Filling the Gaps Through Amplified Asking: Combining Register and Survey Data to Understand Carbon-Intensive Time Use

Statistical institutions collect rich register data and administer valuable surveys, but the latter often cover only a small portion of the population. While combining register and survey data already holds great promise, a new range of opportunities emerges when we can also combine multiple surveys - even when they lack (substantial) overlap. This research project explores how we can leverage Dutch register data to impute survey variables for individuals outside the original sample, thereby effectively expanding the scope of survey-based research. Our initial focus is on time-use data, a uniquely informative but resource-intensive type of data, typically gathered through detailed diaries that are impractical to collect at scale. We are developing a method to generate realistic estimates of how individuals across the Netherlands spend their time, using demographic and household-level register data. This approach may make it possible to link time-use data to other survey domains, such as income, health, or consumption, and thus study previously inaccessible patterns, including intersectional forms of gender inequality or individual-level carbon emissions. In this presentation, we will introduce a preliminary approach for producing such large-scale time-use estimates by combining various sources of register data. As the project develops, particular attention will be given to addressing methodological challenges, including the generalizability of predictions across social groups, potential biases in imputed data, and the broader implications of applying such methods in survey research. Ultimately, the project aims to develop a scalable and responsible framework for expanding the reach of survey research. By enabling new combinations of survey content and increasing representativeness, we hope to contribute to more inclusive, data-driven understandings of societal dynamics in the Netherlands.