Previous studies have examined item nonresponse as a predictor of household nonresponse. The usefulness of item nonresponse and unit nonresponse in examining potential nonresponse bias in multiple surveys is explored. Models of item nonresponse and household nonresponse will be related to survey estimates. The theory that households vary on a continuum of cooperation will be examined through a scale from their unwillingness to respond to the survey at all to their unwillingness to respond to certain items. Comparing the relationships across surveys gives insight into the generality of survey and item nonresponse models.