In many countries, age dependency has been on the rise placing pressures on the working age populations to provide for the socio-economic needs of the young (ages 0–16) and elderly (ages 65+). This dissertation attempts to address the policy problems generated by age dependency problems using predictive analytic methods.
The first paper micro-simulates the effects of a cash transfer program (both conditional and unconditional) of the same individual amount on educational enrollment in basic education for children aged (7–15) and the intensity of poverty for Nigeria. This study shows that increasing program coverage and providing smaller conditional transfers per child in school is more effective at increasing educational enrollment more than increasing the transfer amount. The second study projects the demand for dementia care using a novel microsimulation approach in the United States. The results suggest that the age distribution of the population is and will continue to be an important factor driving dementia prevalence. The final study uses a machine algorithm to predict dementia status across individuals without using any measures of cognition. We show that a modelling solution like this may be used as a first step to eliminate individuals that might not need the more traditional brain scans that are usually more expensive. Therefore, there may be cost savings in the future attributed to this approach but further methodological research is needed.
Governments in both the Western countries and Africa have significant upcoming challenges to deal with as age dynamics evolve over the next 50 years in their respective countries. As cash transfer programs that seek to increase educational enrollment become more prevalent, African governments need to define clearer child labor policy. African development, in part, is dependent on the ability of policymakers to be create the environment for academic innovation as there are no clear cut solutions to deal with the large proportion of young dependants that require education but come from poor households. In the countries with old age dependency, policymakers will have to grapple with the rising prevalence of age-old diseases amongst the overall population. This dissertation is an attempt to begin supporting this line research.
Table of Contents
Micro-simulating the Education Enrollment Effects of Conditional and Unconditional Cash Transfers in Nigeria
Predicting the Demand for Dementia Care: Simulating Dementia Incidence in the United States
Towards a Simple Machine Learning Tool for Predicting Dementia Status