No full-text available
To read the full-text of this research,
you can request a copy directly from the authors.
Microsimulation Methods for Population Projection
Abstract
Microsimulation differs from traditional macrosimulation in using a sample rather than the total population, in operating at the level of individual data rather than aggregated data, and in being based on repeated random experiments rather than average numbers. Here are presented the circumstances in which microsimulation san be of greater value than the more conventional methods. It is particularly relevant when the results of the process being studied are complex whereas the forces driving it are simple. A particular problem in microsimulation results from the fact that the projections are subject to random variation. Various sources of random variations are examined but the most important is the one we refer to as specification randomness: the more explanatory variables are included in the model, the greater the degree of random variation affecting the output of the model. After a brief survey of the microsimulation models which exist in demography, a number of the essential characteristics of microsimulation are illustrated using the KINSIM model for projecting the future size and structure of kinship networks.
... dying or giving birth) determined by an external statistic. This approach seems straightforward, but simulation design is highly flexible and there is a veritable host of implementation options for demographic events (for overviews of demographic microsimulations and their options, see: Birkin and Wu, 2012;Li and O'Donoghue, 2013;Mason, 2014;Morand et al., 2010;Spielauer, 2011;Van Imhoff and Post, 1998;Zagheni, 2015;Zaidi and Rake, 2001). Even when using high-quality exogenous variables, microsimulations can produce unexpectedly divergent outcomes (Li and O'Donoghue, 2014). ...
... When faced with a deluge of data sources and decision-making for microsimulation design, it is tempting to interpret statistics as the probability of experiencing an event for people in the displayed categories (e.g. women aged 20-24) (Van Imhoff and Post, 1998), and to use alignment procedures to correct divergence (Li and O'Donoghue, 2014). As we demonstrate, taking an intuitive interpretation of UN statistics in a microsimulation can produce very large population divergences in some countries, depending on prevailing population dynamics. ...
... CCM and simulation approaches differ in how they assign the probability of experiencing an event at a given point in time (Van Imhoff and Post, 1998). In a simulation approach, agents have an individual probability of experiencing each event based on their characteristics (e.g. ...
Even though standard cohort component models are relatively easy to comprehend, designing simulations to match expected population estimates and projections in multiple countries and over long spans of time is surprisingly challenging. We identify microsimulation design options that replicate United Nations population estimates and projections in three countries (Norway, United States, and India) over a 150-year time span. The design adapts the United Nations cohort component model, which implies a certain ordering of demographic events, and uses either cohorts or age groups to assign risk. We design four simple microsimulations that use United Nations demographic statistics as exogenous variables. One model adjusts event ordering and assignment of risk by age, called the Split Fertility design, and one model does not. Each model operates in either one- or five-year steps. The Split Fertility design has less than one percent divergence in total births, deaths, and population in all three countries. The simpler design produces varying magnitudes of divergence, as large as 20% in India. The Split Fertility design is suitable for simulations that seek to maintain population dynamics in multiple countries, whether operating in one- or five-year steps. The Split Fertility design is ideal for comparative simulations that adapt demographic statistics from cohort component models. Development of the design highlights the flexibility of simulations and the importance of careful interpretation of exogenous statistics in simulation design.
... Another option, especially when individual data are lacking, is the usage of transition matrices or conditional distributions (Li 2014;Li and O'Donoghue 2013). Based on these transition probabilities, it is stochastically decided whether a change of state occurs in an individual using the inverse transformation method (Galler 1997;Van Imhoff and Post 1998). ...
... Since events can only occur once in a fixed time interval for discrete-time microsimulations, an explicit ordering of possible demographic and non-demographic events is necessary, which must also be considered in the modeling process Van Imhoff and Post 1998). However, discrete-time models are less computationally expensive, easier to model, more data-friendly and easier to align with given macro benchmarks (Li and O'Donoghue 2013). ...
... Similarly, the working status influences other events such as fertility in the next iteration, the likelihood of forming partnerships and the probability of separations. This way microsimulations can create their own explanatory variables (Van Imhoff and Post 1998). ...
Political decision-making related to future challenges, for example in the fields of medical care, the housing market or education highly depend on valid estimates of the future population size and structure. However, such developments are usually heterogeneous throughout a country making subnational projections necessary. It is well-known that these regional differences are highly influenced by both internal and external migration processes. In this paper, we investigate the consequences of different migration assumptions on regional development in Germany using a spatial dynamic microsimulation. We find that migration assumptions have a strong direct influence on the future population and composition at the regional level and, therefore, require special attention. Depending on the scenario selected, very different socio-demographic trends may emerge in specific districts or even district types. We also demonstrate that migration assumptions affect non-demographic indicators such as the participation rate, albeit to a lesser extent. The findings are relevant to understanding the sensitivity of population projections to migration assumptions both on the national and regional level. This also paves the way to analyze how potential political interventions behave under those assumed future migration processes.
... The research presented here is part of an ongoing project that seeks to model multiple religious characteristics and behaviors over a long span of time and in multiple countries (including Norway) using a simulation model that operates with a nationally representative population of agents. Given the extended time span and anticipated complexity of agent characteristics and behaviors in the model, using a top-down microsimulation approach to determine individual demographic event risk is attractive (van Imhoff & Post, 1998;Zagheni, 2015). In this article, we analyze the process of creating a microsimulation model to generate population projections, paying particular attention to complexities associated with validating the microsimulation using population projections generated through the cohort-component method (CCM). ...
... While CCMs have been in general use by demographers for many decades, the CCM technique is ill-suited to handling multiple dimensions of individual difference and clarifying the extent to which projections are sensitive to specific conditions (Spielauer, 2011;van Imhoff & Post, 1998;Zagheni, 2015). To illustrate, UN CCM projections take account of the age group (in 5-year cohorts) and period (also in 5-year intervals) as well as gender. ...
... This advantage of microsimulations has routinely been noted but CCMs are still in wide use because demographers are often only concerned with fertility, mortality, and migration by age (or birth cohort), period, and gender. CCMs are still tractable at that level of geometrically expanding complexity, so the alternative method of microsimulation is rarely considered in those cases (van Imhoff & Post, 1998). ...
Social scientists generally take United Nations (UN) population projections as the baseline when considering the potential impact of any changes that could affect fertility, mortality or migration, and the UN typically does projections using the cohort-component method (CCM). The CCM technique is computationally simple and familiar to demographers. However, in order to avoid the exponential expansion of complexity as new dimensions of individual difference are added to projections, and to understand the sensitivity of projections to specific conditions, agent-based microsimulations are a better option. CCMs can mask hidden assumptions that are surfaced by the construction of microsimulations, and varying such assumptions can lead to quite different projections. CCM models are naturally the strongest form of validation for population projection microsimulations but there are many complexities and difficulties associated with matching microsimulation projections and CCM projections. Here, we describe our efforts to tackle these challenges as we validated a microsimulation for Norway by replicating a UN CCM projection. This provides guidance for other simulationists who seek to use CCMs to validate microsimulations. More importantly, it demonstrates the value of microsimulations for surfacing assumptions that frequently lie hidden, and thus unevaluated, within CCM projections.
... Microsimulation is a powerful tool that can be used to create population projections when the number of dimensions becomes large. Such a model is very flexible and characterised by the stochastic simulation of individual life courses based on derived parameters and individual characteristics (Van Imhoff and Post 1998). Until the late 90s, computer power was not sufficient to use microsimulation for very complex population projection. ...
... Though microsimulation methods have been conceptualised for decades and used for other purposes (Orcutt 1957), their application for population forecasts is quite new. For an exhaustive description of microsimulation for population forecasting and its properties, compared to multistate cohort-component methods, see Van Imhoff and Post (1998). ...
... At the national and subnational levels, a sample size higher than 500,000 cases is generally large enough to have a marginal Monte Carlo error and accurate simulation outcomes with only one single run Marois et al. 2020;Van Hook et al. 2020). When large samples are not possible for the base population (for instance, because of limited computing power or because the base population is built from a survey), multiple runs of the microsimulation may be required (Van Imhoff and Post 1998). In this book, since the base population we build is synthetic, we can decide a priori the number of cases to simulate. ...
This open access book provides a step-by-step overview on how to build a microsimulation model with SAS. It shows how to convert an already existing multistate projection by age, sex, education and region into a microsimulation model. Two new dimensions are then added, either the labor force participation and the sector of activity, and/or some examples of outputs and alternative scenarios that would not be possible with standard demographic methods. The book also describes how to adapt the model for other countries or other purposes. It also provides details on how to extend and adapt the model for other purposes as well as other use of microsimulation with SAS. The book suggests codes that are easy to understand, so they can be replicated or adapted for other purposes. As such, this book provides a great resource for people with beginner to intermediate knowledge in SAS.
... Since demographic developments are usually non-homogeneous and decisions often have to be made at the regional or local level, accurate spatially disaggregated projection results are essential. However, if complex, multivariate outcomes are of interest or many explanatory variables should be included in the modelling, the projections become infeasible with traditional macro approaches due to the exponential growth of the state space when additional variables are included [3,4]. In such instances, dynamic microsimulation is the method of choice. ...
... Since the number of cells in the cross-tables grows exponentially with the number of variables included, macro models quickly become infeasible when many variables are included due to the size of the state space. In fact, even for a few variables beyond age and sex, the number of cells may be larger than the number of persons in the population [3,4]. Additionally, microsimulations are less restrictive on the modelling procedure for behaviour, variable types (continuous or categorical) or process types (Markovian or non-Markovian). ...
Dynamic microsimulations are tools to stochastically project (synthetic) microdata into the future. In spatial microsimulations, regional discrepancies are of particular interest and must be considered accordingly. In practice, the probabilities for state changes are unknown and must be estimated, usually from survey data. However, estimating such models on the regional level is often not feasible due to limited sample size and lack of geographic information. Simply applying the model estimated at the national level to all geographies leads to biased state transitions due to regional differences in level and distribution. In this paper, we introduce a model-based alignment method to adapt predicted probabilities obtained from a nationally estimated model to subregions by integrating known marginal distributions to re-introduce regional heterogeneity and create more realistic trajectories, particularly in small areas. We show that the model-adjusted transition probabilities can capture region-specific patterns and lead to improved projections. Our findings are useful to researchers who want to harmonise model outputs with external information, in particular for the field of microsimulation.
... The SF design produces under 1% divergence in total population, births, and deaths over a 150-year period when compared to the UN targets in the United States, India, and Norway [5]. This is true when we initialize the model with 100,000 agents and use the sorting method [1] to reduce stochasticity in demographic events. Details of the SF design are in Bacon et al. [5]. ...
... This problem is less severe in India, because initially high fertility and rapid population growth increase the number of agents in the model very quickly, making it possible for rare events to affect at least one agent in subsequent time intervals. The need for large samples when using variance reduction techniques like such as sorting method is already known, but permitting stochasticity has its own cost and cannot solve the problem in very small samples either [1]. ...
We explore microsimulation design options as a source of divergence in total population when using demographic statistics from the United Nations to model population dynamics in three countries between 1950 and 2100. We compare 176 unique model designs, which toggle options such as the time step, the initial sample size of agents, variance reduction, ordering of demographic events, and adjustments to risk assignment as appropriate to each statistic. Results indicate that small population samples and 1-year time steps can produce particularly high divergence from UN targets, even when other options known to reduce divergence are implemented. Small sample 1-year models with low divergence are possible, but the specific combinations of options interact with a country’s population dynamics in unpredictable ways, which prevents the design from being used in other country contexts. These findings are important for balancing efficiency, accuracy, realism, and generalizability in demographic microsimulation design.
... In the included studies, populations were modelled for countries, regions and cities in Europe (20), Asia (16), Africa (14), North America (12), Oceania (9) and South America (6). Twelve studies covered multiple populations. ...
... However, the number of covariates included in each demographic process varies substantially between the included IBM studies. Given the predicted probabilities, a random number generator is used to determine whether an individual experiences the event and the individual's attributes are updated accordingly [20]. This makes it possible to track the life course of each individual. ...
Background
An increasing number of infectious disease models consider demographic change in the host population, but the demographic methods and assumptions vary considerably. We carry out a systematic review of the methods and assumptions used to incorporate dynamic populations in infectious disease models.
Methods
We systematically searched PubMed and Web of Science for articles on infectious disease transmission in dynamic host populations. We screened the articles and extracted data in accordance with the guidelines of the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA).
Results
We identified 46 articles containing 53 infectious disease models with dynamic populations. Population dynamics were modelled explicitly in 71% of the disease transmission models using cohort-component-based models (CCBMs) or individual-based models (IBMs), while 29% used population prospects as an external input. Fertility and mortality were in most cases age- or age-sex-specific, but several models used crude fertility rates (40%). Households were incorporated in 15% of the models, which were IBMs except for one model using external population prospects. Finally, 17% of the infectious disease models included demographic sensitivity analyses.
Conclusions
We find that most studies model fertility, mortality and migration explicitly. Moreover, population-level modelling was more common than IBMs. Demographic characteristics beyond age and sex are cumbersome to implement in population-level models and were for that reason only incorporated in IBMs. Several IBMs included households and networks, but the granularity of the underlying demographic processes was often similar to that of CCBMs. We describe the implications of the most common assumptions and discuss possible extensions.
... Microsimulation methods that deal with the demographic projection of individuals (or synthetic representations of individuals) rather than population groups/cohorts are becoming more prevalent in the small area projection domain. The strengths of taking a micro over a macro approach for population projection are addressed by Van Imhoff and Post (1998), namely that microsimulation allows for the inclusion of a large number of individual attributes (which impact on demographic behavior) and are capable of producing richer output than macro models in the form of a database of individuals. Examples of implementation include a model for Ireland (Ballas, Clarke, and Wiemers 2005), for Britain , and for the London borough of Tower Hamlets (Lomax and Smith 2017). ...
... Examples of implementation include a model for Ireland (Ballas, Clarke, and Wiemers 2005), for Britain , and for the London borough of Tower Hamlets (Lomax and Smith 2017). Further advantages of micromodels identified by Van Imhoff and Post (1998) are that they are better at dealing with interaction effects between variables, and between individuals; however, it could be argued that if these are requirements of a model then they are more appropriately dealt with within an Agent Based framework (e.g. see Wu and Birkin 2012). ...
The size, composition, and spatial distribution of both people and households have a substantial impact on the demand for and development and delivery of infrastructure required to support the population. Infrastructure encompasses a wide range of domains including energy, transport, and water, each of which has its own set of spatial catchments at differing scales. Demographic projections are required to assess potential future demand; however, official projections are usually not provided at a high level of spatial resolution required for infrastructure planning. Furthermore, generating bespoke demographic projections, often incorporating a range of scenarios of possible future demographic change is a specialist, resource intensive job and as such is often missing from infrastructure development projects. In this paper we make the case that such demographic projections should be at the heart of infrastructure planning and present a set of open‐source models which can be used to undertake this demographic projection work, thus providing the tools needed to fill the identified gap. We make use of a case study for the United Kingdom to exemplify how a range of scenarios can be assessed using our model.
... This uncertainty was then transparently propagated through all stages of the analysis, from the estimation of the mean number of child speakers per adult speaker to the projection of the populations by five-year age groups. The use of a Monte Carlo method further allowed us to consider the impact of chance on a language's prospects and to estimate its dormancy risk [35]. ...
UNESCO launched the International Decade of Indigenous Languages in 2022 to draw attention to the impending loss of nearly half of the world’s linguistic diversity. However, how the speaker numbers and dormancy risks of these languages will evolve remains largely unexplored. Here, we use Canadian census data and probabilistic population projection to estimate changes in speaker numbers and dormancy risks of 27 Indigenous languages. Our model suggests that speaker numbers could, over the period 2001–2101, decline by more than 90% in 16 languages and that dormancy risks could surpass 50% among five. Since the declines are greater among already less commonly spoken languages, just nine languages could account for more than 99% of all Canadian Indigenous language speakers in 2101. Finally, dormancy risks tend to be higher among isolates and within specific language families, providing additional evidence about the uneven nature of language endangerment worldwide. Our approach further illustrates the magnitude of the crisis in linguistic diversity and suggests that demographic projection could be a useful tool in assessing the vitality of the world’s languages.
... We fill this gap and investigate prospective demographic trends over the next two decades by utilizing a microsimulation technique and executing a series of Monte Carlo (MC) stochastic simulations. The latter is a widely recognized approach for modelling demographic changes and a crucial instrument for analysing structured population models (Van Imhoff and Post, 1998;Mielczarek and Zabawa, 2021). 2 Our simulation exercise includes a stochastic population sub-model that embraces demographic uncertainty. 3 Key parameters such as age-and gender-specific mortality rates, age-specific fertility rates, as well age-and gender-specific propensity to migrate (as refugees fleeing the war) and return are modelled as stochastic processes. ...
This paper outlines the impact of the war on Ukraine's economy, sets out the main challenges for reconstruction and EU integration, and provides a set of policy recommendations to "build back better" within the EU accession process. The economic and demographic shock caused by the war has been severe. However, we find that Ukraine can feasibly follow the EU-CEE EU integration path once the war ends, and that it will bring a great deal to the EU. Ukraine will not create an unmanageable extra strain on the EU budget, but the integration of its competitive agricultural sector must be carefully managed. Policy priorities for Ukraine and the EU include measures to mitigate the demographic shock, rebuild infrastructure in a way that integrates Ukraine more tightly into the EU economy, support the regions most impacted by the war, and use trade and FDI policies to maximise the benefits of EU integration.
... This simulation will replicate the evolution of the Ukrainian population under several assumptions on the duration and escalation of the war. We apply a microsimulation technique and run a series of Monte Carlo (MC) stochastic simulations, which is widely used for modelling demographic developments and is a vital tool for analysing structured population models4 (Van Imhoff & Post, 1998;Mielczarek & Zabawa, 2021). The stochastic population sub-model utilized in our simulation exercise incorporates demographic uncertainty, as the primary parameters, such as fertility, mortality, and migration, can be considered as stochastic processes. ...
The chapter analyses characteristics of Ukraine’s economic reconstruction following the intensive phase of the war. We first cover the demographic situation, which is indeed very problematic given a long-run population decline and the significant shock of war-induced migration and displacement of persons. Then, we analyse patterns of the comparative (specialisation) advantages of Ukraine’s economy, identifying the potentials Ukraine will have in order to attract FDI and integrate into cross-border production linkages. Next, we cover the varying situation across Ukraine’s regions, their highly uneven potential for recovery, and evaluate their future possibilities for production and trade specialisation. We discuss Ukraine’s position in the context of its accession to the EU: this covers how Ukraine compares in terms of a multitude of indicators with other Central and Eastern European countries before they acceded to the EU, as well as with candidate countries in Southeast Europe. Finally, we discuss how Ukraine’s accession will impact existing member countries – with an emphasis on countries in Central and Eastern Europe – in macroeconomic and structural terms, and also with respect to implications regarding Ukraine’s participation in EU budgetary programs.
... We randomly generate artificial life histories which we aggregate up into populations classified by age and sex. The strategy of using computer power as an alternative to mathematical theory is common in contemporary applied statistics (Gelman and Vehtari, 2021), though there is also a long tradition in demography of using micro-simulation to study demographic processes (Wachter et al., 1978;Van Imhoff and Post, 1998;Zagheni, 2015). ...
When studying a national-level population, demographers can safely ignore the effect of individual-level randomness on age-sex structure. When studying a single community, or group of communities, however, the potential importance of individual-level randomness is less clear. We seek to measure the effect of individual-level randomness in births and deaths on standard summary indicators of age-sex structure, for populations of different sizes, focusing on on demographic conditions typical of historical populations. We conduct a microsimulation experiment where we simulate events and age-sex structure under a range of settings for demographic rates and population size. The experiment results suggest that individual-level randomness strongly affects age-sex structure for populations of about 100, but has a much smaller effect on populations of 1,000, and a negligible effect on populations of 10,000. Our conclusion is that analyses of age-sex structure in historical populations with sizes on the order 100 must account for individual-level randomness in demographic events. Analyses of populations with sizes on the order of 1,000 may need to make some allowance for individual-level variation, but other issues, such as measurement error, probably deserve more attention. Analyses of populations of 10,000 can safely ignore individual-level variation.
... Microsimulation population methods, by definition, operate at the scale of individuals rather than populations groups. 21 When modelling at the individual level, we can apply a probability that an event will happen, rather than deterministically stating that it will happen to a certain share of the population. If we take the 30-yearold couple from before, we model that the couple has a 9 percent probability of having a child in the year they are 30, instead of modelling that they have 0.09 children. ...
This study presents an analysis of population trends in rural Norway, with a specific focus on the potential implications for rural populations in a long-term perspective. We highlight the significant outward migration of young people from rural areas, which has led to a demographic shift happening at a faster pace and on a larger scale in rural areas compared to central areas. Using North-Varanger, the most north-eastern part of the country, as an example, the article shows that the region has experienced net outward migration for the past 50 years, and if current trends persist, the population in the region could decrease by up to 80 percent by 2100. North-Varanger is an area of strategic importance to Norway. It is also an area of vital importance for the Sami people. This study shows that there is a risk that the population in one of the most strategically important regions will see a dramatic reduction. The only way to reverse this trend is to change the migration patterns and encourage more young families to settle in North-Varanger. This stresses the urgent need for policymakers to re-evaluate current public measures aimed at attracting migrants to rural areas in order to make the measures more effective and to ensure sustainable settlement in the region.
... Migration studies using event history analysis have been began, among others [12], [7]- [10], and [38]. Migration is one of the events that occurred on the individual demographics such as change of marital status, changes in the level of education, type of economic activity and other changes. ...
Migration is the process of moving people from one region to another. There are two aspects that follow the process, those are individuals and regions. The individual data or the micro data requires specific modeling to the individual characteristic related to the decision to migrate. While the region data or macro data requires different modeling to characterize the region, as the origin and the destination of migration. The population mobility become more complex at district/city levels compared to inter-provincial or even international due to the completeness data. The East Java Province is known as the province that send migrants to other provinces in Indonesia and even internationally. Several theories reveals that economic is one of the dominant factors in migration, but it does not comply in this cases. This province can be categories as strong economics as Jakarta, but out migrations from East Java is still high. The high economic growth in East Java province cannot resist the out migration flows. The imbalance of economic growth evenly among districts/cities should be one of major factors that have to be overviewed closely.
... This study makes extensive use of probabilistic based modelling, a common approach which aims to reproduce observed distributional characteristics (Li, 2011). Monte-Carlo simulation is used to assign characteristics, (van Imhoff & Post, 1998;Tanton, 2018), a procedure where randomly 14 generated numbers are compared to rates of incidence in a population, and used to assign characteristics (Lomax & Smith, 2017 of Statistics, such as the proportions of students within different school types (e.g. government schools), a randomly generated number with a uniform distribution between 0 and 1 was assigned to each modelled agent. Where this number fell below the value indicating the incidence within the population, the agent was assigned this characteristic (e.g. ...
Driven by the internationally observed phenomenon that poorer children perform worse on standardised tests than their more affluent counterparts, this study attempts to understand the key links between household income and educational outcomes at a system level, through the development of the Parental Income and Child Academic Achievement (PICAA) microsimulation model. This model was developed with the goal of being used by policymakers to test the magnitude of expected outcomes of educational and welfare interventions for student outcomes. To do so, this study develops and employs a novel targeted review method to identify key pathways in existing education literature. Following this, it shows how this information can be mapped into a systems diagram, which can then be quantified and operationalised as a microsimulation model based on academic literature and aggregate national data. Finally, it outlines how this model can be validated with de-identified individual data from nationally representative samples, and finally how it can generate results when applied to policy questions. This study represents one of the first applications of modelling to investigate educational inequality as a function of income in an Australian context, one of the first applications of microsimulation to the question of income and educational outcomes, and the first application to the Australian education system. The completion of the validated PICAA model represents the key contribution of this work. This model has high potential to be used for the evaluation of education policy interventions, which is illustrated through a series of income-based policy intervention scenarios. However, the development process of the PICAA microsimulation model has also resulted in a number of key findings and contributions. This study reinforces the importance of the early years as a period of significant developmental growth and a time period in which income-based interventions may provide the best value-for-money for policymakers looking to increase achievement scores or reduce inequality in achievement. This study has also identified an increasing correlation between income and achievement with increasing time between measurements, first identified using the PICAA model but also identified in analysis of NAPLAN achievement data. This newly identified trend has implications for both the interpretation of results of policy analysis and for the importance of longitudinal studies in the study of inequality. Additionally, this study has identified a likely rise in formal care costs in real terms within the Australian context over the past twenty years. This increase has made formal care as much as three times more expensive in real terms, after accounting for inflation and allowing for increases in real wages and hours of use. Quantifying the scale of the cost increase over this period is a further contribution of this study, especially in relation to potential policy interventions to address income and educational inequality. The key recommendations for policy arising from this study include that: there is likely to be greater potential to reduce inequalities in educational outcomes through interventions implemented earlier in life; sustained income interventions (for example, increases to welfare payments for low-income households with young children) are more likely to have impact than one-off income interventions (e.g. cash injections); and that interventions that reduce the negative effects of poverty in early childhood are likely to have significant positive effects on later educational inequalities. In summary, this study has illustrated the value of modelling as an interdisciplinary synthesis tool, and specifically the value in the application of microsimulation to the modelling of education systems, with a particular emphasis on the generation of policy-relevant insights.
... The projection model developed for this paper uses a microsimulation approach (Van Imhoff & Post, 1998). A microsimulation model starts from a baseline population that consists of individual actors whose characteristics represent the composition of a given population across chosen dimensions. ...
While India is entering the period of demographic dividend, female labour participation rates remain very low. This paper aims to provide the very first labour force projections for India and its regions up to the year 2060. Projections are achieved using a discrete-time microsimulation model in which changes in population size and composition come from the interaction between demographic characteristics, educational attainment, and secular tendencies. Labour force participation rates are estimated at the individual level using personal characteristics as predictors. Results show that under constant labour force participation rates, the labour force dependency ratio (non-workers/workers) is very unlikely to attain favourable levels, which compromises the potential demographic dividend that the country could gain from its favourable age-structure. At the subnational level, the forecast yields the most favourable dependency ratio in 2060 in the regions that combine both a low-age dependency ratio and a higher participation of women. Results moreover suggest that female labour force participation is a better driver of the labour force dependency ratio than the age composition.
... Microsimulations are also useful for analyzing heterogeneity or in cases where the number of variables and the number of attributes that these variables can take is very large (Van Imhoff and Post 1998;Spielauer 2011). The heterogeneity of kinship networks can be easily observed from simulated population microdata. ...
Kinship relations play a crucial role in structuring populations and shaping individual outcomes. Differences in kinship among individuals, cohorts, and subpopulations are one important aspect of these structures. Demography and related disciplines have proposed sophisticated approaches to study kinship in recent years. We argue that the development of a demography of kinship that centers on these processes will help advance the field of demography as a whole. Here, we review four key substantive areas of kinship research in demography: (1) kin supply and intergenerational transfers; (2) demographic change; (3) kin loss; and (4) social stratification. For each area, we identify important gaps in the literature and avenues for future research. We then review available methods and data sources to advance each of these areas, and conclude with an agenda to foster the study of the demography of kinship in general and kinship inequalities specifically.
... A methodological limitation relates to the degree of model detail. Although failure to model correlation between events may lead to less variation in key output variables (Ruggles 1993), model complexity does not go hand in hand with prediction power due to the stochastic nature of microsimulations (van Imhoff and Post 1998). The greater the number of processes modelled, the higher the number of Monte Carlo experiments involved during the simulation. ...
... The Microsimulation Model. The projection model developed for this research uses a microsimulation approach (27). A microsimulation model starts from a baseline population that consists of individual actors whose characteristics represent the composition of a given population across chosen dimensions. ...
Significance
China has the world’s largest national population and is rapidly catching up with the United States in terms of having the status as the world's largest economy. In this context, recent reports about unexpectedly low levels of fertility have given rise to speculation that the resulting population stagnation/decline and rapid aging may pose a major obstacle to continued prosperity in the future. We show that, depending on the indicator of demographic dependency used, the future may look very different. When associated with rapid increases in human capital, low fertility rates may not pose such a significant obstacle to continued development over the coming decades. Whether this will be the case may have profound geopolitical and global economic consequences.
... At the national and subnational levels, a sample size higher than 500,000 cases is generally large enough to have a marginal Monte Carlo error and accurate simulation outcomes with only one single run (Bélanger et al. 2019;Caron-Malenfant et al. 2017;Marois et al. 2020;Van Hook et al. 2020). When large samples are not possible for the base population (for instance, because of limited computing power or because the base population is built from a survey), multiple runs of the microsimulation may be required (Van Imhoff and Post 1998). In this book, since the base population we build is synthetic, we can decide a priori the number of cases to simulate. ...
This chapter sets the stage before building the microsimulation model. First, we describe proprieties of the microsimulation model that will be built. The model is time-based, discrete-time and stochastic. We then describe properties of a multistate model that will be converted into a microsimulation model and we show how building a synthetic base population that consists of the individuals that will be projected. We finally explain how to set up the workspace in SAS.
... In contrast to most small area population forecasting methods, microsimulation models, by definition, operate at the scale of individuals rather than populations. Consequently, they require considerably more input data and data preparation than macro-scale models, but possess several beneficial features, including rich output detail across many population characteristics (van Imhoff & Post, 1998). They also avoid the need for constraining between geographical scales because all output is aggregated from the individual scale. ...
Small area population forecasts are widely used by government and business for a variety of planning, research and policy purposes, and often influence major investment decisions. Yet the toolbox of small area population forecasting methods and techniques is modest relative to that for national and large subnational regional forecasting. In this paper we assess the current state of small area population forecasting, and suggest areas for further research. The paper provides a review of the literature on small area population forecasting methods published over the period 2001-2020. The key themes covered by the review are: extrapolative and comparative methods, simplified cohort-component methods, model averaging and combining, incorporating socio-economic variables and spatial relationships, ‘downscaling’ and disaggregation approaches, linking population with housing, estimating and projecting small area component input data, microsimulation, machine learning, and forecast uncertainty. Several avenues for further research are then suggested, including more work on model averaging and combining, developing new forecasting methods for situations which current models cannot handle, quantifying uncertainty, exploring methodologies such as machine learning and spatial statistics, creating user-friendly tools for practitioners, and understanding more about how forecasts are used.
... Microsimulation models are built on individual-level determinants of events, offering the possibility to consider a large number of different sub-groups and to estimate the potential interactions among several dimensions. In microsimulation, however, outcomes are subject to random variation, which makes the method subject to "specification randomness"; the degree of random variation affecting the output of the model increases with the number of explanatory variables included in the model (Van Imhoff & Post, 1998). ...
We developed an innovative method to break down official population forecasts by educational level. The mortality rates of the high education group and low education group were projected using an iterative procedure, whose starting point was the life tables by education level for Italy, based on the year 2012. We provide a set of different scenarios on the convergence/divergence of the mortality differential between the high and low education groups. In each scenario, the demographic size and the life expectancy of the two sub-groups were projected annually over the period 2018–2065. We compared the life expectancy paths in the whole population and in the sub-groups. We found that in all of our projections, population life expectancy converges to the life expectancy of the high education group. We call this feature of our outcomes the “composition effect”, and we show how highly persistent it is, even in scenarios where the mortality differential between social groups is assumed to decrease over time. In a midway scenario, where the mortality differential is assumed to follow an intermediate path between complete disappearance in year 2065 and stability at the 2012 level, and in all the scenarios with a milder convergence hypothesis, our “composition effect” prevails over the effect of convergence for men and women. For instance, assuming stability in the mortality differential, we estimated a life expectancy increase at age 65 of 2.9 and 2.6 years for men, and 3.2 and 3.1 for women, in the low and high education groups, respectively, over the whole projection period. Over the same period, Italian official projections estimate an increase of 3.7 years in life expectancy at age 65 for the whole population. Our results have relevant implications for retirement and ageing policies, in particular for those European countries that have linked statutory retirement age to variations in population life expectancies. In all the scenarios where the composition effect is not offset by a strong convergence of mortality differentials, we show that the statutory retirement age increases faster than the group-specific life expectancies, and this finding implies that the expected time spent in retirement will shrink for the whole population. This potential future outcome seems to be an unintended consequence of the indexation rule.
... The use of microsimulations for population projections is relatively new, even though the advantages of such models compared to the traditional macro-level population projection models have been discussed for decades 31 . Microsimulation methods have been used by population scientists to model demographic processes, to make detailed and realistic population projections encompassing various population dimensions, and to gain insights on life course transitions 55 . ...
The extent of the challenges and opportunities that population ageing presents depends heavily on the population’s health. Hence, for the development of appropriate strategies that enable countries to adopt the emerging demographic and epidemiological realities, information on future health trajectories of elderly population is a natural requirement. This study presents an innovative methodological framework for projecting the health of individuals using a dynamic microsimulation model that considers interactions between sociodemographic characteristics, health, mortality, bio-medical and behavioral risk factors. The model developed, called ATHLOS-Mic, is used to project the health of cohorts born before 1960 for the period 2015–2060 for selected European Countries using SHARE data to illustrate the possible effects of some selected risk factors and education on future health trajectories. Results show that, driven by a better educational attainment, each generation will be healthier than the previous one at same age. Also, we see that, on average, an individual of our base population will live about 18 more years since the start of the projection period, but only 5 years in good health. Finally, we find that a scenario that removes the effect of having a low level of education on individual health has the largest impact on the projected average health, the average number of years lived per person, and the average number of years lived in good health.
... A methodological limitation relates to the degree of model detail. Although failure to model correlation between events may lead to less variation in key output variables (Ruggles 1993), model complexity does not go hand in hand with prediction power due to the stochastic nature of microsimulations (van Imhoff and Post 1998). The greater the number of processes modelled, the higher the number of Monte Carlo experiments involved during the simulation. ...
Background: Family patterns in Western countries have changed substantially across birth cohorts. The spread of unmarried cohabitation, the decline and postponement of marriage and fertility, and the rise in nonmarital births, partnership instability, and repartnering lead to an increasing diversity in family life courses. Objective: In this paper we demonstrate how to set up a tool to explore family life trajectories. This tool models the changing family patterns, taking into account the complex inter-relationships between childbearing and partnership processes. Methods: We build a microsimulation model parameterised using retrospective partnership and childbearing data. The data cover women born since 1940 in Italy, Great Britain, and two Scandinavian countries (Norway and Sweden), three significantly different cultural and institutional contexts of partnering and childbearing in Europe. Results: We guide readers through the modelling of individual life events to obtain a set of aggregate estimates, providing information on the power, technical structure, and underlying assumptions of microsimulations. Validation of the simulated family life courses against their real-world equivalents shows that the simulations not only closely replicate observed childbearing and partnership processes, but also provide high quality predictions when compared to more recent fertility indicators. Conclusions: Using observed population estimates to systematically validate the results both validates our model and increases confidence that microsimulations satisfactorily replicate the behaviour of the original population. Contribution: We create and validate a microsimulation model that can be used not only to explore mechanisms throughout the family life course but also to set up scenarios and predict future family patterns.
... Indeed mechanistic constraints, working at the individual level, are classically modeled via Individual-Based Models (IBMs; also called agent-based models) because they allow tracking each specific being during every step of its life trajectory (see for instance an IBM investigating CoR in ungulates by Proaktor et al. (2008)). Thanks to their level of details, such models can be considered as more precise and more flexible population projectors than matrices (Van Imhoff and Post, 1998). However, contrary to PPMs that project the population as a whole, they make it difficult to demonstrate the generalization of simulation results and to qualitatively ponder the weights of the various parameters that influence fitness (Caswell and John, 1992). ...
Assessing the role played by purifying selection on a susceptibility allele to late-onset disease (SALOD) is crucial to understanding the puzzling allelic spectrum of a disease, because most alleles are recent and rare. This fact is surprising because it suggests that alleles are under purifying selection while those that are involved in post-menopause mortality are often considered neutral in the genetic literature. The aim of this article is to use an evolutionary demography model to assess the magnitude of selection on SALODs while accounting for epidemiological and sociocultural factors. We develop an age-structured population model allowing for the calculation of SALOD selection coefficients (1) for a large and realistic parameter space for disease onset, (2) in a two-sex model in which men can reproduce in old age and (3) for situations in which child survival depends on maternal, paternal and grandmaternal care. The results show that SALODs are under purifying selection for most known age-at-onset distributions of late-onset genetic diseases. Estimates regarding various genes involved in susceptibility to cancer or Huntington’s disease demonstrate that negative selection largely overcomes the effects of drift in most human populations. This is also probably true for neurodegenerative or polycystic kidney diseases, although sociocultural factors modulate the effect of selection in these cases. We conclude that neutrality is probably the exception among alleles that have a deleterious effect in old age and that accounting for sociocultural factors is required to understand the full extent of the force of selection shaping senescence in humans.
... Among the advantages of microsimulation models are that they can provide richer outputs, and they can be used to build individual-level determinants of events. Thus, such models can provide better results when the number of dimensions included is relatively large, and when handling interactions between multiple variables (van Imhoff and Post 1998). The CEPAM-mic microsimulation model captures both the educational and the migrant characteristics of individuals, and its dimensions include age, sex, educational attainment, region of origin, duration of stay in the host country, and several others detailed in section 3.2. ...
Building on the well-established knowledge on fertility differentials by education and nativity/migration status, we employ microsimulation modelling to demonstrate the effect of accounting for such differences in population projections. We consider fertility differentials by educational attainment, enrolment in full-time education, region of birth, age at immigration, and duration of stay in the host country, which we introduce step-wise into the microsimulation model for the EU28. Results on projected TFRs and births by 2060 illustrate the importance of accounting for several sources of population heterogeneity. In the context of future educational expansion, modelling education differentials for students and for women with completed education is needed to capture the postponement effect of education on childbearing. Future migration assumptions that include migrant fertility differentials lead to widely varying projected numbers of future births. At fixed fertility differentials and a fixed composition of immigrant flows, the net effect of immigrant fertility on the overall TFR in the EU28 is projected to increase from the estimated 0.12 in 2015-2019 to 0.17 in 2055-59 in the scenario with baseline migration, and to 0.25 in 2055-59 in the scenario with doubled migration.
... Data availability for the statistical models and especially lack of comparable data for all EU28 countries sets limits to the interlinkages that can be captured and incorporated in the model. Microsimulation methods are also very flexible in the sense that they allow the creation of scenarios 10 combining different hypotheses concerning the future evolution of the stochastic parameters that drive the component of population changes (Van Imhoff and Post 1998;Spielauer 2010). ...
EUR 30146 EN This publication is a Technical report by the Joint Research Centre (JRC), the European Commission's science and knowledge service. It aims to provide evidence-based scientific support to the European policymaking process. The scientific output expressed does not imply a policy position of the European Commission. Neither the European Commission nor any person acting on behalf of the Commission is responsible for the use that might be made of this publication. For information on the methodology and quality underlying the data used in this publication for which the source is neither Eurostat nor other Commission services, users should contact the referenced source. The designations employed and the presentation of material on the maps do not imply the expression of any opinion whatsoever o n the part of the European Union concerning the legal status of any country, territory, city or area or of its authorities, or concernin g the delimitation of its frontiers or boundaries.
... Fundamental to almost all spatial microsimulation models is the creation of a micro-dataset containing the spatial distribution of demographic features. This field is a subset of a wider field of demographic microsimulation modelling and population projection (Van Imhoff, and Post, 1998;Van Imhoff and Keilman. 1991). ...
... Indeed mechanistic constraints, working at the individual level, are classically modeled via Individual-Based Models (IBMs; also called agent-based models) because they allow tracking each specific being during every step of its life trajectory (see for instance an IBM investigating CoR in ungulates by Proaktor et al. (2008)). Thanks to their level of details, such models can be considered as more precise and more flexible population projectors than matrices (Van Imhoff and Post, 1998). However, contrary to PPMs that project the population as a whole, they make it difficult to demonstrate the generalization of simulation results and to qualitatively ponder the weights of the various parameters that influence fitness (Caswell and John, 1992). ...
It is increasingly recognized that incorporating life history trade-offs into evolutionary demography models requires trade-offs to be decomposed into fixed (a.k.a genetic) and individual (a.k.a dynamic) components. This is fundamental in order to understand how trade-offs are related to fixed and dynamic components of individual heterogeneities and generate variance in individual trajectories. Therefore, embedding such trade-offs into population projection matrices usually requires three categories: a life-history determining trait (e.g., age or stage), a fixed trait incorporating the genetic trade-off, and a dynamic trait modeling the individual component. This has proved a complex exercise until the recent advent of Multitrait Population Projection Matrices (MPPMs). Recent developments of Trait-Level Analysis (TLA) tools for MPPMs now allow us to study the demographic and evolutionary consequences of each component of a life history trade-off. Here, we illustrate this by constructing and analyzing an evolutionary demography model that implements both dynamic and fixed components of the costs of reproduction, the trade-off between current/early reproduction and future/later fitness. In particular, we explain and describe the use of the TLA to measure the effects of this trade-off on individual fitness. Here, we focus on the variance of lifetime reproductive success between models implementing the individual costs and asymptotically-equivalent matrices from which they are absent. This allows us to show that dynamic costs decrease that variance and more so for slow organisms. Therefore, accounting for this component of the costs, instead of classically focusing solely on fixed costs of reproduction, is paramount in order to correctly assess the relative importance of the "neutral" and "adaptive" components of individual heterogeneity.
The cohort-component method is the standard model for producing population projections in official statistics. It is straightforward to compute, requires minimal input data, and is widely recognised by demographers. However, cohort-component projection models are limited in their ability to capture complex demographic processes and provide detailed output for individual-level outcomes. To address these shortcomings, Statistics Austria has developed the dynamic microsimulation model STATSIM for its official national population projection, which was previously computed by the cohort-component method. We have opted for a gradual transition, starting by replicating the results of past cohort-component projections using microsimulation. As a first extension, we have implemented a model of international migration that takes into account the relationship between emigration risk and length of stay, as well as country of birth. By comparing the results of STATSIM's retrospective projections with counterfactual cohort-component projections, we show that STATSIM's projections are more consistent with observed emigration patterns. In the future, STATSIM can be further developed by adding modules for education, employment, health and other socio-economic characteristics.
In a climate-constrained world, understanding the energy required to achieve universal access to modern energy is critical. This requires making assumptions on future population trajectories. Although access to modern energy can affect population dynamics, this feedback has not yet been accounted for in demographic models. Access to modern energy leads to fertility declines as it reduces child mortality, improves health, increases women’s access to information, education and employment. In this paper we present a demographic model that endogenizes the effect of increased access to modern energy on population dynamics and estimates the size of this effect on total final energy use by households for the case of Zambia. To do so, we built a microsimulation model to project future population size and composition, accounting for how fertility depends on access to modern energy and education. We used these population projections to then estimate household energy demand of the Zambian population until 2070, under different scenarios. We found that in 2070, while electricity consumption is higher in a universal access scenario compared to a baseline scenario, total energy demand is 29% lower, partly due to a strong decline in the use of inefficient traditional cooking fuels. We also found that reduced population growth due to universal energy access contributes to lowering the energy demand by 56% by 2050, compared to a more limited expansion in energy access, and this contribution increases over time. Although the challenge of achieving universal access to modern energy seems daunting, our results suggest that this could have co-benefits with achieving climate goals. Our study also reveals that accounting for the energy–population dividend in energy models will scale down the currently assumed energy needs to ensure a decent life for all.
Demographers use microsimulation for studying individual life courses and their attributes; events are specified to be the result of stochastic processes based on predetermined probabilistic rules.
In this study, we developed and validated a microsimulation model to reconstruct individual’s life courses and their interactions in different regions of the European Union. One of the main objectives of this study was to track migrants’ pathways in the context of three population systems, namely Sweden, the Netherlands and Spain, from 2014 to 2018.
We used official datasets as data inputs to analyze and project future population dynamics. A revised version of the MicSim package which is part of the statistical software R has been used to model mobility and migration patterns at large scale. For this purpose, among other things new functions were added to make the software more efficient concerning runtimes and data handling.
By analyzing the modelling results we conclude that MicSim has the potential to be applied for modelling migration and also more general population movements at a large scale.
The application of the MicSim package would provide policy makers with a valid instrument for the governance of migration accounting for the demographic and social patterns of migrants and their origin-destination contextual environments.
There is a large variety of different kinds of models. However, we think that they all have in common to represent something beyond themselves: they are representations of parts of the world. As scientists, we are driven to select only a few aspects of the phenomena studied. However, these aspects will be characterized with great precision. This explains why models are only considering parts of the world as we are driven to select only a few aspects of the phenomena studied. We will first give a historical presentation of models to show their usefulness in the past. We will then develop agent-based models, which are most used in demography, but also in historical demography. Some of them are not able to solve this incompleteness. Finally, we will show how a deeper philosophical approach to these problems may permit a true scientific treatment of explanation in demography.
There is a large variety of different kinds of models. However we think that they all have in common to represent something beyond themselves: they are representations of parts of the world. As scientists, we are driven to select only a few aspects of the phenomena studied. However these aspects will be characterised with great precision. This explains why models are only considering parts of the world as we are driven to select only a few aspects of the phenomena studied. We will first give an historical presentation of models to show their usefulness in the past. We will then develop agent-based models, which are most used in demography, but also in historical demography. Some of them are not able to solve this incompleteness. Finally, we will show how a deeper philosophical approach of these problems may permit a true scientific treatment of explanation in demography.
Demographers use microsimulation for studying individual life courses and their attributes; events are specified to be the result of stochastic processes based on predetermined probabilistic rules.
In this study, we developed and validated a microsimulation model to reconstruct individual’s life courses and their interactions in different regions of the European Union. One of the main objectives of this study was to track migrants’ pathways in the context of three population systems, namely Sweden, the Netherlands and Spain, from 2014 to 2018.
We used official datasets as data inputs to analyze and project future population dynamics. A revised version of the MicSim package which is part of the statistical software R has been used to model mobility and migration patterns at large scale. For this purpose, among other things new functions were added to make the software more efficient concerning runtimes and data handling.
By analyzing the modelling results we conclude that MicSim has the potential to be applied for modelling migration and also more general population movements at a large scale.
The application of the MicSim package would provide policy makers with a valid instrument for the governance of migration accounting for the demographic and social patterns of migrants and their origin-destination contextual environments.
synthACS is an R package that provides flexible tools for building synthetic microdatasets based on American Community Survey (ACS) base tables, allows data-extensibility and enables to conduct spatial microsimulation modeling (SMSM) via simulated annealing. To our knowledge, it is the first R package to provide broadly applicable tools for SMSM with ACS data as well as the first SMSM implementation that uses unequal probability sampling in the simulated annealing algorithm. In this paper, we contextualize these developments within the SMSM literature, provide a hands-on user-guide to package synthACS, present a case study of SMSM related to population dynamics, and note areas for future research.
Demographic methods have been evolving ever since the birth of demography in response to changes in the field's research contents and theoretical orientations. An early core mission of finding regularities underlying macro-level population phenomena and a later interest in explaining population changes inductively facilitated the development of formal demographic techniques. A more radical methodological shift occurred after the 1960s, with the increasing availability of micro-level survey data and a shift of theoretical focus toward causal mechanisms, leading to the widespread adoption of regression-based models and methods from other social science disciplines. The future development of demographic methods will likely continue to incorporate new methods first developed in other disciplines, including techniques for analyzing unstructured “big” data, but formal demographic techniques will still play a role in population forecasting, measurements improvements, and correction of faulty data, providing foundational knowledge for other social science disciplines.
Zusammenfassung
Um die Potentiale der Mikrosimulation als ein Instrument zur Analyse zukünftiger Versorgungslagen in verschiedenen (sozialpolitischen) Bereichen aufzuzeigen, wurde am Beispiel der Stadt Trier eine Mikrosimulationsstudie durchgeführt, bei der der Einfluss unterschiedlicher Wanderungsszenarien auf die Situation der Grundschülerinnen und Grundschüler im Stadtgebiet untersucht wurde. Die Datenbasis stellt eine synthetische Grundgesamtheit der Bevölkerung auf Basis der Zensusergebnisse 2011 dar, welche mit Hilfe von Zensusgitterzellen geografisch im Stadtgebiet verortet und modellbasiert in die Zukunft fortgeschrieben wurde. Außerdem liegen für Trier für die Jahre 2011 bis 2018 präzise Zahlen zu Schülerinnen und Schülern der Elementarstufe vor, an denen nicht nur die Simulationsgesamtheit konfiguriert werden kann, sondern die auch als zentrale Benchmarks für die Validität der Fortschreibungsmodule und –Parameter dienen können.
Small area population forecasts are widely used by government and business for a variety of planning, research and policy purposes, and often influence major investment decisions. Yet, the toolbox of small area population forecasting methods and techniques is modest relative to that for national and large subnational regional forecasting. In this paper, we assess the current state of small area population forecasting, and suggest areas for further research. The paper provides a review of the literature on small area population forecasting methods published over the period 2001–2020. The key themes covered by the review are extrapolative and comparative methods, simplified cohort-component methods, model averaging and combining, incorporating socioeconomic variables and spatial relationships, ‘downscaling’ and disaggregation approaches, linking population with housing, estimating and projecting small area component input data, microsimulation, machine learning, and forecast uncertainty. Several avenues for further research are then suggested, including more work on model averaging and combining, developing new forecasting methods for situations which current models cannot handle, quantifying uncertainty, exploring methodologies such as machine learning and spatial statistics, creating user-friendly tools for practitioners, and understanding more about how forecasts are used.
Mathematical models of sexually transmitted disease (STI) are increasingly relied on to inform policy, practice, and resource allocation. Because STI transmission requires sexual contact between two or more people, a model's ability to represent the dynamics of sexual partnerships can influence the validity of findings. This ability is to a large extent constrained by the model type, as different modeling frameworks vary in their capability to capture patterns of sexual contact at individual, partnership, and network levels. In this paper, we classify models into three groups: compartmental, individual-based, and statistical network models. For each framework, we describe the basic model structure and discuss key aspects of sexual partnership dynamics: how and with whom partnerships are formed, partnership duration and dissolution, and temporal overlap in partnerships (concurrency). We illustrate the potential implications of accurately accounting for partnership dynamics, but these effects depend on characteristics of both the population and pathogen; the combined impact of these partnership and epidemiologic dynamics can be difficult to predict. While each of the reviewed model frameworks may be appropriate to inform certain research or policy questions, modelers and consumers of models should carefully consider the implications of sexual partnership dynamics for the questions under study.
From origins in economics and financial analysis, microsimulation has become an important technique for spatial analysis. The method relies on conversion of aggregate census tables, sometimes complemented by sample data at the individual level, to synthetic lists of people and households. The individual records generated by the microsimulation can be aggregated flexibly to small areas, linked to create new attributes, and projected forward in time under stable conditions, or in the context of ‘what-if’ policy scenarios. The chapter outlines the basic building blocks of microsimulation and shows how these are combined within a representative practical application. It is argued that further progress can be expected through advances in computation, assimilation of data into models, and greater capacity to handle uncertainty and dynamics. We also expect the creation of more sophisticated architectures to reflect the interdependence between population structures at the micro-scale, and the supply-side infrastructures and urban environments in which they evolve.
Microsimulations make use of quantitative methods to analyze complex phenomena in populations. They allow modeling socioeconomic systems based on micro-level units such as individuals, households, or institutional entities. However, conducting a microsimulation study can be challenging. It often requires the choice of appropriate data sources, micro-level modeling of multivariate processes, and the sound analysis of their outcomes. These work stages have to be conducted carefully to obtain reliable results. We present a generic business process model for conducting microsimulation studies based on an international statistics process model. This simplifies the comprehensive understanding of dynamic microsimulation models. A nine-step procedure that covers all relevant work stages from data selection to output analysis is presented. Further, we address technical problems that typically occur in the process and provide sketches as well as references of solutions.
Significance
COVID-19 has created a mortality shock throughout the world, and it may yield a second wave of population health concerns tied to bereavement and social support reductions. We created the COVID-19 bereavement multiplier, an indicator that clarifies one downstream impact of COVID-19 mortality and can be applied to different epidemiological projections of death counts: How many people are at risk for losing a grandparent, parent, sibling, spouse, or child for each COVID-19 death. In the United States, we estimate that on average, under diverse epidemiological circumstances, every death from COVID-19 will leave approximately nine bereaved. Studying how acute mortality crises reverberate through a population in the form of bereavement multipliers expands understandings of the social impacts of health crises.
This chapter begins by defining computer simulation briefly and distinguishing it from two related techniques. The next section discusses existing uses of computer simulation in understanding family practices. Three main approaches can be identified: microsimulation for demography and policy analysis, rule-based models in anthropology and agent-based models of the emergence of social complexity. I will use these three cases to discuss the implications of the computer simulation approach in more detail. In particular, I shall try to show how simulation avoids the excessive simplifying assumptions of other formal approaches like Rational Choice while managing to make effective use of rich qualitative data. In doing this, I hope to suggest that formality and excessive abstraction need not be synonymous. In the following section, I will list aspects of family behaviour that seem essential to plausible simulation and argue that current research explores these only in a fragmentary way. The next section illustrates the requirements of an integrated simulation of family behaviour using the theory of gendered moral rationalities (Duncan and Edwards 1999) as a case study. The chapter has two deliberate limitations. Firstly, it is not intended as an exhaustive literature survey. Secondly, although the proposed simulation is described in sufficient detail that it could be built, it does not yet exist and results will not be reported. Neither of these objectives could be effectively satisfied in the space available so they have not been attempted.
This chapter describes how a spatial microsimulation method could be used as input in a Spatial Decision Support System (SDSS) to provide simulations for planners and communities, showing what their community might look like given different economic, social and demographic parameters and constraints. We find that spatial microsimulation models can be used to create the synthetic person-level base file that can then be used by the SDSS to project potential future scenarios. The advantage of using spatial microsimulation to create the base file is that it can include other variables through imputation or statistical matching techniques based on the individual-level data. The SDSS with a synthetic population from a spatial microsimulation model can implement a number of complex boundaries, including physical boundaries, environmental boundaries and economic boundaries. Further, indicators like educational attainment, income and household type can also be projected, allowing for complex simulations of potential futures for a city under different scenarios.
In this paper the impact of consistency requirements upon the formulation of multidimensional population projection models is dealt with. The problem of internal consistency arises when certain output variables of the model have to be consistent with one another, as in models containing marital status (two-sex problem). External consistency problems arise when certain output variables have to satisfy externally specified restrictions. For instance, regional population projections may have to add up to national projections. The
paper gives internal and external consistency algorithms, which control both for stock and for flow variables. Two approaches are discussed: a simple proportional adjustment for each relevant demographic component and an approach that minimises deviations in age-specific patterns of the input rates. Both algorithms are illustrated using data from the Netherlands. In the first illustration the internal consistency algorithm is applied to the age, sex, and marital status-specific model of the official 1980 national population forecasts. In the second, external consistency is considered within the context of the multiregional projection model which was constructed for the regional population forecasts for the Netherlands.
The author discusses problems involved in modeling the effects of demographic factors on historical kinship patterns with a focus on microsimulation models. "Since microsimulations of kinship ignore the correlations in demographic behavior within kin groups they ordinarily understate the variance of kinship distributions; for many kin types they also underestimate the expected number of kin." The author concludes that "those who design demographic models of kinship should be sensitive to the potential for systematic error." (EXCERPT)
Numerous marriage formation models have been proposed, and previous attempts to select among them empirically have produced
equivocal results. Herein two models, which had been appraised as most promising on the basis of their logical properties,
are analyzed to identify conditions under which they yield predictions sufficiently different as to permit unambiguous determination
that one or other is wrong. The conditions thus identified involve a relatively large cohort rising through successive age
groups, conditions which will be empirically observed during the next several years. Implications of the models' differences,
when used to anticipate future fertility rather than marriages per se, are also considered.
Demographic projection models describe the development over time of the population in terms of events. A consistency problem arises if projected numbers of events are required to satisfy certain constraints; the consistency problem can be seen as a generalization of the well-known two-sex problem in nuptiality models. This paper presents a very general characterization of consistency problems, using matrix notation, as well as a slightly less general algorithm to solve them. The preferred specification of the objective function to be minimized by the algorithm leads to a solution that can be interpreted as a generalization of the harmonic-mean approach.
We assess the statistical reliability of microsimulation models in two ways: by comparing simulated outcomes with survey ‘actuals’, and by calculating asymptotic confidence intervals for a variety of summary measures. The confidence intervals we derive take account of re-weighting for differential survey response, and also the effects of imposing revenue-neutrality. They are calculated for a version of one of the most widely-used U.K. tax-benefit simulation models. The results suggest that baseline simulations are reasonably accurate, but that some widely-used measures of the effects of policy changes may be very imprecise estimates of population effects.
Abstract This Monte Carlo model for simulating the reproductive history of a cohort of women is described in detail. The model provides for patterns of survival, sterility, formation and dissolution of sexual unions, fecundability, lactation, foetal wastage, family planning practices etc. Natality indices specific for marital status, for duration of marriage and for age, as well as analyses of birth spacing patterns are among the results that may be obtained. In the model, the experimental unit is an individual woman. The complete life history of a woman is generated and recorded before the history of the next woman is generated. The data for the whole cohort are analyzed at the end of the programme. The model includes two kinds of states into which a woman may pass, namely: (1) permanent changes of status such as death, sterility, or becoming a family planner, and (2) temporary states, each with a probability distribution of length of stay. The probabilities of the various events or changes of state may vary from age, parity, and other features of a woman's status or history. Natural fecundability at any age may also vary from woman to woman. In this programme natality patterns and specific indices such as age-specific fertility rates are produced, in a quasi-realistic fashion, by the interplay of the demographic and biological parameters postulated for any cohort. Consequently, the effect of changes in anyone factor can be studied, as well as the interaction resulting from changes in several factors. The purposes and potentials of the model are both substantive and methodological. As an illustration, a series of computer runs attempting to simulate the reproductive patterns of Indian women is presented. These results, as well as some additional ones, indicate some effects of changes in marital patterns, levels of fecundability, duration of post-partum non-susceptibility, age incidence of sterility and foetal wastage. In the final section of the paper, the advantages and possible applications of the model are discussed together with the limitations encountered to date in the efforts to apply the model.
The ability of classical stable population theory to determine the equilibrium growth rate and age structure of a population
from its vital rates in a single period depends on assuming that the observed maternity rates are equilibrium rates. This
paper resolves the two-sex problem by replacing the fixed, age-specific fertility schedule of classical stable population
theory by two basic relationships: a “birth matrix” and a “mating rule.” Placing certain restrictions on the birth matrix
and the mating rule (BMMR), I establish that under certain plausible conditions, the BMMR model solves the two-sex problem
by allowing matings and births to adjust to changes in population structure. The BMMR model thus provides an equilibrating
mechanism in place of a fixed maternity schedule of classical stable population theory.
This paper provides a geometric-mean solution to the consistency problem of multi-dimensional demographic projection models, based on the constrained minimization of an entropy function. A comparison with the existing harmonic-mean solution yields many similarities and almost no differences: both solutions satisfy the properties of availability, monotonicity, homogeneity, competition, and symmetry; and, for both solutions, there is a convenient one-to-one relationship between adjustments in aggregate numbers of events, on the one hand, and age-specific numbers of events, on the other hand. However, one major advantage of the geometric mean is that its corresponding distance function is firmly based on (information) theory.
Population projections for household and family size and characteristics in the Netherlands are presented. "The household forecasts are based on a model in which mothers can be paired with their children.... The model is used for the birth generation of women 1930-1995.... According to the household forecasts the Netherlands had 6.2 million households in 1992. In 2010 [there] will be 1.1 million more.... The percentage of single households will rise from 30 to 36 while the percentage of two-person households will be stable at 31%. The percentage of households with three or more persons will decrease from 39 to 33." (SUMMARY IN ENG)
excerpt
"The Dutch period- and age-specific hazard rates of marriages, cohabitation, [and] dissolution of cohabitation are studied. Also, family formation according to living arrangement of the mother is examined. The number of cohabiting persons has risen markedly.... Rates of direct marriage have declined.... The increased period of cohabitation is mainly linked with a delay in marriage among young cohabitors; it does, however, not coincide with a decreasing risk dissolution.... Fertility rates of married women have risen across the birth cohorts, while fertility rates of cohabiting women and women living without partner have remained fairly stable." (SUMMARY IN ENG)
excerpt