Income inequality has been on the rise in many industrialised countries since around the 1980’s. In Sweden the increase of income inequality has been particularly large. This in spite of Sweden’s extensive redistribution system and public policy that prioritize equality among its population. This paper investigates a potential factor for the rise in inequality that is yet fairly unexplored, namely immigration. As inequality has increased in Sweden, so has also immigration. Sweden experienced large refugee inflows after the 1970’s, the largest flow consisting of circa 100 000 Yugoslavs during the Bosnian war. This study provides indications on what way immigration shapes the income distribution and lays the ground for prospective studies. Results show that the inflow of new migrants during the early 1990’s in Sweden raises income inequality and it is almost entirely due to increased dispersion in the lower tail of the income distribution.

Paired comparison models in loglinear form are generalised linear models and can be fitted using the IWLS algorithm. Unfortunately, the design matrices can become very large and thus a method is needed to reduce computational load (relating to both space and time). This paper discusses an algorithm for fitting loglinear paired comparison models in the presence of many nuisance parameters which is based on partition rules for symmetric matrices and takes advantage of the special structure of the design matrix in Poisson loglinear models. The algorithm is implemented as an R function. Some simple examples illustrate its use for fitting both paired comparison models and (multinomial) logit models. (author's abstract) / Series: Research Report Series / Department of Statistics and Mathematics

Climate change is being observed through increased average temperatures world-wide, as well as through increased frequency of extreme events, such as floods and droughts. As climate is an uncontrollable yet essential input in the agriculture industry, the impact of climate change may have on crop production in Saskatchewan is of importance. The main objective of this study is to investigate how farmers adapt to climate change by switching their crop mix, and how this crop mix may change under future climate change scenarios. A fractional multinomial logit (FMNL) model was used to assess how total area of cropland has changed over a thirty year time period. The panel data included variables to represent the land characteristics of Saskatchewan (i.e. the three major soil zones - Black, Dark Brown and Brown), climatic variables to represent average monthly temperature and precipitation, and price and policy variables in order to assess how average seeded area of each crop group changed. With these results, a simple simulation model was developed to evaluate how the area of each crop group in a base year comparison (2000) would change under future climate scenarios for each soil zone. The results from the FMNL model indicate that crop allocation depends largely on the price of other crop groups and temperatures in the spring (April) and summer (July). Climate plays and important role in the major crop groups, such as wheat, canola and pulses. Cool, dry springs are the ideal conditions when choosing nearly all crops, while hot, wet summers increase the choice to leave land to summerfallow. Policy and the different soil zones also play a significant role in area allocation decisions. Changes in policies such as the removal of the Crow’s Nest Pass Agreement, and the removal of oats from the Canadian Wheat Board (CWB) marketing, had a negative impact on the choice to grow wheat, as expected. The different soil zones in Saskatchewan played an important role in area allocation for a majority of the crops, having a negative effect on the choice of wheat over every other crop group except pulses and summerfallow. Three climate change scenarios were simulated for each soil zone and compared to a base area (year 2000 area seeded) of crop groups. The findings from the projected changes in climate indicate that the area allocated to wheat will continue to decrease into the future, following current trends. The average projected decline in wheat area from the base years by 2099 ranges between 3.5% to 4.6% in the Black soil zone, between 2.7% and 2.9% in the Dark Brown and 2.7% to 4% in the brown soil zone, depending on climate change scenario. Interestingly, the area left to summerfallow is projected to increase over the future climate change scenarios. The choice of wheat is preferred over pulses, feed and forages, while the choice of specialty oilseeds (flaxseed, mustard seed and canary seed) are projected to become preferred over wheat in the future. The major conclusion from this research are: (i) following current trends, the area devoted to spring wheat and durum wheat would continue to decline into the future; (ii) Area devoted to wheat remains a preferred choice over pulses, feed and forages while specialty oilseeds represent a viable alternative choice to wheat and (iii) most significantly, summerfallow area would increase. This is in contrast to the current trend of declining summerfallow area as a result of tighter crop rotations. This finding was observed throughout all three soil zones as well as for all three climate change projection periods. This will have major implications on individual farmers as well as the economy in Saskatchewan, as summerfallow does not produce a crop in the year it is chosen. It is therefore important to determine a possible new crop mix that would benefit from the projected change in climate. This study could be improved by including a measure of profitability for each crop group and introducing a new crop group that is better suited to the projected change in climate in Saskatchewan.

The thesis comprises five papers and an introductory overview of applied models and methods. The papers concern interdependences and interrelations in models applied to empirical analyses of various problems related to production, consumption, location and trade. Among different definitions of 'structural analysis' one refers to the study of the properties of economic models on the assumption of invariant structural relations, this definition is close to what is aimed at in lire present case. Although the subjects cover widely differing aspects of the economic system, applied models and methods, i.e. entropy maximizing (information minimizing) models and random utility maximizing models, are in many cases closely connected. Tlic first paper reports on a regional input-ouput study applied to Norrbotten, Sweden. The paper is mainly concentrated on developing and estimating an econometric model, describing the structural interdependences in the Norrbotten economy. The chapter is composed of three parts. The first part concerns the theoretical basis of the model, the main fields of application and principal problems in connection with the estimation. The core of the estimated model is defined by the intersectoral dependences in the Norrbotten economy. This model can be viewed as a part of a more general model of the regional economy, and such a general model is briefly outlined. The second part reports on the collection and arranging of data, and the methods used for the estimation of the model. In the third part the results are presented. A special interest concerns the effects of production changes in the basic industries in the county, as to the expected impact on different industries and occupational groups. The second paper concerns some aspects of the problem of predicting trade flows in the forest sector. The model, based on information theory, is predicting current trade flows by adjusting the historical, a priori, trade flows to satisfy current export and import totals. In the third paper an entropy model is employed to decompose the interregional and intraregional employment change in Sweden and Stockholm, during the period 1960 - 1980, into effects attributed to regions (zones), industries, occupations and interaction effects. The fourth paper presents an empirical analysis of housing choice, based on individual data of households in Stockholm. The consumer choice is regarded as a complex choice from a finite set of discrete alternatives and a probabilistic choice mode! (multinomial logit) is employed, where secondary dwelling is included in the housing choice decision. In the final paper spectral analysis is used for identifying the significant components of cycle behaviour in time series of Swedish exports of forest products over a twenty year time period. / digitalisering@umu

There has been a noticeable swing towards school pupils being driven to and from school, and away from active modes like walking and cycling, in recent decades. This has had a number of side effects. Less reliance on active modes of transport has been a contributing factor in the reducing levels of physical activity for school children. Traffic volumes associated with school trips have also increased. This increased has tended to contribute to an increase in traffic congestion, adverse environmental effects and reductions in levels of sustainability. School trip traffic contributes specifically to congestion at school gates. Schools have been identified as having significant effects on the transportation system adjacent to them. Schools which seek Resource Consents for new or changed activities are often being required to take measures to mitigate their adverse effects The purpose of this study is to explore the factors contributing to primary school pupils' travel choices. This will help to identify travel choice patterns which may, in turn, be useful in developing policies and planning initiatives which contribute to achieving an efficient and sustainable transport system. A range of literature relevant to school and general commuting travel demand was reviewed. A case study involving the pupils of twenty two Christchurch primary schools was carried out. Pupils and their parents were surveyed to establish mode choices and the factors influencing those choices. The study found that between 55% and 60% of pupils surveyed travel to and from school by car. 30% to 35% walk or scooter, and 5% to 7% cycle. This compares with 34% travelling by car in the late 1980s. In addition, a greater proportion of those pupils who walk, scooter or cycle to school are accompanied by an adult than in the past. The results of the study also suggested that School Travel Plans, when combined with the energy and commitment to implement them can have a significant effect on school travel choices. As part of the case study, parents were asked to rank the importance of a number of factors which could influence choices regarding their children's school travel. The responses from parents identified safety concerns, regarding both road and personal safety, as the major factor behind decisions regarding their children's travel choices. Time constraints coupled with the complexity of travel requirements of many families were identified as significant factors. Multinomial Logit Models for both mode choice and pupils travel independence were then produced for both the journey to and from school. These models were based on the results of the case study. The models produced indicate that, at a school level, there is a correlation between increasing school roll and an increasing proportion of pupils travelling by car. A slight negative correlation between school decile and car usage was also indicated. This is contrary to the normally accepted understanding that in most transport situations there is a positive correlation between increasing affluence and car usage. Superior model results were obtained at a disaggregated individual level, using nine variables relating to the school, the neighbourhood, and the home, than the results obtained using the school based variables of. However, it is not considered that the effort required to obtain information on the additional variables is justified when estimating mode choices of pupils at an individual school. It is therefore recommended that a model using Decile, Average Age, and School Roll variables be used to estimate mode choices at an individual school. At a family level, there was a strong positive correlation between distance from school, age of the pupils, and the number of major roads between school and home, and car usage. It became apparent that the decisions made regarding children's school travel are very complex. Families juggle a number of factors, many of which are in conflict with one another. For example a desire to care for the environment may be in conflict with the demand to get the children to school, and get to work on time. This complex interrelationship between factors has resulted in some instances where normally accepted "Rules of Thumb", such as the understanding that increased car usage is generally associated with increasing wealth, do not appear to be applicable to school travel. The complexity of interrelationships has further meant that it has not been possible to quantify the impact of any one factor on its own.

Fethiye-G&ouml / cek area is one of the nine coastal Specially Protected Area (SPA) in Turkey. Since mid-80&rsquo / s G&ouml / cek town has developed to be a yachting center, and the bays of G&ouml / cek have acquired a well-earned international fame as a paradise for boating vacations. However, the uncontrolled yachting in this bay area presents a growing pressure on the environment, and the coastal and marine ecosystem. In this thesis a computer model for simulating the movements of yachts in G&ouml / cek Bays is developed. The computer model uses the Multinomial Logit Model (MNL) to find the probabilities for the boaters to select the next bay to visit. The model predicts the number of boats in each bay at the end of a day, the number of boats visited each bay during the day and the distribution of boater categories among the bays throughout the simulation time. In order to get the data needed for the inputs, a questionnaire was formed, and a detailed survey was carried out in G&ouml / cek Bays. In addition to the questionnaires, the number of the boats anchored were also observed in the field studies. The model is applied to the G&ouml / cek Bays and the results obtained are compared with the data obtained in the field. In the following years, the yacht movements and distributions at various anchor locations can be predicted with this model. These predictions will be useful in a future management plan that aims to control of yacht movement and anchoring.

abstract: It has been suggested that directed forgetting (DF) in the item-method paradigm results from selective rehearsal of R items and passive decay of F items. However, recent evidence suggested that the passive decay explanation is insufficient. The current experiments examined two theories of DF that assume an active forgetting process: (1) attentional inhibition and (2) tagging and selective search (TSS). Across three experiments, the central tenets of these theories were evaluated. Experiment 1 included encoding manipulations in an attempt to distinguish between these competing theories, but the results were inconclusive. Experiments 2 and 3 examined the theories separately. The results from Experiment 2 supported a representation suppression account of attentional inhibition, while the evidence from Experiment 3 suggested that TSS was not a viable mechanism for DF. Overall, the results provide additional evidence that forgetting is due to an active process, and suggest this process may act to suppress the representations of F items. / Dissertation/Thesis / Ph.D. Psychology 2011

Statistical association tests of quantitative traits have been widely used in the past decade, to locate loci associated with a disease trait. For instance, Genome Wide Association Studies (GWAS) have led to tremendous success in finding susceptible genes or associated loci. However, most of the past studies were based on unrelated samples focusing on quantitative or qualitative traits. The analysis of polychotomous traits in family samples is very challenging. This dissertation describes three projects related to methods to conduct association tests beyond continuous traits, such as multinomial traits, bivariate traits, and tests involving haplotypes. The first project focuses on developing a statistical approach to test the association between common or low-frequency variants with a multinomial trait in family samples. It is an important issue because there is no computer efficient software available for this type of question. We employ Laplace approximation in conjunction with an efficient grid-search strategy to obtain an approximate maximum log-likelihood function and the Maximum Likelihood Estimate (MLE) of the variance component. We also successfully incorporate the kinship matrix to adjust for the familial correlation, based on a regression framework. Extensive simulation studies are performed to evaluate the type-I error rate and power in scenarios with causal variant with different Minor Allele Frequency (MAF). In the second project, we propose an approach to test the association between a genetic variant and a bivariate trait arising from a combination of a quantitative and a binary trait in family samples, based on Extended Generalized Estimating Equations (EGEE). Multiple phenotype-genotype association tests are often reduced to univariate tests, decreasing efficiency and power. Our approach is shown to be much more powerful and efficient than univariate association tests adjusted for multiple testing. The third project involves the development of a general framework for meta-analysis of haplotype association tests, applicable to both unrelated and family samples. Although meta-analysis has been widely used in single-variant and gene-based tests, there are few existing methods to meta-analyze haplotype association tests. A predominant advantage of our novel approach is that it accommodates cohort-specific haplotypes as well as haplotypes common to all cohorts. The cohort participants may be either related or unrelated. Our approach consists of two stages: in the first stage, each cohort performs a haplotype association test, reports the estimates of effect size, variance, haplotypes, and their frequency. In the second stage, a generalized least square method is applied to combine the results of all the cohorts into one vector of meta-analysis coefficients. Our approach is shown to have the correct type-I error rate in scenarios with different between and within cohort variation. We also present an application to exome-chip data from a large consortium. Through the three projects, we are able to tackle the problem of conducting association tests for non-continuous traits in family samples. All the approaches achieve the correct type-I error rate and are computationally efficient. We hope these approaches will not only facilitate analyses of categorical traits in family samples, but will also provide a basis for future methodological development of statistical approaches for non-continuous traits.

In this paper we study data on discrete labor market transitions from Austria. In particular, we follow the careers of workers who experience a job displacement due to plant closure and observe - over a period of 40 quarters - whether these workers manage to return to a steady career path. To analyse these discrete-valued panel data, we apply a new method of Bayesian Markov chain clustering analysis based on inhomogeneous first order Markov transition processes with time-varying transition matrices. In addition, a mixtureof- experts approach allows us to model the probability of belonging to a certain cluster as depending on a set of covariates via a multinomial logit model. Our cluster analysis identifies five career patterns after plant closure and reveals that some workers cope quite easily with a job loss whereas others suffer large losses over extended periods of time.

abstract: Mixture experiments are useful when the interest is in determining how changes in the proportion of an experimental component affects the response. This research focuses on the modeling and design of mixture experiments when the response is categorical namely, binary and ordinal. Data from mixture experiments is characterized by the perfect collinearity of the experimental components, resulting in model matrices that are singular and inestimable under likelihood estimation procedures. To alleviate problems with estimation, this research proposes the reparameterization of two nonlinear models for ordinal data -- the proportional-odds model with a logistic link and the stereotype model. A study involving subjective ordinal responses from a mixture experiment demonstrates that the stereotype model reveals useful information about the relationship between mixture components and the ordinality of the response, which the proportional-odds fails to detect. The second half of this research deals with the construction of exact D-optimal designs for binary and ordinal responses. For both types, the base models fall under the class of Generalized Linear Models (GLMs) with a logistic link. First, the properties of the exact D-optimal mixture designs for binary responses are investigated. It will be shown that standard mixture designs and designs proposed for normal-theory responses are poor surrogates for the true D-optimal designs. In contrast with the D-optimal designs for normal-theory responses which locate support points at the boundaries of the mixture region, exact D-optimal designs for GLMs tend to locate support points at regions of uncertainties. Alternate D-optimal designs for binary responses with high D-efficiencies are proposed by utilizing information about these regions. The Mixture Exchange Algorithm (MEA), a search heuristic tailored to the construction of efficient mixture designs with GLM-type responses, is proposed. MEA introduces a new and efficient updating formula that lessens the computational expense of calculating the D-criterion for multi-categorical response systems, such as ordinal response models. MEA computationally outperforms comparable search heuristics by several orders of magnitude. Further, its computational expense increases at a slower rate of growth with increasing problem size. Finally, local and robust D-optimal designs for ordinal-response mixture systems are constructed using MEA, investigated, and shown to have high D-efficiency performance. / Dissertation/Thesis / Doctoral Dissertation Industrial Engineering 2016