To gauge how the private schooling sector has fared amid the coronavirus pandemic, Cato’s Center for Educational Freedom launched the COVID-19 Permanent Private School Closures tracker. The tracker attempts to capture all private schools that announced permanent closure—not just temporary closure of a building to in‐​person education—with the effects of the COVID-19 pandemic having at least partially driven the decision. The tracker includes school names, closure announcement dates, tuition levels, school type, student demographics, and more. In addition, the Center for Educational Freedom has collected data on private schools that have closed without the pandemic as a contributing factor and has attempted to gauge the effect of COVID-19 on overall private school enrollment.

This paper provides data on private school closures one year after the first reported COVID‐​19‐​connected closure, covering March 18, 2020, to March 17, 2021. It also discusses overall private schooling vitality, including why the pandemic’s effects on the sector were less dire than many people likely expected, which schools and areas were most affected, and what policies going forward can preserve private schooling for families that desire it.

Few Americans alive in March 2020 will likely forget how quickly and totally life changed as COVID-19 went from an obscure illness popping up in faraway places to a pandemic that saw the country and world enter lockdown. People were urged and often mandated to remain at home and distanced from anyone outside their households to slow the spread of the disease. Schools of all types—traditional public, charter, private—quickly halted in‐​person education, and instruction either moved online, became packets of worksheets to be completed at home, or did not occur at all.

The halting of in‐​person instruction for basically all schools lasted through the end of the 2019–20 school year, but by June 2020, there was optimism about many schools allowing it for the next school year. COVID-19’s spread had been slowing, hitting a national seven‐​day moving average low of 4 percent positive tests on June 9.¹ But then positive rates rose until July 23, and optimism about in‐​person education dwindled among public school districts. Many of them changed plans from fully in‐​person schooling to either hybrid—some days in school each week, some at home—or entirely online.² Such changes often did not occur smoothly, with many alterations of plans.

While public schools tended to eschew in‐​person schooling, private schools typically embraced it. A survey conducted by the journal Education Next in November and December 2020 found that while 60 percent of private school students were receiving in‐​person instruction, only 24 percent of traditional public school students were.³ In some places, such as Los Angeles, private schools were prohibited by law from opening in person.⁴ The share of private students physically in school would have likely been even higher otherwise.

As the school year proceeded through the winter holidays and into 2021, the country saw a steep increase in cases,⁵ and public schooling tended to remain virtual. When positive testing rates hit a long decline, however, and vaccines started rolling out, pressure rose on public schools to reopen. Large districts that had been almost entirely virtual, such as Chicago Public Schools, began to reopen to in‐​person classes, at least for younger children,⁶ though others remained virtual as teacher unions objected to returning to classrooms and many parents remained concerned about sending their children back to school.⁷

Private schools are an important part of American elementary and secondary education, providing instruction and cultures that are often different from what public schools supply. The COVID-19 pandemic appeared to be an especially dire threat to those schools, because unlike public institutions, private schools do not receive dedicated taxpayer funding. They rely on voluntary funding from families and donors, and if the economic effects of COVID-19 crippled such funding, which appeared quite possible at the outset of the pandemic, many private schools seemed doomed.

To track private school closures, the Cato Institute’s Center for Educational Freedom (CEF) monitored news reports via Google alerts and education news aggregation websites. The CEF also solicited closure reports from the public via the COVID-19 Permanent Private School Closures website, the Cato blog, and Twitter. Once informed of a closure, the CEF attempted to ascertain whether the real or feared financial effects of the COVID-19 pandemic played a role in the closure decision, either by seeing the pandemic cited as a factor in media reports or by contacting closing schools or their parent organizations, such as Roman Catholic dioceses. The CEF eventually also compiled non‐​pandemic closure data, both by retroactively searching for closures that had arisen since March 18, 2020—the date of the first pandemic‐​related closure that the CEF documented—and by adding ones it saw reported thereafter.

Once the CEF confirmed a COVID-19 closure, it compiled the following information for each school:

To gather the most recent data available, the CEF used numerous sources, including not only media reports and direct contact but also school websites, the websites Private School Review and Niche​.com, and the U.S. Census Bureau. CEF’s entire database can be accessed from the COVID-19 Permanent Private School Closures webpage.⁸

CEF’s goal is to catalog and assess closures as they occur, and data collection will continue as long as the COVID-19 pandemic may be impacting schools. Table 1 provides summary data for all the categories for which the CEF could consistently obtain information. Excluded is the share of students receiving financial aid, which the CEF could only ascertain for 25 schools. To put these data into greater context, where applicable, Table 1 also includes data for non‐​COVID‐​19 closures and private and public schools nationwide.

The CEF identified 132 private schools that announced closures due at least in part to the COVID-19 pandemic, with that number reduced to 122 when accounting for school mergers. The CEF also saw a handful of schools announce closures and then reverse those decisions, usually after schools and parents undertook special efforts to revive the institutions.⁹ Those schools are not reflected in the closure numbers. In addition, the CEF cataloged 28 private schools that announced closures during the allotted time frame unrelated to the pandemic. As Figure 1 shows, most pandemic‐​related closure announcements occurred between April and August 2020, and there were essentially none after August until an uptick starting in January 2021.

Looking at the COVID-19 closures, the most obvious statistic is that most schools are Roman Catholic, accounting for 84 percent of the total while constituting only about 22 percent of private schools nationwide. The next largest group is independent schools, numbering 12 closures and about 9 percent of the total. All others appear to be one‐​offs, and only Episcopal schools appear to be proportionate to their national share of private schools. Data for non‐​COVID‐​19 closures are similar, with Catholic schools taking up an even larger share—89 percent—and non‐​denominational Christian schools accounting for 7 percent.

Geographically, closures are concentrated in the Northeast, which accounts for 66 percent of COVID-19 closures while having less than 22 percent of all private schools. (It holds about 17 percent of the United States’ population.¹⁰) The Northeast accounts for a bit above half of non‐​COVID‐​19 closures. All other regions are underrepresented, though the Midwest is proportionate for non‐​COVID‐​19 closures. The Midwest also accounts for a much higher share of non‐​COVID‐​19 closures than COVID-19 closures.

Private schools that closed because of the pandemic have a bit larger average enrollment than private schools nationally, averaging 160 students versus 151, while non‐​COVID‐​19 closures come in at 195 students. All are much smaller than the average public school enrollment of 528 students. The range of COVID-19 closure enrollment for K–12 students—some schools include pre‐​K students as well—is 5 kindergarteners at an independent nursery and kindergarten in Rhode Island to 573 students at a Christian school on Long Island. The average age of the schools that went out of business both because of the pandemic and not was 75 years and ranged from 5 to 277 years old. The oldest was Saint Francis Academy in Bally, Pennsylvania, founded in 1743; it shuttered all but its pre‐​kindergarten and kindergarten classes.

The schools on the COVID-19 closure list were typically low‐​cost, averaging tuition of $7,066, and those closed for non‐​pandemic reasons were just a bit pricier, averaging $7,754 (see Figure A). The national average private school tuition is $11,173,¹¹ while the United States spends $14,891 per public school pupil. The range of COVID‐​19‐​closed private school tuition was from free to $26,367 per year. The highest non‐​COVID‐​19 closure tuition was $18,428. The high ends of those ranges may seem steep, but to put them in perspective, 2020–21 tuition at Sidwell Friends in Washington, DC, which has been the school of several presidential children, was about $46,000 (there is variation by grade level).¹² Overall, the closed private schools had significantly below‐​average tuitions, which may explain the next finding: 111 schools on the COVID-19 closure list reported having financial difficulties prior to the pandemic.

Demographically, the average student body of all closures was appreciably less white and more black and Hispanic than private schools nationally (see Figure B). COVID-19 closures had almost the same proportion of black students as public schools, and non‐​COVID‐​19 closures were 12 percentage points higher. Both had lower Hispanic enrollments than public schools and roughly equal Asian shares. They also had smaller Asian enrollment than all private schools.

To get a sense for the income of families that might have used closed private schools, we recorded the median family income in the census tracts where each school was located. (Direct data on families’ incomes were not available.) The average of the median family incomes for COVID-19 closures was $89,953 and non‐​COVID‐​19 was $85,589. The national median income for family households is $88,149.¹³ Median incomes for the high‐​ends of where schools were located were very different for COVID-19 and non‐​COVID‐​19 closures: $270,341 and $150,258, respectively.

Private school closures were not as numerous as some people feared early on. With schools and so many other facilities physically closed and unemployment skyrocketing, it seemed probable that private schools would hemorrhage tuition‐​paying families and supplemental funding from such sources as church collections and fundraisers. The school choice advocacy organization EdChoice estimated the likely costs to public schools if private school students had to return to public schools, with director Robert Enlow stating, “If only [emphasis added] 10 percent of private school students return to the public system, the combined state and local cost would be $6.7 billion.“¹⁴ Enlow noted that during the 2007–2009 Great Recession, 40 percent of private school enrollment “in ‘high‐​severity’ metropolitan areas” was lost, and the study he cited found that “the recession decreased private school enrollment by approximately one‐​third.“¹⁵

Were enrollment to drop by one‐​third in a year and schools disappear proportionately, the loss would have far exceeded the 132 COVID-19 closures and 160 total closures that the CEF cataloged. As of the 2017–18 school year—the latest with national data—there were 32,461 private schools serving K–12 students.¹⁶ A one‐​third loss would be 10,712 schools. Closures have been nowhere near that, though the tracker is almost certainly missing some.

The CEF tracker’s main shortcoming is that it relies primarily on media reports. As a result, it may over‐​represent schools in large media markets (the disproportionately represented Northeast has many such markets) and schools that are part of larger networks, such as Catholic dioceses. Catholic schools also tend to be relatively large for private schools—nationally, the average Catholic school enrolled 278 students, while nearly 14,000 private schools enrolled fewer than 50 students—which might draw greater media attention.¹⁷ The CEF also solicited reports of closures, however, and the few received were for schools that the CEF had already cataloged.

To get a sense of how likely the closure list was to capture the universe of private school closures, several efforts were made to assess enrollment changes between the 2019–20 and 2020–21 school years. The CEF made a quick attempt to assess enrollment movement toward the beginning of the new school year, surveying a randomly selected national sample of 400 private schools.¹⁸ The center received only 62 responses, for a +/– 12 point margin of error. Some 57 percent of schools reported enrollment losses, including pre‐​K students, whose enrollment can contribute important funds to some schools. About 19 percent saw no enrollment change, and 24 percent experienced increases. The average change was a 6 percent loss.

CEF’s survey was small, but roughly similar results were reported by groups representing specific subsets of private schools. In December, the Association of Christian Schools International (ACSI), which has about 2,300 members in the United States, found that approximately half of 730 respondent schools had lost students from the previous year, a third grew, and about 15 percent were unchanged. The average change was a 3.2 percent decline.¹⁹ The National Catholic Educational Association (NCEA) announced a Catholic school enrollment drop of 6.4 percent.²⁰ For the previous five years, enrollment in Catholic schools had decreased steadily by about 2.5 percent per year, suggesting that COVID-19 had an independent, depressing effect on enrollment.²¹ Consistent with the closure list, the NCEA reported that losses tended to be focused more in working‐​class neighborhoods, while schools in wealthier areas often gained students as families fled closed public schools. Closures were also concentrated in the Northeast and Midwest. The NCEA reported that more than 200 Catholic schools had closed or consolidated in the 2019–20 school year, which would include part of the COVID-19 time frame. Finally, the Mid‐​South Independent School Business Officers (MISBO) found that about half of schools in a survey of member institutions saw enrollment increases between the 2019–20 and 2020–21 school year, about 22 percent were unchanged, and about 30 percent lost students.²² Responding schools tended to be heavily in the South and more expensive than the schools that typically closed.²³ MISBO did not collect data on the average enrollment change.

CEF’s survey and the NCEA found roughly 6 percent enrollment losses, and the ACSI found about 3 percent. Suppose a rough split—a 5 percent enrollment drop—were applied to school loss. Closures would amount to 1,625 schools, about 10 times what the CEF has recorded for COVID-19 and non‐​COVID‐​19 closures. While the tracker is doubtless missing some closures, it is likely that a 5 percent average enrollment loss does not seal the fate of 5 percent of schools. The percentage is likely to be smaller because closing a school is permanent and typically undesired, and people expected the pandemic to be temporary.

Bolstering the likelihood of lower‐​bound closures, the random sample of schools that the CEF contacted for its survey, based on the 2017–18 universe of private schools, yielded only four closures in 2020. Of those, only one had an explicit connection to COVID-19, one occurred before the pandemic, and two had unknown COVID-19 connections.²⁴ Overall, the CEF had 362 schools with confirmed statuses as of September 2020 that could have had COVID-19 effects.²⁵ Assuming the two closures with unknown causes and the one with a known pandemic connection were all connected to the pandemic, they would represent 0.83 percent of confirmable private schools in the CEF survey. This percentage would extrapolate to 269 private schools nationally. If only one closure was due to COVID-19, the extrapolated number would be 90 schools. This range suggests that the CEF count is in the ballpark of all closures.²⁶

Moving beyond overall numbers, the CEF data indicate that the private schools most likely to permanently have closed due to the COVID-19 pandemic have been Roman Catholic, low‐​cost, and—reflecting a long‐​standing trend²⁷—serving middle‐​income students demographically similar to public school enrollment. They also have tended to be schools that were struggling financially before the pandemic and to have faced greater enrollment losses under COVID-19 than other private schools. As the National Catholic Register illustrated in March 2021:

One reason many schools may have been able to absorb enrollment losses was the federal Paycheck Protection Program (PPP), created under the Coronavirus Aid, Relief, and Economic Security (CARES) Act, that passed in late March 2020 and was subsequently expanded. The PPP made about $659 billion in forgivable loans available to small businesses and nonprofits.²⁹ COVID Stimulus Watch estimated that about $4.5 billion of that went to private schools.³⁰ Additional PPP funding that went to churches also may have helped private schools.

The CARES Act also allowed private schools to receive “equitable services” from $13.5 billion in grants to states for elementary and secondary education, though that funding was disrupted because of a lengthy dispute with the Trump administration over how the money was to be shared with private schools. U.S. Secretary of Education Betsy DeVos issued a rule that services be allocated to private schools (school districts would control the funds) based on their share of total enrollment, while many state officials and members of Congress said the law required that funds be allocated commensurate with shares of low‐​income students. The latter eventually prevailed in court.³¹

Based on an estimate of the share of low‐​income students enrolled in private schools, those schools would be eligible for about $715 million in equitable services.³² The CARES Act also allotted $3 billion total for governors to use at their discretion on early‐​childhood, K–12, and higher education. Assuming that money was evenly split among those areas, private schools would have received funding proportionate to their total share of K–12 enrollment (about 10 percent), roughly $100 million.

The Coronavirus Response and Relief Supplemental Appropriations (CRRSA) Act, enacted in late December 2020, also had funds for private schools—including $2.75 billion specifically for use by non‐​public schools and prioritizing schools serving “low‐​income students” that were “most impacted by the emergency”—in addition to more PPP money.³³ However, the Congressional Budget Office estimated that the bulk of the act’s funds would be spent in 2022 and 2023, so little CRRSA Act help had likely hit private schools as of March 2021. That said, the prospect of eventually receiving some aid may have encouraged some endangered schools to stay open.³⁴

The American Rescue Plan Act (ARPA), signed into law on March 11, 2021, allocates another $2.75 billion for private schools, prioritized the same as was under the CRRSA Act.³⁵ That funding has had no direct effect on closures, but as with the CRRSA Act, the prospect of ARPA funding could have encouraged some schools to stay in business.

The CARES Act PPP likely provided important funding for preserving many private schools. It came in the very early part of the pandemic when schools, and many churches that support them, appeared highly vulnerable. A rough estimate (it is very difficult to pinpoint where every dollar has actually gone) suggests that the CARES Act made about $1,100 available per private school student and that about 87 percent of that was from the PPP.³⁶ That was significantly greater than what public schools, which were not eligible for the PPP, received under CARES: roughly $290 per student. In subsequent legislation, public schools received appreciably more per student—approximately $1,100 to $900 for private schools in the CRRSA Act and $2,500 to $500 for private schools under the ARPA—but that early funding may have been crucial to keeping many private schools afloat and able to do the next thing that likely kept closures down: offer in‐​person schooling for the 2020–21 school year.

At the beginning of the school year, reports started emerging of private schools providing in‐​person education while nearby public schools remained closed.³⁷ Soon after, stories appeared of private schools gaining enrollment as families, disaffected with online‐​only instruction in public schools, moved their children.³⁸ Such movement was likely insufficient to erase net drops but probably prevented even greater losses.

On the other side of the ledger, movement to homeschooling and charter schools—public schools typically run by private groups and free from many public schooling rules and regulations—likely hurt brick‐​and‐​mortar private schools. A Gallup survey conducted between July 20 and August 12, 2020, suggested that private schools would go from enrolling 11 percent of students to 8 percent; public schools would go from 83 to 76 percent; homeschooling would rise from 5 to 10 percent; and charter schooling from 2 to 5 percent.³⁹ An EdChoice survey conducted in mid‐​August predicted smaller losses for private schools—about a 2 percentage point drop—while homeschooling would rise about 3 percentage points and charter schools would go unchanged.⁴⁰ Finally, a U.S. Census Bureau “Household Pulse Survey” found that between April–May 2020 and September–October 2020 the percentage of households with school‐​age children reporting that they were homeschooling rose from 5.4 percent to 11.1 percent. Before COVID-19, about 3.3 percent of students were homeschooled.⁴¹

A potential problem common to all surveys about homeschooling is ensuring respondents clearly understand the difference between homeschooling—education at home not connected to a formal school—and instruction provided by a school that is received at home.⁴² Had private schools accounted for 11 percent of all enrollment and dropped to 8 percent, it would have meant a loss of approximately 1,560,000 students, or 27 percent of all private school enrollment.⁴³ No survey has indicated an enrollment loss of such magnitude, so it is likely that homeschooling estimates are overstating movement to true homeschooling.

Perhaps a major reason private schools appeared to weather the COVID-19 pandemic better than initially feared is that these typically smaller and more autonomous schools responded more efficiently and effectively than public schools. Two early‐​April 2020 surveys revealed that private schools responded to initial shutdowns with greater speed and depth of content than public schools. An American Enterprise Institute survey found that as of April 14—well into lockdowns—almost 20 percent of school districts offered no remote instruction, and only 54 percent had started remote instruction within two weeks of their buildings closing.⁴⁴ In contrast, an EdChoice survey of private schools between April 1 and April 17 found that almost 88 percent had “shifted to online learning with formal curricula [emphasis added].“⁴⁵ In July, the journal Education Next reported that the institutions attended by 88 percent of private school students had introduced new material during lockdowns compared with only 72 percent for public school students.⁴⁶

After lockdowns, private schools were much more likely to return to in‐​person schooling. Polling conducted in November and December 2020 found that 60 percent of private school students were learning in person full time versus 24 percent of traditional public school students. Likely reflecting private schools’ superior responsiveness, private school parents have typically reported fewer negative effects—academic, emotional, physical—on their children from the pandemic than traditional public school parents while expressing a greater degree of overall satisfaction with how their children’s schools responded to COVID-19. In the November–December Education Next poll, parents of 83 percent of private school students reported being either “somewhat” or “very much” satisfied with their educational experience versus 71 percent for traditional public schools.⁴⁷ A February 2021 EdChoice survey found that 48 percent of private school parents felt their children were progressing very well academically, 44 percent emotionally, and 50 percent socially versus 25 percent, 22 percent, and 22 percent, respectively, of public school parents. Some 62 percent of private school parents were “very satisfied” with their children’s school experience versus 37 percent of district public school parents.⁴⁸

In addition to benefiting from early PPP assistance and superior dexterity, private schooling likely took less of a closure hit than some feared because the economy performed better faster than many had predicted. The economy suffered under expansive lockdowns early in the pandemic, with gross domestic product dropping at an annualized rate of about 5 percent in the first quarter of 2020 and a staggering 33 percent in the second quarter. But it rebounded by 38 percent in the third quarter and 6 percent in the fourth.⁴⁹ Similarly, unemployment skyrocketed from below 3.5 percent in February 2020 to 14.7 percent in April, but by February 2021, it had returned to 6.2 percent.⁵⁰ This likely enabled many families to continue paying for private schooling or to start doing so. This may explain why the CEF cataloged 116 COVID‐​19‐​related closures in the months of April through August but only 19 after. It is also likely, though, that those months are when schools naturally assess their viability for the upcoming school year.

One final problem that might have constrained private school enrollment was lack of space, especially given student distancing guidelines, that may have prevented highly responsive private schools from enrolling many students who wanted to attend. MISBO reported, for instance, that if it counted schools that had “excess demand” among those that said their enrollment had increased, the share in that category would have risen from 48.4 percent to 57.2 percent.⁵¹ The NCEA reported that nearly 40 percent of Catholic schools had waiting lists.⁵²

The full effect of the COVID-19 pandemic, and the policies enacted in response, on private schools will not be known until the pandemic ceases to have a meaningful impact on daily life and until funding enacted to deal with those impacts has been exhausted. As of this writing, significant progress toward ending the pandemic is being made, including nationwide vaccine distribution. Biden administration chief medical adviser Anthony Fauci has suggested that life will return to “normal”—people en masse able to return to movie theaters, indoor dining, and other activities—by mid‐​fall or early winter 2021.⁵³ Exhaustion of federal COVID-19 relief will be much longer in coming, with the Congressional Budget Office estimating that the bulk of ARPA elementary and secondary funds will not be tapped into until 2022 and 2023 and not fully spent until 2028.⁵⁴

What policies should we look at to enable private schools to meet potential demand and, as a result, survive? The answer lies at two levels of government, with one needing to do less and the other more. Federalism and equal treatment of private schools lie at the heart of the best policies.

The federal government, through the PPP, likely helped to stabilize many private schools. That beneficial effect notwithstanding, the federal government constitutionally has no authority to run such programs: nowhere among the specific, enumerated powers given to Washington is there anything about emergency lending. That said, to the large extent that government forced businesses, schools, and churches to shut down, government assistance, though not necessarily federal, may have been justifiable.

Despite the benefit of early help, the trend for federal COVID-19 funding since the CARES Act has not been favorable to private schools from a competitive standpoint. The ARPA, especially, is slanted toward public schools. This is notably problematic because it exacerbates private schooling’s greatest problem: private schools must compete against public schools that are free to families because those schools receive huge, annual subsidies from local, state, and federal taxpayers. It is extremely difficult to compete against free.

The primary justification for more federal funding going to public than private schools is that the former serve higher shares of low‐​income students. However, COVID-19 relief money, while allocated according to shares of low‐​income students, can be used for goods and services that benefit all children in a school or district, including cleaning supplies, upgrading ventilation systems, and more. Public schools are likely receiving more public support for all students than private schools. Of course, COVID-19 affected everyone.

The ARPA passed, so the lopsided benefit to public schools cannot be stopped. What the federal government should do moving forward is nothing: enact no new relief for K–12 education. This would be consistent with the Constitution and avoid giving public schools an even more yawning financial advantage. It also makes fiscal sense: the Centers for Disease Control and Prevention estimated in December 2020 that only $442 per pupil would be necessary to bring all schools up to safe standards,⁵⁵ not the $3,900 estimated per student that public schools will receive under the three relief acts.⁵⁶ And since the economy has performed much better than once feared, state and local governments actually saw increased revenue in 2020.⁵⁷

Constitutionally, the level of government at which private schools should be helped is the state. Ordinarily, state and local governments are also the primary funders of K–12 education, with the federal government contributing only about 9 percent of funds.⁵⁸ The good news is that there has been a groundswell of legislative interest in increasing school choice—vouchers, scholarship tax credits, education savings accounts—at the state level, which would move states closer to a level financial playing field for private schools. As of March 15, 2021, 29 states had school choice bills in their legislatures, at least partially driven by dissatisfaction with public schools not providing in‐​person instruction during the pandemic.⁵⁹ Governors also control some money under the CARES Act, CRRSA Act, and ARPA and could look for ways for families to make decisions about how the money is used.

Since the onset of COVID-19, the United States has seen no fewer than 132 private schools announce that they would be closing permanently, at least partially because of the pandemic’s economic effects. The country has also seen perhaps a 5 percent drop in private school enrollment. In keeping with the nature of the economic downturn, which especially hurt lower‐​income people who could not easily telecommute, it appears that disproportionately low‐​cost private schools were most adversely affected. Closed schools also tended to be financially unstable before the COVID-19 pandemic.

Because COVID-19 has mainly shuttered already vulnerable institutions, a sustainable solution to the problem of closing private schools is not more one‐​off infusions of federal cash. It is for state and local governments to change how they fund education, moving away from putting dollars directly into public schools and allowing them to follow children to the educational options that families choose. Ultimately, what would be best for private schools and society would be a level financial playing field.

McCluskey, Neal. “Private Schooling after a Year of COVID-19: How the Private Sector Has Fared and How to Keep It Healthy,” Policy Analysis no. 914, Cato Institute, Washington, DC, April 13, 2021. https://​doi​.org/​1​0​.​3​6​0​0​9​/​P​A.914.