Summary.  Antithetic time series analysis is the solution to a most perplexing problem in mathematical statistics. When a mathematical model is fitted to serially correlated data, the parameters of the model are unavoidably biased. All forecasts that are obtained from the model are unavoidably biased and therefore diverge. The forecast reliability worsens with the forecast horizon. It is shown that the forecast bias can be dynamically reduced. This is made possible by the entirely counterintuitive discovery of antithetic time series theory that permits unbiased forecast error convergence to a constant, independent of forecast origin. The forecast error variance in each time period is the same.

Summary.  Comparing assessments on a subjective scale across countries or socio-economic groups is often hampered by differences in response scales across groups. Anchoring vignettes help to correct for such differences, either in parametric models (the compound hierarchical ordered probit (CHOPIT) model and extensions) or non-parametrically, comparing rankings of vignette ratings and self-assessments across groups. We construct specification tests of parametric models, comparing non-parametric rankings with rankings by using the parametric estimates. Applied to six domains of health, the test always rejects the standard CHOPIT model, but an extended CHOPIT model performs better. This implies a need for more flexible (parametric or semiparametric) models than the standard CHOPIT model.

Summary.  This study investigates whether and to what extent further unemployment experience for youths who are already long term unemployed imposes a penalty on subsequent labour market outcomes. We propose a flexible method for analysing the effect on wages aside from transitions from unemployment and employment within a multivariate duration model which controls for selection on observables and unobservables. We find that prolonging unemployment drastically decreases the chances of finding employment but hardly affects the quality of subsequent employment. The analysis suggests that negative duration dependence in the job finding rate is induced by negative signalling and not by human capital depreciation.

Summary.  The paper analyses the data resulting from the Italian campaign for newborns’ parents, ‘GenitoriPiù’, and focuses on the assessment of healthcare workers’ knowledge about sudden infant death syndrome. Considering two different response sets (dichotomous and polytomous), we used a Rasch model and a logistic quantile regression to analyse which demographic and professional backgrounds influenced the degree of knowledge of this topic. Significant differences between regions are evident, and the effect of training initiatives is proven as a way of rectifying these differences. With regard to professional background, the best-prepared healthcare workers are paediatricians and healthcare workers working in birth centres and family planning clinics.

Summary.  Using nationally representative survey data for Finnish employees linked to register data on their wages and work histories we find that wage effects of high involvement management (HIM) practices are generally positive and significant. However, employees with better wage and work histories are more likely to enter HIM jobs. The wage premium falls substantially having accounted for employees’ work histories, which suggests that existing studies’ estimates are upwardly biased owing to positive selection into HIM. Results using standard regression techniques are robust to propensity score matching and instrumental variables estimation. The premium also rises with the number of HIM practices and differs markedly across different types of HIM practice.

Summary.  Using data from the child supplement of the US National Longitudinal Survey of Youth, and fitting three-level random-effects models of child health and cognitive development, we test whether left-handed children have different outcomes from those of their right-handed counterparts. The health measures cover both physical health and mental health, and the cognitive development test scores span vocabulary, mathematics, reading and comprehension. Overall we find little evidence to suggest that left-handed children have a significantly higher probability of experiencing injury, illness or behavioural problems. In contrast, we find that left-handed children have significantly lower cognitive development test scores than right-handed children for all areas of development with the exception of reading. Moreover, we find no strong evidence that the left-handedness effect differs by gender or age.

Summary.  A gap in cognitive skill between richer and poorer children is evident from a very early age. Some studies have also suggested that highly able children from disadvantaged homes are overtaken by their rich but less able peers before the age of 10 years, in terms of their cognitive skill. This finding has become a widely cited ‘fact’ within the academic literature and has had a major influence on public policy and political debate. We show that this finding is vulnerable to a spurious statistical artefact known as regression to the mean and we propose the application of an alternative methodology to address this problem. After applying some simple adjustments for regression to the mean to data from the Millennium Cohort Study, we no longer find convincing evidence that able but disadvantaged pupils fall behind their more advantaged but less able peers.

Summary.  The paper investigates the link between area-based socio-economic deprivation and the incidence of child pedestrian casualties. The analysis is conducted by using data for small spatial zones within major British cities over the period 2001–2007. Spatial longitudinal generalized linear mixed models, estimated by using frequentist and Bayesian approaches, are used to address issues of confounding, spatial dependence and transmission of deprivation effects across zones (i.e. interference). The results show a consistent strong deprivation effect across model specifications. The incidence of child pedestrian casualties in the most deprived zones is typically greater than 10 times that in the least deprived zones. Modelling interference through a spatially auto-regressive covariate uncovers a substantially larger effect.

Summary.  We show that changes in assortative mating patterns along the dimensions of age, ethnicity, religion and education are not responsible for the increasing marital instability over the last four decades in Austria. Without the rise in the age at marriage, divorce rates would be considerably higher. Immigration and secularization, and the resulting supply of spouses with diverse ethnicity and religious denominations had no overall effect on divorce rates. Countervailing effects—in line with theoretical predictions–offset each other. The rise in the incidence of divorce is most probably caused by increased social acceptance of divorce.

Summary.  Composite indicators aggregate a set of variables by using weights which are understood to reflect the variables’ importance in the index. We propose to measure the importance of a given variable within existing composite indicators via Karl Pearson's ‘correlation ratio’; we call this measure the ‘main effect’. Because socio-economic variables are heteroscedastic and correlated, relative nominal weights are hardly ever found to match relative main effects; we propose to summarize their discrepancy with a divergence measure. We discuss to what extent the mapping from nominal weights to main effects can be inverted. This analysis is applied to six composite indicators, including the human development index and two popular league tables of university performance. It is found that in many cases the declared importance of single indicators and their main effect are very different, and that the data correlation structure often prevents developers from obtaining the stated importance, even when modifying the nominal weights in the set of non-negative numbers with unit sum.

Summary.  Financial institutions and regulatory agencies direct much effort and expertise towards estimating probabilities of bankruptcy. By comparison, the techniques that are used to evaluate probability estimates have attracted little attention and remain somewhat ad hoc. The most common approach is to count misclassifications, based on an arbitrary classification threshold. Some use is made of conventional probability score functions, such as the Brier score (from meteorology) but these are not standard practice. Our purpose is to introduce a family of economic probability score functions designed to capture the utility obtained by a user, with a specified utility function, who uses the estimated probabilities to make hypothetical bets against a rival forecaster or model. The conceptual appeal of these score functions is that probability forecasts are evaluated neither in abstract, nor in isolation, but by whether they would hypothetically have ‘made money’ for a given user, with specified risk aversion, against comparable forecasts or market betting prices.

Summary.  A large proportion of electoral analyses using geography are performed on a small area basis. In each new election there are always modifications to the previously existing polling units. The use of past voting results in small area aggregate data electoral forecasting models and political analyses therefore requires establishing a correspondence between old and new polling units. Traditionally, the task of tracking changes to assign an electoral history to the new units properly has been carried out by hand, comparing unit codes and census figures. This is an extremely cumbersome task that cannot always be performed, as when a massive (geographically intense) reorganization of polling unit boundaries takes place. Nowadays, however, assisted by the increasing availability of geographical data, this chore could be easily automated and even improved with the help of spatial statistical software. The paper suggests several methods for allocating votes by using geographical information systems tools and shows the effectiveness of spatial strategies. These approaches will permit electoral pollsters and forecasters to solve the issue efficiently and to apply the most successful electoral forecasting techniques that are currently in use and will help electoral geographers with the problem of comparing spatial aggregate electoral data from different elections. The relevance of the analysis, nevertheless, goes beyond electoral data, as the reallocation of data from one set of administrative units onto another arises in many applications. The geometric approach is proposed as a natural substitute for the classical approach and three additional approaches (centroid, surface and compositional) are also suggested, exploiting the spatial patterns that electoral outcomes display. The relative performance of the various methods is assessed in three real data instances. The results suggest that the surface approach, which obtains past voting outcomes in each polling unit by averaging their vote proportion interpolations, is the most suitable procedure.

Summary.  To study the effects of out-of-school learning we use data on children who stayed in boarding houses while attending public elementary schools in Sweden in the 1940s. The out-of-school environment at the boarding houses could be considered as being more conducive to learning than the home environment: the pupils at the boarding houses had daily scheduled time for doing their homework with the assistance of a junior school teacher and, in addition, they had access to a small library. The placement at boarding houses was based on the distance from their place of residence to the nearest school and thus had no direct connection to the pupils’ skills, which simplifies the empirical analysis based on register data. We find that the more conducive learning environment equalized skills at school leaving age, and the effect was greater for children with poor initial ability.

Summary.  Although surveys routinely ask respondents to evaluate various aspects of their life on an ordered scale, there is concern about interpersonal comparability of these self- assessments. Statistically, the problem is one of identification in ordered response models with heterogeneous thresholds. As a solution to the identification problem, King and his colleagues proposed the use of anchoring vignettes, namely brief descriptions of hypothetical people or situations that survey respondents are asked to evaluate on the same scale as they use to rate their own situation. Although vignettes have been introduced in several social surveys and are increasingly employed in a variety of fields, the reliability of this approach hinges crucially on the validity of the assumptions of response consistency and vignette equivalence. The paper proposes a joint test of these key assumptions based on the fact that the underlying statistical model is overidentified if the two assumptions hold. Monte Carlo results show that the test proposed has good size and power properties in finite samples. We apply our test to self- assessment of pain by using data from the first wave of the Survey of Health, Ageing and Retirement in Europe. We find that, when using only one of the three available vignettes, or when the test is carried out separately by subgroups of respondents, the overidentifying restrictions are less likely to be rejected.

Summary. In an effort to meet its obligations under the Kyoto Protocol, the European Union has introduced a cap-and-trade scheme where mandated companies are allocated permits to emit carbon dioxide. Financial markets have developed that allow companies to trade these carbon permits. Several recent studies have attempted to model their prices. There are many institutional features that potentially impact on carbon prices and associated financial futures, making such an undertaking quite different from modelling conventional financial assets traded in mature markets. We forecast the carbon markets by using dynamic model averaging, which is a recently developed statistical method which has three advantages over conventional approaches. First, it allows the coefficients on the predictors in a forecasting model to change over time. Second, it allows for the entire forecasting model to change over time. Third, it surmounts statistical problems which arise from the large number of potential predictors that can explain carbon prices. Our empirical results indicate that there are both important policy and statistical benefits with our approach. Statistically, we present strong evidence that there is substantial turbulence and change in the carbon markets. We find that dynamic model averaging can model these features and forecast accurately compared with conventional approaches.

Summary.  We use univariate and multivariate singular spectrum analyses to predict the rate of inflation as well as changes in the direction of inflation time series for the USA. We use consumer price indices and realtime chain-weighted gross domestic product price index series in these prediction exercises. Moreover, we compare our out-of-sample, h-step-ahead moving prediction results with other prediction results based on methods such as the activity-based non-accelerating inflation rate of unemployment Phillips curve, auto-regressive AR(p) model, the dynamic factors model and random-walk models with the last as a naive forecasting method. We use short-run (quarterly) and long-run (1–6 years) time windows for predictions and find that multivariate singular spectrum analysis outperforms all other competing prediction methods. Also, we confirm the results of earlier studies that prediction of the rate of inflation in the USA during the period of the ‘Great Moderation’ is less challenging compared with the more volatile inflationary period of 1970–1985.

Summary.  There is a debate surrounding the implications of big charities’ increasing dominance of total charitable income, but no empirical work which assesses whether indeed big charities are becoming increasingly dominant. We provide this assessment from both cross-sectional and longitudinal perspectives, using a panel data set with information on charities’ income in England and Wales between 1997 and 2008. From a cross-sectional perspective, examining trends in income concentration ratios, there is no evidence that the biggest charities account for a growing share of total charity income over the period of analysis. However, the longitudinal perspective, which relates income growth over the period to initial size, shows that initially large charities have significantly higher median relative growth rates than the initially small. Substantively, these results are relevant to government plans for the ‘Big Society’, which rest in part on the ability of smaller, community-based charities as well as the bigger voluntary bodies to thrive and grow. Methodologically, for studies which examine trends in the distribution of income, these results illustrate the additional insights that are provided by the longitudinal perspective which cannot be inferred from repeated cross-sectional information.

Summary.  Linguistic corpora are databases of text which are linguistically marked up or otherwise structured and designed to be representative of a specific language. The growing availability of such corpora has brought with it opportunities for statistical analysis. The paper develops and uses statistical approaches to address questions pertaining to an important linguistic phenomenon: the use of different syntactic alternatives. We present a model-selection-based approach for determining possible driving attributes affecting verb complementation for written sentence constructions using the verb ‘give’ in three varieties of English. We are interested in explaining the choice of alternatives in terms of a variety of sentence level linguistic features such as the meaning of the verb, in addition to the country of origin.

Summary.  There are several challenges to testing the effectiveness of group-therapy-based interventions in alcohol and other drug use treatment settings. Enrolment into alcohol and other drug use therapy groups typically occurs on an open (rolling) basis. Changes in therapy group membership induce a complex correlation structure between client outcomes, with relatively small numbers of clients attending each therapy group session. Primary outcomes are measured post treatment, so each datum reflects the effect of all sessions attended by a client. The number of post-treatment outcomes assessments is typically very limited. The first feature of our modelling approach relaxes the assumption of independent random effects in the standard multiple-membership model by employing conditional auto-regression to model correlation in random-therapy-group session effects associated with clients’ attendance of common group therapy sessions. A second feature specifies a longitudinal growth model under which the posterior distribution of client-specific random effects, or growth parameters, is modelled non-parametrically. The Dirichlet process prior helps to overcome limitations of standard parametric growth models given limited numbers of longitudinal assessments. We motivate and illustrate our approach with a data set from a study of group cognitive behavioural therapy to reduce depressive symptoms among residential alcohol and other drug use treatment clients.