Nonresponse rates have been used as a proxy for survey quality since they indicate the relative potential for nonresponse bias. Recently the R-index (Schouten) has generated interest in an alternative approach that better represents the potential for bias by focusing more on coverage than nonresponse. The patterns of nonresponse rates (e.g.; seasonal, time in sample) and the R-index can provide insight into the usefulness of nonresponse rates and representativeness. The current study uses different measures of nonresponse bias, nonresponse rates, and the R-index to see if there are patterns for bias and representativeness which might be different than for response rates alone. Two surveys, the Current Population Survey (CPS), and the Consumer Expenditure Quarterly Survey (CEQ) will be used in this analysis.