Abstract. Household income and expenditure survey is one of the most important statistical surveys in Iran. It has many applications in economic planning and decision makings, such as studying the household consumption pattern, studying the trend of changes in goods and service consumption, assessing the economic policies and calculation of regional and national accounts. Obviously, because of various applications of this survey in government’s planning and decision making its data quality is very important. Precision is one of the essential components of data quality. To determine precision of the data, sampling and non-sampling variances should be estimated.

Numerous estimators could be defined due to the fact that a variety of information is collected in the survey. In this paper, some basic estimators are considered. Furthermore, sampling and non-sampling variances are computed using the response error model for the results of two methods of information collection. The precisions of the methods are compared.