International Review of Business Research Papers

Vol. 8. No.6. September 2012. Pp. 113 – 125

Demand for Child Healthcare in Cote d'Ivoire: A Multinomial

Probit Analysis
Tito Nestor Tiehi1
In Côte d'Ivoire, despite priority given to child care by governments, child
mortality remains high and malnutrition and anemia among children
reached alarming proportions. Our study tests influences of variables that
condition demand for child healthcare. The sample covered 1,023
children who used a medical consultation and data are from Living
Standards Measurement Survey of 2008. We obtain that tariff of
consultation has negative impact while distance positively influences
decision to consult modern health services; households headed by men
are less likely to utilize modern services and as well as wealthy and
highly-educated, households have a preference for private care.

JEL Codes: C25, D13 and I14

1. Introduction
Health indicators in Côte d'Ivoire contrast with economic situation of this country
previously prosperous. Yet, Côte d'Ivoire has made the development of modern health
care a priority objective for governance. Funds have been mobilized with support of
development partners for building health centers and implementation of programs
intended for vulnerable people. However, health indicators are not better; supervision of
pregnancy and birthing process is low and use of maternal health services by pregnant
women decreases with progress of pregnancy. Consequently, child mortality (12.3 per
1000 births) and maternal mortality (470 deaths per 100,000 live births) remains high,
malnutrition and anemia among children who are five years old or less, reached alarming
proportions (MSP, 2008).

This paradox is the result of a prestigious health policy adopted by Côte d'Ivoire since its
independence. Thus, in this country health system is characterized by an oversized
tertiary sector, a secondary sector particularly reduced and a primary supply clearly
insufficient: about 300 public functional health centers throughout the country. In parallel,
private medical supply, particularly abundant in Abidjan, is highly deficient in provinces.
Moreover, Ivorian health system remains dependent on a network of an unstructured
traditional care but highly active. Therefore, for a given episode of illness, people use
concomitantly or sequentially several forms of therapies and resources (Haddad, 1994),
so that use of modern medicine has stiff competition from other subsystems of
healthcare, such as self-medication and traditional care (tradithérapy).

In Côte d’Ivoire, despite establishment of programs for maternal and child health (e.g.
national program of reproductive health and child health program) aiming reducing
morbidity and mortality of mothers and children, health situation of children remains
highly marked by eminence of neonatal disease, malaria, acute respiratory infections and
malnutrition. According to Ivoirian Public Health Ministry, these four causes explain 94%

1
Dr. Tito Nestor Tiehi, Department of Economic and Management, University of Cocody, Côte d’Ivoire.
Email: titotiehi@gmail.com
 Tiehi
of infant mortality and one-third of children five years old or less suffer from chronic
malnutrition or stunted growth in Côte d'Ivoire. In consequence, child mortality (119 ‰ in
2009) remains fairly high, reflecting that health policies had no significant effect on infant
mortality (Tiehi 2011).

Unfortunately, despite these alarming health statistics, few studies have focused on
demand for child care in Côte d’Ivoire. Therefore, it seems important to
understand determinants of children's survival process in Cote d’Ivoire. Such an
investigation should drive to not only design in favor of children a suitable health policy,
but also look for ways for implementation and resource mobilization in a timely manner.

The objective of our study is to capture influences of variables that condition demand for
child healthcare and suggest policies to improve provision for health services to Ivorian
households. For that, we ask following set of interrelated questions. What are the
determinants of primary healthcare use? How important are income, tariff, access,
gender, education level and others in choice of health care provider? How do the poor
and rich households make decision about treatment of their ill child in response to the
tariff?

To conduct this analysis, we first present major analysis that looked at demand of
healthcare in developing countries (section 2); then we describe the data and the
econometric model, basis of our research (section 3); we present and discuss in following
session our results (section 4); and finally, in light of these results, we draw conclusions
(section 5).

2. Literature Review
Since work of Arrow (1963) on the unstable, irregular, and unpredictable individual
characters of demand on the market for health care, the utility of care and the
opportunity cost of the disease, analysis of demand has declined in several ways. Among
these approaches, we have those of Grossman (1972), who first proposed a model of
health demand in the context of the theory of human capital and redefines the concepts
of demand for health care and demand for health. Beyond theoretical construction,
Grossman’s approach is original in that it lays foundation for a conceptualization of
demand for medical care as derived from demand for health.

Subsequently, theoretical analysis of the demand for health care extended Grossman
model so that, concept of "health capital" is a tangible translation into measures
developed to assess utility of health states. These measures extend from the 80's to the
issue of choice of care provider and theoretical framework of these measures is
presented by Acton (1975) and Christianson (1976). For these authors, demand for care
is a complex process, covering different decisions, starting with the declaration of an
episode of illness and continuing through choice of a mode of treatment. Then, analysis
moves not only to the study of discrete choice and their determinants, but also to
decision-making processes from which these choices are made (Cissé et al 2004).

The first work (Heller 1982) analyzes behavior of care consumption, to a change in tariff
of public health services. It concludes that there is an inelasticity of demand relative to
the price of care and patient income. Accordingly, there would be no significant
correlation between applied tariffs, income of households and demand for health
services. On other hand, Gertler and Gaag (1990) showed a significant correlation

114
 Tiehi
between demand and hospital tariffs and this correlation varies from child to adult and
according to income and age of individual. For this purpose, they conclude that for adults,
elasticity varies between – 0.38 (for individuals with high income) and -1.83 (for poor),
whereas for children, this elasticity varies between -0.31 (for children from rich
households) and -2.82 (for those from poor households). More recently, Yates et al.
(2006) and Kermani (2008) in their works on Uganda and Iran, showed existence of a
negative relation between price of consultation and utilization ratio of the health service
(public or private).

Overall, research which studied the decisions of individuals regarding disease had to also
analyze influence of factors other than price of medication and income. The choice of
these factors which does not follow a particular rule of selection rather relies on the
demand for healthcare as for any other good. Such an approach is likely to be affected
by parameters other than price and income. Thus, in the beginning, Dor et al. (1987) and
Gertler and Gaag (1990) assess discriminating effect of sex of the head of household (or
decision maker), influence of number of persons in the household, age of the head of
household, etc in decision of health service use.

Thus, age (of patient or decision maker) was analyzed as a quantitative variable.
However, with regards to its nonlinear effects on probability to use health services, Sahn
et al. (2003) considered a discrete measurement of age to assess its impact on choice of
healthcare provider. On other hand, more recently, Sarma (2009) and N’Tembe (2009)
capture this nonlinear effect on decision to use care services by introducing the square of
variable « age ». Moreover, size of households and level of education (Mariko 2003)
were often used as explanatory variables of decision of households. In the same way,
area of residence influences decision of households and it is shown that rural households
declare less morbid episodes (Chima et al. 2003).

In Côte d’Ivoire, Dor et al. (1987) estimated and analyzed determinants of health care
demand and choice of provider in rural areas. They showed that health status, income,
and price of consultation constitute the most determinant factors of the decision of an
individual to resort to modern care. More specifically, Gertler and Gaag (1990) found that
use of health services is inelastic to the price of health care in rural area in Côte d’Ivoire.
Audibert et al. (1998), using a multinomial probit model, studied access to care by
considering three possible choices (self medication, traditional system, modern system)
and identified variables which determine probability of resorting to a given type of care
provider. Furthermore, Perrin (2001) evaluated the impact on demand following a
modification care tariffs and other factors, such as quality of care and income of the
patients (or decision maker). She found that riches individuals are more sensitive to the
quality of the care, whereas poor are significantly affected by the raising of prices of the
care.

In sum, we deduce from last studies that effectiveness of cost recovery policies depends
on socio-economic structure of the country. Therefore, necessity of introducing or
increasing the cost of access to care (especially the tariff) as a means to mobilize
resources for financing of health system should take into account situation of each
countries and particularly situation of vulnerable individuals. To this end, our study aims
to identify factors influencing decision of household in choosing child healthcare provider
in Côte d'Ivoire.

115
 Tiehi
3. The Methodology and Model
Data are from Living Standards Measurement Survey 2008 (LSMS 2008) of National
Institute of Statistics of Cote d’Ivoire. This survey covered 16,200 households for a total
of 59,700 people of all ages living at home. The survey covers a number of socio-
demographic subjects (household structure, level of education, place of residence,
income), a description of living conditions (home furnishings, access to technical
infrastructures), as well as data on individual’s health condition and access to the health
care. From this data pool of surveyed individuals, we extracted children five years old or
less, that is 9,954 individuals. The number (2,340) of ill children in this group is reduced
to 1,196 children having been declared sick during the four last weeks preceding the
survey. While excluding from this group (1,196), individuals imperfectly surveyed, sample
is restricted to 1,023 children who used a medical consultation.

The medical recourses are gathered in categories of medical alternatives representative

of principal options available to the head of household when one of the children of family
is sick: (i) self medication, (ii) private hospital, (iii) public dispensary, (iv) public hospital.
Specifically, "self medication" alternative categorizes care received outside of formal
structures. Thus, it is care received by an individual without any preliminary doctor visit
and about care offered by traditional caretakers. As for "private hospital" alternative, it
includes private clinics and establishments of care whose capital is held by private
actors. "Public dispensary" alternative represents health centers (urban and rural) of first
contact. Lastly, "public hospital" alternative categorizes public sanitary institutions of
second contact which are located in districts and exempts primary care. These medical
alternatives represent dependant variable in estimated model, comprising four modalities
not coordinated and mutually exclusive.

The explanatory variables are those which are most commonly used in analysis of the
determinants of care demand resultant of seminal work of Getler and Gaag (1990) and
Dor et al. (1987). These variables consist of a range of socio-demographic factors
(distance to the care center, gender of the head of household, age of head of household,
level of education of head of household, place of residence of household, and size of
household) and economic factors (income of household, tariff of consultation). In
following table, explanatory variables and assumptions on expected effects are briefly
described.

116
 Tiehi
Table 1: Specification of the variables
Variables Description Expected Impact

Tariff Tariff of consultation (ref)

(-)
(-)
(-)

Income Monthly income approximated to (ref)

consumer expenditure (+)
(-)
(-)
Distance Distance between place of residence (ref)
and provider (-)
(-)
(-)
Gender (Man) =1 if head of household is a man (ref)
(+)
(+)
(+)
Area (urban) =1 if household lives in urban area (ref)
(+)
(+)
(+)

Education Number of years of study to obtain the (ref)

last diploma by head of household (-)
(-)
(-)

Age Age of head of household (ref)

2
Age Square of the age of head of household (-)
(+)
(+)

Household size Number of individuals living in household (ref)

(-)
(+)
(-)

Our approach follows, work of Gertler and Gaag (1990) and the econometrics model is
inspired by works of Akin et al. (1993) and Mwabu et al. (2003). On basis of assumption
that individual (or decision maker) is rational, it will choose alternative which gives the
highest level of utility U i to him among whole of p possible choices. Formally, we note
U ij utility of an individual i when he receives care of a type of provider j. In absence of
specific information on each alternative (such as tariff of consultation for example), this
utility is defined as follows:

U i  X i  j   ij
(1)
1 j  p

117
 Tiehi
In this equation X i is the matrix of specific explanatory variables relative to each
individual. Matrix X i does not vary accross alternatives, and it represents observable
component of the function of utility. Term  ij represents the stochastic component which
captures the unobserved share of the utility. The matrix form of the equation arises as
follows:

U i1   X i 0 ... 0   1   i1 
      
U i 2    0 Xi ... 0    2   i 2 
  (2)
 ...   ... ... ... ...   ...   ... 
       
U ip   0 ... 0 X i    p   ip 

Given that utility cannot be observed, one is interested in probability that an alternative j
is selected compared to other alternatives. An individual chooses alternative j if utility he
withdraws is higher than that of all other alternatives.

1 if U ij  U ik  k  j
y ij   (3)
0 if no

For an individual i, the probability to choose alternative j is equal to:

P( yij  1)  P(U ij  U ik ,  k  j ) (4)

This probability (equation 4) is conditioned by nature of the distribution followed by the

disturbance  ij . Assuming that  ij is normally, identically, and independently distributed
(  ij  (0, ) with  a matrix of covariance p x p without restriction of independence of
disturbance between alternatives), the defined model characterizes multinomial probit
model (Hausman and Wise 1978).

This model (multinomial probit model) is more adapted in cases where alternatives are
similar (like choice of a medical alternative), and Independence of Irrelevant Alternatives
hypothesis (IIA hypothesis) proves implausible, insofar as absence of an alternative is
likely to favor a similar one (Hausman and Wise 1978). Under assumptions of normality
of error terms and homoscedasticity (Alvarez and Nagler 1995; Rudolph 2003), the
model is rewritten in terms of differential of utilities by taking “self-medication” alternative
and is referenced as follows:

U i*  U ij  U i1  X i ( j  1 )  ( ij   i1 )  X i  *j   ij* (5)

Where
 0   1  23  24 
 
( ,  ,  )   0,  
* * *
   23 1  34 
2 3 4 and
 0   24  34 1 
  

118
 Tiehi
Coefficients in discrete choices models do not have direct economic interpretation,
because of problem of standardization of residual variance. This is why we calculate
marginal effects as follows:

Pi  j

j   Pj  j   Pk  k   Pj  i    (6)
xi  k 0 

The marginal effects describe sensitivity of probability of event (yi = 1) compared to

variations of the explanatory variables to give a sense to the obtained results.

4. The Findings
Table 2: Distribution of households according to medical alternatives
Observations Frequency Percentages

Self medication 1023 44 4.30%

Private hospital 1023 336 32.84%
Dispensary 1023 520 50.83%
Public hospital 1023 123 12.03%

In the sample of this study, 4.30 percent of households had recourse to self-medication
and 32.84 percent of them profited from services of private hospitals. Public dispensaries
are used with a rate of frequentation of 50.83 percent whereas public hospitals record
12.03 percent of recourse (Table 2).

Table 3: Socio-demographic and economic characteristics of the households

Self medication Private Hospital Public Dispensary Public Hospital
Variables Obs. Average Min Average Min Average Min Average Min
(std-dev) (Max) (std-dev) (Max) (std-dev) (Max) (std-dev) (Max)

Tariff of 1023 1396 200 5898 100 1728 100 47 30 200

consultation (2701) (30000) (3020) (30000) (995) (10000) (1049) (10000)

Income 1023 66942 6033 66942 6033 66942 6033 66942 6033
(63573) (916666) (63573) (916666) (63573) (916666) (63573) (916666)
Distance 1023 0.55 0 12.13 1 10.89 1.50 13.25 1.50
(3.27) (50) (10.35) (60) (8.22) (50) (10.03) (56.20)
Gender 1023 1.47 1 1.47 1 1.47 1 1.47 1
(man) (0.499) (2) (0.499) (2) (0.499) (2) (0.499) (2)
Age of head 1023 40.51 21 40.51 21 40.51 21 40.51 21
of household (12.85) (99) (12.85) (99) (12.85) (99) (12.85) (99)
Area (urban) 1023 1.49 1 1.49 1 1.49 1 1.49 1
(0.500) (2) (0.500) (2) (0.500) (2) (0.500) (2)
Household 1023 6.2 1 6.2 1 6.2 1 6.2 1
size (3.641) (37) (3.641) (37) (3.641) (37) (3.641) (37)
Education 1023 3.95 1 3.95 1 3.95 1 3.95 0
(diploma) (4.47) (22) (4.47) (22) (4.47) (22) (4.47) (22)

Table 3 reveals that average tariff for consultations in case of self-medication is 1,395
francs CFA (3.1 USD), tariff is 5,897 francs CFA (13.1 USD) for private hospitals, of
1,727 francs CFA (3.83 dollars us) in public dispensaries and 4729 francs CFA (10.5
USD) when public hospital is chosen. Maximum tariffs in case of self-medication and

119
 Tiehi
consultation in a private hospital are approximately 30,000 francs CFA (66.67 USD),
whereas this one is of 10,000 francs CFA (22.22 USD) in public dispensary and public
hospital. We also note that average distance is 0.55 km for self-medication, of 12.13 km
for the private hospital, of 10.89 km for public dispensary, and of 13.25 km for public
hospital. Maximum distance for all alternatives varies between 50 km and 60 km.

Moreover, 52.98 percent of households are headed by men, while 47.02 percent are by
women. In total, 50.24 percent of households are located in urban areas. In addition,
majority of surveyed heads of household have a level of education which does not
exceed primary school (4 years of study on average). The average number of individuals
in a household is rather high (6.2 individuals). The average age of heads of household is
40.51 years and average monthly income of households is 66,941 francs CFA. (148.75
USD)

Table 4: Coefficients and marginal effects of the demand for infant care
Private Hospital Public Dispensary Public hospital
Prob.= 0.310 Prob.= 0.5912 Prob.= 0.0759
Variables coefficient elasticity coefficient elasticity coefficient elasticity
*** *** *** *** *** ***
Tariff of consultation - 0.2593 - 0.0755 - 0.2593 - 0.1087 - 0.2593 - 0.052
(0.000) (0.000) (0.000) (0.000) (0.000) (0.000)
*** *** *** *** ***
Income (Log consumption) 0.1010 0.1335 - 0.3509 - 0.1044 - 0.4622 - 0.0455
(0.516) (0.000) (0.008) (0.000) (0.002) (0.002)
*** *** *** *** *** ***
Distance 0.5259 0.1533 0.5259 0.2206 0.5259 0.1040
(0.000) (0.000) (0.000) (0.000) (0.000) (0.000)
*** *** **
Gender (man) - 0.5269 - 0.1139 - 0.1419 - 0.0846 - 0.1357 - 0.0177
(0.006) (0.001) (0.367) (0.018) (0.432) (0.326)
Age of head of household 0.0637 0.0114 0.0260 - 0.0073 0.0226 - 0.0021
(0.106) (0.135) (0.416) (0.333) (0.520) (0.561)
2
Age of head of household - 0.0006 - 0.0001 - 0.0002 - 0.0000 - 0.0002 - 0.0000
(0.117) (0.124) (0.483) (0.294) (0.575) (0.563)
** *** ** *** * *
Area (urban) 0.4267 0.2281 - 0.3826 - 0.2048 - 0.3694 - 0.0368
(0.040) (0.000) (0.033) (0.000) (0.059) (0.051)
**
Household size - 0.0136 - 0.0082 0.0119 0.0028 0.0356 0.0060
(0.648) (0.133) (0.646) (0.614) (0.180) (0.022)
* *** * ** ***
Education (diplôma) 0.0365 0.0154 - 0.0119 - 0.0073 - 0.0432 - 0.0085
(0.091) (0.000) (0.513) (0.091) (0.057) (0.001)
*** ***
Constant -0.8617 …… 4.576 …… 5.8367 ……
(0.642) …… (0.003) …… (0.001) ……
Integration points: = 200 Wald chi2(23) = 94.65
Log simulated-likelihood = -756.3201 Prob > chi2 = 0.0000
The numbers in bracket are the probabilities;
Reference alternative: self-medication
(***) significance at 1%,
(**) significance at 5%,
(*) significance at 10%.

Self-medication (Prob=0.022) is medical alternative least preferred as a first recourse

when a child falls ill in Ivorian households; probability of resorting to public hospitals
(Prob=0.076) is slightly higher. Private hospitals (Prob=0.311) and public dispensaries
(Prob=0.591) are medical alternatives most desired for infant care. These results reflect
importance of public sector in offering primary care and noteworthy place of private
health centers in Ivorian medical system. Our results contradict with those of Juillet
(1999) and Mariko (2003) which, based on samples of a heterogeneous population,
reveal preeminence of self-medication over formal alternatives in choice of health
provider. Indeed, although it is possible to extend their conclusions to Cote d'Ivoire in
120
 Tiehi
terms of demographic, economic, and cultural similarities, our results nevertheless
demonstrate that when health of their children is concerned, Ivorian households prefer to
address modern physicians.

Considering self-medication as reference alternative, estimation emphasizes that

probability of using formal care (private hospital, public dispensary and public hospital)
drops significantly with tariff of consultation. So, an increase of 1 percent of tariff of
consultation reduces probability of using private hospital by 7.55 percent. As well, an
increase of tariff of consultation of 1 percent reduces probability of using public
dispensary by 10.87 percent and probability to use public hospital by 5.02 percent. This
result is consistent with work of Fournier and Haddad (1997) and of Creese (1997), which
showed a negative correlation between demand for modern care and price of
consultation. Also, our results reflect that of Nyonator and Kutzin (1999), Chawla and
Ellis (2000), Buor (2004), and Kermani (2008) which reveals that increase of tariffs has a
negative effect on use of modern health services, with a prejudice against the most
vulnerable groups.

Household’s income has a positive effect on use of private care. However, the use of
public services is negatively affected by income of households. When household income
increases by 1 percent, decision to use private establishments raises 13.45 percent,
whereas in governmental establishments, it decreases to 10.62 percent in dispensaries
and 4.47 percent in public hospitals. Richest households have a preference for care
provided by private institutions, unlike poor households that prefer public health facilities.
In other words, increase of resources of households is an important determinant of
access to health care provided by modern private services. This result confirms one of
conclusions of Shaw and Griffin (1995), according to which, wealthy individuals are more
willing to pay for more expensive services.

Apart from financial parameters of access to health care, our study was interested in
socio-demographic characteristics of households. Contrary to our expectations and work
of those such as Carrin et al (1993) and Ensor and Cooper (2004), our results show that
distance of access to healthcare positively influences the demand for care. As
paradoxical as it may seem, the positive impact of distance on probability of choosing
formal care by households in Côte d'Ivoire can be explained by proximity of health
facilities.

Hence, distance is not decisive in the decision of care demand. In this respect, although
significant (with a 1 percent threshold), increase of the probability to choose modern care
following an increase of 1% of distance is very weak (0.015 percent; 0.02 percent and
0.01 percent respectively for private hospitals, public dispensaries and public hospitals).
This result is in line with that of Sauerborn et al. (1994) which show that in case of grave
or serious diseases, distance does not constitute an obstacle in utilization of modern
health services. Mariko (2003) also obtains similar results in his study on use of health
services in Bamako (Mali). So, we can conclude that in Cote d'Ivoire, distance is not a
limiting factor in the decision of parents to use modern care services for their children.

With respect to gender, our results show that, in comparison to female heads of
households, male are less likely to require care for the ill children. This inclination is
stronger in case of private providers, so that decision to choose a private hospital drops
by 11.3 percent, whereas probability of resorting to public dispensary decreases by 8.46
percent, when one passes from a household headed by a woman to a household headed

121
 Tiehi
by a man. These results are contradict with those of Nonvignon et al. (2009), who show
that in a study on demand for care for infants with malaria in Ghana, probability of using
modern care increases when payment of cost of care is covered by men. However, our
findings remain in conformity with previous works (such as those of Sahn et al. 2003)
which concluded to the existence of a negative effect of gender on demand of care in a
household headed by a man.

Also, we note that compared to rural areas, probability of using private care
establishments in urban areas increases significantly by 22.8 percent, whereas
probability drops by 20.5 percent for public dispensaries and 3.67 percent for public
hospitals. These results reflect possibility of a larger substitution between private
providers and public providers in urban areas. So urban households are less constrained
in their choice of health provider, and, as Bitran (1995) specified, in the absence of
information on quality, private establishments are regarded as being much better in
quality than public care services.

Education level of household head affects positively and significantly the demand for
child healthcare in private clinics. In contrast, frequentation of public care service
decreases when education level of household head is high. This negative impact is
significant only for public hospitals. The preference for private health centers increases
with level of education of head of household so that probability of resorting to private care
increases significantly by 1.53 percent for every additional year of education. As level of
education rises, households tend to deviate from care offered by governmental services
so that probability of choosing a public dispensary drops by 0.7 percent, whereas
probability of choosing a public hospital falls by 0.88 percent. These results are partly in
conformity with those of Strauss and Thomas (1998), for whom parent’s education level
has a positive and significant impact on child health status as it increases their level of
efficiency in evaluating health of their children and use of available health services.
Heads of households which have high levels of education give more credit to efficiency of
modern private care.

5. Summary and Conclusions

Concept of health care demand is closely related to search for healthcare of the
individuals. In their step, endogenous and exogenous parameters influence household
decision to be seen at one provider rather than another. The investigation of these
parameters is important, particularly when it is question of looking after children, i.e.
those who are most vulnerable due to higher rate of morbidity. The objective of this study
is to measure and understand influence of factors on the demand of infant care in Cote
d'Ivoire.

Our study uses a model of discrete choice inspired by the estimation approach of Akin et
al. (1993) and Mwabu et al. (2003), and is useful in examination of situations where
individuals have to choose between several medical alternatives. Our regression point
out that tariff of consultation negatively affects decision of using modern health services
(public or private) so that an increase in tariff significantly reduces probability of
demanding care for ill children. Also, as expected, income affects recourse to healthcare
services when a child is sick. Thus, as shown, by Gertler et al. (1990), and more recently
by Sarma (2009), we find that tariff of consultation is the principal limiting factor of access
to modern health services and increase in income positively increases demand for health
care for ill children.

122
 Tiehi

Otherwise, our study assesses effects of socio-demographic characteristics of

households. In this respect, our estimates show, firstly, that distance from a service
provider increases probability of resorting to modern care and that households headed by
men are less likely to require infant care. Secondly, we find that compared to rural areas,
urban households have a very significant preference for private hospitals, whereas
probability of using public services in town drops. Lastly, our study reveals that probability
of using private care increases with level of education.

There are a number of limitations with this study and its results. Among other we have
the bad quality of data. Thereby, generalizing study results to all countries of West Africa
on basis of homogeneity of household behavior should be done with extreme caution.
Also, our study doesn’t take into account the characteristics of medical alternatives.
However, it is clear that quality of care, patient intake and other characteristics of health
services are important in the household’s decision. Nevertheless, this study is a
preliminary attempt in the direction of modeling demand for child healthcare in Côte
d’Ivoire and can motivate future research.

References
Acton, JP 1975, ‘Non-monetary factors in the demand for medical services: some
empirical evidences’, Journal of Political Economy, vol. 83, n°3, pp. 595-614
Akin, JS Guilkey, DK and Denton, HE 1995, ‘Quality of services and demand for health
care in Nigeria: a multinomial probit estimation’, Social Science and Medicine,
vol.40, n° 11, pp. 1527-37.
Alvarez, RM and Nagler, J 1995, 'Economics, issues and the Perot candidacy: voter
choice in the 1992 presidential election', American Journal of Political Science,
vol.39, n° 3, pp. 714-44.
Arrow, KJ 1963, ‘Uncertainty and the Welfare Economics of Medical Care’, American
Economic Review, vol.53, n° 5, pp. 941-73.
Audibert, M, Mathonnat, J, Nzeyimana, I and Henry, MC 1998, ‘The determinants of
health care demand amongst the senufo of Côte d'Ivoire’, Health and System
Science, vol.2, n° 1-2, pp. 111-26.
Bitran, R 1995, ‘Efficiency and quality in the public and private sectors in Senegal’,
Health Policy and Planning, vol.10, n° 3, pp. 271-83.
Buor, D 2004, ‘Gender and the utilization of health services in the Ashanti Region,
Ghana’, Health Policy, vol.69, pp. 375-88
Carrin,G, Perrot, J and Sergent, F 1993, ‘The Influence of Financial Participation on
Health Care Demand: An Analytical Tool for Countries in Greatest Need’, WHO,
Geneva.
Chawla, M and Ellis, RP 2000, ‘The impact of financing and quality changes on health
care demand in Niger’, Health Policy and Planning, vol.15, n° 1, pp. 76-84.
Chima, RI, Goodman, CA and Mills, A 2003, ‘The economic impact of malaria in Africa: a
critical review of the evidence’, Health Policy, vol. 63, pp. 17-36.
Christianson, JB 1976, ‘Evaluating locations for outpatients medical care facilities’, Land
Economics, vol.52, n° 3, pp. 299-313.
Cissé, B, Luchini, S and Moatti, JP 2004, Recouvrement des coûts et demande de soins
dans les PED’ Revue Française d'Economie, vol.18 n° 4, pp. 111-49.
Creese, A 1997, ‘They don’t reduce costs, and they increase inequity’, British Medical
Journal, vol.315, pp. 202-03.

123
 Tiehi
Dor, A, Gertler, P and Van Der Gaag, J 1987, ‘Non-price rationing and the choice of
medical care providers in rural Cote d’Ivoire’, Journal of Health Economics, vol.6, n°
3, pp. 291-304.
Ensor, T and Cooper, S 2004, ‘Overcoming barriers to health service access: influencing
the demand side’, Health Policy and Planning, vol.92, n° 2, pp. 69-79.
Fournier, P and Haddad, S 1997, ‘Dimensions de la qualité des services de santé en
Afrique au Sud du Sahara’, In BRUNET-Jailly, (ed.), Innover dans les systèmes de
santé : Expériences d’Afrique de l’Ouest, Karthala, Paris, pp. 275-90.
Gertler, P, Van Der Gaag, J 1990, The willingness to pay for medical care: Evidence from
two developing countries, Baltimore: Johns Hopkins University Press.
Grossman, M 1972, ‘On the concept of health capital and the demand for health’, Journal
of Political Economy, vol.80, n° 2, pp. 223-35.
Haddad, S 1994, ‘Utilisation des services de santé en pays en développement. Une
étude longitudinale dans la zone de santé rurale de Nioki au Zaire’, In G DURU
(ed.), Analyses et évaluations en économie de la santé, Hermes, Paris, pp. 227-52.
Hausman, JA and Wise, DA 1978, ‘A conditional probit model for qualitative choice:
discrete decisions recognizing interdependence and heterogeneous preferences’,
Econometrica, vol.46, n° 2, pp. 403-26.
Heller, P 1982, ‘A model of demand for medical and health service in peninsular
Malaysia’, Social Science and Medicine, vol.16, pp. 267-84
Juillet, A 1999, ‘L’impact des tarifs des services de santé et des revenus sur les
décisions de recours aux soins des malades à Bamako. Estimation d’un modèle
logit emboîté’, Revue d’Economie du Développement, vol.4, pp. 69-89.
Kermami, M, Ghaderi, H and Yousoufi, A 2008, ‘Demand for medical care in the urban
areas of Iran: An empirical investigation’, Health Economics, vol.17, pp. 849-62.
Mariko, M 2003, ‘Accès aux soins et qualité: résultats d’une étude empirique menée à
Bamako Mali’, In M Audibert, J Mathonnat and De Roodenbeke E (ed.), Le
financement de la santé dans les pays d’Afrique et d’Asie à faible revenu, Karthala,
Paris, pp. 5-37.
Ministère de la Santé Publique 2008, Annuaire des statistiques sanitaires 2007-2008.
Direction de l’Information, de la Planification et de l’Evaluation, Abidjan, Côte
d’Ivoire.
Mwabu, G, Wang’ombe, J and Nganda, B 2003, ‘The demand for health care in Kenya’,
African Development Review, vol.15, n° 2-3, pp. 439-53.
N’Ttembe, AN 2009, ‘User Charges and Health Care Provider Choice in Cameroon’,
International Review of Business Research Papers, vol.5, n° 6, pp. 33-49.
Nonvignon, J Aikins, MKS Chinbuah, MA., and others. 2010, ‘Treatment choices for
fevers in children under-five years in rural Ghanaian district’ Journal of malaria,
vol.9, pp. 1-8.
Nyonator, F and Kutzin, J 1999, ‘Health for some? The effects of user fees in the Volta
Region of Ghana’, Health Policy and Planning, vol.14, pp.329-41.
Perrin, H 2001, Justice sociale et santé: l’accès aux soins à Abidjan, Thèse de Doctorat
es Sciences Economiques, université de Toulouse IRD. Unpublished
Rudolph, TJ 2003, ‘Who's responsible for the economy? The formation and
consequences of responsibility attributions’, American Journal of Political Science,
vol.47, n° 4, pp.698-713.
Sahn, DE Younger, SD and Genicot, G 2003, ‘The Demand for Health Care Services in
Rural Tanzania’, Oxford Bulletin of Economics and Statistics, vol.65, n° 2, pp.241-
60.
Sarma, SK 2009, ‘Demand for outpatient health care: Empirical findings from rural India’,
Applied Health Economics and Health Policy, vol.7, n° 4, pp.265-77.

124
 Tiehi
Sauerborn, R Nougtara, A and Latimer, E 1994, ‘The elasticity of demand for health care
in Burkina Faso: differences across age and income groups’, Health Policy and
Planning, vol.9, n° 2, pp.185-92.
Shaw, RP and Griffin, CC 1995, Financing health care in Sub-Saharan Africa through
user fees and insurance, Washington, D.C: World Bank.
Strauss,J and Thomas, D 1998, ‘Health, Nutrition, and economic development’, Journal
of Economic Literature, vol.36 n° 2, pp.766-57.
Tiehi, TN 2011, ‘Population’s health status in WAEMU countries : an analysis according
to the theory of convergence’, In K KONDLO, and C EJIOGU (ed.), Africa in Focus:
Governance in the 21st century, HSRC Press, Cape Town, pp. 299-308,
WHO 1988, Economic support for national health for all strategies, WHO, Geneva,
Yates, J Cooper, R and Holland, J 2006, ‘Social protection and health: experiences in
Uganda’, Development Policy Review, vol.24, pp.339-56.

125