The Intrahousehold Communications Study: Family Cohesion and the Level of Knowledge about Expenses

Government surveys often accept reports from proxy respondents. For example, the Current Population Survey (CPS), used to measure unemployment, uses proxy information about the employment status of other household members. The Consumer Expenditure Program, which provides the cost weights in the Consumer Price Index, accepts proxy reports about the expenditures of other family members. Because many govemment surveys rely on proxy reporting, the accuracy of important estimates from Federal surveys often depend on how much information family members communicate to one another. Recent work by Tanur (1990) on the problems of proxy reporting in the Current Population Survey (CPS), particularly with respect to the activities of young adults, has led to an interest at the Bureau of Labor Statistics (BLS) in the pattems of communications within families. It is neither reasonable to expect nor practical to seek individual reports from each family member in every household. Some members cannot be contacted, others would refuse, and all of these individual reports would require a much greater expenditure of resources, especially interviewer time. Instead, it might be wise to find some method for estimating the accuracy of proxy responses from specific households. Interviews with every family member would only be attempted in those households with a very low probability of act~urate proxy knowledge. Even if this information were not used to screen families, it would provide a means of evaluating the quality of survey estimates. In order to explore the possibility of collecting such information, BLS contracted with Oak Ridge National Laboratory (ORNL) to conduct an experimental study of the level of proxy knowledge and the communication pattems within selected households conceming the employment history and expenditures of individual members. The project is called the Intrahousehold Communications Study (IHCS). Miller and Tucker (1993) developed a typology of family cohesion using data from the IHCS, and examined the relationships between cohesion and family characteristics. This paper uses the typology of cohesion, as well as information about the family characteristics, to better understand the causal process leading to proxy knowledge about expenditures. No attempt will be made to explain the creation of the cohesion measure. For a detailed account of the creation of the cohesion measure refer to Miller and Tucker (1993).