Comparability — Argentina [top]

Samples prior to 2001 report de facto status, including consensual unions. The 2001 marital status question records only the person's formal legal status. Information on consensual unions in 2001 is recorded separately in CONSENS.

Separated and divorced persons are combined in the 1970 and 1980 samples, but are identified separately in 1991 and 2001. The 1991 census included separations from a union or marriage, while the 2001 sample explicitly counts only legal separations.

The 2010 sample does not include this variable.

Comparability — Armenia [top]

In all samples, the "Divorced" category includes both legally registered and unregistered divorces. These categories are reported separately in the unharmonized source variable.

Persons younger than age 15 are coded as "Single".

Comparability — Austria [top]

All samples report legal marital status only. Unmarried persons in MARST could be in a consensual union, recorded in CONSENS.

Separated persons are considered married. A separate variable identified if married persons cohabitated with their spouse, and IPUMS constructs "Married, spouse absent" to identify these de facto separated persons.

Comparability — Bangladesh [top]

The category "married" may include persons who are legally or religiously married.

Comparability — Belarus [top]

"Separation" included consensual unions.

All persons younger than age 15 are coded as "never married/single".

Comparability — Benin [top]

Persons younger than age 10 are coded as "Never married/single" in all samples.

All samples identify polygamous and monogamous marriages.

In 1979, persons in consensual unions were recorded as married, while those separated were included with divorced persons. The 1992, 2002, and 2013 samples have "consensual union" and "separated" as additional responses to this census question.

Comparability — Bolivia [top]

In both 1976 and 1992, the samples include joint categories for persons in a consensual union/marriage, and for separated/divorced persons. Those four categories are reported separately in 2001 and 2012. In 1992 and 2012, the categories separated and widowed include persons separated from or widowed in a consensual union. The enumeration instructions do not specify how to categorize persons separated from or widowed in a consensual union in 1976 and 2001.

Persons younger than age 12 in 1976 and younger than age 15 in 2001 and 2012 are coded as "Single". There is no age restriction in 1992.

Comparability — Botswana [top]

In 2011, enumeration instructions indicated that persons who are no longer living together are considered to revert to their original marital status and should not be registered as separated. Previous samples for Botswana do not include this distinction.

Persons younger than age 12 are coded as "Never married/single" in the 1991, 2001, and 2011 samples.

Comparability — Brazil [top]

All censuses separately identify formal marriages and consensual unions and include persons widowed in consensual unions in the category "Widowed". From 1960 to 1991, the samples distinguish between legal and de facto separations. A similar category is inferred from questions on legal marital status and previous partnerships in 2000 and 2010. The 1991, 2000 and 2010 samples distinguish between separation from a consensual union and marriages. In earlier years, enumeration instructions do not specify whether informal separations include persons separated from consensual unions.

The 1991, 2000 and 2010 data are derived from multiple census questions on current and previous unions. There are some inconsistencies among the marriage variables in 2000 and 2010, particularly in the reporting of religious marriages. Users should refer to the source variables if they wish to fully understand the interrelationships. Persons younger than age 10 are coded as "Never married/single" in all samples.

Comparability — Burkina Faso [top]

All samples identify monogamous and polygamous marriages but only the 1976 sample distinguishes this for men and women. Marriages may include both civil marriages and those conducted according to local custom.

In 1985 and 1996 the question was asked of residents 12 years old and over whereas in 2006 the question was asked of all persons 12 years old and over. Persons younger than 12 years of age are coded as "Single/never married."

Comparability — Cambodia [top]

In the 1998-2013 samples, marriages include de facto unions, as well as couples who are contractually married but not yet living as man and wife. The category of "separated" includes both legal and de facto separations. The 2019 sample does not have instructions for enumerators.

Marital status is not reported for absent persons in the 2013 sample.

Comparability — Cameroon [top]

All samples identify monogamous and polygamous marriages for men and women. In the 1976 and 1987 samples divorces and separations are combined in a single category. Marriages may include both civil marriages and those conducted according to local custom. The 2005 sample distinguishes marriage from consensual unions as well as divorces from separations.

In 1976 and 1987 the question was asked of persons 4 and 11 years old and over, respectively. In 2005, the question was asked of women ages 12 years old and over and men ages 15 and older. All persons younger than the age universe for marital status are coded as "Single/never married."

Comparability — Canada [top]

The samples are comparable over time. They include both common-law and legal marriages. The 1991-2001 samples have separate questions on legal marital status and common law marriage status. Those data are available in the unharmonized variables. The 2011 sample refers to the de facto marital status, thus including common law marriages.

Comparability — Chile [top]

In the 1960 sample, there was no age restriction on the marital status question. In other years all persons younger than the following ages are coded as "never married/single": age 12 in 1970, age 15 in 1982 and 2002; age 14 in 1992.

All the above samples, excluding 1970, identify marriages that have been legally annulled. In the 1970 sample, this category was combined with the "separated" one.

The 2017 sample does not have a marital status variable. The marital status of some persons can be derived from RELATE (relationship to the head of household).

Comparability — China [top]

In all samples, a union need not have been legal for a person to respond as "Married".

Persons younger than age 15 are coded as "Never married/single" in all years.

Comparability — Colombia [top]

All samples explicitly identify consensual unions.

In the 1964 sample, divorce was not recognized in the census. In subsequent years, the "Separated" and "Divorced" categories are combined and cannot be disaggregated.

Persons younger than age 10 in the 1985-2005 samples are coded as "Never married/single."

Comparability — Costa Rica [top]

In 1973, there was no age restriction on the marital status question. In other years, persons younger than the following ages are coded as "Never married/single": age 12 in 1963, 2000 and 2011, age 10 in 1984.

Comparability — Cuba [top]

The 2002 and 2012 samples identify both formal marriages and consensual unions. Separated persons are distinguishable from both married and divorced persons, and include persons separated from consensual unions. The category "widowed" includes persons widowed from consensual unions.

Persons younger than age 15 in 2002 and younger than age 12 in 2012 are coded as "single/never married".

Comparability — Côte d'Ivoire [top]

Persons younger than age 12 are coded as "Never married/single" in the 1998 sample.

In 1988, the monogamous versus polygamous distinction for married persons is only available for men.

Comparability — Dominican Republic [top]

Consensual unions are included in marital status in all years except 1970. In 1970, information on cohabitation is collected in a separate question, and MARST records only formal marital status.

The definitions of separation change over time. The 1960 and 1970 censuses identify only legal separation, while the 1981 census includes a general separation category. The 2002 and 2010 samples distinguish between separation from marriage and separation from consensual unions. The 1970 sample identifies annulments.

In 1970, the marital status question is asked of all persons. In 1960 and 1981, marital status is asked only of persons age 12 and above. The universe is further restricted to age 15 and above beginning in 2002. Persons under these ages are coded as never-married.

Comparability — Ecuador [top]

Persons younger than age 12 are coded as "Never married/single" in all years.

The 1962 sample allowed no response for "Separated". These persons were most likely included with "Married" or "Single" individuals.

Comparability — Egypt [top]

The Egypt samples identify persons who were contractually married but not living together ("Engaged"). Separated persons are not identified. Thus it appears the data identify legal marital status only.

The polygamous category is created from a separate question on the number of wives. Women in polygamous unions are not identified as such.

Females younger than age 16 and males younger than age 18 were considered "single".

Comparability — El Salvador [top]

Persons younger than age 12 in 1992 and younger than age 10 in 2007 are coded as "Never married".

In both samples, the marital status question asks about formal marriage and consensual unions. The samples distinguish between persons who are formally married and those in consensual unions. Persons who were separated or widowed in consensual unions are included in the categories for "separated" and "widowed". The "single/never married" category only includes persons who have never married and never lived in a consensual union.

Comparability — Ethiopia [top]

The 1984 sample was only asked to persons who reside in the household. The 1994 sample was asked to both present and absent household members. The 2007 sample was asked only to those responding to the long form.

Persons who are separated or divorced are only combined in the 1994 sample. Persons who are married or living together are combined in the 1984 and 1994 samples.

The question was asked only to persons age 10 and older in all samples.

Comparability — Fiji [top]

In 1966, 1976 and 1986, "married" may include persons who are legally or religiously married. In 2007 and 2014, the census question identifies separately legal marriages from consensual unions.

Separated is not identified in 1966; while divorced and separated are combined in 1976 and 1986. The 1996, 2007, and 2014 samples distinguish between divorces and separations.

In all samples except 1976, the question was asked of all persons. In 1976, persons born in 1962 or before are coded as "Never-married/single".

Comparability — Finland [top]

The 2010 sample reports legal marital status. The enumeration documents indicate that divorce regulations in Finland do not recognize the concept of legal separation; thus, people who are legally separated on the basis of old divorce provisions (prior to 1 January 1988) and are living apart were considered to be "married". The categories of "married", "divorced", and "widowed" include persons from registered partnerships.

Comparability — France [top]

Starting in 1990, the census stipulated that the marital status question applied only to legal marital status. We have coded the 1990-2006 marriage responses identically to other years on the assumption that those earlier responses also overwhelmingly refer to legal status, although that was not made explicit in the questionnaires.

Consensual unions prior to 2006 can be inferred using the RELATE variable, but only pertain to unions that include the head of household.

Comparability — Germany [top]

None of the samples allow a response for "separated". These persons were most likely included with "married" individuals.

Comparability — Ghana [top]

Marital status codes are consistent over time. The category "married" includes civil, traditional and common law marriages. Persons in consensual unions are identified separately. "Separated" indicates a fractured union, not simply living apart.

In all years, the marital status question was asked only of persons age 12 and above. Persons younger than age 12 are assumed to be never-married.

Comparability — Greece [top]

All Greece samples record the person's legal status.

Consensual unions are not recognized in the 1971-2001 samples. Registered consensual unions are recorded in 2011.

Separations are only recognized in the 2001 and 2011 samples. In earlier samples, persons who did not formally divorce are considered "Married".

In 2011, divorced and widowed include those from registered partnerships.

Comparability — Guatemala [top]

"Separated" is not identified in 1964 or 1973; separated people are included with those who are married. In the 1981, 1994, and 2002 samples, divorced and separated are combined in a single category. Separations include those from consensual unions in 1981, while the 1994 and 2002 samples do not refer explicitly to this possibility.

The category of "single/never married" includes persons below age 12 in 1994 and 2002, persons below age 13 in 1973, and persons below age 14 in 1964, for whom the information about marital status was not collected by the corresponding census question.

Comparability — Guinea [top]

In 1983-1996 samples, the "polygamous" category is only available for men. The 2014 sample identifies polygamous unions for men and women.

The 1983 sample does not identify consensual unions, nor do the instructions indicate how they were to be recorded (i.e., if they should be included in the "married" category). The 1996 and 2014 samples identify consensual unions.

Non-resident visitors were not asked the marital status question in the 1996 sample. All resident persons younger than age 12 are coded as "single" in 1996 and 2014. There is no age restriction in the 1983 sample.

Comparability — Haiti [top]

All samples identify marriages and consensual unions. The 1982 and 2003 samples distinguish divorces, separations from marriage and separations from consensual unions.

In 2003, persons younger than age 10 are coded as "never-married/single".

Comparability — Honduras [top]

Persons younger than age 10 in 1988 and younger than age 12 in 2001 are coded as "never married/single".

The category "separated" is identified only in the 1988 and 2001 samples.

Comparability — Hungary [top]

All samples report the legal marital status only.

The 2001 and 2011 samples has a de facto "spouse, absent" category, which was included with the "married" category in previous years. In 2011, registered cohabiting partners are grouped with "married"; while divorced and widowed persons include also the correspondent categories from those who were in registered cohabiting partnerships.

All samples report the legal status only in MARST. In 2001 and 2011, separate information on consensual unions is recorded in the CONSENS variable.

Comparability — Indonesia [top]

In all samples, the category "married" includes common-law marriages, as well as couples who are living apart temporarily but consider themselves married. The 1976 sample distinguishes between divorced and separated persons. In all other samples, persons who are legally divorced are combined with those who are not legally divorced, but consider themselves divorced.

In 2010, all persons younger than age 10 are coded as "never married." In previous samples, marital status was asked of all persons.

Comparability — Iran [top]

The category married in the 2006 and 2011 samples include persons who are contractually married but do not live together, as well as people who separated from their spouses but are not divorced. Couples who are engaged are not included in the "married" category. Polygamous marriages are not identified.

Comparability — Iraq [top]

The meaning of divorce differs for Christians and non-Christians. "Divorce" indicates the legal severing of the marriage bond for non-Christians. Persons in de facto separations are considered "married". Christians are considered divorced if they have separated. There is no legal status requirement.

The "polygamous" category is created from a separate question on number of wives. Women in polygamous unions are not identified as such.

Persons younger than age 12 are coded as "single".

Comparability — Ireland [top]

The 2016 sample includes same-sex marriages in the "married" category.

The treatment of separated and divorced persons changes over time. The 1971 and 1979 samples must have included divorced and separated persons in either "married" or "single" categories. The 1981 census includes a category for "other marital status", which refers to persons who obtained a divorce in another country. Separated persons are not identified. The category "separated/divorced" is introduced in the 1986 sample. It is available in all subsequent samples. Divorced persons are never separately identified, even after divorce became legal in 1997.

Persons younger than age 14 in 1971 and younger than age 15 in later samples are assumed to be in the "never married" category.

Comparability — Israel [top]

The 1972-1995 samples indicate the person's de facto marital status, as reported by the respondents. In 2008, separations are introduced as a response category.

Comparability — Italy [top]

The 2001 and 2011 samples contained no instructions on the recording of consensual unions.

Comparability — Jamaica [top]

In all samples, this variable records the formal marital status. Because cohabitation is very common, the percent of persons married is relatively low in all samples. A large number of persons have unknown marital status in the 1982 sample.

The 1982 sample collected information on current cohabitation only for reproductive-age women (ages 14 to 45). This information is available in the unharmonized source variable JM1982A_0440. Most consensual unions involve the household head, and can be inferred in the 1980 sample based on the presence of a "Spouse/partner" in the harmonized RELATE variable.

Persons younger than age 14 in 1982 and younger than age 16 in 1991 and 2001 are assumed to fall in the "Never married" category.

Comparability — Jordan [top]

The data report formal marriages, not consensual unions.

All persons younger than age 15 are coded as "single".

Comparability — Kenya [top]

The instructions state that the enumerator should accept what people say about their marital status, and that the census is not trying to determine who is legally married. Even the category for "divorce" does not necessarily imply a particular legal status.

The 1989, 1999 and 2009 samples identify polygamous and monogamous marriages. These samples also distinguish divorces and separations.

Persons younger than age 12 are coded as "never married/single" in the 1989 sample.

Comparability — Kyrgyzstan [top]

Persons younger than age 15 are reported as "married" only if they insisted. The categories "separated" and "widowed" include consensual unions as well as marriages.

Comparability — Laos [top]

In 1995 and 2005, people who live together as spouses without being formally married, were recorded as married. The 2015 census question includes an additional category for persons in a consensual union.

Comparability — Lesotho [top]

All samples distinguish between "monogamous" or "polygamous" marriages.

The category "consensual union" is separately identified only in 2006.

Comparability — Liberia [top]

Marital status detail expanded over time. The 1974 sample does not distinguish between monogamous and polygamous marriages, while the 2008 sample includes separate categories. Second wives are consistently identified as being in polygamous marriages. Polygamous status is reported inconsistently for first wives and husbands, many of whom are coded as though they are in a monogamous union.

Although consensual unions are identified in 2008 only, the category of married in 1974 includes people who consider themselves as married, whether or not they are legally married. The 2008 census distinguished between separated and divorced persons; the 1974 census included a combined separated/divorced category.

Persons younger than age 10 are coded as "never married/single" in both censuses.

Comparability — Malawi [top]

Marital status categories are largely comparable over time, with divorce and separated persons combined into a single category. In the 2008 sample, couples who are not legally married but live together as man and wife are coded as "Married". The following persons are coded as "Never-married/single": persons younger than age 10 in 1987 and 1998; persons younger than age 12 in 2008.

Comparability — Malaysia [top]

The samples are comparable over time. They indicate de facto marital status.

Comparability — Mali [top]

Prior to 2009, men in polygamous marriages are identified as such, but women are not. Instead, these samples collected information on the total number of marriages formed by each woman, which is recorded in the MARRNUM variable. In 2009, information on polygamy was collected as part of marital status for both men and women.

"Separated" and "divorced" persons are explicitly combined into one category in the 1998 and 2009 samples, while the 1987 sample only includes a category for "divorced".

Persons younger than age 12 are coded as "single/never married".

Comparability — Mauritius [top]

The category of "unmarried parents" are coded as "single/never married" in 2000-2011, while single refers in all samples specifically to persons who have never cohabited.

In 1990-2011, a distinction is made between: civil marriages, religious marriages, and civil and religious marriages. Marriages are separately identified from consensual unions. Furthermore, "widowed" and "separated" include those persons previously married as well as those who were in a free union.

Comparability — Mexico [top]

All samples explicitly identify consensual unions. The marital status question differentiates between persons married in civil ceremonies, religious ceremonies, or both for most of the Mexico samples.

For some samples, the "separated" and "widowed" categories explicitly include events from both legal marriages and consensual unions.

Persons younger than age 12 are coded as "never married/single" in the 1960-2000, 2010, 2015, and 2020 samples.

The 2005 sample did not include a question on marital status.

Comparability — Mongolia [top]

Consensual unions are reported as "married" in the 1989 sample. The categories are reported separately in 2000.

Persons younger than age 15 are coded as "single".

Comparability — Morocco [top]

The category "married" includes persons who are legally or religiously married, even if the marriage has not yet been consummated. Both legal and religious divorces are included in the "divorced" category. There is no category for persons who are separated.

Comparability — Mozambique [top]

Only residents age 12 and older were asked the question.

Comparability — Myanmar [top]

In 2014, the questionnaire identifies separately persons who have renounced marriage such as "monks", "nuns", or "yawgi", which are coded as single/never married.

Comparability — Nepal [top]

In 2001 and 2011, information on multiple spouses is collected for both men and women. The proportion of women reporting more than one husband is relatively small, especially in 2001. Marriage is any union that is considered as such by the respondent, indicating that this category is not restricted to legal marriages only.

Persons who are re-married can be identified in the corresponding source variable in 2001 and 2011. These are included under the "monogamous" category in MARST.

Persons younger than age 10 are coded as "single/never married".

Comparability — Netherlands [top]

All samples identify legal marital status.

"Divorced" is combined with "Widowed" in the 1960 and 1971 samples. This combination poses an incompatibility at the first digit of the MARST variable.

Comparability — Nicaragua [top]

Children are coded as "Single/never married" if they are younger than age 10 in 1995, or younger than age 12 in 2005. Marital status was asked of all persons in 1971.

In all samples, consensual unions are identified separately from legal marriages. The category "Widowed" includes persons who experienced the death of spouse or partner. In 1995 and 2005, the sample distinguishes between persons who are legally divorced and those who are separated from a marriage or consensual union. In the 1971 sample, persons who are divorced or separated from a marriage are combined into a single category. The 1971 enumeration instructions do not specify how to categorize persons separated from a consensual union.

Comparability — Pakistan [top]

Marital status codes are consistent across samples. Persons who are formally married but live separately from their spouse are combined with married persons in each sample. Divorced persons are separately identified.

Polygamous marriages for the household head are not separately identified. However, they can be inferred in the 1973 and 1998 samples based on the number of spouses present in the household. That inference is not made for the MARST variable.

Persons younger than age 10 are coded as "Single/never married" in the 1973 and 1998 samples. Persons younger than age 12 also are coded as "Single/never married" in the 1981 sample.

Comparability — Palestine [top]

The 1997 sample distinguishes between divorced and widowed persons, while separated persons are included in the "married" category. In the 2007 sample, "divorced", "widowed", and "separated" are combined into one category. In 2017, "divorced", "widowed", and "separated" are all separate categories.

The 1997 and 2017 samples include a category for "legally engaged", which indicate that a person has a legal marriage document but is not yet married by the customs and traditions of the country. In the 1997 and 2017 samples, this is interpreted as a subcategory of "single/never-married", but it may include persons who were in a previous union. In 2007, there is no "engaged" category. Since the 2007 instructions state that previously married people should not be included in the "engaged" category, then engaged persons are likely coded as "never married".

In 1997 and 2007, persons younger than 12 are coded as "single/never married". In 2017, persons younger than 14 are coded as "single/never married".

Comparability — Panama [top]

Indigenous persons were not asked the question in the 1960 sample.

Separated persons are considered "married" in the 1960 sample. The 1970-2010 samples distinguish between marriages and consensual unions as well as persons separated from a legal marriage and those separated from consensual unions. The category "widowed" includes persons who experienced the death of a spouse or partner in consensual unions.

Persons younger than age 15 are coded as "single" in the 1980-2010 samples.

Comparability — Papua New Guinea [top]

The Papua New Guinea samples do not explicitly indicate how to treat consensual unions. However, enumeration instructions suggest that persons living together should be considered to be married.

The 1980 sample includes widowed, divorced, and separated all as a single category. Widowed is separately identified in the 1990 sample. The 2000 sample distinguishes these three categories separately

Comparability — Paraguay [top]

Category availability differs across samples. The 1962 sample differentiates between "civil" and "civil and religious" marriages. Marriages and consensual unions are distinguished for all samples. Divorced and separated categories are combined for only the 1962, 1972, and 1982 samples.

The age of those enumerated for this question differs across samples. All persons were asked in 1962, persons 12 and older were asked in 1972 and 1982, and persons 10 and older were asked in 1992 and 2002.

Comparability — Peru [top]

The census identifies both formal marriages and consensual unions. The "separated" category includes persons separated from either formal or consensual unions. They are separately identified from persons who are formally divorced. The "widowed" category includes persons who experienced the death of spouses or partners in consensual unions.

Persons younger than age 12 are coded as "single/never married" in the 1993, 2007, and 2017 samples.

Comparability — Philippines [top]

In all samples, persons younger than age 10 are coded as "Never married/single". Permanently separated persons were recorded as such, regardless of whether the separation was legal or not.

Comparability — Poland [top]

The 1978-1988 samples refer to a de facto marital status, while Poland 2002 and 2011 report legal marital status.

In 1978 and 1988, the "married" category includes those persons who entered into marriage even if they are not formally married. Similarly, for these two samples, persons not living together in a marriage are registered as divorced regardless of formalities.

The Poland 1988 and 2002 distinguish between married couples living together from other cases living apart or with an absent spouse.

In 2002 and 2011, information about common-law partnerships is reported in a separate source variable (de facto marital status) that is recorded in CONSENS.

Persons younger than age 15 are coded as "Never married/single" in the 2002 and 2011 samples.

Comparability — Portugal [top]

All samples record de facto marital status as reported by the respondent. The 1981 sample combines legal and de facto marriages, while the later samples distinguish between them. The 1981 sample only reports legal separations. In the 1991 and 2001 samples, "Separated" include separations from consensual unions. For these samples, the enumeration instructions do not specify how to treat persons who experienced the death of an unmarried partner. The 2011 sample distinguishes between divorce and de facto separation (still married).

Comparability — Puerto Rico [top]

The 1970-1990 samples collect information on persons in either formal or consensual unions. Consensual unions can be inferred in the 2000, 2005, 2010, 2015, and 2020 samples from the RELATE variable only for the household head and his/her unmarried partner.

In the 2005 sample, persons who have an annulled marriage are coded as "Never married".

Comparability — Romania [top]

The marital status question addresses legal status only. There is no category for de facto separations in any of the samples. In these situations, respondents decide for themselves whether to say that they were married or divorced.

The 2002 and 2011 samples ask a separate question about consensual unions (see the CONSENS variable). This information is not incorporated in MARST because it would mask the legal marital status information for the person.

Persons younger than age 14 were presumed to be "Never married/single" during enumeration, according to instructions documents in the 1992-2002 samples.

Comparability — Russia [top]

The 2010 sample distinguishes between "separated" and "divorced" categories, while the 2002 sample contains no such distinction and both cases are included in the "separated or divorced" category.

Persons younger than age 16 are coded as "never married/single" in 2002 and 2010.

Comparability — Rwanda [top]

Persons under age 12 are coded as "Single/never married" in 2012.

The 1991 and 2002 samples report the de facto marital status. The 1991 sample distinguishes between customary and legal marriages. In 2012, the enumeration instructions suggest that persons in consensual unions should be coded as married.

Only men are identified as polygamous in 1991 and 2012. Both sexes can be coded as "Polygamous" in the 2002 sample. The samples record additional detail for persons in polygamous marriages in the POLYGAM variable, including number or wives and rank order of each wife.

Divorced and separated are combined in 1991 and 2002, while they are separately identified in 2012.

Comparability — Saint Lucia [top]

Both samples identify formal marital status. Cohabitation is very common, resulting in unusually low marriage rates in MARST. Persons in consensual unions are identified separately in the 1991 census (see CONSENS).

The 1980 sample collects information on current cohabitation only for reproductive-age women (ages 14 to 45). This information is available in the unharmonized source variable LC1980A_0437. Most consensual unions involve the household head, and can be inferred in 1980 based on the presence of a "Spouse/partner" in the harmonized RELATE variable.

Persons younger than age 13 are coded as "Single/never married" in 1980, as are persons younger than age 14 in 1991.

Comparability — Senegal [top]

The 1988-2013 samples identify polygamous marriages for men and women. Marriages include both civil marriages and those conducted according to local custom.

In 1988 and 2002 samples, persons living in informal unions and persons who married but live separately from their spouses are coded as "other cases". Because it is impossible to determine marital status for these persons, they are included in the "unknown" category of MARST. In the 1988 sample, the "unknown" category includes additional persons with unknown marital status, but these can be distinguished from "other cases" in the unharmonized source variable. In 2013, separations and consensual unions are separately identified.

Persons younger than age 6 are coded as "single/never married" in 1988, as are persons younger than age 12 in 2002. In 2013, residents younger than age 10 in ordinary households are coded as "single/never married".

Comparability — Sierra Leone [top]

The samples includes a category for "engaged". This has been coded as a subcategory of "never-married/single" in the integrated variable. Persons younger than age 10 are assumed to be "never married".

The source variable for 2015 distinguishes between cohabitation for less than five years and equal to or more than five years.

Comparability — Slovenia [top]

The samples report legal marital status.

Comparability — South Africa [top]

In 1996 and 2007, only persons in private households are asked the question on marital status. Persons living in collective dwellings are excluded.

Polygamous marriages for men are identified in the 2001 and 2007 samples. Women in polygamous marriages must give one of the other "married" responses. Polygamous marriages are not identified in the 1996, 2011, or 2016 for either gender.

"Traditional" marriages include persons married according to African traditions or customs. If persons married according to traditional customs also have a religious or civil marriage, they are reported as "civil/religious" in the 2001 sample.

In 2016, a distinction is made between single persons who "have never lived together as husband, wife or partner before" and those who have. The latter is coded as "single/never married"; persons younger than 12 and those who never cohabited are coded as "never married and never cohabited."

Comparability — South Sudan [top]

The category "married" may include persons who are legally or religiously married. The "divorced" category may also include "separated".

Persons younger than age 12 are coded as "single/never married".

Comparability — Spain [top]

All samples indicate the person's legal status. The 2001 sample has separate information for consensual unions (see CONSENS). This information is not incorporated in MARST because it would mask the legal marital status information for the person.

In 2001, only persons who reside in the household received the question on marital status.

Persons who are separated or divorced are combined in 1981, but identified separately in later years.

Comparability — Sudan [top]

Persons younger than age 12 are coded as "never-married/single".

Comparability — Suriname [top]

In 2004 and 2012, married is interpreted as formally married. Separations are not identified in these samples.

In 2012, persons younger than age 15 are coded as "single/never married".

Comparability — Switzerland [top]

All samples collect information on formal marital status. Separated persons are included in the "married" category.

The 2011 sample includes a separate category for registered partnerships, which is classified as "consensual union". For this sample, dissolved partnerships are included within "divorced".

Comparability — Tanzania [top]

The 2002 and 2012 samples contain greater marital status detail than the 1988 sample. Persons in consensual unions are identified only in 2002-2012. Separated and divorced persons are combined in the 1988 sample, while they are identified separately in 2002 and 2012.

Polygamous unions are not explicitly identified, but can be inferred for the head of household based on the presence of multiple spouses. That inference is not made in MARST.

Comparability — Thailand [top]

The samples collect marital status for all persons except Buddhist monks. For nuns, enumerators asked whether they had ever married, but did not identify the reason the marriage ended. The unharmonized source variables for each sample include a code for "monk", as well as a code for persons who were ever married but whose current marital status was unknown. This latter category contains many persons with unknown marital status in addition to nuns. It is coded as "unknown" in the MARST variable.

Married couples who live separately, but still consider themselves married, are included in the "married" category. All samples distinguish separated from divorced persons.

Persons younger than age 11 in 1970 and 1980, and younger than age 13 in 1990 and 2000 are coded as "never married".

Comparability — Togo [top]

The 1960 and 2010 samples identify persons in consensual unions. The 1970 sample does not explicitly identify consensual unions; it is unclear if persons in consensual unions are included in the single or married category.

All Togo samples identify polygamy. Information on the number and order of spouses is available in unharmonized source variables. Polygamous responses recorded in MARST include both men (married to multiple women) and women (married to a man with multiple spouses) in 1960 and 2010, but are provided only for men in 1970 (inferred from the number of spouses).

In 2010, marital status information was not reported for visitors or children younger than 12 years old. For this sample, children under age 12 are coded as "never married/single".

Comparability — Trinidad and Tobago [top]

All samples refer to legal marital status. Married persons include those not living with their partners, unless they are legally separated or divorced. In the 1970 sample, persons with "other" marital status are coded as "unknown".

Persons younger than age 15 in the 1970 sample and younger than age 14 in other Trinidad and Tobago samples are coded as "never married/single".

Information about marital status was not collected for the 1980 sample for persons age 14 and older who are attending primary or secondary school in a full time basis. For this sample, persons attending these school levels full time are coded as "never married/single" if they are between 14 and 18 years old and as "unknown" if they are 19 years old or more. Users can refer to the unharmonized source variable for 1980 if they wish to perform an alternative coding of these cases.

Comparability — Turkey [top]

In all samples, children younger than age 12 are coded as "single/never married".

Marital status codes are consistent across samples and reflect de facto rather than legal status. The category "married" includes both common-law and religious marriages. The samples do not allow a response of "separated". These persons are most likely included in the "married" category.

Comparability — Uganda [top]

Persons younger than age 10 are coded as "never married/single" in the 1991-2014 samples.

Persons in consensual unions are considered "married". This category includes polygamous unions, which are separately identified in 2002 and 2014. Separated and divorced persons are identified separately in 1991 and 2014, but are combined in 2002.

Comparability — Ukraine [top]

The marital status question in the Ukrainian census collects information on legal and informal relationships. The census includes a category for people currently in an "unregistered marriage", which is coded as a cohabiting union. In addition, the categories for separated and widowed include both formal and unregistered marriages.

Persons under age 15 are assumed to be never-married.

Comparability — United Kingdom [top]

The 1961-2001 samples ask for legal marital status, but in the 2001 sample there is a separate question to identify de facto separations. In the 1991 and 2001 samples, persons who are in their second or subsequent marriage are identified in the unharmonized source variable. In 1961, 1971 and 1991, if a person was legally separated but not divorced, they were considered to be married.

Comparability — United States [top]

Beginning in 1980, couples who live together but are unmarried are allowed to report the marital status they consider most appropriate. Persons in common-law marriages are considered "married" in all samples.

Beginning in 1990, consensual unions can be inferred using the RELATE variable, but would only pertain to unions that include the head of household. Such unions cannot be identified in earlier samples.

Persons younger than the following ages are coded as "never married/single": age 14 in the 1960-1970 samples; age 15 in the 1980-2020 samples.

Comparability — Uruguay [top]

Children younger than age 12 in 1996 and 2011 and younger than age 14 in 2006 are coded as "single/never married." There is no age restriction in earlier samples.

In 2006 and 2011, MARST is constructed based on two questions: current type of union (marital or cohabiting) and marital status. The 2006 marital status question reflected formal marital status only and was asked only of person who were not currently married, while the 2011 asked marital status of persons who were not living with a spouse or partner and included categories of separation and widowhood from consensual unions.

Persons in consensual unions are identified in every sample, but the treatment of separated and widowed persons varies across samples. In 2006 and 2011, persons separated from legal marriages are to be included with persons in intact marriages--but they can be identified as not living with a partner in the source variables. It is likely that separated persons are treated similarly in 1963. While there is no separated category in the 1963 and 2006 samples, the 2011 sample includes a category for persons separated from consensual unions. The 1975 sample provides the most detailed information on marital status, distinguishing separated by union type (marriage or cohabitation). In 1985 and 1996, the category for "separated" includes both marriage and consensual unions.

In 1963 and 2006, the category "widowed" in both samples appears to refer only to persons widowed in a legal marriage. In 1985 and 1996, this category also includes persons widowed from consensual unions. The 1975 and 2011 censuses distinguish between persons widowed in formal marriages and consensual unions. The 2011 census further distinguishes between persons in same-sex and different sex-unions. This distinction is preserved in the unharmonized source variable.

Comparability — Venezuela [top]

The 1971 sample allows no response for "separated". These persons are included in the "married" category.

The 2001 sample specifies that "separated" and "widowed" statuses can refer to either a previous marriage or a consensual union. Previous samples do not discuss this point or explicitly restrict these categories to formerly married persons.

Comparability — Vietnam [top]

The instructions concerning married status are liberal. They state that the "married" category includes persons joined by custom or law, as well as those living together as husband and wife. Nevertheless, "divorce" refers strictly to the legal status of marital dissolution.

Persons younger than age 13 are coded as "never married/single" in 1989 and 1999. Persons younger than age 15 are coded as "never married/single" in 2009 and 2019.

Comparability — Zambia [top]

In 1990, persons who were living together as man and wife but who had not been formally married are included in the married category. Beginning in 2000, consensual unions are separately identified. Separated, divorced, and widowed persons are identified in each year.

In all years, children under age 12 are assumed to be never-married. In 2010, no information was collected on marital status for usual residents of the household who were absent at the time of the census.

Comparability — Zimbabwe [top]

Persons younger than age 10 are coded as "single/never married" in 2012. In this sample, enumeration instructions indicate that persons living together should be recorded as married. The sample combines separated and divorced persons into a single category.