Comparability — Argentina [top]

The 1991 sample identifies squatters.

Comparability — Armenia [top]

The 2001 and 2011 samples indicate only if the dwelling was owned or not owned by a member of the household -- not the specific terms of non-ownership.

Comparability — Austria [top]

The data are comparable across samples. See the enumeration text for detailed discussion of the tenure distinctions.

Comparability — Bangladesh [top]

This variable is generally comparable across samples. However, in 1991 the question was asked of all households, while it was only asked of permanent households in later years. Additionally, the 1991 sample includes an "unknown" category.

Comparability — Belarus [top]

In 2009, dwellings that are owned by citizens must be fully paid.

The 1999 sample distinguishes between dwellings rented from other persons and those rented from local councils. In 2009, the distinction is between dwellings that are rented from the state, rented from non-governmental private developers, and rented from other persons. Dwellings that are rented from other persons for all samples were simply coded as "renting."

Comparability — Benin [top]

The 2002 and 2013 samples provide further detail regarding types of dwelling tenure for free; "family property" is classified in these two samples as "not owned, provided by family or friend."

Comparability — Bolivia [top]

The categories are largely consistent across samples. The 1992, 2001, and 2012 samples identify dwellings ceded by family members. It isn't clear where this relatively large category would have been coded in 1976.

Loan-backed habitations involve an upfront loan to the owner in exchange for no subsequent payments for a period of time.

Comparability — Botswana [top]

The 1981 and 1991 census questions refer to acquisition of the housing unit or plot, whereas the 2001 and 2011 samples asked about the housing unit only.

In 1981 and 1991, households that have plots allocated by tribal authorities, landboards, or acquired through the Self Help Housing Agency Scheme (SHHA) are listed as other type of ownership.

The samples consistently distinguish between various forms of household rental agreements. Households that are occupying land where no allocation was made by any authority are listed as squatters.

Comparability — Brazil [top]

The distinction between renting and owning is consistent across samples, but the level of detail within those categories varies over time.

Comparability — Burkina Faso [top]

This variable is fully comparable across samples.

Comparability — Cambodia [top]

The data are fully comparable between samples.

Comparability — Cameroon [top]

All samples have categories for rented, owned, and housing provided by employer. However, there is some variation in other categories across samples. See source variables for more detail and explanation.

Comparability — Canada [top]

The data are fully comparable across samples.

Comparability — Chile [top]

Some Chilean samples distinguish fully-owned from partially-owned homes.

The 1960 sample is a sample of individuals, not households. Only heads of households have most household-level information.

Comparability — Colombia [top]

The category detail differs across years.

Comparability — Costa Rica [top]

All samples distinguish fully-owned from partially-owned homes.

The 2000 sample identifies squatters.

In 2011, the category "precarious condition" refers to dwellings that have been occupied without using legal mechanisms of possession. These dwellings could have been built by their inhabitants but, because the terrain upon which the dwelling was built does not belong to the inhabitants, the dwelling cannot be considered "owned."

Comparability — Côte d'Ivoire [top]

The 1998 sample distinguishes between dwellings that are owned from those where the household is still paying back a loan on the dwelling ("term owner") and that are classified as "owned, still paying". The 1988 sample only includes "renters", while 1998 differentiates between a simple lease from a lease-purchase.

Comparability — Dominican Republic [top]

The data are largely comparable across samples with minor category differences.

Comparability — Ecuador [top]

The Ecuadorian samples consistently identify multiple types of usufruct arrangements, with 1962 and 1974 including categories for sharecropping.

The 1962 sample is a sample of individuals, not households. Only heads of households have most household-level information.

The 2010 sample provides more detail than previous samples on payment arrangements for dwellings that are owned, distinguishing between those that are fully paid, being paid for, and those that were acquired as a gift or inheritance.

Comparability — Egypt [top]

The source variable for 2006 distinguishes between rental property under the old law versus the new law, but the data are combined in OWNERSHIP. The 2006 sample differentiates households that fully own their dwelling from those that are still paying to own. The 1986 and 1996 samples distinguish "ordinary" rentals from publicly owned.

Collective households do not have responses in 2006; there are no collectives in 1986 or 1996.

Comparability — El Salvador [top]

The unharmonized source variable for the 2007 sample distinguishes those still paying for ownership by the type of recipient (public institution, private, or NGO). The 2007 census also adds a category for free occupancy without any reciprocal work/services, which is presumably contained in the "other" category in 1992.

Comparability — Ethiopia [top]

The samples are largely comparable over time.

In all samples there are many categories for rented property. "Rented from Kebele" refers to the Kebele Local Administration and has been categorized as "Local authority." "Rented from public housing agency" has been categorized under Government. If a dwelling is rented from a private household it has been categorized as "Private." Any dwellings that are rented from another organization not otherwise listed is categorized as "Other rented". Finally, individuals may pay a rent difference if the rent of the housing unit is which they are currently residing is higher than the rent of the housing unit in which they used to reside. These cases have been categorized as "Other rented".

For more information see the unharmonized source variables.

Comparability — Fiji [top]

This variable is generally comparable across samples. However, the 1986 census separately identifies squatters. The 1996 has a general category as well as more detailed categories for usufruct arrangements, while the 2007 and 2014 samples have a "caretaker" category. Both were categorized as "free/usufruct (no cash rent)" in the harmonized variable.

Comparability — Finland [top]

In 2010, the variable distinguishes between ownership of the dwelling and only a share of the dwelling, where the latter is classified as "owned, other."

Comparability — France [top]

Available detail differs across French samples, with owned, rented, and usufruct situations constituting the minimum detail identified in any year. There are separate categories for furnished and unfurnished dwellings.

The 2011 sample differentiates between a renter of an unfurnished housing project and unfurnished non- housing project dwelling.

Comparability — Germany [top]

The data are largely comparable across censuses. The data for 1971 are constructed from separate responses for the first and second households within the dwelling. Higher order households, which are rare, are coded to unknown.

The data for 1981 are available only for dwellings containing a single household (comprising roughly 93% of all private households). There is an unharmonized variable providing information for second households within dwellings, but they cannot be distinguished in the data.

Comparability — Ghana [top]

The "other" category includes a temporary living arrangement called "perching." Further ownership and tenancy information is contained in a separate unharmonized source variable.

Comparability — Guatemala [top]

The 1981 and 1994 distinguish between dwellings that are "owned, already paid" and those "owned, still paying" categories; these two categories are not available in the 1964 and 2002 samples. The 1981, 1994, and 2002 samples identify dwellings that are occupied for "free/usufruct", but this response is seemingly combined with "other" in 1964.

Comparability — Guinea [top]

The 1996 sample identifies subsidized renters who pay partial rent with their employers paying the rest.

Comparability — Haiti [top]

All Haiti samples have categories for owning, renting, usufruct, and other. The 1982 sample has a separate category for tenant farmers, and the 2003 sample has a separate category for squatters.

The 1982 sample only has data for households in specific communes (Port-au-Prince, Pétion-Ville and Croix des Bouquets) which may affect its comparability with other years.

Comparability — Honduras [top]

The 1961 sample only has information for households consisting of a head.

The 1988, 2001, and 2013 samples distinguish between owners who had fully paid and those paying in installments.

Comparability — Hungary [top]

The 1970-1990 samples identify co-tenants who used only a part of the dwelling exclusively and shared other parts. They were combined with tenants in 2001.

The 1990 sample has more detail than the others; there was a response option for owned by relatives, which is considered as owned per previous samples. This response might be included with usufruct in other countries.

The 2011 census question asks about the type of ownership but has significantly less detail compared to other Hungary samples. For this sample, housing units owned by a local government or other institution/organization are simply listed as "not owned."

Comparability — Indonesia [top]

The category "provided by employer" includes when the dwelling is provided by a private or public employer regardless of whether there are payments. The unharmonized source variables preserve detail on whether the dwelling was contracted (rent paid for set time period) or rented (month-to-month). The censuses delineate whether the household was paying installments to own or already owned, except in 2005 and 2010 when the categories were combined. The 2005 census does add categories for rent-free parent/relative-owned dwellings (coded into "provided by family or friend") and for free rent (non-relative, non-employer owned). The 2010 census "other" category includes free rent, employer-provided, etc.

Comparability — Iraq [top]

See the enumeration text for more information on the categories.

Comparability — Ireland [top]

The pre-2002 samples provide categories of renting that are not identified in later years. The 1981 and 1991 samples distinguish between households still purchasing their dwelling and those that completed all payments. The unharmonized source variables for the pre-2002 samples identify households in the process of purchasing their homes from local authority.

There are only minor universe differences between samples.

Comparability — Israel [top]

The 1972 census asked the question only of urban households.

The 1983, 1995, and 2008 samples included information on renters entitled or not entitled to key money. This distinction was not incorporated into OWNERSHIP. The 1983, 1995, and 2008 samples also distinguish between dwellings rented from private, parastatal, or other.

The 2008 sample identifies a separate category for dwellings owned by relatives where no rent is paid on it; these cases were likely included in "owned" in 1972-1995.

Comparability — Italy [top]

In 2001 and 2011, the "owned" category is defined as owning fully, partially, or when the household has the right to use the house, even if part of the dwelling is rented out. For both samples, "other title" is classified as "free/usufruct" given it refers to tenure free of charge or in exchange for services.

Comparability — Jamaica [top]

The samples are comparable across categories, aside from the category for "Government rented" dwellings in 1982 and 1991.

The 1982 sample includes responses for all households, while the 1991 and 2001 samples include only private households.

Comparability — Kenya [top]

The Kenya samples are comparable with each other, but they classify owned housing units differently than other countries, specifying whether they were purchased, constructed, or inherited.

The Kenyan censuses also provide more detail about renting than the other censuses do. Users should note that "rented" would also refer to housing units that are "provided/donated" (that we interpret as "free of rent"); this is more explicitly indicated in the enumeration instructions for 1999, 2009, and 2019.

Comparability — Liberia [top]

This variable is only available for the 2008 sample.

Comparability — Malawi [top]

The data are generally comparable across all samples.

The 1987, 1998, and 2018 censuses collected separate information on dwelling ownership for multiple dwellings inhabited by a household. Only information for the first dwelling is reported in OWNERSHIP, but the unharmonized source variables retain the details for higher-order dwellings.

The 2018 sample includes a response option for "institutional" ownership, which the questionnaire refers to as "employer", and is classified as "provided by employer".

Comparability — Malaysia [top]

The samples differ considerably in terms of number of categories.

Comparability — Mali [top]

The 1998 sample identifies more categories of tenure, including free lodging and co-ownership (condominium) status. The unharmonized source variables for 1998 and 2009 also distinguish between owners with and without official documentation and renters versus (long-term) lessees. Those distinctions are not maintained in OWNERSHIP.

Comparability — Mexico [top]

The 1960 sample is a sample of individuals, not households. Only heads of households have most household-level information. The 1960 and 1970 censuses made only the basic distinction between owned and not owned housing.

The 1995 and 2000 censuses distinguished between owners who had fully paid and those with mortgages.

The 1990, 2010, 2015 and 2020 censuses identified renters versus other non-owners.
In 2015 and 2020, any household that responded to the question indicating dwelling is "owned by a relative or is borrowed" is assumed to be living in the dwelling for free.

Comparability — Mongolia [top]

In 2010 and 2020, "other owned" refers to households that are owner occupied, but where the owner is the government or a private company; furthermore, the category of "occupied free of rent" but owned by "individual" are interpreted as "provided by family or friend".

Comparability — Morocco [top]

The responses "in process of being owner" and "mortgage" were combined into the category "owned, still paying" in 1982; the mortgage category was eliminated in subsequent censuses. The categories of "owner, already paid" and "in the process of becoming owner" are combined into the "owner" category in the 2014 sample.

The category "provided by employer" is documented as provided "with a small payment" in 2004, but no other censuses provide more detail, and free rent, employer-provided dwellings are explicitly excluded from the "free" category.

Comparability — Myanmar [top]

In 2014, the categories of "government" and "private company" are interpreted as "free/usufruct" since they are separate from "renting".

Comparability — Nepal [top]

Dwelling ownership is based on legal ownership except in rural areas where it is based on land ownership. Responses indicating that the dwelling is occupied by virtue of an agreement with an institution are coded as "free, public". The 2001 sample distinguishes a (rent) free option, which is combined with "other" in 2011.

Comparability — Nicaragua [top]

The 2005 census adds a free rent and an installments/mortgage category that were presumably contained within "other" in previous censuses; there was also a response option for "borrowed," but as it is unclear what this term means, it is classified in "other" not owned in this variable.

Comparability — Palestine [top]

The data are comparable between samples, with a minor universe difference.

Comparability — Panama [top]

The 1960 sample has less category detail than the later samples, which are consistent with one another.

Comparability — Papua New Guinea [top]

The 2000 census includes "free/usufruct", which is not identified in 1980 or 1990.

Comparability — Paraguay [top]

The data are comparable between samples. There are several more response categories in the 1992 and 2002 samples. In the 2002 sample, the response "Borrowed or caretaking" is categorized under other.

Comparability — Philippines [top]

The Philippines samples include categories for public and private subsidized housing.

Comparability — Poland [top]

The census question for Poland refers to the tenure status or legal basis for occupancy of the dwelling. The unharmonized source variables for these samples include related data on the building ownership.

In 1978 and 1988, "allocation of the dwelling" is considered as "not owned, free/usufruct (no cash rent)", given the enumeration instructions do not clarify whether rent is paid in these cases.

For the 1978, 1988, and 2002 samples, responses indicating that the household has a relationship with the owner or main tenant (parents, children, or other related persons) have been classified as "not owned, free/usufruct, provided by family or friend," even though some of these cases may pay rent for the occupied dwelling.

In 1988, workers' hotel is coded as "not owned, free/usufruct, provided by employer."

"Other" forms of tenure for all the Poland samples may include households residing in religious institutions, residing in return for agricultural work or other activity, or squatters.

Comparability — Portugal [top]

The variable is consistent across samples. The 2011 census lists several more categories for non-ownership than the previous samples. In 2011, the original response category "State, autonomous public institutes or other non-profit institutes" is categorized as Government ownership.

Comparability — Puerto Rico [top]

The variable is comparable across samples.

Comparability — Rwanda [top]

The 2002 and 2012 samples have a response for mortgage/leasing, but the number of cases is very small; nevertheless, we have coded other owners as having fully paid for their dwelling.

The 2002 unharmonized variable also identifies persons in refugee camps, which are included in "other" not owned in this variable.

Comparability — Saint Lucia [top]

The general "renting" category in 1980 identifies persons with leases as opposed to short-term renters from government and private persons.

Comparability — Senegal [top]

The 1988 sample identifies co-tenants (co-locataire), while the 2002 and 2013 samples identifies co-owners (co-proprietaire). The 2013 sample identifies "rent to own", while the 1988 sample includes the option of subletting.

Comparability — Sierra Leone [top]

"Purchased" is coded as "already paid" but might include cases where residents are still paying. Rental from a quasi-government entity is included with rental from government, but the two categories can be separated using the unharmonized source variable. Likewise, the unharmonized variable identifies different types of employer-provided housing.

Comparability — South Africa [top]

The 1996 sample only has categories for rented and owned. The 2001, 2007, 2011, and 2016 samples distinguish fully paid off from not yet paid off homes. The addition of the large usufruct category starting in 2001 substantially alters the distribution of values between 1996 and other South Africa samples. In 2016, there is an additional distinction between "rented from private individual" and "rented from other".

Comparability — Spain [top]

All samples distinguish fully-owned from partially-owned homes, and all samples are comparable over time.

Comparability — Switzerland [top]

The variable is consistent across samples. In 1970-2000, the unharmonized source variables in each sample distinguish between owners of buildings and owners of a floor or dwelling, but that distinction is not preserved in OWNERSHIP. For the same samples, the unharmonized variables also distinguish tenants from lessees.

Comparability — Tanzania [top]

The 1988 sample indicates only if the dwelling was owned by a member of the household, rented, or in other situation. The 2012 census question provides further details. In 2012, if the dwelling was owned by the employer, but the household was paying rent then the dwelling was categorized as being rented from a private company.

Comparability — Thailand [top]

The variable is consistent across samples.

Comparability — Togo [top]

In 2010, owners with or without a title are considered to own the dwelling.

Comparability — Trinidad and Tobago [top]

The data is comparable across samples. The 2011 responses distinguish between homes owned fully and those still subject to a mortgage.

Comparability — Turkey [top]

The 1985 and 1990 censuses only distinguish whether the household owned the dwelling or not. The 2000 census provides more detail; the "provided by employer" category does not detail whether payments were made to the employer or not.

Comparability — United Kingdom [top]

The 1961, 1971, and 1991 samples provided more detailed information than 2001 on the types of renting. The 1961 sample identified dwellings received by virtue of employment, whether rent is paid or not; these cases are classified as "free/usufruct", even though information about rent is not available. The 2001 sample identifies households paying part rent and part mortgage (shared ownership).

Comparability — United States [top]

The 1960-2020 U.S. samples distinguish between renters who paid cash rent and those who did not, where the latter response is classified as "free/usufruct". The 1990-2020 responses also allow users to distinguish between homes "owned free and clear" and those still subject to a mortgage. The 1990, 2000, 2010, 2015, and 2020 censuses explicitly placed owned units resting on leased land (for example, mobile trailers in some parks) in the "owned" category.

Comparability — Uruguay [top]

Because of lack of any definition, the category "inherited" was coded into "other" in 1985. Response categories vary throughout the years -- see the unharmonized source variables for more detail and explanation.

Comparability — Venezuela [top]

The 1971 sample categorizes subletting and usufruct situations not identified in later samples.

Comparability — Vietnam [top]

The categorization is relatively consistent between samples. The 1989-2019 samples include responses for "collective" or "cooperative" ownership, where a dwelling is built by an organization (including corporations or religious groups) and is assigned to the organization members. The 1999 and 2009 samples identify "joint government and individual" ownership, where a dwelling was built with funds from both sources and it is assigned to private individuals. Both the "collective/cooperative" and "joint government and individual" are classified as "renting", even though documentation does not explicitly indicate whether a rent is being paid.

Households in "simple houses" were not asked the question on ownership in 1999; these are dwellings constructed of makeshift or plant materials. The 2009 sample asked this question to households that have a dwelling, defined as a place intended for human habitation and that has walls, roof, and floor. In 2019, this question is asked of households living in a house/dwelling, as opposed to those in a temporary structure or in places lacking walls, roof, or floor.

Comparability — Zambia [top]

The variables are generally comparable.