According to this article, the Joint Committee on Taxation (JCT) has scored a Senate homebuyer tax credit at $16.7 billion. How does the JCT $16.7 billion cost estimate square with my previous back-of-the-envelope calculation of the cost of the tax credit of $73.9 billion?
For my calculation, I assumed a tax credit of $15,000, available for one year. The Senate proposal scored by JCT is for an $8,000 tax credit. I also assumed the tax credit would be available for one year, whereas the Senate proposal scored by JCT is for seven months (December 1, 2009 to June 30, 2010).
In my article, I computed a range of estimates, each assuming different parameter inputs. But let’s consider one of my estimates, which is based on a price elasticity of -0.65, baseline sales of 5.5 million houses annually, take-up of 85 percent (since high-earners are not eligible), and a median home price (for non-high-earning buyers) of $180,000. Using these assumptions, I arrived at an estimate of 253,000 additional houses sold due to the credit at a cost of $73.9 billion, for a cost-per-additional home of $292,000.
If we use the same parameter inputs, but change the credit to $8,000 from $15,000, and assume baseline sales of 2.8 million rather than 5.5 million (based on existing and new homes sales from December 2008 through June 2009), then we get an estimate of 69,000 additional sales at $19.6 billion. The $19.6 billion is higher than JCT’s estimate of $16.7 billion, but I would say within the ballpark. Part of the difference is due to JCT’s estimate that there will be a revenue gain of $1.8 billion from 2012 to 2014, since homebuyers must pay back the credit if they sell their house within three years. Even assuming the $1.8 billion in additional revenues, the point remains the same – an estimated $258,000 per additional house sold (i.e., $17.8 billion divided by 69,000) is a poorly targeted subsidy!