Research Question:

• Can policy practitioners predict the preferences of those they serve?

  • Are they biased towards their own views?

  • Does familiarity with a context lead to better forecasts?

• Using dynamic programming, I modelled the labour supply decision of US workers under an unconditional cash transfer program

• The model distinguishes from other dynamic labour supply models in the following dimensions: 

  • • Used regularisation technique (ridge estimation) for the estimation of spline coefficients of the utility functions instead of the common collocation technique. 

    • Incorporated limited forward-looking capacity by explicitly modelling myopia as a rolling window optimisation. 

    • Relaxed hyperbolic discounting initial period limitation from 1 to n.

I studied state-level legalisation in the US between 2005 and 2014. I found that same-sex marriage decreases wages for gay workers and has no detectable impact on lesbian workers. The results also show a weak, imprecise, but positive effect on the employment rate for both gay and lesbian couples.