The Springer Series on Demographic Methods and Population Analysis 36 Yi Zeng · Kenneth C. Land Danan Gu · Zhenglian Wang Household and Living Arrangement Projections The Extended Cohort-Component Method and Applications to the U.S. and China Household and Living Arrangement Projections THE SPRINGER SERIES ON DEMOGRAPHIC METHODS AND POPULATION ANALYSIS Series Editor KENNETH C. LAND Duke University Inrecentdecades,therehasbeenarapiddevelopmentofdemographicmodelsandmethods and an explosive growth in the range of applications of population analysis. This series seekstoprovideapublicationoutletbothforhigh-qualitytextualandexpositorybookson modern techniques of demographic analysis and for works that present exemplary applications of suchtechniques tovarious aspectsof populationanalysis. Topicsappropriate fortheseries include: (cid:129) General demographic methods (cid:129) Techniques of standardization (cid:129) Life table modelsand methods (cid:129) Multistate and multiregional lifetables, analyses andprojections (cid:129) Demographic aspects ofbiostatistics and epidemiology (cid:129) Stable populationtheory and its extensions (cid:129) Methods ofindirect estimation (cid:129) Stochastic population models (cid:129) Event historyanalysis, duration analysis,and hazard regressionmodels (cid:129) Demographic projection methods andpopulation forecasts (cid:129) Techniquesofapplieddemographicanalysis,regionalandlocalpopulationestimatesand projections (cid:129) Methods ofestimation and projectionforbusiness andhealth care applications (cid:129) Methods andestimates forunique populationssuch asschoolsand students Volumes in the series are of interest to researchers, professionals, and students in demography,sociology,economics,statistics,geographyandregionalscience,publichealth and health care management, epidemiology, biostatistics, actuarial science, business, and related fields. For furthervolumes: http://www.springer.com/series/6449 Yi Zeng (cid:129) Kenneth C. Land (cid:129) Danan Gu Zhenglian Wang Household and Living Arrangement Projections The Extended Cohort-Component Method and Applications to the U.S. and China YiZeng KennethC.Land CenterforStudyofAgingandHuman DepartmentofSociologyand DevelopmentMedicalSchool CenterforPopulationHealthandAging DukeUniversity PopulationResearchInstitute Durham,NC,USA DukeUniversity Durham,NC,USA NationalSchoolofDevelopment CenterforHealthyAgingand ZhenglianWang DevelopmentStudies CenterforPopulationHealthandAging PekingUniversity PopulationResearchInstitute Beijing,China DukeUniversity Durham,NC,USA DananGu PopulationDivision HouseholdandConsumptionForecasting,Inc. UnitedNations ChapelHill,NC,USA NewYork,NY,USA ISSN1389-6784 ISBN978-90-481-8905-2 ISBN978-90-481-8906-9(eBook) DOI10.1007/978-90-481-8906-9 SpringerDordrechtHeidelbergNewYorkLondon LibraryofCongressControlNumber:2013954374 ©SpringerScience+BusinessMediaDordrecht2014 Thisworkissubjecttocopyright.AllrightsarereservedbythePublisher,whetherthewholeorpart of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation,broadcasting,reproductiononmicrofilmsorinanyotherphysicalway,andtransmissionor informationstorageandretrieval,electronicadaptation,computersoftware,orbysimilarordissimilar methodologynowknownorhereafterdeveloped.Exemptedfromthislegalreservationarebriefexcerpts inconnectionwithreviewsorscholarlyanalysisormaterialsuppliedspecificallyforthepurposeofbeing enteredandexecutedonacomputersystem,forexclusiveusebythepurchaserofthework.Duplication ofthispublicationorpartsthereofispermittedonlyundertheprovisionsoftheCopyrightLawofthe Publisher’s location, in its current version, and permission for use must always be obtained from Springer.PermissionsforusemaybeobtainedthroughRightsLinkattheCopyrightClearanceCenter. ViolationsareliabletoprosecutionundertherespectiveCopyrightLaw. The use of general descriptive names, registered names, trademarks, service marks, etc. in this publicationdoesnotimply,evenintheabsenceofaspecificstatement,thatsuchnamesareexempt fromtherelevantprotectivelawsandregulationsandthereforefreeforgeneraluse. While the advice and information in this book are believed to be true and accurate at the date of publication,neithertheauthorsnortheeditorsnorthepublishercanacceptanylegalresponsibilityfor anyerrorsoromissionsthatmaybemade.Thepublishermakesnowarranty,expressorimplied,with respecttothematerialcontainedherein. Printedonacid-freepaper SpringerispartofSpringerScience+BusinessMedia(www.springer.com) Preface This book presents an innovative demographic toolkit known as the ProFamy extended cohort-component method for the projection of household structures andlivingarrangementswithempiricalapplicationstotheUnitedStates,thelargest developed country, and China, the largest developing country. The ProFamy methodusesdemographicratesasinputstoprojectdetaileddistributionsofhouse- holdtypesandsizes,livingarrangementsofallhouseholdmembers,andpopulation by age, sex, race/ethnicity, and urban/rural residence at national, sub-national, or smallarealevels.Itcanalsoprojectelderlycareneedsandcosts,pensiondeficits, andhouseholdconsumption.Thebookconsistsoffourparts.Thefirstpartpresents themethodology,data,estimationissues,andempiricalassessments.Thenexttwo parts present applications in the United States (Part II) and China (Part III), concerningdemographic,social,economic,andbusinessresearch;policyanalysis, including forecasting future trends of household type/size, elderly living arrangements,disability,andhome-basedcarecosts;andhouseholdconsumptions, including housing and vehicles. The fourth part includes a user’s guide for the ProFamy software to project households, living arrangements, and home-based consumptions. TheveryinitialideaoftheresearchpresentedinthisbookbeganwhenIwasa Ph.D.studentatBrusselsFreeUniversityandattendedtheInternationalUnionfor ScientificStudiesofPopulation(IUSSP)1983familydemographyseminar.Atthat seminar, I was especially interested in a paper presented by Professor John Bongaarts on “The projection of family composition over the life course with the familystatuslifetable.”Withstrongsupportfrommysupervisors,ProfessorsFrans WillekensandRonLesthaeghe,andstimulatedbyProfessorJohnBongaarts’initial nuclearfamilystatuslifetablemodel,IconductedmyPh.D.thesisresearchatthe Netherlands Interdisciplinary Demographic Institute (NIDI) in 1984–1986 to develop a general family status life table modelincludingboth nuclear andthree- generation families, with an empirical application to China. During my study at NIDI, I learned a great deal of multistate demography from Professor Frans Willeken, who not only supervised my Ph.D study but also helped my long-term professionalcareerdevelopmentincludingworkonthisbook.AsaFrankNotestein v vi Preface Post-doctoralFellowin1986–1987,Ifurtherstudiedthisdemographictopicatthe Office of Population Research at Princeton University, under the supervision of Professors Ansley Coale and Jane Menken. My research at Brussels Free Univer- sity,NIDI,andPrincetonUniversityenabledmetowinthePopulationAssociation of America 1987 Dorothy Thomas Award, and my paper on “Changes in Family Structure inChina: ASimulation Study” was published inPopulation and Devel- opment Review (Zeng 1986). I greatly appreciate what I learned from Professors Willekens,Lesthaeghe,Coale,Menken,andBongaartsduringmyPh.D.andpost- doctoralstudies,knowledgewhichledtothenewresearchreportedinthisbook. After I became a faculty member at the Institute of Population Research at PekingUniversityinAugust1987,Icontinuedmyresearchinfamilydemography andtriedtoexpandittohealthyagingandpopulationpolicyanalysis.Iparticularly enjoyedandlearnedalotfromlong-termproductivecollaborationswithProfessor James Vaupel, who I first got to know and started to collaborate with at the International Institute of Applied Systems Analysis in 1985, a summer young scientists’programwhichwassupervisedbyProfessorsNancyKeyfitzandJames Vaupel. Among many collaborative research projects and co-authored peer- reviewedpublishedpaperswithJamessince1985,ourjointarticles(Zeng,Vaupel, andWang1997,1998)aremostrelevanttothisbook.Inour1997–1998papers,we initiallydevelopedthehouseholdandlivingarrangementprojectionmodelknown as ProFamy, which includes two sexes and time-varying demographic rates as input,andappliedittoChina,basedontheone-sexfamilystatuslifetablemodels (withconstantdemographicrates)ofBongaarts(1987)andZeng(1986)mentioned above. In a 1992 interview entitled “Talk to Demographer: Chinese demography startstogototheworld”publishedinthejournalPopulationbyFudanUniversityin Shanghai,Itoldtheinterviewerwhowasoneofthejournal’seditors:“Ilearneda lotfrommycollaborator JamesVaupel;hismindislikeacomputer.WhenIvisit him every year, we start to discuss research immediately after saying hello and shakinghands.Sometimeshewouldbedrivingbutcontinuingtotalkwithmeabout themathematicaldemographicformulasandcomputerprogramsofourjointpapers (whichwasoneofthepriorresearchbasesofthisbook),andIhadtoremindhimto becarefultoavoidacaraccident.IamsurethatmanyChinesepeopleweredeeply impressedfromProfessorJamesVaupel’sstyleofscientificresearchbyreadingthis publishedinterviewreport. Sincelate1998,IhavebeenafacultymemberatDukeUniversity,whileIkeep my faculty position at Peking University and divide my work time between these twouniversitiesacrosstheocean.ProfessorKenLand,whoisthesecondauthorof thisbook,pickedmeupattheairportwhenIfirstcametoDuke,anoccasionthatI will remember for life. Since then we began to frequently discuss our work and closely collaborate on research. Our joint work includes substantially extending and developing the ProFamy extended cohort-component model, software, and applications to the United States and China. Ken’s extremely broad mind and strongmathematical, statistical,anddemographicskillshavehelpedmealot.For example, he innovatively developed and summarized the basic procedures of the ProFamymodelintofour“coreideas”(refertoSect.2.2ofChap.2ofthisbook), Preface vii which are remarkably helpful for readers to easily understand the relatively com- plicated structure of the new method. He presented these four “core ideas” at the American Statistical Association 2006 Annual Meeting, a presentation that was wellreceivedbystatisticalanddemographiccolleaguesatthemeeting.Duringthe rather long period of preparing the manuscript of this book, Ken repeatedly reminded me that “We need to take our time to work this manuscript over and overagain,sothatitisaproductofthehighqualitythatwewant.”Clearly,Ken’s solidscientificresearchstyleandknowledgesignificantlycontributedtothemerit ofthisbook. I am of course very grateful for the outstanding contributions of the other two co-authorsofthisbook, Dr. Danan Gu andDr. Zhenglian Wang; thisbook would never have been produced without their hard work and close collaborations. We would like to sincerely thank Dr. Jessica Sautter, who very carefully edited, questioned, and commented on the entire manuscript to help us to significantly improve the quality of this book. We also would like to thank the supports from DukeUniversity,PekingUniversity,andtheMaxPlankInstituteforDemographic Research,andourcolleaguesattheseinstitutionslistedintheAcknowledgments. ProfessorofDukeUniversityand YiZeng PekingUniversity Acknowledgments TheresearchreportedinthisbookwasmainlysupportedbyNIA/NIHSBIRPhaseI and Phase II project grants, which received the best possible evaluation score (zero percentile rank), and China National Natural Science Foundation grants (71110107025 and 71233001). We also thank the Population Division of U.S. Census Bureau, NIA (grant 1R03AG18647-1A1), NICHD (grant 5 R01 HD41042-03), Duke University, Peking University, and the Max Planck Institute forDemographicResearchforsupportingrelatedbasicandappliedresearch. Wesincerelythankourcolleagueswhocontributedtoourpreviouscollaborative research or supported our research related to this book, including the following researchers (listed alphabetically): John Bongaarts, H. Booth, O’Neil C. Brain, Shuai Cao, Huaming Chai, Huashuai Chen, Lingguo Cheng, Simon Choi, Ansley Coale, Harvey J. Cohen, Ted Chu, Paul Demeny, Zhiheng Ding, Qiushi Feng, Changqing Gan, Ning Gao, Linda K. George, Daniel Goodkind, Todd Graham, Zhigang Guo, Helin Huang, Angang Hu, Nicole Keilman, Nancy Keyfitz, Cheng Jiang,LeiwenJiang,WeipingJiang,RonLee,RonLesthaeghe,ChunhuaLi,Faju Li, Jianxin Li, Lan Li, Ling Li, Qiang Li, Shuang Liang, Zhiwu Liang, Nan Lin, Justin Yifu Lin, Yuzhi Liu, Jiehua Lu, Wolfgang Lutz, Andrew Mason, Jane Menken, Philip S. Morgan, Jessie Norris, Diane Parham, Xiezhi Peng, Alexia Prskawetz, Jessica Sautter, Melania Sereny, Ke Shen, Weiheng Shi, Wenzhao Shi,StanleyK.Smith, EricStallard,Paul Shultz,BeckyTesh, FengWang,James W.Vaupel,FransWillekens,JohnWilmoth,DeqingWu,QiongWu,ZhenyuXiao, YuXie,ChangliYang,AnatoliYashin,XuejunYu,ChunyuanZhang,ZhenZhang, Zhongwei Zhao, Zhenzhen Zheng, Zhiheng Ding, Liquan Zhou, Yun Zhou, and Hania Zlotnik. Thanks also go to many scholars who participated in the software usability tests and provided helpful comments to make ProFamy software a user- friendlypackage. Notethatmostofthecontentsofthisbookarebasedonourupdated,extended, and improved previous research related to the ProFamy extended cohort- component method and its applications to the United States and China, published since 1997 in Demography, Population and Development Review, Population ResearchandPolicyReview,DemographicResearch,JournalofAgingandHealth, ix