Dynamic linear models for policy monitoring. The case of maternal and neonatal mortality in Ghana. (October 2022)

• Record Type: Journal Article
• Authors: Adebanji, Atinuke; Rios Insua, David; Ruggeri, Fabrizio
• Is Part Of: Socio-economic planning sciences. Number 83(2022)
• Publication Date: 2022-10
• Subjects: Policy monitoring -- Bayesian forecasting -- Dynamic linear models -- Millennium development goals -- Maternal mortality -- Neonatal mortality; Planning -- Periodicals; Economic policy -- Periodicals; Social policy -- Periodicals; Planification -- Périodiques; Politique économique -- Périodiques; Politique sociale -- Périodiques; ECONOMIC PLANNING; SOCIAL PLANNING; DECISION-MAKING; 361
• DOI: 10.1016/j.seps.2022.101348
• Languages: English
• ISSNs: 0038-0121

Abstract

Abstract: Monitoring is a major step in policy analysis used to assess whether a policy is actually working as desired. We provide a general policy monitoring approach based on Bayesian forecasting models. These are employed to predict the evolution of relevant monitoring variables over time and support expected utility calculations to assess the efficiency of the policy. We illustrate the approach by monitoring the Free Maternal Health Care and MDG5 Acceleration Framework policies aimed to reduce maternal and neonatal mortality in Ghana, using dynamic linear models for forecasting purposes. Despite major investments, results at national level suggest no significant improvement in maternal and neonatal survival between pre- and post-policy periods. However, regional analyses show that gains have actually been attained in certain regions, suggesting possible directions for improvements nationwide. Highlights: General policy monitoring approach based on Bayesian forecasting models. Monitoring of policies aimed to reduce maternal and neonatal mortality in Ghana. No significant improvement at national level despite major investments. Directions for improvements based on gains attained in some regions.