Study with the several resources on Docsity
Earn points by helping other students or get them with a premium plan
Prepare for your exams
Study with the several resources on Docsity
Earn points to download
Earn points by helping other students or get them with a premium plan
Community
Ask the community for help and clear up your study doubts
Discover the best universities in your country according to Docsity users
Free resources
Download our free guides on studying techniques, anxiety management strategies, and thesis advice from Docsity tutors
A scholarly article presented at the annual meeting of the Population Association of America in April, 2010. The authors, J. Trent Alexander and Annemarie Steidl, challenge the long-held belief that women were more migratory than men based on E.G. Ravenstein's 'Laws of Migration'. Using data from various countries and demographic analysis, they argue that the apparent over-representation of women among internal migrants was not due to their higher propensity to move but rather to the higher mortality and emigration rates among men.
What you will learn
Typology: Study notes
1 / 23
This page cannot be seen from the preview
Don't miss anything!
Gender and the 'Laws of Migration': A reconsideration of nineteenth-century patterns
To be presented at the annual meeting of the Population Association of America April, 2010
J. Trent Alexander Minnesota Population Center University of Minnesota Annemarie Steidl Institute for Economic and Social History University of Vienna
Scholars across the social sciences have observed a long-term "feminization" of migration, where international migrant streams formerly dominated by men have gradually become gender balanced or even majority-female (e.g., Donato et al. 2006; Gabaccia 1996; Houston 1984; Simon and Brettell, 1986; United Nations 2006; Yinger 2006). The earliest works making this case were seen as revolutionary, reversing long-held beliefs about the role of men in undertaking long-distance, high-risk moves (New York Times 1985). Subsequent scholarship has shown the pattern to be significant and widespread. Current work on the topic is producing even better data on timing and has begun attempting to explain the consequences of the long-term shift (e.g., Donato et al. 2009). Scholarship on internal migration has generally been less concerned with global patterns in the sex-balance of migrants. This is unfortunate and in a sense ironic, since the earliest and boldest statements about gender and migration concerned internal migrants in England and Wales in the late- nineteenth century. Geographer E.G. Ravenstein's influential "Laws of Migration" stated in no uncertain terms that women were simply more migratory than men, particularly among short-distance, within-country movers. This paper uses census microdata to take a fresh look at the relationship between gender and internal migration in late-nineteenth century Europe and North America. At the most basic level, we argue that there was a significant flaw in Ravenstein's key finding on gender and migration. Our analysis suggests that women were rarely more migratory than men in the countries we consider, not even in Great Britain in 1881. The data Ravenstein presented was technically correct, but his source materials did not allow him to probe beyond basic bivariate relationships. In short, we argue that the apparent over-representation of women among internal migrants was driven by an over-representation of women among the
Why revisit an article from 1885? E.G. Ravenstein's two papers on the "Laws of Migration" are probably the most-often cited works in migration studies. Using "Web of Science" to search for citations to Ravenstein's work, we found that his "Laws of Migration" articles have been cited more than 350 times since 1974.^1 And it is worth noting that the Web of Science excludes all books, as well as articles appearing in edited books, conference proceedings, working papers series, and many humanities journals. The complete number of citations to Ravenstein's work was surely much higher. According to the Web of Science citation index, journals containing the most citations to Ravenstein's work during this period were, in order, International Migration Review , Demography , Social Forces , Social Science History , American Sociological Review, Annals of the Association of American Geographers , Geografiska Annaler , Population and Development Review , Social History , and Transactions of the Institute of British Geographers. It is difficult to imagine any other single piece of work being regularly cited in such a broad range of venues. And this is just in the past 35 years. Various literature reviews from previous decades suggest that the Laws of Migration articles were recognized as path- breaking long before the 1970s. This is not to suggest that every mention of Ravenstein over the years has been in- depth or even positive. Even in 1885, after Ravenstein's original presentation of his first paper, two members of the audience expressed serious concerns with his use of the term "laws." According to a written description of the proceedings, Ravenstein was gracious and
(^1) We are grateful to Johanna Leinonen and Elizabeth Zanoni for allowing us to use results from an extensive literature review they conducted in Summer 2008. Their review searched for citations to Ravenstein'swork in the "Cited Reference Search" available via the "Web of Science" tab at the ISI Web of Knowledge site (http://apps.isiknowledge.com/). They identified all works citing the "Laws of Migration" articles, covering allmajor journals from 1976-2008. They also produced summaries of many recent citations that considered Ravenstein in more than a superficial way. We hope to incorporate some of that work into the final draft of this paper.
conceded that his findings "lack the rigidity of physical laws" (Ravenstein 1885: 235). Even so, his use of that term and the more general clarity of his ideas probably go very far in explaining why his papers have stimulated many other researchers. Social scientists of all stripes have clearly seen value in Ravenstein's approach and findings. The two of Ravenstein's laws relating to gender asserted that (1) short-distance migrants far outnumbered longer distance ones, and that (2) within-country moves were dominated by women. Ravenstein argued that, while men may have been more likely to move long distances, women were the ones making the majority of the short-distance moves that made up most of all migratory activity. The net of these two tendencies led Ravenstein to conclude that women were on the whole significantly more migratory than men. In his typically straightforward language, he stated that "Woman is the greater migrant than man...This may surprise those who associate women with domestic life, but the figures of the census clearly prove it" (Ravenstein 1885: 196). Ravenstein's findings dealing with gender have fared particularly well over the years. There have been at least two systematic literature reviews of research dealing using Ravenstein's laws. In the early 1960s, two sociologists published a U.S.-focused literature review arguing that Ravenstein's laws on gender and distance "have a high degree of predictive validity" (Macisco and Pryor 1963: 221). Their conclusion argued that "the review of the literature undertaken here has shown that, with certain exceptions, the following tentative uniformities do exist: (1) For rural-urban streams, females are more migratory than males, and (2) Women predominate in short-distance moves." Geographer D.B. Grigg conducted a much broader-ranging literature review in 1977. Much like the sociologists before him, Grigg said of Ravenstein that "his laws on sex and age differentials have been
have used it. With data only on county and sex, Ravenstein naturally focused on more strictly geographic questions.
Figure 1 A page from the published tables of the 1881 British census
Source: Census Office, Census of England and Wales, 1881, Volume 3 (London: Eyre andSpottiswoode, 1883), 266.
We rely on a microdata file of 26 million individual records from the 1881 British census. Microdata is individual-level data, where each record represents a person and contains all of each persons characteristics as they were enumerated in the census. We obtained the 1881 census microdata from the North Atlantic Population Project (NAPP), a data dissemination project housed at the University of Minnesota and created in partnership with the U.K. Data Archive and several other national statistical agencies. Microdata usually
permit a great deal more analytical flexibility than the published books or tables released by national census offices. The downside of using microdata is that they usually provide only a sample of the population, since data entry is expensive and the benefits of increased sample size are small relative to what could be gained from entering a sample from another census. The NAPP microdata is extremely unusual in that it provides the complete population enumerated in the 1881 U.K. census. Compete data was entered by the Church of Jesus Christ of Latter Day Saints (LDS), for genealogical purposes. The NAPP project has optimized the files for academic researchers, mainly by coding alphabetic variables and providing good documentation. NAPP also provides the complete population from the U.S., Norway, Canada, and Sweden during approximately the same period. Our analysis of these data proceeds in two steps. First, we verify that Ravenstein's findings can be replicated with microdata from the 1881 British census. Next, we investigate whether basic age-standardization can account for the apparent sex-differentials among internal migrants in England and Wales. In our conclusion we briefly consider whether similar patterns were apparent in Norway and the United States. This analysis uses data from six additional censuses from Norway and the United States. Table 1 lists the datasets that we use in our work.
Table 3 presents the same statistics as derived from the microdata, confirming that these data are indeed comparable to those used by Ravenstein in 1885. This table also uses the adjacency of counties to get a general sense of the relationship between gender and distance. Women were over-represented among all internal migrants, but they were particularly over-represented among those who left their county of birth for an adjacent county. Those who moved to a county that did not share a border with their county of birth were slightly more likely to be men than were the shortest-distance movers. These data thus suggest some support for Ravenstein's contention that women's predominance among migrants declined with distance moved. When NAPP provides identifiers on parish-of-birth, we expect to be able to work with much finer-grained measures of distance moved.
Table 2 Sex ratios among residents of the United Kingdom (produced from aggregate data)
Source: E.G. Ravenstein, "The Laws of Migration," Journal of the Royal Statistical Society, 1885: 197.
Table 3 Sex ratios among residents of England and Wales (produced from microdata)
Females per 100 malesamong natives of
England and Wales Residing in county where born 104 Migrants from an adjacent county 112 Migrants from a non-adjacent county 109
Other countries Migrants from Scotland or Ireland 92 Migrants from other countries 86 Total 108
Source: North Atlantic Population Project (NAPP), Minnesota Population Center
The second step of our analysis is to consider the extent to which these apparent sex differentials among internal migrants could be driven by demographic processes other than internal migration itself. Our basic hypothesis is that the over-representation of women among internal migrants was not be due to their higher propensity to move, but rather to the much higher rate at which male migrants left the population, through either death or emigration. In this way, men were probably just as likely to make internal moves as women were; the difference was that men did not remain in the population to be counted when the decennial census was conducted.
Figure 2 Number of women per 100 men, internal migrants in England/Wales
0
20
40
60
80
100
120
140
160
180
200
(^0 5 10 15 20 25 30 35 40) Age 45 50 55 60 65 70 75 80 85 90
Source: North Atlantic Population Project (NAPP), Minnesota Population Center
Figure 2 by itself does not demonstrate that women were more likely than men to move at any particular age. For instance, if sex ratios for the non-migrant population followed the same curve, that would suggest that women and men had the same rate of migration at all ages. In order to determine the ages (and magnitude) of true migrant sex differentials, we applied men's and women's age-specific migration rates to a standard population. In this case, we chose to standardize on the population of native-born women in England and Wales. Figure 3 reproduces the information in Figure 2, but also includes an additional curve showing the age-standardized sex ratio for migrants, by single years of age. To make the ratios for the solid line in Figure 3, we first determined the number of male migrants that would be expected if men had the same population structure as women. We obtained this
"age-standardized" number of male migrants by multiplying native-born men's migration rate at each year of age by the absolute number of native-born women at each year of age. These "expected" numbers of male migrants at each year of age, along with the known number of female migrants at each year of age, were then used to calculate a new set of age-specific migrant gender ratios. This direct age-standardization effectively controlled for the different age-structures of men and women in 1881 England and Wales.
Figure 3 Number of women per 100 men, internal migrants and the entire population, England and Wales
Source: North Atlantic Population Project (NAPP), Minnesota Population Center
The difference between the dashed and solid lines in Figure 3 shows the extent to which apparent migrant sex differentials were due to differences in the men's and women's
rates of mortality than women beginning at about age 35 (Lancaster 1990: 376). Adult men also seem to have been over-represented among out-migrants.^3 The British population below age 16--free from sex differentials in mortality and out-migration--was sex-balanced. Probably because of sex-differentials in both mortality and out-migration, there were more women than men at every single year of age from 16 up. Many adult men were "missing" from the migrant population not because they had failed to move, but rather because they had left the country or died. The effect of these differences between adult men and women were even further exaggerated by several unique attributes of the younger British population. First, the 1881 British youth were a very large group. As the population pyramid in Figure 4 suggests, forty percent of the native-born population was under the age of 16 in 1881. Second, the young population was sex-balanced. The two main components of population change--mortality and out-migration--did not much effect people below the age of 16 (except for infants). Finally, British children had very low rates of internal migration. Taken together, these patterns meant that there was a massive amount of overlap between the youth population and the non- migrant population. In 1881 England and Wales, statistics on youth and statistics on non- migrants were hopelessly entangled; this was the root of Ravenstein's problem. Stated in the most baldest terms, any statistics Ravenstein generated on "the non- migrant population" were in fact statistics mostly about children. Statistics about "the migrant population," on the other hand, were mostly about adults. Statements Ravenstein made (^3) We have compiled data on the British-born population living in Canada, the United States, Scotland, New Zealand, the six Australian colonies, India, and in the armed services. In all cases, men outnumber women. On thewhole, of about 1.5 million British-born persons living in one of these areas, 950,000 were men. As was true for internal migrants, the age-structure of the international migrant population was skewed upwards, with childrenunder-represented. Data on English/Welsh-born in Canada, the United States, and Scotland were generated from the North Atlantic Population Project. Data on the England/Welsh-born in New Zealand, Australia, and India comefrom published census reports from those countries.
comparing non-migrants to migrants were at risk of instead simply comparing children and adults. This is exactly what we think happened when Ravenstein was formulating his laws on migration and gender. The core of his finding was not that women were more migratory than men; it was that adults were more migratory than children.
Figure 4 Population pyramid of the native-born, England and Wales, 1881
Migrants are in black, non migrants in gray
‐ 8 ‐ 6 ‐ 4 ‐ 2 0 2 4 6 8
0 ‐ 4 yrs.
5 ‐ 9 yrs.
10 ‐ 14 yrs.
15 ‐ 19 yrs.
20 ‐ 24 yrs
25 ‐ 29 yrs
30 ‐ 34 yrs.
35 ‐ 39 yrs.
40 ‐ 44 yrs.
45 ‐ 49 yrs.
50 ‐ 54 yrs.
55 ‐ 59 yrs.
60 ‐ 64 yrs.
65 ‐ 69 yrs.
70 ‐ 74 yrs.
75+ yrs.
% of women in each group % of men in each group Source: North Atlantic Population Project (NAPP), Minnesota Population Center
moves are male dominated. With county-to-county migration data, we did not feel that our measures of distance were fine-grained enough to confidently address this topic. It is our understanding that the U.K. Data Archive plans to release additional variables that will allow us to identify parish-to-parish moves. There are only 60 counties in England and Wales, but there are more than 15,000 parishes (see Figure 5). Good data on parish-to-parish migration will allow us to assess the applicability of the laws on distance to 1881 England and Wales. We have also begun to collect similar data from more countries, and compare and interpret differences over time and place. This work is only preliminary so far. Figure 6 presents the ratio of women to men for inter-municipality migrants in Norway for 1865, 1875, and 1900. All figures are age-standardized, following the same procedures we used to construct Figure 3. Our preliminary findings suggests that, controlling for age, migration within Norway shifted from being majority female to majority male over the 1865- period. Figure 7 shows the same data for the United States. In the U.S., internal migrants were majority-male for the entire 1870-1900 period. The differentials show no change over time, and virtually no variation by age. This work is obviously still very preliminary, and we welcome any advice about whether either of these approaches (or another approach that we haven't thought of) would be most useful.
Figure 5 Counties and parishes in England and Wales, 1881
Counties (N=60) Parishes (N=15,021) Source: U.K. Data Archive. Note: We are grateful to Kevin Schurer of the U.K. Data Archive for providing us with GIS- ready maps of counties and parishes in 1881 England and Wales.
