Migration, health status and utilization of health services.

Evidence concerning the relations of migration to health status and health care utilization is inconclusive. This paper outlines the theoretical positions on these relations and reviews the empirical findings supporting varying positions. The paper also presents the findings of a survey of a probability sample of the U.S. population concerning these issues. The analysis is organized around comparisons among immigrants from other societies, internal migrants, and nonmigrants on important dimensions of health status and the utilization of different types of health services. Comparisons were also made among first generation, second generation, and `old stock.' Finally, immigrants were grouped by country and area of origin to assess the influence of cultural differences on health and the utilization of services. Attempts were made to control the influence of demographic characteristics while assessing the relations between migration and health status. The controlled analysis showed immigrants to enjoy better health conditions followed by migrants and then the nonmigrants. Significant differences in physical performance were manifested among immigrants from varying areas of origin, and also among generations of nativity. Controlling for both demographic characteristics and health status, immigrants were consistently the least utilizers of services and internal migrants the highest. An increase in utilization was associated with generational residence in the U.S. First generation were least utilizers, followed by second generation with the `old stock' being the greatest utilizers. Immigrants from different countries and regions of the world also exhibited Significant differences in the patterns of utilization of health services. The Research Problem Health is a revealing indicator of well-being and adjustment, and an important variable in migration research. The present large scale crossnational migration in Europe has attracted considerable public and * This research was in part supported by grants from the Rehabilitation Services Administration, and the Social Security Administration, U.S. Dept. of Health, Education and Welfare. Sociology of Health and Illness Vol. 2 No. 2 July 1980 R. K. P. 1980 0141-9889/80/0202-0174 $1.50/1 Migration, Health Status and Utilization of Health Services 175 research interests, especially in the health conditions of immigrant groups (Bagley, 1970; Castles and Kosack, 1973 ; Cochrane, 1971; Dalgard 1967; Haavio-Mannila and Stenius, 1974,1976, and 1977; and Invandrarutredningen 1974). Nevertheless, the long massive experience with both external and internal migration has made the United States the prime laboratory of the West in studies of the relationships between health and migration. The findings of studies on this topic have been often inconclusive or contradictory, changing over time, and primarily focused on mental illness. Concerning mental illness, several studies support the thesis of its high incidence among the immigrant population of the United States (Lazarus et. al., 1963; Lee, 1963; Locke et. al., 1960; Malzberg and Lee, 1956; Malzberg, 1962; Murphy, 1965; Srole, 1962; odegaard, 1932; and see Sanua 1969 for a detailed survey of the literature). Murphy (1965) attributes the greater incidence of mental illness he found to exist among immigrants to the U.S. and Australia, compared to the natives of the two countries, to a repressive assimilation policy and the the greater scattering of migrant groups in these two countries than in others he studied such as Canada, Israel, and Singapore. In these latter States, the rates of mental illness among immigrants were even lower than those among the indigenous populations. According to some reports, internal migrants in the U.S. have higher rates of hospitalization for mental illness than foreign-born and non-migrants (Lee, 1963; Malzberg and Lee, 1956; and Malzberg, 1962). While data from Norway (Dalgard, 1967) and from Britain (Rwegellera, 1970 and Bagley, 1970) lend support to the association between mental illness and geographic mobility, other studies in the U.S. tend to negate or weaken that thesis (Hollingshead and Redlich, 1958; Jaco, 1960; and Leighton, 1963). Particularly important are Jaco's (1960) findings of no differences between migrants and non-migrants in the incidence of schizophrenia. In studies of migration, physical health received far less attention than did mental health. In support of an association between migration and physical health problems are the findings of a controlled analysis of the Scandinavian Survey (Allardt, 1972) linking frequent migration to health problems measured by a composite index consisting of long-term illness, psychosomatic illness, symptoms of anxiety, and the use of medications. Compared on several health indicators, Finnish immigrants to Sweden were found to have poorer health than both native Swedes and Finns in Finland (Haavio-Mannila and Stenius, 1977). European migrant workers in France, Switzerland, Germany and Britain were reported to have higher rates of tuberculosis and venereal diseases than characteristic of natives (Castles and Kosack, 176 Nagi and Haavio-Mannila 1973: 325-326). Also, compared to native Swedes, foreign workers in Stockholm during the period 1967-72 had higher rates of absenteeism due to illness (Carlson and Taghavi, 1973). Differences in this latter study diminished as a result of controlling over occupational distributions. Evidence to the contrary comes in the form of relatively low absenteeism due to illness among foreign workers in West Germany (Castle and Kosack, 1973: 324) which may be due to selectivity in recruitment by the German Commissions abroad. Cochrane also has shown that, in times of unemployment in South Wales, it was the more physically fit who tended to seek work elsewhere (McDonnell, 1975: 35). In the U.S., in an analysis of the 1967 survey of economic opportunity, Wan and Tarver (1972) found only slight differences favoring migrants over non-migrants in the prevalence of work-limiting morbidity. However, migrants with a rural background had substantially higher rates than both non-migrants and urban migrants. Furthermore, considering the origin and destination, rural to rural migrants were in the poorest health conditions of all. Explanations offered for relations between migration and health have been systematized by Murphy (1966) who pointed out Odegaard's emphasis on `rootlessness', and the hypothesis of exploitation of migrants advanced by Miller and expounded by Faris, and the newer perspective calling for the specification of conditions under which migrants develop higher rates of illness. Murphy suggests three constellations of variables as links between mental illness and migration: (a) those related to the culture of origin, (b) circumstances of migration, and (c) factors related to the society of resettlement. Cochrane (1971) proposes differential selection in migration as an explanation for differences in the rates of admission to mental hospitals in England and Wales by place of birth (within and outside Britain). He concludes that `where migration is relatively easy the less stable members of a population self select for migration, but where migration is relatively difficult only the most stable individuals can achieve migration' (Cochrane, 1971: 11). The theme of adjustment to changing environments and the stresses engendered as the links between health problems and migration receive mixed support. On the one hand, an increased risk of coronary heart disease has been attributed to the condition where an `individual finds himself in a new social situation where he is unable to anticipate the social consequences of his behaviour. . moving from a world where the rules of the game are clear and known to a new world where different and unknown rules apply' (Benyoussef et. al., 1974). On the other hand, Choldin (1965: 172) sees migration now being perceived as a normal routine of life, and Burchinal and Bauder (1965: 198-199) Migration, Health Status and Utilization of Health Services 177 maintain that in the United States most individuals and families adapt reasonably well to new circumstances. These latter investigators see residential mobility to be perceived by people as means by which they improve their personal and family situations in life. The paucity of data on migration differentials in patterns of utilization of health care has already been mentioned. Stromberg (1974: 317) found that it takes some time for migrants in Iran to learn about services sufficiently to reach the same levels of utilization as nonmigrants. In Sweden, Finnish immigrants in a middle sized town learned to use health services fairly quickly (Haavio-Mannila, 1975). However, the Finns were more likely to use the public health services offered in hospitals and clinics compared to the Swedes who tended to rely more on private physicians. And, in an analysis of internal migration and access to medical care in Honduras, it was concluded that `while migrants as a group did have less access to medical care when ill, when socioeconomic status was controlled, the differences virtually disappeared' (Teller 1972: 192). In a recent critique of research on these topics, it was concluded that the `evidence supporting the viety that migration is either deleterious or beneficial to health, and results in an increase or reduction in utilization behavior, is fragmentary and equivocal' (McKinlay, 1975: 587). This paper is intended to expand available evidence and to further the specifications of health and utilization measures. It compares immigrants, internal migrants, and non-migrants in a probability sample of the United States population on two dimensions of health status (physical and emotional) and on patterns of health care utilization. Before turning to a description of the types and sources of data used in the analysis, we will first briefly review some important elements in the picture of immigration and migration in the U.S. Immigration and Migration in the U.S. Until 1890 the majority of the immigrants to the U.S. came from Western and Northern Europe; during the period 1891-1920, from Southern and Eastern Europe; and since 1921, from the Western Hemisphere (The Immigration Acts of 1924 and 1952 exempted immigrants from independent countries of the Americas from quota restrictions) (Eckerson, 1966: 6-11). In 1924, in order to preserve the traditional ethnic composition of the nation, quotas were established on the basis of national origin of white persons as recorded in the Census of 1920. Although these quotas authorized about 150,000 admissions annually, countries given large quotas, mainly those of 178 Nagi and Haavio-Mannila Northern and Western Europe which received 80% of the total, seldom used half of their quotas. Even in the years of highest immigration, only two-thirds of the quotas could be used. In addition to immigration from the Western Hemisphere, spouses and children of United States citizens, as well as some refugees and displaced persons were not barred from entry by the quota system. Thus, the selection of immigrants proportionately according to national origins was never successful (Eckerson, 1966). The immigration Act of 1965 abolished the national origins quotas providing instead a numerical annual limitation of 120,000 on Western Hemisphere immigration. For the other countries immigrant visas were limited to 170,000 exclusive of parents, spouses, and children of Untied States citizens and a ceiling of 20,000 visas for any one country. The 1965 Act also created preference categories for the purposes of family reunion, professional and skilled workers, and for refugees (Kennedy 1966: 148-149). Already in 1952 a special preference of 50% of each quota had been set aside for the skilled workers, whose services and education would be useful to the nation (Bennett, 1966: 131). Especially significant from the health point of view is the exclusion of physically or mentally ill persons from entry. The 1882 law already provided for the exclusion of persons who were ill , citing specifically persons with loathsome and contagious diseases or mental deficiencies, as well as criminals and paupers. During the depression years consular officers who issued visas interpreted the exclusion clause of the Immigration Act of 1917 `likely to become a public charge' with unusual strictness (Rubin, 1966: 17). The Immigration Act of 1965 states that the following classes shall be ineligible to receive visas and shall be excluded from admission into the United States: aliens who are mentally retarded, or insane, or who have had one or more attacks of insanity; who are afflicted with psychopathic personality, or sexual deviation, or mental defect; who are narcotic drug addicts or chronic alcoholics; who are afflicted with any dangerous contagious disease; and others certified by the examining surgeon as having physical defect, disease, or disability which may affect their ability to earn a living (Immigration and Nationality Act, 1965: 33-34). Currently, the Public Health Service inspects every entrant to the United States prior to the receipt of a visa, as well as at the port of entry, according to standards issued by the Department of Health, Education, and Welfare (Murphy and Blumenthal, 1966: 119). Since the 1940's internal migration in the United States has been fairly steady. Every year about 18-20% of the population changes residence, 6-7% migrating from one county to another, and some 3% over state borders (Bureau of the Census, Current Population Reports, 1971). The personal characteristics of migrants differ from those of a Migration, Health Status and Utilization of Health Services 179 typical cross-section of the population. For example, they comprise greater proportions of professional, technical, managerial, administrative, and kindred workers who are young, better educated, and highly mobile geographically (Bureau of the Census, Current Population Reports, 1974). Migrants are also occupationally more upwardly mobile than non-migrants (Blau and Duncan, 1967: 250-275), and are more likely to be white than non white (Wertheimer, 1970: 14). Nevertheless, the traditional image of migrants as special risk groups in regard to health persists. This is perhaps because a visible part of the migrant population, from the point of view of health policies and programs, comprises those who come from and go to poverty areas (Slavin, 1971: 193). Types and Sources of Data Data to be reported here were derived from a survey addressed to broader objectives. The survey was conducted in 1972 under the direction of the senior author as a Principal Investigator. Sampling operations and data gathering were carried out under contract by the University of Michigan's Survey Research Center. The survey included 8090 households constituting a probability sample of continental United States excluding Alaska. One person (18 years of age and over) in each household was selected at random for personal interviews. The survey yielded 6493 (80.3%) completed interviews; 92% of the respondents were either household heads or their spouses. Compared to persons 18 years of age and over reported in the 1970 U.S. Census, this sample includes 5.5% more females, 6% less white males, 5.7% less single persons, and 2.9% more persons 65 years of age and older. Reasons for the 19.7% non-completed interviews were: (a) refusals by designated respondents or on their behalf (11.2%), (b) no contacts made with any members of the household (2.8%), (c) selected respondents unavailable (2.8%), and (d) other miscellaneous reasons (2.9%). Data were collected through personal interviews. For the purpose of this analysis, respondents are divided into three groups according to migration status: (1) Immigrants persons born outside the United States (N = 428); (2) Migrants persons born in the United States who have lived in the present county for less than five years (N = 1044), and (3) Non-migrants persons born in the United States who have lived in the present county for at least five years (N = 5004). Although some of the immigrants (79 persons or 18.5%) were also internal migrants, they were included only in the immigrant group. That the proportion of immigrants in the sample (6.6%) was 180 Nagi and Haavio-Mannila somewhat larger than in the total population of the United States in 1970 (4.7%), can be attributed to the exclusion of persons below 18 years of age from this survey. As many as 17.3% of the sample had migrated in five years. The classification of internal migrants, which excludes external migrants, diminishes the proportion of migrants in the analysis to 16.1%. The total proportion of internal migrants in this sample (17.3%) is nearly identical to the 17.1% five-year intercounty mobility of population (5 years old and over 1965-70) reported by the U.S. Bureau of the Census in its Subject Report `Migration Between State Economic Areas'. The four year mobility of persons 18 years and over in 197073 reported in Current Population Reports (1974) was 12.6%. The Migration Groups A comparison of immigrants, migrants, and non-migrants in this survey on the standard socio-demographic characteristics reveals important differences (Table 1). Using X 2 tests, all variables show statistically significant differentials (p<.001) except for gender (p = .21). Immigrants display a clear tendency to be older, less often married, and to comprise fewer whites. Migrants depict the opposite pattern with similar strength. While migrants were more highly represented among those with higher educational backgrounds indicating a linkage between education and internal migration, 52% of the immigrants did not complete high school. The immigrants' low educational attainments are primarily a function of age. More of the migrants were in white collar occupations and, again in large part because of age, more of the immigrants were not connected with the labor force. Of the three groups, migrants seem to be in the better economic conditions as measured by both personal and family incomes; this can be expected in view of their higher educational attainments. It is interesting to note that while immigrants are the least represented in the lowest levels of personal income, immigrant family incomes show them in the worst situation. This demonstrates the influence of loss of income because of the lack of secondary employment. More of the immigrants were unemployed housewives with no earnings. Migration and Health Status As mentioned earlier, two dimensions of health status were included in this analysis: physical and emotional performance. A `Physical Migration, Health Status and Utilization of Health Services 181 Table 1 Socio-demographic Characteristics and Migration Status Total Sample (N=6493) Immigrants (N=428) Migrants (N=1044) Non-Migrants (N=5004) -% 34.9 20.8 62.9 30.3 47.2 46.3 31.7 50.5 17.9 32.9 5.4 19.2 42.1 41.4 44.5 41.6 57.9 58.6 55.5 58.4 68.0 65.4 77.1 66.3 32.0 34.6 22.9 33.7 85.7 69.6 89.8 86.2 14.3 30.4 10.2 13.8 40.3 52.1 23.7 42.7 59.7 47.9 76.3 57.3 27.1 24.5 33.3 26.1 26.3 24.1 24.4 26.9 46.6 51.4 42.4 47.1 35.7 43.8 26.8 36.9 36.7 33.2 40.7 36.2 22.3 17.6 25.6 21.9 5.4 5.4 6.9 5.0 37.7 43.4 31.8 38.5 41.2 37.6 47.8 40.2 21.0 18.9 20.3 21.3 Socio-demographic Characteristics Age 18-34 years 35-64 years 65 years and over