No abstract

In this paper we support the proposition that the output-unemployment relationship as represented by Okun's law is asymmetric. Okun's coefficients are defined based on a dynamic model that allows for asymmetry in the relationship between cyclical output and unemployment. Using data from the United States for the post-war period, our results show that the short-run effects of positive cyclical output on cyclical unemployment are quantitatively different from those of negative ones, and the data are consistent with the proposition that cyclical unemployment is more sensitive to negative than to positive cyclical output. Several theoretical explanations of asymmetry rationalize the findings. JEL Classification: C13, C22, E32Asyme´trie dans la loi de Okun. Ce me´moire fournit de bonnes raisons pour croire et des re´sultats empiriques pour supporter la proposition que la relation produitchoˆmage telle que repre´sente´e par la loi d'Okun est asyme´trique. On de´finit les coefficients d'Okun a`partir d'un mode`le dynamique qui permet l'asyme´trie dans la relation entre le niveau cyclique du produit et du choˆmage. A partir de donne´es ame´ricaines pour la pe´riode de l'apre`s guerre, les analyses montrent que (i) les effets a`court terme d'une pousse´e cyclique a`la hausse du niveau du produit sur le niveau cyclique du choˆmage sont quantativement diffe´rents de ceux d'une pousse´e ne´gative et (ii) les donne´es sont compatibles avec la proposition que le choˆmage cyclique est plus sensible aux variations cycliques ne´gatives qu'aux variations cycliques positives du niveau du produit. Ces re´sultats sont supporte´s par quelques explications the´oriques de l'asyme´trie.

No abstract

No abstract

Eight determining variables of FDI inflows are examined by applying extreme bounds analysis to a cross-sectional sample encompassing data on 140 countries. With GDP per capita serving as the free variable, seven variables are tried as the variables of interest in combination with three other variables. The results reveal that only two variables are robust: exports as a percentage of GDP and telephone lines per 1000 of the population. It is shown that a parsimonious model with a reasonably good predictive power contains the free variable, the two robust variables and two dummies.