Various versions of the wild bootstrap are studied as applied to regression models with heteroskedastic errors. It is shown that some versions can be qualified as "tamed," in the sense that the statistic bootstrapped is asymptotically independent of the distribution of the wild bootstrap DGP. This can, in one very specific case, lead to perfect bootstrap inference, and leads to substantial reduction in the error in the rejection probability of a bootstrap test much more generally. However, the version of the wild bootstrap with this desirable property does not benefit from the skewness correction afforded by the most popular version of the wild bootstrap in the literature. Edgeworth expansions and simulation experiments are used to show why this defect does not prevent the preferred version from having the smallest error in rejection probability in small and medium-sized samples. It is concluded that this preferred version should always be used in practice.

International audienceWe examine the statistical performance of inequality indices in the presence of extreme values in the data and show that these indices are very sensitive to the properties of the income distribution. Estimation and inference can be dramatically affected, especially when the tail of the income distribution is heavy, even when standard bootstrap methods are employed. However, use of appropriate semiparametric methods for modelling the upper tail can greatly improve the performance of even those inequality indices that are normally considered particularly sensitive to extreme values

International audienceIn regression models, appropriate bootstrap methods for inference robust to heteroskedasticity of unknown form are the wild bootstrap and the pairs bootstrap. The finite sample performance of a heteroskedastic-robust test is investigated with Monte Carlo experiments. The simulation results suggest that one specific version of the wild bootstrap outperforms the other versions of the wild bootstrap and of the pairs bootstrap. It is the only one for which the bootstrap test gives always better results than the asymptotic test

A random sample drawn from a population would appear to offer an ideal opportunity to use the bootstrap in order to perform accurate inference, since the observations of the sample are IID. In this paper, Monte Carlo results suggest that bootstrapping a commonly used index of inequality leads to inference that is not accurate even in very large samples, although inference with poverty indices is satisfactory. We find that the major cause is the extreme sensitivity of many inequality indices to the exact nature of the upper tail of the income distribution. This leads us to study two non-standard bootstraps, the m out of n bootstrap, which is valid in some situations where the standard bootstrap fails, and a bootstrap in which the upper tail is modelled parametrically. Monte Carlo results suggest that accurate inference can be achieved with this last method in moderately large samples