From Annual to Monthly Simulation of Social Assistance in Sweden

FASIT is a microsimulation model primarily used by the Swedish Parliament and Government to calculate the effects of various regulatory changes. Statistics Sweden (SCB) is responsible for the maintenance and development of the model. This presentation explores experiences moving from annual to monthly data when simulating social assistance.

The purpose of the change has been to improve the simulation of social assistance for households that only require support during part of the year. Relying on annual data results in an underestimation of households that receive social assistance only intermittently throughout the year. A transition to monthly simulation has been made possible by the availability of monthly information on wages and other taxable incomes.

Social assistance is a means-tested benefit provided by municipalities to households that cannot support themselves. Households must actively apply for the benefit. FASIT includes different models for this (take-up), which will also be reviewed in the presentation.

In the simulation of social assistance, both the total amount of assistance paid out and the number of households receiving assistance are important to consider. Depending on how underlying incomes vary for each household throughout the year, and the likelihood of applying for assistance during short periods of reduced income, the monthly model may risk overestimation of the total number of households receiving assistance over the year.

We will describe how the simulation was performed before the transition, the challenges encountered during the shift, how these challenges were addressed, and the differences in outcomes between the two simulation approaches.