Bias corrected estimates in multivariate student t regression models

The t distribution has proved to be a useful alternative to the normal distribution especially When robust estimation is desired. We consider the multivariate nonlinear Student-t regression model and show that the biased of the estimates of the regression coefficients can be computed from an auxiliary generalized linear regression. We give a formula for the biases of the estimates of the parameters in the scale matrix, which also can be computed by means of a generalized linear regression. We briefly discuss some important special cases and present simulation results which indicate that our bias-corrected estimates outperform the uncorrected ones in small samples.     

Loss prediction based on run-off triangles

Loss reserving is an important subject of actuarial mathematics. It aims at the prediction of future losses caused by claims which have incurred in the past but have not yet been closed. The problem of predicting such losses is particularly important in liability insurance. More generally, it is most relevant with respect to the new regulatory requirements for insurance companies operating in the European Union, which are known as Solvency II.     

Maximum-likelihood and marginal-sum estimation in some particular collective models

In the present paper we consider collective models in risk theory and their thinning and decomposition. We give three applications to reinsurance, multiplicative tariffs, and loss reserving. For each of these applications we show how maximum-likelihood and marginal-sum estimation can be used to estimate the parameters.     

Demographic change, endogenous labor supply and the political feasibility of pension reform

Options for reforming unfunded public pension schemes that are now being discussed all share the feature that the burden induced by demographic change would be shifted towards presently living and away from unborn generations. Existing models of the political economy of pension reform can not explain why such reform options are being discussed at all. We present an alternative model in which the possibility of evasion of workers from payment of social security taxes is taken into account by modelling a labor supply function. It turns out that the burden of demographic change may fall completely or at least predominantly on the pensioners. Thus this type of model can much better explain recent trends in legislature on unfunded public pension systems in industrial democracies. Received: 7 January 1999/Accepted: 20 December 1999     

Children,Kitchen, Church: does ethnicity matter?

Ethnic differences in the allocation of non-market time are important, as they may shed more light on the integration level of ethnic minorities and on the factors that affect both household productivity and ethnic identity. In this paper we examine the role of ethnicity and gender by analyzing differences in the time spent on a range of activities employing the 2000 UK Time Use Survey. Based on the economics of religion and identity economic models, we hypothesize that if ethnic minority women have lower opportunity costs of time and a strong ‘ethnic’ or ‘traditionally female’ identity, they will engage more in ‘traditional’ home activities. Double-hurdle regression results indicate that while the effect for childcare is not significant when estimated for parents only, non-white women spend significantly more time on food management and particularly religious activities than white women, with the greatest effect of the latter being for Pakistani and Bangladeshi women.     

On (n-l)-wise and joint independence and normality of n Random variables: an example

An example is given of a vector of n random variables such that any (n-1)-dimensional subvector consists of n-1 independent standard normal variables. The whole vector however is neither independent nor normal.     

Invited article: Face,voice, and body in detecting deceit

Studies based on mean accuracy of a group of subjects suggest that most observers do no better than chance in detecting the lies of others. We argue that a case-by-case methodology, like that used in polygraphy studies may be more useful. Three behavioral measures (two kinds of smiles and pitch) were used to make predictions about the lying or truthfulness of each of 31 subjects. A case-by-case analysis of the hits and misses achieved in this way yielded an over-all accuracy of 86%. The effect on lie detection accuracy of individual differences in the use and control of different behavioral channels is discussed.Paul Ekman's work is supported by a Research Scientist Award from the National Institute of Mental Health (MH 06092).     

Robust Bayesian methods in simple ANOVA models

A robust Bayesian analysis in a conjugate normal framework for the simple ANOVA model is suggested. By fixing the prior mean and varying the prior covariance matrix over a restricted class, we obtain the so-called HiFi and core region, a union and intersection of HPD regions. Based on these robust HPD regions we develop the concept of a ‘robust Bayesian judgement’ procedure. We apply this approach to the simple analysis of variance model with orthogonal designs. The example analyses the costs of an asthma medication obtained by a two-way cross-over study.