International Journal of Social Economics

DISEASE AND ECONOMIC DEVELOPMENT: The Impact of Parasitic Diseases in St. Lucia
Burton A. Weisbrod Ralph L. Andreano Robert E. Baldwin Erwin H. Epstein Allen C. Kelley Thomas W.
Helminiak
Article information:
To cite this document:
Burton A. Weisbrod Ralph L. Andreano Robert E. Baldwin Erwin H. Epstein Allen C. Kelley Thomas
W. Helminiak, (1974),"DISEASE AND ECONOMIC DEVELOPMENT", International Journal of Social
Economics, Vol. 1 Iss 1 pp. 111 - 117
Permanent link to this document:
http://dx.doi.org/10.1108/eb013761
Downloaded on: 28 June 2016, At: 10:40 (PT)
Downloaded by Purdue University Libraries At 10:40 28 June 2016 (PT)

References: this document contains references to 0 other documents.

To copy this document: permissions@emeraldinsight.com
The fulltext of this document has been downloaded 50 times since 2006*

Access to this document was granted through an Emerald subscription provided by emerald-
srm:281668 []
For Authors
If you would like to write for this, or any other Emerald publication, then please use our Emerald
for Authors service information about how to choose which publication to write for and submission
guidelines are available for all. Please visit www.emeraldinsight.com/authors for more information.
About Emerald www.emeraldinsight.com
Emerald is a global publisher linking research and practice to the benefit of society. The company
manages a portfolio of more than 290 journals and over 2,350 books and book series volumes, as well
as providing an extensive range of online products and additional customer resources and services.
Emerald is both COUNTER 4 and TRANSFER compliant. The organization is a partner of the
Committee on Publication Ethics (COPE) and also works with Portico and the LOCKSS initiative for
digital archive preservation.

*Related content and download information correct at time of download.

 DISEASE AND ECONOMIC
DEVELOPMENT
The Impact of Parasitic Diseases
in St. Lucia*
by
Burton A. Weisbrod, Ralph L. Andreano,
Robert E. Baldwin, Erwin H. Epstein, Allen C. Kelley,
assisted by Thomas W. Helminiak†

This study sought to measure the impacts of five parasitic diseases on

(1) mortality and natality, (2) school attendance and academic
performance of children, (3) labor productivity on a rural plantation, and
in an urban light-manufacturing plant —all in St. Lucia, West Indies.
Downloaded by Purdue University Libraries At 10:40 28 June 2016 (PT)

The study focused on schistosomiasis, a water-borne disease carried

by snails. The disease is prevalent throughout much of the tropical
world, (1) and is being expanded by irrigation projects. These projects
extend the potential habitats of the snail that transmits schistosomiasis,
a disease that has been called "the greatest unconquered parasitic
disease now afflicting man and animals." (1)
The present study has as its goal the development of a methodology
for quantitative measurement of disease effect and the empirical
application of this methodology to the five parasites prevalent in
St. Lucia.

St. Lucia
St. Lucia is a tropical isle in the Windward Island group of the Lesser
Antilles. The population is about 110,000, of which blacks of African
origin constitute 70 percent and whites of European ancestry less than
than one-half of one percent. The remainder consists of a small number
of East Indians and a much greater number of racially mixed less dark
people.
During the 1950's the island's small, inefficient sugar industry was
supplanted by the rapid expansion of banana cultivation for export to the
U.K. market. During the period of our data-gathering activities in St.
Lucia (1967-1969) the banana "boom" reached its peak, contributing
over 80 percent of total export receipts. (2) A World Bank study has
indicated that per capita income in St. Lucia was $155 in 1964 and $380
in 1971.
Malaria was eradicated in St. Lucia during the mid-1950's. St. Lucia
is now also relatively free of yellow fever, yaws, cholera, and typhoid'
all of which had been important diseases previously.
* Copyright © 1973 by the Board of Regents of the University of Wisconsin System. Disease
and Economic Development: The Impact of Parasitic Diseases in St. Lucia; Edited by Weisbrod,
Andreano, Baldwin, Epstein and Kelly (to be published by the University of Wisconsin Press in
June 1973).
† Department of Economics, University of Wisconsin.

Vol. 1, No.1 111

 Disease and Economic Development

The Data
We mounted four separate field surveys during our enquiry. The first was
a questionnaire administered to households in two St. Lucian agricul-
tural valleys. Some 3,000 persons were surveyed in one valley and 4,800
were surveyed in the other. These people constituted about 74 percent
and 93 percent of the total population in each valley. The total of 7,800
rural persons covered in the two surveys amounted to more than 7
percent of the total population of St. Lucia and about 13 percent of the
rural population.
A Rockefeller Foundation-supported biological research station on
St. Lucia, the Research and Control Department, concurrently requested
stool specimens for parasitological analysis from household members in
both valleys. Specimens, which were tested for the presence of each of
five parasites (Schistosoma, hookworm, Ascaris, Trichuris, and Stron-
gyloides), were obtained from 4,668 persons, a response rate of 60
percent. (3)
The second survey was an island-wide study involving schoolchildren
Downloaded by Purdue University Libraries At 10:40 28 June 2016 (PT)

aged 13-14. (The prevalence of schistosomiasis seems to "peak" around

this age.) A total of 45 schools were involved, from which a sample of
162 pupils were randomly selected. In addition, an intensive study was
made for one school in a high schistosomiasis prevalence area where
267 children, aged 9-14, were studied.
The third survey was an occupational survey of workers employed by
one of the largest banana-growing estates on St. Lucia. Data were
assembled for a sample of 466 men and women who performed a variety
of manual labor tasks on the plantation. Information was obtained on
daily output and earnings for eight different piece-work jobs, and daily
attendance and earnings information were also obtained for laborers
performing non-piece work. The data covered two 18-month periods
between January 1, 1966 and December, 1968. A clinical examination
was conducted to determine the prevalence and the intensity of
schistosomiasis as measured by schistosome egg counts, and by liver
and spleen enlargement. A supplementary questionnaire was admini-
stered to these 466 people to get more detailed information on their
work patterns.
Finally, another occupational survey was taken, this one of a group of
114 women who worked in an urban light-manufacturing plant. This
work involved folding and assembling paper advertising material, work
that required manual dexterity but not physical endurance. Data on
absenteeism, physical productivity, and weekly salary were obtained
from plant records; and, as in the rural banana plantation study, a
clinical examination and a questionnaire were administered to these
workers.

Summary Tests and Results

Natality. We considered first the effects of schistosomiasis and the four
other parasitic diseases on the natality or birth rate. Two models, an
interaction model and an additive model, postulating specific relation-
ships between the diseases and the birth rate were separately explored.

112 IJSE
 Disease and Economic Development

The interaction model is one which considers the impact on the birth
rate of parasitic diseases working in combination as well as singly. The
second formulation, the additive model, assumes that the impacts of the
combinations of diseases is simply the summation of the impacts of the
individual diseases. In both formulations we controlled not only for the
presence of each of the five parasites, but also for the woman's age and
education.
The statistical results using the interaction model yielded the expected
negative relationships between the birth rate and both age and education.
However, the hypothesis concerning a negative effect of parasitic
diseases on the birth rate was not supported. In the two valleys, only four
of the disease interaction terms were significant, and in each of these
cases the sign was counter to what was expected if the diseases
depressed fertility.
The additive model - where the impact of each disease on the birth
rate is assumed to be the summation of the individual impacts of each
disease - was next tested. Again we found no evidence of a significant
Downloaded by Purdue University Libraries At 10:40 28 June 2016 (PT)

negative impact on the birth rate for schistosomiasis, Trichuris, or hook-

worm. Strongyloides, however, was shown to have a significant and
negative association with fertility in both valleys. It was estimated that
the elimination of Strongyloides would increase the birth rate among
those mothers infected with it by 240 percent; however, taking into
account the relatively low prevalence of this disease, the impact of its
elimination would result in a smaller, though still substantial, increase in
the aggregate birth rate - about 16 percent.
Mortality. We tested the basic hypothesis that schistosomiasis and
other parasitic infections decrease an individual's life expectancy - or, in
other words, increase the death rate.
Because the mortality data from the household survey were of poor
quality, we were restricted to explore the hypothesis at the group, rather
than the individual level, utilizing death rate statistics for sub-regions of
the two valleys. The findings are again surprising. None of the inde-
pendent variables (age, sex, disease prevalence) exerted a significant
impact on death rates, regardless of which regression form was used. A
seriously constraining factor in the analysis, however, was the limited
number of sub-group observations and, hence, the limited number of
degrees of freedom.
Academic Performance. During an island-wide survey of school-
children, a standardized reading test was given to all children in the study.
Using the data from this survey, we tested the hypothesis that pupils'
achievement scores would be negatively affected by disease infection,
by the use of hazardous sources of water (to drink, bathe in, and launder
in) and by lack of knowledge of germ theory. We also used data from the
Babonneau school, located in a high prevalence area (described above),
to test the hypothesis that schistosomiasis, alone or in combination with
other parasitic diseases, exerts a negative effect on the child's vitality and
physical condition. Thus, we expected to find from the Babonneau data
that the presence of parasitic infection would be associated with low

Vol.1 No.1 113

 Disease and Economic Development

weight of child, low height, low rank in class, and high absenteeism. In
both studies we controlled in our multiple regressions for several
variables, such as urban vs. rural residence, that might affect childrens'
school performance independently of disease conditions.
In fact we found that the effects of the parasitic infections were not
significant, either in explaining variations in pupil's academic per-
formance, for the island as a whole, or in explaining variations in
schoolchildren's class rank in the Babonneau area. Schistosomiasis was
significantly related (at the .05 level) to absenteeism among pupils
attending the Babonneau school, but surprisingly in the direction
opposite to that expected - children with that disease were absent less
than other children. We might speculate that this finding is the result of
infected children being too sick or weak to work efficiently in the fields,
but healthy enough to attend - and perform equally well - in school. It
is true that children in St. Lucia are often kept home from school to work
in the fields.
Rural Labor Productivity. Our main hypothesis is that part of the
Downloaded by Purdue University Libraries At 10:40 28 June 2016 (PT)

empirical analysis was that schistosomiasis and other parasitic diseases

debilitate their victims and thereby diminish productive potential. By
"productive potential" we mean the worker's productivity capacity, for a
given number of hours worked and with a given stock and quality of
capital, land, and other complementary resources.
The fact that a worker can adjust to any debilitating effects of disease
by shifting to work that is less demanding physically or may alter the
amount of time worked, confounds the attempt to assess the effects of
disease on productive potential. Debilitation may be thought of as a
change in relative prices confronting a worker - that is, a change in the
relative wages he can earn at various types of work and at work versus
leisure. Thus, in order to understand better the effects of disease on
productive potential, it is necessary to examine not only the effects on
total weekly marginal productivity (as estimated by earnings), but also
the worker's adjustments in the type and amount of work performed.
Therefore, in addition to testing our main hypothesis that disease leads
to reduced earnings, we also tested three hypotheses that underlie it:
(1) infected workers work at jobs that are physically less demanding;
(2) they are less productive per-day-worked on any given type of work;
and (3) they work fewer days per week. A variety of control variables -
including the person's age, education, housing circumstances, and
outside work activities were included in the regressions that attempted
to explore the associations between these labor supply and productivity
variables and the presence of parasitic infections.
The hypothesis that parasitic infection leads to reduced weekly
earnings for the banana estate workers was not supported. Only one
disease variable, Strongyloides, was significant - and that one was
significant only for females. For the male workers the coefficient signs
for all five parasites were negative, as expected, although not statistically
significant. An F-test for the combined effect of all five parasites, how-
ever, disclosed that the total effect on weekly earnings of the entire
group of parasitic infections was significantly negative at the 05 level.

114 IJSE
 Disease and Economic Development

Individual workers on the banana estate in our study can, to a con-

siderable extent, decide for themselves whether to do "task" work (for
which piece rates are paid) or "day" work. Although task work is
physically more demanding than day work, a highly-productive worker
can earn considerably more per day on a task job. Thus, we used the
distinction between task work and day work as equivalent to the choice
between more-demanding and less-demanding work in testing our
second hypothesis. The dependent variable used in our test was the
proportion of days worked at task work. The test results failed to support
the hypothesis that infected workers would choose to work at physically
less demanding jobs more often than non-infected workers. None of the
disease or other variables was significant in explaining the observed
variation among workers in the percentage of task work performed.
The third hypothesis, that infected workers' productivity per day is less
than that of non-infected workers, was tested using two alternative
dependent variables: earnings per day for all types of day and task work
taken as a whole, and earnings per day for selected specific task jobs.
Downloaded by Purdue University Libraries At 10:40 28 June 2016 (PT)

These tests showed that schistosomiasis was indeed associated with

lower daily earnings for males, and presence of Strongyloides was signifi-
cantly related to lower earnings for females. The estimated effect on
earnings of being infected with schistosomiasis represented a reduction of
about 30 percent from the average daily earnings of male plantation
workers; the indicated effect of Strongyloides infection was a 15 percent
reduction of average daily earnings by female plantation workers. Among
the tests for individual tasks, no consistent pattern appeared.
The hypothesis that infected workers have higher absentee rates than
healthy workers was tested by regressing the average-number-of-days-
worked-per-week (during the 18 month period) on the same set of
disease and other independent variables as above. Again, none of the
five parasitic infections was associated for males with a significantly
smaller number of days worked, while for females, only one parasite,
Strongyloides, was so related. For males the sign of the coefficient
indicating intense schistosomiasis infection (that is, "many eggs") was
actually significantly positive. Our interpretation of this finding is that
workers respond to a decrease in their daily productive capacity by
working more days per week in order to minimize the effect of their
debility on their total weekly earnings. In summary, it seems that
schistosomiasis infection is reducing the productive potential for males
by a rather substantial 30 percent, but is having essentially no effect on
actual market output.
Urban Labor Productivity. Labor productivity hypotheses were also
tested for workers (all female) at a firm located in St. Lucia's capital city,
Castries, who were employed in the relatively sedentary tasks of folding
and assembling paper advertising material. (4)
The hypothesis that infected workers are absent more was not
supported. Neither schistosomiasis nor any of the other parasitic
infections was associated significantly with the percentage of days
worked.
Physical output data for three separate jobs were used to test the

Vol. 1, No. 1 115

 Disease and Economic Development

hypothesis that parasitic infection reduces actual productivity. Results

for two of the jobs produced no support for the hypothesis. For the third
job, presence of one parasitic disease, Ascaris, was associated with
significantly lower productivity, but presence of hookworm was asso-
ciated with significantly higher productivity.
Next, the question was examined: is parasitic infection associated
with lower level of weekly earnings from this firm? The results of
regressing weekly salary on the five disease variables and the other
independent variables show that none of the independent variables was
significant at even the 10 level.
In another test we compared the estimated coefficient signs for the
disease variables in all of our various studies. That is, we compared the
frequency with which each disease was associated with the expected
effect on any of the demographic, school performance, or labor pro-
ductivity variables, with the frequency with which an unexpected effect
was found, regardless of the magnitude or significance level of the effect.
Among the five parasites, only Ascaris showed a statistically significant
Downloaded by Purdue University Libraries At 10:40 28 June 2016 (PT)

excess of expected associations for the various kinds of effects taken as

a whole. The results, when pooled over all five parasitic diseases and all
bodies of data do indicate a statistically significant preponderance of 91
expected effects of the diseases as compared with 63 non-expected
effects. Schistosomiasis and Ascaris are the greatest contributors. That
is, on the whole, there do appear to be adverse effects of disease
infections, but the effects are generally quite small.

Conclusion
There are a number of possible explanations for our findings, and these
are discussed in some detail in our book. In our judgment the quality, the
quantity and the variety of our data were quite satisfactory, on the whole.
The possibility of model-misspecification always exists, but given the
consistency of our independent results, we have some confidence that
this is not at the root of our findings of few significant effects of the
diseases. In addition, our models — developed separately by each of
several researchers - examined a number of alternative functional forms
for the models but found no evidence that the general lack of disease-
impact was due to non-linearities; this leads us to believe that our
findings are probably not attributable to incorrect identification of
functional form. We conclude that the evidence is rather strong that
schistosomiasis and the other parasitic diseases studied in St. Lucia are
actually having little or no effect in any of the three broad forms we were
investigating: birth and death rates, performance of children in school,
and productivity of adult workers. These judgments, however, remain
hypotheses that require further testing. Similarly, the questions of
whether schistosomiasis and the other parasites have larger impacts
outside St. Lucia, or will have larger impacts later on, remain open. We
hope that the methodology presented in this study, by directly con-
fronting demographic and productivity impacts, will stimulate future
studies in other geographic areas and for other diseases.

116 IJSE
 Disease and Economic Development
References

(1) Estimates of the number currently infected range from in excess of 100 million to in excess of
200 million persons. (1) (2)
(2) During the past four years, however, a large expansion in tourist hotel construction has drawn
laborers away from the now troubled banana industry. With the entry of the U.K. into the
European Common Market, St. Lucian banana producers are also threatened by competition
from French Overseas Departments.
(3) The 93 percent response rate to our Cul-de-Sac survey, compared to the 74 percent response
in Roseau, noted above, represents a more intensive effort to obtain personal attribute data
for households which failed to cooperate with the associated parasitological survey. This
additional information for Cul-de-Sac—household attribute data in the absence of matching
parasitological data—assisted our determination that there was no response bias in the
parasitological survey such as would tend to understate our disease impact findings.
(4) The relatively uniform level of physical exertion among the jobs at this firm precluded our
testing the hypothesis that disease causes workers to shift to physically less-demanding jobs.
(1) Weir, John M. "The Unconquered Plague", Bulletin of the Atomic Scientists, October, 1966,
22 (8), 46-48
(2) Wright, W. H. "Schistosomiasis as a World Problem," Bulletin of the New York Academy of
Medicine, XLIV, pp. 301-312.
Downloaded by Purdue University Libraries At 10:40 28 June 2016 (PT)

Vol.1. No. 1 117

 This article has been cited by:

1. Melissa Parker. 1993. Bilharzia and the boys: Questioning common assumptions. Social Science &
Medicine 37:4, 481-492. [CrossRef]
2. Melissa Parker. 1992. Re-assessing disability: The impact of schistosomal infection on daily activities
among women in Gezira Province, Sudan. Social Science & Medicine 35:7, 877-890. [CrossRef]
3. Yixin Huang, Lenore Manderson. 1992. Schistosomiasis and the social patterning of infection. Acta
Tropica 51:3-4, 175-194. [CrossRef]
4. Michael Drummond, Anne Mills. 1989. Developing economics expertise for the evaluation of
primary health care. Public Administration and Development 9:1, 83-96. [CrossRef]
5. Martine Audibert. 1986. Agricultural non-wage production and health status. Journal of Development
Economics 24:2, 275-291. [CrossRef]
6. Nicholas M. Prescott. 1979. Schistosomiasis and development. World Development 7:1, 1-14.
[CrossRef]
Downloaded by Purdue University Libraries At 10:40 28 June 2016 (PT)