Comparability — Argentina [top]

The 1980 sample distinguishes piped systems independent of the source, and specifies the location of the piping. Because the 1991 and 2001 samples do not identify piped systems, the integrated data identify the water source independent of the location, with "public system" (1991) and "running water" (2001) classified as piped. All three samples preserve water source and location information in the unharmonized variables.

The 2001 sample specifies the source from which water is obtained for cooking. The 2010 sample distinguishes between piped water inside the dwelling (or property) and outside it, regardless of the source.

Comparability — Armenia [top]

The 2001 and 2011 data indicate access to piped water from a distribution network.

Comparability — Austria [top]

All samples identify piped water inside the dwelling.

Comparability — Belarus [top]

Piped water is defined as a water tap in the dwelling connected to the water supply network. A household does not have access to piped water if there is a water pump or hydrant outside the dwelling.

Comparability — Benin [top]

All samples refer to household source of water supply. Running water (at home or elsewhere) is classified as piped water. Fountain and public faucet are classified as public piped water. Other options are considered to be no piped water.

Comparability — Bolivia [top]

The data are mostly comparable, indicating access to piped water, independent of the source. The 2001 sample does not further distinguish the location outside the dwelling (on- or off-property) as the other samples do.

Comparability — Botswana [top]

The 1981 sample includes piped water indoors and piped water standing in the yard as "piped water." The 1991 sample identifies separately piped water indoors from public piped water. Further, the 2001 and 2011 samples also distinguish between piped water inside and outside the dwelling. Water from a neighbor's tap is considered to be equivalent to piped shared supply with other households in 2011.

Detail on non-piped sources is available for all Botswana samples in the unharmonized source variables.

Comparability — Brazil [top]

The samples for Brazil consistently distinguish between piped water inside the dwelling and outside it, regardless of the source.

Comparability — Cambodia [top]

The question in all censuses asked for the household's main source of drinking water. Only piped water from the public network is classified as piped; the unharmonized variables also include categories on water piped from wells.

The Cambodia 2013 and 2019 samples include information on the location of piped water.

Comparability — Canada [top]

The sample identifies piped water inside the dwelling.

Comparability — Chile [top]

All samples, except 1960 and 2017, distinguish between water piped into the dwelling or outside it.

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

The 1970-2002 samples ask explicitly about piped water, while the census question in 2017 refers to water source from the "public system"; this may create classification differences in the access to piped water in 2017 with respect to previous censuses.

The 1992 and 2002 samples differ in that they classify water obtained off-property as not having piped water, whereas the previous two samples explicitly include all types of piped water, even if obtained off-property.

Comparability — Colombia [top]

The samples consistently identify if the dwelling had piped water. Only the 1973 sample notes whether the household had exclusive access or whether it had access to a public tap.

Comparability — Costa Rica [top]

The 1963, 1973 and 1984 samples focus on whether the piped water is exclusively used by the household, while the 2000 sample distinguishes between water piped into the dwelling or outside it.

In 2011, piped water is defined as water piped into the dwelling, which has at least one tube or faucet for supplying water and does not require travelling outside the dwelling to obtain it.

Comparability — Cuba [top]

The 2002 and 2012 samples identify the presence of piped water within or outside of the dwelling.

Comparability — Côte d'Ivoire [top]

The 1988 sample only identifies running water, while the census question in 1998 also indicates its location.

Comparability — Dominican Republic [top]

The data are mostly comparable across samples. The 2010 sample distinguishes between private and shared piped water. The 2002 and 2010 source variables offer more detail on non-piped water sources.

Comparability — Ecuador [top]

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

The 1962 sample focuses on whether the piped water is exclusively used by the household, while the later samples distinguish between water piped into the dwelling or outside it.

Comparability — Egypt [top]

Samples consistently and explicitly indicate piped water from the public system and distinguish piped water inside the dwelling from water piped outside the dwelling. In the 1986 and 1996 census, the question was asked to all households but only to private households in the subsequent census.

Comparability — El Salvador [top]

The 1992 data comes from a question on the location of water installations, which are defined as "faucets" or "streams of water." There is a separate question that asks about the water source. Although for most cases of households with installations, the water source was from a community piped system, there are also cases where the original source was from a public or private well, river/stream/lake, spring, or other source (and vice versa). These cases are coded by the location of the installation, regardless of the source.

The 2007 sample offers more clarity and distinguishes between piping inside the dwelling, from a neighbor (shared), and outside the dwelling, on property.

Comparability — Ethiopia [top]

The data are mostly comparable across samples. In all samples, water supply is asked about potable water only. In 1994, the census questionnaire collected information on urban and rural households separately. IPUMS combines urban and rural data in the integrated variable. Information collected from rural households was less specific, focusing on whether the household had general access to piped water. See the unharmonized source variables for more detail.

Comparability — France [top]

The 1968-1975 data are comparable and collect information on water for domestic use only. The samples distinguish piped water as connected to a water main serving the building, or connected to a well, cistern, fountain, or private spring by piping. The 1968-1975 samples additionally collect information on the location of piped water.

The 1982 sample defines piped water as a connection of the building complex through a common distribution network or as a connection by pipes to a water supply; it does not collect information on piped water location.

Comparability — Germany [top]

The variable is available for the 1971 and 1981 censuses and represents piped water supply. The universe for the variable differs in the two samples.

Comparability — Ghana [top]

The data are drawn from source variables that identify the household's main source of drinking water. The 2010 sample includes an additional source variable that identifies the household's main source of water for domestic purposes.

Comparability — Greece [top]

The samples consistently identify piped water into the dwelling or building. The 2011 sample includes water supply outside the dwelling but connected to a public or private network if it is in the yard or on land.

The question on water supply was asked to all households in 2011, but only to regular dwellings in the previous censuses.

Comparability — Guatemala [top]

The 1964 census distinguishes whether the household had piped water inside or outside the dwelling; the 1981, 1994, and 2002 censuses further identify whether access was exclusive or shared with other households.

Public tap or faucet are considered as access to public piped water.

Comparability — Guinea [top]

Both samples identify piped water on the property. The 1983 sample contains a category for a public tap/fountain, whereas the 1996 sample only specifies a "faucet outside the compound," which is classified as "piped outside the building or lot."

Comparability — Honduras [top]

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

Comparability — Hungary [top]

Piped water includes both public and private networks. The samples are generally comparable with minor changes in the response categories. See the unharmonized source variables for more information.

Comparability — Indonesia [top]

The data are drawn from question(s) on drinking water. In some samples, there are also question(s) on the source/location of water used for bathing/washing -- these can be found in the unharmonized source variables. Some samples also contain a question on the distance between the drinking water source and the nearest cesspool -- these are also not represented here.

The 1971 and 1976 samples specify whether the location of the drinking water source is inside or outside the "yard", which are classified as inside or outside the dwelling, respectively. The 2010 sample also distinguishes these categories, but differs in wording as "in-house piped water system" or "piped water through retail." The 2005 and 2010 questions contain responses indicating bottled drinking water which are classified as non-piped, despite the fact that they may actually receive piped water but do not use it for drinking.

Comparability — Iran [top]

Piped water is explicitly from the public network and inside the dwelling.

Comparability — Iraq [top]

The Iraqi census reported the source of drinking water from various sources. Piped water indicates water from a public network into the dwelling or at a public tap.

Comparability — Ireland [top]

The 1971-1991 samples distinguish where the water tap was located. In these samples, piped water inside the building is coded as inside the dwelling but in reality may include households with piped water outside the dwelling and within the building. This categorization is independent of the source.

The 2002, 2006, 2011 and 2016 samples are not comparable with previous samples, because there are no data on households without piped water nor on the location of the supply. For these samples, all the sources reported are piped water.

Comparability — Italy [top]

The Italian census identifies whether the household got its drinking water from a public water network. Water from a well or other source indicates no piped water.

Comparability — Jamaica [top]

The universe differs across samples. All samples identify piped water inside the dwelling, from public or private sources, and distinguish piped water outside the dwelling from a public standpipe (public piped).

Comparability — Jordan [top]

The Jordan census indicates if the household's source of water was a public network.

Comparability — Kenya [top]

The census questions were similar across samples, asking whether the household's main source of water was piped.

Comparability — Kyrgyzstan [top]

The sample indicates whether the dwelling has running water from a network or well.

Comparability — Laos [top]

The 1995-2015 census questions refer to the household's main source of water for drinking (and cooking). The 1995-2015 samples also provide details on the distance of the piped water source from the dwelling.

Note that availability of piped water decreases over the years, which might be related to the census questions source of water for drinking and cooking and not necessarily the availability of piped water in the dwelling.

Comparability — Lesotho [top]

The community supply is classified as access to piped water shared with other households.

Comparability — Malawi [top]

In all samples, the source of drinking water was asked separately for the dry and wet seasons and for multiple dwellings. Only information about the first dwelling is included in this variable, and households receiving piped water in either season were coded to piped water in this variable. The majority of households reported a common water supply in both seasons, but users should be aware that a piped water response may not indicate year-round access to piped water in all cases. Further information can be obtained in the source variables for each sample.

The 2008 sample explicitly indicates piped water into yard/plot of the household as opposed to piped water outside the dwelling as in previous samples.

Comparability — Malaysia [top]

The meaning of piped water differs between samples. In 1991, piped water included water from any source delivered through pipes. The 2000 census specified that piped water meant treated water only.

Comparability — Mexico [top]

There are minor universe changes across samples. The 1960 sample is a sample of individuals, not households. Only heads of households have most household-level information.

The samples for Mexico consistently distinguish between piped water inside the dwelling and outside it. Most samples identify water from a public tap. The 2000-2010 samples also delineate water from another dwelling (shared). The 2015 and 2020 ask specifically about the use of pipes; separate questions refer to the water source.

Comparability — Mongolia [top]

The data are comparable across samples, indicating piped water in the household.

Comparability — Morocco [top]

The samples are comparable. It is indicated whether the household had a public water supply connection, and if so, whether the access was private or shared.

Comparability — Mozambique [top]

The data are comparable across samples. In all samples, water supply is asked about for potable water only. All samples contain a category for a public tap/fountain, though the 2007 sample defines fountains as consisting of "one or more faucets" and "usually made of cement."

Comparability — Myanmar [top]

The sample specifies access to drinking water; information about non-drinking water is available in unharmonized source variables.

Comparability — Nepal [top]

The census question in 2001 and 2011 indicates the household's main source of water for cooking and drinking.

Comparability — Nicaragua [top]

The 1971 question distinguished dwelling-specific piped water connections from public or private networks, which are both classified as piped inside the dwelling. The 2005 question specifies that the connection asked for is the public supply, while for 1995 there is no clarification on this matter.

Comparability — Pakistan [top]

Hand pumps and wells are not interpreted as having piped water, even when they were present in the dwelling.

Comparability — Palestine [top]

The universe differs across samples. In both samples, piped water is defined as from a public water network.

Comparability — Panama [top]

The 1960, 2000 and 2010 samples indicate if the source of water was from a piped source, regardless of the water supply location. The 1980 and 1990 samples distinguish between piped water inside and outside the dwelling.

Comparability — Papua New Guinea [top]

Water piped to central unit (or another central unit) is classified as shared with other households.

Comparability — Paraguay [top]

The 1972, 1982, and 1992 data are mostly comparable. In 1972 and 1982, piped water is classified as water from a public network (i.e. Corposana or Senasa). The 1992 sample distinguishes between private and public networks for piped water.

In 2002, if households received water from neighbor(s), vendor(s), or other source(s) but indicated the water supply as Corposana/Senasa or a private network, they were recoded as "access to public piped water".

Comparability — Philippines [top]

All samples identify exclusive or shared access to piped drinking water from a community water system.

Comparability — Poland [top]

The Poland 1978, 1988, and 2002 samples distinguish between piped water within the dwelling and within the building.

Comparability — Portugal [top]

The data are comparable across samples and include piped water from public or private sources.

Comparability — Puerto Rico [top]

The 1970-1990 samples consistently identify piped water in the dwelling. Between 2000 and 2010, the question was either omitted or reworded in a manner that made it difficult to determine whether dwellings had access to piped water. The 2015 and 2020 samples include a question about availability of "running water" in the dwelling, referring to water accessible inside the housing unit through pipes.

Comparability — Romania [top]

The 1977 and 1992 data are from the building-wide questions where the household resides. The data reflect if the building was connected to the public water network for the household(s) residing in it. Household information on the water supply location can be found in the unharmonized source variables.

The 2002 and 2011 samples are not comparable with the previous two, and reflect if the household obtained water from the public water network, regardless of the location, which could be outside the building. In 2002, "own system" installations, which may or may not be piped, are classified as "no piped water."

Comparability — Rwanda [top]

The data are mostly comparable across Rwanda samples, identifying running water (tap) access. In the 1991 sample, the question distinguishes between a tap outside the dwelling and a tap on the street (public piped). For the 2002 and the 2012 samples, the question identifies a tap inside the compound (outside the dwelling), but then combines the street tap and tap outside the compound responses into one category, which is coded as "public piped." Therefore, this frequency may be overestimated.

Comparability — Saint Lucia [top]

The data are comparable across samples, consistently identifying different means of access to piped water.

Comparability — Senegal [top]

The data are consistent across samples. Piped water is defined as water from a tap, but there is no requirement that it be from a public system. Water from wells and other sources, even if present within the house, are not considered piped water.

Comparability — Sierra Leone [top]

The sample distinguishes between piped water inside the dwelling, inside the compound, and further from a public tap. Households that indicated obtaining water from a neighbor's tap are coded into the "no piped water" category.

Comparability — Slovenia [top]

The data indicate piped water from a network or private facility.

Comparability — South Africa [top]

The samples are largely comparable and consistently distinguish between piped water in and outside the dwelling.

The 2007 and 2011 samples provide a category for "piped water from access point outside the yard." Given that the previous samples offered a category for public piped water, and 2007 and 2011 did not, it is presumed that these cases are public access and are classified as such for comparability.

Further details on the access to public piped water can be found in the unharmonized source variables.

Comparability — Spain [top]

The 1991 sample indicates whether water was piped into the dwelling or the building, whereas the 2001 and 2011 samples provide no such distinction. The original 2011 source variable distinguishes between water that was supplied by a public source and water that was supplied by a private or local source. See the unharmonized source variable for more information.

Comparability — Sudan [top]

The data come from a question on main source of drinking water. All responses that indicated a "common network" are coded as piped water.

Comparability — Suriname [top]

The sample indicates if households have access to piped water and distinguishes between access to piped water inside and outside the dwelling. The information corresponds to drinking water only.

Comparability — Tanzania [top]

All samples identify piped water for drinking. The 1988 and 2012 samples distinguish between piped water inside and outside of the dwelling, but the 2002 sample does not.

Comparability — Thailand [top]

All samples distinguish piped water within the dwelling from piped water outside the dwelling and are comparable. They do not indicate that the water must be from a public distribution system. In all years the variable indicates the main source of water for the dwelling; the 1980-2000 samples also asked specifically about the source of drinking water in a separate census question. These data are available in the unharmonized source variables for those samples.

Comparability — Togo [top]

The 1960 and 2010 census questions differentiate between piped water inside or outside the compound: the former is considered to be piped inside the dwelling, while the latter to be outside the dwelling.

Comparability — Trinidad and Tobago [top]

The samples are comparable across years, except minor universe changes in 2000 and 2011. Piped water is distinguished as being located inside or outside the dwelling.

Comparability — Turkey [top]

The census asks for whether the household has a piped water connection, public or private, inside or outside the dwelling.

Comparability — Uganda [top]

The 1991 and 2014 samples distinguish whether water was piped inside or outside of the dwelling.

Comparability — Uruguay [top]

The 1963-1996 and 2011 samples are mostly comparable with slight differences from sample to sample. They represent whether the water was piped, regardless of the original source. The 1975, 1985 and 2011 samples are identical and distinguish location; the category "piped outside the building or lot" was further segmented into greater or less than 100m away from the property in the unharmonized source variable. The 1996 sample only offers the distinction of inside or outside the dwelling, whereas the 1963 sample does not offer any distinction on location in some cases.

The 2006 data draw from a question on whether the water was obtained from the public network, and does not include any differentiation on piping -- although public network supply is assumed to be piped, there may be cases where the household obtained water from a non-public source that was also piped.

Comparability — Venezuela [top]

In all samples, piped water indicates connection to the public water system or access to water from the system. The 1981 sample distinguished whether water was piped inside or outside the dwelling. The 1990 and 2001 samples also provide a category for a public fountain/tap (public piped). The 2001 sample has a separate category for piped water from a well, which is interpreted as no piped water for consistency with other Venezuela samples.

Comparability — Vietnam [top]

The 1989 sample differentiates between piped water inside or outside of the dwelling. The 1989 and 2009 sample both distinguish water from a public tap, whereas the 1999 and 2019 samples do not.

Many households in the Vietnam 1989 data lack dwelling information, including all households in some provinces.

Comparability — Zambia [top]

The data are comparable across samples. Piped water is distinguished as located inside or outside of the housing unit. Detail on non-piped and communal water sources is available in the unharmonized source variables.

Comparability — Zimbabwe [top]

The census question refers to the household source of water for drinking and cooking.