University of Southern California Dissertations and Theses

A study concerning the relevance of using quality measures of education in economic growth models of sub-Saharan African nations

(USC Thesis Other)

A study concerning the relevance of using quality measures of education in economic growth models of sub-Saharan African nations

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content A STUDY CONCERNING THE RELEVANCE OF
USING QUALITY MEASURES OF EDUCATION
IN ECONOMIC GROWTH MODELS OF
SUB-SAHARAN AFRICAN NATIONS
by
Jeffrey Michael Ishii
A Thesis Presented to the
FACULTY OF THE GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
MASTER OF ARTS
ECONOMICS
May 2001
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
UMi Number: 1406452
___ ®
UMI
UMI Microform 1406452
Copyright 2001 by Bell & Howell Information and Learning Company.
All rights reserved. This microform edition is protected against
unauthorized copying under Title 17, United States Code.
Bell & Howell Information and Learning Company
300 North Zeeb Road
P.O. Box 1346
Ann Arbor, Ml 48106-1346
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
U NIVERSITY O F S O U T H E R N CALIFORNIA
THE GRADUATE SCHOOL
UNIVERSITY PARK
LOS ANGELES. CALIFORNIA 8 0 0 0 ?
This thesis, written by
J e f f r e x J l i t-k ag, f ______ _ _
under the direction of hiik Thesis Committee,
and approved by all its members, has been pre
sented to and accepted by the Dean of The
Graduate School, in partial fulfillment of the
requirements for the degree of
_______________ ffet§JtS £-Q £.is£ta___ ___________ _
£
* rv/1 r -
B m
D ate—
THESIS COMMITTEE
Jf Jf /
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
ii
Dedication
I have great appreciation for my wife and children.
This thesis is dedicated to Vivian, Jordan, Jaret, and Kaitlyn.
Their support for my work is always unconditional and unwavering.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Table of Contents
lii
I. Introduction 1
II. Macroeconomic Growth Models with Human 9
Capital
HI. Microeconomic Schooling-Retums Models 21
IV. Educational Research Concerning Quality 32
of Education
V. Current Situation of Sub-Saharan Africa 52
VI. Conclusion 59
VII. Bibliography 66
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
iv
Abstract
Recent literature within the field of economics has studied the
relationship between human capital and economic growth. Using the
quantity of education as a proxy for human capital, these studies have
led to an unclear understanding of the mechanism between education
and development. Many economists support the need for a complete
specification of this relationship. Included in these approaches is the
application of measurements of educational quality. This paper will
provide theoretical support for two models to explain the relationship
between improvements in educational quality and economic
development. The first is a macroeconomic model attempting to
demonstrate this relationship by using variations in test scores on
international exams. A second microeconomic model is discussed that
accounts for variation in inputs as well as contextual factors and
pedagogical differences in the classroom.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
1
I. Introduction
For many years, the original Solow model o f economic growth received most of
the attention regarding the mechanism behind the progress o f developed and
developing nations. Utilizing physical capital as the major input in economic growth
and assuming the rate o f population growth and savings rates as exogenous, the
Solow model was consistent with many empirical observations o f economic growth.
More than half of the cross-country variation in income has been found to be
explained by the exogenous variables of savings and the rate of population growth.
However, as pointed out by ManMw, Romer, and Weil (1992) this finding only
predicts the direction and not the magnitude o f the relationship. For this reason, they
find a need for a human capital component to be used in addition to physical capital.
The assumption of decreasing returns to scale o f physical capital, the main input,
led to the prediction o f convergence of the income levels o f different nations. It was
this finding o f the Solow model that seemed to be inconsistent with the results of
many developing nations into the 1980s. The income o f many developing nations
began to diverge from the richer nations. This apparent shortcoming of the Solow
model led to the development o f the endogenous growth models of the late 1980s
and 1990s. Using a model which assumes constant or increasing returns to scale of
the factors o f production, these models were able to explain the divergence o f many
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
2
developing nations. In doing this, they utilized the variation o f human capital as an
additional determinant of economic growth.
Dating back to the 1966 study of Nelson and Phelps (1966), the effect o f human
capital on the economic growth of a nation has received considerable attention
within the field o f economics. However, as pointed out by Gemmell (1996), the
endogenous growth literature o f the late 1980s involved the first significant attempt
to model economic growth as explained by human capital. Throughout the 1990s,
this consideration within endogenous growth models and attempts such as that by
Mankiw, Romer, and Weil to model human capital within the framework o f a
textbook Solow model have led to numerous seemingly conflicting studies. This
paper will seek to frame the relationship between education and economic growth in
a way that accounts for the contradictions that result from these various viewpoints.
The first section o f this paper will study valuable findings from the recent growth
literature that focus on the relationship between economic growth and human capital.
As mentioned, the endogenous growth literature will be central to these models.
Various theoretical and empirical models will be discussed in terms o f their
relevance. This section will demonstrate the lack o f consensus among economists
regarding the mechanism between human capital and economic growth. A key
finding in this section is the difficulty that economists have experienced in agreeing
on the proxy to include as a measure o f human capital. The predominant usage of
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
3
measurements of the quantity o f education have led an incomplete picture in
describing the relationship between human capital and economic growth.
The result of this finding is the desire to design human capital measurements that
are more reflective o f the relative effectiveness o f the schooling systems o f the
nations that are studied. The assumption seems to be that simply increasing the
quantity o f education will lead to a more productive workforce and corresponding
economic growth. The inconsistent findings of economic theorists might be partly
explained by the inability o f the quantity measurements to capture the variance in the
actual educational products being delivered across different societies. This lack of
consensus among economic models and theoretical intuition based on educational
methodologies guide this study to search for measures of human capital that also
reflect the quality of education that is being delivered in the classroom.
This desire to use quality measurements o f education as a proxy for human capital
is supported by recent developments within the field. The second section o f this
paper will clearly demonstrate that microeconomic evidence supports the
relationship between improvements in the quality o f education and various returns to
these factors. Furthermore, these returns are shown to result from more than a
simple signaling effect. Pritchett (1997) cites the use ofMincerian regressions to
demonstrate the fact that workers with higher levels o f education earn higher wages.
He cites other studies to demonstrate that the relationship does not exist simply
because education is a signal. The contradiction that remains is clear. If the returns
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
4
to education are a result o f more than just a signal concerning workers, then there
most be some corresponding increase in their abilities and accumulated human
capital. The issue that needs to be solved is, if workers are experiencing an increase
in their abilities and productivity, then why is this relationship not reflected
consistently in empirical studies of economic development?
Barro and Lee are among the greatest contributors o f data in this field. Recently,
they have begun to discuss the importance o f alternative measures of human capital
that are more difficult to measure, but provide additional insights into the
relationship between education and growth. (Barro and Lee (2000)) In a recent
paper, they discuss the fact that their commonly used data on educational attainment
“... does not take account of differences in the quality o f schooling across countries.”
(p. 12) In this same paper, they discuss the importance, o f capturing the interactions
between the quantity and quality o f education within a large enough sample to
conduct empirical studies. They support using measures such as international test
scores, differences in adult literacy rates, and estimates of various labor-market
outcomes to model the variation in the quality o f education across countries.
One o f the clearest arguments in the recent literature supporting this quality of
education issue is explained by Temple (1999). As he states, “Another problem to
emerge is that changes in human capital appear to explain little o f the variation in the
changes in output.” (p. 139) This seeming contradiction is also supported by
Pritchett (1997) in his recent paper. He discusses the dramatic increases in the
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
5
enrollment rates of education in developing nations. Among developing nations,
gross primary and secondary enrollment rates have increased from 66 and 14 percent
to nearly 100 and 40 percent respectively since 1960. However, during this time, the
incomes of these nations have diverged from the developed countries of the world.
This lack of consistency between the theoretical relationship and actual experiences
o f developing nations is also discussed in the work of Schultz (1999). In his recent
paper, he supports studies that also consider quality improvements in education.
Pritchett (1997) and Temple (1999) both feel that the use of enrollment ratios as
used by ManMw, Romer, and Weil is problematic. Pritchett finds that the
relationship between estimates of the growth o f the human capital stock and primary
and secondary enrollment rates is strong and negative since 1960. Temple supports
using human capital indices that are more reflective o f changes in the quality of
education within the developing nations. Regarding the apparent trade-off, he states,
“What matters most, breadth o f access or school quality?” (p. 139)
The discussion concerning the quantity and quality o f education in developing
nations provides the primary motivation for this study. With the lack of relevant
studies involving educational quality in the field of macroeconomics, the focus turns
to microeconomic studies. As mentioned, the second section o f this paper will
utilize contributions from the field o f microeconomics in order to understand the role
o f educational quality. This section will study various methods previously used to
measure the quality of education and quantify the returns in economic terms. The
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
6
work of economists in studying the effect o f education on microeconomic returns is
more diverse than the macroeconomic applications. Among these measurements, the
most numerous proxies for the quality o f education are pupil-teacher ratios, average
term length, textbook provision, and teacher training and salaries. This section will
demonstrate that there are still difficulties with the empirical studies within this field.
The most significant difficulty for economists has been conducting studies in a
randomized and controlled fashion without the influences of other factors. In this
regard, the section will focus on the strength o f such studies as the Tennessee
Student/Teacher Achievement Ratio experiment. (Krueger, 1999)
Clearly there is an argument against the use o f the microeconomic studies in the
second section. The relationships that are being studied are often measuring the
private returns to education as opposed to the social returns. Pritchett (1997)
discusses the significance o f the difference between private and social returns to
education. He hypothesizes that, at least with some types o f education, the private
return is greater than the return to the society as a whole. As mentioned, these two
measures will also clearly diverge if education serves primarily as a signal with no
corresponding increase in the ability o f the student/worker. However, if one accepts
that education does improve the ability o f the student/worker, which is then
translated into private returns, then the connection between microeconomic effects
and macroeconomic effects becomes significantly more plausible. This issue will be
considered in more detail in this section.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
7
Another issue of using quality measurements is found in the same paper by
Pritchett (1997). In discussing the lack o f evidence to explain the inconsistencies
found between increasing school enrollments and poor economic growth, he
considers whether this is a simple issue of excluding quality factors at the
macroeconomic level. He cites several sources in stating, “...the general expectation
is that the exclusion o f quality o f education measures will bias the estimate of the
returns to the quantity of schooling upwards.” (p. 32) The implication is that
including quality measurements will not strengthen the findings. However, this
belief is based on one key assumption. He assumes that there is a strong, positive
covariance between the quantity and quality of schooling. However, he provides
little evidence to support this assumption. In fact, in a finite period o f time with the
severe budget constraints o f developing nations, there are reasons to believe this to
be an inverse relationship. With his reasoning, he theorizes a decreased role for the
quantity o f education in an empirical model when quality measures are included.
Even if one believes that previous measures o f school enrollment ratios were poor
proxies o f human capital, he still feels there is little reason to include both measures.
The next step in designing a study demonstrating the relationship between
educational quality and economic growth is determining which quality factors are
most important. The third section o f this paper will address this issue. Where the
previous section seeks to study quality relationships resulting from empirical studies
within the field o f economics, this section will broaden the scope. Most of the
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
8
literature in this section comes from the field of education. A key finding in this
section is the difference between quality o f education factors and important
contextual issues of a society. This section will demonstrate that there is a need to
not only model quality measurements o f education, but to also consider political,
societal, and cultural factors that might create additional variance between nations.
A recent paper by Lee and Barro (1997) demonstrates how this can be done in an
empirical study. In modeling the differences in school quality across nations as
measured by test scores and dropout and retention rates, they include an “East Asia”
dummy variable that accounts for “.. .the cultural and religious features unique to the
East Asian countries.” (p. 25) This section, as well as the future course o f studies
implied by this paper, will support this type o f complete specification of the
education production functions of nations.
The final section of this paper seeks to apply the most recent experiences of the
Sub-Saharan African nations. If one accepts the finding o f the previous section that
the unique experiences and situation o f a society will affect their educational needs,
then this type o f understanding is extremely important to the formulation o f any
model or policies that might result. The goal o f this section will be to discuss the
relevant characteristics o f these nations today in suggesting future methods to model
the relationship between the educational policies of these nations and their
corresponding economic development.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
9
II. Macroeconomic Growth Models with Human Capital
The 1966 study of Nelson and Phelps (1966) was among the first to include
human capital as an important factor in the economic growth of nations. Among
their observations, they caution economists from treating all workers the same in
assuming that highly educated workers are perfect substitutes for less educated
workers. A second conclusion is that education is more relevant to some jobs than it
is to others. They also argue that education increases a worker’s ability to utilize
information and that this ability to process and interpret information is important for
workers in many settings. This apparent externality of human capital will be
discussed in other applications throughout this paper.
After recognizing the deficiencies o f neoclassical growth models in explaining the
poor performance o f developing nations throughout the 1980s, Lucas joined other
economists in formulating a new explanation. The resulting endogenous growth
models sought to explain the lack o f empirical evidence for convergence during this
period. Lucas supported the importance o f human capital. In explaining why poor
countries do not converge to the income levels of richer countries, he theorized the
discrepancies to result from differences in human capital between nations. (Lucas,
1990) Among other early endogenous growth models that supported the application
o f human capital, theoretical models by Buiter and Kletzer discuss the work of Lucas
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
10
concerning the usage o f human capital with increasing returns to scale. They support
the modeling of the externalities o f education that were originally discussed by
Nelson and Phelps. (Buiter and Kletzer, 1991) This study, and a second theoretical
model by the same authors, further support the importance o f human capital in
explaining economic growth and persistent differentials between nations. (Buiter and
Kletzer, 1992)
Among the first empirical studies o f this relationship was the paper by Mankiw,
Romer, and Weil (1992). This study was important for a variety of reasons. They
support the implications o f previous theoretical models in finding a positive role for
human capital on economic growth. However, their model is counter to the
immediately preceding models that discounted the assumptions of the original Solow
model. They specify an augmented Solow model that possesses a consideration for
human capital measured in terms o f the percentage o f the population that is enrolled
in secondary school. This augmented Solow model supports the importance of
human capital in finding that its inclusion along with differences in savings rates and
population growth explain cross-country differences in income per capita.
A central result of the Mankiw, Romer, and Weil study is its attempt to align the
Solow model with actual economic experiences. This issue is important as the lack
of convergence played a large role in the creation of the endogenous growth models.
Barro (1991) finds that convergence does occur under certain circumstances. In an
empirical model using school enrollment ratios at the secondary and post-secondary
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
11
levels, he finds a positive role for initial human capital as related to the growth rate
o f real per capita GDP. Important in this discussion are the externalities inherent in
education such as those discussed by Nelson and Phelps, Romer, and Lucas. In a
later study, Barro and Lee (1993a) find a positive role for human capital, as
measured by secondary school attainment and life expectancy. In this study as well,
he supports his conditional convergence effect. The argument is that the
convergence predicted by the original Solow model does occur if a nation’s stock of
human capital is large compared to what is expected from its corresponding level of
income. In this way, the country experiences a sort of catching-up effect in which
the larger than expected stock o f human capital combined with the relevant
externalities allows the country to take advantage o f physical capital and investment
domestically and from abroad.
A later study makes a point that will be returned to later in this paper. Gemmell
(1996) argues that there is one key problem with empirical models that seek to
measure the impact o f human capital. He demonstrates that the process of
accumulating human capital does not lend itself easily to empirical data for testing.
Consistent with this, he argues that using only school enrollment ratios as a measure
of human capital is inherently inaccurate because o f their tendency to combine the
effects of the stock and accumulation o f human capital. Using data beginning in
1960, he estimates the initial stock o f human capital by using school enrollment
ratios and labor force data. He then creates a new measure o f human capital by
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
12
including both this initial stock and the growth o f the variable over time in a
regression equation. With this new measure, he finds an even stronger relationship
between human capital and economic growth.
A different study by Hail and Jones (1996) uses the average educational
attainment o f workers to emphasize the importance o f the supportive infrastructure
o f the nation. This infrastructure includes governmental, cultural, and natural factors
that improve the accumulation of physical and human capital and has the resulting
externality o f raising Total Factor Productivity. Using a Solow model across
countries, rather than over time, they emphasize differences in output per worker that
result from differences in human capital, physical capital, and total factor
productivity. They find that three-fourths o f the variation in output per worker
results from differences in capital accumulation, skill acquisition, invention, and
technology transfer. These are all factors that are theorized to be related to the
various measures o f human capital.
In studying the relationship between human capital and economic growth, it is
possible to find studies with insignificant results under certain conditions. A study
by Benhabib and Spiegel (1994) models human capital in various ways. They draw
a distinction between treating human capital as an “ordinary input” in the production
function or as an input with significant externalities resulting from its effect on
technological progress and total factor productivity as in the endogenous growth
models. In particular, they discuss problems with the typical measurements of
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
13
human capital as enrollment ratios or literacy rates. Treating human capital as an
“ordinary input” by allowing it and physical capital to enter as factors of production
in a Cobb-Douglas production function, they find an insignificant role for human
capital. However, using a specification that allows human capital to influence the
relative growth rates o f technology and a convergence component based on the
initial stock, they find that human capital has a significant impact on the economic
growth o f a country.
A recent study focusing on nine West African countries by Sacerdoti,
Brunschwig, and Tang (1998) draws a conclusion that will be important to the
findings of this paper. First, they find that school enrollment ratios are poorly
correlated with human capital accumulation. For this reason, they use the average
years of schooling and a wage-weighted measure o f relative labor productivity
resulting from education. However, even using what they feel is an improved
measure o f human capital, they find that there is no significant relationship between
human capital and economic growth.
In staying consistent with themes suggested earlier by Pritchett, these authors
consider the apparent contradiction in schooling that brings about higher cognitive
skills and increased private returns from education but does not lead to economic
growth. They give three possible explanations for this result. In explaining this
finding, they cite the dependence in West Africa on the agricultural sector that may
not benefit significantly from schooling, the lack o f technological progress in the
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
15
the existing stagnant demand has the effect of lowering the rate of return to
education; and 3) Education does indeed raise productivity in the presence of
adequate demand, but it does so in ways that are not productive to the economy as a
whole. Again, this issue o f how a society effectively uses its human capital
resources does not correspond directly to recent attempts to measure the variance of
educational quality. However, this type of understanding might also be quantified to
explain the variation in economic results that are not captured by purely educational
inputs. An additional belief stated in this paper is that the quality measurements
discussed in the next section would be difficult to implement and study on a
macroeconomic level.
Another study that casts doubt on the effects o f human capital argues that the
relationship is not schooling to growth, but growth to schooling. In this study, Bils
and Klenow (1998) replicate Barro’s 1991 model to find less than half of his
relationship between schooling and growth. Similar to other findings, they believe
growth is caused more by the possible externalities o f education. As with the
economists who stress the externalities o f human capital, they suggest there might be
a role for education through the development and diffusion of technology that
positively affects growth and might result from schooling.
A more recent study by Krueger and Lindahl (2000) seeks to bring the
microeconometric and empirical macroeconomic growth models together. They first
conclude that average schooling data across countries are poorly measured because
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
15
the existing stagnant demand has the effect o f lowering the rate of return to
education; and 3) Education does indeed raise productivity in the presence of
adequate demand, but it does so in ways that are not productive to the economy as a
whole. Again, this issue o f how a society effectively uses its human capital
resources does not correspond directly to recent attempts to measure the variance of
educational quality. However, this type of understanding might also be quantified to
explain the variation in economic results that are not captured by purely educational
inputs. An additional belief stated in this paper is that the quality measurements
discussed in the next section would be difficult to implement and study on a
macroeconomic level.
Another study that casts doubt on the effects o f human capital argues that the
relationship is not schooling to growth, but growth to schooling. In this study, Bils
and Klenow (1998) replicate Barro’s 1991 model to find less than half of his
relationship between schooling and growth. Similar to other findings, they believe
growth is caused more by the possible externalities of education. As with the
economists who stress the externalities o f human capital, they suggest there might be
a role for education through the development and diffusion o f technology that
positively affects growth and might result from schooling.
A more recent study by Krueger and Lindahl (2000) seeks to bring the
microeconometric and empirical macroeconomic growth models together. They first
conclude that average schooling data across countries are poorly measured because
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
16
they are often a result o f enrollment flows. They also conclude that country-level
data is no more reliable than individual-level data. With these assumptions, they
explain that there are severe measurements errors that occur in educational data that
have the result o f diluting the estimated effects o f human capital on the growth of
GDP. They give three arguments concerning these assumed measurement errors: 1)
The use o f enrollment rates as a proxy to estimate average years of schooling may
not apply to all countries and time periods; 2) Enrollment rates are often measured
imprecisely; and 3) Enrollment rates do not correspond accurately to the workforce.
Accounting for this measurement error, they find a significant and positive role for
education on economic growth. A final important conclusion o f this study is the
importance o f using aggregate data in macroeconomic studies. In particular, they
believe using aggregate data can explain possible returns to education that are not
emphasized by microeconometric literature.
Many o f the papers cited in this section point to a definite weakness in the field of
macroeconomics as it relates to education. To this point, most of the research has
attempted to demonstrate a direct relationship from some form o f education to
development. However, most of these studies do not go beyond this effort to explain
the underlying mechanism in this relationship. A recent paper on development
within Latin American countries by Campos and Nugent (1998) creates a model that
does go beyond simply allowing for a quantitative relationship between education
and economic growth. Using both cross-sectional and time-series data, this paper
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
17
also attempts to explain the precise mechanism that allows education to provide a
positive stimulus to the development of a nation.
There are a number of important findings from this study that are relevant for this
paper. First, using a model similar to Mankiw, Romer, and Weil, they use the rate of
change in average years of schooling of the population over 25 years of age as a
proxy for human capital growth. They first find that human capital has an
insignificant impact on real per capita GDP. Then, they apply an index of
institutional development. Their competitive index of institutional development
(CUD) provides an accounting for various measures of the executive branch of the
government, the competitiveness and regulation of participation in the economy, the
process o f selection and the effectiveness of the legislative branch of the
government, and limitations on the role of government in the economy. They
believe that countries with high levels o f the CHD would also have individuals who
feel a greater level o f involvement in the system. In this situation, rent-seeking
activities would be reduced with a corresponding increase in the level of
participation in the system and more resources being used in productive economic
activities. When they introduce this index to the model, the index itself and their
measure of human capital growth enter the equation with positive and significant
relationships. Using another analysis, they demonstrate a positive effect from the
level of human capital to the index. These findings point to a possible explanation
for some of the aforementioned studies that do not demonstrate a significant role for
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
18
education in the process o f development. As these authors believe, these
relationships support the idea that human capital is an important link allowing
institutions within the economy to affect growth. Their finding is similar to
Pritchett’s belief that, in some countries, the private return to education exceeds the
social return thus causing agents to participate in activities that lead to insignificant
or even negative growth in the economy. Among other suggestions for
improvements in future studies, these authors support the development of better
measures of human capital.
A final issue for this section deals with the important question o f which data to
use in macroeconomic models. Specifically, the question here that pertains to
macroeconomic growth models involves the type o f data that should be collected to
serve as a proxy for human capital. The next section will introduce the possibility of
using quality measurements. For now, the discussion will continue to consider
quantity measurements. Barro and Lee along with Mulligan and Sala-i-Martin have
contributed significantly to this issue. In an earlier paper, Mulligan and Sala-i-
Martin (1995a) explain the issue clearly when they state, “.. the first stage o f an
empirical study must aggregate the heterogeneous workers.” (p. 1) This issue of
accounting for heterogeneous workers will be discussed in the next section.
Mulligan and Sala-i-Martin (1995b) support a measure of human capital that does
not only identify schooling as human capital. Their aggregate value measure is the
ratio between total labor income per capita and the wage o f the person with zero
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
19
years of schooling. Another method by Barro and Lee (1993b) uses male and female
attainment at four levels o f schooling: 1) No-Schooling; 2) Primary; 3) Secondary;
and 4) Post-Secondary. As part of their continuing attempts to improve on both the
quantity and quality measurements of human capital, Barro and Lee (2000) support
the usage o f attainment levels while accounting for additional missing data and the
different length o f school years across countries.
A number o f relevant issues that apply to this paper can be drawn from the field
o f macroeconomics. First, the support for the existence of a relationship between
schooling and growth is fairly strong. Many studies find a positive relationship. The
ones that do not find a clear, positive relationship provide caveats as to their
findings. In addition, they actually provide more guidance for further work than the
studies finding positive results. Benahabib and Spiegel demonstrate the need for
capturing the entire role of human capital with a complete specification. Bils and
Klenow also discuss the need for completely specifying the role of education.
Sacerdoti, Biunschwig, and Tang support the importance of considering the
particular situation o f each country. Despite the lack o f a positive role for human
capital in his model, Pritchett clearly states that his results should not be used by
governments to cut education. He also discusses the possibility that the negative
finding is a result o f education not leading to the most productive usage o f the
workforce. Important in this discussion is the relationship between the private and
social returns to education. Campos and Nugent first apply a model with an
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
20
insignificant role for human capital. However, by introducing their index which
accounts for variations in institutional development, these authors give a more
complete picture of the role of human capital and find a significant and positive
result.
These issues are not simply answered by introducing a quality measurement for
education. However, they do introduce important issues surrounding the various
roles for human capital in a society. Any study of the relationship between human
capital and economic growth must seek to completely capture the effects o f
education. A final point with relevance for this paper concerns the importance of
data for a quantitative study. Krueger, Mulligan and Sala-i-Martin, and Barro and
Lee emphasize the importance and difficulty o f finding accurate empirical data for
these types o f studies.
In keeping with the original purpose o f this paper, am important theme must be
emphasized. The studies to this point have used quantity measurements o f education
as a proxy for human capital. The purpose o f this study is to reconcile these
experiences with the possibility o f using quality measurements o f education as the
relevant proxy. As stated by Schultz (1999), “In principle, improvements in both the
quantity and quality can be used to improve educational performance.” (p. 83) The
next section will address this question of quality measurements by using the relevant
microeconomic literature.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
21
III. Microeconomic Sehooling-Returns Models
In the absence o f literature relating the effects o f schooling quality to economic
growth, this section turns to the findings of microeconomic models that demonstrate
the relationship between the quality of education and various forms o f returns. One
aspect of microeconomic models is apparent from the available literature. The
diversity o f variables, models, and methods is much larger and more difficult to
simplify than the macroeconomic models. This makes intuitive sense, however, as
the macroeconomic models use only a finite set of quantity variables such as
enrollment ratios, educational attainment, and literacy rates. As will be
demonstrated in the next section, the set o f theorized quality improvements in the
field of education is extremely large. Furthermore, the mixture of “qualities” that
can be used within schools is seemingly infinite. In fact, some economists attempt to
simplify this question to be a relationship between expenditures and returns to
schooling. This attempt does the field a disservice in not accounting for the large
number of quality inputs that these expenditures can fund. Resources spent in one
use might be more or less productive in another application.
The diversity of these models is further complicated by the relevant output
variables to measure. The macroeconomic studies are consistent in using some form
o f national income such as GDP. The microeconomic literature is divided between
numerous measures o f private returns such as earnings, probability o f employment,
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
22
retention rates, and test scores. The purpose o f this section will be to draw relevant
conclusions for the situations of developing countries from this diverse literature.
These conclusions will then be combined with the research from the field of
education discussed in the next section.
A recent paper by Krueger (2000) makes a relevant point in interpreting the
findings of microeconomic studies. Kreuger, referencing a quote from Galileo,
discusses the importance and relevance o f a study on class size conducted in a
randomized fashion with over 11,600 students from kindergarten to third grade. He
states that, “Insofar as sample size and strength of design are concerned, I would
argue that Tennessee’s Project STAR is the single Barbary steed in the class size
literature.” (p. 4) He quotes statistician Frederick Mosteller as observing that Project
STAR “...is one of the most important educational investigations ever carried out
and illustrates the kind and magnitude o f research needed in the field of education to
strengthen schools.” (p. 4) Krueger’s point is extremely relevant. The majority of
the literature in the field either suffers from a small sample size or an experiment that
is not randomized and possesses possible selection bias. These and other issues will
be discussed as they relate to recent studies.
There are a small number o f writers who have debated the importance of
schooling quality. Among these critics o f the relationship, the strongest case is made
by Heckman, Layne-Farrar, and Todd. In three important papers from the mid-
1990s, they argue against some o f the findings o f various authors. Two key
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
23
assumptions are made in most studies that they do not accept. The first key
assumption is the absence o f selective migration on the basis of earnings in models
that study regions within a country. The second key assumption is the assumption of
linearity of the eamings-school relationship. (Heckman, Layne-Farrar, and Todd,
1995a) Other issues they raise are the absence o f supply and demand relationships in
previous studies, the question of whether quality produces the same results in all
markets, and whether or not the aggregation of data is appropriate. (Heckman,
Layne-Farrar, and Todd, 1996) An important finding for this study, however, is that
issues affecting the marginal returns to schooling quality explain why some authors
find much stronger effects in developing countries, (Heckman, Layne-Farrar, and
Todd, 1995b) This is a concept that proves to be a common theme throughout the
discussion on school quality. Improvements in the quality of education tend to have
more of an effect on low-ability, high-needs students.
A final important point comes from a later study. Cameron and Heckman (1998)
find that policies seeking to attract students to school on average attract less able and
less productive students. This is a relevant issue for the situations of developing
countries. The main result o f these studies is to discount any relationship between
quality and earnings with the exception of a marginal difference for attending
college.
A later study conducted to demonstrate the relationship between textbooks and
test scores in Kenyan primary schools also shows less than definitive results. The
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
24
key finding is that providing textbooks only increases the test scores of the students
in the top quintile o f the testing distribution. (Glewwe, Kremer, and Moulin, 1998)
This finding, although the test was structured in a randomized fashion, demonstrates
the difficulty in holding all other variables constant. In addition to collecting data on
textbook provision and test scores, the authors also collected information on the way
the textbooks affected pedagogical practices in the classroom. They find that the
provision o f textbooks decreased the likelihood o f the teacher to use the blackboard
or visual aids and increased the likelihood of the students to be using the textbooks.
These findings are likely to explain why only students in the top quintile of the
range benefited from textbooks. First of all, it is these students who were best suited
to read the textbooks. Surely, the students at the lower part o f the testing range
would not benefit significantly from the provision because they probably were not
able to read the textbooks without significant difficulty anyway. Furthermore, the
two pedagogical differences that resulted from the provision o f the textbooks, more
time reading and less visual aids, would certainly benefit the higher students more
than the lower students. In fact, one could argue that the decreased use o f visual aids
actually hurt the lower achieving students as these are the students most likely to
need additional support aside from the textbook and teacher lectures. Combining
pedagogical information concerning the usage o f textbooks in the classroom with the
valuable findings o f how the textbooks changed teaching practices in the classroom,
it is no surprise to find that the textbooks benefited only the top quintile o f students.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
25
This finding is contrary to the results o f other improvements in quality that find that
the lower achieving students benefit the most. However, the usage o f textbooks in
the classroom and the resulting altered pedagogical practices of the teachers seem to
explain this conclusion.
An early seminal work in this field was the study o f Card and Krueger (1992)
relating pupil-teacher ratios, average term length, and relative teacher pay to rates of
return to education. Finding a problem with previous studies that allowed family
background variables to affect education expenditures and labor market earnings,
they purge the rate o f return of these effects. Using data for white men bom between
1920 and 1949 in the 48 mainland states, they find school quality improvements lead
to higher economic returns. A later study by these same authors applies a theoretical
model to study the effects o f school quality and educational attainment on labor
market earnings. They find that the estimated payoff to a year o f education increases
as the average quality o f schooling increases. The effect is greater for those students
who have attended school the longest. Higher school quality leads to higher earnings
by increasing the payoff per year of schooling and by encouraging students to stay in
school longer. They also consider the issue o f using aggregate data and find no
significant difference in results based on aggregation. (Card and Krueger, 1996)
A number o f studies take account o f the issue regarding the endogeneity of school
choice and inputs as discussed by Heckman. The work o f Case and Deaton (1999) '
uses data from the South African Living Standards Survey conducted in 1993. This
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
26
study controls for the issue o f endogeneity in that the Black South Africans in the
study were extremely limited in their residential choice, and decisions regarding the
schools were primarily made by the White centralized government. Within the
distribution o f schools for Black South Africans, there was considerable variation in
class size. They regress pupil-teacher ratios, whether the primary school has a
library, whether the secondary school has a library, and whether the secondary
school has a laboratory on enrollment rates, educational attainment, and test scores
for numeracy. Their key findings are that the primary school library and the
secondary school laboratory have an insignificant effect. The pupil-teacher ratio and
the secondary school library are both found to have significant coefficients which
support the importance of school quality.
A second study by Case and Yogo (1999) has the same benefit of controlling for
the endogeneity of school choice and inputs. This study uses data from the 1996
South African census and two national surveys o f school quality. They use a two-
stage estimation procedure to quantify the effects of pupil-teacher ratios on
educational attainment, probability o f employment, and returns to education. In
response to Heckman’s previously mentioned concerns, they allow for the retum-to-
education-by-district-of-origin coefficient to vary between provincial labor markets
to account for the issue of selection bias. They also allow log income to be non
linear in years o f completed schooling. They find a significant role for school
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
27
quality on the rate o f return to schooling, educational attainment, and the probability
of employment.
Further studies by Ashenfelter and Rouse attempt to control for possible
endogenous variables by gathering data from samples o f twins. An earlier study
finds the average return to schooling to be nine percent. They also find that this
return is slightly higher for less able workers, a conclusion that is important for
developing countries. (Ashenfelter and Rouse, 1997) A later study finds that an
omitted correlation between ability and schooling does not cause the increased return
to schooling. This is an important finding as many critics of the schooling-retums
debate feel that higher educated people earn more income simply because they
inherently have more ability. They also use instrumental variables similar to the
methods discussed below to account for the endogenous nature o f their variables.
(Ashenfelter and Rouse, 1999)
The previous authors have dealt with the endogeneity of schooling variables by
studying the extreme case of Apartheid in South Africa, studying sets o f twins, or
using exogenous determinants o f schooling to create instrumental variables. Two
other studies find a positive relation between class size and the returns to education
by using a technique employing instrumental variables. A study by Akerhielm
(1995) discusses the non-random assignment o f low-achieving students to classes
with low pupil-teacher ratios. This is usually done through various types of special
education formats. She uses an instrumental variable procedure to predict class size
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
28
to deal with this endogenous relationship. A second paper by Angrist and Lavy
(1999) predicts Israeli class sizes by using a procedure first developed by the twelfth
century rabbinic scholar Maimonides regarding the ideal pupil-teacher ratio. They
use instrumental variables estimates to explain the effects of class sizes on test
scores. Akerhielm finds that reducing class size will benefit students in math,
english, science and history classes with the greatest effects in the latter two subjects.
Angrist and Lavy find test score increases for the students in fourth and fifth grades,
but no effect for those students in third grade. Akerhielm brings up a point relevant
to this paper when she warns that a cost-benefit analysis must be performed to
determine if the added costs are justified by the increased benefits. This is an issue
that will be discussed in the next section.
In attempting to model the variation o f educational quality across countries, Lee
and Barra (1997) gathered data on educational outcomes and inputs and built a panel
data set separating the results into different subjects and age groups. Their research
differs from previous studies in allowing the effects o f family inputs. They find that
family inputs and school resources have a positive impact on student test scores and
a negative relationship with repetition rates and dropout rates. Other factors leading
to improved education returns include the income and education o f parents, smaller
class sizes, higher teacher salaries, and greater school length. The issue concerning
the importance o f family inputs in different cultural settings will be discussed further
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
29
in the next section o f this paper regarding improvements in the schooling quality of
developing nations.
As mentioned earlier, perhaps the most statistically sound experiment on
school quality was conducted in Tennessee beginning with the 1985-1986 school
year. From kindergarten to third grade, 11,600 students were randomly assigned to
one o f three class types. The smaller had between 13 and 17 students. The larger
class had between 22 and 25 students. A third category was added by using the
larger class size with an aide. The students in small classes continued in these small
classes each year. The remaining students were randomly assigned between the
larger classes with and without aides. There was minimal switching between small
and large classes throughout the experiment. Krueger (1999) analyzes the data to see
if school effects, parent selectivity, movement of students in and out o f the program,
reassignment o f students after kindergarten to please parents, and variations in actual
class sizes affected the validity of the experiment. He finds that these flaws did not
affect the main results of the study. He concludes that an overview o f the literature
supports results consistent with his findings. Students benefit about four percentile
points on standardized test scores during the first year. Subsequent years lead to one
percentile point increases. He also finds that the addition o f an aide has no
significant benefit outside o f first grade. An important result in regards to
developing countries is that minority students and students who receive free lunch
receive the greatest benefit from the reductions in class size. A consistent finding in
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
30
the field is that low-achieving students benefit the most from improvements in the
quality of education. Krueger cites other studies by Word et al. and Summers and
Wolfe that produce this same finding regarding high-needs students. He also
discusses studies by Folger and Breda, Nye et al., and Finn and Achilles that support
his findings. (Krueger, 1999)
Krueger (2000) recently completed a meta-analysis o f numerous published
studies regarding the effects of schooling quality similar to a 1997 survey conducted
by Hanushek. Hanushek’s study found an insignificant relationship between school
resources and school outcomes. Krueger discusses many issues regarding
Hanushek’s methodology. He disagrees with the method Hanushek used to weight
the various studies. He finds the papers that broke data sets into multiple studies
received more weight as each data set was counted separately. This meant that less
rigorous studies with much smaller samples received equal or greater weight than
more reliable studies. He reweights the study in a more equitable fashion and finds
that class size is a determinant of student achievement. The final section of
Krueger’s paper is important in that it discusses how a cost-benefit analysis should
be conducted to determine whether or not a reduced class size should be adopted.
He argues that if an input is shown to have positive outcomes, it should only be
implemented if the benefits outweigh the costs using a method similar to that
suggested by Akerhielm. This application o f a cost-benefit analysis will be
discussed in the next section o f this paper.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
31
The recent microeconomic studies demonstrating the relationships between
increases in school quality and the returns to education provide many insights to be
used in the next section. Despite the diversity of studies within the field, the findings
are fairly consistent. Increasing expenditures and providing an improved quality
education will lead to numerous future benefits to students. In particular, the finding
that is most relevant for the next section is the result that low achieving students tend
to benefit the most from improvements in the quality of education. With the
exception of the Kenyan study regarding textbooks, this finding held. Furthermore,
in the Kenyan study, there is a clear explanation why low achieving students did not
benefit. The emphasis on providing a cost-benefit analysis before implementation
will be extremely important in the next section in dealing with developing nations.
With the severely constrained budgets o f developing nations, it is important to
determine which improvements in quality will be most effective. A final theme that
results from this discussion is the question o f how to build an empirical model that
will demonstrate the actual relationship between economic growth and
improvements in the quality o f education.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
32
IV. Educational Research Concerning Quality of Education
With a better understanding of the efforts o f economists to quantify the effects of
education on macroeconomic income growth and microeconomic returns to
education, it is now necessary to integrate the contributions of the field of education.
In particular, it is important to understand the quality relationships as they apply to
developing countries. The relevant findings for this paper from the field of
education will fit into two distinct categories. First, specific factors will be discussed
that have been theorized to lead to an improved quality o f education. However, in
studying individual countries, another important finding results. This is the
importance o f country-specific factors that will affect the delivery of education in a
nation. In particular, cultural, political, and societal factors have been found to
impact the implementation o f educational policies.
In maintaining the clarity o f this discussion, the majority of these contextual
factors will be discussed in the beginning of this section. However, in discussing
these factors, one important relationship is clear. The contextual factors that are
unique to a society are often imbedded within the quality improvements. An
excellent example o f this relationship is the provision o f textbooks within
classrooms. Many theoretical and empirical papers have discussed the provision of
textbooks as a method to improve school quality. As will be pointed out later in this
section, to discuss the provision o f these textbooks without considering contextual
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
33
factors o f the country would be leaving out a large part o f the equation. In this way,
there will be parts o f this discussion that will combine the discussion of country-
specific factors and quality o f education inputs.
An excellent starting point in the discussion about the importance of considering
contextual factors in studying the educational process of developing countries is a
paper by Fuller and Clarke (1994) concerning the effects o f culture in the classroom.
These two writers are clear on how the studies o f educational quality apply. As they
state, “We argue that simply accumulating more evidence from production-function
studies - without specifying local conditions and without linking inputs to teaching
practices - is becoming a less useful exercise.” (p. 135) They seek to combine the
work of two common approaches in educational research. First, they discuss the
work of the “policy-mechanic tradition” that attempts to identify the specific school
inputs that raise student achievement. In this category, they place writers in the field
of education and economics who have conducted the type o f quantitative studies
discussed in the previous sections. The second group o f contributions is from the
“classroom culturalists” who provide an understanding o f cultural, political, and
societal factors that impact the educational process of developing countries. They
cite various studies that reveal important factors to be: 1) The expected role of
teacher authority; 2) Understood rules regarding student participation relative to the
teacher and fellow students; and 3) The logistical structure o f the classroom
including instructional tools, specific tasks demanded of students, and the amount of
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
34
individual or cooperative work. Fuller and Clarke do not feel that either group’s
contributions can exist without the other. In particular, they believe the policy-
mechanic approach will not be useful until it links cultural conditions with inputs and
outcomes, and the work o f the classroom culturalists will not be applicable until they
allow for a clear link between classroom processes and effects.
These authors provide a number o f examples o f how the policy-mechanic and
classroom culturalist approaches can both be considered to facilitate an
understanding o f the educational production function in developing nations. As an
example, they discuss several studies that demonstrate the impact o f family inputs on
education to be greater in developing countries. Another cultural consideration they
discuss is the pedagogical manner in which indigenous forms o f knowledge are
treated in relation to the textbooks and other materials that are used. They cite
studies that find teachers not accounting for the prior knowledge of the students
when it is precisely this connection between new knowledge and previously
understood knowledge that makes learning so valuable. The importance o f this
paper by Fuller and Clarke in relation to this section is its implication that any
empirical relationship between quality o f educational inputs and the resulting
outcomes must also consider contextual issues o f the developing nation.
An additional example in the literature that illustrates the impact o f cultural
factors on educational processes is a paper by Tsuneyoshi (1994) concerning the
usage of small group settings in Japanese schools. She discusses the differences
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
35
between the United States and Japan in the application of informal teaching o f the
“hidden curriculum” which includes the teaching of attitudes and cultural norms. In
Japan, non-academic activities such as lunch and cleaning periods are used to teach
portions o f the official curriculum in small group settings. The national guidelines
for the curriculum include formal, small group instruction in: 1) classroom activities;
2) student government meetings; 3) club activities; and 4) school events. The
emphasis on being part of a community and contributing to the product o f the whole
actually becomes part o f the formal, taught curriculum. This discussion is not meant
to imply that the educational system o f Japan is inherently better because of this
application. It simply points out the importance of understanding the way contextual
factors will invariably affect the schooling processes o f different cultures.
This same belief in the importance o f culture is supported by Kai-ming (1997) in
the book regarding research in developing countries edited by Crossley and
Vulliamy. This author supports the role of qualitative methods of research that are
able to address and reflect the contextual factors o f a society as opposed to the
shortcomings o f a purely quantitative study in this area. Kai-ming states, “Education
is practiced differently in different countries and the same indicator may reflect very
different stories.” (p. 68) He supports qualitative approaches that account for
differences in cultures across different groups of people. As mentioned in the
previous section in regards to Pritchett and Sacerdoti, Brunschwig, and Tang, these
contextual factors are not true measures o f educational quality. Rather, they are
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
36
factors that may impact the effectiveness o f inputs in the educational process. The
final section o f this paper will return to these contextual issues in supporting a role
for their application in quantitative studies.
A recent work by Crossley and Murby (1994) demonstrates how a quantitative
study can consider quality improvements, in this case textbook provision, while also
considering contextual factors. They emphasize the importance o f textbooks ih the
educational process. The four basic approaches to textbook provision that they
discuss clearly have different implications in developing nations - who are usually
receiving the knowledge from developed countries - and in the developed countries
themselves. These approaches include: 1) The question o f which textbooks to adopt;
2) Adaptation o f the textbook to the particular cultural application; 3) The issue of
“home-grown” texts within the national ministries o f education; and 4) Local
publishing o f materials, often in collaboration with outside agencies. In discussing
findings o f the World Bank, they believe that textbooks must facilitate effective
teaching techniques, apply to the particular culture in a relevant manner, and be
constructed well. They later conclude that there are many factors that affect
textbook adoption and effectiveness. According to these authors, an effective
textbook must meet local needs and the unique needs o f a diverse group o f societies.
The importance o f textbooks in relation to cultural factors is also emphasized by
Harber (1997) in the aforementioned book edited by Crossley and Vulliamy. In this
book, Harber discusses the importance o f textbooks in affecting the parameters o f the
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
37
classroom. Harber’s main concern is that developing nations have spent so many
resources on other inputs that they have given little attention to designing a strong
and comprehensive curriculum. Furthermore, they have devoted few resources to
ensuring the alignment of the actual content within the textbooks to the official
curriculum. This alignment o f the curriculum and materials is important in
increasing the efficiency o f resources devoted to education.
These issues related to textbook provision are further supported by the findings of
a World Bank Discussion Paper by Wolff, Schiefelbein, and Valenzuela (1994). In
studying the state o f primary education in Latin America and the Caribbean, they
find textbooks to be deficient in supporting the education of students. According to
the authors, textbooks suffer from: 1) Poor management; 2) Inadequate evaluation
and selection; 3) Poor physical quality as many last only one year; 4) Poor content
quality o f the textbook; 5) Books lacking modem pedagogical practices that are not
linked with curriculum and provide little guidance for teachers; and 6) Unclear
definition o f the roles o f the public and private sectors in the supply o f textbooks.
They suggest that all students have at least one textbook in reading and one in math.
More importantly, they support a dynamic teaching pedagogy that does not rely
solely on the textbook. This is consistent with the point made in the previous section
discussing the weakness o f the Kenyan textbook study and in the pedagogical
changes that occurred as the textbooks were introduced into the classroom. In fact,
the textbook should be used in combination with workbooks, library materials, and
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
38
increased means for teachers to create their own materials that meet the specific
needs of their individual students. These same authors stress the importance in
developing nations o f creating a stronger system to provide pre-schooling for all
children. This is especially important as the early years are consistently found to
contribute to the formation o f intelligence, personality, and social behavior. An
additional policy suggestion that will be returned to later in this section involves
methods to improve the classroom behavior o f teachers.
The previous discussion concerning textbook provision has important
implications regarding empirical studies. It is the belief o f this author that a strong
academic curriculum is not created by the textbook in isolation. Dynamic teaching
that meets the needs o f different groups of students - even in developed countries
such as the United States there are cultural, political, and societal diversities in
relation to the needs of individual students - must transcend the confinements o f the
textbook and use the textbook as one input among many in the means instead o f the
end result o f the educational process. Studies such as the Kenyan example in the
previous section often do not take account for these factors. The belief seems to be
that the simple provision of textbooks is going to improve learning within the
classroom as if the students will automatically learn more by possessing the
textbook. In fact, in the case o f the Kenyan study and its excellent analysis o f how
pedagogical strategies were affected by the provision o f textbooks, it is the belief of
this author that the textbooks might have actually hurt the education of certain
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
39
students as the teachers relied less on other visual aids and more on the textbook.
The dynamic teaching implied by combining the information from the textbook with
other learning methods such as visual aids is precisely the pedagogical strategy that
is going to appeal to the learning needs o f low-achieving students in developing
countries.
Changes within the educational system of California have been extensive in the
past five years. Chief among these changes has been the collection and application
o f data. From standardized test scores to one-on-one assessments in the classroom,
forms of data are being collected that were previously not applied. As a classroom
teacher, the author of this paper can verify the importance and value o f collecting
this type of information. The data has been valuable for the state, school districts,
school officials, and individual classroom teachers in designing instruction that
meets the needs of every student in the classroom. A consistent slogan has become
part o f the common language in educational circles, “Assessment drives instruction.”
This type of data collection and application has great importance in developing
nations as well.
An additional study completed for the World Bank provides support for these
types of data collection within the educational systems o f developing nations. As
supported by other research mentioned later in this section, Heneveld and Craig
(1996) encourage a greater focus on what happens inside the school. They stress the
importance o f collecting, interpreting, and applying data that is also emphasized in
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
40
the book edited by Chapman and Mahlck (1993). Consistent with the importance of
understanding the inputs and outcomes o f schools, these authors support an
improved understanding of what makes education successful by taking account of
two central issues in effectiveness studies. The first is the fact that generalization
across such a broad group o f countries is difficult. This concept is certainly
supported by the earlier discussion regarding the importance o f considering
contextual factors. However, they go even further in stating that the accepted
econometric input-output model o f schooling is inherently flawed because it does not
reflect the dynamic nature o f schools. Heneveld and Craig (1996) believe, “Schools
are organic social systems with significant day-to-day autonomy, and that changes
only evolve according to each school’s individual contexts and conditions.” (p. 3) In
addressing this concern, their final implication is that policies should be adopted in a
sectoral manner that empower local schools and communities to affect the education
o f their own children.
A final issue discussed by these authors is one that is related to data collection,
but actually receives little attention in the literature. Quantitative studies seem to
assume that the result o f implementing a new policy for quality improvement implies
immediate application by teachers. In fact, the actual adoption of a new educational
policy does not guarantee that teachers will actually implement the change in the
classroom. These authors accept the idea that there are various factors that will
affect the success of creating meaningful change in school systems. Data collection
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
41
can be used to study how much of the improvement is actually being implemented in
the classroom. It can also improve the understanding o f the dynamics behind these
factors that must also be understood and accounted for in creating any policy
adjustments.
The previously mentioned study edited by Chapman and Mahlck (1993) addresses
data collection in addition to many of the previously mentioned methods for
improving the quality o f education in developing countries. First, the editors
themselves support the design o f Education Information Systems (EIS) in developing
nations to collect and analyze appropriate data to lead to the implementation of new
programs. Within this context o f improved data collection and analysis, various
authors support themes previously mentioned. Windham (1993) supports
decentralized control in developing countries to make decision making more
efficient, allow for quicker response to educational problems, and make lines of
accountability clear. In this regard, Education Information Systems would support
decentralization in encouraging data based management and allowing for detailed
educational indicators. Kemmerer (1993) supports a system o f teacher incentives
that is clear in goals and standards, collects relevant information, provides
procedures for adjustment, and is adequate and equitable. Bude (1993) supports the
increased empowerment and professionalization of teachers by increasing horizontal
sharing o f information between teachers. Welsh (1993) addresses the importance of
not just assuming that the implementation o f a quality improvement will lead to
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
42
changes in the classroom. He mentions that resistance to change among many factor
inputs are crucial to understand, and he supports the fact that information systems
can be helpful in monitoring implementation.
Kudjoh and Mingat (1993) discuss the importance o f considering contextual
factors as they emphasize that successful programs in one country or region may not
actually be successful in another. More importantly, they also cite an important
study in Togo to demonstrate that large-scale empirical studies can be conducted in
developing countries and the findings from these studies can be applied successfully.
This study measured various factors that are theorized to affect the test scores within
the classrooms of developing nations. In this way, they provide guidance as to how
studies can be conducted either within nations with great diversity or across nations
using a cross-sectional approach. They also support using a cost-benefit type of
analysis for decision-making in continuing the theme that these empirical findings
should not be the only factors consulted in formulating new policies. This is an
example o f how the field o f economics and education can come together to
maximize the education resources and returns o f developing nations.
As quality improvements apply specifically to developing countries, the early
work of Fuller (1986) remains extremely comprehensive. Fuller performed a meta
analysis on 72 empirical studies that preceded his work. The size o f his study and
the application it has to the specific contexts o f developing nations are the strengths
o f his paper. Fuller found that in developing nations, expenditures per pupil,
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
43
instructional materials, school library activity, teacher training at the tertiary level,
length of instructional program, and teacher’s social class were consistently related
to achievement. He also found that class size, laboratories, and individual teacher
salary levels were not related to achievement. The finding o f no effect from
laboratories is consistent with the findings of Case and Deaton in the previous
section. The class size result is contrary to the findings o f the previous section, but
the previous empirical studies were conducted mainly on developed nations. It may
actually be that the needs o f developing nations are such that adjusting the class size
does indeed have minimal effects compared to other factors.
The conclusion that individual teacher salaries have no effect on student learning
is discussed in another paper by Farrell and Oliveria (1993). They find that teachers
should be paid as much as is necessary to attract and maintain people with strong
qualifications for the occupation. They find that teachers in developing countries
earn incomes on average more than auto mechanics but slightly less than the incomes
o f stenographers/typists. This paper has many important suggestions for providing
quality teachers in the schools o f developing countries. The issue o f incentives that
is discussed in many studies, including the book by Hanushek (1994), is supported
by these authors. However, they favor rewards for groups of teachers or entire
schools rather than incentives for individual teachers. This belief can be supported
by the informal observations o f the author o f this paper as the state o f California has
implemented teacher incentives for test score increases in recent years. In informal
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
44
discussions with many teachers, it is the belief o f this author that individual
incentives possess numerous possible negative consequences including division of
staff, lowered morale, and increased incentives to cheat on testing results.
Most valuable from the Farrell and Oliveira study is the call for developing
nations to empower teachers and school officials. When considering the incredible
diversity existing among Sub-Saharan African nations that will be discussed in the
next section, there are clear issues with educational policies that are dominated by
national ministries o f education. These authors call for on-the-job supervision of
teachers by principals and more in-service training o f teachers. The alternative to in-
service training would be pre-service training for teachers to obtain credentials.
These authors support a system in which educated individuals from other
occupations could begin teaching while being trained through in-service techniques.
This is similar to the “emergency credential” program that the state o f California has
used in recent years to deal with the shortage of teachers. The advantage of this
program is that it opens the doors o f the occupation to more people who might not
have the technical certifications but do possess the temperament and talent for
teaching and otherwise would not enter the profession with the requirements of
additional education. This method is also valuable in that teachers get to leam on the
job as they work with students in the classroom. They site an enlightening finding
that teachers leam the majority of their pedagogical skills during the first five years
of teaching practice and not during specialized pre-service training. One downfall to
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
45
a program o f this nature is that it might increase turnover o f teachers as more
individuals move in and out o f the profession. Secondly, there is clearly the risk of
highly unqualified individuals teaching in the classroom. However, both of these
concerns can be addressed in providing more control and autonomy by local officials
as the authors support.
Other programs supported by Farrell and Oliveira include methods for
empowering teachers. These encourage more involvement o f teachers in designing
and delivering training programs. They support methods that will increase the
professionalization of the occupation such as allowing more peer-group regulation
rather than the large bureaucracy that exists. Important to their study are efforts to
strengthen the knowledge and pedagogical skills o f poorly qualified teachers. Again,
these improvements are ones that are being supported by the state o f California to
improve education in this state. As an elementary school teacher, the author of this
paper completely supports efforts to increase the professionalization of the teaching
profession. With this professionalization will come teachers who are more
empowered in their profession and take more pride in their efforts.
Universities deserve special attention as they are often responsible for training the
members o f a society most likely to create and facilitate the transfer o f new
technologies from developed countries. Two studies for the World Bank document
and provide prescriptions for the improved state o f universities in Africa. The study
edited by Ransom, Khoo, Selvaratnam (1993) states that the key issue facing
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
46
institutions o f higher learning in M ica is the condition of rising enrollments with
stagnant or decreased financing. They support the role o f government in facilitating
greater institutional autonomy, flexibility, and innovation within the institutions to
lead to improved quality. This emphasis on decentralized management and
increased local control is a recurrent theme in developing nations. Verspoor (1993)
supports the need to consider cultural factors in stating that country conditions must
shape the nature of policy analysis in institutions of higher learning. Castells (1993)
supports the role o f higher education in developing the science and technology o f a
society. This development o f universities is crucial in allowing the country to
facilitate technology transfer and create endogenous technological innovations.
An earlier study by Saint (1992) also emphasizes the potential to build upon the
strengths o f the traditional cultures to improve the universities. He supports
expanding the access of the universities by creating a differentiated system with
universities, professional, trade, and vocational schools. Of particular importance to
Saint is the autonomy o f the institutions in conducting and utilizing self-studies for
improvement.
A comprehensive text edited by Chapman and Carrier (1990) emphasizes
many of the previously discussed common themes. In particular, Windham (1990)
addresses an important issue regarding the findings from Hanushek later in this
paper. He stresses the differences between educators and economists on the issue of
educational quality. Chief among these concerns is the shortcoming o f economic
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
47
analysis to provide measures to account for nonmonetary costs and benefits. He
reemphasizes the concern that findings in one country cannot necessarily be applied
to the system of another country. He concludes that a responsive cost-effectiveness
analysis is an approach that satisfies both educators and economists. In improving
the quality o f education in developing countries, he supports an approach that
includes both educators and economists in making decisions.
In separate chapters, the editors o f the book themselves discuss the need for
following-up and monitoring to ensure implementation o f quality improvements.
Carrier (1990) stresses the power that is held by the stakeholders in the education
system itself including the bureaucracy, teachers, and unions. Chapman (1990)
identifies organizational, individual, and program factors that influence the
implementation o f programs. Significant in his concerns are the factors that affect
teachers since they are the inputs responsible for delivering the majority of
improvements.
The work of Hanushek is important in keeping the focus of this study on
economic considerations in order to promote the improvements in quality that are
most likely to lead to economic growth in the developing nations. Similar to the
discussion by Windham in the previous paragraph, Hanushek (1994) emphasizes the
use of economic principles in a cost-benefit analysis. This type of analysis allows
educational organizations with limited budgets and seemingly infinite decisions to
narrow the focus to only the most efficient inputs that will lead to educational
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
48
outcomes. The earlier paper by Fuller (1986) also supports this analysis. Fuller
suggests three key issues in this process: 1) A comparison must consider the levels
o f costs and benefits; 2) There must be an accounting o f marginal costs and benefits
to maximize returns; and 3) Making decisions must also based on the social and
private rate of return to education. Hanushek supports this type o f analysis in
addition to performance incentives for teachers and schools. Similar to another
continuing theme, Hanushek emphasizes the need for continuous learning and
adaptation within the current educational system. In discussing the case for low-
performing students and programs in other settings, he also supports the concept of
considering contextual factors in designing educational programs.
A later paper by Hanushek (1995) summarizes the issues confronting developing
nations in regard to education. Facing rising populations and budget constraints, in
many countries, the question has become one o f providing broad access to education
or limiting access in order to improve quality. In his response to Hanushek, Kremer
(1995) warns against making this apparent quality for quantity trade-off. The
findings o f this section point to the hope that improvements will be found within the
field of education that do not involve this difficult decision.
Upon studying the state of education in developing countries and the
corresponding efforts to improve quality, many themes explain why there was some
difficulty in finding consensus in the earlier sections o f this paper. First o f all,
education itself possesses a multi-input production function with numerous outputs.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
49
Furthermore, these inputs and outputs are viewed and valued differently by various
groups o f people. In addition, many of these inputs and outputs are difficult to
quantify. An additional difficulty is the issue of studying the effect o f these inputs in
isolation. Few purely randomized experiments have been conducted studying even
one of the inputs. In addition, the random experiments that have been conducted
have had difficulty accounting for the impact of other variables. As the shortcoming
of the Kenyan study on textbooks shows, changing one input such as textbooks will
have effects on other inputs, in this case pedagogical approaches.
A deficiency that does exist in the field is the lack of findings regarding two
levels of education. The majority of studies that have been conducted focus on the
primary schools in developing countries. A small number discuss the roles and
needs of institutions o f higher learning. Few studies have been conducted regarding
the secondary schools o f these regions. Even more significant, virtually no
documentation discusses the methods used to improve the pre-school inputs of
children. This is important since so much intellectual, social, and cognitive
development takes place in the years before students enter the traditional primary
schools in these countries.
At various points in this section, a number o f policies have been discussed that
have been theorized to lead to improving the educational system in developing
nations. Issues that relate to the actual teachers themselves include providing
incentives, more in-service and less pre-service training to open the profession to
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
50
skilled individuals, higher pay to attract quality candidates, empowering teachers and
increasing their professionalization, and training teachers to use a more dynamic
teaching pedagogy that meets the individual needs of all children. A significant
factor in meeting the needs o f all children is the importance o f collecting and
analyzing data in order to adopt necessary improvements. This data collection is
also important in monitoring and providing teachers assistance in implementing
these new programs. The importance o f aligning the textbooks with the curriculum
delivered in the classroom has also been supported. An additional concern is the
strengthening o f pre-school programs during the critical period of child development
and university programs that are important for technology transfer and diffusion.
Finally, an issue that relates to the next paragraph is the empowerment of local
officials. Education is a process that is often affected by factors that are not
considered by centralized decision makers. Furthermore, the time to implement new
strategies from this level is often significant. In developing nations, the
empowerment o f local educational officials has numerous benefits that outweigh any
significant costs.
In considering educational factors in developing countries, the question of
applying findings across cultural, political, and societal barriers is important. In
cross-country studies, there are important contextual norms that are going to affect
the education production function. Even within nations, there are sure to be
substantial regional variations. This is especially true in Sub-Saharan Africa with the
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
large amount o f ethnic diversity. Do these studies suggest that it is impossible to
study the impacts of improvements o f educational policy in order to formulate more
effective programs? The findings o f this paper do not support this limitation. The
goal that remains is gathering data which accounts for the endogeneity of inputs and
the contextual factors o f the society being studied that can provide valuable insights
concerning the education production function for individual countries. The next
section o f this paper will begin this process as it applies to Sub-Saharan Africa.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
52
V. Current Situation of Sub-Saharan Africa
The previous sections of this paper point to the importance o f understanding the
contextual factors of a country or region in order to specify the education production
function. The purpose o f this section is to understand the aspects o f the current
situation in Sub-Saharan Africa that provide assistance for this question. An earlier
study immediately stresses the difficulty o f studying and impacting a region o f this
size. In discussing the difficulties o f African nations to keep pace with other
developed and even other developing nations, Easterly and Levine (1996) find five
factors, including poor governmental policies, inefficient educational systems,
political instability, inadequate infrastructure, and weak institutions. However, when
studying the inability o f these nations to grow compared to the East Asian countries
that were in a similar situation in the 1960s, the authors focus on the ethnic diversity
throughout this region. They find that ethnic diversity in the region accounts for
30% of the difference between the growth rates o f Sub-Saharan Africa and East
Asia.
Collier and Gunning (1999b) find a similar relationship in a recent study
comparing the African nations to other developing countries. The importance of this
finding for this paper is two-fold. First, the impact o f this issue on economic growth
certainly hampers the ability o f these nations to supply their future generations with a
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
53
quality education. Furthermore, this ethnic diversity makes studies of the region less
practical as pointed out previously by numerous authors. More practical are the
studies within particular countries or regions such as the one discussed concerning
Togo in the previous section. If the larger, cross-sectional studies are conducted,
these authors support the usage o f a dummy variable to reflect the unique situation of
the Sub-Saharan African nations.
Regardless o f the method followed to study the education production function
empirically, an understanding of the current situation is an important starting point.
In a recent study o f 14 o f the least developed nations in the world, most o f which are
in Sub-Saharan Africa, Postlethwaite (1995) found conditions shocking compared to
what students and teachers experience in this country. He found that the conditions
o f most schools were not conducive for effective teaching. Class sizes were 40 to 70
students per classroom on average. Among the lowest performing countries was
Zambia. Over 43% o f the students did not have a chalkboard in the classroom, and
over 87% lacked a wall chart. Even more shocking, 100% o f the students were in a
classroom without a world map, 86.3% had no cupboards, 99.7% had no
bookshelves, 94.1% had no class library, 63.9% were in a classroom without a
teacher’s desk, 96.4% had no textbooks in their native language, 12.2% had no
writing utensil, and 95% had no electricity. The shortcomings of empirical studies to
account for cultural factors was noted as an important issue in the previous section.
Nevertheless, it is hard to imagine a country in which cultural factors were such that
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
54
12.2% o f the students could learn effectively without pencils. These findings point
to the possibility o f cross-sectional studies in relation to the most necessary aspects
o f education such as materials for the students and more specialized studies with
consideration for cultural factors when accounting for factors with less clear effects
such as teacher salaries.
A recent policy paper from the World Bank by Ridker (1994) discusses many of
the factors that are particular to this region in regards to education. First, the policies
o f the 1960s and 1970s were not supportive o f developing human capital in that they
did not create demand for the types o f workers the region could provide. In
particular, the large-capital and import-intensive operations that were followed did
not encourage the growth o f small-scale organizations that would have demanded the
domestically trained workers. The resulting demand for managers and technicians
with advanced skills and significant experience could only be supplied from abroad
and stunted the growth of domestically skilled workers. The operation of domestic
and international labor markets has led to many expatriates from abroad being hired
in place o f Africans with similar skills. This resulted in the latter group seeking
employment in international capacities. These are similar to the issues discussed by
Pritchett and Sacerdoti, Brunschwig, and Tang involving the efficient application of
educational skills in the society.
In addition, he states that the education system following the independence of
most nations in the 1960s and 1970s was created along high cost lines that could not
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
55
be funded with the stagnant growth o f the 1980s. Combining this with the
population explosion created an almost impossible situation in attempting to increase
or even keep quality the same. The policy suggestions from this study support
themes from the previous section. First, policies relating to African education should
consult issues that are specific to this region. Ridker also warns against dropping
adopted policies too quickly before their designed impacts have taken effect. In
considering the significant growth externalities resulting from an educated work
force, he suggests a greater focus on institutions o f higher learning in order to train
workers that can help the African nations take advantage of new technologies.
The importance o f taking advantage o f these new technologies is supported by the
recent World Bank book, Can Africa Claim the 21st Century? This book discusses
the declining role o f Africa in the global economy as evidenced by their declining
ability to stimulate exports. This diminished role is one factor that is threatening to
exclude Africa from the information revolution at the beginning o f this century.
(International Bank for Reconstruction and Development, 2000) In keeping with the
study by Easterly and Levine, this book also sites the importance of improving
governance and resolving conflict among groups in the region. In discussing the
importance o f female education, they declare that women are the hidden growth
resource. In supporting this, they emphasize the relationship between female
education and fertility reduction.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
56
An additional concern relating to educational policies in Africa is mentioned by
Ishikawa (1999) in a recent study regarding assistance provided by Japan. In terms
of literacy rates and enrollment rates at the primary and secondary levels, he shows
the most deficient nations to be those ofW est Africa. However, the most striking
finding is the continuing disparity between the education o f males and females.
Depending on the region, he finds the educational attainments and enrollments of
females to be about half that o f males. In a recent paper, Schultz (1999) supports the
same issue when he explains the benefits o f female education. Higher female
education is correlated with the health gains o f children and reductions in fertility,
both results that would substantially improve conditions in the region.
Of particular importance to the African nations in this century is how to create
economies that will facilitate the connection between education and economic
growth. In regards to human capital in a recent paper, Temple (1999) discusses the
importance o f educating workers whose skills are rewarded in the job market in ways
that also support the growth o f the nation. This is a theme that was discussed by
Pritchett and Sacerdoti, Brunschwig, and Tang as they theorized that the connection
between economic growth and education is not clear because workers are being
educated in ways that are not productive in the economy. This is an important issue
that probably needs to be studied on a case-by-case basis.
The recent developments among African nations discussed in this section add to
the findings of the previous section regarding quality improvements in the education
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
57
o f developing nations. A chief finding in the previous section was the difficulty in
generalizing educational studies across countries with cultural, political, and societal
differences. This is especially true o f the African nations who differ in many ways.
Collier and Gunning (1999b) cite the divergent recent histories of Uganda, Ethiopia,
and Mozambique when compared with the Democratic Republic of the Congo and
Sierra Leone. They believe that efforts to categorize the Sub-Saharan African
nations as a group are becoming less valuable.
In their related paper, Collier and Gunning (1999a) provide insights into how
future studies o f the region can account for the great variation. They divide the
nations into groups with the necessary criteria for economic growth based on: 1)
Stability o f the country; 2) Emphasis on sound macroeconomic policies; and 3)
Strength o f allocative efficiencies. They find seven countries that do not meet the
criteria o f stability as they experienced civil wars during the 1990s. Of the
remaining nations, they find twelve that do not meet their criteria for sound
macroeconomic policies. Eight countries do not satisfy their final category of
resource allocation. These twenty-seven nations that they feel do not meet the
criteria for growth possess 67% of the region’s total population. According to these
authors, only Benin, Burkina Faso, Cape Verde, the Ivory Coast, Ethiopia, Gambia,
Ghana, Guinea Bissau, Mali, Mauritania, Senegal, and Uganda meet these three
important criteria. Nevertheless, these authors see potential in the current situations
o f the nations within the region when they state, “The wave of democratization in
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
58
Africa during the 1990s has, however, weakened the power o f the old elites and so
paved the way for politically sustainable reforms.” (p. 106)
There are many issues for education in Sub-Saharan Africa that are called for by
this paper. First, is the need to determine more closely the relationship between
education and economic growth. Specifically, assuming that quality improvements
do lead to economic growth, which improvements are most valuable to each
particular situation? Which levels of education will lead to the greatest impact on
growth? What government policies can be followed to facilitate the transmission of
improved abilities through education into greater income growth? Other issues
include the improved education of females which has been shown to have dramatic
externalities on the nation. Finally, one finding is fairly clear. As a result of the
education production function, the large-scale educational studies grouping culturally
diverse countries together are certainly ignoring the diversity among nations. The
need is for the type o f studies that Fuller suggested in the previous section that can
bring the policy-mechanics and the classroom-cuituralists together.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
59
VI. Conclusion
The goal of this paper was to find factors that would lead to a more complete
understanding o f the relationship between education and economic growth within the
African nations. First, the relationship between educational inputs and economic
growth was studied. All of these papers used quantity measurements of educational
inputs to serve as a proxy for human capital. The results were somewhat inconsistent
in emphasizing the need for governments to support education to facilitate economic
growth. These studies lead to a number o f conclusions relevant for this paper. The
contextual situation o f the country being studied must be considered in
understanding the relationship. In addition, the question was raised as to whether or
not the educated workers were being used in the most productive inputs that lead to
economic growth. Another issue is capturing the entire role o f education considering
its likely externalities. An additional issue is the need for accumulating large and
accurate data sets.
The competing theories attempting to explain the role o f education on economic
growth provided the primary motivation for the next section. This section was a
response to the issue o f quality measurements versus quantity measurements. The
theory behind this is that quantity o f education does not by itself lead to a more
productive workforce. There is the need to model the variation in educational
quality across nations as well. The lack o f macroeconomic studies concerning
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
60
quality o f education caused this section to focus on microeconomic relationships.
The relationship is supported as various quality improvements are found to increase
returns to education. This finding increases the need to demonstrate the mechanism
between quality improvements and economic growth. It was also demonstrated that
low achieving students benefit more from improvements in the quality of education.
This is a finding that has clear implications for developing nations. Finally, the
question was asked again as to which improvements in the quality o f education
would be most valuable on a microeconomic and macroeconomic level.
The third section o f this paper explored the role of quality improvements in
developing nations. One key finding of this section was that education is a multi
input function with many outcomes. In addition, there is a need to conduct studies
that are randomized or account for the relationship between the inputs and outputs.
A major deficiency in developing nations is the development o f pre-school programs
to ensure that children receive stimulation during the most critical developmental
years before formal schooling begins. Finally, the most important finding of this
section deals with the importance o f contextual factors in conducting these studies.
The final section o f this paper provides guidance for future studies. One question
that needs to be answered is, which improvements in a society are the most
productive using a cost-benefit analysis? A related question is, which level(s) o f
education correspond to the most productive usage o f each improvement? Certainly,
either the government or the market needs to perform the role o f using educated
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
61
inputs in a more productive manner. The externalities regarding female education
also cannot be ignored for the African nations.
From the beginning, the purpose of this paper has been to understand the needs of
Sub-Saharan African nations in relation to quality improvements in education that
will support economic growth. The most important conclusion from this paper,
therefore, is the development o f a theoretical model explaining the relationship
between human capital (but more specifically a measurement of education as a
proxy) and economic growth. Many of the writers cited are calling for better
measures o f education to represent human capital. As this paper demonstrates, there
is great interest in applying not just quantity measurements o f human capital, but
also measurements that capture differences in quality as well. Therefore, the
proposed theoretical model is two-fold. First, a macroeconomic model will be
constructed that demonstrates the relationship between variations in the quality of
education and economic development. Then, a microeconomic model will be used
to study the various ways to improve the quality o f education.
The macroeconomic model will apply various lessons learned in this study. In
order to develop the model, the issue o f the large level o f ethnic diversity in Sub-
Saharan Africa must be a point o f departure. As Easterly and Levine (1996) pointed
out, any attempt to study this region without accounting for the negative impact of
the high degree of ethnic diversity would be incomplete. The theoretical model thus
possesses two ways to account for this. First, similar to the efforts o f Lee and Barra
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
62
(1997) to account for heterogeneity in their model by using the East Asia dummy
variable, this author proposes a similar variable in cross-sectional studies. The
application of this variable would have importance whether the study was applied
completely within the region or by using data from other regions.
A second method to deal with this ethnic diversity deals with the issue of
contextual factors and the degree of centralization within the educational system. As
was pointed out by this paper, any quantitative model involving education must
account for the significant impact o f cultural, political, and societal factors on the
education production function o f a region or nation. In considering this factor, many
writers within the field o f education support the application of a more decentralized
structure in delivering education. The belief is that local officials, especially with the
great diversity in Sub-Saharan Africa, will be more knowledgeable o f and more
responsive to the needs o f the students. This will allow them to apply more effective
methods o f education. By placing a greater degree o f autonomy in the hands of these
officials, the resources devoted to education would be more productive. Therefore,
this author proposes the development of an index of decentralization that would
account for the variation in these policies. However, an important point to consider
in the development o f this index is that there might be a certain point at which too
much decentralization might have a negative impact. Some aspects o f education
might be better left in the hands o f national officials who have access to different
types of information.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
63
The model would also account for variations in the quantity and quality of
education. Pritchett provided a brief argument against the relevance o f both quantity
and quality measurements in discussing studies that found an insignificant role for
the former measurement on economic growth. He based this belief on a positive
covariance between the quantity and quality of education. The argument is that
countries spending more money to increase the quantity o f education would also be
improving the quality. This is a relationship that does not seem to apply to the
situations o f many developing countries. In considering the severely limited budgets
of these nations, one could easily argue that attempts to increase the quantity of
education have diluted the resources to the point of dropping the quality of education
in the system. Temple (1999), Hanushek (1995), and Kremer (1995) support this
idea when they discuss the apparent trade-off between increased quantity and
increased quality.
This macroeconomic model possesses four important components. The first is a
measure o f the quantity o f education such as the average years of schooling. Barro
and Lee provide great guidance in selecting the data o f this type. The second
component captures the variance in quality across countries or over time. This
measure o f educational quality will be a data set of test scores on an international
exam. The dummy variable for Sub-Saharan Africa will capture the unique
challenges o f this region in terms o f ethnic diversity. Finally, the index of
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
64
decentralization will measure the ability of a country to deal with this ethnic
diversity by allowing local officials to make more decisions.
The suggested first stage o f this study involves the macroeconomic model that
uses test scores as an input variable. In order to bring the relevance of this model
into the classroom, this author supports the development o f a microeconomic model
that uses test scores as the output variable. This study would be similar to many of
the studies cited in the section on educational quality in the field o f microeconomics.
The difference would be the attempt to model contextual factors in the study. The
independent variables to affect test scores would include average term length,
teacher-pupil ratios, teacher training, teacher salaries, and the provision o f supplies
including textbooks and computers.
Two important lessons from this paper will provide guidance for the application
o f this study. First, the discussion concerning contextual factors guides this author to
search for variations across regions and nations that would impact efforts to improve
education. For example, the discussion concerning the role and application of
textbooks in various societies is an important factor to consider. The Kenyan
textbook study o f Glewwe, Kremer, and Moulin (1998) provides an important lesson
about changes in one input factor affecting another. As was demonstrated, the
provision of the textbooks led to pedagogical changes in the classroom that
theoretically would damage the education o f the lowest performers in the classroom.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
65
Therefore, the microeconomic model studying the relationship between test scores
and various input factors must also consider these two important issues.
This type of proposed study satisfies the objectives o f this paper. First, the
macroeconomic model provides a clear picture o f the positive role between
education and economic growth. This type of result will provide guidance for
government officials to invest in the future o f their countries through education.
This policy suggestion leads to greater resources being devoted to education. At the
same time, the microeconomic study provides guidance for educational policy
makers in applying these increased resources. With this type o f a study, the needs of
the society and the needs of the individual can be brought together in the
development o f the nation.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
6 6
Bibliography
Akerhielm, Karen. “Does Class Size Matter?” Economics o f Education Review.
14, no. 3 (Summer 1995): 229-241.
Angrist, Joshua D. and Victor Lavy. “Using Maimonides’ Rule to Estimate the
Effect of Class Size on Scholastic Achievement.” The Quarterly Journal of
Economics. 114, no. 2 (May 1999): 533-575.
Ashenfelter, Orley and Cecilia Rouse. “Income, Schooling, and Ability: Evidence
from a New Sample of Identical Twins.” National Bureau of Economic
Research Working Paper No. 6106, July 1997.
Ashenfelter, Orley and Cecilia Rouse. “Schooling, Intelligence, and Income in
America: Cracks in the Bell Curve.” NBER Working Paper 6902, January
1999.
Barro, Robert J. “Economic Growth in a Cross-Section o f Countries.” The
Quarterly Journal o f Economics. 106, no. 2 (May 1991): 407-443.
Barro, Robert J. and Jong-Wha Lee. “Losers and Winners in Economic Growth.”
Proceedings o f the World Bank Annual Conference on Development
Economics, (1993a): 267-297.
Barro, Robert J. and Jong-Wha Lee. “International Comparisons of Educational
Attainment.” Journal o f Monetary Economics. 32 (August 1993b): 363-394.
Barro, Robert J. and Jong-Wha Lee. “International Data on Educational Attainment
Updates and Implications.” National Bureau o f Economic Research
Working Paper No. 7911, September 2000.
Benahabib, Jess and Mark M. Spiegel. “The Role o f Human Capital in Economic
Development: Evidence from Aggregate Cross-Country Data.” Journal of
Monetary Economics. 34 (May 1994): 143-173.
Bils, Mark and Peter J. Klenow. “Does Schooling Cause Growth or the Other Way
Around?” National Bureau of Economic Research Working Paper No. 6393,
February 1998.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
67
Bude, Udo. “Strategies for Using Information to Improve Learning Conditions and
Instructional Practices at the School Level,” in From Data to Action:
Information Systems in Educational Planning, edited by David W. Chapman
and Lars O. Mahlck. (Oxford, England: Pergamon Press), 1993.
Buiter, Willem H. and Kenneth M. Kletzer. “Persistent Differences in National
Productivity Growth Rates with a Common Technology and Free Capital
Mobility.” National Bureau o f Economic Research Working Paper
No. 3637, February 1991.
Buiter, Willem H. and Kenneth M. Kletzer. “ Permanent International Productivity
Growth Differentials in an Integrated Global Economy.” National Bureau
o f Economic Research Working Paper No. 4220, November 1992.
Cameron, Stephen V. and James J. Heckman. “Life Cycle Schooling and Dynamic
Selection Bias: Models and Evidence for Five Cohorts.” National Bureau of
Economic Research Working Paper No. 6385, January 1998.
Campos, Nauro and Jeffrey B Nugent. “Institutions and Growth: Can Human
Capital be a Link?” CEPAL Review. 64 (April 1998): 7-27.
Card, David and Alan B. Krueger. “Does School Quality Matter? Returns to
Education and the Characteristics ofPublic Schools in the United States.”
The Journal o f Political Economy. 100, no. 1 (February 1992): 1-40.
Card, David and Alan B. Krueger. “Labor Market Effects o f School Quality: Theory
and Evidence.” National Bureau of Economic Research Working Paper No.
5450, February 1996.
Carrier, Carol A. “Evaluating Instructional Improvement Programs,” in Improving
Educational Quality: A Global Perspective, edited by David W. Chapman
and Carol A. Carrier. (New York: Greenwood Press), 1990.
Case, Anne and Motohiro Yogo. “Does School Quality Matter? Returns to
Education and Characteristics o f Schools in South Africa.” National Bureau
of Economic Research Working Paper 7399, October 1999.
Case, Anne and Angus Deaton. “School Inputs and Educational Outcomes in South
Africa.” Quarterly Journal o f Economics. 114, no. 3 (August 1999): 1047-
1084.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
68
Castells, Manuel. “The University System: Engine of Development in the New
World Economy,” in Improving Higher Education in Developing Countries.
edited by Angela Ransom, Siew-Mun Khoo, and Viswanathan Selvaratnam.
(Washington D C.: The World Bank), 1993.
Chapman, David W. “Monitoring Implementation,” in Improving Educational
Quality: A Global Perspective, edited by David W. Chapman and Carol A.
Carrier. (New York: Greenwood Press), 1990.
Chapman, David. W., and Carol A. Carrier, Editors. Improving Educational Quality:
A Global Perspective. (New York: Greenwood Press), 1990.
Chapman, David W. and Lars O. Mahlck, Editors. From Data to Action:
Information Systems in Educational Planning. (Oxford, England: Pergamon
Press), 1993.
Collier, Paul and Jan Willem Gunning. “Explaining African Economic
Performance.” Journal o f Economic Literature. 37 (March 1999a): 64-111.
Collier, Paul and Jan Willem Gunning. “Why Has Africa Grown Slowly?” Journal
o f Economic Perspectives. 13, no. 3 (Summer 1999b): 3-22.
Crossley, Michael and Myra Murby. “Textbook Provision and the Quality o f the
School Curriculum in Developing Countries: Issues and Policy Options.”
Comparative Education. 30, no. 2 (1994): 99-114.
Easterly, William and Ross Levine. “Africa’s Growth Tragedy: Policies and
Ethnic Divisions.” The World Bank, April 1996.
Farrell, Joseph P. and Joao B. Oliveira. Teachers in Developing Countries:
Improving Effectiveness and Managing Costs. (Washington, D.C.: The
World Bank), 1993.
Fuller, Bruce. “Raising School Quality in Developing Countries: What Investments
Boost Learning?” World Bank Discussion Paper No. 2, 1986.
Fuller, Bruce and Prema Clarke. “Raising School Effects While Ignoring Culture?
Local Conditions and the Influence of Classroom Tools, Rules, and
Pedagogy.” Review o f Educational Research. 64, no. 1 (Spring 1994):
119-157.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
69
Gemmell, Norman. “Evaluating the Impacts of Human Capital Stocks and
Accumulation on Economic Growth: Some New Evidence.” Oxford Bulletin
o f Economic Statistics. 58, no. 1 (1996): 9-28.
Glewwe, Paul, Michael Kremer and Sylvie Moulin. “Textbooks and Test Scores:
Evidence from Prospective Evaluation in Kenya.” Yale University mimeo,
1998.
Hall, Robert E. and Charles I. Jones. “The Productivity ofNations.” National
Bureau of Economic Research Working Paper No. 5812, November 1996.
Hanushek, Eric. Making Schools Count. (Washington, D C.: The Brookings
Institution), 1994.
Hanushek, Eric. “Interpreting Recent Research on Schooling in Developing
Countries.” The World Bank Research Observer. 10, no. 2 (August 1995):
227-246.
Harber, Clive. “Using Documents for Qualitative Educational Research in Africa,”
in Qualitative Educational Research in Developing Countries, edited by
Michael Crossley and Graham Vulliamy. (New York: Garland Publishing,
Inc.), 1997.
Heckman, James, Anne Layne-Farrar, and Petra Todd. “The Schooling Quality-
Eamings Relationship: Using Economic Theory to Interpret Functional
Forms Consistent with the Evidence.” National Bureau o f Economic
Research Working Paper No. 5288, October 1995a.
Heckman, James, Anne Layne-Farrar, and Petra Todd. “Does Measured School
Quality Really Matter? An Examination o f the Eamings-Quality
Relationship.” National Bureau o f Economic Research Working Paper
No. 5274, September, 1995b.
Heckman, James, Anne Layne-Farrar, and Petra Todd. “Human Capital Pricing
Equations with an Application to Estimating the Effect of Schooling Quality
on Earnings.” The Review o f Economics and Statistics. 78, no. 3 (August
1996): 562-610.
Heneveld, Ward and Helen Craig. “Schools Count: World Bank Project Designs and
the Quality o f Primary Education in Sub-Saharan Africa.” The World Bank,
Africa Technical Development Series No. 303, 1996.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
70
International Bank for Reconstruction and Development. Can Africa. Claim the 21s t
Century? (Washington D C.: The World Bank), 2000.
Ishikawa, Kaoru. Nation Building and Development Assistance in Africa. (New
York: St.Martin’s Press, Inc.), 1999.
Kai-ming, Cheng. “Qualitative Research and Educational Policy-Making:
Approaching the Reality in Developing Countries,” in Qualitative
Educational Research in Developing Countries, edited by Michael Crossley
and Graham Vulliamy. (New York: Garland Publishing, Inc.), 1997.
Kemmerer, Frances. “ Monitoring Teacher Incentive Systems: A New Use for the
EMIS?” in From Data to Action: Information Systems in Educational
Planning, edited by David W Chapman and Lars O. Mahlck. (Oxford,
England: Pergamon Press), 1993.
Kremer, Michael R. “Research on Schooling: What We Know and What We Don’t
- A Comment on Hanushek.” The World Bank Research Observer. 10,
no.2 (August 1995): 247-254.
Krueger, Alan B. “Experimental Estimates o f Education Production Functions.”
The Quarterly Journal o f Economics. 114, no. 2 (May 1999): 497-532.
Krueger, Alan B. “Economic Considerations and Class Size.” Princeton University
Industrial Relations Section Working Paper No. 447, September 2000.
Krueger, Alan B. and Mikael Lindahl. “Education for Growth: Why and For
Whom?” Princeton University Economics Department mimeo, 2000.
Kudjoh, Ayele and Alain Mingat. “Toward a Better Understanding o f the
Functioning o f School Systems for Better Decision-Making: The Case o f
Primary Schools in Togo,” in From Data to Action: Information Systems in
Educational Planning, edited by David W. Chapman and Lars O. Mahlck.
(Oxford, England: Pergamon Press), 1993.
Lee, Jong-Wha, and Robert J. Barro. “Schooling Quality in a Cross Section of
Countries.” National Bureau o f Economic Research Working Paper
No. 6198, September 1997.
Lucas, J r. Robert E. “Why Doesn’t Capital Flow from Rich to Poor Countries?”
The American Economic Review. 80, no. 2 (May 1990): 92-96.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
71
Mankiw, N. Gregory, David Romer, and David N. Weil. “A Contribution to the
Empirics of Economic Growth.” The Quarterly Journal o f Economics.
107, no. 2 (May 1992): 407-437.
Mulligan, Casey B. and Xavier Sala-I-Martin. “Measuring Aggregate Human
Capital.” National Bureau o f Economic Research Working Paper No. 5016,
February 1995a.
Mulligan, Casey B. and Xavier Sala-I-Martin. “A Labor-Income-Based Measure of
the Value o f Human Capital: An Application to the State of the United
States.” National Bureau of Economic Research Working Paper No. 5018,
February 1995b.
Nelson, Richard R. and Edmund S. Phelps. “Investment in Humans, Technologiccal
Diffusion, and Economic Growth.” The American Economic Review. 56,
no. 1-2 (March 1966): 69-75.
Postlethwaite, Neville. “The Conditions of Primary Schools in Least-Developed
Countries.” International Review o f Education. 42, no. 3-4 (1995): 289-317.
Pritchett, Lant. “Where Has All the Education Gone?” World Bank Policy Research
No. 1581, 1997.
Ransom, Angela, Siew-Mun Khoo, and Viswanathan Selvaratnam. Improving
Higher Education in Developing Countries. (Washington D.C.: The World
Bank), 1993.
Ridker, Ronald G. The World Bank’s Role in Human Resource Development in
Sub-Saharan Africa. (Washington, D . C.: The World Bank), 1994.
Sacerdoti, Emilio, Sonia Brunschwig and Jon Tang. “The Impact of Human Capital
on Growth: Evidence from West Africa.” International Monetary Fund
Working Paper No. 162, November 1998.
Saint, William S. “Universities in Africa: Strategies for Stabilization and
Revitalization.” World Bank Technical Paper No. 194, 1992.
Schultz, T. Paul. “Health and Schooling Investments in Africa.” Journal of
Economic Perspectives. 13, no. 3 (Summer 1999): 67-88.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
72
Temple, Jonathan. “The New Growth Evidence.” Journal of Economic Literature.
37 (March 1999): 112-156.
Tsuneyoshi, Ryoko. “Small Groups in Japanese Elementary School Classrooms:
Comparisons with the United States,” Comparative Education. 30, no.2
(1994): 115-129.
Verspoor, Adriaan M. “Improvement and Innovation in Higher Education,” in
Improving Higher Education in Developing Countries, edited by Angela
Ransom, Siew-Mun Khoo, and Viswanathan Selvaratnam. (Washington D C.
The World Bank), 1993.
Welsh, Thomas. “The Politics o f Valuing in Information System Construction,” in
From Data to Action: Information Systems in Educational Planning, edited
by David W. Chapman and Lars O. Mahlck. (Oxford, England: Pergamon
Press), 1993.
Windham, Douglas M. “Economics, Instructional Development, and the
Enhancement o f Educational Efficiency,” in Improving Educational Quality:
A Global Perspective, edited by David W. Chapman and Carol A. Carrier.
(New York: Greenwood Press), 1990.
Windham, Douglas M. “Strategies for Decentralizing Data Use,” in From Data to
Action: Information Systems in Educational Planning, edited by David W.
Chapman and Lars O. Mahlck. (Oxford, England: Pergamon Press), 1993.
Wolff, Laurence, Ernesto Sehiefelbein and Jorge Valenzuela. “Improving the
Quality of Education in Latin America and the Caribbean: Toward the 21s t
Century.” World Bank Discussion Paper No. 257, 1994.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Economic analysis of ground lease-based land use system

PDF

Growth, human capital, and technological change in a centrally planned economy: Evidence from Cuba

PDF

An enquiry into the effectiveness of export assistance organizations serving southern California SMEs

PDF

A model of student performance in principles of macroeconomics

PDF

Calculating economic damages in litigation matters

PDF

Development and secretions of salivary glands using mouse models

PDF

Debt reduction by way of inflation: The case of Lebanon

PDF

Education in Athens Co., Ohio: A clash of cultures

PDF

Cost analysis of three pharmacy counseling programs for diabetics in a health maintenance organization

PDF

Economic valuation of impacts to beneficial uses of water quality in California: Proposed methodology

PDF

Belize pegged exchange rate: Heading towards devaluation or maintaining credibility

PDF

Business group affiliation in India

PDF

Does intense probation monitoring of truants work? An empirical econometric analysis of its effect on school and individual attendance, grades, and behavior

PDF

Head injury and dementia: A co-twin control study of Swedish twins

PDF

Decentralizing of public finance: Centralizing forces in developing countries

PDF

Imputation methods for missing items in the Vitality scale of the MOS SF-36 Quality of Life (QOL) Questionnaire

PDF

How financial statements in Mexican firms reflect changes in the financial/economic environment from 1978 to 1996

PDF

Income inequality and economic growth: A theoretical and empirical analysis

PDF

Detection of anti-Hu antibodies, a possible key to early diagnosis of small cell lung cancer

PDF

Explanations, the availability heuristic, and biases of subjective likelihood in economics

Asset Metadata

Creator Ishii, Jeffrey Michael (author)

Core Title A study concerning the relevance of using quality measures of education in economic growth models of sub-Saharan African nations

School Graduate School

Degree Master of Arts

Degree Program Economics

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag Economics, General,education, sociology of,OAI-PMH Harvest

Language English

Contributor Digitized by ProQuest (provenance)

Advisor [illegible] (committee chair), [illegible] (committee member)

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c16-37880

Unique identifier UC11341426

Identifier 1406452.pdf (filename),usctheses-c16-37880 (legacy record id)

Legacy Identifier 1406452.pdf

Dmrecord 37880

Document Type Thesis

Rights Ishii, Jeffrey Michael

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the au...

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus, Los Angeles, California 90089, USA