Parameter selection in modified histogram estimates

A multivariate modified histogram density estimate depending on a reference density g and a partition P has been proved to have good consistency properties according to several information theoretic criteria. Given an i.i.d. sample, we show how to select automatically both g and P so that the expected L ₁ error of the corresponding selected estimate is within a given constant multiple of the best possible error plus an additive term which tends to zero under mild assumptions. Our method is inspired by the combinatorial tools developed by Devroye and Lugosi [Devroye, L. and Lugosi, G., 2001, Combinatorial Methods in Density Estimation (New York, NY: Springer–Verlag)] and it includes a wide range of reference density and partition models. Results of simulations are also presented.     

Somebody versus nobody: An exploration of the role of celebrity status in an election

This study examines the role celebrity status may play in potential voters’ evaluation of a political candidate presented in a newspaper article. Participants indicated greater intention to vote for a candidate who was a recognizable Hollywood actor than an unknown candidate in a political race, regardless of how substantive the political information provided about the candidate was. This suggests that familiarity with a celebrity can act as a heuristic in peripheral processing. Younger people were more likely to vote for a celebrity candidate than older voters, but how liberal or conservative participants are was not a significant factor in the decision to vote for the celebrity. Nor did participants’ need for cognition or level of political involvement predict intention to vote for the celebrity, suggesting that celebrity status is meaningful to motivated and thoughtful voters as well as those who are less motivated and informed. The possibility is raised that this could be an indication of celebrity status being used as a component of deliberate political decision-making, and future research in this direction is suggested.     

Lorenz non-consistent welfare and inequality measurement

Typical welfare and inequality measures are required to be Lorenz consistent which guarantees that inequality decreases and welfare increases as a result of a progressive transfer. We explore the implications for welfare and inequality measurement of substituting the weaker absolute differentials and deprivation quasi-orderings for the Lorenz quasi-ordering. Restricting attention to distributions of equal means, we show that the utilitarian model – the so-called expected utility model in the theory of risk – does not permit one to make a distinction between the views embedded in the differentials, deprivation and Lorenz quasi-orderings. In contrast it is possible within the dual model of M. Yaari (Econometrica 55 (1987), 99–115) to derive the restrictions to be placed on the weighting function which guarantee that the corresponding welfare orderings are consistent with the differentials and deprivation quasi-orderings respectively. Finally we drop the equal mean condition and indicate the implications of our approach for the absolute ethical inequality indices.     

On edge orienting methods for graph coloring

We consider the problem of orienting the edges of a graph so that the length of a longest path in the resulting digraph is minimum. As shown by Gallai, Roy and Vitaver, this edge orienting problem is equivalent to finding the chromatic number of a graph. We study various properties of edge orienting methods in the context of local search for graph coloring. We then exploit these properties to derive four tabu search algorithms, each based on a different neighborhood. We compare these algorithms numerically to determine which are the most promising and to give potential research directions.     

A regional solution to the strategy and structure of multinationals

The transnational solution developed by Bartlett and Ghoshal is shown to be suitable for only a few special cases of multinational enterprise (MNE) strategy and structure. As MNEs have most of their assets and sales within their home region, they are in need of regional, not transnational strategy and structure. Here we provide data on the regional dimension of assets and sales of the world’s largest 500 multinationals. We explore how the empirical reality of a regional concentration of assets and sales imposes a regional solution, rather than the transnational solution.     

Multistate life table with duration dependence: An application to Hungarian female marital history

Multistate life table methods have been recognized as an excellent tool in the analysis of many types of transitions. Yet demographers have never been completely satisfied with the Markovian assumption and have stressed the importance of population heterogeneity. In marital status analysis, for example, the time spent in the current status is thought to be of first importance in determining transitions, but computational problems linked to the introduction of duration prevented any estimation of the bias in life table calculations arising from its omission. Building on recent developments in multistate demography, and using data from the 1984 Hungarian microcensus, this paper analyzes the impact that the introduction of duration-specific transitions has on the results of a multistate life table analysis of marital dissolution. The results show that the inclusion of duration has its greatest impact on the distribution of the stationary population between ages 25 and 35.Les tables de survie multi-états ont été reconnues comme un très bon instrument pour l'analyse de nombreux types de transitions. Cependant les démographes n'ont jamais été complètement satisfaits des hypothèses markoviennes et ont insisté sur l'importance de l'hétérogénéité des populations étudiées. Dans l'analyse par état matrimonial, par exemple, le temps passé dans l'état étudié est considéré comme de première importance dans la détermination des transitions. Cependant des problèmes de calcul liés à l'introduction de cette durée avaient empêché toute estimation des biais venant de son omission dans le calcul des tables de survie. Utilisant des développements récents en démographie multi-état et utilisant des données du micro-recensement hongrois de 1984, cet article analyse l'impact que l'introduction de transitions dépendant de la durée, a sur une analyse multi-état des dissolutions de mariages. Les résultats montrent que l'inclusion de la durée a un effet maximum sur la distribution de la population stationnaire, entre les âges de 25 at 35 ans.     

Corporate communications managers in large firms: new challenges

The new organisational form of large corporations (horizontal organisation), is affecting the communications job. This corporate transformation leads to new challenges for the corporate communications manager and his team. On the one hand, he has to develop a fully integrated strategy which enables the different forms of communication to contribute to the success of the firm's mission (a common goal). On the other hand, he has to provide small business units, led by strong front-line managers, with professional services. The analysis of several companies will show that the position of the corporate communications department is subject to various pressures and that the development of the `communications function' is calling into question the corporate communications manager's `profession'.     

Normalité asymptotique d'estimateurs convergents du mode conditionnel

We state sufficient conditions for asymptotic normality of convergent estimates of the conditional mode, irrespective of data dependence, and give an application to α-mixing stationary processes.     

Sharing beliefs and the absence of betting in the Choquet expected utility model

Choquet expected utility maximizers tend to behave in a more “cautious” way than Bayesian agents, i. e. expected utility maximizers. We illustrate this phenomenon in the particular case of betting behavior. Specifically, consider agents who are Choquet expected utility maximizers. Then, if the economy is large, Pareto optimal allocations provide full insurance if and only if the agents share at least on prior, i. e., if the intersection of the core of the capacities representing their beliefs is non empty. In the expected utility case, this is true only if they have a common prior. Received: July 2000; revised version: May 2001