Taking Sides: The Effects of Supply and Demand Driven Policies on Parent Decision Making and Education Investment in Brazil

Introduction

With near universal primary enrollment levels achieved in Brazil, the focus of education reform has shifted from purely access issues to increased attention on improving learning outcomes for all children. Yet the term "universal" confuses issues of access and quality. Although most of Brazil's children have access to primary school, an important question is: access to what? Many children receive teaching of low quality, based on an outmoded curriculum, in what could only be called a façade of a school. Enormous disparities in the distribution of educational resources and opportunities lead to gaps in attainment, particularly between income groups. While 96 percent of 7-14 year olds enroll in grades 1-8, less than two-thirds finish the primary cycle and only about a third continue on to upper secondary school.

Reform efforts are now centered on those marginalized students still out of school or caught in a cycle of low-attendance, repetition and eventual drop out. Two types of policies have emerged in response to these disparities, each based on different theories of the causes of low school attainment. Traditional supply-side policies that build schools and equip them with a minimum set of resources schools are complemented by a new approach-involving parents in school management and decision-making. These policies compete for scarce resources with a new set of demand-side programs, designed to reduce the direct and indirect costs of schooling to poor families through payment of stipends or subsidies.

Ongoing experiments with both supply and demand side strategies make Brazil an interesting case for study. On the supply side, policymakers have developed programs that include parents in the day-to-day management of schools. One example is the PDE (School Development Program) in which a school leadership committee (including parents) undertakes an evaluation of their school, develops objectives, and formulates a plan to achieve those objectives. On the demand side, the federal Bolsa Escola (School Scholarship) program awards cash grants to families as long as their children maintain a minimum level of attendance.

These policy options are each supported by an empirical research base. Supply-side proponents, citing production function and school effectiveness literature, assert that by increasing parent involvement and participation in school processes, parents will grow to understand the benefits of schooling and schools will become more accountable to parent needs. Demand-side proponents, citing household survey literature, assert that high opportunity costs and ignorance of the long-term benefits of schooling constrain families from fully investing in schooling. In examining these approaches it became clear to me that both policymakers and researchers look at only one side of the equation-either schools or households-and fail to see the interdependence of the two.

The idea that school quality impacts household demand for schooling does not, at first glance, appear to be a revolutionary idea. Yet, as the following pages show, a significant gap in both the empirical literature and policy initiatives reveals a sharp division in how researchers and policy makers think about improving learning outcomes for children. Reviewing the Literature

The majority of the literature addressing low attainment and achievement can be divided into two categories: literature evaluating the supply of schooling (school quality and school effectiveness literature), and literature assessing the demand for schooling (family characteristics and child labor). Studies in the production function tradition (Fuller and Clarke 1994) address supply-side issues and relate student achievement levels to school inputs (including teacher education, textbooks and other physical inputs). The demand-side literature, primarily consisting of research utilizing household survey data, assesses household demand for schooling (expressed in years of attainment). In the following pages, each of these will be reviewed in turn, as well as a third group of studies: those that attempt to combine school and household factors and model their interaction. In addition to summarizing the evidence, I assess whether the findings contribute to understanding how school quality affects parent decision-making regarding the schooling of their children.

Production Function Literature and the Supply of Schooling

In economic research, production functions have long been used to assess the efficiency of the input mix of firms in producing a given output. In the case of education, schools produce output in the form of achievement, using both school inputs (teachers, facilities, equipment) and family inputs (parental education, income level, race, language spoken in the home). The beginnings of this research "paradigm" can be traced to the Coleman Report (1968), which established that in U.S. schools the most important determinants of student performance were family background and the student's peer group (Levin 1995). In developing countries, researchers have sought support for an "anti-Coleman" hypothesis. Numerous studies have demonstrated that characteristics of schools in developing countries explain a greater share of the variance in student achievement than do family characteristics (Heyneman and Loxley 1983; Fuller 1986; Velez, Schiefelbein et al. 1993; Behrman, Khan et al. 1997).

Other approaches have included meta-analysis of existing production function literature. Hanushek (1986) reviewed the literature on the impact of school spending, finding that typical input studies do not show a consistent relationship to student achievement. This conclusion was immediately challenged and subjected to re-analysis in Hedges, Laine et al. (1994) who countered that "money might matter" somewhere. In a similar review (but with the opposite result) of 100 studies in developing countries, Hanushek (1995) reports that 22 of 34 studies reported a positive link between school facilities and student performance.

Production functions present serious methodological flaws, principal of which is the limited and often biased nature of information collected only in schools. The method behind most production functions is to apply a student assessment to capture learning output, while collecting information on both family and school inputs. Schooling, however, has multiple outputs, including difficult to quantify results such as psychological development and socialization. Sampling is also a measurement problem, as information is not collected on those students not in school that day. Systematic measurement error frequently occurs when asking children to assess their parents' education and income levels. Finally, production functions do not examine the relationship between demand for school and school quality, and almost always dismiss the fact that parents make choices about where, how often, and how long their children will attend school. Failure to account for differences in optimizing behavior by parents could lead to false inferences about the effectiveness of different education investments (Glewwe 1999).

In addition to these methodological flaws, education has peculiar characteristics that simply do not fit into the standard economic production function and cost minimization literature (Levin 1995). School characteristics such as organizational structure, degree of autonomy, and pedagogical processes are not easily captured by education production functions. One branch of studies that have attempted to capture more of these education processes is school effectiveness literature, as described below.

Redefining School Quality: Effective Schools in Developing Countries

Effective schools research, which gained popularity in the U.K. and U.S. in the early 1980s, grew out of a realization that schools with similar student populations were achieving very different results. In comparing high-achieving schools with low-achieving schools (each enrolling similar students), researchers identify operational differences, called "effective school correlates" (Levin 1995). If ineffective schools could adopt these correlates, researchers asserted, they would become effective. Edmonds, considered the father of school effectiveness research, identified five ingredients of an effective school: strong administrative leadership, high expectations for children's achievement, an orderly atmosphere conducive to learning, emphasis on basic skill acquisition, and frequent monitoring of student progress (1979).

Studies of school process factors in developing countries are few (Riddell 1989), but have increased substantially in recent years (Fuller and Clarke 1994). Riddell (1997) calls for expansion of research into how schools work in developing countries, in part to investigate the legitimacy of the notion that school factors are more important than family factors in determining achievement. Examples include a Botswana study of junior schools combining classroom observation and longitudinal student achievement results (Fuller, Hua et al. 1994) and a study of 16 factors measuring dynamic conditions of effective schools in India (Heneveld 1994).

Riddell's and similar studies, however, miss the mark, as each focuses only on one aspect of the school-family interaction. As is the case of production function literature, school effectiveness literature does not examine why parents choose certain schools, or the contribution parents make to the learning process. Methodological problems in what are basically outlier studies are also prevalent. While school effectiveness studies have become popular among educators, numerous criticisms surround the methods employed. The major strategy of identifying outliers on a predicted mean achievement score based on socioeconomic conditions of the school presents statistical problems such as selection and omitted variables biases. That is, particularly effective schools may simply have more "motivated" students or teachers, or a similar characteristic that will not be captured by the studies. Schools found to be effective are highly sensitive to in-school variation between grade levels and subject areas, and can demonstrate high variations in results between years. Finally, there is little evidence that schools can be "taught" to be effective (Levin 1995).

Production Functions as a Tool for Policy: The Case of Brazil

These studies provide support for a long history of investment programs in school inputs. In Brazil, an exemplar and oft-cited study of the education production process is that of Harbison and Hanushek (1992). In a detailed account of the inequities that plague primary schools in the Northeast the authors report a link between school facilities and student outcomes. The direct results of these inadequate facilities, the authors state, are high levels of repetition and age-grade distortion, as well as correspondingly low achievement levels. How is this link determined? Based on four biennial applications resulting in panel data of some 6,000 students in 600 schools, researchers assess student school inputs, student achievement levels and progress, and teacher and community variables. The study, one of the most ambitious for a developing country at that time, nevertheless makes no effort to assess parent responses to increased resources in schools. Further, despite widespread acclaim for their studies, the authors themselves acknowledge that "simply collecting readily available data on basic resources does not allow assessment of effects . . . there is no substitute for investigations into the fundamental educational relationships that lie behind student performance" (p. 186). Production function literature, while offering substantial insight into the distribution of resources among schools, does little to reveal what really goes on in the education process.

Despite this caveat, the research of Hanushek and others have been used to support continued investment in standard inputs in schools with little relation to the processes that influence learning and other student outcomes. Although World Bank-financed federal level projects are a small part of total education expenditures in Brazil, they nevertheless represent a not inconsequential share of non-salary expenditures. They are also perhaps the easiest interventions to assess due to relatively accurate project expenditure estimates. Over the lifetime of the Northeast Basic Education project, some US $500 million was spent upgrading school facilities, providing textbooks, and training teachers. The effect of these inputs on demand for schooling is largely unknown (Silveira 2000), yet the politically popular input intensive projects continue to be an important part of the Bank loan portfolio in Brazil and throughout the developing world.

Household Survey Literature and the Demand for Schooling

In attempting to understand why some families strive to obtain additional schooling for their children while others do not, researchers have focused on household characteristics as predictors for the level of investment of education. The development of household survey techniques and their use in policymaking is a fairly recent innovation. Beginning in the 1980s, a number of developing countries applied household surveys and Living Standards Measurement Surveys (LSMS) in order to better understand responses to current and proposed government polices (Grosh and Glewwe 1995). The two types of surveys are similar, however the LSMS is actually a composite of surveys assessing families, communities and local price conditions, often introduced with World Bank support to those countries without an established household survey.

In education, studies that use household survey data compare years of attainment to family characteristics. Much of the justification for the approach comes from human capital theory. The concept of education as an investment, like physical capital, has dominated economics of education since the publication of Gary Becker's Human Capital (1964, Third Edition 1993). Human capital refers to the knowledge, skills and abilities possessed by people. Human capital theory asserts that individuals can increase future earnings through enhancement of productive capacities. As is the case of physical capital, these investments require up-front costs. Comparing these costs to benefits is the focus of estimations of the rate of return to investment in education, where wages and increased productivity (private and social benefits from schooling) are compared to total costs of schooling (social and private costs). Calculations of this "profitability" of different levels of education have sparked considerable debate among economists and education policymakers. The main conclusion of this literature is that parents will invest in their children's schooling until the point at which costs exceed benefits (Woodhall 1995). That is, if the rate of return to an additional year of schooling is less than the current interest rate, then the resources that would have been devoted to schooling could be better used in other investments.

Literally hundreds of econometric studies have calculated rates of return to different levels of education, most showing a positive and substantial impact of schooling on earnings. Psacharopoulos' rates of return studies, among them Peru, Uruguay, and Brazil (Psacharopoulos and Arriagada 1987; Psacharopoulos et al. 1994; Psacharopoulos and Ng 1994; Psacharopoulos and Velez 1994) have provoked considerable debate among educators and economists. Reviewing private and social rates of return to investment in education in 32 countries (1973; with subsequent updates 1981; 1985), Psacharopoulos reveals four underlying patterns (1981, p. 326):

a. The returns to primary education are higher than those of other education levels.

b. Private returns exceed social returns, particularly at the university level.

c. All rates of return to investment in education exceed the 10 percent benchmark of the opportunity cost of capital.

d. The returns to a given level of education in developing countries are higher than those of more developed countries, reflecting the scarcity of human capital in many developing countries.

While many of these assertions have held, research into rates of return over time has demonstrated that where there is expansion of schooling toward universalization, rates of return decline over time, beginning with the primary level (Carnoy 1994).

Linked to these calculations of the rates of return are studies that attempt to understand why, if rates of return to primary education are so high, families differ in their levels of investment in education. The most widely cited source of under-investment in schooling is parents' level of education (Lockheed, Fuller et al. 1988; Canagarajah and Coulombe 2000). Other causes include unequal distribution of income (Lam 1999) and access to (particularly agriculture) technology (Rosenzweig 1992). In Nicaragua, after controlling for differences in the supply of school places, family background was found to affect the probability that children enroll in, attend, and complete various levels of schooling (Wolfe and Behrman 1984). Similar studies for Brazil (Psacharopoulos and Arriagada 1987) and other countries have found that parent literacy is strongly associated with the number of years that children attend school.

School attendance and attainment are also tied to children's activities when they are not in school. Much of the research utilizing household survey data points to high incidences of child labor as an indication that high opportunity costs preclude poor families from investing in higher levels of education for their children (Gootaert and Kanbur 1994; Basu 1999). A common belief is that "working prevents children from benefiting fully from school and may thereby condemn them to perpetual poverty and low wage-employment" (Ray 2000: p. 348). Conversely, some have argued that education and child labor may not in fact be mutually exclusive activities. Patrinos and Psacharopoulos' (1997) research using Peruvian household survey data reveals that child labor is not detrimental to schooling and in some cases "working actually makes it possible for the children to go to school" (p. 398). Recently, authors have documented the impact of ILO, Unicef, and government campaigns on child labor. In his comprehensive review of child work in Mexico, Peru and Chile, David Post describes the impact of changing regulations and resulting heterogeneity of the "child labor problem"(2001). Although it is not within the scope of this paper to review the entirety of child labor research, it is important to note, as in the case of other household survey research, the lack of research into school quality as a factor in predicting higher incidences of child labor.

A recent review of child labor literature is that of Basu (1999). Though Basu indicates that the provision of quality schooling can play a big part in reducing child labor, he finds few studies that attempt to attempt to link school quality factors to levels of child work. Rather most researchers limit their analysis to testing relationships between different household characteristics and schooling outcomes. Rosenzweig and Evenson (1977) found that higher child wages lead to increased child labor and decreased enrollment levels. High demand for child labor was also found to increase drop-out rates (King, Orazem et al. 1999). One typical example is that of Ray (2000) who tests whether there is a positive association between child labor and poverty and a negative association between child schooling and poverty in Pakistan and Peru (both relationships confirmed in Pakistan, but only the latter in Peru). Higher levels of parent education, he determines, have a positive effect on children's schooling in both countries, as parents with more education are "better able to see the value to their children's education and to resist the temptation to pull them out of school" (Ray 2000: p. 361).

Support for the Eradication of Child Labor: Student Stipend Programs

In Brazil, a combination of external pressures and internal politics has generated considerable interest in and growing consensus regarding the relationship between child schooling and work (Schwartzman 2000). In addition to legislative approaches, including prohibitions against certain forms of child work and the creation of the child and adolescent code, Brazil has adopted a targeted enrollment subsidy policy designed to reduce relatively high (27 percent of 10-14 year olds are considered economically active) indices of child labor. In exchange for ensuring that their children are enrolled in school, families receive a monthly stipend, designed to offset both direct and indirect costs of schooling. The underlying assumption of this policy is that child labor displaces schooling. That is, by decreasing child labor, families will invest in additional schooling for their child. Yet this assumption may not be true in all cases. Where students can both work and go to school, subsidies are certainly likely to reduce short-term poverty but may have little impact on long-term schooling outcomes.

Ravillion and Wodon (2000) test this assumption in examining the effects of a targeted enrollment subsidy on labor force participation and school enrollment in Bangladesh. Based on the results of the Bangladesh program, the authors conclude that the subsidy increased schooling by far more than it reduced child labor. As is this case of many developing countries the school day in Bangladesh is divided in two four-hour shifts. It is not unreasonable to assume that children who work 20 hours per week can still attend primary school. The authors found that the stipend program was enough to assure nearly full attendance among participants. Nevertheless, child labor was only decreased by a small fraction (one eighth for girls, one fourth for boys) of the increase in enrollment rates. These results lead the authors to question whether there is a one to one tradeoff between schooling and child labor.

Other studies have raised doubts regarding the efficiency of stipend programs compared to other investments in education. In an examination of the impact and costs of fifteen prevention and social reintegration programs targeting at risk students in Latin America, Schiefelbein et al. (1999) find that only four of these programs generate benefits greater than their costs. Prevention programs are those that seek to raise the achievement levels of at risk students, while reintegration programs promote reintegration of children that have already dropped out of school. Programs analyzed include MECE (Chile), Escuela Nueva (Colombia), EDUCO (El Salvador), as well as EDURURAL, the Accelerated Learning Program (PAA) and two student stipend programs (Campinas and Brasília) in Brazil. The highest cost-effectiveness ratios are attributed to program with a rural focus, including EDURURAL, Escuela Nueva and EDUCO, while the costs associated with each of the stipend programs substantially exceed the benefits. While the authors acknowledge the estimated nature of their findings, it is important to note the growing research base surrounding the impact of school stipend programs.

Hybrid Research: A Dynamic Approach to Supply and Demand

Integrated approaches to mapping education and family decision-making are few, but those studies that attempt to understand the interaction between school quality and parent decision-making are increasing in number. In the U.S., Mora documents the influence of opportunity cost of school attendance, educational quality attributes, and SES on educational demand of minorities in the U.S. (1997). Her results suggest that a school's attendance rate is sensitive to education quality and student characteristics. The impetus for this and other research, Mora states, stems from the observation that African- and Mexican- Americans consistently attain less education on average than non-Hispanic whites. To this Mora adds a crucial observation: that low schooling investments serve as one of the primary sources of minority African- and Mexican-American/white earnings gaps.

Similar relationships have held in the few studies that exist for developing countries. Initial attempts to study the interaction between demand for and supply of schooling utilized existing household survey data in combination with information about local schools. Using household survey data, researchers integrated individual and household data with school information to examine the importance of school inputs in Pakistan (Behrman, Khan et al. 1997). As interest increased and data improved, researchers became more sophisticated in their analyses. Hanushek and Lavy (1999) study the effects of school quality on the dropout rate, and reveal that children are strongly influenced in their schooling decisions by the quality of their prospective school. Glewwe et al. use an expanded production function (derived from a combination of household survey and school data)-including pedagogical processes and school organization in addition to the standard inputs mode-to evaluate the determinants of achievement in Jamaica (Glewwe, Grosh et al. 1995).

In Honduras, Bedi and Marshall (1999) estimate a model of primary school attendance, attributing the problems of high repetition rates to two factors: high opportunity costs and low school quality. High opportunity costs in terms of loss of labor on the farm or in the household can disproportionately impact low-income families. "Attendance will suffer when parents perceive that the return to time spent in school does not justify the loss of additional labor, which, in turn, increases the likelihood of grade failure, repetition, and eventual desertion" (p. 658). Low school quality affects this calculation by reducing the expected gains to schooling, in effect lowing parent expectations. Their school attendance model, applied to a cohort of students in 33 randomly selected schools supports the hypothesis that expected achievement gains influence the demand for primary schooling.

For Brazil, the work of Fuller et al. (1999) has documented the influence of family, teacher and classroom on children's early literacy in Northeast Brazil. Their study focuses on which teaching practices observable in Brazilian classrooms are empirically related to high student achievement. The greater political context of the organization of schools and the ongoing democratization debate (including comments on participatory learning as conceived by Freire) are also the subject of study, demonstrating a depth of interest by the authors not commonly seen in studies of the factors that influence student outcomes. While the study does not attempt to map how differences in school quality affect parent demand for schooling, two important observations are noted. The first is that interactions between parents and students through reading at home or other forms of engagement are suggested as a way of raising children's early literacy. A second point is that the condition of school facilities is often an indicator of school quality in the eyes of parents, a hypothesis confirmed in Glewwe and Jacoby's research in Ghana (1994). These and other studies attempt to bridge the gap between traditional studies of either school supply or household demand for schooling.

Families as Inputs to and Managers of the Education Process

The decision by families to educate their children through formal schooling is the central issue of this paper. Accordingly, a review of the literature cannot neglect the influence of parents, both as passive inputs to and dynamic actors in the education system process. Investigation of the influence of family background on educational outcomes has long dominated the field of sociology of education (Coleman 1968; Jencks 1972; Marjoribanks 1979). In Britain for example, longitudinal data on some 10,000 families during the last half-century (The Nuffield Studies) have revealed the dearth of social mobility between classes (Halsey 1977; Halsey, Heath et al. 1980). More recently, researchers have focused on the processes by which these outcomes are distributed between different (racial, socio-economic and gender) groups (Carvalho 2000; Lareau 2000). The following pages briefly discuss the role of families both as inputs to and managers of the education process.

Interpretations of the structural reasons for the unequal distribution of education tend to focus on the family as a (usually passive) factor or input into the education process. Policies shaped by the dominant class or group then certify the students for their respective societal roles (Bowles and Gintis 1976). For some theorists, the purpose of education is this certification or signaling process; schooling serves to classify students according to a hierarchy of ascriptive characteristics such as social class, ethnicity or religion (Groot and Hartog 1994). These theorists see schools as reflecting a class struggle that arises in the production system and becomes embodied in the institution of schooling (Carnoy 1981). One shortcoming of this literature is its neglect of the interaction between families and schools: the influence of parents in schooling activities is typically excluded from these analyses.

Cultural capital theories nearly bridge the gap between passive input and dynamic process oriented understandings of the contribution of families to the education process, but fall short of explaining the dynamics of parent-school relations. As outlined by Bordieu and Passeron, families acquire and impart to their children a set of values, norms and language called cultural capital (1977). Cultural capital, like economic capital, can be used to leverage wealth and privilege. Schools recognize and reproduce the structure of the distribution of cultural capital, perpetuating existing inequalities. Bordieu criticizes (neo-classical) economic theory and its distinction between economic and non-economic forms of capital:

[economic theory] defines as disinterested [i.e. non-economic] those forms of exchange which ensure the transubstantiation whereby the most material types of capital-those which are economic in the restricted sense-can present themselves in the immaterial form of cultural capital or social capital or vice versa (Bordieu 1997).

That is, those processes that fall outside of (traditional) economic analyses are actually the processes that contribute most to structural inequalities in education and other spheres. Critics of the cultural capital theories point out that for Bordieu and Passeron there is no source of power relations, nor is there an analysis of how schooling changes. Carnoy (1981) asks: if reform occurs but the system stays the same, why do the "dominant powers" bother with reform? The authors are notably silent on this point.

More recently, the dynamic processes by which families influence the distribution of education outcomes have been the focus of a growing body of research. Together with the movement for greater school accountability and choice (a movement Brown (Brown 1997) has labeled the "Parentocracy"), "the value of parental involvement has become an acceptable truism across a wide spectrum of political positions in the U.S." (Cassanova 1996). Partly in response to the growing number of policies aimed at increasing parent involvement in education, a group of critical scholars began to question the social, economic and educational implications of this call for increased parent influence in schooling. "Parent involvement" in this case is assumed to be beyond that of the home role, in which parents meet the obligations of feeding, clothing and sending their children to school. Under the expanded concept, parental involvement also includes participation in school activities (such as volunteering in the classroom) and decision-making (e.g. requesting that a child be placed in gifted education) (Carvalho 2000). This focus on parent involvement is not, however, without risks. The current policy focus on family involvement, warns Carvalho, disregards differences in family and cultural conditions between social classes (including feelings about schooling), overlooks historical differences in the ideal of the "community school," and neglects the point that "school relations are relations of power-and most families are powerless" (p. 6).

In a study of two elementary schools in the U.S., Lareau "americanizes" Bourdieu's concept of cultural capital by documenting the processes by which middle-class parents manage the education of their children (Lareau 1987). Through careful ethnographic research the author is able to show that while both working-class and middle-class parents value education, working class parents seldom try to influence the education system, leaving academic matters to their children's teachers. Similarly, Baker and Stevenson (1986) document the specific strategies used by a sample of 41 mothers of eighth graders transitioning into high schools. Their findings indicate that parents "actively manage their child's school career in ways that can have direct consequences for their children's educational achievement" (Baker and Stevenson 1986). The authors uncover similar evidence in a nationally sample of elementary schools, finding that mother's education is positively related to the degree of parental involvement in school activities (Stevenson and Baker 1987).

Outside the U.S., researchers have recorded similar results, though system and structural differences affect the degree to which parents can influence the education process. The lack of flexibility in the German system, for example, allows for very little of the negotiation and lobbying that characterizes parent involvement in the American system (Baker and Stevenson 1989). A substantial number of studies continue to focus on more traditional forms of parent involvement, including: help with homework (Nadon and Normandeau 1997) and parent-adolescent interaction and achievement (Deslandes and Potvin 1999) in Quebec; parent acceptance of teacher decision-making in Honduras (McGinn, Reimers et al. 1992); parental involvement with homework in Portugal and Luxembourg (Villas-Boas 1998); and the frequency of parent reading in the Netherlands (De Graaf, De Graff et al. 2000).

In the case of Brazil, analysis of parent involvement is the topic of a small but growing body of literature. Interviews with 18 parents in six public schools in Bahia reveal substantial levels of participation in school activities, including school cleaning and preparation of meals, making of small repairs, and participation in school councils (Portela and Barretto Bastos 1997). Involvement in academic decisions and help with homework is minimal, which researchers attribute to low levels of parent education. In a survey of parents of "successful" students, researchers document the absence of school-parent communications, of which parents seem to be proud; school-parent communication is seen as a negative occurrence (Gurgel 1998). In other studies, the "gap" between school and community is frequently cited as an impediment to children' learning (Gurgel 1997; MEC/Projeto Nordeste, World Bank et al. 1997). Criticism of the traditional or closed "school culture" has increasingly been the focus of education reformers. Citing evidence from a survey of 954 parents from a single primary school in Brazil, researchers reported that:

The parent's interest expressed in relation to school topics and the schooling process of their child contradicts an idea that apparently is part of the culture of Brazilian schools: the one that says that parents do not worry too much about the school subjects and that they are not interested in talking or discussing about it (Reali, Reyes et al. 2000).

Understanding how families act both as passive inputs to and dynamic managers of the education of their children is important to understanding how the decision to educate children is made. Allocation of children's time between work and play is the most basic of decisions a family must make. Equally important to student outcomes such as achievement, attendance and attainment are such issues as: quantity and quality of assistance with homework, acceptance of teacher's decisions regarding grades and student progress through school, and parent expectations for success and failure. Capturing the dynamic of the parent-child-school relationship is a considerable challenge, but one that should at least be attempted when modeling student outcomes. Conclusions

The central focus of this article is the decision by parents to enroll and maintain their children in school. As the previous pages have documented, even the rudimentary mechanics of this decision-making process are unknown to policy-makers and researchers. This remains the case in spite of the fact that low completion and high repetition and dropout rates are an important concern to many governments from both a human capital and efficiency perspective. The human capital perspective concentrates on the loss of investment opportunities though under-investment in schooling. Despite the questioning of standard rates of return estimates to education, the pay-off to completion of primary schooling is thought to be high, from both a social and economic perspective. From an efficiency perspective, high repetition and dropout rates increase the costs of producing a graduate. For parents, the main payoff to completing primary school with "good scores" is likely to be the option value of completing secondary school and university.

For these reasons, examination of the factors that contribute to parents' decisions to retain their children in school, and to help them do well in school, is important from a policy perspective. Demand for schooling has been measured in relation to expected earnings, but with the exception of a few studies (Glewwe 1999; Hanushek and Lavy 1999), most research makes no attempt to link the quality of school supply with demand.

References

Baker, D. P. and D. L. Stevenson (1986). "Mothers' Strategies for Children's School Achievement: Managing the Transition to High School." Sociology of Education 59(3): 156-166.
Baker, D. P. and D. L. Stevenson (1989). Parent's Management of Adolescents Schooling: An International Comparison. The Social World of Adolescents. U. Engel: 339-50.
Basu, K. (1999). "Child Labor: Cause, Consequence, and Cure, with Remarks on International Labor Standards." Journal of Economic Literature 37(3): 1083-119.
Becker, G. S. (1993). Human Capital. Chicago, National Bureau of Economic Research.
Bedi, A. S. and J. H. Marshall (1999). "School Attendance and Student Achievement: Evidence from Rural Honduras." Economic Development and Cultural Change: 657-682.
Behrman, J. R., S. Khan, et al. (1997). "School Quality and Cognitive Achievement Production: A Case Study for Rural Pakistan." Economics of Education Review 16(2): 127-142.
Bordieu, P. (1997). The Forms of Capital. Education: Culture, Economy, and Society. A. S. Wells. Oxford, Oxford University Press: 46-58.
Bordieu, P. and J. C. Passeron (1977). Reproduction in Education, Society, and Culture. Newbury Park , CA, Sage.
Bowles, S. and H. Gintis (1976). Schooling in Capitalist America. New York, Basic Books.
Brown, P. (1997). The 'Third Wave': Education and the Ideology of Parentocracy. Education: Culture, Economy, and Society. A. S. Wells. Oxford, Oxford University Press: 393-408.
Canagarajah, S. and H. Coulombe (2000). Child Labor and Schooling in Ghana. Working Paper. Washington DC, World Bank.
Carnoy, M. (1981). Education, Economy and the State. Cultural and Economic Reproduction in Education: Essays on Class, Ideology and the State. A. Michael. London, Routledge & Keegan Paul.
Carnoy, M. (1994). Rates of Return to Education. International Encyclopedia of Education. M. Carnoy, Pergamon: 4913-18.
Carvalho, M. E. P. d. (2000). "Relações Entre Família e Escola e Suas Implicações de Gênero." Cadernos de Pesquisa 110(July): 143-155.
Carvalho, M. E. P. d. (2000). Rethinking family-school relations: a critique of parental involvement in schooling. Mahwah, New Jersey, Lawrence Erlbaum Associates.
Cassanova, U. (1996). "Parent Involvement: A Call for Prudence." Educational Researcher 25(8): 30-32, 46.
Coleman, J. S. (1968). The Evaluation of Equality of Educational Opportunity, Johns Hopkins Univ., Baltimore, Md. Center for the Study of Social Organization of Schools.: 41.
De Graaf, N. D., P. M. De Graff, et al. (2000). "Parental Cultural Capital and Educational Attainment in the Netherlands: A Refinement of the Cultural Capital Perspective." Sociology of Education 73(2): 92-111.
Deslandes, R. and P. Potvin (1999). Autonomy, Parenting, Parental Involvement in Schooling and School Achievement: Perception of Quebec Adolescents. Montreal, Presented at AERA 1999.
Edmonds, R. (1979). "Effective Schools for the Urban Poor." Educational Leadership 37: 15-27.
Fuller, B. (1986). Raising School Quality in Developing Countries: What Investments Boost Learning? World Bank Discussion Papers 2, World Bank, Washington, D. C.: 106.
Fuller, B. and P. Clarke (1994). "Raising School Effects while Ignoring Culture? Local Conditions and the Influence of Classroom Tools, Rules, and Pedagogy." Review of Educational Research 64(1): 119-57.
Fuller, B., L. Dellagnelo, et al. (1999). "How To Raise Children's Early Literacy? The Influence of Family, Teacher, and Classroom in Northeast Brazil." Comparative Education Review 43(1): 1-35.
Fuller, B., H. Hua, et al. (1994). "When Girls Learn More than Boys: The Influence of Time in School and Pedagogy in Botswana." Comparative Education Review 38(3): 347-76.
Glewwe, P. (1999). The Economics of School Quality Investments in Developing Countries : an Empirical Study of Ghana. New York, St. Martin's Press.
Glewwe, P., M. Grosh, et al. (1995). "An Eclectic Approach to Estimating the Determinants of Achievement in Jamaican Primary Education." The World Bank Economic Review 9(2): 231-258.
Glewwe, P. and H. Jacoby (1994). "Student Achievement and Schooling Choice in Low-Income Countries: Evidence from Ghana." Journal of Human Resources 29(3): 843-64.
Gootaert, C. and R. Kanbur (1994). Child Labor: A Review. Background Paper for the 1995 World Development Report. Washington DC, World Bank.
Groot, W. and J. Hartog (1994). Screening Models of Education. International Encyclopedia of Education. M. Carnoy, Pergamon: 5350-55.
Grosh, M. and P. Glewwe (1995). A Guide to Living Standards Measurement Study Surveys and Their Data Sets. Living Standards Measurement Study Working Paper No. 120., World Bank, Washington DC.: 50.
Gurgel, P. R. H. (1997). Ditos Sobre a Evasão Escolar: Estudo de Casos No Estado da Bahia, MEC/Projeto Nordeste.
Gurgel, P. R. H. (1998). Ditos Sobre o Sucesso Escolar: Estudo de Casos No Estado da Bahia, MEC/Projeto Nordeste.
Halsey, A. H. (1977). Towards Meritocracy? The Case of Britain. Power and Ideology in Education. A. H. Halsey. New York, Oxford University Press: 173-185.
Halsey, A. H., A. F. Heath, et al. (1980). Origins and Destinations: Family, Class, and Education in Modern Britain. Oxford, Clarendon Press.
Hanushek, E. A. (1986). "The Economics of Schooling: Production and Efficiency in Public Schools." Journal of Economic Literature 24: 1141-71.
Hanushek, E. A. (1995). "Interpreting Recent Research on Schooling in Developing Countries." The World Bank Research Observer 10(2): 227-46.
Hanushek, E. A. and V. Lavy (1999). School Quality, Achievement Bias, and Dropout Behavior in Egypt. Living Standards Measurement Study Working Paper No. 107. Washington, D.C, World Bank.
Harbison, R. W. and E. A. Hanushek (1992). Educational Performance of the Poor: Lessons from Rural Northeast Brazil, Oxford University Press.
Hedges, L. V., R. Laine, et al. (1994). "Does Money Matter? A Meta-analysis of Studies of the Effects of Diffferential School Inputs on Student Outcomes." Education Researcher 23: 5-14.
Heneveld, W. (1994). Planning and Monitoring the Quality of Primary Education in Sub-Saharan Africa. Technical Note no. 14. Human Resources and Poverty Division, Technical Department, Africa Region, Washington DC, The World Bank.
Heyneman, S. P. and W. A. Loxley (1983). "The Effect of Primary-School Quality on Academic Achievement across Twenty-nine High- and Low-Income Countries." American Journal of Sociology 88(6): 1162-94.
Jencks, C. (1972). Inequality. New York, Basic Books.
King, E., P. F. Orazem, et al. (1999). Promotion With and Without Learning: Effects on Student Dropout. Impact Evaluation of Education Reforms, Working Paper No. 18., World Bank, Washington DC.
Lam, D. (1999). Generating Extreme Inequality: Schooling, Earnings, and Intergenerational Transmission of Human Capital in South Africa and Brazil. Unpublished Paper.
Lareau, A. (1987). "Social Class Differences in Family-School Relationships: The Importance of Cultural Capital." Sociology of Education 60(2): 73-85.
Lareau, A. (2000). Home Advantage: Social Class and Parental Intervention in Elementary Education. Lanham, Maryland, Rowman and Littlefield.
Levin, H. M. (1995). Raising Educational Productivity. The Second International Encyclopedia of the Economics of Education. M. Carnoy, Pergamon: 283-291.
Lockheed, M. E., B. Fuller, et al. (1988). Family Background and Student Achievement. Policy, Planning and Research Working Papers, WPS 27. Washington, DC, World Bank.
Marjoribanks, K. (1979). Families and Their Learning Environments: An Empirical Analysis. London, Routledge and Kegan Paul.
McGinn, N., F. Reimers, et al. (1992). Why do Children Repeat Grades? A Study of Rural Primary Schools in Honduras. BRIDGES Research Report Series No. 13. Cambridge, MA, Harvard University.
MEC/Projeto Nordeste, World Bank, et al. (1997). Chamada à Ação: combatendo o fracasso escolar no Nordeste. Brasilia.
Mora, M. T. (1997). "Attendance, Schooling Quality, and the Demand for Education of Mexican Americans, African Americans, and Non-Hispanic Whites." Economics of Education Review 16(4): 407-18.
Nadon, I. and S. Normandeau (1997). Can Parents' Involvement with Homework Moderate the Relation between Children's Cognitive Abilities and School Achievement? Washington, DC, Paper presented at the Biennial Meeting of the Society for Research in Child Development.
Patrinos, H. A. and G. Psacharopoulos (1997). "Family Size, Schooling and Child Labor in Peru." Journal of Population Economics 10(4).
Portela, A. L. and E. S. Barretto Bastos (1997). Educação, Escola e Comunidade: Estudo Piloto no Estado da Bahia. Brasília, MEC/Projeto Nordeste.
Post, D. (2001). Children's Work, Schooling, and Welfare in Latin America. Boulder, Colorado, Westview Press.
Psacharopoulos, G. (1973). Returns to Education: An International Comparison. Amsterdam, Elsevier.
Psacharopoulos, G. (1981). "Returns to Education: An Updated International Comparison." Comparative Education 17: 321-41.
Psacharopoulos, G. (1985). "Returns to Education: A Further International Update and Implications." Journal of Human Resources 20(4): 583-604.
Psacharopoulos, G. and A. M. Arriagada (1987). School Participation, Grade Attainment, and Literacy in Brazil: A 1980 Census Analysis. (Discussion Paper EDT 86). Washington, DC, Education and Training Department, World Bank.
Psacharopoulos, G. and et al. (1994). "Education and Earnings in Paraguay." Economics of Education Review 13(4): 321-27.
Psacharopoulos, G. and Y. C. Ng (1994). "Earnings and Education in Latin America." Education Economics 2(2): 187-207.
Psacharopoulos, G. and E. Velez (1994). "Education and the Labor Market in Uruguay." Economics of Education Review 13(1): 19-27.
Ravillion, M. and Q. Wodon (2000). "Does Child Labor Displace Schooling? Evidence on Behavioural Responses to an Enrollment Subsidy." The Economic Journal 110(March): C158-C175.

Ray, R. (2000). "Child Labor, Child Schooling, and Their Interaction with Adult Labor: Empirical Evidence for Peru and Pakistan." The World Bank Economic Review 14(2): 347-67.
Reali, A. M. d. M. R., C. R. Reyes, et al. (2000). "Let's Help Our Children": A Brazilian Experience Involving the Home - School Interaction. São Carlos, SP, Brazil, Federal University of São Carlos. Paper Presented at AERA 2000.
Riddell, A. R. (1989). "An Alternative Approach to the Study of School Effectiveness in Third World Countries." Comparative Education Review 33(4): 481-497.
Riddell, A. R. (1997). "Assessing Designs for School Effectiveness Research and School Improvement in Developing Countries." Comparative Education Review 41(2): 178-204.
Rosenzweig, M. R. (1992). "Educational Subsidy, Agricultural Development, and Fertility Change." Quarterly Journal of Economics February: 67-88.
Rosenzweig, M. R. and R. Evenson (1977). "Fertility, Schooling and the Economic Contribution of Children in Rural India: An Econometric Analysis." Econometrica 45(5): 1065-1079.
Schiefelbein, E., J. Swope, et al. (1999). Cost Effectiveness of Education Programs for Reducing Child Labor, http://www.ust.cl/framearea.asp?cod=13. 2001.
Schwartzman, S. (2000). Evolução e Características do Trabalho Infantil no Brasil, Unpublished Paper.
Silveira, M. L. (2000). O Projeto Nordeste. Background Paper for the Implementation Completion Report. Brasilia, Brazil, FUNDESCOLA/MEC.
Stevenson, D. L. and D. P. Baker (1987). "The Family-School Relation and the Child's School Performance." Child Development 58(5): 1348-57.
Velez, E., E. Schiefelbein, et al. (1993). Factors Affecting Achievement in Primary Education. Human Resources Development and Operations Policy Working Paper. HROWP 2. Washington, DC, World Bank.
Villas-Boas, A. (1998). "The Effects of Parental Involvement in Homework on Student Achievement in Portugal and Luxemboug." Childhood Education 74(6): 367-71.
Wolfe, B. L. and J. R. Behrman (1984). "Who Is Schooled in Developing Countries? The Roles of Income, Parental Schooling, Sex, Residence and Family Size." Economics of Education Review 3(3): 231-45.
Woodhall, M. (1995). Human Capital Concepts. The Second International Encyclopedia of the Economics of Education. M. Carnoy, Pergamon.