In just a matter of days, President Donald Trump went from expressing the desire that we get the economy going in just a few weeks to announcing that “social distancing guidance” would remain in effect until April 30 to help us battle COVID-19. That was a welcome and important change, but neither decision was truly in his hands. Thanks to our long  tradition of decentralized decision-making, that power lies primarily with our state governors and local leaders.

We can take some lessons from that tradition to understand how our governments have responded thus far to the pandemic. More importantly, these lessons can inform our thinking—and give us encouragement—about how our elected officials can and should act in the weeks and months ahead.

Though there’s much talk about the horizontal separation of powers among our federal branches, we can take for granted America’s vertical distribution of authority—among Washington, states, counties, cities, towns, non-governmental bodies, and individuals. But decentralization has a profound influence on how we go about the public’s business and solve problems. Thanks in part to our settler, revolutionary, pioneer, and immigrant roots, we have developed habits and institutions related to freedom and proximate self-government. So for more than 200 years, Americans have been acclimatized to liberty, local democracy, and civil-society activity.

Everyone wants a measure of control over their own lives and to be able to help shape their communities. Decentralization offers that by providing efficacy and agency—a sense that we can make a difference and the tools to do so. But in a highly diverse, continental republic like the United States, decentralization is even more important. Our citizens and their associations will always possess a vast array of needs and priorities. Close-to-home decision-making maximizes our capacity to have government action reflect this diversity.

We are willing to shift some power to farther-away levels of government if we’re convinced there’s a good reason. But we recognize that there are costs to doing so. More power in Washington or a state capital might help solve a stubborn problem, but your ability to govern yourself and your community decreases; a centrally created, uniform policy is less likely to fit the needs of your community; policies crafted by distant “experts” can ignore the accumulated wisdom of trial and error-shaped local practice; and centralized policies are notoriously difficult to adjust and produce scads of unintended consequences.

Beyond their general, day-to-day authorities, the executives of the federal and state governments are also empowered to take bold action in times of domestic crises (e.g. states of emergency, public-health emergencies). In these extreme cases, we accept that central authority will expand to match the scope of the problem. We—and the courts—tolerate faraway leaders making huge decisions that massively influence, even injure, our lives. The COVID-19 pandemic is a case in point. Its ultimate human toll will be staggering. President Trump is now bracing the nation for more than 200,000 Americans lives lost. But without dramatic action, it could be more than 2 million.

We should bear in mind, however, that just because conditions have temporarily placed greater weight on the centralization side of the scale, that does not mean that the weight on the decentralization side has disappeared: It has merely been outweighed for the time being. The reasons for power distribution still exist and the costs of not distributing power still exist. And those reasons have been on display since the crisis started.

After downplaying the risks for entirely too long, the Trump administration has mobilized some of the federal government’s authorities. Many governors acted earlier and more energetically—for example, by limiting the size of public gatherings and closing non-essential businesses.

One of the boldest moves has been the mandatory closure of schools. Ohio Gov. Mike DeWine announced on March 12 that schools in his state would close; the governors of Maryland, Kentucky, New Mexico, and Michigan followed suit the same day. Though local districts have much power, under state constitutions in most places state governments have ultimate control. That authority (often located in state boards of education), plus governors’ power to take extraordinary action to protect citizens’ safety, enabled such action. As of this writing, almost all states have closed their schools, with variation in duration.

So our leaders determined that the centralized decision-making side suddenly had enormous weight, enough to tip the scale. Such a decision was probably priceless, but it was not costless. Those familiar with K-12 education policy and operations immediately understood how disruptive such a dramatic decision would be. Because schools have been locally operated and governed for so long and because they are so integral to how communities function, this swift, central decision implicated countless activities.

For example, many low-income students get their most reliable meals at school; many homeless students rely on schools for a host of services. Many young people live with grandparents and others with health risks. Many families cobble together complicated work and childcare routines based on reliable school schedules. Few districts are equipped to suddenly deliver education online and at scale. State and federal policies that mandate equitable provision of educational services raise questions about the legality of offering online content if not all students can access computers at home or if that content can’t be delivered in a way that meets the needs of students with disabilities. And very, very few parents are able to both suddenly work at home and immediately learn how to homeschool adeptly.

Now briefly consider the economic costs associated with our leaders’ bold decision to prevent gatherings and shutter places of work—unemployment, curtailed consumer spending, businesses that may never open again. The economy lost a shocking, record-breaking 3.3 million jobs in one week. Goldman Sachs forecasts the economy could shrink by as much as 24 percent in the second quarter alone; former Trump economic adviser Kevin Hassett worried about the possibility of a depression if economic life doesn’t return to normal.

For the time being, Americans seem to trust that our leaders’ bold, centralized decision-making is necessary. But at some point, the weights on the two sides of the scale will begin to change and public sentiment will shift. And that will be complicated by the fact that different communities—given their vastly different conditions and priorities—will start to assess the scale differently.

So what are we to do?

***

There are lessons to learn from the way we handle an issue that is near and dear to many citizen’s hearts: Public education. People love their kids and want to influence their local schools. So America has almost 14,000 (that’s right) school districts. Most are quite small—averaging about seven schools apiece. Each district makes most of its own decisions on personnel, contracts, curriculum, and operating hours. Parents and citizens feel reassured when they personally know local teachers, administrators, and school board members. And because we know that even a hyper-local district will not suit everyone perfectly, we also have about 35,000 (that’s right) K-12 private schools. In short, because decisions about kids’ education are so important, we push power down to the local level as much as possible.

At the same time, we want some commonality among a state’s schools, districts, and educators. So we have state-level rules on what kids should learn, rules for teacher certification, and graduation requirements. And there are federal rules guaranteeing that students with special needs receive the services they need and that states aim to help all students reach proficiency in key subjects.

But we monitor this balance assiduously. We tolerate some centralization only if we’re convinced there’s enough benefit. Initial advocates of national or state-based initiatives like the No Child Left Behind act, high-stakes standardized tests, Race to the Top, teacher-evaluation reform, and Common Core were able to convince us that the balance tipped in favor of the pro-centralization side. But over time, many citizens reassessed and decided we’d consolidated too far. We ended up with backlash against all of these centralized initiatives.

The lesson here is that we distribute power in America for clear, compelling reasons, especially in domains of life that are of the most importance to us, where individual control and community agency are invaluable. We give our elected officials the chance to convince us that centralized decision-making is necessary, and we acquiesce from time to time. But we’ll take our power back once we see that individuals, families, local bodies, and civil society better understand the problem and can better address our particular conditions.

Indeed, we should all keep in mind over the next few weeks, as we hopefully turn the corner on this crisis, that people tend to rebel against centralized decision-making for two primary reasons: They determine the overarching uniform policy presents fewer benefits or greater costs than the default decentralized approach; or they assess that the uniform policy—even if it has value in aggregate—is poorly tailored to the needs of their particular community.

These are both consequences of a general problem: A central body is typically incapable of knowing everything necessary to make smart, timely decisions about complex matters affecting a vast array of communities. Even if central authorities are brilliant and fair, they simply can’t collect and analyze every bit of information related to every variable associated with every location. They can’t know the full picture, much less know how to respond to each corner of it.

This causes central decision-makers to limit their field of vision. Since they can’t know everything, they focus outsized attention on a very few things. This can be thought of as the Inverse Law of Metrics: As decision-makers’ distance from people’s day-to-day lives increases, the number of indicators the decision-makers deem critical decreases. This helps explain why central administrators charged with increasing lumber yields assessed forests based narrowly on timber production; why some central city planners focused narrowly on efficient design; why some economists focused narrowly on GDP; why some central school authorities focused narrowly on standardized test scores.

The blessing is that central figures typically pick a metric whose value is unquestionably important—GDP is valuable, test scores are valuable. The curse is that such limited metrics don’t reflect the entirety of the systems we care about. There is more to a forest than timber, to city life than efficiency, to economic health than GDP, to schools than test scores. But central leaders can, with justification, defend the importance of their chosen variable. They can then focus the public’s attention on and orient the government’s behavior around that measure and downplay or neglect other valuable considerations.

This is one way to think about “flattening the curve.” This swiftly became the summum bonum of our COVID-19 response. For very good reason, public-health experts have focused the nation’s attention on this metric. It is of indisputable importance. Indeed, it is probably the best way to think about health success when viewed from the national level. It will enable our health care system to save lives because hospitals can be supplied, treatments can be tested, and the total case load will spread over time. It will prevent untold misery among patients, their loved ones, and among those suffering other ailments who can’t get the care they need. No one should doubt the extraordinary cost of allowing this virus’s spread to overwhelm a geography’s health-care capacity.

But in order to truly flatten the curve, extraordinary measures—along the lines of shutting down the economy and social interactions—are required, measures that dramatically affect education, income, childcare, mortgages, jobs, consumer spending, and much more. And those factors become more and more real—and more and more terrifying—when you move from viewing today’s situation telescopically from a central location to microscopically from inside individuals’ daily lives. This helps explain President Trump’s unexpected declaration that our response to the virus shouldn’t be worse than the virus itself and his gradually floating the possibility of easing restrictions, culminating in his encouragement “to have the country opened up and just raring to go by Easter.”

But that approach suffers the same problem. Looking exclusively at the national economic picture misses that cases are spiking in particular locations with potentially catastrophic effects. One lesson from abroad is that easing restrictions too quickly can erase hard-fought gains.

So how can we possibly balance these seemingly mutually exclusive goals—applying a flatten-the-curve social brake and an up-and-running economic jolt?

With prudence: informed, careful, compassionate planning and decision-making. And this is where decentralization becomes essential.

So far, perhaps unsurprisingly, America’s governors have generally shown the kind of leadership we need. Because they are aware of the risks of COVID-19 and knowledgeable about the conditions of their cities, towns, schools, industries, and hospitals, they are positioned to balance competing priorities. It is why many of them moved quickly to require social distancing and close workplaces—and now avoid setting arbitrary timelines for changes while holding the line on their tough policies despite the president’s oscillation. It’s why they are focused on tactical matters that can save lives now, like acquiring ventilators, hospital beds, and protective equipment. It’s also why they are continuously adjusting policies on gatherings, testing, and what constitutes an “essential” business. It’s also why different governors are charting modestly different courses.

People will trust their leaders’ consolidation of power during these tough times, and they will welcome ameliorative strategies like the economic relief legislation’s personal rebates, unemployment benefits, and support for hospitals. But they will also monitor whether the overarching policy direction is serving the nation well and whether those policies are the right fit for their particular communities. Neither a tepid response that allows this crisis to explode nor a uniform, heavy-handed response that poorly matches an area’s conditions will suffice. Governors will track new cases, hospital capacity, job losses, bankruptcies, foreclosures, public sentiment, and much more. They are likely to experiment with a variety of approaches to social tracing, mass testing, selective containment, temperature checks, and masks. Their mayors and county commissioners will provide an ongoing stream of information. States will make continuous and responsive course-corrections—evidenced in the last several days by changes in school closures and online learning.

In the days and weeks ahead, we may well see differentiation among communities within states—related to sheltering in place, business closures, schools, and more. National figures should certainly keep an eye on these changes to ensure they don’t cause unnecessary risk while remembering that local leaders will know far more about their communities than someone hundreds or thousands of miles away. Let us not underestimate the knowledge, wisdom, energy, and compassion of people outside the Beltway. And let us not underestimate the vital details of individuals’ and communities’ lives that are missed by good-hearted, well-meaning but faraway leaders who possess only a synoptic understanding of conditions.

And herein lies the ultimate challenge for leaders in Washington now and leaders in state capitals soon. They will feel an extraordinary responsibility to solve this crisis, and they possess extraordinary power in this moment. In these first weeks, they need a certain set of skills and dispositions: foresight, courage, decisiveness, and resolve. This is essential for making and sticking with tough, uniform pronouncements.

But as this crisis eventually stabilizes and slowly abates, a different set of skills will be essential: humility, judiciousness, and flexibility. Although we’ve all been holding on tight for weeks, these leaders will need to find the strength to loosen the reins. That means distributing a degree of decision-making authority and trusting the great American strengths of localism, pluralism, philanthropy, and social entrepreneurialism. Communities are capable of developing astonishingly complex and locally responsive courses of action—plans that can be nimbly adjusted as conditions change. Our leaders will need to allow individuals and their communities to do so. Most governors probably have the equivalent of a war room right now, where the state’s executive can gather information and make final calls. But governors should be ramping up a room next door, staffed by a team with a different mindset: how to go about rolling back our firm decisions.

At this point, the most admirable public figures have been those who’ve understood why and how our system facilitates the consolidation of power as a matter of necessity in periods of crisis. In time, however, we will recognize that the extraordinary leaders of this era were those who also understood why and how our system widely distributes power as a matter of principle.