Epistemic note: This is a long post that applies two models of work to the government context. I've found those models to be useful in that context and elsewhere, but the arguments I present are very much debatable, and I'm very much open to discussing them.
I’m going to tell you a story. It is not a true story. Please ignore that fact; you can argue with me later if it seems relevant.
In The Games I Play, I mentioned that I used to play Cities: Skylines, a game whose objective is to build as large of a city as possible. It’s a difficult game to play.
Your city begins with a single highway exit. You connect a road to it, a simple two-lane road because that’s all you have the money to build. You zone that road for houses. People move into those houses, but those people need work, so you build commercial and industrial regions nearby. Commercial zoning might be right along your main street, but you can’t put your industry right next to people’s homes because they’ll complain about noise pollution if you do.
This is your first rule.
Your city develops. You add a water pump, then a sewage pipe, which has to be far downstream of the pump because your population will get sick otherwise. You unlock elementary schools and build one for every hundred kids in the city. Your main intersection gets blocked by traffic because it’s used for industrial trucks and residents’ cars, so you give each new industrial area a highway exit. You destroy a couple homes to add parks in one neighborhood because its residents are unhappy. You add fire stations every ten blocks. There’s traffic again between the residential and commercial parts of the city, so you add metro lines in a clockwise and counterclockwise loop around the city. You try adding high-density housing, but it gets too loud and happiness levels decline.
Your residents still need jobs. Not the commercial jobs—those need higher education levels and you don’t have a university in the city yet. But you can’t build an industrial area. One area might work, but it’s not near any major highways and trucks couldn’t get out of the city. The city’s only other undeveloped area would end up polluting a residential zone. You can’t condense that residential zone because any more high-density housing will make homeowners unhappy. Besides, doing that would cut off your existing metro stations, and you’re not sure you have enough money to move them.
As a result, people start leaving the city. You try to bring them back—you add recycling service, build more parks, allow them to smoke weed—but nothing seems to work. Abandoned homes line your blocks; new residents move in every so often, only to demand industrial jobs that don’t exist. Tax revenue declines, and your population plummets. Frustrated, you turn off the game and go back to your math homework.
If only you could build an industrial zone.
How Stuff Gets Done
I recently read One Day Sooner and Never Drop A Ball, a pair of LessWrong posts that present a dichotomy in how organizations function. They’re worth reading in full if you have time. If you don’t have time, I’ll briefly summarize each one.
One Day Sooner is a mode of working that aggressively pursues a single goal. The canonical example is vaccine trials, where organizations like 1DaySooner work to get vaccines out as quickly as possible. The idea is that some goals, like getting vaccines to the public or eliminating a bottleneck in a project, are so important that they should be pursued immediately. The post describes how to get to things One Day Sooner, typically involving pushing through unnecessary slowdowns like “let’s schedule this meeting three days from now” or “I need a break from work tonight.”
Never Drop A Ball is a mode of working that pursues several goals or projects and ensures none of them are lost in the shuffle. Where a startup founder aiming for One Day Sooner might spend a sixty-hour week working on their core project, a founder aiming to Never Drop A Ball will carefully keep track of every bit of work, ensuring things like taxes and floor cleaning are properly handled. Notably, the failure mode of Never Drop A Ball work is that balls need to be dealt with; if they take too much work to resolve or there are too many to handle, balls will begin to drop.
The difference between these modes is how many balls they handle. Aiming for One Day Sooner means juggling one or two balls, typically ones that are very difficult to resolve, and trying very hard to resolve them. Aiming to Never Drop A Ball means juggling many balls—and, importantly, juggling new balls that come up—which tend to be easy to resolve and thus easy to miss. The number of balls and the likelihood of being assigned new balls are sliding scales, but it’s often easier to think of them as a binary. As anyone who has worked on a project can tell you, dealing with a small number of balls is associated with a low likelihood of being assigned new balls because new balls tend to be distracting.
In the rest of this essay, aiming for One Day Sooner will be referred to as “mission mode,” while aiming to Never Drop A Ball will be referred to as “fallback mode.”
How Stuff Gets Done in Government
Governments are intended to do things. These things can be thought of as balls.
There are a lot of balls floating around in government.
Governments as a whole act in fallback mode. This is because governments do many, many things and can never make a mistake. A single misapplication of power—say, a police officer making an unusually dangerous arrest and accidentally killing George Floyd—could result in nationwide protests and shake the foundation of the government’s power. In the words of Thomas Jefferson:
Governments are instituted among Men, deriving their just powers from the consent of the governed, [and] whenever any Form of Government becomes destructive of these ends, it is the Right of the People to alter or to abolish it, and to institute new Government…
Because of this fragility, governments create all sorts of rules to prevent making mistakes. This is the origin of the endless bureaucracy of the U.S. government. It is extremely frustrating. It is also written in blood. Every rule, regulation, and system is created for a reason; changing it might make something more efficient, but it might cause something else to fail, which is the absolute worst thing that can happen to a government.
Unfortunately, this falls squarely into the failure mode of never dropping a ball. The government becomes consumed with keeping things maintained and functional rather than making progress on new ideas.
Imagine you’re an official working on energy policy in West Virginia. A declining population means fewer people to fill jobs in power plants. Ancient coal plants are failing left and right as infrastructure becomes dilapidated. You’re trying to get new gas plants online, but it’s hard to find the money or the people. You’re so focused on keeping energy flowing that installing clean energy and reducing emissions aren’t even on your radar. (Recent budget cuts certainly aren’t helping.) You can’t fix climate change, and neither can anyone else. It’s easier to simply not care about it at all. So you let the clean energy ball drop. So does everyone else in the department. And no one ever picks it up.
Now, Congress could certainly pass a law requiring that employees conform to climate change standards—say, implement 100% renewable energy by 2050. But that doesn’t change the underlying situation where the employee has far too much going on already. If they’re required by law to hold this ball, something else might drop. That something else might be their main goal, which is providing and maintaining energy in the country. Congress will claim to be making progress; they are only delaying the problem.
The standard business solution to this problem is to hire more people to spread out the workload. But this new workload might require two, five, or twenty new employees, which requires new meetings for coordination, new managers, new company infrastructure…all the things. It’s inefficient and expensive.
Alternatively, you could leave those people alone and create a new set of people—Department #2—dedicated to reaching the clean energy goal without hurting Department #1’s efficiency. They work on their own, keeping Department #1 updated on their progress with recommendations for preparing for the transition. Department #1 can focus on their project and communicate their needs to Department #2. Because Department #2 is entirely focused on meeting the climate change goal, they will never drop the ball on it.
Department #1 is in fallback mode; Department #2 is in mission mode.
Why Federalism Works
You might recall from history class that the original thirteen U.S. states were originally mostly self-governing colonies. The operative term here is salutary neglect, a policy under the British government where, as long as the colonies fed the British government money, they would be left to their own devices. This is a surprisingly effective policy, and it’s why many empires (the Aztec, Mongols, Inca, Sumer, Panem) operated under the same principles: they collected tributes from the territories they governed but left them mostly alone otherwise.
Consider a local government that has to submit a tribute to its empire. That’s a single ball that they need to care about dropping, and it’s a relatively easy one to maintain—just subtract 10% from your yearly revenue and operate as if it doesn’t exist. It’s annoying, certainly, but you can live with it. A human tribute is even easier to manage—just designate a few people to organize the selection process when it comes around. It doesn’t affect your work because you’re not the ones to blame.
In exchange, there are some messy parts of governance that the local government would prefer not to take care of. The early U.S. government had just a few purposes: manage a military, manage a treasury, and deal with international affairs, among others. Militaries are annoying to maintain, rarely used, and better in large numbers, so it makes sense to spread them across a larger population. Commerce is easier when it’s uniform across a large population, so the states are happier when the national government runs finances. It’s hard to manage relations with countless countries across the world, especially for a small government; better to leave that to a larger government that can designate more resources to it.
When the U.S. Constitution was created, it expressly gave the government these powers and nothing more:
The powers not delegated to the United States by the Constitution, nor prohibited by it to the States, are reserved to the States respectively, or to the people.
– Tenth Amendment to the U.S. Constitution
In this sense, state governments act as Department #1, and the national government acts as Department #2. Department #1 juggled a lot of balls and found that a few were quite annoying, so it created Department #2 to take care of those. Department #2 takes care of those balls and basically nothing else, allowing Department #1 to do its job with relatively little interruption.
The states are in fallback mode; the national government is in mission mode.
Why Federalism Doesn’t Work
The above section discusses how federalism is supposed to work. Unfortunately, that’s not what happens in the U.S. political system.
The federal government has massively grown in its power since its inception, largely because it has been given more balls to manage or it has decided to manage them itself. The passage of the 13th and 14th Amendments explicitly gave the government the power to regulate race issues. The 15th and 19th Amendments allow the government to protect minority and women voters in election practices, which were left to the states under the Constitution’s Election Clause. The New Deal and its accompanying Supreme Court decisions vastly expanded the Constitution’s Interstate Commerce Clause to include commerce within states (!!), getting the national government intimately involved in agriculture and other economic regulation. Programs like Medicare, Medicaid, and the Affordable Care Act get the government involved in healthcare. The Clean Air Act and Clean Water Act get the government involved in environmental regulation through the EPA.
I’d imagine many Americans are surprised to hear that the Ninth Amendment exists. We hear about the national government much more than the state government, and it’s generally assumed that major policies will be decided at the national level. Typically, policy-related protests ask Congress to take on more balls and add new regulations—legalize/ban abortion, decriminalize marijuana use, end/allow affirmative action. Recent bills have even proposed simply banning state legislation, such as the proposed 10-year moratorium on state AI legislation. States tend to be focused on localized day-to-day items like providing energy and education, but even these are subject to national oversight.
This is part of why scholars like the writers of Abundance think it’s so hard to do things in the U.S. So many blanket national regulations exist—limit to this environmental impact, protect these union interests, subject to these inspections—that complying with all of them is expensive and nearly impossible. To the best of my knowledge, that’s much of what has delayed California’s high-speed rail program and New York’s Second Avenue Subway.
Interlude: How Stuff Gets Done as Gradient Descent
Policymakers are trying to find the set of policies that produces the optimal outcome for some reward function, typically net happiness for their constituents. Most policy changes come with some tradeoff—for example, increasing defense spending might make people safer and thus feel happier, but will force them to pay more in taxes and thus feel sadder. When policymakers act in good faith, they debate these tradeoffs to find the ones that will maximize their reward function.
Anyone who has studied gradient descent will tell you that the obvious failure mode is entering a local minimum. With climate policy, for example, people will be happier if greenhouse gas emissions slow down and CO2 levels are reduced, but there’s an annoying intermediate step where people have to live with high CO2 levels while also suffering the increased costs of reducing emissions. In a more familiar example, drivers will always be happier after road repairs are made, but they have to live with the negative impacts of construction in the meantime. Each of these examples is one dimension of the enormous optimization game that policymakers play.
I have two examples of this effect. Note that the first contains spoilers for Jaws. If you haven’t seen it, go watch it—it’s fifty years old.
One of the deaths in Jaws is caused by the mayor’s decision to open Amity Island’s beaches on July 4th, knowing that the killer shark likely hadn’t been caught. The mayor made that decision because he considered the decision to close the beaches unthinkable. Doing so would cause so much backlash and protest that it simply wasn’t an option, especially for a mayor desperately trying to be re-elected. The optimal decision, prioritizing safety, was so uncomfortable that the mayor simply deluded himself into making it easier. That discomfort was the price of a better outcome.
Soon after the release of Jaws, the Soviet Union began its collapse. When it finally dissolved in 1991, Russia and many former Soviet countries adopted a policy of “shock therapy”: they knew that a transition to capitalism would cause some damage, so they tried to do it as fast as possible to avoid that intermediate step. After significant declines in economic productivity, every former Soviet country rebounded and began growing under the capitalist model. That decline was the difficult intermediate step preceding greater long-term stability.
TANGENT: This doesn’t change the validity of the model, but I should note that the transition failed to produce stable capitalist democracies. The countries that saw the greatest “shock” elected populists and autocrats who abandoned reformers’ democratic aims. Vladimir Putin came to power after ten years of sustained economic decline in Russia, the country arguably hit hardest by reforms. In contrast, the greatest economic success stories come from countries like China and Poland that underwent slow change with stable institutions throughout. (That argument comes from Andrew Walder, one of my professors this quarter. I think he’s probably right.)
Making Federalism Work
Back to government structure. As a reminder, I’m arguing that the federal government is spread too thin to act on important missions, as it was intended to do.
I wrote the above sections of this post without having a solution in mind. As such, the rest of this essay is relatively weak. I’d love to discuss this section more and find better solutions.
We find effective mission-style governance in many other contexts, especially in international governance. Despite its many flaws, the U.N. has been shockingly effective at establishing international standards that improve the world’s quality of life. The Montreal Protocol has turned the depletion of the ozone layer into a complete non-issue. The Geneva Conventions and Universal Declaration of Human Rights have led to massive reductions in war crimes and human rights violations. UNICEF, UNHCR, and the World Food Programme have done wonders in protecting the world’s most vulnerable populations. The U.N. hasn’t stopped great power wars or solved poverty, but it’s made progress on many missions that wouldn’t be addressed otherwise.
Similarly, many single-focus organizations have absolutely accomplished their missions. Consider IATA, the International Air Transport Association, which has standardized international air travel to an incredible degree. Or the International Telecommunication Union, which manages the Internet, satellite orbits, and radio frequencies so well that communication across the world is always possible.
I’d like to apply those ideas to domestic governance.
I recently came across Saikat Chakrabarti’s campaign for the House seat currently held by Nancy Pelosi. I don’t think that campaign will be successful, and Manifold doesn’t think so either. However, I think he’s going about governance the right way. His plan for governance comes from his time with a think tank called New Consensus, where he developed the so-called Mission for America, a New Deal–type program intended to provide an economic boost and avert global warming. The plan proposes “national missions” in twenty different sectors, from EVs to shipping to geothermal energy, that would provide the energy needed to modernize those sectors and advance the economy.
Following the gradient descent idea, Chakrabarti’s thesis is that the U.S. is stuck in a local minimum. Policymakers avoid making changes because they don’t want to deal with the uncomfortable transition period that precedes a new equilibrium. Making well-informed changes as fast as possible allows the country to avoid that middle ground.
I think the Mission for America does climate policy the right way. Most current climate policy involves regulations—use x% clean energy, reach net zero emissions by 2040 or 2050 or 2075, pay a x% carbon tax on fossil fuels. These get the ball rolling in the right direction, but they’re unpopular because they draw on sustained goodwill from everyone they impact. Carbon pricing in Australia and Canada has produced massive resistance, including from my own grandmother. It also draws significant resistance from corporate interests, who can use sustained influence to slow down and build opposition to small and potentially unpopular policies.
In contrast, a “national mission” style of policy uses its energy all at once to get past sustained resistance and make real progress. Because it moves faster, it’s harder to oppose, and the effects it brings can be seen much sooner. (Relatedly, I’m not a fan of recall elections because I’m a fan of letting policies play out. Missions commit to policies so that it’s much harder to recall them. That failed in post-Soviet Russia, and look what’s happened since.) Having energy available means policymakers can use tools other than basic regulations and create systems that work with existing ones rather than create friction.
This mission framework is very much applicable outside of climate policy. For example, it’s roughly what Zohran Mamdani is doing to New York City. His administration has an enormous amount of energy behind it, which allows him to make policy changes that would rarely happen normally—building massive new housing blocks, providing free city services, freezing rent. Doing all that requires breaking some “rules” of policy, e.g., allowing the market to determine prices on its own.
It remains to be seen what Mamdani’s impact will be. He hasn’t made sweeping changes yet, but he’s made impressive progress on day-to-day items by treating them as missions. His administration had an impressive response to a snowstorm a couple months ago, it’s been on a pothole-fixing spree in recent months, and it’s secured guarantees from the state government to expand child care offerings. Despite his meetings with Trump, his administration isn’t trying to address nationwide political issues (which, to be fair, is true for many city governments).
I’m very much okay with that. It’s arguable that many of the pitfalls in American governance stem from its structure, with separate local, state, and national governments that can each make policies on the same issues. Every level of government needs to focus on not dropping any balls, despite not having the resources to do so. Regulations and funding sources appear arbitrarily at every level of government without much coordination. As a result, potholes never get fixed, high-speed rail never gets built, and climate issues go unaddressed.
I don’t support all of Mamdani’s policies, but I was excited to see him be elected because I hoped to see New York as a laboratory for active Democratic governance. So far, it looks like that style of governance has proven successful. I hope that’s not the last we’ll see of it. I’d like to see every level of government using missions to address their own issues: city governments fixing potholes and clearing streets, state governments providing education and social protections, and the national government addressing international economic and climate issues. Maybe then we can finally fix America’s failing bureaucracy.
A cross-post from my personal website, Chasing Sunsets.
Epistemic note: This is a long post that applies two models of work to the government context. I've found those models to be useful in that context and elsewhere, but the arguments I present are very much debatable, and I'm very much open to discussing them.
I’m going to tell you a story. It is not a true story. Please ignore that fact; you can argue with me later if it seems relevant.
In The Games I Play, I mentioned that I used to play Cities: Skylines, a game whose objective is to build as large of a city as possible. It’s a difficult game to play.
Your city begins with a single highway exit. You connect a road to it, a simple two-lane road because that’s all you have the money to build. You zone that road for houses. People move into those houses, but those people need work, so you build commercial and industrial regions nearby. Commercial zoning might be right along your main street, but you can’t put your industry right next to people’s homes because they’ll complain about noise pollution if you do.
This is your first rule.
Your city develops. You add a water pump, then a sewage pipe, which has to be far downstream of the pump because your population will get sick otherwise. You unlock elementary schools and build one for every hundred kids in the city. Your main intersection gets blocked by traffic because it’s used for industrial trucks and residents’ cars, so you give each new industrial area a highway exit. You destroy a couple homes to add parks in one neighborhood because its residents are unhappy. You add fire stations every ten blocks. There’s traffic again between the residential and commercial parts of the city, so you add metro lines in a clockwise and counterclockwise loop around the city. You try adding high-density housing, but it gets too loud and happiness levels decline.
Your residents still need jobs. Not the commercial jobs—those need higher education levels and you don’t have a university in the city yet. But you can’t build an industrial area. One area might work, but it’s not near any major highways and trucks couldn’t get out of the city. The city’s only other undeveloped area would end up polluting a residential zone. You can’t condense that residential zone because any more high-density housing will make homeowners unhappy. Besides, doing that would cut off your existing metro stations, and you’re not sure you have enough money to move them.
As a result, people start leaving the city. You try to bring them back—you add recycling service, build more parks, allow them to smoke weed—but nothing seems to work. Abandoned homes line your blocks; new residents move in every so often, only to demand industrial jobs that don’t exist. Tax revenue declines, and your population plummets. Frustrated, you turn off the game and go back to your math homework.
If only you could build an industrial zone.
How Stuff Gets Done
I recently read One Day Sooner and Never Drop A Ball, a pair of LessWrong posts that present a dichotomy in how organizations function. They’re worth reading in full if you have time. If you don’t have time, I’ll briefly summarize each one.
One Day Sooner is a mode of working that aggressively pursues a single goal. The canonical example is vaccine trials, where organizations like 1DaySooner work to get vaccines out as quickly as possible. The idea is that some goals, like getting vaccines to the public or eliminating a bottleneck in a project, are so important that they should be pursued immediately. The post describes how to get to things One Day Sooner, typically involving pushing through unnecessary slowdowns like “let’s schedule this meeting three days from now” or “I need a break from work tonight.”
Never Drop A Ball is a mode of working that pursues several goals or projects and ensures none of them are lost in the shuffle. Where a startup founder aiming for One Day Sooner might spend a sixty-hour week working on their core project, a founder aiming to Never Drop A Ball will carefully keep track of every bit of work, ensuring things like taxes and floor cleaning are properly handled. Notably, the failure mode of Never Drop A Ball work is that balls need to be dealt with; if they take too much work to resolve or there are too many to handle, balls will begin to drop.
The difference between these modes is how many balls they handle. Aiming for One Day Sooner means juggling one or two balls, typically ones that are very difficult to resolve, and trying very hard to resolve them. Aiming to Never Drop A Ball means juggling many balls—and, importantly, juggling new balls that come up—which tend to be easy to resolve and thus easy to miss. The number of balls and the likelihood of being assigned new balls are sliding scales, but it’s often easier to think of them as a binary. As anyone who has worked on a project can tell you, dealing with a small number of balls is associated with a low likelihood of being assigned new balls because new balls tend to be distracting.
In the rest of this essay, aiming for One Day Sooner will be referred to as “mission mode,” while aiming to Never Drop A Ball will be referred to as “fallback mode.”
How Stuff Gets Done in Government
Governments are intended to do things. These things can be thought of as balls.
There are a lot of balls floating around in government.
Governments as a whole act in fallback mode. This is because governments do many, many things and can never make a mistake. A single misapplication of power—say, a police officer making an unusually dangerous arrest and accidentally killing George Floyd—could result in nationwide protests and shake the foundation of the government’s power. In the words of Thomas Jefferson:
Because of this fragility, governments create all sorts of rules to prevent making mistakes. This is the origin of the endless bureaucracy of the U.S. government. It is extremely frustrating. It is also written in blood. Every rule, regulation, and system is created for a reason; changing it might make something more efficient, but it might cause something else to fail, which is the absolute worst thing that can happen to a government.
Unfortunately, this falls squarely into the failure mode of never dropping a ball. The government becomes consumed with keeping things maintained and functional rather than making progress on new ideas.
Imagine you’re an official working on energy policy in West Virginia. A declining population means fewer people to fill jobs in power plants. Ancient coal plants are failing left and right as infrastructure becomes dilapidated. You’re trying to get new gas plants online, but it’s hard to find the money or the people. You’re so focused on keeping energy flowing that installing clean energy and reducing emissions aren’t even on your radar. (Recent budget cuts certainly aren’t helping.) You can’t fix climate change, and neither can anyone else. It’s easier to simply not care about it at all. So you let the clean energy ball drop. So does everyone else in the department. And no one ever picks it up.
Now, Congress could certainly pass a law requiring that employees conform to climate change standards—say, implement 100% renewable energy by 2050. But that doesn’t change the underlying situation where the employee has far too much going on already. If they’re required by law to hold this ball, something else might drop. That something else might be their main goal, which is providing and maintaining energy in the country. Congress will claim to be making progress; they are only delaying the problem.
The standard business solution to this problem is to hire more people to spread out the workload. But this new workload might require two, five, or twenty new employees, which requires new meetings for coordination, new managers, new company infrastructure…all the things. It’s inefficient and expensive.
Alternatively, you could leave those people alone and create a new set of people—Department #2—dedicated to reaching the clean energy goal without hurting Department #1’s efficiency. They work on their own, keeping Department #1 updated on their progress with recommendations for preparing for the transition. Department #1 can focus on their project and communicate their needs to Department #2. Because Department #2 is entirely focused on meeting the climate change goal, they will never drop the ball on it.
Department #1 is in fallback mode; Department #2 is in mission mode.
Why Federalism Works
You might recall from history class that the original thirteen U.S. states were originally mostly self-governing colonies. The operative term here is salutary neglect, a policy under the British government where, as long as the colonies fed the British government money, they would be left to their own devices. This is a surprisingly effective policy, and it’s why many empires (the Aztec, Mongols, Inca, Sumer, Panem) operated under the same principles: they collected tributes from the territories they governed but left them mostly alone otherwise.
Consider a local government that has to submit a tribute to its empire. That’s a single ball that they need to care about dropping, and it’s a relatively easy one to maintain—just subtract 10% from your yearly revenue and operate as if it doesn’t exist. It’s annoying, certainly, but you can live with it. A human tribute is even easier to manage—just designate a few people to organize the selection process when it comes around. It doesn’t affect your work because you’re not the ones to blame.
In exchange, there are some messy parts of governance that the local government would prefer not to take care of. The early U.S. government had just a few purposes: manage a military, manage a treasury, and deal with international affairs, among others. Militaries are annoying to maintain, rarely used, and better in large numbers, so it makes sense to spread them across a larger population. Commerce is easier when it’s uniform across a large population, so the states are happier when the national government runs finances. It’s hard to manage relations with countless countries across the world, especially for a small government; better to leave that to a larger government that can designate more resources to it.
When the U.S. Constitution was created, it expressly gave the government these powers and nothing more:
In this sense, state governments act as Department #1, and the national government acts as Department #2. Department #1 juggled a lot of balls and found that a few were quite annoying, so it created Department #2 to take care of those. Department #2 takes care of those balls and basically nothing else, allowing Department #1 to do its job with relatively little interruption.
The states are in fallback mode; the national government is in mission mode.
Why Federalism Doesn’t Work
The above section discusses how federalism is supposed to work. Unfortunately, that’s not what happens in the U.S. political system.
The federal government has massively grown in its power since its inception, largely because it has been given more balls to manage or it has decided to manage them itself. The passage of the 13th and 14th Amendments explicitly gave the government the power to regulate race issues. The 15th and 19th Amendments allow the government to protect minority and women voters in election practices, which were left to the states under the Constitution’s Election Clause. The New Deal and its accompanying Supreme Court decisions vastly expanded the Constitution’s Interstate Commerce Clause to include commerce within states (!!), getting the national government intimately involved in agriculture and other economic regulation. Programs like Medicare, Medicaid, and the Affordable Care Act get the government involved in healthcare. The Clean Air Act and Clean Water Act get the government involved in environmental regulation through the EPA.
I’d imagine many Americans are surprised to hear that the Ninth Amendment exists. We hear about the national government much more than the state government, and it’s generally assumed that major policies will be decided at the national level. Typically, policy-related protests ask Congress to take on more balls and add new regulations—legalize/ban abortion, decriminalize marijuana use, end/allow affirmative action. Recent bills have even proposed simply banning state legislation, such as the proposed 10-year moratorium on state AI legislation. States tend to be focused on localized day-to-day items like providing energy and education, but even these are subject to national oversight.
This is part of why scholars like the writers of Abundance think it’s so hard to do things in the U.S. So many blanket national regulations exist—limit to this environmental impact, protect these union interests, subject to these inspections—that complying with all of them is expensive and nearly impossible. To the best of my knowledge, that’s much of what has delayed California’s high-speed rail program and New York’s Second Avenue Subway.
Interlude: How Stuff Gets Done as Gradient Descent
Policymakers are trying to find the set of policies that produces the optimal outcome for some reward function, typically net happiness for their constituents. Most policy changes come with some tradeoff—for example, increasing defense spending might make people safer and thus feel happier, but will force them to pay more in taxes and thus feel sadder. When policymakers act in good faith, they debate these tradeoffs to find the ones that will maximize their reward function.
Anyone who has studied gradient descent will tell you that the obvious failure mode is entering a local minimum. With climate policy, for example, people will be happier if greenhouse gas emissions slow down and CO2 levels are reduced, but there’s an annoying intermediate step where people have to live with high CO2 levels while also suffering the increased costs of reducing emissions. In a more familiar example, drivers will always be happier after road repairs are made, but they have to live with the negative impacts of construction in the meantime. Each of these examples is one dimension of the enormous optimization game that policymakers play.
I have two examples of this effect. Note that the first contains spoilers for Jaws. If you haven’t seen it, go watch it—it’s fifty years old.
One of the deaths in Jaws is caused by the mayor’s decision to open Amity Island’s beaches on July 4th, knowing that the killer shark likely hadn’t been caught. The mayor made that decision because he considered the decision to close the beaches unthinkable. Doing so would cause so much backlash and protest that it simply wasn’t an option, especially for a mayor desperately trying to be re-elected. The optimal decision, prioritizing safety, was so uncomfortable that the mayor simply deluded himself into making it easier. That discomfort was the price of a better outcome.
Soon after the release of Jaws, the Soviet Union began its collapse. When it finally dissolved in 1991, Russia and many former Soviet countries adopted a policy of “shock therapy”: they knew that a transition to capitalism would cause some damage, so they tried to do it as fast as possible to avoid that intermediate step. After significant declines in economic productivity, every former Soviet country rebounded and began growing under the capitalist model. That decline was the difficult intermediate step preceding greater long-term stability.
TANGENT: This doesn’t change the validity of the model, but I should note that the transition failed to produce stable capitalist democracies. The countries that saw the greatest “shock” elected populists and autocrats who abandoned reformers’ democratic aims. Vladimir Putin came to power after ten years of sustained economic decline in Russia, the country arguably hit hardest by reforms. In contrast, the greatest economic success stories come from countries like China and Poland that underwent slow change with stable institutions throughout. (That argument comes from Andrew Walder, one of my professors this quarter. I think he’s probably right.)
Making Federalism Work
Back to government structure. As a reminder, I’m arguing that the federal government is spread too thin to act on important missions, as it was intended to do.
I wrote the above sections of this post without having a solution in mind. As such, the rest of this essay is relatively weak. I’d love to discuss this section more and find better solutions.
We find effective mission-style governance in many other contexts, especially in international governance. Despite its many flaws, the U.N. has been shockingly effective at establishing international standards that improve the world’s quality of life. The Montreal Protocol has turned the depletion of the ozone layer into a complete non-issue. The Geneva Conventions and Universal Declaration of Human Rights have led to massive reductions in war crimes and human rights violations. UNICEF, UNHCR, and the World Food Programme have done wonders in protecting the world’s most vulnerable populations. The U.N. hasn’t stopped great power wars or solved poverty, but it’s made progress on many missions that wouldn’t be addressed otherwise.
Similarly, many single-focus organizations have absolutely accomplished their missions. Consider IATA, the International Air Transport Association, which has standardized international air travel to an incredible degree. Or the International Telecommunication Union, which manages the Internet, satellite orbits, and radio frequencies so well that communication across the world is always possible.
I’d like to apply those ideas to domestic governance.
I recently came across Saikat Chakrabarti’s campaign for the House seat currently held by Nancy Pelosi. I don’t think that campaign will be successful, and Manifold doesn’t think so either. However, I think he’s going about governance the right way. His plan for governance comes from his time with a think tank called New Consensus, where he developed the so-called Mission for America, a New Deal–type program intended to provide an economic boost and avert global warming. The plan proposes “national missions” in twenty different sectors, from EVs to shipping to geothermal energy, that would provide the energy needed to modernize those sectors and advance the economy.
Following the gradient descent idea, Chakrabarti’s thesis is that the U.S. is stuck in a local minimum. Policymakers avoid making changes because they don’t want to deal with the uncomfortable transition period that precedes a new equilibrium. Making well-informed changes as fast as possible allows the country to avoid that middle ground.
I think the Mission for America does climate policy the right way. Most current climate policy involves regulations—use x% clean energy, reach net zero emissions by 2040 or 2050 or 2075, pay a x% carbon tax on fossil fuels. These get the ball rolling in the right direction, but they’re unpopular because they draw on sustained goodwill from everyone they impact. Carbon pricing in Australia and Canada has produced massive resistance, including from my own grandmother. It also draws significant resistance from corporate interests, who can use sustained influence to slow down and build opposition to small and potentially unpopular policies.
In contrast, a “national mission” style of policy uses its energy all at once to get past sustained resistance and make real progress. Because it moves faster, it’s harder to oppose, and the effects it brings can be seen much sooner. (Relatedly, I’m not a fan of recall elections because I’m a fan of letting policies play out. Missions commit to policies so that it’s much harder to recall them. That failed in post-Soviet Russia, and look what’s happened since.) Having energy available means policymakers can use tools other than basic regulations and create systems that work with existing ones rather than create friction.
This mission framework is very much applicable outside of climate policy. For example, it’s roughly what Zohran Mamdani is doing to New York City. His administration has an enormous amount of energy behind it, which allows him to make policy changes that would rarely happen normally—building massive new housing blocks, providing free city services, freezing rent. Doing all that requires breaking some “rules” of policy, e.g., allowing the market to determine prices on its own.
It remains to be seen what Mamdani’s impact will be. He hasn’t made sweeping changes yet, but he’s made impressive progress on day-to-day items by treating them as missions. His administration had an impressive response to a snowstorm a couple months ago, it’s been on a pothole-fixing spree in recent months, and it’s secured guarantees from the state government to expand child care offerings. Despite his meetings with Trump, his administration isn’t trying to address nationwide political issues (which, to be fair, is true for many city governments).
I’m very much okay with that. It’s arguable that many of the pitfalls in American governance stem from its structure, with separate local, state, and national governments that can each make policies on the same issues. Every level of government needs to focus on not dropping any balls, despite not having the resources to do so. Regulations and funding sources appear arbitrarily at every level of government without much coordination. As a result, potholes never get fixed, high-speed rail never gets built, and climate issues go unaddressed.
I don’t support all of Mamdani’s policies, but I was excited to see him be elected because I hoped to see New York as a laboratory for active Democratic governance. So far, it looks like that style of governance has proven successful. I hope that’s not the last we’ll see of it. I’d like to see every level of government using missions to address their own issues: city governments fixing potholes and clearing streets, state governments providing education and social protections, and the national government addressing international economic and climate issues. Maybe then we can finally fix America’s failing bureaucracy.