Introduction
In the Venezuela of today there is concern being shared by different sectors of society with regard to the dramatic status of the country's education. Although the educational system's structure met the objective of providing universal access for elementary education, it looks as if the current system were incapable to produce the results required by a modern society. The fundamental problems, in our opinion, does not lie in the amount or resources allotted to the sector, but rather on the structure of a system lacking the mechanisms to guarantee an efficient and fair investment.

The Venezuelan public educational system assigns resources on the basis of the input, rather than on the services provided. There is not in the current system any relation between the quantity and quality of the education being provided by a school and the amount of resources it gets. The purpose of this paper is to present thoughts on an alternate mechanism to assign resources known as the school vouchers. Although vouchers would not solve all the problems affecting our damaged educational system, we believe that a better structure of incentives should propitiate a more positive dynamics than the current one.

In this short paper we begin by offering, under quite general terms, the results having been produced by the Venezuelan educational sector, the level of expenditures having been assigned to the sector and some of the current system's most evident problems. We then present the main elements of a voucher system. Then we discuss the pros and cons of using this mechanism for resource allotment. The fourth section presents a couple of recent international experiences with the use of the voucher mechanism. The fifth section discusses some considerations on the application of school vouchers to the Venezuelan educational system, ending with the conclusions.

I - A brief description of the Venezuelan educational system
Between the years 1960 and 1980, the Venezuelan educational system reached massive expansion. Currently, 97 percent of the children have access to basic education and approximately half of the 5 year old children are in kindergarten. Student participation at the level of higher education is relatively high if we compare with other Latin American countries --roughly 27 person of the young between the ages of 18 to 27). These indicators, however, are far from reflecting a successful educational system; the rate of recurrence at school level is high, almost 30 percent of registered children are repeating. What is worse, from the total of children having access to the educational system, only 47 percent go through the nine levels of elementary education. This has as a consequence that student participation at high school level is only at the level of 23 percent of the young between 15 and 17 years of age, quite a low percentage as compared with the average of Latin American countries,

The problem does not only lie on the rates of recurrence and conclusion at the elementary education level, but also on the quality of education. The Venezuelan educational system has not been able to reach satisfactory quality levels. In a recent international test aimed at measuring the reading abilities of a group of students from 32 countries, the Venezuelan nine year old children got the lowest percentage and those at the 13 year old level placed themselves only above the children of Nigeria, Zimbabwe and Botswana. In the tests, only 5 percent of the Venezuelan students performed above the international average.

There are discrepancies also between the access to education of children according to their income level and their geographical location. In the year 1991, children between the ages of 15 and 19 from 10% of the poorest homes had less than 13% school participation than those from "non poor homes". What is even worse, in rural zones the proportion of children within this age bracket actually going to school was less than half of that in urban zones.

According to 1992 figures, Venezuela is one of the Latin American countries that spends more in education. Education expenditure by person is the region's second --something to be expected by reason of the per capita income. However, due to the high proportion of the educational budget allotted to higher education, in terms of expense by elementary and high school student makes of us the country that spends the least in the region (see table I.1)

Sources: CEPAL, UNDP, UNESCO.

* PPA means that the dollar figure was adjusted under purchase power parity factor.

Although the Venezuelan government spends substantially more than other Latin American countries in education, the net school attendance levels at the elementary and high school bracket are way below those of countries with a per capita income similar to Venezuela's. (See table I.2)

Table I.2 - Net School Rates of Attendance, 1990

Source: UNESCO

To a great extent, the problem is that the educational expenditure is inefficient, a great proportion (almost 95%) is tied to the payment of teachers and administrative staff, leaving only roughly 5 percent for infrastructure, teaching material and teacher training. What is more, of this 5 percent a substantial share goes to infrastructure repairs. This leads to the consequence that most of the children have no access to educational materials, with the learning process entirely depending on the teachers' ability and quality.

Approximately 50 percent of the educational budget is assigned to higher education and students at this level are only 10 percent of the country's total student population.

The low performance of Venezuelan students and a relative high expenditure shows that the use of educational resources is highly inefficient. It thus becomes necessary to look for alternatives in order to improve the allotment or resources, the fairness and quality of education.

II - What is a voucher system?
Seen in a wider context, the vouchers are presented as an alternative within a set of market-based reforms of the educational system. An educational voucher is some kind of certificate by means of which the governments binds itself to make a payment to the school chosen by the beneficiary child's parents. The general idea is as follows: The parents of children in school age get a voucher from the government; this may be used to pay for tuition expenses or other admission charges at any school participating in the program. These schools may be private or public. The parents present their vouchers to the schools chosen by them. Then, each school presents the vouchers to the competent government agency and gets a corresponding sum of money.

In practice the educational voucher system has numerous versions. Some governments decide to provide them to any interested family, while others decide to give them only to low income families. There are even cases of educational voucher to quite specific groups, aiming at emphatically promoting education in them. There are interesting examples in Guatemala and Bangladesh; these countries use the vouchers for girls at elementary school ages. In some cases, the government inspects the schools getting the vouchers in order to ensure that minimum standards prevail. Governments may impose regulatory measures also for enrolled schools such as requiring that teachers in them have a special teaching license --generally government issued.

As to the vouchers' monetary value, there are also several modalities. Some governments assign a fixed amount, while others establish a value being inversely proportional to the family income. Governments issuing fixed amount vouchers may assign a value to them being the equivalent of the average expense by student in public schools or a lower one.

From the point of view of the procedure used to establish the voucher system, some governments deliver the vouchers directly to the parents while others do it to the schools chosen by the parents of the children participating in the program.

III - Arguments in favor and against an educational voucher system

A. Arguments in favor:
The vouchers' most immediate benefit is that it provides the children's parents with an opportunity to choose. The parents, given the role they play in their children's education, may be considered as the educational services' consumers. Under a voucher system, the government is financing the education consumers (parents) instead of the education providers (schools). In words of economists, the vouchers are a demand financing system, not an offer's.

One of the strongest arguments in favor of the vouchers is that they promote competition, The idea is that, when the voucher system is introduced, the "good" schools attract many students and will collect many vouchers allowing them to keep prospering. The "inferior" schools will be avoided by the parents, will get fewer funds and accordingly will have incentives to improve. These schools must improve if they do not want do disappear. Given the fact that public schools are frequently monopolies, the voucher system may introduce competition entailing lowering of costs, improvement in the quality of education and promotion of innovation.

A third benefit from the educational voucher system is that it promotes greater access to private schools. Gary Becker (1995), among others, argues in favor of a selective voucher system allowing low income classes to have access to private schools. Part of the argument is base on the fact that North American private schools have proved to have better performance than public ones and that students with a disadvantaged background "are those who take most advantage from private schools". Studies in developing countries have shown the relative superiority of private schools over public ones (see Jiménez and Lockheed, for example).

Another argument in favor of the vouchers is that they help in changing the relationship between the children's parents and schools. When being able to choose the school for their children, parents may begin to see the schools as providers of education as a good and are in a better position to demand better service.

B. Arguments against:
A most common argument against the vouchers is that one should not expect that parents adopt optimal decisions for their children (Carnegie Foundation, 1992; Levin, 1991; Bridge, 1978). Some of this argument's opponents hold that parents will learn to take this kind of decisions with some experience. Others argue that there are empirical studies showing that parents with modest educational levels adopt rational decisions for their children (Fossey, 1994)

The possibility of corruption and fraud in the voucher system has raised some skepticism among some experts in this issue. If the voucher system were to be implemented addressed to a specific group (children with lower resources, for example), the possibilities of corruption would sharpen in view of the incentives that the families would have not to reveal their true financial situation.

Teachers' unions and managers have sustained that a voucher system would destroy the public schools system. The argument is that, once the vouchers have been introduced, the middle class people would abandon the public schools migrating to the private ones, where low income individuals and minorities are discriminated. Thus, the latter would be left abandoned in a system of decadent public schools with continuously lower budgets (Krashinsky, 1985). This argument supposes that public schools are not able to react to the competition generated by private schools. From the premise that the objective of an educational reform is to provide a good education for all the students, said polarization would be damaging.

Another argument against the voucher system is that it would not introduce the benefits of competition in those rural zones where there is hardly a public school. In these zones, that system's benefit would be to act as an incentive for existing schools in their keeping the children within the educational system, reducing desertion.

Last, when consumers are able to choose, they may indicate to the organization --the school in this case-- the need for change. This is quite positive; the signal, however does not necessarily tell the organization how to change.

IV - The international experience

The purpose of this section is to present a summary of the existing experience in the use of voucher systems in elementary and high school education. During the nineties some 20 experiences have been reported on a world level, of which 8 are in developing countries and 12 in developed ones (West, 1996). These cases, although they meet a wide definition of the voucher mechanism, in the sense that Government finances schools proportionately to the number of students, do differ substantially under the terms of beneficiary student population eligibility, coverage, regulatory frame and value or percentage of Government subsidy.

Five cases are known in Latin America: Colombia, Chile, Guatemala, Belize and Puerto Rico. By reason of the relevance for Venezuela as a reference point and because of the availability of information, a summary follows of the Chilean and Colombian experiences:

A. The Chilean experience
Chile's educational reform was part of a wide decentralization policy that resulted in a substantial change in Government's structuring and functioning, assigning new financial sources to it and new responsibilities to the municipalities. In 1980 both the property (schools and high schools) and the teachers were transferred to the municipalities.

One of the Chilean educational reform's most innovating elements was the massive introduction of the use of vouchers at elementary and high school level in 1987. Under this new scheme a formula was established to calculate the subvention as directly proportional to the number of students attending school, to be adjusted by means of cost differential according to educational level and geographical location. This new financing mechanism is used to finance not only the children attending public municipal schools, but also those studying at non profit private schools. The voucher mechanism was used in 1993 by 90% of the elementary and high school students.

Under the voucher system the schools' income is determined by the number of pupils, thus creating incentives for the schools to attract students. System design elements and political restrictions, however, have hampered the scheme's optimal operation. The Ministry of Education keeps providing physical infrastructure to municipal schools, while subventioned private ones are forced to finance all the investments by themselves. Additionally, municipal schools often get transfers from the municipalities' overall income, thus covering the deficit created by the reduction of pupils. In this sense, municipal schools have a mild budgetary restriction. These schools, however, are bound to accept any student wishing to register, a restriction that does not apply to private schools. Another important difference in the degrees of freedom of management in both kinds of schools is the handling of the teaching staff. In 1991, labor negotiation with public sector teachers went back from the municipalities to the Ministry of Education. In conclusion, one may say that the rules of the game have not been the same for both kinds of institutions, and this makes it difficult to make any comparison as to cost-efficiency of public versus private managing.

Two years ago, with the reduction of the voucher's real value and in view of the willingness to pay by some sectors, subventioned private schools were authorized to charge an amount to the parents as tuition. However, schools deciding to make use of this source of financing loose a percentage of the voucher they get from the Government.

Having operated for more than 10 years, the Chilean educational reform has raised great interest and a many reviews have been made of their results (Winkler, 1993; Aedo, 1996). Obviously, there is no consensus among those who study the issue, but there are certain non objectable conclusions:

1. The municipalities diverse fiscal capacity has led to unfairness in the expense by student. The richest municipalities have been able to add more resources to the financing provided by the central level than the poorer ones.

2. The high proportion of total financing represented by the voucher guarantees that all students receive this amount of resources as a minimum.

3. Subventioned private schools have sustainedly increased their participation in student roster, going from 13% of the overall student roster in 1979 to 33% in 1993.

On another hand, the considerations as to the reform's results in terms of quality of education are less clear. In order to monitor the quality of education in Chile, a standardized testing system was introduced in 1982 to be used at the 4th. and 8th. grade level and known as SIMCE.

1. Results of the SIMCE tests show, first, that municipal schools and private subventioned schools have lower grades than plain private schools and, second, that municipal schools have lower grades than subventioned private ones. One must consider, however, that there are other factors that may explain the differences in school performance, such as the pupils' socio-economic, the level of rurality, the selection of pupils made by private schools, etc.

2. Since 1988 a gradual improvement has been observed in the three groups of schools' average performance, something being construed by some others as a systematic improvement in the quality of education. One must be cautious, however, with these readings because, first, it is possible that the teachers may be teaching for the test and, second, in view of the existence of a conflict of interests for the Ministry of Education, when it administers a test indirectly evaluating its work.

B. The Colombian Experience
Prior to presenting the Colombian experience one should bear in mind that Colombia has been one of the Latin American countries having least invested in education for the past thirty years. In secondary education, specifically, the private sector has played a most significant role. Forty percent of high school students attend private institutions.

In 1991, Colombia initiate an experimental voucher program at high school level. The program's immediate objectives were:

i) to close the gap between the relatively high elementary school attendance rates and the relative low ones at high school level, through use of private sector infrastructure,

ii) to widen the alternatives available to poor families as to their children's secondary education

iii) to improve the educational system's quality. On the long term, the authorities thought that the voucher system would increase the efficiency when fostering competition between secondary schools. As different from the Chilean case, the program implemented in Colombia did not introduce substantial changes in the public system's financing mechanism.

The voucher system did not become the tool used to finance public schools, as it was in the case of Chile, rather, it introduced the possibility that students could attend a private institution. In other words, only private schools may accept the vouchers. The Colombian program, also, is much more restricted than the Chilean experience, both in terms of coverage as of elligibility of beneficiaries, The program began in 1991 with 18,000 vouchers and 80,000 were delivered in 1995, which represents roughly 4% of the overall high school roster. Students having access to the program are those from low income families and they must have concluded elementary school at a public school.

In view of the fact that the voucher program has been operating for less than five years, no thorough evaluations have been made on its results (Molina, 1993, Van der Gaag, 1996). Until now, the following problems have been identified:

1. Quality of the establishments. According to results of students having taken the ICFES tests (at the end of high school) the average performance of the vouchers' beneficiaries is lower than the average performance of the overall public high school population. The problem has been more notorious in the case of Bogota's low income students.
2. Substitution effect. One of the objections been made to the program is that it financed students who could have had access to private schools. However, the evaluation on this aspect does not seem to support such assumption. As long as the voucher allows an increase in the overall roster, either because financing becomes available to a student who otherwise would not have attended school or because it is provided to another one who could have attended a public school but that now is opening a spot for another student, then the program's first objective is being met.

Generally, the program's approach appears to have been successful. Around 89% of the vouchers have been delivered to students belonging to the lowest strata, and this is much higher than these strata's weight in overall population.

V - The voucher system and its applicability to the Venezuelan educational system
Having discussed the main elements of a voucher system and the experience of two Latin American countries, it is important to point out what is the problem of the Venezuelan educational system that could be solved by means of a vouchers' mechanism. In our opinion, one of the fundamental conditions to improve the efficiency of education is to have a mechanism available to guarantee that the resources invested by Government will be assigned as a function of the quantity and quality of the service being provided by each school. This such a basic principle to stimulate improved performance does not exist in our present educational system. The vouchers are a tool to assign public resources aiming at schools that, in the parents' opinion, are doing a better job.

This notwithstanding, as we saw in the previous section, there are many factors to be borne in mind when having to decide if the voucher system shows a positive balance. Our position is that the voucher system by itself may not be taken as an integral solution to the Venezuelan educational system's problems. In fact, in order that the voucher system may operate in a more or less efficient way, changes are required in our present educational system.

First, a larger number of tools must be supplied to public schools. In concrete, this implies decentralization of decision making in matters of education, going to a much lower level. There is no way to ask a public school to compete with a private one if the former has no power to decide which input blend (teachers, material, etc.) is optimal, in its opinion, to operate with. If public schools are not entitled to hire or dismiss a teacher without the Ministry of Education's authorization, or if they are barred from establishing their own systems to give better compensation to better teachers, their ability to compete is going to be seriously reduced. It is important to stress on the need to decentralize both the taking of decisions as the financial resources. It is obvious that these changes would meet resistance by the teachers' unions.

The voucher system starts from the premise that the parents will be willing to choose the best schools for their children. In order to be able to choose adequately, the parents require information. Government may (should?) then play a role in providing this information, by creating a school evaluation system allowing to classify the schools according to the value added they provide to their students. Undoubtedly, there are difficulties in drafting this system but it is imperative that the consumer --parent or representative-- should have information at the time of choosing his child's school. Perhaps this information role could be played by a private entity, similar to "Consumer Reports" in the United States. To say, however, that a car's engine is more powerful than another one's is something quite easier than saying that this school offer more value added than the other. It could be possible to provide a list of criteria (change in average results of a standardized test, absenteeism ratios, average number of students per classroom, teaching staff's qualification, etc.) used to gauge the schools' quality, and then let the parents decide according to the importance given by them to each one of these criteria.

In most of the existing voucher systems, it is expected that private school should play an important role when forcing the public schools to compete. In this context, an aspect deemed by us as vital when analyzing the implementation of a voucher system is the physical capacity and the incentives of present privates schools in order to accept new students. Today, in Venezuela there are serious regulatory hurdles (permits, price controls, etc.) by the Ministry of Education and local governments that have limited private sector participation in education.. In many countries, the private schools system is operating practically at full capacity. As an anecdote, in Uruguay one sees pregnant mothers queuing in order to be able to register their future child in a private school. The first possible solution would be to think of the creation of new private schools or of the widening of existing ones' capacity, looking forward to attracting the group of students migrating from "inferior" schools. We believe that the fact that incentives exist for this to occur is something quite debatable. If these incentives actually existed, then we ask: why is it that some of the existing private schools reject applications filed by students willing to pay for their education? Is it because these students do not meet the necessary academic standards required for admission in these schools? If this is so, is it then that the voucher system does not generate a structure where children with lesser resources otherwise may not have access to private schools?

One of the important aspects to be considered is if parents really have alternatives to choose from. The relevant question would be: How many children attending public schools could, if they had no registration restriction, have access to a private school? The fact of failing to take into account indirect costs, such as transportation, uniform and food, could set a barrier for the parents to be able to choose and change their children to another school. In rural environments, the distances from home to school are also and additional important obstacle.

VI - Conclusions
In our opinion, the vouchers are not an integral solution for Venezuela's educational problems. As we have seen they are not a panacea and it is even dangerous to see them as such. When deciding, however, if a system of educational vouchers should be implanted or not in Venezuela, we believe that it is not convenient to think in terms of seeing it as a solution to all the present Venezuelan educational system's problems. In this sense, the voucher system constitutes an alternative mechanism to assign resources that could help to link the schools' budget to the quality of the service by them provided.

We are of the opinion that in order that a voucher system may operate efficiently, it is imperative that greater autonomy should be given to public schools, that minimum standards should be established, to be met by the schools sharing the program, ensuring that parents may have access to information on the schools' quality, and that the physical capacity and the incentives currently offered to private schools be analyzed.

The Chilean and Colombian experiences allow us to become aware of the importance of Government's role when regulating and drafting policies for the sector, even more when market reforms are introduced in the educational system.

At a theoretical level it is difficult to reach conclusion as to how a voucher system would operate properly in Venezuela. If it were decided to implement this system it would be convenient to start with a pilot program in one or several states and, after the results should have been reviewed, another decision should be taken as to implementing the system to the rest of the country. An additional advantage of a pilot program is that it would be possible to detect possible implementation errors and then correct them when starting the program at a national level. The introduction of this kind of reforms generally has led to confrontation with the teaching profession, and this is an additional reason to try a gradual implementation making it easier to negotiate with the unions.

1 Defined as homes with a per capita income at least twice higher than the poverty line.

2 The nets school attendance rate measures the percentage of children from the corresponding age bracket attending schools.

We express our thanks for the conversations held with Ana Acevedo, Laura Rawlings, Myriam Waiser and Carlos G. Molina, who enriched our understanding of the issue.

• Bibliography
  Aedo, C. 1996. Organización Industrial de la Prestación de Servicios Sociales en Chile. ILADES-BID. Proyecto Red de Centros (por publicar).
• Becker, G. 1995. Human Capital and Poverty Alleviation. HRO Working Paper No. 52. The World Bank. Human Resource and Operation Policy. Washington, D.C.
• Bridge, G. 1978. Information Imperfections: The Achilles Heel of Entitlement Plans. School Review. 86.
• Calderon, A. 1996. Voucher Program for Secondary Schools: The Colombian Experience. HCD Working Papers No. The World Bank. Washington D.C.
• Carnegie Foundation for the Advancement of Teaching. 1992. School Choice. Princeton: Carnegie Foundation.
• Fossey, R. 1994. Open Enrollment in Massachusetts: Why Families Choose. Education Evaluation and Policy Analysis. 16 (3).
• Krashinksy, M. 1986. Why Educational Vouchers May be Bad Economics. Teachers College Record. 88 (2).
• Levin, H. 1991. The Economics of Educational Choice. Economics of Education Review. 10 (2).
• Lockheed, M. and E.Jiménez. 1994. Public and Private Secondary Schools in Developing Countries. HRO Working Paper No. 43. The World Bank. Human Resource and Operations Policy. Washington, D.C.
• Molina, C., Alviar, M. y D. Polania. 1993. El Gasto Público en Educación y Distribución de Subsidios en Colombia. Fedesarrollo. Bogotá.
• Murnane, R. and F. Levy. 1996. What General Motors can teach US schools about the proper role of markets in education reform. Phi Delta Kappan. October.
• UNESCO. 1995. World Education Report. Van der Gaag, J. 1995. Private and Public Initiative: Working Together in Health and Education. The World Bank. Washington D.C.
• West, E. 1996. Education Vouchers in Practice and Principle: A World Survey. HCD Working Papers No. 64. The World Bank. Washington D.C. Winkler, D. 1993. Municipal and Private Sector Response to Decentralization and School Choice: The Case of Chile, 1981-1990. HCO Working Paper No. 8. The World Bank. Washington D.C.
• World Bank. 1993. Venezuela 2000: Education for Growth and Social Equity. Report No. 11130-VE. Latin America and the Caribbean Regional Office. Washington D.C.