Identifying Profiles of Support for Legal Abortion Services in Zambia: A Latent Class Analysis

Abstract

Relative to neighboring countries, Zambia has among the most progressive abortion policies, but numerous sociopolitical constraints inhibit knowledge of pregnancy termination rights and access to safe abortion services. Multistage cluster sampling was used to randomly select 1,486 women aged 15-44 years from households in three provinces. We used latent class analysis (LCA) to partition women into discrete groups based on patterns of endorsed support for legalized abortion on six socioeconomic and health conditions. Predictors of probabilistic membership in latent profiles of support for legal abortion services were identified through mixture modeling. A three-class solution of support patterns for legal abortion services emerged from LCA: (1) legal abortion opponents (∼58 percent) opposed legal abortion across scenarios; (2) legal abortion advocates (∼23 percent) universally endorsed legal protections for abortion care; and (3) conditional supporters of legal abortion (∼19 percent) only supported legal abortion in circumstances where the pregnancy threatened the fetus or mother. Advocates and Conditional supporters reported higher exposure to family planning messages compared to opponents. Relative to opponents, advocates were more educated, and Conditional supporters were wealthier. Findings reveal that attitudes towards abortion in Zambia are not monolithic, but women with access to financial/social assets exhibited more receptive attitudes towards legal abortion.

Keywords: abortion laws; latent variable modeling; population‐based study; sub‐Saharan Africa; termination of pregnancy.

FIGURE 1

Weighted distribution of support for legal abortion services in Zambia, by socio-demographics (N = 1,486).

FIGURE 2

Conditional item-response probabilities of support for legal abortion services, by posterior probability of latent class membership (N = 1,486).