1
RETURN MIGRATION FROM THE USA TO BRITAIN, 1815-60
By John Killick. School of History, Leeds University, Ret’d.
This paper is preliminary. Please do not cite without permission from the author. E-mail address for contact:
It has been conventional wisdom for many years that very few migrants—possibly only one per cent—returned to Europe in the sailing ship era because of the hardships of the voyage. Raymond Cohn summarized the best sources in his recent book. Mass Migration under Sail (2009), pp. 10-11, and concluded that migration ‘in virtually all cases was permanent’. The long and hazardous outward voyage made return unlikely, and often there was little to go back to.
Nearly all migration historians agree with this low rate, but a minority have suggested there was a larger return migration. In particular, Wilbur Shepperson, Emigration and Disenchantment (1965) researched the biographical accounts of 50 returnees – including many famous names – Cobbett, Kemble, Nuttall, Trollope etc. to show what they thought of the US, and why they left. This is convincing as social analysis, but is too small a number for statistical conclusions. Shepperson shows how they came in many varieties – farmers, mechanics, professionals etc. – each with personal reasons for return, but it is arguable they were in the main a special literary group.
The aim of this paper is to buttress Shepperson’s case from more mundane sources– first from the Cope packets which kept passenger lists from all their eastbound packets – about 25000 names over 50 years; Second by checking the British press for return passenger details more fully– which probably Shepperson was not able to do,- and thirdly from some British government reports in the mid and late 1850s, not yet thoroughly utilised. For this see the final section of Killick, ‘Transatlantic steerage fares’.
My thesis is that the return migration was larger than previously thought, and was strongly tied to general trade and migration conditions. A small proportion of those attracted by the booms, fled immediately during the following financial crisis, and more in the subsequent depressions. This has implications for what we think about the transatlantic crossing pre steam. This was less arduous and horrific than often painted - the new packet ships after 1818 made return relatively easy, and even re-emigration possible. Similarly frequent returns suggest the Anglo-American contrast were less marked, and the economic gains from emigration less obvious than often suggested at least in the east. The social and political contrasts remained of course, and there were huge differences between groups.
:
1. RETURN MIGRATION IN THE LATE 19TH CENTURY:
It has been long established that there was substantial return migration from America to Europe in the late 19th and 20th centuries, as there was between other sending and receiving countries such as Europe and Latin America or Europe and Australasia, numbers being determined by the rhythm of boom and slump, and by ethnic characteristics in each case. See Gould, ‘European Intercontinental Migration’.
Figure 1: Return Migration, 1854-1889.
Source: For 1854-61, showing British emigration from the US.- See Killick, ‘Transatlantic Fares…’ For 1862-1890 showing all migration from the US - see Simon, United States Balance of Payments, 1861-1900, Table 13.
Figure 1 shows return migration from the US to Britain and Europe rising after the economic crises of 1866 and 1873, but in any one year the numbers of say British or Scandinavians returning was very much higher than that of say European Jews who had obvious special reasons not to go back. The general reasons however for the growing return migration, which had doubled by the depression of the mid-1890s were the great improvements in travel offered by the fast new steamships and continental railroad systems. Most migrants still went to the US intending to stay for ever, but increasing numbers went with short-term plans to work hard in the US and make sufficient money to establish themselves back in their own country after a relatively short sojourn.
By 1900 some tradesmen were crossing the Atlantic just to take advantage of short-term or seasonal American building and infrastructure projects that they saw advertised in the commercial press. Italian workmen visiting New York or Argentina came so regularly that they began to be called swallows. After the hiatus of the of the wars and the depression, it became normal for young people even students to visit North America in the summer to make some cash and see the country, (as I did myself in 1958 in the first ever BUNAC - British Universities North America Club charter flight to Canada, and my son in 1992).
2. RETURN MIGRATION IN THE 1850s:
It is reasonable to assume that the late (19th pattern began in the late 1850s after the great migration of the late 1840s – early 1850s. The first steamships only carried cabin passengers westwards, but none the less some migrants may initially have returned in the cabin. However by the mid-and late 1850s fares on the sailing ships had fallen to such an extent that many migrants returned in sailing ship steerage. In the late 1850s and early 1860s, the steamships also moved down market and began to offer steerage passage to return migrants. The increase in return migration therefore started well before the steamship era had properly begun. See Killick, ‘Transatlantic fares…’
One of the great problems of studying return migration is that unlike the migration to the United States which can be documented in great detail in the passenger lists completed by captains on arrival at American ports, there were no equivalent leaving passenger lists from America before 1860, and no arrival passenger lists in Britain until the late 19th century. However the British Emigration Commissioners did keep records of the return migration to Britain from America (and Australia) which reached substantial numbers in the late 1850s because of the American depression. Consequently as immigration into the United States fell from 1854 onwards, so the percentage of migrants returning increased from 9% in 1854 to 55% in 1861, when only 43,000 migrants from Britain entered the country See Killick, Transatlantic fares…
Table 1: Return Migration from the United States, 1854-1861.
1854: 138361858: 16957
1855: 165621859: 13268
1856: (14154)1860: 17798
1857: 154151861: 23838
Source: Killick ‘Transatlantic Fares.’; Reports of the Emigration Commissioners, 1855-62.
The Emigration Commissioners and other commentators noted that the return migrants from America were poorer than from Australia since few travelled in the cabin, and suggested reasons such as the Crimean War, the Know Nothing anti migrant political pressure in 1855, unemployment and the US commercial crisis in 1857, faster steam travel and fear of the US Civil War. See Killick, ‘Transatlantic Fares’, p.18
3. RETURN MIGRATION BEFORE 1850:
The possibility that immigration might have been higher than normally understood before 1855 comes initially from study of the Cope packets line records in the Historical Society Pennsylvania. (Henceforth HSP) The Cope packet line was started in 1822 by Thomas Cope with an agreement with Alexander Brown of Baltimore to run four ships as packets from Philadelphia to Liverpool. The Cope records are very complete and include unique eastward passenger lists for all their voyages between 1822 and 1870. Over this period they carried about 20,000 passengers eastwards compared to 70,000 passengers westwards. However more interestingly – see Figure 1 - these proportions were not uniform through time and in the 1820s and 1830s they carried more passengers eastwards than westwards. The much larger numbers of westward passengers appeared with the massive Irish migration of the mid-1840s and early 1850s. Killick, ‘Cope Line…’
Figure 2: Cope Passenger Numbers per Year.
Source: Killick, ‘Transatlantic Fares…’
Clearly Copes’ figures was representative of Philadelphia migration as a whole which is obviously positive. The anomaly - see Table 2 below – is because Copes only carried a small proportion of the British westward emigration to Philadelphia especially in the 1820s and 1830s. In fact most British migrants to Philadelphia in the 1820s and 1830s were carried by transient ships, which could offer a cheaper service, rather than by the first class packets. In the late 1850s Copes share rose much higher as the British passenger acts excluded low quality ships from the migrant trade. Copes secured a much larger share of the eastward trade because many of the transient ships returned to Britain with cargo rather than passengers, often via southern United States ports where they loaded with tobacco or cotton. The last two rows of the table compare Copes’ eastwards trade with their westwards trade which is especially high in the 1820s and 30s, and their eastwards trade compared to total British migration to Philadelphia which is substantial in the 1820s and 1830s. See Williams, ‘The Shipping of the North Atlantic Cotton Trade’; US Customs ship arrival and departure records in National Archives, Philadelphia.
Table 2: Copes’ share of British Emigrants to Philadelphia, 1822-60
Source: Killick, ‘Transatlantic Fares…’
Figure 3 below contains three quite striking hypothetical estimates, using overlapping sources, of return migration as a per cent of immigration. The latter part of this article assesses if this could be reasonable. According to these estimates, return migration peaked in the early 1820s, episodically from the mid-1830s to the early 1840s, and in the late 1850s. It is important to note that although the return migration proportion may have been surprisingly high in the early 1820s, the actual numbers returning then were very small compared to the numbers in the late 1850s. See Figure 10 below.
Figure 3: Estimates of British Return Migrants as % British Immigrants.
Sources: Albion, Square Riggers; Cutler, Queens…; Killick, ‘Transatlantic Fares…
The three estimates are all derived from Cope returns for eastward migration, but are compared to different total figures. The first compares Cope’s eastward passengers with total British immigration to Philadelphia. The second and third compare Copes’ eastward passengers multiplied by estimates of their share of the total United States packet fleet, as a per cent of total westward migration. The second estimate is based on Robert G. Albion’s lists of New York packet ships in his classic book Square Riggers on Schedule, from 1822 to 1860, with an allowance for the packet ships belonging to the other ports. The third is based on Carl Cutler’s estimates of the number of American packet ships in his survey, Queens of the Western Ocean which lists the packet lines in all the leading ports. See Figure 4 below.
Figure 4: Estimates of US Packet Ship Numbers, 1815-60
Source: Cope eastward passenger lists; Albion, Square Riggers; Cutler, Queens,
Figure 4 estimates the total number of American packet ships carrying steerage passengers eastwards between 1820 and 1860. The first packet line carrying passengers, the Black Ball line of New York, was formed in 1817, but at that time most transatlantic passengers, both cabin and steerage, sailed in ‘regular’ or transient ships. Cope for instance, began sending his first ship, the Lancaster, to Liverpool as soon as the war ended, but initially only as a transient, sailing only when he had assembled a cargo, but usually crossing the Atlantic twice yearly. He built a second ship the Tuscarora in 1819, and then joined Alexander Brown and Sons of Baltimore to form a fully organized four ship packet line, departing on schedule, in 1822. Several New York lines also began in 1822. Thereafter it is reasonable to assume that until the mid-1850s the packet lines probably carried most eastward passengers, but from about 1855 steamships began to cut into their business. Albion contains a careful list of all the packet ships based in New York, but does not include lists for the other cities. Cutler, on the other hand, lists the lines in New York, Boston, Philadelphia and Baltimore with some indication of ship numbers, but does not give a complete list. See Killick, ‘Nineteenth Century Shipping Line…’
From 1822 to 1845 most ships sailed from New York, and Boston, Philadelphia and Baltimore did not add much to the total. Albion lists 22 ships in 1823, 32 ships in 1840, and 36 ships in 1845. The other ports added six ships – four were Copes’ – in 1823, and ten in 1845. Copes ran four packet ships through the whole period, and so between 1823 and 1845 their share of the total declined from about a seventh to about a ninth. Cutler’s estimates are rather higher. Occasionally, Copes also-ran one to three older ships as reserves adding to their total, but this was usually because one of their main ships was delayed or was being repaired, and was not able to sail on time. So it is reasonable to assume that their use of the older ships did not add much to their total. The New York, Boston and Baltimore lines also sometimes use older ships occasionally. It is also possible that some eastward passengers returned on transient cargo ships which were still often sailing from the East Coast ports- but this would have the effect of increasing the return ratio.
From 1845 it becomes increasingly difficult to estimate the total number of ships involved because, once the Irish migration began in earnest, new lower class ships crowded into the business. See Figure 4. Hence many lines listed in Cutler – especially to Boston – were in fact only ‘immigrant lines’ - that is sailing nominally as highly advertised ‘packets’ from Liverpool – but then reverting to cargo ships for their return either via the South with cotton, tobacco, etc. or from the St Lawrence with timber. This vagueness however matters less then might have been expected because the British Emigration Commissioners’ reports mentioned above, provide an 1850s anchor for the total number of eastward passengers, which then can be compared to Copes lists. See Cutler, Queens: Williams, ’Shipping of North Atlantic cotton..’
Hence in 1840, Copes ships were about one ninth of the total fleet. By 1854 the Commissioners reports suggests they were about one tenth of the total fleet, and in 1857 one fifteenth. Thereafter the ratio changes more rapidly partly due to Copes’ declining share of the sailing ship passenger business, but also because more passengers were returning by steam. So in 1861 Copes only took one thirtieth of the total number of eastwards steerage passengers, but – on Albion’s cautious criteria -appear to have about one thirteenth of the US sailing packet fleet. Therefore one can make reasonable approximations for the total number of eastward passengers between 1845 and 1853 when the total number of packet ships is not so clear. In 1854 and 1855 Copes took exceptionally large numbers back - respectively one twelfth and one tenth of the total eastward movement. It is suggested therefore in the early 1850s they took back about one fourteenth of the total, and in the mid-and late 40s respectively between one tenth and one thirteenth of the total.
Figure 3 reveals three periods of high return immigration – in the early 1820s, the mid-1830s to 1844, and the late 1850s as already discussed. The figures before 1822 are misleading. Thomas Cope did not begin a comprehensive passenger service with four ships in conjunction with Alexander Brown until mid-1822. Figure 3 therefore understates the return migration from 1819 to 1822.The trade cycle history of the period and narrative accounts of immigration suggest that the post-war boom broke in 1819, and therefore there was likely to have been a high return migration in 1819, 1820 and 1821. British migration to the east coast ports fell from about 6700 in 1820 to 4000 in 1822. Total emigration to Philadelphia fell from about 2000 in 1822 to 460 in 1823, the British share being approximately 80% of that, and these figures produce the high return ratio in 1823. British migration to Philadelphia rose again to 1300 in 1824 thereby reducing the return ratio more than changes in Cope’s passenger numbers which remained at about 250 through the late 1820s. There is a similar phenomenon in the late 1850s when the falling immigration figures combine with rising return figures to temporarily produce very high ratios, but overall Copes eastward steerage was about 12% of British immigration to Philadelphia. See Table 1 last row.
During the 1830s and 1840s, Copes’ eastward passengers peaked in 1830, 1833 and 1834, 1837 and 1838, and 1843-6. The three return proportions are generally well correlated. The two – Albion + Cutler - Cope: US return proportions - peaked - that is were 20% or over - in 1830, 1833, 1838, and 1843, but the Cope: Philadelphia return proportion also peaked in 1835 – see Figure 3 – suggesting local factors were at work. There followed the long dip in all three return proportions in the late 1840s when the massive increase in Irish emigration swamped the eastward movement.