Simulating the low fertility hypothesis

Daniel Hallberg, Uppsala University
Thomas Lindh, Institute for Futures Studies
Jovan Zamac, Uppsala University

Several countries are experiencing low fertility, often defined as TFR below 1.5. Recently it has been put forward, e.g. Lutz et al., that these countries may be caught in a low fertility trap. The hypothesis put forward consists of three main elements: a demographic, a sociological, and an economic. The social interaction mechanism is especially hard to capture in a standard economic framework. For this reason we propose to use an agent based simulation model to investigate the economic consequences of low fertility and its feedback on the fertility decision. By using simple thumb rules for agent behaviour with regards to their fertility decision we want to investigate if it is possible for the aggregate level of fertility to stay below 1.5 even if the wanted number of children of each individual is at the replacement level. This would be in line with the observations from the Eurobarometer survey where the reported ideal family size is substantially above the observed outcome, especially for the low fertility countries. An agent based model has the possibility to display such emergent properties since the outcome is a result of agent interaction and cannot be deduced directly from the thumb rules. To validate the used thumb rules we will use the Swedish data which has substantial variation in TFR together with rather stable completed cohort fertility. We can also simulate policy changes, too for instance the Swedish child care system, and investigate if the fertility rates will fall in the low region and for how long these rates will be low.