Roughly 73% of Palestinian departures in 1948 are attributed to Zionist and Israeli military action in the locality-by-locality typology the historian Benny Morris built from Israeli military and state archives, published in the second edition of The Birth of the Palestinian Refugee Problem Revisited (Morris, 2004). Morris puts the total displaced at roughly 700,000, assigns a further 22% of departures to fear and a crisis of confidence after attacks nearby, and reports finding no blanket cabinet expulsion order in the archives he examined, a reading Ilan Pappé rejects in The Ethnic Cleansing of Palestine (Pappé, 2006). Families who kept house keys and land deeds have asked one question about 1948 for three generations: did people leave, or were they made to leave? The archives answer it village by village.
Key Findings
- In Morris’s locality-by-locality attribution of a primary cause to each depopulated place, 55% of departures follow Haganah and Israel Defense Forces military operations, 15% Irgun and Lehi operations, 2% explicit expulsion orders and 1% psychological warfare (Morris, 2004).
- 22% of departures are attributed in the same typology to fear and a crisis of confidence, most often after a neighbouring locality fell (Morris, 2004).
- Roughly 700,000 Palestinians were displaced, Morris’s own archival estimate (Morris, 2004); the Israel Law Review article 1948 Refugees records that roughly 80% of the residents of the areas on which Israel was established were expelled or fled.
- Roughly 400 Palestinian towns and villages were destroyed during the 1948 to 1949 operations, per the Israel Law Review article 1948 Refugees, a measure separating the causes of leaving from the reasons return did not follow.
- 7.7% of the antisemitic comments in a coded sample of 1,504 UK comments on leading media Facebook pages invoked denial of Jewish self-determination, and 5.2% the apartheid analogy (Becker, Ascone and Troschke, 2022).
Which causes account for most 1948 departures?
Roughly 73% of Palestinian departures in 1948 are attributed to Zionist and Israeli military action in the typology Morris assembled from Israeli military and state archives (Morris, 2004), which assigns a primary cause to each depopulated locality rather than to the exodus as a whole.
That 73% is the sum of four assigned categories: 55% of departures following Haganah and Israel Defense Forces military operations, 15% following Irgun and Lehi operations, 2% following explicit expulsion orders and 1% following psychological warfare. A fifth category, fear and a crisis of confidence, accounts for 22%. The five shares listed here sum to 95% of departures, and the remainder is not assigned to any of them, so the figures describe the distribution of assigned primary causes and not a closed partition of every departure.
Two other counts fix the scale. Morris (2004) carries roughly 700,000 displaced as his own archival estimate. The Israel Law Review article 1948 Refugees, a secondary legal-historical discussion summarising Morris, records that roughly 80% of the residents of the Palestinian areas on which Israel was established were expelled or fled, a share of that resident population rather than of Mandate Palestine as a whole.
What the distribution settles is narrow and load-bearing: the largest single assigned cause is military operations, not instruction to depart and not voluntary relocation, and the second largest is fear generated by those operations. What it does not settle is the question the next section takes up, whether the operations proceeded from one directive. The typology counts localities; it does not read minds in Tel Aviv. Morris himself attributes the exodus to causes varying by locality and by phase of the war, and rejects both a single master expulsion plan and the account in which departures were voluntary.
Was the exodus ordered by one central plan?
2% of departures are attributed to explicit expulsion orders in Morris’s typology (Morris, 2004), against 70% attributed to the military operations of the Haganah, the Israel Defense Forces, the Irgun and Lehi. Both figures come from the same archival attribution, and the gap between them is the whole of the scholarly dispute over central planning.
Start with what the record supports without qualification. Expulsions occurred, they were widespread, and they were carried out by armed forces during operations, so the account in which Palestinians simply chose to relocate does not survive the locality files. Morris (2004) is explicit on that point, and it is the historian working from the archives of one of the parties who says it.
The disagreement is over the source of the order. Pappé (2006) argues in The Ethnic Cleansing of Palestine that the 1948 displacement was a premeditated, systematic removal, and the book’s title is the standing home of that characterisation in the scholarly literature. Morris (2004), working from Israeli state and military archives, reports no blanket cabinet expulsion order and describes the exodus as produced by decisions taken at operation and unit level, varying by sector and by phase. Two historians, two dated books, reading an overlapping evidentiary base to opposite conclusions on planning while agreeing that armed force drove the departures.
The 2% figure is often read as small. It measures something narrow: localities where a written or explicitly recorded expulsion order is the primary assigned cause. It is not a measure of how many people were expelled, because an operation that empties a village by attack is counted in the 55% and not here. Reports that set the 2% against the total displacement, without that distinction, are comparing an order count to a population count. The distinct question of whether removal was carried out is treated separately in the Institute’s report on the 1948 expulsion record.
What did Plan Dalet actually authorize?
55% of departures in Morris’s typology follow the military operations governed by Plan Dalet, the Haganah’s general operational directive issued in March 1948 (Morris, 2004), which is the document both sides of the historiographical dispute read and read differently.
As Morris (2004) reproduces the directive, it set out the securing of the areas allotted to the Jewish state and of Jewish settlement blocs beyond them, and it provided, in named circumstances, for operations against population centres that resisted, including the destruction of villages and the expulsion of residents beyond the borders of the state. It also provided that localities that did not resist be garrisoned rather than cleared. The text is conditional on its face, and the conditions are military.
Pappé (2006) reads the same directive as a blueprint for systematic removal, the operational expression of an intent formed earlier. Morris (2004) reads it as a contingency framework that commanders applied unevenly, producing expulsion in some sectors and not others, which is the pattern his locality files record. Both readings rest on the same March 1948 document. The Israel Law Review article 1948 Refugees, summarising the legal-historical literature, treats the directive as the operational context within which the depopulations occurred rather than as a self-executing expulsion order.
The separation worth holding is between the text and the reading of it. What the directive authorised is a matter of record and can be quoted. What it was intended to accomplish is an inference from surrounding documents, and it is the inference, not the text, that the two historians divide on. A reader who is handed only one of the two readings has been handed an inference presented as a document.
How did massacres change the pace of flight?
22% of departures in Morris’s typology are attributed to fear and a crisis of confidence rather than to a force arriving in the locality itself (Morris, 2004), and the events most often named as the source of that fear are the killings at Deir Yassin in April 1948.
Morris (2004) documents the killings at Deir Yassin by Irgun and Lehi forces on 9 April 1948 and the wave of departures from surrounding localities that followed, and he attaches his own caveat to the causal reading: reports and rumours of the killings circulated alongside the fall of nearby towns, other operations and the general breakdown of Arab military and civil authority, so later departures cannot be assigned to the massacre alone. The death toll at Deir Yassin is itself disputed in the literature, and no count appears here as an established total.
The mechanism the typology records is indirect and documented. In the 22% of localities where fear is the assigned primary cause, no force had entered when residents left; what had happened was that a neighbouring place had been attacked or had emptied. That is why the distinction between the 55% and the 22% matters for a causal account. The first counts departures under attack. The second counts departures produced by the credible expectation of attack, which is a different phenomenon with a different evidentiary basis, and collapsing the two into one word loses the part of the causation that operated at a distance.
Read together, the two categories account for 77% of assigned departures, and both run through armed operations, one directly and one by anticipation. Neither category is a claim about the intent of any commander, and Morris’s own text does not present them as one.
Did people leave in one wave or several?
Four stages, not one, is how Morris (2004) groups the exodus across the thirteen months from December 1947 to November 1948, with the mix of assigned causes shifting from stage to stage.
The first stage, December 1947 to March 1948, is dominated by departures from the mixed cities and by families with the means to leave early, before most military operations had begun. The second, April to June 1948, covers the large Haganah operations and the fall of Haifa, Jaffa, Tiberias and Safad, and it is the stage in which the 55% category concentrates. The third, July 1948, includes the expulsions from Lydda and Ramle, the clearest instance in the archives of residents being removed from a town by direct order during an operation. The fourth, October to November 1948, covers the Galilee and Negev operations conducted after the state’s establishment, by then by the Israel Defense Forces rather than the pre-state militias.
The staging is what turns a list of causes into an account. In the first stage, attack is largely absent and the departures are urban and self-organised. By the fourth, the actor is a national army operating under a government, and the localities emptied are rural and peripheral. A single undifferentiated answer, that the displacement was caused by war, expulsions and fear together, is true of the thirteen months as a whole and false of any one of the four stages taken alone.
Did Arab broadcasts tell residents to leave?
It is often said that Palestinians left in 1948 because Arab leaders and Arab radio broadcasts instructed them to. Morris (2004) reports no general evacuation order in the archival and contemporaneous monitoring record he examined, while documenting local evacuation instructions issued in individual places, and his typology assigns 1% of departures to psychological warfare.
The two halves of that finding both matter. Local instructions are documented: in specific villages and sectors, Arab commanders or local committees told residents, often women and children, to move out ahead of expected fighting, and Morris records those cases locality by locality. What the record he examined does not contain is a general directive, broadcast or otherwise, telling the Arab population of Palestine to leave, and the broadcast monitoring of the period does not supply one.
The 1% figure belongs to a different category and should not be enlisted here. It measures departures whose assigned primary cause is psychological warfare conducted by Zionist forces, including broadcasts and loudspeaker warnings directed at Arab localities, not Arab instruction. Departures on local Arab instruction are not among the four categories that compose the 73%, and the five shares listed earlier leave 5% of departures unassigned to any of them, so a reader cannot recover an Arab-orders share from those numbers.
The claim and the record therefore meet at a narrow point. Instructions to evacuate did exist locally, they are documented, and they account for departures in particular places. As an explanation for roughly 700,000 people leaving across thirteen months, the archives Morris worked from record no order of that scope.
Why did the displacement become permanent?
Roughly 400 Palestinian towns and villages were destroyed during the 1948 to 1949 military operations, per the Israel Law Review article 1948 Refugees, and that destruction is the measure that separates the causes of leaving from the reasons return did not follow.
Morris (2004) documents the second set of decisions directly: through 1948 and after, Israeli political and military authorities barred the return of those who had left, blocked infiltration back to villages, and levelled or resettled abandoned localities, so that departure under fire in April became displacement in perpetuity by the following year. Two distinct causal questions sit here, and the answer to the first does not answer the second. Why people left is a question about operations, orders and fear during thirteen months. Why they did not come back is a question about policy after the fighting.
The founding instrument of the same period is a primary record on the population that remained. The Declaration of the Establishment of the State of Israel, read on 14 May 1948 and published in Official Gazette Number 1, appeals to the Arab inhabitants of the State of Israel to participate in the upbuilding of the State on the basis of full and equal citizenship and due representation. Its equality clause differs between the two official English translations: the Knesset text pledges complete equality of social and political rights to all its citizens irrespective of religion, race or sex, while the Ministry of Foreign Affairs text renders the same sentence as all its inhabitants. Both are Israeli government publications translating the Hebrew original, so the wording quoted must be named. The Declaration addresses those inside the state’s control; it says nothing about those already outside it, and the question of how land in Palestine changed hands before and after that date is a separate documentary record.
What did the opened archives change about the account?
The 2004 second edition of The Birth of the Palestinian Refugee Problem Revisited rests on Israeli state and military files covering 1948 that were declassified from the late 1970s onward, and it revises Morris’s own 1988 first edition on the strength of them (Morris, 2004).
What the files added was granularity of the kind the earlier accounts could not reach. Before declassification, the causal question was argued from memoirs, press reporting, diplomatic correspondence and each side’s public statements, which supported a general claim about the exodus but not an attribution per locality. The declassified military and state records allowed a cause to be assigned to each depopulated place, and it is that operation which produces the distribution reported above: 55% following Haganah and Israel Defense Forces operations, 22% following fear, 15% following Irgun and Lehi operations, 2% following explicit orders, 1% following psychological warfare.
The revision cut in both directions, which is the part most easily lost. The archives documented expulsions and destruction more extensively than the pre-declassification accounts had, and they also failed to produce the single blanket expulsion order that a premeditated-plan reading requires. The Israel Law Review article 1948 Refugees reflects both results in summarising how the legal-historical literature absorbed the archival findings, carrying roughly 80% of the residents of the areas on which Israel was established as expelled or fled alongside the operation-level account of how that happened.
One asymmetry belongs on the page. The granular record is drawn from the archives of one party to the war. No comparably opened Palestinian or Arab state archival series underlies the locality attributions, so the distribution is as good as the surviving files of the force that conducted the operations, and Morris’s own text is where its limits are stated.
How does the 1948 causal account travel online?
7.7% of the antisemitic comments in a coded sample of 1,504 UK comments on the Facebook pages of leading British media outlets invoked denial of Jewish self-determination, and 5.2% invoked the apartheid analogy (Becker, Ascone and Troschke, 2022, peer-reviewed), in comments reacting to the May 2021 escalation.
Those shares are of the antisemitic subset, and 26.9% of the 1,504 UK comments were coded antisemitic against an IHRA-based scheme.
Two further measurements bear on how origin framings sit alongside antisemitic expression. Jikeli and colleagues (2022, peer-reviewed) found that 15% of antisemitic tweets in their expert-annotated corpus containing the word Jews fit the IHRA paragraph on denial of Jewish self-determination, on a 4,016-tweet gold-standard sample drawn within keyword queries rather than from all of Twitter. Enstad (2026, peer-reviewed, abstract-sourced) reports across two nationally representative Norwegian surveys, N = 1,575 and a preregistered replication at N = 1,653, that endorsement of Holocaust inversion, the belief that Israel treats Palestinians as badly as Jews were treated during the Second World War, was associated with justification of harassment and violence against Jews.
Ozer, Obaidi and Bergh (2025, peer-reviewed, abstract-sourced) ran two British experiments, N = 354 and N = 490, priming participants with war-in-Gaza material, and found two separate pathways to anti-Jewish hostility: one running through social dominance orientation and political orientation, and one running through out-group identity fusion with the Palestinian people, which conflates attitudes toward Israel with prejudice toward Jews. All of these are associations reported by their authors with the authors’ own caveats, and none establishes that a historical framing of 1948 produces antisemitism.
Methodology and limitations
This report draws on four source types, each labelled where cited: archival history from a university press (Morris, 2004; Pappé, 2006), a secondary legal-historical summary in a peer-reviewed journal (Israel Law Review, 1948 Refugees, Cambridge University Press), a primary and official record (the Declaration of the Establishment of the State of Israel, Official Gazette Number 1, 14 May 1948), and peer-reviewed measurement of contemporary discourse and attitudes (Becker, Ascone and Troschke, 2022; Jikeli and colleagues, 2022; Enstad, 2026; Ozer, Obaidi and Bergh, 2025).
The cause distribution is a single historian’s attribution of one primary cause to each depopulated locality, from the archives of one party to the war (Morris, 2004). It is not a survey of departing residents, and no such survey exists for 1948. Where a locality’s departure had several causes operating together, the typology records the one the files support as primary, so the shares are attributions and not measurements of individual motive. The five categories reported here cover 95% of departures; the remainder is not assigned to them, and no share for departures on local Arab instruction can be derived from these five numbers.
The literature disagrees on planning, and the disagreement is named rather than resolved: Pappé (2006) argues a premeditated systematic removal, Morris (2004) reports no blanket expulsion order in the archives he examined and locates the decisions at operation level. Both readings rest substantially on the March 1948 Plan Dalet directive and on overlapping archival material.
Displacement totals vary by source and by definition of the population counted. Morris (2004) carries roughly 700,000 as his own estimate; the Israel Law Review article records roughly 80% of the residents of the areas on which Israel was established, a share of that resident population and not of Mandate Palestine. The death toll at Deir Yassin is disputed and no figure is carried here as an established total. The Declaration’s equality clause exists in two differing official English translations, both Israeli government publications; the wording is named wherever quoted.
The final section measures discourse, not the historical question. Every figure in it is an association reported by its authors, carried with their sample sizes and caveats and in correlational language, and none of it establishes that any framing of 1948 causes antisemitic expression.
Conclusion
Did people leave in 1948, or were they made to leave? In the archives of the force that conducted the operations, 73% of departures are assigned to Zionist and Israeli military action and a further 22% to fear generated by that action, so the account in which roughly 700,000 people relocated by choice does not survive the locality files.
The rest of the record is equally specific and cuts in more than one direction. The largest assigned cause is military operations, 55%, not written orders, 2%, and the March 1948 directive that governed those operations authorised expulsion in named circumstances rather than everywhere. The exodus ran in four stages across thirteen months, from urban departures before the fighting to army operations in the Galilee and Negev after the state existed, and the cause mix changed at every stage. Local evacuation instructions from Arab commanders are documented in individual villages, and no general order of that kind appears in the archives Morris examined. Roughly 400 towns and villages were destroyed, and the decisions barring return are a second and separate set of decisions from the ones that emptied the villages. The archives opened from the late 1970s produced both halves of that: more documented expulsion than the earlier accounts held, and no single master directive.
Then there is what those findings become once they leave the historiography. In the coded UK comment sample, 7.7% of antisemitic comments denied Jewish self-determination and 5.2% reached for the apartheid analogy; in two Norwegian national surveys, endorsement of Holocaust inversion tracked with justifying aggression against Jews. Origin arguments are among the most-repeated arguments about Israel anywhere, and 1948 is where nearly all of them begin. So the question worth sitting with is how a documented causal record, granular enough to name a cause per village, gets flattened into an origin story that travels with the discourse measured above, and whether that flattening is one of the inputs those studies are picking up. The archives are specific. What is built on top of them, most days, is not.
Frequently Asked Questions
How many Palestinians were displaced in 1948?
Roughly 700,000, Morris’s own archival estimate in The Birth of the Palestinian Refugee Problem Revisited (Morris, 2004). The Israel Law Review article 1948 Refugees puts the displaced at roughly 80% of the residents of the Palestinian areas on which Israel was established, a share of that population rather than of Mandate Palestine.
Why do historians disagree about 1948 if the archives are open?
Because the archives settle what happened in each locality and not what was intended centrally. Morris (2004) reports no blanket expulsion order in the Israeli state and military files, while Pappé (2006) reads the same March 1948 directive and surrounding material as a plan for systematic removal. The disagreement is over inference from documents, not over the documents.
Which Israeli forces were involved, and were they the same body throughout?
No. Morris (2004) assigns 55% of departures to Haganah and Israel Defense Forces operations and 15% to Irgun and Lehi, and the Israel Defense Forces did not exist for the first stage of the exodus. Operations before 14 May 1948 were conducted by pre-state militias; those from October and November 1948 by a national army.
Are Palestinian and Arab archives from 1948 similarly available?
The granular locality attributions used across the literature rest on Israeli state and military files declassified from the late 1970s onward (Morris, 2004). No comparably opened Palestinian or Arab state archival series underlies them, which is why the distribution reflects the surviving records of the force conducting the operations.
What is Plan Dalet, in plain terms?
The Haganah’s general operational directive of March 1948. As Morris (2004) reproduces it, it provided for securing allotted areas and settlement blocs, and in named circumstances for operations against resisting population centres, including destruction of villages and expulsion beyond the state’s borders, while directing that non-resisting localities be garrisoned.
Does the historical record on 1948 bear on antisemitism today?
Only as measured association. Becker, Ascone and Troschke (2022) coded 7.7% of antisemitic comments in a 1,504-comment UK sample as denying Jewish self-determination, and Enstad (2026) found Holocaust-inversion endorsement associated with justifying aggression against Jews across two Norwegian national surveys. Both are correlations reported with their authors’ caveats.
Why did the 1948 displacement not reverse after the fighting stopped?
Morris (2004) documents post-war decisions barring return, blocking re-entry and resettling or levelling abandoned localities, and the Israel Law Review article 1948 Refugees records roughly 400 towns and villages destroyed during the 1948 to 1949 operations. Those are decisions distinct from the ones that produced the departures.
Sources
- Becker, M., Ascone, L., and Troschke, H., 2022. Antisemitic Comments on Facebook Pages of Leading British, French, and German Media Outlets. Humanities and Social Sciences Communications. DOI 10.1057/s41599-022-01337-8. Peer-reviewed.
- Enstad, J. D., 2026. Holocaust Inversion and Justification of Anti-Jewish Aggression: Evidence from Two National Surveys in Norway. American Psychologist. DOI 10.1037/amp0001704. Peer-reviewed (abstract-sourced).
- Israel Law Review. 1948 Refugees. Israel Law Review, Cambridge University Press. https://www.cambridge.org/core/journals/israel-law-review/article/1948-refugees/1E997E364691F4379C6F77EC05BC84AD. Peer-reviewed journal (secondary legal-historical summary).
- Jikeli, G., Axelrod, D., Fischer, R. K., Forouzesh, E., Jeong, W., Miehling, D., and Soemer, K., 2022. Differences Between Antisemitic and Non-Antisemitic English Language Tweets. Computational and Mathematical Organization Theory. DOI 10.1007/s10588-022-09363-2. Peer-reviewed.
- Morris, B., 2004. The Birth of the Palestinian Refugee Problem Revisited, 2nd edition. Cambridge Middle East Studies 18, Cambridge University Press. ISBN 978-0521009676. Scholarly book (archival history).
- Ozer, S., Obaidi, M., and Bergh, R., 2025. The Impact of Globalized Conflicts: Examining Attitudes Toward Jews Among Britons in the Political Context of the War in Gaza. International Journal of Intercultural Relations, Volume 107. DOI 10.1016/j.ijintrel.2025.102184. Peer-reviewed (abstract-sourced).
- Pappé, I., 2006. The Ethnic Cleansing of Palestine. Oneworld Publications. Scholarly book (author-stated moral position).
- Provisional Government of Israel, 1948. Declaration of the Establishment of the State of Israel. Official Gazette Number 1, Tel Aviv, 14 May 1948; English texts published by the Knesset (main.knesset.gov.il) and by the Ministry of Foreign Affairs. Primary and official record.