasantries covered, he jumped into substance and asked rather pointedly how his readers should understand the Trump Administration’s trade war with his country.

Imagine you are a schoolyard hulk, I said, the sort who by virtue of being on an accelerated hormone schedule finds oneself bigger than one’s middle-school classmates. You relish your relative power.

But with time, puberty comes to those around you, and you find your dominance withering away. Already it is more difficult than it once was to get your way with your classmates. One, in particular, is growing too soon. You decide, somewhat questionably, to act before it is too late—you will show them who is boss, once and for all. A fist fight, perhaps?

You are aware that this fight is not without costs: With the classmate now a little bigger, you might have to take a few punches, perhaps even a black eye. But you are confident that you will emerge victorious, because you still have the size advantage—the fist fight will leave your classmate in a worse state. Wait much longer, and a fight will not be an option. So you act.

As bizarre as this analogy may be, it reflects how a good fraction of America sees the country navigating the world we live in. How did we get here?

America’s evolving trade war is an expression of a broader frustration in the country with free markets. It’s not just Trump voters who are fed up. A 2020 Institute of Politics poll of Americans aged 18 to 29, who tend to skew Democratic, found fewer than half with a favorable view of capitalism. Where once America was synonymous with free markets—because they embodied individualism, that that most archetypal of American values—today we see the specters of economic nationalism and even socialism, although the precise meaning of these terms, even among their proponents, remains contested.

Much of the popular debate and antipathy on free markets has focused on growing wealth concentration in the top one percent. But focusing on wealth inequality alone can be misleading. Even in the heydays of les trentes glorieuses, the lower 90^th percentile’s share of total wealth in the United  States amounted to only about 35 percent. Moreover, wealth inequality data can obfuscate reasonably comfortable working-class lifestyles. For instance, Americans across the board have consumed ever more gadgetry and services over 50 years.

But such consumption for many has come with rising personal debt. For Americans in the bottom quintile of earners in 2016, median debt was over one-and-a-half times median net worth. This is up from debt for that group being about 75 percent of net worth in 1989—so the underlying concerns raised by rising wealth inequality fester.

New research I have conducted with Timon Forster provides a fresh perspective on the debate. We focused on wage-market returns to human capital in America. After all, for most people, all they have is their human capital, both as their primary source of earning potential and of self-worth. And, ideally, free markets are about rewarding such human capital.

So how well have free markets done on this metric in recent years? Quite unevenly, it turns out.

Just as there is a distribution of financial wealth in society, there is a distribution of human capital—not all of us can be rocket scientists, and moreover, not everyone has access to opportunities that would allow them to become a rocket scientist. Where you find yourself on this distribution, it so happens, determines how well your lot has improved over time.

For those in the lower third of the distribution, inflation-adjusted wages have stagnated or even declined since 1980. Those in the upper two-thirds have seen some improvements in inflation-adjusted wages, with more impressive improvements the higher up the distribution (see figure 1).

(Figure 1)

In effect, those who win the human capital game, determined largely by how cognitively demanding their occupations are, have done well over time; those who have lost the game are worse off than they would have been nearly four decades ago.

Consider Joe, a typical freight handler in 1980, in the middle of his career. At the time, Joe’s occupation would have put him in the ninth percentile of skills in the U.S. population, and he could have expected to earn $17.27 an hour in 2017 dollars. Joe’s 1980 wages were about 60 percent of those in the 95^th percentile of skills in the United States, such as physicists and astronomers.

Were Joe the freight handler instead to find himself in the middle of his career in 2017, he would be in a less favorable position. While his occupation would still place him in the ninth percentile of U.S. skills, his real wages would have fallen to $14.57 per hour and would represent only about 40 percent of those in the 95^th percentile.

Of course, these trends do not account for non-wage benefits, such as health insurance or other perquisites. But such benefits, to the extent that they have grown with time, tend to have accrued more generously further up the distribution. Non-wage benefits are unlikely to have alleviated the decline or stagnation in real wages at the lower end; they might well have exacerbated it.

To understand why the schoolyard hulk is willing to engage in a self-injurious fist fight, consider the situation. Americans have experienced an economic system that, over a fairly long run, has left about one-third of the population worse off. I suspect even Milton Friedman, who was not too hassled with inequality so long as the economy was lifting all or most boats, would find this scenario somewhat bothersome.

Automation is implicated here. Computerization advances over the past four decades have reduced demand for routine cognitive tasks, just as the Industrial Revolution reduced demand for routine physical labor. To compound inequality, computerization increases demand for higher-order cognitive tasks that complement the output of computers. Tasks such as inductive problem-solving, implicit pattern recognition, and organizing and communicating across unpredictable situations are now more valuable since they can be leveraged many fold with advances in data analytics.

In 2013, Oxford academics Cary Frey and Michael Osborne produced a study that identified jobs in the U.S. economy being most vulnerable to computerization “over the next decade or so.” They estimated about 47 percent of jobs as vulnerable. In sorting those jobs across the distribution of human capital in the U.S. workforce, I find that the vast majority of these occupations appear in the bottom half of the distribution. Accountants and auditors, at about the 85^th percentile of cognitive skills, are the notable exception (see Figure 2).

(Figure 2)

The celebrated physicist Stanislaw Ulam is said to have on occasion challenged Paul Samuelson—perhaps the 20^th century’s most widely regarded economist—to name one theory in all of the social sciences that is both non-trivial and true. Samuelson took his time to construct a response (years, by his own account), but when he did, he offered up David Ricardo’s theory of comparative advantage in international trade. Put simply, the theory argues that trade benefits all countries because it promotes efficiency by driving countries to focus on their relative advantages rather than their absolute advantages.

This non-intuitive result has been the intellectual backbone of public policy driving globalized free markets. And beyond automation, globalization— and offshoring thereof—has been core to America’s low-skill wage stagnation.

Consider the offshoring of manufacturing jobs that emerged in earnest in the 1980s, after container standardization and faster shipping dramatically changed the cost calculus on having bulky items fabricated in distant low-cost destinations. In 2002, economist Lori Kletzer identified manufacturing jobs in 1980 that were most vulnerable to offshoring. Perhaps not surprisingly, my research shows that the majority of those jobs were below the fiftieth percentile of the U.S. workforce’s human capital distribution (see Figure 3).

But the story doesn’t end there. The offshoring of manufacturing jobs meant greater demand, and a wage premium, for service skills that enable offshoring—skills in IT, back-office finance, logistics, and so on. The wage premium on these jobs helped drive the service boom of the late 1990s. But that wage premium also created the incentive for new technologies that would allow those jobs to be themselves offshored. Again, free-market policy carried the day, enabling such offshoring.

In 2009, economist Alan Blinder identified service-sector occupations vulnerable to offshoring in the mid-2000s and beyond. Remarkably, I find that the majority of those occupations are not in the lower half of the U.S. workforce’s human capital distribution; they are largely in the middle of that distribution (see also Figure 3). Globalization and offshoring revved up in the 1980s at the lower end of the human capital spectrum but with time have begun to pinch those in the middle.

(Figure 3)

Another free-market policy that appears implicated in U.S. low-skilled wage stagnation is immigration. The proportion of native-born U.S. workers in the first 40 percentiles of the U.S. human capital distribution has declined since at least 1994, when reliably precise data become available. This decline mirrors the stagnation and decline of inflation-adjusted wages in that segment of the population. Immigrant workers appear to be increasing labor supply at the lower end of the distribution, resultantly depressing wages overall (see Figure 4).

(Figure 4)

High-skilled immigration tells a different story. The U.S. workforce has also experienced an increase in non-native workers in the top five percentiles of its human capital distribution. But a number of studies, including by economist William Kerr, have shown high-skilled immigrants boost commercial patent activity, creating demand for other workers, both native and foreign-born.

Free-market policy in America catalyzed the perfect storm of offshoring, low-skill immigration, and automation that together have decimated wages of those in the lower third of America’s human capital distribution. With time, this decimation is spreading like a cancer to the middle third. Given the record, it is fair to ask if this was ever an idea worth pursuing.

This challenge can be quickly put to bed with evidence of the impressive improvements in wages across previously socialist economies that have since embraced market liberalization. In fact, I suspect my youthful Chinese journalist’s sharp questioning of emerging American protectionism stems from his own experience of the sheer advantages wrought by an export-driven trade policy in his homeland. Most Chinese, I imagine, prefer little pink iPhones to Little Red Books.

India is another case in point—a country that started to deregulate its markets about a decade after China, most convincingly after 1991. Contrary to the American experience, wages across the entire distribution of human capital in India have seen improvements in inflation-adjusted terms since the 1980s (see Figure 5).

(Figure 5)

The recent global success of free-market policy is, however, unlikely to convince many Americans at whose expense these wins have been secured. Not surprisingly then, they have turned to populist alternatives to free markets—economic nationalism on the right or outright socialism on the left. Witness how Donald Trump and Bernie Sanders have transformed American politics.

What is so difficult about our current situation is that it pits two moral rights against each other. The first is the moral right of technological progress and globalization, with their potential to lift billions out of poverty worldwide; the second is the moral right of aiding the working and middle classes in the West who have seen their earnings potential decimated by advances in technology and globalization.

If we uniquely choose either right as the way forward in public policy, then we will create profound harm amongst the right not chosen, and thus even further the outrage. The defining policy challenge of our time is to temper the unequal effects of free markets without falling victim to populism. Admittedly, this is a task perhaps analogous to reasoning with a frustrated schoolyard hulk eager to recapture a fleeting power that is quickly slipping away.

On socialism, it suffices to say that it bodes poorly for individual liberty —that core American ideal at the heart of why we have a free-market system. One need travel no further than Venezuela to see how a once-vigorous polity has been decimated by a despot and his heir, who cloaked their own power grab in deceitful promises of economic parity.

But the same is also true of economic nationalism. Such policies cut down those on the home team who are winning from free trade, as foreign states reciprocate with sanctions. Then, predictably, nationalist governments must respond with compensatory giveaways to the erstwhile winners.

We are living this dystopia as the Trump Administration has tried to (further) placate an irate farmer lobby that is feeling the pinch from Chinese retaliation on U.S. agricultural exports. Soon, hitherto globally competitive businesses become addicted to government favors. It is difficult to bite the hand that feeds you, and eventually meritocracy loses to mediocrity, self-reliance to sycophancy.

If populists cannot be trusted to address the failures in free markets whilst preserving personal liberties, what are our options?

The more moderate response has been to preserve the core of our free-market system but with significant adaptations. These adaptations can be broadly categorized as either consumer subsidies or producer inefficiencies. Are these good ideas? Let’s take them in turn through their two most visible manifestations: a universal basic income (UBI), and greater corporate social responsibility (CSR).

Amongst our new Brahmin class in Palo Alto and Seattle, where dwell some of free markets’ greatest winners, a more generous welfare system is being conceived, now rebranded for the 21^st century as UBI.

At first glance, UBI is an intriguing idea—a “citizen’s dividend” to support consumption and quality of life in the face of the obviation of human productivity by technological advances. Even if such advances have stagnated wages, the argument goes, the march of technology is welfare improving and should not be impeded. Making the UBI work would require a profound re-imagination of the social contract—today, so much of an individual’s identity in society is tied to the work he or she does. How do we reconstruct this dignity for those who are to simply live off giveaways?

This question begets another: How to avoid the net recipients of a UBI becoming vassals of the net contributors? A UBI can only be financed by high taxes that can obligate takers to givers. And, moreover, these taxes can thwart new entrepreneurship, protecting incumbent producers and preventing further mobility of takers into givers. It is only a short road from a cash-based UBI to one that provides vouchers for Amazon only, meaning that tech titans could use the UBI to perpetuate their own monopolies.

As crucial questions about UBI remain unanswered, the boardrooms of Main Street Corporate America are resurrecting mid-20^th-century ideas that corporations have social responsibilities to citizens, customers, and employees. These ideas had fallen out of fashion in America during the heyday of free markets in the 1980s. Contrary to conventional wisdom, concerns about inequality now run deep in Corporate America, even if largely for instrumental reasons—the affected, after all, are their paying customers. So it is worthwhile to consider corporations’ rediscovered interest in CSR.

As I have written more extensively in the California Management Review, the idea that corporations have responsibilities to citizens, customers, and employees is eminently sensible—those responsibilities are usually met if corporations focus on creating long-term value for shareholders. The real challenge with CSR, however, is posed when purported corporate responsibilities to citizens, customers, and employees are at odds with those to shareholders, even over the long run. For instance, what should happen when it is in the best interest of shareholders to shut down a cash-bleeding factory that is a major local employer?

On the surface, the objective to make the corporation more humane by focusing away from profits to social goals is laudable. But we’ve long been cautioned against this, from Adam Smith to Milton Friedman, not just because it is wasteful of investors’ capital, but more so because it quickly morphs into a system that allows corporate managers to use that capital to advance their own private political philosophies. Imagine, for instance, what would happen if corporate political spending were not constrained, at least nominally, by the principle that it should be value-accretive to shareholders—if that spending could be driven by managers’ own idiosyncrasies.

CSR, if deployed to its purest intent in the U.S. setting, would likely catalyze our decline into authoritarian capitalism.

Faced with a deeply illiberal menu of populist responses to free-market failures and serious challenges amongst even the more moderate alternatives, perhaps we are forced to confront the hard reality: Despite its shortcomings, we can do no better today than free markets. This is akin to Winston Churchill’s defense of democracy—admittedly hardly inspiring, but credible and sobering nonetheless. Over the long run, the only way the American experiment can keep going is by growing, if not spatially then through intellectual investments it makes in its citizens.

The wage-stagnation data suggest that the human capital difference between the top and the bottom segments of the American population has been widening over nearly four decades. In fact, between 1980 and 2017, we have seen an almost 60 percent increase in the cognitive-skill differential across the first and hundredth percentiles of human capital, as implied in real wages.

With further coming advances in computing technology and increased global competition from varieties of capitalism that are less free, the state and progression of human capital inequality in America is unsustainable if this society hopes to remain free.

America must respond boldly by redoubling investment in the entirety of its human capital. And this is not simply a call for more vocational education at the lower end: More for-profit colleges for cosmeticians are unlikely to rebuild the American middle class.

Rather, we must invest in the kinds of education that enable those at even the lower end of the skill distribution to add sufficient value through their labor so as to sustain and grow their wages in the face of automation and global competition. This means a public education program that enables creativity and innovation across the human capital distribution.

Somewhat promisingly, the human capital wage gap is less severe in countries that have made such public education a prerogative, such as Finland (see Figure 6).

(Figure 6)

The Finns treat human capital development as a national public good, not simply as an individual investment. Their policy acknowledges the interdependencies between a skilled workforce and a free society. And this policy has profound consequences. For instance, teaching at all levels, including early education, is genuinely respected in society. Admissions to Finnish primary-school teacher programs are reportedly amongst the most competitive, and the training itself is argued to be as rigorous as that for U.S. doctors. With great talent in the teaching system, decisions can be decentralized and localized to specific pupils, fostering creativity.

Further, the Finns’ collective investment in their human capital means that they start early. Building on research that inadequate prenatal and neonatal care impedes cognitive development, the system invests in supporting expectant and new parents in providing critical nutrition and stimulation to young minds. Irreversible nurturing errors in early life are mitigated, providing better fodder for the school system.

Certainly, there are substantive differences between Finland and the United States that make broad comparisons tricky. For one, Finland is, in relative terms, more culturally and ethnically uniform. But the relevant lesson for America cannot be obscured by Finnish homogeneity. Public education is an urgent national security priority in America.

It might seem Panglossian to propose a future where most of our human capital is engaged in the creative economy, but this kind of skill transformation is not without precedent.

When the Industrial Revolution started to rev up in England in the mid-to-late 18^th century, most of the country was employed as domestic and farm workers, as they had been for centuries. These were cognitively undemanding tasks by today’s standards. In the course of two generations, however, the country transformed its human capital to meet the skills necessitated by industrialization —by the early 19^th century, many Britons were engaged in occupations that were unfathomable just decades prior.

Admittedly this transformation ushered in tremendous social upheaval, but the transformation itself is proof of concept that human societies can fundamentally reinvent the nature of productive activity.

If we are to preserve our free society, we must make every effort to reduce the skill differential between those at the top and bottom of the human capital distribution. Of course, not all of us can be rocket scientists, but more should have access to opportunities that would allow them to become one if they are naturally so inclined.