Data Cart

Your data extract

0 variables
0 samples
View Cart

Sample Description:
Population Census of Egypt, 1868

Overview
The individual-level samples of the 1848 and 1868 Population Censuses of Egypt were digitized, created, and codified by Professor Mohamed Saleh, with the help of a data entry team, from the original handwritten Arabic census registers preserved at the National Archives of Egypt. The digitization project took place at the premises of the National Archives of Egypt between 2010 and 2012. The census registers, the census sample design, and the sample characteristics are described in detail in Saleh (2013). Professor Saleh is the copyright holder of the dataset. Hence, users who wish to download and use this dataset in their work are required to use the following citation:

Saleh, Mohamed (2013). A Pre-Colonial Population Brought to Light: Digitization of the Nineteenth Century Egyptian Censuses. Historical Methods: A Journal of Quantitative and Interdisciplinary History 46 (1): 5-18. DOI: 10.1080/01615440.2012.678807

The sample description in this document draws on Saleh (2013). This document and the public dissemination of the census sample have been done in collaboration between Professor Saleh and IPUMS International.

Census definition
The 1848 and 1868 Egyptian censuses are among the earliest precolonial individual-level population censuses from any non-Western country, and the earliest in the Middle East and North Africa to enumerate females, children, and slaves, and not only adult free men or household heads.

In 1845, the autonomous Ottoman viceroy of Egypt, Muhammad Ali Pasha (1805-1848) ordered the beginning of a nationwide population census (known in Arabic at the time as Ti'dad al-Nufus). The translated original 1848 census order can be found in Cuno and Reimer (1997, 213-6). The census operations started in rural provinces in early 1846 and were then extended to urban provinces in 1847. However, most of the census registers belong to 1848.

Although there were several enumerations that took place in the 1850s under the viceroys Abbas I (1848-1854) and Said (1854-1863), these were geographically scattered and did not constitute a nationwide census. The second nationwide population census of Egypt took place 20 years later between 1865 and 1869, during the reign of Ali's grandson, Khedive Ismail (1863-1879). Most of the second census registers belong to 1868. The 1868 census did not cover the whole country.

The breakdown of the enumeration by province for both censuses is discussed in detail in Saleh (2013). The population universe for both the 1848 and 1868 censuses includes all individuals (Egyptians and foreigners) who were present within the political boundaries of Egypt at the time, including the Nile delta and valley, Sinai, and the Western and Eastern Egyptian deserts. However, the geographic coverage varied across the two censuses.

The 1848 and 1868 population census were never published in census reports with aggregated tabulations, which is the standard practice in modern censuses, probably due to the lack of a statistical bureau. Instead, only the individual-level census returns were recorded in over 7,000 handwritten census registers that have been preserved at the National Archives of Egypt since then. Because they were never published, while the subsequent 1882 census was published, it has been long believed (wrongly) that the first population census of Egypt was conducted in 1882, following the British occupation of Egypt in May 1882, because the 1882 census was published. The 1848 and 1868 censuses were long forgotten before they were "discovered" by Egyptian historians in the 1980s.

Enumeration technique
The enumeration procedure in both the 1848 and 1868 censuses was direct and followed the de facto concept. Enumeration was done by census takers in cities and by village headmen in rural provinces. The 1848 census was conducted over three years (1846-1848), whereas the 1868 census was conducted over five years (1865-1869). In both censuses, certain geographic units were enumerated more than once, perhaps due to under-enumeration in the earlier attempts. For each geographic unit, only the most recent enumeration (census register) is employed in the sampling frame of the censuses.

No questionnaire form was used in either census. Instead, the census returns were recorded directly by scribes in handwritten registers of household and individual characteristics as well as housing conditions, as. In both the 1848 and 1868 censuses, the smallest geographic unit is the street address in Cairo and Alexandria, the quarter (shiyakha) in other urban provinces, and the village sub-section (hissa) in rural provinces. The dwelling unit definition employed in enumeration is any unit used as a residence, including marginal living quarters not intended as such (e.g., cemeteries). The household definition used is a person or a group of persons, Egyptians or foreigners, connected or unconnected with blood relation, sharing the same living quarter, food and spending the night together. This includes slaves and free people (including servants). The exception to this are military personnel residing in military barracks, who were enumerated by the ministry of defense, foreign non-Ottoman subjects, who were enumerated by their consulates, and Cairo notables in 1848 (the latter were enumerated in 1868), who were enumerated at the household level only. However, military personnel residing outside barracks are enumerated. Also, foreign non-Ottoman subjects are actually recorded in the census registers at the household head level, with a note that they are to be enumerated by their respective consulates. Arab Bedouin tribes whether settled or nomadic are enumerated in separate registers.

Census coverage
The 1848 census covered nearly 92% of the true population in 18 provinces and 52% in 1868 in 10 provinces, where the true population is estimated based on the 1882 census. Both censuses enumerated slaves and Arab Bedouin tribes (settled or nomadic).

Sample description
The 1848 sample is a 1% sample (8-10% in Cairo and Alexandria), totaling 80,028 observations; and the 1868 sample is a 1% sample (10-12% in Cairo and Alexandria) totaling 78,415 observations. Both samples are stratified by province. The target sample size in each province was decided based on the predetermined sampling rate of 8-10 percent in Cairo and Alexandria and 1 percent in all the other provinces. Because there is not a priori information about the population size of each province in 1848 and 1868, the province population is estimated based on the population share of the province out of Egypt's total population in the 1882 population census. Egypt's total population size in 1848 is 4,476,439, based on Alleaume and Fargues (1998)'s estimate. The population size in 1868 is estimated assuming a constant annual growth rate between 1848 and 1882.

Systematic sampling by page is applied to the entirety of the registers of each province. Specifically, an initial page within a range of pages (x) is randomly chosen, and the successive pages every x pages are selected until the end of the province's registers. The page range (x) is determined a priori based on the target sample size and the average number of individual records in the page for each province.

Person sampling weights are needed to adjust for the different sampling rates across provinces, and for the non-enumeration of certain provinces. Details about their calculation are described below.

Household variables

Year _ _ _ _

Census year is 1868, although there were data collection in the years 1865 to 1869. But most of the data was collected in 1868.

Register code _ _ _ _ _ _ _ _ _ _ _ _ _ _

It is the code of the source census register from which the individual record has been digitized. It is the code used in the bookkeeping of the physical census registers at the National Archives of Egypt. Census registers are organized by province, and within each province, by district (qism) in Cairo and Alexandria, quarter (shiyakha) in other urban provinces, and village (nahiya) in rural provinces.

Page number _ _ _ _

The page number within the original census register.

Province _ _

The governorate (muhafaza) in urban Egypt or the province (mudiriya) in rural Egypt. It consists of two digits. The province coding follows the administrative division of the 1882 census.

District _ _

The qism in urban provinces and the markaz in rural provinces. It consists of four digits, the first two digits denote the province and the second two denote the district. The district coding follows the administrative division of the 1882 census.

District modified for 1848 _ _

The same as district code but is modified to harmonize Alexandria, where Alexandria in 1848 had four districts and in 1868 had 8. The seven districts in 1868 are matched to the 1848 districts in this variable.

Quarter _ _

The shiyakha in urban provinces and the qarya in rural provinces. It consists of six digits, the first two digits denote the province, the second two denote the district, and the last two denote the quarter. The quarter coding follows the administrative division of the 1882 census.

Dwelling identification code _ _ _ _ _ _ _ _ _ _

The dwelling identifier. Consists of ten digits; the first six denote the province, the district and the quarter, respectively, and the last four denote the dwelling.

Household identification code _ _ _ _ _ _ _ _ _ _ _ _

Household unique identifier. Consists of twelve digits; the first six denote the province, the district and the quarter, respectively, the next four denote the dwelling, and the last two denote the household.

Street name or village section ________

The lowest available administrative division. String variable in Arabic. It is the street name in urban provinces and the village section (including Arab tribes) in rural provinces.

Building address number _ _ _

The building address number in the street, at the time of the census. Only available in Cairo and Alexandria.

Dwelling type _ _

Type of dwelling in which the respondent is living in.

Dwelling legal status

Legal status of the ownership of the dwelling. "Owned" could be by a private individual that is not a household member. The endowment or Waqf refers to a donation without transfer of ownership. All dwellings in rural provinces are recorded as "owned"; this variable exhibits more variation in urban provinces.

[] Owned
[] Endowed waqf
[] Private endowment
[] Charity endowment
[] Owned endowment
[] Mosque endowment
[] State ownership
[] Guardianship
[] Leased
[] Suspended

Name of owner ____

The name of the private owner of the dwelling or the holder of the endowment.

Number of households within dwelling _ _ _

The number of households living within a dwelling

Number of persons within dwelling _ _ _ _

Number of individuals living in the dwelling.

Person variables

Individual identifier _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _

Individual unique identifier. Consists of fourteen digits; the first six denote the province, the district and the quarter, respectively, four digits denote the dwelling and two denote the household, and the last two identify the person.

Full name ________

Respondent's full name in Arabic.

Age _ _ _

Respondent's age. Women in Alexandria and Cairo reported only "juvenile" or "adult". The original text included information in years and months for children, which was rounded up here to a whole year if they reported six or more months.

Age in days _ _

Children's age in days, for respondents who are 2 years old or younger.

Legal status

The legal status of the respondent.

[] Free
[] Slave
[] Emancipated slave/Mamluk

Nationality

The nationality of the respondent. "Under government control" in the original text means Egyptian.

[] Egyptian
[] Non-Egyptian

Occupation _ _ _ _ _ _

Respondent's occupation coded using the Historical International Standard Classification of Occupations (HISCO). The first digit represents major groups classification. Minor groups classification corresponds to the first two digits of the code. Occupations specific to the Egyptian context that do not have an equivalent in HISCO were assigned special codes under a given HISCO major or minor group.

Province of origin _ _

The respondent's province of origin, which could coincide with the location currently living in. This information is self-identified. Thus, for example, it could correspond to their place of birth, place where the respondent's parents were born, or where most of the respondent's family members currently reside.

District of origin _ _

The respondent's district of origin, which could coincide with the location currently living in. This information is self-identified. Thus, for example, it could correspond to their place of birth, place where the respondent's parents were born, or where most of the respondent's family members currently reside.

Quarter of origin _ _

The respondent's quarter of origin, which could coincide with the location currently living in. This information is self-identified. Thus, for example, it could correspond to their place of birth, place where the respondent's parents were born, or where most of the respondent's family members currently reside.

Religion _ _

Respondent's religion.

Relationship to the household head _ _

Relationship to the household head.

Gender

Respondent's gender.

[] Male
[] Female

Ethnicity _ _

Respondent's ethnicity.

Disability status (infirmities) _

Respondent's disability status.

[] Without disability
[] Disabled

Census registration status (enumeration)

Denotes whether the individual was enumerated for the census or not. The value 0 ("not enumerated") includes people who may not be living in the household at the time of enumeration, such as emigrants, incarcerated individuals, individuals residing in the military barracks, or individuals protected by their respective consulates.

[] Not enumerated
[] Enumerated

Person sampling weight

The person sampling weight is the inverse of the sampling probability: it is equal to the (estimated) population size of the province divided by the actual sample size. The sampling probabilities and person sample weights vary across provinces, because of two reasons. First, Cairo and Alexandria were sampled by design at a higher rate (8-12%) than the other provinces (1%). Second, certain provinces are assigned higher person weights in the post-sampling phase to account for the non-enumeration of neighboring or geographically similar provinces. Specifically, the weights include the following adjustments:
(1) Rosetta in both 1848 and 1868 receives a higher weight to account for the missing Damietta.
(2) Al-Qusayr in 1848 is over-weighted to represent the missing Suez.
(3) Al-Arish in 1868 is over-weighted to represent al-Qusayr and Suez.
(4) Al-Daqahliya and al-Sharqiya in 1868 are over-weighted to represent the other missing Lower Egypt provinces (al-Qalyubiya, al-Gharbiya, al-Minoufiya, al-Buhayra).
(5) Giza, Bani-Souayf, Fayuum, and Asyut in 1868 are over-weighted to represent the other missing Upper Egypt provinces (Minya, Girga, Qena, Aswan).