Figure 6 - uploaded by Arnab Bhattacharjee
Content may be subject to copyright.
Observed versus Estimated Fertility rates by age group ( = 5 6 7)

Observed versus Estimated Fertility rates by age group ( = 5 6 7)

Citations

... The spatial lagged variable DLNStudents is positive, meaning that the variation in the number of students in the neighbouring municipalities has a positive influence on the variation in the number of primary schools of a given municipality (Table 3). This stresses the spatial nature of this kind of process, and the need to consider the broader structures of spatial dependence for understanding variations in the number of schools at the local scale (the spatial autocorrelation in school demand has already been observed by Torres and Prior, 2015, while the spatial autocorrelation in demographic variables has been analysed quite extensively by authors such as Carioli et al., 2021, or Castro et al., 2015. ...
Article
Full-text available
The adjustment of services of general interest to ageing and shrinking populations is a significant challenge in many European regions. This article analyses the response in the number of primary schools to changes in the student population in the municipalities of mainland Portugal between 2007 and 2016. It focuses on the time lag of this adjustment, the territorial factors that influence it and the role of spatial spill-over effects. The main finding is that the relationship between the number of schools and the number of students is not straightforward and can only be understood considering the broader geographic contexts and spatial dependence structures. El ajuste de las escuelas primarias a la disminución de la población: un enfoque de modelado espacial Resumen: La adaptación de los servicios de interés general al envejecimiento y la disminución de la población es un reto importante en muchas regiones europeas. Este artículo analiza la respuesta del número de escuelas primarias a los cambios en la población estudiantil en los municipios de Portugal continental para el período de 2007-2016. Se centra en el desfase temporal de este ajuste, los factores territoriales que influyen en él y el papel de efectos de contagio espaciales. El principal hallazgo es que la relación entre el número de escuelas y el número de estudiantes no es directa y solo puede entenderse considerando los contextos geográficos más amplios y las estructuras de dependencia espacial. Palabras clave: Crecimiento poblacional; escuelas primarias; modelado espacial; Portugal, dependencia espacial. Clasificación JEL: R53; R58.
... Research related to the SAE method to estimate ASFR and TFR was conducted at the district level in Portugal [9]. In that study, SAE was carried out without the use of auxiliary variables because no suitable auxiliary variables were found. ...
Article
Full-text available
The large population in Indonesia has an impact on providing basic services for population which is not optimal so the condition and distribution of the population in a country must be addressed through fertility control methods. Total Fertility Rate (TFR) is one of fertility measures used in Indonesia. The estimation of TFR at the district level is very important, especially for the Nusa Tenggara Timur (NTT) Province as the province with the highest TFR in Indonesia. The availability of TFR data up to the district level is difficult to obtain every year due to data limitations. This study uses the National Socio-Economic Survey to address these problems. TFR estimation through survey data (direct estimation) generally results in a large Relative Standard Error (RSE) value, so it is necessary to estimate using an indirect estimate in the form of Small Area Estimation (SAE). By using SAERestricted Maximum Likelihood (REML) procedure, TFR with an RSE that is lower than the direct estimate is obtained. There are 5 district that have a medium-high TFR, namely: Sumba Barat Daya, Sumba Tengah, Sabu Raijua, Sumba Barat, and Manggarai Barat. The government is recommended to focus more on that 5 districts to suppress the high TFR in NTT.
... Assunção et al. (2005) developed empirical Bayes methods for small area estimation, where the data from spatially neighboring regions is used to improve the precision of estimates, particularly for a region with small sample size. Castro et al. (2015) developed Bayesian small-area inference by spatial clustering of observed fertility rates. The fertility rates were modeled as determined by a collection of covariates, plus a regionspecific random effect that has spatial dependence following a conditional autoregressive (CAR) model Besag (1974). ...
... This approach is also related to Billari et al. (2012Billari et al. ( , 2014 who developed methods for Bayesian stochastic population forecasting based on expert evaluations. In this article, we consider the approach of Castro et al. (2015) and extend this by developing methods for spatial clustering of curves, applied to time trends of fertility rates across different regions. ...
... In sharp contrast, this new branch of the literature treats these weights as unknown and potentially endogenous, and in itself an object of inference; see Bhattacharjee and Jensen-Butler 2013;Bhattacharjee and Holly 2013;Bailey et al. 2016) for a representative selection, and Bhattacharjee et al. (2014Bhattacharjee et al. ( , 2016 for reviews. Castro et al. (2015) propose Bayesian estimation of spatial weights, or equivalently the spatial contiguity matrix, by spatial clustering of fertility rates in a specific time-point. This article extends the methodology by developing Bayesian methods for spatial clustering of curves, which are then applied to time trends of fertility rates in different regions. ...
Article
Full-text available
It is important for demographic analyses and policy-making to obtain accurate models of spatial diffusion, so that policy experiments can reflect endogenous spatial spillovers appropriately. Likewise, it is important to obtain accurate estimates and forecasts of demographic variables such as age-specific fertility rates, by regions and over time, as well as the uncertainty associated with such estimation. Here, we consider Bayesian hierarchical models with separable spatio-temporal dependence structure that can be estimated by borrowing strength from neighbouring regions and all years. Further, we do not consider the adjacency structure as a given, but rather as an object of inference. For this purpose, we use the local similarity of temporal patterns by developing a spatial clustering model based on Bayesian nonparametric smoothing techniques. The Bayesian inference provides the uncertainty associated with the clustering configurations that is typically lacking in classical analyses of large data sets in which a unique clustering representation can be insufficient. The proposed model is applied to 16-year data on age-specific fertility rates observed over 28 regions in Portugal, and provides statistical inference on the number of clusters, and local scaling and shrinkage levels. The corresponding central clustering configuration is able to capture spatial diffusion that has key demographic interpretations. Importantly, the exercise aids identification of peripheral regions with poor demographic prospects and development of regional policy for such places.
... Following from Castro et al. (2015) and Zhang et al. (2021), we extend quantitative research in demography along several dimensions. First, the literature acknowledges substantial spatial diffusion in fertility (Tolnay, 1995;Weeks et al., 2000). ...
... Socio-cultural distances are difficult to measure. Castro et al. (2015) and Zhang et al. (2021) extended recent literature in spatial econometrics and statistics (Bailey et al., 2016;Bhattacharjee et al., 2016;Bhattacharjee & Holly, 2013;Bhattacharjee & Jensen-Butler, 2013) that treats drivers of spatial diffusion as unknown parameters and developed methods for socio-cultural diffusion. Then, estimated diffusion reflects drivers that may not be closely related to geographic distances and contiguity, but socio-cultural, ideational or coreperiphery relations. ...
... Regional data reflect spatial heterogeneity due to macrostructures, together with complex spatial diffusion due to endogenous social networks. By allowing diffusion to be based on a variable number of estimated spatial clusters, methodology in Castro et al. (2015) and Zhang et al. (2021) allows socio-cultural diffusion to shape fertility outcomes. This enriches spatial analysis and allows accurately modelling of social and cultural diffusion embedded in a spatial context. ...
Article
There are missing links between research and policy that can be provided by better information on the real world. This is important not only to evaluate the contribution of research to the policy makers and to society, but also to design policies based on evidences. Three models for meeting such objectives are presented, emphasizing the role of (a) journals, (b) government and (c) the researchers. We provide outline examples of the three approaches, focussing on regional demographic policies in Portugal.
... Fortunately, there are also researchers trying to bridge the gap, such as Wall (2004) andVer Hoef et al. (2018), who explain the differences and similarities between CAR and SAR processes. Readers may also be interested in other recent demographic work attempting to draw structural interpretations from CAR models with unknown spatial weights matrix (e.g., Castro et al., 2015;Zhang et al., 2020). Second, the paper by Epifani et al. (2020, in this issue) is quite unique because of its focus on the spatial Durbin error model (SDEM), which is a model containing spatially lagged explanatory variables, a spatially correlated error term (CAR process), but not a spatially lagged dependent variable. ...
Article
Full-text available
This editorial summarizes the papers published in issue 15(2) so as to raise the bar in applied spatial economic research and highlight new trends. The first paper combines a conditionally autoregressive process from the spatial statistics literature with a spatial Durbin error model from the spatial econometrics literature. The second paper feeds a multistage and multilevel data envelopment analysis with a microeconomic foundation. The third paper provides empirical evidence that Flegg’s location quotient combined with a gravity model produces the most accurate interregional input–output multipliers. The fourth paper investigates the impact of inventor networks on the number of patents per capita in Brazil.
... We assumed the increment v s;t admits the conditional autoregressive (CAR) structure (Besag et al. 1991), which is frequently adopted to capture the spatial correlation for areal data. We employed the CAR specification with a spatial dependence parameter (Castro et al. 2015) to alleviate spatial confounding: v s;t $ N 0;t 2 t M À g t W ð Þ À1 , where W is the binary adjacency matrix, M is the diagonal matrix with number of adjacent counties for individual county, t 2 t captures unmodeled sources of variation by Poisson distribution to account for overdispersion, and g t measures the spatial dependence (0 ¼ no dependence). We assumed both parameters t 2 t and g t to be year-specific as the increments v s;t can vary by scale across a different year t. ...
Article
Vehicular collisions with large ungulates pose serious challenges for managing and conserving large ungulates throughout the world. Despite the global frequency, mitigation efforts are mostly limited to localized hotspots and not effective on broad scales. Our goal was to determine whether dynamic, regional attributes could inform broader focus for mitigation efforts. We applied a spatiotemporal dynamic model to examine the regional influences on white-tailed deer (Odocoileus virginianus)–vehicle collisions (DVCs) throughout the Midwest United States from traffic, abundance of deer, and composition and configuration of the landscape during 2000–2011. The regions included eco-zones representing landscape dominated by shelter-forage habitats with ubiquitous and abundant distribution of deer (i.e., forest-agriculture matrix), landscape dominated by agriculture with sparse refugia (i.e., agriculture), and landscape dominated by forests with seasonal migration for deer (i.e., northern forest). We found little fluctuation in the factors affecting collisions through time but substantial differences among regions. In the forest-agriculture matrix eco-zone, fragmentation of the landscape was the most important predictor of collisions. In the agriculture eco-zone, traffic and abundance of deer best predicted collisions. In the northern forest eco-zone, the predictors of collisions were variable and likely related to winter severity and deer migration. This research provides new justification for broadening the focus of current mitigation measures to regional extents. In regions dominated by forest and agriculture, new policies that reduce habitat fragmentation should be the primary focus for reducing collisions. Reducing abundance of ungulates will have the most direct effect in regions dominated by agriculture. Finally, a variety of seasonal and local mitigation measures will be most effective in northern forests where large ungulates migrate. © 2018 The Wildlife Society.
... Zhang et al. (2014) developed a VT-based Bayesian hierarchical model for identifying spatial extreme in cancer disease. Castro et al. (2015) adopted VT-based Bayesian clustering model to estimate the neighborhood matrix for modeling areal count data. Feng et al. (2016) compared the performance of the VT-based spatial clustering with traditional approaches such as Poisson-CAR model for spatial count data, and demonstrated the merit of VT-clustering using several simulation studies and real data examples. ...
Article
Full-text available
In climate change study, the infrared spectral signatures of climate change have recently been conceptually adopted, and widely applied to identifying and attributing atmospheric composition change. We propose a Bayesian hierarchical model for spatial clustering of the high-dimensional functional data based on the effects of functional covariates and local features. We couple the functional mixed-effects model with a generalized spatial partitioning method for: (1) producing spatially contiguous clusters for the high-dimensional spatio-functional data; (2) improving the computational efficiency via parallel computing over subregions or multi-level partitions; and (3) capturing the near-boundary ambiguity and uncertainty for data-driven partitions. We propose a generalized partitioning method which puts less constraints on the shape of spatial clusters. Dimension reduction in the parameter space is also achieved via Bayesian wavelets to alleviate the increasing model complexity introduced by clusters. The model well captures the regional effects of the atmospheric and cloud properties on the spectral radiance measurements. The results elaborate the importance of exploiting spatially contiguous partitions for identifying regional effects and small-scale variability.
Article
Full-text available
The shrinking population in many European regions has led to a debate on the relation between the local provision of services of general interest and the demographic growth of different territories. Primary schools have been a frequent focus of this debate, given the significant number of school closures in recent years, as well as their social, economic and symbolic importance for local communities. However, the relationship between primary schools and population growth at the local scale has not been clearly established and is a challenging subject given the potential circular causality between them. This paper analyses this relationship for the mainland municipalities in Portugal between 1999 and 2016, considering three variables—total fertility rates, net migration rates and the number of primary schools—in a panel vector autoregressive (PVAR) model, where all variables are simultaneously considered endogenous and exogenous. Through this approach, it was possible to conclude that, although there is a mutual influence between these variables, the impact of school closures on the growth prospect of a municipality is limited. The main goal of policy decisions regarding the provision of primary schools should, therefore, be the quality of life of local communities, and not so much their role in countering depopulation.
Presentation
Network models are widely used to represent relations between actors or nodes. Some times we see transitivity in the network data which means if two nodes have ties with a third node then they are more likely to be tied. Sometimes the interest lies on finding the clusters in the network. Here we define a model where the adjacency structure is unobserved and the probability of tie between two nodes depend on their latent positions in the unobserved Euclidean social space. To identify the clusters within the network we develop a latent space clustering algorithm and propose a fully Bayesian estimation method which used Reversible Jump MCMC to perform the posterior inference. Finally we implement this algorithm to obtain the Network Structure and the clustering configuration of US business firms.