One of the most enduring critiques of school vouchers appears in Albert O. Hirschman’s modern classic, Exit, Voice, and Loyalty (1970). Responding to Milton Friedman’s economic case for educational competition, Hirschman argued that one of the potential pitfalls of a school voucher program is that it risks removing the most quality-conscious parents from an existing public school that is on the decline.

But what would happen to public schools if vouchers caused the most racist parents to leave? That’s not an entirely hypothetical question. It was relevant during the stormy period of the late 1950s and early 1960s in the American South, when white parents who wanted their children in segregated schools saw vouchers as a way out. Had they used these vouchers to get their kids out of public schools, then the public schools could have well have improved. Thus my term “the Inverse-Hirschman scenario.”

Hirschman’s Critique

One of Hirschman’s major contributions in Exit, Voice, and Loyalty is his concept of “voice.” Voice, in Hirschman’s lexicon, is a way to influence institutions without leaving them—that is, without “exit.” Voice means what the term suggests: speaking out, calling people to account, and actively participating.

Hirschman observed that quality-conscious parents serve an important role in improving the quality of public education through the tool of their voice. They participate in the PTA, vote in schoolboard elections, and regularly call their child’s teachers or principal to address their concerns about the quality of the education being offered. They pressure administrators to offer new and better programs for students. And if a school’s quality declines, they alert other parents, teachers, or school board officials to their concerns. Today, such parents spearhead fundraising drives for new computers.

One of the problems associated with vouchers is that these quality-conscious parents are also the most likely to leave if a private school offers their children something better. Their departure becomes especially problematic if a school is already on the decline. As Hirschman put it:

[I]f and when there is a deterioration in the quality of public school education these schools will lose the children of those highly quality-conscious parents who might otherwise have fought deterioration.1

In other words, if the quality-conscious parents leave on a voucher, a public school is left with largely apathetic parents who exert no similar pressures, fail to participate in the PTA, ignore school board meetings, etc. The voucher system triggers a cycle that degrades the important feedback mechanism once fulfilled by quality-conscious parents. As the active parents leave, the public school’s deterioration accelerates. Hirschman, therefore, predicted that vouchers would introduce a scenario with a compounding decline of voice and become a persistent disadvantage to the deteriorating public school.

Inverting Hirschman

Hirschman’s argument above pertains to vouchers under the normal circumstances of a minimally functional public school system. Let’s investigate another scenario with an added complication: racial segregation.

For more on racial discrimination, see Discrimination by Linda Gorman in the Concise Encyclopedia of Economics. See also the EconTalk podcast episode Michael Munger on Slavery and Racism; and “The Economics of Discrimination,”, by Robert P. Murphy, Library of Economics and Liberty, August 2, 2010.

For background on Albert O. Hirschman’s perspectives, see the EconTalk podcast episode Bent Flyvbjerg on Megaprojects. For educational choice from an international perspective, see “Education Vouchers,” by Pedro Schwartz, Library of Economics and Liberty, June 3, 2013.

In recent years, segregation has become a flashpoint of controversy in the national school choice debate.2 Several recent arguments from voucher opponents have sought to link the history of school choice to the segregationist resistance to Brown v. Board of Education. In this telling, early economist proponents of vouchers, such as Milton Friedman, James M. Buchanan, and G. Warren Nutter, either inadvertently or intentionally handed segregationists a powerful tool to keep black children out of white-majority schools.3 As these critics point out, several states and localities experimented with rudimentary tuition grant programs in the 1950s and 1960s that had the effect—and in some cases intent—of allowing segregationist parents to pull their children out of public schools in response to integration orders from the courts.

To bolster this criticism, many voucher opponents point to the notorious case of Prince Edward County, Virginia, which shuttered its public school system for five years, starting in 1959, to avoid integration. The replacement was not actually a voucher system—the county set up a “private” school for white children and subsidized their attendance while leaving black children with no education. There was no actual “school choice” in the matter because the county closed its public school, leaving only a private option. Federal court rulings against the county even expressly enjoined it from taking advantage of a state-run voucher system in operation at the time. Nonetheless, the dubious legacy of the segregation era has become an increasingly common argument against vouchers.

Interestingly, this particular line of argument can be evaluated economically in ways that actually draw upon the insights of Hirschman’s original criticism. To see how, let’s consider a second case that I call “the Inverse-Hirschman scenario.”

The Inverse-Hirschman Scenario

In this scenario, we start with a school system that is racially segregated but facing external pressures to integrate. Let’s assume that this school is in the 1950s South. Although there are many vocal parents, they do not use “voice” to improve the school’s quality but, rather, to maintain racial segregation. These vocal parents do not call upon the school board to provide new textbooks, band instruments, books for the library, or classroom supplies. They do attend PTA meetings, but to espouse white supremacy and other noxious racial ideas. They vote for school board candidates who pledge to resist desegregation by changing the boundaries of the district or by fighting the NAACP in court. In some instances, they even physically protest integration and attempt to block black parents from registering their students at the school. This is Hirschman’s voice with a twist.

Owing to their prejudices, these parents cannot countenance having a single black student in the same classroom as their child. And if the option to exit for a whites-only private “segregation academy” becomes available, these racist parents will be the first ones to leave.

The harmful effects of this exercise of segregationist “voice” are many. As a result of their racial and political preferences, scarce school district resources are diverted away from the classroom and towards hiring attorneys to fight integration. The segregationist parents’ exercise of voice also harms other students—both inside and outside of the public school. First, they are seeking to impose segregation upon all other students in the same school, including those whose parents do not share their racial prejudices. Second, and more significantly, they are harming black students who cannot attend the school due to segregation. In the worst scenarios, they are even subjecting black students to harassment and discrimination. In short, these voice-active parents are an unambiguous “bad” for their child’s school.

In our Inverse-Hirschman scenario, something interesting happens as a result of the existence of segregation. Voice is no longer a mechanism for improving the quality of a public school education. Instead, it becomes the very source of that school system’s greatest ills. Holding all else equal, the departure of a segregationist family from a public school actually becomes a net benefit for everyone else currently in that school who doesn’t share their segregationist beliefs. It also benefits the black students who previously couldn’t attend the school or were harassed and made to feel unwelcome.

Vouchers and Segregation

I call attention to the Inverse-Hirschman scenario’s twist for a specific historical reason. It basically describes another historical situation—this one in Charlottesville, Virginia in the spring of 1959, amidst ongoing court orders to desegregate.

The Charlottesville school system’s attorney at the time was John S. Battle, Jr.—a committed segregationist politician and son of a former Virginia governor with close ties to the political machine of Senator Harry Flood Byrd, Sr. Virginia’s legislature was debating the adoption of a tuition grant program that had some characteristics of a rudimentary voucher system. Although ostensibly race-neutral, these tuition grants could be used to send white children to “segregation academies” in place of the public school. This proposal is what modern voucher opponents use to link vouchers and segregation.

Not all segregationists supported the tuition grants, though, and one of their vocal opponents was none other than John S. Battle, Jr. Speaking at a Charlottesville PTA meeting on March 23, 1959, Battle denounced the program and implored white parents to exercise solidarity by registering their children in the public schools. His rationale was simple. By keeping segregationist parents around as a social buffer against integration, the white-controlled public school system could limit the admission of black students to only a few families, despite the court orders to integrate. They would do so by keeping white-majority schools at constant full capacity, leaving no space for more black students. Battle explained to the segregation-supporting parents at the PTA meeting:

No Negro child is going to force any white child out of his desk at Venable [Elementary School]. But for every white child who vacates his desk there is a great risk that a Negro child will occupy it.4

“In short, defeating the use of tuition grants became a key plank in Battle’s strategy for preserving segregation.”

Battle ominously referred to this process as “Negro Engulfment,” appealing to the racist instincts of the most hardline segregationists in attendance. His resistance strategy proposed two additional features, both of which depended on retaining the same vocal segregationist parents. First, the public schools would draw geographical districts, or zones, that intentionally restricted enrollment at certain schools to majority-white neighborhoods. Second, they would adopt enrollment restrictions that capped elementary schools at 30 students per classroom and high schools at 23 students. So long as the white parents remained, Battle insisted, they could “prevent mass integration” and keep black student enrollments at white-majority schools “to a bare minimum.” They would accept a small number of black students who could not be excluded from the geographic zones. This would allow them to comply with Brown, or so Battle argued. In practice, it would maintain de facto segregation through other tools—and through the segregationist buffer that his PTA audience provided. In short, defeating the use of tuition grants became a key plank in Battle’s strategy for preserving segregation.

We can take this exercise one step further and ask what would have happened next if the segregationist parents had departed from the public schools in spite of Battle’s anti-voucher strategy. These parents would almost certainly have sought out private, all-white “segregation academies” for their children. At the same time, the public schools would have lost the most vocal proponents of the former segregationist voice buffer and would have begun to integrate. Although the social harm caused by the segregationist parents would have ended, private segregation would still have persisted in these alternative institutions—and, in fact, it did for another decade or more in the Virginia desegregation fight.

As economists, we may ask another question, though: would such institutions have persisted over time? Friedman’s arguments, also drawing upon the 1955 dissertation of his former student Gary S. Becker, suggest that we might expect them to have diminished in number as well. The reason stems from the costliness of maintaining racial prejudices when these prejudices are not propped up by the formal segregation laws of the state.

By setting itself up as an all-white “segregation academy” catering only to racist parents, a private school of this type forgoes a sizable segment of potential students, including people who are put off by racists and anyone who is not white. This voluntary forfeiture of a segment of its student base is not an immediate death sentence for the “segregation academy,” but it will almost certainly take an efficiency toll over time compared to non-segregated schools—provided that competition exists. And that will be particularly true if social patterns and pressures are moving against racism, or if other external policies (including at the federal level) can be used to deny public tuition grant eligibility to institutions that racially discriminate. Friedman, therefore, likely expected both the stature and the resources of private, segregated institutions to diminish, relative to other institutions, over time.

Looking back at 1959 with the benefit of history and hindsight, we might fully admit that this expectation oversimplifies the problem. Racist beliefs were deeply engrained, and the legal institutions supporting them often extended well beyond the legal separation of white and black students into different schools. Prince Edward County stands out as a chilling example of extreme and pervasive racism, even if it was not a true school choice system. A Supreme Court decision put an end to the county’s public school closure in 1964, but only after a substantial toll on the affected children. Thus the Inverse-Hirschman scenario described above is predicated on there being both public and private schools, but, at the same time, so is the pressure of competition that threatened to undermine Battle’s scheme for preserving segregation.

What we see in the history of vouchers is that the posited link to segregation, so carelessly repeated as “fact” by voucher opponents, is not nearly so simple. Segregation also has profound political economy implications that voucher opponents have largely ignored—namely the Inverse-Hirschman scenario that I have described.

As Battle’s speech in Charlottesville illustrates, some of the most die-hard segregationist politicians of the era perceived school choice—and the exit associated with it—as a threat that would undermine their entire social order. Friedman and the economists who advocated school choice almost certainly recognized this outcome, but from the other side of the question as Battle. The Inverse-Hirschman scenario shows why this is so: an exit of die-hard segregationist parents is itself a way to alleviate the harm that they otherwise perpetuate in the public school system with their voices and pressures on school administrators. And that is why an economist in the Brown era might legitimately have believed that vouchers would lead to a substantial decline in segregation over time.


Albert O. Hirschman. Exit, Voice, and Loyalty: Responses to decline in firms, organizations, and states. Harvard University Press, 1970, p. 51.

For recent examples that make this argument see: Chris Ford, Stephenie Johnson, and Lisette Partelow, “The Racist Origins of Private School Vouchers,” July 12, 2017, Center for American Progress; Nancy MacLean, Democracy in Chains. Viking Books, 2017.

Economic arguments for vouchers in this era include: Milton Friedman. “The role of government in education.” In R. Solo (ed.), Economics and the Public Interest, Rutgers University Press, New Brunswick, 1955; G. Warren Nutter and James M. Buchanan, “The Economics of Universal Education,” Thomas Jefferson Center for Political Economy, University of Virginia, February 10, 1959.

“Battle cautions against move to private schools.” Charlottesville Daily Progress, March 24, 1959. For elaborations of this same segregationist position, see: John S. Battle, Jr. “Scholarship Program May Result in Even More Integration,” Lynchburg News, April 22, 1959; “Private schools are linked to ‘Engulfment’ by Negroes,” Richmond News-Leader, March 24, 1959.


*Phillip W. Magness is a Visiting Assistant Professor of Economics at Berry College. He researches the economic history of the United States, with a dual emphasis upon the political economy of slavery and the history of taxation. He holds a PhD and MPP from George Mason University.