Page 172 - GHES-3-3

Global Health Economics and
Sustainability
Path model of child marriage in Africa

in 54 African countries. In this study, we examined the and females aged 15 – 24 who are affected by HIV (World
selected hypothesized socioeconomic and reproductive Bank, 2014; World Bank, 2014).
determinants including poverty, economy, literacy rates In 2014, adolescent fertility referred to the number
for youth males and females, HIV prevalence for youth of women per 1,000 who gave birth between the ages of
males and females, adolescent fertility, and maternal 15 and 19. The United Nations Population Division and
mortality to child marriage. The data used in the World Population Prospects compiled the rates procured
analysis were sourced from the World Bank, UNICEF, from the World Bank. Adolescent fertility rates were
and various United Nations agencies, which are widely derived from registered live births in vital registration
regarded as the most reliable sources for the variables systems, or, in their absence, from censuses or sample
under investigation. surveys. Estimated rates are generally thought to be

2.2. Variables and measures reliable indicators of past fertility. In cases where age-
specific fertility rates are unknown, a model was utilized
The data for child marriage in this study refers to the to estimate the proportion of births by adolescents. For
proportion of females aged 20 – 24 who got married or countries lacking vital registration systems, fertility rates
entered into a union before turning 18 between 2005 and are typically extrapolated from trends noted in earlier
2013. UNICEF compiled the indicators for child marriage census or survey data (World Bank, 2014).
based on surveys such as the Demographic and Health
Survey (DHS), Multiple Indicator Cluster Survey (MICS), The World Bank provided the maternal mortality ratio
and other national household surveys (UNICEF, 2014). data for 2013, which was collected by the World Health
Organization (WHO), UNICEF, United Nations Population
The data for the poverty variable in 2014 were sourced Fund, and the World Bank. This data encompassed trends
from the World Bank’s International Comparison in maternal mortality from 1990 to 2013, aligning with
Program, the World Development Indicators database, the Millennium Development Goals of 2015, which have
and the Eurostat Organization for Economic Co-operation since been replaced by the Sustainable Development
and Development (OECD) purchasing power parity Goals aimed at reducing the maternal mortality ratio. The
(PPP) program. This variable is calculated using the gross maternal mortality ratio is determined as the number of
national income (GNI per capita) adjusted for PPP. GNI women who die from pregnancy-related causes while
is transformed into international dollars using PPP as the pregnant or within 42 days of pregnancy termination per
conversion rate. GNI represents the total value added by 100,000 live births (World Bank, 2013).
all resident producers along with any product taxes not
accounted for in the output valuation, and net receipts of 2.3. Statistical analysis
primary income from abroad (World Bank, 2014). The Statistical Packages for Social Sciences was used to
The data on the economy were obtained from the 2014 conduct descriptive statistics, Spearman rho correlation,
national accounts data files of the World Bank and the and stepwise linear regression. In the descriptive statistics,
OECD. It was determined by dividing the gross domestic each variable was examined to determine the mean,
product (GDP per capita) by the mid-year population and minimum, maximum, and standard deviation. Spearman
was presented in United States dollars. GDP is the total rho correlation was used to identify the nature of the
of the gross value added by all resident producers in the bivariate correlation between the variables. Stepwise linear
economy plus any product taxes minus any subsidies not regression was employed to establish a linear equation that
included in the value of the product. This calculation does represents the relationship between the variables. Finally,
not account for the depreciation of fabricated assets or the EQS was used to conduct a path analysis to assess how well
depletion and degradation of natural resources (World the data fits the model and determine the coefficients of the
Bank, 2014). significant predictors for the outcome variables.
The 2015 youth literacy rates were obtained from 3. Results
the United Nations Educational Scientific and Cultural
Organization (UNESCO). These rates refer to the 3.1. Variable characteristics
proportion of males and females between 15 and 24 years Table 1 presents the mean, minimum, maximum, standard
old who possess the ability to read, write, and comprehend deviation, and sample size for both of the outcome variables
a brief statement about their daily life (UNESCO Institute and the possible determinants. This table shows that on
of Statistics, 2015). The 2014 HIV prevalence data were average, 32.9% (SD 17.2%) of child marriages happened
sourced from the World Bank, based on estimates from between the years 2005 and 2013. In addition, the region
UNAIDS. This variable represents the percentage of males that had a maximum of 76% of child marriages occurred

Volume 3 Issue 3 (2025) 164 https://doi.org/10.36922/ghes.7117