INATIONAL CENTER Series 4For HEALTH STATISTICS Number 2

VITAL and HEALTH STATISTICS

DOCUMENTS AND COMMITTEE REPORTS

National

Vital Statistics Needs

A Report of the United States National

Committee on Vital and Health Statistics

An analysis of the role of the national vital statis-

tics agency, with recommendations concerning the

procurement and production of regular annual data,

the need for special studies, provision of technical

services to the States, and other policy questions.

Washington, D.C. September 1965

U.S. DEPARTMENT OF

HEALTH, EDUCATION, AND WELFARE Public Health Service

Anthony J. Celebrezze Luther L. Terry

Secretary Surgeon General

This report was originally published as Vital Statistics—Special Reports,Vol. 45, No. 11 (July 1957). Because its recommendations for meetingneeds for national vital statistics are still of interest, the report isnow being reprinted without change in the National Center for HealthStatistics Vital and Health Statistics series.

Public Health Service Publication No. 1000 -Scri es 4-No. 2

NATIONAL CENTER FOR HEALTH STATISTICS

FOR’REST E. LINDER, PH. D., ~Z’7%C~07

THEODORE D. WOOLSEY, Ikputy Director

O. K. SAGEN, PH. D., ~Jf;f~(zn~ ~i?’eC~O~

WALT R. SIMMONS, M. A., stati~tical Advi~or

ALICE M. WATERHOUSE, M. D., Medical Advisor

JAMES E, KELLY, D.D.S., Dental zfdui$o?’

LOUIS R. STOLCIS, M.A., Executive O@cer

OFFICE OF HEALTH STATISTICS ANALYSIS

IWAO M. MORXYAMA,PH. D., Chbf

DIVIS1ON OF VITAL STATISTICS

ROBERTD. GROVE, PH. D., Chief

DIV1S1ONOF HEALTHINTERVIEWSTATISTICSPHILIPS. LAWRENCE, Sc. D., Chief

DIVISION OF HEALTH RECORDS STATISTICS

MONROE G. SIRKEN, PH. D., Chief

DIV1S1ONOF HEALTHEXAMINATIONSTATISTICSARTHURJ. MCDOWELL, Chid

DIVISION OF DATA PROCESSING

SIDNEY BXNDER,Chief

Public Health Service Publication No. 1000-Series 4-No. 2

CONTENTS

SUMMARYOF REPORT

PageIntroduction ----------------------------------------------------------------------------------- 221The role of tbe National Office of Vital Statistics .-- —-. -— -------------------------------- 221Major recommendations and observations ------------------------------------------------------- Z2ZOther recommendations ------------------------------------------------------------------------ 223

Completeness of &b ------------------------------------------------------------------------- 223uIIifOXIUi& Of &ta---------------------------------------------------------------------------- 223

Scope and detail of &ti---------------------------------------------------------------------- 223Explanatory and interpretive tefi ----------------------------------------------------------- 223Cyclical schedule of tabulations ------------------------------------------------------------- 223

Utilization of data by consumers ----------------------------------------------------------------- 224

DETAILED REPORT

“---- --- —_ .._. . . -L. s–. ---iicu~ aim meumu ox SIULIy— -------- ------------------------------------------------------------

The present national vital statistics system ----------------------------------------------------Legislative authority ------------------------------------------------------------------------Functims of the National Office of Vital Statistics --------------------------------------------Organtaation, budget, and personnel of the National Office of Vital Statistics ------------------

Findfng$ and recommendations --------------------------------------------------------- ---------Regular production of basic national vital statistics ------------------------------------------Statistical studies of public health and population chuacteristics ----------------------------Coordination of the national vital statistics system ---—--— ------------------------Observations on economy and efficiency of operations ----------------------------------------Methods of producing national vital statistics -----------------------------------------------

APPENDIXES

A, Consumer views on national vital stitistic&--------------------------------------------------Use of vital stithtics ----------------------------------------------------------------------Scope or detil of subject matter, timeliness, and accuracy --------------------------------Reduction in quantity of matertal published by the National Office of Vital Statistics -------Tables A-1 through A-6-------------------------------------------------------------------

B, Consumer views on national marriage s~tistic=--------------------------------------------Major ffndfngs of survey -------------------------------------------------------------------Description of smvey ----------------------------------------------------------------------Analysis of questionnaire returns ---------------------------------------------------------Tablee B-1 through B-10 -----------------------------------------------------------------

C. Survey of methods of producing national vital statistics --------------------------------------Methods of productng tdional vital statisftcs ---------------------------------------------The problem of duplication -------------------------------------------------------------The transcript metkd -------------------------------------------------------------------The punched-card metiti ------------------------------------------------------------------The pretabulation method ----------------------------------------------------------------Tablea C-1 and C-2------------------------------------------------------------------------

D, Technical services to Statee ---------------------------------------------------------------Types of sefice&------------------------------------------------------------------------

219

~L[

22’7228228230230231237238239240

243243244244245248248248249251256

256257258259262

264268266

National Vital Statistics Needs ~

A Report of the United States National Committee on Vital and Health Statistics

SUMMARY OF REPORT

Introduction

The Committee was requested by the U. S. Pub-lic Health Service to make a comprehensive studyof the objectives and the program of national vitalstatistics, to report its findings, and to make recom-mendations for the future development of the pro-gram. It was asked, also, to undertake a specialstudy of the needs for marriage statistics and theextent to which existing data satisfy these needs.

The Committee’s findings and recommenda-tions are based ox

1. An extensive questionnaire survey of theprincipal users of the various kinds of vital statis-tics and a special questionnaire survey of the majorusers of marriage statistics.

2. Partici@tion in apaneldiscussion at a meet-ing of the American Association of RegistrationExecutives on the subject, “A National System ofVital Statistics.”

3. Intensive studies of particular phases of theproblem.

4. Frequent long discussion meetings of thesubcommittees.

This report is in three parts-the summary,the detailed report, and the appendixes containingthe detailed information upon which the report isbased.

The Committee has examined the major areasof responsibility of the National Office of Vital Sta-tistics (NOVS) and for each area has outlined the

goals to be achieved. This is followed by a state-ment of the principal deficiencies in the presentsystem with respect to these goals, and finally bythe recommendations.

The Role of the National Office ofVital Statistics

In this country, the registration of vital eventsis the responsibility of the States. Each State alsohas the responsibility for tabulating and publishingits own data so as to meet its special needs. Thetabulation and publication of mtional data on births,deaths, notifiable diseases, marriages, and divorcesand annulments are functions of the National Officeof Vital Statistics, a branch of the Division ofGeneral Health Services, Bureau of State Serv-ices, Public Health Service, U. S. Department ofHealth, Education, and Welfare.

The functions of the NOVS should not be re-garded as confined to the routine processing andpublication of historic series of data. The justifi-cation for publishing data is that they serve someuseful purpose. Vital statistics are essential foradministrative purposes in both business and gov-ernment, and also for research. Used in conjunc -tion with population data, they provide a backgroundfor public health programs and medical researc~for making projections of the size and compositionof the population; sndfor the study of trends regard-ing family formation, composition and dissolution,and other aspects of our society. The NOVS hasmajor responsibilities both for making available the

221

222 VITAL STATISTICS-SPECIAL REPORTS

kind of data that will be most useful in these areasand for stimulating widespread and fruitful applica -tion of these data. These responsibilities call forforesight and flexibility to meet changing needs.Further, although many of the uses of vital statis-tics lie in the field of public health, the data alsohave important applications to demography.

The fact that the NOVS is a part of the PublicHealth Service insures full recognition of the healthinterests in vital statistics. This implies an obli-gation and responsibility to maintain a national vitalstatistics program that is not overshadowed byhealth inte~ests but also meets the justifiable needsin the demographic and other areas of interest.

Major Recommendations and Observations

A complete list of the recommendations is givenin the main body of this re~rt. Of these, the Com-mittee wishes to draw particular attention to the fol-lowing which appear to deserve the highest priority.

1. Immediate priority should be given to theprocessing and procurement of data so as to achievea schedule of processing which permits the releasefor printing of final data within 15 months after theclose of the data year, and this schedule should bemaintained or improved upon in the future.

2. Statistical studies that contribute to knowl-edge in public health, fertility, marriage, divorce,and mortality should be developed as part of theregular program of the NOVS. In particular, atten-tion should be given to the influence of ~pulationcharacteristics, such as sex, age, marital status,and occupation. The program should include col-lection of new data and development of new methodsof analysis that are needed for program planning,administrative, and research purposes.

3. The NOVS should be given sufficient fiscalresources to provide more adequate technical serv -ices to the States in the following areas: profession-al advisory and consultant service on technical andmanagement problems; expanded information andclearinghouse services on technical subjects andmethodology; educational programs directed to-wards improvement of source data; national pro-gram for recruiting and training professional vitalstatistics personnel; and work with State health de-partment program divisions in promoting effectiveutilization Oi vital statistics data.

4. The Committee has also examined two alter-native methods of producing national vital statistics,

in particular the use of data pretabulated by theStates and of punched cards supplied by the States,

Recommendations with regard to these alternativesare as follows:

A. The NOVS should reject any data procurementmethod that limits or restricts the use of vitalstatistics records for research or that preventsit from controlling the quality of national vitalstatistics tabulations. Specifically, the NOVSshould procure transcripts of the individual vitalrecords for processing.

B. The NOVS should reject the proposal to producenational vital statistics from data pretabulatedby the States because:

(1)

(2)

(3)

the available national ditta would shrink tothose obtainable from a minimum rigid ltstof tabulations;it would become very difficult to make na-tional studies based upon information ob-tainable only from transcripts of the indi-vidual records ok from individual punchedcards;it would add to the cost of State operationswithout a compensatory decrease in the costto the NOVS.

C. The NOVS should reject the punched-card meth-od of collecting vital statistics data as a generalmethod applicable to all States because:

(1)

(2)

(3)

(4)

(5)

it restricts the freedom of action both of theState offices of vital statistics and of theNOVS, thus making vital statistics less use-ful rather than more useful;effective leadership in national and interna-tional vital statistics will be sacrificed dueto the loss of technical skills and knowledgein the NOVS;the scope and detail of national vital statis -tics cannot be greater than those of the par-ticipating State ranking lowest in these re -pects;the possible savings in cost are outweighedby the dieadvantages of the method;the overlapping of State and Federal’ vitalstatistics procedures, tabulations, and needsis not great enougirto make this method pos -sible without serious disadvantages to bothparties.

Vol. 45, No, 11 NATIONAL VfTAL STATISTICS NEEDS 223

D. The NOVS should abandon any further experi-mentation with the punched-card method in thecollection of mortality statistics and if the useof this method to collect birth statistics is con-tinued at all, should limit it to the few Stateswhere it can be shovm to be mutually advanta-geous and where it will not adversely affect thescope, quality, continuity, and usefulness of na-tional vital statistics.

E. The NOVS should continue to cooperate with theStates in studies of ways to improve the divisionof labor in the national vital statistics system.

Changes in the system should be made only afterfull consideration of:

(.1) the basic requirements for national vital sta-tistics specified on page 231 of this report;

(2) safeguards for the continuity of operation;(3) opportunities to ixqprove the final product

and to render maximum services.

5, The Committee also notes that marriageand divorce (including annulment) statistics are ina highly unsatisfactory state as regards geographiccoverage, uniformity of reporting, accuracy andamount of detail. A program for the improvementof these data should be consistent with the aboverecommendations and receive the same level ofpriority, but it is recognized that it is not possibleto carry out such a program under the presentbudgetary position, The Committee recommendsthati

A,

B,

c.

The NOVS work actively to est.ablisha MarriageRegistration Area through implementation ofthe four criteria established for admission tothe area, and to extend it until every State isincluded.Methods of collecting marriage statistics beexplored with a view to assu~ing (a) completecoverage of the United States, (b) uniform databoth between and within States, (c) comparabilityof time series, and (d) accuracy of information.The NOVS obtain copies of marriage recordsfrom the States and local areas to be-processedin the NOVS,

Similar actions are needed for national statis -tics on divorces and annulments. However, specificrecommendations are deferred until a thoroughstudy of this problem can be made.

Other Recommendations

Completeness of data

The NOVS should stimulate and assist in spe-cial studies of birth registration completeness inareas where completeness of registration is stilla problem.

The NOVS should endeavor to develop methodsfor measuritig incompleteness of death and late fetaldeath registration adaptable to those areas wherethis p>oblem is important. If practical techniquescan be found, they should be applied to obtain moreinformation on the extent of incompleteness and onmeans for attaining more complete reporting.

Effort? toward improvement of notifiable dis -ease reporting should be concentrated first on thosediseases that are major health problems and areamenable to control procedures.

The NOVS should cooperate with the States inthe development and application of tests to deter-mine the completeness of the reporting of mar-riages and divorces and annulments to the centraloffices, and the accuracy of the information givenon the report forms.

Uniformity of data

The NOVS and the State vital statistics officesshould continue their efforts to obtain completecoverage of all items on the standard certificates.

Scope and detail of data

The NOVS should keep in closer touch with theusers of its data when planning the scope and detailof its tabulations and publications and making deci-sions on fundamental issues.

Explanatory and interpretive text

The analytical functions of the NOVS should bestrengthened to make available more explanatoryand interpretive material useful to consumers.

Cyclical schedule of tabulations

The concept of cyclical collection and tabulationof data should be firmly established as a part of theregular program of the NOVS with adequate safe-guards for the continuity of the various tabulationcycles.

224 VITAL STATISTICS—SPECIAL REPORTS

Utilization of Data by Consumers

Questionnaires were sent to 369 persons se-lected for their presumed interest in some aspectof vital statistics data. Replies were receivedfrom 254. Although the Committee was aware” of

the limitations of a questionnaire of this type, in-formation was sought regarding the relative f re -quency with which various types of vital statisticsare used and the different sources from which

these data are obtained, the extent to which therespondents regard available vital statistics asadequate for their uses, and ways of reducing the

quantity of material published by the NOVS yet

still meeting the important needs of consumers.The results of the questionnaire may be sum-

marized as follows (see Appendix A for details):

1. Frequency of use of types of data. —Mor -tality and natality statistics are the data issuedby the National Office of Vital Statistics which areused most frequently by the persons replying. Nextin order of frequency of use are the statistics onmarriages and notifiable diseases. The consumerssurveyed showed the least interest in fetal death

statistics and divorce statistics. This generally

corresponds to the requests for data received bythe NOVS.

2. Geographic categories used. —An over-whelming majority of the consumers surveyed in-dicated use of statistics for the country as a wholeand for States. Also, the majority of resgmdents

appeared to have need for data on all of the other

geographic areas listed.3. Use of State and local publications .—Ex-

cluding State registrars of vital statistics, rela-tively little reference appears to be made to State

and local publications. However, this may be due,in part, to the composition of the list of persons

and agencies surveyed.

4. Secondary references. —Of the secondaryreferences, the Statistical Abstract of the UnitedStates published by the Bureau of the Census ismost frequently used, followed by the United Na-

tions Demographic Yearbook and by the WorldAlmanac.

5. Unpublished data . —About one-half of therespondents stated that they have requested un-published data from the NOVS.

6. Timeliness .—The dissatisfaction expressedmost frequently and most strongly concerning theNOVS publications related to the lack of timelinessof their issuance. The release of unpublished datato those requesting them appears to have softenedthe reactions of some of the respondents. Consid-

erable understanding and appreciation of the prob-lems faced by the NOVS in the early publication ofdata were evident, but the general opinion was that

the data would be much more useful if they could

be issaed on a more timely schedule.‘7. Accuracy. —Most of the respondents did not

express any opinion concerning accuracy, exceptto indicate general satisfaction with the quality of

data. Most of the defects mentioned were thosethat have been already recognized by the NOVS,namely, errors in residence allocation of birthsand deaths, and incompleteness of notifiable dis-ease statistics.

8. Scope and detail.—A large proportion ofrespondents indicated satisfaction with the existing

scope and detail of the published data. There wererelatively few suggestions for reducing the quantityof data published.

Another questionnaire devoted exclusively tomarriage statistics, was sent to 539 persons andorganizations to obtain information concerning usesand needs of consumers of marriage statistics. Re-

plies were received from 262 or 49 percent.Since users of marriage statistics were se-

lected as respondents it is to be expected that use

would be reported in ways proportionate to the kindof persons included in the mailing list, namely,business men, university people, private research-

ers, and government officials. In the judgment ofthe Committee the mailing list of business usersstressed national organizations, and the surveymay have failed to reach local users in businessfields. While one question dealt with potent ial usesif statistics were available, it is doubtful that thefull effect of supply on demand is revealed. Per-haps improved statistics must actually be available

in order to call out full expression of demand from

consumers. Probably the questionnaire surveywas most effective in revealing the varied uses ofmarriage statistics, in indicating the scope of thedesired data, and in eliciting new suggestions (seeAppendix B for details).

Vol. 45, No, 11 NATIONAL VfTAL STATISTICS NEEDS 225

There are four general findings of the surveycm marriage statistics:

1, There is evidence of extensive use of mar-riage statistics in large business organizations, incolleges and universities, in government, and inprivate research.

2. The returns indicate use by individuals inbusiness organizations for estimating householdformation and for predicting demand for consumergoods . Persons in research and academic fieldsexpressed a preference for data useful in socio -logical and demographic research.

3. With respect to the time-space aspects ofstatistical reporting, the broader categories werefavored. Although there was strong support formonthly totals and for annual national and Statereporting of more detailed figures, there seemedto be a relatively limited demand for reporting ofthe monthly figures in advance of present publi-cations.

4. Strong support was expressed for completenatioml and State coverage on an annual basis sndfor more specific data, with cross -tabulations bybride and groom, and informationavailable from registration records,

on items not

FIIWIRE 1.. -----

FLOW OF VITAL RECORDS AND STATISTICS IN THE UNITED STATES

RESPON~LE FETAL DEATH

PEWN OR BfRTH CERTIFICATENOTIFIABLE DISEASE

DIVORCE, C

DEATH CERTIFICATE CERTIFICATEREPGRTfNG

AGENCY (Stlilbirfh)RSFORT OFFfCIALS

bfARRIAGE RECORD ANNULMFARJXORD

Completes entire 1. Completes medfcaf1. Receives applfcatfcm

Physician certUlcate in consuf.certtflcation and

Certifies to the came Reports each ca8e by for marriage ftcense.

or Other tation with parent(s).signs certificate.

of fetaf death and telephone or by maff

ProfessionalFifes certificate with SISI16certificate. Re - on swclal form tolocal registrar cd

2. Verifies information

Attendant 2. Returns certificateturns it to funeral di- local or State health

district in whichfrom aerologlcaf

to funeraf director.rector. department.

birth occurs. Clerk of LCWaftests.

Government 3. Issues marriage 11-

1

cenae.

1. Obtains personaf 1. 0k4afns facts almutfacts about the de- the fetid death. 4. Sends completed rec-

ceased. ord of marriage to2. Takes certificate to State registrar.

2. Takes certificate toFuneral

physician for entrypbyslcian for medf - of causea of fetid

Director caf certification. death.1. Perf0rm8 the mar-

riage ceremony.

3. Delivers completed 3. Deffvers completedMarrfage

certificate to locaf Officimt 3. Certffies to facts ofcertificate to local

registrar and obtainsmarriage and sends

registrar and obtainsburial permit.

the recurd to Ucenseburlaf permit. clerk.‘ 1

1. Verifies cOmpiete- 1. Verifies cOmplete-1. Provides for

neas and accuracy. ness and accuracy.

- 1

report to pet

Makes coPy, ledger fdakes copy, ledgeror attorney, c

entry, or index for entry, or index forpetftion for d<

Verifie8 complete, Iocaf use. Sends Iocaf use. Sendsmake entries

nem and accuracy. certificates to localform.

Makes copy, ledgercertificates to local

Lc+al Registrar health department or heafth department or Clerk

ofentry, or index for to State registrar. to State registrar.

3. Verffies entrof

locaf use. Sendsreturned fern

vital Sfatfstfcs certificates to localcourt

3. Issues burlaf permit 2. 1ssue8 burial permithealth department or to ftmeraf director.

3. Enters finalto funeral dfrector,

to State registrar.fact%

3. Verlfie* returns of 3. Verifies returns ofburiaf permits. bwial permits.

4. Sends compleport to State

+ —J---+ +

trar.

1. Enter8 prmn

city or cOuntJ 1. Uses certificates in alfocatfng medical and nursing services, follow up of fnfectioua diseases,

Heafth

relative to sw

plannlng programs, and meammfng effectiveness of activities. Attorney for

L3cprtment 2. Forwards certificates and case repofis to State registrar. PeUtioner 3. Returns formof court.

+

1. Qeries incomplete or Inconsistent information. w-State Health 2. Maintafns fifes for permanmt reference and source of certffied copies.Dqutment 3. Compiles statistics for fite and civil divisions of State for use of the health department and other interestedBureau of agenciem or groups.

vital Statfatfcs 4. Prepsres transcripts or micro[ilm copies of birth, death, and fetal death certificates, and summary reportsof marriage, dfvorce, and notifiable disease records for transmissim to Naticiml Office of VIM Statistics.

Publkc Health+

service 1. Prepares and pubffshes Iuttonal statistics of bids, deafbs, fetal de~ths, marriages, divorces, and mdffiable

Naticaml Office diswes for officfal and voluntary consumers.C4 2. Publisbes znafyses of data as they refate to PLMIC bedtb and sc-cial problems.

vim Statistics 3. Provides .mwices needed to fcdter more cmmplcte and uniform registrattcm.

XGZZ.-IB _ S-t,’ certifimt,, of blti, dutl,,u+ fetal d,mth d rqmrt. of mtiflabl, dl,use, mmmt routed thmu&b led hcaltb .ie~t,; in ah,rm, there 1. IJOcemtrm.1 fle for ~m.se 4 .UWXWt m? state level.

DETAILED REPORT

Scope and Method of Study

The U. S. National Committee on Vital andHealth Statistics was asked by the U. S. PublicHealth Service to make a critical study of the ob-jectives and the program of national vital statis-tics, and to report its findings, and to make rec -ommendations on the following topics:

1. The effectiveness of the National Office ofVital Statistics in meeting the major consumerneeds in all fields of use, with respect to timeli-ness, degree of accuracy, periodicity, et cetera.

2. Types of data and analytical studie 3 to beproduced routinely and as special studies, withSuggestions on proportion of total budget that shouldbe allocated to special studies.

3. Program of technical assistance to States.4. Coordination of vital statistics activities at

various levels of government-Federal, State, andlocaI, and interrelation between these levels.

5. General blueprint outlining goals to beachieved by the national vital statistics system.

In addition to making various studies of dif-ferent aspects of the program and operations of theNOVS, an inquiry was conducted among 369 princi-pal users of national vital statistics, by means ofa questionnaire designed to discover what types ofdata published by the NOVS are most frequentlyused and what deficiencies, if any, the users havefound in these data (see Appendix A). Informationalong the same lines was ‘also sought from businessnews services.

At the request of the special Subcommittee onVital Statistics Needs, the American Associationof Registration Executives held a panel discussionon’ ‘A National System of Vital Statistics” at itsOmaha meeting, June 14, 1955. Three members ofthe Subcommittee participated and obtained a help-ful sounding of the views of registration officials.

The National Committee also undertook a spe-cial study of the status of national marriage sta-tistics with emphasis on (a) the uses of marriagedata by business, government, research, and otherinterests, and (b) the extent to which existing sta-tistics on marriages meet consumer needs.

A questionnaire survey of 539 organizationsand individuals to determine present uses of sndneeds for marriage statistics was conducted by thespecial Subcommittee on Utilization of MarriageStatistics. The survey findings were presented atmeetings of the Working Group on Marriage andDivorce and of the Council of the Public HealthConference on Records and Statistics.

Although it was recognized that a similar studyon statistics of divorces and annulments is needed,the investigation of this subject was deferred. TheCommittee suggests that a survey of consumerneeds for divorce and annulment statistics be con-ducted.

The Present National Vital Statistics System

This section presents a brief account of cer -tain aspects of the present national vital statisticssystem, as a background for the discussion leadingto the findings of the Committee.

227

228 VITAL STATISTICS-SPECIAL REPORTS

Legislative authority

The method of producing nationaI vital sta-tistics in the United States differs from those of alarge majority of other coontries. Registrationof vital events in the United States has been’ estab-lished as part of the public health responsibilitiesthat are included within the general powers re -served to the States. The legal authority for theproduction of national vital statistics is found in alaw of Congress (1904) which requires the U. S.Bureau of the Census to collect “statistics of thebirths and deaths in registration areas annually,the data for which shall be obtained from and re-stricted to such registration records of such Statesand municipalities as in the discretion of the Di-rector possess records affording satisfactory datain necessary detail . . . .“ This authori~ VZISt.HIS -

ferred to the Federal Security Agency, now theDepartment of Health, Education, and Welfare, in1946. The language of the reorganization plan thatmade this transfer, in referring to vital statistics,included marriages, divorces, and annulments alongwith births and deaths. However, there is no sps -cif ic statutory authority to pay for marriage, di-vorce, and annulment records. The Surgeon Gen-eral of the Public Health Service has statutoryauthority to collect morbidity and mortality sta-tistics. All of the present activities of the NOVSderive from these authorizations.

Under this system, a major problem faced bythe NOVS is to develop procedures that will pro-duce adequate national vital statistics in a situationwhere the legal authority over the registration orcollection of the original information rests with theStates.

Functions of the Natioml Office ofVital Statistics

Collection of data. —This office tabulates andpublishes data on births, deaths (including fetaldeaths), notifiable diseases, marriages ~d di -vorces (including annulments). Figure 1 taken fromVolume I of Vital Statistics of the United States,~ shows the flow for each of these items fromthe original record to its processing in the NOVS.

Copies of individual records of births, deaths,and fetal deaths filed in the State registration of-fices are received by the NOVS in the form ofmicrofilm reproductions of the record or as tran-scriptions. In addition to the microfilm copies,

punched cards for births were received in 1954 fromi 1 reporting areas, and punched cards for deathswere obtained on a trial basis from California andOregon. The numbers of records collected for 1954from the State vital statistics offices in the variousforms are as follows:

PunchedMicro- cards and Tran-

T&A f~m microfilm scripts

Total --5,717,854 4,274,795 1,028,036 415,023

Births --4,119,313 2,923,551 904,043 291,719Deaths-- 1,502,364 1,262,853 123,993 115,518Fetal

deaths- 96,177 88,391 7,786

Data on marriages and divorces and annulmentsare provided by the registration areas on tableforms prepared by the NOVS. The number of Statesand other areas providing tables for 1954 datavaried for marriages from 37 areas giving data bymonth to 19 areas providing. data by ages of brideand groom and number of previous marriages. Fordivorces, the number of areas varied from 38 givingdata by county to 7 providing statistics by ages ofhusband and wife.

Publications. —The publications issued by theNOVS may be classified into three types: current,annual, and special reports.

The Morbidity and Mortality Weekly Repor~and an accompanying press release CommunicableDisease Summa ry, are issued at the end of eachweek following the week to which the data refer.The Monthly Vital Statistics Report, issued about 6weeks after the end of the month to which the datarefer, gives summary data for each State and theentire country on total births, deaths, infant deaths,and marriages. Divorce and annulment data areincluded for slightly more than one-half of theStates. In addition, mortality statistics for thecountry are classified by cause of death, age, sex,and color, for a 10-percent sample of the deathcertificates. All data in these publications areprovisional.

The regular, comprehensive compilation ofvital statistics for this country is the publication,Vital Statistics of the United States, which is issuedannually by the NOVS. It contains the most com -prehensive and detailed vital statistics for theUnited States and individual States and local areas.

Vol. 45, No. 11 NATIONAL VITAL STATISTICS NEEDS 229

It is the basic statistical reference source for re-search workers concerned with the study of mor -tality, natality, fetal mortality, marriage, and di-vorce and annulment.

Sp?cial reports on particular subjects are pub-lished in Vital Statistics —Special Reports. Onevolume of this series comprises 20 or so separatereporti issued for the year, and subtitled NationalSummaries. Typical titles in such a volume are:“Births by Age of Mother, Race, and Live-BirthOrder: United States, 1954”; “Divorces and An-nulments: Detailed Statistics for Reporting Areas,1954’ ‘; “Infant Mortality: United States, 1954.”While these reports are devoted primarily to datafor the most recent year, they frequently containsummary tables with rates or other index figuresfor earlier years.

Another volume of the Vital Statistics –Special-s series is subtitled Selected Studies. Itis devoted entirely to special analytical studies ofvital statistics subjects in which detailed analysisand interpretation of the statistics are emphasizedas much as the presentation of the data. An exampleof this type is “Weight at Birth and Its Effect onSurvival of the Newborn in the United States, Early1950,” Vol. 39, No. 1, 1954,

Occasionally, an entire volume of the V~Statistics-Special Re~rts is devoted to one majorsubject, An example is Volume 32, “Cancer Mor-tality in the United States, 1900 -1945.”

Technical services. —The NOVS has two con-tinuing programs of technical service to the Statevital statistics offices (see Appendix D). One pro -gram includes the training of cause-of-death codersthrough quarterly comparisons of State and NOVScause -of -death codes; monthly publication of acolumn on cause-of-death coding problems in ‘llQRegistrar, a NOVS publication for State and localregistrars; and a monthly issue of 10 coding prob-lems and answers for cause -of -death coders. Theother program pertains to serving the Public HealthConference on Records and Statistics and the work-ing groups of the Conference in a secretariat andconsulting capacity.

Other technical services are rendered uponrequest. These include ‘‘ioan” of personnel tointernational and State organizations, as well asconsultation on various problems in registration,statistical processing, machine tabulation pro-cedures, and analytical methods. Requests fortechnical opinions are generally handled by corre -spondence,

Coordination and standards. -In order to pro-duce uniform national vital statistics, the NOVSmust coordinate registration and statistical prac -tices of the States insofar as they affect interstateand international comparability of data. The prin-cipal mechanism utilized by the NOVS for coordi-nation is the Public Health Conference on Recordsand Statistics. In collaboration with the PublicHealth Conference on Records and Statistics, theNOVS develops, for adoption by States, recommen-dations on basic items such as the model vital sta-tistics law, and standard certificates of live birth,fetal death, death, marriage, and divorce. The NOVSparticipates in the development of uniform defini-tions on other specific aspects such as motor-ve-hicle accidents and home accidents, for nationaluse. In the international field, the NOVS contributesto the formulation of international standards andassumes the responsibility for interpretation ofthese standards and practices for national use. TheNOVS participation in the development of the Inter-national Statistical Classification of Diseases, In-juries, and Causes of Death, and of the internationalrecommendations of definitions of live birth andfetal death are examples of its activities in theinternatioml field.

Status of national marriage statistics. —In re-cent years many efforts have been made to improvenational marriage data. Transcripts of marriagerecords for the years 1939 and 1940 were purchasedfrom the State vital statistics offices in those Stateswhich maintained central files of marriage records.Some of the data obtained from these transcriptswere published in reports showing data on mar-riages by various characteristics of brides and ofgrooms. This program was discontinued in theearly years of the war.

Beginning with data for the year 1948, a co-operative program was inaugurated under whichcertain States provide the NOVS with completedtables on marriages classified by age, previousmarital status, race, and other characteristics ofindividuals at the time of marriage. There hasbeen considerable expansion in this program, butthe coverage remains far from complete. ManyStates lacking central files of marriage recordsprovide no detailed data on marriages. The Statesthat cooperate in the program complete varyingnumbers of tables so that coverage is not uniformeither within one year or from year to year, Onereason for this variation is the lack of uniformityin State marriage records. This is being changed.

230 VITAL STATISTICS—SPECIAL REPORTS

In cooperation with the Public Health Conferenceon Records and Statistics (PHCRS), an organizationof State statistical and registration officials, theNOVS has developed a Standard Record of Mar-riage. To further encourage uniform tabulations,the NOVS and the Public Health Conference havedrafted a Manual for Registration of Marriages toserve as a guide to the States on definitions ofterms, registration procedures, and coding, punch-ing, and tabulation’ of data. It is expected that aMarriage Registration Area will be established topromote improvement in marriage statistics andstimulate all States to provide for central Stateregistration of marriages. This would follow thepattern of the Birth and Death Registration Areas.

To obtain information on important sociM andeconomic characteristics of recently married per-sons for the United States, the facilities of theCurrent Population Survey (CPS) of the Bureauof the Census have been used. In the first of thesesurveys (April 1953), data were obtained for theUnited States on demographic and economic char-acteristics of persons who married between 1950and the survey date.1 The second survey (June1954) provided data on education, occupation, num -ber of marriages, and residence status of personsmarrying between 1947 and the survey date.

Status of national statistics on divorces andannulments .—The development of national statisticson divorces and annulments has proceeded alonglines similar to marriage statistics, Transcriptsof divorce records for the years 1939 and 1940were obtained from the State vital statistics officesin those States which maintained central files ofdivorce records, and a repok”t was issued.

Beginning with the data year 1948, a cooper-ative program was started where certain Statesprovide the NOVS with completed tables on divorcesclassified according to certain characteristics.While there has been expansion in this program,the coverage is even less complete than for mar-riages, The States that participate in the programprepare varying numbers of tables so that data arenot entirely comparable from year to year, or fromState to State. One reason for this is the lack ofuniformity in the State records of divorces and

lN~ti~n~l Office of Vital statistics, Vital Sta-

tietics —Special Reports, Vol. 39, No, 3, ‘ ‘Demo-graphic Characteristics of Recently Married Persons:United States, April 1953, ” 1954; and No. 5, “Eco-nomic Characteristics of Recently Married Persons:United States, April 1953,1 ~ 1955.

amulments. The recent development of the Stand-ard Record of Divorce or Annulment will promotemore uniform tabulations. Also, the Manual forRegistration of Divorces which is being draftedwill contribute further to uniformity in tabulations.It is anticipated that a Divorce and AnnulmentRegistration Area will stimulate States to providefor central registration.

Organization, budget, and personnel of theNational Office of Vital Statistics

The NOVS is organized in several large serv-ice sections and in professional sections establishedon subject-matter lines. About three-fourths ofthe approximately 200 employees are in the Statis-tical Processing Section which is responsible forthe operations involving receipt, checking, andprocessing of data from the States, including coding,punching, tabulating, computing, table posting, edit-ing, and typing for reproduction. There is an ad-ministration section with about 16 employees whichhandles psrsonnel,budget, data procurement, travel,supplies and equipment, correspondence files, mes -senger service, and routine inquiries for infor -mation. The remaining sections are staffed withfrom 2 to 6 persons, chiefly of professional grade,who are responsible for planning statistical tabu-lations and published reports, conducting analysesof data, and planning and conducting special studies.These sections are: Mortality, Natality, Marriageand Divorce, Actuarial Analysis, Morbidity, andState Consulting Service. overall technical mdadministrative coordination and relations with thePublic Health Conference on Records and Statisticsare the responsibilities of the Office of the Chief,staffed by 8 persons.

The total annual budget for the fiscal year 1955has been about $1,300,000. Of this sum, about$740,000 is used for personnel salaries; $200,000for purchase of data; $160,000 for communicationcosts, chief ly penalty mai~ $70,000 for equipmentrental; $60,000 for printing; $35,000 for travel; andthe remainder for miscellaneous items.

Findings and Recommendations

In its directive, this Committee was asked toprepare findings on the five distinct topics men-tioned in the section on Scope and Method of Study(p. 22’7). As a matter of presentation, it was de-cided to use topic 5 -general blueprint outlininggoals to be achieved by the national vital statistics

NATIONAL VITAL STATISTICS NEEDS 231

system -as a frame into which to fit the discussionof the other topics.

In the detailed consideration of these goals,the Committee recognizes four major types of prob-lems which are pertinent to work of the NOVS:

1. Production of national vital statistics.2. Statistical studies of public health and popu-

lation characteristics (including the developmentof new data),

3. Coordimtion of the national vital statisticssystem.

4. Economy and efficiency of operations.

Subsequently in this report, each type of prob-lem is examined in turn. Under each, the Commit-tee presents a statement of the goals to be achievedas it sees them, a statement of the principal defi-ciencies in the present system with respect tothese goals, and its recommendations.

Two other items were discussed. These werethe level of the NOVSin the organizational hierarchyof the Public Health Service and the channel ofcommunications to the State vital statistics offices.While it was agreed that these were problems thathave an important bearing on the effectiveness ofoperation of the NOVS, they were considered to beoutside of the scope of the Committee’s assignment.

Regular production of basic nationalvital statistics

The responsibility for producing national vitalstatistics carries an obligation to make them ofmaximum value for administrative, public health,research, and other pu%poses which they shouldserve. In the postwar years the importance ofthe need for detailed national vital statistics forresearch in the social sciences and in the medi-cal and allied sciences has increased greatly. Forexample, detailed birth, death, and marriage sta -tistics are basic for the study of determinants ofpopulation trends and for making population pro -jections, Government at its several levels is con-cerned with population projections in planning forschools, for public health services, for housingdevelopment, for power development, for the na-tional defense, for agricultural needs, and in ad-ministering its social security program. Populationpro jections are used by business for gauging poten-tial markets and determining the locations fQr

plants. With the greatly intensified efforts by bothprivate and public agencies to promote health andreduce mortality, the medical and allied scienceshave increased their demands for detailed nationalvital statistics in order to detect problems requiringstudy, to design. health studies or programs, toevaluate the results of such studies or programs,and also to study the relation of disease to environ-ment.

In the production of national vital statisticsthe responsible agency should make every effortto publish statistics which

1. Cover the entire country.2. Are based on as complete reports as it is

feasible to obtain, with measurement of the degreesof completeness achieved.

3. Are based on as accurate reports as it isfeasible to obtain, with measurement of the degreeof accuracy achieved.

4. Are based on records which are essentiallyuniform with respect to all items of statistical in-formation needed for natioml vital statistics.

5. Are compiled according to uniform classi-fications and classification rules.

6. Give sufficient detail, in terms of cross-classification of items, to serve adequately not onlyadministrative needs, but also research interests:

7. Are available for public use at the earliestpracticable time.

8. Are accompanied by adequate explamtionsof factors bearing upon the interpretation of thestatistics.

The current status of national vital statisticsdata with respect to these eight objectives is asfollows:

Geographic coverage. —The U.S. Death Regis -tration Area was established in 1900 and the U. S.Birth Registration Area in 1915, in each case with10 States and the District of Columbia. AdditionalStates were added year by year so that nationwidecoverage was achieved in 1933. The situation withrespect to marriage and divorce statistics is simi-lar to that for birth and death statistics prior tothe establishment of the registration areas, exceptfor the publication by the NOVS of data tables show-ing marriages performed or divorce or annulmentdecrees granted which are based on tabulationssupplied by some States.

VIT

AL

ST

AT

IST

ICS

—S

PE

CIA

LR

EP

OR

TS

Vol. 45. No. 11 NATIONAL VITAL STATIS’HCS NEEDS 233

In order to extend coverage there is need for

the organization of registration areas for mar-

riage and divorce at an early date. One necessarystep in this process is the centralization of mar-riage and divorce records in each State. Thereare now 30 States, 4 Territories and one indepen-dent reporting city with central registration of bothdivorces and marriages. In addition, there are 6States with central registration of marriages butnot divorces. (See figure 2.) It would be parti-cularly desirable for the latter group of States toestablish central records of divorces and annul-ments in the near future because of the close inter-relationship between the two subjects, Eventually,all States should have central files of both mar-riage and divorce records.

RECOMMENDATION. The NOVS should in-crease its efforts to obtain nationwide mar-riage and divorce statistics through the es-

tablishment of registration areas based oncentral files of marriage and divorce rec -orals which contain specified items.

Completeness of data, —Excellent progress has

been made toward the goal of complete birth regis -tration for the country as a whole. The 1950 testshowed that the problem of unrecorded births is

now limited to a few States and to certain ruralareas and population groups in those States.

RECOMMENDATION The NOVS should stim-ulate and assist in special studies of birth reg-istration completeness in areas where com-pleteness of registration is still a problem.

In death and fetal death registrations, thereare indications that underregistration, particularly

of infant deaths, is significantly high in ,certain

rural areas, as in the case of birth registration.For fetal deaths, a study made in New York Cityzindicates that reporting is seriously incomplete.Even for late fetal deaths (28 weeks or more) re-porting may be more than 20 percent incompletefor the country as a whole.

It is recognized that there are technical diffi-

culties involved in measuring completeness of deathregistration. The problems are even more diffi-cult in testing completeness of fetal death regis -tration. Nevertheless, the increased attention beinggiven to the reduction of fetal and early neonatalmortality emphasizes the need for improvement ofdeath snd late fetal death registration. There seems

to be little possibility of obtaining complete regis -tration of early fetal deaths.

RECOMMENDATION: The NOVS should en-deavor to develop methods for measuring in-

completeness of death and late fetal deathregistration adaptable to those areas wherethis problem is important. If practical tech-niques can be found, they should be appliedto obtain more information on the extent ofincompleteness and on means for attainingmore complete reporting.

In notifiable disease reporting, most, if not.aIl, notifiable diseases are incompletely reported.In many instances, the administration of therapeu-

tic measures before a definitive diagnosis is mademakes it very difficult to obtain complete reportingof illness due to certain diseases. k other in-stances, failure to report arises from oversight on

the part of the physician or from lack of interestor appreciation of the need for reporting.

The purpose of notifiable disease reporting isto provide data that will aid in the control of com-municable diseases. Some of these diseases havebecome of minor significance as public health prob-lems in this country. The importance of havingcomplete da’a is greatest for diseases such as PO-liomyelitis, typhoid fever, and tuberculosis, whichare still major public health problems amenable to

control procedures, or for potentially dangerousepidemic diseases such as smallpox.

RECOMMENDATION. Efforts toward im-provement of notifiable disease reporting

should be concentrated first on those (i@-eases that are major health problems andare amenable to control procedures.

The degree of completeness of marriage ~dZB~U~~*~t~~~,Leona, Wallace, Helen M., Landsberg, divorce registration is unknown. The NOVS has

Eva, and Pessin”, Vivian, , ,The Inadequacy of Routine

Reporting of Fetal Deaths.’ ‘ American Journal of

prepared p“bns to cooperate with the States in

Public Health, Vol. 39, No. 12, December 1949? checking the completeness with. which certificates

PD. 1549-1552. of marriage and divorce are actually filed. The

234 VfTAL STATISTICS-SPECIAL REPORTS

successful execution of these plans will constitutean important step toward the establishment of asound base for national statistics on marriagesand divorces.

RECOMMENDATION The NOVS should co-operate with the States in the development

and application of tests on the completeness ofthe reporting of marriages and divorces tothe central off ices.

Accuracy of data .—Accuracy of the informationon the records, and of the statistics based on it, isof crucial importance. It is also difficult to mess-

ure. There is evidence of significant errors in thereporting of certain items—cause of death, placeof residence, and age. For other items such asrace, nativity, and marital status of the deceased(or of the parents in the case of births), there isvirtually no knowledge of the quality of reporting.Several general methods of evaluating statements

on vital records may be used. One is to comparethe entries on the certificates with the correspond-ing information for the same persons on anotherrecord, such as the population census schedule.

For example, a comparison of the occupation in-

formation on birth and death certificates with thaton the census schedule is now being done by theScripps Foundation for Research in Population

Problems. The correspondence of other items

should be measured. Another method is to reinter-view the informant who gave the information on thecertificate, or to interview another person who maybe able to make a morti accurate report than thefirst informant. This method was used with at leastpartial success in the Census Post-EnumerationSurvey.

RECOMMENDATION: Studies designed tomeasure the accuracy of the information onthe original records should be an essentialpart of the regular NOVS program.

Uniformity of data. —It is essential that all

State certificates contain the items needed for na-tioml statistics and that these items be defined insubstantially the same form. In order to accomplishuniformity in the report forms, the NOVS has issuedstandard certificates of birth, death, and fetal death.More recently, standard records of marriage andof divorce or annulment have been recommendedby the NOVS and adopted by several of the States.For national statistics, the information must also

be processed according to uniform classificationswith uniform rules for classifying the specific factsgiven on each record.

The NOVS and the States are to be congratu-lated on the impressive degree of uniformity whichhas been achieved for most of the items used innational birth, death, and fetal death statistics. Thefew exceptions occur mainly on the birth certificate,where the legitimacy item is omitted by 14 States,chiefly owing to prohibition by State laws, and birthorder and birth weight are omitted from the Massa-chusetts certificate

Because the standard records of marriage

and divorce have been available only recently, con-

siderable diversity exists among the record forms

in use. Uniformity in coverage and definition ofitems is essential for production of satisfactorynational statistics.

RECOMMENDATION The NOVS and the Statevital statistics offices should continue theirefforts to obtain compiete coverage of allitems on the standard certificates.

Uniformity of classification. —Uniform classi-fication of the information on the birth, death, and

fetal death certificates has been achieved for theentire country by having copies of all of these vitalrecords available in one central office and classified

according to a uniform set of instructions and pro-cedures. This has meant that the data for all partsof the country have been coded in the same way.

The classification procedures used by the

various States in compiling tables for consoli-dation in national statistics of marriages and divor-

ces are not uniform. Greater uniformity could beobtained if the NOVS and the participating Statesdevelop and adopt uniform definitions and clae.si-fication procedures. However, it is much moredifficult to maintain uniform practices when the

operations are performed in many separate offices,The value of national vital statistics depends

greatly upon uniformity and consistency in theclassification of the various items. This is parti.

cularly true for interstate comparisons of data.

RECOMMENDATION Uniformity of classi-fication of the items is essential in the pro-duction of natioml vital statistics, and con-tinuous efforts should be made to maintainand increase the degree of uniformity. Spe-cial attention should be given to this prob-

lem in statistics of marriages and divorces.

Vol. 45, No, 11 NATIONAL VITAL

Scope and detail. —There are certain generalprinciples which, in the opinion of the Committee,should be followed in determining the detail andscope of tabulations. These are:

a. Vital statistics should be available for the

b.

c.

d.

United States, each State, each county, andeach city above a specified population size

(e. g., 10,000 or 25,000).Statistics for the country as a whole shouldbe in the maximum detail, with less detailfor the individual States and large citiesand least detail for counties and small cities.This gradation is justified by three con-siderations: National data are subject to themost extensive use; the higher frequenciesat the national level than for the geographicsubdivisions allow more detailed classifi-cation for the former; and in intercensalyears annual population data, which arenecessary for computing vital statisticsrates, are available in detail only for the

country as a whole.For all geographic areas, the most detailedtabulations requiring a population base foranalysis should be made for the decennial

census year (or 2-or 3-year period center-ing on the census) when the maximum amountof population data is available.Certain statistics should be tabulated an-nually; for others the needs may be satis-fied by less frequent tabulations (see Cycli-cal schedule of tabulations, p. 239). Forintercensal years, there is now some evi-dence that a 50 -percent sample of the birthsmay be adequate for most purposes. Whereless -than-annual tabulations are scheduled,

they should be given the same priori~ asthose scheduled annually, in order to avoideventual loss of information.

Xn addition to these general principles appli-cable to all vital statistics, there are specific sug-gestions for much needed improvement in the scopeand detail of national marriage statistics. For cin-stance, annual data on marriages showing the char-acterktics of brides related to corresponding char-acteristics of grooms should be made available.‘This is especially important for information on age,race, previous marital status, and number of timespreviously married. In addition, data on occupa-tion, attained years of schooling, and number ofchildren under 16 years of age from prior mar-riages should be available periodically. It is recog-

STATISTICS NEEDS 235

nized that some of this information would not beavailable from the Standard Record of Marriage.

Information on marriages by place of resi-dence of bride and groom, as contrasted with placeof occurrence of the marriage, is also important,especially if linked to the characteristics of thepersons who are married in States other than theone of usual residence. The possibility of tabulat-ing marriages by place of intended residence after”marriage should be explored because of the obviousvalue of such data to-numerous groups, includingmarket research organizations.

In the past, the NOVS has ‘relied on the fol-lowing guides in making decisions regarding thescope and detail of its tabulations and publications:advice of professional leaders in the various sub-ject fields, interests and needs of administrativeand research programs, specific requests for datafrom importsnt organizations, and occasional mailsurveys of representative consumers of the statis-tics. There have also been instances where propo-sals for the discontinuation of a publication serieswere submitted to the consumers on the mailinglists for the publications, snd action taken on thebasis of the responses received.

RECOMMENDATION. The NOVS should keepin closer touch with users of its data whenplanning the scope and detail of its tabulationsand publications and making decisions on fun-damental issues.

Timeliness .—The weekly and monthly publi-cations of the NOVS giving current vital statisticsare being issued on a satisfactory schedule, andpresent no special problems in regard to timeliness.

For some types of data, useful particularly for

research studies or in long time series, the e~s -

tence of an annual repetitive series is more im-portant than the time at which the figures becomeavailable. On the other hand, the effective appli-cation of data pertinent to public health snd admin-istrative problems depends on the promptness withwhich the figures reach the consumers.

If the records are received within the speci-fied time from the States, the special reports and theamuai volumes (Vital Statistics of the United States)of the National Office of Vital Statistics can be re-leased for printing from 12 to 18 months after theend of the calendar year to which the data apply.Starting with the data year 1949, however, increasingdelays occurred in the issuance of the annual pub-lications. The 1949 annual volumes were released

236 VITAL STATISTICS—SPECIAL REPORTS

to the printer 25 months after the close of 1949,whereas the usual schedule prior to that period wasabout 18 months after the year. Thereafter, the

situation became worse. The poorest record wasset with the 1951 a.mmal volumes which were re -leased for printing 32 months after the close of theyear. Three factors contributed to the delays inthe annual processing schedules, namely: unchangedbudget appropriations during a period of increasingpersonnel and material costs, the diversion of per-sonnel to conduct several special census-related

projects, and the increase in the number of recordsto be processed.

After the data year 1952, a special effort wasmade to catch up to the normal processing schedule.The release for printing of death statistics in theannual volume (Volume II) for 1953 was on normalschedule (18 months) but the release of birth data

(Volume I) occurred 23 months after the end of thedata year. For 1954 and thereafter, it is expectedthat the annual volume on mortality will be releasedto the printer 15 months after the close of the data

year, and on births, 18 months. The projected

processing schedule, if maintained, will be as

PromPt as it has ever been. The principal prob-lem now in maintaining operations on scheduleappears .to be in the procurement of data from theState vital statistics offices according to the spec -ified time schedule.

The contract schedule for delivery of tran-scripts of birth and death records from the Statevital statistics off ices to the NOVS calls for de-livery of data for each month within 90 days afterthe end of the month. For example, data for De-cember are due on March 31 of the year following.

Several reporting areas usually fall behind in thedelivery schedule. The following table shows thedelays in delivery of transcripts as of December

1954:

One.ched-ule

Births:Reporting axeas 40

Deaths:Reporting areas 41

Fetal deaths:Reporting areas 36

MONTHS EEHINO DELIVERYSCHEDUIE

—

7—

1

—

The schedule for obtaining statistical tablesgiving 1954 data on marriages and divorces (includ-ing annulment@from the State office was as follows:Table forms were mailed by the NOVS early in Feb-

ruary 1955 with a requested return date of June 15;and followup letters were sent to the States in Julywith a suggested deadline of November 1. Actually,

the 1954 tables were received from some States as

early as April 1955 and as late as January 1956.An effort will be made by the NOVS to obtain 1955data for all participating States by October 1,1956.If this goal is met, it should be possible to publish

consolidated reports early in 1957.The delay in the issuance of NOVS publications

was the most frequent criticism of consumers re -spending in the questionnaire survey. Most of the

consumers indicated considerable understandingand appreciation of the difficulties faced by the NOVSduring the period of maximum delay. Nevertheless,it was felt that the data would be more useful ifmade available more currently.

RECOMMENDATION Immediate priority

should be given to the processing and pro-curement of data to achieve a schedule ofprocessing which permits the release forprinting of final data within 15 months after

the close of the data year; and this scheduleshould be maintained or improved upon in thefuture.

Explamtion of factors bearing on interpr e-tation of data . —Statistics may be relatively com-plete, accurate, comparable, detailed, and timelybut still fall short of their potential value, It is

also necessary to explain the various factors whichmay affect the meaning of the figures so that thestatistics can be correctly used and interpreted,Population coverage, definitions, classifications,rules for allocating individual events to classes,known biases in reporting, changes in reporting

practices or rules which may affect comparability,

and any other sources of error or misunderstand-ing need to be fully discussed.

In addition, there are many professional and

lay users of vital statistics who have limited knowl-edge of the ways in which statistics can be cor-

rectly used. For this large group, it is importantthat the tables be accompanied by a clearly writtentext describing the trends or difference indicatedby the statistics and, where possible, pointing out

Vol. 45, No, 11 NATIONAL VITAL STATISTICS NEEDS 237

their implications in terms of the relevant subject,that is, health, population changes, social problems,economic considerations, et cetera. Such explan -ations should be provided regularly by the NOVS.

RECOMMENDATION The analytical func-tions of the NOVS should be strengthened tomake available more explanatory and inter -pretive material useful to the consumer.

Statistical studies of tmblic health andcopulation characteristics

The annual repetitive series of publicationsmake available a considerable body of statistics.However, not all of the data tabulated can be orneed to be published regularly. Also, there arecertain valuable items of information on the vitalrecords that have not been fully exploited for sta-tistical purposes. Examples are” the occupationitems on the birth, death, and fetal death certifi-cates; the hospital and medical items on the birthand fetal death certificates; and the associatedcause information on the death certificate. Bothuntabulated data and unpublished tabulations shouldbe exploited periodically, and the results madeavailable in the form of special studies.

Two kinds of studies maybe distinguished. Thefirst type of study is a program-oriented analysiswherein data relating to a specific program arebrought together. The series of reports on cancermortality and on cardiovascular-renal disease mor -tality, prepared in collaboration with the NationalCancer Institute and with the Division of PublicHealth Methods, respectively, are emtmples ofprogram-oriented studies.

The second type relates to the presentationand analysis of data with primary emphasis onmaking available new information which may servea variety of interests, Examples are studies ofmortality by marital status and by occupation andindustry; neonatal mortality by cause, race, andsex; the fertility of birth cohorts of women by ageand number of children borne; births according tobirth weight; and special studies based on matchedmarriage and divorce or amulment snd other vitalrecords.

The development of a program of special stud-ies should be an important goal of any vital statis-tics office. The vital statistics offices are in aunique position to make impmtant contributions toknowledge regarding health and population growth,

and to the furtherance of the various public healthprograms. The program-oriented studies in parti-cular represent valuable services to public healthactivities. Such studies should also have impor-tant side effects in pointing to needed improvementsin the data, to new areas for study, and to regulartabulations which should either be eliminated orreshaped.

The National Office of Vital Statistics has con-ducted a number of such studies in the past but noton a regular basis.

A program of more complete interpretation ofvital phenomena will require from time to time thedevelopment of new data not obtainable from thepresent records. The decennial revisions of thestandard certificates by the NOVS provide an op-portunity to introduce new items on the vital rec -orals, although this approach can be taken only ona limited number of items that are needed for atleast a 10-year period. A more flexible methodis to follow back on a sample of vital records foradditional information. By use of this follow-backprocedure, it may be possible to obtain additionalmedical information from hospital records andlorfrom the physician. The supplemental data mayrelate to illness history, verification of diagnoses,and other facts about medical care and hospitali-zation. Certain other pertinent social facts, suchas economic status, medical costs, education, num-ber of dependents, et cetera, may be secured fromthe family. Although much will have to be learnedabout the practical application of this approach, itappears to us potentially very useful.

Surveys of the general population protideanother source of new data. The National Office ofVital Statistics has already utilized the CurrentPopulation Survey for obtaining supplemental dataon child spacing as related to certain socioeconomicfactors, and on the various characteristics of new-ly married persons.

Surveys of the general population possess thegreat advantage of providing the population baseat the same time that new data, which should berelated to this base, are being collected. On theother hand, the sample size may not be adequatefor measurement of the characteristics under studyand there may be problems of underreporting. Ifa population base is not needed, the use of a sam-ple of existing vital records may provide a moreefficient means than a survey for collecting addi-tional data.

238 VITAL STATISTICS—SPECWL REPORTS

Further research is needed in developing newmethods of analysis of data, including statisticalmeasur’es of mortality, fertility, and marriage anddivorce. For example, more intensive research isneeded in methods of projecting birth, death, andmarriage rates which are useful in planning healthand social programs and are essential to populationprojections.

RECOMMENDATION: Statistical studies thatcontribute to knowledge in public health, fer -tility, mortality, marriage, and divorce shouldbe developed as part of the regular programof the National Office of Vital Statistics. Inparticular, attention should be given to theinfluence of population characteristics, suchas sex, age, marital status, and occupation.The program should include collection of newdata and development of new methods of ana-lysis that are needed for program planning,administrative and research purposes.

Coordination of the national vitalstatistics system

Responsibilities of the NOVS.-The productionof national vital statistics in the United States bythe NOVS is dependent on the cooperation of theStates, since the collection of primary data bymeans of registration is a State responsibility.Neither the legislation enacted by the States norby tfie Federal Government defines the responsibi-lity of the States to the Federal Government in vitalstatistics matters since the provisions are permis-sive rather than mandatory.

Over the yezrs there has developed a fair de-gree of uniformity in the State laws on birth anddeath registration with respect to authorizing theState to furnish “copies of vital records to the Fed-eral Government for national statistics. This uni-formity stems from the fact that the States haveenacted into their laws the essential portions ofthe various ‘‘ uniform” and “model” vital statis -

tics acts. The State laws usually specify that theircertificates of birth and death are to contain as aminimum the items on the standard certificatesrecommended by the Federal Government.

The subjects of marriage and divorce regis -tration were not included in the “model” acts until1942. Furthermore, no standard certificates forthese events were issued by the Federal Government:mtil 1954. For these and other reasons, State lawson marriage and divorce registration vary greatly.

In view of these considerations, the nationalvital statistics system must be based on closecooperation between the Federal Government andthe States through mutual understanding of the na-tional and State interests and responsibilities forvital statistics. Enlightened self -interest is nec-essary on all sides and agreements must be en-forced by moral suasion.

The Committee stresses the essential impor-tance of minimizing the intrinsic complexity of thenational vital statistics system and its o~rationby developing a clearer understanding and appre-ciation by all concerned of the respective aims,functions, and responsibilities of the NOVS and theState vital statistics offices.

The production of national vital statistics re-quires an effective central office at the nationallevel. Therefore, it is essential that the necessaryauthority and responsibility of that office in rela-tion to the statistical services desired be clearlydefined. There is not now, in fact, an agreementor uniform understanding concerning the responsi-bilities of the national office.

The following functions are recommended asnecessary duties (in addition to its usual operatingfunctions) of the national vital statistics office, ifreasonably adequate national statistics and relatedservices are to be regularly availablrx

1. Determination of the scope and content ofthe mtional vital statistics tabulations and publi-cations.

2. Recommendation and promotion of uniformdefinitions of vital events and stsndard forms forreporting these events.

3. Determination of the classifications andclassification rules used in preparing national vi-tal statistics.

4. Assistance to the States in improving andmaintaining the quality and timeliness of the pri-mary vital records.

5. Measurement of the degree of accuracy ofall of the procedures which enter into the produc-tion of national vital statistics; completeness ofreporting, accuracy of reported information) ac-curacy and consistency of classification of infor-mation, and accuracy of statistical compilation.

6. Maintenance of direct regular contacts withnational groups representing the major interestsin vital statistics.

7. Maintenance of control over the accuracyof processing of records used in national vital sta-tistics tabulations.

Vol. 45, No. 11 NATIONAL VITAL STATE3TICS NEEDS 239

These are basic responsibilities of the nationaloff ice, regardless of whether all of the specific op-erations involved are performed by the nationaloffice personnel or whether some are done by Stateoff ice personnel. Certain operations have alwaysbeen and will continue to be done by the States-the securing of complete and accurate registrationof vital events, direction and instruction of localregistrars and other persons responsible for theorighal reports, and transmission of data to thenational office.

There are many other operations which areperformed by State vital statistics offices to serveState health department needs or to provide thelegal registration services for which they are re-sponsible. fnsofar as these operations affect orare r@.ated to the services necessary to producenational vital statistics, their proper performanceis of direct interest to the national office.

Technical services to States-As mentionedon page 229, the NOVS maintains programs of tech-nical services to the States. The services that aregiven and that might be given are discussed morefully in Appendix D. In summary, such programsare of great value to the national system of yitalstatistics in two ways. The system benefits di-rectly from improvements in the quality of datacollection and processing at State and local levels.Of equal importance is the more effective utiliza-tion of these data by State and local agencies in theplanning of health programs or in special studieswhich correlate vital statistics with other health,social, or demographic data.

RECOMMENDATION. The NOVS should begiven sufficient fiscal resources to providethe States upon request with more adequatetechnical services in the following areas:professional advisory and consultant serviceon technical and management problems; ex-panded information and clearinghouse serv-ices on technical subjects and methodolo~educational programs directed toward im-provement of source data; national programfor recruiting and training professional vitalstatistics personnel; and work with Statehealth depar~ment program divisions in pro-moting effective utilization of vital statisticsdata.

Observations on economy’ andefficiency of operations

Increasing efficiency of processing meth-@. —Fullachievement of the preceding objectives,within the usual limitations of available facilities,requires that the NOVS have an active programfor the development of more efficient and flexiblestatistical processing methods. Such a programshould include experimentation with the high speedelectronic computers, the new record-reading ma-chines, methods and applications of sampling, meth-ods of obtaining better information on the sourcedocuments, and more efficient coordination of rec -oral-collecting procedures at the local, State, andnational levels.

Cyclical schedule of tabulations. —f.nan annualvital statistics tabulation and publication program,consideration must be given to the various con-sumer needs and uses. Certain basic tabulationsare essential each year, whereas data of morespecialized interest can be made available peri-odically, perhaps every 3. to 5 years. There arealso data that change slowly over the years. Itwould seem unnecessary to tabulate and publishsuch data every year.

The usefulness of almost all vital statistics isenhanced to a great degree by the existence of cor-responding population data to permit the compu-tation of rates. The usefulness of some items de-pen& entirely on the possibility of computing rates,and for such items, tabulation is warranted only inthe years when population data are availabh c-can be estimated reliably.

There are certain statistics that are of valueon an annual basis, but are so expensive that theycan be prepared only at infrequent intervals.

These factors—the need for data, the availa-bility of ppulation data, and the cost of produc -tion-shouldbe basic considerations in determiningthe frequency with which various statistics are pro-duced. Ideally, a complete cyclical program ofcollection of data and tabulation of them should beplanned so as to maintain a level budget operation.In practice, this is difficult to achieve. New itemscannot be introduced and old ones deleted from theoriginal vital records at frequent intervals. Evenif this were possible, it would be very unlikely thatthe vital records filed for any particular year wouldbe returned on the same form. Cyclical shifts in

240 VITAL STATISTICS—SPECIAL REPORTS

items to be collected and/or tabulated entail thedanger of forever losing the information on an itemonce it is drqpped. Another limiting factor is thatlong range planning is needed, with an annual re-view in order to assure that changing conditionsand demands are being met. This is not an insur-mountable clifficulty although serious problemsmay arise if there is loss of flexibility of operationthrough decentralization of statistical processing.Perhaps a more serious problem relates to thefact that a long range program is difficult to exe-cute because of the uncertainties of annual appro -priations. Anticipated funds needed for a plannedproject may not materialize at the appointed time.This could throw certain phases of the cyclicalprogram out of line.

Despite the problems and difficulties of car-rying out a cyclical program of tabulations, suchan approach to tabulation represents the most ef-ficient use of resources consistent with meetingthe major consumer demands,

RECOMMENDATION. The concept of cycli -cal tabulation of data should be firmly estab-lished as a part of the regular program of

the NOVS with adequate safeguards for thecontinuity of the various tabulation cycles.

Methods of producing national vital statistics

As part of its study, the Committee investi-gated (a) various methods of producing national vi-tal statistics from the standpoint of relative costsand the scope and quality of the end product, and(b) the question of duplication of effort between theState vital statistics offices and the NOVS. Thefindings are as follows (for details, see Appendix C):

1. Similarity and overlapping in State and Fed-eral vital statistics procedures, tabulations, publi-cations, and needs have not in the past and do notnow cause sufficient” duplication’ 1so that this mer-its consideration as a major or controlling factorin pkuming the program of the NOVS.

2. The transcript m&hod for producing na-tional vital statistics now in operation in the NOVSmeets the basic requirements for national vitalstatistics specified on page 2310f this report. It isefficient to administer, and, because of the large

volume of records processed, operates at a low unitcost per record.

3. The NOVS needs a transcript of each in-dividual vital record in order to maintain controlof the quality of national vital statistics tabulationsand to provide a source of data for research inmedicine, public health, and population growth.

4. The punched-card method for producingbirth statistics was initiated with the belief thatsignificant savings to the NOVS would result fromthe elimination of an assumed duplication in proc-essing between the States and the NOVS, Existingduplication alone is not sufficient to make it a de-cisive factor in planning the program of the NOVS.Precise statements of savings by the eliminationof such duplication are not available and existingestimates vary widely. In the judgment of the Com-mittee, any savings which are likely to occur willbe insufficient to counterbalance the following dis-advantages of the punched-card method: (a) thepunched-card method undesirably restricts thefreedom of action both of the States and of the NOVSin planning their programs; (b) the scope and de-tail of national vital statistics cannot be greaterthan those of the participating State with the leastscope and detail; (c) the NOVS would gradually loseits position of leadership in national and interna-tional vital statistics due to the loss of technicalskills and knowledge and no other agency wouldgain a corresponding leadership; and (d) there isno way of ensuring that a State may not suddenlywithdraw from the plan at a time when the NOVSis unprepared to expand its work.

The Committee has been unable to discoverany advantages that cannot be realized by methodsthat are free from the defects of the punched-cardmethod. It can foresee no way to obviate the dis -advantages of the punched-card method in the pro-curement of mortality statistics. It recognizes thatit may be possible for the NOVS to work out a mu-tually advantageous punched-card program for birthstatistics with a few State or city vital statisticsoff ices that will obviate most of the defects of themethod as originally proposed. It sees no possi-bility of extending this program to all, or even amajority, of State and city vital statistics officeswithout jeopardizing. the scope, quality, and use-fulness of national vital statistics.

Vol. 45, No. 11 NATIONAL VITAL STATISTICS NEEDS 241

5. The pretabulation method of collecting na-tional vital statistics has all of the disadvantagesmentioned above on the punched-card method. Inaddition, the pretabulation method is inflexible. Ifit were applied to births and deaths, the availablenational data would shrink to a minimum and to arigid list of tabulations. With pretabulation, it wouldbecome extremely difficult to make studies of im~portance to the country’s health and welfare thatare based upon information which has been obtain-able only from copies of the original birth and deathrecords. The pretabulation method would increasecosts of State operations without a compensatorydecrease in costs to the NOVS.

State pretabulated data have been used by theNOVS to provide statistics on marriages and di-vorces. In the absence of resources necessary toprocess copies of the individual records, use ofpretabulated data has been necessary. However,use of this method over a 6-year period has dem-onstrated clearly its serious limitations. No datapresented in a uniform way for all States are avail-able. Additional tabulations of the data to answerrequests or make special studies have not beenpossible.

RECOMMENDATIONS: The NOVS should—A.

B.

Reject any data procurement method that limitsor restricts the use of vital statistics recordsfor research or that prevents it from control-ling the quality of national vital statistics tab-ulations. Specifically, the NOVS should procuretranscripts of the individual vital records forprocessing.Reject the proposal to produce national birthand death statistics from data pretabulated bythe States because:

(1)

(2)

(3)

the available mtional data would shrink tothose obtainable from a minimum rigid listof tabulations;it would become very difficult to make na-tional studies based upon information ob-tainable Only from transcripts to the indi-vidual records or from individual punchedcards;it would add to the cost of State operationswithout a compensatory decrease in the costto the NOVS.

c.

D.

E.

Reject the punched-card method of collectingvital statistics data as a general method ap-plicable to all States because:

(1)

(2)

(3)

(4)

(5)

it restricts the freedom of action both ofthe State offices of vital statistics and ofthe NOVS, thus making vital statistics lessuseful rather than more useful;effective leadership in natioml and inter-national vital statistics will be sacrificeddue to the loss of technical skills and knowl-edge in the NOVS;the scope and detail of national vital sta-tistics cannot be greater than those of theparticipating State ranking lowest in theserespects;the possible savings in cost are outweighedby the disadvantages of the method;the overlapping of State and Federal vitalstatistics procedures, tabulations, and needsis not great enough to mAke this methodpossible without serious disadvantages toboth parties.

Abandon any further experimentation with thepunched-card method in the collection of mor-tality statistics and, if the use of this methodto collect birth statistics is continued at all,limit it to the few States where it can be shownto be mutually advantageous and where it willnot adversely affect the scope, quality, con-tinuity, and usefulness of national vital sta-tistics.Continue to cooperate with the States in studiesof ways to improve the division of labor in themtional vital statistics system. Changes in thesystem should be made only after full consid-eration ofi

(1)

(2)(3)

6.recommendations in its efforts to develop ade-

the basic requirements for national vitalstatistics specified on page 231 of this re-port;safeguards for the continuity of operation;opportunities to improve the final productand to render maximum service.

The NOVS should observe the preceding

quate national statistics of marriages and divorces,

242 VITAL STATISTICS—SPECIAL REPORTS

while recognizing that existing conditions may notpermit immediate adoption of all procedures thatare now feasible in the case of births and deaths.

Specifically:

Similar actions are needed in the field of na-tional divorce statistics. However, specific rec-ommendations are deferred until a thorough studyof this problem can be made.

A.

B.

c.

The NOVS should work actively to establish aMarriage Registration Area through imple-mentation of the four criteria established foradmission to the area,3 and to extend it untilevery State is included.Methods of collecting marriage statistics shouldbe explored with a view to assuring completecoverage of the United States, uniform data bothbetween and within States, comparability of timeseries, and accuracy of information.The NOVS should obtain copies of marriage rec -orals from the States and local areas to be proc -essed in the NOVS.

3Th~ Novs ~’~d the p“bli~ Health COnferen~e on

Records and Statistics have agreed that the Mar-riage Registration Area will be comprised of thoseStates meeting the following criteria:

1.

2.

3.

4.

Central record files for marriages in Stateoffice containing items on Standard Recordof Marriage form.

Adoption of report form of marriage contain-ing the items on the Standard Record of Mar-riage form.Reporting by all local areas regularly tothe State office.Agreement between State office and NationalOffice of Vital Statistics on joint testingof reporting for completeness and accuracy.

APPENDIX A

Consumer Views on National Vital Statistics

A consumer survey was conducted on the usesmade of national vital statistics and on their ade -quacy. Questionnaires were mailed to 369 individ-uals including (a) persons or agencies selectedfrom the membership lists of the American Sta-tistical Association, the American Marketing As-sociation, and the Population Association of Ameri-ca; (b) actuaries; (c) State registrars of vital sta-tistics; and (d) various Federal agencies concernedwith health and demography.

Opinions were solicited on the following ques-tions:

1. The frequency with which the principal typesof vital statistics data are used.

2. How well the published NOVSstatistics meetthe consumer requirements in terms of content,timeliness, accuracy, and frequency of publication.

3. Suggestions for reducing the quantity ofmaterial published by the National Office of Vitalstatistics.

The number of responses from the variousgroups were as follows:

QQ!L Sent Returned4

Act&raries --------------------- 12 9Economists ------------------- 96 56Federal agencies ------------- 42 42American Statistical

Association ----------------- 104 64Population Association -------- 55 35State registrars -------------- 60

m 4

q~ncludes 30 questionnaires returned uncom-

pleted; of theee, s indicated that the respondentused vital statistics, and 22 that he “had made nouse of vital etatistice.

The persons or agencies to which question-naires were sent were selected for their knowninterest in order to secure informed opinion on thesub ject. In general, the respondents appeared tohave given careful thought to the questions pre-sented to them. Many of them seemed to have agenuine appreciation of the problems involved inthe production of national vital statistics. Tabularsummaries of the responses to the various ques-tionnaires are shown in tables A-1 through A-6 atthe end of this section.

Use of vital statistics

Mortality and mtality statistics (table A-1) are .the data issued by the National Office of Vital Sta-tistics which are used most frequently by the re-spondents. Next in order of frequency of use arethe statistics on marriages snd notifiable diseases.The consumers surveyed showed the least interest

in fetal death statistics and divorce statistics. Thisgenerally corresponds to the requests for data re-ceived by the NOVS. However, there are indica-tions that if certain inadequacies could be correctedin the data for notifiable diseases, marriages, di-vorces, and fetal deaths, more use would be madeof them.

An overwhelming majority of the consumerssurveyed indicated use of statistics for the countryas a whole and for States (table A-2). Also, themajority of respondents appear to have need forall of the other geographic data listed. A numberof respondents stated that they were interested indata only for a specific county or city.

Question I(c) together with question I(d) (tables

A-3 and A-5) was designed to determine the extentof use of sources other than the NOVS for vital sta-tistics data. k? may be expected, the meet frequentusers of publications of State and local agencies

were the State registrars of vital statistics. Theyconstituted one -half of the “fre~ent’1 users of

State and local publications. A comparison of ta-

243

244 VITAL STATISTICS —SPECWL REPORTS

bles A-2 and A-3 with respect to all the people re-porting the use of State data indicates that only

abcut one-quarter of them refer “frequently” toState publications as the source, less than one-halfdo so “occasionally, ~t and the remainder ‘‘prac -

tic ally never.” However, this may be due, in part,to the composition of the list of persons and agen-

cies surveyed.Of the secondary references, the Statistical

Abstract of the United States, published by the—.—— -------- .-—Bureau of the Census, is most frequently used (ta-

ble A-5). This is followed by the United NationsDemographic Yearbook and the World Almanac.These yearbooks and the almanac are apparentlyconvenient references. Because of the natureof yearbooks and almanacs, it is likely that ref -erences to these secondary sources are more fornational figures than for State, city, and county data.

About one -half of the respondents stated that

they have requested unpublished data from the NOVS(table A-4). A number of those that had not triedto obtain unpublished data either were not aware ofthe availability of unpublished data or stated that

they did not consider their needs important enoughto justify a special request.

Sco~e or detail of subject matter, timeliness,——. -——..and accura~

Scope or detail of subject matter. —A large

proportion of respondents indicat=d–satisfactionwith the existing scope and detail of the publisheddata (table A-6). There were, however, a numberof unmet needs mentioned. Some of the more fre-

quently mentioned ones are as follows:

IiUxME W (a) interval since marriage, (b) in-terval since last previous birth, (c) educational

attainment of one or both parents, (d) order ofbirth by single years of mother’s age, and (e)occupation of father.Deaths by (a) occupation and (b) in more detail

, for local areas.Fetal deaths by (a) late fetal deaths by causeand (b) data on early fetal deaths.Notifiable diseases by (a) age, sex, and other

personal characteristics; and (b) in greaterdetail with respect to cause. There were a

number of suggestions that morbidity statisticsbe extended to include disabling illnesses ofvarious kinds.Marriages and divorces by characteristicssuch as age, previous marital status, literacyor level of education, occupation, et cetera.The lack of complete geographic coverage was

mentioned frequently as a serious defect inpresent statistics.

Timeliness .—The dissatisfaction expressed

most frequently and most strongly concerning theNOVS publications related to the lack of timelinessof their issuance (table A-6). The release of un-published data to those requesting them appearsto have softened the reactions on the part of someof the respondents. Considerable understanding andappreciation of the problems faced by the NOVS in

the early publication of data were evident. How-

ever, the general opinion was that the data wouldbe much more useful if issued on a more timelyschedule. It seems unlikely that any reasonablecurrent schedule of the NOVS operations would meet

some of the needs indicated.Accuracy . —Most of the respondents did not

express any opinion concerning accuracy except toindicate general satisfaction with the quality of data

(table A-6). The defects mentioned were those thathave been already recognized by the NOVS, niamely,errors in residence allocation of births and deaths,

and accuracy of notifiable disease statistics.

Reduction in quanti ty of material publ.khed by

the National Office of Vital Statistics

There were relatively few suggestions for re-ducing the quantity of data published. In most cases,

the respondents appeared satisfied with the datanow being published and questioned the desirability

of cutting out data. In instances, suggestions weremade to increase the volume of published data.

In a sense, the question on reducing the quan-tity of material now published by the NOVS was a

difficult one to answer. The respondent is morereadily able to see the value of data in his specific

area of interest, but appears to have less appre-c iation of other data.

Vol. 45, No. 11 NATIONAL VITA1, STATISTICS NEEDS 245

Table A-1, Question I(a): DO IOU refer to ths publtcations of the Natloflal (l ff[ce of V( tat Stat tst tcs foF znformat ton on btrths,deaths, fetul deaths, morbidl ty, morrioges, and dlucrces?

SOMMARYOF B2PL12S

(F. , Frequently; O. . Cccasionelly;P.N. , PI’scticdJ.Ynever; 1/.S.. Not stated)

—

GRcUP

TOTAL----------------------------------

Im’rican Stltiaticul.L;.ociatlon---.---------,..,cia~ un~ ~,2Jic~l-------------------------

Other--------------------------------------Population Aosoclation-----------------------Ftid,ralng,.ncico-----------------------------Wit.. m~fllutrnrs-----------------------------Smiety uf Actuaries-------------------------EKKwni!3t-.,-----------------------------------

GRoUP

TOTAL----------------------------------

American statisticalAssociation-------------SaCial and medical-------------------------War--------------------------------------

Population k.ociation -----------------------federal agencies-----------------------------State registrars-----------------------------ZacLety of Actuaries-------------------------Ekonmists-----------------------------------

BIRTSS F2TAL D2A!CRS

7F. O.

95 103

P.N.

19

P.N.N.S.

U

—F.

38

87163

192

P.N.

15

F.

1(26

24168

201634

4a

F.

o.

89

N.s.

10

0. 1.s.—

33—

67 86

I5233

G

422

1

5

167 t9 17

53212

16

734

1111

311219EL1413

1:

Ll92

1.51021

37

3212201216

52

19

1349113114

22 u11 1630 17

I —

MORBIDITY MABBIAG2S

N.S.I

F. o. I P.N. o.

99 TP.N. N.S.

65 12

62 26 46-

21 1.7 31 18 1

F. 0.

f

P.N. N.S.

99 23

33 819 214 612-22 19 43 120 9

I

48 78 I 70 28 48 29 73

I

26 1715 4u. 138 21

131125

1:13

826

15113

918

125

10210

2

2104

1021

21.u.1012

325

1:

13 B18 767 1;

Table A-2. Question I(b): lb uou use the following geographic categories of data: Nat tonal, State, countu, Citu, cities groupedbg size of populat ion, and urban-rural?

SUMMARYOF BSPLIXS

(N.S. = Not stated)

I NATIONAL I STATE I COONTY I CITYGROOP

Yes

meal -------------------------- 218——

American .%atistf.calAs60ciati0n----- 61EmcitiLend medical----------------- 30other ----------------------------- 31

Population AsB0cL9tion--------------- 34Federe.1agencies---------------------- 30Stntc registrars--------------------- 4sSatiety of Actuwice ---------------- 8EcnnamintB--------------------------- 39E

10 N.S. Yes

2 4 192

1 541 : 29

1 25.- 27

. 24;14s-- 7

1 34

—

70

—

20=

615641

3—

.s. Yes No N.S. Yes

12 128 68 2C 147

4 19 l-l 422 2 7 5 232 14 12 6 191 20 12 2 262 15 12 3 151 33 12 3 351 2 4 2 33 24 9 7 26

—

fo 1.S.] YesINo ]N.S.i Yes/No ~.S.m54=

144

106

11939

26 8 - 30 4 -4 16 9 5 21 5 44 29 17 2 31 ls 22 3 3 2 6 1 1.3 20 u 9 23 98

246 VITAL STATISTICS—SPECIAL REPORTS

Table A-3. Question I(c): no uou refer to puol icat ions of Stateund laeal a~etici~s as sourcw for uital statistics riata?

SOMWARYOF REPLIES

(F. . Frequently;O. . Occasionally;P.N. . Practicallynever;N.S. . Not stated)

GROUP F. o. P.N. N.S.

TOTAL----------------------- 50 95 76 3

AmericanStatisticalAesOciatiOn-- 16 26 21 1Socialend medical--------------- 10 16 5 1Other --------------------------- 6 10 16 -

PopulationAssociation------------ 5 17 12 -Federal~encies------------------ 2 I.1 17 -State registrars------------------ 25 18 ~ 1Societyof Aclwnries-------------- 1 1 6 -Economists------------------------ 3 22 14 1

Table A-4. Question I(e): ,%ZWyou ever tried to nhtgtn un,nub-lished data from the Nattonal Office of Vital Statistics?

SWY OF RsPL03S

GROUP Yes No NCItstated

NN!lm---------------------------

PmericanStatisticalAssociation------.%cialend medicti------------------other-------------------------------

Populationksociation----------------Federala~ncies----------------------Stateregistrars----------------------Sbcietyof Actusies------------------Economists----------.->---------------II

1.13 109 2

29 34 118 14 -M. 20 118 16 -27 3-29 18 14 4 .6 34 -

Table A-5. Question I(d): Do you usv swondarg references (e. g., olmanacs, .uearbooks, encyclopedias) as soUrces for Ui tal Sta-tistics data?

SUMM4RYOF REPLINS

(F. = Frequently;O. . Occasionally;P.N. . Practicallynever;N.S. . Not etated)

GROUP

TOTAL -----------------

AmericanStatisticalAesoc.-Socinland medical--------Gther---------------------

PopulationAssociation------FederelAgencies------------State re@strsrs------------.%Cietyof Actuaries--------Econmists------------------

STATISTICAL u

ANSTRACT

F. 0. P.N. N.S. F.

17 35 2 ,- 12

5 10 - - 32 4 - - 13 6 24 6 i : 63 4 - - 22 7 - - 1

2 - - -3 6 - - -

ED NATIoNSKwAlwIc WORID AIM4NAC C1’sERGMRSOOK

IO. IP.N

t

9-

2-2---4-2-1----- mF. 0. P.N. N.S.

4 131-

15 :1 -131-

z--il ..-

2--13--

l--;l -- ml?. 0. P.M. N,S,

31 54 82 1

11 14 24 -6711 -5713-459.4312 -4926-

23-8 2181

NOI’ STATED

f

F. O.

2

----

i1

------

:

?.N. N.S.

9s

1313-.2.21321

:

Vol. 45, No. 11 NATIONAL VITAL STATISTICS NEEDS 247

Table A-6. QuestionIk Are the &ta published bg the National office of Vital Statistics adequate for bwur uses in scope ar de-tai z of subject matter, t imel tnMS, and acCuraCLl?

GROUP

--------------------- ------

Anwicm StatisticalA6s0ciat10n------socia.1 and medicd ------------------other-------------------------------

populationAssociation----------------Federalagencies----------------------state r&*tr~L+ ---------------------societyof Actuaries-----------------Economists----------------------------

. --------------------------

AmericanStatlstic@J.Association------~CI~ ~d medic~------------------Other-------------------------------

PopulationAssociation----------------Feder~ ~n~iee ----------------------State registrars----------------------Society of Actumies ------------------Econcmrhta----------------------------

. . . . . . ------------------- ---

AmericanStatiatlcalAssociation------go~i~ ~ ~~cd------.--------.-.Other-------------------------------

FmpUhtion Aseoclation ----------------Federal 4@C@e ----------------------fJtata regtatrus ----------------------Society of Actuaries --- ---------------ECon&~t*---------------------------

SWMARY OF NsPIms

(N.S. . Not eta+-ed)

BIRTSS DEAIHS FETAL DBAmS MoRSIOITY MAsslAGEs DIVORCES

(es No NOS. Yee No N.S. Yes No N.S. Yee NO N-s.

SCOPEOR INmAIL OF SUBJECTw-m

T162 30 32

40 6 M23271741122932262426-

6-23037

—,09=

3211151521155a— T_

70 45

16 169s7 10

13 64530 3126 13

I.47 10

38 121-17 121 122 140 26-20 5

T150 3s

40 820 520 323 61s u39 962

24 2

L02

=

3013171420137

1s

IS7 143

25 383,2. 2214 1612 207 206 4027

15 1s

3s=

167951

14

U6 14 94 99 40 S5 IJ.9 32

1 I I

27 3 34 22 15 27 32 420 3 9 12 12 s 17 27 - 25 10 3 19 15 2

17 1 16 12 4 $ 13 y17 2 .u 17 s s37 8 3 33 9 6 35 55 - 3 5 1 2 113 - 27 10 3 27 2:1 3

II I I I I I

mmwss

II73 49 79

M 16 21.13 s 155 10 6IL91O731532 3 1s

15i 17 10

IIU 70 94

3 23 2410 18

3 13 614 12

4 Sll2634

142 20 9

II36 109 S9

10 33 259 8 161“ 25 94 20 91 14 M20 10 24

351 29 S T

30 105

II 2s972 212 2339

12 1.23

230

ACCUF.4CY

m12 IJ.8 66 26 112

3 37191. I.342 121.2S121 2573223 1993223 16 15 3 122 12 X4. 4 13

4521;307330

F73 104 25 95

2S 272=13 17 1 1415 109 14

15 10s=2

1 2110 104 165 9

IIi-ii11 15 3

%+

95 40 S9 m

30727241? 3 12 1613415s10 IL 13 10II Z1712

1161 i31 19196-25

19 4 17 12

322

—

T_37 105

7333 134 2010 141 171s 13

33 25

T!r103 16 103 92

2s 234=16 - 1S 1512 2 1s 1012 6 16 128 319s

34-14335-34

16 5 19 .10

13=

1

1s3

3—

—

119

3s172116191.5427

-

APPENDIX B

Consumer Views on National Marriage Statistics

A survey was conducted of persons who mightbe expected to use marriage statistics in their work.The major purpose was to obtain information onuses and needs for marriage statistics. The sur-vey was directed to persons in four major areas ofactivity: business, colleges or universities, privateresearch, and government. Information was ob-tained from individuals representing widely diver-gent groups and interests.

Major findings of survey

The major findings of the survey on marriagestatistics are:

1. More than four-fifths of all respondents(84 percent) reported that they had used marriagestatistics in their work within the last year.

2. Respondents fell into 2 major groups em-phasizing 2 distinct types of needs. Respondentsin business organizations stressed the need forcomplete coverage of marriage and for data by age.Respondents in research fields stressed needs fordata on characteristics of brides and grooms.

3. More than two-thirds of all users reportedthat they used marriage statistics for estimatingthe number of households; more than one-half, inanalyzing social trends; and two-fifths, in esti-mating future markets for products.

4. Additonal marriage data for the UnitedStates requested most frequently by users were:(a) previous marital status of brides and grooms,(b) age at marriage, and (c) various cross-clas-sifications of characteristics of brides and grooms.

5. Comments on the national program fell intobroad categories the most important of which were:(a) expressions of need for more data, (b) requestsfor expanded coverage, and (c) statements of use-fulness of present data.

248

Description of survey

Mailing lists.. -Mailing lists totaling 539 nameswere, for the most part, provided by members ofthe Subcommittee on Utilization of Marriage Sta-tistics. A few names were suggested by privateindividuals who had received questionnaires. Aboutone -half of the persons on the mailing lists wereindividuals in business organizations, mostly headsof research departments or officers of the com-pany. In general, the survey mailing list of busi-ness users was limited to those in national organi-zations. The important group of users of localdata in the business field were not included.

Questionnaire style. —The questionnaire wasdesigned insofar as possible to provide for a checktype of response.

Questionnaire tally .—The numbers of question-

naires mailed, returned, and not returned are asfollows:

Questionn~res mailed -----------’539Questionnaires returned --------- 262

Ql]estionnaires tabulated-------- 256Reporting use----------------- 214Reporting nonuse ------------- 42

Questionnaires not tabulated --- 6Questionnaires not returned -----277

Tabulated questiomaires represented 47.5 percentof the total mailed, 51.4 percent were not returned,and 1.1 percent were received too late to be in-c luded in the final tables. It is possible that a high-er response rate could have been achieved but nofollowup was made of the nonrespondents. Further-more, no information was obtained to indicate rea-sons for failure to respond.

Vol. 45, No. 11 NATfONAL VITAL STATISTICS NEEDS 249

Analysis of questionnaire returns

Of the 256 questionnaires tabulated, 84 percentwere from persons who had used marriage statis-tics during the preceding year and 16 percent werefrom nonusers. Summaries of the responses tothe questionnaires are shown in tables B-1 throughB-10 at the end of this section.

Types of users of marriage statistics. —TableB-1 shows the users and nonusers of marriage sta-tistics classified by type and size of organizationand title of the person answering the questionnaire.Close to one-half of the 256 responses were frompersons in business organizations; one-third frompersons in colleges or.universities; a tenth, in gov-ernment; an eighth, in private research; and 1 in50 failed to indicate the type of organization.

Most of the respondents indicated they usedmarriage statistics. This was true of 9 out of 10respcmdente working in colleges, universities, andin government; 8 out of 10 persons in business or-ganizations; and 7 out of 10 in private research.Two -thirds of the respondents in business organi-zations and in private research were connectedwith organizations employing 100 persons or more,and 19 respondents were in organizations with fewerthan 100 employees.

The largest single group of responses (71) werefrom “directors or managers of research depart-ments” of whom 4 out of 5 reported that they usedmarriage statistics in their work. The next largestgroup (54) represented professors in educationalinstitutions of whom 96 percent were users. Nearlyall economists were users, as were most statisti-cians, and a majority of the persons classified as‘‘officer of compariy.t’

Of the 42 respondents who were nonusers ofmarriage statistics, 15 were “directors or man-agers of research departments.” The remaining27 nonusers were scattered among the other cate-gories listed under “position of respondent” intable B-1,

Specific uses of marri&ge statistics. —Thesecover a wide range as indicated in table B-2. Manyof the 214 users indicated more than one use of thedata. Estimating the number of households was theuse repmted by the largest number of users; andthis was followed by use of marriage statistics in

analyzing social trends. Third, and including two-fifths of the users, was the use of marriage sta-tistics to estimate future markets for products.One-fourth of the users utilized the data for teach-ing materials, lectures, sociological research, andmiscellaneous purpaes. Other specific uses ofmarriage statistics include: estimating futureschool needs, determining needs for public utilities,actuarial studies. and analvzing insurance costs., .

Potential uses of marriage statistics. -Usersof marriage statistics were asked what uses theywould make of statistics on marriages, assumingthat all data needed were available. Uses listed bythe 214 respondents are given in table B-3 clas-sified by type of use and by type of organizationwith which the respondent was affiliated.

The two largest groups of users indicated twowidely divergent groups of uses. Of the individualsin business organizations, 38 indicated that the datawould be used in making housing estimates, and 35,in market analysis which includes estimating formerchandising of household appliances and homefurnishings, and sales promotional plans. Of theindividuals working in colleges or universities,31 listed uses of the data for materials for teach-ing and lectures, 29 mentioned sociological re-search, and 21, demographic research. Possibleuses were also listed by persons in governmentand in private research.

Degree of usefulness of published tables ofmarriage statistics. —Respondents were asked tocheck 1 of 3 degrees of usefulness (very useful,moderately useful, not useful) for each of the prin-cipal tables on marriages published in the repertsof the National Office of Vital Statistics (tables B-4and B-5). A large number of respondents answeredthis question in part, indicating the degree of use-fulness of some tables and leaving others blank.These omissions were classified as “not stated:’

All of the tables listed in this question werechecked as very useful by some of the respondents.Four out of 5 persons checked as very useful an-nual totals of marriages in the United States; and3 out of 5 placed in this category data available forsome States on marriages by age, race, and sex,and on marriages by previous marital status andsex. About one -half indicated as very useful tablespublished annually on marriages by State; mar-

250 VITAL STATISTICS—SPECIAL REPORTS

riages by number of previous marriages, age, and

sex; and marriages cross -tabulated by age of brideand groom. Also indicated as very useful weremonthly data on marriages for the United Statesby 3 out of 5 respondents and cm marriages or mar-riage licenses by State and marriage licenses inmajor city areas by 1 out of 3 respondents. All ofthe published tables were found moderately usefulby some users. Monthly data on marriages or mar-riage licenses by State were reported moderatelyuseful by close to one-half of those who checkedthis table. The remaining tables listed were foundmoderately useful by one-fifth to two-fifths of thepersons who checked the degree of usefulness. Anegligible number (less than 1 percent) checked theannual total of marriages in the United States asnot useful. Annual data on marriages by county andmonthly data on marriage licenses in major cityareas were found not useful by about one-third of thepersons who checked these tables. The remainingtables were checked as not useful by one-fifth ofthe respondents or less.

Type of national data on marriages requestedby respondents. —In response to the question ofwhat useful information on characteristics of bridesand grooms should be collected for the United States,many suggestions were received. Table B-6 showsthese suggestions by type of user making them.The data most frequently requested by all types ofusers —business, private research, college or uni-versity, government-were cross -tabulations ofvarious characteristics of brides and grooms, asfollows :

(1) Age by previous marital status.(2) Race of bride by race of groom.(3) Age of bride by age of groom.(4) Previous marital status by age and race.(5) Previous marital status of bride by pre-

vious marital status of groom.

Other requests for cross -tabulated data includereligion, occupation, and education tabulated withvarious factors such as age, race, previous maritalstatus, and residence. The most frequently re-quested single items of data for’ the United Stateswere for mtional data on marital status and ageat marriage. A large number of requests were

also received for data on occupation or industryof brides and grooms, their education, and m-religiousdenomination.

The categories listed under “single items”also include a few requests for related data. Thesecategories and related data are:

(1)

(2)

(3)

(4)

(5)

Income: savings at marriage and financialstatus of couple.Marital status: year of last marriage,length of time since dissolution of lastmarriage, and number of times widowedor divorced.Nationality: nativity, national culturalgroup, place of birth, and ethnic origin.Religious denomination type of ceremonyand degree of agreement on religion.Residence: urban, rural-farm, and rural-nonfarm residence, census tracts, andsize of place.

Frequency of publication. -Respondents wereasked to indicate their preferences for frequencyof publication of data on number of marriages. An-nual publication was the first preference of thelargest number of respondents (table B-7). Monthlypublication was the second choice of close to one-half of the respondents, and weekly publication, thethird choice. Quarterly and semiannual publica-tions were preferred by a few persons, either asfirst or as second choice.

Geographic categories. —Respondents wereaskeu to number, in order of importance, theirneeds for data on marriages by various ‘geographiccategories. Table B-8 shows for each of the areasthe choices of the respondents and the medians ofthese choices. Data for the entire United Stateswas the first choice of three-fourths of all personswho indicated a geographic preference. Data byState was the second choice of slightly less thanone -half of the respondents. Other data by order cdchoice were data by region, county, standard met-ropolitan area, and major city areas.

Advance release of monthly data. —Respondent’.were asked to state whether it was importmt forthem to receive monthly data on marriages 2 weeksprior to their publication in the Monthly VitalStatistics Report. More than four-fifths of the

Vol. 45, No. 11 NATIONAL VITAL STATISTICS NEEDS 251

respondents to this question (172 out of 202) indi-cated that they did not need an advance release.Table B-9 shows both the distribution by @e oforganization of the 30 respmdents who indicatedthat an advance releas~ was important to them, andthe distribution of those who indicated no need foran advance release.

Comments on the national program of marriagestatistics , -Respondents were asked to make com-

ments or observations on the national program ofmarriage statistics. Table B-10 shows these com-ments classified under broad categories and bytype of organization with which the respondent isaffiliated. They emphasize three points: (a) pres-ent available statistics are useful, (b) complete ge-ographic coverage is needed, and (c) more detailedinformation on characteristics is needed.

Table B-1, RESPONDENTS REPORTING USE OR NON ISE OF MARRIAGE STATISTICS, BY TYPEAND SIZE OF ORGANIZATION, AND 3 SITION OF RESPONDENT

TYPN AND SIZE OF ORGANIZATIONAND POSITION OF RESPONDENT

TYPE OF ORGANIZATION

Total respondents----------------------------

Rwinem -------------------------------------------College or university----------------------------Governtuent---------------------------------------Private research---------------------------------Not stated---------------------------------------

SIZE OF ORGANIZATION=

Total respondents----------------------------

Less than 5 employees----------------------------5-99 employees-----------------------------------100 employeesor more----------------------------Notistated---------------------------------------

POSITION OF Respondent!

Total.respondents-----------------------------

Officer of’company-------------------------------Professor (educatione3institution)---------------Director or msnager of research department-------Statisticisn-------------------------------------Economist----------------------------------------Other-------------------------- ------------------Not stated----------------------------------------

%winess and private research only.

mtal.

256

1218426205

141

3169230

256

205471X5304818

Use

214

1007623141

114

3148314

214

135256I-328391.3

Nonuse

42

218364

27

-

29

16

42

7

G2295

PERCENT DISJ?RIBU!FION

Total

100.0

100.0100.0100.0100.0100.0

100.0

100.0100.0100.0100.0

100.0

100.0100.0100.0100.0100.0100.0100.0

83.6

82.690.588.570.020.0

80.9

100.087.590.246.7

83.6

S3.o96.37%.986.793.381.372.2

Nonuse

16.4

17.49.5

I.1.s30.080.0

19.1

012.59.s

53.3

16.4

35.03.721.113.36.718.727.8

252 WTAL

Table B-2. SPECIFIC USES OF

STATISTICS-SPECIAL

MARRIAGE STATISTICS AS IJSTED BY 214 RESPONDENTS

TYPE OF USE

Estimating number of households-------------------------------------------------Analysis of social trends--------------------------------------------------------Estimating future markets for products------------------------------------------Estimating f’utureschool needs--------------------------------------------------De%emining needs for public utilities------------------------------------------Other---------------------------------------------------------------------------University teaching and lect~~es----------------------------------------------Sociologicalresearch and tisce~-eous ---------------------------------------All other---------------------------------------------------------------------

Numberof

users

149IJ.687301859182615

Fercentof

respondents

69.654.240.714.08,427.68.412.17.0

TableB-3. POTENTIAL USES OF MARRIAGE STATISTICS AS LISTED BY 214 RESPONDENTS, BYTYPE OF ORGANIZATION

TYPE OF USE

Housing estimates-------------------------------------Market analysis---------------------------------------Demographicresearch----------------------------------Lectures and teachingMaterials-----------------------Articles and textbooks--------------------------------Sociologicalresearch---------------------------------Other-------------------------------------------------Communityneeds-------------------------------------Public utilities------------------------------------Credit and insurance--------------------------------Special studies-------------------------------------All other-------------------------------------------

Total

4841493321414171058n

Business

383513

132239523

?ollegem ud.-versiti

532131172971

24

Govern-ment

3112

1583

.14

Privateresearch

2232244

1

3

Vol. 45, No. 11 NATIONAL VITAL STATISTICS NEEDS 253

Table B-4. DEGREE OF USEFULNESS OF MARRIAGE STATISTICS PUBLISHED BY THE NATION-AL OFFICE OF VITAL STATISTICS AS REPORTED BY RESPONDENTS, BY TYPE OF DATA

m OF DA!m

Annual.dataforthe UnitedStates:Totalumrrieges------------------------------------------Marriageeby State---------------------------------------Mwriages by county--------------------------------------

Annualdatafor reportingStates:FY.r6tmartiagesand remafiegesby age,race,end sex----Mard.egesby age,previousmaritalstatus,and sex-------Mwrlages by age,numberof previousmarriages,and sex--AM. marriagesby age of brideby age of groom------------llrstuwzlages by age of brideby age of groom----------

Montlilydt+z:Totalmawieges in tie UnitedStates---------------------Mwriages Q. rtiegelicensesby State------------------Mu-riagelicensesin w~or ci~ areas--------------------

Total

209209208

209209209209209

209209209

Veryuseful

1547741

8083667069

1015142

!$3der-atel.yMeful

375549

334041

5144

606857

NotMefUl

11347

2621312423

142946

No%stated

I-7647’2

7065716473

346164

Table B-5, PERCENT DISTRIBUTION OF DEGREE OF USEFULNESS OF MARRIAGE STATISTICSPUBLISHED BY THE NATIONAL OFFICE OF VITAL STATISTICS, AS REPORTED BY RE -

SPONDENTS

(Percentsbasedon data shownin tableBA, omitting“notstated”)

TYPEOF DATA

Annualdatafor the UnitedStates:Totalwrtiages-----------------------------------------------Fiwriagesby Stwte--------------------------------------------Mwriages by county-------------------------------------------

Annualdata for reportingStates:Firstmarriagesand remarriagesby age,race,and sex---------Msrriagesby age,previousmaritalstatus,and sex------------Msrrlagesby age,numberof previousmwriages, and sex-------AIL marriagesby age of brideby age of groom-----------------‘It&atmarriagesby age of brideby age of groom---------------

Monthlydata:Totalmarriagesin the UnitedStates--------------------------Marrlageaorumwiage licensesby State-----------------------Murl.agelicensesin maJorci+jyareas-------------------------

Total

100.0100.0100.0

100.0

100.0

100.0100.0100.0

100.0100.0100.0

Veryuseful.

80.253.129.9

57.657.647.848.350.7

37.734.529.0

Moder-atelyuseful

19.337.935.8

23.727.829.735.232.4

34.345.939.3

Notuseful

0.59.0

34.3

18.7

14.622.516.616.9

8.019.631.7

254 VITAL STATISTICS-SPECIAL REPORTS

Table B-6. TYPE OF NATIONAL DATA ON MARRIAGES REQUESTED BY 214 RESPONDENTS, BYTYPE OF ORGANIZATION

TYPEOF DATA

Cross-tabulationsof vm’iouscharacteristics----------

Singleitems:&e-------------------------------------------------Coloror race---------------------------------------Education-------------------------------------------Income----------------------------------------------mrital status--------------------------------------Militarystatus-------------------------------------Natiotii~-=---------------------------------------Occupationor industry------------------------------Religiousdenomination------------------------------Residence-------------------------------------------Othersocioeconomicdata----------------------------All other-------------------------------------------

Total

173

391730164210143227221741

=+=

CollegeJ3asinessoruoi-

versity

22 98

1317895

311129

1712177202

11202215424

Zovern -ment

35

41416316

512

PrLV2rteresearch

18

532

7-2341

6

TableB-7. RECOMMENDED FREQUENCYOF PUBLICATIONOF MARRIAGE STATISTICS BYORDEROF RESPONDENTS’ PREFERENCE

FREQUENCYOFPUBLICATION Total First Second Tllira Notchoice choice choice stated

ti*---------------------------------------------- 164 123 26 6~n~-----------------------------------------------

9135 68 55

Weekly5

------------------------------------------------ 61 1 2 5:@rterQ ---------------------------------------------

3n 3 5

&timuay ------------------------------------------S

2 1 L

TableB-8. RECOMMENDED GEOGRAPHIC CATEGORIES FOR PUBLICATION OF MARRIAGE STA-TISTICS BY ORDER OF RESPONDENTS’ PREFERENCE

GEOGRAPHICCATEGORY

National-----------------------Regional-----------------------State--------------------------county-------------------------Majorcityarea----------------Standardmetropolitanarea-----T

Total Firstchoice

187 108112 1161 27113 7108 5124 u

I 1

Second Third Fourthchoice choice choice

132950141313

62025262124

—

5816251230

I I IFW?th sixth Not Medianchoice choice stated choice

,

615513309

8222

17Il.14

411736IL1623

1.03.42.23.74.13.6

Vol. 45, No. 11 NATIONAL VITAL STATISTICS

Table B-9. RESPONDENTS UWING NEED FOR ADVANCE

NEEDS 255

RELEASE OF MONTHLY DATA ON

MARRIAGES BY TYPE OF ORGANIZATION

TYPE OF ORGANIZATION

TM------------------------------------------------------------

~winesB-----------------------------------------------------------------Collegeor tiversi*----------------------------------------------------Mvement ---------------------------------------------------------------Privatemsearti---------------------------------------------------------tier ---------------------------------------------------------------------

=-II==!=-172

977120X51

Table B-IO. COMMENTS OF RESIWNDENTSON NATIONAL PROGRAM OFTICS, BY TYPE OF ORGANIZATION

I III Total,

ToTAL. . . . . . . . . . . ------------- -------------------- I 108

Expandedgeographiccoverage--_----F------------------

r

22Mme bti---------------------------------------------34Improveddata------------------------------------------6Presentstatisticswe~----------------------------- 24AU other.-------------------------------------------- 22

Business

43

U_12 858 636 144 9

1

STATIS-

Collegeor wi- IIGovern- Private

versityment reseereh

43 I 13 I 9

APPENDIX C

Survey of Methods of Producing National Vital Statistics

Methods of producing national vital statistics

Three such methods were considered by theCommittefx

1. Transcript method. This involves the com-plete processing by the NOVS of data on copies ofindividual records received from State vital statis-tics offices. Recently the use of microfilm hassuperseded handwritten individual transcripts inall but 5 States, Puerto Rico, and the Virgin Islands.In this report, the term t?transcript method” ap-

plies to both the use of handwritten transcripts andmicrofilm copies. The transcript method has beenin use in the United States for more than 50 years.

2. Punched-card method. The coding of dataand card punching is done in the State vital statis-tics offices; copies of punched cards are sent tothe NOVS for verification and tabulation. Verifica-tion is effected by comparison with transcripts ormicrofilm copies of the original vital records whichalso are sent to the NOVS by the States, This pro-cedure was introduced experimentally in 1951. In1955, 18 State and independent registration areasparticipated in the State-produced birth punched-card program, and two States in the State-produceddeath punched-card program.

3. Pretabulation method. Experience withthe application of this method in the processing ofbirth and death records is not available, since noState has yet submitted pretabulated data. TheNOVS has been receiving pretabulated statistics ofmarriages and divorces for years and commentson this experience are presented later,

The pretabulation method would involve stnlfurther decentralization of the production of na-tional birth and death statistics. The State vitalstatistics offices would furnish the NOVS with pre-

256

tabulated data and/or summary punched cards.From these, the NOVS would prepare natioml tab-ulations. If the States were to furnish pretabulateddata in tabular form, the NOVS would preparepunched cards from them in order to make nationaltabulations and tables for publication most effi-ciently. By this method, the NOVS would not haveavailable either copies of the individual records orthe detailed punched cards from which the pretab-ulated data were prepared,

These three methods of producing national vi-tal statistics will be evaluated primarily, but notexclusively, with reference to the preparation ofnational birth statistics, since no comprehensiveexperience in the use of the punched-card methodfor the compilation of national mortality statisticsis available, The problems in processing deathstatistics are more complex than those of birthstatistics; hence, any negative conclusions con-cerning the birth punched-card program will applyat least equally to a punched-card program fordeaths. There is no experience in the use of State-prepared punched-cards for production of nationalstatistics of marriages or divorces. However, theprocessing problems are comparable in complexityto those for births.

The proposals for the alternative methods ofproducing national birth and death statistics arosefrom the view that considerable duplication of ef-fort in processing and tabulating data existed be-tween the NOVSand the State vital statistics offices.It was further believed that the total national ex-penditure for the production of vital statistics couldbe reduced materially if the processing that theStates do for their own purposes could be modifiedso as to simultaneously produce the data requiredfor the natioml vital statistics program.

Vol. 45, No. 11 NATIONAL VITAL STATISTICS NEEDS

The problem of duplication

The problem of duplication is often oversim-plified. Careful examination is needed to ascertainwhether the duplication is apparent or real, andalso whether any existent duplication is unneces-sary or undesirable. While some’ aspects of Stateand Federal needs overlap, the complete elimi-nation of all duplication which may now exist mayactually increase the total cost and even result ina loss in efficiency if both State and Federal needsare to be met satisfactorily.

Studies of duplication .—Two comparative stud-ies have been made of the birth and death statisticsin State publications. The purpose of the f irst studywas to determine the number of States which pub-lished each of 14 selected basic tables for birthsand deaths. The study was conducted in 1947 by acommittee of the Conference on Vital Records andVital Statistics under the chairmanship of Dr. JosephV. De Porte, Director of Statistical Services andResearch, New York State Department of Health,and was based primarily on State reports for theyears 1940 to 1945.

The Committee headed by Dr. DePorte con-CIuded, inter alia, “We feel, in general, that nation-

al statistics should be compiled by the NOVS forthe whole United States. Secondly, we feel that theNOVS should compile and publish tables of basicfigures for States for comparative purposes. Thereare certain deviations in every State and we cannotcompare figures compiled by one State with thoseof another. Thirdly, we feel that States should com-pile and publish their own local statistics . . . . ‘‘

The second study, made by the NOVS in 1954,analyzed in detail the birth and death statisticspublished by 42 State offices and the District ofColumbia. Most of the reports considered relatedto years from 1950 to 1952. The results of thisstudy, presented in tables C-1 and C-2 at the endof this section, are not directly comparable withthose of the first study; however, the general pic-ture is similar.

Of 18 basic birth topics in table C-1, no Statepublication contained more thsn one-half and only14 States published data on as many as one-third.For mortality, 2 States published statistics on 11of the 16 topics in table C-2 but the published re-

ports of only 9 States contained more than one-halfof these topics. .

257

The most common birth classification, se-was found in the reports of 29 of the 42 States; dataon month of birth, and city and county of birth werepublished by 25. Statistics about other basic topicswere published by still fewer States, that is, age ofmother by only 12 and birth order by only 6. Corr-esponding variations existed for the specific mor-tality topics. Only 35 States published statisticsfor cause of death for counties and cities and only28 States published tix?m by age. ‘13velve Statespublished the number of deaths by month.

The two comparisons show that most of thestatistics published by NOVS are not duplicated bythose published by the States.

Qualifications to these studies. —Both studieswere focused on the extent of similarity of the top-ics covered in State publications. Although no studywas made of the similarity between the tabulationsprepared by the States and those of the NOVS, thedetail and scope of the latter undoubtedly exceedby far those of any State.

The summary data presented in tables C-1 andC-2 exaggerate any apparent duplication. For ex-ample, table C-2 indicates that 35 States publisheddata by cause of death for cities and counties. How-ever, the number of causes given by the States forthis tabulation ranged from 6 to 185. Similarly,any of the items shown in tables C-1 and C-2 maybe presented in a varieiy of classifications basedon different editing or coding procedures. In par-ticular, there are rather wide differences in com-mon items such as the geographic code and clas-sification, and the causeof-death classification.

In general, tables C-1 and C-2 lead to the con-clusion that the needs of individual States for pub-lished data vary over a wide range and that wide-spread systematic duplication behveen the tablespublished by the States and those published by theNOVS does not exist.

Processing practices. —Another source of pos-sible duplication may be found in some of the pre-liminary steps, for example, coding and punching.A thorough survey of the latter would entaila studyof the editing, coding, and punching procedures ofeach State. Although such a survey has not beenmade, some evidence is provided in the responsesto a letter from the NOVS to the States concerninguniform codes and procedures. For each inquireditem regarding births, a number of States differedfrom the NOVS in their codes or coding rules to

258 VITAL STATISTICS-SPECIAL REPORTS

such an extent that further discussion was requiredfor reconciliation. The items for the 31 States thatprotided detailed information are:

No. ofSubject States Comment

differing

Geographic code --- 15

sex ---------------- 5

Month of birth ----- 6

Race --------------- 10

Age of father ------- 3

Age of mother ----- 3

Nativity of mother - 12

Number of children 12

Attendant ---------- 7

Length of pregnancy 3

Birth weight ------- 8

Legitimacy -------- 12

Plural births ------ 8

Other States also haddifferences but indi-cated that these couldbe reconciled byadoption of proposeduniform procedures.

This includes 1 Statewithout this item.

This includes 7 Stateswithout this item.

None of these Stateshad the item in thesame form as NOVS.

.

This includes 1 Statewithout this item.

This includes 3 Stateswithout this item.

None of these Stateshad &is item in sameform as NOVS.

The foregoing evidence, though not conclusive, in-dicates wide differences in State and Federal prac-tices.

On the other hand, the 18areas participating inthe experimental punched-card program for birthsin 1955 have brought their procedures into essentialconformity with those of the NOVS. If the existingvariation among the remaining States reflects validdifferences based upon individual needs, then it isnot desirable for States to change their procedures

so that they will conform with those of the NOVS.h fact, a few States already have decided not tomake the chsnges required for the punched-cardprogram. The Committee did not conduct any in-quiry among the remaining States with regard tothe desirability of conformity.

Conclusion regarding duplication. —At the pres-ent time, overlapping in State and Federal vitalstatistics procedures and tabulations is not greatenough to m&e such apparent “duplication” a ma-jor or controlling factor in planning the programof the NOVS.

The transcript method

For more than 50 years, the compilation of na-tional statistics has been based on copies of each@ividual vital record sent to the NOVS (or itspredecessor, the Division of Vital Statistics in theBureau of the Census) by the separate States orregistration cities. This method may be evaluatedfrom the viewpoint of the basic requirements fornational vital statistics (specified on page 231 of thisreport), efficiency of administration, and cost.

Meeting the basic requirements for nationalvital statistics. —By central processing of tran-scripts or microfilm copies of individual records,national tabulations can be prepared more easilyaccording to uniform classifications and rules thanby either the punched-card or the pretabulationmethod. Central processing also makes it possibleto measure inexpensively the degrees of complete-ness and accuracy achieved in producing the sta-tistical end-results. The knowledge thus acquiredprovides the background for the NOVS to explainand interpret the statistics it publishes. Such ex-planations and interpretations should be incorpo -rated in the publication of national vital statistics.The availability of copies of the basic records en-ables the NOVS to cross-classify its tabulations insufficient detail to meet not only administrativerequirements, but also the interests of research.Central processing has the further advantage off legibility and adaptability to meet unanticipatedneeds.

Efficiency of administration. —Compared tothe punched-card and pretabulation methods, theuse of transcripts places the minimum burden uponthe 56 State and city vital statistics offices for theproduction of national vital statistics. With thetranscript method, the timetable for compiling na-tional vital statistics depends mainly upm the re -

Vol. 45, No. 11 NATIONAL VITAL STATISTICS NEEDS 259

sponsible national agency, provided that the tran-scripts are received when due. It is subject in alesser degree to the varying timetables of 56 Stateand city off ices, to their varying degrees of ef -ficiency and accuracy, and to their administrativevicissitudes. In particular, except when changesare required on the original record, it is not nec-essary to enter into, extensive negotiation with tie56 vital statistics offices preliminary to makingchanges in the national vital statistics program inorder to meet new needs. By fixing at a singiepoint the res~nsibility for producing adequate andtimely national vital statistics, the possibility of aserious interruption or breakdown of the system ismintmized,

C-. —Considered as a data processing pro-cedure, the transcript method is inexpensive andefficient. At present, the rate of payment to theStates for individual copies of vital events is 3 centsper record. This is somewhat more than the unitcost of making a microfiim transcript. The unitcost for processing the record, which includes cod-ing, punching, verification, correction, ad super-vision (all stages up to tabulation) is about 3 cents.Thus, the total cost of getting the birth record readyfor tabulation is a fraction under 6 cents per record.The low costs are made possible by the large vol-ume of records (about 4,000,000 annuaily) processedin a continuous and efficient operation. The volumeis sufficient to require the fuil-time attention of

specialized clerks, continuous and expert super-vision, and the development of exact productionrecords and controls.

The punched-card method

The experimental plan initiated in 1951 forbirths involved not only the preparation of punchedcards by the State vital statistics offices for theuse of the NOVS, but also the shipment to the NOVSof microfilm copies of ail records from which thecards were punched. By this arrangement, theNOVS was able to verify the results of the Stateoperations of coding, punching, and reproducing thecards, and also to make any necessary corrections.

C&-Cost estimates of the punched-card

method should not be related to the total cost to theStates for the preparation of the punched cards butonly to the extra cost to the States preparing punchedcards as required by the NOVS. This includes the

cost of coding and punching items required by theNOVS which a State does not need for its own useas well as the cost of reproducing the punched cardsfor the NOVS. Counterbalancing these extra coststo the States, are the savings to the NOVS in codingthe data and punching the cards, together with su-pervising these actual operations.

As yet, the NOVS has not reimbursed all ofthe States that have submitted punched cards. Es-timates of the extra cost vary widely. The Sub-committee on Compensation to States for Copiesof Vital Statistics Data of the Association of Stateand Territorial Health Officers recommended thatStates be reimbursed for punched cards at the rateof $7.5o per 1,000 cards. To this must be addedthe cost of the card stock, shipping charges, veri-fication, correction of errors, and the field znd ad-ministrative work necessary to maintain the sys-tem. The best estimate available to this Subcom-mittee on these extra costs is about $17.00 per1,000. The estimated total cost to the NOVS forpunched cards prepared by the States would beabout $24.50 per 1,000 compared to about $31.00per 1,000 punched cards prepared by the NOVS.In 1954 the punched cards furnished the NOVS bythe States represented slightly less than one-fourthof the total birth records filed in the United States.Assuming that the number of State -prepared punchedcards couid be increased to 1,000,000, the esti-mated annual savings to the NOVS would be about$6,500 during the years when all births are tabu-lated. The estimated annual savings would be cutapproximately one -half for the years when the 50-percent sample of births is tabuiated.

Advantages .—The decentralized punched-cardmethod may raise the general level of quality ofstatistics produced by the State vital statistics of-fices. Each State preparing to enter the punched-card program necessarily must make a careful,detailed review of procedures in order to deter-mine if they conform to the requirements of theNOVS. At present, many States do not have written

detaiied procedures. Participation in the programmay result, as some States have already found, iRmore efficient processing procedures. However,this increase in efficiency is a by-product that canbe secured in other ways. Another advantage isthe improved consistency and comparability re-sulting from the continuous process of verificationof the State punched cards by the NOW.

260 VITAL STATISTICS-SPECIAL REPORTS

Disadvantages to the States. —The national re-quirements for data place the individual State vitalstatistics offices in a “strait - jacket” in terms ofthe classifications used and, in certain instances,in terms of the scope of what a State can do in itsown program. Indeed, some States have declined toenter the plan for this reason. It is already clearthat the uniform geographic code developed underthe “minimum core” program has increased thecost of coding and tabulation to both the NOVS andthe States. In order to accommodate the needs ofboth, a 7-digit geographic code is required. Thisis one digit more than is needed by the NOVS forits national tabulations; and 3 or 4 digits more thanis usuaily needed by the States for their own pur-poses. Moreover, the national system of allocationof geographic residence requires each State toclassify nonresident events according to the de-tailed geographic subdivisions of other States withwhich it is not familiar and in which it has no in-terest. Another example of the rigidity of the sys-tem arises when a State is using its punched cardto full capacity for its own program. Such a Statehas the alternative of not accommodating fully itsown program needs or of preparing a second punchedcard. A second punched card may be advantageousbut the total cost obviously is not reduced.

Disadvantages to the production of nationalvital statistics .—The punched-card method has thefollowing defects insofar as the production of highquality national vital statistics is concerned.

a. It is relatively inflexible. At no time hasthe NOVS ever completely tabuiated all ofthe information entered on the original birthand death records or even all of the codedinformation on punched cards. The maxi-mum utilization of the available data canbe obtained most efficiently by a cyclicaltabulation plan whereby certain items arecoded and tabulated annually while otheritems are coded and tabuiated at periodicintervals. In addition, demands for pre-viously uncoded or untabulated data ariseas new health programs are developed, andmore socioeconomic problems requirestudy. To introduce changes in coding spec-ifications, especially those requiring modi-fication of or additional columns on thepunched card, would be regarded as im-

practicable by many States. Changes incoding specifications often would requirethe training of coders in all States. In view

b.

c.

of the difficulties in making a change, it isalmost inevitable that the list of items codedas well as the details of the codes wouldshrink to the minimum level acceptable toall States cooperating in the program. Forexample, it is inconceivable that the cur-rent study of occupational and social classmortality now being made by the NOVS couldhave been undertaken if it had been neces-sary to depend upon the States to code andpunch the data.TheYe is no way of ensuring that a Statemay not suddenly withdraw from the plan.A State may find it impossible or undesir-able from its point of view to continue thepreparation of punched cards for nationaltabulations. Since the staffing and trainingof the processing unit of the NOVS are tiedto a budget prepared more than a year inadvamce, the withdrawal of a State wouldimpose an additional workload for whichthe NOVS would be unprepared either withrespect to budget or, more serious, wtthrespect to trained personnel.Loss of leadership. The continued highstanding of the national vital statistics sys -tern in the United States depends on thequality of leadership and technical skillsprovided by the NOVS. A complete con-version to the State-produced punched-cardsystem would result in a loss of direct con-tact with the original data and the loss ofcoding knowledge and skills. The NOVSwouid have less control over the scope ofthe national program for it could not intro-duce any changes beyond those acceptableto all cooperating States, many of whichmight have no interest in the informationdesired.

Modifications of punched-card methods. —Jnreviewing costs, the Advisory Committee on VitalStatistics Methods5 noted that so long as the NOVSpurchased microfilm copies of all records for veri-fication purposes, iittle or no savings would resultfrom the punched-card method. A significant re-duction in total cost would be possible only if the

5This committee, composed of State office rep-resentatives, was appointed by the NOVS in 1951 toassist in the study of methods by which the Statesand the NOVS might cooperate more efficiently inthe production of national vital statistics.

Vol. 45, No, 11 NATIONAL VITAL STATISTICS NEEDS 261

verification of State punched cards were on a sam-ple basis. Two methods of sample verification

have been proposed, one based on certificates fora fixed 3-month period and the other on a 10-per-cent random sample of certificates.

A substantial reduction in cost undoubtedly

would result from verification on a sample basis.However, this proposal provides only for meas-urement of error. Sample measurement of errorwithout specific provision for replacement of in-correct cards in the rejected lots is not quality con-

trol, Provision for such replacement then becomesa matter of negotiation with the States. Therefore,any sampling scheme of verification, such as theproposed 3-month sample or 10-percent randomsample, cannot yield a product of comparable qualityto that derived by the centralized microfilm proc -essing method or by a decentralized punched-cardmethod with access to all the individual records.

Copies of the original vital records are im-portant to the NOVS for many purposes. Amongthese are special investigations of relatively rarediseases. An example is furnished by the usesbeing made and to be made of. death certificates bythe National Institute of Neurological Diseases andBlindness to help estimate the prevalence of themany disabling neurolo~ical and sensory disorderssuch as multiple sclerosis, amyotrophic lateralsclerosis, and cerebral palsy; it is expected thatnatality information will be used in a proposed studyof mental retardation that may develop during preg-nancy or at birth. The National Cater Institute isusing copies of original death certificates in a study

currently in progress with the Railroad RetirementBoard and the Brotherhood of Railroad Trainmento trace cancer mortality among groups of workersin covered industries. The copies of the original

death certificates in the NOVS are being used inanother study by the National Cancer Institute basedupon a suggestion that Mexican women may havehigher death rates from cancer of the lung thanother white women. The basic records in the NOVShave provided, in addition, detailed information forstudies of selected types of accidents, poisonings,and deaths from new drugs; the last is of consider-able interest to the Food and Drug Administration.The original records have also been used for aspecial study of the cause-of-death classificationsystem and will be used for a study of multiple

causes of death. The National Heart Institute isinterested in the latter in connection with its stud-ies of cardiovascular disease.

From these examples, it is evident that thecentral collection of copies of individual vital rec -

orals can significantly assist medical and publichealth investigations. The cost of maintaining sucha file in the NOVS is small compared with the. healthneeds served and the. investment in those needs.

Opportunity for wide variety of detailed studies fromindividual punched cards and copies of original rec -o rds has given the NOVS a knowledge of coding andtabulation problems to which is largely due itspresent influence in promoting improved vital sta-

tistics, both nationally and internationally.Conclusions regarding the punched-card meth-

0>. —Although the punched-card method was startedin 1951, the Committee has been unable to obtainreliable figures, based upon broad experience, fromwhich it could determine whether or not a signifi-cant decrease in the cost of preparing national birthstatistics has resulted. Based upon the rate ofreimbursement recommended by the Subcommitteeof the Association of State and Territorial HealthOfficers, the estimated annual savings might be$6,500 as the program is now operated. This sumwould be considerably larger if verification werelimited to a 10-percent sample and no provisionwere made for the correction of errors. This po-tential saving would be false economy since thevalue and efficiency of national vital statistics wouldbe impaired if the NOVS did not receive a copy ofthe original vital record and if complete qualitycontrol were abandoned.

H cost is to be the decisive factor in determin-ing the method of preparing national vital statistics,then consideration should be given to the centrali-zation in the NOVS of the editing, coding, punching,and tabulation of all vital records for both Stateand Federal uses. The utilization of modern, high-

speed electronic equipment made economical by acentralized operation in combination with the ad-vantages of mass production almost certainly wouldyield more extensive and detailed statistics for

both the States and the NOVS at a substantiallysmaller total cost than the decentralized productionof punched cards by the States.

Although cost is important, the differences inthe amounts of money involved by the adoption ofalternative methods are not large enough to be de-cisive. The decisive factor is the effect of alter-native methods upon the scope, qwW~ cont~uity~

and usefulness of mtional and State vital statis-tics. The Committee has noted a number of seri-ous disadvantages of the punched-card “method when

262 VITAL STATISTICS-SPECIAL REPORTS

judged by this criteriow disadvantages that far out-weigh monetary considerations. It has been unableto discover any advantages that cannot be realizedby methods that are free from the defects of thepunched-card program,

The Committee can foresee no way to obviatethe defects of the punched-card method in the pro-curement of mortality statistics. The decision isnot so clear cut with respect to birth statistics.The preparation of birth punched cards is largelya keypunch code operation. The advantages of acyclical program of processing are not so great.The Committee recognizes that it may be possiblefor the NOVS ta work out a mutually advantageouspunched-card program with a fewState or city vitalstatistics offices. If undertaken, this should bedone on aw individual basis anti only after appro-priate safeguards against the disadvantages notedabove. The Committee sees no possibility of ex-tending this program to include all State and cityvital statistics offices without jeopardizing thescope, quality, and usefulness of national vital sta-tistics.

There has been no experience with the use ofprepunched cards for marriages and divorces.However, the statistical processing problems forthese subjects are similar to those for births.

The pretskndation method

In addition to the essentials of the pretabulatedmethod, as described earlier, there is the require-ment that each State, conforming to nationwide uni-form specifications,, edit snd code the individualvital records and punch a card with standard items.The experience of every developed country of theworld indicates that national vital statistics havebeen prepared effectively only when the responsiblecentral off ice, or the branch offices under its directadministrative control, work from an individualrecord for each event.

Where the subnational offices are not admin-istratively dependent upon the national office, andthe national office is dependent on pretabulateddata, the available national statistics are highlydeficient, and characterized by a paucity of cross-classifications, irregularity of publication, and arigidity of types of information available.

The pretabulation method will be consideredfrom the point of view of the basic requirementsfor national vital statistics specified on page 231 ofthis report, efficiency of administration, and cost.

Meeting the basic requirements for nationalvital statistics. —The adventages and disadvantagesof the punched-card method previously mentionedare equally applicable to the pretabulation method,Only the additional disadvantages of the latter meth-od will be mentiQned here.

a,

b,

c.

The pretabulation method is even more in-flexible than thepunched-card system. Sincethe tabulations would be pre~red for na-tional purposes, they would not meet the di-rect operating needs of the States and hencewould represent extra, unproductive work.The tendency to reduce the tabulations to aminimum rigid list would be difficult toovercome.The NOVS would lose control of the qualityof national vital statistics. Without accesseither to original records or to individualpunched cards, the NOVS would have nomeans of ascertaining andl controlling theaccuracy of the pretabulated data, More-over, since the States could not directlyuse much of the data for their own needsthere would be little incentive for them tospend much time and effort in maintaininga system of quality control,National vital statistics would become lessuseful rather than more useful. It is im-possible to foresee all of the useful pur-poses which national vital statistics mayserve and to devise a tabulation programto provide the desired information. Con-sequently, the NOVS continually uses theindividual punched cards to answer requestsfor information not provided by the regu-lar annual tabulations from other Federalagencies, the Congress, and other publicand private organizations and individuals,The retabulation of mortality punched cardsfor 1949-51 now under way is an example ofthis use. The last Congress authorized thePublic Health Service to undertake studiesof the health hazards of air pollution. Asone part of this program, the individualpunched cards for 1949-51 are being retab-ulated in order to obtain mortality data forresidents of the central part of large met-ropolitan areas, for residents of the adja-cent densely populated areas, and for resi-dents of nonmetropolitan areas. The tabu-lation is also being done in a manner to

Vol. 45, No. 11 NATIONAL VITAL STATISTICS NEEDS 263

permit the analysis of geographic varia-tions of mortality and of variations betweenvarious economic regions of the UnitedStates. Studies of this nature could not bedone if the NOVS did not have individualpunched cards available. This is only oneof many special studies requiring the useof individual punched cards that have beencarried out. Requests for such studies willcontinue to be received with the expansionof research into health problems.

d. Some of the defects of the pretabulationmethod are exemplified by the u~e of thismethod to collect statistics of marriage anddivorce. The method was adopted by theNOVS in order to start the development ofnational data concerning marriage and di-vorce. So far, however, no consistent bodyof data is available either for a specifiedgroup of States or for the United States as awhole. There is no uniformity of definitions,of classification of items tabulated, nor ofamount of detail tabulated; and the numberof States for which data can be publishedvaries from item to item, Although the datamade available by the pretabulation methodare better than no data at all, this is not anappropriate standard for judging the nationalvital statistics of the United States. Withrespect to marriage and divorce statistics,the United States ranks among the less de-veloped nations of the world.

Efficiency of administration. -The pretabu-lation method is administratively cumbersome.The timeliness of the national tabulations woulddepend entirely upon the timetable of the slow-est State. If, for any reason, a State should wish towtthdraw, the NOVS could not easily reinstate theprior system since it would have lost its budget,personnel, and administrative machinery. Whenmodification in the pretabulation method is necess-itated by new needs, the NOVS would be required

to enter into time-consuming negotiations with 56vital statistics off ices.

~. —The individual States would incur avery large increase in cost since none edits, codes,punches, and tabulates for its own needs in the de-tail required for national purposes. Each Statewould be required to make tabulations or to preparesummary punched cards not only for all vital eventswithin the State, but it also would be required to re-peat this procedure for the vital events among non-residents, preparing summary cards for each Stateof residence separately, This additional detailwouldbe needed by the NOVS since its publications pre-

sent State data on the basis of place of residence..

At present, the allocation of vital events to place ofresidence is made by the NOVS.

Since pretabulation for national vital statistics

would entail a large increase in cost to the individ-

ual States over their own requirements, they un -

doubtedly wou?d expect to be reimbursed by the

NOVS. On the other hsnd, the NOVS would bear thecost of consolidating the tabulations or summarypunched cards of the 56 vital statistics offices. Ifany of these offices should send its data in tabulatedform, then the NOVS would almost certainly preparesummary punched cards from them. Thus, if therewere savings in cost to the NOVS by pretabulation,they would be offset by the reimbursement it wouldhave to make to the States.

Conclusions concerning pretabulation. —The.—pretabulation method has disadvantages of its ownas well as those inherent in the punched-card method.Although precise cost data are not available, it isclear that pretabulation would add to th~ costs ofState operations without a compensatory decreasein the costs to the NOVS. National studies based oninformation obtainable only from copies of the in-dividual records or from individual punched cardswould be impossible. The available national vitalstatistics would shrink to those provided by a mini-mum rigid list of tabulations. This method shouldnot be considered further by the NOVS.

Table C-1. BIRTR STATISTICS PUBLISHED IN STATE REPORTS 42 STATES, SPECIFIED YEARS

I—

#*

;z !— —

29

9 x

9 x

9 x. —

9 x

8 x

8 x

8 x

8 x

7 x

7 x

6 x

6 x

6 x

6 x

5 x

5 x

5 x

5

5 x

5 x

4 x

4 x

=

j

22!

—6.

x—

x—

x—

—

.

x

—

x

—

—

—

—

x—

—

—

—

—

—

—

COONc

Awlr

A!ITEllmm Al! iIJMITIMAT2 ?E2uc;Esw

d:

2

BI.

31RTs —

ii—6

—

STATS Year

F!—

T-------------------------------

Connecticut----------------------------

New York-------------------------------

Washington-----------------------------

Wisconsin------------------------------

Nebraska-------------------------------

North Carolina-------------------------

south Ca?ml.ina-------------------------

Vemwnt --------------------------------

Michigan-------------------------------

Vimda -------------------------------

Ma fionti-------------------------------

New mti~--------------------------

Oregon---------------------------------

West Virginia --------------------------

Imisicula------------------------------

Mississippi----------------------------

Newda --------------------------------

New Mtico-------------—--------------

Oklahoma-------------------------------

Pen?lBylmulia---------------------------

FlOrlaa-------------------------------

Idaho—-----------------------—-------

8 17

—

16.

x

x—

x

15

x

x

x

12.

x—

x—

x

x

12—

—

8 6 3— — —1950

1949

1952

1951

1952

4J- X

x—

x—

—

—

x—— —

i=

x x

x x

x.

—

—x— .

x

—c—

. —x

—

—

—

—

—

—

—

—

—

.

—

—

—

—

—

-P- X.

x—

x— — —

Ex x

x

x x

x x

x—x

x

x

x

x x—x—1951

1951

x

x

x—x

— —x x— — —

1952

1951

1951

x

x

x

x

x—

—

x— .x—

— — —x

x x

x

x x

x

x

x

x x

x

x x

x

x x

x

x—x— x x— .

x— —

194s

1951

x— x x— —xx

—

x—

—

x—

—x.

—

—

—x

x— — —

950-51 x — —1950

1952

x — — —x x x—

—

—1952

949-50

x— — — —x—

.

x

x

—

— —x.

— — —1950

+

1952

951-52

x x—x—

— —x— —

—

— —x— —

--lL951

1952

x— —x—

x— — — — —

x— . — — — —

Table C-1. BIRTH STATISTICS PUBIJSHED IN STATE RE~RTS 42 STATES, SPECIFIED YEARS-Continued

Year

—

i!s

2F!

—

—

x.x—

.

—

.

—

.

.

—

—:Omm m

CITYk!rmmm AT

IuRmmmicixm—

38

OFOli—

ii%

ia

— — —

~ii

x

x

.

x—

x

—

—

—

sTA!m

—

x—

—

—x—x—

—

x—x—x—

—

—x

—x

.

—.

—

—

$x

x

x

Iowa---------------— -----------------

-8---— ---------------------------

KentucW—-----------------------------

M9ssachuaetts--------------------—----

Minnesota ------------------------------

l!ememee ------------------------------

(korgia-------— -----------------------

Indiena --------------------------------

Maine-------—----------— -------------

New Jersey -----------------------------

Ohio---------------------—-----------

south Dakota--------------— ----------

Alabama-------------------------------

A1-lzona-------------------------------

Arkansas -----------------------— -------

mfornia -----------------------------

Delaware -------------------------------

!lexm ----------------------------------

Illinois -------------------------------

Maryland-------------—----------------

%mlsioml data.

1952 4 x

x— —

19531

1951

4

4

x

x

x x— —

1951

1952

4—

x4 . —x

x

— —1951 4 x x — — .1951

1952

1951

3

3

3 x

x

x x— —x

x—

—1949

1950

3—

3*

x x—

. —,

+

x1952

19521

1952

3—

2—

2

x—x—x—

x—

.

— —

1949 2 x — — —1950

L949-50

1949

2—

2—

2

x

x

— —x— I

x——x

— —1952 1—19531 1 x

— — —

Table C-2. DEATH STATISTICS PUBLISHED IN STATE REPORTS 43 STATES, SPECIFIED YEARS N

INF’4h’l D2Am7s

Year

TwJ!ALD2Ams

t I t I

$

~ ;

~ g

33

2 x— .

3 x

2 x

2— —

2 P

2 F

.2

2 F

1 F

1 P

3 P

1

2 P

3 x

1 x— —

2 F

2 .@

3 x

2 x

1

1

‘!g $

;

8 $1.3

Il. x

II

10

9 x

9 x

9 x

8

8

8 x

7 x

7

7

7 x

-1

7

7 x

7 x

7

7 x

7

6

CaUnty I C8ue.e of deathLedtng c1

of deal

~

3j ‘!

$ ~ $2 8 8 ii

10 20 5 6

x x x

x x x

x x

x

x

x x

x x

x x

x x

x

x

x

x

x .

x

x

x

x

x x

;es I

STAm

--------------

Connect icut-------------

New York----------------

New Jersey-— -----------

California--------------

North carOlina----------

Virginia----------------

14msachusetts-----------

Oklahoma----------------

south carOlixla----------

319 19

x x

x x

x x

x x

26—

x

x?

16—1950

1949 x x—x1949 x

x—

x

x x

x x

x

x

x

x #

x x

x

x

x

x x

x x

xI

1950

1951

x x

d- X x—1951

1951

+

x

x

x

x

x

x

x—x

—x—

1952

1951

1949

x x x

x x x x

x x

x x x

x x x

x= x x

x x x

x

x x x

x x

x x x

x x x

x x x

x x x x

x x—

—x xArke.!lsas ----------------

Plorma-----------------

Kentucky----------------

Louisiana---------------

Louisiana--------------Michigan----------------

Minnesota---------------

Mississippi-------------

14isclouri----------------

Nw sampEMre ----------

Ohio--------------------

Wimmldn---------------

Ihlne-—--— -----------

1951

1951

x—

— x

1952

1951

1952

x

=

x x

x x

x

x x

x x—1952

1948

—x x=—

5=x

x x

x

x

x— x

x

1951 x—1950 —

—x+

x x x1951 —1951

I

‘Sex ad age. ‘Cause of death by age and race. ‘County - city date by race. 4Pace and age. 5Sme, sex, end age.

Table C-2. DEATH STATWITCS PUBLISHED IN STATE REPORT& 43 STATES,SPECIFIED YEARS-Coatinued eo5

Iwmn! DxA!cm.%

Yea

‘ml!ALIEA!cF5

:j g.)-+

$!!

8 $’6

6

6 x

5

5

5 x

5

5

5

4—

4 x

4

4

4

4

3

3

2

2

2

2 ,

1

Cause or daatbcounty—

$’

—

i!—

—

;

r!—x—

fj5!—

x—

—

Nebrask&---------------

&uth DakOti----------

Teonemee-—-----------

Delaware---------------

Idaho-------------------

Nevada--------—--------

0re60n-------—-----

WMrin@On------------

West Virginia----------

Arizona----------------

In&iana----------------

Iowa--------------------

Hew lkxlm -------------

mXaa----------—----

Vemcmt-----------------

GcOrgla-----------------

Pennsylvania------------

)d,nbma-----------------

IUinOis----------------”

Canaan------------------

IJm’yle.na---------------

Districtof Columbia----

1952

1952

1951

1250

i252

949-50

x—x—x—x—x

— — —x—x—x—x—x—

x x—x

— —x x—

—x—x

—x—x

—

—x x—

—

—

x—x—

—

x—x—x

—

—X351 x x x x

x—

1952 Xe—

x x—

—x

—

—x

—1950 x x x x— — —

x1952

1952

1952

1950

1949

x—x

— —

—

—

—

—

—

—

—

—

—

—x—x

x

x— x x—x—

—

x x x— — — —x x x— x—

x— —

—

—

—

—

1.252 x— x— x— —x x1.651 x—

—

—

x—

—

—

— —x

—1952 x

x

x—x

—19526

1952

x — — —x— — — —

x—x—

—

1953S

19536

1950

x—

—

—

—

—

—

— — —x —

—x

— — —%unty ad citydata by r8Cr3. ‘Race, sex, and age. %.mtisiom.l data.

APPENDIx D

Technical Services to States

The primary authorized function of the NationalOffice of Vital Statistics is the collection of copiesof vital records from the States and the productionfrom these records of national vital statistics meet-ing national requirements. In consequence, firstpriority in the NOVS program for providing techni-cal assistance to the States should be given to activ-ities related to this function. More specifically,emphasis should be laid on projects which

1. Improve the quality of original data both incompleteness and in content.

2. Keep the basic information collected inline with that required in the national interest.

3. Decrease the time lag between the eventand the publication of reports.

4. Increhse the utility of national vital statis-tics,

In carrying out these, technical assis~nceshould concentrate heavily on the problems of dataimprovement and analysis. At the same time, theassistance program could very well touch onlylightly on the operations aspects of records man-agement as they relate to legal amendments andother aspects of the registration office that arelargely custodial in nature.

In recent years, the subject-matter specialistsin the NOVS have been used extensively for con-sultative aid to States. This is largely because gen-eral consultants have been in short supply, budgetshave been small, and needs have changed. Thisshould not be interpreted as implying that NOVSshould no longer use general consultants. SomeStates have indicated that they feel routine visitsby general consultants wouldbe helpful. An impor-tant advantage to the NOVS of a routine visitingplan would be a better understanding of the limi-

268

tations of the basic data which would make possiblea more informative analysis of national statistics.

Types of services

The meaiis which may be used for providingtechnical assistance include professional advisoryand consultant service on technical and manage-ment problems; personnel development through in-service training institutes; expanded informationand clearinghouse service on technical subjectsand methodology; national educational programsdirected toward improvement of source data; anda national program for recruiting and training pro-fessional vital statistics personnel. The primarysource of this technical assistance should be theNational Office of Vital Statistics, which, however,should utilize and make available to State and localoffices trained personnel from other Federal agen-cies and from universities and foundations.

To provide more adequate technical assistance,the National Office of Vital Statistics should be giv-en sufficient fiscal reswrces to make a specificbudget allocation for the purpose.

Consultative service on technical problems .-Many vital statistics offices have indicated theirdesire for help in dealing with the increased sta-tistical demands upon them. In particular, sev-eral would like assistance in special studies uti-lizing sampling methods. Others feel they couldprofit from consultation on the planning of proce-dures and reports. Interest has been expressed ingetting aid to set up the registration system andoffice procedures for a new function such as mar-riage and divorce statistics. The State registrarsare usually fully occupied in keeping abreast oftheir current work and find it very difficult to si-

Vol. 45, No. 11 NATIONAL VITAL STATISTICS NEEDS 269

multaneously develop a new program, even whenthey are fully conversant with the techniques andfacilities required for the new program. It has beensuggested that in addition to consultative servicefrom the NOVS personnel, provision be made forthe exchange of technicians between States, underarrangements similar to the international assist-ante programs.

In-service training institutes. —Since 1949,the National Office of Vital Statistics has conductedseveral 1-week institutes for training cause-of-death coders in different States. These instituteshave been very successful and particularly well re-ceived, Similar institutes on other technical prob-lems would undoubtedly be of great aid to the Stateand local vital statistics offices and confer bene -fits on the national vital statistics. Aid from theprogram divisions of the Public Health Servicemight be arranged on special topics, such as ac -cident statistics, cancer statistics, fetal mortality,natality statistics, et cetera. Population estimationpresents a difficult problem for most State and lo-cal areas, and could very well be the subject of in-stitutes conduct ed in cooperation with the Bureauof the Census.

Clearinghouse and information service.—Inthe past, the National Office of Vital Statistics hascollected a large amount of technical information anddisseminated it to vital statisticians. Rapid tech-nological change and fast moving developments inpublic health have increased the difficulty of keep-ing up to date. But the NOVS is handicapped in de-veloping a better organized information service bya shortage of personnel. To some extent T-Registrar and Vital Statistics—Special Reportsmeet this need for technical information. However,the concensus of registrars is that these are notsufficient, and the solution would be to develop aseries of technical bulletins similar to those usedby large industrial and business concerns to keeptheir widely spread organizational units informedon technical matters.

Leadership of national programs for improvingstatistics and their utilization. —The States and 10-cal areas have great difficulty in doing much withthe medical profession, medical examiners, andcoroners, toward improving cause -of -death cer -tification. Few, if any, of the States have expertsto draw upon for this purpose. Even the medicalpersonnel of the health department are, for the mostpart, not well informed on cause-of-death clas-sification. This is a national and international

problem which merits the full-time attention of atleast one statistician with medical training in thenational vital statistics agency.

Work in morbidity statistics—hospital, clinic,survey, and notifiable disease —is so closely re-lated to mortality statistics that better coordina-tion should be brought about. This represents anarea where the States and local areas need help,but which requires personnel at present unavail-able in the NOVS. More leadership and aid to Stateson demographic problems is desirable., Much couldbe done by working out a closer relationship be-tween the vital statisticians in State and local areasand demographers. The recent regional studies ofrural-urban migration, undertaken under the aus -pices of the U. S. Department of Agriculture, raninto difficulty because of the deficiencies in Statevital statistics data in all but a few areas. TheStates are interested in such matters, but are illequipped to deal with them.

Advisory assistance on management prob-lems, —Several States have drawn upon the NOVSfor aid in studying the organization, ‘management,and administration of their vital and health sta-tistics services. Independent help in drawing upproposals for improved statistical services canbe an invaluable aid to the State or local depart-ments.

Technical assistance in the national vital sta-tistics system should be directed toward helpingState and local vital statistics offices maintain andimprove the production and utilization of vital sta-tistics. National vital statistics benefit not onlyfrom direct success in data improvement but alsofrom effective utilization of vital statistics at Stateand local levels. In particular special studies whichcorrelate vital statistics with other health and so-cial data are assuming ever increasing importanceand are in strong demand. Ways should be soughtof furthering the more effective use of vital sta-tistics data in relationto program planning of Stateand local agencies. This calls for study of the fun-damental problem of the utilization of vital sta-tistics data in health department policy making andadministration. For the most part, State and localoffices lack the resources to do much more thanto carry on the present vital statistics job, whichin many areas is rather perfunctory. Technicalassistance of high quality should be made availableto the States to improve the personnel, skills, pro-duction, operations, quality of original data, anddata utilization.

THE UNITED STATES NATIONAL COMMITTEE ON VITAL AND HEALTH STATISTICS

Lowell J. Reed, Ph. D., Chairman,

President, The Johns Hopkins University,

Baltimore, Md.

Philip M. Hauser, Ph. D.,Professor and Chairman,Department of Sociology,

University of Chicago,Chicago, Ill.

Vice Chairman,

I. M. Moriyama, Ph. D., Secretary,Chief, Mortality Analysis Section,National Office of Vital Statistics,Public Health Service

Department of Health, Education, and Welfare,Washington, D. C.

Edwin L. Crosby, M. D.,

Director, American Hospital Assmiation,Chicago, Ill.

Edwin F. Daily, M. D.,Deputy Medical Director,Health Insurance Plan of Greater New York,New York, N. Y.

Paul M. Densen, D. Sc.,Director, Research and Statistics,Health Insurance Plan of Greater New York,New York, N. Y.

Harold F. Dorn, Ph. D.,

Chief, Biometrics Branch,National Institutes of Health,Public Health Service,Department of Health, Education, and Welfare,Bethesda, Md.

Halbert L. Dunn, M. D.,Chief, National Office of Vital Statistics,Public Health Service,Department of Health, Education, and Welfare,Washington, D. C.

Eugene L. Hamilton,Ch;ef, Medical Statistics Division,Office of the Surgeon General,Department of the Army,Washington, D. C.

John P. Hubbard, M. D.,Professor of Preventive Medicine,University of “Pennsylvania School of Medicine,Philadelphia, Pa.

Robert H. Hutcheson, M. D.,

Commissioner of Public Health,Tennessee Department of Public Health,Nashville. Term.

O. K. Sagen, Ph. D.,Chief, Bureau of Statistics,Illinois Department of Public Health,Springfield, Ill.

P. K. Whelpton,Director, Scripps Foundation for Research in

Population Problems,Miami University,Oxford, Ohio

270

The Subcommittee on National Vital Statistics Needs

William G. Cochran, Chairman,Professor of Biostatistics,School of Hygiene and Public Health,The Johns Hopkins University,Baltimore, Md.

Harold F. Dorn, Ph. D.,Chief, Biometrics Branch,National Institutes of Health,Public Health Service,Department of Health, Education, and Welfare,Bethesda, Md.

Forrest E. Linder, Ph. D.,Chief, Demographic and Social Statistics Branch,Statistical Office,United Nations, New York, N. Y.

O. K. Sagen, Ph. D.,

Chief, Bureau of Statistics,Illinois Department of Public Health,Springfield, Ill.

Mortimer Spiegelman,Associate Statistician,Metropolitan Life Insurance Company,New York, N. Y.

Robert F. Lenhart, P. K. Whelpton,Secretary, Director, Scripps Foundation for Research inCommittee for Economic Development, Population Problems,Washington, D. C. Miami University, Oxford, Ohio

The Subcommittee on Utilization of Marriage Statistics

Conrad Taeuber, Ph. D., Chairman, Charles B. Reeder, Ph. D.,

Assistant Director, Bureau of the Census, Associate Economist,

Department of Commerce, E. I. Dupont de Nemours and Company,

Washington, D. C. Wilmington, Del.

Ann Dillon,Director of Statistical Service,Tennessee Department of Public Health.Nashville, Term.

Clifford Kirkpatrick, Ph.Professor of Sociology,

University of Indiana,Bloomington, Ind.

Myron S. Silbert,

Vice President, Federated Department Stores, Inc.,New York, N. Y.

Louis Weiner,

Chief, National Income, Money Flows, andD., Labor Section,

Division of Research and Statistics,Board of Governors of the Federal Reserve System,Washington, D. C.

271

* U.S. GOVERNMENT PRIMT!NG OFFICE : 19650-783-884

OUTLINE OF REPORT SERIES FOR VITAL AND HEALTH STATISTICS

Public Health Service Publication No. 1000

SeYies 1.

Series 2.

Series 3.

Series 4.

Series 10.

SeYies 11.

Series 12.

Series 20.

Series 21.

Series 22.

Programs amt collection procedures. —Reports which describe the general programs of the NationalCenter for Health Statistics and its offices and divisions, data collection methods used, definitions, andother material necessary for understanding the data.

Reports number 1-4

Qata evaluation and methods veseavch. —Studies of new statistical methodology including: experimentaltests of new survey methods, studies of vital statistics collection methods, new analytical techniques,objective evaluations of reliability of collected data, contributions to statistical theory.

Reports number 1-10

Analytical studies. — Reports presenting analytical or interpretive studies based on vital and health sta-tistics, carrying the analysis further than the expository types of reports in the other series.

Reports number 1-4

Documents and committee veports.— Final reports of major committees concerned with vital and h@thstatistics, and documents such as recommended model vital registration laws and revised birth anddeath certificates.

Reports number 1 and 2

Data Fvom the Health Interview Su?wey. —Statistics on illness, accidental injuries, disability, use ofhospital, medical, dental, and other services, and other health-related topics. based on data collected ina continuing national household interview survey.

Reports number 1-22

Data From the Health Examinationmeasurement of national samples ofcific diseases, and distributions ofmeasurements.

Reports number 1-10

Survey .—Statistics based on the direct examination, testing, andthe population, including the medically defined prevalence of spe-the population with respect to various physical and physiological

Data From the Health Records Su~vey. —Statistics from records of hospital discharges and statisticsrelating to the health characteristics of persons in institutions, and on hospital, medical, nursing, andpersonal care received, based on national samples of establishments providing these services andsamples of the residents or patients.

Reports number 1 and 2

Data an movtality. —Various statistics on mortality other than as included in annual or monthly reports—special analyses by cause of death, age, and other demographic variables, also geographic and timeseries analyses.

No reports to date

Data on natality, marriage, and divorce. —Various statistics on natality, marriage, and divorce otherthan as included in annual or monthly reports —special analyses by demographic variables, also geo-graphic and time series analyses, studies of fertility.

Reports number 1-6

Data Fvom the National Natality and Mortality Surveys. —Statistics on characteristics of births anddeaths not available from the vital records, based on sample surveys stemming from these records,including such topics as mortality by socioeconomic class, medical experience in the last year of life,characteristics of pregnancy, etc.

Reports number 1

For a list of titles of reports published in these series, write to: National Center for HealthU.S. Public Health Service

Statistics

Washington, D.C. 20201