Near the end, the conversation focuses on philosophical questions about what makes a good life. Moyn believes that the old need to have their money and resources taken away, and be forced to rely on the state. This raises the question of whether we should try to make people as wealthy as possible so they can pursue their own ends, or rather have the government take more of a role in making sure they find creativity and meaning. Hanania brings up Jimmy Buffett-themed retirement communities as an embodiment of the kind of preference-driven lifestyle choices that would be much less likely to exist under what Moyn calls “utopian socialism.”

For more on the ideas discussed in this podcast, see also Hanania’s review of Gerontocracy in America here.

This discussion explores the structure of the North Korean state, its evolving legal system, ideological shifts, and everyday life under Kim Jong Un, including stand-up comedy, sports, pornography, religion, and the ways in which the government has allowed markets to function.

The conversation spends significant time on what may be the most consequential shift ever in North Korean ideology: Kim Jong Un's abandonment of Korean reunification as a goal. For generations, reunification was the animating myth of the North Korean state. Kim has now declared South Korea a corrupted and permanently hostile country, in many ways little different from any other enemy. Ward explains why this happened and how it changes the geopolitical situation. As it turns out, there is both a more optimistic and a less benign interpretation of this change.

Ward talks about the emergence of the category of “unsocialist” behavior, a concept that is increasingly used, surprisingly, to crack down on cultural deviance. What makes this paradoxical, as Hanania points out, is that the regime has simultaneously become more tolerant of markets. He asks about BR Myers’ book The Cleanest Race, and whether it is correct to say that North Korea has moved away from Marxism-Leninism and toward a racialized view of the world.

Perhaps the sharpest edge of the regime's legal apparatus is directed not at "hostile" foreign culture generally but at South Korean culture specifically. Ward explains the legal distinctions: disseminating South Korean cultural content carries harsher penalties than disseminating content from other officially hostile countries. A North Korean caught with American media potentially faces a less stringent punishment than one found watching a K-drama. The regime understands that South Korean culture is uniquely threatening because it comes from a society that shares the same ethnocultural background.

The episode closes with the question of succession. Kim Ju Ae, Kim Jong Un’s teenage daughter, looks to be his heir apparent. Ward says that whether she is the successor or not, we know that her father wants us to think she is. This at the very least tells us something about the impression the regime is trying to create internally and for the rest of the world. Yet we can only guess whether she actually is the successor, and what exactly the North Koreans believe they are communicating.

Links

Peter Ward, “The Legal Regulation of Unsocialist Activities in North Korea: Over-criminalization, Political Control and Human Rights.”

Peter’s work at the Sejong Institute, NK Pro (NK News), and Google Scholar

Peter’s recent appearance on the NK News podcast

Hanania and Arm disagree strongly about JD Vance, and what he says about the future of the GOP. Recently, Rod Dreher made waves by suggesting that perhaps 30% to 40% of Republican staffers are Groypers. As Arm’s job involves dealing with this class of people, he has particular insight into that issue. Hanania is interested in not only whether they are personally fans of Nick Fuentes, but also what their views are about Ronald Reagan and how they think about political issues more generally.

Arm also explains the dynamics of the recent NYC mayoral race. How did Mamdani end up winning? And why couldn’t the Republicans field a serious challenger? The conversation provides much to consider for those who believe that embracing markets and new technology is the best way to achieve economic growth but feel alienated from both major parties right now. Arm and Hanania also handicap the 2028 Democratic race, discussing how Alexandria Ocasio-Cortez might rise to the top due to the moderate wing being split, in the same way that Biden had a lane open to him in 2020.

For more articles by Arm and other Manhattan Institute scholars, subscribe to the City Journal Substack.

Related Links

Jesse Arm, “Ezra Klein’s Blindspot”

Richard Hanania, “The Broad Lessons of YIMBY”

Richard Hanania, “Boomer Liberalism Must Be Overcome” (conversation with Derek Thompson)

Richard Hanania interviews State Senator Scott Weiner on the YIMBY victory in California

Timothy B. Lee and Kai Williams, “Waymo’s Next Five Cities Are All in Red States”

JD Vance on the problems of Appalachia

Beginning in the 2010s, the salience of the immigration issue took off. Voters have been flocking to populist candidates as a result ever since. As much as many academics and political leaders would like to deny this fact, this provides the most straightforward explanation for recent trends.

The conversation focuses on what to make of the data. Does this mean that to defeat populism, all politicians have to do is become more restrictionist? How does this square with the rise of populism being a global phenomenon, even in countries where immigration isn’t an issue? What to make of research showing that populism is bad for the economy? Are there examples of European parties that have headed off the rise of populism in this way? Guenther discusses the well-known case of Denmark, but also brings up the surprising example of Hungary, where Orbán took the mainstream conservative party, made it more populist, and has dominated the politics of his nation ever since.

Another perhaps surprising finding is that there is much less of an elite-public gap on economic issues. If anything, politicians are more inclined to support free markets. Hanania asks whether the gap might be even greater than the paper suggests, given the way questions are framed. This helps explain why left-wing populism hasn’t been as successful as its right-wing equivalent, whether in the US or Europe.

Abortion pops up as the one social issue on which there is no notable difference between elites and the public. This is reminiscent of experiences in American red states, where politicians take abortion rights away while voters nearly everywhere support the pro-choice position in referendums. The conversation discusses why this might be.

The rise of right-wing populism is the political story of our time. By grounding the discussion in empirical data, Guenther sheds light on why mainstream parties have struggled to respond, and what strategies may or may not work to blunt the populist surge. This conversation helps provide a clear and evidence-based perspective on one of the defining issues of our era, while providing hints of possible future productive avenues of research.

• Lack of legal basis for martial law: Yoon’s justification — political crisis and opposition party tyranny — did not meet the constitutional or legal criteria.

• Procedural violations: Yoon bypassed required Cabinet deliberation, failed to properly announce martial law details, and did not notify the National Assembly.

• Obstruction of the National Assembly: The president used the police and military to block and remove lawmakers, infringing their constitutional rights and immunities.

• Suppression of political opponents: Yoon ordered intelligence officials to track opposition leaders for possible arrests.

• Misuse of the military for political purposes: Military forces were deployed to interfere with democratic institutions.

• Undermining democracy and the separation of powers: The martial law orders barred political and legislative activity nationwide.

• Violations of civil liberties: Citizens’ political rights, freedoms of assembly and occupation were unjustly restricted.

• Illegal raid on the National Election Commission: Yoon ordered a warrantless military operation against the NEC, breaching its independence.

• Threats to judicial independence: He tracked former justices, creating a chilling effect on the judiciary.

Following the ruling, Yoon was immediately removed from office, thus mandating a new presidential election within 60 days.

Far from resolving the polarizing debate over Yoon’s martial law declaration of December 3-4, the unanimous ruling may have exacerbated it. The dominant progressive Democratic Party (DP) saw the ruling as affirming its view of Yoon and his conservative People’s Power Party supporters as legacies of the nation’s authoritarian past. International media, such as the BBC, reinforced this narrative by interviewing reasonable, younger, English-speaking progressive activists and politicians, and translating the angry Korean outbursts of older, conservative protestors.

Conversely, conservatives saw the ruling as affirming progressive dominance of the nation’s major institutions and marginalizing conservative voices; specifically, conservative lawyers argued, the ruling relied on contradictory and unreliable testimonies that the President ordered the arrest of political leaders or attempted to obstruct the National Assembly, despite Yoon’s claim to the contrary. Others made less-substantiated claims about Chinese and North Korean influence.

What we have then is a mainstream media narrative on the impeachment ruling that selectively highlights certain forms of information that support the progressive party, along with an alternative media narrative that makes both reasonable and poorly founded claims. All of this illustrates a general pattern of what we may term asymmetric polarization in South Korea and other countries, where two sides act in ways that threaten liberal democratic norms and individual rights through different methods, depending on their levels of representation within elite institutions.

In a 2022 Pew Research survey, South Koreans ranked the most polarized among 19 countries, with 49% of respondents endorsing the idea of “very strong conflicts between people who support different political parties.” The US was ranked second, with 41% claiming the same.

Vigorous disagreement among supporters of different ideologies and parties is not necessarily problematic for liberal democracy. What is problematic is when partisans view and treat their opponents as illiberal, harmful actors, who do not deserve the protection of liberal procedural rights and norms (PRN), which include both formal-legal rights such as free speech, due process, and equal treatment under the law, and non-legal norms such as journalistic impartiality and academic objectivity, which are associated with fair procedures in a liberal democracy.

When competing groups of partisans exercise differential power in mainstream bureaucratic institutions, they may violate liberal PRN in different ways. To the extent that supporters of one party or ideology dominate institutionally, and view members of the opposing party as illiberal, they can exercise institutional power to marginalize opposing viewpoints and information. Conversely, to the extent that members of the opposing party view mainstream institutions as biased and oppressive, they reject them and turn to alternative outlets. Non-mainstream media offer both substantive information suppressed in elite institutions, and unsubstantiated information and conspiracy theories that fuel grassroots support for radical disruption.

In America’s post-1960 culture war politics, partisan identity is significantly defined by one’s view of the nation’s political and religious founding, as heroic and inspiring or oppressive. Progressive-leaning Democrats often view the nation’s white, male, Christian-dominated past as problematic and champion its challengers, while conservatives take the opposite position.

In the past two decades, self-identified progressives in the US have gained unprecedented dominance in colleges and universities. From 1989 to 2016-2017, the ratio of progressive to conservative faculty more than doubled from 2.3:1 to 5:1, with 60% identifying as liberal and only 12% as conservative; among campus administrators, the ratio was 12:1, as only 6% identified as conservative. In the latest 2021-22 survey, the ratio of liberal to conservative faculty skewed further to 7:1.

Progressive faculty and administrators often frame conservative viewpoints as illiberal and harmful, and exercise their institutional power to marginalize them. Sixty-three percent of liberal faculty believe it can be acceptable to shout down speakers, compared to 47% of moderates and only 12% of conservatives. Additionally, 31% of liberal faculty think blocking students from attending a speech is sometimes acceptable, while only 16% of moderates and 5% of conservatives agree. One survey showed half of academics in the social sciences and humanities expressing a willingness to discriminate against Evangelical Christians in hiring decisions.

Fearing progressive ostracism, scientists avoid certain issues, especially race and gender, or only allow members of a victimized group to speak on a topic. Cathy Young writes, “Researchers and clinicians who depart from the orthodoxy on youth gender dysphoria have been targeted by intense pressure from activists, leading to severe career repercussions and, in one case, the retraction of a large study of adolescent ROGD [rapid onset gender dysphoria].” Thus, progressive-dominated academia and media ignore or suppress inconvenient information, such as the use of puberty blockers being questionable, data that as many as half of transgender female prisoners are sex offenders, or that key scientists were uncertain about the origins of Covid-19.

To the extent that progressives dominate knowledge-producing institutions and view conservatives as illiberal, they exercise institutional power to marginalize information that can be weaponized by those they consider harmful actors; this violates procedural liberal norms on objective scholarship and nonpartisan journalism. Conversely, to the extent that conservatives view mainstream institutions as biased and oppressive, they turn to alternative news sites and social media accounts that are disconnected from and even hostile to mainstream institutions, with these sources usually lacking traditional journalistic procedural norms and practices used to distinguish between corroborated and non-corroborated claims. Alternative media popular with MAGA Republicans such as X and Truth Social offer both true information suppressed in mainstream institutions, and fake news and conspiracy theories, including those that come from Russian influence campaigns.

Mainstream media also disseminate some unsubstantiated claims such as hidden mass graves of indigenous children, but the dissemination of misinformation, without basic fact checking procedures, occurs at much higher rates in conservative and alternative media.

Finally, to the extent that partisans access mutually isolated information markets, they diverge in their worldviews and policy preferences. If the information market on one side is flawed, partisans advance policies based on false assumptions. Some of the Trump Administration’s steps taken on Ukraine, tariffs, and DOGE budget cuts, among other topics, exemplify reliance on false information, such as that President Zelensky is at a four percent approval rating.

The pattern of asymmetric polarization and diverging (mis)information between the progressive-dominated mainstream and the conservative alternative is replicated in other OECD countries such as Germany. As in the US, the far-right media in Europe disseminates more misleading and uncorroborated claims than do mainstream outlets.

In South Korea, asymmetric polarization is both more illiberal, as conservative and progressive partisans use legislation and judicial rulings to imprison opponents who challenge their favored narratives; and unique, as the progressive-dominated mainstream matches the ‘far-right’ alternative in disseminating unsubstantiated claims, and only the progressive mainstream disseminate such claims from their autocratic neighbor, in this case China, and censor serious academics.

Even more than in the US, the major parties in South Korea divide over the legacy of the nation’s major founding presidents: Rhee Syngman (1948-1960) and Park Chung-hee (1961-1979). Conservatives, led by the current People Power Party, generally laud the founders’ opposition to communism, and communist-party led North Korea and China, and support of the US and free market capitalism. Conversely, progressives, led by the Democratic Party, strongly criticize the human rights violations and social injustices associated with the founding administrations’ anticommunist and pro-capitalist policies, and support those more favorable to labor unions at home and to North Korea and China abroad.

Since South Korea’s first leftist government, that of Roh Moo-hyun (2003-08), the communist-capitalist divide has overlapped with another one based on religion. Many Korean Evangelicals laud the Christian values of founding President Rhee, and oppose same-sex marriage and other challenges to traditional family values. Conversely, progressive Presidents Roh and Moon Jae-in (2017-22) have appointed judges who support the rights of historically victimized groups, including members of the LGBT community.

Before democratic reforms and the political transition of the 1990s, conservatives used their dominance of the government and media to frame their opponents as pro-North Korea leftists and collaborators. They vigorously enforced the 1948 National Security Law’s Article 7 prohibition against pro-communist speech to silence progressives.

Since the 1990s, however, progressive partisans have effectively promulgated an equally Manichaean narrative of the Rhee and Park regimes, and their conservative successors, as pro-Japan collaborators and authoritarian fascists. Korean progressive media producers partner with US providers of content like Apple TV and Netflix to globally disseminate narratives that dramatize or fantasize the evils of the Japanese colonial (1910-45) and anticommunist regimes (1947-1987), which effectively stigmatize pro-Japan or anti-communist viewpoints and justify their legal suppression. Works in this genre include the film Exhuma (2014), and the series Mr. Sunshine (2018-) and Taxi Driver (2021-).

Although progressive icon Kim Dae-jung praised free speech, his successors in the Democratic Party have used their growing political and judicial influence to punish speech that defames former victims of colonial or authoritarian regimes, including obtaining a six-month prison sentence for Song Dae-yup, a professor who suggested that most Korean comfort women had volunteered for the Japanese military and two years for Jee Man-won, a commentator who claimed that North Korea helped direct rioters during the 1980 pro-democracy movement in Gwangju.

In December 2020, the DP-dominated Assembly passed the May 18 Distortion Punishment Act to imprison up to five years those who make false claims about the 1980 movement. If progressives regain the presidency, they promise to extend the Act to cover “false” claims, including those about North Korean agents, in other democratic uprisings.

By suppressing supposedly harmful far-right or pro-Japanese viewpoints and information, progressive-dominated media and schools become an echo chamber of uncorroborated claims, such as that the Japanese military kidnapped as many as two hundred thousand Korean girls and that North Korean government agents did not communicate with democracy activists, and that arguments to the contrary are those of fascists.

Korean academia and media such as the Korea Times and Hankyoreh also partner with government-controlled institutions in China. In 2014, China’s Jilin Provincial Archives provided the South Korean media with 25 documents showing that the comfort women did not sign up voluntarily in order to make money – as Japan claims – but had actually been officially mobilized by the Japanese imperial army. US-based academics and journalists repeat the sexual slavery narrative and ostracize dissenting scholars such as Harvard Law School’s Mark Ramseyer.

Korea- and US-based academics support South Korean legislative efforts to further criminalize far-right speech. They even target creative fiction: in 2021, activists pressured corporate sponsors to drop the Disney plus television show Snowdrop because of its fictional portrayal of a North Korean spy during South Korea’s democracy movement in the 1980s. US-based mainstream media have supported such activists by labeling claims about North Korean involvement as “conspiracy theories”, despite some documented evidence. Kookmin University historian Andrei Lankov writes that he “personally knows [three] people who once commuted to Pyongyang via submarine” (Lankov, The Real North Korea, p. 31).

Shaped by progressive-dominated, knowledge-producing institutions, Democratic Party leaders condemn scholars who challenge the progressive narrative. After Sejong University professor Park Yu-ha published her book, Comfort Women of the Japanese Empire, which won top academic awards in Japan, Democratic Party chief Lee Jae-myong posted the following on Facebook

This professor…
Is she still holding her position as a professor?
How did I end up breathing under the same sky as someone like her… sigh.
These remnants of pro-Japanese collaborators must be cleared away.

Since former conservative President Park Geun-hye’s 2016 impeachment, the Democratic Party has dominated the National Assembly. Yoon, the conservative party candidate, barely won the 2022 presidential election, but the DP parliamentary majority paralyzed the administration with an unprecedented array of budget cuts, along with 22 motions to impeach government officials, before Yoon declared martial law on December 3, 2024.

To justify such paralysis, DP partisans framed the Yoon Administration’s policies of reconciliation with Japan as pro-Japan collaboration. The initial December 4 impeachment motion stated: “Yoon has neglected geopolitical balance, antagonizing North Korea, China, and Russia, adhering to a bizarre Japan-centered foreign policy, and appointing pro-Japan individuals to key government positions.”

Grassroots conservatives, in turn, rejected mainstream media, whose “fake news” defamed their political heroes as pro-Japan collaborators, whether former Presidents Rhee Syngman and Park Chung-hee or current President Yoon. Instead, they turned to social media, especially YouTube, which has become a primary source of information for 53% of South Koreans, including 70% of right-wing rally participants. By disconnecting from mainstream media, conservative media became an echo chamber of uncorroborated theories, such as that progressive lawmakers are secretly collaborating with North Korea or that North Korean hackers infiltrated the National Election Commission.

In the aftermath of the December 3 martial law declaration, grassroots conservatives praised President Yoon for performatively standing up to the Democratic Party and drove out institutional conservatives who disagreed. After resigning as PPP chief on December 16, Han Dong-hoon warned party members, “If we sympathise with extremists like the conspiracy theorists and extreme YouTubers, or if we are consumed by their commercially produced fears, there is no future for conservatism.”

Even as institutional conservatives felt excluded in their party by radicalized conservatives, they felt excluded in mainstream, knowledge-producing institutions by dominant progressives. CM, a female graduate student at a Seoul-based university, criticized the May 18 Act:

Asking a question like “Aren’t they rioters?” is not allowed despite the fact that protesters during the May 18 Gwangju Uprising broke into armories and armed themselves with stolen weapons and ammunition. Some people around me argue that the May 18 Special Act suppresses freedom of expression. I would go further and say that it is a draconian law that also suppresses the freedom to seek the truth.

PJ, a male graduate student, opposed Yoon’s martial law declaration as “too radical,” but also opposed the National Assembly’s swift impeachment of the president, before any police investigation or assembly hearing. PJ preferred the months-long process of investigation and deliberation that led to bipartisan consensus on President Nixon’s resignation. But he declined to openly voice this argument in his university, fearing not just condemnation from progressive faculty and student groups, but also imprisonment from a future DP administration. In his view, DP-appointed judges could interpret existing defamation laws, or legislators might pass a new law (“December 3 Act”) modeled on the May 18 Act, to criminalize statements justifying Yoon’s “fascist” martial law declaration.

Korea-based conservatives such as PJ risk more suppression, including imprisonment, for expressing their views than do American conservatives. But unlike their US counterparts, Korean conservative radicalism was largely limited to supporting President Yoon’s (allegedly) performative, and brief, martial law declaration. They have not yet sought to defund and dismantle mainstream, knowledge-producing institutions (e.g., the education ministry, national research foundations, elite universities), as in the US . Neither do conservative Korean students reject elite universities in favor of conservative alternatives like Hillsdale.

Influenced by Confucian norms, practically all Koreans, conservative or progressive, valorize elite, mostly Seoul-based, universities such as Hanyang and higher-ranked universities abroad. Like the older, Bush-style Republicans in the US, Korean conservatives desire mainstream institutions to be more open to conservative perspectives, not to destroy them. Unlike many MAGAs, most Korean conservatives do not consider reforming the elite universities or the education and scientific ministries to be a lost cause, but a worthy even if difficult mission.

A key difference is that matriculation into elite Korean universities, major newspapers, and government bureaucracies is predominantly determined by standardized exams, not subjective and politicized essays, letters of recommendations, and extracurriculars. Seemingly non-meritocratic, Claudine Gay-style educational elites do not exist in Korea. Therefore, Korean conservatives challenge the progressive majority, but not the core legitimacy of mainstream institutions.

Among OECD democracies, the US and Korea exemplify the challenges of polarization and especially asymmetric polarization. Progressive dominance of mainstream information sources does not justify the self-imposed isolationism and disruptive radicalism of conservatives, culminating in the December 3 martial law declaration in South Korea and the defund, dismantle, and deportation agenda of the second Trump Administration. But it does help us understand that conservatives’ procedural illiberalism and misinformation, and policies based on such misinformation, do not appear in a vacuum, and that any infringements of liberal-procedural norms can polarize and destabilize democracy.

Breaking this cycle requires not more suppression or radicalism, but reaffirming the liberal-procedural norms of open debate and respect for legal rules and customs.

In this conversation, Hanania and Wallach review the historical role of Congress in American politics, focusing on key events from the FDR administration to the present. Wallach explains the important role Congress played in making sure that Roosevelt did not take the country too far down the path of economic statism, a history that Hanania points out is in tension with some of the ideas of Curtis Yarvin.

Throughout the conversation, Wallach emphasizes the importance of Congress in shaping policy and maintaining a balance of power between the branches of government. They explore the impact of significant legislation, the dynamics of the Civil Rights Movement, and changes in Congressional structure during the 1970s. The discussion also touches on the influence of Newt Gingrich, the ongoing challenges of governance in a polarized political environment, the impact of Trump, and the potential for bipartisan cooperation on policy issues. Wallach argues for a more responsive Congress that can find compromises between political factions and effectively address the concerns of the American public.

Hanania plays devil’s advocate, and asks whether it might be better for Congress not to legislate, using European states and the EU as a comparison. Near the end, Hanania asks who the best people are to follow in order to stay informed about Congress. You can find Wallach’s preferred list here.

On December 22, then President-elect Trump announced the appointment of Sriram Krishnan, a successful venture capitalist, as a senior adviser on artificial intelligence. A model immigrant, Krishnan came to the US at 23 to work at Microsoft, later becoming a naturalized citizen. What should have been a discussion of AI policy and a celebration of the success of high-skilled immigration soon turned into a civil war on the right, after nationalist influencers highlighted Krishnan’s past comments endorsing more high-skilled immigration. The H-1B visa program became the focus of the debate, which exposed a deep division among the Trump coalition, between the techno-libertarians led by Elon Musk and the national-populists led by figures like Steve Bannon. Trump eventually came out in favor of the H-1B visa, but it remains to be seen what his administration does on this issue.

At the heart of the objection by the national-populists are arguments based on what economists have called the “fixed pie fallacy,” but in this case can be referred to as the “lump of labor fallacy.” “It’s about taking American jobs,” Bannon has claimed. “A tech company used the H-1B visa to replace [American workers],” JD Vance bemoaned when criticizing the visa as he was running for Senate in 2022. From the other side of the political spectrum, Bernie Sanders chimed in against the program, arguing it replaces “good-paying American jobs with low-wage indentured servants from abroad,” forming an unholy alliance of socialists and nationalists.

According to this logic, there is a set, or at least highly inelastic, number of jobs in the United States. This is true in tech, just like it presumably is in other fields like retail or oil extraction. If more people are allowed in the country, then more Americans are going to either be unemployed or expect much lower wages for their work.

This fallacy has long been known to economists. Adam Smith in The Wealth of Nations (1776), challenged zero-sum assumptions prevalent in mercantilist theories. Later, David Ricardo and John Stuart Mill further explored how trade, innovation, and productivity could expand the economic “pie.” And in the 20th century, President Reagan’s economic guru Milton Friedman famously taught that “most economic fallacies derive from the tendency to assume that there is a fixed pie, that one party can gain only at the expense of another.”

Yet this is not how the economy works. Newcomers do in a sense compete for jobs within one industry. But they also buy homes, food, clothing, and shelter within the United States. If employers prefer the labor of immigrants, whether due to them being better workers, cheaper, or some combination of the two, then it must mean that firms are likely to become more efficient after hiring them, further stimulating economic growth. If a tech firm pays less money for labor, even assuming no gain in efficiency otherwise, they do not simply burn the money but invest it or put it back into the economy by returning it to shareholders. Even if all the savings go to CEO compensation, and executives spend that money rather than place it in the stock market, this still increases demand in the economy.

The lump of labor fallacy assumes that the demand for labor is static: there is a set number of jobs and that introducing more workers must reduce employment opportunities for those already in the workforce. American history is filled with counterexamples. When the US gradually allowed women into the workforce, it didn’t cause massive job losses among men; when segregation was abolished in the South, black workers didn’t cause mass unemployment among whites; when the 13 colonies established free movement of people among the states with the Articles of Confederation in 1781, this did not cause economic upheaval. The vast majority of native-born Americans are descended from immigrants, including the national-populists vehemently opposed to H-1B visas. Did their ancestors take away American jobs? Or did they contribute to the economy by working with the natives?

Even setting aside their roles as consumers, what restrictionists fail to realize is that high-skilled immigrants not only take existing jobs, but create new ones by becoming entrepreneurs themselves. The most salient statistic is perhaps the fact that 45% of Fortune 500 companies were founded by immigrants or children of immigrants. In AI, a report from the National Foundation for American Policy concluded immigrants also founded or co-founded 65% of the top companies. Elon Musk, maybe the greatest entrepreneur in history, was once on an H-1B visa. All of this has happened in spite of the restrictiveness of the immigration system. In many ways, high-skilled immigrants are the lifeblood of American innovation. Even low-skilled immigrants, by taking the most unglamorous and unpleasant jobs, enable natives to be productive in other ways by lowering the costs of labor. One may have other arguments against immigration, but the economic case based on a fixed number of jobs makes no sense.

The fixed pie fallacy fails not only in the context of the labor market. As alluded to above, immigrants raise demand for goods and services, including those that natives are uniquely positioned to provide, such as education, legal services, entertainment, and management. If the pie were truly fixed, a virus that randomly killed half of Americans would make everyone twice as rich.

Housing is an area where the pie is close to being fixed in many regions, but this is only the result of zoning laws that restrict supply. There are an endless number of reasons why we should make housing easier to build, and the payoffs to doing so are potentially massive. But to restrict immigration on that basis alone would mean responding to one broken policy by hurting the economy in another.

What all this means is that not only are H-1B workers highly-paid, with median earnings at around the 90th percentile of wages, but each H-1B worker is estimated to create 1.83 jobs for native-born Americans. This is not surprising at all. As high earners, H1-B recipients must be a net benefit to the economy. In no other context would anyone argue that workers in the top decile of productivity are making Americans poorer.

Callum Williams, senior economics editor at The Economist, has argued that having a much larger common market goes a long way towards explaining why Americans are so much wealthier than Europeans. In a tiny village where individuals only trade with members of their own community, compared to a large developed country, it is less likely there will be enough demand to make being the owner of a boutique bookstore, a Substacker, or an exterminator a realistic career path. This is why, for example, American restaurants will commonly have different jobs for hostesses and waiters, while in Europe one employee will take on both roles. More people means greater economies of scale. When newcomers are well above average in productivity, as H1-B recipients are, opening the borders to them becomes a no-brainer.

What if you only wanted to focus on wages for tech workers? One might acknowledge that highly productive workers making businesses more efficient are good for the economy as a whole, but can we at least say that a certain class of worker they compete with is worse off?

That is far from certain. Noah Smith has an excellent recent summary of papers arguing that H-1B visas don’t even hurt workers that are in direct competition with new arrivals. We have a strong basis for believing this because H-1B visas are distributed based on a lottery system, so we can treat the process as a random experiment and see what happens to the wages of individuals working for firms that win the lottery versus those that don’t. Workers in the firms that get immigrant labor end up hiring more tech workers and paying them more, because they become more efficient and sometimes scale up.

Of course, immigration policy shouldn’t be determined on the basis of what is good for workers in one industry, but what is good for the nation as a whole. So even if you do not find such studies convincing, you should want more productive workers to come into the country because a welcoming policy still makes society better off. This is not controversial among economists, as the question gets to the basic assumptions of their field. A 2016 poll of economists showed exactly 0% disagreeing with the idea that letting employers hire more immigrants with STEM degrees would raise per person income in the US in the long run.

To be fair, an overwhelming majority also said that such a policy would at least temporarily lower the premium earned by workers with similar degrees. However, it must be noted that this survey was conducted before most of the studies cited by Noah Smith in his recent article. It would be interesting to see what economists who are familiar with the literature that has come out over the last few years think. But, again, even if it is true that immigrants directly suppress the wages of workers they compete with, public policy being designed to help subsets of society rather than the entire country is the path to stagnation. If we try to “protect” the workers in every industry from competition, we will all end up poorer.

Unfortunately, zero-sum thinking tends to be common when individuals consider economic issues. People often misjudge market interactions as win-lose rather than win-win, especially when foreigners are part of the equation. Irresponsible politicians and pundits take advantage of such biases, demagoguing bad policies into existence, to the detriment of society.

Yet letting in more immigrants is not a form of charity on the behalf of Americans. It’s an arrangement that expands the pie, benefiting both newcomers and natives. America is blessed with being the top talent magnet in the world. It would be a shame to waste one of its greatest assets.

One thing that is highlighted in this conversation is the problem with the “just taking questions” posture adopted by the anti-vaxx movement. Most of the questions they ask have been answered. Those that remain do not create enough doubt to suggest a person should not get the vaccines recommended by public health professionals. For more along these lines, see Richard Hanania, “Vaccines and the Tightrope of Progress.” This interview originally appeared here.

The themes touched on here will be familiar to those who have read about the “invisible graveyard” that the FDA is responsible for. Yet even listeners who know about the utter lack of interest in patient well being normally shown by federal agencies will find themselves shocked by the degree to which bureaucratic procedures with few plausible benefits govern the lives of sick individuals who want nothing but to get some extra time on this earth and help move science forward. For dealing with the clinical trial system in its current state, Bess is currently trying to figure out ways to assist oncologists and patients in being able to navigate the process at HelpMeFindAClinicalTrial.com. And hopefully by telling her story, she can help inspire much needed reforms to the system.

Romina joins the podcast to discuss available paths to deal with the coming entitlement crisis. One potential way to get politicians out of making tough choices is to create a debt commission that takes responsibility for unpopular reforms. Romina has written about using the model of the BRAC commission, which was relied on to close down military bases at the end of the Cold War.

The conversation also touches on the politics of debt, how policymakers are thinking about these issues, Paul Ryan as an unappreciated hero of our time, and much more. Near the end, Romina reflects on her career as a DC policy-wonk, and why she is motivated to help ensure that America continues to be the land of opportunity. If we don’t get entitlements under control, it could potentially degrade our entire way of life. For more discussion on this topic and the difficult choices our leaders will soon be facing, see the previous CSPI podcast with Brian Riedl.

Listen to the podcast with Romina here or watch the video on YouTube.

He joins the podcast to discuss why he’s not worried about the alignment problem, where he disagrees with “doomers,” the accomplishments of ChatGPT versus DALL-E, the dangers of regulating AI until progress comes to a halt in the way it did with nuclear power, and more. With his background in computer science, Brian takes issue with many of those who write on this topic, arguing that they think in terms of flawed analogies and know little about the underlying technology. The conversation touches on a previous CSPI discussion with Leopold Aschenbrenner, and the value of continuing to work on alignment.

Brian’s view is that AI doomers are making people needlessly pessimistic. He believes that this technology has the potential to do great things for humanity, particularly when it comes to areas like software development and biotech. But the post-World War II era has seen many examples of government hindering progress, and AFF is dedicated to stopping that from happening with artificial intelligence.

Listen to the conversation here, or watch the video here.

Links

Donate to AFF

AFF manifesto

Brian on diminishing returns to machine learning, and discussing AI with Marc Andreessen

Vaswani et al. on transformers

Limits of current machine learning techniques

• All across the world, there is a negative correlation between IQ and fertility.

• This means that populations are getting less intelligent as time goes on. Through some basic assumptions, across 62 nations for which we have data, one can estimate a decline in IQ per decade that ranges from as low as 0.01 points in Estonia and Switzerland to as high as 0.65 in Panama, Romania, and North Macedonia.

• In general, more developed countries see a weaker decline in intelligence than less developed nations.

• Selection for lower IQ does not appear to be a major long-term problem in the richest countries today. If at some point even 20% of the population uses embryo selection, it would be more than enough to reverse the current trends that we are seeing.

• There may be more reasons to be concerned about developing countries. The fact that the rate of decline is so fast implies that even if IQ differences between nations are completely environmentally determined today, over the coming decades there may still be a significant divergence between them. That said, the data here may exaggerate the scale of the problem if large family size depresses phenotypic IQ in the developing world.

Introduction

In the 2006 film Idiocracy, a young American soldier of average intelligence wakes up 500 years after a botched hibernation experiment and finds himself to be the smartest man in the world. A narrator explains that over the years, humanity got progressively dumber as the less intelligent outbred smarter individuals in one generation after another.

While a work of fiction, it is reasonable to ask whether the dystopia described in the film gets at an issue with real world significance. There is certainly a stereotype that more intelligent people have fewer children, and as we know, stereotypes generally have a basis in fact. Whether this one is true is certainly worth looking into, given how important intelligence is for human flourishing. Moreover, if we are concerned with humanity as a whole, it is important to analyze differential fertility by intelligence not only in United States and other Western countries, but across the world.

Along with some co-authors, I have gathered relevant data on this question from 62 countries. In general, our research shows that humans are likely getting dumber, with the problem being much more pronounced in the developing world. If the data is correct and reflects a decline in genotypic intelligence, it may cause problems for poor states trying to catch up to first world nations, and ultimately threaten their economic health and political stability.

Past Work on IQ and Fertility

Scholars have for generations argued that the fears expressed in works like Idiocracy are supported by empirical data. Early research on the relationship between fertility and intelligence found that the variables were negatively correlated, albeit weakly.1 Subsequent work vindicates these early studies, with a recent meta-analysis finding much the same thing.2 Despite this previous research, there is very little agreement on the magnitude of the decline in IQ per decade, with estimates ranging from 0.14 to 0.64 points.3 In addition, most of the data on fertility and intelligence comes from the USA and Britain, and few studies have been done on other regions.

Other critics of the differential fertility literature argue that the Flynn effect, the global increase in IQ scores of about 2-3 points per decade,4 is evidence against differential fertility resulting in decreases in intelligence in future generations. However, IQ tests do not measure the same construct across time,5 and a portion of Flynn effect increases can be attributed to increases in guessing.6 Besides these issues, genetic predisposition to educational attainment fell in Iceland by about 0.1 standard units, the equivalent of 1.5 IQ points, in the 20th century, confirming the hypothesis that differential fertility leads to changes in a population’s genetic makeup.7

Other researchers have estimated that average levels of intelligence in the West have fallen by the equivalent of 13 IQ points since the Victorian era based on declines in reaction time.8 This is much larger than what could be reasonably inferred based on differential fertility. Instead, the primary cause of this decline is hypothesized to be mutational load. However, this study suffers from certain methodological issues, notably that some of the early samples were not representative of the general population and that only 14 samples were available.9 Woodley et al. (2013) responded to some of these critiques by arguing that the difference in reaction time between 19th century and modern students is similar to the overall decline in reaction time observed from their samples.

Fortunately, there is a better study on this topic that compares secular trends in the number of digits people can remember (the ‘digit span’ test), depending on whether they must be repeated forwards or backwards.10 The authors involved gathered over 1,700 effect sizes in total and controlled for the effect of moderators, so this study does not share the same concerns as the previous paper mentioned. It found that performance increased for forward digit span tasks, but decreased for backwards digit span tasks in recent years. The latter correlates more with IQ because it is a more cognitively complex task, so this is suggestive evidence that general intelligence has fallen with time. This also holds true for corsi tasks, where the test taker is asked to tap a given sequence of blocks forwards or backwards. Using the summary statistics from the previous study, I calculated that, based on the change in backward digit span, the magnitude of the decline in general intelligence is about 2 points per century.

Moreover, a decline in intelligence should result in falls in quantitative and verbal abilities of students if the quality of education has stayed the same over time. Fortunately, the National Assessment of Educational Progress has tracked the achievement of American students in math and reading. Over time, Americans have hardly changed in their reading or mathematical abilities. In 1970, 13-year-old students scored 255 out of 500 in reading and 266 out of 500 in mathematics, and in 2023 they scored 256 in reading and 271 in mathematics. This is confounded a bit by immigration, but one might suspect that if anything that makes the arguments against a drop in IQ among Americans who have lived in the country for generations even stronger, given that most new residents over the past several decades have been low-skilled.

If mutational load had such a large effect on IQ, then we might suspect that there would be stronger evidence for its existence by now. However, trends in mathematical performance, reading comprehension, and backwards digit span suggest that the average IQ in the United States has not fallen by a large amount. Sure, these tests do not measure intelligence directly, but if intelligence really was dropping, then performance on at least some of them would be expected to fall as well.

There is a lot of uncertainty in the literature regarding the exact effect of mutational load on the decline in IQ per decade (95% CI: [0.14, 1.5]), based on statistics taken from Woodley (2015). Also, recent research on mutational load finds an effect that is severely attenuated and not statistically significant after controlling for birth order.11 Based on theory, mutational load should result in decreases in IQ over time, but the magnitude of the effect is difficult to measure.

New Cross-National Data

To add to the literature on this topic, I collaborated with Bryan Pesta, George Francis, and Emil Kirkegaard on a study that assessed the relationship between intelligence and fertility in as many countries as possible, and projected the change in the average IQ of each nation and the entire world based on the available data. We were able to gather 156 effect sizes (n = 419,444) from a combination of the academic literature and public data, in particular the datasets of the National Longitudinal Study of Youth (NLSY), General Social Survey (GSS), Programme for International Student Assessment (PISA), Progress in International Reading Literacy Study (PIRLS), and Program for International Assessment of Adult Competencies (PIAAC). Not all of these sources directly test IQ  — proxies such as performance in mathematics or reading had to be used in most cases. In some sources, such as the PISA and PIRLS datasets, the IQ of the children was tested instead of the IQ of the adults.

Unfortunately, only 62 countries had data on the relationship between intelligence and fertility, so imputations had to be made for the others. These imputations were estimated using the correlation between educational attainment and fertility, with, depending on the country, national IQ, socioeconomic development, or geographic estimates. For the sake of transparency, all of the figures in this article use data from countries that tested intelligence or a reasonable proxy for it (e.g., mathematical ability, reading comprehension). These projections of the average IQ of each country were then combined with the United Nations’ population projections to estimate the change in the world’s IQ from 2023 to 2100. Note that this analysis does not take into account the possibility that IQ can improve due to socioeconomic development or cultural changes. This may matter in developing nations, but is much less likely to in wealthier countries.

To measure the relationship between intelligence and fertility, we used a statistic called the selection differential: the difference between a reference population and the parents of that population weighted by the number of children each parent has. Then, this number is multiplied by the heritability of IQ to generate a different statistic called the response to selection, which is the expected genetic decline in a trait in one generation. So if the selection differential for IQ is 2 points, and the additive heritability of IQ is 80%, then the response to selection is 1.6, which is the expected genetic decline in that trait in the next generation. As the correlation between the average IQ of two parents and their child’s IQ is about 0.6,12 selection differentials found were shrunk by a factor of 0.6 to estimate the expected decline in genetic intelligence within the next generations. To calculate the projected decline in genotypic IQ per decade in each country, these selection differentials were divided by the average maternal age within each country to calculate the decline per year, and then that decline was multiplied by 10.

This may be an underestimate or overestimate of the true future decline depending on several factors, notably mutational load, environmental shifts, measurement error, and class based differences in generational turnover. Unfortunately, these variable cannot be modeled given the information that we have, so they have not been taken into consideration when projecting the future IQs of different countries.

While I was collecting the data, I noticed that the selection differentials for IQ varied considerably depending on the source. For example, the selection differential for IQ in the USA is about -2.4 based in the NLSY datasets, which were considered the best sources for that country due to the fact IQ was measured very accurately, they are large samples, most individuals were close to completed fertility (ages of 51-64 in NLSY79, 35-38 in NLSY97), and the samples are roughly representative of the nation as a whole. In comparison, the selection differential in the international datasets is -1.2.

It is unclear why this is the case. Perhaps the tests used are of lower quality, or the United States was not sampled in a representative fashion. The NLSY was relied on for the “regional” study of the US, meaning that it did not apply to other countries. It uses a very exhaustive measure of cognitive ability, the ASVAB, while the international datasets only have rough proxies for cognitive ability, which are comprised of 3 subtests in the case of the PISA dataset, one subtest in the case of the PIRLS, and 2-3 subtests in the case of the PIAAC dataset. In addition, the PIAAC dataset had uniquely high estimates of the selection differentials, with the average country having a selection differential for IQ 1.07 points higher than what would be normally expected. The differences in the magnitude of the selection differentials between datasets hold up well between countries, so this effect is not an artefact of different countries being present in each dataset.

To correct for the bias introduced by the international (PISA, PIRLS, PIAAC) datasets, two different methods were considered. One assumed that the bias is multiplicative, that is, the international datasets shrink effect sizes by a factor of 0.5; or that the effect is additive, and the international datasets reduce the selection differentials for IQ by 1.2 points. Because it’s not clear which method is better, the results of both were averaged into one final estimation. Besides this adjustment, the effect of specific datasets was also considered, so the effects of individual datasets wouldn’t bias the estimation of the decline between countries.

If the reason why the international datasets have attenuated effect sizes within the United States is sampling error, correcting for this bias will potentially lead to deflated selection differentials within smaller countries, which are easier to sample. The reasoning is as follows: if sampling error is leading to attenuated effect sizes within the international datasets, then correcting for the effect observed within a large country like the United States will overcorrect for the effect of sampling error within other countries that are not as large.

Based on our data, we found that the average country is expected to decline in IQ by 0.35 points per decade, while the whole world is expected to decline by 1.1 points per decade during this century. The reason that the global decline is larger than the decline within the individual countries is that lower IQ nations are expected have higher growth rates. Fortunately, the rate at which global intelligence is falling is slowing down; by the end of the century the decline will only be 0.8 points per decade.

Countries also vary considerably in differential fertility for intelligence. In Iceland, Finland, Denmark, Estonia, and Switzerland, fertility and intelligence are negligibly related, while Latin American and certain Middle Eastern countries have a stronger negative relationship. Part of this is because the latter countries have younger mothers, so selection for intelligence occurs faster independent of the relationship between fertility and intelligence within a generation. But this doesn’t explain the whole relationship, as countries only vary so much in their average maternal ages; the number ranges from 25.7 to 33.4, while the decline in IQ per decade varies much more, from 0.01 in Switzerland to 0.65 in Panama.

Selection differentials for intelligence correlated with both national IQ (r = 0.51) and socioeconomic development (s-factor) (r = 0.48). As s-factor and national IQ are highly correlated (r = 0.8 in this dataset), it is difficult to tell which of these variables is more likely to be driving the relationship.

Prior to conducting the study, I had hypothesized that socioeconomic development past industrialization had a positive effect on selection differentials for IQ, because developed countries have more resources, so a slower life history. In addition, wealthier nations have more variance in years spent in the education system,13 so if education has a negative effect on fertility, the relationship between the two variables should be stronger in developing countries because of lower restriction of range. According to our data, socioeconomic development and the correlation between fertility and education are correlated between countries (r = .27, p < .01), though the relationship is weak and potentially spurious.

Another reason that less developed countries have more negative relationships between intelligence and fertility is that, at the individual level, early births tend to be disproportionately from less intelligent individuals, while later births follow the opposite pattern.14 As developed countries tend to have higher average ages of births (r = .37, p < .001), differential fertility for IQ and socioeconomic development should positively correlate.

The cause of the relationship between national IQ and the selection differential for IQ is opaque — it could be because the average IQ of a nation also measures its social development, or because there is some causal relationship between the two variables.

Alternatively, the relationship between socioeconomic development and the selection differential could be an artefact that arises out of using the IQs of children to estimate the selection differential instead of the parents. If family size has a more negative causal effect on the IQs of children in the developing world, then a positive relationship between the estimated selection differentials and social development will be found, but only because the causality occurs in the opposite direction — that is, larger families cause lower IQs, rather than the other way around. Because the heritability of fertility is very low,15 this environmental effect won’t be transmitted in any large degree to future generations, so the decline in IQ of these nations may be overestimated.

Implications and Conclusion

My main takeaway from the study is probably going to surprise most readers: future reductions in intelligence due to differential fertility for IQ are not going to be a major issue in the developed world. This is because the process is happening too slowly for it to have a significant effect; by 2123, the most concerning drop in the developing world is the US falling from an IQ of 97.5 to 93.7. Most European countries will have hardly budged by 2123, besides a few in the Balkans: Romania, Macedonia, and Albania are projected to drop from 87 to 80.5, 80 to 73.4, and 82.1 to 77.5 respectively. All of this of course assumes that whatever rates of differential fertility by IQ exist now will continue into future generations, which will be true to a greater or lesser extent depending on the country.

Besides the fact that the effect is weak within most developed nations, embryo selection for intelligence, and potentially other technologies, can attenuate secular decreases in intelligence, though I doubt the practice will be adopted by most of the general population. The average member of the public estimates their likelihood of using embryo selection to boost intelligence at 35%, with 30% morally opposed to the practice.16 Data from Pew Research (2018) suggests that favorability among the public may be even lower, with only 19% of Americans saying that selecting embryos for intelligence is an appropriate use of medical technology, while 80% think that it takes science too far. Embryo selection is also very expensive at the moment, with clinics requesting between $4,000 to $20,000 for the service, so some of the population will be priced out.17 Nonetheless, it is historically common for a practice to be rejected by the public at first and then have opinion towards it shift as its availability increases and its benefits become clear. Therefore, polls should be seen as reflecting a lower bound estimate for how many people might use a technology.

Assuming that 20% of Westerners in the future select their embryos for intelligence, half of those 20% only select for intelligence, and 20% of the causal variance in cognitive ability can be estimated with genes, then embryo selection will boost the average IQ of the population by approximately 0.73 points per generation (math done with gwern’s help). This is roughly equivalent to an increase of 0.24 IQ points per decade, which is enough for most developed countries to maintain their current levels of intelligence, assuming that the relationship between intelligence and fertility remains the same. Theoretically, genetic predictors should be able to explain up to 50% of the variance in intelligence, so this may be an underestimate of the impact of embryo selection.

Some commentators like Anatoly Karlin argue that concerns about national IQ will not be as relevant once strong artificial intelligence has been developed, as lots of white collar work will then be automatable. This is true, but there are other reasons why national IQ would continue to be useful in the future; barring rapid technological breakthroughs, manual labor will still need to be done by humans, and job performance and IQ correlate at about 0.4, even within categories like rifleman (r = .59) and military policeman (r = .46).18

Across individuals, low IQ is a predictor of marital instability,19 obesity,20 and criminality,21 and the relationships are likely to be causal as they cannot be ascribed to confounders. This means that declines in national intelligence will lead to lower levels of wealth and social peace, and potentially civilizational collapse if the fall is too severe.

What is most concerning is the strong negative relationship between fertility and intelligence that is observed in some developing countries, like Turkey and Brazil. By the end of this century, these countries will have dropped by 5.5 and 6.4 points respectively. The effectiveness of embryo selection for intelligence drops with genetic distance from the reference population.22 This technique may therefore be effective for populations that are genetically close to Europeans like white Brazilians or Turks, but less so for people from Central America or Southeast Asia. Even worse, countries that are less developed are less likely to adopt embryo selection and similar practices in the first place, and will need to use the technology more to sustain their current levels of intelligence.

Figures for changes in IQ within nations and across the world should not be taken as exact estimates. They depend on many assumptions: for example, that the rate of differential selection by IQ doesn’t shift over time, the heritability of IQ stays constant, and nations do not see a rise or fall in IQ anywhere due to cultural or technological changes unrelated to patterns of fertility. It is certain that not all of these assumptions will hold for all countries indefinitely into the future. Nonetheless, given the importance of intelligence for accomplishing practically all human goals, it is worth undertaking back-of-the-envelope calculations to give us some idea of how worried we should be. The exercise is enough to determine that selection for low IQ is a minor problem in wealthier states, and potentially much more concerning in developing countries.

The conversation begins with a discussion of different counterfactuals regarding ways in which Napoleon might have been able to stay in power, which leads to Roberts explaining his view that the wars of the era could be understood at least in part as resulting from a rejection of free trade. Other topics include:

• Meritocracy as a guiding principle of the French Revolution and a justification for Napoleon’s regime

• Napoleon’s personal magnetism and why men were willing to follow him

• The relationship with Josephine, and whether or not it influenced any of his political decision

• Whether Napoleon was in fact the greatest general of his time

See also Hanania’s audio review of the Ridley Scott film, and Roberts’ reviews in Commentary and The Times. For an edited transcript of this conversation, see here.

He joins the podcast to talk about the financial future of the United States, with a special focus on entitlements. Medicare is projected to run out by 2031, and Social Security only two years later. Because of politicians kicking the can down the road for so long, this will mean that the federal government will at that point have to either implement massive benefit cuts for seniors or significantly raise taxes across the board.

Brian talks about his experience in Washington, the history of negotiations over the debt, and what politicians say when you bring up these facts. We appear to be in an undesirable equilibrium, where everyone’s incentive is to ignore the issues involved, while the status quo is leading us towards disaster. Despite liberals wanting to tax the rich and conservatives calling for a cut to foreign aid and non-entitlement forms of domestic spending, the numbers for such proposals don’t add up. We will either get entitlement spending under control, or become taxed at the level of Europeans.

In one important way, we will actually be worse off than Europe, because their welfare states pay for services and benefits that go to families across a wide section of the population. We are potentially building a US welfare state that will have high taxes primarily to funnel money to the elderly. The fact that older Americans are richer than those who will be supporting them makes the future we are moving towards even more absurd.

Links

Brian Riedl: chart book on spending, report on the limits of taxing the rich, CNN op-ed on interest rates, NYT op-ed on Biden’s promises on entitlements

Brian’s X page, Manhattan Institute website

This was quite an optimistic conversation. The title of the podcast comes from the last thing Niklas said, which was that you don’t actually need attention or to talk about grand projects, but just to show the world what you can do.

Niklas is part of the charter city movement, which seeks to build hubs of innovation and progress while bringing the rule of law and economic development to poorer regions of the world. In this eventful conversation, Richard and Niklas touch on

• The mechanics of governance in Próspera

• Getting around red tape and becoming a hub of medical innovation

• Amenities and quality of life in the city

• Upcoming conferences and events

Despite a new government in Honduras that is hostile to charter cities, Niklas is optimistic that they will be able to continue operating. He and Richard also talk about potential medical breakthroughs that Próspera might help bring about, like bacteria that remove cavities from your mouth, and a currently available gene therapy that may make your muscles and bones stronger.

Links

Niklas on X, his Substack, RSS for his podcast

The Ultimate Guide to Próspera

Alex Ugorji on X

Próspera website

Ciudad Morazán

Infinita Manifesto

Scott Alexander on Próspera, Part I and Part II

Mark Lutter on the CSPI podcast

Marc Andreessen, The Techno-Optimist Manifesto

Documentary on medical tourism in Próspera; DW report, with appearance from Niklas

Upcoming Events

Nov 3-5: Crypto Futurism & Legal Engineering 2023 - A Próspera Builders’ Summit

Nov 17-19: DeSci & Longevity Biotech 2023 - A Próspera Builders' Summit

Jan 6-Mar 1: Vitalia - Starting the Frontier City of Life

The winner of the $25,000 prize, and the Salem Center fellowship for 2023-2024, is Jonathan Zubkoff, better known to other participants as zubbybadger. We also had a $5,000 award for the person who most improved their portfolio between February 15 and July 31, and the winner was Malte Schrödl, a college student studying in Aachen, Germany. Winners of the Gold, Silver, and Bronze prizes have all been notified, as have the top 20 finishers, who will be invited to a forecasting tournament taking place at the Salem Center in the fall.

The second place finisher in the tournament was at the end going by the handle “guyin2ndplace.” He’s written a very interesting account of his experience. See Part I and Part II. For other commentary on the tournament, see Astral Codex Ten. One can browse the history of all of the markets here, and get historical data related to activity on the site here.

This report is meant to provide a summary of what we have learned. Part I deals with the issue of calibration. How good were our betting markets at actually predicting the future? The rest of the report looks at the participants themselves. Part II gives summary statistics on their backgrounds and political views. Finally, Part III discusses what variables correlated with being a good or bad forecaster.

Data used for the calibration analysis is being made available at the bottom of this report. Most of the data related to the characteristics of bettors is being kept under wraps due to privacy concerns, but we are open to collaborating with those who would like to explore the results. All of the analysis below should be taken as preliminary. We invite others to look for further insights that can be found in the data.

I. How Good Was the Market?

I start by taking the beginning of day probability for each market and each day, and then finding an average for each market across the entire time period. From that, I calculated a Brier score for each individual market. The average across all markets was 0.133. The biggest misses were on the following questions, listed with their average probability and eventual outcome.

1. Republicans favored in Senate on Election Day?: 34.6% chance, settled YES

2. Paul Vallas mayor of Chicago?: 63.9%, settled NO

3. Biden cancels student debt in 2022?: 38.1%, settled YES

4. Will PredictIt survive?: 38.6%, settled YES

5. Michigan abortion rights landslide?: 57.1%, settled NO

I don’t see much of a pattern here. One might’ve thought that the market would be biased in the direction of the political leanings of participants, which are discussed below. But they are more conservative than liberal, and their biggest miss was not believing Republicans would be favored in the Senate on election day, when in fact they were. I also assume that participants wanted PredictIt to survive, given that they were presumably supporters of prediction markets. Overestimating Paul Vallas could’ve been a mistake based on their political and class orientation, but the upset in the Chicago mayoral race seems to have been something a lot of people missed.

For the next analysis, we take the first probability of each market every day across the whole tournament. This gives us 12,521 observations. Since there were 91 markets, that means each market existed for an average of just under 138 days.

I divide the observations into 99 bins, from 1% to 99%, and then calculate the average outcome. For example, the market on whether California would ban flavored tobacco was at 58% at the beginning of the day on August 23, and that market ended up settling as YES. Thus, the 58% likelihood was one observation and the dependent variable connected to it was 1. That observation was grouped in with all other observations that rounded to a 58% likelihood, and this was done for each observation in the dataset.

The results are in Figure 1 below. Dotted lines show 95% confidence intervals as calculated using the Adjusted Wald Technique developed for data with binomial outcomes.

Before the start of the tournament, it was known that by having prizes given at the end, individuals would have an incentive to bet on longshot probability events. Since there’s no advantage to being at, say, the 50th percentile over the 10th percentile, participants would have been smart to take large risks.

Figure 1 shows that this is exactly what they did. Of the 2,504 datapoints for which the market gave the outcome a 17% chance or less of happening, the market resolved YES exactly 0 times. For outcomes that the market gave a 20% chance of happening, the outcome was YES only 6% of the time.

The orange line in Figure 1 represents perfect calibration. Ideally, the blue line would hover around it. But it doesn’t at any point cross the perfect calibration (orange) line until the predicted probability of events is 47%. For high probability events, at every point on the x-axis beginning at 70% each bin had outcomes that were more likely to settle as YES than what the market predicted.

Basically, the market lets you divide events into four categories. Things that are given a 20% or less chance of happening have close to a 0% chance of happening in the real world. Between 20-45% the market becomes more useful but systematically underestimates the likelihood of events. Only between around 45%-70% does the market behave as one would wish. Anything that is given a higher probability than that is close to a sure thing. The problem with betting on longshots seems to be worse at the left side of the graph. In other words, people like the idea of a longshot YES bet over a longshot NO bet. There is probably a reason for this out there in some psychological literature. Or is this perhaps due to non-randomness in how we framed the questions, that is, which outcomes we chose to call YES and which we called NO?

This is pretty bad relative to real prediction markets. Here’s one study of PredictIt, based on prices one month before resolution.

One sees the same problem, with the calibration line being below the perfect calibration line at low prices and above it at higher prices. Nonetheless, this is clearly nowhere near as bad as our own chart. Markets that are based on prizes, like ours was, are clearly not useless. Since the errors are somewhat predictable, they provide good information. One can assume that things are correctly priced near the center of the probability distribution, overpriced at the low end, and underpriced at the high end. We are therefore left with data supporting the well-established but nonetheless important conclusion that prize markets are better than nothing, with markets based on money, or some adequate substitute that doesn’t over incentivize longshot bets, being much superior for predicting high and low probability events.

II. The Bettors

We had 548 individuals take the survey. Of those, 95.1% were male, compared to 4.3% female, and less than 1% other. It was a young crowd, with 48.8% aged 28-42, and another 35.7% aged 27 or younger. 88.9% were straight, with 5.5% bisexual.

Just under 74% were from the United States, and close to 85% identified as white. The sample was extremely secular compared to the general population, with 53% identifying as atheist or agnostic. 77.8% had graduated from college, and 39.7% had some kind of post-graduate degree.

The most common fields were tech (24.7%), academia (grad student or research fellow) (11.4%), government (7.7%), academia (professor) (6%), entrepreneur (non-tech) (6%), academia (other) (5.8%), nonprofit (5.6%), banking (5.2%), tech (business owner) (5.2%), politics (2.8%), and think tank (2.8%).

Individuals were asked to describe themselves politically on a five-point scale, from very liberal (1), to moderate (3), to very conservative (5). The survey takers rated themselves on average 3.26 on economic issues, 2.79 on social issues, and 3.07 overall. In other words, the sample can be described as broadly libertarian leaning.

More precisely, the field can be described as pro-democratic capitalism, of a techno-optimist flavor.

We know this because we asked respondents to rate 78 different ideas, people, and concepts, from Donald Trump to bringing back dinosaurs to embryo selection to banning pornography.

The four statements and ideas they rated most favorably were “I believe people are better off than at any other point in history” (4.07 on a 5-point scale); “I believe capitalism is a better economic system than all available alternatives” (4.07); “I believe anthropogenic climate change is real” (4.01); and “Democracy” (3.72). The least popular were “There should be no law against a boss pressuring an employee for sex” (1.82); “It is more likely than not that the Democrats stole the 2020 election” (1.85); “I believe humans have made contact with aliens” (1.85) (note this was before that topic started to be in the news); and “I believe people in developed countries should have fewer children” (1.89).

As found in previous studies of this kind, participants were more likely than chance to be first borns. 55.1% of respondents reported having no older siblings, compared to 34% who reported no younger siblings. With a sample size of 492 individuals answering both questions, this result passes all conventional thresholds of significance.

III. What Makes a Good Forecaster?

Figure 3 shows the twenty statements with the strongest correlation, positive or negative, with being a good forecaster, defined as the balance in S$ an individual ended up with. For this analysis, I only included bettors who placed at least S$50 on 5 different markets and answered all item questions, of which there were 192 individuals.

If there’s one takeaway from the chart above, it’s that the best forecasters liked left-wing ideas, people, and movements. The best predictor was agreeing with the statement that liberals are more honest than conservatives. Since I’ve written articles to this effect, this warmed my heart. But that result was just part of a general pattern of liberals being better forecasters than conservatives. A univariate regression likewise shows that those who identified as more conservative on the five-point scale ended up with lower balances, with the result just missing the conventional threshold of significance (p = .056).

There are two plausible theories here, other than the possibility that liberals are just superior at predicting the future. First, since this was a highly selected group, it may have been that liberals who participated in the Salem/CSPI tournament were of a higher quality, because they were familiar with and willing to participate in an event that was coded as being ideologically opposite of them. Recall that the crowd leaned more conservative overall, albeit slightly. For reference, participants rated BLM (2.46) higher than Biden (2.34) and Trump (2.1), but lower than libertarianism (3.21) or DeSantis (2.86). The Democratic Party (2.06) and the Republican Party (2.08) were rated about equally negatively.

The second possibility is that the liberal advantage is explained by the results of the midterms. Of our 91 markets, 18 had to do with the American elections held in November 2022. Imagine if each side tended to bet on how they wanted things to turn out. Then liberals would have ended up doing much better, because Democrats outperformed the conventional wisdom. PredictIt gave Republicans a 74% chance of winning the Senate on the day before the election, and the FiveThirtyEight model ended up at 59%.

This led me to wonder, what happens if you simply take out the 18 midterm markets, and see how people did in the other 73 markets? Figure 4 shows the results for the top 20 items most strongly related to forecasting outcomes not having to do with the midterms, otherwise using the same methodology as described for Figure 3.

Suddenly, conservatives are the better forecasters. If you want someone’s advice, you should look for an individual who thinks theology should dominate public life, incest definitely shouldn’t be legal, and Trump was robbed in 2020. Unless, that is, you want to predict the 2022 midterm results. The results in Figure 3 and Figure 4 are not driven by outliers. Removing the biggest winners of the tournament does not change the pattern of liberal respondents doing better overall and conservatives performing better in non-midterm elections.

For both Figure 3 and Figure 4, there is an obvious potential multiple comparisons issue. When a Benjamini-Hochberg FDR correction is applied to the p-values, 7 items remain significant in Figure 3 and 5 in Figure 4.

Is the conclusion that nobody is better than anyone else? Mostly, yes. The relationships found are quite weak, with very few items having a correlation of at least r =+/- 0.2 with forecasting results. Yet I think we do have to give the advantage to liberals. It would be unfair to discount their accomplishment just because much of it was driven by the 2022 midterms. If conservatives were too optimistic about their own side winning, then that is their fault. Maybe Democrats would’ve been too optimistic about their chances in a different year, but we can’t just assume that they would have been. So credit to the left leaning participants, but keep the two caveats in mind that they didn’t do as well in non-midterm markets, and the ones who participated in this tournament are a highly selected group. While the midterm story can explain the overall results, it doesn’t provide an obvious hypothesis as to why conservatives would be better in non-midterm markets.

Data

I’d like to encourage anyone who is interested to explore the data for themselves. Please reach out if you do at contact@cspicenter.org, as we would be interested in hearing about and publicizing your results. All of the analysis in this report was done in Stata or Excel. The spreadsheets to reproduce the data in Part I can be found below, along with the codebook, explaining what’s in each file. As mentioned above, we’re withholding most of the data relied upon in Part II and Part III, as it can be used to identify individuals and gain access to personal information. We are, however, making available the ratings of various items and political self-identification in the file Salemratings below. If you would like to test hypotheses about any other individualized data, contact us and we can possibly discuss some kind of collaboration or non-disclosure agreement.

All files listed below can be found here, in addition to a .json file with all the bets.

Salemcodebook

197KB ∙ PDF file

Download

Download

Salemratings

142KB ∙ XLSX file

Download

Download

Calibration

21.3KB ∙ XLSX file

Download

Download

Allobs

416KB ∙ XLSX file

Download

Download

Does higher education value policy and administrative reforms likely to save black (and ideally other) lives? Or does higher education care more about virtue signalling and promoting a broad leftist agenda to enhance the power and status of academics and their institutions, whatever the impacts on most other human beings?

The genesis of this essay can be found in a 2019 academic conference in which one of us (Maranto) presented a paper, later published in Public Integrity,1 which ranked police departments in 21 of the 25 largest US cities by their success in 2015 in keeping homicide rates low and not killing civilians, while adjusting for poverty rates, since high poverty makes policing more difficult. The New York City Police Department (NYPD) topped the rankings, with very low homicide rates and the lowest rates of police killings of civilians despite the city’s relatively high poverty. (Our 2020 analyses of 50 cities again ranked NYPD first.) As with other attempts at bureaucratic reform, ranking police departments based on their success on these important metrics has the potential to drive improvements by encouraging less successful cities to copy their more successful peers. Since both police killings of civilians and much more common civilian-on-civilian homicides disproportionately take black lives, progress on these indicators would indeed make black lives matter.

Almost any American without a PhD can see the value in this kind of research. But as recounted by one of us in the Wall Street Journal,2 other social scientists at the 2019 conference expressed confusion and at times astonishment as to why anyone interested in Black Lives Matter might study ways to improve policing. Instead, BLM researchers were interested in using the movement to increase progressive power. As of June 2020:

Using Google Scholar, the five most cited research articles and book chapters with “black lives matter” in the title include “a herstory of the #blacklivesmatter movement,” an article on “the evolving role of social media,” another on “mass struggle,” a commentary on racism and public health, and a piece on environmental justice. Other widely cited works study news coverage of the movement, youth activism, and the “migrant crisis” in Europe. Judging by their titles, none of the twenty most cited articles or chapters on Black Lives Matter directly seek ways to reduce police killings of black civilians.3

Similarly, as of May 5, 2022, there was just one Web of Science citation for Franklin Zimring’s 2017 When Police Kill,4 a serious empirical analysis suggesting ways to reduce civilian deaths at the hands of law enforcement. This (mis)allocation of research effort reflects the takeover of large sections of the social sciences by critical theory supporting activists seeking power (and grant funding), who are relatively uninterested in seeking to understand and ameliorate social problems like violent crime and police brutality.5

Here we repeat the June 2020 Wall Street Journal analysis in a somewhat different way, focusing not on titles but, instead, on generating rankings by conducting a general keyword search on the Web of Science Core Collection for “black lives matter” and sorting by publications with the most citations as of April 20, 2022.6 The Web of Science is more restrictive than Google Scholar, yielding far fewer citations, in our experience over 95% fewer.

The 25 most cited works listed in the above table have a combined total of 1,704 citations. As in the 2020 search using Google Scholar, the vast majority of the Web of Science citations seemingly have little to do with either studies or commentary likely to save black lives. At least from their titles, the top five works, those with 98 or more Web of Science citations, resemble, and often are, the most influential publications of 2020 using Google Scholar, involving public health, millennial activism, and the migrant crisis in Europe. In addition, the second most widely cited piece discusses how political science as a field treats marginalized communities of color.7 As its authors explain: “Focusing on policing, we seek to unsettle the mainstream of a subfield that rarely inquires into governmental practices of social control and the ways ‘race-class subjugated communities’ are governed through coercion, containment, repression, surveillance, regulation, predation, discipline, and violence.” Such analyses seem unlikely to decrease either homicide or police violence against civilians. They are unlikely to help make black lives matter on this planet in any foreseeable time frame.

Generally, the highly cited articles listed in the table explore social media use and activism (4, including one piece involving Ben and Jerry’s ice cream and BLM), racial activism and white attitudes (3), immigration and migrants (2), anti-Blackness in higher education, “democratic repair,” radically re-imagining law, anti-Blackness of global capital, urban geography, counseling psychology, research on K-12 schools, BLM and “technoscientific expertise amid solar transitions,” BLM and “evidence based outrage in obstetrics and gynecology,” and BLM and differential mortality in the electorate. Again, we see little evidence of author (or citer) intent to save black lives in the works that frequently get cited.

Of the 25 most cited scholarly articles, only one, ranked 19th, from Public Administration Review, presents research that has the potential to save black and other lives in ways that are relatively immediate and related to law enforcement. It is titled “Preventing the Use of Deadly Force: The Relationship between Police Agency Policies and Rates of Officer-Involved Gun Deaths.”8 This fascinating article uses cross sectional data from 1,107 law enforcement agencies over 15 years, standardized by population. The authors, Jennings and Rubado, find no evidence that either community policing or having police officers who demographically resemble their cities reduces the number of fatal police shootings of civilians. In contrast, roughly half of departments require that officers file a report whenever they point their guns without shooting. (Virtually all require such reports when weapons are discharged.) Empirical evidence indicates that this simple reporting requirement substantially reduces police shootings of civilians, while managing not to increase shooting deaths of officers.

Most proper science is tentative, and sometimes reforms have unintended consequences. That said, the paper empirically evaluates practical ways to save lives, prescribing a reform with the potential to do so. This article accounts for just 44 (2.6%) of the 1,704 Web of Science citations of the 25 most cited articles found in the “black lives matter” keyword search. The Bearfield, Maranto, and Wolf9 ranking of large police departments to identify those most successful at preserving life was cited just twice by Google Scholar as of May 2022 and not at all in the Web of Science. We doubt this article will ever rank among the top 1,000 cited BLM-related publications.

One force that sets a positive example other cities could follow is the aforementioned NYPD. Each year, its internal affairs division pushes dozens of officers to resign, retire early, or move to routine administrative duties like vehicle impoundment in which they are unarmed. Others are prosecuted, as longtime NYPD Internal Affairs Bureau Chief Charles Campisi documented in his lively Blue on Blue: An Insider’s Story of Good Cops Catching Bad Cops.10 An NYPD source indicates that about 300 officers, nearly 1% of the force, are on internal watch lists. When told that NYPD should feel satisfied having so few problematic officers, the source retorted that they had trouble sleeping at night since every one of those officers presented a “potential” George Floyd or Freddie Gray “situation.” In retrospect, Minneapolis and Baltimore might have been better off had more of their police administrators thought this way. Two years after the Floyd murder, observers believed that, along with civil service rules, the Minneapolis police union contract still posed undue burdens on police managers who wanted to hold their officers accountable.11

In short, much can be done to improve policing in ways that would make all lives, but especially black lives, matter. Alas, currently, we cannot much rely on academia to help. If anyone intends research to save black lives rather than merely to promote radical causes and the careers of professors, then we need a major reorientation of social science research, and arguably, of the whole higher education enterprise.