THE POPULATION OF PALESTINE
THE INSTITUTE FOR PALESTINE STUDIES SERIES
Muhammad Y. Muslih. The Origins of Palestinian Nationalism. 1988
The population of Palestine : population history and statistics of the late Ottoman period and the mandate / Justin McCarthy.
CONTENTS
List of Statistical Tables viii
List of Maps and Charts xiii
Abbreviations xv
Preface xvii
Acknowledgments xxi
CHAPTER ONE: PALESTINE IN THE OTTOMAN EMPIRE 1
I. The Ottoman System of Population Registration 2
II. The Population of Palestine, 1878-1914 5
III. Migration into Palestine 16
CHAPTER TWO: PALESTINE UNDER THE MANDATE 25
I. Population in 1918 25
II. Population Statistics of the Mandate 27
III. Migration into Palestine 33
IV. The Population of Palestine, 1922-1946 35
V. Summation 37
Notes 39
APPENDIXES
1. Ottoman Statistics 45
2. European Statistics of the Ottoman Period 55
3. Mandate Statistics: Population 63
4. Mandate Statistics: Age Distribution 97
5. Mandate Statistics: Births and Fertility 107
6. Mandate Statistics: Deaths and Mortality 119
7. Mandate Statistics: Marriages, Divorces, and Widowhood 139
8. Mandate Statistics: Urban, Rural, and Population Density 154
9. Mandate Statistics: Migration 169
10. Zionist Statistics 217
Index 237
LIST OF STATISTICAL TABLES
CHAPTER ONE: PALESTINE IN THE OTTOMAN EMPIRE 1
1.1. The Population of Palestine. Estimates of Various Scholars 1
1.2. Administrative Divisions of Ottoman Palestine. Sanjaks and kazas 6
1.3. Published Summaries from Ottoman Population Records 6
1.4. The Ottoman Citizen Population of Palestine by Religious Group, Corrected Figures,
1878 to 1914 7
1.5. Refugees in Palestine, According to Records of the Ottoman Refugee Commission,
1302 (1884-85) 11
1.6. Christian Population of Jerusalem, Acre, and Nablus Sanjaks, Corrected, 1304/1306
to 1330/1331 12
1.7. Jewish Ottoman Subjects in Palestine as Recorded in Ottoman Registers 14
1.8. Ben-Arieh's Estimates of the Population of Large Towns in Palestine 15
1.9. Semseddin Sami's Estimates of Urban Population, 1308 to 1316 15
1.10. Total Population of Palestine in 1914 as Given by Barron, Compared to Actual Ottoman
Data 18
1.11. Population Figures of Arthur Ruppin Compared with Official Ottoman Figures.
Kazas in Which Ruppin Lists a Sizeable Jewish Population 18
1.12. Ruppin's Estimate of Jewish Population in Palestine, 1914 19
1.13. Ruppin's Estimates of the Growth of the Jewish Population of Palestine, 1800 to
1933 19
1.14. Representative Estimates of Jewish Immigration, 1850 to 1914 20
CHAPTER TWO: PALESTINE UNDER THE MANDATE 25
2.1. Jewish Migration into Palestine, 1919 to 1922 26
2.2. Population of Palestine in 1918 26
2.3. Population Loss from 1914 to 1918 by Religion 26
2.4. Proportion of Total Palestinian Population in each Ottoman Sanjak, 1914 and
1922 27
2.5. Comparison of Ottoman and Mandate Population Statistics. Total Population
as Recorded in 1895 and 1922 29
LIST OF STATISTICAL TABLES ix
2.6. Recorded Population of Palestine in 1922 by Sex and Age Group 29
2.7. Population of Palestine in 1922, as Recorded and Corrected 30
2.8. Recorded Population 0-15 in 1931, by Religion and Sex 30
2.9. Recorded and Corrected Population in 1931, by Religion 31
2.10. Muslims, Recorded Births and Proportion of Deaths Under Age 1, 1923 to 1931 31
2.11. Recorded Muslim Birth and Death Rates (per thousand per year), 1935 to 1946 32
2.12. Adjusted Muslim Births, Deaths, and Natural Increase, 1935-46 32
2.13. Comparison of Jewish Agency and Government Figures on Jewish Immigration 34
2.14. Population of Palestine by Religion, 1922 to 1931, End of Year Estimates 35
2.15. Population of Palestine by Religion, 1931 to 1946, End of Year Estimates 35
2.16. Proportion of Main Religious Groups, 1931 to 1946 36
2.17. Population of Palestine by Ethnic Groups, 1931 to 1946 36
2.18. Population of Palestine by Religion, 1870 to 1946 37
APPENDIX 1. OTTOMAN STATISTICS
A1.1. Number of Urban Households ca. 1868-69 47
A1.2. Urban Male Population of the Cities of Acre and Nablus, ca. 1873 47
A1.3. Male Population ca. 1877, Various Kazas 47
A1.4. Male Population of Acre and Nablus sanjaks, 1878-79 48
A1.5. Urban Male Population of the Kasabas of Acre and Nablus in 1878-79 48
A1.6. Urban Total Population of the Kasabas of Acre and Haifa, ca. 1905-6 48
A1.7. Population of Jerusalem, Acre, and Nablus Sanjaks, 1886-92 49
A1.8. Population of Jerusalem Sanjak, 1884-85 50
A1.9. Population of Acre and Nablus Sanjaks, 1888-89 50
A1.10. Population of Acre Sanjak, 1881-82 50
A1.11. Population of Jerusalem Sanjak and Beirut Vilayet by Religion, 1893-94 51
A1.12. Population of Acre and Nablus Sanjaks ca., 1896 51
A1.13. Population of Acre and Nablus Sanjaks, ca. 1905-6 52
A1.14. Population of Jerusalem Sanjak, 1911-12 53
A1.15. Population of Acre and Nablus Sanjaks, 1911-13 53
A1.16. Population of Jerusalem Sanjak and Beirut Vilayet by Age Group, 1893-94 54
A1.17. Population Density in 1893-94 54
A1.18. Aliens (not registered in the Population Registers), 1893-94 54
APPENDIX 2. EUROPEAN STATISTICS ON OTTOMAN PALESTINE
A2.1. Statistics of G. Schumacher. The City of Acre 57
A2.2. Statistics of Vital Cuinet: Population by Sanjak in Ottoman Palestine 58
A2.3. Statistics of Vital Cuinet: Cities in Palestine 59
A2.4. Statistics of Moses Friedlander (based on Luncz), Jewish Population in 1897 60
A2.5. Statistics of Davis Trietsch, 1906. Cities in Palestine 60
A2.6. Statistics of Davis Trietsch, 1906. Cities in Palestine in Other Years 61
A2.7. Statistics of Arthur Ruppin. Total Population by Kaza in 1915 62
A2.8. Statistics of Arthur Ruppin. City Population, 1915 62
A2.9. Statistics of Arthur Ruppin. Jewish Population in 1915 62
APPENDIX 3. MANDATE STATISTICS: POPULATION
A3.1. Settled Population by Sex and Religion, 1922 to 1946 65
A3.2. Total Population by Sex and Subdistrict, 1922 and 1931 66
A3.3. Settled Population by Sex and Subdistrict, 1922 and 1931 66
A3.4. Nomadic Population by Sex and Subdistrict, 1922 and 1931 67
A3.S. Population by District and Religion and Variation, 1922 and 1931 67
A3.6. Proportional Strength of Religious Groups by Subdistrict, 1922 and 1931 68
A3.7. Total Settled Population by Religion and Subdistrict, 1931 and 1944 68
A3.8. Settled Population by Religion, Subdistrict, Towns, and Villages, Estimate of
December 31, 1946 69
x LIST OF STATISTICAL TABLES
A3.9. Population by Sex, Religion, and Subdistrict, 1922 70
A3.10. Population by District, with Percentage in each Religion, 1922 73
A3.11. Christian Population by District and Sect, 1922 74
A3.12. Population by Religion, Sex, and District, 1931 75
A3.13. Population by Religion, Sex and Subdistrict, 1931 76
A3.14. Christian Population by Subdistrict and Sect, 1931 78
A3.15. Nomadic Population by Subdistrict, 1931 80
A3.16. Population by Mother Tongue, 1922 81
A3.17. Inhabitants of Municipalities by Mother Tongue, 1922 82
A3.18. Population by Mother Tongue, Sex, and Religion, 1931 84
A3.19. Birthplaces of Settled Population by Religion, 1931 85
A3.20. Birthplaces of the Settled Population by Religion, 1931 87
A3.21. Birthplaces of the Urban Population by Religion, 1931 88
A3.22. Citizenship by Sex, Religion, and Ethnic Group, 1931 90
A3.23. Indices of Population of Each Religion, 1922 to 1945 95
A3.24. Percentage of Each Religion in the Settled Population, 1922 to 1945 95
APPENDIX 4. MANDATE STATISTICS: AGE DISTRIBUTION
A4.1. Population by Age Group, Sex, and Subdistrict, 1922 99
A4.2. Urban Areas by Sex and Age Group 99
A4.3. Population of Palestine in 1931 by Age, Sex, and Religion 100
A4.4. The Population of Palestine by Age, Sex, and Religion 104
A4.5. Age Distribution per Hundred by Religion and Sex, 1926 to 1944 105
APPENDIX 5. MANDATE STATISTICS: BIRTHS AND FERTILITY
A5.1. Births by Year, Sex, and Religion, 1922 to 1946 109
A5.2. Recorded Birth Rate by Religion, 1922 to 1946 110
A5.3. Masculinity of Births by Religion, 1922 to 1946 111
A5.4. Live Births by Place of Residence, 1943 112
A5.5. Births by Religion and Age and Sex of Parents, 1943 113
A5.6. Mothers of Children Born in 1938 by Religion, Duration of Marriage, and Number
of Children Born Alive 114
A5.7. Births in 1940 by Religion and Age of Parents 116
A5.8. Mothers of Children Born in 1940 by Age and Number of Children Born Alive 117
A5.9. Mothers of Children Born in 1940 by Number of Children Born Alive to Them
and Religion 118
APPENDIX 6. MANDATE STATISTICS: DEATHS AND MORTALITY
A6.1. Deaths by Religion, Sex, and Year, 1922 to 1946 121
A6.2. Death Rate by Religion, 1922 to 1946 122
A6.3. Deaths by Religion and Age Group, 1926 to 1943, As Recorded in the Blue
Books 123
A6.4. Deaths by Age Group, Sex, and Religion, 1923 to 1931, As Recorded in the
1931 Census 130
A6.5. Deaths by Age Group and Religion 1928 to 1944 133
A6.6. Life Expectancy at Birth by Religion and Sex, 1926 to 1944 134
A6.7. Mortality (per 1,000 at the Beginning of Each Year of Life) 134
A6.8. Infant Death Rate by Religion, 1922 to 1946 135
A6.9. Infant Deaths, 1932 to 1946 135
A6.10. Infant Mortality by Region, 1926 to 1938, Towns 136
A6.11. Infant Mortality by Region, 1926 to 1938, Rural Areas 136
A6.12. Infant Deaths and Infant Mortality in Each Medical Subdistrict. Towns and Rural
Areas, 1939 and 1940 137
LIST OF STATISTICAL TABLES xi
APPENDIX 7. MANDATE STATISTICS: MARRIAGES, DIVORCES, AND WIDOWHOOD
A7.1. Conjugal Condition by Sex and Subdistrict, 1922 141
A7.2. Conjugal Condition in Municipal Areas by Sex, 1922 142
A7.3. Marriages and Divorces Registered in Palestine by Religion, 1935 to 1946 143
A7.4. Marriages and Divorce Rates in Palestine by Religion, 1935 to 1946 143
A7.5. Age, Sex, and Conjugal Condition, 1931 144
A7.6. Conjugal Condition (per 1,000) by Sex and Subdistrict, 1922 and 1931 147
A7.7. Conjugal Condition (per 1,000) by Age Group, Sex, and District, 1931 148
A7.8. Age at Marriage of Brides and Bridegrooms, 1944 and 1945 149
A7.9. Unmarried (per 10,000) by Sex, Religion, and Main Age Group, 1931 149
A7.10. Females Widowed (per 1,000 Females) by Religion and Main Age Group, 1931 150
A7.11. Nomads, Population, and Conjugal Condition by Subdistricts, 1931 151
APPENDIX 8. MANDATE STATISTICS: URBAN AND RURAL, AND POPULATION DENSITY
A8.1. Urban and Rural Population by Subdistrict, 1922 to 1931 155
A8.2. Total Population by Religion, Sex, and Subdistrict, 1922, Urban 156
A8.3. Total Population by Religion, Sex, and Subdistrict, 1922, Rural 157
A8.4. Population in Municipal, Village, and Tribal Areas, 1922 158
A8.5. Population of Municipal Areas by Religion, 1922 158
A8.6. Population of Towns by Religion, Sex, and Subdistrict, 1931 159
A8.7. Population of Rural Areas by Religion, Sex, and Subdistrict, 1931 160
A8.8. Proportion of Urban Population by Religion and Subdistrict, 1931 161
A8.9. Proportion Females, Proportion Foreign Born, and Density, 1931 162
A8.10. Urban Population by Religion and Town, 1931 and 1944 163
A8.11. Rural Settled Population by Religion and Subdistrict, 1931 and 1944 164
A8.12. Population According to Geographic Regions, 1922 and 1931 164
A8.13. Estimated Settled Population of Towns and Villages, 1938 164
A8.14. Estimated Population of Four Main Towns by Religion, 1936 165
A8.15. Estimated Population of Four Main Towns by Religion, 1937 165
A8.16. Four Main Towns, Population in 1922 and 1931, Estimated Population in 1939 166
A8.17. Proportionate Distribution of Urban and Rural Population by Religion, 1931 166
A8.18. Land Area, Population, and Density, by Districts, 1922, 1931, 1938, and 1944 167
A8.19. Population, Area, and Number of Houses, Towns, and Villages, by Subdistrict,
1931 167
A8.20. Total Population and Area by Subdistrict, 1931 168
APPENDIX 9. MANDATE STATISTICS: MIGRATION
A9.1. Immigrants by Year, Religion, and Ethnic Group, 1920 to 1946 171
A9.2. Authorized Immigration into Palestine, 1920 to 1945 171
A9.3. Natural Increase and Migratory Increase by Religion, 1922 to 1946, and by Sex,
1932 to 1946 172
A9.4. Countries of Destination for Emigrants, 1943 to 1945 173
A9.5. Immigrants by Ethnic Group and Country of Origin, 1 April 1939 to 31 December
1945 174
A9.6. Travellers Entering and Registered as Immigrants, 1924 to 1945, by Religion and
Ethnic Group 175
A9.7. Persons Born Outside of Palestine Resident in 1931 by Religion. Proportion per
1,000 175
A9.8. Internal Migration, 1931 176
A9.9. Principal Sources of Jewish Immigration, 1922 to 1944 176
A9.10. Travellers Recorded as Remaining Illegally, 1933 to 1945 177
A9.11. Jews Deported from Palestine, 1941 to 1946 177
A9.12. Persons Other Than Jews Deported from Palestine, 1941 to 1946 178
xii LIST OF STATISTICAL TABLES
A9.13. Number of Illegal Jewish Immigrants Recorded by the Mandatory Government from
1 April 1939 to December 1943 178
A9.14. Arrivals and Departures, 1923 to 1946 179
A9.15. Movement of Residents of Palestine, 1926 to 1937 179
A9.16. Provenance of Persons Registered as Immigrants, 1928 to 1942 180
A9.17. Age Discrimination per 10,000 Immigrants by Ethnic Group, 1935 to 1945 181
A9.18. Proportion of Females per 1,000 Males among Immigrants by Ethnic Group, 1935
to 1945 181
A9.19. Countries of Destination of Residents of Palestine Departing Permanently by Ethnic
Group, 1937 to 1945 182
A9.20. Provenance of Immigrants, 1926 to 1946 18S
A9.21. Previous Occupations of Immigrants by Jews and Non-Jews 216
APPENDIX 10. ZIONIST STATISTICS
A10.1. Estimates of Jewish Population at Various Periods, 1170 to 1929 219
A10.2. Growth of the Population of Tel-Aviv, 1911 to 1929 219
A10.3. Population of Palestine by Religion, 1922 to 1945, Government and Jewish Agency
Figures 220
A10.4. Natural Increase of the Settled Population, 1922 to 1944 220
A10.5. Jewish Population by Age and Sex, 1943 221
A10.6. Population of Places with Large Jewish Populations, 1922 and 1929 221
A10.7. Historical Estimates of the Jewish Population of Jerusalem, 1170 to 1939 222
A10.8. Jewish Population of Principal Places, 1922, 1931, and 1943 222
A10.9. Growth of Jewish Urban Population by Town and Year, 1170 to 1945 223
A10.10. Growth of Jewish Urban and Rural Populations by Settlements and Population,
1170 to 1945 224
A10.11. Distribution of Jewish Urban and Rural Population by Region, Settlements, and
Population, 1945 224
A10.12. Jewish Population of Jerusalem by Communities, Sex, Conjugal Condition, and Age
Groups, 1939 225
A10.13. Jewish Population of Jerusalem by Age and Sex, 1939 226
A10.14. Jewish Immigration, 1919 to 1945. Government and Jewish Agency Figures 227
A10.15. Jewish Immigration by Various Criteria, 1919 to 1923, 1924 to 1931, 1932 to 1939,
1940 to 1945 227
A10.16. Summary of Jewish Immigration by Various Criteria, 1919 to 1945 229
A10.17. Jewish Immigration by Jewish Communities, 1919 to 1934 230
A10.18. Country of Last Residence of Jewish Immigrants, 1925 to 1934 230
A10.19. Country of Last Residence of Jewish immigrants, 1925 to 1931, 1932 to 1939,
1940 to 1945 231
A10.20. Country of Origin of Jewish Population in Towns and Villages, Various Years 231
A10.21. Country of Origin of Jewish Rural Population for Various Periods of
Immigration 232
A10.22. Citizenship of Jewish Immigrants by Year, 1919 to 1934 233
A10.23. Citizenship of Jewish Immigrants by Period, 1919 to 1923, 1924 to 1931, 1932
to 1939, 1940 to 1945 234
A10.24. Citizenship of Travellers Registered at the Jewish Immigration Office, 1927 to
1934 235
LIST OF MAPS AND CHARTS
CHART 1. Population, 1850-1946 xxiii
CHART 2. Proportion, Muslim, Christian, and Jewish, 1922 to 1944 xxiv
CHART 3. The Growth of Population in Selected Cities, 1895 to 1944 xxv
MAP 1. Ottoman Palestine xxvi
MAP 2. Mandate Palestine xxvii
ABBREVIATIONS
1313 istatistik Ottoman Empire, Nezaret-i Umur-i Ticaret ve Nafia, Devlet-i Aliye-i Osmaniye'nin 1313 Senesine Mahsus Istatistik-i Umumisi ("General Statistics of the Ottoman Empire for the year
1313 1895"), Istanbul, 1315.
1330 Niifus Ottoman Empire, Dahiliye Nezareti, Sicil-i Nufus Idare-i Umumiyesi Miidiiriyeti, Memalik-i
Osmaniye'nin 1330 Senesi Niifus litatistiiii ("Statistics of the Ottoman Empire in the Year
1330"), Istanbul, 1330M (1914).
1922 Census Palestine, Report and General Abstracts of the Census of 1922 (compiled by J. B. Barron,
Superintendent of the Census), Jerusalem, 1923.
1931 Census Palestine, Census of Palestine, 1931, Population of Villages, Towns, and Administrative Areas
(by E. Mills, Superintendent of Census), Jerusalem, 1932.
Palestine, Census of Palestine, 1931, Volume 1, Report (by E. Mills, Superintendent of Census),
Alexandria, 1933.
Palestine, Census of Palestine, 1931, Volume 2, Tables (by E. Mills, Superintendent of Census),
Alexandria, 1933.
Abstract (year) Palestine, Department of Statistics, Statistical Abstract of Palestine, Jerusalem.
Bachi, Israel Roberto Bachi, The Population of Israel, Jerusalem, 1976.
Beyrut (year) Ottoman Empire, Beirut Province, Salname-i Beyrut, (hijra year) or Beyrut Salnamesi, (hijra
year), ("Yearbook of Beirut Province"), Beirut.
Blue Book (year) Palestine, Blue Book (year), Jerusalem, yearly from 1926-27 to 1945.
Bulletin, 1939 Palestine, Office of Statistics, Jerusalem, Vital Statistics Bulletin (Annual), no. 1, June 1939,
Jerusalem, 19391?]
Bulletin, 1940 Palestine, Office of Statistics, Annual Bulletin of Vital Statistics, 1940, Jerusalem, 1941.
Bulletin, 1947 Palestine, Department of Statistics, General Monthly Bulletin of Current Statistics, December
1947, Jerusalem, 1947.
xvi ABBREVIATIONS
Department of Statistics of the Jewish Agency for Palestine, Statistical Data on Palestine,
Jerusalem, 1945.
Gurevich, David, Statistical Handbook of Palestine, Jerusalem, Karen Hayesod, 1930.
Palestine, Department of Health, Annual Report for the Year . . . , Jerusalem.
Ottoman Empire, Maliye Nezareti, Ihsaiyat-i Maliye, 1326 ("Statistics of the Finance Ministry,
1326"), Istanbul, 1329.
Palestine, A Survey of Palestine Prepared in December 1945 and January 1946 for the Information of the Anglo-American Committee of Inquiry, Jerusalem, 1946.
Gurevich, David, The Jewish Population of Jerusalem: Based on the Jerusalem Jewish Census,
September 1939, Department of Statistics of the Jewish Agency for Palestine, Jerusalem, 1940.
Gurevich, David, 15 Years of Jewish Immigration into Palestine, 1919-1934, The Jewish Agency
for Palestine, Statistical Bulletin No. 24, 1935, Jerusalem, 1935.
Gurevich, David, Statistical Handbook of Jewish Palestine, 1947, Published by the Department
of Statistics of the Jewish Agency for Palestine, Jerusalem, 1947(?]
Jewish Agency for Palestine, Economic Research Institute, Statistical Handbook for Middle
Eastern Countries, Jerusalem, 1945.
Palestine, Department of Migration, Annual Report (to 1939) and The Statistics of Migration
and Naturalization for the Year . . . (after 1939), Jerusalem.
Istanbul University MS TY 4807.
Istanbul University MS TY 947.
Trietsch, Davis, Palaestina-Handbuch
Palestine Liberation Organization, Economic Department, Central Bureau of Statistics, Palestinian Statistical Abstract, 1983, no. 5, Damascus, 1983.
Great Britain, Palestine Partition Commission, Report, London, 1938.
Michael Roof, Detailed Statistics on the Population of Israel by Ethnic and Religious Group
and Urban and Rural Residence, 1950 to 2010, Center for International Research, U.S. Bureau
of the Census, 1984.
Ottoman Empire, Suriye Vilayeti, Salname-i Vilayet-i Suriye, (hijra year) or Suriye Vilayeti
Salnamesi, (hijra year), ("Yearbook of Syria Province"), Damascus.
Palestine, Supplement to the Survey of Palestine: Notes Compiled for the Information of the
United Nations Special Committee on Palestine, Jerusalem, 1947.
Data
Handbook
Health (year)
Ihsaiyat
Inquiry
Jerusalem
Jewish Immigration
Jewish Palestine
Middle Eastern
Migration (year)
Ottoman Census I
Ottoman Census II
Palaestina-Handbuch
Palestinian Abstract
Report
Roof, Israel
Suriye (year)
Survey Supplement
PREFACE This analysis of population numbers and population change in Palestine during the Late Ottoman and Mandate Periods contains a collection of demographic data on Palestine. For those who wonder how a book with more statistical tables than text can ever be justified a few words of explanation may be appropriate.
If one is interested in the people of a land, and not only in political, diplomatic, and military maneuverings, knowledge of population numbers is valuable in itself. The history of a place cannot truly be known without knowing who lived there, to which ethnic and religious groups they belonged, and how their numbers changed over time. To the student of population history, statistics on mortality and fertility also tell an absorbing tale of what was most important in the lives of all the people. That knowledge is sufficient unto itself. However, many other branches of history also rely on accurate demographic statistics.
Economic and social history are both particularly dependent on population figures: A tax increase divided among 10,000 people is quite different from one divided among 30,000; an immigration of 1,000 potentially affects a settled community of 5,000 much more than one of 20,000. Yet often our knowledge of the historical population of the Middle East is so poor that we do not know whether 10,000 or 30,000 lived in a region. Any improvement in our knowledge of population numbers thus potentially improves our knowledge of many sorts of history.
In Palestine, where demographic change -particularly migration -has shaped the entire political history of the land, population figures take on added importance. The question of who lived in Palestine in Ottoman and Mandate times holds interest for Palestinian Arabs and Israelis alike. More than seventy years after the British took Palestine from the Ottomans, the character of the Palestine population is still a subject of debate -one clouded by questionable statistics, long-accepted "facts" that do not pass scrutiny, and not a little prejudice. Nowhere is this seen so well as in the question of the migrations that altered the population of Ottoman and Mandate Palestine. Some hold that Palestine under the Ottomans was a stable and overwhelmingly Arab society, which was altered by the migration primarily of European Jews. Others contend that Palestine was essentially a desert before Jewish immigration began and that Jewish migration into Palestine was matched by Arab immigration, which was attracted by Jewish economic development. The one evaluates the history of Palestine as disruption and conquest by a group of foreigners who entered the land in great numbers and eventually conquered it. The other sees the history as a conflict between two immigrant forces, one of which triumphed. Obviously, what is needed to resolve the debate is accurate demographic data. That the argument is politically motivated is all the more reason for sound statistical analysis.
It is my hope that this exposition of the population numbers of Palestine will help readers understand the larger history of the region, and answer a few of the question now open to debate. The detailed collection of demographic statistics should allow others more easily to include demographic factors in their own work. While no one is free from prejudices and preconceptions, particularly where the history of Palestine is concerned, historical analysts should at least be guided by a desire for accuracy, no matter the political implications. I have tried my best to provide accurate demographic statistics in this book, and I hope that I have contributed to an accurate picture of the people of Palestine.
PLACE NAMES One of the skills most needed by authors, especially authors who wish to see their arcane works of historical demography published, is the ability to acquiesce gracefully to the wishes of editors. I have acquiesced, although not too gracefully, in the matter of place names. My preference was to transliterate Ottoman place names as they appeared in the sources (e.g., Kuds-i $erif Vildyeti, not Jerusalem Vilayet), but I succumbed to superior wisdom and superior force.
Standard English place names, as they appeared on the documents of the Palestine Mandate, have been used throughout this volume. For reference, the equivalents of Ottoman district names in other languages are as follows: English Ottoman Arabic Jerusalem Kildus al-Quds Jaffa Yafa Jaffa Gaza Gazze Ghazza Hebron Haliliirrahman al-Khalil Acre Akka 'Akka Haifa Hayfa Hayfa Salad Safat Salad Nazareth Nasira al-Nasira Tiberias Tabarya Tabartyya Balqa Belka Balqa Nablus Nablus Nablus Bani Saab Benisap Bani Sa'b Jemain Cemayin Jama'in Jenin Cenin Janin Because usages such as "sub-province" are clumsy and sometimes confusing, Ottoman names for administrative units have been retained: vilayet (province), sanjak (sub-province), kaza (district).
ON APPROXIMATIONS AND ESTIMATIONS
Most of the numbers in the first chapters are approximations or estimations -in many cases the result of statistical calculations, which are explained in the notes for those who enjoy such things. Some feel that all such estimations should be rounded, perhaps to the thousands place, so that the reader knows that they are not actual census or registration figures and so that the reader is conscious of the margin of error in all calculations of population. This system has one difficulty -the reader can never check to see if the calculations have been done correctly, but is instead asked to take the writer at his or her word and assume that all is well. Better, I think, that the figures be published as they came off the computer, complete to the last digit. This allows the critic to check both calculations and procedure, not simply the latter. It also allows the critic to exclaim, with the great pleasure that comes when finding a colleague's error, "Aha, a mistake in column four!" The only figures that have been rounded here are those that are not the result of mathematical calculations, but of estimates pure and simple. One hopes that the estimations are always precise and as accurate as the statistical manipulations. It still seems better to notify the reader which is which. Both are estimations and should always be regarded as such.
DEMOGRAPHIC TECHNIQUE Certain techniques of demographic analysis lie behind the population estimates in chapter 1. They allow the projection of population numbers forward
PREFACE xix and backward in time from a particular date as well as the correction of underenumeration of certain age groups in statistical sources. The techniques have been explained elsewhere, both theoretically and as they apply to Middle Eastern populations, and the references to these techniques are to be found in the notes to chapter 1. However, a brief explanation of how they work is not out of place here.
The most important demographic technique is that of projection. A simple formula has been applied throughout this work to projections of population from one year to another: P2 = Pier' where P2 is the unknown population, PI is the known population, e is the base of the natural logarithms, r is the rate of increase per year, t is the number of years between PI and P2.
The formula is universally accepted among demographers. It projects populations asymptotically.
For those right-thinking individuals who have lived long lives without knowing what "asymptotically" means, the formula simply tells what a population will be if it increases (or decreases) at a constant rate for a given number of years.
The second demographic technique is a tool for estimating the degree of undercounting of young children in censuses and other population lists, a condition that existed in both Ottoman and Mandatory population counts, as well as in modern Middle Eastern censuses. Ansley Coale and Paul Demeny have created a detailed set of tables that show approximately what proportion of the population should be in each age group in various populations with stable rates of mortality and fertility. One can take a recorded population, such as that of Palestine, and compare the age groups in the recorded population with those in the tables. While the comparisons are never exact, they do allow estimations of the numbers of the population at the younger ages who have not been recorded. To take a general example: If a population census for a high-mortality area where fertility is assumed to be the same every year (approximately true in traditional Middle Eastern societies) shows the same number of males in the ages 0-4 as in the ages 5-9 group, we know that something is wrong, because half the males born in the high mortality area should have died before they reached age S. Knowledge of certain demographic conditions (fertility, rate of increase, etc.) and comparison to the more accurate statistics for the upper ages allow us to select a table from the Coale and Demeny collection that estimates how many children must actually have been in the 0-4 age group. The population estimate can then be corrected by adding in the missing children.
As must be the case with all such nontechnical explanations, many more factors affect the estimation procedure than have been given above. Migration and possible changes in mortality and fertility patterns must particularly be taken into consideration.
The demographic questions asked in the text are simple ones: "What was the population of Palestine?" "What was the relative size of the different communities in Palestine at different times?" "How did migration affect the population?" As is often the case in population history, simple questions demand complicated answers. Whenever possible, I have explained my analyses without resorting to the arcana of demographic formulae, which have been relegated to the notes.
ACKNOWLEDGMENTS Demographic history, like all history, is built on the work of scholars who have done previous work in the field. In particular, the collection and publication of Ottoman statistics by Professors Stanford J. Shaw and Kemal Karpat and the demographic analyses of Professors Roberto Bachi and U.O. Schmelz must be noted. The statistical materials in the book have been collected over a long period under grants from the Social Science Research Council, the National Science Foundation, the Institute for Palestine Studies, and the University of Louisville. For their assistance in bringing the manuscript to completion, I thank Linda Butler, Rita Jones, Rita Hettinger, Dolores Locke, and, in particular and as always, my wife, Beth.
2.0 1.5 1.0 0.5 1850 1880 1900 1914 1922 1931 1946 aim 1. Population, 1850-1946
1922 1931 OGIIST IPA RIO 1944 OR/ST IPA Carl Chart 2. Proportion, Muslim, Christian, and Jewish, 1922 to 1944
Chart 3. The Growth of Population in Selected Goes, 1895 to 1944
Map 1. Ottoman Palestine
Map 2. Mandate Palestine
CHAPTER ONE:
PALESTINE IN THE OTTOMAN EMPIRE
Scholars have provided a great number of estimates of the population of ancient and medieval Palestine.
As is often the case with such estimates, there is little agreement among them. One partial list of estimates for the period immediately before the Roman-Jewish Wars destroyed classical Palestine offers widely varying figures. The disparity among the estimates serves to underline the lack of accurate figures available on the population of ancient Palestine, a situation that holds true for medieval Palestine as well. Many of the estimates listed in table 1.1, especially those for the beginning of the Christian era (most of which are in the range of 2.5 million), seem impossibly high.' One can easily suspect the authors of uncon
Table 1.1. The Population of Palestine: Estimates of Various Scholars Year Estimate 960 B.C. 750,000-1,800,000 ca. SO A.D. 1,000,000-6,000,000 ca. 1300 225,000 ca. 1550 200,000 ca. 1800 275,000 SOURCE: Baal, Israel, pp. 4 and S.
sciously adjusting population numbers upward because of their conviction that ancient Palestine was an important place and thus must have possessed a large population. The estimates for the period after 1500 seem a bit low, although not nearly as unreasonable as those for classical times. However, it seems doubtful that there was a net gain in population between 1550 and 1800. Judging by the political history of the sixteenth century, there most probably was a gain in population following the Ottoman takeover of the area from the Mamluks in 1516.
During the reigns of Suleyman the Magnificent (152066) and Selim II (1566-74) there was a temporary revival of the economy of Palestine, including the new Jewish settlements sponsored in Tiberias by Don Joseph Nasi, a banker and adviser to Suleyman.
However, the revival was short and any gain in population would have been slight: the economic conditions and public security needed for sustained population growth were poor during the seventeenth and eighteenth centuries. Although there is no way to be certain, it seems likely that the population of Palestine throughout the seventeenth and eighteenth centuries was not much smaller than it was in 1850 (ca.
340,000), after which the population began to increase (table 1.4).
The problems in evaluating ancient and medieval
2
population arise from a lack of population registra
tion data. The only way to know accurately the population of any region is to count the people individually. This is the principle upon which all modern
population statistics rest. Indeed, the best (although
still deficient) of the estimates in table 1.1, that for
ca. 1550, is based on partial Ottoman registration
lists. The others are guesses, based not on population
counts but on information on grain consumption,
cultivated land, and the like. Relatively complete
enumeration of the population of Palestine began
only in the late nineteenth century.
THE OTTOMAN SYSTEM OF POPULATION REGISTRATION
Ottoman registration of the population of Palestine began soon after the Ottoman conquest in 1516. As in other parts of the Empire, the purpose of registration was primarily fiscal. The Ottomans wished to know whom they could tax. Those who could not be taxed and those, such as the Bedouin, who could not be persuaded to submit to taxation were not counted.
Although the surviving Ottoman registers from this early period can provide informative data on many aspects of Palestinian life and finances,2 they are of limited use as sources for demographic analysis.
Registration of population, as well as of land and taxes, is both a cause and a result of government control. The greater the degree of government control, the greater the effectiveness of registration. Conversely, accurate registration increases governmental power, because it allows the government to know whom it can tax and whom it can conscript. Ottoman registration of population was thus best when the Empire had most control over its citizens. After the sixteenth century, Ottoman power in the provinces declined. The central government did not have the effective local control that was necessary for population registration, which effectively disappeared.
It was only during the reign of Sultan Mahmud II (1808-1839), when Ottoman provincial power began to increase, that registration of Ottoman citizens resumed. By that time, however, the purpose of enumeration had gone beyond the merely fiscal. During the reigns of Mahmud II and his successors, the Ottomans realized that to rebuild their internal and international power they had both to improve their military and increase their revenue. Both steps were necessary and each depended on the other. New soldiers were needed to enforce tax collections as well as to defend the Empire, and new taxes were needed to support the soldiers as well as to buy the new machines and advisers needed for economic development. An efficient system of population registration was needed for both conscription and taxation.
After Mahmud II had destroyed the Janissaries the decayed traditional standing army of the Ottomans-in 1826, conscription became a pressing issue. Collection of information on males as a source of military manpower was critical, and registration for conscription provided the main impetus for the first modern Ottoman population registration, begun under Mahmud II. Muslims, the only religious group subject to conscription, were registered by their military status and availability for service. Non-Muslims, who did not serve in the armed forces, were registered by their economic status, the usual division being "rich," "average," and "poor." Thus Muslims were registered primarily because of their usefulness as soldiers and non-Muslims primarily for their taxpaying ability. Other registers were taken for purely fiscal purposes and the taxes of all groups were entered in registers of land use and taxable animals.
There was, in addition, an underlying political reason for population registration or census. As the nineteenth century advanced, the Ottomans carried the reform of the Empire into new areas. Educational, social, and economic development were added to the original military and political reforms. Knowing the structure and numbers of the population were important for endeavors as diverse as staffing elementary schools and building railroads. And as revolutionary movements among Christian minorities began to threaten the integrity of the Empire, the Ottomans found that information on the ethnic makeup of their provinces was vital. Finally, it should be noted that as modernization of the Empire progressed, the Ottomans began to take what can only be called a scientific interest in their population. Under Abdiilhamid II (1876-1909), data on population characteristics were recorded and published without any immediate governmental justification beyond advancing knowledge of the Empire.
The traditional Ottoman registration system in
the fifteenth and sixteenth centuries usually identified
by name only the household head (i.e., the head male
3
or occasionally a lone female household head), which
was sufficient for most matters of taxation and land
tenure. By the middle of the nineteenth century, the
information recorded generally included the name of
each male and his place in the family (e.g., "Ahmed
son of Mehmet, grandson of Ali" or "Ahmed's cousin
Mahmud"), his age or date of birth, and other relevant characteristics (e.g., "blind," "crippled," etc.).
At first only males were registered, but women began to be included in the 1870s. In early records, each religious group was kept in separate registers.
Later registers kept the religious groups together, but population was always identified by religion.
The Ottoman government kept what amounted to a census bureau (nufus dairesi) in each province and in Istanbul. Rules were sent out from the central bureau, and occasionally from the Sultan himself, setting the guidelines that all population registrars were to follow. At the local level, population officials (niifus memurlan) appointed by the government entered records for each individual by household in registers (tahrir-i nufus), which were kept in larger towns and district capitals. Village leaders and others were required to notify the nufus memuru in the district town of all births, deaths, and migration, and the registers were updated accordingly. When a person died or moved from the village, "dead" or "moved" was written over his name. Those who had permanently moved (as opposed, for example, to merchants away on business or soldiers) were recorded in the register of their new region.
When data from a district or province were found to be inadequate through lack of timely updating or missing registers, an official inquiry was launched and efforts were made to correct the deficiencies.
While not up to the standards of a modern census bureau in a developed country, the Ottoman system was founded on detailed rules of procedure and a relatively competent bureaucracy. Given the financial and educational resources of the Empire, the results were remarkably good.
No actual census was ever taken in the Ottoman Empire, which had neither the vast organizational resources nor the large number of trained personnel required to count the entire population in its farflung territories at once, contemporaneity being a necessary condition of a true census. Thus, what are referred to by many authors as censuses are actually compendia of data from population registers of most (never all) of the provinces in the Empire. When the call for a "census" went out, the district registers available at the time were often simply compiled. In some districts, new registers had to be drawn up and were then compiled when they reached Istanbul. The whole process could take years: Population records from an "advanced" province in Western Anatolia might arrive in Istanbul soon after they were requested by the central government, while new registers from Eastern Anatolia could take as much as five years to be drawn up and dispatched to the capital.
Thus, unlike a true census, whose tables give a "snapshot" view of a country at a specific time, the Ottoman statistics represented the population of sections of the Empire taken at a different time over a span of some years.
The Ottoman government printed only two "censuses" for public consumption. The first was part of a general volume of Empire-wide statistics, the Devlet-i Aliye-i Osmaniye'nin 1313 Senesine Mahsus Lctatistik-i Umumisi ("General Statistics of the Ottoman Empire in 1313 [mali,* 1895-96 Gregorian]," hereafter abbreviated as 1313 Istatistik.) The second was compiled in the last years of the Empire, an updating of a general population collection begun in 1905, Memalik-i Osmaniye'nin 1330 Senesi Ntifi4s istatistiki ("Statistics of the Ottoman Empire for the Year 1330 [mali, 1914-15 Gregorian]," hereafter abbreviated as 1330 Niifus.) The 1330 Ntifus gave total (male and female combined) population statistics for each vilayet (province), sanjak (sub-province), and kaza (district) in the Ottoman Empire, with the exception of some provinces such as Hijaz and Yemen in which enumeration had not been completed. The 1313 Istatistik listed the population only by the larger administrative units-by vilayets and independent sanjaks which were not attached to vilayets (such as Jerusalem), with no breakdowns for smaller administrative subdivisions. The 1313 istatistik was nonetheless valuable in that it listed population by age group and sex.
At various times the Ottoman government also
compiled detailed population lists for its own internal bureaucratic use. These lists, some of which are
available in archival collections, were often as detailed as the published "censuses" if not more so. This term refers to the fiscal year used by the Ottoman administration; it differs from the hijra year used in other contexts.
4
Two of these have been used here: Istanbul University MS TY 947 (hereafter called Census II), which
was discovered in the Istanbul Library by Professor
Stanford J. Shaw, and "The Census of 1881/2-1893"
(hereafter called Census I). Two copies of the latter
have been found, one in the library of Istanbul University by Professor Shaw and another in the Bal
Bakanlik Ar§ivi ("Prime Minister's Archives" in
Istanbul) by Professor Kemal Karpat. Both "censuses" were compiled in the 1880s and 1890s. Professor Karpat has published Census I with detailed
comments.3 Many more of these compilations were
undoubtedly assembled than have been found to date:
perhaps 95% of the millions of documents in the
Ottoman Archives remain uncatalogued despite a recent acceleration in the cataloguing process.
In addition to the detailed Empire-wide compendia, provincial governments published tables of population in the salnames ("yearbooks") of the provinces. They often provide very detailed information on population; sometimes they included statistics on migration and other demographic variables. Unlike the censuses, they were published locally and closer to the sources of the data. Beyond serving various governmental uses, the salnames seem to have been published with the intent of informing the educated Ottoman public of the status of the province. From the inclusion of pictures of new buildings and descriptions of modernization work being carried out, it is obvious that the salnames were a point of pride for the provincial governments.
Ottoman population records were extensive and consistent. They were part of a regular system of population registration that intended to provide the central government with an accurate picture of the Empire's population, and are valuable demographic resources. However, this does not mean that the Ottoman statistics were correct to the last digit, and present-day researchers must realize that the Ottomans were satisfied with close approximations. In addition to problems of lack of contemporaneity already discussed, the Ottoman data also lacked the uniformity expected in modern censuses. Categories of entries changed (e.g., Samaritans might be counted as Jews in one list, as a separate religion in another) and administrative boundaries shifted. If one remembers that the data were intended primarily for administrative use, many of these anomalies make sense.
Administrators and bureaucrats, the main readers of the reports, would have known when kazas were switched from one sanjak to another or kaza names changed, and they would have been well aware of the process of population registration.
Above and beyond the lack of uniformity and contemporaneity, the two specific problems with Ottoman population statistics are dating and undercounting. With the exception of a few population tables in Anatolian salnames, Ottoman statistics were published without listing the dates on which the data were collected, and one can almost always be sure that data published in one year were actually collected two or more years earlier. For statistics printed in salnames, two or three years delay before publication is the rule. For larger compilations made by the central government, statistics may be many more years out of date. The only way to date the larger compilations is by comparison to the data in the salnames, for which years of compilation are known.4 In undercounting the population, Ottoman population records contained the same type of errors found in all Middle Eastern censuses to this day. Because of the social and economic conditions of the region, Ottoman population registrars uniformly underenumerated women (once they were entered at all) and children. The reason for undercounting women is obvious-the difficulty of penetrating the sacredness and privacy of the home. As for children, it should come as no surprise that they were sometimes overlooked in a system that was based on conscription, their existence hidden by their parents. Counting children has proved to be a difficult task for census takers throughout the world, even where conscription is not a factor. Luckily, such undercounts are correctable. Female numbers can be assumed to be roughly the same as the numbers of males. While this is not a completely satisfying procedure (selective loss of males in warfare could mean slightly more females) it does not produce gross errors. Numbers of children can be estimated accurately through recourse to standard population tables which calculate the number of children when the number of adults is known. This is possible because a certain number of children are necessary if the adult population is to reach a certain size. Given a stable mortality schedule, x children must be born and go through the mortality schedule before the cohort can arrive at age y. This type of analysis necessitates treating the
5
population as if it were a stable population, which it
was not; but the deviation from stability was small
enough to have had little effect on the broad analysis
made here.
Taking these factors into account, a set of correction factors for the population of the Palestine provinces of the Ottoman Empire has been devised.5 The population as recorded is adjusted by multiplying the uncorrected figures by the correction factor, thus allowing for undercounting of females and children.' The correction factors for data in which male population is given (males x 2 x factor) or when only total population is given (total x factor) are:
For Males For Totals Acre and Nablus 1.1778 1.1977 Jerusalem 1.0463 1.0751 The correction factor for Acre and Nablus is the factor for their home province, the Beirut Vilayet.
Factors cannot be drawn for individual sanjaks because age-specific data, essential to the process of correction, are available only for vilayets, not smaller administrative subdivisions.7 These difficulties notwithstanding, Ottoman statistics are the best source on Ottoman population. It has repeatedly been demonstrated that the only ones who can properly evaluate population numbers are those who count the population. For the Ottoman Empire, it has been shown that no population statistics but those of the Ottoman government provide usable demographic data. Since the nineteenth century researchers who have wished to study the population of Palestine have recognized the necessity of using Ottoman data; as will be seen below, some with political points to make have even falsified data and then identified it as Ottoman, because they recognized the intrinsic power of population figures published by the only ones actually able to collect those data, namely, the Ottoman government.
THE POPULATION OF PALESTINE, 1878 TO 1914
Palestine was not part of the initial population registration undertaken by Mahmud II, since it was not a region in which either conscription or taxation was important. The Ottoman army at that time was made up primarily of ethnic Turks, who lived mainly in Anatolia and Ottoman Europe, the areas which were thus recorded. Few Palestinian Arabs served in the conscripted forces of the Ottoman Army until some time after. Concerning taxation, contributions to the central government from Palestine during the first half of the nineteenth century were effectively in the hands of tax-farmers, local landlords, and such- none of them particularly well recorded in the central files.
Satisfactory registration of the Palestinian population did not begin until after 1860. As in other Ottoman provinces, registration did not approach completeness until the reign of Abdulhamid II (18761909). By then, effective control by the central government had spread to Palestine and Ottoman population registrars were able to keep registers of the population. As was the case elsewhere in the Empire, at first only males were enumerated, because of the social difficulties of counting females. Although enumeration of females was soon added, it was never as complete as the count of males.
Ottoman registration of the Palestine population was first published in the salname (yearbook) of Syria Province for 1288 (1871-72, table A1.1). The population was broken down by religious group only for the cities: for village areas, only the population as a whole, undifferentiated by religion, was listed. Later salnames gave detailed listings by religious group for each district (kaza).
Analyses of these data are somewhat complicated by changes in administrative borders. At first, all of Palestine was included in the large province of Syria.
By 1886 (table 1.2), Syria had been divided into smaller provinces. Northern Palestine (Acre and Balqa, or Nablus, sanjaks) was included in the new vilayet of Beirut, southern Palestine in the Independent Sanjak of Jerusalem (Kuds-i Serif). The borders of Acre and Jerusalem sanjaks remained stable after that, but Nablus' borders fluctuated, part inside what was later to be Palestine, part outside of Palestine (e.g., Main, now in Jordan), until the late 1880s when its boundaries settled into what was essentially to be the border of Palestine until 1948. In all the statistical calculations in this text, the borders of Nablus are taken to be those that applied after 1890; no region of the River Jordan has been included. (Administrative boundaries have been drawn from the yearbooks of the Ottoman Empire (Salname-i Devlet-i Aliye-i Osmaniye.1).
6
The Ottoman government had no administrative
designation "Palestine." Provincial borders seem to
have been set simply to assure greater administrative
control, not because they were felt in any way to be
natural boundaries. Similar divisions of large provinces into smaller, more manageable ones were made
in other parts of the Empire as well. The creation of
the Jerusalem region as an independent sanjak (i.e.,
not tied to a larger vilayet) was a reaction to the
special position of Jerusalem both religiously and
politically. The Palestine that came under British
Mandate after World War I was, with small differences, made up of the three sanjaks of Jerusalem,
Acre, and Nablus, and it is those three districts (table
1.3) that will be considered here as Ottoman Palestine.8
Table 1.4 is made up of projections from known
Ottoman data, corrected for undercounting of women
and children. For each province, the most accurate
data from the "censuses" and salnames have been
selected and used as base years for projections.* The
Muslim, Jewish, and Christian populations have been
Table 1.2. Administrative Divisions of Ottoman Palestine, ca. 1900. Sanjaks and Kazas' Jerusalem Sanjak Acre Sanjak Balqa Sanjak b Kiidus (Jerusalem) Akka (Acre) Nablus Yafa (Jaffa) Hayfa (Haifa) Benisap (Banisab) Gazze (Gaza) Safat (Safad) Cemayin (Jemain) Haliliirrahman (Hebron) Nasira (Nazareth) Cenin (Jenin) Tabarya (Tiberias) a. Ottoman titles transliterated into Modern Turkish orthography.
b. Titled either Belka (Balqa) or Nablus at different penods.
Table 1.3. Published Summaries from Ottoman Population Records (Uncorrected) Approximate Population Collection Dates SANJAK OF JERUSALEM Census 11 1302 (1884-85) 229,801 Census I 1303 (1885-86) 234,770 1313 fstatistik 1311 (1893-94) 258,860 1330 Niifus 1330 (1911-12) 328,168 SANJAK OF AKAA 1298 Suriye Salnamesi 1296 (1878-79) 36,368 (males only) Census I 1302 (1884-85) 75,882 Census II 1306 (1888-89) 82,668 1318 Beyrut Salnamesi 1313 (1895-96) 95,617 1330 Niifus 1331 (1912-13) 133,877 SANJAK OF NABLUS 1298 Suriye Salnamesi 1296 (1878-79) 53,359 (males only) Census 11 1306 (1888-89) 114,015 1318 Beyrut Salnamesi 1315 (1897-98) 130,212 1326 Beyrut Salnamesi 1321 (1903-04) 138,841 1330 Niifus 1330 (1911-12) 154,563 ' (The populations used in the projections were drawn from: Jerusalem: Census 11(13021; 1313 Istatistik 11311]; 1330 Nufus 1330).
Acre: 1303 Sunye Salname (1299); Census 11(1306); 1318 Beyrut Salname [13131; i 330 Niilus 11.3.311.
Nablus: Census 11 (1306 1; 1318 Beyrut Salname 113151; 1330 Niifus 11.3301.;
7
projected separately and the result for each year added
for the total population. The "Christian" population
includes the designation Kibti Gayri Muslim ("nonMuslim Gypsies") in the Ottoman sources.' Provincial boundaries have been held constant as they
were from the 1880s to the end of the Empire. It
should be noted that the yearly populations in table
1.4 are not absolutely precise. They are projections
based on the assumption that the populations increased evenly from one enumeration to another
(through an excess of birth and in-migration over
death and out-migration), and this is only an approximation, because no population increases at exactly
the same rate every year. The figures have been computed to the single digits so that readers can check
the calculations, but the figures are just as approxi
Table 1.4A. The Ottoman Citizen Population of Palestine by Religious Group, Corrected Figures, 1878 to 1914,
Jerusalem Sanjak
Years Muslims Jews Christians' Total
1295 (1877-1878) 205,016 6,529 21,100 232,645
1296 (1878-1879) 206,749 6,760 21,672 235,181
1297 (1879-1880) 208,497 7,000 22,259 237,756
1298 (1880-1881) 210,260 7,247 22,861 240,368
1299 (1881-1882) 212,038 7,504 23,480 243,022
1300 (1882-1883) 213,830 7,769 24,116 245,715
1301 (1883-1884) 215,638 8,044 24,769 248,451
1302 (1884-1885) 217,461 8,329 25,440 251,230
1303 (1885-1886) 219,299 8,624 26,129 254,052
1304 (1886-1887) 221,153 8,929 26,836 256,918
1305 (1887-1888) 223,023 9,245 27,563 259,831
1306 (1888-1889) 224,909 9,572 28,310 262,791
1307 (1889-1890) 226,810 9,911 29,076 265,797
1308 (1890-1891) 228,728 10,262 29,864 268,854
1309 (1891-1892) 230,661 10,625 30,672 271,958
1310(1892-1893) 232,611 11,001 31,503 275,115
1311 (1893-1894) 234,578 11,390 32,356 278,324
1312 (1894-1895) 237,039 11,815 32,881 281,735
1313 (1895-1896) 239,526 12,256 33,415 285,197
1314 (1896-1897) 242,038 12,714 33,957 288,709
1315 (1897-1898) 244,578 13,189 34,508 292,275
1316 (1898-1899) 247,143 13,681 35,068 295,892
1317 (1899-1900) 249,736 14,192 35,637 299,565
1318 (1900-1901) 252,356 14,722 36,216 303,294
1319(1901-1902) 255,004 15,272 36,804 307,080
1320 (1902-1903) 257,679 15,842 37,401 310,922
1321 (1903-1904) 260,382 16,433 38,008 314,823
1322 (1904-1905) 263,114 17,047 38,625 318,786
1323 (1905-1906) 265,874 17,683 39,252 322,809
1324 (1906-1907) 268,663 18,343 39,889 326,895
1325 (1907-1908) 271,482 19,028 40,536 331,046
1326 (1908-1909) 274,330 19,739 41,194 335,263
1327 (1909-1910) 277,208 20,476 41,862 339,546
1328 (1910-1911) 280,116 21,240 42,542 343,898
1329 (1910-1911) 283,055 22,033 43,232 348,320
1330 (1911-1912) 286,024 22,856 43,934 352,814
1331 (1912-1913) 289,025 23,709 44,647 357,381
1332 (1913-1914) 292,057 24,595 45,372 362,024
1333 (1914-1915) 295,121 25,513 46,108 366,742
a. Including non-Muslim Gypsies of unknown religion ("Kibti Gayri Muslim").
8
mate as figures rounded to the thousands. The estimates for the earlier years (pre-1877, which are listed
only for Palestine as a whole, not the individual sanjaks) are considerably less precise than those for 1877
to 1914. They are based solely on the author's
impression of what the population was, given the
population numbers in 1877. However, the ratios
between the three religious groups before 1877 are
accurate overall; no possible error could change the
relative strengths of the three communities.
MUSLIM POPULATION.
The Arab ethnic and linguistic community in Palestine was made up of Muslims and Christians, the former demographically predominant. After the Arab conquest, completed in Palestine by 640 A.D., conversion of the indigenous
Table 1.4B. The Ottoman Citizen Population of Palestine by Religious Group, Corrected Figures, 1878 to 1914,
Acre Sanjak
Years Muslims Jews Christians' Total
1295 (1877-1878) 55,335 7,033 17,307 79,675
1296 (1878-1879) 56,714 7,061 17,462 81,237
1297 (1879-1880) 58,128 7,089 17,617 82,834
1298 (1880-1881) 59,577 7,117 17,774 84,468
1299 (1881-1882) 61,062 7,145 17,933 86,140
1300 (1882-1883) 62,584 7,173 18,093 87,850
1301 (1883-1884) 64,144 7,202 18,254 89,600
1302 (1884-1885) 65,743 7,230 18,417 91,390
1303 (1885-1886) 67,382 7,259 18,581 93,222
1304 (1886-1887) 69,062 7,287 18,747 95,096
1305 (1887-1888) 70,783 7,316 18,914 97,013
1306 (1888-1889) 72,548 7,345 19,083 98,976
1307 (1889-1890) 73,770 7,375 19,609 100,754
1308 (1890-1891) 75,012 7,405 20,150 102,567
1309 (1891-1892) 76,276 7,435 20,705 104,416
1310 (1892-1893) 77,561 7,465 21,276 106,302
1311 (1893-1894) 78,867 7,496 21,863 108,226
1312 (1894-1895) 80,195 7,526 22,466 110,187
1313 (1895-1896) 81,546 7,557 23,085 112,188
1314 (1896-1897) 83,169 7,765 23,441 114,375
1315 (1897-1898) 84,823 7,980 23,803 116,606
1316 (1898-1899) 86,511 8,200 24,171 118,882
1317 (1899-1900) 88,233 8,426 24,544 121,203
1318 (1900-1901) 89,988 8,658 24,923 123,569
1319 (1901-1902) 91,779 8,897 25,308 125,984
1320 (1902-1903) 93,605 9,142 25,699 128,446
1321 (1903-1904) 95,468 9,395 26,095 130,958
1322 (1904-1905) 97,367 9,654 26,498 133,519
1323 (1905-1906) 99,305 9,920 26,907 136,132
1324 (1906-1907) 101,281 10,193 27,323 138,797
1325 (1907-1908) 103,296 10,475 27,745 141,516
1326 (1908-1909) 105,351 10,764 28,173 144,288
1327 (1909-1910) 107,447 11,060 28,608 147,115
1328 (1910-1911) 109,585 11,365 29,050 150,000
1329 (1910-1911) 111,766 11,679 29,498 152,943
1330 (1911-1912) 113,990 12,001 29,954 155,945
1331 (1912-1913) 116,258 12,332 30,416 159,006
1332 (1913-1914) 118,571 12,672 30,886 162,129
1333 (1914-1915) 120,931 13,022 31,362 165,315
a. Including non-Muslim Gypsies of unknown religion ("Kibti (;ayri Muslim").
9
population to Islam proceeded gradually so that by
Ottoman times the Muslims were the vast majority.
These were almost exclusively Arabic-speakers. NonArabic speakers probably constituted less than 1% of the de jure Muslim population, since the Turkish soldiers and officials and their families stationed in Palestine were recorded not in the local population registers but in their home provinces. (This causes some confusion between Ottoman and Western statements on population, since Western estimates of Muslim population often did not distinguish between Ottoman soldiers and officials on the one hand and locals on the other.) Other than soldiers and officials, there can have been few Turks living in Palestine before World War I. The first Turkish Republican census (1927)10 reported few Turkish refugees from
Table 1.4C. The Ottoman Citizen Population of Palestine by Religious Group, Corrected Figures, 1878 to 1914,
Nablus Sanjak
Years Muslims Jews Christians' Total
1295 (1877-1878) 125,969 380 2,181 128,530
1296 (1878-1879) 127,134 376 2,197 129,707
1297 (1879-1880) 128,310 371 2,213 130,894
1298 (1880-1881) 129,497 367 2,229 132,093
1299 (1881-1882) 130,695 362 2,246 133,303
1300 (1882-1883) 131,904 358 2,262 134,524
1301 (1883-1884) 133,124 353 2,279 135,756
1302 (1884-1885) 134,356 349 2,295 137,000
1303 (1885-1886) 135,599 345 2,312 138,256
1304 (1886-1887) 136,853 340 2,329 139,522
1305 (1887-1888) 138,119 336 2,346 140,801
1306 (1888-1889) 139,397 332 2,363 142,092
1307 (1889-1890) 140,687 328 2,380 143,395
1308 (1890-1891) 141,988 324 2,398 144,710
1309(1891-1892) 143,302 320 2,415 146,037
1310 (1892-1893) 144,627 316 2,433 147,376
1311 (1893-1894) 145,965 312 2,451 148,728
1312 (1894-1895) 147,316 308 2,468 150,092
1313 (1895-1896) 148,678 304 2,487 151,469
1314 (1896-1897) 150,054 301 2,505 152,860
1315 (1897-1898) 151,442 297 2,523 154,262
1316 (1898-1899) 153,196 292 2,571 156,059
1317 (1899-1900) 154,971 287 2,620 157,878
1318 (1900-1901) 156,766 282 2,670 159,718
1319(1901-1902) 158,581 277 2,720 161,578
1320 (1902-1903) 160,418 273 2,772 163,463
1321 (1903-1904) 162,276 268 2,825 165,369
1322 (1904-1905) 164,156 264 2,879 167,299
1323 (1905-1906) 166,057 259 2,933 169,249
1324 (1906-1907) 167,981 255 2,989 171,225
1325 (1907-1908) 169,926 250 3,046 173,222
1326 (1908-1909) 171,895 246 3,104 175,245
1327 (1909-1910) 173,886 242 3,163 177,291
1328 (1910-1911) 175,900 238 3,223 179,361
1329 (1910-1911) 177,937 234 3,285 181,456
1330 (1911-1912) 179,998 230 3,347 183,575
1331 (1912-1913) 182,083 226 3,411 185,720
1332 (1913-1914) 184,192 222 3,476 187,890
1333 (1914-1915) 186,325 219 3,542 190,086
a. Including non-Muslim Gypsies of unknown religion ("Kibti Gayri Muslim").
10
Palestine with the collapse of the Empire after the
war, and only 634 Turkish speakers appeared in the
1922 Palestine census.
More important than the Turks demographically were the Circassians and other Muslim refugees forced from their lands in the Caucasus after Russia's final conquest of the region in the 1850s and 1860s, following decades of warfare. To deal with the massive
Table 1.4D. The Ottoman Citizen Population of Palestine by Religious Group, Corrected Figures, 1878 to 1914,
All Palestine
Years Muslims Jews Christians' Total
1267 (1850-1851) 300,000 13,000 27,000 340,000
1277 (1860-1861) 325,000 13,000 31,000 369,000
1295 (1877-1878) 386,320 13,942 40,588 440,850
1296 (1878-1879) 390,597 14,197 41,331 446,125
1297 (1879-1880) 394,935 14,460 42,089 451,484
1298 (1880-1881) 399,334 14,731 42,864 456,929
1299 (1881-1882) 403,795 15,011 43,659 462,465
1300 (1882-1883) 408,318 15,300 44,471 468,089
1301 (1883-1884) 412,906 15,599 45,302 473,807
1302 (1884-1885) 417,560 15,908 46,152 479,620
1303 (1885-1886) 422,280 16,228 47,022 485,530
1304 (1886-1887) 427,068 16,556 47,912 491,536
1305 (1887-1888) 431,925 16,897 48,823 497,645
1306 (1888-1889) 436,854 17,249 49,756 503,859
1307 (1889-1890) 441,267 17,614 51,065 509,946
1308 (1890-1891) 445,728 17,991 52,412 516,131
1309 (1891-1892) 450,239 18,380 53,792 522,411
1310 (1892-1893) 454,799 18,782 55,212 528,793
1311 (1893-1894) 459,410 19,198 56,670 535,278
1312 (1894-1895) 464,550 19,649 57,815 542,014
1313 (1895-1896) 469,750 20,117 58,987 548,854
1314 (1896-1897) 475,261 20,780 59,903 555,944
1315 (1897-1898) 480,843 21,466 60,834 563,143
1316 (1898-1899) 486,850 22,173 61,810 570,833
1317 (1899-1900) 492,940 22,905 62,801 578,646
1318 (1900-1901) 499,110 23,662 63,809 586,581
1319 (1901-1902) 505,364 24,446 64,832 594,642
1320 (1902-1903) 511,702 25,257 65,872 602,831
1321 (1903-1904) 518,126 26,096 66,928 611,150
1322 (1904-1905) 524,637 26,965 68,002 619,604
1323 (1905-1906) 531,236 27,862 69,092 628,190
1324 (1906-1907) 537,925 28,791 70,201 636,917
1325 (1907-1908) 544,704 29,753 71,327 645,784
1326 (1908-1909) 551,576 30,749 72,471 654,796
1327 (1909-1910) 558,541 31,778 73,633 663,952
1328 (1910-1911) 565,601 32,843 74,815 673,259
1329 (1910-1911) 572,758 33,946 76,015 682,719
1330 (1911-1912) 580,012 35,087 77,235 692,334
1331 (1912-1913) 587,366 36,267 78,474 702,107
1332 (1913-1914) 594,820 37,489 79,734 712,043
1333 (1914-1915) 602,377 38,754 81,012 722,143
a. Including non-Muslim Gypsies of unknown religion ("Kibti Gayri Muslim").
11
influx of these groups, the Ottomans set up Refugee
Commissions, official government bodies which not
only enumerated the refugees but also distributed
property, subsidies, food, and clothing (never enough
of any of these). Many of the Caucasus refugees were
initially settled in Ottoman Bulgaria, but were forced
to move once again when the Russians conquered
Bulgaria in the 1877-78 Russo-Turkish War. Settled
by the Ottoman government in various regions of
Greater Syria, including Palestine, many did not find
the climate and political conditions salubrious, and
either died or migrated to Anatolia where much larger
numbers had been settled. The title of a table of these
refugees in Palestine (table 1.5) as it appears in the
records of the Ottoman Refugee Commission is indicative: "A population Table which gives the refugees who are settled in the vilayet, except for those
who died or went to Anatolia." As the table demonstrates, the number of those settled in Palestine was
small; even if they were greatly undercounted, there
were probably less than 2,000 refugees who remained, although larger numbers were sent originally. More may also have been settled in the area
around Amman, but they do not concern this study
and are not entered in table 1.5. The 1922 census
recorded only 656 Circassian speakers in all Palestine.
Members of other Muslim ethnic groups lived in Palestine as merchants, workers, or students, but there is no evidence that their numbers were anything but minuscule.
Muslim Arabs were unquestionably the largest section of the population throughout the Ottoman period. Despite the Jewish migrations that began at the end of the nineteenth century, the percentage of Muslims in the population changed only slightly, be cause high Muslim fertility balanced Jewish immigration. In 1880, 87% of the Ottoman citizens in Palestine were Muslim; in 1890, 85%; and in 1914, 83%.
Even if non-citizen Jews, considered below, were included in the figures, Palestine in 1914 was still 77% Muslim, almost all of whom would have been Arabic-speaking.
The Muslim population increased at an average rate of .011 a year, which was approximately the same rate observed for other Ottoman Muslim populations in the same period." This rate of increase demonstrates the radical turnaround in the condition of life in Palestine that occurred in the latter half of the nineteenth century, because such a rate of growth could not have been true in earlier years. The reason is purely mathematical. 441,000 persons in 1878, projected back in time at a rate of .011 per year, becomes twenty-eight persons in the year 1000.
NON-SUNNI MUSLIM GROUPS. The vast majority of the Muslims in Ottoman Palestine were Sunnis, the largest theological group of Islam, whose adherents predominated in Anatolia, Greater Syria, and Eygpt. However, two other sects, the Druze and the Shia, were also present in Ottoman Palestine, in each case the southernmost portion of communities based primarily in what is now Lebanon. Although in religious practice the Shia were relatively close to the Sunni Muslims, they differed in their acceptance of the place of the Imams (the Caliph Ali and his descendants) in their theology and in the great deal of authority given to interpreters of the Holy Law of Islam (the Sharia). The Druze, an esoteric sect based on initiation, differed from Sunni and Shia in both religious practice and authority. Both the Shia and the Druze communities had existed for centuries within
Table 1.5. Refugees in Palestine, According to Records of the Ottoman Refugee Commission, 1302 (1884-85) POPULATION HOUSEHOLDS Male Female Total Circassian From Rumeli' Total Acre Sanjak 583 442 1,025 441 13 454 Nablus Sanjak t' 201 107 308 75 21 96 Jerusalem Sanjak none registered a. Refugees from the Ottoman Balkans during and after the 1877-78 Russo-Turkish War.
b. Many of those listed in the Nablus sanjak were settled in regions east of the Jordan River, at the time part of the province. This area has not been considered as part of Palestine in other sections of this study. No females were listed for one group from Rumeli (64 males listed).
SOURCE: Suriye Salname, 1302, pp. 234, 235.
12
a predominantly Sunni governmental structure. Although theologically heterodox, such groups were
generally tolerated by the Ottomans, as long as they
caused no trouble for the state.
The Ottomans did not separately record the population of either group, because of the age-old position of the Sunni Ottoman government concerning the Shia and other groups. Simply put, the Ottomans did not statistically admit that non-Sunni Muslims existed. Consequently, the only listing in Ottoman population records was "Muslim." The term "Diirzi" (Druze) does appear as a column heading in the 1330 Niifus, but no Druzes were enumerated as such for the three Palestinian sanjaks. They were included within the heading "Muslim." The only serious attempts at recording the Druze and Shia population separately seem to have been in the 1303 Suriye Salname and the 1308 Cebel-i Lubnan Salname, in which both Shia and Druze were listed as such, and greatly underenumerated.
European commentators either did not know the distinctions between the Muslim sects or did not care; their attention was usually focused on Christians and Jews. The one real exception was Vital Cuinet who, in Syrie, Liban, et Palestine gave numbers for non-Sunni Muslims.' However, Cuinet's estimates were generally so poor that one must doubt the accuracy of his data on Druze and Shia.
Whether it is possible to count accurately Shia or Druze Palestinians, who as members of nonconformist religious groups had no desire to be recorded anywhere, is doubtful. Both groups had, after all, continued to exist in a Sunni Empire precisely because of their ability to stay out of the way of the government. The numbers of those styled "Muslims" in the Ottoman records who were Shia or Druze will probably never be known. For purposes of statistical analysis, they are best considered simply as Arabicspeaking Muslims, like their fellows.
CHRISTIANS.
Before the Muslim conquest, the population of Palestine was overwhelmingly Christian, divided between the Greek Orthodox, who were mostly in the cities, and the Jacobites (Syriac Orthodox) in the countryside. Although Aramaic (Syriac) was the predominant language, Arabic had begun to infiltrate the area even before the advent of Islam through migrations from the Arabian peninsula. In the centuries following the conquest, the Christian population was Arabicized and many converted to Islam. Nonetheless, an important minority remained.
During the Ottoman period, the largest Christian sect was the Greek Orthodox, but Ottoman Palestine possessed communities representing most of the Christian denominations in the Empire. In addition to the Orthodox, Latins (Roman Catholics) and UniTable 1.6. Christian Population of Jerusalem, Acre, and Nablus Sanjaks, Corrected, 1304/1306 to 1330/1331 (1886-87/1888-89 to 1911-12/1912-13)
JERUSALEM SANJAK ACRE SANJAK NABLUS SANJAK 1304 1330 1306 1331 1306 1330 Greek Orth. 17,264 27,990 7,163 11,459 1,399 2,159 Armenian Ap. 768 1,408 5 Greek Cath. 454 1,168 8,240 13,212 Protestant 577 1,863 481 931 412 495 Roman Cath. 7,274 10,622 1,736 2,815 537 692 Maronite 290 1,463 1,994 Armenian Cath. 14 Syrian 459 Chaldean 12 Gypsy 122 TOTAL 26,337 43,934 19,083 30,416 2,362 3,346 a. Non-Muslim Gypsies. Religion unknown.
SOURCES: Census 11 and -.3.30 Was.
PALESTINE IN THE OTTOMAN EMPIRE 13 ate Catholics (following traditional Eastern liturgies but owing allegiance to Rome) were also present in sizeable numbers.
The Christian communities increased at a greater rate than the Muslim community, especially in the Jerusalem sanjak. In the case of the Greek Orthodox, most of this increase (.018 per year) can be explained by the fact that the community was generally more urban and less likely to reside in areas of high disease mortality. The growth of the other Christian communities was primarily through conversion, more specifically conversion from the Greek Orthodox sect.
The Greek Catholic population grew twice as fast as the Greek Orthodox in the Jerusalem sanjak. They were originally centered in the Acre, Haifa, and Safad kazas of the Acre sanjak and in the Jaffa kaza of the Jerusalem sanjak. By 1911 there was a small group of more than 600 Greek Catholics in the Jerusalem kaza.
The other Christian groups in table 1.6 grew much more quickly than either the Muslims or the Greek Christians. In Jerusalem sanjak, for example, Protestants increased at a rate of .045 per year, largely from conversion, but also from a small immigration from Europe and America. Other groups, such as the Chaldeans, Syrian Orthodox, and Syrian Catholics, did not even appear in the population records of Jerusalem until the end of the Empire. This rapid increase can only have been the result of in-migration. It should also be remembered that Christians who held European passports would not have been included in the Ottoman statistics. However, the migratory increase of Christians made up only a small proportion of the population of Ottoman Palestine, no more than a few thousand inhabitants.
With the minor exception of foreign-born clergy and a few new settlers, these Christians were Arabs.
JEWISH POPULATION.
Insofar as the historic record is known, a small Jewish community continued to exist in Palestine throughout the centuries.
Their numbers grew under Ottoman rule, increased by refugees from Spain who found haven in Palestine, as elsewhere in the Ottoman Empire, in the fifteenth and sixteenth centuries, and by a small but steady infusion of Jewish pilgrims and students who came to the Holy Land to study and pray or to die.
The Jews of the Old Yishuv ("Old Settlement"), then, comprising Ottoman subjects and immigrants settled in the Holy Land for religious reasons, were primarily urban, divided among four towns in Palestine-
Jerusalem, Safad, Hebron, and Tiberias. The largest settlement in the seventeenth and eighteenth centuries was in Safad, where mystically oriented Jews gathered. A small number of Jews lived on the land, but most were either dependent on the charity of the world's Jews (Halukka) or earned a precarious existence as craftsmen or merchants. By any standard, they were peripheral to the economic and political life of the region.
In the nineteenth century, the Jewish population grew slightly through immigration of European (Ashkenazi) Jews, who came primarily to Jerusalem.
Although they enjoyed favored status under European consular protection, they could not be considered to have been a self-supporting community: in the period following the Crimean War (1853-56), charitable contributions from European Jews increased and constituted the main support of the Jews of Palestine.
From analyses of Ottoman statistical documents it appears that the Jewish population of Palestine was approximately 10,000 in the 1860s. It is unlikely that Jewish numbers had changed greatly since the beginning of the century." The first published figures on Jewish residents in Palestine, the 1288 Syria salname (printed in 1871/72), listed 630 Jewish households in the city of Jerusalem, 1,197 in the urban region (kasaba) of Safad, and 400 in the urban region of Tiberias. Some towns that were later to be centers of Jewish population were still almost completely Muslim or Christian. Jaffa listed no Jewish households, Haifa only 8. To those figures must be added an unknown number of Jewish immigrants who had kept their original citizenship and thus were not registered in the Ottoman tahrir-i nufus (population register). However, there would have been relatively few non-citizen Jews at that early date, perhaps one to two thousand (based on the Ottoman records of non-citizens in 1895 in the 1313 istatistik).
The great immigration of Jews began in the 1880s.
It was from Europe that the population changes originated: the Jews of the Ottoman Empire, by and large integrated into Ottoman economic life, cannot be said to have been particularly interested in the backwater region of Palestine except insofar as, like all Jews, they looked religiously to the Holy Land.
The rise of Zionism in Eastern Europe led to the
14 PALESTINE IN THE OTTOMAN EMPIRE creation of new Jewish settlements in Palestine in the late nineteenth century (the New Yishuv) that eventually were to dwarf the Old Settlement. The first wave of the new Jewish settlers (the First Aliyah) began slowly in 1882 and extended to 1903. In numbers, the migration was small, but it was a significant addition to the minuscule Jewish population. Although most of the new settlers were forced by economic circumstances to work in cities, agricultural settlements were begun. The Second Aliyah (19051914), better prepared and better trained, increased the agricultural and general economic stake of the Jews in Palestine as well as the Jewish population.
Fortuitously, they arrived at a time when Ottoman reforms were beginning to take hold in Palestine and new roads, railroads, and telegraph lines (transportation improved progressively and greatly after 1878) made economic development possible.
The New Yishuv differed considerably from the Old. Secular or nationalistic rather than religious in orientation, European rather than Sephardic in culture, the Zionists or proto-Zionists who constituted the New Yishuv aspired to self-sufficiency, with work rather than prayer as their ideal. It could not be seen
Table 1.7. Jewish Ottoman Citizens in Palestine as Recorded in Ottoman Registers Kaza 1885 1912 Jerusalem 7,414 19,556 Jaffa 454 2,263 Gaza 261 Hebron 460 775 Total 8,329 22,856 Acre 130 126 Haifa 944 2,897 Tiberias 2,538 3,794 Safad 3,733 5,516 Nazareth - - Total 7,345 few' 12,332 Nablus Sanjak a. e.g. only 29 registered in the Nablus Kaza in 1912.
SOURCES: Census 11 and 1330 Census.
N.E. The dates are approximate. 1885 is only a date between the two different dates included in Census II. The dates for the 1330 Niifus were for 1330 (1911-12) and for the sanjak of Acre, so these figures were drawn up by assuming the percentage of Jews in each kaza was the same as in the 1326 Beyrut Salname. The results are thus only rough approximations. but this is sufficient for the points being made.
at the time, but they were the bridgehead of the massive Jewish immigration that was to follow.
As the Jewish population grew through immigration and natural increase, it remained essentially urban despite efforts to develop agricultural colonies.
Some regions of Palestine that had not seen resident Jews for many centuries, such as Gaza, attracted small Jewish settlements.
The position of Jerusalem as center of Jewish life in Palestine was enhanced by Jewish immigration.
The sanjak of Jerusalem increased its percentage of the total population of Ottoman Jews in Palestine by only a bit, from slightly under 50% to slightly over 50%, but to this must be added the considerable number of non-citizen Jews who were not registered in the statistics. This was true of Jaffa and Haifa, as well. While immigration had the effect of slightly spreading the Jewish population, it also resulted in its greater relative centralization, primarily in the coastal towns and Jerusalem. The traditional centers of Tiberias and Safad grew in Jewish population, but lost their relative importance. (See tables A10.6 and A10.9.) SAMARITANS. The Samaritans were a small remnant of the community mentioned in the Bible. They had separated from the main Jewish community in theology, religious practice, and customs. Unique among the inhabitants of Palestine, they had retained the Aramaic language, although they more usually spoke Arabic. In their population records, the Ottomans sometimes listed Samaritans as a separate community, sometimes together with Jews. The Ottoman sources that listed Samaritans separately indicated a small community. According to the 1298 Syria Salname, there were 80 male Samaritans in the Sanjak of Nablus in 1296 (1878-79), while 86 Samaritan males were recorded in the 1303 Syria Salname, 97 in the 1318 Beyrut Salname. If these numbers were close to correct, once adjusted for undercounting they indicate a Samaritan population of approximately 230 ca. 1900, almost all of whom were in the city of Nablus. The 1330 Nfifus (corrected) listed slightly less than 200 immediately before the world war and the 1922 Palestine census (uncorrected) gave 163 Samaritans, 155 in the District of Samaria, 8 in the Jerusalem-Jaffa District. This indicates a considerable diminution of the already tiny Samaritan community before and especially during the war. How-
PALESTINE IN THE OTTOMAN EMPIRE 15 ever, as the Samaritan population was increasing ca.
1900, and seemed to increase from 1922 to 1931 as well, there is a reason for caution in this assessment.
URBAN POPULATION. The population of the cities and towns of Ottoman Palestine is particularly difficult to evaluate. From population registers found in the Ottoman Archives, it is known that the Ottomans kept detailed population registers for cities, which were actually far more complete and up to date than their counterparts in rural areas." But the Ottomans did not appear interested in publishing such statistics, and did not even bother to identify city populations as such in most printed tables of population. None of the "censuses" contained information on city population and the salname population lists seldom mentioned cities. Census and salname tables very seldom went below the kaza (subdistrict) level. Occasional references to city population were made in descriptive passages of the salnames, but seldom in tables.
Another difficulty arises from the nature of the urban population of Palestine. The Ottomans kept lists of the local citizen population (yerli-"of the place," citizens actually residing in the area in question). Even the nonlocal Ottoman citizens (yabana -Ottoman citizens legally registered elsewhere in the Empire) who were in fact living in Palestine were often not listed. The thousands of non-native residents of Palestine, Jews of foreign nationality and others, were by definition not registered and not listed.
In cities such as Haifa, in which many non-citizen Jews were resident, the Ottoman figures represent a real undercount of the actual population.
Therefore, we must fall back on secondary sources of city population. Most of these are European sources.
The Europeans were generally more reliable for urban population than for rural or total population, although their statistics were usually only rough estimates. The best of these have been compiled and analyzed by Yehoshua Ben-Arieh.15 His estimates (table 1.8) are based on European figures. Comparing Ben-Arieh's figures to what Ottoman sources exist, one can see some close similarities. For example, the 1326 Beyrut salname listed 6,945 persons in Haifa ca. 1906, consistent with Ben-Arieh's estimate of 6,000 in 1880. The 1298 Syria salname gave a population for Nablus of 6,625 males for ca. 1880; Ben-Arieh gave 12,500 for both sexes. However, for the city of Acre in 1880, the Ottomans listed only 2,138 males,16 not Ben-Arieh's 8,500 total population. Part of the difficulty may arise from differing definitions of what constituted a city (i.e., how much of the suburban area was included) and from the inclusion of soldiers in the European estimates. However, the numbers still appear quite disparate. By 1906, the Ottomans counted 9,279 in the total population of the city of Acre," indicating either large-scale urbanization or an undercount in the 1880 figures, perhaps both.
The closest thing to an Ottoman governmental estimate of city populations was the set of statistics printed by Semseddin Sami in his monumental work, Kamusidalam." Because he was a government official as well as a great scholar, Semseddin Sami had access to most of the available Ottoman statistics.
Table 1.8. Ben-Arieh's Estimates of the Population of the Large Towns in Palestine 1860 1880 Jerusalem 19,000 30,000 Acre 10,000 8,500 Haifa 3,000 6,000 Jaffa 6,520 10,000 Gaza 15,000 19,000 Hebron 7,500 10,000 Nablus 9,500 12,500 Nazareth 4,000 6,000 Safad 6,500 7,500 SOURCE: Ben-Arieh, "The Large Towns." Table 1.9. 5emseddin Sami's Estimates of Urban Population, 1308 (1890-91) to 1316 (1898-99) Town Population Publication Year Jerusalem 43,000 1314 (1896-97) Nablus 16,000 1316(1898-99) Gaza 16,000 1314 (18%-97) Hebron 11,000 1308 (1890-91) Jaffa 10,000 1316 (1898-99) Acre 8,000 1314 (1896-97) Nazareth 7,000 1316 (1898-99) Haifa 6,000 1308 (1890-91) Safad 5,000 1311 (1893-94) SOURCE: Kan:milk/am.
16
The difficulty with his work, as has been established
elsewhere," is that he did not date his estimates:
there is no way of knowing whether they are for a
date very close to that of the publication of the volume in which the statistics appear or for an earlier
period. Nevertheless, they are worth considering, especially as many of them are close to the estimates of
Ben-Arieh and the available Ottoman official figures
on urban population. Table 1.9 gives Semseddin Sami's
estimates, plus the date they were published.
MIGRATION INTO PALESTINE: MUSLIM IMMIGRATION.
In considering Muslim immigration into Palestine one cannot reasonably avoid the so-called "desertification thesis," which holds that Palestine was largely a wasteland under the Ottomans and only became a truly living land after Jewish settlers arrived. The demographic component of the thesis is that when Jewish immigration began Palestine was an underpopulated area with few Arabs in residence, and that Arabs migrated to Jewish areas in Palestine because of the economic benefits of Jewish settlement. In other words, that the Arab refugees of 1948 were themselves immigrants, or the children of immigrants, and not inhabitants of the land "from time immemorial." 20 The question of the relative economic development of Palestine in Ottoman times is not a matter to be discussed here, but the question of underpopulation should be considered. By modern standards, Palestine was indeed underpopulated. In a world where rapid population growth is endemic, all areas appear "underpopulated" compared to later times. The real issue is immigration of Arabs into Palestine, an issue that resolves into two questions: Was there demographically significant Muslim (Arab) immigration into Palestine in late Ottoman times, and did Muslims migrate into areas of Jewish immigration following the Jews?
First, real evidence for Muslim immigration into Palestine is minimal. Because no Ottoman records of that immigration have yet been discovered, one is thrown back on demographic analysis to evaluate Muslim migration. From analyses of rates of increase of the Muslim population of the three Palestinian sanjaks, one can say with certainty that Muslim immigration after the 1870's was small. Had there been a large group of Muslim immigrants their numbers would have caused an unusual increase in the population and this would have appeared in the calculated rate of increase from one registration list to another.
For example, an increase of one-eighth of the population over a twenty-year period would have caused the observed yearly rate of increase to grow by S0%.* Such an increase would have been easily noticed; it was not there.
The other alternative is a slow in-migration of Arabs over many years. An increase of the Arab population by one-eighth over SO or 75 years would not have been noticeable in the observed rates of increase. However, to postulate such immigration- thousands of Arabs arriving in Palestine each year, during good years and bad-stretches the limits of credulity. Moreover, the phenomenon would have to have gone unnoticed, because it is not mentioned in any of the sources. An increase of many thousands in good economic years, seeking employment in new factories, etc., might be barely believable, except that there were few very good years, there were few factories, and there is no evidence in the statistics. (There was unquestionably seasonal Arab labor in Palestine.
However, these Arabs do not enter the immigration equation, as they were not counted in the Palestinian population registers, but rather in the registers of their own provinces. At least theoretically. Those who might have come from the other side of the Jordan were unlikely to have been registered anywhere.) Second, there is the question of Muslim internal migration. A number of authors have maintained that Muslims migrated to Jewish areas because of better economic conditions, etc.2' The answer is to be found in the economic history of the Eastern Mediterranean in the nineteenth and early twentieth centuries. Some areas of Palestine did experience greater - (Postulate a population of 400,000 with a natural increase rate of .0116 per year. the rate for Palestine from 1295 to 1333 (1878-1914/14 Twenty years later the population would be 504,448. If 50,000 persons (1/81 were added to the population at the beginning of the eleventh year and they too experienced the same rate of natural increase, the population at the end of year 20 would be 560,598. When calculations were made, the rate of increase would appear to be .0169 per year. If 100,000 (1/41 were added, the rate would be .0216. The differences between these figures and .0116 would be very noticeable, i.e., if the rate for one period were .0116, then for the next period .0169, then for the third period .0116, an analyst would suspect in-migration in the second period.
17
population growth than others, but the explanation
for this is simple. Radical economic change was occurring all over the Mediterranean Basin at the time.
Improved transportation, greater mercantile activity, and greater industry had increased the chances for employment in cities, especially coastal cities. At the same time, a population increase, fueled by the same improved security that had contributed to a better economy, had caused the presence of "spare manpower," which could go to the cities for work. Differential population increase was occurring all over the Eastern Mediterranean, not just in Palestine.
The increase in Muslim population had little or nothing to do with Jewish immigration. In fact, the province that experienced the greatest Jewish population growth (by .035 annually), Jerusalem Sanjak, was the province with the lowest rate of growth of Muslim population (.009). The province that experienced the highest Muslim growth, Acre Sanjak (by .020), showed no effect of the supposed drawing power of Jewish immigration. The kaza of Acre, which had little Jewish immigration, had almost the same rate of increase of the Muslim population as did the kaza of Haifa, which was the center of Jewish immigration (.017 per year for Acre as opposed to .018 per year for Haifa, seen by comparing the figures in Census 1 and in the 1330 Nsifus). The major Jewish centers of the kazas of Tiberias and Safad actually experienced lower rates of Muslim population growth than the kaza of Nazareth, which had almost no Jews.
JEWISH IMMIGRATION.
Since the advent of Zionism the question of the number and condition of Jews in Palestine has been one of the most analyzed subjects in Middle East demography. Standard texts on the history of Palestine and Israel present an established view of Jewish population movement to Palestine, indicating a small but relatively stable Jewish population until the mid-nineteenth century, a slow increase as of that time, and larger influxes with the first and second Aliyahs (starting 1882 and 1905, respectively). These points in their broad outlines are not contested. The controversy centers on numbers.
No Ottoman records concerning Jewish immigratin into the Ottoman Empire have been discovered, although it is known from the 1313 Istatistik and from administrative law codes that the Ottoman government registered travelers and immigrants coming into its ports. Nor are there any Jewish records of the early, relatively haphazard migration of Jews; the type of accurate immigration data collected after World War I by the Jewish Agency was not available before that time. Standard evaluations of Jewish immigration have therefore usually been drawn from synthetic analysis.
The starting point of traditional analyses of Jewish immigration is invariably a population figure of almost 85,000 Jews resident in Palestine in 1914, which is then compared to a very small Jewish population fifty years before. The difference between the two figures is assumed to be the amount of Jewish immigration. The figure of 85,000 Jews in Palestine in 1914 derives from J.B. Barron, the first director of the census of Mandatory Palestine, who in turn took it from Arthur Ruppin, the head of the Jewish colonization office in Palestine. As the Barron-Ruppin statistics are central to population research on Palestine, a closer look at them is in order.
THE BARRON/RUPPIN STATISTICS
J.B. Barron was the Superintendent of the census of Palestine taken by the British Mandatory regime in 1922. In the following introduction to the census volume he analyzed previous attempts at enumerating the population, including Ottoman population registration:
The information available in regard to prewar population is accessible from Turkish (i.e. Ottoman] sources, and, through not strictly accurate, it provides data of an interesting nature. The following record was taken in 1914: Mutessariflik of Jerusalem (Southern and Jerusalem-Jaffa Districts 398,362 Mutessariflik of Nablus (District of Samaria) 153,749 Mutessariflik of Acre (Northern District) 137,164 689,275
It is difficult to obtain reliable figures regarding the division of this population into religions, but from other sources (Syrien als Wirtschaftsgebiet, von Dr. A. Ruppin, 1916) the Jewish population in 1914 was calculated to be 84,660, of whom 11,660 were in Jewish colonies and the remainder in towns (1922 Census, p. 3).
Students of Palestinian population have assumed, given Barron's position and the official character of
18
the 1922 Census, that the overall figures he gave for
total population were those actually recorded by the
Ottoman government. However, closer analysis reveals flaws in his assertions and his statistics for
1914. He could not have seen an actual Ottoman
statistical tabulation or population register: he stated
that he had not seen the population broken down
into religious groups (1922 Census, p. 3), and tabulation by religion was always provided in late Ottoman population records. Moreover, the statistics
Barron presented as being Ottoman were at variance
with the 1914 statistics as they were published by the
Ottomans in the 1330 Niifus. The difference between
the real and the putative Ottoman statistics was particularly great for the Jerusalem sanjak, a difference
that could not be explained by scribal error or a
slightly different time of compilation (table 1.10).
In fact, Barron's so-called Ottoman statistics were taken directly from Dr. Arthur Ruppin's book, Syrien als Wirtschaftsgebiet,22 not from any Ottoman Table 1.10. Total Population of Palestine in 1914 as Given by Barron, Compared to Actual Ottoman Data Barron Ottoman' Sanjak of Jerusalem 398,362 328,168 Sanjak of Nablus 153,749 154,563 Sanjak of Acre 137,164 133,877 TOTAL 689,275 616,608 a. As in source, uncorrected.
SOURCES: 1922 Census and 1930 Census.
source. Even though the complete Ottoman archives of the Jerusalem Sanjak were at his disposal and the Ottoman government had presented their 1914 population statistics (in a French version of the 1330 Niifus) to the Versailles Peace Conference, he did not consult Ottoman sources.23 He simply copied Ruppin's statistics and omitted their real source. Barron did not have a tabulation by religion because Ruppin did not give one. Barron indicated that he had used the Ruppin book, although only for statistics of Jewish population, so the deception must have been deliberate.
Leaving aside Barron's inaccuracies, the statistical question is whether Ruppin's figures were accurate.
Ruppin had little understanding of the Ottoman registration system: he stated in Syrien that "since 1902 there is a law compelling all Ottomans to record their names in the official register," whereas in fact the Ottoman law on population registration was passed in 1830 although not applied to Palestine until the latter half of the nineteenth century. Nonetheless, and despite his stated reservations ("No reliable census has ever been taken in Turkey"), he based his book on Ottoman data and called the figures that he used "Ottoman figures." 24 But from the comparison in table 1.11 it is fairly obvious that Ruppin tampered with his "Ottoman" statistics by simply adding population to some kazas. The two kazas to which he added the most population were those to which he attributed the largest Jewish population, Jerusalem and Jaffa, and this was probably no accident. Ruppin was an official of the Zionist Organization and was in charge of its colonization program.
By increasing the total population of certain districts,
Table 1.11. Population Figures of Ruppin Compared with Official Ottoman Figures.25 Kazas in which Ruppin Lists a Sizable Jewish Population TOTAL POPULATION JEWS Ruppin Ottoman Ruppin Ottoman Jerusalem 123,017 120,921 45,000 h 18,190 Safad 31,735 30,561 8,000 4,644 Tiberias 13,102 12,027 4,000 3,194 Jaffa 81,490 72,206 10,000 2,105 a. Place names as they appear in Ruppin.
b. Ruppin's figures are rounded in the text.
N.B. Ruppin's Jewish figures for Jerusalem and Jaffa are for those within the city limits. The actual Ottoman figures for Jerusalem and Jaffa are for the kazas (which include cities).
SOURCES: Ruppin, Syrien and 1330 Was.
19
Ruppin made more plausible his own high estimate
of Jewish population in those districts and in the
whole of Palestine. It might be assumed that Ruppin
was adding to the number of local Jews foreign national Jews not otherwise recorded. However, from
his comments on population it is clear that, at least
in his book Syrien, he did not mean the figures to
include foreign nationals who were not registered in
the Ottoman records, since he indicated that foreign
national Jews, as well as an estimate for an undercount, were to be added to the figures in the table
(p. 9).
Ruppin seems to have come to his basic conclusions about Jewish population and immigration without reference to Ottoman statistics, despite his mention of them. He first stated in 1913 that 85,000 Jews had lived in Palestine in 1908, which he gave as 14% of the total population. His source was given as "Our own calculation," no other reference." In his next book, the much-quoted Syrien als Wirtschaftsgebiet, he had decreased slightly the number of Jews to 83,000 and set the date of the population at 1914 (table 1.12). His final figures were given in The Jews in the Modern World, published in the 19305,27 in which his estimate for the number of Jews in Ottoman Palestine had reverted to his original figure of 85,000 but the date for the figure was set at 1914.
From the figure of 85,000, Ruppin drew up a table of the Jewish population in Palestine in the nineteenth century, and a statement on migration (table 1.13).
It seems obvious that Ruppin had fixed fairly early on a figure of 85,000 Jews and had constructed other figures to match. He had even (in Syrien) "adjusted" Ottoman official statistics to be more in keeping with his own estimates. His estimation was actually a piece of guesswork, grounded on his own ad hoc assumptions. He may have thought that he was adjusting what he felt to be an Ottoman undercount of the Jews in Palestine, or he may have had his own agenda.
Whatever the case, he was not accurately relaying Ottoman data.
Nevertheless, Ruppin's figures, in his own work and as quoted by Barron, have become almost universally accepted28 and have formed the basis for future estimates of the growth of the Jewish population of Palestine through immigration. Statistics such as those given by Bachi (table 1.14) were based on Ruppin, as were those of all the others who have assumed large-scale Jewish immigration in Ottoman times. While the assumption of both Ruppin and Barron that Ottoman statistics were the basic source on Ottoman population was true, the numbers they listed were false. It is better to consult the Ottoman statistics as they were recorded.
STATISTICAL PROBLEMS AND THE WARTIME LOSSES CONTROVERSY
Accepting Ruppin's figure of 85,000 for the Jewish population in 1914 has led to two major problems.
The first relates to the size of the immigration. By assuming a Jewish population of close to 85,000 in Palestine in 1914 (table 1.12) and a very small Jewish population before 1850, writers have concluded that there must have been a disproportionately large Jewish migration before 1914 (table 1.14). In effect, the number of Jewish immigrants was assumed to be whatever number was needed to make up the large 1914 population. What was known of the relative
Table 1.12. Ruppin's Estimate of the Jewish Population in Palestine, 1914 Jerusalem 45,000 Safad 8,000 Tiberias 4,000 Jaffa 10,000 Haifa 3,000 Hebron 1,000 Various Agricultural Colonies 12,000 TOTAL 83,000 SOURCE: Ruppin, Syrien.
Table 1.13. Ruppin's Estimates of the Growth of the Jewish Population of Palestine, 1800 to 1933 1800 10,000 1850 12,000 1880 25,000 1890 35,000 1900 55,000 1910 80,000 1920 75,000 1930 170,000 1933 240,000 SOURCE: Ruppin, Modern World.
20
strength of the various waves of migration was used
to divide up the migration into discrete periods. Because relatively more was known of later (18901914) migrations, the writers were forced to assume
large migrations in the middle of the nineteenth century. This presents a logical flaw: the migrations were
postulated to have been greater in the early years of
bad security and bad transportation in Palestine than
in later years when the situation had greatly improved on both fronts. Moreover, they neglected the
very history of the Jews of Europe, which indicated
that large-scale migration to Palestine began in the
1880s, much later than 1850.
Going to the original Ottoman population statistics one sees a much smaller Jewish population in Palestine. The Ottoman figures indicate only 19,000 (corrected figure) Jewish Ottoman citizens in Palestine in 1893 (Acre Sanjak: 7,563; Nablus Sanjak: 304; Jerusalem Sanjak: 11,390; Total: 19,257) and 38,000 Jewish Ottoman citizens in 1914 (Acre Sanjak: 12,672; Nablus Sanjak: 300; Jerusalem Sanjak: 24,806; Total: 37,778). In 1893, the only year for which such data are available, the Ottomans registered 5,457 resident foreign nationals in the Jerusalem Sanjak and 2,742 in the entire Beirut Vilayet.29 Obviously all the enumerated foreigners were not Jews in Palestine, but even if they had been it would have added only 8,199 Jews to the 1893 figure, bringing the total number of Jews (Ottoman subjects plus foreigners) in Palestine to 28,000. This contrasts sharply with the usually accepted estimates (of which table 1.14 is representative) which puts Jewish immigration prior to 1893 at about 35,000, not including the Jews already resident in Palestine.
Table 1.14. Representative Estimate of Jewish Immigration, 1850 to 1914 Years Jewish Immigrants 1850-1880 25,000 1881-1903 20,000 1904-1910 20,300 1911-1914 14,000 TOTAL 79,300
N.B. One of Bachi's two estimates for 11181-1903, 30,000, has been omitted as unlikely, since even the 20,000 estimate is too high.
SOURCE: Bachi, Israd, p. 79.
The second problem created by the 85,000 figure is the wide discrepancy between it and the widely accepted figures for the number of Jews in Palestine immediately after the war (based on information in the 1922 census of Palestine)-slightly less ihan 60,000. Either the 85,000 figure was in error or something had happened to the Jews not accounted for at the end of the war. The former possibility was not seriously considered. Instead, the statistical loss of close to 25,000-about 30% of the Jewish population-was explained by assuming that great numbers of Jews died or were permanently deported during World War I.
The demographic questions are, first, was there mass migration of Jews from Ottoman Palestine and, second, did the Ottomans cause great mortality among the Jews of Palestine? Many Jews did in fact leave Ottoman Palestine when the war began. Some 600 had been deported from Jaffa to Egypt by the end of 1914, later to be joined by their families, who were transported on the American warship Tennessee." The deported Jews were considered political threats by the Ottoman government because they were subjects of Russia (at war with the Ottomans) or because they were Zionists who, it was believed, advocated the separation of Palestine from the Ottoman Empire. For a time it appeared as if all Jews who had retained their Russian nationality would be deported. However, the German and American governments prevailed upon the Ottomans to allow the Russian Jews to become Ottoman subjects. The Ottomans complied: the Jews were given thirteen months to become naturalized, and the usual naturalization fees were waived first for indigent Jews, then for all Jews.3' Undoubtedly, deficiencies in Ottoman administration caused suffering among the deportees. As Arthur Ruppin, the Zionist official in Palestine, reported: Although these expulsions are no worse than the things being done by all the European nations now at war, and they may even be carried out with greater moderation and decency, the authorities here are particularly incompetent, and this causes much hardship and anger.31 To the Jaffa deportees and their families were added a number who left by their own choice. The Ottomans allowed the Americans to transport Jews who wished to leave Greater Syria, including Palestine, for Alexandria, Egypt. There they joined Jewish
21
refugees from other regions including, somewhat surprisingly, thousands of Jews from countries controlled by the Allies. The British kept no count of the
Jewish migrants to Alexandria, but estimated that
between six and seven thousand persons had come
from "Syria and Palestine" by January 1915, a figure
that included both Jews and non-Jews. By March of
1915, Jews had come to Alexandria from all over the
Mediterranean. An estimated 2,000 Jews had come
from French North Africa alone. The only official
British estimate at the time stated that "some 8,000
to 12,000 Jew [sic] refugees came to Alexandria."33
In this figure were counted the 2,000 from North
Africa, as well as Jews from Anatolia, the Balkans,
and Syria.
Obviously, Jews from Palestine were only part of the refugees in Alexandria. Yet a mythology has arisen that counts all the refugees as emigrants from Palestine. Howard M. Sachar, representing the common belief, has written, "By March 1915 some 10,000 Palestine Jews had found asylum in Egypt."34 The basis of such statistics is easily understood-all Jewish migrants from all over the Mediterranean were listed as coming from Palestine. There is little need to comment on such statistical procedures.
During the war, the English and American press was rife with reports not only of mass expulsions of Jews from Palestine but also of Ottoman mistreatment of the Jews who remained. While few were as extravagant in their accusations as Vladimir Jabotinsky, who contended that "more than 25,000 [ Jews] had died of starvation and disease" in the city of Jerusalem alone,35 the reports were insistent enough to cause considerable consternation in Jewish and Western circles. Accusations of such mistreatment peaked when Palestine became an active war zone.
At that time, Jews were indeed suffering, as were all sections of Palestine society. Food was scarce, although not as scarce as in Lebanon of the same period, and a major locust plague destroyed crops.
Farm animals were seized by the military, young farmers were conscripted. All this led to shortages and increased mortality. In the newspapers of the European Allies and America, however, the Jewish situation was portrayed as being the result of persecution rather than of wartime conditions.
The complaints of ill-treatment of Palestine's Jews that originated in the region (as opposed to those created in foreign newspapers) came from the "Al exandria Palestine Committee."36 The makeup of this committee was not mentioned in their communications, but the fact that they wrote from Alexandria, under the aegis of the British at war with the Ottomans, cannot have been coincidental. Nor can it have been coincidence that the other geographical source of newspaper reports of alleged anti-Jewish disruptions in Palestine was Cairo.37 A sample of an Alexandria Committee telegram reads as follows:
We are informed authentic source with documentary proof holy war organized by military authorities Palestine to break out soon. Inflammatory documents exciting fanatic sentiments of Arab population against English, French, Russian, Christians and all Jews prepared to be spread at moment fixed by leaders of this movement. Documents state every Muslem [sic] obliged at given signal to kill at least three or four Jews and Christians."
The Alexandria Committee reports were printed in Jewish publications in Great Britain, particularly in the Jewish Chronicle. The Chronicle was edited by L.J. Greenberg, whose visit to Lord Newton, the Under-Secretary for Foreign Affairs, produced a Foreign Office report that contained the following: Mr. Greenberg, who appears to be an intelligent and able man, is anxious to know what line he should take in dealing with the matter in the two publications which he edits. He is ready to conform with any indication of policy which may be given to him from the F.0."
The anti-Ottoman propaganda from Cairo and Alexandria, filtered through Greenberg and disseminated throughout the English-speaking world, was widely believed. Nonetheless, the seriousness of the charges prompted calls for independent investigations. In the spring of 1917, the American ambassador in London accused the Turks of treating Jews badly and of organizing massacres on the strength of a report in the Jewish Chronicle, picked up in New York newspapers, that "a massacre is impending in Jerusalem." Asked by his government for the source of the article upon which he based his charges, the ambassador stated that he assumed it had come from British spies. He later found that the report had come from the British High Commissioner in Egypt, hardly a neutral source, who in turn had telegrammed it to the Jewish Chronick.40 Responding to the charges, the United States Department of State asked the Istanbul Legation of neutral Sweden to investigate:
22
"Request Swedish Government to please instruct
Minister Constantinople to protest earnestly for United
States in behalf of humanity against deportation or
massacre Jews."' The U.S. subsequently asked a number of other neutral countries to investigate as well.
The investigations of Ottoman activities extended through July of 1917. It became clear that the charges centered not on massacres but on the Ottoman evacuation of the Jews from the cities of Gaza and Jaffa, and that the facts were very different from what was written in the Jewish Chronicle. According to the Swedish Minister in Istanbul, "Jewish population was sent away without ill treatment, not deported." The Spanish Consul in Jerusalem and Spanish Minister in Constantinople reported no killing or persecutions of Jews, but stated that the Jews and Christians had been evacuated for military purposes and were being allowed to return. Quoting sources in Palestine, the Norwegians found that no anti-Jewish actions had occurred. The Danish Legation said the same thing.42
The final report came from the Swedish Chargé d'Affaires in November 1917. It stated that Jews had been evacuated from Jaffa, most of whom had been sent to Jewish settlements around Tiberias. "Apart from the difficulties and hardships inherent in the situation, there was no rioting nor systematic illtreatment of the Jewish population and above all that there was no such thing as massacres," the report said. Jewish colonies in the vicinity of Jaffa were not evacuated, only those in the city where there was danger of naval bombardment. The greatest suffering for Jews came from the very high cost of living and cessation of business. "There was never a general measure of evacuation applied to the civil population of Jerusalem," the November 1917 Swedish Report continued. "Some 16 to 20 families of Ottoman Jews were sent away to various places in Asia Minor as an administrative measure, either because some of the heads of these families were accused of having dealt in gold illegally or for police reasons. A number of Zionistic [sic] leaders, both Ottoman and foreign, were also obliged to leave Jerusalem and a number of them are now residing in Constantinople."43
The Ottoman Legation at Berne also issued a report, which was disseminated by the Zionist Organization in Berne, entitled "Official Details of the Evacuation of Gaza and Jaffa." The report stated that the evacuation of Gaza and Jaffa had been for military reasons and the safety of the population, the need of which was demonstrated by the widespread destruction in Gaza from British artillery. It said that all groups were evacuated-Muslims, Christians, Jews, Ottoman subjects, and non-subjects. The evacuees were allowed to transport their belongings to Jerusalem or to leave them in their homes. Soldiers were delegated for the protection of personal property and to prevent looting. Jewish evacuees were installed on Jewish farms and villages in the interior." Both the Ottomans and Germans insisted that Gaza and Jaffa had been evacuated to protect the civilian population from expected naval bombardment. Jemal Pasha, the Ottoman military governor, invited German Zionists to tour Palestine to see that no harm had come to Jews. While Ottoman and German statements could be seen as self-serving, the same cannot be said of the assertion of Arthur Ruppin, who reported, "no massacres or cruelties happened and that the Jews of Jaffa were to be allowed to return." 45
Ottomans, Germans, Ruppin, and all the investigating neutral powers agreed that the Ottomans had acted with propriety toward the Jews. Indeed, Jewish lives had been saved by the evacuation of Gaza. The British, whose bombardment of Gaza demonstrated little concern for civilian life, were the source of the accusations against the Ottoman government. The charges were wartime propaganda intended to discredit the enemy." Western opinion, conditioned by centuries of prejudice against the Turks, was all too ready to accept without question any evil attributed to them.
As to those Jews who were jailed and deported, a distinction must be made between oppressing Jews as such and acting against an organized political movement. The Ottomans unquestionably were worried that the Zionists intended to divide off a part of the Ottoman Empire into an independent Jewish state and intended to do all they could to prevent this from happening. This was the cause of the deportations and arrests of small numbers of Zionists during the war. Some were released, others were deported to Anatolia and Istanbul.47 As treatment of suspected separatists in wartime, this was a very light punishment. For example, Ruppin, perceived as the leader of the separatists and personally disliked by Jemal Pasha, was exiled to the Pera Palas Hotel in Istan-
23
bul.48 Arab separatists, on the other hand, were often
hanged, with no uproar in Europe or America over
their fate.
The assertion of great population loss among the Jews in Palestine during World War I rests upon two "pillars"-the migration of approximately 10,000 Jews to Alexandria and high mortality resulting from ill-treatment of the Jews. Neither is true. Because approximately 6,000 Jews came to Alexandria from all of Greater Syria, it is doubtful if more than three to four thousand, at most, could have been from Palestine." The majority of these returned to Palestine with the British conquerors, so their numbers cannot be counted as population loss. The claim of Ottoman persecution of the Jews was wartime propaganda, as attested to by neutral sources. When the spurious justification for Jewish population loss is excluded, it can be seen that it is the claim of population loss that is erroneous. The corollary is that the figure of 85,000 Jews in Palestine is also wrong.
CALCULATION OF JEWISH POPULATION.
If the corrected Ottoman statistics for the number of Ottoman-citizen Jews in Palestine are accepted, 39,000 Jews were legal residents of Palestine in 1914. To these must be added a sizable community of Jews who were not Ottoman citizens. Official statistics of resident noncitizens were published only in 1895 (for the year 1893). From these it appears that the foreign Jews in Palestine cannot have numbered much more than 4,000 of the 8,200 total foreigners in Beirut Vilayet and Jerusalem Sanjak together, assuming arbitrarily that about half the foreigners resided in the north (including present-day coastal Lebanon) and half in three Palestine sanjaks (Acre, Nablus, and Jerusalem). Professor Roberto Bachi has estimated that 34,000 Jewish immigrants came to Palestine from 1903 to 1914 and to these can be added 10,000 more from 1895 and 1903, 44,000 in all. Bachi (p. 79) lists possible figures of 20,000 or 30,000 for the 1881 to 1903 period. A figure of 10,000 as an estimate for 1895-1903 is indicated, because the 30,000 estimate for 1881-1903 is far too high, especially in light of the Ottoman figure, which indicated that approximately 4,000 noncitizens had come (and stayed until 1895) during all the previous period. The numbers may be exaggerated; it is unlikely that Bachi's estimates can be considered as anything but the up per limit of the number of immigrant Jews.5° In any case, his figures must be considered as the number of Jews who arrived in Palestine, not the number who remained. We know from the statistics of Jewish migration during the Mandate period that many Jewish immigrants in the early years of the Mandate did not remain." It would be odd if many Jewish immigrants did not leave the less hospitable conditions in Ottoman Palestine. What contemporary evidence exists indicates this to be true: David Ben-Gurion: "Half the immigrants who came to Palestine in those early days took one look and caught the same ship home again." Indeed, more. Possibly 80% of the Second Aliyeh returned to Europe or continued on to America within weeks or months of their arrival.s2 Conservatively applying the post-war Jewish reemigration rate to the immigrants who arrived before World War I, 11,000 of the 44,000 immigrants would have left again, a very conservative estimation of reemigration. The result is approximately 33,000 Jewish immigrants from 1895 to 1914 who remained in Palestine.
Many of the 33,000 immigrants obviously took Ottoman citizenship or were otherwise counted as yerli ("of the land" or legal resident) and thus were included in the Ottoman population registers. This is the only way to explain the phenomenal rise in Jewish yerli population observed in the Ottoman sources.
Between 1895 and 1914, the resident Jewish population rose from 20,000 to 39,000 (table 1.4). If there had been no immigration, the Jewish population should have risen through natural increase to at most approximately 24,000, not the 39,000 the Ottomans registered. (The figure of 24,000 is based on a rate of .008 per year for 20 hijra years, so that the 1895 figure of 20,117 becomes 23,608. This rate is somewhat higher than the rate of natural increase experienced by Jews in Mandate Palestine [i.e. not counting immigration.] However, to adopt a lower rate would ultimately mean fewer Jews who were counted as non-citizen residents of Palestine [see the analysis in the text] and thus fewer Jews in Palestine. It seems better to err on the other side. No way to correct this is evident.) It is reasonable to assume that the 15,000 "extra" Jews were immigrants, part of the approximately 33,000 who arrived in the period.
The 18,000 immigrants remaining were non-citizen
24
Jews living in Palestine in 1914.53 To these can be
added an estimate of 3,000 for those pre-1895 noncitizen Jews who were still alive and for children of
the non-citizens who were born in Palestine, but not
registered as yet*. (This is an estimation. Many of
the pre-1895 non-citizen Jews may have left, many
surely died, and there is no way to estimate the number of children.) The total Jewish population of Palestine in 1914 was thus approximately 60,000.
(Combining the corrected Ottoman figure with the estimates made here, we have: 38,754 + 18,000 = 56,754. This analysis has throughout intended, whenever exact figures were not attainable, to overestimate rather than underestimate Jewish numbers.
This has been done because the overall calculations presented diminish the usually accepted and exaggerated estimates of Jewish population. It should be noted that the process of correction of Ottoman statistics for undercounting also resulted in an exaggeration of Jewish numbers.
The Palestine Zionist Office took a census of Jews in Palestine during the war which confirms that an estimate of 60,000 Jews in Palestine in 1914 is at least reasonable. The census arrived at a figure of 56,000 Jews. While there is a question as to how accurate a census made by a minority community during wartime could have been, later statistics published by the Jewish Agency were remarkably accurate and this gives some cause to rely on the earlier estimate.54
CHAPTER TWO
PALESTINE UNDER THE MANDATE
Palestine was a combat zone in World War I. After abortive Ottoman attacks on the Suez Canal in 1915 and 1916, the British invaded from Egypt in 1917.
In a series of quick battles, all of Palestine was conquered between October 1917 and September 1918.
Beersheba fell on October 30, 1917, followed by Gaza on November 7 and Jaffa on November 16.
Jerusalem was taken on December 9 and the British line extended to north of Jericho by February, 1918.
After a respite and a limited attack across the Jordan, the British completed the conquest of Palestine in September. By October 1, all of Palestine was effectively in their hands.
Due to the speed of the British conquest, Palestine was spared the sort of slow, grinding, and murderous campaigns that characterized the world war in Anatolia and, to a lesser extent, Iraq. In other areas of conflict between the Ottomans and their enemies, regions had been conquered by one side, lost, and reconquered many times during the war years. Civilian losses, exacerbated by intercommunal conflicts in the war zones, disease, and the impossibility of sowing crops, had been among the most horrible in modern warfare. The Palestine conflict, while it was particularly awful for the Ottoman soldiery, did not cause this wholesale loss of civilian life. Of the cities of Palestine, only Gaza was bombarded and de stroyed, and then only after the Ottomans had evacuated the civilian population. Thus the residents of Palestine were not so subject to killing, but they did suffer the privations of wartime life, and these privations had a significant demographic impact.
POPULATION IN 1918
As stated in chapter 1, there is some debate over the effects of World War I on the population, particularly the Jewish population, of Palestine. To find the demographic impact of the war, one must compare the population of Palestine before and after World War I. After ascertaining the corrected population of Palestine from the 1922 census, calculated below, the population of the major religious groups in 1918 can then be calculated (see table 2.2). The only practical method for doing so is to project the 1922 population back to 1918 at the same rate (.0186189 per year) that applied from 1922 to 1931, subtracting migrants who came to Palestine from 1918 to 1922.
The equation is complicated by the presence in the 1922 population of a large number of immigrants who had arrived since 1918. The largest group of these were Jewish immigrants, who were counted by the Zionist Organization and the figures later published by the Jewish Agency (table 2.1).'
26 Others were those who had come as a result of the British occupation -Anglican and Presbyterian Christians, Hindus, and Sikhs. The figures (corrected) for those excluded are Christians: 5,486 (Church of England 5,084; Presbyterian 402); Hindu 1,620; Sikh 455. The original (uncorrected) figures are in tables A3.9 and A3.11. This procedure does ignore the small number of Anglicans and Presbyterians who might have been in Palestine in 1918 and remained until 1922. These must be subtracted before the population can be projected back, and this has been done in table 2.2. There is no way to estimate the amount of in-migration of Muslims from 1918 to 1922. Judging by recorded immigration after 1922, the number would have been small (see table A9.1 and others in appendix 9).
A comparison of the population of Ottoman Palestine in 1918 and the population in 1914 relates demographically the losses suffered by the Palestinians in World War 1 (table 2.3). During the war years the overall population declined by over 6%. Because
Table 2.1. Jewish Migration into Palestine, 1919 to 1922 Year Immigrants 1919 1,806 1920 8,223 1921 8,294 1922 8,685 TOTAL 27,008'
a. Because these immigrants were figured into the 1922 population as corrected, their number must be multiplied by the 1922 correction factor (1.114164) before being subtracted from the 1922 population. The result is 30,091. This is purely a statistical procedure and does not indicate anything about the Jewish Agency's figures on immigration.
SOURCE: Jewish Agency. Table 12-15.
PALESTINE UNDER THE MANDATE the statistics are imprecise, some of the loss may be an artifact of the statistical method, and losses may be somewhat less than indicated in the table. It must be understood that these statistics are "population loss," not necessarily deaths. Population loss is the result of subtracting those present in Palestine in 1918 from those present in 1914. Such statistics by their very nature are greatly affected by migration.
For example, Muslim families residing in Palestine in 1914 who crossed into Syria or over the Jordan to escape the battles in Palestine and who didn't return after the war would be counted as part of the 6% Muslim population loss. Jews who were deported during World War I and did not return would also be included in the loss figures.
At present, there is no way to ascertain the extent of permanent out-migration, as opposed to death, of those who lived in Palestine in 1914. Certainly, migration of Christians is the only thing that can explain the great population loss -some 13% -experienced by that community. This was not usually true of the Jewish refugees, many of whom were in camps in Egypt and thus would have been likely to return to Palestine after the British conquest, although an
Table 2.2. Population of Palestine in 1918, Projected Back from 1922 Census Figures, As Corrected Muslim 611,098 Jewish 58,728 Christian 70,429 Druze 7,268 Shii 162 Other 443 TOTAL 748,128 Table 2.3. Population Loss from 1914 to 1918 by Religion Muslim' Jewish Christian Total b 1914 657,377' 60,000 d 81,012 798,389 1918 618,528 58,728 70,429 748,128 Difference 38,849 1,272 10,583 50,261 Proportion Lost (.059) (.021) (.131) (.063) a. Including Druze and Shia.
b. Including Other Religions.
c. Including the Ottoman estimate of 55,000 nomads.
d. Including non-citizen Jews.
SOURCES: Table 1.4 with additions and Table 1.20.
27
unknown number did migrate to Europe and, especially, America.
If an attempt is made to estimate wartime mortality, as opposed to population loss, it appears that Christians and Muslims may have experienced a mortality on the order of 4% and the Jews of 2%, but it must be stressed that these estimates are very imprecise.
The Ottoman sanjak of Nablus seems to have suffered greater loss of population than other areas of Palestine, but all regions were affected. The nature of the calculations needed to produce corrected populations and the probability of migration between the regions of Palestine make it impossible accurately to assess relative mortality in the various regions.
(The infusion of new Jewish and other immigrants into the 1922 population affected the results.) However, a comparison of the proportionate population of the three regions of Ottoman Palestine does give an indication of relative loss (table 2.4).
Because so little is known of wartime emigration, it is statistically preferable to consider population loss rather than mortality. The causes of population loss in Palestine were, in addition to emigration, warfare, disease, and hunger -factors shared with much of the rest of the Middle East during World War I.
Palestine did not suffer the great famine that occurred in areas such as Lebanon and Anatolia, but there was hunger and some starvation arising from military confiscation of food and farm animals and a plague of locusts in the spring of 1915. There was, of course, warfare in Palestine and many Palestinian Arabs died while serving in the Ottoman Army. Disease, particularly typhus, was present in wartime Palestine. Therefore, a high level of mortality and emigration are explicable. A loss of 6% was not the worst in the Middle East during the war; Anatolia, for example, suffered a loss of more than 20%.
Table 2.4. Proportion of Total Palestine Population in each Ottoman Sanjak a 1914 1922 Jerusalem Sanjak b .57 .61 Acre Sanjak .20 .21 Nablus Sanjak .23 .18 a. For 1922, the same geographic area as in the Ottoman sanjaks.
b. Including Beersheba nomads (55,000).
Nevertheless, 6% is still a great loss. For comparison, it can be noted that only 1% of the French population was lost in World War I. In that war the English and Germans suffered no loss of total population.
THE POPULATION STATISTICS OF THE MANDATE
Palestine was captured by the British in 1917-18 and administered under military occupation. The San Remo Conference created the boundaries of Palestine and awarded it to the British as a mandate in 1920, confirmed in the League of Nations. On July 1, 1920, the British created a civilian mandatory government in Palestine with Sir Herbert Samuel as the first High Commissioner. A British administration was put into place and the new administrators began to transform much of the governmental system to approximate a British standard. In some areas, such as family law, the British were content to leave the Ottoman system virtually intact.
This was not true of population registration. Other than an occasional archival reference to Ottoman registration practices, the British seem to have taken no notice of past records or procedures in enumerating the population. Indeed, the first superintendent of the Palestine Census, J.B. Barron, had never seen an Ottoman population count, as was clear from his introductory remarks to the census volume (see chapter 1). The British system of population enumeration was thus created de novo. At first, it depended on estimations drawn up by military, then civilian district officials. In 1922, a census was taken and registers of births, deaths, and migration were created.
Yearly figures on population were drawn up by adding the net increase from natural causes and immigration to the 1922 figures, updated after a census taken in 1931, and continuing until 1946.
The population statistics of the Palestine Mandate have always been accepted as accurate, in many cases undeservedly. Like the Ottoman statistics, they are in need of revision. The British Mandatory Power, operating, at least at first, in an environment very similar to that of the Ottoman population registrars, arrived at similar underenumerations of the total population. Nevertheless, the Mandate statistics are far more accessible, and surely more detailed, than
28
the Ottoman statistics. They allow a more sophisticated analysis of population numbers and population
change than can be achieved with Ottoman data.
THE PALESTINE CENSUS OF 1922. The 1922 census was the first actual census ever taken in Palestine. It was designed for a political purpose -the enumeration of Palestine residents by religious group as the basis for proportional voting for a projected Legislative Council. Although the census incorporated some Western techniques of census-taking, the primary difference between it and Ottoman population counts was that all parts of Palestine were enumerated at the same time. It also differed from the Ottoman "censuses" in that it was a de facto compilation, or list of all residents, no matter their usual abode (as opposed to a de jure compilation, a list of legal residents by usual place of residence).
For the census, towns and villages were divided into census districts of approximately 500 houses.
The population of each district was registered by an Enumerator, and the Enumerators were supervised by Revising Officers, each of whom oversaw the work of a group of Enumerators and then compiled their statistics. Results were forwarded to the office of the Superintendent of the Census. In all, 533 Enumerators and 38 Revising Officers were employed.
Although the system of the 1922 Palestine census was superficially similar to that of Western censuses, a number of deficiencies detracted from the census results. It was not possible to name government officials to all the Enumerator positions, so questions of the Enumerator's reliability arose. The counting of a number of Palestine residents was not entrusted to supervised Enumerators at all, but to "heads of ecclesiastical institutions, hospitals, orphanages, boarding schools, etc."2 Military authorities, harbor masters, railway officials, and others were trusted to count those "under their jurisdictions," including civilians who were not members of their groups but living in areas under their control. None of the returns submitted by the heads of institutions, military authorities, et al. were checked in any way. Once again, there is question as to the reliability of such methods.
The main problem of the 1922 census was one of supervision. Enumerators were normally expected to count the inhabitants of 500 houses in six days, an average of 83 houses per day. No matter what assistance they were able to commandeer, the Enumera tors could not have spent much time in checking whether information provided by household heads was correct. In fact, they often seem to have accepted the word of village leaders as to population in the villages.; The result, as might be expected, was an undercount of women and children -the type of undercount seen in all countries where the sanctity of family privacy was of great importance. Manpower was not available for the sort of repeat enumerations (counting some sample areas twice to evaluate accuracy) and cross-checks that help ensure the accuracy of modern censuses. Adding to the difficulties of the census was what the Superintendent of the Census described as the "suspicion" of the inhabitants and the "unpopular" nature of the census.4 The Bedouin of the Southern District (Beersheba Region) would not agree to be counted, so the Bedouin numbers were simply estimated. Such deficiencies have gone unrecognized. It has long been asserted that the census taken by the British in 1922 was the first accurate enumeration of the Palestine population. It was assumed to have been qualitatively different from the Ottoman population counts, because it properly enumerated the Palestinian population, while the Ottoman registration presumably (evidence was never given) did not. The census has been unquestionably accepted as a base for all demographic study of the historic population of Palestine.
In reality, the first British census produced results very much like the Ottoman population enumerations that preceded it. The errors seen in the 1922 census were the same types of errors seen in the Ottoman records. The 1922 census was, as might be expected, superior to the Ottoman counts, but only marginally so. In other words, the results of the 1922 census were exactly what might be expected in an improving series of enumerations. Although the census takers were British civil servants rather than Ottoman civil servants, the populace was seemingly little more forthcoming with correct demographic information. Like the Ottomans, the British undercounted women and children, as demonstrated in table 2.5. The table compares Ottoman and Mandate statistics for certain often undercounted age groups with an estimate of what statistics for those age groups should be. The table indicates that Mandate statistics, although not as imprecise as Ottoman figures, were also deficient. As they had in Ottoman times, Palestinian males overestimated their reported ages
PALESTINE UNDER THE MANDATE 29 in 1922. Numbers of children were underreported in both sets of statistics.
Nevertheless, despite all its difficulties, the 1922 census produced usable data. Its figures were far better than the data available for most other Middle Eastern countries of the time, but those figures are in need of correction. Because of the undercounting of women and children, the population tables in the 1922 census, like their Ottoman counterparts, must be adjusted. The process for correcting for undercounting of children is essentially the same as that followed for Ottoman statistics -comparing the recorded figures to standard tables to ascertain the degree of undercounting, then adding the missing population to the total. However, the process for correcting the undercount of females is a bit different. Unlike the Ottomans, the British did not seriously undercount females beyond the childhood years.
Therefore, only the numbers of younger females need to be corrected. (The 1922 census gave only abbreviated age groups, but it is possible to use them for purposes of correction.) This has been done by doubling the corrected number of males aged 0-15 (table 2.6) to obtain the total population aged 0-15.
The assumption of equal numbers of males and females aged 0-15 is statistically accurate. In a typical population (assuming a population corresponding to a Coale and Demeny Model 8-East), the numbers of males and females alive during the interval 0-5 is Table 2.5. Comparison of Ottoman and Mandate Population Statistics. Total Population as Recorded in 1895 and 1922
Ottoman Mandate Ages Expected a 1895 b 1922 Males 0-15 .44 .35 .40 Males 65 + .02 .04 .04 Proportion Male, 0-5a .50 .53 .53 Proportion Male, All Ages .48 .51 .51 a. In standard tables.
b. Males in Jerusalem Sanjak.
c. Male Populationfrotal Population.
Non: The table used is Coale and Demeny East-Level 8, GRR = 3.S.
The Proportion Male figures are for the stationary population and thus will be marginally different than in a population in which the population is increasing. It is difficult to postulate conditions that would have resulted in more females than males being lost during World War I in Palestine. (See Ansley J. Coale and Paul Demeny, Regional Modal Life Tables and Stable Populations, Princeton, 1966).
SOURCES: 1922 CENSUS and 1313 Istatistik.
almost exactly the same, with only a slight bias towards females (less than 1% if the normal 105 male births per 100 female births are assumed). Of course, systematic biases such as considerably better care and feeding of male children could have affected survival, but there is no evidence of such practices.
The corrected total population is thus the corrected population 0-15 added to the recorded population 15 + .
By applying the correction factor evenly to all segments of the population, the process of correction undoubtedly overestimates the numbers of Jews and some other groups, such as Anglicans and Hindus, because those groups were surely better counted than Muslims and indigenous Christians. (The Jews were mainly urban and more attuned to governmental procedures, which would have ensured a more complete count.) However, the single correction factor must be used to retain comparibility with Ottoman data as corrected. Moreover, the 1922 census does not give age-specific information by religion, so the population must be corrected as a whole. If one assumes that the entire undercount was one of Muslims and Christians, the numbers would be Muslims 667,825 and Christians 82,532, which would surely be too high for both groups. Many Christians were urban and thus well-recorded, and other groups such as the Druze, were probably underenumerated, as well. The real population of the Muslims was probably higher than in table 2.7 and lower than the above.
Table 2.7 lists the corrected population by religion.
THE 1931 CENSUS. Most of the problems that affected the accuracy of the 1922 census were satisfactorily corrected in the 1931 enumeration. The
Table 2.6. Recorded Population of Palestine in 1922 by Sex and Age Group Ages Males Females Total 0-5 67,391 58,820 126,211 5-14 85,778 69,202 154,980 15-24 51,826 62,689 114,515 25+ 182,123 179,353 361,476 TOTAL 387,118 370,064 757,182 SOURCE: 1922 Census.
30
number of houses assigned to each Enumerator was
significantly decreased. Only 50 to 80 houses were
given to each, and twenty blocks were under a Supervisor. A Superintendent in turn oversaw the Supervisors and, most importantly, "tested a proportion of
the work of his subordinates.r'S More than 3,500
Enumerators were included in a staff of 4,000 assigned to the census. It was assumed that Enumerators could not accurately count even 50 houses in the
one day allocated for the census, so the forms for
each house were filled out in advance, then each
house was checked on the census day to see if the
data entered were correct.
The major lacunae in the 1931 census concerned Bedouin. Although an attempt was made to actually count many Bedouin, the Superintendent of the Census admitted that Bedouin were not accurately enumerated. Therefore, estimates of Bedouin population made on the basis of the actual counts and the opinions of those with close knowledge of each tribe were substituted.6 In form, the 1931 census was similar to censuses taken in Europe in the same period. Its data were what Europeans expected to find in a census -not simply population divided into various economic and social categories. Data on literacy and educational level, profession and occupation, religion, language, birthplace, marital status, infirmities, etc., were included. The population was listed by single years of age and sex, important demographic data that gave much better information than age groups given in the 1922 census. In terms of completeness, in the Middle East only the censuses of Egypt and Turkey could compare with the 1931 census of Palestine.
Although the 1931 census was statistically far superior to the 1922 census, it still undercounted women in the younger age groups.' Males were not significantly undercounted, and correction is a simple matter of assuming the recorded male and female popu lations under 15 years of age were the same size and adjusting the total population accordingly. All of the religious groups seem to have been the subjects of undercounts. As might be expected, the Muslims, with the population most heavily rural, were more greatly undercounted. The process of correction yields an additional 15,469 persons for Muslims, 1,326 for Jews, 1,404 for Christians, and 169 for the others.
Druze cannot be separated from the "Others" because they were not listed separately in the tables.
(The caveats on correction procedure noted above apply to the 1931 corrections, as well.)
THE BRITISH SYSTEM OF REGISTRATION OF VITAL EVENTS.
In addition to actual censuses, the British maintained a system of birth and death registration in which they had great confidence, as they did in many things. They did succeed in registering vital events more completely than was managed in other Middle Eastern countries of the time. Registration of vital events was made compulsory in 1920 and put under the charge of the Palestine Department of Health. In cities and towns, midwives and doctors had to complete notices of birth, and doctors those of death. In villages, mukhtars filled out birth and death notification forms with information provided by midwives, parents, and relatives. Because it was recognized that obtaining ac-
Table 2.8. Recorded Population 0-15 in 1931, by Religion and Sex Religion Males Females Muslim 151,807 136,338 Jewish 29,090 27,764 Christian 16,419 15,015 Other 2,047 1,878 SOURCE: 7931 census.
Table 2.7. Population of Palestine in 1922, as Recorded and as Corrected Muslim Jewish Christian Druze Other Total Recorded' 572,992 83,794 73,024 7,028 2,446 739,284 Corrected 638,407 93,360 81,361 7,830 2,726 823,684 Proportion (.775) (.113) (.099) (.010) (.00.3) (1.000)
2. As recorded, except that Bectshcha nomads have been counted at 55.(X/) (Ottoman estimate) rather than the inflated estimate on the 4:en,us.
PALESTINE UNDER THE MANDATE 31 curate information from villages would be difficult, mukhtars were paid a fixed sum for each notification (Vital events among Bedouin were seldom recorded.) The success of the registration procedure was enhanced by the relative ease with which the Palestinian population could be counted. Mandatory Palestine was small and lacking the impenetrable deserts or mountain ranges that made counting vital events so difficult in other Middle Eastern countries. British officials were thus able to keep relatively close control over the registration process. Despite this, the enumeration of births and deaths in Palestine was not complete.
After 1922, relying on its statistics of birth and death, the Palestine government made updated estimates of population size each year by adding and subtracting vital events from the recorded 1922 population. The method: 1. Take as a base the 1922 census results.
2. Subtract deaths and out-migrants.
3. Add births and in-migrants.
4. Repeat steps 2 and 3 each year for an updated estimate.
When the British compared the population numbers derived from the above system to the accurate numbers in the 1931 census they found the former to be too low. The system had not worked. The British assumed that this was the result of an undercount of births. Had births been properly recorded, the British statisticians believed, the natural increase of the population would have been seen to have been higher.
They thus adjusted their estimates of natural increase between 1922 and 1931.8 The British officials had indeed been mistaken in taking their vital registration system as accurate, but it was not the fault of poor birth registration. When analyzed, the birth statistics for the period between the censuses seem actually to have been fairly good.
Recorded birth rates were reasonable and female births were remarkably well recorded. (Over the period 1923-1931 registration of female births appears to have been only approximately 2% too low.)9 In fact, it was infant mortality that was poorly recorded. As seen in Table 2-10, infant deaths averaged one-third of the total reported deaths. This was much too low.
In the Palestine Muslim population immediately after World War I, approximately 50% of all deaths were
Table 2.9. Recorded and Corrected Population in 1931, by Religion Muslims Jews Christians Others Total Recorded 759,712 174,610 91,398 10,101 1,035,821 Corrected 775,181 175,936 92,802 10,270 1,054,189 Table 2.10. Muslims, Recorded Births and Proportion of Deaths Under Age 1, 1923 to 1931 Year BIRTHS PROPORTION OF TOTAL DEATHS UNDER AGE ONE Males Females Male Female 1923 13,389 11,768 .2977 .2888 1924 14,923 13,429 .3293 .3282 1925 14,771 13,390 .3051 .3008 1926 16,802 15,162 .3304 .3068 1927 15,791 14,825 .3337 .3193 1928 17,656 16,378 .3232 .2965 1929 17,076 15,974 .3408 .3256 1930 18,358 17,157 .3394 .3189 1931 19,161 17,708 .3499 .3359 SOURCES: Table 6.1 and calculated from Table A6.4.
32
those of infants under the age of one.10 Infant mortality was thus underestimated by 30 to 40%. Had
infant deaths been properly recorded, the number of
total recorded deaths would have increased by more
than 10%. Deaths in other age groups must also
have been undercounted, so the increase would have
been slightly greater.
If the British had recognized the deficiency of their death registration system, they would have been in a further quandary, as their system of updating population year-by-year would have yielded even greater deficiencies in total population had deaths been properly recorded. The year-by-year estimates were already too low and adding more deaths in each year would have made them even lower. Their problem, in fact, was their unquestioning acceptance of the 1922 census. If one begins by assuming a census to be accurate when that census was actually more than 10% too low, all projections from that census will naturally also be low. The British statisticians were not only in error, they were in error when they identified their errors. Moreover, both birth and death records were very deficient for the first years of recorded enumeration, 1922 and 1923. The method of adding births and immigration and subtracting deaths and emigration to arrive at the next year's population, which was fairly reliable for the period after 1931, once the data had been corrected, could not be used for the period 1922-1931.
There were also problems with the registration of Table 2.11. Recorded Muslim Birth and Death Rates (per thousand per year), 1935 to 1946 Year Birth Rate Death Rate 1935 53 24 1936 53 20 1937 50 25 1938 47 19 1939 46 17 1940 47 25 1941 49 21 1942 45 20 1943 52 19 194 54 17 1945 54 16 1946 54 15
population in the years after 1931, but these related less to method of registration than to civil conditions. Because of the disruptions that occurred between 1936 and 1948, Muslim birth and death records were often imperfectly kept, although this was not especially true of the other religious groups. In a number of years, the official birth and death rates, listed in table 2.11, were inaccurate. The actual birth rate probably remained at approximately 53-54 per thousand each year throughout the period.
(The great adjustment and correction in recorded births in 1943 was the result of wartime rationing instituted in that year. It became beneficial to record births and therefore rural Muslims made sure that births were recorded.) If the figures in table 2.11 were correct, deaths appear to have decreased when the General Strike and Arab Revolt against the British broke out (at the very time one would expect mortality to have gone up), increased when the revolt was put down, then decreased again when the Palestine statistical system started to suffer the effects of World War II and the Jewish Revolt which followed.
In fact, deaths were underregistered because of the natural inefficiency of any registration system in times of civil unrest, especially in rural areas. Because of underregistraton, both birth and death figures as supplied by the Government must be adjusted. The adjusted births, deaths, and natural increase for 1935 and 1946 are listed in table 2.12. The table assumes
Table 2.12. Adjusted Muslim Births, Deaths, and Natural Increase, 1935-1946 Year Births Deaths Natural Increase 1935 39,918 17,823 22,095 1936 41,543 18,749' 22,704 1937 42,848' 20,091 22,757 1938 43,849' 19,843' 24,006 1939 44,985' 20,337' 24,648 1940 46,317' 21,612 24,705 1941 47,408a 19,133 28,275 1942 48,732' 18,271 30,461 1943 49,621 18,978' 30,643 1944 52,710 19,650' 33,060 1945 55,052 20,371' 34,681 1946 57,194 21,098' 36,096 SOURCES: Tables 6.2 and 7.2. 'Adjusted figures.
33
average figures for birth and death rates in the years
of poor registration, preferring to err on the low side
of each."
As always, it should be remembered that these
figures are not absolutely correct. Some births and
deaths were missed even in the best years. The most
important thing in estimating change in total population is that the underestimations of births and deaths
are approximately the same. When both are slightly
too low they tend to balance each other out. Therefore, although the uncorrected estimates may tend to
underestimate both births and deaths, the result in
natural increase is little affected. However, if the
corrections in table 2.12 were not made, the years
1937 and 1940-1942, in which births alone were
underreported, and 1936 and 1943-1946, in which
deaths alone were underreported, would have shown
considerable inaccuracies.
MIGRATION INTO PALESTINE MUSLIM IMMIGRATION.
There was a small but significant unrecorded Muslim immigration into Palestine from 1931 until the end of the Mandate. In the only scholarly analysis of the Arab immigration, Professor Roberto Bachi has concluded that this migration averaged 900 Muslims a year, a total of 13,500 for the period 1931 to 1945.12 This figure does not include Arab workers who remained in Palestine for a brief time and then returned home. The results of Bachi's closely reasoned analysis have been applied here in estimates of the actual Arab population, and 900 Arabs per year have been added to the figures for the Arab population (table 2.15).
The validity of analyses such as Bachi's has been denied in other sources," which posit a much larger unrecorded Arab immigration. As a result of this immigration, according to these sources, the proportion of Arabs to Jews as published by the British erred to the detriment of the Jewish population. Like the claims concerning Arab immigration into Ottoman Palestine considered in chapter 1, this argumentation intends to demonstrate that the Arabs in Mandatory Palestine were largely newcomers, as were the Jews. The political ramifications of this assertion are not the province of this study, but the demographic confusion over illegal Arab immigration is worth analyzing.
From a statistical standpoint, even if a large unrecorded Arab immigration existed it could not have affected the population totals establishing Arab numerical supremacy, since the illegal Arab immigrants would not have been counted. The commonly accepted (and basically correct) proportions of Arabs to Jews in the years after 1931 were the yearly ratios of legal resident Arabs to legal resident Jews. The British came to their estimates of the Muslim, Christian, and Jewish populations of Palestine after 1931 by adding recorded immigration and births every year to the previous year's numbers and then by subtracting recorded emigration and deaths. No matter how many illegal immigrant Arabs came to Palestine, they would not have been included in the official Arab-Jewish ratio unless they were registered, which, by definition, they were not.
An Arab "illegal" would have entered the population lists only if he or she died in Palestine and the death were recorded. In these circumstances, the dead illegal, even if never entered in the base population, would be subtracted from the total. The result would be a population "illegitimately" decreased by one.
Correspondingly, the illegal immigrant's children, if their births had been recorded, would illegitimately increase the recorded population. Large numbers of recorded births to unrecorded parents would have significantly changed the rates upward. That the latter did not occur to any great extent is evidenced by the relative stability of birth rates among Arabs in Mandate Palestine and among those who remained in post-1948 Israel. Indeed, the birth rates of Palestinian Arabs were actually higher on average in Israel after 1948, where illegal Arab migration was most certainly not a factor, than in Mandatory Palestine, where it has been alleged." A similar argument is that large-scale Arab immigration from surrounding countries camouflaged undercounted deaths among the Arabs. This argument presupposes that Arab mortality could not have decreased so rapidly as to show the high rate of natural increase recorded by British statistics from the early days of the Mandate: in fact, Arab mortality had been declining for the last four decades of Ottoman rule and the decrease obviously continued under the British. Under this second argument, many deaths were not recorded, and therefore what appeared to be a low death rate was actually a combination of unregistered deaths and large in-migration. To test
34 PALESTINE UNDER THE MANDATE the argument, and show the effect on the Arab-Jewish ratio if indeed Muslim deaths had gone unrecorded, one has simply to inflate artificially the number of deaths and then calculate the population on the basis of higher mortality. Thus, if one assumes for the sake of argument the highly unlikely proposition that one-third of the Muslim deaths were unrecorded, the result would be a total of 117,000 additional Muslim deaths between 1931 and 1944.15 If the 117,000 were then subtracted from the 1944 estimates of Muslim population, the percentage of Muslims in the Palestine population at the end of 1944 would have dropped from 62% to 59%. Therefore, even if large numbers of Muslim deaths had gone unregistered, the ratio of Arabs to Jews in the official statistics would have changed little.
That there was unrecorded Arab migration into Palestine is beyond doubt, but it was small. Throughout the Mandate years, Arabs whose ancestors had lived in Palestine in Ottoman times formed a distinct majority of the population. For the unrecorded immigration to have had a significant effect on the ethnic composition of Palestine it would have had to have been immense. For example, the non-Jewish population at the end of the Mandate was double the Jewish population. If Jews were to be considered a "legal majority" (de jure, not de facto), one-half of the non-Jews would have had to have been illegals or illegals' children, not a likely situation. The argument that Arab immigration somehow made up a large part of the Palestinian Arab population is thus statistically untenable. The vast majority of the Palestinian Arabs resident in 1947 were the sons and daughters of Arabs 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.
JEWISH IMMIGRATION. The tables of "migratory increase" published by the Mandatory Government did not include unrecorded illegal immigration, which was a significant factor in the Jewish population in the latter years of the Mandate. In essence, the British counted as actual immigrants all those they recorded as having entered Palestine, subtracted those they recorded having left the country, and took the result as net immigration.16 This was a far better procedure than simply counting those who were legally considered immigrants, because it caught, at least statistically, those who came as tourists and stayed as illegal Jewish refugees and others who entered Palestine without registering with anyone. but the Jewish Agency. The Jewish Agency published statistics on their numbers17 and they have been included in the yearly population estimates (table 2.13) as Jewish immigrants.
The population records of the Jewish Agency were generally very good. Jewish immigrants did not necessarily owe any loyalty to the Mandatory Government and felt no patriotic pressure to register with government officials, but they may very well have felt loyalty to the Jewish Agency, which they were more likely to view as their "government" than the Mandatory. Moreover, there were positive reasons for making their presence known to the Jewish authorities, who had at their disposal settlement lands and educational and other services that aided the welfare of Jews. The Jewish Agency statistics need no correction. The British belatedly recognized this in the 1940s when they abandoned their own estimates of the Jewish population and substituted estimates based on the Jewish Agency statistics.
Table 2.13. A Comparison of Jewish Agency and Government Figures on Jewish Immigration Year Government JEWISH AGENCY Figures' FIGURES b Total Undocumented 1932 9,553 9,553 1933 30,327 30,327 1934 42,359 42,359 1935 61,854 61,854 1936 29,727 29,727 1937 10,536 10,536 1938 12,868 12,868 1939 16,405 27,561 11,156 1940 4,547 8,398 3,851 1941 3,647 5,886 2,239 1942 2,194 3,038 844 1943 8,507 8,507 1944 14,464 14,464 1945 12,751 12,809 58 1946 7,851 n.a. a. See Table A9.1.
b. See Table A10.14.
c. Maapthm refugees, included in Total figures.
PALESTINE UNDER THE MANDATE 35 THE POPULATION OF PALESTINE, 1922-1946 Table 2.14 offers a simple projection of each of the main religious groups from 1922 to 1931 (corrected figures), assuming that the populations increased at the same rate each year. Such a projection can be taken as no more than a guide, since it smooths out all variance. For the Muslim and Christian communities an assumption of relatively constant increase is valid, but it is not so for the Jews, whose population increased more by immigration than by natural increase. Jewish immigration was far greater in some years than others, a factor not shown in the table.
Nevertheless, table 2.14 does give an idea of the rate at which population increased.
Table 2.15 gives the estimated population of Palestine from 1931 to 1946, including both settled population and nomads. Through the simple process of not attempting an adjustment of the initial estimate of Bedouin population (55,000, the Ottoman estimate), the nomadic population in the Beersheba
Table 2.14. The Population of Palestine By Religion, 1922 to 1931, End of Year Estimates Muslim' Jewish Christian Other Total 1922 Census 638,407 93,360 75,875 8,481 816,123 1922 Dec. 31 640,798 94,752 76,194 8,515 820,259 1923 Dec. 31 663,296 102,134 77,905 8,696 852,031 1924 Dec. 31 676,544 113,059 79,653 8,882 878,138 1925 Dec. 31 690,055 137,484 81,441 9,071 918,052 1926 Dec. 31 703,838 149,066 83,270 9,265 945,438 1927 Dec. 31 717,896 153,828 85,139 9,462 966,325 1928 Dec. 31 732,234 158,122 87,050 9,664 987,070 1929 Dec. 31 746,858 164,492 89,004 9,870 1,010,224 1930 Dec. 31 761,775 170,783 91,002 10,081 1,033,641 1931 Census 775,181 175,936 92,802 10,270 1,054,189 a. Including 55,000 Beersheba Bedouin in each year.
Table 2.15. Population of Palestine by Religion, 1931 to 1946, End of Year Estimates Year Total Muslim' Jewish Christian Other 1931 1,057,214 777,403 176,468 93,029 10,314 1932 1.095,602 795,184 193,467 96,415 10,536 1933 1,163,616 815,787 236,297 100,686 10,846 1934 1,234,129 832,560 284,305 106,302 10,962 1935 1,332,587 855,769 356,487 109,131 11,200 1936 1,388,852 879,496 385,408 112,401 11,547 1937 1,427,441 903,699 397,166 114,764 11,812 1938 1,462,249 921,820 412,552 115,869 12,008 1939 1,540,727 949,612 457,943 120,853 12,319 1940 1,593,204 976,119 479,872 124,482 12,731 1941 1,639,757 1,004,989 492,458 129,260 13,050 1942 1,683,178 1,035,249 503,608 131,031 13,290 1943 1,739,695 1,068,623 522,112 135,128 13,832 1944 1,800,995 1,099,432 547,902 139,394 14,267 1945 1,868,597 1,136,851 573,587 143,132 15,027 1946 1,942,349 1,175,196 602,586 148,910 15,657 'Including 55,000 Beersheba Bedouin in each year.
36 PALESTINE UNDER THE MANDATE District has been kept constant in both tables 2.14 and 2.15. Since this population almost surely increased somewhat, this has the effect of causing a slight undercount of Muslim population. However, the evaluation of nomadic population in Palestine is in such a state of confusion that I have felt that to make an estimation is fruitless. H.V. Muhsam, who made an extensive evaluation of the various Mandate statistics on nomads, wrote: The only condusions which, under these circumstances, can be drawn with a reasonable degree of reliability are: (a) that the Bedouins of the Negev may in 1946 have numbered more than 55,000, but they certainly numbered less than 95,000; and (b) that it is likely that their number was between 65,000 and 90,000.... The wary observer might be inclined to agree with the only conclusion 'Aref el 'Aref thought himself entitled to draw from his enumeration, namely, that "Allah alone knows the truth." '8 Using slightly adjusted and corrected numbers from the 1931 census (table 2.9), the population for each religious group has been projected to the end of the census year in 1931, a minor adjustment." The population of each religious group by year has been estimated to be the previous year's population plus natural increase and net immigration 2° All the subsequent estimates are for December 31st of the year. As stated above, the estimates in table 2.15
Table 2.17. Population of Palestine by Ethnic Group, 1931-1946 Year Total Arab' 1931 1,057,214 864,806 82 1932 1,095,601 885,440 81 1933 1,163,616 909,687 78 1934 1,234,129 931,025 75 1935 1,332,587 956,662 72 1936 1,388,852 983,244 71 1937 1,427,441 1,009,491 71 1938 1,462,248 1,028,611 70 1939 1,540,727 1,060,593 69 1940 1,593,204 1,090,287 68 1941 1,639,756 1,123,168 68 1942 1,683,178 1,154,951 69 1943 1,739,694 1,191,976 69 1944 1,800,995 1,226,489 68 1945 1,868,597 1,267,456 68 1946 1,942,350 1,310,866 67
differ somewhat from the yearly population estimates made by the Mandatory Government, because of the inaccuracies in the government statistics.21 Muslim births and deaths have been adjusted to reflect undercounts in the official figures and 900 Muslims have been added each year as an estimate of illegal Muslim immigration. For the Jews, Jewish
Table 2.16. Proportion of Main Religious Groups, 1931 to 1946 Muslim Jewish Christian Other .7353 .1669 .0880 .0098 .7258 .1766 .0880 .0096 .7011 .2031 .0865 .0093 .6746 .2304 .0861 .0089 .6422 .2675 .0819 .0084 .6333 .2775 .0809 .0083 .6331 .2782 .0804 .0083 .6304 .2821 .0792 .0082 .6163 .2972 .0784 .0080 .6127 .3012 .0781 .0080 .6129 .3003 .0788 .0080 .6151 .2992 .0778 .0079 .6143 .3001 .0777 .0080 .6105 .3042 .0774 .0079 .6084 .3070 .0766 .0080 .6050 .3102 .0767 .0081 Jewish Other 174,139 16 18,269 2 191,125 17 19,036 2 233,912 20 20,017 2 281,863 23 21,241 2 353,959 27 21,966 2 382,857 28 22,751 2 394,611 28 23,339 2 409,984 28 23,653 2 455,329 30 24,805 2 477,247 30 25,670 2 489,830 30 26,758 2 500,976 30 27,251 2 519,478 30 28,240 2 545,253 30 29,253 2 570,935 31 30,206 2 599,922 31 31,562 2 Year 1931 1932 1933 1934 1935 1936 1937 1938 1939 1940 1941 1942 1943 1944 1945 1946
a. Including nomads.
PALESTINE UNDER THE MANDATE 37 Agency figures on illegal immigration have been added.
Christian and other populations have not been altered.
RELATIVE POSITION OF THE COMMUNITIES IN PALESTINE. The great Jewish immigration of the 1930s considerably added to the numbers of Jews in Palestine. It was only somewhat balanced by the much higher fertility of the Muslim and Christian Palestinian Arabs. The Jewish population increased from 17% of the total in 1931 to 31% in 1946. Extensive immigration in the year 1947, not shown in table 2.15, increased the Jewish population to approximately one-third of the population of Palestine. Nevertheless, the Muslim Arab community remained considerably larger than the Jewish community, and the Arabs were more than twice as numerous as the Jews.
ETHNICITY. Partly as a holdover from the Ottoman practice of identification by religion, partly from a policy of "divide and rule," the British attempted to keep a division of the Palestine population by religious group. This division was reflected in the 1922 census and statistics derived from it. In the 1931 census, however, the British added a question on ethnicity, allowing for the first time an ethnic identification of Palestine residents by self-identification, rather than language group. All but nomads (who were surely Arabs) were asked to identify themselves as Arabs, Jews, or "Other" (table A3.22).
79.8% described themselves as Arabs, 18.0% as Jews, and 2.2% as "other." 99.6% of the Muslims were Arabs, as were 80.1% of the Christians and 90.2% of the "other religions." When in later statistics the British statisticians estimated ethnicity, they applied the ethnic percentages of each religious group in 1931 to data known only by religion to estimate ethnic group. Although identification may have varied slightly between 1931 and 1946, it is instructive to apply the same procedure to the figures of yearly population in table 2.15. The results are in table 2.17. (The nomads have been figured into the table as Arabs.) SUMMATION Beginning in the 1870s the population of Palestine began to increase rapidly. This was a common phenomenon in the Ottoman Middle East. All over the Ottoman Empire, thanks to a long period without war and an improving economy, the population was growing.22 As did the other subjects of the Empire, the Arab and Jewish population of Palestine suffered
Table 2.18. The Population of Palestine by Religion, 1870 to 1946 Year Muslim % Jewish % Christian % Total 1860-61 325,000 88% 13,000 4% 31,000 8% 369,000 1882-83 408,318 87 15,300 3 44,471 10 468,089 1895-96 469,750 86 20,117 4 58,987 10 548,854 1905-06 531,236 85 27,862 4 69,092 11 628,190 1914-15 602,377 83 38,754 6 81,012 11 722,143 Year Muslim % Jewish % Christian % Other % Total 1914 657,377a 82% 61,000b 8% 81,012 10% 798,389 1918 618,528a 83 58,728 8 70,429 9 443 748,128 1922c 640,798 78 94,752 12 76,194 9 8,515 1 820,259 1931' 777,403 73 176,468 17 93,029 9 10,314 1 1,057,214 1936 879,496 63 385,408 28 112,401 8 11,547 1 1,388,852 1939 949,612 61 457,943 30 120,853 8 12,319 1 1,540,727 1946 1,175,196 60 602,586 31 148,910 8 15,657 1 1,942,349 a. Including Druz* and Shia. Both are in "Other" in 1922 and after. Also including 55,000 nomads of Beersheba.
b. Including noncitizen Jews.
c. All Mandate statistics are for December 31, not for the census date.
38 PALESTINE UNDER THE MANDATE losses during World War I, then rebounded to increase even more rapidly after the war. Unlike the other regions, Palestine also grew through the immigration of Jews, primarily the Jews of Europe. In the end it was this migration that set Palestine apart, ultimately replacing through war its indigenous population with an immigrant population. Table 2.18 traces the population of Palestine from the beginning of its increase until the end of Palestine as it had been.