To complete our current study, we needed to revise and extend our methods. We also used additional data sources and elements to more fully account for the health status of individuals in our models using claims data from the employers participating in our consumer driven health plan analysis.
-
A. Methods
-
There are three major components to our methodological approach: 1) Model Estimation; 2) Choice Set Assignment and Prediction; and 3) Policy Simulation. As illustrated in Figure 1, often more than one database was required to complete the task. Integral to this analysis was the use of consumer directed health plan data from four large employers working with the study investigators.
Figure 1
Figure 1 illustrates the three major components to our methodological approach: 1) Model Estimation; 2) Choice Set Assignment and Prediction; and 3) Policy Simulation. This particular model shows that more than one database was required to complete the task. Several steps were need for this model estimation.
The model estimation had several steps. As a first step, we pooled the data from the four employers offering CDHPs to estimate a conditional logistic plan choice model similar to our earlier work (Parente, Feldman and Christianson, 2004). In the second step we used the estimated choice-model coefficients to predict health plan choices for individuals in the MEPS-HC. In order to complete this step, it was necessary first to assign the number and types of health insurance choices that are available to each respondent in the MEPS-HC. For this purpose we turned to the smaller, but more-detailed MEPS Household Component-Insurance Component linked file, which contained the needed information. The third step was to generate 2006 HSA premiums and benefit designs. The final step was to apply plan choice models coefficients to the MEPS data with premium information as well as 2006 State of union tax treatment adjustments to get final estimates of take up and subsidy costs.
This process was similar to our previous work and described in more detail in the appendix. Two of the most substantial of several changes were inclusion chronic illness into the plan choice model and generation of premiums through an iterative process using prior years claims data to create actuarially fair estimates of premium. Below we describe in more detail specific issues that we addressed in our current analysis.
-
-
B. Baseline Simulation Results
-
Table 2 shows the results of the baseline simulation prior to iterations. We predict that approximately 3.2 million people who are not eligible for employer-sponsored health insurance (ESI) will choose an individual HDHP. This estimate was calibrated to be the same as our previous analysis in order to compare the proposed policy changes in 2004 to the 2006 State of the Union proposals.9 An additional 47,509 people will turn down their employers' offers of coverage to purchase an individual HDHP. This is fewer than the 365,150 people who were predicted to "opt out" of ESI in our original simulation of the MMA effects (Parente, et al., 2005). The large swing was attributable to a difference in the tax treatment of the premium in the original analysis that was clarified and corrected in the current study. As a result, we believe our original analysis significantly overstated the likely opt-out of the employer market into individual HSAs.
We predict that only 67,812 people will choose an employer-sponsored HDHP. This is also fewer than the 334,938 individuals who chose this option in the original simulations. Once again, this change was the result of an inaccurate assessment of the tax treatment of the employer sponsored HDHP. We are confident our current model better reflects the actual economic incentives present in the group market. However, both of these predictions are quite small in relation to the number of people who choose PPOs or HMOs in the group market, reinforcing our earlier result (Feldman, et al., 2005) that HDHPs will not be popular among employees with an employer health insurance offer, primarily because the employer's premium subsidy reduces the attractiveness of HDHPs compared with other types of health insurance plans.
Under both old and new simulations, approximately 13.3 million to 13.5 million people turn down their employers' offers of health insurance but do not purchase an individual HDHP. They may pick one of the other individually-offered policies, but many of them will remain uninsured. There are two explanations for the large changes in the PPO market. First additional employer information allowed us to more accurately identify the structure of low, medium and high PPOs more completely. Second, the new premiums based on prior claims history are significantly different than previous premiums, which were largely adapted from several year-old surveys from secondary sources. Thus the premiums used were a much closer match to results, estimated from claims data, of the actual incentives in place including coinsurance, copayments and provider panel access. As a result, the differences between our current and previous work are not the impact of a policy change as much as they are a refinement of the inputs into the model with more accurate and appropriate data.
Individual Market Original Simulation of MMA Effects New Simulation of MMA Effects Table 2: Comparing MMA Original and New Baseline Populations H.S.A. 3,155,982 3,156,133 PPO High 4,651,023 37,591 PPO Low 310,041 6,046,777 PPO Medium 1,426,040 232,105 Uninsured 27,273,018 27,313,692 Group Market HMO 26,330,531 19,036,514 HRA 1,838,559 2,250,267 Employer-sponsored H.S.A 334,938 67,812 Opt-out H.S.A (indivual fully funds) 365,150 47,509 PPO High 5,951,085 8,528,436 PPO Low 1,575,203 1,014,605 PPO Medium 35,001,278 40,289,118 Turned Down 13,322,842 13,515,131
-
View full report
"report.pdf" (pdf, 321.78Kb)
Note: Documents in PDF format require the Adobe Acrobat Reader®. If you experience problems with PDF documents, please download the latest version of the Reader®
