Description:
We take advantage of a unique experiment that took place in Iceland in 2015. Following the collapse of the country's banking system in 2008, the authorities decided on a program of mortgage relief that in effect lowered the principal of mortgages overnight. What sets our experiment apart from others is that the debt relief was financed by foreign creditors, that we have administrative data on all taxpayers in the country and that the mortgage relief was implemented many years after the end of the economic crisis when household balance sheets had been repaired. We measure the effect of the mortgage relief on the saving of every taxpayer in the country using households that were not eligible for debt relief as a control group. While a negative wealth effect on saving could have been expected, households amortized even more in response to the debt forgiveness. The increased amortization is not only due to lower interest costs but also due to higher saving, mostly by highly leveraged and liquidity constrained households.