Search the archive.

The JGP proposal also augments inflation concerns that are already inherent in MMT’s money-financed approach to fiscal policy. The main problem is a JGP sets a real wage floor for the entire economy, thereby implicitly introducing inflation indexation. As the economy approaches full employment, inflation will tend to rise in accordance with Phillips curve logic. That will trigger a higher nominal JGP wage, and raising the nominal wage floor will tend to spread inflation throughout the economy. If private sector firms fail to match the JGP nominal wage increase, workers will start to drift toward JGP jobs, causing a contraction of private sector output that could even potentially cause stagflation. The JGP is intended to be a non-inflationary automatic stabilizer. Yet closer inspection shows its implicit indexing may also render it an automatic destabilizer.23

Furthermore, to the extent that a JGP delivers quasi-full employment, it will also tend to exacerbate income distribution conflict inflation which emerges at full employment (Kalecki, 1943). That is because it would remove or diminish the threat of labor pricing itself out of work, thereby potentially increasing wage demands. The problem of full employment conflict inflation is generic, but a JGP would exacerbate it. Indeed, a JGP could trigger such inflation even earlier by emboldening wage demands earlier in the business cycle. There is no easy solution to that problem, which explains Lerner’s (1977, 1978) interest in Wage Increase Permits (WIPs). However, MMT proponents are wrong to claim that a JGP is not afflicted by the problem of conflict inflation, and they are doubly wrong to be dismissive of Lerner’s later concerns that prompted his WIPs proposal.

The adoption of the MMT main policy proposals would result in a drastic change in the institutional framework used by governments to make policy decisions and how they define their budget constraint and their relation with the national central bank. In an MMT-dominated world, a central bank would cease to bear any resemblance to a financial institution. Instead, it would become an outpost of the treasury, its behaviour determined in particular by government decisions to spend more or less, thus determining whether to create more or less money. In such a scenario, the central bank would not retain any meaningful degree of independence regarding the decisions affecting the amount of money in the economy or the interest rate charged in its regular lending to commercial banks. The central bank’s only role would be to accommodate fully the financial needs of the treasury, determined by the government’s own economic policies and political goals. This is what economists traditionally refer to as a ‘fiscal dominance’ scenario (see Sargent and Wallace 1981). In the aftermath of the Global Financial Crisis, when reflecting on his experience with the US economy and high inflation back in the 1970s, the former President of the US Fed (1979- 1987), Paul Volker (2011) pointed out how difficult it was to return to more sound and sustainable monetary policies once fiscal dominance becomes entrenched. Adopting a fully-fledged fiscal dominance scenario would lead to a significant policy regime change, with enormous implications for longterm economic outcomes. As pointed out previously in this paper, if the financial constraint on the government were relaxed for too long, let alone eliminated as MMT would entail, there would be virtually no limit on money creation and rising inflation would rapidly become entrenched in the economy.

Currently, central banks are far from being in an ideal ‘monetary dominance’ scenario - in other words, one where an independent central bank determines its own targets following a clear policy rule and is not affected by government policies and decisions. From the outbreak of the Global Financial Crisis, the purchase by central banks of a significant amount of public debt has put them in a difficult position.17 This trend has accelerated since March 2020, when central banks all over the world have drastically increased their public debt holdings and at a faster pace. Effectively, this is making governments more reliant on their national central banks’ direct or indirect purchases of their debt to finance their enlarged spending programmes at affordable (indeed virtually zero) rates. To be clear, the central bank purchase of public debt under QE programmes is not per se tantamount to central bank dependence on the government. As recently explained by the Deputy Governor of the Bank of England, in times of severe crisis it is expected that both the fiscal and monetary policy makers use their respective policies to support spending (Broadbent 2020), one of their tools being asset purchase programmes or the reduction of policy rates. However, central bank independence will be tested in the next 1-2 years, when a likely increase in spending and in CPI inflation will put central banks in an extremely difficult policy position (Goodhart and Pradhan 2020). They will have to choose between pursuing their own targets, which will almost certainly mean tightening monetary policy to avoid future inflation (i.e. monetary dominance), or else keeping very low interest rates and loose monetary conditions for longer to support the government (i.e. fiscal dominance).18 At that point, central banks will need to choose between either surrendering monetary stability and low inflation altogether and effectively moving towards the adoption of an MMT paradigm, or taking the necessary measures to rein in inflation and take back control of the amount of money in circulation. If the former, both historical evidence and sound monetary theory point to an economy entrenched in enlarged government spending, much higher rates of growth in the amount of money, and therefore higher inflation.

The key MMT macroeconomic policy proposal is that government spending should target full employment, and that the central bank should keep interest rates low by monetizing the debt.7 Moreover, the government should not worry about “financing the budget” through taxes or cuts in other spending. This is functional finance with an MMT twist. MMT advocates argue that at full employment, the fiscal authorities should raise taxes or cut spending to drain demand from the economy to maintain full employment without inflation (Wray 2012; Palley 2013, 2015; Tymoigne and Wray 2015). In this rendering, then, macroeconomic policy should be guided by fiscal spending to achieve full employment, and the policy should be subject to an inflation constraint managed by fiscal policy tools. Monetary policy, in turn, should focus on keeping the interest rate low. Palley has addressed the problems of using fiscal policy to fine-tune the economy to reach full employment without excessive inflation (Palley 2013, 2015). These are familiar from the old debates about fine-tuning vs. rules and emphasize uncertain lags, uncertain shocks to the economy, and poor understanding of key parameters such as multipliers associated with different policies and the shape and position of the Philips Curve. In addition, there are many related problems associated with this idea of fine-tuning fiscal policy at the point of full employment: how exactly do we define full employment? What about underemployment? What about variations in labor force participation rates over the cycle? There are additional problems. What happens when the economy crosses the threshold between the unemployment/low taxes world and to the full employment/higher taxes world: Let’s say we do know where to set the full employment threshold. We are now introducing a massive shock into the economy when we cross into full employment. That is, just before full employment taxes are very low. After we cross the threshold, all of a sudden, the government is collecting a significant amount of taxes. How do we avoid a significant destabilization?

Resolution of Supply Constraints Plus Fed Tightening Will Crush Inflation

We expect inflation to fall to normal levels after peaking at 6.2% in 2022. We still think most of the sources of high inflation since the start of the pandemic will abate (and even unwind in impact) over the next few years. This includes energy, autos, and other durables. Aggressive capacity expansion across many areas could turn widespread shortages into gluts within a few years. Combining these factors with monetary policy tightening, we expect inflation to drop to 3.6% for full-year 2023 and average just 1.8% over 2024-27.

Inflation in 2022 reached the highest level in over 40 years. But we expect inflation to fall dramatically over the next year.

Understanding why inflation will fall requires understanding why inflation became so bad in the first place. Over 2021 and 2022, strong consumer demand ran into a brick wall of supply constraints, causing prices to explode in many industries, especially food, energy, and durable goods.

A combination of higher interest rates, easing supply chain constraints and tighter credit has helped tame inflation in 2023. As of May, the annualized rate of CPI inflation has declined for 11 months straight.

In May, core inflation—which excludes volatile food and energy prices—was up 0.4% on a monthly basis and 5.3% from a year ago.

This proposal envisions a generous, open- ended federal match to states for the formula grant. Each states’ grant would be based on states’ FMAP (Federal Medical Assistance Per- centage, or Medicaid match rate). The FMAP for each state is based on the relative per capita income of a state compared with national per capita income. It currently varies between 50 and 82 percent, limited by a statutory maxi- mum of 83 percent (ASPE 2015). Under this sub- sidized employment proposal, each state would receive a minimum (regardless of macroeco- nomic conditions) match equal to its FMAP plus half the gap between the state’s FMAP and 100 percent, resulting in minimum matching rates ranging from 75 percent in the wealthiest states to 91 percent in the poorest. States would be able to count TANF,8 Workforce Innovation and Opportunity Act (WIOA), and other federal funding legitimately utilized for subsidized em- ployment toward their contributions. Based on the TANF Emergency Fund’s subsidized jobs experience and the adoption of the Medicaid expansion under the Affordable Care Act, a re- imbursement rate that is more generous than the existing FMAP for basic Medicaid is likely necessary to ensure that all or nearly all states choose to participate in this program (Kaplan 2009; Ollove 2015). This proposed federal sub- sidized employment match (FSEM) also would rise (never above 100 percent) and fall (never below 75 percent), with each year-over-year per- centage point increase in three-month average state unemployment rates triggering a percent- age increase in the match rate. Unlike the FMAP, the FSEM would be more responsive to economic changes by being updated every cal- endar quarter, though increased match rates would trigger on for a minimum of the balance of the current and subsequent fiscal year to allow for state fiscal, policy, and programmatic planning. (Overly volatile match rates, espe- cially falls in the match rate, would jeopardize the viability of the program’s match structure.)

The competitive grant, administered by HHS, would be administered similarly to exist- ing competitive grant programs.

During periods of high unemployment, subsidized employment programs can boost the economy by increasing wages and employment levels — and should be used to do so now. (See text box, “Subsidized Employment Already Being Used to Fight Pandemic.”)

But subsidized employment programs have a role to play even in expansionary periods, when they can improve the employment opportunities and outcomes for individuals with more significant employment barriers who may need more support and supervision.[28] These barriers, such as low education levels or caregiving responsibilities, make it challenging for many people to find jobs even when the economy is strong.[29] A permanent federal program could create the foundation to support these workers — then scale up rapidly during recessions. Subsidized employment should be a part of a permanent investment in a stronger, more equitable economy.

The total direct compensation costs of serving 20 percent of the full universe of eligible par- ticipants would be $15.9 billion annually to pro- vide subsidized employment for 2.4 million participants—which would benefit more than seven million members of those workers’ households.9 In addition to these direct com- pensation costs for wages would be added costs for payroll taxes and net tax benefits, wrap- around services, and state and local adminis- trative costs, which would further add to the annual costs.

The total additional earnings received would reduce the nation’s overall poverty rate (SPM) from 14.3 percent to 14.0 and thereby lift approximately one million people out of SPM poverty. The reduction in the poverty rate of expected program participants—who are significantly more disadvantaged than the overall population, having a poverty rate nearly 2.5 times the overall rate—would be far greater. For participants, the reductions in the poverty rate and the deep poverty

Nearly 15 million people in the U.S. who would like to work are unable to find a job right now—despite a historically low national unemployment rate.1 This joblessness is not uniformly distributed across the country but concentrated in the most disadvantaged communities, particularly Black and Brown communities. A targeted, large-scale subsidized employment program can help these communities increase employment and reduce poverty.

Last year the national unemployment rate was 3.6 percent.2 The last time the national rate was this low was in 1969.3 But the national average hides the fact that there are many communities that are suffering from persistent, high levels of joblessness. In every state, one can find communities that have been similarly left behind due to deindustrialization, disinvestment, and discrimination related to the legacy of redlining and other policies. These communities tend to have a high share of workers with low incomes or workers of color. For example, last year the unemployment rate in the majority-Latinx town Yuma, Arizona, was 10 percent, nearly three times the national rate. It was 11 percent in both Flint, Michigan, and Magoffin County, Kentucky, predominantly Black and white locales, respectively. It was higher still in Jefferson County, Mississippi, where the unemployment rate is 14 percent and the large majority of residents are Black. It was 16 percent in the Alaska Native-majority Kusilvak area of Alaska.4

Subsidized employment programs, meanwhile, have improved workers’ wages and employment for decades, and can provide economic benefits for participants that last for years. In the recovery from the Great Recession, state subsidized employment programs resourced through the Temporary Assistance for Needy Families (TANF) Emergency Fund placed more than a quarter-million people in public- and private-sector jobs, giving them needed income and work experience. Experience like this, in turn, can connect participants to unsubsidized employment, improving their long-term job prospects.

Subsidized employment programs have improved workers’ wages and employment for decades and can provide economic benefits for participants that last for years.

Bottom Line: Subsidized employment programs provide jobs to people who cannot find employment in the regular labor market and use public funds to pay all or some of their wages. Some programs are designed primarily to provide short-term income support in poor economic times, while others also seek to improve long-term employment outcomes among “hard-to-employ” groups. Evaluations of subsidized employment programs suggest that they are effective at providing jobs in the short term but are less successful at helping participants transition to unsubsidized employment. However, these programs can also have valuable ancillary benefits, including the value of the work produced, reduced welfare receipt, and decreased recidivism among ex-offenders. To address lingering high unemployment rates for the long-term unemployed in the aftermath of the Great Recession, subsidized employment programs may be part of the answer, much as they were in past periods of deep unemployment.

What Do We Know?

It is fair to say that MMT authors tend not to emphasise the potential detrimental effects of high interest payments. As a result of political choices, real resources could de facto be diverted from low-income taxpayers, on whom the general taxation system relies disproportionally in developed countries, to high-income domestic or foreign savers. Of course, this also depends ultimately on who are the holders of government bonds. Yet regressive redistributive effects cannot be ruled out, when these holders are predominantly the representatives of the most affluent portions of the population. Even if one holds the MMT assumption that government expenditure is not funded by taxes, taxes are still used to drain real resources from the private sector: when a citizen pays taxes, the individual relinquishes purchasing power and gives it to the government. Interest payments on government debt correspond to a transfer of this purchasing power to individuals. Again, this could come with bottom-up redistributive effects. It makes a lot of difference whether public money is used to provide public goods (for example, to build schools or hospitals) or to pay interest on government bonds to a specific portion of the domestic or foreign population.

MMT does not currently advocate, nor demand taxation for redistribution of wealth for the purpose of addressing inequality. However, Wray would like to tax the wealthy until they are no longer rich, though this is not a formal part of MMT (Wray 2020). 3 The redistribution of wealth through taxation is a fundamental feature of the neoclassical theory (Samuelson 1956; Mas-Colell et al. 1995; Keen 2011), but such wealth redistribution is absent from MMT (Wray 2014a). Redistribution taxation is no longer practiced as much as it once was, particularly in the USA (FRED 2020d), and despite its foundational position in neoclassical theory, it is rarely mentioned in public discourse today. However, I think that adding redistribution taxation to the MMT system is probably a necessity to stabilize the economy (Weller 2010), and to accomplish political stability (Acemoglu – Robinson 2012). The former reason is that taxation for redistribution lowers Gini, and thus, stabilizes demand growth. While Weller and Rao’s argument is relative to the emerging economies, the same issues appear in the developed ones. Relative to politics, Acemoglu and Robinson’s work shows that the more unequal economic hierarchies produce political instability together with worse economic performance. Barring this kind of taxation, other careful work must be done to prevent rentier gaming of the economy, although Mosler’s concept of eliminating bond sales has a great deal of merit.

Providing assistance to the poor through in-kind transfers, such as vouchers for food and housing, garners more political support than providing assistance in cash. Supporters of such policies believe that providing voucher payments for certain goods (such as groceries) will cause recipients to purchase more of the goods being subsidized, and that recipients will not be able to use public assistance to buy other, less socially desirable goods (such as alcohol or cigarettes). According to canonical economic theory, providing a transfer in-kind should lead to the same outcome as a similar-sized cash transfer as long as program participants are inframarginal. As a result, depending on consumer preferences, the provision of in-kind transfers (relative to cash) may have little to no impact on purchases of the actual goods being subsidized.

Despite strong theoretical predictions about consumer behavior, little empirical evidence has been brought to bear on the impact of providing in-kind transfers on consumer purchases.1 The Food Stamp Program (FSP) is one of the largest transfer programs for the low-income population. In 2004, for example, the program cost $27 billion and served 24 million people. This compares to $25 billion for Temporary Assistance for Needy Families and $33 billion for the Earned Income Tax Credit. It has been very difficult for researchers to isolate the causal impact of the FSP on food spending, nutritional intake, labor supply, and other outcomes. Because the program is national, there is no variation in program parameters (such as stark differences in state benefit levels or eligibility) that are typically exploited by researchers to measure program impacts. In the absence of programmatic variation, most researchers have studied the impact of the FSP by comparing food stamp recipients with eligible nonrecipients. Since we would expect participants and nonparticipants to differ in important, and potentially unobservable, ways (Janet Currie 2006), researchers have employed a variety of methods to control for selection into the program (see Thomas M. Fraker 1990 for a comprehensive review of the literature on food stamps and consumption).

We view a UBI to be a sub-optimal, and possibly harmful, policy response to all three of these challenges. A UBI in its most basic form would be massively expensive yet do little to reduce inequality or advance opportunity. Devoting that level of spending to targeted benefits, focusing on the poorest and those hardest hit by ongoing economic forces, and polices dedicated to human capital development instead of mere redistribution would produce a much greater social return than a UBI.

First, on inequality and redistribution, a UBI is by design not ideal for redistributive purposes. For example, a UBI that paid $10,000 to every U.S. adult would cost about $2.5 trillion per year, well more than half the current federal annual budget. Furthermore, by giving money to everyone, there would be far fewer resources available to redistribute money and/or invest in the human capital of those with the most need. In calculations presented below, we show the practical trade-off between giving more money to a more targeted group of low-income individuals and less money to a more diffuse group including less needy individuals. We cite work by Hoynes and Rothstein (2019) documenting the loss of progressivity that would come from replacing our current system of transfer programs with a UBI. We further argue that programs that are universal do not accomplish as much in terms of generating social benefits as those that are targeted, citing evidence from examples such as child care and early childhood education programs.

However, other assessments of these existing schemes have concluded that existing federal regulatory schemes are inadequate to assess and respond to the prospective risks posed by nanotechnology.278 Significant regulatory coverage gaps (and in some cases, gulfs) would manifest upon any effort to use existing federal regulatory authority, and only that authority, to regulate nanotechnology.279 These gaps can take multiple forms. For example, some existing environmental statutes (such as the Toxic Substances Control Act (TSCA) and the Resource Conservation and Recovery Act (RCRA)) are triggered based on volume or mass.280 Such measures are inappropriate for nanotechnology regulation.281 Alternatively, the fundamental incapacity to detect and measure nanoscale quantities will make regulation difficult.282 In practice, the experience of state and local regulators can ultimately help the EPA decide which regulatory authority works best.

The discussion of potentially applicable federal statutes should make clear that even if nanotechnology is regulated at a centralized level, coordination difficulties will likely arise because nanotechnology arguably falls within the purview of a variety of different statutes and different agencies.283 Therefore, centralized regulation cannot be justified on the basis of improved coordination.

Historically, the risks that have arisen from emerging technologies have been small when compared with their benefits. The potential exceptions are unprecedented risks that could threaten large parts of the globe, or even our very survival1. Technology has significantly improved lives in the United Kingdom and the rest of the world. Over the past 150 years, we have become much more prosperous. During this time, the UK average income rose by more than a factor of seven in real terms, much of this driven by improving technology. This increased prosperity has taken millions of people out of absolute poverty and has given everyone many more freedoms in their lives. The past 150 years also saw historically unprecedented improvements in health, with life expectancy in the United Kingdom steadily increasing by two to three years each decade. From a starting point of about 40 years, it has doubled to 80 years2. These improvements are not entirely due to technological advances, of course, but a large fraction of them are. We have seen the cost of goods fall dramatically due to mass production, domestic time freed up via labour saving machines at home, and people connected by automobiles, railroads, airplanes, telephones, television, and the Internet. Health has improved through widespread improvements in sanitation, vaccines, antibiotics, blood transfusions, pharmaceuticals, and surgical techniques. These benefits significantly outweigh many kinds of risks that emerging technologies bring, such as those that could threaten workers in industry, local communities, consumers, or the environment. After all, the dramatic improvements in prosperity and health already include all the economic and health costs of accidents and inadvertent consequences during technological development and deployment, and the balance is still overwhelmingly positive. This is not to say that governance does or should ignore mundane risks from new technologies in the future. Good governance may have substantially decreased the risks that we faced over the previous two centuries, and if through careful policy choices we can reduce future risks without much negative impact on these emerging technologies, then we certainly should do so. However, we may not yet have seen the effects of the most important risks from technological innovation. Over the next few decades, certain technological advances may pose significant and unprecedented global risks. Advances in the biosciences and biotechnology may make it possible to create bioweapons more dangerous than any disease humanity has faced so far; geoengineering technologies could give individual countries the ability to unilaterally alter the global climate (see case study); rapid advances in artificial intelligence could give a single country a decisive strategic advantage. These scenarios are extreme, but they are recognized as potential low-probability high-impact events by relevant experts. To safely navigate these risks, and harness the potentially great benefits of these new technologies, we must continue to develop our understanding of them and ensure that the institutions responsible for monitoring them and developing policy responses are fit for purpose. This chapter explores the high-consequence risks that we can already anticipate; explains market and political challenges to adequately managing these risks; and discusses what we can do today to ensure that we achieve the potential of these technologies while keeping catastrophic threats to an acceptably low level. We need to be on our guard to ensure we are equipped to deal with these risks, have the regulatory vocabulary to manage them appropriately, and continue to develop the adaptive institutions necessary for mounting reasonable responses. Anthropogenic existential risks vs. natural existential risks An existential risk is defined as a risk that threatens the premature extinction of humanity, or the permanent and drastic destruction of its potential for desirable future development. These risks could originate in nature (as in a large asteroid impact, gamma-ray burst, supernova, supervolcano eruption, or pandemic) or through human action (as in a nuclear war, or in other cases we discuss below). This chapter focuses on anthropogenic existential risks because — as we will now argue — the probability of these risks appears significantly greater. Historical evidence shows that species like ours are not destroyed by natural catastrophes very often. Humans have existed for 200,000 years. Our closest ancestor, Homo erectus, survived for about 1.8 million years. The median mammalian species lasts for about 2.2 million years3. Assuming that the distribution of natural existential catastrophes has not changed, we would have been unlikely to survive as long as we have if the chance of natural extinction in a given century were greater than 1 in 500 or 1 in 5,000 (since (1 – 1/500)2,000 and (1 – 1/5,000)18,000 are both less than 2%). Consistent with this general argument, all natural existential risks are believed to have very small probabilities of destroying humanity in the coming century4. Case study omitted. In contrast, the tentative historical evidence we do have points in the opposite direction for anthropogenic risks. The development of nuclear fission, and the atomic bomb, was the first time in history that a technology created the possibility of destroying most or all of the world’s population. Fortunately we have not yet seen a global nuclear catastrophe, but we have come extremely close. US President John F. Kennedy later confessed that during the Cuban missile crisis, the chances of a nuclear war with Russia seemed to him at the time to be “somewhere between one out of three and even”. In light of this evidence, it is intuitively rather unclear that we could survive 500 or 5,000 centuries without facing a technologically-driven global catastrophe such as a nuclear war. We argue that in the coming decades, the world can expect to see several powerful new technologies that — by accident or design — may pose equal or greater risks for humanity. 1. Engineered Pathogens Pandemics such as Spanish flu and HIV have killed tens of millions of people. Smallpox alone was responsible for more than 300 million deaths in the first half of the twentieth century. As the ongoing Ebola epidemic reminds us, disease outbreaks remain a potent threat today. However, pressures from natural selection limit the destructive potential of pathogens because a sufficiently virulent, transmissible pathogen would eliminate the host population. As others have argued, and we reiterate below, bioengineering could be used to overcome natural limits on virulence and transmissibility, allowing pandemics of unprecedented scale and severity. For an example of an increase in fatality rates, consider mousepox, a disease that is normally non-lethal in mice. In 2001, Australian researchers modified mousepox, accidentally increasing its fatality rate to 60%, even in mice with immunity to the original version5 . By 2003, researchers led by Mark Buller found a way to increase the fatality rate to 100%, although the team also found therapies that could protect mice from the engineered version6 . For an example of an increase in transmissibility, consider the ‘gain of function’ experiments on influenza that have enabled airborne transmission of modified strains of H5N1 between ferrets7 . Proponents of such experiments argue that further efforts building on their research “have contributed to our understanding of host adaptation by influenza viruses, the development of vaccines and therapeutics, and improved [disease] surveillance”8 . However, opponents argue that enhancing the transmissibility of H5N1 does little to aid in vaccine development; that long lag times between capturing and sequencing natural flu samples limits the value of this work for surveillance; and that epistasis — in which interactions between genes modulate their overall effects — limits our ability to infer the likely consequences of other genetic mutations in influenza from what we have observed in gainof-function research so far9 . Many concerns have been expressed about the catastrophic and existential risks associated with engineered pathogens. For example, George Church, a pioneer in the field of synthetic biology, has said: “While the likelihood of misuse of oligos to gain access to nearly extinct human viruses (e.g. polio) or novel pathogens (like IL4-poxvirus) is small, the consequences loom larger than chemical and nuclear weapons, since biohazards are inexpensive, can spread rapidly world-wide and evolve on their own.”10 Similarly, Richard Posner11, Nathan Myhrvold12, and Martin Rees13 have argued that in the future, an engineered pathogen with the appropriate combination of virulence, transmissibility and delay of onset in symptoms would pose an existential threat to humanity. Unfortunately, developments in this field will be much more challenging to control than nuclear weapons because the knowledge and equipment needed to engineer viruses is modest in comparison with what is required to create a nuclear weapon14. It is possible that once the field has matured over the next few decades, a single undetected terrorist group would be able to develop and deploy engineered pathogens. By the time the field is mature and its knowledge and tools are distributed across the world, it may be very challenging to defend against such a risk. This argues for the continuing development of active policy-oriented research, an intelligence service to ensure that we know what misuse some technologies are being put to, and a mature and adaptive regulatory structure in order to ensure that civilian use of materials can be appropriately developed to maximize benefit and minimize risk. We raise these potential risks to highlight some worst-case scenarios that deserve further consideration. Advances in these fields are likely to have significant positive consequences in medicine, energy, and agriculture. They may even play an important role in reducing the risk of pandemics, which currently pose a greater threat than the risks described here. 2. Artificial Intelligence Artificial intelligence (AI) is the science and engineering of intelligent machines. Narrow AI systems — such as Deep Blue, stock trading algorithms, or IBM’s Watson — work only in specific domains. In contrast, some researchers are working on AI with general capabilities, which aim to think and plan across all the domains that humans can. This general sort of AI only exists in very primitive forms today15. Many people have argued that long-term developments in artificial intelligence could have catastrophic consequences for humanity in the coming century16, while others are more skeptical17. AI researchers have differing views about when AI systems with advanced general capabilities might be developed, whether such development poses significant risks, and how seriously radical scenarios should be taken. As we’ll see, there are even differing views about how to characterize the distribution of opinion in the field. In 2012, Müller and Bostrom surveyed the 100 most-cited AI researchers to ask them when advanced AI systems might be developed, and what the likely consequences would be. The survey defined a “high-level machine intelligence” (HLMI) as a machine “that can carry out most human professions at least as well as a typical human”, and asked the researchers about which year they would assign a 10%, 50% or 90% subjective probability to such AI being developed. They also asked whether the overall consequences for humanity would be “extremely good”, “on balance good”, “more or less neutral”, “on balance bad”, or “extremely bad (existential catastrophe)”. The researchers received 29 responses: the median respondent assigned a 10% chance of HLMI by 2024, a 50% chance of HLMI by 2050, and a 90% chance of HLMI by 2070. For the impact on humanity, the median respondent assigned 20% to “extremely good”, 40% to “on balance good”, 19% to “more or less neutral”, 13% to “on balance bad”, and 8% to “extremely bad (existential catastrophe)”18. In our view, it would be a mistake to take these researchers’ probability estimates at face value, for several reasons. First, the AI researchers’ true expertise is in developing AI systems, not forecasting the consequences for society from radical developments in the field. Second, predictions about the future of AI have a mixed historical track record19. Third, these ‘subjective probabilities’ represent individuals’ personal degrees of confidence, and cannot be taken to be any kind of precise estimate of an objective chance. Fourth, only 29 out of 100 researchers responded to the survey, which therefore may not be representative of the field as a whole. The difficulty in assessing risks from AI is brought out further by a report from the Association for the Advancement of Artificial Intelligence (AAAI), which came to a different conclusion. In February 2009, about 20 leading researchers in AI met to discuss the social impacts of advances in their field. One of three sub-groups focused on potentially radical long-term implications of progress in artificial intelligence. They discussed the possibility of rapid increases in the capabilities of intelligent systems, as well as the possibility of humans losing control of machine intelligences that they had created. The overall perspective and recommendations were summarized as follows: • “The first focus group explored concerns expressed by lay people — and as popularized in science fiction for decades — about the long-term outcomes of AI research. Panelists reviewed and assessed popular expectations and concerns. The focus group noted a tendency for the general public, science-fiction writers, and futurists to dwell on radical long-term outcomes of AI research, while overlooking the broad spectrum of opportunities and challenges with developing and fielding applications that leverage different aspects of machine intelligence.” • “There was overall skepticism about the prospect of an intelligence explosion as well as of a “coming singularity,” and also about the large-scale loss of control of intelligent systems. Nevertheless, there was a shared sense that additional research would be valuable on methods for understanding and verifying the range of behaviors of complex computational systems to minimize unexpected outcomes.” • “The group suggested outreach and communication to people and organizations about the low likelihood of the radical outcomes, sharing the rationale for the overall comfort of scientists in this realm, and for the need to educate people outside the AI research community about the promise of AI for enhancing the quality of human life in numerous ways, coupled with a re-focusing of attention on actionable, shorter-term challenges.”20 This panel gathered prominent people in the field to discuss the social implications of advances in AI in response to concerns from the public and other researchers. They reported on their views about the concerns, recommended plausible avenues for deeper investigation, and highlighted the possible upsides of progress in addition to discussing the downsides. These were valuable contributions. However, the event had shortcomings as well. First, there is reason to doubt that the AAAI panel succeeded in accurately reporting the field’s level of concern about future developments in AI. Recent commentary on these issues from AI researchers has struck a different tone. For instance, the survey discussed above seems to indicate more widespread concern. Moreover, Stuart Russell — a leader in the field and author of the most-used textbook in AI — has begun publicly discussing AI as a potential existential risk21.

Amid the continuing revelations about what led to the January 6, 2021, attack on the U.S. Capitol, one aspect of the crisis has received comparatively little attention: how the effort to negate the presidential election outcome was built on the malign use of the United States’ federal system of government. Since the slates of electors that collectively certify the presidential election are chosen at the state level, the January 6 conspirators sought to appoint alternative slates of electors in several states to overturn the results. In the end, Republican state officials in Arizona, Georgia, and other states refused to undermine democracy on behalf of their partisans. But the conspiracy underscored the far-reaching importance of the states in some of the most fundamental decisions of the U.S. government, as well as how much it matters who controls those governments and what interests they serve.

Although it was an extreme case, the January 6 crisis was not the only situation in recent years in which the states have played a crucial role in setting the direction of the country as a whole. In areas as varied as access to firearms, emergency health care, immigration enforcement, cryptocurrency regulation, and the climate crisis, states have been asserting their powers to influence, and in some cases to challenge, U.S. policy. State leaders aggressively litigate to block federal policy and are active in responding to federal developments that contrast with the preferences of state electoral majorities. And some of the largest states—California, Florida, New York, and Texas collectively account for about 37 percent of U.S. GDP—are becoming more involved in foreign affairs, not only on economic and social issues but also through the soft diplomacy of values and culture. In the summer of 2022, even as the Biden administration was reeling from West Virginia’s successful litigation before the U.S. Supreme Court to limit federal regulations on greenhouse gas emissions, the president asked California Governor Gavin Newsom to join a ministerial meeting on climate with Chinese Minister of Ecology and Environment Huang Runqiu.

The best option, it seems, is to maintain existing allocations of authority and attempt incremental reform. The federal-state partnership is not as cooperative as it used to be. It’s actually sometimes disruptive — but finding a more even-keeled balance is critical for the future of environmental protection. The success of our system depends on a working relationship between the federal government and state environmental agencies.

When not functioning, a problem arises for the quality of environmental policy and big, looming issues like climate change and water pollution. What there is a need for within environmental policy is partnerships. Yes, the federal government is important — and so are states. There has to be some collaboration. That’s how our government was designed to function and how it works best, if it’s done right. There needs to be some flexibility, some innovation and a lot of power-sharing.