menage - Millennium Development Goals Indicators

their use in constructing the life table, no attempt has been made in theStatistical Office to evaluate the quality of the values, and hence noneare set in italics.Coverage: Values are available in table 21 for 120 countries. Lack ofreliable base data restricts the number of countries for which life tablescan be constructed but, even so, the coverage has increased notably inthe last few years.Limitations: Expectation-of-life values are subject to all the limitationsof population and mortality statistics.Perhaps the most important specific qualification which can be setforth in connexion with expectation of life values is that they must beinterpreted strictly in terms of the underlying assumption that survivingcohorts are subjected to the age-specific mortality rates of the periodto which the life table refers. Thus, in interpreting !'iO in the 1963-65life table for England and Wales it may be said that, if males reachingage 50 were to experience for the rest of their lives the same agespecificmortality rates that obtained during the period 1963-65, theywould, on the average, live 22.8 years past the age of 50 or to 72.8years of age. Experience during the past decades has indicated,however, that an assumption of unchanging mortality in fact understatesthe expectation of life, since mortality rates are decreasingalmost universally.Table 22Because this issue of the Demographic YearbooK features statisticsof marriage and marriage dissolution, table 22 is devoted to a timetrend of the annual numbers of marriages beginning 1949 and ended1968. This together with table 15 of the 1958 issue and table 27 of the1948 issue provides annual frequencies for a number of countries backto 1932. These are statistics of marriages performed, obtained fromcivil registers, and not of persons marrying.In accordance with the usual practice, the code which indicates thequality of civil-registration data is presented in the first column of thetable in the form C. U and ...• and, in addition data believed to beunreliable (incomplete) or of unknown reliability are set in italics ratherthan roman type. A description of the basis for this quality specificationmay be found on p. 11.Coverage: Marriages are shown for 165 countries. Major areas withdata for at least one year are included and those for sub-nationalsegments of population are presented when statistics are not availablefor the country as a whole. Such data are of minor significance, butthey do serve to identify the countries for which marriage statistics areobtainable on a national basis.Limitations: Data in this trend table are subject to almost all thelimitations of vital statistics set forth on p. 10, as well as those pointedout in connexion with marriage statistics specifically. They suffer fromnational variations in statistical definitions; they are deficient asindicators of de facto family formation to the extent that consensualunions exist; even as measures of formal marriage incidence, they areunreliable to the extent that marriage registration is incomplete.However, it must be emphasized that since marriage is a legal ratherthan a biological event the problem of comparability in the sense ofcomplete and incomplete data takes on a somewhat different aspect.This should be borne in mind in using the quality-of-data code C andU. which appears in the first column of the table.The symbol (t) is used here, as elsewhere in the DemographicYearbook, to indicate that data have been tabulated by date of registration,rather than by date of occurrence or performance, in this case. Aswas pointed out on p. 13,this qualification is not particularly importantin connexion with marriages because, in most cases, registration is aprerequisite for, or equivalent to, performance. However, since thereare some countries in which a potential time interval may exist betweenperformance of a marriage and its registration,65 the implications of thesymbol should be considered in relation to the data it qualifies.Table 23Table 23 is the crude-marriage-rate trend table, based on frequenciesset forth in table 22. Rates for each year, 1953-1968 plus five-yearannual average rates for the period 1920-64 makes available a 48 yeartime trend for a number of countries.Data believed to be unreliable (incomplete) or of unknown reliability,are set in italic rather than roman type. A description of the basis forthis Quality specification may be found on p. 11.65 See Handbook of Vital Statistics Methods. op. cit.. Table 10, pp. 84-86.Coverage: Rates are shown for 161 geographic areas. This coverageis slightly less than the corresponding "numbers" table because of thelack of population figures in some cases.Rate computation:Annual crude marriage rates are the number of marriages performedper 1 000 persons in the population estimated at the mid-point of theyear in question.Average annual marriage rates are the aggregate number of marriagesperformed in the five-year period per 1 000 aggregate population inthe period.The marriage frequencies which form the numerator of the rates arethose given in table 22 and in the Statistical Office's files. Thepopulations-mid-year or other-used in the computation of the ratesare, in general. the figures presented in table 4. Every effort was madeto achieve strict correspondence between the numerator of the rate(marriages) and the denominator (population). and notes are used todescribe instances where correspondence could not be obtained.Limitations: Since rates in this table are based primarily on frequenciesfrom table 22, they are also subject to the limitations described inconnexion with that table.In addition, it should be noted that rates are affected also by thequality and limitations of the population estimates which are used intheir computation. This factor is of special importance because of thepolicy, adopted in 1965 Demographic Yearbook, of computing rateson estimates of population constructed by the United Nations, whenadequate official series were lacking.As will be seen from the footnotes, strict correspondence betweenthe numerator of the rate and the denominator is not always obtained;for example, marriages among civilian and military segments of thepopulation may be related to civilian population only. The effect of thismay be to increase the rates or, if the population is larger than thatfrom which the marriages are drawn, to decrease them but in mostcases, it is probably negligible. The potential size of the effect may bejudged from the rate series for Gibraltar.High incidence of consensual unions will be especially noticeable inthe rates, because the range of formal marriage rates is relativelynarrow and marked divergence is almost always due to a preponderanceof de facto rather than de jure unions. Lack of comparability from thisfactor is especially evident in abnormally low marriage rates for parts ofLatin America and Africa.Finally, it should be emphasized that crude marriage rates-likecrude birth and death rates-are not probability indices, because theytake no cognizance of the age/sex/marital status structure of thepopulations to which they relate. Nevertheless, they do measure levelsand changes in marriage frequency. For more details on the uses ofcrude marriage rates, see chapter I of the 1958 Demographic Yearbook.Table 24Marriages and marriage rates specific for urban and rural residenceare shown for the first time in table 24. These data for the latestavailable year complement those shown in table 22, since they are theurban/rural component of the total marriages performed, shown in thattable.Frequencies shown are obtained from civil registers of marriage andrates based on these frequencies have been computed in the StatisticalOffice of the United Nations by the method described below.Urban is assumed to have been defined to accord with the nationaldefinitions of urban and conversely of rural used in the censuses ofpopulation. These national definitions are set forth at the end of table11.Data believed to be unreliable (incomplete) or of unknown reliability,are set in italic rather than in roman type. A description of the basis forthis quality specification may be found on p. 11.Coverage: Marriages by urban/rural residence are available for 20countries and rates specific for this characteristic have been computedfor 14 of these.Rate computation:Urban and rural marriage rates are the number of marriages reportedas urban and rural per 1 000 urban or rural population at the midpointof the year. The population used as the denominator of the rates is thatshown in table 11 of this issue.Limite.: 'Is: Since this is an extension of tables 22 and 23, data aresubject .0 all the qualifications set forth in connexion with those trendtables.In addition, they are subject to all the limitations imposed by thelack of international comparability among national concepts of urbanity31