Which Irish men and women immigrated to the United States during the Great Famine migration of 1846–54?

Abstract Despite the extensive scholarly literature on both the Great Famine in Ireland and the Famine immigration to the United States, little is known about precisely which Irish men and women emigrated from Ireland in the Famine era. This article makes use of a new dataset comprised of 18,000 Famine-era emigrants (2 per cent of the total) who landed at the port of New York from 1846 to 1854 and whose ship manifests list their Irish county of origin. The data is used to estimate the number of emigrants from each county in Ireland who arrived in New York during the Famine era. Because three-quarters of all Irish immigrants intending to settle in the United States took ships to New York, this dataset provides the best means available for estimating the origins of the United States’s Famine immigrants. The authors find that while the largest number of Irish immigrants came from some of Ireland’s most populous counties, such as Cork, Galway, and Tipperary, surprisingly large numbers also originated in Counties Cavan, Meath, Dublin, and Queen’s County, places not usually associated with the highest levels of emigration. The data also indicates that the overall level of emigration in the Famine years was significantly higher than scholars have previously understood.

James was the fourth member of his nuclear family to immigrate to New Yorkhis older sister Ellen, who he lived with initially, had arrived in New York in 1845, and his younger brothers John and William had made the journey in 1850. Like most New Yorkers of his day, James opened a bank account, a fact documented in the Emigrant Savings Bank records preserved at the New York Public Library. By 1859, he had saved $103.45, the equivalent of about $2,700 today. 1 But there is one way in which the documentation of Garvin's immigration saga is unusual. While the vast majority of ship manifests from this era housed in the U.S. National Archives merely list each Famine refugee as a native of Ireland, the Princeton's passenger roll records in precisely which Irish county he and the other 268 Irish immigrants travelling with him were born. Garvin and seventy-two of his fellow passengers were natives of County Cork, Ireland's most populous county. Twenty-eight were from County Tipperary, twenty-six from County Cavan, and fifteen from County Dublin. In all, twenty-three of Ireland's thirty-two counties were represented among the Princeton's 269 Irish-born passengers. The surprisingly detailed birthplace data recorded in the Princeton's manifest and dozens of others like it provide an opportunity to answer a question that has tantalized historians for more than a century: exactly where in Ireland did the United States's Famine immigrants originate?
Despite the vast literature on the Great Famine in Ireland and the Famine immigration to the United States, few scholars have attempted to determine precisely who left Ireland in the Famine era. Those studying particular Irish-American communities have noted when such places had a disproportionate number of immigrants from a specific Irish locale. Cleveland, for example, had an unusually large number of immigrants from County Mayo; in Savannah, immigrants from County Wexford predominated; and refugees from Counties Sligo and Kerry were especially numerous in New York's Five Points neighbourhood. 2 But for the most part, historians of the Irish-American experience, recognizing that 'hard evidence is scarce', have not been able to determine the origins of the Famine immigrants in any detail. As a result, the scholarship on Irish-American immigration is largely silent on this question. 3 1 Manifest of the Princeton, 11 June, 1852, (New York passenger lists, 1820-1957 Group 36, U.S. National Archives, accessed via Ancestry.com); family 506, 6th district of the Fourteenth Ward, 1855 New York state census (Old Records Division, New York County Clerk's Office, accessed via familysearch.org); test book and deposit ledgers for accounts 12,174,17,871,38,131,and 46,028,Emigrant Savings Bank (Emigrant Savings Bank papers, New York Public Library, accessed via Ancestry.com).
2 Nelson J. Callahan and William F. Hickey, The Irish Americans and their communities of Cleveland (Cleveland, 1978), p. 176; Edward M. Shoemaker, 'Strangers and citizens: the Irish immigrant community of Savannah, 1837-1861' (Ph.D. thesis, Emory University, 1990), p. 106; Tyler Anbinder, Five Points: the nineteenth-century New York neighborhood that created tap dance, stole elections, and became the world's most notorious slum (New York, 2001), p. 98. 3 Quotation from Cormac Ó Gráda, Black '47 and beyond: the Great Irish Famine in history, economy, and memory (Princeton, 1999), p. 228. Among the most important works on the Famine Irish in America are Oscar Handlin, Boston's immigrants, 1790-1880: a study in acculturation (Cambridge, 1941); Robert Ernst, Immigrant life in New York City: 1825-1863(1949reprint ed., Syracuse, 1994); Hasia R. Diner, Erin's daughters in America: Irish immigrant women in the nineteenth century (Baltimore, 1983); There is an equally large historical literature on the Famine itself and many of these works discuss emigration. Many are micro-histories that study the Famine in a single location but say little about their subjects once they board the emigrant ships. Macro-studies of the Famine sometimes devote a bit more attention to the flight to America, but rarely say more than that the Famine emigrants came largely from northern, eastern, and central Ireland in the early years of the Famine, while in the later years more emigrants originated in the south and west. 4 When these authors do consider which parts of Ireland sent out the most emigrants during the Famine, reliance is often placed on the half-century-old work of English geographer S. H. Cousens. To determine the flow of emigration from each Irish county, Cousens compared the population of each county listed in the 1841 census to that in 1851 (when the population of every county except Dublin had declined due to the Famine). Then Cousens examined government reports of famine-related deaths in each county. He attributed population decline greater than the reported famine-related deaths to emigration, and in that manner estimated the emigration 'rate' from each Irish county (not the actual number of emigrants, but the proportion of the county's population decline that seemed attributable to emigration). Since its publication in 1960, Cousens's work has been used by almost every historian seeking to quantify regional variation in the Famine emigration from Ireland. 5 J. Matthew Gallman, Receiving Erin's children: Philadelphia, Liverpool, and the Irish famine migration, 1845-1855(Chapel Hill, 2000; David T. Gleeson, The Irish in the South, 1815-1877(Chapel Hill, 2001. 4 William Forbes Adams, Ireland and Irish emigrants to the New World from 1815 to the Famine (New Haven, 1932); Arnold Schrier, Ireland and the American emigration 1850-1900(Minneapolis, 1958; Cecil Woodham-Smith, The Great Hunger: Ireland 1845(New York, 1962; Joel Mokyr, Why Ireland starved: an analytical and quantitative study of Irish poverty, 1800-1851(Boston, 1983; Christine Kinealy, This great calamity: the Irish Famine, 1845-52 (Boulder, 1995; Robert Scally, The end of hidden Ireland: rebellion, famine, and emigration (New York, 1995); Ciarán Ó Murchadha, Sable wings over the land: Ennis, County Clare, and its wider community during the Great Famine (Ennis, 1998); Gerard Moran, Sending out Ireland's poor: assisted emigration to North America in the 19th century (Portland, 2004); Cormac Ó Gráda, Ireland's Great Famine: interdisciplinary perspectives (Dublin, 2006); Mary Lee Dunn, Ballykilcline rising: from Famine Ireland to immigrant America (Amherst, 2008); Ciarán Ó Murchadha, The Great Famine: Ireland's agony 1845-1852(New York, 2011. For an example of the observation that Irish emigration originated more in the north and east early on and more in the south and west toward the end of the Famine migration, see David Fitzpatrick, 'Emigration, 1801-70', in W. E. Vaughan (ed.) 6 Beginning in May 1851, the British government tracked emigration from Ireland at the county level, and many scholars, including Fitzpatrick, Ó Gráda, and Kerby Miller, make extensive use of this data in their work. But given that we do not know which Irish emigrants relocated to the United States (the remainder moved primarily to Great Britain, Canada, and Australia), we cannot be sure that the Irish emigration from 1851 onward matched the Irish-American immigration during the entirety of the Famine years. As a result, the Irish emigration data that commences in mid-1851 does not provide answers to the questions that historians of the Irish-American immigration experience are seeking. 7 Some scholars have previously examined ship passenger lists in an attempt to better understand the demographics of the Famine immigration. But they focused on age, gender, and occupations and did not address the birthplaces of the Irish immigrants. 8 Economic historians and those interested in the history of trans-Atlantic shipping have also made some profitable use of ship manifest data, but they have also ignored the question of the origins of America's Irish immigrants. 9 Attempts to infer the geographic origins of the United States's  , xxix (1975), pp 143-9. 7 David Fitzpatrick has calculated that only about 40 per cent of the Irish who lived outside of their county of birth in 1851 were residing in the United States. Another 20 per cent, he estimates, were living in other parts of Ireland, and the remaining 40 per cent had relocated to England, Canada, or Australia: David Fitzpatrick, Irish emigration, 1801-1921(Dublin, 1984, p. 6. But by 1851, approximately three-quarters of Irish emigrants were relocating to the U.S., though the proportion could vary significantly from year to year: Miller, Emigrants and exiles, pp 570-9; Ó Gráda, 'Note on nineteenth-century Irish emigration statistics ', pp 143-4. 8 Charlotte Erickson, 'The uses of passenger lists for the study of British and Irish immigration' in Ira A. Glazier and Luigi De Rosa (eds), Migration across time and nations: population mobility in historical contexts (New York, 1986), pp 318-35; Deirdre Mageean, 'Ulster emigration to Philadelphia, 1847-1865: a preliminary analysis using passenger lists', ibid., pp 276-86; Cormac Ó Gráda, 'Across the briny ocean: some thoughts on Irish emigration to America, 1800-1850', ibid., pp 79-94; Ira A. Glazier, Deirdre Mageean, and Barbara Okeke, 'Socio-demographic characteristics of Irish immigrants, 1846-1851' in Klaus Friedland (ed.), Maritime aspects of migration (Cologne, 1989), pp Miller,Emigrants and exiles,p. 295. 9 Terry Coleman, Passage to America: a history of emigrants from Great Britain and Ireland to America in the mid-nineteenth century (London, 1972); Robert Scally, 'Liverpool ships and Irish emigrants in the age of sail' in Journal of Social History, xvii Irish immigrants using the records of a single New York savings bank or the missing persons advertisements in a single Boston newspaper are admirable but limited, and exemplify the desperate situation scholars face in trying to determine where in Ireland the American Famine immigrants originated. 10 Believing that manifests such as the one that recorded the arrival of James Garvin could help create a better estimate of the Irish county origins of the Famine immigrants than any published previously, a sample of Irish immigrants arriving in New York from 1846 to 1854 was constructed from the passenger lists that provide such information. 11 The year 1846 was chosen as the starting point because it is the first year in which observers noted an increase in immigration due to the failure of the potato crop in Ireland. The year 1854 was selected as an end date because after that year immigration plummeted to pre-Famine levels, in part because of a recession in the United States and in part because the rise of the anti-immigrant Know Nothing movement made the United States seem like a less desirable destination. 12 These passenger lists were compiled not by American immigration officials but, as in Garvin's case, by an officer on each emigrant ship and then submitted to customs officials upon arrival.
This sample has several limitations which must be acknowledged at the outset. First, the study was limited to the port of New York, where 74 per cent of Irish immigrants to the United States landed from 1846 to 1854. 13 Relatively few manifests from the other ports are extant in American archives, and very few of these list the Irish county of origin, but it must be noted that the findings might differ somewhat had every American port been included. In particular, not examining immigrants who arrived in Philadelphia probably means a slight undercounting of American immigrants from Ulster, since emigrants from that part of Ireland found Pennsylvania an especially attractive destination. Because ship traffic from Belfast and Londonderry to Philadelphia decreased sharply after 1847, and virtually disappeared after 1851, there was not much of a direct Ulster to Philadelphia migration during the Famine years. But even ships sailing from Liverpool to Philadelphia carried significantly more Ulster emigrants than the Liverpool ships headed for New York. 14 The number of Irish who arrived in this period at each 10 Cormac Ó Gráda, 'The Famine, the New York Irish, and their bank' in Ó Gráda, Ireland's Great Famine, pp 182-3; David Noel Doyle, 'The remaking of Irish America, 1845-1880' in J. J. Lee and Marion R. Casey (eds), Making the Irish American: the history and heritage of the Irish in the United States (New York, 2006), pp 219-21, 245 n. 21. 11 Accessed via Ancestry.com. A total of sixty-nine New York ship manifests for these years provide Irish immigrants' county of birth, and these are listed in Appendix One. Of the sixty-nine manifests, sixty-three are for ships sailing from Liverpool to New York. 12  13 For the 74 per cent figure, see Table One. 14 On the two ships sailing from Liverpool to Philadelphia whose manifests listed the emigrants' county of origin, slightly more than a third of the passengers originated in the six northernmost Irish counties -Antrim, Armagh, Derry, Donegal, Down, and Tyrone. In contrast, only 13 per cent of the Irish emigrants on ships sailing from Liverpool to New York originated in these counties. For the Philadelphia significant American immigration port is difficult to determine because most of the figures can be obtained only in difficult-to-find quarterly customs reports. As a result, they have not been previously published, but they indicate the dominance of the port of New York in American immigration, even fifty years before the opening of Ellis Island. 15 A second limitation of our figures is that they do not account for Irish immigrants who came to the United States via Canadaat the beginning of the Famine, Canadian officials estimated that as many as one-half of the Irish arriving there moved on to the United States. 16 These Irish emigrants, if they Source: see appendix one.
emigrants, see the manifests of the Clara Wheeler, 15 July, 1850, and the George Green, 20 Aug., 1850, ('Philadelphia passenger lists', Ancestry.com). 15 These five major ports account for well over 99 per cent of Irish immigrants landing in the United States on trans-Atlantic vessels. In the twelve months ending on 30 September 1850, for example, 335 Irish immigrants landed at Portland, Norfolk, Charleston, and Savannah. Another 1,400 immigrants of all nationalities landed in those twelve months at Mobile and Galveston, but it is unlikely that more than a few hundred were Irish immigrants who had not previously landed at an East Coast port. The same holds true for the 10,000 or so non-citizens who arrived in San Francisco in that period. Many were probably Irish, but few would have sailed there directly from a non-American port. See Passengers arriving in the United States: letter from the secretary of state transmitting a statement of the number and designation of persons arriving in the United States, finished their journey by sea, were labelled on American manifests as originating in British North America rather than Ireland, and if they came by land (as the majority did) their crossings were not recorded at all. But after 1847, when Canadians passed laws to discourage the immigration of Irish 'paupers', about 85 per cent of Irish emigrants heading to North America sailed directly to U.S. ports, so the lack of Canadian data becomes less of a problem after that date. 17 Finally, we must note that only 2 per cent of the arriving Irish immigrants listed their Irish birthplace informationa larger sample would have been better. But given that 916,000 Irish immigrants arrived at the port of New York in the nine years under consideration, our sample of 18,000 Irish passengers is sound, if not perfect. 18 Despite its limitations, this sample allows us to analyze the sources of Irish immigrants to the United States far more accurately than any previous study. 19 The information from the manifests provides a number of insights into the Famine immigration. First and foremost, it enables us to calculate fairly reliable estimates of the number of immigrants from each Irish county who arrived in New York from 1846 to 1854.
It is not surprising that County Cork, by far the most populous Irish county, sent the most immigrants to the United States in the Famine years, nor that Galway and Tipperary, the second and third most populous Irish counties, also appear near the top of the list. But other counties that rank near the top are somewhat surprising. Cavan, not a particularly populous county, ranks . 19 The entire data set can be found in Excel format at https://dataverse.harvard.edu/ dataverse/anbinder, the website of the Harvard University Institute for Quantitative Social Science. Information derived from this sample will be cited below as 'Irish emigrants manifest database'. fourth in emigrants sent to New Yorka fifth of its pre-Famine population sailed to Manhattan in the Famine years. Little has been written about Cavan in the Famine era, so determining why so many of its inhabitants left is difficult. Its inhabitants were not the most prosperous in Ireland, but neither were they among the very poorest. Perhaps because the county was so densely Source and methodology: see appendices. The populations given for each county do not equal the total given for all of Ireland due to rounding. populated, even by Irish standards, there was also a strong tradition of seasonal labour migration in Cavan. Thousands temporarily left the county each year, travelling to England in search of work, but returning in autumn to harvest their potatoes. When the potato blight struck, Cavan was not one of the very worst hit areas, yet neither was it one of the eastern and north-eastern counties in which suffering and death were below average. It seems likely that the people of Cavan were hit hard enough by the Famine to want to emigrate, but not so badly affected that they lacked the capacity to do so. As a result, once the potato blight struck in 1845 and persisted in subsequent years, residents of Cavan once again walked to Dublin and boarded ships for Liverpool. But now, instead of dispersing across England, they transferred to the trans-Atlantic sailing ships that left Liverpool almost daily for New York. These same factors probably help explain the larger-than-expected emigration from Meath and Queen's County as well. 20 Perhaps the most surprising county near the top of our list is Dublin. Because Dublin was the only one of Ireland's thirty-two counties to gain population during the Famine (due to in-migration from other parts of the island), Cousens and scholars drawing on his work could not detect any emigration from there at all. As a result, Dublin is rarely mentioned in the Famine literature as a major source of American immigrants, even though the proximity of its residents to Dublin's busy port made emigration from there relatively easy. Yet the American ship manifests prove that County Dublin sent large numbers of Irish immigrants to the United States. By the end of 1854, about 45,000 residents of Dublin, one-eighth of its pre-Famine population, had immigrated to the United States via New York. 21 One might wonder if the high number of immigrants recorded in the manifests from Dublin could be accurate given Dublin's population increase during the Famine years. Perhaps ship passengers told officials where they had last lived rather than where they were born. But the evidence suggests that the manifest figures indicating high levels of emigration of Dublin natives are not erroneous. At the Irish-run Emigrant Savings Bank that opened in 1850 in New York, bank employees asked all Irish-born depositors specifically for their parish and county of birth. County Dublin ranked sixth out of Ireland's thirty-two counties in terms of Irish customer birthplaces, just as it ranks sixth in the manifests. It should not surprise us that the fifth most populous Irish county, one with a major seaport, sent the sixth most emigrants to New York. 22 Equally noteworthy are the populous counties that did not rank as high in immigrants sent to New York as they did in total population. Counties Down and Antrim, which converge at the city of Belfast and ranked sixth and seventh in population in 1841, were twenty-sixth and seventeenth respectively in the number of immigrants sent to New York. This fact may stem from the relative prosperity of those counties, perhaps giving their residents less reason Map One: Irish immigrants landing in New York, 1846-1854 by county of birth compiled by Tyler Anbinder in collaboration with economic historians Cormac Ó Gráda and Simone Wegge. A 5 per cent sample of this data is currently available at https://dataverse.harvard.edu/dataverse/anbinder and the entire data set will be placed in the same location after work on it is completed. to emigrate. Emigrants from these north-east counties may also have chosen to relocate to other places including Great Britain or Canada. The Ulster Irish as a whole definitely preferred Pennsylvania to New York, and this factor may account for the low immigration rate to New York from these counties. Many took ships from Belfast directly to Philadelphia. The relative dearth of emigrants from another populous county, Mayo, however, cannot be ascribed to the same causes. 23 The Irish immigration varied not only by place but also over time. It has long been understood that the north and east of Ireland sent more immigrants to North America in the early years of the Famine, while the south and west came to dominate the outflow by about 1851. Our data in some ways confirms these long-held views about the differences between the early and later Famine emigration, albeit with some significant wrinkles. The most striking thing about Tables Two and Three, in our opinion, is less the variation between the two periods, which is usually emphasized in the historical literature on the Famine, and more the striking continuities. Seven of the ten leading sources of Irish immigration in the early years of the Famine were also the leading sources of immigrants in its later years. Cavan, Meath, and Roscommon in the northern midlands, Cork, Tipperary, Limerick, and Galway in the south and west, Dublin in the east, and Donegal in the northwest sent large numbers of immigrants to the United States throughout the Famine years. Only Tyrone in the north fell off the top-ten list from the early to the late years, replaced by Kerry in the southwest. But overall, in terms of Irish county of origin, continuity was more significant than change over the course of the Irish Famine immigration. 24 Comparing Table Four to the official data on Irish emigration that the British government began compiling on May 1, 1851 indicates that our estimates of Irish immigration to New York are fairly accurate. The British records and our manifest data both suggest that the same five counties -Cork, Tipperary, Limerick, Galway, and Kerrysent the most emigrants out of Ireland from 1851 to 1854. But further down the list, our rankings differ from those of port officials tallying all Irish emigration, something to be expected when the difference from county to county is only a few hundred people per year. Both lists also place Cavan and Donegal in the top ten, but while the New York manifests indicate that Dublin, Meath, and Roscommon rounded out the top ten, the British figures for Irish emigration from 1851 to 1854 place Clare, Waterford, and Kilkenny ahead of those three. In the case of Kilkenny, the British government ranked it eighth in emigration from 1851 to 1854 and we ranked it eleventh in New Yorka fairly insignificant difference. The remaining, more significant discrepancies in the top ten could have resulted from the fact that emigrants from Clare and Waterford were less likely than others to choose New York as their new home. Indeed, American data indicates that during the 1850s, there were far more Irish immigrants from Cavan, Dublin, Meath, and Roscommon living in New York City than there were from Clare and Waterford. This suggests that the differences between our findings and the emigration figures the British collected beginning in May of 1851 result not from any problem with our sample or methods, but from the fact that emigrants from different parts of Ireland did not go to New York in equal proportions. The more significant differences between our findings in Table Two and the government emigration data strongly suggest that the  government figures for 1851 onward are not the best source for understanding the sources of the entire Famine emigration, which began in 1846. 25 The data from the manifests helps us to understand not only the total number of New York Famine immigrants who originated in each Irish county, but also provides insight into the proportion of each Irish county's population that was lost to emigration. Twenty or more per cent of the residents of Counties Meath and Cavan left Ireland for New York during the Famine years. King's County, Limerick, Longford, Louth, Queen's County, Roscommon and Tipperary each sent from 15 to 19 per cent of its pre-Famine population to America via the port of New York. Given the fact that more than one quarter of all Irish immigrants in these years landed at other American ports, and still more went to Canada, England, and Australia, the total percentage of residents leaving these districts was significantly higher. The counties that saw the smallest proportion of their populace immigrate to New York were Down, Wexford, Antrim, and Mayo.
Mapping the percentage of each Irish county's pre-Famine population that emigrated to New York brings some of the most striking features of Ireland's Famine-era emigration into sharper focus. The counties that sent the highest proportion of their populations to the port of New York are concentrated in a belt of midland counties stretching from south Ulster (Cavan) to north Munster (Limerick and Tipperary). While populous coastal counties such as Cork, Dublin, and Galway may have sent the largest numbers of emigrants to the United States, these were by no means the counties that experienced the greatest proportion of population loss as a result of the Famine emigration. In fact, what is perhaps most striking when measuring emigration to New York as a proportion of each county's population is that nearly all the coastal counties rank in the bottom third of the list. Examining the emigration in this manner also reminds us that low-population counties such as Kildare, Longford, and Westmeath, whose Famine-era emigrants made up a tiny fraction of the passengers arriving in New York, nonetheless experienced tremendous depopulation in proportional terms.
Comparison of Map Two with Cousens's map of emigration from 1846 to 1851 shows that he too identified central Ireland as the location of the greatest proportional emigration. But in other ways, his findings and those presented here differ significantly. He deduced that there was far more emigration from Fermanagh, Mayo, Sligo, and Wicklow than can be documented in the American passenger manifests. Conversely, he underestimated emigration from Donegal, Kerry, and Waterfordand grossly underestimated the emigration of Dubliners. In the case of Dublin, his mistake stems from the 25  limitations of his methodology. In the other instances, his figures may differ from those presented here because his method did not allow him to consider the significant Famine-related emigration in 1852, 1853, and 1854. It is also possible that in cases where he estimated more emigration than the American manifests document, residents of these counties might have chosen to immigrate to destinations other than New York. But the data presented here, which is based on documented immigrants rather than deduction from unreliable Famine death records, and covers the entirety of the Famine-era emigration, can be considered to present a more reliable set of figures. 26 Even more importantly, our study suggests that Cousens significantly underestimated the overall extent of the Irish Famine emigration. Cousens estimated that the highest county-level rate of emigration from 1846 to 1851 was 17.5 per cent, for County Roscommon. But the ship manifests indicate that three other counties sent more than 17.5 per cent of their pre-Famine populations to New York City alone. For every three emigrants who sailed to New York, another landed at a different American port and still another relocated to Canada, Australia, or England. As a result, the emigration rate from Ireland in the Famine era must have been significantly higher than Cousens led us to believe. 27 Most of the manifests that contain Irish counties of origin also list where in the United States the immigrants intended to settle, information that sheds further light on the immigration experience in these years. Immigrants' intended destinations were just as surprising as their origins. Seventy-one per cent of the Irish immigrants landing in New York stated that they planned to make New York State their home, a figure much higher than was expected. The next most popular destinations were Pennsylvania (7 per cent of the Irish arrivals), Massachusetts (5 per cent), Ohio (3 per cent), as well as Rhode Island, Canada, Connecticut, and New Jersey (all 2 per cent). Another 2 per cent reported that they intended to settle in the slave states (the bulk choosing Virginia, Maryland, and Georgia). Sixty-three per cent of the Irish immigrants specified that they would settle in New York City. 28 We know, of course, that many of the Irish immigrants who initially lived in New York City eventually moved elsewhere. Already the most populous city in the nation in these years, New York would have had twice as many residents had all the Irish and German immigrants who listed that city as their destination actually remained in the metropolis.
A number of factors may explain these unexpected intended-settlement figures. It is possible that the immigrants did not understand the question, and thought they were being asked where in America they were landing, or that the person recording the answers was lazy, and filled in 'New York' for those whose answers he could not obtain or understand. Yet the data in this manifest column was only analysed if the variety of answers indicated that the passengers understood the question and that the compiler of the manifest was diligent in recording the answers, so it does not seem that misunderstanding or laziness could account for most of these responses. Instead, it would appear 27 That Cousens underestimated Irish emigration is not a new observation: Cormac Ó Gráda, 'Some aspects of nineteenth-century Irish emigration', in L. M. Cullen and T. C. Smout, Comparative aspects of Scottish and Irish economic and social history, 1600-1900(Edinburgh, 1977, pp 70-1. 28 Irish emigrants manifest database. Some manifests clearly differentiated between New York City and State, but most merely said 'New York'. In those cases we used the following rules to decide whether or not New York meant the city or merely the state: 1) if the manifest listed five or more cities other than New York in the intendeddestination column, then we interpreted 'New York' to mean New York City because the compiler of the manifest seemed intent on recording cities; 2) if a manifest listed fewer than five cities other than New York, then we did not infer that 'New York' referred to New York City because the manifest compiler had not exhibited much of an effort to record destinations beyond the state level; 3) if the compiler of the manifest listed a city as the destination for 45 per cent or more of the passengers not intending to settle in New York, then 'New York' was interpreted to mean the city of New York.
that the intended-settlement responses indicate that for any number of reasons (lack of money, exhaustion, health, economic opportunity), the majority of Irish immigrants landing in New York did initially intend to live there. It is possible that even if these immigrants planned to eventually move elsewhere, they may have believed that New York was the best place to start life in America. On the other hand, it is possible that the decision to leave New York was only made after the immigrants had already lived in the United States and learned more about the nation and the geographic variation in employment opportunities. Whatever the case might be, while most Irish immigrants did eventually decide that it would be preferable to leave New York, the fact that so many initially intended to settle there has not been previously understood. 29 An immigrant's county of origin in Ireland also affected, to some extent, his or her intended place of settlement in the United States. Immigrants arriving in New York from the five north-eastern counties of Ulster (Armagh, Antrim, Down, Derry, and Tyrone) were nearly three times more likely to continue on to Pennsylvania than other Irish immigrants. It is not clear whether these immigrants moved to Pennsylvania because they could more easily afford to do so, because they considered Pennsylvania more welcoming to Ulster immigrants, or because they sought to join Ulster immigrants who had settled in Pennsylvania before the Famine. All three factors may have been involved. We do know that there were longstanding trade connections between Belfast, Londonderry, and Philadelphia, and that the transportation of large numbers of immigrants from these Irish ports to Philadelphia predated the Famine. 30 The manifests also reveal that a disproportionately large number of the immigrants from the northern half of Ireland (not Ulster alone) settled in Illinois. Again, it is unclear whether this pattern was a result of the financial ability of these immigrants to continue their journeys to the west or if other factors pulled them there. Other than these two correlations, there was not significant variation in settlement patterns among the Famine Irish when controlling for county of birth. 31 The Irish occupations of the immigrants also varied by county of birth. Immigrants from north-east Ireland were much less likely to have worked as labourers before emigrating than those from other counties. About half the male emigrants from Antrim, Armagh, Derry, and Down were labourers, while from the rest of Ireland the proportion was typically about 70 per cent, and grew as high as 85 per cent for the male immigrants from Roscommon, Cavan, and Kerry. The rest of the male emigrants were not primarily farmersonly 6 per cent reported that to be their occupationbut rather artisans. Not surprisingly, the county that sent the highest proportion of tradesmen to New York was Dublin. Forty-three per cent of immigrants arriving in New York from Dublin reported following a trade. Conversely, only 35 per cent of Dublin's adult male immigrants called themselves labourers. Many Ulster counties (Antrim, Armagh, Derry, Down, and Tyrone) also sent an aboveaverage proportion of artisans to New York, but the counties on the southern periphery of that province did not. The proportion of tradesmen among the immigrants from Cavan, for example, was not merely below the average for Ulster, but was below average for all Irish emigrants who sailed to New York. 32 For women, the only significant occupational trend was that women from Dublin and the northeast were much less likely to report working for pay than those from other parts of Ireland. Only about a third of the women from Dublin and about half from Antrim and Down stated that they had an occupation, whereas 80 per cent or more from Queen's County, Galway, Roscommon, and Cavan called themselves domestic servants of one kind or another. Age difference does not explain this disparity, as the average age of the working-age female immigrant from Dublin, Antrim, and Down (twenty-six) was only one year higher than that of the women from Cavan, Galway, Queen's County, and Roscommon. We surmise that because the emigrants from Dublin and the northeast were more likely than other emigrants to have spouses, brothers, and fathers who were business owners and tradesmen, necessity did not push these women into the paid workforce as often as those from other parts of Ireland.
The authors had anticipated conducting further analysis of the passengers' pre-emigration occupations, to look for additional county-of-origin patterns, but came to the conclusion that the occupational data was too unreliable to allow for such fine-grained analysis. Some manifest compilers recorded minute variety for place of birth but seemed to put little effort into the occupation category, labelling nearly every steerage passenger as a 'labourer' when there must have been at least some variety in their occupational backgrounds. As a result, analysis of the occupational patterns was restricted to the broad categories discussed above. With the manifest data on all 18,000 Irish emigrants available online, the opportunity exists for scholars to seek to discover other occupational patterns in the manifest data.
Of course, an immigrant did not necessarily need to follow the same occupation in the United States as he or she had in Ireland. That was one of the attractions America held for the Famine Irish. James Garvin from Castletownroche, for example, reported being a labourer when he arrived in New York in 1852, but by 1855 he had learned a trade and worked as a gas fitter, laying gas lines in the city's booming construction industry. Garvin still followed that trade in 1870, the last time he appears in American census records. 33 The manifests that document the immigration of James Garvin and more than 900,000 other Irish men, women, and children to New York during the era of the Great Famine shed significant new light on the Famine migration to the United States. First, while the Irish would eventually fan out across every American state (especially those without slavery), it appears that initially the vast majority intended to go no further than New York City. Second, a larger proportion of the residents of many Irish counties sailed to New York than has been previously understood. Nearly a quarter of the pre-Famine inhabitants of 32 Irish emigrants manifest database. 33 Irish emigrants manifest database. 'Working-age' was defined as sixteen or higher. For Garvin, in addition to the documents cited in note 1, see family 293, 1st division, 15th Ward, New York County, New York, 1860 manuscript census, and family 888, 14th district, 19th Ward, New York County, New York, 1870 manuscript census (National Archives, Washington, D.C., accessed via Ancestry.com).
County Meath and a fifth from Cavan and Queen's County took trans-Atlantic vessels to New York. One in six inhabitants of Tipperary, Louth, and Longford also made the journey. Because additional residents of these counties immigrated to the United States via Boston, Philadelphia, Baltimore, New Orleans, and Canada, while still others settled permanently in Canada, Australia, or England, the population loss due to emigration is certainly much higher than the figures provided by Cousens that continue to dominate the Famine scholarship. Third, there appears to have been less variation in the Famine emigration from the early to later years than we have previously imagined. The emigration from some northern counties, such as Cavan and Tyrone, did fall significantly as a percentage of the whole in the later Famine emigration years. But that seems less significant than the fact that nine of the ten counties that sent the most immigrants to New York in the early Famine years also made the top-ten list for the later years.
Finally, this study provides the first estimate of Irish county-level immigration to the United States based on hard data. It must be emphasized that what has been provided are only estimates, based on an imperfect two per cent sample, and they should be treated as such. But to better understand the Famine Irish immigrants' experience, more serious consideration of their county of origin is needed. When Famine immigrants met one another in New York or almost anywhere else in North America, presumably one of the first questions one newcomer would have asked the other was where in Ireland he or she was from. They must have identified to some degree with their county of birth, because we know that the Famine Irish often selfsegregated, with those from one county living in one neighbourhood (or part of a neighbourhood) and those from another living somewhere else. Other immigrant groups did the same thing, and today's American immigrants surely do so as well. 34 Understanding the Famine immigrants' origins enables us to better comprehend the effects of the Famine emigrationpart of an unprecedented flow of people out of Europe and into North Americaon both Ireland and the new world the immigrants found and remade in the United States.
United States, Thirtieth Congress, Second Session, House of Representatives, Ex. Doc. No. 10; Passengers arriving in the United States: letter from the secretary of state transmitting a statement of the number and designation of persons arriving in the United States, First Session,Ex. Doc. No. 7;Ibid.,Second Session,Ex. Doc. No. 16;Ibid.,First Session,House of Representatives,Ex. Doc. No. 100 Some points are worth noting about the recording of data at different ports, as used in Table 1. Baltimore: data for Baltimore in these sources is very unreliable. The official who provided the Baltimore figures for these reports at times seems to have made no effort to tally the manuscript manifests and seems instead to have made wild guesses. Instead of relying on these clearly erroneous figures for Baltimore, estimates have been made based on the assumption that two-thirds of the British immigrants landing there were Irish from 1844 to 1846 and that 90 per cent were Irish from 1847 to 1852 (this latter figure based on the reading of actual manifests of ships sailing to Baltimore from Liverpool). Beginning in 1853, when a new customs officer took office, the Baltimore figures seem to become reliable. New Orleans: custom house officials here did not provide data on immigration for the first quarter of 1844, the middle two quarters of 1848, and the last quarter of 1852. The number of Irish immigrants for those quarters has been estimated based on the number documented in the same quarter in the years most similar to the ones missing. From October 1850 to December 1854, New Orleans records differentiate Irish immigrants from English, Scottish and Welsh immigrants, and during that period 85 per cent of the British immigrants were Irish. For the period before October 1850 when the Irish are not differentiated from the rest of the British immigrants, we have estimated that 85 per cent of the immigrants from Great Britain were Irish. Philadelphia: these returns do not differentiate between Irish and English/ Scottish/Welsh immigrants in the fourth quarter of 1844 and the second quarter of 1845. In 1846, 88 per cent of British immigrants were Irish, and we have used that percentage to determine the Irish portion of the British immigration in those two quarters. New York: New York returns do not differentiate between Irish and other British immigrants until May 1847. For the period before that date, the proportion of Irish immigrants among the British immigrants to Boston has been used to estimate the number of New York's British immigrants who were Irish. In 1846, that was 70 per cent, so for 1846, for example, this meant multiplying 60,065 (the number of British immigrants arriving in New York) by 0.7 to arrive at an estimate that 42,046 Irish immigrants arrived in New York in 1846. Boston: Boston immigration returns do not differentiate the Irish from other British immigrants in the middle two quarters of 1844. The proportion of Irish among the British immigrants in the middle two quarters of 1845 was used to estimate the number of Boston's British immigrants who were Irish in the same two quarters in 1844.

Appendix two: methodology for county-level estimates
The estimate of county-by-county immigration figures to New York was arrived at using the following methods and sources. In order to gather the manifests for our sample, all of the 4,000 or so manifests submitted by the officers of ships that sailed from ports that typically carried Irish immigrants and that arrived in New York from 1846 to 1854 were examined. For the earliest years under consideration, not many manifests were found. So in preparing estimates of total immigration from Ireland, the sample was weighted to compensate for the relative dearth of detailed manifests from those early Famine years and for the relative abundance of such manifests from the 1850s. Care was also taken to ensure inclusion of the proper proportion of immigrants from each significant Irish port of embarkation, because one would expect that ships from a given Irish port would be more likely to carry passengers from that part of Ireland than ships from other ports. If vessels departing from Cork were under-represented, for example, the result would likely be to under-count immigrants from Cork and the counties immediately surrounding it.
With the data from these manifests in hand, the following method was used to estimate the number of immigrants from each Irish county. To determine the number and source of Irish immigrants arriving in New York from ports other than Liverpool, we looked at every extant ship manifest from those ports and counted the total number of Irish passengers (the manifests do not distinguish between immigrants and visitors; for the purposes of our study we have assumed that all Irish passengers were immigrants). For some of these ports (Belfast, Cobh, Dublin, Galway, Glasgow/Greenock, Limerick, London, Londonderry, and Newry) manifests were found that listed the Irish immigrants' counties of origin. For example, on ships sailing from Cobh that specified Irish county of birth, 72 per cent of the passengers were natives of County Cork. The total number of passengers sailing to New York from Cobh (18,299) was multiplied by .72 to estimate that 13,175 passengers sailing from Cobh to New York were born in County Cork.
For those ports for which there are no manifests that list Irish county of origin, the passengers' origins were estimated based on the patterns found for similar ports where manifests that specified county of origin are available. Surviving manifests show, for example, that ships from Tralee in County Kerry carried 4,840 passengers to New York in these nine years. Based on the patterns we found in nearby Cork and Limerick, it was estimated that 80 per cent of the Tralee passengers were natives of County Kerry, while another 10 per cent were from Limerick and the remainder was from County Cork. It was therefore estimated that 3,872 natives of County Kerry immigrated to the United States on ships from Tralee from 1846 to 1854. This method was also used to estimate the origins of immigrants sailing from the Irish ports of New Ross, Sligo, Waterford, Westport, and Wexford. County estimates for Irish ports such as Donegal, Drogheda, Killybegs, Kilrush, and Youghal that sent fewer than 500 passengers to New York in the nine years under consideration were not made because the figures were statistically insignificant. For the same reason Irish county estimates for ports in the Caribbean or South America, or for British ports other than Glasgow, Liverpool, and London were not calculated, because no basis for estimating where in Ireland these immigrants came from was available. For present purposes, it was assumed that these immigrants, just 1.2 per cent of the total, came from the same variety of counties as those from Liverpool.
Indeed, Liverpool was the port from which the vast majority of Irish immigrants travelling to New York (approximately 82 per cent) boarded their trans-Atlantic ships. The proportion embarking from that port did vary significantly from year to year, however, from 94 per cent in 1846 and 93 per cent in 1854 to as low as 68 per cent in 1849, the peak year for emigration directly from Ireland to the United States. The estimate of the number of immigrants from each Irish county who arrived via Liverpool is based on the sixty-three ships in the sample that originated in Liverpool, vessels that carried 22,943 passengers to New York (17,013 of these 22,943 were Irish). For seven of the nine years under consideration, the sample contains Irish county of birth for about 2.5 per cent of the Irish passengers who emigrated via Liverpool. But the proportion was only 1.6 per cent for 1850 and 0.5 per cent for 1847.
Because the sample is not large enough to enable us to accurately estimate county-by-county emigration via Liverpool on a year-by-year basis, the following method was adopted. The sample was used to determine the proportion of immigrants emigrating via Liverpool from each county in what was determined to be the Famine migration's early years, 1846-9 (based on the patterns seen when the county of origin from Liverpool was calculated on a year-by-year basis). The total number of Liverpool-emigrating Irish in those four years was multiplied by that proportion that the sample indicated belonged to each county to estimate the number of Irish immigrants travelling via that port to New York. For example, 8.9 per cent of all Irish immigrating to New York via Liverpool from 1846 to 1849 in the sample indicated that they were from County Cavan. The total number of Irish sailing from Liverpool to New York in those years (256,107) was therefore multiplied by 0.088637 (the precise proportion) to estimate that 22,701 of the Liverpool Irish heading for New York from 1846 to 1849 were from Cavan.
The same method was used to determine the number of people from each Irish county who emigrated to New York from 1851 to 1854, the later years of the Famine immigration. The proportion of emigrants from each Irish county in the sample for those years was multiplied by the total number of Liverpool Irish travelling to New York in that four-year period (411,707) to estimate the number of immigrants leaving each county for New York. Finally, the manifests from 1849 to 1851 were used to estimate the proportion of immigrants from each county sailing via Liverpool in 1850, and those proportions were then multiplied by the actual number of Irish emigrants sailing from Liverpool to New York to determine the number from each county in that year.
In terms of the county population figures for 1841 and 1851, it should be noted that the county-by-county population figures from the censuses taken in these two years vary from one government publication to the next. The figures used here are taken from the most detailed account of the census, the countyby-county returns, published in Census of Ireland for the year 1851. Part I. Showing the area, population, and number of houses, by townlands and electoral divisions. [C 1465, 1481, 1486, 1492, 1494, 1503, 1504, 1527, 1542-1555, 1557, 1560, 1563, 1565, 1567, 1570, 1571, 1574, 1575, 1579, 1581], H.C. 1852-3, xci-xcii.