Private schools in a global world.

Globalization is a popular term that, to economists, describes the effects of diminished costs to trade around the world. The technological revolution of the past two decades lowered barriers to trade in the financial sector as well as in the goods and services sectors. As a result, we routinely consume products from China, India, or Latin America, and these and other countries likewise consume goods and services from us. Many argue that a by-product of the technological revolution is the loss of country and regional cultural identity. We find McDonald's and Starbucks in Seattle, Washington, Beijing, China, and small towns around the globe. Others argue that this simply implies wider choice of goods and of cultural identities.

As economists think about globalization, we attribute these changes to efficiency in the marketplace. The technological advances have enabled countries to export those goods and services for which they have a comparative advantage, and the gains from trade have resulted in increased economic growth worldwide. While we see the benefits from the enhanced flow of goods and services across markets, economists have paid less attention to the nonmarket aspects of globalization. The increased flow of information and people across national boundaries implies increased competition across all sectors of the economy. Similar to the market for goods and services, the political institutions that survive the competitive race in the long run presumably will be those that enhance efficiency. This is not to intimate that political systems will replicate the efficiency of the marketplace but to suggest that competitive forces tend to lead to more efficient structures and, ultimately, toward more efficient outcomes. As these public administrative institutions become more competitive, politics should play a diminished role and economic efficiency should play an even greater role in future structures.

If globalization ultimately leads to more efficient institutions, the value of examining current international differences in institutional structures is immense. Through understanding past and current differences, we gain insight into our future. Today I want to examine one specific institution within this context--private provision of elementary and secondary schooling. I want to examine private schooling by exploring the institutions that govern the provision of private schooling, the role of politics and economics in the evolution of these institutions, and, finally, the efficiency implications of the various institutional arrangements. At each step in this examination, I shall be assessing the role that globalization has played, and is expected to play, in the delivery of schooling around the world.

1. Private Provision around the World

Economists have studied education since the days of Adam Smith. We have estimated the demand for schooling, provided normative assessments of the role of the state in education, examined the effects of public expenditures for education on housing markets, estimated the returns to schooling, and estimated the relationship between public schooling and economic development. But most of all, economists have estimated education production functions that relate public sector inputs such as per pupil expenditures, pupil--teacher ratios, or teacher salaries to educational achievement. And, as last year's Southern Economic Association Distinguished Guest Lecturer so eloquently demonstrated, economists, policymakers, and the public have been sorely perplexed by the lack of positive association between these inputs and the achievement of public school students (Hoxby 2004).

The private side of schooling has received increasing attention from economists as data have become more available (James 1993). Private schooling occupies a significantly smaller sector of the economy in the United States than does public schooling. According to UNESCO (2000) data, approximately 12% of students enrolled in primary schools in this country attend private schools. At the secondary level, the numbers are slightly less at approximately 10%.

The 12% of students in private primary schools in the U.S. schools places the United States at 55th in the world with respect to reliance on the private sector for provision of schooling. For perspective, let us examine Table 1. which lists the top 20 countries in terms of proportion of enrolled students in private primary schools. The small Asian country of Macau (China) ranks number 1 in the world with 94% of its students enrolled in private schools. Zimbabwe (Africa) ranks second at 88% of students in private schools at the primary level. As we move down the list, we see Caribbean countries, European countries, and Middle Eastern countries. The 20th-ranked country, Spain, enrolls more than 33% of its students in private primary schools.

A comparison of enrollment at the secondary level (see Table 2) provides only a slightly different image. First, the United States falls to number 72 in the world at the secondary level. The Caribbean and European countries remain on the list, but there is significant shuffling of the rankings between the two levels. Some countries fall out of this ranking because of lack of provision of secondary schooling. Others such as Zimbabwe fall from number 2 to number 21.

Before considering these numbers further, I want to clarify some definitional issues. In the United States, we think of private schools as those that are governed and managed by an institution or organization outside the public sector and for which attendance fees are required to cover the costs of operating the school. But I shall illustrate this afternoon that this definition does not extend to much of the rest of the world. Private schools are virtually always defined as those for which a private sector authority holds some degree of governance responsibility and rights and has managerial authority over the schools. The public sector, however, may exert significant degrees of regulatory authority over the "private" schools. Furthermore, financing of privately governed schools may be entirely from public funds. As we progress, I shall return often to these definitional issues.

The relatively small numbers for private schooling in the United States have not always described our situation. From the pre-Revolutionary War period into the mid-1800s, education in this country was largely privately provided and privately financed. For the most part, families paid fees to tutors to visit their homes or, in more urban areas, paid fees to teachers for instruction in a school setting. Churches and other charitable organizations sponsored the education of children whose families were too poor to contribute to teachers' fees. As education became more widespread, the public contributed to the schooling for the poor, but full taxpayer financing of schooling was not common until well into the 1800s. Education was not universal, but E. G. West (1994) demonstrated that literacy rates were surprisingly high and private education was a growth industry. This same success in literacy occurred with private schools in England and France as well through this same period in time.

Private schooling dominated initially in our country's history in large part because the public sector was simply not developed. Scarce resources limited taxing capacity and service provision largely to defense. In part because of the success of schools in producing human capital, and in part because of general economic growth, the capacity of both the private and the public sectors grew rapidly in the mid-1800s.

With general economic growth came an increased demand for goods and services, including those provided by the public sector. The collective provision of goods and services in general began to grow relative to its historical rates. The population of towns was small by modern standards, and the composition of the population in each town was relatively homogeneous. Efficiency gains from collective provision of goods were shared among the townspeople without severe allocation costs that accompany provision to persons with heterogeneous tastes.

The local public sector was growing in numerous functional areas both for economic and for political reasons. Several factors directly affected the growth of public schools. To spur settlement in the West, the federal government began to grant land to new townships that was to be used specifically for the purpose of building schools (Tyack, James, and Benavot 1987). Around the same period in time, philosophical debates about the role of the state in education ensued. Some professional educators believed schools were the single place where immigrants could be taught the common values and common language of a new country. Some educators thought schools for the public were the only way to ensure socialization of humanity. The public choice answer to these debates resulted in a collectively funded, publicly operated system of schools available to all children.

In these early days of public schooling, towns were dominated by Anglo-Saxon Protestants. The curriculum of the public schools reflected this in that they provided instruction steeped in Protestant teachings. As Catholic immigrants began to move to the United States in large numbers, many towns found themselves with clusters of Catholic families who wanted an education based on Catholic teachings. Protestant families, at the same time, did not wish to use the collective funds to pay for schooling that conflicted with their own religious beliefs. In some instances, the divisions between Catholics and Protestants resulted in full-scale riots with historical recordings of at least one Catholic Church that burned to the ground in a school riot (Coulson 1999). The outcome that emerged gave the Protestant majority the "public" schools financed by taxpayers and Catholics formed their own schools financed through tees and subsidies from the churches. Today we continue to have public schools financed through tax revenues, but no longer can these schools provide religious education. Catholics, for the most part, continue to provide their own, separate schooling financed by fees and church subsidies.

So while the state's ability to provide education and its efficiency in doing so grew relative to its pre-Revolutionary War days, politics also played a substantial role in the move to state funding of some schools and private fees for other (Catholic) schools. Of great interest from an evolutionary perspective, during this same period of time, other countries following roughly the same path of economic development as the United States chose different paths for private schools. Although the stories differ, the current outcomes reflect some of the political culture across the countries. Certain European countries, for example, developed a public sector for schooling but continued to fully subsidize private schooling (Walford 1989). Many Western European countries allow freedom of school choice and, in almost all cases, school choice originated from political struggles in the 19th century (Dronkers 1995; Dijkstra, Dronkers, and Karsten 2001). Belgium, where approximately 70% of secondary enrollment is in private schools, constitutionally guarantees students the right to attend the school of their choice, private or public, with full funding from the state. Private schools have received public funding since 1914. Religious affiliation of the school is not an issue in Belgium, but the state specifies language of instruction depending on the location of the school. The Netherlands has a similar policy of fully funding schools whether they are public or private. Germany and Austria allow school choice and provide subsidies to private schools under certain conditions. France directly ties the degree of public subsidy for the school to the regulations the state imposes on the school.

Countries outside of Europe have equally interesting financial arrangements for private schools, and some have evolved relatively recently. Chile, where almost 50% of primary students enroll in private schools, has received much attention for a voucher program that it implemented in 1981 (McEwan 2001). Budgets to schools, private and public, depend upon enrollment. Argentina has a subsidy program in which the state subsidizes the salaries of teachers and principals of private schools. Argentine law guarantees private school teachers a salary similar to that of their counterparts in the public sector. If private schools do not charge tuition, teachers' salaries can be subsidized 100% (McEwan 2000). New Zealand also allows state subsidies of teacher salaries in private schools. These subsidies do not go directly to students but presumably influence the tuition price confronting the students who attend the private schools.

The institutions we currently observe for private schooling around the world reflect an accumulation of past economic, political, and cultural influences. Of particular interest, the world of private schooling in many places resembles the contracting out and privatization that we see in other sectors in the United States.

Not only in parts of the developed world, but also today when we look specifically at the developing world, we see private schools playing the role of major provider of schooling in many countries. Many parts of Africa, for example, enroll less than a third of their school-age population in primary schools of any type. In these countries, the governments often do not have sufficient resources at their disposal to offer schooling on a publicly financed basis. The developing countries lack both the rule of law necessary for private market success and the infrastructure necessary for provision of all public services, including education. Water, public health, and transportation typically rank higher in demand than schooling at the earliest stages of development. Public expenditures for schooling are low or nonexistent because economic development is in an infant stage and publicly provided education is not the top priority (Cox and Jimenez 1990; Lockheed and Verspoor 1991). But education is provided in these countries in spite of the obstacles. It occurs in the private sector. In Macau, a former Portuguese colony and now part of China, 94% of primary students enroll in private schools. In Zimbabwe, approximately 88% of students at the primary level enroll in private schools. In Belize, 87% of enrollment is in private schools. Eighty percent of children in the urban areas of some parts of India enroll in private schools. These statistics are not unlike those of the early stage of development in the United States and parts of Western Europe. Adult literacy rates vary but in some countries approach that of the pre-Revolutionary War days in the United States. The private sector in these countries serves as the feasible supplier of schooling.

The enrollment patterns in private schools that I illustrated in Tables 1 and 2 reflect to some degree either the stage of economic development or the subsidy policies that have emerged over the years. These and other factors influence the availability of schools in both the public and the private sectors. The fact that Belgium, for example, enrolls more than 50% of its students in private schools whereas the United States enrolls roughly 12% is not surprising given that private schools charge fees for service in the United States and students attend at no cost beyond the tax dollars they pay for all schools in Belgium.

While the public-private structure in the United States has been relatively unchanged over the past 100 years, the question of the efficiency of the public sector in providing schooling has become a salient policy issue in the United States over the past couple of decades. To gain insight into whether change is on the horizon with regard to institutional structures for schooling, we must assess whether the structure of schooling matters in terms of efficiency of output. More specifically, do these wide variations in structures for private schooling and the subsequent variations in enrollment patterns affect efficiency in the provision of schooling?

2. Relative Effectiveness of Private Schools

If globalization implies that, at the margin, efficiency will influence the choice of institutions in the future, it is important to understand whether there are efficiency differences in the two types of schooling around the world. As I shall argue, it is theoretically conceivable that there are always efficiency advantages for private schools, efficiency advantages at some stages of economic development but not others, or no efficiency advantages at all for private schools. Fortunately, evidence exists that can enlighten us.

I want to begin with the United States and then again place the American experience into an international context. I might add that data availability has increased in this area like other fields in economics over the past decade. Before looking at numbers, think conceptually about private and public school provision in the United States. For the simple fact that consumers pay for private schools in addition to the tax dollars they contribute for the support of public schooling, we can anticipate that private schools provide some aspect of schooling that is more valuable to parents who choose to enroll their children than the public schools do. If academic test scores or achievement is the margin valued by parents, the private schools will produce higher test scores lest they go out of business. The earlier discussion about enrollment in the United States, however, raises interesting questions. If parents choose private schools because of their special character with regard to religious education, for example, it is possible that academic achievement as measured by standardized test scores in private schools does not exceed that of public schools. Stated differently, the utility from attending private schools can exceed that of attending public schools even if some narrowly defined measure of quality is not superior.

For the past 20 years, researchers have examined the effectiveness of private schools in the United States by examining standardized test scores. Because these scores are the only quantifiable measure of output that currently exists, I shall use them as well. Data availability also limits most of the private school research in the United States to Catholic schools. Since the first study by Coleman and coauthors in the 1980s, the evidence has been fairly consistent when examining average performance levels (Coleman, Hoffer, and Kilgore 1982). On average, a student in a private school performs at higher levels on achievement examinations than students at public schools. (1)

Just as with the examination of public school effectiveness, averages tell a story that requires qualification. In the United States, for example, private school attendance relates positively to income and is related to religious affiliation, and, as I have just discussed, particularly to being affiliated with the Catholic Church (Long and Toma 1988). Almost all empirical studies of school achievement indicate that socioeconomic status of the family is one of the chief determinants of student performance. If socioeconomic status or religious affiliation influences who attends private schools, then this must be taken into account when assessing the effects of school type on educational achievement.

Labor economists have devised a variety of ways to control for the selection bias that may affect the estimates of the effects of school type on school achievement. Some of these methods are better than others. Identifying instruments that influence the probability of choosing private schools but that do not influence the educational achievement on the part of the student presents a challenge to researchers.

Although the results differ somewhat across studies, the most consistent findings of the relative effectiveness of private and public schools relate to minority students in inner cities. This is the single group of students that shows significant advantages to enrolling in private schools in terms of test score gains. These students score higher on achievement exams when enrolled in private, Catholic schools than when enrolled in public schools. More robust across studies, minorities in Catholic schools do better in terms of high school graduation rates and college attendance rates than they do in public schools (Evans and Schwab 1995: Neal 1997).

Most notable from the U.S. results is that white suburban children do not perform significantly different on standardized exams whether they are in private or public schools. Put in context of school enrollment decisions in the United States, this should not be surprising. Families choose their public school for the most part based on residential location. Upper-income households, who can afford to pay the double price of public and private schooling, can move to the location that offers them the best public schools. The children of these families perform well in school whether the school is public or private.

Especially in light of the institutional differences with respect to financing of private schools that I described earlier, it is most informative to look beyond the United States to examine the private-public performance differentials. The OECD makes available a set of data, Programme for International Student Assessment (PISA), that allows comparisons across countries. These data divide students into schools of three types--schools that are private and independent of the government in terms of funding, schools that are private but dependent on the government for financing, and government or public schools that are both publicly operated and financed. Schools that are privately managed but receive more than 50% of their core funding from government agencies are considered government-dependent. Controlling for no inputs but looking at raw averages alone proves most interesting. Table 3 presents results based on a survey of 40 countries. The highest scores on reading literacy are from students in government-independent private schools, and the lowest scores are from students in government or public schools. Table 4 indicates some of the range of enrollment percentages in these schools as well as an index of socioeconomic status. Here, not surprisingly, the socioeconomic status of students in schools that are private and government-independent exceeds that of the other two groups. Private, government-independent schools charge fees to cover the costs of operation. Because families must pay tuition in addition to tax dollars to support the public schools, we would expect to find a positive relationship between financing of the school and the socioeconomic status of those who attend the school. Also, and perhaps not surprisingly, the socioeconomic status of students in the government-dependent private schools does not differ significantly from that of students in the public schools. The price paid by the individual household to enroll in the two types should not differ significantly.

This raises the selection bias issue once again. Although I cannot give results from controlling all inputs across students in the 40 countries because of lack of data, I do have data from specific country studies. A data set from the International Association for the Evaluation of Educational Achievement (IEA) (2) consists of observations of more than 20,000 students in Belgium, France, New Zealand, and the United States. As stated earlier, Belgium offers constitutional choice of schools, public or private, sectarian or nonsectarian, at zero tuition to all students. Regulation of the private schools deals with subjects taught and language of instruction. It does not extend to religious instruction or pedagogy. Since 1959, funds have been allocated to these schools on an enrollment basis, and schools receiving the public funds cannot charge tuition.

France has a more complex system of subsidies in which the degree of regulation imposed on the school influences the size of the subsidy. A school can choose to accept government requirements for curriculum and testing in exchange for staff salaries or also receive support for operating expenses if it allows some government control over pedagogy and selection of teachers. New Zealand subsidizes salaries of private school teachers but the private schools continue to charge tuition. Schools operated by the Catholic Church have been fully "integrated" into the public system and charge no tuition.

These three countries offer interesting contrasts to the United States in terms of the institutional structure surrounding the governance and financing of private schools. Controlling for socioeconomic factors of the students' families, characteristics of the teachers and the schools, and controlling for socioeconomic characteristics of their peers, private students in Belgium and New Zealand, like those in the United States, outperform their public counterparts. In France, on the other hand, students in private schools perform at significantly lower levels than do students in public schools. Sociological descriptions of the French private schools now portray them as the residual claimant of troubled children who cannot succeed in the public schools.

Table 5 illustrates these differences. (3) The test results are based on a 40-question exam. These are predicted scores for the average student after controlling for relevant inputs regarding families, teachers, and schools. As indicated, in the four countries listed here, students from Belgium scored highest on this exam whether they were in private or public schools whereas students from New Zealand scored lowest. The last column of Table 5 lists the score gain during an academic year (eighth grade) for these students. Here the differences in public and private schools are indicated. The gain in score over the year for the private school students exceeded that of the gain for students in the public schools for students in Belgium, New Zealand, and the United States, but did not differ statistically for students in France. These results from Belgium suggest that public financing of private schools does not erase their performance advantage over the public schools. Given the large percentage of students enrolled in private schools in Belgium, it also suggests that expanding private schooling to a broader socioeconomic mix of students does not erase the comparative advantage of the private schools. The regulations applying to the private schools that choose public funding in France and the achievement results indicate that the distinction between public and private can become so blurred as to be meaningless.

I want to shift focus again to developing countries. Here the issue of school performance is qualitatively different than in the developed world. Generally, the quality of schooling is lower in developing countries than in developed ones whether we are assessing private schools or public schools. For example, on the International Assessment of Educational Progress exam in which U.S. students achieve a composite score of 46.77, students in Algeria score 28.06, students in Mozambique score 27.94, and Egyptian students score 26.43 (Hanushek and Kimko 2000).

In developing countries, not only is the quality of schooling generally poor, smaller percentages of children enroll in school. In some countries, as few as one-third of the school-age population is enrolled in school. (4) The reasons for the weak enrollment in school are numerous, but fundamentally the countries are too poor to educate their children. Consider the schooling decision relative to income levels in these countries. In the United States, per capita GNP is approximately $29,000. (5) By comparison, GNP per capita in Mali is $260. In Ethiopia, income per capita is $110. People in these countries are concerned about the consumption of food and water. Education is a luxury good that a relatively small proportion of the population demands.

In these countries, the problem of education is an entirely different one than that in the countries we normally reference. Fortunately, researchers have conducted case studies within developing countries that provide comparisons on school effectiveness in the public and private sectors. Consider India, for example, a developing country with a large population that proportionately allocates its resources roughly similar to the United States. More than 42% of its secondary students enroll in private schools. (6) Indian private schools are of two types: those that are unaided, self-financed, and autonomous: and those that are entirely funded by the public sector but also heavily regulated. The latter type of private schools cannot charge tuition or make their own hiring decisions regarding staff. Finally, there are the regular government-run, government-financed schools. On an exam designed specifically to test student achievement in India, achievement of students in all three types of schools has been scientifically estimated. A study that corrected for selection bias associated with type of school chosen found that private, unassisted schools outperformed the private, government-assisted and the government schools (Kingdon 1994, 1996). In comparing private, unassisted schools to government schools, the study found an 18% advantage for the private school effect; the unassisted private schools were found to be 27% more effective than the assisted private schools in teaching math.

In another case study of five countries--Colombia, Dominican Republic, Philippines, Tanzania, and Thailand--researchers controlled for family background and found that students in private schools outperformed students in public schools on both mathematics and language standardized exams (Lockheed and Jimenez, 1996: see Table 6). The finding was not only statistically significant but also was a meaningful advantage. For example, in Colombia, a student with an average public school background would perform at a level 13% higher if placed in a private school (Lockheed and Jimenez 1996). The private advantage ranges from one-half to two standard deviations above that of the public schools in these countries.

In these and other case studies of developing countries, authors also examined cost differences between the public and private sectors. Although the information is typically imperfect, the calculations revealed unit costs in private schools to be significantly lower than those in public schools. In one study, "For the same unit cost, private schools provide as much as three times more learning as the public schools do" (Lockheed and Jimenez 1996). In almost all cases, private schools tend to choose higher student-teacher ratios than do public schools. Scholars also have provided evidence that the private schools adopt different and better internal management strategies than do the public schools (Hannaway 1991). Private schools tend to invest more in libraries and textbooks, whereas public schools tend to invest more in personnel.

Others who have examined the cause of the private school advantage in the United States have pointed to the more active role of the principal in the private schools and the bureaucratic rules that interfere with a school leader's discretionary powers in the public schools (Chubb and Moe 1990). Even though the private schools in the United States and much of the developed world operate within a nonprofit world, there appear to be some efficiency advantages to the private sector even when funding comes from the public sector. Given the studies thus far, I cannot assess whether this advantage comes from external constraints, internal constraints, or some combination of the two.

3. Policy Implications

The school effectiveness literature suggests that, worldwide, private schools offer some advantages over public schools in terms of test scores and, with more limited data, also in terms of costs. This cursory survey of countries suggests that the efficiency advantages may be absolutely greater for developing countries than developed ones. I have surveyed a limited number of countries and have not covered every possible stage of economic development. Certainly, there may be stages of development under which the efficiency advantages of private schools may be nonexistent. Of interest, the public subsidization of private schools does not erase their efficiency advantage, although in some countries, subsidies do appear to decrease the advantage of privates over publics, and specific countries such as France illustrate that public financing with stringent regulations attached can render meaningless the differences between private and public schools. The relationship between public funding and private school performance is especially interesting given that the numbers and socioeconomic composition of students attending private schools expand with public funding.

I have implied throughout this paper that some economics and more politics played a role in the evolution of public schooling in the Western world and the subsequent policies chosen for financing private schools. I have also suggested that globalization means that politics will play a weaker role vis-a-vis the role of economics in forming institutions and public policies in the future. Thinking about the efficiency advantages in both the developing and developed countries returns us to these globalization issues. Will competition across national boundaries alter the set of institutions that govern the provision of schooling? Begin with the developing countries. As economic growth expands to more parts of the developing world, the public sector is likely to increase its capacity to provide goods and services generally just as it did in the United States, in Europe, and in other parts of the Western world in the 1800s. Strictly from the perspective of efficiency in provision, it is unclear that increased capacity is sufficient to warrant public institutions for schooling. But increased public sector capacity does enable countries to address issues such as families' ability to pay for schooling. Developing countries will be able to choose between public management of schooling or private provision with public funding. It is likely that those developed countries with lesser propensities toward regulating the private market will find greater gains from public funding for private provision of schools. Those countries with a strong cultural propensity toward stringent regulating of markets will find less reason to publicly fund private schools. As globalization pushes the latter countries toward more marketlike arrangements generally, then it seems the same will occur for schools.

Turn now to the U.S. schools. There is evidence that our school structures are already evolving just as other institutions are changing in our public sector. Forty-two states have introduced charter schools that allow full public funding but private governance and management. The states have chosen varying degrees of regulatory authority over these schools. In limited cases, such as Milwaukee and Cleveland and now Washington, D.C., we are funding private schools through vouchers. The federal government has mandated accountability measures that carry sanctions when standards are not met. The sanctions allow children to choose among public schools and, in limited cases, between public and private schools. Private for-profit management companies are operating public schools. The lines between public and private in education are growing less distinct as they are blurring in other sectors of the economy. And if globalization and competition constitute the source of this blurred distinction, then these changes presumably are efficient.

In the end, only a crystal ball can predict what an efficient institutional structure for schooling will be tomorrow or a few decades hence. That indeed is the magic of the marketplace. Globalization seems to imply that efficiency will dominate politics in the evolution of our institutions. I can think of no reason for the exemption of schools from the evolutionary process. Transitions are costly and our population is heterogeneous. We may move slower than developing countries that will experience higher returns from marginal changes to their institutions of schooling. But I predict we will see a continuation toward market-based institutions of schooling.

Table 1. The Top 20 Enrollment Rates for Private Schools
at the Primary Level

Country                  %

Macau                   94.1
Zimbabwe                88.1
Belize                  87.1
Aruba                   82.4
Netherlands Antilles    74.8
Netherlands             68.4
Lebanon                 63.6
Belgium                 54.4
United Arab Emirates    46.5
Chile                   45.5
Togo                    39.7
Qatar                   39.3
Bangladesh              38.7
Liberia                 38.4
Antigua and Barbuda     38.2
Malta                   35.9
Brunei                  35.1
Pakistan                34.8
Bermuda                 33.8
Spain                   33.4

Source: UNESCO, Institute for Statistics. Various years. Summarized by
Natiomaster.com.

Table 2. The Top 20 Enrollment Rates for Private Schools at the
Secondary Level

Country                              %

Aruba                               88.9
Belize                              79.1
Netherlands Antilles                78.0
Kiribati                            77.5
Netherlands                         75.1
Mauritius                           74.6
Tonga                               74.5
Belgium                             69.3
United Kingdom                      52.8
Comoros                             52.1
Lebanon                             52.0
Guatemala                           47.7
Rwanda                              46.4
Saint Vincent and the Grenadines    46.3
Madagascar                          45.0
Chile                               44.0
Dominica                            43.5
Tanzania                            43.1
Indonesia                           42.4
Liberia                             41.1

Source: UNESCO, Institute for Statistics. Various years. Summarized by
Natiomaster.com.

Table 3. Average Country Scores: Reading Literacy by School Type

Government or     Government-Dependent    Government-Independent
Public Schools      Private Schools          Private Schools

466                      489                      504

Source: OECD/UNESCO 2003, table 7.13.

Table 4. Average Occupational Status Index Across Countries and School
Type

Government or                         Government-Dependent
Public Schools                           Private Schools

Percent                              Percent
Participating     Occupational    Participating    Occupational
Students             Index          Students          Index

84                     45              12               47

    Government-Independent
       Private Schools

   Percent
Participating     Occupational
  Students           Index

      9                57

Source: OECD/UNESCO 2003, table 7.13.

Table 5. Predicted Score Gain and Achievement Level, by Country

                 Achievement Level    Score Gain
Belgium
  Public               21.36             3.64
  Private              23.65             5.55
France
  Public               20.68             6.73
  Private              20.16             6.46
New Zealand
  Public               16.44             1.67
  Private              17.46             2.60
United States
  Public               20.03             4.40
  Private              20.58             4.91

Source: Toma 1996, p. 143.

Table 6. The Private School Advantage

                                                   Relative    Effect
Country               Indicator of Achievement    Advantage     Size

Colombia              Average math and verbal        1.13       0.55
Dominican Republic    Mathematics (O-type)           1.31       0.89
                      Mathematics (F-type)           1.47       2.16
Philippines           Mathematics                    1.00      -0.09
                      English language               1.18       0.33
                      Filipino language              1.02       0.25
Tanzania              Average math and verbal        1.16       0.97
Thailand              Mathematics                    2.63       1.69

Source: Lockheed and Jimenez 1996.

(1) See Neal 2002 for a summary of this literature.

(2) The data here are from the second international study of math and science (SIMMS). The third wave of this test across countries by the IEA collected information about school type--whether public or private--but the data on school type information are not available to the public.

(3) See Toma 1996 for a full discussion of the methods and results using these data.

(4) In almost all these countries, the enrollment rate for females is less than that of males. In 1996, for example, Ethiopia enrolled 55% of its primary school-age males but only 30% of its females. Mali enrolled 58% of its age eligible males in primary school but only 40% of its females in the same year (UNESCO 2000).

(5) This is 1997 dollars from UNESCO (2000).

(6) Tooley 2001. Statistics at the primary level are not available.

References

Chubb, John, and Terry Moe. 1990. Politics, markets, and America's schools. Washington, DC: Brookings Institution.

Coleman, James S., Thomas Hoffer, and Sally Kilgore. 1982. High school achievement: Public, Catholic, and private schools compared. New York: Basic Books.

Coulson, Andrew J. 1999. Market education. The unknown history. New Brunswick, NJ: Transaction Publishers.

Cox, Donald, and Emmanuel Jimenez. 1990. The relative effectiveness of private and public schools: Evidence from two developing countries. Journal of Development Economics 34:99-121.

Dijkstra, AnneBert, Jaap Dronkers, and Sjoerd Karsten. 2001. Private schools as public provision for education: School choice and marketization in the Netherlands and elsewhere in Europe. National Center for the Study of Privatization in Education, Occasional Paper No. 20.

Dronkers, Jaap. 1995. The existence of parental choice in the Netherlands. Educational Policy, 9:227-43.

Evans, William N., and Robert M. Schwab. 1995. Finishing high school and starting college: Do Catholic schools make a difference? Quarterly Journal of Economics 110:941-74.

Hannaway, Jane. 1991. The organization and management of public and Catholic schools: Looking inside the 'black box'. International Journal of Educational Research 15:463-81.

Hanushek, Eric A., and Dennis D. Kimko. 2000. Schooling, labor force quality, and the growth of nations. American Economic Review 90:1184-208.

Hoxby, Caroline M. 2004. Productivity in education: The quintessential upstream industry. Southern Economic Journal 71: 209-31.

James, Estelle. 1993. Why do different countries choose a different public-private mix of educational services? Journal of Human Resources 28:571-92.

Kingdon, Geeta Gandhi. 1994. An economic evaluation of school management-types in urban India: A case-study of Uttar Pradesh. D.Phil. thesis, University of Oxford.

Kingdon, Geeta Gandhi. 1996. The quality and efficiency of private and public education: A case-study of urban India. Oxford Bulletin of Economics and Statistics 58:55-80.

Lockheed, Marlaine E., and A. M. Verspoor. 1991. Improving primary education in developing countries. New York: Oxford University Press.

Lockheed, Marlaine E., and Emmanuel Jimenez. 1996. Public and private schools overseas: Contrast in organization and effectiveness. In Who chooses, who loses? Culture, institutions, and the unequal effects of school choice, edited by Richard Elmore and Bruce Fuller. New York: Teacher College Press, pp. 138-53.

Long, James E., and Eugenia F. Toma. 1988. The determinants of private school attendance, 1970-1980. Review of Economics and Statistics 70:351-57.

McEwan, Patrick J. 2000. Private and public schooling in the southern cone: A comparative analysis of Argentina and Chile. National Center for the Study of Privatization in Education, Occasional Paper No. 11.

McEwan, Patrick J. 2001. The effectiveness of public, Catholic, and non-religious private schools in Chile's voucher system. Education Economics 9:103-28.

Neal, Derek. 1997. The effects of Catholic secondary schooling on educational achievement. Journal of Labor Economics 15(1):98-123.

Neal, Derek. 2002. How vouchers could change the market for education. Journal of Economic Perspectives 16:25-44.

OECD/UNESCO. 2003. Literacy skills for the world of tomorrow: Further results from PISA 2000. Paris: OECD/UNESCO.

Toma, Eugenia F. 1996. Public funding and private schooling across countries. Journal of Law and Economics 39:121-48.

Tooley, James. 2001. The global education industry. Lessons from private education in developing countries. London: Institute for Economic Affairs.

Tyack, David, Thomas James, and Aaron Benavot. 1987. Law and the shaping of public education: 1785-1954. Madison, WI: University of Wisconsin Press.

UNESCO. 2002. Worm education report. Paris. France: UNESCO.

UNESCO. Institute for Statistics. Various years. World education indicators. Montreal, Canada: UNESCO. Summarized by Natiomaster.com.

Wallord, Geoffrey, ed. 1989. Private schools in ten countries: Policy and practice. London: Routlege Publishing.

West, E. G. 1994. Education and the state: A study in political economy. 3rd edition. Indianapolis, IN: Liberty Fund Press.

Eugenia Froedge Toma is Professor, Martin School of Public Policy and Administration, University of Kentucky, Lexington. This paper was prepared as her presidential address at the annual meeting of the Southern Economic Association, held in New Orleans. Louisiana, November 22, 2004.

I thank J. S. Butler, David Salisbury, Mark Toma, and Ph.D. students at the Martin School for comments on earlier versions of this paper.