Effect of wealth inequality on child and infant mortality in Togo

For our estimates, we use data from the Multiple Indicator Cluster Survey (MICS6) conducted in Togo in 2017 by the National Institute of Statistics and Economic and Demographic Studies (INSEED). This is a household survey containing a section of questions for household and children under five years. Among the 8,065 households found for the information-gathering sessions, 7,916 were surveyed in order to obtain information on the situation of children and women throughout the country. In these households, 5,030 children were under 5 years old and 4,942 of them has been successfully interviewed. A large number of variables were filled in as part of this survey. However, for the purposes of this article, only the variables presented in Table 1 have been retained. A large number of variables were filled in as part of this survey.

The variables retained for our article are chosen on the basis of literature and theory. These variables are recorded in Table 1 where we also mention the expected sign of each of the variables.

Through this method, we estimate the distribution of the survival function, i.e. the distribution over time of the probability of not having experienced infant and child mortality. To do this, we take into account indicators such as:

We therefore deduce two empirical models that correspond respectively to the estimation of the effects on infant and child mortality and infant mortality:

Model 1: infant and child mortality

Model 2: Infant mortality

Table 2 presents the overall socio-demographic characteristics of the households included in the analysis. The chi-square test (Coron, 2020) significant at the 1% threshold confirm the association between ICM and another variable introduce in the model. The number of children who were the subject of the analysis is 11,731. These children are both “boys or girls” and come either from rural or urban areas. Our statistical analysis shows that the health outcomes of those children who are under age 5 years depend on the education of their parents. The data show with regard to infant mortality that, it’s boys who have the highest quotient “44 children out of 1000 live births die between their birth and their first birthday” against a quotient of 39 for girls. This quotient is lower when the level of education increases “25 deaths per 1000 live births for a secondary level and more against 48 for those with no level. Happiness statistics show that men are more unhappy than women (27.5% versus 18.8% respectively). Concerning child mortality, no difference is revealed in terms of the quotient recorded for girls as for boys. The number of children aged between 1 and 4 years per 1000 surviving children is higher when with uneducated parents. This quotient is 42 against 13 surviving children for children whose parents have at least secondary education. Observation of the data recorded in Table 2 also shows that the infant and child mortality rate is higher in rural areas (82) than in urban areas (53 deaths per 1000 live births. concerning the wealth inequality variable, our results are in percentage of survival. The statistics also show concerning wealth inequality that, in poorest households, respectively 89.94% and 76.39% of children survive beyond their first and fifth anniversary. However, these rates rise when it comes to the wealthiest. These are respectively 95.21% and 93.84% children who manage to celebrate their first and fifth anniversary.

We present through Fig. 3, the evolution of the probability of survival of children during the first eleven months and the fourth years following their delivery. The stairs in this figure represent the probability of survival during the different periods of death “four years or eleven months of observation”. Each height of a stair describes the risk of a child to die at some point in the analysis. The observations are made per year for the figure on the left (Fig. 3A) and per month for the figure on the right (Fig. 3B). This figure shows us, taking into account the speed of occurrence of deaths, the probability of survival of children.

Concerning the probability of survival of children under 5 years, the evolution is more or less regular. Over the four years following the birth of the children, they have at each birthday almost the same probability of survival. This probability follows a steady trend during the first four months after childbirth before registering a drop in the fifth month. This month represents the modal class of children under the age of one who record a rapid but not regular fall. Beyond this fifth month, the function of survival continues its regular fall until the 11th month.

The econometric estimates made it possible to highlight, among the variables that were integrated into our model, the factors that explain infant and child mortality (Table 3). These show that the coefficients are mostly significant at the 1%, 5% and 10% threshold. Differences in the wealth held by households affect the death of children under five. We based our interpretation on the distribution of wealth by using the deciles. this distribution goes from the poorest 10% to the richest 10% through the median. Our results suggest that when a household is in the median wealth decile, the risk that a child in that household die before reaching their fifth birthday decreases by 4.29%. it is the same with children living in 10% of wealthiest household. These children’s chances of surviving are 14.41 higher than those of the poorest 10%. This chance is 33.41% for those under one year old. These results are consistent with those of [7, 29]. This can be explained by the fact that wealth enters into the acquisition of health capital which improves with the purchase of adequate health services. Recently in 2019, nearly half of the Togolese population lived below the poverty line (i.e., 45.5% of Togolese) (INSEED, 2020). Those living on less than $1.9 a day will not be able to truly acquire the health capital necessary for their children, unlike those who live with a much higher level of wealth. The health capital that deteriorates over time with age will quickly depreciate and cause poor health in children and consequently death before they can celebrate their fifth birthday.

The analysis of our results also shows that the risk for a boy to die during the five years following childbirth decreases by 14.2 than that of a girl.

The level of education contributes to infant and child health. According to the results of the regression, the risk of infant and child mortality is lower when the mother of the child has a primary or secondary level of education and more compared to those who have no level of education. That said, children born to mothers with at least secondary level of education are 9.6% less likely to die than those born to mothers with no schooling at all. This result is similar to some works by [30]. In our context, the springs find foundation in the fact that the education of an individual has an advantage which is much more economic. Education leads to greater awareness of health rules,

Among these economic benefits of education is insurance coverage. Our results show that the risk of infant and child mortality decreases with the possession of health insurance. A child born to a woman who has health insurance has 35% less likely to die than those colleagues whose mothers have no health insurance. This result is consistent with those of [31] in Rwanda.

The results obtained through the estimation of the effect of water suggest that using drinking water in the concession does not reduce the risk of infant and child mortality. This result is counterintuitive and does not corroborate with the results found by [31] for whom water, sanitation and even hand washing reduce the risk of death in Ethiopia. This is a result that may be due to the non-representativeness of the individuals in our analysis who use spring water in both urban and rural areas (24.6%). In contrast to the effect of water on the health of children under five, the Model 2 estimate proves that using clean water in the home for basic needs helps reduce the risk of mortality children under one year old. Our analysis shows that the chances of a child surviving beyond eleven months postpartum are 0.0429 higher than those using water which is not safe to drink [31, 32]. With the advancement of the rules of medicine, the seventh month of the link between the type of water and the seventh month of the child explains this importance insofar as this child will have to start drinking water, which affects his organization directly.

Depending on the size of the household, the risk of infant and child mortality increases proportionally. We also find that a child born in a household where there are more than four children in charge has 4,36% more likely to die before celebrating his fifth birthday than those with a maximum of four children. This is explained by the fact that the importance that should be given to each child decreases as they become numerous; which is consistent with the results of the work of [33, 34].

The changing world of technology is also helping to improve child health. Our study shows that new information and communication technologies such as the use of smartphones or tablets by one of the parents improves the health of the latter. When a parent has a mobile phone or tablet, children have 0.7% more likely to be able to celebrate their fifth birthday more than those whose parents do not have such a digital tool. These tools allow appointments to be made, for example for vaccinations and reminders, as we have seen with the booster doses against the Covid-19 pandemic in Togo and everywhere else.

Our results also prove that having already procreated and made a prenatal visit reduces the risk of infant and child mortality. The risk that a child who is less than five years old dies is 0.021 lower when the mother has previously had at least one child. This result is nothing more than the experience that mothers acquire before, during and after previous deliveries to better take care of future births. This is felt much more through the domestic assistance of mothers to their daughters ("young mothers") which are in need at the first Birth. From the second childbirth, the mother-in-law or the mother of the one who gave birth may no longer be present because of the experience already acquired by the latter. Also having followed the four prenatal visits of the WHO during pregnancy has advantages for the child under five years of age. These has 10% less likely to not celebrate their fifth birthday alive when his mother has made more than four visits than their counterpart whose mothers have made less than four visits. This result is consistent with those of Ahmed and al., 2010 in developing countries.

Place of residence being one of the controversial factors in the two models, we determine through a third model the effect that an interaction of wealth inequality and place of residence may have on infant and child mortality.

The estimation of the interaction model is presented in Table 4. This part consists of highlighting the effects of wealth inequality while taking into account the place of residence of people. We add in this section an aspect of African society particularly Togolese, that of the treatment accorded to twin children. Our estimate shows that wealth inequality has a different influence on infant and child mortality depending on whether one living in an urban or rural environment. The results obtained following this estimation prove that when a child is born in median 10% wealthy households in urban areas, the probability to die before celebrating their fifth birthday decreases by 7.3 percentage points compared to the poorest in rural areas. This probability is lower for children living in top 10% wealthiest household in urban areas. For children in this category, the probability to die before celebrate their fifth anniversary decreases by 8,1 percentage points than the poorest 10% in rural areas. These results seem similar to those of [35] and contradict the findings of [36]. They find no evidence of a statistically significant difference in the risk of infant and child mortality between households in the poorest and poorest wealth quintiles in urban areas and their rural counterparts in Sub-Saharan Africa. Our result can be explained in this context by the fact that in urban areas, there’s easy access to the medical technical platform necessary for emergency intervention in the event of a child's illness, easy mobility due to available infrastructure, reduced proximity to health centers which all contribute to reducing the risk of infant and child mortality. Besides this result, the estimation of this third model gives credence to the two previous ones with regard to the control variables used in the analysis. Among other things, it emerges from this estimate that the probability that a child dies before celebrating its fifth birthday decreases with the level of education of one of its parents. The probability for a child born to a mother with primary education to die before 5 years decrease by 3.9 percentage points than those with mothers not educated. The education of the mother makes it possible to better control the rules of hygiene, to better apply them, to follow the indications of the prescriber of drugs for a better health of the child. Similarly, factors such as a parent's health insurance and prenatal visits reduce the risk of infant and child mortality. These factors contribute to reduce the probability of infant and child mortality respectively by 30,86 percentage points and 1.87 percentage points more than those who do not have health insurance and those who make less than four prenatal visits during their pregnancy. The prenatal visits allow the mother to collect a certain number of essential essentials for the survival of the child as well as for her own health. This estimate also shows that children whose parents are Muslims or animists run less risk of dying during the 4 years following childbirth. The probability for a Muslim child to die before celebrating his fifth birthday is 15.79 percentage points lower than for Christians. Similarly, practicing a traditional religion or being an animist seems to reduce the risk of infant and child mortality in Togo by 21.1 percentage points [37, 38]. The cultural attachment, the bad interpretation of the biblical writing by the Christians can explain the increase of the risk these last make run with their children.

These aforementioned factors are for the most part of the order of humans who can, because of their choices, ask for less or more to contribute to the increase in the health capital of the child. Genetic factors that escape humans cannot be captured in the same way. So, we take into account the fact of being born alone or twins. Our result shows that being born twins increases the risk of infant and child mortality compared to children born alone. A child who is born twin has 14.07 percentage points more like than their counterparts who are born alone to die before their fifth birthday. This result seems to agree with that of [39]. Having twins requires a lot more effort to care for them, a lot more attention to them, and even minimal neglect is fatal for these children. In African traditions particularly in Togo, this can also be explained by the fact that the death of one of the twins leads to that of the other provided that he does not see him or is aware of it.