Controlling the size of multivariate outlier tests with the MCD estimator of scatter

Multivariate outlier detection requires computation of robust distances to be compared with appropriate cut-off points. In this paper we propose a new calibration method for obtaining reliable cut-off points of distances derived from the MCD estimator of scatter. These cut-off points are based on a more accurate estimate of the extreme tail of the distribution of robust distances. We show that our procedure gives reliable tests of outlyingness in almost all situations of practical interest, provided that the sample size is not much smaller than 50. Therefore, it is a considerable improvement over all the available MCD procedures, which are unable to provide good control over the size of multiple outlier tests for the data structures considered in this paper.     

A Note on the Inoperability Input-Output Model

In this note, we propose some comments and some extensions of the inoperability input-output model (IIOM), as recently proposed by Santos and Haimes (2004). In particular, we propose the use of some analytic tools capable of providing information on the reaction of sectors subsequent to a terrorist attack on infrastructure service sectors. These tools, namely, the field of influence and the multiplier product matrix, provide information on the way sectors react to a shock on the aggregate demand and/or to a (temporary or permanent) change of production function coefficients. Finally, using the 2003 65 sectors input-output matrix for the U.S. economy, a simple empirical example is presented.     

Sabotaging Potential Rivals

This paper studies sabotage in a contest with non-identical players. Unlike previous papers, we consider sabotage in an elimination contest and allow contestants to sabotage a potential or future rival. It turns out that for a certain seeding of players there is a pure-strategy equilibrium in which only the most able contestant engages in sabotage while less able contestants do not. The most able contestant may therefore prefer a situation where sabotage is allowed to one where sabotage is not allowed. For another seeding of players, there is a unique equilibrium in which none of the players invests in sabotage.     

Social Networks and Individual Perceptions: Explaining Differential Participation in Social Movements

This paper seeks to explain differential participation in social movements. It does so by attempting to bridge structural-level and individual-level explanations. We test a number of hypotheses drawn from the social networks and the rationalist perspectives on individual engagement by means of survey data on members of a major organization of the Swiss solidarity movement. Both perspectives find empirical support: the intensity of participation depends both on the embeddedness in social networks and on the individual perceptions of participation, that is, the evaluation of a number of cognitive parameters related to engagement. In particular, to be recruited by an activist and the perceived effectiveness of one's own potential contribution are the best predictors of differential participation. We specify the role of networks for social movements by looking at the nature and content of networks and by distinguishing between three basic functions of networks: structurally connecting prospective participants to an opportunity to participate, socializing them to a protest issue, and shaping their decision to become involved. The latter function implies that the embeddedness in social networks significantly affects the individual perceptions of participation.     

A Country on the Move: International Migration in Post‐Communist Albania1

Albania is a country on the move. This mobility plays a key role in household‐level strategies to cope with the economic hardship of transition. With the relaxing of controls on emigration at the beginning of the 1990s, international migration has exploded, becoming the single most important political, social, and economic phenomenon in post‐communist Albania. Based on the 1989 and 2001 population censuses we estimate that over 600,000 Albanians live abroad, mostly in nearby Greece and Italy, with the vast majority coming from a limited number of districts located at the coastal and transport gateways to these destination countries, as well as Tirana. The available data also suggest that a similar number have considered migrating, and of these, half have tried and failed. Almost one‐half of the children who since 1990 no longer live with their parents are now living abroad, a number of almost exodus proportions. This article also identifies clear patterns of temporary migration, with Greece being by far the most important destination and rural areas from the Center and North‐East of the country being the primary origins of these flows. Although migration, with the resulting remittances, has become an indispensable part of Albanian economic development, there is increasing consensus on the necessity to devise more appropriate, sustainable strategies to lift households out of poverty and promote the country's growth.     

Family structure and children's achievements

In this paper we estimate the relationships between several outcomes in early adulthood (educational attainment, economic inactivity, early childbearing, distress and smoking) and experience of life in a single-parent family during childhood. The analysis is performed using a special sample of young adults, who are selected from the first five waves of the British Household Panel Survey (1991–95) and can be matched with at least one sibling over the same period. We also perform level (logit) estimation using another sample of young adults from the BHPS. We find that: (i) experience of life in a single-parent family is usually associated with disadvantageous outcomes for young adults; (ii) most of the unfavourable outcomes are linked to an early family disruption, when the child was aged 0–5; and (iii) level estimates, whose causal interpretation relies on stronger assumptions, confirm the previous results and show that, for most outcomes, the adverse family structure effect persists even after controlling for the economic conditions of the family of origin. Received: 24 August 1998/Accepted: 21 January 2000     

Migration Between States and Markets

Liège has always been a city of passage, of migration, of intercultural encounters. Due to its location at the core of Europe and to its economic and industrial structure, Liège has for decades attracted immigrants. In the nineteenth century, migrants were coming mainly from Flanders, which at the time was an underdeveloped rural area. In the interwar period, many migrants from Poland and other Eastern European countries settled in the region. Right after the Second World War, the mining industry needed an additional labor force. It came from Italy and later from Morocco and other countries. Nowadays, refugees and asylumseekers from Africa and Asia live in the city and in the region. Liège is a multicultural, multiethnic and multiracial society with a long tradition of integration and toleration. Of course, Liège is no paradise. In this changing city, there are serious social and economic problems that sometimes find an expression in the field of ethnicity. But, contrary to other Belgian cities, racist and fascist political parties do not play a significant role in local politics.     

Where are the motives? A problem with evidence in the work of Richard Thaler

In this journal (Mackenzie, 1997. Economic Psychology 18(1), 123–135), Mackenzie argues that there is a problem with Thaler's work: in failing to provide evidence about motives, Thaler's claims are vulnerable. In this comment, I present a different view. Thaler merely pinpoints some stylized facts, such as the endowment effect. Models of economic behaviour should be able to account for these. Mackenzie's discussion of the wine owner anecdote gives ad hoc reasons for apparently irrational behaviour. Thaler's work focuses on systematic effects, which is a much more fruitful approach. Mackenzie fails to do justice to the work that has been done to document the endowment effect.