Mexico’s XIII Population and Housing Census, 2010 A proposal under study. Population Census • • A population census is the total process of collecting, compiling, evaluating, analysing and publishing or otherwise disseminating demographic, economic and social data pertaining, at a specified time, to all persons in a country or in a well-delimited part of a country. Population is basic to the production and distribution of material wealth. In order to plan for, and implement, economic and social development, administrative activity or scientific research, it is necessary to have reliable and detailed data on the size, distribution and composition of population. The population census is a primary source of these basic benchmark statistics, covering not only the settled population but homeless persons and nomadic groups as well. Data from population censuses may be presented and analysed in terms of statistics on persons, married couples, families and households and for a wide variety of geographical units ranging from the country as a whole to individual small localities or city blocks. Principles and Recommendations for Population and Housing Censuses Revision 1 Housing Census • • A housing census is the total process of collecting, compiling, evaluating, analysing and publishing or otherwise disseminating statistical data pertaining, at a specified time, to all living quarters and occupants thereof in a country or in a well-delimited part of a country. The census must provide information on the supply of housing units together with information on the structural characteristics and facilities that have a bearing upon the maintenance of privacy and health and the development of normal family living conditions. Sufficient demographic, social and economic data concerning the occupants must be collected to furnish a description of housing conditions and also to provide basic data for analysing the causes of housing deficiencies and for studying possibilities for remedial action. In this connection, data obtained as part of the population census, including data on homeless persons, are often used in the presentation and analysis of the results of the housing census. Principles and Recommendations for Population and Housing Censuses Revision 1 Uses in an integrated programme of data collection and compilation 1. Uses of population censuses – (a) Uses for policy-making, planning and administrative purposes – (b) Uses for research purposes – (c) Uses for business, industry and labour 2. Uses of housing censuses – (a) Uses for development of benchmark housing statistics – (b) Uses for the formulation of housing policy and programmes 3. Relationship between the population census and the housing census 4. Relationship of population and housing censuses to intercensal sample surveys 5. Relationship of population and/or housing censuses to other types of censuses and other statistical investigations – – – – – (a) Census of agriculture (b) Census of establishments (c) Census of buildings (d) System of current housing statistics (e) Civil registration and vital statistics 6. Relationship of the population census to continuous population registers CRITERIA • • • • • Number of topics Depth of treatment Geographical detail Cost Frequency Conflict quantity-quality vs. cost • During the planning stages of the XII Population and Housing Census, 2000: – consultations with users – revision of national commitments. and iternational • Identification of large number of topics and items. Conflict quantity-quality vs. cost • When translated into census forms, high volume: – Exceedingly high collection costs and, – High response burden, low quality. • Conclusion: impossible and useless to ask everything to everyone. Conflict quantity-quality vs. cost • Should renegotiation take place, minor reductions occur. • Conclusion: Same as before • However: users’ needs not satisfied. • No-win situation: high costs – poor quality, as before, but poor analytical capabilities. Conflict quantity-quality vs. cost • Option No. 1: – One-sided reduction of covered topics and items. – Expected results: • manageable costs; • improvement in information quality; • reduction of study capabilities due to superficial treatment and/or elimination of topics; • greatest geographical detail Conflict quantity-quality vs. cost • Option No. 2: – Use two sets of questions: • Basic, to be given to everyone; short form; • Complementary, to a sample; long form. – Expected results: • • • • • • reasonable increase in costs new topics included and/or improved depth gains in analytical capabilities; total geographical detail for a subset of topics only minor losses in quality of response; unfulfilled users’ expectations. Conflict quantity-quality vs. cost 14 items for housing 3 for households REST SHORT FORM TOTAL POPULATION SAMPLE LONG FORM Up to 29 for population same+3+0+20* Conflict quantity-quality vs. cost Census forms • Purpose: “seek balance to optimize use of resources for information collection and analysis”. • In particular, long form improves supply of information keeping costs in check. • In other words: improve on quantityquality vs. cost relationship (QQ-C). (¿cost-benefit?) Issues Number of topics and depth of treatment • Traditional methodologies currently in use do not allow for much improvement. • Additional questions (in long form) fail to go deep enough. Issues Geographical detail • Adequate for limited number of topics • Not so for the rest (sample). Issues Dissemination of results: • Descriptive statistics for topics covered by census and sample. • Both exercises treated inedependently. Issues • Under current analysis methodologies, difficult to improve on QQ- C relationship. • necessary to explore new alternatives not available until recently. Purpose: to improve on QQ-C. • missing oportunity: use of long forms for statistical modelling and analysis; results applied to enrich information supply. Issues Results could be used to estimate: • Individual answers; or • Totals or averages for: – – – – – – – households, living quarters, city blocks, census tracts, towns, municipalities, states, depending on standard deviations. Issues • Ideally, to complement the census data base from the sampling exercise. A number of models are fitted using data obtained from the sample. Their results are used to estimate answers from the population segment which received only the short form. 14 questions for dwellings+3 for households+ up to 29 for individual residents 3d+0h+ up to 20i 14d+3h+ up to 29i e. g.: E(y)=f(x) LONG FORM SHORT FORM •Based on the enlarged data base, long-form topics could be studied in greater geographic detail. Side advantages Development of infrastructure • Human resources • Analytical tools. – Methodological – Computational available for other projects. Proposal No. 3 • Concerning only census operations – Short form (kept growing, early XX century) – Short form - Long form. (keep growing, late XX century) – Short form - many tematic forms. • Statistical systems (XXI Century). – Administrative records – Surveys – Censuses Depending on success • Proposal No. 3 (for 2010): – Universally applied battery of questions. – As many additional batteries as identified relevant topics, applied to separate samples. – Estimation of results geographical level. at “best” Proposal No. 3 Proposal No. 3 • Proposal No. 3(Cont): – Expected results: • marginal cost increase • same quality of data gathered • greater number of covered topics • greater depth of coverage • greater geographical representation – In other words, almost the ideal situation. Required actions Census topics • Consultations with users – Analysis of past experiences – Analysis of legal requirements – Analysis of international initiatives • In order to reach consensus as to relevance of topics Future actions Study and statement of relationships • Study of theoretical relationships put forward by experts on the subject. • “Letting information speak”. • In both instances, analysis of data from many sources: – Census and sample, year 2000 – Reglar surveys – Other surveys Future actions Study and statement of relationships • With a view to determining optimal sets of explanatory variables – Use models at different levels of geographical reprsentation. • Determine also “unpredictable” variables. Future actions Determining sets of questions • Integration of basic and thematic forms on the basis of statistical considerations (predictive capacity and/or non-response rates). • Ideally (and practically) limited to a small number of questions. Future actions Determining sets of questions • Statement and testing of census forms. • Establishment of expected quality of answers, including non-response. Future actions Sampling design • Simple random sampling is ignorable, in most instances • however, same dwelling may be selected for application of more than one type of questionnaire. • Therefore, more complex designs may be necessary.