A Proposal to Decentralize Higher Education in Taiwan:

The Voucher System


            The aim of this paper is to present a proposal for a radical change in Taiwan’s higher education system. Under the current system, the Ministry of Education formally controls all aspects of demand and supply. It formally determines which student applicants will be allowed to enter the system and assigns them to the departments and colleges. It also formally approves hirings of all administrators and teachers and their assignments, although the practical exercise of this power is typically delegated. Most importantly, the Ministry pays for higher education by authorizing government money to be sent to the colleges and universities. About 90% of the total funding for “public colleges” comes from the government through the Ministry. The amount is less for so-called “private colleges.” However, all colleges and universities are subsidized and controlled by the Ministry.

            This paper proposes to change the subsidization method. It proposes to issue vouchers to all students who pass an entrance examination. The vouchers would allow a student to enter a college or university of his choice. The school could then redeem them, through the Ministry, for cash.

            Under the voucher system, all colleges and universities would be privatized. The ministry would sell off all existing government-owned colleges and universities to the current teachers and administrators at below-market prices. In order to receive additional funds from the government, the schools would then have to attract voucher or tuition-paying students in competition with other schools. In addition, the Ministry would gradually open the market to new competition from accredited schools.

            The paper argues that a switch to this system will eliminate the current shortage of college and university spaces in Taiwan, greatly improve quality, and greatly increase student and parental choice. In addition, it is likely to raise the level of economic efficiency. Finally, the paper describes a practical plan for gaining the approval of the voucher system by current teachers and administrators.

A Proposal to Decentralize Higher Education in Taiwan:

The Voucher System

            Taiwan's education system at all levels has been highly centralized since 1950, when it was instituted under the regime of Chiang Kai-shek. The centralized system was consistent with the two functions of education under the one-party Leninist system of government that Chiang had imposed: (1) to prepare citizens to perform societal functions and (2) to indoctrinate them into the party philosophy. Although the task of preparing citizens to perform societal functions can be delegated to the private sector, indoctrination, including the identification and modification of dissident behavior, must be administered by trusted party loyalists. Such an important function cannot be delegated to the universities but must ultimately be the responsibility of a loyal ministry that controls the entire system.

            The Taiwan Ministry of Education (MOE) was employed to perform these tasks. The basic structure has not changed, despite political liberalization beginning in the late 1980s. As in the early days of one-party rule, the MOE today possesses direct control over all elements of higher education. Footnote It stipulates detailed regulations for the schools regarding hiring, firing, promotion, and pay of teachers and administrators. It has the duty of reviewing and approving all changes in curriculum and programs. And it is responsible for selecting and assigning students to the various colleges and universities. It also has independent power expel students and to reject expulsions. Finally, all higher education funding is channeled through the MOE. Thus the amount of money received by a college or university depends on an MOE decision.

            Today, the MOE no longer rules with an iron hand. Its decisions no longer take account of faculty, administrator, or student political views within colleges and universities, although the ruling party still appears to be over-represented in the top administrative positions. In addition, the MOE delegates most of its power regarding personnel and curriculum to the colleges and universities. Nevertheless, it retains the ultimate power of approval of all such decisions and requires detailed reports of college and university meetings and activities concerning such issues. We shall discuss some of these remnants of the earlier system below. The important points at this stage are that under the current system the MOE (1) selects students for colleges and universities (2) controls budgets, and (3) subjects colleges and universities to a myriad rules and red tape inherited from the monitoring and control philosophy that prevailed prior to the end of martial law in 1987.

            While the structure of the higher educational system has not changed much since 1950, Taiwan society has undergone radical changes. During the last thirty or forty years, Taiwan has experienced economic development similar to that of the other three Asia tigers: Hong Kong, Singapore, and South Korea (6-8% per year). Development has been accompanied by a decentralized local market economy, relatively free exports, and gradual privatization. Throughout, the government respected individual property rights and allowed private citizens freedom in their choice of business and jobs. As a result, citizens’ incomes have continually risen. These incomes have been used partly to travel and to spend on foreign goods. Taiwan is well on the way to joining the ranks of the "developed countries."

            Political change following the end of martial law has been rapid. The emergence of party politics in 1987 was followed by the first legislative elections in 1992 and the first Presidential election in 1996. During the last eleven years, the ruling party has gradually, if not intentionally, lost power. Today (late 1998), the ruling party still commands a small majority in the legislature. However, the mayors of the two largest cities are from the main opposition party and it is possible that no party will be able to gain a majority in the next legislature. Modern Taiwan. is truly a pluralistic, democratic society whose laws respect basic human rights and where people enjoy the typical freedoms associated with democracy.

            These radical economic and political changes contrast sharply with the stagnant nature of higher education. Although the MOE has drastically reduced its activities in behalf of the ruling party, its dictatorial control over school budgets, over who can be a student, and over who can operate a college or university seems incongruous with the freedom and increased citizen participation that is occurring in other parts of the society. Just as important is that the centralized system is grossly inefficient for pretty much the same reasons that central planning of other services is typically inefficient in a democracy.

            The aim of this paper is to present a radical alternative to the highly centralized higher education system of Taiwan. Its name is the voucher system. A detailed description of how the system would work is provided later in the paper. At this stage it is enough to say that the system would abolish government-funded colleges and universities. Under the voucher system, existing public colleges and universities would be fully privatized and they would have to compete for students. Subsidies for higher education would be paid directly to each eligible student in the form of a voucher, which he could redeem only at accredited private colleges and universities. Entry into the business of supplying higher education would be totally free. Anyone could start a college or university. The only controls retained by the MOE would be its testing of students and its determination of minimum standards for schools to be eligible to redeem student vouchers.

            The paper begins in part one by describing higher education in Taiwan and its problems. Part two presents the theory of the voucher system. Part three shows how a voucher system would deal the problems identified in part 1. Part four shows how the system would be administered. Part five devotes special attention to the difficulty of achieving acceptance on the part of the teachers and administrators who now supply higher education in Taiwan. It presents a plan to compensate them for their losses. The paper ends with a conclusion. The appendix presents other voucher proposals and actual cases in the U.S. elementary and secondary schools, where variants of the voucher system have occasionally been adopted.

1. Higher Education in Taiwan

            This section begins by describing the organization of higher education in Taiwan. It goes on to identify some of the consequences and problems of this system.


            Higher education in Taiwan is a two-tiered system of (1) national universities and colleges and (2) private universities and colleges. The difference between a university and college is based mainly on size, with a university consisting of two or more colleges. Footnote Thus, in subsequent discussion, we refer to both as colleges. The term "private," when used to describe colleges in Taiwan, must be used with caution. Today's "private" colleges are really non-profit organizations. Formally, this means that excess revenue is absorbed into the cost structure and any shortage of revenue must be offset by reductions in costs. However, it is likely that, as in other non-profit organizations, college directors are often able to convert some of the profits into benefits for themselves. More importantly, private colleges are subject to most of the same regulations as public colleges.

The Demand Side

            The MOE fully regulates demand, supply and price of higher education. On the demand side it decides how many students each college, both public and private, will be allowed to accept and it determines which students will be allowed to attend. The basis for its control over demand is the nationwide entrance examination (Joint College Entrance Examination, hereafter JCEE). The MOE assigns students to colleges and to the departments within them, based on their test scores on the JCEE and their preference rankings. The highest scoring student in a particular subject (e.g., science) gets her pick of the departments in the subject that she has listed. The lowest scoring student is typically assigned to the college that was lowest on her list and less preferred by other test-takers.

            Student rankings of colleges are, in large part, determined by price (i.e., tuition). Tuition at private colleges are set higher by the MOE than it is at public colleges. Since, other things equal, students prefer lower-tuition education, those who sit for the national college entrance exam tend to rank private colleges lower than national colleges. Of course, if private colleges were free to use their higher tuition to offer a higher quality of education, this preference for low tuition might be offset by a preference for higher quality. However, due to regulation of supply conditions (discussed below) and to the MOE's discrimination in subsidy between public and private colleges, Footnote administrators of private colleges lack both the incentive and the means to engage in the kind of vigorous quality competition that is observed in less regulated systems. Apparently, the greater cost pressures on private colleges has resulted in much higher student-teacher ratios. Footnote

            How does the MOE decide how many students to admit to each college? In earlier days of the military dictatorship under the Chinese Nationalist party, it seems likely that such decisions were made on the basis of party policy. Today, however, it seems safe to say that the MOE follows a policy of incrementalism. Allowing for some change in the number and size of particular departments, it allows each college to grow more or less proportionally.

The Supply Side

            On the supply side, the MOE's regulation appears at first to be pervasive. It decides college teacher qualifications; teacher and administrative salary structures, including salaries corresponding to each rank. It also has the power of approval over all teaching and administrative appointments and the type of permissible specialized departments, institutes, and programs in each college. Footnote The result of its control over salary has been a rigid system that equates the salaries for each rank regardless of the real demand for and supply of teachers in different fields. It responds to perceived future shortages of teachers in particular fields either by controlling demand or by creating new graduate institutes, departments, and even colleges. Surpluses of teachers in some fields are accompanied by vigorous competition for jobs.

            At an earlier point in history we might have expected the MOE to only approve new hires and proposals for increases in department size, new departments, and other changes according to whether they meet the above-stated party goals for higher education. Indeed, an active MOE would set the agenda for change by recommending to college presidents and deans that they institute various changes that it had decided upon. Since the beginning of democratization in 1987, however, the MOE seems to have adopted an increasingly passive role. It has tended become reactive rather than pro-active. The bulk of changes are now initiated by the departments, schools and colleges themselves. At the lowest level, proposals pass through the democratic filtering process. As a result, the higher education system has been tailor-made to serve the special interests of college teachers, college administrators, administrators of the MOE bureaucracy, and suppliers of resources used by colleges. One way to characterize the supply of higher education in Taiwan is as a set of democratically-run colleges subject to centralized, authoritarian rules administered by a large government bureaucracy.

            Because the MOE controls college budgets; assigns students to departments; determines teacher and administrative salaries; and decides whether a department, school, or college can expand; neither the private nor the public colleges have much incentive to compete. It is broadly true that no one can earn money profits by increasing enrollment, improving quality, or offering innovative service in accord with consumers' money demands. The main benefits an administrator or teacher can hope for are personal prestige and a possible advance in the unlikely event that his actions are recognized by superiors who care.

Consequences and Problems

            In this section, we identify several consequences and problems that seem to be associated with this form of organization. The first is due to a combination of the MOE's subsidy and restrictive control over supply. The result has been a shortage of available spaces. In 1995, altogether 125,490 high school graduates competed in the JCEE, but only 55,604 spaces were made available (Ministry of Education 1997a: p. 35). This means that 69,886 applicants (56%) had no opportunity to attend colleges in Taiwan. No matter how much money a student was willing to pay or how competent she might be in the eyes of others, she was excluded by the MOE. The only alternative was to study abroad. It is difficult to know how many of those excluded students would have been willing to pay the full price of an efficiently supplied education. But presumably some would have.

            A second consequence is low quality relative to that which is achievable. Because spots at colleges are centrally determined by the MOE, there is little incentive for colleges to compete in terms of quality. This explains why the average failure rate plunged to the lowest level ever, 0.64%, for the spring semester in 1996 school year. Footnote Once a student is admitted to a college, he or she will almost always finish with a degree by meeting minimal academic standards.

            This does not mean that students have no incentive to study. If they want to obtain a government job after graduation, they must pass the government employees' exam, which typically contains questions from material covered in college. Also, if they want to continue their education in Taiwan, they must pass the graduate institute exam in the particular program for which they apply. Such exams typically are based on material covered in undergraduate schools. The point is only that the college itself provides little incentive for students to perform well.

            Perhaps the most important impact of centralization in Taiwan's higher education is its effect on parental and student choice. Under the current system, the colleges are not responsive to citizens wishes. They mainly serve the interests of college teachers, administrators and resource suppliers and, to a lesser degree, the interests of the ruling party and MOE officials. As the democratization process matures, the ruling party and MOE will have to become more responsive to citizens' demands or it will lose power. At present, however, the wishes of students, parents, or citizens in general seem to play an insignificant role in MOE decisions.

 2. The Theory of the Voucher System

            The voucher system proposed in this paper would phase out national colleges and eliminate most regulations on private colleges. In its place, the MOE or some other organization would be responsible for deciding whether a totally private college meets the accreditation requirements. In addition, the ministry would have to decide which students are eligible for vouchers and it would have the task of controlling fraud. The aim of this part is to describe the general theory of the voucher system.

The Voucher System as a Means of Causing Specific Goods and Services to be Supplied

            The voucher system is a means through which a government can use its revenues, obtained through taxes or some other means, to cause a specific good or service to be supplied to selected consumers. Government agents select a target population of consumers and proceed to give the members of the population purchase vouchers with a given dollar face value. They promise suppliers of the good to redeem the voucher at its face value so long as the suppliers have supplied the specified good. Assuming that the amount of money represented by the voucher is high enough or that consumers add to it with their own money, entrepreneurs will cause the good to be supplied to consumers in exchange for the voucher. For example, suppose that the service supplied is a measles inoculation. The fiscal agent might set the voucher amount equal to, say, $1000. He would decide on a target group of eligible recipients, say all children who are four years old. Then he would distribute the $1000 vouchers to their parents. The parents would use the vouchers to buy the inoculations from authorized suppliers. Eventually, the suppliers would send the vouchers to the government fiscal agent, who would pay $1000 for each one. The voucher system has been used for a variety of healthcare services. In addition, it has been used for food and housing services to the poor.

            It should be noted that the voucher system distributes some good or service, not money. Government decision-makers have decided that the service is worthy of distribution. If the government wanted to give money to the target population, it would have no need for a voucher.

            As suggested above, the agency administering a voucher system may permit individuals to supplement the voucher by paying an additional amount. For example, suppose that government decision-makers wanted to increase the consumption of garlic. It might distribute garlic vouchers of, say, $200 per bottle for standard garlic. If a voucher recipient wanted to purchase a $300 bottle that was enhanced with vitamins, he could do so by only paying the difference of $100.

The Voucher System as an Alternative to Government Supply and Contracting Out

            Most of the recent literature on the voucher system treats it as a means of privatization. The voucher system is not complete privatization since the government still subsidizes the demand for the good or service and it still must monitor the supply. Compared with a system of government direct supply, it is partial privatization. Subject to the government requirements of supply, the voucher system privatizes the decision about quality and ancillary services by requiring suppliers to meet the demands of consumers, instead of politicians and bureaucrats, if they want to earn profit. There are two alternatives: direct government supply and contracting out. We consider each in turn.

Direct Government Supply

             The voucher system is an alternative to two other means of supply: (1) direct government supply and (2) contracting out. Consider each in turn. In direct government supply, government officials not only decide on the services, they also hire the factors of production needed to supply them. National defense services are the best examples. At present, Taiwan's colleges are also examples, since teachers and administrators must ultimately be approved by the MOE. This means that the teachers and administrators are basically government bureaucrats. The modern theory of public bureaucracy recognizes that when government officials supply marketable goods and services themselves, they lack incentives to discover consumer wants, to satisfy them at the lowest cost, to identify and utilize new technologies, and to invent novel ways of combining the goods and services they supply with others.(Mitchell &. Simmons 1994) Moreover, government bureaucracies are often employed partly to serve special interests, who contribute in one way or another to the ruling political party. The voucher system is regarded as a means of shifting supply from presumably inefficient, bloated, and possibly corrupt government bureaucracies and contractors to cost-minimizing, profit-seeking private companies.

            Advocates of the voucher system envision a vigorous competition among private firms either along a quality or along an ancillary service dimension. Consider the hypothetical example of measles inoculations. We might expect private suppliers to compete in a variety of ways. Some firms might offer consumers convenience of location and time. Some might offer very high quality control. Some might offer special psychological counseling to minimize the trauma that some children experience. Some might specialize in offering short waiting times and in minimizing red tape. And some might combine the measles inoculations with other complementary services like physical examinations and other inoculations.

            Compare this with government-established clinics. We might expect the locations and times to be less convenient, lack of preparedness for trauma, long waiting times, and substantial red tape. In addition, we might expect that the costs of supplying the basic service would be higher since managers of government hospitals are likely to be less cost conscious.

Contracting Out

            In contracting out, government officials decide which services will be supplied and proceed to find a private contractor to supply them at a negotiated price or at a price that is determined through a bidding process. Footnote A typical example is road-building, in which competing private contractors bid for the government contract to build a road. Although the private contractors themselves have incentives to reduce costs, their performance is shaped by the government bureaucrats who choose them and are responsible for monitoring them. Such bureaucrats lack incentives to be hard bargainers with taxpayers' money and to monitor in the consumers' interests. Unlike business firms, bureaucrats cannot earn profit by causing quality products to be supplied at cheap prices. They have other goals besides profit, such as prestige, power, desires for promotion and political office, and service to their political party. Such goals can often best be achieved if they try to maximize their budgets. Also, as in the case of direct government supply, the bureaucrats are subject to pressure from special interests. Since they are typically appointed by elected officials, their actions can easily be influenced by special interest groups seeking favors from the officials. Finally, there is a prospect for corruption.

            Democracy enables citizens to maintain some control over the bureaucrats through the election of politicians who do not have such goals and who are not easily influenced by special interests. However, voter control over bureaucrats is indirect and weak because of the relative costs of organizing effective special interest groups to compete with the rival special interest groups of bureaucrats and resource suppliers. One result of this is collusive price-fixing among otherwise rival contractors. Another is shoddy quality.

            We can regard the voucher system as a partial solution to the oversight problem. If the system is properly administered, voucher recipients will use the voucher in the same way that they use money. They will want the highest quality good or service for the voucher. Private suppliers must satisfy consumer demands if they want to succeed vis-a-vis competing suppliers. If the voucher amount is high enough, suppliers will offer packages of goods or services that they believe are more suited to consumers tastes than the packages of their rivals. The prospect for profit also gives firms an incentive to discover new technologies and to use them in order to outbid their rivals.

The Special Problem of Contracting in Taiwan

            An especially important problem in a new democracy like Taiwan is corruption. Freed from the stern discipline of the Leninist dictatorship, prospective contractors and the politicians that decide on the contracts can reap high profits from the lax enforcement of regulations. Under the dictatorship of earlier times, a corrupt contractor or government official could face the severest penalty for a transgression. However, under the new democracy, such controls are lax. On the one hand, many of the rewards can be channeled into the coffers of the ruling party or party officials. This is regarded as politically acceptable to most party members since the party faces fierce competition from rival parties. On the other hand, complaints by the opposition about corruption can be written off as having a political motivation. The corruption is facilitated by the legacy of secret dealings by the former dictatorship. One can expect these factors to diminish in influence as the democracy becomes more transparent and as elections become more competitive. For the time being, however, the problem of corruption is quite serious. One type of corruption is payoffs by the contractors themselves to the government officials responsible for deciding contracts or monitoring. A second type consists of threats of bodily harm to such officials.

Voucher System Efficiency in Taiwan Higher Education

            The only country in the world that currently uses a voucher system for higher education is Finland. Other countries have experimented with the voucher system in lower education. However, the system has only be used on a temporary basis or in isolated, rather unique situations. Moreover, it has often been used to achieve purposes other than that of subsidizing education for qualified students. Accordingly, there are no empirical studies that are directly relevant to the problem of comparing the efficiency of the voucher system with its alternatives for the case of higher education in Taiwan. Nevertheless, we can consider some indirect, suggestive studies and we can present some additional theoretical considerations.

            To do this it will be convenient to identify two types of efficiency. We can call the first one "supplier efficiency." This refers to the relationship between the price, or dollar value, of output supplied by a single supplying institution and the dollar cost of the inputs. An efficient supplying institution would maximize the difference between the dollar value of output minus the dollar cost. The second can be called "allocative efficiency," following Liu (1996). It refers to the dollar value of one type of educational service compared with the dollar value of another kind of service. Allocative efficiency exists when it is not possible to increase the total value of educational service by shifting a dollar's expenditure from one field of study to another -- for example, from science to history.

Judging Supplier Efficiency

The Statistics

            Supplier efficiency refers to the dollar value of output per dollar value of input. In higher education, there is no unambiguous definition of the output of a college education. Test scores can be employed to compare the knowledge of different individuals or to compare a single individual's knowledge at different points in time. However, test scores are imperfect measures of students' abilities. More importantly, test scores are influenced not only by the supply inputs but also by the inputs provided by students themselves and other factors over which the supplier has no control. Other suggested measures of output include drop-out rate, school attendance rate, and the rate at which students pursue further schooling (e.g. college). Like test scores, none of these is necessarily related to the abstract notion of output. And each is influenced by other factors that cannot be controlled by the supplier.

            If we knew which teaching techniques were efficient, however defined, we might be able to compare the voucher system with other forms of supply by comparing the techniques used in each. If schools under a voucher system used techniques that were proven to be efficient, we would conclude that the voucher system was superior. However, we do not know this. Consider the technique of maintaining low student-teacher ratios. Most people probably believe that lower student-teacher ratios lead to better learning, although they would be more costly. However, Hanushek's 1986 survey of United States public schools found that only 23 out of 112 of the schools studied showed a statistically significant relationship between student-teacher ratio and achievement by students. Apparently, it is uncertain whether even low student-teacher ratios cause greater achievement.

            If low student-teacher ratios only occasionally cause higher achievement, what about other methods? These include the experience of the teacher; school organization; and resources such as facilities provided, size of the library, services of the computer center, etc. Hanushek reported the same statistically insignificant relationship between these variables and education output, as measured by achievement.

Putting Ourselves in the Shoes of Suppliers

            Because we lack a good way to determine the efficiency of alternative methods of supplying education services, we have no choice but to try to put ourselves in the shoes of educators in order to understand how they might behave, given the incentives that they face. On the basis of research on bureaucracy in general, it seems reasonable to argue that when education is supplied and regulated by the government, the incentives to maximize the output for a given set of inputs would be low, since it is difficult for a college president or other manager to profit by being efficient. Such people are likely to want prestige and to enhance their prospects for future employment and political office. But these desires are not high motivators of efficiency. More importantly, these goals can often be best achieved by not being efficient. Presidents at government-supported colleges in a democracy can often best achieve their goals by pursuing a policy of unrelenting growth or by catering to the aims of influential special interests or members of the ruling political party.

            However, suppose that supply by government bureaucracy is replaced by a voucher system. Then the president of each competing college would have an incentive to maximize the difference between the sum of the voucher money and the money costs. So long as the colleges compete, and so long as the overseer (MOE) does not attach an unrealistically high value to the voucher, the voucher money will represent the value of output to the voucher recipient. Supply efficiency would be approximately achieved. Roughly speaking, a college president's aim would be like that of most business managers -- to attract as many students as possible in order to maximize profit for stockholders.

            Suppose that during a particular period of time, the president of a national college under the current system in Taiwan had the same inputs as the president of a fully private, competitive, college under the voucher system. To maximize his budget, the national college president would use some of the inputs to persuade the MOE to increase its budget. He would encourage his administration and faculty to do the same. Moreover, because he could not earn profit, he would have little incentive to determine the value of his output or the efficiency with which the inputs were used to produce output. The private college president under the voucher system would want to maximize his profit in order to serve the interests of the college’s owners. He would try to determine which output is most valued by holders of vouchers, he would try to find the dollar costs of producing alternative outputs, and he would choose to produce the outputs that he believed would result in the greatest dollar profit. The efficient use of inputs to supply educational services would be a central concern. The aim of the president will be reflected by the actions of the rank and file. Administrators and teachers at a fully private college would help the president earn a profit; otherwise they would risk losing their jobs.

            Consider the situation in Taiwan. Statistics from MOE show that, on average, per student expenditure of the national colleges in 1995 was $233,471. This is 2.5 times higher than that of private colleges ($93,335). Footnote In other words, the cost of educating a student at the public colleges are on average 2.5 times as high as the cost of educating a student at the so-called private colleges. For public colleges to be as efficient as private colleges in this sense, they would, roughly speaking, have to provide 2.5 times better education than that of the private colleges. Given the unequal treatment of these two types of schools under the law, there is no fair way for us to determine whether this is so. However, it seems unlikely that most observers would claim that the public colleges provide 2.5 times the "amount of education services" as the private colleges. Footnote

            Private colleges under a voucher system would have the goal of maximizing profit. With this fact in mind, some analysts might be critical of the system on the grounds that some of the government's expenditures on higher education would wind up in the hands of the owners and stockholders of private companies. Although this is true, it is important to recognize the benefits of market competition. First, so long as the government does not restrict competition, high profit for a particular college would be a sign that the college is either unusually good at identifying consumer demands, unusually efficient in supplying the demands, or both. Second, high profits are usually temporary since they provide a signal to competitors that the services of the profit-making firm are more desired by consumers or that the methods of production are more efficient. This signal ordinarily indicates that competitors can increase profit by trying to copy the high-profit firms' actions. Such copying -- i.e., such competition -- not only reduces the profit of the high profit firm, it improves the efficiency of overall supply. Thus, one can argue that the high profits would often go into deserving hands. In the voucher system we shall propose below, existing teachers and administrators of public colleges will initially own all the stock. This means that any profits would go to them. Since they are presumably the ones responsible for the superior service offered to students, we might deduce that they are deserving.

Judging Allocative Efficiency

            Supplier efficiency refers to the costs to a single firm of supplying given output values. Allocative efficiency, as described by Liu (ibid.), refers to the types of education services provided by schools as a whole. It refers to the allocation of the totality of resources among different types of education outputs. In higher education it refers to the allocation of money among the different subject areas, such as science, engineering, business, government administration, social science, history, political economy, the professions, and so on. Theoretically, there is one best or one set of best possible allocations for the education system as a whole.

The Problem of Determining a Standard

            The concept of allocative efficiency is a theoretical standard against which any actual allocation can be compared. In using this norm, we ask: Is it reasonable to argue that government supply or a voucher system is more likely to approach the standard?

            To construct such a standard, we must make some judgment about which college subjects are most important. The usual procedure in economics is to use the concept of consumer demand. A system is said to be allocatively efficient if it best satisfies consumer demand. For example, we might argue that the Chinese communist system of producing food for the citizens of Mainland China was inefficient because it resulted in mass starvation by either not using valuable farm resources efficiently or by using them for other purposes. This was clearly against consumer demands. A market system, which provides profit incentives for farmers to specialize and trade, is more allocatively efficient. Under such a system, profit-maximizing farmers have an incentive to use their resources to produce goods which consumers are most willing and able to buy.

            To apply this standard to higher education in Taiwan, we must first identify a target group of consumers. In ordinary economic analysis of a free market, economists assume that the consumers who have the highest demands for the products are the target group. For example, in applying the concept of efficiency to the restaurant industry, economists would ask whether it supplies the demands of those who are willing and able to pay for restaurant meals. Economists would not try to second guess consumers by suggesting, for example, that efficiency requires fewer American restaurants and more Japanese ones than consumers themselves would want. However, in the case of Taiwan education, the traditional assumption in the education ministry is different.

            During the military dictatorship, of course, the consumer was the President and his assistants. The function of the MOE was to carry out the government’s directives and provide it with information and advice. Its main aim was to (1) to prepare students for various specializations regarded as important to the President and party cadres and (2) to indoctrinate students into the party philosophy. One result is that, compared to Western democracies, there was little opportunity to study subjects like political economy, the social sciences, psychology, and history (in the Western sense).

            Since democratization, the MOE has had greater autonomy. Today, the MOE’s predominant assumption appears to be that the MOE experts are more capable of deciding which programs a college should have and which courses students should take than the students themselves or their parents. As opposed to the MOE officials, typical distinguished Taiwan academics and administrators have their own viewpoints. Not surprisingly, they believe that the professors and administrators of the various colleges should decide programs and course offerings. In their view, the MOE should gradually reduce its influence over such decisions. The important point is that in both of these cases, the assumption is that consumers in the market should not be able to determine for themselves which kinds of programs and courses are most desirable. Given these views, it would be a mistake to apply the usual procedure in economics of determining allocative efficiency according to whether the service satisfies consumer demand. Instead, we should assume that the efficiency standard is determined by some other designated agency or group. Given this assumption, we can ask whether the current system designed to achieve efficiency according to a standard that might be established by such an agency or group.

Incrementalism and Red Tape

            Even though we do not know exactly which standard such an agency or group might impose, we can use our knowledge of the existing system to make a judgment whether it is likely to be able to adjust to the standard. Consider how decisions about programs and course offerings are currently made. Under the current system, the MOE decides how many students will be allowed to study each subject. For example, it may decide that X slots will be available for engineering students and Y slots will be available for public administration students. How does it decide? By far the most important force influencing decisions is bureaucratic inertia, as evidenced by incrementalism. This refers to the tendency to approve the same subjects and programs that were approved in previous years. Of course, new programs and courses are approved and even new colleges are opened. However, the bureaucratic red tape and delays involved in changing the status quo are extreme. These factors lead new programs and courses to mainly be proposed by individuals who attach a low value to their time (or who have been able to obtain assistants for research grants that they can put to work on the projects) and/or who plan to spend a long time in the program or teaching the course. Some of these people may be capable of efficient innovation. However, many innovative teachers and administrators of colleges simply avoid the process.

            Why incrementalism? The answer illustrates one of the fundamental characteristics of bureaucratic decision-making in a democracy: risk aversion. The importance of this is apparent when we consider the likely impact on the MOE of radical change. Suppose that the MOE decided to eliminate the mainland Chinese history and philosophy program in favor of a more cosmopolitan approach to those subjects or in favor of a program that focuses on the island of Taiwan. Existing programs and professors would have to be replaced by new ones. The arrival of competent professors and the institution of new programs to teach these classes would disturb existing relationships and programs throughout the college. Administrators who had achieved their positions through skillful politicking would have to build new ties. Changes of this sort are practically always resisted by established bureaucrats. Such resistance might begin with an appeal by existing teachers, backed by nearly universal support from school administrators, teachers, and students. MOE officials would have to respond or face public criticism through the media. Since major politicians would have to take a stand, the MOE would face a series of inquiries from both the ruling and opposition parties. From the viewpoint of the MOE, the best that could happen is that a major controversy would be decided in its favor. The worst is that the director and top staff would be required to resign. To avoid the risk associated with such public scrutiny, the MOE director would be wise to act conservatively.

            Risk aversion is often associated with avoidance of responsibility. By adopting the same policies as past MOE directors, the current director can blame those directors for any defects in the current system that are discovered during his tenure. But if he adopts a new policy, only he can be held responsible.

            It should be evident that bureaucratic red tape and incrementalism are contrary to what one would hope and expect on the basis of allocative efficiency considerations, no matter what standard the MOE established. Even if we agree that the MOE experts are more capable of deciding programs and courses than the students and their parents, the current system is inefficient. Whatever reasons there may be for the MOE to decide how many students to support in each subject area, those reasons should not be based on the subjects and fields that colleges have offered in the past. They should be based on some reasoned judgment about current needs. The current method of making decisions based on incrementalism should be abandoned. In addition, red tape should be minimized.

Substituting Consumer Preferences for Those of the MOE

            Beyond this, there is a good reason to take the MOE completely out of the role of deciding which studies to subsidize. One of the principles of democracy in a market economy is that unless there is some reason to believe that there is "market failure," citizen-consumers are the best judges of which specializations are most productive in serving their demands. Just as it would be unwise to appoint a ministry to decide the foods that people should be served at restaurants, so also does it seem unwise to employ a ministry to decide the programs and courses to be taught at colleges.

            The voucher system is an alternative. Suppose that the MOE selects which students to subsidize according their scores on a single standardized test. Then it gives them vouchers, allowing the students, in consultation with their parents and advisors, to choose their colleges, programs, and courses. In this event, the students and parents could choose a private college that they believed would best satisfy their demands. Assuming that the MOE monitored the supply, they could avoid abuses. Under the voucher system, the presidents of private colleges would have an incentive to supply the subjects that voucher recipients most demanded. In pursuit of profit, they would quickly replace obsolete and superfluous programs and courses with ones that are regarded as most useful by parents and students.

A Greater Potential for Fraud

            A disadvantage of the voucher system relative to direct supply and contracting is that the government has less control over the quality of the service supplied and the conditions of supply. This leads to a greater potential for fraud. Because some recipients may prefer money or some other commodity instead of the service desired by the a ministry or by the legislature, they have an incentive to trade the voucher for cash. For example, instead of taking her child to get an inoculation, a voucher-receiving mother may decide to exchange her $1000 voucher for, say, $500 in cash. If we disregard various costs of transfer, both the mother and the supplier of the inoculation would make a $500 gain, even though no inoculation was supplied. Another possibility is that ineligible people would deceive government administrators into giving them an undeserved voucher. For example, a family whose four-year old child was living with her grandparents abroad, where inoculations are free, might try to receive the inoculation voucher and then exchange it for cash. A third possibility is service substitution. Instead of supplying an inoculation, a supplier may substitute a different service, such as a $500 bottle of cosmetics.

            In the absence of watchdogs and strong penalties for fraud, both the recipient and the supplier of the service have an incentive to circumvent the government's intentions. To deal with this problem requires proper administration. The government would want to pass laws and to set up monitoring mechanisms. To monitor, it could establish an agency to determine whether a particular supplier does indeed supply the service according to specifications. Such an agency could play a reactive role by accepting complaints and then appointing agents to investigate them. Or, it could play a pro-active role of carrying out unannounced, random inspections and by using undercover agents to pose as either suppliers or consumers wanting to benefit from fraud. To deter fraud by consumers, the government can conduct follow up investigations of voucher recipients. A random sample of voucher recipients could be further investigated in order to verify that they deserve to receive the vouchers, according to the government's standards. In the case of higher education in Taiwan, the MOE would presumably be the watchdog agency.


            In summary, a broad outline of a voucher system is as follows. First, a target group is identified as eligible voucher recipients. Second, an amount to be given to each recipient is decided. Third, money is set aside in the government budget to finance the vouchers and other parts of the program. Fourth, an agency is appointed to administer the program. The agency causes the vouchers to be printed and distributed to eligible recipients. It also takes on the responsibility of detecting and prosecuting fraud.


3. How the Voucher System Would Help Solve

Taiwan's Higher Education Problems

            In part one, we identified three problems with Taiwan's Higher Education: (1) unsatisfied demand, (2) low quality, and (3) lack of parental and student choice. We now try to show methodically how the voucher system, as described in part two, would deal with these problems.

Satisfying the Demand for Education Services

            Currently, students who fail to meet the MOE cutoff standards for public and “private” colleges have only two alternatives: go abroad or study for the next MOE test. Under the voucher system, opportunities to study in Taiwan would be widely available to such students, since the MOE would drop its requirement that a student must pass the MOE test to attend a college. To see how students who failed to pass the MOE test would be dealt with by colleges, let us divide them into two classes: (1) those who do not meet the admission standards of accredited colleges and (2) those who do.

Students Who Do Not Meet the Admission Standards of Accredited Colleges

            We begin by considering a student who does not meet the admission standards of any accredited college. If the student was wealthy enough, she still might persuade the college to admit her by paying a high tuition. In this case, the college could afford to give her special treatment, perhaps by setting up special classes or tutoring sessions for her. But consider a student who is not so wealthy. Then the accredited college would be unwilling to supply special services. And in the absence of these, it would be unwilling to admit her. The reason is that if adopted a policy of mixing students of her type with more competent students, it might risk losing the latter to other colleges that limited their enrollments to better students. This would mean a loss of voucher money. In addition, if a college admitted too many less competent students, it might risk losing its MOE accreditation. However, there would still be opportunities at unaccredited colleges.

            It is possible, of course, that students would be wrongly judged by the accredited colleges. By attending an unaccredited college, such students would have an opportunity to both prove their competence and to obtain a college education. Also, a college that had developed a specialized skill of identifying good prospects that were misjudged by accredited colleges would have an opportunity to prove its worth and perhaps to gain future accreditation.

Students Who Meet Accredited Colleges’ Admission Standards

            Now consider a student who is judged by an accredited colleges to be competent even though he did not pass the MOE cutoff point for receiving a voucher. Assume that the student is willing to personally pay the voucher amount. If a college president believed that the admission of particular student would not reduce the overall quality of the education service, he would have every reason to accept the application. His only worry would be that the MOE accrediting body would regard the admission of such students as evidence of poor quality. The MOE would have to be careful to recognize the limitations of its testing.

            We see, then, that the demand for education services for those who failed to meet the MOE standards would be widely available. Also both accredited and non-accredited colleges would have an incentive to identify competent students that the MOE testing had excluded.

Improving Quality

            Competition among colleges would improve quality. To understand how, we must recognize each college's desire to earn profit. To earn a profit, a college would have to please voucher recipients and other demanders of education services (i.e., the consumers of education). Since many voucher recipients have a high demand for high quality, colleges would compete in the quality dimension. Other things equal, a college that supplied high quality education would obtain more vouchers -- thus higher profit -- than its rivals. Such competition would be dynamic. Each college would have an incentive to find ways to improve its quality since doing so would give it an edge over its competitors. Thus, we would not only expect high quality colleges to emerge, we would expect a continuing effort to improve quality.

            Would an accredited college retain voucher-paying students whose performance fell far below that of their peers? Ordinarily it would not. To allow such students to continue would reduce demand for the college’s services by better students. In addition, if it catered to too many poorly-performing students, the MOE might take away the college’s accreditation. The exception would be for students who were willing to supplement the voucher amount with a substantial amount of their own money. In this case, some accredited colleges might be willing to take these risks. Thus the current problem of poorly performing students continuing to received the MOE subsidy would be solved. It would be in the interest of most accredited colleges to fail students who do not meet its standards.

            The quality might improve in another way. Instead of expelling students who do not meet the standards, a college might offer to allow the students to stay, so long as they (1) pay additional tuition to supplement the voucher and (2) participate in remedial education supplied by the college.

The Need for Monitoring

            The quality level of accredited colleges depends in some measure on MOE monitoring. The reason is that without supervision, colleges may decide to substitute cheaper non-education services. Students who receive vouchers demand other services besides education, such as a pleasant environment for studies, entertainment, travel, good housing, high-quality food, fashionable clothing, and other things. Modern colleges sometimes provide some of these services because the services complement their education services. In other words, it is cheaper for the college to supply them than for outsiders to do so. For example, a heated and air-conditioned library makes it more convenient for students to borrow books and to study there. Otherwise, the students might have to rent space at a local private reading library. Colleges also typically provide cafeterias, living accommodations, and recreational facilities. Sometimes these are cheaper when provided by the college; sometimes not. In the absence of monitoring, some colleges may increase the amount and quality of these other services so much that they substantially reduce the quality of their education services. For example, we can imagine that a college would offer a foreign studies program in which it supplied voucher-eligible students with travel to popular tourist destinations. Similarly, it might construct a campus designed to provide entertainment like free video games, movies, and concerts.

            It is the task of the MOE to decide which services to allow under its accreditation program. The MOE wants to avoid having vouchers used for services that do not facilitate education in the strict sense. However, it must give colleges some freedom to experiment with different methods of improving their education services.

Parental and Student Choice

            Perhaps the strongest argument for the voucher system concerns the additional choice that it would give to consumers of education services. The demands for educational services are typically derived from other demands, such as the demand for skills to earn money after graduation. In a planned socialist system, the education ministry would take it upon itself to choose which services would be supplied. Its judgments would determine how the education subsidy would be allocated among the different services. For example, it might decide that because it predicted a future shortage of doctors, it would increase its subsidy to national medical colleges and reduce it to other colleges. As pointed out above, decisions about which services to supply in Taiwan are currently made by particular active faculty and administrators, subject to the MOE policy of incrementalism and red tape. In a true democracy, the voters and taxpayers should make such decisions. Assuming that the people want to subsidize education, Footnote the best way to achieve this is through a supervised voucher system.

            How would the voucher system deal with an expected shortage of doctors? The answer is that if students and parents expected a shortage, they would, in a market system, also expect the salary paid to doctors to rise relative to other occupations. The demand for a medical education would rise. Accordingly, some voucher recipients would shift their voucher choices to colleges that provided medical education and away from other colleges. This would give competing colleges an incentive to shift some of their resources away from other kinds of education and toward a medical education. Thus the supply of medical education would rise to meet the higher demand; while the supply of other education would fall due to the relatively reduced demand. Indeed, managers of private colleges would have an incentive to be ready for the higher demand at the precise time that it occurs.

            Thus parent and consumer choice is catered to by means of the decentralized decision-making of the private colleges. Changes in demands for different types of education are quickly met by profit-oriented colleges, whose goal is to attract voucher-eligible students.

4. Putting the Theory into Practice

            The long-term goal of a voucher system in Taiwan higher education is to completely eliminate all direct payments to colleges. Except for employees of the agency that administers the voucher system and monitors the colleges, presumably the MOE, nobody who supplies higher education services in Taiwan should be paid by the government or by a government agency. In addition, the government should own no college assets, such as land or buildings. All teachers, administrators and maintenance personnel should be employed by private college companies. Those same companies would individually own or lease the privately-owned land and other physical assets used in higher education. Footnote

            The system as it would ultimately exist is fairly easy to describe. Money is set aside in the budget and the MOE presides over the money's distribution. The MOE establishes a target population of students and a voucher amount to be given to each student. It accredits colleges. And it sets up procedures to detect fraud. The criminal justice system would also have to aptly prosecute, judge and punish voucher system cheats.

            However, before this ultimate system can be implemented, a transition must occur. Since the Taiwan government already has in place a system for supplying higher education and since it owns most of the assets and employs most of the human resources, the transition must be a critical consideration of any voucher system proposal. Constructing a viable transition requires sensitivity to those who stand to lose as a result of the privatization. We shall discuss this in length in part five. But first, we deal with the easier parts of the system.

Identifying the Target Population

            Identifying the target population is perhaps the easiest part of the voucher system to implement in Taiwan. The current system identifies a target population by means of a series of national examinations (JCEE). Using a general examination comprised of parts of the several examinations that are already being used, the MOE would choose a passing score such that the number of students exceeding that score would equal the number chosen to be subsidized, in accord with legislative mandates and the size of the budget. Each student would receive a voucher with the same face value.

            If the MOE wanted a graduated, or incremental subsidy, it would have to construct a more complex method. For example, it might want a two-tiered system that would correspond to the current system of highly subsidized national college students and lowly subsidized private college students. It could divide subsidized students into two groups. The first group would consist of those whose JCEE scores are high enough to be admitted to a college under the current system. They would receive the full value of the voucher. The second group would consist of students whose JCEE scores are lower but still higher than some cutoff point. They would receive, say, fifty percent of the full value. The system could be made more complex by dividing the students into additional groups. Footnote

Deciding the Voucher Amount

            Any decision regarding the amount of the voucher must be arbitrary. In 1996, the MOE listed the amount of money spent per undergraduate student enrolled in national private colleges, taken together, at $198,611 (Ministry of Education, 1997b, p. 14). This is an important index and may be recommended as the voucher amount. Footnote

Setting Aside the Money

            The Government already allocates enough money to the MOE to finance the voucher system proposed here. The MOE uses some of that money in its own administration, while it disburses the rest to the colleges for operating expenses. Administration of the new MOE is discussed below. The MOE's function of disbursing money to colleges would undergo radical change. Instead of disbursing it piecemeal depending on the particular schools, programs, departments, projects, etc.; the MOE would have the simple task of exchanging a fixed sum of money for each voucher turned in by the college. This would drastically simplify the MOE's task and should lead to substantial MOE downsizing and cost savings.


            The MOE would presumably have the task of administering the voucher system. There are three parts to administration: (a) distributing the vouchers and disbursing money to the colleges that turn in vouchers, (b) monitoring and punishing fraud, and (c) determining the eligibility of applicant suppliers to participate in the program (i.e., accrediting colleges).

Distributing the Vouchers and Paying Suppliers

            As we have pointed out in the previous section, a two-tiered or graduated voucher system could easily be devised, with the face value of the voucher reflecting the relative ranking of the student on the national examination. The MOE would have the relatively simple tasks of choosing the tests, determining eligibility according the test performance, and then paying out the money when suppliers turned in the vouchers. The voucher system does not seem to introduce any new problems that the MOE does not now face in its testing. Indeed, the testing and selection process under the voucher system would be less complicated.

Monitoring and Punishing Fraud

            The deterrence of fraud would require a major rethinking of the means used by the MOE to control college education. Up to now, the mode of control has been almost totally bureaucratic. To obtain funds for any purpose, an individual in a college must fill out forms and send accompanying documentation, describing in detail how the funds are to be spent. Later the applicant must provide receipts for all expenditures. Applications are handled according to rigid rules, so that there is always a paper trail to facilitate any future investigation of fraud. Such procedures would become more or less obsolete. The only documentation required would be the one that demonstrates an identity match between the student who scores above the cutoff point on JCEE and the student who is issued the voucher. The MOE would have to verify that the voucher turned in for redemption by a college matches the one issued to the student. The task of determining whether a college is qualified would belong to the accreditation aspect of the revamped MOE, to which we now turn.

Accreditation: Selecting Qualified Colleges

            Currently, no specific department in the MOE is charged with determining the quality of colleges. As a result, a means of accreditation would have to be developed. An advising committee and a new department would have to be created for this mission. Local scholars and experts as well as accrediting professionals invited from abroad can serve as the committee members to devise principles and standards for proper accreditation. The accreditation department would examine each proposed college in order to determine whether it meets the minimum standards for initial, provisional accreditation. Initial accreditation standards could be established on the basis of a desire to provide a supply of education for X number of redeemable vouchers. As we shall see, it is desirable from the standpoint of making an easy transition to the voucher system to maintain existing colleges intact. Thus, existing colleges would be automatically accredited for, say, the first three years. The main problem faced by the accrediting department during the three years would be to decide on new applicants. Such new applicants should be subject to no higher standards, at first, than the existing institutions. To establish such standards, it would be necessary to thoroughly examine and rank existing colleges. Footnote

            It should be recognized that much of the accreditation problem would be solved through the choices of parents and eligible students. Broadly speaking, these consumers of education services have an incentive to choose colleges that they expect to offer higher quality service for the money than their competitors. However, this is not their only motivation. Some students may prefer a college in which they can play rather than study. Also, as mentioned above, there is an incentive to commit fraud. Accordingly, some amount of government accreditation and monitoring is necessary.

            Whether a college passes the test of accreditation should not depend on the size of the salary paid for administration, faculty, or maintenance. It should also not depend on the methods used for education. The primary question to be asked by the accrediting body is this: Is there a reasonable chance that students in this college will receive an adequate education service? One of the virtues of a well-constructed voucher system is the incentive it gives to private colleges to experiment with new methods of providing service. We really do not know which methods of providing service are best and, even if we did, we would expect them to change due to the innovations of cost-minimizing entrepreneurs. We know from our experience with other government-run businesses that profit-oriented colleges are much more likely than government-run colleges and non-profit colleges to adopt cost efficient and high quality education without corresponding cost increases. Thus, the accrediting body decided upon by the MOE would want to make a special effort to avoid forming preconceptions about what constitutes a good education. Rather than strict accreditation rules, the best system is one that encourages dialogue between representatives of colleges and the accrediting body. "Make your case, the accrediting body should say, and we will consider it with an open mind."

5. Making the Transition to the Existing System

             The only research on the voucher plan for higher education of which the author is aware devotes the bulk of its analysis to the problem of transition. Sontheimer (1994) believes, correctly in the author’s view, that the most important factor to consider in discussing a shift from a centralized education system to a voucher system is the potential political opposition. The problem is that although there are likely to be huge net gains from the voucher plan, some individuals are likely to be big losers unless special provisions are added to the plan to provide a cushion, or compensation plan.

            Sontheimer recognizes that most of the teachers, administrators, and employees of existing government-subsidized colleges are firmly entrenched in their current positions. With this in mind, he suggests that a proposal to adopt the voucher system could not receive enough votes from a democratic legislature unless it includes an additional transition plan that would potentially make the employees of existing colleges better off after the shift. His plan is to allow employees to collectively buy the capital assets of existing government colleges at subsidized prices, provided that they do not resell them for some specified period of time.

            We can call this plan the Sontheimer plan. It has two characteristics. The first is a transition period during which the new plan is gradually phased in. The second is that existing employees of a college are offered a "preferential employee stock ownership plan (PESOP)." Here we discuss the PESOP.

The Preferential Employee Stock Ownership Plan

            All of the national colleges in Taiwan have appreciable capital assets. These include land, buildings, furniture, auxiliary administrative and teaching equipment, extensive libraries, maintenance equipment, and so on. All of this property is owned by the government. Privatization requires that it be sold off.

            If citizens wanted to obtain the maximum amount of money for such items, they would probably want to auction them off, piece by piece, to the highest bidders. Perhaps one company would bid for all of the items associated with a particular college as a block. It would plan to use them, along with newly hired employees, to supply its own brand of higher education in exchange for vouchers and tuition. On the other hand, the assets may be more efficiently used in a different enterprise. An example might be a national college with old buildings located on valuable city land. A company may want to buy the land in order to build a high-rise office building or hotel. Sontheimer argued in favor of a different procedure for disposing of the assets. He recommended PESOP. Under PESOP, employees would be placed first on a priority list of prospective buyers. They would be permitted to collectively buy the capital assets as a block, at a discount, provided that they continue to operate the college for a given period of time. He recommended ten years. Footnote If the employees did not want to buy the assets at the discount price, the assets should be sold to the highest bidders.

            According to Sontheimer, PESOP would work as follows. After the decision is made to adopt the voucher system, the capital assets of each college would be appraised by one or more reputable accounting firms. The resulting appraisal would provide potential bidders with a reference for their bidding. The employees would be offered the assets if they agreed to pay only 20 percent of the appraised value.

            One might ask how employees could raise the funds to pay for even 20 percent of the appraised value. The answer is that they would form a corporation and borrow the money by using the college assets as collateral, or security, for the loan. Since the appraised value is five times the size of the loan and since they would be banned from selling the assets for ten years, Footnote a bank would probably be most willing to lend the money. If the employees fail to meet the payments, the bank could foreclose on the property.

            How would the ownership and control by employees over the capital assets be organized? Employees would first have to form a corporation and issue stock. Then they would have to decide on some kind of stock distribution plan. They might apportion shares on a one-person-one-share basis. Or they might apportion them in a way that is proportional to the salary earned prior to the adoption of the voucher system. Or they might use some other system based on current position or rank in the college hierarchy. Whatever they decide, they would have to appoint a board of directors who would appoint a president, or chancellor, to manage the college in the stockholders' interests.

            With collective ownership of the private college corporation, employees could receive payment in two ways: (1) wages or salary and (2) stock dividends or capital gains if the stock rises in price. For the first ten years, no tangible capital gains would be available, since employees would be barred from selling their stock. This means that if a college that performed extremely poorly relative to others, the employees’ total income might be low. However, an employee could quit and take another job. Or he could use the stock certificates as collateral for his borrowing. (The lender would, of course, be barred from taking control of the certificates until the end of the ten-year period.

6. Conclusion

            This paper has proposed a voucher system for higher education in Taiwan. It first pointed out some major problems in the higher education under the current centralized system and how such a system has become incompatible with recent political and social changes. It then illustrated the general nature and functions of a voucher system and showed how it could deal with the recognized problems. Next it showed how to implement the system in Taiwan. It devoted the last part to a key issue in implementing the system -- achieving acceptability by the current suppliers of education services. It described a preferential employee stock ownership plan that could achieve this goal.

             The essential question addressed in this paper is whether a voucher system would result in an improvement. Through an analysis of what we called supply and allocative efficiency, we argued that the answer is that it would. In the opinion of the author, the voucher system, based on the spirit of choice, would better meet the demands of students and parents in modern democratic Taiwan. Furthermore, the system tends to decentralize and deregulate the current educational system and thus provides a solution to the existing problems in Taiwan's higher education.



History of the Voucher Proposals in Education

            As mentioned above, the concern of this paper is the voucher system as it applies to the higher education in Taiwan. However, the voucher system is much broader than this. It applies to any good or service that government decision-makers want to encourage people to buy. Its most extensive treatment as a method of encouraging education has been in the U.S., where the system has been proposed as a means of reforming lower education. Although the literature on this subject is only indirectly and partly relevant to the case in which we are interested, it is worth reviewing.

            The voucher system idea as a replacement for public schools was introduced in the U.S. over forty years ago (Friedman, 1955, 1982, 1995; Friedman & Friedman, 1980; Peacock & Wiseman, 1964; Jencks, 1970; Sontheimer, 1994; Hakim, Seidenstat, & Bowman, 1994; Bast, Harmer, & Dewey, 1997). The first and most influential educational voucher system was proposed by the Nobel-winner economist Milton Friedman in 1955. As originally presented, the system was mainly designed for elementary and secondary schools. Friedman proposed that instead of using tax revenue to supply schools directly, the government should give parents of each school-age child a voucher. Parents could then send their child to any school of their choice as long as that school met certain basic standards relating to safety and curriculum. Parents would use the voucher as payment for tuition to that school. After collecting the vouchers, the school would redeem them for cash from the government. Under the Friedman proposal all schools would be private, although the government would require them to meet certain minimum standards. Also schools would have the right to reject an application and could charge tuition and fees in addition to the voucher amount.

            Since its emergence in the 1960s, a number of modifications and variants have been proposed. These have been designed mainly to address the financial difference between the rich and the poor; racial problems between black and white (i.e. segregation issues in education); and the special problems arising from the U.S. Constitution, which requires a separation of church and state. Among the variants is the plan delineated by Peacock and Wiseman (1964), who argued that vouchers should not be of uniform value. Children from lower-income families should receive higher-valued vouchers.

            The earliest voucher program in the U.S. was implemented in 1954 for reasons that had nothing to do with economic efficiency. Four states -- Virginia, Mississippi, Alabama, and Louisiana -- used the voucher program in response to the Brown vs. Board of Education racial desegregation verdict by the Supreme Court. The program enabled the children of white parents to use government funds to send their children to existing and newly-created private schools designed to serve only whites. The program was later declared unconstitutional and stopped. In 1962, Friedman proposed his educational voucher plan to oppose the increasing monopolization of education by local public school districts.

            In 1969, CSPP (the Center for the Study of Public Policy) published its final report: "Educational Vouchers: A Report on Financing Elementary Education by Grants to Parents." This report presented various models for voucher programs, analyzing their advantages and disadvantages. It proposed that an experimental voucher program be put into practice (Jencks, et al., 1970). With parental agreement, the proposal was implemented in Alum Rock, a school district near San Jose, California. Alum Rock at the time was a school district with 15,000 elementary students, with 55% Hispanic and 12% Black. Under the program, each student had the right to choose which school to attend (Jencks, et al., 1970). However, the scheme was very restrictive, involving only public schools. In addition, it imposed a number of mandates on the schools participating in the experiment.

            In 1990, Wisconsin became the first state in which the legislature instituted a constitutionally acceptable tuition voucher plan (the Milwaukee Voucher Plan, see Lieberman, 1994). In 1994, a bill that would have instituted a voucher system in California was voted down due to the strong opposition from the teachers' unions (Lieberman, 1994).

            The studies and programs discussed above are applicable only to the elementary and secondary education in the U.S. Only one theoretical study has been done to explore the voucher system in higher education.(Sontheimer, 1994) The main reason for the lack of studies in higher education is that in the countries where such studies have been done, governments have played a relatively smaller role in higher education than in elementary and secondary education. Accordingly, although writings on the theory underlying the voucher system are relevant to our concerns in this paper, reports of cases in which the voucher system has already been applied, mainly elementary schools in the U.S., are not.


Bast, J. L., Harmer, D. & Dewey, D. (1997). “Vouchers and educational freedom: A

debate.” Policy Analysis. 269 (March 12): 1-48.

China Times (1997). “The Failure Rate in College Education” (June, 12)

Coleman, J., Hoffer, T. & Kilgove, S. (1982). High school achievement: Public, Catholic and Private Schools Compared. New York: Basic Books.

Cox, D. & Jimenez, E. (1987). Private-public differences in secondary school performance: The role of selection effects in Colombia and Tanzania. Washington D.C.: World Bank Education and Training Department.

Friedman, M. (1955). “The role of government in education.” In Robert, S. (ed.) Economics and the Public Interest: (pp. 124-29) New Jersey: Rutgers University Press.

Friedman, M. (1982). Capitalism and Freedom (new ed.). Chicago: The University of Chicago Press.

Friedman, M. (1995). “Public schools: Make Them Private.” Washington Post (February 19). Reprinted as Cato Institute Briefing Paper No. 23: Washington, D.C.: Cato Institute.

Friedman, M. & Friedman, R. (1980). Free to choose: A Personal Statement. New York: Avon Books.

Hakim, S., Seidenstat, P., & Bowman, G. W. (eds.). (1994). Privatizing education and educational choice: Concepts, Plans, and Experiences. Westport, Conn.: Praeger.

Hanushek, E. (1986). “The Economics of Schooling: Production and Efficiency in Public Schools.” Journal of Economic Literature. 24 (3): 1141-77.

Jencks, C. et al. (1970). Education Vouchers: A Report on Financing Elementary Education by Grant to Parents. Cambridge, MA: Center for the Study of Public Policy.

Jimenez, E., Lockheed, M. & Wattanawaha, N. (1988). “The Relative Efficiency of Private and Public Schools: The Case of Thailand.” The World Bank Economic Review. 2 (2): 139-64.

Lieberman, M. (1994). “The School Choice Fiasco.” Public Interest: 114: (17-34).

*Liu, P. W. (1996). Education vouchers and Education Subsidy in Hong Kong. Hong Kong: Hong Kong Institute of Asia-Pacific Studies, The Chinese University of Hong Kong.

Bureau of Statistics, Ministry of Education, R.O.C. (1997a). Education Statistics of the Republic of China Taipei, Taiwan: Taiwan Ministry of Education.

Bureau of Statistics, Ministry of Education. (1997b). Key Indices in Education Statistics. Taipei, Taiwan: Taiwan Ministry of Education

Mitchell, W. C. & Simmons, R. T. (1994). Beyond Politics: Markets, Welfare and the Failure of Bureaucracy. Boulder, Colorado: Westview Press.

Peacock, A. T. & Wiseman, J. (1964). Education for Democrats. London: Institute of Economic Affairs.

Sontheimer, K. C. (1994). “Privatizing higher education.” In Hakim, S., Seidenstat, P., & Bowman, G. W. (Eds.). Privatizing Education and Educational Choice: Concepts, Plans, and Experiences (pp. 145-159). Westport, Conn.: Praeger.

Gunning’s Address
J. Patrick Gunning

Professor of Economics/ College of Business
Feng Chia University
100 Wenhwa Rd, Taichung

Taiwan, R.O.C.
Please send feedback  

Email: gunning@fcu.edu.tw

Go to Pat Gunning's Pages

Mirror Site of Pat Gunning's Pages