The Effect of Food Stamps on Food Expenditures: An Assessment of Findings From the Nationwide Food Consumption Survey
利用1977-78年美国农业部全国食品消费调查数据，发现使用加权与未加权数据估算的食品券边际消费倾向差异很大，加权估计值约为未加权的两倍，且少数高权重样本是主要驱动因素。
Abstract Estimates of the marginal propensity to consume food out of food stamp benefits based upon the low‐income supplement to the U.S. Department of Agriculture's “Nationwide Food Consumption Survey” are very sensitive to whether weighted or unweighted data are used in the analysis. Based on data from the 1977–78 survey, we estimated identical models of household food expenditures using unweighted and weighted data. Weighted estimates of the marginal propensity to consume food out of food stamps are approximately double the unweighted estimates, even when sample stratifiers are included in the model. A small number of cases with large sampling weights are responsible for almost all of the difference between the weighted and unweighted estimates.