Essays in Long-run Development and Spatial Economics

This thesis tries to further our understanding for why some countries today are more prosperous than others. It establishes that part of today's observed variation in several proxies such as income or gender inequality have been determined in the distant past. Chapter one shows that 450 years of (Catholic) Portuguese colonisation had a long-lasting impact in India when it comes to education and female emancipation. Furthermore I use a historical quasi-experiment that happened 250 years ago in order to show that different outcomes have different degrees of persitence over time. Educational gaps between males and females seemingly wash out a few decades after the public provision of schools. The male biased sex-ratios on the other hand stay virtually unchanged despite governmental efforts. This provides evidence that deep rooted son preferences are much harder to overcome, suggesting that a differential approach is needed to tackle sex-selective abortion and female neglect. The second chapter proposes improvements for the execution of Spatial Regression Discontinuity Designs. These suggestions are accompanied by a full-fledged spatial statistical package written in R. Chapter three introduces a quantitative economic geography model in order to study the peculiar evolution of the European urban system on its way to the Industrial Revolution. It can explain the shift of economic gravity from the Mediterranean towards the North-Sea ("little divergence"). The framework provides novel insights on the importance of agricultural trade costs and the peculiar geography of Europe with its extended coastline and dense network of navigable rivers.

Geographic weighted regression: applicability to epidemiological studies of leprosy

Abstract: INTRODUCTION: Geographic information systems (GIS) enable public health data to be analyzed in terms of geographical variability and the relationship between risk factors and diseases. This study discusses the application of the geographic weighted regression (GWR) model to health data to improve the understanding of spatially varying social and clinical factors that potentially impact leprosy prevalence. METHODS: This ecological study used data from leprosy case records from 1998-2006, aggregated by neighborhood in the Duque de Caxias municipality in the State of Rio de Janeiro, Brazil. In the GWR model, the associations between the log of the leprosy detection rate and social and clinical factors were analyzed. RESULTS: Maps of the estimated coefficients by neighborhood confirmed the heterogeneous spatial relationships between the leprosy detection rates and the predictors. The proportion of households with piped water was associated with higher detection rates, mainly in the northeast of the municipality. Indeterminate forms were strongly associated with higher detections rates in the south, where access to health services was more established. CONCLUSIONS: GWR proved a useful tool for epidemiological analysis of leprosy in a local area, such as Duque de Caxias. Epidemiological analysis using the maps of the GWR model offered the advantage of visualizing the problem in sub-regions and identifying any spatial dependence in the local study area.

Maps of residuals from regression: their characteristics and uses in geographic research.

Bibliography: p. 59-60.

Estimation of average annual daily traffic (AADT) using geographically weighted regression (GWR) method and geographic information system (GIS)

Annual average daily traffic (AADT) is important information for many transportation planning, design, operation, and maintenance activities, as well as for the allocation of highway funds. Many studies have attempted AADT estimation using factor approach, regression analysis, time series, and artificial neural networks. However, these methods are unable to account for spatially variable influence of independent variables on the dependent variable even though it is well known that to many transportation problems, including AADT estimation, spatial context is important. ^ In this study, applications of geographically weighted regression (GWR) methods to estimating AADT were investigated. The GWR based methods considered the influence of correlations among the variables over space and the spatially non-stationarity of the variables. A GWR model allows different relationships between the dependent and independent variables to exist at different points in space. In other words, model parameters vary from location to location and the locally linear regression parameters at a point are affected more by observations near that point than observations further away. ^ The study area was Broward County, Florida. Broward County lies on the Atlantic coast between Palm Beach and Miami-Dade counties. In this study, a total of 67 variables were considered as potential AADT predictors, and six variables (lanes, speed, regional accessibility, direct access, density of roadway length, and density of seasonal household) were selected to develop the models. ^ To investigate the predictive powers of various AADT predictors over the space, the statistics including local r-square, local parameter estimates, and local errors were examined and mapped. The local variations in relationships among parameters were investigated, measured, and mapped to assess the usefulness of GWR methods. ^ The results indicated that the GWR models were able to better explain the variation in the data and to predict AADT with smaller errors than the ordinary linear regression models for the same dataset. Additionally, GWR was able to model the spatial non-stationarity in the data, i.e., the spatially varying relationship between AADT and predictors, which cannot be modeled in ordinary linear regression. ^

Population subdivision in marine environments: the contributions of biogeography, geographic distance, and discontinuous habitat to genetic differentiation in a blennioid

The relative importance of factors that may promote genetic differentiation in marine organisms is largely unknown. Here, contributions to population structure from biogeography, habitat distribution, and isolation by distance were investigated in Axoclinus nigricaudus, a small subtidal rock reef fish, throughout its range in the Gulf of California. A 408 basepair fragment of the mitochondrial control region was sequenced from 105 individuals. Variation was significantly partitioned between many pairs of populations. Phylogenetic analyses, hierarchical analyses of variance, and general linear models substantiated a major break between two putative biogeographic regions. This genetic discontinuity coincides with an abrupt change in ecological characteristics (including temperature and salinity) but does not coincide with known oceanographic circulation patterns. Geographic distance and the nature of habitat separating populations (continuous habitat along a shoreline, discontinuous habitat along a shoreline, and open water) also contributed to population structure in general linear model analyses. To verify that local populations are genetically stable over time, one population was resampled on four occasions over eighteen months; it showed no evidence of a temporal component to diversity. These results indicate that having a planktonic life stage does not preclude geographically partitioned genetic variation over relatively small geographic distances in marine environments. Moreover, levels of genetic differentiation among populations of Axoclinus nigricaudus cannot be explained by a single factor, but are due to the combined influences of a biogeographic boundary, habitat, and geographic distance.

Socio-economic, migrant and geographic differentials in coronary heart disease occurrence in New South Wales

Objective: This study examines the variation in coronary heart disease (CHD) mortality and acute myocardial infarction (AMI) by socio-economic status (SES), country of birth (COB) and geography (urban/rural) in the total population of New South Wales (Australia) in 1991-95. Method: CHD deaths and AMI are from complete enumerations of deaths and hospital admissions, respectively; and population denominators are from census information. Data are examined separately by sex, and comparisons of SES groups (based on municipalities), COB and region are analysed using Poisson regression, after adjustment for age. Results: The study identified higher risk for AMI admissions and CHD mortality in lower SES populations with significant linear trends, for both sexes, adjusted for age, region and COB. According to the population attributable fractions (PAF), 23-41% of the risk of CHD occurrence is due to SES lower than the highest quartile. The higher age-adjusted risk for CHD occurrence in rural and remote populations for both sexes, compared with urban communities, was lessened by adjustment for COB, and all but abolished when also adjusted for SES. COB analysis indicated significantly lower age-adjusted AMI admissions and CHD mortality compared with the Australian-born, Conclusions: Higher risks for CHD in rural populations compared with the capital city (Sydney) are due, in part, to lower SES, lesser migrant composition. Implications: Strategies for reducing CHD differentials should consider demographic factors and the fundamental need to reduce socio-economic inequalities, as well as targeting appropriate prevention measures.

Trends in the risk of mortality due to cardiovascular diseases in five Brazilian geographic regions from 1979 to 1996

OBJECTIVE - To analyze the trends in risk of death due to cardiovascular diseases in the northern, northeastern, southern, southeastern, and central western Brazilian geographic regions from 1979 to 1996. METHODS - Data on mortality due to cardiovascular, cardiac ischemic, and cerebrovascular diseases in 5 Brazilian geographic regions were obtained from the Ministry of Health. Population estimates for the time period from 1978 to 1996 in the 5 Brazilian geographic regions were calculated by interpolation with the Lagrange method, based on the census data from 1970, 1980, 1991, and the population count of 1996, for each age bracket and sex. Trends were analyzed with the multiple linear regression model. RESULTS - Cardiovascular diseases showed a declining trend in the southern, southeastern, and northern Brazilian geographic regions in all age brackets and for both sexes. In the northeastern and central western regions, an increasing trend in the risk of death due to cardiovascular diseases occurred, except for the age bracket from 30 to 39 years, which showed a slight reduction. This resulted from the trends of cardiac ischemic and cerebrovascular diseases. The analysis of the trend in the northeastern and northern regions was impaired by the great proportion of poorly defined causes of death. CONCLUSION - The risk of death due to cardiovascular, cerebrovascular, and cardiac ischemic diseases decreased in the southern and southeastern regions, which are the most developed regions in the country, and increased in the least developed regions, mainly in the central western region.

Regional abundance of on-premise outlets and drinking patterns among Swiss young men: District level analyses and geographic adjustments.

INTRODUCTION AND AIMS: This study investigated the associations of alcohol outlet density with specific alcohol outcomes (consumption and consequences) among young men in Switzerland and assessed the possible geographically related variations. DESIGN AND METHODS: Alcohol consumption and drinking consequences were measured in a 2010-2011 study assessing substance use risk factors (Cohort Study on Substance Use Risk Factors) among 5519 young Swiss men. Outlet density was based on the number of on- and off-premise outlets in the district of residence. Linear regression models were run separately for drinking level, heavy episodic drinking (HED) and drinking consequences. Geographically weighted regression models were estimated when variations were recorded at the district level. RESULTS: No consistent association was found between outlet density and drinking consequences. A positive association between drinking level and HED with on-premise outlet density was found. Geographically weighted regressions were run for drinking level and HED. The predicted values for HED were higher in the southwest part of Switzerland (French-speaking part). DISCUSSION AND CONCLUSIONS: Among Swiss young men, the density of outlets and, in particular, the abundance of bars, clubs and other on-premise outlets was associated with drinking level and HED, even when drinking consequences were not significantly affected. These findings support the idea that outlet density needs to be considered when developing and implementing regional-based prevention initiatives. [Astudillo M, Kuendig H, Centeno-Gil A, Wicki M, Gmel G. Regional abundance of on-premise outlets and drinking patterns among Swiss young men: District level analyses and geographic adjustments. Drug Alcohol Rev 2014;33:526-33].

Demographic and geographic vascular risk factor differences in European young adults with ischemic stroke: the 15 cities young stroke study.

BACKGROUND AND PURPOSE: We compared among young patients with ischemic stroke the distribution of vascular risk factors among sex, age groups, and 3 distinct geographic regions in Europe. METHODS: We included patients with first-ever ischemic stroke aged 15 to 49 years from existing hospital- or population-based prospective or consecutive young stroke registries involving 15 cities in 12 countries. Geographic regions were defined as northern (Finland, Norway), central (Austria, Belgium, France, Germany, Hungary, The Netherlands, Switzerland), and southern (Greece, Italy, Turkey) Europe. Hierarchical regression models were used for comparisons. RESULTS: In the study cohort (n=3944), the 3 most frequent risk factors were current smoking (48.7%), dyslipidemia (45.8%), and hypertension (35.9%). Compared with central (n=1868; median age, 43 years) and northern (n=1330; median age, 44 years) European patients, southern Europeans (n=746; median age, 41 years) were younger. No sex difference emerged between the regions, male:female ratio being 0.7 in those aged <34 years and reaching 1.7 in those aged 45 to 49 years. After accounting for confounders, no risk-factor differences emerged at the region level. Compared with females, males were older and they more frequently had dyslipidemia or coronary heart disease, or were smokers, irrespective of region. In both sexes, prevalence of family history of stroke, dyslipidemia, smoking, hypertension, diabetes mellitus, coronary heart disease, peripheral arterial disease, and atrial fibrillation positively correlated with age across all regions. CONCLUSIONS: Primary preventive strategies for ischemic stroke in young adults-having high rate of modifiable risk factors-should be targeted according to sex and age at continental level.

Reconsolidation of the soil surface after tillage discontinuity, with and without cultivation, related to erosion and its prediction with RUSLE

Site-specific regression coefficient values are essential for erosion prediction with empirical models. With the objective to investigate the surface-soilconsolidation factor, Cf, linked to the RUSLE's prior-land-use subfactor, PLU, an erosion experiment using simulated rainfall on a 0.075 m m-1 slope, sandy loam Paleudult soil, was conducted at the Agriculture Experimental Station of the Federal University of Rio Grande do Sul (EEA/UFRGS), in Eldorado do Sul, State of Rio Grande do Sul, Brazil. Firstly, a row-cropped area was excluded from cultivation (March 1995), the existing crop residue removed from the field, and the soil kept clean-tilled the rest of the year (to get a degraded soil condition for the intended purpose of this research). The soil was then conventional-tilled for the last time (except for a standard plot which was kept continuously cleantilled for comparison purposes), in January 1996, and the following treatments were established and evaluated for soil reconsolidation and soil erosion until May 1998, on duplicated 3.5 x 11.0 m erosion plots: (a) fresh-tilled soil, continuously in clean-tilled fallow (unit plot); (b) reconsolidating soil without cultivation; and (c) reconsolidating soil with cultivation (a crop sequence of three corn- and two black oats cycles, continuously in no-till, removing the crop residues after each harvest for rainfall application and redistributing them on the site after that). Simulated rainfall was applied with a Swanson's type, rotating-boom rainfall simulator, at 63.5 mm h-1 intensity and 90 min duration, six times during the two-and-half years of experimental period (at the beginning of the study and after each crop harvest, with the soil in the unit plot being retilled before each rainfall test). The soil-surface-consolidation factor, Cf, was calculated by dividing soil loss values from the reconsolidating soil treatments by the average value from the fresh-tilled soil treatment (unit plot). Non-linear regression was used to fit the Cf = e b.t model through the calculated Cf-data, where t is time in days since last tillage. Values for b were -0.0020 for the reconsolidating soil without cultivation and -0.0031 for the one with cultivation, yielding Cf-values equal to 0.16 and 0.06, respectively, after two-and-half years of tillage discontinuation, compared to 1.0 for fresh-tilled soil. These estimated Cf-values correspond, respectively, to soil loss reductions of 84 and 94 %, in relation to soil loss from the fresh-tilled soil, showing that the soil surface reconsolidated intenser with cultivation than without it. Two distinct treatmentinherent soil surface conditions probably influenced the rapid decay-rate of Cf values in this study, but, as a matter of a fact, they were part of the real environmental field conditions. Cf-factor curves presented in this paper are therefore useful for predicting erosion with RUSLE, but their application is restricted to situations where both soil type and particular soil surface condition are similar to the ones investigate in this study.

Geographic variation of the greater white-toothed shrew (Crocidura russula Hermann, 1780 Mammalia, Soricidae)

A multivariate morphometric study of the Greater white-toothed shrew (C. russula) throughout its Palearctic range was carried out to search for patterns of geographic variation within the species boundary. Burnaby's and multiple group principal component analysis allowed the adjustment of raw data with respect to within-sample allometric variation. Multivariate 'size-free' results show a stepped dine with the phenotypical trait reduction and shape change from the eastern to the western Maghreb. Pleistocene fossil mandibles proved to have low phenetic distances with eastern populations (Tunisia, east Algeria) and it is argued that their character set is the primitive condition. The ancestral Mid-Pleistocene shrews lived in a relatively more humid climate. Gee-climatic changes in the north African range during the Quaternary provoked phenetic variation of C. russula and, it can be argued, evolution of the modern western C.r. yebalensis. A historical process can thus be assumed as the main cause of this categorical variation, by segmentation of the species range due to gee-climatic events. Morphometric discontinuity within the C. russula Maghreb range is shown to be congruent with karyological and biochemical studies. Moroccan and Tunisian shrews differ, for example, in NFa chromosomes and electrophoretical traits. A stasipatric process should be invoked to explain categorical variation in the Maghreb range. Colonization and divergence of insular populations results in more or less differentiated geographic races. The populations of Ibiza and Pantelleria are close to the species threshold (Nei's D greater than or equal to 0.1). The process of speciation undergone by the Greater white-toothed shrew results in a complex pattern of geographic variation, including both allopatric and non-allopatric modes.

Computational Techniques for Spatial Logistic Regression with Large Datasets

In epidemiological work, outcomes are frequently non-normal, sample sizes may be large, and effects are often small. To relate health outcomes to geographic risk factors, fast and powerful methods for fitting spatial models, particularly for non-normal data, are required. We focus on binary outcomes, with the risk surface a smooth function of space. We compare penalized likelihood models, including the penalized quasi-likelihood (PQL) approach, and Bayesian models based on fit, speed, and ease of implementation. A Bayesian model using a spectral basis representation of the spatial surface provides the best tradeoff of sensitivity and specificity in simulations, detecting real spatial features while limiting overfitting and being more efficient computationally than other Bayesian approaches. One of the contributions of this work is further development of this underused representation. The spectral basis model outperforms the penalized likelihood methods, which are prone to overfitting, but is slower to fit and not as easily implemented. Conclusions based on a real dataset of cancer cases in Taiwan are similar albeit less conclusive with respect to comparing the approaches. The success of the spectral basis with binary data and similar results with count data suggest that it may be generally useful in spatial models and more complicated hierarchical models.

Predicting childhood sexual or physical abuse: a logistic regression geo-mapping approach to prevention.

This study investigates the degree to which gender, ethnicity, relationship to perpetrator, and geomapped socio-economic factors significantly predict the incidence of childhood sexual abuse, physical abuse and non- abuse. These variables are then linked to geographic identifiers using geographic information system (GIS) technology to develop a geo-mapping framework for child sexual and physical abuse prevention.