See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/262851296

Economics of Education. The Need to go Beyond Human Capital

Theory and Production-Function Analysis

Article in Educational Studies · July 1999

DOI: 10.1080/03055699997864

CITATIONS
READS
38
830

1 author:

Vincent Vandenberghe
Université Catholique de Louvain - UCLouvain
115 PUBLICATIONS 1,391 CITATIONS

All content following this page was uploaded by Vincent Vandenberghe on 30 July 2019.

The user has requested enhancement of the downloaded file.

 Economics of Education.
The Need to go Beyond Human Capital Theory and
Production Function Analysis.

V.Vandenberghe$
July 1998
Published in Educational Studies, Vol. 25, No 2, pp 129-144.

Abstract

Human capital theory takes for granted that an individual's demand of education will
automatically be transformed into real human capital: there is no supply constraint or,
said differently, educational systems can be assimilated to "neutral" black boxes
mechanically transforming input into output. This optimistic view of human capital
production was rapidly abandoned. In the late 60's some economists began to revise the
assumption that the black box's functioning was neutral with regard to human capital
production. Supply-Side Economics of Education appeared first under the form of
production-function analysis. It was initially believed that the information delivered by
such input/output approach would help policy makers and administrators choose the
most productive 'mix' of inputs. But the major conclusion of this line of research is that
there is apparently no clear and undeniable relation between both expenditure per
student and the specific resources they can buy (teachers' degrees and experience,
smaller student-teacher ratios…) on the one hand, and student achievement on the other.
This probably means that further conceptual development is necessary to overcome the
current analytical — and also political — deadlock. This paper argues that one
promising way consists of introducing organisational assets as well as non-monetary
inputs into the production function paradigm.

Key words: Education, Human Capital, Production Function Analysis, Organisational

Efficiency, Peer Effects.

$ Assistant Professor in Economics of Education. Institute Recherché Economiques et

Sociales (IRES) and Department of Education (EDUC), Université Catholique de Louvain. Postal
address: 3, place Montesquieu, UCL, B-1348 Louvain-la-Neuve, Belgium; Tel (+32) 10 47 41 41;
Fax(+32) 10 47 39 45; E-mail: vandenberghe@ires.ucl.ac.be.
JEL Classification: I210 (Analysis of Education), I280 (Education: Government
Policy)

Introduction

Originally, economists neglected the analysis of co-ordination problems inside the

educational sector. Human capital theory — the first contribution to the economics of
educational issues — had almost nothing to say about educational systems and the way
they function. It was solely concerned with the individual demand for education. In its
first version (Becker, 1964), human capital theory presented a cost-benefit analysis
carried out by individuals. Education amounts to an investment that generates a
particular form of capital: human capital. But education policy can no longer be solely
based on human capital theory. The reason for this is twofold. From a purely logical
point of view, the economic analysis of education is bound to incorporate the supply
side in its analysis of human capital production that was previously totally centred on
the demand side. But there are also empirical reasons to this necessity to renew the
economic analysis of education.

1. The theoretical factor

Human capital theory essentially develops a 'black box' approach to production of

education issues. The implicit vision conveyed by this model is that educational systems
mechanically respond to their — private or public — clients 1. One could argue that
human capital theory is excessively based on the implicit assumption that schools and
teachers do not have objectives of their own. As a result educational institutions are
supposed to be passive and not submitted to any constraints.

More and more observers (Gravot, 1993; Jarousse, 1991) consider today that individuals
or governments investing in the educational system will not automatically get 'the best
value for money'. Production of education services is exposed to asymmetry of
information problems, quality control challenges and non-trivial co-ordination strains.
Human capital accumulation is more than individual effort accomplished by pupils and
students who expect some financial return on their investment during the rest of their
life cycle. Both demand side and supply side have to be taken into account and be seen

1
This viewpoint is shared by other social scientists, mainly sociologists, who focus on
social stratification, particularly the role of education in status attainment. The primary interest
of these studies regards the consequences of schooling — the years of schooling being the key
independent variable — for occupational and social mobility (Bryk, Lee & Smith, 1990). In these
contributions, the organizational structure of school is also conceived as a 'black box' whose
internal organization is not central to the analysis. Neither economists nor sociologists during
the 60's offered insight into how the process of schooling actually produces the observed
(desirable or undesirable) outcomes.

2
as a source of regulatory difficulties. A positive economic analysis of education should
at least focus on the outcome of supply and demand interaction. Similarly, a more
normative analysis ought to conceive educational policy instruments aimed at
coordinating both categories of actors.

2. The empirical factors

Human capital theory is optimistic. It promotes the idea that education is a very
powerful individual and social lever. Better educated people and nations will earn more
and prosper at a faster rate. Public investment in education can reduce income inequality
and eradicate poverty. But education did not keep all its promises over the last two
decades. In addition, the public finances' crisis has progressively persuaded decision-
makers that each dollar or franc of tax receipt ought to be spent more efficiently. Quite
logically, those factors (and others) led most observers to the conclusion that supply-
side factors deserved greater attention than in the past.

Several socio-economic trends during the late 70's and the 80's have invalidated human
capital theory's claims in its aggregate version. The automatic connection between the
aggregate level of education and economic growth showed its weakness when growth
rates and productivity gains began to decline. Income inequality exploded dramatically
(Murphy & Welch, 1992, 1993) although educational achievement differentials —
measured by the highest grade completed — considerably shrunk during the same
period (Hanushek, 1992). Economists and other social scientists also viewed education
as the solution to many social challenges including productivity, inequality, health
status, over-population and unemployment (Levin & Kelly, 1994). But again, expected
results didn't show up.

It is true that extra-school parameters have simultaneously favored earnings inequality.

One of them is the skilled-biased technological progress which tends to be particularly
detrimental to individuals with lower skills (Kremer, 1993; Piketty, 1994). Yet
educational outcome per se measured by standardized test score differentials also
revealed disappointing. Some elementary school programs specially designed for 'at
risk' children for example generated encouraging results but those gains vanish rapidly
when these children enter higher levels of the educational process (Glazer, 1986). Other
empirical studies, centred on the educational system's internal performance, contributed
to alter the human capital optimistic prophecy. Hanushek (1986) highlighted a relative
decline in US pupils' standardized test scores: the famous SAT scores. The drop began
in 1967 and lasted until the mid-80's2. More astonishingly, this reversal coincided with a

2
According to Peltzman (1993), the decline was remarkably pervasive, affecting many
different types of students in most grades, in all regions of the United States, in Catholic as well
as public schools and even in Canadian schools. The drop was apparent in the results of
differ-

3
Substantial rise of per-pupil expenditure which was essentially due to a significant
reduction of the average class-size (Hanushek & Rivkin, 1994).

Growing concern about the actual return of educational public expenditure has also to
do with the crisis of the public finances. The mid-70's coincided with the first big public
finances' crisis since the end of World War II. Keynesian economic policies which were
implemented during the late 70's and the early 80's, combined with the economic growth
slow-down and the rise of real interest rates, led to an explosion of public deficits and
gave birth to large public debts. Between 1970 and 1994, most advanced and
industrialized western nations (G7 countries) doubled their public debt/GDP ratio: from
an average of 40 percent, the latter inflated to 70 percent (The Economist, 1995). In
addition, as the population is growing old in most western countries, retirement costs
and health costs are bound to rise significantly over the next decades (Wolfe, 1994).
Especially in Europe, this means growing transfers from active population towards
elderly people. It will thus become more and more difficult to increase the part of the
budget devoted to educational investment (Shoven, Topper & Wise, 1994) unless there
are spectacular growth rates and productivity gains.

I. Inside the black box first — production function analysis

Limitations of the human capital theory and dramatic shifts in the political and
economic context have significantly contributed to a renewal in Economics of
Education. Supply-side Economics of Education appeared first under the form of
production-function analysis (Cohn & Geske, 1990). The basic idea underlying this field
of research was the 'production possibility frontier' commonly exposed in every
undergraduate micro-economics textbook.

1.1. Production function — the concept

Most of these studies — surveyed by Hanushek (1986) and Monk (1992) — are based
on a very simple assumption: education corresponds to some technology which has to
be identified and then efficiently used. Identification necessitates input/output cross-
section empirical analysis. Correlation coefficients between all sorts of educational
inputs (teacher salaries, class-size, capital expenditure…) and outputs (typically
standardized test scores) are computed on a nation-wide scale. Econometric results are
supposed to provide a significant basis for educational decision making. Using the
information delivered by the production function analysis, education policy makers and
administrators can supposedly choose the most productive 'mix' of inputs.
Complementary to human capital theory focusing on the demand side, the production-

rent kinds of tests covering many subject areas. We have not come across information sugges-
ting that a similar test score decline has occurred in Europe or elsewhere in the world.

4
function stream of work represents thus a first step towards a better understanding of the
supply side of the educational process.

1.2. Results

The first well-known empirical study exploiting the production function idea was
carried out by Coleman et al. (1966) 3 its most striking result was that a child's
educational attainment appears essentially correlated to his socio-economic origin.
Additional explanatory variables — essentially schools' monetary resources — seem to
be of little statistical significance. Coleman and his colleagues did not find the expected
positive connection between per-pupil expenditure or average class-size and educational
outcome generally measured by standardized test scores, controlling for initial human
capital endowment (i.e. value-added measures). This very controversial result triggered
off an impressive number of similar studies. In his 1986 survey, Hanushek comments
more than 147 papers containing education production-function results over the last two
decades4. His discussion tends to temper Coleman's very distinct conclusion. Yet, the
same central idea remains. The idea of socio-economic determinism is restated by
Hanushek: a child's socio-economic background is of central importance to predict his
level of educational attainment. Children whose parents are well-off or better educated
get — on average — better academic results. Contrary to Coleman, Hanushek does not
question the idea that school matters. Indeed, he concludes that 'some' schools and
teachers are systematically more productive than others. Nonetheless he heavily insists
on the fact that this observation is not statistically related to the level of inputs with a
monetary expression, particularly the per-pupil expenditure.

This striking result is as controversial as Coleman's initial statement about the absence
of 'school effect'. Since 1986, new estimations have been carried out. Methodological
considerations have come to light and have challenged Hanushek's meta-analysis
approach (a census-like study). Greenwald, Hedges & Laine (1994) dispute his way of
3
Coleman is not the first to consider what goes on in schools in an input output
framework. British economists like Burkhead (1967) carried out production function estimation
in the early 60's. American literature also contains numerous studies on school effectiveness
from the 20's on, but these are not due to economists.
4
It is worth stressing that cross-section production function analyses are logically more
frequent in the USA than anywhere else in the world. The reason for this is twofold. First,
education is heavily decentralized. Per-pupil expenditure, teacher salaries… present substantial
variance (Cohn & Geske, 1990). Thousands of American districts finance their educational
systems with local property tax and this represents an impressive set of individual experiments
that can be used by an econometrician to explore educational technology. Second, standardized
test score measures are available on a nation-wide scale. Econometricians can thus use those
results to create their dependent variables. In most European countries, education is financed
centrally: teacher-to-pupil ratio, wage scales… are determined centrally and this means that
variance is extremely limited. Hence, econometric studies must necessarily be carried out at
international level. But this raises other difficulties: data is not always available or standardized.

5
interpreting successive production function analyses and conclude that monetary inputs
'might' have some influence on educational achievements. They basically use a null test
hypothesis argument to reject Hanushek's point. They argue that the 147 production
function empirical analyses must be interpreted as a set of data randomly sampled. A
relatively small proportion of positive and significant coefficients does not represent a
sufficient condition to reject the null hypothesis5.

Card & Krueger (1992) examined further Hanushek's perplexing conclusion. In their
cross-states empirical study, they regressed post-graduation wages (earnings) on per-
pupil educational expenditure in the state of origin. Contrary to Hanushek, they
conclude that higher public expenditure (a proxy for school quality) does matter: it is
synonymous with higher remuneration at adult age. Their statistical models show a
positive relation between average per-pupil school expenditure in the state and later
earnings. But recent theoretical work (Hanushek, Rivkin & Taylor, 1996) stresses the
potential upward bias affecting regression coefficients when the level of aggregation
becomes higher. Card & Krueger (1992) typically use much aggregated data to create
their per-pupil expenditure variable. This theoretical point finds some echo in empirical
studies using less aggregated data i.e. data collected at school and district levels.
Attempts to replicate Card & Krueger's results at that level have not shown similar
patterns (Betts, 1995).

The debate about education production functions is particularly complex. A first reason
is that researchers do not use the same proxy of educational achievement. Some
researchers, such as Hanushek, focus on test results while others, like Card & Krueger,
exclusively refer to wages. It is true that human capital theory predicts some connection
between education achievement, productivity and individual wages. Yet, we know that
labor market mechanisms — be they internal or external to firms — have their own
rules and interfere with the outcomes of the educational sector. The second source of
difficulty is simply that successive empirical studies come to opposite conclusions
despite their relative homogeneity in terms of data and methodology.

1.3. Conclusion

There is apparently no clear and undeniable relation between expenditure per student
and for example student achievement. The only well-established result is that socio-
economic origin is decisive (Glennerster, 1991). Schools differ dramatically in 'quality'
but this fact cannot be connected statistically or econometrically to rudimentary factors
that many researchers have examined. Differences in quality do not seem to reflect only
variations in expenditure, class size, or other commonly measured attributes of schools
and teachers. Instead, they appear to result from differences between teachers' skills that

5 For a response to this argument, see Hanushek (1994)

6
defy detailed description and empirical causal analysis (Hanushek, 1986). We believe
that this is partly because production function research relies upon a too simple
conception of the educational black box. Regression techniques are more and more
sophisticated but the conceptual background is still very similar to the technological
conception of production conveyed by micro-economics textbooks. Because of data
availability too, most regressions still focus on ratios of outcome proxies (test scores or
wages) to monetary input.

II. Inside the black box second — the need for an enriched conception
of education production

The list of variables that can help us re-conceptualize schooling is rather large. We shall
focus here on those that have been recently explored either empirically or conceptually,
that relate somehow to the aforementioned co-ordination problem and are non-monetary
by nature i.e. cannot be purchased on the marketplace.

2.1. Intra-school organisational efficiency or x-efficiency

Several case studies (Monk, 1992), but also nation-wide empirical research (Hanushek,
1986, 1992) tend to confirm the critical role played by intra-organizational attributes.
The technological relation between inputs and outputs could be conditional to the
presence of organizational assets: the so-called x-efficiency (Leibenstein, 1966; Levin,
1997). These cannot be directly related to the amount of monetary resources made
available by the public authority and cannot be purchased on the marketplace like a
teacher with a certain degree or some formal qualification, facilities, textbooks or
computers. Although organization seems to be very important, its clear comprehension
and control appears problematic. At least two questions must be answered. The first one
relates to the nature of organization. The second corresponds to its production and
diffusion system-wide.

2.1.1. The nature of school organization: coordinating professionals

Schools, like firms or government agencies, require a certain level of co-ordination.

This diagnosis seems reinforced by the fact that education is a joint-product (i.e.
produced by several teachers). In addition, student learning is a process which occurs
over time in a multilevel structure (Bryk, Lee & Smith, 1990). Of course, the
importance of external (e.g. bureaucratic) control can considerably limit the intra-school
organization problem. But in any case at least some residual organizational
responsibility is given to schools. Intra-school co-ordination — typically the school
head's responsibility — supposes some administration: information must be
communicated to teachers and pupils, resources must be allocated… He must also
establish a certain number of rules that

7
decrease school disruption, increase students' safety or protect teachers from excessive
parental interference (Bidwell, 1965). Beyond administration, mediation or 'buffering',
school organization essentially amounts to curricular arrangement and teaching. The
process of learning is actually central to a school's life. It is quite invariably organized
by age and grade level. Contrary to primary education, secondary education also means
subject specialization. This rather specialized learning process is carried out by teachers.
All this highlights the necessity of some co-ordination of individual decisions.

But teachers like to consider themselves as professionals (Weiss, 1990; Maroy, 1992).
The work teachers do is like that of other professionals (doctors, lawyers…): it is
intellectual, cannot be standardized or reduced to routines, and requires preparation
through advanced training. This professional component of school life generates a
situation where teachers expect to have broad control of their daily task in the
classroom. Weiss (1990) judiciously indicates that several crucial domains of
educational decisions are left virtually undisturbed by other control mechanisms in
schools. The day-to-day practices of instruction, evaluation of pupil performance and
maintenance of order are usually left in the hands of individual teachers. This
observation raises questions about the scope of the school organization idea. Maroy
(1992) insists on the fact that schools are generally characterized by a certain level of
'structural looseness'. Co-ordination attempts — be they internal or external — are
always a priori limited in magnitude by the presence of professionals who tend to fight
to preserve their independence.

Hence, organizational efficiency in the educational sector implies quite invariably that
rulers find the right balance between co-ordination requirements and professional
autonomy. Schools where such a balance was successfully implemented apparently
exist. The whole question is to determine whether these schools are ruled by
outstanding, but quite uncommon, principals or whether they present distinct
organizational — structural — features, combined in a very specific and identifiable
manner which ensure a certain efficiency, no matter the personality of the staff in
charge.

Hanushek seems to support the first assumption. In his words 'organization defies both
description and prescription'. Cousin (1993) brings in nuances. Efficient organizational
practices can be identified. What is at stake is essentially the capacity to induce a
significant proportion of the staff to partly abandon the grade-classroom-subject
reference and structure their work by reference to the whole school. But Cousin
concludes that factors permitting this kind of mobilization are still relatively unknown.

Levin (1994) is more optimistic. He argues that necessary conditions for a good
educational organization largely correspond to the conditions defining efficient firms. In

8
brief, the more 'productive' schools Hanushek (1986) presented as 'outliers' could be
viewed as schools that have managed to combine some of the five organizational
attributes. First, schools must be clear about what they are attempting to achieve. This
supposes that there is a widespread acceptance and agreement by all participants. This
objective must be associated with measurable outcomes in order to appraise what the
school is doing. Second, teachers and principals must get incentives tied to student
success. Incentives can be intrinsic (e.g. a sense of accomplishment) or extrinsic (e.g.
financial reward or recognition either by hierarchy or peers). They can be individual or
collective but their very purpose is to stimulate effort and increase accountability. Third,
information must be made available concerning existing pedagogical possibilities, test
scores and their evolution. The absence of rapid feed-back can also prevent the
implementation of a trial and error innovation strategy. Fourth, schools evolve in a
changing environment and must constantly adapt to meet new individual and social
demands. Adaptability is thus of great importance to achieve organizational efficiency.
Fifth, schools must be able to adopt the most productive teaching technology consistent
with budgetary constraints.

2.1.2. How to diffuse efficient organizational arrangements?

To define and implement the appropriate balance between professional autonomy and
co-ordination procedures à la Levin for example in a particular school is a non-
negligible challenge. But its reproduction and dissemination system-wide is even more
complicated. In the US, several school renovation programs for at-risk pupils are
currently experimented. There is the accelerated schools project promoted by Professor
Levin from Stanford or the one sponsored by the John Hopkins University (Slavin & al.,
1990). These programs aim at a gradual dissemination of educational practices that have
proved particularly adapted to at-risk pupils' needs. First results seem very promising.
However, the key point is to appreciate the 'survival' prospect of those results once their
very enthusiastic, skilled and dynamic advocates disappear or are forced to delegate
because, almost by definition, their capacity to control their project is limited beyond a
certain level of development.

Traditionally, economists argue that the generalization of 'good' organizational features

heavily depends on a nexus of external incentive and co-ordination mechanisms
(Laffont & Tirole, 1993; Milgrom & Roberts, 1992). In other words, a regulatory
principle is necessary to organize individual actions and orient independent decision-
makers towards a certain end. Information circulation is important but most economists
apparently believe that 'incentives' are more important. The new theories of regulation
(Baron, 1989) tend to focus on 'effort' incentive problems. Several external incentive
mechanisms are studied by this literature. Some rely on output-based remuneration or

9
promotion schemes defined by the regulator. Other regulatory approaches are more of
the market type.

Many analysts (Chubb & Moe, 1990; Lankford & Wyckoff, 1992) believe that school
choice is crucial to improve the overall efficiency of education systems but the
empirical evidence regarding the impact is still contrasted suggesting that further work
is necessary. Glenn (1989) claims, at the end of his multi-country investigation (France,
Belgium, the Netherlands, Canada, UK, and Germany), that the case for choice-driven
efficiency is weak. In his view, the danger of extended segregation dominates any other
benefit. He underlines the crucial role played by some regulatory requirements in order
to limit the propensity of choice to generate segregation. Hoxby (1994) by contrast
concludes that more school choice causes better educational achievement with no
increased sorting or segregation. Those results are interesting but problematic because
the US have no real 'voucher' or quasi-market system 6. Voucher plans have not yet been
adopted on a state-wide scale in the US. Hoxby argues that is it already possible to
estimate the effect of easier parental choice. Indeed, variation in the ease of choosing
among public schools already exists because metropolitan areas differ greatly in their
public school enrolment concentration. But this sort of choice is very much constrained.
Some parents must live in a particular district simply because their employer is there.
School choice through residential mobility is also heavily influenced by real estate
markets. Educational choice in the US is thus wealth-contingent.

2.2. Peer effects

We have argued above that intra-school organization potentially enhances the

productivity of monetary inputs. A lot of work is still necessary to identify the
mechanisms through which organizational assets can emerge and be disseminated
system-wide. The point we put forth here is that organization and money cannot
accomplish everything. A large body of recent research stresses the importance of social
interactions. The latter, if properly mobilized, can considerably buttress human capital
production and usefully complement monetary input and organization.

2.2.1. Peer effects: the idea

In the school context, these social interactions are called 'peer' effects — some people
also use the term 'contextual' effects. The underlying idea is that the knowledge a child
assimilates during a school year depends directly on the characteristics or actions of his
comrades. In other words, education is characterized by social spillovers. Note that the

6 More than 85% of US pupils attend public schools at elementary and secondary level.
This figure is relatively constant since the end of world-war II.

10
concept is far from being specific to educational problems. Recent empirical evidence,
highlighting the importance of social interaction, has developed in several contexts:
teenager pregnancy, drug addiction, inter-generation and ghetto poverty (Jencks &
Meyer, 1987; Corcoran, Gordon, Laren & Solon, 1990; Dynarski, Schwab & Zampelli,
1989; Evans, Oates & Schwab, 1992). Case & Katz (1991) provide evidence that the
probability of social ills in one neighborhood increases with the prevalence of the same
ills in adjacent neighborhoods.

In the educational context, Coleman et al. (1966) were the first to defend the social
interaction idea. Most observers retained their controversial conclusion that objective
attributes (monetary inputs) of school have little impact on achievement. The latter's
most significant determinant is simply the student's socio-economic background. It is
less known that Coleman and his co-authors also insisted on the importance of peer
characteristics. Since Coleman, several empirical studies have come to the same
conclusion: the quality of social interactions heavily influences educational achievement
(Summers & Wolfe, 1977; Henderson, Mieszkowski & Sauvageau, 1978; Duncan 1994;
Dynarski, Schwab & Zampelli, 1989).

2.2.2. Peer effects: conceptual status and regulatory implications

Conceptual and political challenges raised by the idea of peer effects should not be
underestimated. From a theoretical standpoint, peer effects amount to externalities
(spillovers). This notion was already present in the human capital theory: the more
educated a worker, an employee or a staff member, the higher his productivity and his
colleagues' productivity. In most public finance textbooks, education is also
synonymous with richer social, cultural and political life, better public health, less crime
(OECD-CERI, 1998).

Yet, the peer effect concept bears some innovation with regard to the concept of
externalities as it has been traditionally presented and exploited by human capital
theory. First, peer effect is an externality that operates during the production of human
capital and not only later, when individuals as adults get involved in socio-economic
life. Second, and more essentially, peer effects are local. This means that the basic co-
ordination problem is to control the allocation of individuals between schools. Social
interaction is a local phenomenon and takes place in bounded entities (schools) that are
separated from each other. Individuals attending classes somewhere in the educational
system possess different human capital endowments. Some originate from well-off
families while others come from poor families, have parents with no or poor educational
records. When grouped in a particular school or classroom, what level of externality do
they benefit from? The question would be senseless in a world of permanent and
boundless social interaction. In that very unrealistic situation, each individual would

11
permanently be exposed to the sum of externalities dispersed by the rest of mankind.
Real life is slightly different. A 'rich' individual attending a school and generating some
(positive) externality in that particular school is 'lost' for the other schools. Externalities
conveyed by individuals are almost by definition spatially limited. Their diffusion is not
universal. In most situations, it is limited by the size of the entities they choose (or are
obliged) to live in.

Consequently, allocation of heterogeneous individuals between strictly delimited

entities thus becomes a critical issue (Vandenberghe, 1996, 1998). Educational policy
must bring some answer to at least two problems. The first one is to identify the optimal
allocation of individuals across schools. The second problem is to get the individuals
(the pupils and their parents) to accept this allocation.

Should individuals be sorted between homogeneous entities (i.e. schools wherein all
pupils have the same ability) or should they be placed in heterogeneous ones (i.e.
schools mixing ability levels)? Allocation of heterogeneous individuals relates here to
'productive' efficiency problems: the production of human capital is directly affected by
the way heterogeneous individuals are allocated. A social objective consisting of
maximizing the average level of human capital can be compromised if individuals are
inappropriately allocated among schools. The same is true with an egalitarian objective
aiming at equalizing educational achievement. The cost of this policy is potentially
influenced by the way peer effects are allocated among schools (Vandenberghe, 1996).

In addition, a social planner who would want to mobilize peer effects in a certain way
(segregation or mixing) could be confronted with a preference-incompatibility problem
among pupils and parents. If individuals are sensitive to peer effects (i.e. are aware of
their importance), they are also sensitive to the allocation of relevant human
characteristics among schools. For example, the desire of 'rich' individuals to segregate
from the 'poor' in order to maintain the peer effect component at its highest level can be
challenged by the desire of 'poor' people to benefit from this social input. Some co-
coordinating mechanism must exist to ensure minimal compatibility between conflicting
individual preferences. This raises a co-ordination problem that basically consists of
finding ways to make sure that decentralized decision-makers (schools, teachers, parents
in the school context) properly 'internalize' the existence of peer effects.
Correspondence between social priorities and each individual decision-maker cannot be
taken for granted.

Conclusion

We essentially retain from this discussion that the supply side of the educational process
can no longer be represented as a simple (and neutral) black box. Our survey of

12
production function analyses — a first attempt to overcome the black box assumption
implicitly made by the human capital theory — has led to the conclusion that no clear
and indisputable relation exists between both expenditure per student and specific
resources they can buy on the one hand, and student achievement on the other. The only
well-established result is that socio-economic origin is decisive. Schools differ
dramatically in 'quality' but this appears to result from differences between teachers'
skills that defy detailed description and empirical causal analysis.

We have argued that the reason for this deadlock could be that traditional production
function research in Economics of Education relies upon a too simple — actually too
mechanical — conception of the production process. It is common to present production
possibility frontiers as a purely technical relationship, void of any economic content.
From our point of view, some conceptual development around the idea of human capital
production is necessary to overcome the analytical — and also political — limitations
illustrated for example by the endless 'school quality' debate in the US.

We have attempted to achieve this objective by focusing on two ideas: intra-school

organization or x-efficiency and social interaction synonymous with social and local
spillover among pupils. Several case studies tend to confirm the critical role played by
intra-organizational attributes. But internal organization of schools is a very puzzling
issue wherein numerous variables play a role (administration, curricular arrangements,
scheduling, tracks….). In addition, the generalization of 'efficient' intra-school
organizational attributes constitutes a rather uneasy task. Regarding this issue, one idea
that gains in importance in the academic and political debate is the key role played by
the governance structure in which schools are embedded, in particular a structure that
incorporates market-oriented mechanisms (Glennerster, 1991; Vandenberghe, 1996).

The second idea we put forward is that intra-school organization is only one face of the
coin. The capacity to mobilize social input is probably as important. In the school
context, these social inputs amount to social interactions called 'peer' effects— some
people also use the term 'contextual' effects. In other words, education is one of those
numerous human activities characterized by social spillovers. This means the
implementation of socially optimal education policy is conditional to the identification
of the optimal allocation of individuals (by type or ability level) across schools. It also
heavily depends on the ability of decision-makers to get the individuals (the pupils and
their parents) to accept this allocation.

13
Bibliography

Baron, D.P. (1989) Design of Regulatory Mechanisms and Institutions, in: R.

Schmalensee & R.D. Willig (Eds) Handbook of Industrial Organization (The
Netherlands, Elsevier Science Publishers).
Becker, G. S. (1964), Human Capital (New York, Columbia University Press).
Betts, J.R. (1995) Does School Quality Matter? Evidence From the National
Longitudinal Survey of Youth, The Review of Economics and Statistics, 77(2), pp.
231-247.
Bidwell, C.E. (1965) The school as a formal organization, in: J.G. March (Ed.),
Handbook of Organizations (Chicago, Rand McNally).
Bryk, A., Lee, V. & Smith, J. (1990) High School Organization and its Effects on
Teachers and Students" in: W.H Clune. & J.F Witte. (Eds), Choice and Control in
American Education, Vol. 1&2 (London,The Stanford Series on Education & Public
Policy, The Falmer Press).
Burkhead, J. (1967) Input-output in large city high schools (Syracuse, NY, Syracuse
University Press).
Card, D. & Krueger, A.B. (1992) Does School Quality Matter? Return to Education and
the Characteristics of Public Schools in the United States, Journal of Public
Economics, 100(1), pp. 1-40.
Case, A.C. & Katz, L.F. (1991) The company you keep: The Effects of Family and
Neighborhood on Disadvantaged Families, National Bureau of Economic Research
Working Paper, 3705, Cambridge, Ma.
Chubb, J.E. & Moe, T.M. (1990) Politics, Markets and America's Schools (Washington
D.C, The Brookings Institution).
Cohn, E. & Geske, T.G. (1990), The Economics of Education (Oxford, UK, Pergamon
Press).
Coleman, J.S., Campbell, E.Q., Hobson, C.F., Mc Partland, J., Mood, A.M., Weifeld,
F.D. & York, R.L. (1966) Equality of Educational Opportunity (OE-38001,
Washington, DC: US Office of Education, US Department of Health, Education and
Welfare).
Corcoran, M., Gordon, R., Laren, D. & Solon, G. (1990) Effects of Family Community
Background on Economic Status, American Economic Review, 80(2), pp. 362-366.
Cousin, O. (1993) L'effet établissement. Construction d'une problématique, Revue
française de sociologie, 34, pp. 395-419.
Duncan, G.J. (1994) Families and Neighbors as Sources of Disadvantage in the
Schooling Decisions of White and Black Adolescents, American Journal of
Education, 103, pp. 20-53.
Dynarski, M., Schwab, R. & Zampelli, E. (1989) Local Characteristics and Public
Production: The Case of Education, Journal of Urban Economics, 26, pp. 250-263.
Evans, W.N., Oates, W.E. & Schwab, R.M. (1992) Measuring Peer Group Effects: A
Study of Teenage Behavior, Journal of Political Economy,100(5), pp. 966-991.
Glazer, N. (1986) (+ comments by Jencks, C.) Education and Training Programs and
Poverty, in: S. Danziger S. & D.H. Weinberg (Eds), Fighting Poverty (Madison,
Washington, Madison's Institute for Research on Poverty and Department of Heath
and Human Services).

14
Glenn, C.L. (1989), Choice of school in six nations: France, Netherlands, Belgium, UK,
Canada, West-Germany (Washington, DC, US Department of Education).
Glennerster, H. (1991) Quasi-Markets for Education?, Economic Journal, 101, pp.
1268-76.
Gravot, P. (1993), Economie de l'Education (Paris, Economica).
Greenwald, R., Hedges, L.V. & Laine, R.D. (1994) When Reinventing the Wheel is Not
Necessary: A Case Study in the Use fo Meta-Analysis in Education Finance,
Journal of Education Finance,20, pp. 1-20.
Hanushek, E. (1986) The Economics of Schooling: Production and Efficiency in Public
Schools, Journal of Economic Literature, 24, pp. 1141-1177.
Hanushek, E. (1992) Improving Educational Outcomes While Controlling Costs,
Carnegie-Rochester Conference Series on Public Policy, 37, December, pp. 205-
238.
Hanushek, E. & Rivkin, S.G. (1994) Understanding the 20 th Century Explosion in U.S.
School Costs, Journal of Human Resources, 32(1), pp. 35-68.
Hanushek, E. (1994) Money Might Matter Somewhere: A Response to Hedges, Laine,
and Greenwald, Educational Researcher, 23(4), pp. 5-8.
Hanushek, E., Rivkin, S.G & Taylor, L.L. (1996) Aggregation and the Estimated Effects
of School Resources, Review of Economics and Statistics, 78(4), pages 611-27.
Henderson, V., Mieszkowski, P. & Sauvageau, Y. (1978) Peer Group Effects and
Educational Production Functions, Journal of Public Economics, 10(1), pp. 97-106.
Hoxby, C. M. (1994) Does Competition between Public Schools Benefit Students and
Taxpayers? National Bureau of Economic Research Working Paper, 4978,
Cambridge, Ma.
Jarousse, J. P. (1991) L'économie de l'éducation: du 'capital humain' à l'évaluation des
processus et des systèmes éducatifs, Repères Bibliographiques. Perspectives
documentaires en éducation, 23, pp. 79-101.
Jencks, C. & Meyer, S.E. (1987) The Social Consequences of Growing Up in Poor
Neighborhood, in: L.E. Lynn L & M.G.H. MacGeary (Eds), Inner-City Poverty in
the United States (Washington, D.C., National Academy Press).
Kremer, M. (1993) The O-Ring Theory of Economic Development, Quarterly Journal
of Economics, 108(3), pp. 551-575.
Laffont, J.J & Tirole, J. (1993) A Theory of Incentives in Procurement and Regulation
(Cambridge, Ma, MIT Press).
Lankford, H.L. & Wyckoff, J. (1992) Primary and Secondary School Choice Among
Public and Religious Alternatives, Economics of Education Review, 11(4), pp. 317-
337.
Leibenstein, H. (1966) Allocative Efficiency and X-Efficiency, The American Economic
Review, 56, pp. 392-425.
Levin, H.M. (1994) Economics of school reform for at-risk students, in: Hanushek et
alii. (Eds), Making Schools Work: Improving Performance and Controlling Costs
(Washington D.C., The Brookings Institution).
Levin, H.M. & Kelly, C. (1994) Can Education do it Alone?, Economics of Education
Review, 13(2), pp. 97-108.
Levin, H.M. (1997) Raising School Productivity: An X-Efficiency Approach,
Economics of Education Review, 16(3), pp. 303-311.

15
 Maroy, C. (1992) L'école à la lumière de la sociologie des organisations, Etudes &
Documents, février, Unité de sociologie, UCL, Louvain-la-Neuve.
Milgrom, P. & Roberts, J. (1992) Economics Organization and Management
(Englewood Cliffs, N.J., Prentice Hall).
Monk, D.H. (1992) Education Productivity Research: An Update and Assessment of Its
Role in Education Financial Reform, Education Evaluation and Policy Analysis,
14(4), pp. 307-332.
Murphy, K.M. & Welch, F. (1992) The Structure of Wages, Quarterly Journal of
Economics, 152(1), pp. 285-326.
Murphy K.M. & Welch F. (1993), Inequality and relative wages, American Economic
Review, 83(2) pp. 104-109.
Peltzman, S. (1993) The Political Economy of the Decline of American Public
Education, Journal of Law & Economics, 36, pp. 331-370.
Piketty, T. (1994) Inégalités et redistribution ; développements théoriques récents,
Revue d'Economie politique, 104(6), pp. 769-800.
Shoven, J.B., Topper, M.D. & Wise, D.A. (1994) The Impact of Demographic
Transition on Government Spending, in: D.A. Wise (Ed.), Studies in the Economics
of Ageing (Chicago, The University of Chicago Press).
Slavin, R.E., Madden, N.A., Karweit, N.L., Dolan, L. & Wasik, B.A (1990) Success for
All (Baltimore, M.D., Center for Research on Effective Schooling for Disadvantaged
Students, John Hopkins University).
Spence, M.A. (1974) Model 2: Education as a signal, Chapter 3 of Market Signalling.
Information Transfer in Hiring and Related Screening Processes, Harvard
University Press, Cambridge, Ma.
Summers, A.A. & Wolfe, B.L. (1977) Do Schools Make a Difference?, American
Economic Review, 67(4), pp. 639-652.
Taubman, P.J. & Wales, T.J., (1973) Higher Education, Mental Ability and Screening,
Journal of Political Economy, No 81, pp. 28-55.
The Economist (1995) May 27th-June 2nd, page 17.
Vandenberghe, V. (1996) Functioning and Regulation of Eductional Quasi-Markets
(Louvain-la-Neuve, CIACO, Nouvelle série No 283).
Vandenberghe, V. (1998) Educational Quasi-Markets: the Belgian Experience, in: W.
Bartlett, J.A. Roberts & J. Le Grand (Eds), A Revolution in Social Policy. Lessons
from developments of Quasi-Markets in the 1990s (Bristol,The Policy Press).
Weiss, J. (1990) Control in School Organizations: Theoretical Perspectives, in: W.H.
Clune & J.F. Witte (Eds), Choice and Control in American Education, Vol.1 &2
(London, The Stanford Series on Education & Public Policy, The Falmer Press).
Wolfe, J.R. (1994) Review of the coming health crisis: Who will pay for the aged in the
twenty-first century?, Journal of Economic Literature, 32(3), pp. 1254-1255.

16

View publication stats