The Smoking Gun: Arab Immigration into Palestine, 1922-1931
by Fred M. Gottheil
Middle East Quarterly[/indent]
Winter 2003, pp. 53-64
['........]
Evidence for Arab Migration
There are several problems associated with estimating Arab immigration into Palestine during the 1920s, the principal one being that
Arab migration flows were, in the main, illegal, and therefore unreported and unrecorded.[17] But they were not entirely unnoticed.
Demographer U.O. Schmelz's analysis of the Ottoman registration data for 1905 populations of Jerusalem and Hebron
kazas (Ottoman districts), by place of birth, showed that of those
Arab Palestinians born outside their localities of residence, approximately HALF represented intra-Palestine movement—from areas of low-level economic activity to areas of higher-level activity—while the other half represented Arab immigration into Palestine itself, 43% originating in Asia, 39% in Africa, and 20% in Turkey.[18] Schmelz conjectured:
The above-average population growth of the Arab villages around the city of Jerusalem, with its Jewish majority, continued until the end of the mandatory period. This must have been due—as elsewhere in Palestine under similar conditions—to in-migrants attracted by economic opportunities, and to the beneficial effects of improved health services in reducing mortality—
just as happened in other parts of Palestine around cities with a large Jewish population sector.
[19]
While Schmelz restricted his research of the 1905 Palestinian census to the official Ottoman registrations and used these registrations with only minor critical comment, he did acknowledge that "stable population models assume the absence of external migrations,
a condition which was obviously not met by all the subpopulations" that Schmelz enumerated.
[20]
Like U.O. Schmelz, Roberto Bachi expressed some reservation about the virtual non-existence of data and discussion concerning migration into and within Palestine. He writes:
Between 1800 and 1914, the Muslim population had a yearly average increase in the order of magnitude of roughly 6-7 per thousand. This can be compared to the very crude estimate of about 4 per thousand for the "less developed countries" of the world (in Asia, Africa, and Latin America) between 1800 and 1910.
It is possible that part of the growth of the Muslim population was due to Immigration.
[21]
Although Bachi did not pursue the linkage between undocumented immigration into Palestine and the 6 (or 7) to 4 per thousand differential in growth rates between Palestine and the other less developed countries (LDCs), the idea that at least one-third of Palestine's population growth may be attributed to immigration is—using Bachi's own growth rate differentials—not an entirely unreasonable one.
Lacking verifiable evidence did not prevent Bachi from stating the obvious concerning internal migration within Palestine:
The great economic development of the coastal plains—largely due to Jewish immigration—was accompanied both in 1922-1931 and in 1931-1944 by a much stronger increase of the Muslim and Christian populations in this region than that registered in other regions. This was probably due to two reasons: stronger decrease in mortality of the non-Jewish population in the neighborhood of Jewish areas and Internal migration toward the more developed zones.[22][/B]
In the footnote accompanying this quote, Bachi writes: "As no statistics are available for internal migration, this conclusion has been obtained from indirect evidence."[23] Bachi's footnote is instructive. The "indirect evidence" he referred to no doubt included his understanding of the important role economics plays in explaining demographic movements. While appreciating the value of Ottoman registrations and British mandatory government censuses in providing estimates of Palestinian demography, they were, in his judgment, still crude and incomplete.
Reference to Arab immigration into Palestine during the 1920s is made as well in the British mandatory government's annual compilation of statistical data on population. The Palestine Blue Book, 1937, for example, provides time series demographic statistics whose annual estimates are based on extrapolations from its 1922 census.[24] The footnote accompanying the table on population of Palestine reads:
There has been unrecorded illegal immigration of both Jews and Arabs in the period since the census of 1931, but it is clear that, since it cannot be recorded, no estimate of its volume is possible.[25]
The 1935 British report to the League of Nations noted that:
One thousand five hundred and fifty-seven persons (including 565 Jews) who, having made their way into the country surreptitiously, were later detected, were sentenced to imprisonment for their offence and recommended for deportation.[26]
The number who "made their way into the country surreptitiously" and undetected was neither estimated nor mentioned.
Historian Gad Gilbar's observation on Ruth Kark's contribution to his edited volume Ottoman Palestine, 1800-1914, touches on the issue of Arab immigration into and within Palestine. He relates her ideas in "The Rise and Decline of Coastal Towns in Palestine" to Charles Issawi's thesis concerning the role of minority groups and foreigners in the development of Middle Eastern towns. Explaining why NO other Palestinian cities grew as rapidly as Jaffa and Haifa did during the final three decades of the Ottoman rule, Gilbar writes: "Both attracted population from the rural and urban surroundings and immigrants from outside Palestine."[27]
Each piece of the demographic puzzle by itself may reveal no identifiable picture. But given a multiplicity of such pieces, an image does begin to appear. The Royal Institute for International Affairs adds another piece. Commenting on the growth of the Palestinian population during the decades of the 1920s and 1930s it reports: "The number of Arabs who have entered Palestine illegally from Syria and Transjordan is unknown. But probably considerable."[28] And C.S. Jarvis, governor of the Sinai from 1923-36, adds yet another:
This illegal immigration was not only going on from the Sinai, but also from Trans-Jordan and Syria, and it is very difficult to make a case out for the misery of the Arabs if at the same time their compatriots from adjoining states could not be kept from going in to share that misery.[29]
Estimating Real Numbers
The derivation of Palestine migration estimates in this section is based on an uncomplicated imputation theory. Migration becomes a residual claimant for numbers not explained by a population-estimating model based on known initial population stocks and known sets of birth and death rates for that population. In this way, expected population stocks can be derived for any set of subsequent years.
The value of the model depends, of course, on the reliability of the estimates given for initial population stocks and for the rates associated with natural increase. Therein lies the problem with estimating Arab immigration into Palestine. The model itself may be simple and applicable, but its usefulness—as with all estimating models—is contingent upon the quality of the data inputs. That quality in the case of Palestinian migration is compromised by the explicit neglect of illegal entrants. If illegal migrants and subsequently illegal residents escaped the census taker, how could the census account for them? It couldn't and didn't.
It is not surprising then that the British census data produce an Arab Palestinian population growth for 1922-31 that turns out to be generated by natural increase and legal migrations alone. Applying a 2.5 per annum growth rate[30] to a population stock of 589,177 for 1922 generates a 1931 population estimate of 735,799 or 97.6% of the 753,822 recorded in the 1931 census. Does the imputation model then "prove" that illegal immigration into Palestine was inconsequential during 1922-31? Not at all. A footnote accompanying the census's population time series acknowledges the presence in Palestine of illegal Arab immigration. But because it could not be recorded, no estimate of its numbers was included in the census count.[31] Ignoring illegal migrants does not mean they don't exist.
Setting illegal immigration into Palestine aside, the imputation model does generate substantial migrations of Arab Palestinians within Palestine itself and confirms what many demographers, historians, government administrators, and economists have alluded to: the migration of Arab Palestinians from villages, towns, and cities of low economic opportunity to villages, towns, and cities of higher economic opportunity.
Which towns, villages, and cities offered the higher economic opportunity? Analyzing the 1922 and 1931 demographic data by sub-district and separating those sub-districts of Palestine that eventually became 1948 Israel—that is, sub-districts that had relatively large Jewish populations (with accompanying Jewish capital and modern technology)—from those that were not designated as part of 1948 Israel, identified not only the direction of Arab Palestinian migration within Palestine but its magnitude as well.[32]
The Arab Palestinian populations within those sub-districts that eventually became Israel increased from 321,866 in 1922 to 463,288 in 1931 or by 141,422. Applying the 2.5 per annum natural rate of population growth to the 1922 Arab Palestinian population generates an expected population size for 1931 of 398,498 or 64,790 less than the actual population recorded in the British census. By imputation, this unaccounted population increase must have been either illegal immigration not accounted for in the British census and/or registered Arab Palestinians moving from outside the Jewish-identified sub-districts to those sub-districts so identified. This 1922-31 Arab migration into the Jewish sub-districts represented 11.8% of the total 1931 Arab population residing in those sub-districts and as much as 36.8% of its 1922-31 growth.
That over 10% of the 1931 Arab Palestinian population in those sub-districts that eventually became Israel had immigrated to those sub-districts within the 1922-31 years is a datum of considerable significance. It is consistent with the fragmentary evidence of illegal migration to and within Palestine; it supports the idea of linkage between economic disparities and migratory impulses—a linkage universally accepted; it undercuts the thesis of "spatial stickiness" attributed by some scholars to the Arab Palestinian population of the late nineteenth and early twentieth centuries; and it provides strong circumstantial evidence that the illegal Arab immigration into Palestine, like that within Palestine, was of consequence as well.
Denying the Evidence
As compelling as the arguments and evidence supporting consequential illegal immigration may be to some scholars, they are clearly unconvincing to others. The single most cited contemporary publication on Palestinian demography is Justin McCarthy's 1990 The Population of Palestine. Of McCarthy's 43 pages of descriptive analysis—supplemented by 188 pages of demographic tables copied directly from Ottoman, European, and Jewish source materials—slightly more than one and a half pages are devoted to Arab immigration into and within Palestine during the Ottoman period, and a similar one and a half pages are devoted to Arab immigration during the succeeding mandate period.[33] According to McCarthy, these few pages offer enough critical analysis to close the lid on the "infamous" immigration thesis.
Consider first McCarthy's analysis of Arab immigration during the Ottoman period. That he finds no illegal immigration of consequence is not surprising because McCarthy uses official Ottoman registration lists that, by the nature of its classifications, take no account of the unreported, illegal immigration. That is to say, if you look in a haystack for a needle that wasn't put there, the probability is high you won't find it. It is strange that that idea had not occurred to McCarthy. Choosing to focus on the official registration lists allows him to write:
From the analysis of rates of increase of the Muslim population of the three Palestiniansanjaks [Ottoman sub-provinces], one can say with certainty that Muslim immigration after 1870 was small.[34]
Reflecting elsewhere on the possibility that the immigration may have occurred over an extended period of time, McCarthy writes: "To postulate such an immigration … stretches the limits of credulity."[35]
McCarthy's treatment of the linkage between economic disparities and migration impulses appears to be even more disingenuous. He writes: "The question of the relative economic development of Palestine in Ottoman times is not a matter to be discussed here."[36] Nor is it considered anywhere else in his book. That is to say, McCarthy does not contest the linkage so much as ignore its relevance to the Palestinian situation.[37]
His dismissal of Arab immigration into Palestine during the mandate period is based on a set of assumptions concerning illegal immigration that is both restrictive and unsubstantiated. He contends that even if the illegal immigrants were unreported on entry, their deaths in Palestine would have been registered. So too, he argues, would their children born in Palestine. Deriving estimates based on such registrations, he arrives at this conclusion: immigration was minimal.[38] But he provides no evidence to show that these supposed registrations of births and deaths were actually made. Had McCarthy considered the fact that detection of illegal immigration during the mandate period resulted in imprisonment and deportation and that immigrants, aware of this, may have avoided any formal registration of deaths and births, he would have had to revise his assessment of illegal immigration.
Perhaps the more serious charge against McCarthy's analysis of Arab immigration is his use of Roberto Bachi's estimates. McCarthy's numbers are based, in part, on Bachi's reporting of 900 illegal Arab immigrants per year over the period 1931-45.[39] But McCarthy misrepresents what Bachi's estimate is meant to show. Bachi is careful to identify his 900-per-year illegal Arab immigration estimate as only those discovered by the mandatory authorities. Illegal Arab immigration that went undetected and unreported is not included. He writes:
A detailed analysis presented in Appendix 6.5B on the basis of the registration of part of the illegal migratory traffic, discovered by the Palestine police, shows that legal movements (as reflected in Tables 9.4-9.7) constituted only a small fraction of total Muslim immigration.[40]
To emphasize this point, Bachi writes: "It is hardly credible that illegal movements which were actually discovered included all the illegal entrances which actually occurred, or even the majority of them."[41] As a result, Bachi can only conclude that "in the present state of knowledge, we have beenunable to even guess the size of total immigration."[42]
Such a cautionary comment finds no place in McCarthy's analysis or conclusions. Using Bachi's estimates inappropriately, deriving estimates based solely on registration lists, and ignoring completely the linkages between regional economic disparities and migratory impulses, McCarthy confidently concludes,
the vast majority of the Palestinians resident in 1947 were the sons and daughters who were living in Palestine before modern Jewish immigration began. There is no reason to believe that they were not the sons and daughters of Arabs who had been in Palestine for many centuries.[43]
Every Reason to Believe
Therein lies the ideological warfare concerning claims to territorial inheritance and national sovereignty. Contrary to McCarthy's findings or wishes, there is every reason to believe that consequential immigration of Arabs into and within Palestine occurred during the Ottoman and British mandatory periods. Among the most compelling arguments in support of such immigration is the universally acknowledged and practiced linkage between regional economic disparities and migratory impulses.
The precise magnitude of Arab immigration into and within Palestine is, as Bachi noted, unknown. Lack of completeness in Ottoman registration lists and British Mandatory censuses, and the immeasurable illegal, unreported, and undetected immigration during both periods make any estimate a bold venture into creative analysis. In most cases, those venturing into the realm of Palestinian demography—or other demographic analyses based on very crude data—acknowledge its limitations and the tentativeness of the conclusions that may be drawn.
Fred M. Gottheil is a professor in the department of economics, University of Illinois.
