Transatlantic wage gaps and the migration decision: Europe–Canada in the 1920s


As has been seen in other contexts, workers in similar occupations earned much higher wages in Canada than Europe during the 1920s. This observation and related aspects of immigration are addressed with a life-cycle model of the migration decision. The characteristics of immigrants from five European countries: Ireland, Italy, the Netherlands, Poland, and Sweden, are explored in a way that sheds light not just on those population flows but on the process of immigration generally. We draw on passenger manifests from immigrant ships for information on the individual migrants. Simulations, based partly on the these manifests, reveal that the costs of migration, borrowing constraints, concern for status, and country-specific taste preferences account for key immigrant characteristics, and help explain the large wage differentials that persisted despite Canada’s relatively open immigration policy.

This is a preview of subscription content, access via your institution.

Fig. 1
Fig. 2


  1. 1.

    Hatton and Williamson (1994, 35) also document the change in the ratio of real wages in sending and receiving countries over the period 1850–1913. Their results for the 1870s are similar. For a discussion of antebellum immigrants, see Ferrie (1999).

  2. 2.

    At about this time Liverpool to New York fares for passengers travelling steerage were about $15–$20, and the voyage averaged 44 days (Babak 1931, 166; Gould 1979, 612–613). Annual earnings of lower-skilled urban workers in the USA were roughly $300 (U.S. Bureau of the Census 1975, 164, 165, 168 [based on 1880 with an allowance for unemployment]). The fare of $20 as a percentage of annual earnings in the sending countries would have been about 10, 15, and 30 % for Britain, Germany, and Italy, respectively.

  3. 3.

    In this formulation, income in both the home and receiving country is uncertain, but the discount rate, moving cost, and future lifetime are known. In most models, as in ours, migration is treated as an individual decision. Certainly there were those who emigrated as a family, but adding this complication should not affect the underlying insights. It might be noted that in our sample of immigrants to Canada, just 7 % were accompanied by a spouse, and only 5 % had accompanying children (Table 3).

  4. 4.

    Chiswick’s (2000) approach is much the same, although his model does not include uncertainty and future lifetime is treated as infinite.

  5. 5.

    While shedding light on the question of self-selection, our paper emphasizes the role of tastes in determining the wage incentives required by potential migrants and the effect of their position on the wage distribution.

  6. 6.

    For a review of research on happiness that includes the impact of place on the income distribution see Frey and Stutzer (2002).

  7. 7.

    In their chapter on Italian emigration in the Hatton–Williamson edited volume, Faini and Venturini (1994) also emphasize the role of financial constraints in limiting migration. An inability to borrow was even more restrictive in earlier years. During the seventeenth and eighteenth centuries more than half the free immigrants to the Thirteen Colonies came as indentured servants, an arrangement that emerged to deal with a borrowing constraint faced by low-wage workers in Europe. Unable to pay for their passage, workers signed indenture contracts that were sold in America. See, for example, Galenson (1981); among his many publications on indentured servitude, Grubb (1985, 1994).

  8. 8.

    For a discussion of chain migration from Italy to North America that has a focus on Canada, see Sturino (1990).

  9. 9.

    Employment agencies were in operation in addition to the more informal arrangements for those nationalities who already had established communities in Canada. Still, unlike the indication in Rosenbloom (2002) for the USA, a large majority of immigrants to Canada paid their own travel costs. The breakdown for the five countries in our sample over the period, 1925 to 1929, is: Ireland—61 %; Italy—96 %; Netherlands—95 %; Poland—85 %; Sweden—94 %.

  10. 10.

    This statement corresponds quite well to Borjas (1987), which finds that the wages of immigrants with the characteristics described in the Canada Yearbook converged more rapidly to the wages of natives. Similar findings for immigrants from English-speaking countries to Canada during the early twentieth century are reported in Inwood et al. (2015).

  11. 11.

    Among the prohibited groups were Mennonites, Hutterites, and Doukhobors.

  12. 12.

    Kelley and Trebilcock (1998, 195) estimate that under the Railway Agreement, 185,000 Europeans entered Canada between 1925 and 1929.

  13. 13.

    On the impact of US immigration policy on immigration to Canada see Kelley and Trebilcock (1998, 210) and Lew and Cater (2002).

  14. 14.

    Since moving to a city conflicted with the official objective of attracting agricultural workers, a system emerged whereby prospective immigrants would obtain, via a family or acquaintance in Canada, an offer of employment from a Canadian farmer. The offer would then be used to gain entry into Canada even if the immigrant had no intention of taking the job. Since the supply of legitimate invitations fell short of the demand, a market arose in which a farmer would sign a fraudulent invitation in exchange for a payment of between $25 and $75. In time, even farmers with bona fide requests for workers demanded these kickbacks (Ramirez 1989; Sturino 1988).

  15. 15.

    Immigration from all countries was 1,264,000 or 14.9 % of the 1920 Canadian population (Leacy 1983, A350). There was little return migration among these groups.

  16. 16.

    The records were retrieved from the Canadian Library and Archives Immigration Passenger List database through the website, Records for each individual arrival are searchable by name, nationality, data of arrival, and other criteria. The search results provide a link to the original passenger list record which has the information to construct the data set. We initially searched for records by year and nationality and drew roughly equally sized samples for each year (1925–1929) based on the alphabetized order of the search results. This should insure that the samples for each country are random. In the case of observations of immigrants from the Netherlands, fewer than 1000 legible records could be retrieved for each year, which accounts for the smaller sample size for that country.

  17. 17.

    Students and visitors are excluded.

  18. 18.

    The assignment of occupations to wage classes is based on weekly wages reported in the Labour Gazette (Canada) in the period 1925–1929. The wage classes are: low - less than $17.50, medium low - $17.50–$22.50, medium high -  $22.50–$30.00, high - over $30.00. Ideally the assignment would be based on wages in the originating countries, but we have less occupational data for Europe. The Canadian data, however, are a good substitute for our purposes (Armstrong and Lewis 2012, 747).

  19. 19.

    Part of the difference in average savings between the Dutch and the other nationalities is attributable to the fact that the Dutch immigrants were far more likely to be travelling with families. This influences the amount of savings directly through the added cost of fares and indirectly through a greater cash requirement after arrival.

  20. 20.

    On Italian immigration and the farm labour system, see Sturino (1988).

  21. 21.

    An approach that solves the model with a single lifetime optimization equation leads to the same results.

  22. 22.

    To simplify the figure the discount rate, r, is set equal to the pure rate of time preference, ρ. In the simulations, r exceeds ρ.

  23. 23.

    The International Labour Review reports wage rates in local currency. As well it provides detailed price data along with weights for a standardized consumer budget. We have used these series to develop PPP measures. Our results are generally similar to those in Williamson (1995). The calculations are for 1929, but any year from 1925 to 1929 would give similar results. Details are in the Appendix of ESM.

  24. 24.

    The high ratio for Italy possibly reflects too low a PPP measure. The exchange rate in 1929 was 19.0 lira to the (Cdn) dollar. We derive a lower PPP rate of 17.3, based on the International Labour Review, but it may have been still less (Williamson 1995, 90).

  25. 25.

    As shown in Table 1, a negligible share of the sending country’s population emigrated to Canada. Although larger, the impact of immigration on the Canadian population (and workforce) was also very small. Thus, the general equilibrium effect of migration on wage rates could hardly have been significant. Treating the wage ratios as exogenous therefore seems appropriate.

  26. 26.

    The medium–low wage class includes unskilled workers and labourers as described in Table 4 (the low wage class is made up mainly of farm workers). The medium–high wage class comprises skilled workers. The allocation of the occupations in Table 4 to the two groups is detailed in the Appendix of ESM.

  27. 27.

    The fare is assumed to include the ocean passage (Table 2) plus $10 for rail transport, which was the rate to Ontario. We assume cash holdings on arrival do not include that rail cost. The ocean crossing itself took about one week. There was the additional time, more significant for those from Eastern Europe, to reach the point of embarkation.

  28. 28.

    These values are for a single male, family contact, non-spring arrival, intending to go into agriculture.

  29. 29.

    Where there is uncertainty, δ, can be interpreted as relative risk aversion. In our model there is no uncertainty. Still, the term, δ, is important in that it affects the “utility” cost of migration arising from the unequal consumption streams as described in Fig. 1. A higher value of δ increases the cost.

  30. 30.

    It has been observed that immigrants to America were not generally drawn from the poorest groups. Selection tended to be from the middle, or lower-middle part of the income distribution. There may have been a variety of reasons, but the need to save enough to provide for subsistence was an important factor. This point is made by Chiswick and Hatton (2003) who discuss migration to the USA in the late nineteenth century: “In…least developed countries and regions, despite the large incentive to emigrate, those who had the most to gain were simply too poor to finance the move”. Faini and Venturini (1994) argue that the growth in wages in Italy helps explain the surge in emigration from that country.

  31. 31.

    The weights selected are well within the range of reported elasticities, although somewhat at the low end. Higher elasticities with respect to per capita income would tend to strengthen our conclusion about the role of subsistence in migration decisions. This treatment of subsistence is in keeping with the sort of advice prospective immigrants were given by friends and family who had already come to Canada. For example, Ganzevoort’s (1999) translation of letters home from Dutch immigrants to Canada in the late 1920s includes: “…please note that those who wish to emigrate would do well to bring along a small sum of money, about 100 guilders (p. 39)”. This equalled subsistence in the Netherlands for two months, based on our weighting.

  32. 32.

    In general, occupations listed as unskilled labourer are in the unskilled category. Higher-paid occupations are in the skilled category. The precise division is given in the Appendix of ESM.

  33. 33.

    Annual hours are put at 2150, close to the value derived by Huberman and Minns (2007, 548) for 1929. Although there was some variation, we apply this value across countries. The Irish unskilled wage is based on the average wage of labourers in Mechanical Engineering, Transport, and Local Authorities (Table 4).

  34. 34.

    In the simulations, therefore, the Irish unskilled wage is 1, and the Canadian unskilled wage is 1.71.

  35. 35.

    For a discussion of the earnings of immigrant farm workers see Armstrong and Lewis (2012, 741–742).

  36. 36.

    There is a large literature on the rate of convergence of immigrant earnings associated especially with George Borjas and Barry Chiswick. A summary of some of this literature is in Borjas (1994). Abbott and Beach (1993) have a detailed analysis of wage convergence in Canada in the 1970s. We assume a quadratic wage path that approximates their findings. There are few Canadian studies for the early twentieth century. A recent paper by Inwood et al. (2015) explores the impact of the Great Depression on relative immigrant earnings, finding a large negative impact. But their results for the decade 1911–1921 can be seen as roughly consistent with a 10-year adjustment period for those from English-speaking countries. The adjustment period for those from other countries appears to have been longer. We have assumed the same period of adjustment for all countries, which may account in part for the higher taste parameter estimates for immigrants from Italy and Poland. Although most immigrants began as farm workers in Canada, there is a considerable literature indicating that they quickly shifted to work more suited to their training. Green and Green (2014) find that despite the high proportion of immigrants in the 1920s who began as farm workers, the occupational distribution of those immigrants corresponded very closely to that of the native population by 1931. On the relative earnings of immigrants to the USA in the late nineteenth and early twentieth centuries, see Minns (2000). His results strengthen “the ‘optimistic’ view of immigrant progress in America (p. 349)”.

  37. 37.

    Based on the values in Table 7 and the assumed parameter values, a Fortran programme is used to derive the optimal migration time for a given value of τ. At this value the lifetime utilities of the emigrant and non-emigrant are compared. The programme then searches for the τ that equates the lifetime utility of the emigrant and non-emigrant.

  38. 38.

    Equivalent consumption is derived from: \(\frac{{\left( {c^{E} - s_{j} } \right)^{1 - \delta } }}{1 - \delta } = \tau + \frac{{\left( {c_{j} - s_{j} } \right)^{1 - \delta } }}{1 - \delta }\) where \(c^{E}\) is the consumption in Canada equivalent to consumption, \(c_{j}\) in home country, \(j\). For unskilled workers in Ireland the values are: \(c_{j} = .91\), \(\tau = .72\), and \(s_{j} = .37\) (see Tables 7, 8); \(\delta = 2\). Since \(r > \rho\) consumption is increasing over time which would affect the comparisons in other years slightly.

  39. 39.

    A breakdown of age at migration by both skill and country of origin reveals that the skilled immigrants from Italy and Poland were in fact slightly younger than the unskilled.

  40. 40.

    The assumed elasticity, δ, of 2 implies sharply diminishing marginal utility of consumption. As well, because of the subsistence constraint, the effect on utility of increases in consumption diminishes with consumption. Because of these two factors, the same (additive) taste parameter gives rise to a much larger consumption effect on high-wage than low-wage workers.

  41. 41.

    For unskilled (skilled) workers in Sweden the values are: c j  = .93 (1.23), τ = 0.45 (.47), and s j  = .42 (.42). See Tables 7 and 8. Note that the greater required increase in consumption for skilled workers is due primarily to the concavity of the utility function. The subsistence constraint also has an impact, since for skilled workers the effect of a wage increase on net consumption is less.

  42. 42.

    Although his reference is to US immigration in the late nineteenth century, John Higham (1985, 65–66) singles out Italians and Poles as groups that were especially looked down on: “The Italians were often thought to be the most degraded of the European newcomers”. He also points to a Polish colony in Illinois that was shunned by those in the area.

  43. 43.

    It should be emphasized that the greater importance of these factors to skilled workers arises not from a difference in their utility functions, as the earlier example of Swedish workers illustrates. Rather, it is due to the concavity of the utility function and the subsistence constraint, which implies that with rising consumption non-pecuniary factors are given increasing importance.

  44. 44.

    It should be pointed out that the wage and price data are drawn mainly from the International Labour Review for 1929. The advantage of this source is that it provides detailed information for the same (reported) occupations and the same consumption goods. Refinement and additional sources might lead to somewhat different wage comparisons and, of course, different simulation results. The wage ratio for skilled workers from Italy seems high, for example. The objective here, however, has not been to develop a new set of international wage comparisons, but rather to ask what implications can be drawn about migration from the international wage data, possibly imperfect, that currently exists. Williamson (1995) presents a long-term review of international wages. For the 1920s his sources and results are similar to ours.

  45. 45.

    This reflects the difference in subsistence between Poland, 0.60, and Canada, 1.09 (see Tables 7 and 9).

  46. 46.

    The implied consumption difference due to taste for the home country falls from 69 to 46 % for Italian unskilled workers and from 65 to 42 % for Polish unskilled workers. These values are not much different from the taste effects derived for the other countries (see Appendix of ESM). It should be noted, though, that convergence may have been incomplete for workers from the preferred countries as well.

  47. 47.

    Immigration to the European countries in 1924, for example, was: Ireland—333 (British North America); Italy—495 (Canada); Netherlands—269 (Americas); Sweden—142 (Americas excluding the USA). Ferenczi (1969, 735, 754, 760, 840). To the extent that return migrants were not counted as immigrants, the numbers would have been higher. Return migration to Poland over the period 1925–1929 was 3407, which was less than 5 % of Polish immigration to Canada during that time (Reczyńska 1996, 93; and Table 1).

  48. 48.

    An early estimate of the intertemporal elasticity of substitution by Hall (1988) put it close to zero, implying almost no willingness to substitute consumption between periods. The value we assume is within the range of Epstein and Zin (1991) who put the value between 0.2 and 0.8 (δ between 1.25 and 5). If, for example, we assume δ = 5, and apply it to unskilled Irish workers, the implied equilibrium taste effect is reduced from 26 to 18 % of consumption at migration time, implying less emigration. It is generally recognized that those who are more risk averse are less likely to emigrate. There is no uncertainty in our model; instead a higher value of δ discourages migration because of the greater impact of an unequal consumption stream on utility.

  49. 49.

    These calculations assume the borrowing constraint applies only to the migration cost. If instead new migrants are unable to borrow after arrival, they would be forced to consume less over the 10-year adjustment period. This would increase the impact of the borrowing constraint and also reduce the implied taste preference effect.


  1. Abbott M, Beach C (1993) Immigrant earnings differentials and birth-year effects for men in Canada: post-war—1972. Can J Econ 36:505–524

    Article  Google Scholar 

  2. Abramitzky R, Boustan L, Eriksson K (2012) Europe’s tired, poor, huddled masses: self-selection and economic outcomes in the age of mass migration. Am Econ Rev 102:1832–1856

    Article  Google Scholar 

  3. Armstrong A, Lewis F (2012) International migration with capital constraints: interpreting migration from the Netherlands to Canada in the 1920s. Can J Econ 45(2):732–754

    Article  Google Scholar 

  4. Babak FL (1931) Spanning the Atlantic. Alfred A. Knopf, New York

    Google Scholar 

  5. Borjas G (1987) Self-selection and the earnings of immigrants. Am Econ Rev 77(4):531–553

    Google Scholar 

  6. Canada, Department of Labour (various), The Labour Gazette. Ottawa

  7. Canada, Library and Archives Canada. Canadian Immigration Passenger Lists. Data accessed at

  8. Canada Yearbooks, 1926–1930. Ottawa

  9. Canadian Pacific Railway (1925a) Canadian Pacific sailings to and from Europe, April 17 (pamphlet)

  10. Canadian Pacific Railway (1925b) Canadian Pacific: to and from Europe third class fares, June 8 (pamphlet)

  11. Chiswick BR (2000) Are immigrants favourably self-selected? In: Brettell CB, Hollifield JF (eds) Migration theory: talking across disciplines. Routledge, New York, pp 63–82

    Google Scholar 

  12. Chiswick BR, Hatton TJ (2003) International migration and the integration of labor markets. In: Bordo M, Taylor A, Williamson J (eds) Globalization in historical perspective. University of Chicaogo Press, Chicago, pp 65–120

    Google Scholar 

  13. Easterlin R (1974) Does economic growth improve the human lot? Some empirical evidence. In: David PA, Reder MW (eds) Nations and households in economic growth: essays in honor of Moses Abramovitz. Academic Press Inc, New York

    Google Scholar 

  14. Epstein L, Zin S (1991) Substitution, risk aversion, and the temporal behavior of consumption and asset returns: an empirical analysis. J Polit Econ 99:263–286

    Article  Google Scholar 

  15. Faini R, Venturini A (1994) Italian immigration in the pre-war period. In: Hatton T, Williamson J (eds) Migration and the international labor market. Routledge, London, pp 1850–1939

    Google Scholar 

  16. Ferenczi I (1969) International migrations, vol 1, demographic monographs, vol 7. Breach Science Publishers (first published 1929), New York

  17. Ferrie J (1999) Yankees now: immigrants in the antebellum U.S., 1840–1860. Oxford University Press, New York

    Google Scholar 

  18. Frey B, Stutzer A (2002) What can economists learn from happiness research. J Econ Lit 40:302–335

    Article  Google Scholar 

  19. Galenson D (1981) White servitude in colonial America: an economic analysis. Cambridge University Press, New York

    Google Scholar 

  20. Ganzevoort H (1988) A bittersweet land: the Dutch experience in Canada, 1890–1980. McClelland and Stewart, Toronto

    Google Scholar 

  21. Ganzevoort H (1999) The last illusion: letters from Dutch immigrants in the “land of opportunity”, 1924–1930. University of Calgary Press, Calgary

    Google Scholar 

  22. Gould JD (1979) European inter-continental emigration, 1815–1914: patterns and causes. J Eur Econ Hist 8:593–679

    Google Scholar 

  23. Green AG, Green DA (1993) Balanced growth and the geographical distribution of European immigrant arrivals to Canada, 1900–1912. Explor Econ Hist 30:31–59

    Article  Google Scholar 

  24. Green AG, Green DA (2014) Immigration and the Canadian earnings distribution in the first half of the 20th century. University of British Columbia (unpublished)

  25. Green AG, MacKinnon M (2001) The slow assimilation of British immigrants in Canada: evidence from Montreal and Toronto, 1901. Explor Econ Hist 38:315–338

    Article  Google Scholar 

  26. Grubb F (1985) The incidence of servitude in transatlantic migration, 1771–1804. Explor Econ Hist 22:316–339

    Article  Google Scholar 

  27. Grubb F (1994) The end of European immigrant servitude in the United States: an economic analysis of market collapse, 1772–1835. J Econ Hist 54:794–824

    Article  Google Scholar 

  28. Hall R (1988) Intertemporal substitution in consumption. J Polit Econ 96(2):339–357

    Article  Google Scholar 

  29. Hatton TJ, Williamson JG (1994) International migration 1850–1939: an economic analysis. In: Hatton T, Williamson J (eds) Migration and the international labor market. Routledge, London, pp 1850–1939

    Google Scholar 

  30. Hatton TJ, Williamson JG (1998) The age of mass migration: causes and impact. Oxford University Press, New York

    Google Scholar 

  31. Higham J (1985) Strangers in the land: patterns of American nativism, 1860–1925. Atheneum, New York

    Google Scholar 

  32. Huberman M, Minns C (2007) The times they are not changin’: days and hours of work in old and new worlds, 1870–2000. Explor Econ Hist 44:538–567

    Article  Google Scholar 

  33. International Labour Review (ILR), various issues

  34. Inwood K, Minns C, Summerfield F (2015) Reverse assimilation? Immigrants in the Canadian labour market during the great depression. University of Guelph (unpublished)

  35. Johansson O (1967) The gross domestic product of Sweden and its composition, 1861–1955. Almquist and Wiksell, Stockholm

    Google Scholar 

  36. Kelley N, Trebilcock M (1998) The making of the mosaic: a history of Canadian immigration policy. University of Toronto Press, Toronto

    Google Scholar 

  37. Kennedy K (1971) Productivity and growth: the Irish experience. Clarendon Press, Oxford

    Google Scholar 

  38. Kilpatrick R (1973) The income elasticity of the poverty line. Rev Econ Stat 55:327–332

    Article  Google Scholar 

  39. Leacy FH (ed) (1983) Historical statistics of Canada, 2nd edn. Statistics Canada, Ottawa

    Google Scholar 

  40. Lew B, Cater B (2002) The impact of United States immigration quotas on migration to Canada during the 1920s. Working paper, Trent University

  41. Maddison A (2003) The world economy: historical statistics. Organization for Economic Co-operation and Development

  42. Massey DS, Arnango J, Hugo G, Kouaouci A, Pelligrino A, Taylor E (1993) Theories of international migration: a review and appraisal. Popul Dev Rev 19:431–466

    Article  Google Scholar 

  43. Minns C (2000) Income, cohort effects, and occupational mobility: a new look at immigration to the United States at the turn of the 20th century. Explor Econ Hist 37:326–350

    Article  Google Scholar 

  44. Mitchell B (1981) European historical statistics, 1750–1975. Second revised edition. Macmillan, London

    Google Scholar 

  45. Perin R (1989) Introduction: the immigrant: actor or outcast. In: Perin R, Sturino F (eds) Arrangiarsi: the Italian immigrant experience in Canada. Guernica, Montreal, pp 9–37

    Google Scholar 

  46. Ramirez B (1989) Workers without a cause: Italian immigrant labour in Montreal. In: Perrin R, Sturino F (eds) Arrangiarsi: the Italian immigration experience in Canada. Guernica, Montreal, pp 1880–1930

    Google Scholar 

  47. Reczyńska A (1996) For bread and a better future: emigration from Poland to Canada, 1918–1939. University of Toronto Press, Toronto

    Google Scholar 

  48. Rosenbloom JL (2002) Looking for work, searching for workers: American labor markets during industrialization. Cambridge University Press, Cambridge

    Google Scholar 

  49. Roy A (1951) Some thoughts on the distribution of earnings. Oxford Econ Pap 3:135–146

    Article  Google Scholar 

  50. St. Jean E (2004) Swedes on the move: politics, culture and work among Swedish immigrants in British Columbia, 1900–1950. University of Victoria, unpublished dissertation

  51. Sturino F (1988) Italian immigration to Canada and the farm labour system through the 1920’s. In: Potestino J, Pucci A (eds) The Italian immigrant experience. Canadian Italian historical association, pp 81–97 (Originally published in Studi emigrazione XXII, 77, 1985)

  52. Sturino F (1990) Forging the chain: a case study of Italian migration to North America, 1880–1930. The Multicultural Society of Ontario, Toronto

    Google Scholar 

  53. U.S. Bureau of the Census (1975) Historical statistics of the United States, colonial times to 1970, Bicentennial Edition, Part 1. Washington, DC

  54. Von Bochove C, Haitker T (1987) Main national accounting series, 1900–1986. CBS occasional paper no. 17. The Hague

  55. Wegge S (1998a) Migration decisions in mid-nineteenth-century Germany. J Econ Hist 58:532–535

    Article  Google Scholar 

  56. Wegge S (1998b) Chain migration and information networks; evidence from nineteenth-century Hesse-Cassel. J Econ Hist 58:957–986

    Article  Google Scholar 

  57. Williamson JG (1995) The evolution of global labor markets since 1830: background evidence and hypotheses. Explor Econ Hist 32:141–198

    Article  Google Scholar 

Download references


We thank Stan Engerman and participants at meetings of the Canadian Network for Economic History and the Canadian Economics Association, along with two referees. We gratefully acknowledge the research assistance of Scott Edmonds, Laura Swan and Odette Melvin.

Author information



Corresponding author

Correspondence to Frank D. Lewis.

Electronic supplementary material

Below is the link to the electronic supplementary material.

Supplementary material 1 (DOCX 26 kb)

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Armstrong, A., Lewis, F.D. Transatlantic wage gaps and the migration decision: Europe–Canada in the 1920s. Cliometrica 11, 153–182 (2017).

Download citation


  • Immigration
  • Wage differentials
  • Borrowing constraints
  • Status

JEL Classification

  • J61
  • N32
  • N34