Sideline demographic methods and data

Advances in projecting international migration have been hindered by a lack of adequate data. Consequently, international projection-making agencies commonly use simplistic assumptions of net-migration measures derived as residuals from demographic accounting. However, past net migration can be often volatile and are known to introduce inaccuracies when projecting populations (Rogers, 1990). This paper presents a set of global population projections to 2060, focusing on alternative international migration assumptions. Expert-based assumptions about fertility, mortality and migration developed for the new edition of the Wittgenstein Centre (WiC) global population projections are combined to project each country’s population. We overcome the limitations of using net-migration models and zero convergence assumptions by drawing on a first-of-a-kind set of estimated quinquennial bilateral migration flows developed by Abel (2012). Using a multiregional cohort-component projection model, alternative future migration trends are explored based on a set of ‘what-if?’ scenarios. Our results highlight differences in the future level and distribution of populations around the globe between a constant-rates, a convergence to zero-net, a zero flows and two ‘what-if’ scenarios.

Using the 2000 Census of the Philippines Microdata for the Autonomous Region of Muslim Mindanao, we examined the age-sex distribution at the barangay level, the smallest geographical and political unit in the Philippines, to determine data quality. Despite the sluggish economy in this region, and peace and order problem ARMM registered the highest intercensal growth rate in the country. Preliminary analysis showed that such growth was highly irregular and the barangay level age-sex distributions were atypical. In most areas there were more 18 year olds than those in the younger ages. Some areas had no infants, boys, girls, adult men, and elderly population. Analysis of age ratio between an older age group versus younger age group (such as 35-39 vs. 30-34) and/or sex ratio help identify problematic census data.

In India, consumer expenditure data collected by National Sample Survey (NSS) is used for the poverty estimation in the absence of income data in the sample surveys. There have been debates about the comparison of expenditure estimates from the consumer expenditure schedule (CES) data and employment and unemployment schedule (EUS) data. Therefore, this study aims to compare the estimates of poverty using unit level data from the 66th quinquennial round (2009-10) of NSS from CES and EUS for all the states and union territories of India and tries to present a new methodology to estimate poverty. We construct variable MPCE percentiles by distributing households into hundred equal percentile groups for both the schedules (schedule 10 and schedule 1.0). We further merged the both the schedules using the key variables state, place of residence and percentile groups (100 percentile). We further compare the mean MPCE of from as well as Poverty Head-count Ratio for both the schedules using regression method. Preliminary results suggest that it could be argued based on preliminary findings that comparison of poverty estimates could be made using the consumer expenditure and employment & unemployment schedules. It will be helpful in understanding the relationship between poverty and other socio-economic factors.

Although the modeling of age specific fertility schedules has been exposed in a lot of researches, modeling fertility by birth order has received little attention. Several works have attempted to fit observed parity specific schedules, but none of them expands to models that could be used to estimate parity specific schedules when only data on all births combined fertility are available.

We propose to derive parity specific fertility rates from an overall age-specific (all births combined). Based on model proposed by Peristera and Kostaki for fitting fertility curves, we explore the assumption that there is a relationship between all the parameters of their model for all births combined and for parity specific, and derive several sets of coefficients to estimate age and parity specific fertility schedules from age and all birth combined fertility schedules.

The analysis of the fertility transition in Syria allows us to distinguish several phases. The first phase was when fertility reached world records, and resisted any change. Followed by the phase of rapid decline in the mid 1980s, and lastly by a phase of slow decrease or a phase of quasi-stagnation in fertility. Despite the interest to represent more accurate estimates of recent Syrian fertility levels and trends, very few studies have been devoted to this question. This paper addresses this issue by trying to review and analyze trends in fertility over the last three decades. Fertility rates are estimated by the own-children method using data from the SMCHS 1993 and the SFHS 2001 surveys as well as the multiple indicator cluster survey MICS3 conducted in 2006. The own-children estimates show that fertility has declined from almost 7.5 in the first half of the 1980s to about 6.3 in the second half, before declining to about 5.31 in the early 1990s and to about 3.12 in the mid-2000s. Despite this significant decline, the Syrian fertility transition is in a phase of stall or temporary blockage above 3 children per woman. Our arguments suggest continued stall of fertility decline at least in the visible future.

The paper deals with 3 topics: 1) Overview of existing sources of information on population. It covers databases: UN world population prospects, Eurostat, World health organization, Statistics Sweden, Institut National Etudes Démographiques, Human Mortality Database, and a series of others. The sources are analyzed using more than twenty quality criteria – user interface, list of countries and indicators covered, age groups of population, data formats and precision, periodicity of updating, etc. 2) Description of the Database of demographic indicators for countries of the world and regions of Russia under development at the Institute of Demography of the National Research University Higher School of Economics: the principles of data collection from different sources, methods of formation of the data cubes, metadata for these arrays, the set of indicators currently covered, the reference files for coding, ... 3) New features distinguishing this Database from another ones, examples of its use for specific data queries and modes of work. They display several know-how options, including formation of query result from the several sources specified by the user and elimination of erroneous values “on-the-fly”, recalculation of values of the indicator in the case of diverse units of measurement used in different sources.

National Sample Survey (NSS) provides national and sub-national level information on morbidities and health care at regular intervals since its inception on 1953-54. In these surveys, information for all members of sample household was gathered either from head or a key informant of the household. This paper aims to explore the effect of proxy-reporting on population estimates of morbidity prevalence and to identify patterns of biases in these estimates due to proxy responses based on recent 60th round NSS data on morbidity and health care. In this round, information for 68 percent sample population was collected from proxy respondents. Proxies underreported morbidities with an overall morbidity prevalence of 81 per 1000 adult population compared with self-reported morbidity prevalence of 148. Proxy-responses in NSS on morbidity and health care introduce systematic biases, affecting national and regional estimates of morbidity prevalence. Suitable adjustment for proxy-responses should be made while estimating realistic population risks from NSS data.

A mixture probability density functions is proposed to model mortality data. We propose to model the deaths function (dx) rather than probability or rates, since the normalized dx function is a probability density function. In this way we expect to have a relatively goog fit with a low number of parameters. We fit the mixture model for a large variety of mortality pattern, showing that the model defined is flexible to have a good fit in many cases.

When mortality forecasts for subpopulations are treated as independent, historic relationships among subpopulations may not be reflected in the forecasts. The product-ratio method of coherent forecasting (Hyndman, Booth and Yasmeen, Demography, online 2012) takes account of the relationships among subpopulations based on a single criterion such as sex or state. The coherent forecasts have been shown to improve overall accuracy and to equalise accuracy across subpopulations. The aim of this paper is to compare the accuracy of mortality forecasts when the subpopulations are based on sex and state (or country). The product-ratio method is applied to male and female populations of a group of n states, producing mortality forecasts for 2n subpopulations defined by sex and state. Two forecasts per subpopulation are made: the first set are sex-coherent forecasts for each country, while the second set are country-coherent forecasts for each sex. The accuracy of the two sets of forecasts are compared. Examples include four Nordic countries (Norway, Sweden, Denmark and Finland), three constituents of the UK (England & Wales, Scotland, Northern Ireland) and four states of Australia.

In response to growing evidence that official population statistics based on the decennial census are inaccurate and not fit-for-purpose at the Local Authority level - the fundamental administrative unit of the UK, the author has established an innovative method of using locally available administrative data sets for estimating populations. The method combines different data sources with different population coverage according to a defined set of rules. The result is timelier and geographically flexible data which is more cost-effective to undertake than a survey-based census. The ONS in the UK are now committed to exploring the use of administrative data as a replacement for the traditional survey-based census, and their findings are to be given in 2014. Key to this is the linkage of disparate datasets that are captured for other purposes to determine when the same entity (person) is present. This paper presents the author’s experience of linking administrative datasets and investigates how geography is not only the common link, but also an inherent bias that affects matching rates in terms of neighbourhood effects such as population turnover and ethnic diversity. An optimised algorithm for linking local administrative datasets is put forward, based on the results of sensitivity analyses.