Estimating the Cost of Healthcare Reform: Round 1, the Kennedy Bill
The CBO has issued a preliminary analysis of the Kennedy healthcare reform plan. The following points are useful in understanding the estimates.
1) 30% of the direct cost is assumed to be offset by increased wages from reduced healthcare coverage. About 14 million people who would be covered by their employer under current law will not have employer based coverage under the Kennedy bill. The CBO believes that their wages will rise to maintain competitive compensation. I think this is a bad assumption for budgeting purposes.
- It is highly uncertain and therefore an example of budget arbitrage, although it comes from within the CBO itself.
- 10 of the 14 million will voluntarily decline coverage that is offered to them. It is not plausible that those individuals will receive additional compensation. We can’t know for sure if the $257b in increased revenues over six years comes from wage increases on just the remaining 4 million individuals.
- Pricing the value of health insurance as a “compensating differential” is difficult. I would suggest that many of the current features of employer based health insurance are poorly understood in such a model, thereby casting doubt on the validity of using that approach to project future revenues.
- I am concerned about long-run slack in the labor market and “paying” for healthcare reform by assuming that employees are in a strong negotiating position may not be prudent.
2) The direct cost of this bill was surprisingly low. Premium subsidies were granted to those with incomes of more than $110,000 for a family of four. Presumably, these subsidies must grow with healthcare costs. We don’t know the assumed growth rate in those costs. Also remember that these subsidies must make relatively rich federally-defined plans affordable. The CBO assumes that the subsidies will be effective at doing this, causing 10 million people to decline their current employer-based coverage. Much of the increase in healthcare costs in 2013-2015 will be due to this legislation and the corresponding stress on our health care delivery system. I therefore want to see more specifics on the average size of the subsidy.
3) The “clawbacks” within this approach are going to be a significant deterrent to work. I don’t see this effect estimated. To illustrate, let’s assume that a family of four earning $40,000 in 2014 gets free healthcare insurance (not free healthcare since there is cost-sharing in the federal plans). Let’s also assume full-phase out of the subsidy by a $120,000 income. This subsidy is likely worth something on the order of $15,000-20,000 (for the entire family; again in 2014). This means that any increase in pay between $40,000 and $120,000 will result in an effective 18%-25% tax as the subsidy is gradually withdrawn. This feature also has the possibility of significantly affecting family structures, the desirability of two-earner households, and the willingness to save and invest for the future.
I am opposed to the Kennedy bill for many reasons, include some expressed by Keith Hennessey. I find this cost estimate unfortunate not because of the low price tag but because I believe the assumptions and modeling structure to be very, very optimistic running real fiscal risks. I hope these issues and others are fixed before final scoring.