For rich or desperate there was no forgetting that there was a war on. Salazar was playing a careful game of fence sitting. But the neutrality could only be maintained because it benefitted both the Axis and the Allies to keep Portugal out of the war. Neither side wanted the obligation to defend her if called to. For Britain, Portugal kept supply lines open to Europe and tungsten and trade flowed to Germany.
While Hitler saw the possibility that Britain might mount a new European front from the Portuguese coast, he also saw Portugal as as potential launching pad for an attack on Gibraltar and North Africa. Franco and Hitler discussed plans for the invasion of Portugal. Naturally, the plan itself was top secret but after Germany had so easily annexed Austria at the start of the war, the Portuguese were well aware of the danger, from a Spain bolstered by Germany, of being swallowed up by a larger neighbour.
Transient refugees probably only saw Portugal as an oasis, as any theoretical threat here would seem insignificant after having experienced the war first hand. Initially the influx of foreigners was good for the economy, and ordinary people did benefit from the increase in demand for everything that was in short supply in the rest of Europe. However, by Portugal was in the midst of a food shortage and the scarcity of fuel caused uneven distribution of the food that was available. The influx of refugees and the increase in trade drove up inflation, doubling the price of staples like sardines and bacalhau since the start of the war.
Black marketeering, crime and corruption peaked. For the Portuguese, neutrality did not necessarily mean peace. Sami, Agree. Very interesting. Emma knows almost every hidden secret of Portugal. Very objective in her analysis…I like that. Very well written. I noticed you mentioned there was freedom, which is interesting considering Portugal was ruled by a dictator. Freedom has many facets and there was indeed plenty of freedom of all kinds but political, cultural and economical to a lesser extent. Make no mistake, it was a dictatorship indeed, however it certainly was a much better place than many other dictatorships.
In spite of all his darkness, Salazar does deserve a great deal of credit for having brilliantly maintained our neutrality. Thanks Rui. Freedom is relative indeed, and I meant it from the perspective of the refugees. Perhaps at no other time in the next 30 years would foreigners have felt freedom, in Portugal. Another good post, Emma. Thank you Emma, that was really interesting. I am going to pass this on to my daughters who are studying this period in history class.
Is this only part 1? You whet my appetite to learn more. Thanks to you and also Pamela for the reading references. I found this article really interesting. My father who was a merchant seaman at this time spent some time shore time in Lisbon. He said the dockside bars and cafes were full of the agents and spies of every nation trying to pick up information from drunken seamen. He was only seventeen when he was involved in a big brawl with some Germans from another ship. It was an explosive situation as they were at war but in that place they were expected to behave.
The portuguese police for the area were understandably the toughest available and he got literally caught up in a net between two police horses which they used to round up the trouble makers.
He was imprisoned and his captain contacted the British Consul as was the custom, he came down with money for dad until he got out. Dad considered the jailors to of been more sympathetic to the British and indeed was treated really well, as long as he had money they used to take him to a cafe every day for lunch where he used to supply wine for them. This was a surprise to him as he had been warned by officers prior to shore leave that Portuguese loyalties were likely to be with Germany and to be very careful because of this.
He remaind fond of all things Portuguese all his life. I have always wondered about the Portuguese involvement in the war. For a country that was considered neutral, I think we were quite involved. We need books that tell us the whole story history , the good and the bad. I loved your articles, very informative. I am from a little town called Adao-Lobo, Cadaval. In the same area as Bombarral, Caldas da Rainha, Obidos. Beautiful area with good fruit and wine.
Once again thank you. Very interesting all the information gathered. No more bussiness…. Great info for a chapter in the novel I am writing. I did know that Portugal was important to the Allies and to the refugees who were lucky to get there. Anyone who knows more… please contact me re. The Literary Life of Janet S. This made fascinating reading. Your blog post gave an excellent background. Not being able to afford their own passage or not being sponsored by a close family member could have been an indication of lower socioeconomic status, which might also be reflected in shorter stature.
Table 4 b shows results for females.
Similar to males, female refugees were 2. Adding controls for time of migration H2 erases this effect, which is also not present in the joint test. This category obviously does not reflect female human capital, but their traditional role within the family. Though not significant the skill coefficient is negative possibly indicating that the skilled women were the relatively poor who needed to support their families.
Table 4b. Correlates of individual height females. Contrary to males, when testing the migration hypothesis we find that females that were prior migrants before leaving Europe were associated with a height advantage of 1. These early migrants were taller, regardless of refugee status, when compared with females who lived in their country of birth prior to traveling to the United States.
Females with visas issued later had a height penalty of 1. Testing the wealth hypothesis does not confirm the inverse relationship between travel class and height that we found in men, but we still find that females with passage paid by a third party were shorter than the otherwise sponsored. Again, in the s women were rarely the bread earners in the household and as such being sponsored by a family member usually the husband or the father was the norm.
Indeed, we find a negative height correlation in the range of 1. Refugee interactions in this test are not statistically significant. Again, we run this augmented regression in four models separately for gender and refugee status to obtain estimated height values after controlling for migration initiative. We then compare the new estimated height averages by nationality, gender and refugee status, with the computed height averages resulting from regression 1. If the augmented estimation strategy changes the height gap between migrants and home population we may explain some of the observed migrant selection with differences in migration initiative rather than differences in general selection.
On average, adding controls for migration initiative, as outlined above, does not substantially alter the selection of male nonrefugees; only the values of some female nonrefugee samples change notably.
The Lisbon Route tells of the extraordinary World War II transformation of Portugal's tranquil port city into the great escape hatch of Nazi Europe. Royalty. The Lisbon Route tells of the extraordinary World War II transformation of Portugal's tranquil port city into the great escape hatch of Nazi Europe.
Table 5. Changes in estimated height by passenger group in cm. Since nonrefugees are unaffected by the changes in the estimation method, we believe that the observed selection in tables 3 a and 3 b is partially the result of changes in migration initiative, and does not entirely reflect socioeconomic differences. The differences in selection between nonrefugee and refugee males are reduced, but not eliminated for Belgians, Germans, Hungarians, and Swiss, while virtually eliminated for French nationals.
Conversely, the height premium for male Italian refugees raises, and the selection differentials for male Poles and Dutch do not experience substantial changes. For females, the selection differentials are smaller for Poles, Russians, and almost disappear for Czechoslovakia, Belgium, France, and Germany. Finally, the existing refugee height advantage for Hungarians and Swiss increases slightly. We assess the human and health capital of Europeans arriving New York City between and and originating in Lisbon, the only European port with regular passenger traffic to the Americas after mid when most of Europe was engulfed in war.
The majority of these individuals were Jewish refugees and their families escaping Nazi persecution and expanding war. Many Jews left Europe ever since the Nazis took power in Germany in , but the closure of all other European gateways to passenger traffic after the outbreak of the war, made these Lisbon passengers the last to escape the Holocaust.
We investigate how these migrants compared to fellow nationals remaining in Europe, whether they were different from non-Jews traveling in the same vessels, and if such differences were accounted by observable characteristics. We construct a novel dataset based on ship manifests from the United States Records of the Immigration and Naturalization Service, which contain microlevel information on all alien passengers making the journey from Lisbon to New York between July of and June of Detailed information on personal and socioeconomic characteristics, in addition to anthropometric indicators, allows for a comparative study of the patterns of selection of these wartime migrants.
We use adult stature as a proxy for the human and health capital these migrants carried. For most of the observed nationalities, the last Holocaust refugees were positively selected relative to the source populations, a pattern that was more pronounced in women, reflecting the war context. The degree of selection we observe reflects the immense difficulty of traveling from European source regions to the south-western tip of the continent in the early s; those who succeeded were not only fortunate, but also well-off.
Despite reported records that refugees had to overcome more hurdles than nonrefugees to escape Europe in the early s, we find that the latter were taller than the former and, therefore, more positively selected relative to the populations in the source countries. We assess three hypotheses potentially behind the height gap between refugees and nonrefugees: refugees may come from a different socioeconomic background, have different migration initiative, or come from different sectors of the wealth distribution.
We find that refugees are no different from refugees in terms of wealth or skill, but that migration initiative, proxied by visa date and migrant status prior to traveling to Lisbon, plays a significant role in explaining height differences between refugees and nonrefugees. Earlier arrivals were more positively selected than later migrants, but even the latest of these Holocaust refugees were positively selected relative to the source country populations.
Our present findings suggest not only that earlier Jewish refugees escaping Europe between — might have carried even higher levels of human capital. More generally, refugees from long lasting conflicts can still be positively selected relative to their place of origin long after living conditions started deteriorating in conflict zones. Unfortunately, the ship manifests are not informative about the motives behind the timing of migration, so we can only speculate about the late departures.
It seems reasonable however, that owning a house, a shop, or a factory would make individuals hold out for longer than if they work for somebody else or rent a house. We leave the important discussion of the determinants of the observed migrant selection for future work. The early s marked the infancy of commercial aviation. The outbreak of the war in September halted service to Southampton and Marseilles, so Foynes in Ireland and Lisbon in Portugal became the terminals for the northern and southern routes, respectively.
Winter brought the suspension of the northern route in October of and it was never resumed, which consolidated air service across the Atlantic on the New York-Lisbon line Trippe : After the US joined the war in December of , clippers carried military personnel and equipment on the Atlantic and Pacific routes under the orders of the US Army Transport Command, though the planes were still flown by Pan Am crews.
The end of the war retired the clipper, which had become technologically obsolete with the development of new planes that could travel longer distances and land on runways a legacy of wartime building programs , making flying much safer. Among the dead and the seriously injured were the prominent American author and war correspondent Benjamin Robertson and American singer and actress Jane Froman.
Also killed in a plane shot down by the Luftwaffe was English film star Leslie Howard when flying from Lisbon to Bristol in June of The high profile of the passengers involved in these accidents suggests that flying in the s was the form of travel of the upper elites. Commercial air travel became more generalized in the late s with the development of jet technology, which considerably reduced travel times and effectively replaced passenger ships in long-distance transportation. The dominant form of international travel in the early s was therefore the ocean liner, which is the reason we focus on the passengers thus carried.
Steamers connected Lisbon and New York in 9 days and carried the bulk of the transatlantic passengers. Several shipping lines from multiple nations had been crossing the Atlantic since the nineteenth century. The former provided cargo and passenger service to Mediterranean ports from Gibraltar to Haifa , while the latter directed cargo, passenger, and mail operations to ports further north from Le Havre in France to the Free City of Danzig, today Gdansk in the Baltic. In Portugal, the National and the Colonial Navigation Companies operated in the Atlantic since and , respectively, mostly on routes connecting Lisbon to Africa and Brazil.
After the US joined the war on 8 December, , these ships went into service under the US Army for the transport of troops. In , there were only two voyages by Portuguese vessels between Lisbon and New York in , in August and November, but service picked up in with two or three vessels each month. Of the vessels crossing to New York between July of and June of , 66 were American, 28 Portuguese, and six had other nationalities: four Greek and one Japanese all in , and one Swedish ship carrying the last diplomats out of Europe in May of Ocean liners on the Lisbon—New York route differed substantially in size.
Originally luxury ships, the Four Aces had smaller capacity than the larger Portuguese passenger ships. The SS Excalibur and its sister ships carried first class passengers, whereas the SS Serpa Pinto could carry a total of passengers first class, 86 second class, third class, and steerage.
Passenger ships out of Lisbon were often overbooked and oversold. Cargo vessels operated by the same shipping companies occasionally carried very few passengers 5—13 if any. The header on each manifest page contains name of the vessel, date of departure from Lisbon or any other intermediary port of call, date of arrival in New York, and class of travel.
Manifests were filled out by officials of the shipping company. Each manifest page contains up to 30 passengers on separate lines numbered 1 to The information asked of each passenger figures in numbered columns, now transcribed. If native of United States insular possession or if native of the United States, give date and place of birth city or town and state. If naturalized, give name and location of court which issued naturalization papers and date of papers.
Data concerning verification of landings, etc. The name and complete address of nearest relative or friend in country whence alien came, if none there, then in country of which a citizen or subject. Final destination, state, city or town Intended future permanent residence : Foreign country via port of departure , in the United States its territories or possessions State, city or town. By whom was passage paid whether alien paid his own passage, whether paid by relative or any other person, or by any corporation, society, municipality, or government.
Whether going to join a relative or friend; state name and complete address, and if relative, exact relationship. Purpose of coming to United States: Whether alien intends to return to country whence he came after engaging temporarily in laboring pursuits in the United States; Length of time alien intends to remain in the United States; Whether alien intends to become a citizen of the United States.
Ever in prison or almshouse or institution for care and treatment of the insane or supported by charity, if so, which? Whether a person who believes in or advocates the overthrow by force or violence of the Government of the United States or all forms of law, etc.
Whether coming by reason of any offer, solicitation, promise or agreement expressed or implied to labor in the United States. The Armstrong index considers five categories that classify occupations according to the required amount of training. In skilled we consider occupations with solid training and skills, such as merchants, nurses, and skilled industrial workers.
Semi-professional occupations include students, engineers, teachers, economists, chemists, and other white collar occupations indicating higher education. Professionals constitute the upper end of the occupational scale, which include diplomats, physicians, and university professors. Figures A1 and A2 provide the top occupations for men and women 16 and older in our data. Male refugees were mostly merchants and students in contrast with other male passengers who were mostly students and diplomats.
These females could have been educated, but such is not captured by the Armstrong index. Language regressions in table A1 report differences in language skills by gender, refugee status and origin after controlling for nationality. The dependent variable is the number of languages an individual is able to speak; accordingly, coefficients are interpreted as the average number of languages an individual is able to speak conditional on all control variables.
We find that male and female nonrefugees spoke 1. For instance, 47 percent of French female refugees spoke an additional language compared with female nonrefugees. This trait is in line with the general finding in the migration literature that a larger migrant stock reduces selection.
Languages by gender and origin. Table A2. United States immigration quotas by country of origin in Government Printing Office, Oxford University Press is a department of the University of Oxford. It furthers the University's objective of excellence in research, scholarship, and education by publishing worldwide. Sign In or Create an Account. Sign In. Advanced Search. Article Navigation.
Close mobile search navigation Article Navigation. Volume Article Contents. Jewish outmigration and US immigration policy. The passenger data. Human and health capital. Appendix A. Editor's Choice. Escaping Europe: health and human capital of Holocaust refugees 1 Matthias Blum. Oxford Academic. Google Scholar. Claudia Rei. Vanderbilt University. Cite Citation. Permissions Icon Permissions.
Abstract The large-scale persecution of European Jews during the Second World War generated massive refugee movements. Figure 1. Open in new tab Download slide. Jewish arrivals in New York by country of departure. Source : Ancestry. Figure 2. Figure 3. Figure 4. We evaluate the impact of this outmigration in terms of health and human capital by estimating average heights of adult European passengers, separately by nationality, gender, and refugee status.
Table 3a. Table 3b:. We address this question by testing whether the height gap of Jews in our data is robust in a multivariate regression setting. To do so, we test whether the Jewish height gap is associated with differences in skills, in the timing of migration, and in wealth. We just showed that observed differences in height between refugees and nonrefugees can partially be explained by differences in the time of migration of these groups.
Generally, taller individuals tended to migrate earlier, and nonrefugees in our data tended to migrate before refugees see figure A3 in the Appendix. As a consequence, migration initiative, captured by migration timing and prior migrant status, may explain some of the migrant selectivity in general. See Kaplan for a thorough discussion. New York was the major entry port into the United States for Portuguese ships. Baltimore, Boston, and New Orleans also received Jewish migrants but not in large numbers.
Between and , Baltimore received a total of Jews from Portugal , ports in Massachusetts 1, none from Portugal , and New Orleans none from Portugal. Routes to reach Lisbon varied. Most famous is probably the one described in the initial credits of the movie Casablanca where refugees went from Paris to Marseille, crossed the Mediterranean to Oran in Algeria and from there traveled to Casablanca in French Morocco where they would wait for exit visas to Lisbon.
See Halperin , Lochery , and Weber for accounts of refugee journeys to Lisbon during this problematic period, and Morewitz and Lieberman , Redel and Remarque for fictionalized novels of these dramatic events based on true stories. For a thorough review of immigration regimes in US history, see Abramitzky and Boustan See table A2 in the Appendix for exact quota numbers after the amendment.
For detailed visa procedures, see Morse : — Increased visa restrictions sealed the fate of many who sought to escape Nazi occupied or threatened territory after this date SDD Such was the case of Otto Frank father of Anne Frank , residing in Amsterdam since , who sought American visas for his family in April of Breitman et al. There were also vessels departing from other Portuguese ports in the mainland or the Atlantic Islands, but these vessels overwhelmingly carried cargo. Sometimes ships departed from Lisbon and stopped at intermediate ports such as Casablanca, Bermuda, or Havana, to drop off and pick up passengers before getting to New York.
See Data Appendix for the different sets of questions asked to US citizens and alien passengers. Given the war context, some refugees could have tried to hide their Jewish origins, potentially creating an undercounting problem. A two-sample t -test and a two-sample Wilcoxon rank-sum test for the differences in age of refugees and nonrefugees suggest that these age differences are statistically significant. Approximately 2, passengers in our sample were declared stateless so we assigned them a nationality based on country of birth. While this was formally correct after the Anschluss, we code all individuals born in Austria as Austrian.
See Appendix for multipart question 24 on the purpose of the trip and intended length of stay. The couple declared a length of stay of 6 months but their status must have changed after arrival since they remained in the US for 8 years. This case shows that some of the nonrefugees with temporary but not necessarily short stays did not return to Europe any time soon. As such we consider them as part of the nonrefugee outmigration movement.
Since literacy tests for incoming migrants were in place as of the Immigration Law, we interpret this question simply as language knowledge. We investigate whether this difference is due to refugees from bilingual countries and find that refugees of quite a few nationalities had better language skills than nonrefugees even after controlling for country fixed effects. See table A1 in the Appendix. See Baten and Blum for details of its construction. We calculate female heights based on the concept of sexual dimorphism, which considers differences across genders within the same species, such as height.
We thus apply the formula relating male and female heights in Holden and Mace An alternative formula in Gustafsson and Lindenfors results in lower benchmark heights for women, which yields even higher positive selection so we report the more conservative estimates only. See Spitzer and Zimran : 32—33 , for a thorough discussion of this matter.
We also adjust heights of individuals 50 and older for shrinking, using estimates that the elderly male and female English populations experience an annual decline in physical stature of approximately 0. This adjustment does not dismiss the importance of age dummies to control for cohort effects.
The pattern suggests refugees got visas later than other passengers. For example, a visa issued on the 10th of May corresponds to 1. Robustness checks including vessel fixed effects confirm these findings. Most refugees were Jews so the statistically significant coefficient could potentially result from kosher diet. Such cultural factors are difficult to quantify but late twentieth century Israelis are relatively tall given their national income Blum Surely not all Israelis keep kosher, but devout Christians also fast and abstain from meat once a week, which can also have implications in adult stature.
We therefore cannot associate a kosher effect with the negative coefficient. All models control for nationality, eliminating the potential concern that taller nationalities left Europe earlier.