Counting Palestinian Refugees Part III

Counting Palestinian Refugees Part III


Howard Adelman

Yesterday, I included in my blog Richard Mather’s “Op-Ed: How many Arab refugees were there in 1948? Maybe 300,000 – or less,” (20 January). That meant denying the accuracy of the usual figures used of 650,000 to 711,000. I promised to offer a critical analysis of Mather’s piece. An accurate and responsible figure should range between 650,000 at a minimum and 725,000 as a maximum with, I believe, 711,000 as the most accurate estimate of ethnic Palestinian refugees resulting from the Arab-Israeli War.

There have been efforts to diminish the numbers of Palestinian Arabs living in Mandatory Palestine as well as to decrease the number of Palestinian refugees for political purposes to minimize the claims to the land and for compensation or return for the refugees. The most infamous works along these lines was Joan Peters From Time Immemorial which even a right-wing Zionist like Daniel Pipes agreed used statistics sloppily, ignored any data that might contradict her argument and even distorted quotations. At the same time, there have been efforts to minimize the number of Jews in Mandatory Palestine by excluding so-called “illegal” Jewish immigrants and otherwise playing with figures and statistics. Given the ambiguity of and uncertainties about some of the data, it is even more important that partisan perspectives be bracketed while analyzing data.

Accurate counting should begin by obeying the following guidelines:

  1. Examine all reputable claims in light of the best evidence available;
  2. Demographic counts should be done indifferent to the political and moral claims of contending positions;
  3. The analysis should also be carried out disregarding whether Zionism is a legitimate ideology or whether the Palestinian arguments about colonialism and a nahba have any merit.
  4. The analysis should follow standard demographic norms of determining population figures even when it may be impossible to get a precise figure because of variations in original counts that were not subjected to rigorous verification.
  5. A common definition should be used for who is to be included – in this case, “persons whose normal place of residence was Palestine during the period 1 June 1946 to 15 May 1948, and who lost both home and means of livelihood as a result of the 1948 conflict.”
  6. When there are differences among authoritative data, then an effort should be made to reconcile differences and explain discrepancies.

It is clear that Mather followed none of the procedural norms. Below, I include his claims in quotation marks.

  1. “The conventional figure of Palestinian refugees who fled is 650,000.” [Later he cited the figure as 600,500 but I assume this was a typo.]


  1. The current conventional figure of the total number of refugees in 1948 is 750,000.
  2. The current accepted number of ethnic Palestinian refugees living in camps in 1949 is said to be 711,000. (Many Palestinian scholars and institutions use a figure of 726,000 or more.) The discrepancy between the two figures results from:
  1. The original figure was based not on ethnic Palestinians, but on definition of refugees from Palestine that included about 30,000 Jews who lost their homes in what became known as the West Bank, East Jerusalem and the Old City; “Palestine refugee” were defined as both Arabs and Jews whose normal place of residence was in Mandatory Palestine who had lost homes and their livelihoods as a result of the 1948 war;
  2. Just over 1%, that is, up to 9,000 individuals who lost their homes and livelihoods were neither ethnic Palestinians nor Jews.
  1. The figure used in 1950 was 914,000, not 650,000. Mather begins with the discounted figure that was 78% of the original figure and then he discounts again. The main reason for the original overcount was because local residents registered as refugees to obtain access to ration cards. Double counting and not delisting deaths tended to become significant later. By starting with a figure of 650,000, already a bit low, and then discounting that figure again for the very same reasons the original figure of 914,000 was discounted, results in an immediate major problem. Hence the inapplicability of the claim that, “UNRWA (the United Nations Relief and Works Agency), either through incompetence or deliberate manipulation, handed out multiple identity cards to the same persons, some of whom were not refugees at all but permanent residents who took advantage of the aid offered by UNRWA. This is attested by UNRWA officials.” The depiction is accurate, but it applies to the figure of 914,000, and then initially only to the problem of local residents registering as refugees.
  2. “There were fewer than 660,000 Arabs living in the part of Palestine that eventually became Israel.” This was a result of his claim that, “The Statistical Abstract of Palestine in 1944-45 set the figure for the total Arab population living in what would become the Jewish-settled territories at 570,800. Another set of figures based on a census taken in 1944 suggests there were 696,000 Arabs living in what would become Israeli-controlled territory.” This figure is contradicted by the following sources:
  1. Itamar Rabinovitch and Jehuda Reinharz (eds,) (2008) Israel and the Middle East: Documents and Readings on Society, Politics, and Foreign Relations, Pre-1948 to the Present report that there were 543,000 Jews and 1,267,037 non-Jews in Mandatory Palestine. In 1948, there were 716,700 Jews in Israel and only 156,000 non-Jews. Ignoring that the figure of non-Jews included Circassians, Armenian and other Christians who were not Arab, but whose numbers were small, as well as Bedouins who were both difficult to count and often not counted because they were nomadic, if the number Palestinian refugees is considered to be 711,000, then the number of Palestinians in what became the West Bank, Gaza and East Jerusalem totaled just over 400,000 in 1948 before the influx of refugees from the area that became Israel. (1,267,037 – 711,000 – 156,000 = approx. 400,000.) [Note that Itamar Rabinovitch is a very highly regarded Israeli scholar who also served as the negotiator with the Syrians between 1993-1996, was Israeli ambassador to the U.S. and President of Tel Aviv University.]
  2. To obtain the pre-1948 Arab population of the West Bank and Gaza, there is a table included in the Supplement to a Survey of Palestine (pp. 12-13) that excluded nomadic Bedouins providing the population by town and district in 1946 prepared for the United Nations and UNSCOP in 1947. As I will show, an initial analysis might appear to support Richard Mather The key districts of concern which in 1946 were virtually 100% Arab were:

East Jerusalem and Arab parts of the Old City                152,000

Gaza                                                                                  150,540

Hebron                                                                              93,120

Baysan                                                                              24,950

Jenin                                                                                  62,210

Nablus                                                                               98,600

Ramallah                                                                           49,930


Total                                                                                  631,350

This is 231,350 more than the resident non-Jewish populations should have been according to the above calculations. So if the 231,350 are subtracted from the 711,000 claimed refugees, then there should have been about 475,000 ethnic Palestinian refugees (more precisely 474,650). This is about half way between Mather’s figure of less than 300,000 and the figure of 650,000 which he claimed to be the generally accepted figure. However, there is a problem. The 231,000 are estimated to be about the number of local residents who registered as refugees to obtain rations. The totals were as follows:



711,000 actual refugees

30,000 Jewish refugees

231,000 local residents who registered as refugees

972,000 was the total, about 60,000 more than the original figure of 914,000. Double counting and non-recording of deaths would come later. This supports a figure of about 650,000 Palestinian refugees.

  • The Israeli Statistical Abstract from 1941 reported 1,111,398 Arab Palestinians in Mandatory Palestine. Allowing for an increase of 25% by 1947, that would yield 1,389,350 Arab Palestinians in Mandatory Palestine (versus Mather’s figure of 1.2 million). Deduct from that the 631,350 living in the West Bank and Gaza, that would mean the Arab Palestinian population in what became Israel before the war was 758,900. If 156,000 were left after the war, then the refugee population must have been 612,900. This would be a minimum figure because there is a general consensus that the figures inherited from the British Statistical Abstract undercounted both Jews and Arabs. The figures provided are:


1941 474,102 1,111,398 1,585,500 29.902%
                   1950 1,203,000 1,172,100 2,375,100 50.650%
  1. The Report of the Anglo-American Committee of Inquiry of 1946, which provides the foundation for most estimates, used the British 1945 survey based on both the 1922 and 1931 census data extrapolated by rates of natural increase for the population, without considering that both census figures were most likely undercounts. At the same time, there is a discrepancy between the number of ethnic Palestinian refugees yielded by this calculation (650,000) and the numbers scholars have considered authoritative (711,000). The reason for the difference is attributed to three factors:
  1. undercounting in the 1922 and 1931 census data;
  2. excluding in-migration from Syria and Transjordan resulting from the relative economic boom in Mandatory Palestine, the agricultural revolution underway in Mandatory Palestine both in the Jewish and Arab sectors (new technologies, use of fertilizers and pesticides, drainage of swamps), urbanization and the fact that Arabs in Palestine enjoyed a standard of living double that in Lebanon, Syria and Transjordan;
  3. underreporting of both births and infant deaths;
  4. Adding the Bedouin population that became refugees to the 650,000 figure.


In any case, the count of ethnic Palestinian refugees ranged between 650,000 and 711,000.







  1. The British 1945 Survey
Year Source Total Moslems Jews Christians Others
(No.) (%) (No.) (%) (No.) (%) (No.) (%)
1922 Census 752,048 589,177 78.34 83,790 11.14 71,464 9.50 7,617 1.01
1931 Census 1,033,314 759,700 73.52 174,606 16.90 88,907 8.60 10,101 0.98
1937 Estimate 1,383,320 875,947 63.32 386,084 27.91 109,769 7.94 11,520 0.83
1945 Survey2 1,845,560 1,076,780 58.35 608,230 32.96 145,060 7.86 15,490 0.84
19471 Projection 1,955,260 1,135,269 58.06 650,000 33.24 153,621 7.86 16370 0.84


    1. Justin McCarthy (1990), The Population of Palestine, Population History and Statistics of the Late Ottoman Period and the Mandate. New York Columbia University Press. (see also his article in Encyclopedia of the Palestinians, Philip Mattar, ed.) who concluded, based on his demographic studies, that of 873,00 ethnic Palestinians residing in what became Israel after the 1948 war with only 155,000 remaining, there were 718,000 Palestinian refugees and fatalities as a result of the war.
    2. Both Walid Khalidi (ed.) (1992) All That Remains: The Palestinian Villages Occupied and Depopulated by Israel in 1948 (Washington, D.C.: Institute for Palestine Studies and Nur Masalha (1992) Expulsion of the Palestinians: The Concept of “Transfer” in Zionist Political Thought, 1882–1948 (Washington, D.C.: Institute for Palestine Studies, insist that the basic figure (they use 726,000 or more) should be increased by Palestinians who were outside Palestine when the 1948 war broke out but were prevented from returning to their homes, as well as those who did not register with the United Nations Relief and Works Agency (UNRWA).
  • Janet L. Abu Lughod (1971) “The Demographic Transformation of Palestine”, in Ibrahim Abu Lughod (ed.) The Transformation of Palestine: Essays on the Origin and Development of the Arab-Israeli Conflict, Northwestern University Press, 139-163, concludes there were 720,000 ethnic Palestinian refugees.
  • Mather cites Dr. Walter Pinner’s 1966 claim that the totals of refugees included both 109,000 who had been resettled in 1948 and 225,000 who subsequently settled elsewhere and became self-supporting to bolster his argument for a 1948 Palestinian refugee population under 300,000. However, the UNRWA definition is based on refugees who lost their homes and livelihoods and, therefore, includes refugees resettled elsewhere and those who became self-supporting; using the UNRWA definition, this would bring the totals back up to over 600,000. Further, given the definition that included those who lost their homes and livelihoods, it should also be noted that the Arab Palestinians (as well as Jewish refugees in Palestine) only had to have lived in Mandatory Palestine for two years. They did not have to be registered permanent residents. There was no requirement that those who settled outside the countries of first asylum ceased being refugees in accordance with the UNRWA definition, a definition which preceded the creation of UNRWA.



Further, there is a tendency (Mather is an example) of asserting that the inflated figures (originally 916,000) were matters of fraud. This is far too simplistic. In fact, in almost all surveys of refugee camps, the figures are estimated to be inflated by 10%, an inflation that increases to 25% when “military” units control the camps.

Finally, poor scholarship does not help, whatever side one is on, but rather undermines it, in this case, the position of Israel in dealing with the refugees, just as original inflated figures used by UNRWA lost UNRWA credibility in the international sphere in its first decades. In the above I have not cited studies in which I myself was involved, such as the study undertaken by the Centre for Refugee Studies at York University in the nineties of the number of Palestinian refugees. Nor have I cited the work of the Canadian government that gaveled the refugee multilateral talks. Though my figures were initially a bit lower than the 711,000, I subsequently concluded that the 711,000 was more accurate.

Whatever our differences, we should at least learn to count accurately, or as accurately as possible.

With the help of Alex Zisman


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s