Elsevier

Social Science & Medicine

Volume 59, Issue 12, December 2004, Pages 2603-2616
Social Science & Medicine

Migration, community context, and child immunization in Ethiopia

https://doi.org/10.1016/j.socscimed.2004.04.009Get rights and content

Abstract

This paper examines the relationship between parental migration status and child immunization in Southern Ethiopia, a region characterized by high mortality and morbidity. Using the 1997 Community and Family Survey and a multilevel modeling approach, we find that children born to rural–rural migrant mothers have significantly less chance of receiving full immunization coverage than children born to non-migrant mothers. The social mechanism that explains this huge disparity is that rural–rural migrant women have limited social networks in the host community. In addition, significant variation in receiving complete immunization is found by age of child (a likely period effect), mother's education, and distance to nearest health center. Marked child immunization differentials are also observed by ethnicity. The results from the multilevel analysis confirm the persistence of substantial community effects, even after controlling for a standard array of personal and household characteristics. Given the low levels of vaccination among children born to migrant women, health policy interventions and information campaigns might be effectively augmented to reach such migrant women and their children. Community and ethnic group effects suggest that further targeting of health activities could be efficient and effective.

Introduction

Infant and child mortality rates in many parts of Africa have been decreasing in recent decades, but still remain among the highest in the world (Hill & Pebley, 1989; Hill, 1992; Delaunay, Etard, Preziosi, Marra, & Simondon, 2001). Numerous studies have demonstrated the considerable importance of preventive health measures such as vaccination in reducing child mortality, morbidity, and disability (Feachem & Jamison, 1991; Kaute Defo, 1994; Foster, 1995; Brockerhoff & DeRose, 1996; Nyarko, Pence, & Debpuur, 2001). Indeed, vaccines have significantly reduced or, in some areas, eliminated diseases such as smallpox. The six diseases (measles, pertussis, diphtheria, tuberculosis, tetanus, poliomyelitis) of the expanded program of immunization (EPI) continue to seriously effect morbidity and mortality in Africa (Feachem & Jamison, 1991), and Ethiopia is no exception. UNICEF estimates that worldwide, these infectious diseases kill 10 children with each passing minute and disable 10 more. In Ethiopia, in the early 1990s, about 300,000 children died per year from immunizable diseases and diarrhea (Lidetu & Okubazgi, 1993). Providing full immunization coverage to substantially reduce infant and child mortality is also a high-priority national public health objective in Ethiopia (Office of the Prime Minister, 1993). In order to most efficiently allocate scarce public health resources, it is crucially important to identify groups with low vaccination coverage and uncover the behavioral process of traits associated with low immunization.

Migration is likely to be implicated in the coverage and success of immunization. Migration is a salient feature of the social life of developing countries (Zarate, 1983; Evans, 1987), and a variety of political and economic forces have operated to redistribute population in Ethiopia over the last several years (Berhanu & White, 2000). Still, the role of population redistribution in health outcomes is not well understood. On the one hand, migrants are acknowledged to be new—and often less integrated—residents. On the other hand, migrants are generally highly motivated individuals, even innovators. Perhaps this motivation or receptivity to change may counteract the lack of social integration in the new community. We also consider that that the migration of women due to marriage may be implicated. It is often the case that neighbors, friends, or relatives exchange information regarding their experiences with health workers and their children's vaccination. In the case of marriage migration, disruption of these social ties may reduce access to such information for women who do not reside in their natal community. This paper investigates the empirical evidence on the nature of the relationship between migration and health.

The investigation of barriers to full immunization has important policy implications. Indeed, examining how migrants adapt in terms of their health seeking behavior as they move among various social settings provides a good opportunity to gain insight into the roles of environment and culture in shaping human behavior. The redistribution of population affects the migrants themselves, their children, and the origin and destination communities, as well. For example, children may be particularly vulnerable to disruptions caused by migration, yet our knowledge in this area is limited. Research assessing the impact of migration on child immunization is woefully inadequate.

In recent studies, researchers have established a link between migration and child survival, and they have documented the health advantage of children living in urban areas compared to rural areas (Bender, Rivera & Madonna (1990), Brockerhoff (1995); Bender, Rivera, & Madonna, 1993). These studies, however, focus on rural–urban migration and have ignored rural–rural migration. In sub-Saharan Africa, rural–rural migration constitutes the largest share of total internal migration (Oucho & Gould, 1993). Further, these studies have not made a link to child immunization. Hence, empirical evidence on the relationship between parental migration status and child immunization is very limited.

This study uses a multilevel framework to examine the effect of a set of individual and community traits on child immunization, with a special focus on the influence of parental migration. We examine this association through the analysis of the 1997 Community and Family Survey (CFS) data collected in the Southern Nations, Nationalities and Peoples Region (SNNPR) of Ethiopia. This multilevel approach allows us to simultaneously account for individual, household, and community-level influences on the chance that a child is immunized.

Section snippets

Background

Ethiopia, with 65 million people, is the second most populous country in sub-Saharan Africa. Infant and child mortality rates in Ethiopia remain at very high levels, even relative to many other countries in sub-Saharan Africa. According to the recent 2000 Ethiopia Demographic and Health Survey (Central Statistical Authority & Macro International, 2001), infant mortality in 2000 was 97 deaths per 1000 live births, while the child mortality rate was 166 deaths per 1000. By contrast the infant

The study area

The Southern Region of Ethiopia has a land area of 117,506 km2 and a population of 12 million (CSA, 1998). The region accounts for about 10% of Ethiopia's total land area and approximately 20% of its population. The region has the highest population density in all of Ethiopia with an average of about 95 persons/km2, while, at the same time, being one of the least urbanized regions of the country with only 7% of the population living in urban areas (CSA, 1998). The region's peasants rely heavily

Social mechanisms

The influence of women's social networks ties or networks on child health in the developing context has recently gained attention (Castle, 1994; Adams, Madhavan, & Simon, 2002). Collective notions about child vaccinations could emerge when members of social networks, relatives, or neighbors exchange ideas and information about the quality of health services provided in the community, treatment of health workers, side-effects associated with vaccinations, and their experiences related to

Data

The data used for this analysis come from the 1997 CFS conducted by the Demographic Training and Research Center of Addis Ababa University and the Population Studies and Training Center of Brown University. The sampling design combined stratified sampling with simple random sampling proportional to population size. The CFS was conducted between May and July 1997. Five zones (Gurage, Hadiya, KAT, North Omo, and Sidama) in the Southern Region were covered by the survey. The 1994 Census found that

Complete immunization according to parental migration

We ascertain migration status of adults by asking adults questions about whether they were born in the community of current residence, and if not, where they were born. Migrants are those residing outside their community of birth. We also ascertain duration of residence in the community in years and months. Of all residents aged 10 years and older living in the study communities, 15.3% of males and 29.8% of females are migrants. In order to get some insight about the motives for migration,

A multilevel analysis

We employ multilevel logistic regression models to examine the effect of parental migration on child immunization, while also identifying factors that influence child immunization at other levels of aggregation. In the 1997 CFS, the data have a three-level structure—children (level 1), within families (level 2), within communities (level 3). The multilevel structure of the data used in this analysis is displayed in Fig. 1.

In addition to the observed covariates considered at each of these three

Conclusion

In this study we have examined the impact of parental migration on immunization of children in a country where war, famine, and economic deterioration have influenced both health resources and population distribution in the last three decades. We used a multilevel approach to examine determinants of immunization among children born between 1991 and 1996 and of ages between 12 and 59 months. There have been repeated international and national exhortations to improve child immunization, yet

References (52)

  • Brockerhoff, M., & Hewett, P. (1998). Ethnicity and child mortality in sub-Saharan Africa. Population Council Working...
  • J.C. Caldwell

    Education as a factor in mortality declineAn examination of Nigerian data

    Population Studies

    (1979)
  • J.C. Caldwell

    How is greater maternal education translated into lower child mortality?

    Health Transition Review

    (1994)
  • S. Castle

    The (re)negotiation of illness diagnoses and responsibility for child health in rural Mali

    Medical Anthropology Quarterly

    (1994)
  • Central Statistical Authority (CSA). (1992). The analysis of local moves and internal migration, Vol. II. Addis Ababa,...
  • Central Statistical Authority. (1993). The 1990 National Family and Fertility Survey, Addis Ababa,...
  • Central Statistical Authority (CSA). (1998). Analytical report of the 1997 population and housing census of Ethiopia....
  • Central Statistical Authority & Macro International Inc. (2001). Ethiopia demographic and health survey. Addis Ababa:...
  • L.C. Chen et al.

    Sex bias in the family allocation of food and health care in rural Bangladesh

    Population and Development Review

    (1981)
  • M. Das Gupta

    Death clustering, mother's education and the determinants of child mortality in rural Punjab, India

    Population Studies

    (1990)
  • V. Delaunay et al.

    Decline of infant and child mortality rates in rural Senegal over a 37-year period (1963–1999)

    International Journal of Epidemiology

    (2001)
  • S. D’Souza et al.

    Sex differentials in mortality in rural Bangladesh

    Population and Development Review

    (1980)
  • DTRC/PSTC. (1998). The 1997 Community and Family Survey. Demographic Training and Research Center, Institute of...
  • J. Evans

    IntroductionMigration and health

    International Migration Review

    (1987)
  • R. Feachem et al.

    Disease and mortality in sub-Saharan Africa

    (1991)
  • Foster, S. O. (1995). Potential health impact of immunization. In H. Rashad, R. Gray, & T. Boerma (Eds.), Evaluation of...
  • Cited by (89)

    • Childhood vaccination timeliness following maternal migration to an informal urban settlement in Kenya

      2022, Vaccine
      Citation Excerpt :

      Scholarly interest in investigating the root cause of health inequities, including disparities in timely vaccine receipt, is growing [10]. Given its relationship to vaccination coverage in some African nations [14,15] maternal migration may be a significant predictor of timeliness disparities, though this relationship has not been explored in sub-Saharan Africa or other regions. Migration may influence vaccination timeliness via disruption of a woman’s social, instrumental, and economic resources.

    • The influence of maternal migration on child vaccination in Kenya: An inverse probability of treatment-weighted analysis

      2021, International Journal of Infectious Diseases
      Citation Excerpt :

      Extended family members may provide emotional, financial and logistical support crucial to child health that may be compromised by distance and isolation from family (Clark et al., 2018; Madhavan et al., 2018). This disruption may directly influence vaccination by impeding maternal knowledge of vaccination programming or the ability to finance indirect costs (e.g., transportation), resulting in decreased vaccination of migrant children compared to their non-migrant peers (Kiros and White, 2004; Kusuma et al., 2010). Migrants also require time to adapt to their new environment (e.g., integrate into the community, find employment) before they can effectively find and use local healthcare (Smith-Greenaway and Madhavan, 2015).

    • Internal migration and child health in Malawi

      2019, Social Science and Medicine
    View all citing articles on Scopus
    View full text