Too Many Zeros: A Method for Estimating Long-Term Time-Use from Short Diaries

Various issues of public policy and individual wellbeing, ranging from physical exercise to access to cultural activities and leisure more generally, require reliable estimates of the distribution of relatively infrequent (much less than daily) activities across populations. Questionnaire items (such as those asking "How often do you (engage in ... activity)?", suffer from respondent recall, classification, observation period and social desirability problems; own-words diaries are therefore the preferred method of measuring time use. But the respondent burden imposed by diary instruments is considerable, and the normal advice is to limit the length of diaries to a single day. This paper sets out a method for combining questionnaire "habit" indicators and diary methods to provide reliable long-term estimates of time use. It concludes that the usefulness of exercises such as the American Time Use Study and the Harmonised European Time Use Study would be greatly enhanced by the collection of habit measures from their diary respondents. And the use of this method might substantially reduce the average cost of collecting high quality time use evidence.