Employment, Health Insurance and Income Based Subsidies

The CBO Budget Outlook (2014-24) is out today, and much of the focus is on their revised estimate of the ACA’s impact on the labor market (Appendix C, pp. 118-27). CBO is projecting that “the ACA will cause a reduction of roughly 1 percent in aggregate labor compensation over the 2017-2024 period, compared with what it would have been otherwise” (p. 117). This is basically CBO’s estimate of the impact of the marginal tax rate on labor income created by the various policy structures of the ACA that Casey Mulligan has written about. As I wrote in this post, you get less of something if you tax it, so if that is not your goal (it is with tobacco taxes, for example) then you are left to decide whether the reduction is worth achieving an alternative goal(s). That can be a difficult question to answer, because you are trading off important things–labor market participation, rates of uninsured, and the system reform provisions of the ACA.

Is reduced labor market participation a sign that we should move away from the structure of the ACA? If yes, to where?

Republican Senators Burr, Coburn and Hatch introduced their health reform proposal PCARE last week to much fanfare (this contains links to my previous four posts on PCARE). Many Republicans were understandably glad that key leaders in their party had finally taken this step of offering a coherent health reform strategy. While the focus last week was on what PCARE would do to the tax treatment of employer sponsored health insurance and coverage rates, the CBO report should prompt those who are critical of the ACA’s labor market impacts to ask how PCARE would impact the marginal tax rate of labor? I asked Casey Mulligan on twitter last Saturday about this:

Casey answered:

While the numbers haven’t been run yet, Casey says that the marginal tax rate of labor income in PCARE will be similar to those in the ACA. Why is this? It is primarily because PCARE provides tax credits from 100% to 300% of poverty (the ACA goes from 100% to 400%).

The intuition of wanting to target and not waste precious resources that leads us to design means tested programs creates higher marginal tax rates on labor income. This reduces incentive to work more hours, and you get results like what CBO predicts for labor compensation (Casey predicts even higher impacts than does CBO). Casey expects a similar impact from PCARE, because of its means tested subsidy structure (it is worth remembering that it was the income based means test structure of tax credits in the ACA [Table 1, p. 48]  that created the largest work disincentive in Casey’s paper, not the employer mandate). The Republican reform plan PCARE shares this means tested structure.

Where does that leave us?

• A consensus that the pre-ACA status quo is not acceptable
• An expectation that the ACA will reduce work incentives
• Strong reason to believe (someone needs to run the numbers) that the leading Republican alternative, PCARE, will do the same

It is relatively easy to sketch a distant health reform approach, be it left, right or center, if you can control all of the variables. However, you cannot control them all, and we can only begin to transition from where we are now. If we decide that minimizing negative work incentives is paramount, then means tested subsidies are a bad idea, and some sort of universal approach, or flat financing mechamism seems warranted; but that will undermine employer sponsored insurance rapidly if flat subsidies are large, or lead to concerns about what can be bought with them if they are small (The Patients’ Choice Act from 2009 had flat subsidies, that were small as compared to the cost of insurance).

If we decide that we want to minimize crowd out of employer sponsored health insurance and to minimize disruption, then that favors means tested subsidies, while seeking ways to keep employer sponsored health insurance as stable as possible (employer mandates). However, today we are worrying about the impact on the labor market of the ACA, that does just these things.

I am drawn to the question “how do you want your children (or grand children) to get health insurance in 20 years?” I think that a distant system in which employment is not a means of obtaining health insurance would be a better way to do it, and could lead to more risk taking in the economy. And the ACA as eventually modified in some way by the Republican PCARE proposal can be the first steps to this, assuming near continuous tweaking for the rest of my life. The alternatives to this muddled through approach strike me as politically impossible (big bang single payer, ESI level of flat subsidy provided to everyone), or are unknown to me.

Update: I am unsure of the magnitude of the disincentive effect of ACA v. PCARE (meaning, which one is more disruptive to labor market). Someone needs to run the numbers, and I have suggested to Doug Holtz-Eakin and Steve Parente that if they did this in a dispassionate manner, it would be a great service to the debate. I think the density of workers from 100-300% of poverty v 100-400% is important, and am also uncertain of how the high risk pool structure in PCARE and the tax treatment changes proposed in PCARE will impact these calculations.