Friedman or Jencks? Early childhood education vouchers in practice

: Market-based educational reforms, including vouchers, have been a highly contested topic in education policymaking over the last three decades. Vouchers have been used in practice to promote educational choice and student success by facilitating "free" market competition and eliminating the monopolistic privileges of public schools. Nonetheless, market-based reforms, particularly education vouchers, are not commonly used in early childhood education (ECE). This article examines the arguments for and against market-based education reforms and education vouchers and provides a brief overview of five voucher programs that aimed to improve ECE delivery. Comparisons of these programs demonstrate that neither Friedman's liberal market nor Jencks' social policy approaches are definitively superior to one another. The effectiveness of an ECE voucher program depends on the specifics of the policy text and the context in which it is introduced.

Direct governmental intervention in education has often been considered expensive and ineffective in tackling the problem of academic underachievement of students and producing a competitive workforce in the face of rapid globalization. As a result, since the 1980s, governments in the West have turned to market strategies to reform their education systems (Gois, 2010;Taylor, Rizvi, Lingard, & Henry, 1997). By devolving their responsibility to parents, these governments have relied on the aggregate of parental choices to provide discipline, accountability, and demand, which schools cannot escape. This shift to market-based reforms has been driven by a belief that competition among schools will lead to greater efficiency, innovation, and improvement in educational outcomes. Despite much theoretical and practical work on market reforms in school education, little has addressed the multi-sectorial, voluntary, and devolved nature of early childhood education (ECE). Meanwhile, education vouchers have received considerable interest over the years as a manifestation of the neoliberal doctrine. However, in examining the studies put forth by advocates, little consistency or consensus appears to emerge (Lubienski & Brewer, 2016;Usher & Kober, 2011).
Public vouchers explicitly designed for ECE are particularly rare, even though education and care services received in early childhood (generally defined as the years between 0 and 8) have been identified as crucial factors in determining children's future success (Rao & Wong, 2018). In many countries, preschool attendance is not mandatory, and governments do not feel obliged to provide free ECE. Instead, services for young children are often provided by charitable and religious organizations (Daniels & Trebilcock, 2005). It was not until the last two decades that developed economies acknowledged the importance of early investment in human capital and considered utilizing market forces to enhance ECE delivery.
Given the limited empirical evidence, this article takes a closer look at ECE vouchers, specifically focusing on the debate between Milton Friedman's liberal market approach and Christopher Jencks' social policy approach. By analyzing and comparing several ECE vouchers, this paper aims to shed light on the effectiveness of such vouchers and to determine if one approach is definitively superior to the other. Ultimately, this paper argues that the efficacy of an ECE voucher relies on the details of the policy text and the context in which it might be introduced. But first, it begins by presenting a brief overview of marketbased education reforms and education vouchers.

Market-based education reforms and education vouchers
Market-based education reforms have been enacted based on the neoliberal assumption that the state's older bureaucratic structures and practices are inefficient, expensive, and unable to respond quickly to rapid societal changes. At the same time, commercial activities in the market have been viewed as the most efficient methods for producing and supplying goods and services (Gois, 2010;Taylor et al., 1997). Bowe, Ball, and Gold (1992) assert that the market is composed of three essential elements: (a) the individual choices of consumers; (b) the reactive and proactive strategies of producers; and (c) the constant adjustment of producers' products and services, resulting in equilibrium. In theory, the market is a mechanism that produces its own order (Gewirtz, Ball, & Bowe, 1995). In the education market, it is assumed that parents (the consumers) will choose schools that will provide the maximum advantage to their children. At the same time, as producers/providers, principals and teachers must make decisions and provide services to ensure that their schools prosper or, at least, survive in the marketplace. The combination of selfmanagement of schools and market forces is generally assumed by neoliberals to improve the efficiency of service delivery and the quality of education (Bowe et al., 1992;Gois, 2010;Taylor et al., 1997).
Nonetheless, the vast difference in tuition fees between public and private schools has deterred the development of such market forces. Education vouchers are thus premised on giving parents the freedom to choose the school for their children, by providing grants, subsidies, or reimbursements to cover private school costs. They have been theorized to enhance parental choice and permit competition among schools to develop, leading to more cost-efficient educational outcomes (Daniels & Trebilcock, 2005;Friedman, 1962;Lubienski & Brewer, 2016). However, critics argue about the extent to which this choice is truly free. For example, the limited availability of schools that accept vouchers in certain areas can restrict parents' options, particularly those in lowincome areas who may not have access to a wide range of private schools (Jongbloed & Koelman, 2000;Witte, 2000). Moreover, some critics point out that education vouchers may lead to a further stratification of education, as families with more resources may be better able to take advantage of the voucher system. In addition to the above debate, critics worry that education vouchers may lead to a more fragmented education system, impacting service efficiencies. They draw attention to the possibility that schools may become more inclined to invest in marketing and advertising to attract students rather than improve their education quality. Some also maintain that segregation may occur as schools may "cream skim" students from better financial backgrounds and with a higher academic ability (Jongbloed & Koelman, 2000). In other words, these critics contend that education vouchers would not necessarily improve parental choices or the quality of education but would almost certainly affect the equality and stability of education provision (Carswell, 2007;Loomis, Rodriguez, Honeycutt, & Arellano, 2006;Sutton & King, 2011;Usher & Kober, 2011). Even so, proponents argue that the market is a self-regulating mechanism formed by the aggregate of parental choices, and it is not necessarily fair and equal. In effect, as noted by Bowe et al. (1992), if it did generate equality, "it would remove incentive and competitive drive"(p.25).

Comparing Friedman's and Jencks'education voucher approaches
Milton Friedman and Christopher Jencks are two prominent scholars who have proposed different approaches to implementing education vouchers (Abrams, 2019;Lee & Wong, 2002). While both approaches aim to increase competition and choice, their key differences highlight the tensions between market-based efficiency and social equality.
Friedman's liberal market approach, which is centered on individual choice and competition, with an emphasis on market-driven efficiency, is closely aligned with the neoliberal ideology behind education reforms in the USA and the UK over the past 30 years. However, it originated in the 1950s in the USA, where public schools were observed to be segregated, and the quality of education varied depending on the location of the schools. According to Friedman (1962), public schooling was a monopoly that increased societal stratification and failed to provide equal educational opportunity. The primary goal of introducing education vouchers was to break this monopoly and give parents the freedom to choose the best educational options for their children, whether public or private (Usher & Kober, 2011). This approach assumes that competition between schools will ultimately improve education quality and efficiency for all students (Lee & Wong, 2002). Friedman's (1962) approach has four defining characteristics. First, the vouchers should be universal and available to all parents. Second, the value of the voucher should be enough to cover the average cost of a place in a state school or a proportion of that cost. Third, "topping up" is allowed, meaning schools may charge fees higher than the voucher value, and parents could pay the remaining portion of the tuition using their own money. Fourth, parents and schools are unconstrained: parents can spend the voucher at any public or private school of their choice, and schools have complete freedom in choosing students and prioritizing their waiting lists.
In contrast, Jencks' (1970) social policy approach prioritizes equalizing educational opportunities and social mobility. While it still embraces market competition, it does so intending to create a more equitable educational landscape. Jencks' approach is more focused on ensuring that disadvantaged students have access to quality education by providing them with additional resources and limiting schools' freedom to select students based on their ability to pay additional fees (Jongbloed & Koelman, 2000).
Jencks ' (1970) approach also has four main characteristics. First, it favors market competition among service providers but with appropriate regulation to prevent the negative consequences of a completely free education market. For instance, only schools that meet specific standards or adopt certain curricula can redeem vouchers. Second, the basic vouchers should cover the total average cost of state education. Third, disadvantaged parents receive a complementary increment to the basic voucher, and topping up is not allowed. This additional funding diverts resources to schools with disproportionate numbers of children from deprived backgrounds. Schools accepting children from better-off families receive less income than schools with predominantly low-income students. On the contrary, the fact that disadvantaged students receive a complementary increment on top of the basic voucher would motivate schools to admit such students (Abrams, 2019;Jongbloed & Koelman, 2000). Fourth, schools have limited freedom in selecting students and must accept students if vacancies exist. If the demand for school places exceeds the supply, they must randomly allocate at least half of the available places.
The debate between these two models revolves around balancing individual choice and social equity. Advocates of Friedman's model argue that a free market approach, with minimal government intervention, will lead to the most efficient allocation of resources and the highest overall quality of education. They also claim that individuals are the best judges of what is suitable for their children and should be free to make those choices without interference. On the other hand, supporters of Jencks' model argue that a purely market-based approach can exacerbate existing inequalities and perpetuate social stratification. They contend that some level of government intervention is necessary to ensure equal access to quality education for all students, regardless of their socioeconomic background. These proponents believe that education is a public good and that society as a whole benefits when all children have the opportunity to reach their full potential (Sparkes & West, 1998). In other words, Friedman's less-regulated approach may encourage innovation but risks exacerbating inequities, while Jencks' approach attempts to address inequality directly but may hinder market dynamics.

Early childhood education vouchers in practices
Friedman's and Jencks' approaches to education vouchers reflect contrasting views on the role of markets and government intervention in education. However, it is essential to note that real-world voucher programs often do not fully embody the key characteristics of either approach. Nevertheless, the underlying ideological perspectives can be discerned by examining their objectives, particularly with respect to promoting social equality, and the degree of market and government involvement in their implementation.
To illustrate the varied outcomes of implementing voucher programs in different contexts, below presents the experiences of five real-world ECE voucher or voucherlike programs. While these programs are not the only public vouchers for ECE, they are chosen because they demonstrate how applying the same voucher concept can lead to different outcomes in different contexts, ranging from the East to the West.

The United Kingdom
Before introducing the Nursery Voucher Programme in 1996, English's preschool sector was never part of any coherent and sustained government policy. In the 1990s, preschool education became a prominent issue in policy debates due to recent social and economic changes, including the rising participation of mothers with young children in the workforce, the growth of two-parent working families, and an increase in lone-parent households (Sparkes & West, 1998). In 1994, former Conservative Prime Minister John Major pledged to offer a preschool place to all fouryear-olds whose parents desired it. The Nursery Education and Grant Maintained Schools Act of 1996 subsequently introduced the education voucher program for fouryear-olds, emphasizing the importance of high-quality publicly funded provision that encouraged diversity and parental choice. The voucher, valued at £1,100 for three terms before the compulsory school age of five, operated essentially following Friedman's approach (Lee & Wong, 2002; Table 1). It could be redeemed in any independent, voluntary, or state sector institutions providing nursery education; in a reception class in any state primary school; in any playgroup; or a combination of them (Lee & Wong, 2002).
However, since its implementation, the program had been heavily criticized, particularly for the universal flat rate design and the low value of the voucher. Critics, such as Coulter (1995), argued that the voucher value was often insufficient to cover the cost of a private nursery place, creating inequality of choice among parents, as only those from better-off families could afford to top up the vouchers and send their children to private nursery schools. The program's original intention of a market-driven system in removing the direct link between parental choice and school funding was also compromised by allowing statemaintained providers to receive advance funding. Sparkes and West (1998) evaluated the voucher program and concluded that it failed to meet its three key objectives. The program was formally abolished in 1999 (Lee & Wong, 2002).

Taiwan
Unlike many countries, public kindergartens and nurseries were preferred over private ones in Taiwan due to their abundant resources, which enabled them to offer higherquality education at affordable prices (Li, Wong, & Wang, 2008;Wu, 2002). In Taipei, for example, though only 1/9 of all the early childhood settings were publicly-run, the amount of subsidies they obtained from the government was 14 times as much as that received by their private peers (Li et al., 2008). However, the limited number of public preschools meant that many parents had to pay additional fees to send their children to private preschools if they failed to obtain a place at public preschools. Moreover, as many as 24% of the kindergartens were not legally registered, raising concerns about their quality (Xu, 2001).
In 1998, the Taiwanese government launched the education voucher program for ECE in Taipei and Kao Hsiung, extended throughout Taiwan in 2000. Even though the Taiwanese voucher program aimed to enhance quality, equality, and diversity in education provision and reduce the financial burden on parents (Chen & Li, 2017), its design mimicked Friedman's more than Jenck's approach (Table  1). Every child aged 5 or above was eligible for a voucher worth NT$10,000 per year, but the voucher could only be used to reimburse parts of school fees at registered private kindergartens or nurseries. However, studies suggest that the low value of the vouchers only reduced parents' tuition expenses by 20% and did not change their preference for public schools (Wu, 2002). After a year of full provincial implementation, in 2001, 74% of parents surveyed did not know about the existence of the voucher program or did not bother to apply for the vouchers (Xu, 2001). Khoo and Wu (2007) found that the implementation of the voucher program increased the workload of school administrators and had a limited impact on the quality of ECE in Taiwan. The most noticeable impact of the voucher program, as found by Chiu and Wu (2003), was encouraging unregistered early childhood settings to become legally registered.

Georgia, USA
Georgia's education system has historically ranked among the lowest in the USA on measures of academic achievement due to underfunding. In 1990, for example, Georgia high school students had the second-lowest Scholastic Aptitude Test (SAT) scores in the nation, and over 30% of Georgia's adults did not graduate from high school (Raden, 1999). In response, the former governor of Georgia, Zell Miller, proposed a comprehensive, family-oriented, state-wide, and state lottery-funded Pre-Kindergarten (Pre-K) program in 1989. Based on market competition, Georgia's Pre-K functions like a regular voucher program. Parents may choose among public and private service providers in any district, and the state pays providers directly, covering the entire cost of tuition and fees (Levin & Schwartz, 2007). Adopting Jenck's social policy approach, initially, the program only served low-income four-year-olds when it began full implementation in 1993. However, thanks to the exceptionally profitable Georgia Lottery, the program expanded in 1995 to provide free Pre-K education to all four-year-olds, making Georgia the first state in the USA to offer a universal, voluntary Pre-K program (Raden, 1999;The Southern Education Foundation, 2008). Individual service providers may determine their own enrolment policies. To prevent cream-skimming of children from wealthier families and ensure equal opportunities for lowincome and disadvantaged children, no children can be designated as "low-income" or "disadvantaged" at the time of registration. However, once admitted, low-income and disadvantaged children receive additional support, such as transportation reimbursement and food vouchers (Georgia Department of Early Care and Learning, 2011; Table 1).
Research by Levin & Schwartz (2007), Raden (1999), Soliday Hong et al. (2021), and The Southern Education Foundation (2008) highlights the success of Georgia's Pre-K program in enhancing young children's school readiness, increasing parental choice of schools, and effectively serving children from families living in poverty. The program has become a model for state-run universal early education.
Nonetheless, in recent years, the program has struggled to keep up with inflation and the increase in school tuition costs. Despite Georgia increasing its budget to raise teachers' salaries in 2016, it has been unable to fulfill its promise of providing a place for every four-year-old child (Roelcke, 2018).

Cleveland, Ohio, USA
The Cleveland Scholarship and Tutoring Program (CSTP) was enacted in 1995, making it the first education voucher program in the USA to allow the participation of religious schools, following the Supreme Court's ruling in the case of Zelman v. Simmons-Harris (Marks, 2002). The program aims to provide educational opportunities for students from low-income families residing within the Cleveland school district to attend private schools in Cleveland or public To provide quality education for the next generation  To ease the financial burden of parents Eligibility All 4-year-olds All children aged 5 or above All 4-year-olds  All students in kindergartens through 8th grade  Priority will be given to low-income students  Government expenditure on ECE increased by 217.6%  90% of teachers were trained with a certificate level or above  All voucher kindergartens were quality assured Negative:  There were increasing amounts of kindergartens reaching their tuition ceilings, declining participation resulting from a variety of deficits, and a high teacher turnover rate The PEVS was replaced with a new free kindergarten policy in 2017   Table 1. Summary of the five real-world ECE voucher experiences schools in adjacent districts. The program follows the voucher approach outlined by Jencks (Lee & Wong, 2002; Table 1) and is open to all students enrolled in kindergarten through twelfth grade in a public or private school within the Cleveland Municipal School District. However, priority is given to students whose family incomes are less than 200% of federal poverty guidelines, and scholarships are awarded to kindergarten students first, followed by students in first through eighth grade, and then high school students (Plucker, Muller, Hansen, Ravert, & Makel, 2006). As of 2022-23, the maximum voucher value is US $5,500 for students in grades K-8 and US $7,500 for high school students (Ohio Department of Education, 2022). Parents can redeem the voucher at all private schools in Cleveland and public schools in adjacent districts, provided they meet specific minimum standards prescribed by the state.
Schools must select students by lottery if the number of applicants exceeds the available voucher spaces (United States General Accounting Office, 2001).
The results from different studies evaluating the CSTP have been inconsistent (United States General Accounting Office, 2001). Greene, Howell, and Peterson's (1997) study found that the voucher program effectively increased the choices of low-income families, and participating schools were successful in retaining students in the program.
Voucher-receiving students also demonstrated significant gains in mathematics and reading achievement. However, Pathak, Holmes, Mincberg, and Neas (2001) found that many students receiving vouchers were not from lowincome families and had already been attending private schools before their application. Furthermore, there was no credible evidence demonstrating that vouchers offer an academic advantage over public schools, leading them to conclude that the voucher program failed to fulfill its objective. Plucker et al. (2006) agreed with Pathak et al.'s findings that only a small percentage of public school students used the CSTP to enroll in private schools. Additionally, Plucker et al. found that scholarship recipients who attended private schools were less likely to be racial-ethnic minorities than their public school peers. When controlling for minority status, student mobility, and prior achievement, Plucker et al. found no significant difference in overall achievement scores between scholarship recipients and students in the two public school comparison groups by the end of sixth grade. However, significant differences (p <0.05) were observed in three subject areas: language, science, and social studies.

Hong Kong
The Pre-primary Education Voucher Scheme (PEVS) of Hong Kong, executed between 2007 and2017, was a controversial program considered a "hybrid" of Friedman's and Jencks' models (Wong & Rao, 2015). It was implemented in a private education market wherein "winning at the starting line" was commonly heard, and parents sought a curriculum that focused on academic components from a young age to prepare their children for primary school (Cheung, 2009;Pearson & Rao, 2006). Although ECE was not free or compulsory in Hong Kong, almost all children over three years attended kindergarten programs (Wong & Rao, 2015;2022). However, with limited financial support from the government, kindergartens struggled to operate with limited resources and meet the academic-oriented and teacher-centered pedagogy needs of parents, which often conflicted with the core values the government promoted (Chan & Chan, 2002).
In 2007, the government finally committed to investing HK $2 billion per annum to implement the PEVS and a series of new initiatives to provide "quality education for our next generation" and "[ease] the financial burden of parents" (Education and Manpower Bureau, 2006, pp. 15-16; Table 1). The PEVS worked similarly to other voucher programs based on liberal market principles, empowering parents to exert market power. However, for kindergartens, it required them to charge below the tuition ceiling set by the government, follow the government's curriculum framework, and meet professional qualification standards for teachers and principals, resembling a social policy approach. Despite the usual agenda of diminishing authority through decentralization and the endorsement of market forces, the PEVS pushed the once-free ECE market towards increasing government interventions. It restricted eligibility to a specific subsection of the private sector and explicitly tied service quality to benchmarks set by the government with the mandatory quality assurance mechanism (Wong & Rao, 2015;2022). Over ten years, government expenditure on ECE increased by 217.6%, while the overall education expenditure increased by 45.8% only (Census and Statistics Department 2016). Overall, 90% of teachers were trained with a certificate level (A certificate level is equivalent to two-year fulltime study or the first two years of a full-time Bachelor's program) or above, and all voucher kindergartens were quality assured, improving the general accountability of kindergartens (Audit Commissions, 2013).
Despite these improvements, many criticized the PEVS for curtailing parents' choices, and it was ultimately replaced with a new free kindergarten policy in 2017. The decision was made due to increasing amounts of kindergartens reaching their tuition ceilings, declining participation resulting from a variety of deficits, and a high teacher turnover rate (Wong & Rao, 2022).

Discussion
Although all five voucher programs evolved from the same concept that the extension of the market system to education would improve parental choice and quality of education provision, the experiences above suggest that there is, in fact, no conclusive evidence to justify these claims. For example, the Nursery Voucher Programme in England showed limited success in improving education quality, with concerns raised about the impact of marketization on educational equity. Similarly, Taiwan's voucher program did not produce the intended effects of enhancing parental choice and improving the quality of ECE. On the other hand, Cleveland's and Georgia's voucher programs showed some positive results. Still, meaningful gaps have emerged in Georgia's Pre-K design and quality in recent years. Hong Kong's hybrid voucher program succeeded in improving education quality due to its focus on quality and accountability measures. On the other hand, it faced criticisms due to restricting parents' choices and increasing government intervention in the ECE market.
The analysis also reveals that the effectiveness of voucher programs is dependent on policy details and context rather than their underlying theoretical approaches. For instance, the implementation of England's voucher program was complicated by the fragmentation of the education system and the lack of regulation of private providers. The low voucher values of both England's and Taiwan's voucher programs seemed to lead to their demise, and Georgia's Pre-K appears to go downhill due to financial issues. The debate between the liberal market and social policy approaches in the implementation of voucher programs in ECE has not yielded a clear winner, prompting a rethinking of the nature and applicability of vouchers in ECE. The Hong Kong voucher experience, in particular, challenges the viability of traditional forms of vouchers, such as those proposed by Friedman and Jencks, as an effective policy tool. It also highlights the need for a new form of education vouchers that can be implemented in a free market while being responsive to governments' short-and long-term political agendas. These discussions underscore the importance of theorizing and examining educational policies in context, as a range of contextual factors, including political, social, and economic factors, influence the success of voucher programs. Policymakers need to be cautious when introducing ECE voucher programs and consider the policy details and context in which they are implemented to maximize their potential benefits. They also need to consider alternative education reforms that can achieve the goals of voucher programs while minimizing their limitations. For instance, increasing funding for ECE programs and improving the training and development of ECE providers may be more effective strategies for improving educational outcomes in early childhood.

Conclusion
Much theoretical and practical work has taken place regarding reforms in primary, secondary, and tertiary education, but little has addressed the multi-sectorial, voluntary, and devolved nature of ECE. This paper has reviewed and compared five ECE vouchers, including England's Nursery Voucher Programme, Taiwan, Georgia's Pre-Kindergarten, Cleveland Scholarship and Tutoring Program, and Hong Kong's Pre-primary Education Voucher Scheme. It provides good references for the theoretical discussion of how to go about reforming and financing early childhood services in an era of urgent need, sometimes constrained resources, and highly individualized policy contexts and regimes globally.
Ultimately, the success of voucher programs will depend on careful design and implementation that considers the education system's specific needs and circumstances and the broader political and social context in which it operates. Future research should continue to explore the effectiveness and limitations of voucher programs and alternative approaches to ECE reform. The debate between Friedman and Jencks is ongoing, but it is clear that there is no one-size-fits-all solution to education reform.