Constructing Confidence Intervals Using the Bootstrap: An Application to a Multi-Product Cost Function
用自助法为387家银行的边际成本、产出弹性等指标构建置信区间，发现自助法标准误与传统估计差异显著，并扩展了成本函数正则性分析。
A multi-product cost function system is estimated for 387 banks in states that allow branch banking. The bootstrap resampling method is used to construct confidence intervals for marginal costs, output-cost elasticities, economies of scale and scope, and Allen elasticities of substitution. Confidence intervals for these measures are usually constructed using a first-order variance approximation under a normality assumption, but such confidence intervals are inexact if the measures are not normally distributed or the variance approximations are imprecise. We find that the bootstrap standard error estimates can differ significantly from the usual estimates. Furthermore, we use the bootstrap to expand the analysis of cost function regularity properties. Copyright 1990 by MIT Press.