Elsevier

Social Science & Medicine

Volume 102, February 2014, Pages 41-48
Social Science & Medicine

Healthy migrant and salmon bias hypotheses: A study of health and internal migration in China

https://doi.org/10.1016/j.socscimed.2013.11.040Get rights and content

Highlights

  • We use panel data to study the healthy migrant and salmon bias hypotheses.

  • We find support for both hypotheses in the China internal migration setting.

  • Healthier people are more likely to migrate and to move further away from home.

  • Migrants with poorer health are more likely to return or to move closer to home.

Abstract

The existing literature has often underscored the “healthy migrant” effect and the “salmon bias” in understanding the health of migrants. Nevertheless, direct evidence for these two hypotheses, particularly the “salmon bias,” is limited. Using data from a national longitudinal survey conducted between 2003 and 2007 in China, we provide tests of these hypotheses in the case of internal migration in China. To examine the healthy migrant effect, we study how pre-migration self-reported health is associated with an individual's decision to migrate and the distance of migration. To test the salmon bias hypothesis, we compare the self-reported health of migrants who stay in destinations and who return or move closer to home villages. The results provide support for both hypotheses. Specifically, healthier individuals are more likely to migrate and to move further away from home. Among migrants, those with poorer health are more likely to return or to move closer to their origin communities.

Introduction

The relationship between migration and health is an important topic of investigation for two reasons. First, it offers a better understanding of various factors that can play a part in the migration decision-making process. Second, it enhances our knowledge about the health profiles of the migrant population and how it may change over the course of migration and adjustment. Previous studies on migration and health, almost all conducted in the context of international migration to the U.S., have documented a clear immigrant health paradox: despite having low levels of socioeconomic status and limited access to health services, immigrants in the United States, especially Hispanic immigrants, generally enjoy better health outcomes than the native population in terms of mortality rate, chronic conditions, and mental health (Halliday and Kimmitt, 2008, Hayward and Heron, 1999, Markides and Eschbach, 2011, Riosmena and Dennis, 2012, Rubalcava et al., 2008, Villa et al., 2012). This health advantage is especially evident in the early stages of migration, though it tends to diminish over time in tandem with acculturation (Antecol and Bedard, 2006, Castro, 2007, Smith, 2003).

Two explanations are commonly offered to understand this health paradox. The first is the “healthy migrant” hypothesis, which states that migrants represent a positively selected group of individuals with respect to health, relative to the general population in origin societies (Chiswick et al., 2008, Palloni and Morenoff, 2001). This selection process makes immigrants stand out in terms of health levels when compared with the general population in destination countries. This positive selection even holds for migrant populations from relatively under developed origins experiencing high rates of mortality and morbidity (Abraído-Lanza et al., 1999, Markides and Eschbach, 2011).

In order to endure the demanding journey, migrants often require considerable physical stamina. Moreover, migration causes disruption in an individual's life and requires adaptation to a new environment and lifestyle. The process of migration is so stressful and challenging that an individual with relatively poor health is less likely to undertake migration. In addition, migrants often have physically demanding jobs. This means that those who can endure such jobs and succeed economically are more likely to migrate than others. The theory of the New Economics of Labor Migration (Stark, 1984) suggests that migration is a household survival strategy. Within this framework, migrants are likely to be selected within the household among eligible members—that is, those who are more capable, including having good health status, are more likely to migrate because of a higher propensity of achieving economic success. Formal health screening, especially in the context of international migration, may further select healthy people. This holds particularly true with respect to infectious diseases (Chiswick & Miller, 2008).

Facilitated by the availability of suitable data on migrants and the origin population (including pre-migration characteristics), there has been a growth of systematic empirical research testing the healthy migrant hypothesis. Rubalcava et al. (2008) study the healthy migrant effect in Mexico-US migration by using nationally representative longitudinal data obtained from following a sample of the Mexican population over time (some migrants moved to the U.S. between waves of the survey). Though they find some evidence for the health selection of migrants, it seems to apply to females and rural males only and is quite modest in magnitude. In theory, international migration entails greater barriers to migration and more disruption of social networks than internal migration. This could lead to a higher degree of health selection in the process of international migration. Empirical research, however, does not necessarily support such a speculation.

Studies based on internal migration generally find support for the healthy migrant hypothesis, but they also show the complexity of the relationship. Using longitudinal data from Indonesia, Lu (2008) provides some support for the healthy migrant effect in the context of internal migration. The study further highlights the complex process of health selection, which varies across different dimensions of health as well as by age. Young people with chronic diseases are less likely to move. By contrast, older people with chronic diseases have a higher propensity to move. Presumably, this is because the young often migrate for work-related reasons. For older people, migration may be a response to health deterioration and be taken as a means of receiving better health care services or family care.

A few studies have examined the healthy migrant hypothesis in China. Chen (2011) uses survey data in Beijing that include a sample of migrants and urban residents, and finds the healthy migrant effect for migrants on the basis of self-rated physical health, but not psychological distress. Tong and Piotrowski (2012) utilize longitudinal data over a decade and contrast migrants and non-migrants at the origin. They show that migrants are positively selected for health, especially self-reported health. However, they find that the level of the impact of health on migration declined over time as migrants become less selective in general.

The second explanation for the better health of migrants is the salmon bias hypothesis, or selective return migration, which postulates that unhealthy migrants or migrants who experience deteriorating health have a greater tendency to return or to move closer to their origin communities than healthier migrants (Abraído-Lanza et al. 1999). Migrants facing health problems are less capable of achieving high productivity in destination labor markets. This could lead to decreasing earnings and a lowered standard of living. These factors, along with constraints in affording and effectively utilizing health care services, separation from family, lack of a supportive network, and various sources of work-related stress in the destination, make returning or moving closer to home a more attractive option to migrants.

Empirical investigation of the salmon bias hypothesis has been very scarce because it requires data tracking migrants who return and who stay. Ullmann, Goldman, and Massey (2011) provide some useful findings in their study, which compares returned migrants and non-migrants in Mexico while controlling for their early-life health profiles. It shows worse health for return migrants compared to non-migrants in terms of obesity, smoking, and psychological health. Turra and Elo (2008) present one of the first direct assessments of salmon bias in Hispanic migration to the U.S., and find very modest evidence for the hypothesis. Their study also calls for the use of longitudinal surveys that follow individuals in and out of migrant destinations for a better evaluation of selective return processes. By contrast, Sander (2007) examines return migration from Germany to Turkey and suggests that for male migrants, those with poorer health are less likely to return than those with better health. This observation is used to explain the declining health of immigrants in destination societies.

Examining the presence and magnitude of both selective migration and selective return migration is a crucial step in fully understanding the health consequences of migration. Research on the effects of migration on health is potentially plagued by these processes of selection. Observed effects are not necessarily evidence of the consequences of migration, but may arise from the selective feature of migrants on health. For example, if migrants are favorably selected on health, one may conclude with a positive migration effect on health even when the health of migrants remains or slightly deteriorates after migration. Additionally, if return migrants are negatively selected with respect to health, migrants who remain in the destinations will appear healthier than the destination population. This could lead us to reach a spurious conclusion of a positive effect of migration on health.

The present research seeks to examine both questions in the context of internal migration in China using longitudinal data. Since the early 1980s, an estimated 220 million migrants have moved from rural areas to work in Chinese cities (National Bureau of Statistics [of China], 2011). At the same time, the existence of a long-standing bifurcated social institution—the household registration [hukou] system, which categorizes all citizens into a rural or urban dichotomy—has resulted in various structural and social barriers that preclude migrants from becoming full urban citizens. There is thus limited access to desirable jobs and limited provision of social services for migrants and their families, including health care and education (Hesketh et al., 2008, Solinger, 1999). Whereas most of the existing work is based on international migration to the U.S., we expect a positive health selectivity to be present in the context of internal migration in China because of its unique hukou system that has resulted in migrants' precarious conditions. Indeed, previous research has drawn parallel between the conditions of internal migrants in China and international migrants to the U.S. (Roberts, 1997). Therefore, health is likely to be a crucial factor in determining migrants' ability to move and make a living in Chinese cities. In the present study, we adopt an appropriate research design (health measure prior to migration or return) and the appropriate groups for comparison (non-migrants for the healthy migrant hypothesis, and migrants for the salmon bias hypothesis), and provide evidence for the health selection of migration and return migration in internal migration in China.

Section snippets

Data

Data used are from a national representative longitudinal survey spanning 2003–2007 (The China Rural Production Survey [CRPS]) conducted by the Ministry of Agriculture of China (Qin, Qin, & Jiang, 2012). The survey used multistage stratified sampling and covered 16 provinces (out of the 31 provinces in the mainland of China) representing different geographical regions and levels of economic development (based on per capita annual net income of rural residents). There were 5, 6 and 5 provinces

Descriptive statistics

The descriptive statistics for variables used in the analysis are shown in Table 1. On average, the sample consists of 29% migrants. About 56% of the sample report excellent health. The average household size is about 4. In terms of demographic characteristics, men account for 52% of the sample. Around 13% of the sample are between age 16 and 25, 19% are between age 26 and 35, 24% are between age 36 and 45, and 29% are between age 46 and 55. The 56–65 age group accounts for almost 16% of the

Discussion

Over the past several decades, rural-to-urban migration has become a prominent feature in China. Set against this background, the present study examines the presence of selective migration and selective return migration on the basis of health, using nationally representative longitudinal data. We employ measures of health status prior to migration or prior to return, and restrict the analyses to the appropriate sample. To test the healthy migrant hypothesis, we use the non-migrants sample (at

Acknowledgment

The authors would like to thanks to the support of the National Social Science Foundation of China (13CGL102), the Ministry of Education of China (12YJC790152), the National Post-doctoral Council of China (2012M510316) and the special support of the National Post-doctoral Council of China (2013T60058).

References (32)

  • T. Halliday et al.

    Selective migration and health in the USA

    Population Studies

    (2008)
  • M.D. Hayward et al.

    Racial inequality in active life among adult Americans

    Demography

    (1999)
  • T. Hesketh et al.

    Health status and access to health care of migrant workers in China

    Public Health Reports

    (2008)
  • E.L. Ider et al.

    Self-related health and mortality: a review of twenty-seven community studies

    Journal of Health and Social Behavior

    (1997)
  • G. Li et al.

    The application and research of mobile phone on solving rural migrant worker related problems

    Symposium of the Fifth China Digital City Construction

    (2011)
  • K. Markides et al.

    Hispanic paradox in adult mortality in the United States

    International Handbooks of Population

    (2011)
  • Cited by (0)

    View full text