mcmsupply: An R Package for Estimating Contraceptive Method Market Supply Shares

TL;DR

The mcmsupply R package provides Bayesian hierarchical models to estimate and project contraceptive market supply shares at national and subnational levels, incorporating diverse datasets and offering uncertainty quantification.

Contribution

It introduces the first annual estimates and projections of contraceptive market supply shares with uncertainty at multiple administrative levels using Bayesian models.

Findings

First to produce annual supply share estimates and projections with uncertainty.

Utilizes diverse datasets from IPUMS and DHS microdata.

Supports analysis of family planning market stability.

Abstract

In this paper, we introduce the R package mcmsupply which implements Bayesian hierarchical models for estimating and projecting modern contraceptive market supply shares over time. The package implements four model types. These models vary by the administration level of their outcome estimates (national or subnational estimates) and dataset type utilised in the estimation (multi-country or single-country contraceptive market supply datasets). mcmsupply contains a compilation of national and subnational level contraceptive source datasets, generated by IPUMS and Demographic and Health Survey microdata. We describe the functions that implement the models through practical examples. The annual estimates and projections with uncertainty of the contraceptive market supply, produced by mcmcsupply at a national and subnational level, are the first of their kind. These estimates and projections have diverse applications, including acting as an indicator of family planning market stability over time and being utilised in the calculation of estimates of modern contraceptive use.