This report presents descriptive analysis of the trends in rate increases in periods before and after ACA rate review, but there is no way of knowing what would happen absent the ACA, as its provisions apply to all states. NORC did not conduct multivariate analyses to test the impact of factors unrelated to the ACA that may affect premium increases.
In both the individual and small group markets, we cannot explain why the number of fillings sometimes fluctuates dramatically from year to year for a given state.
For some data fields in some filings, data were either missing or seemingly implausible. For example, some filings were missing either requested premium increases or approved rate increases; in these cases, we were unable to assess whether state regulators modified the rate originally proposed by the carrier, and so these observations had to be omitted from analysis of that question. In other instances, available data seemed implausible. For example, in some cases the total reported enrollment in multiple filings from the same year by a single carrier summed to a figure much greater than that carrier’s entire enrollment listed in the NAIC April Supplemental Report, suggesting that some enrollees may have been double-counted in the filings. Where enrollment data is missing or implausible, the weighting methodology we use employs the data from NAIC on state insurer enrollment in the small group and individual markets to cap the maximum possible weight such filings can receive. From sensitivity testing conducted for a prior ASPE study of similar data, we believe that measures of central tendency in this report are robust to the particulars of the weighting method used.
Another limitation is the comparability of the current study’s findings to the findings from the Trends study, as the study sample and data collection methods differed. The current study includes a modified panel of states, with six states that were included in the Trends study sample replaced by five states with publicly available websites. The six states replaced did not have public websites. For each state included in this study, NORC did not extract data for insurers outside of the carrier sample (sampled carriers were either those with at least one percent market share in the state or the five largest carriers, with the more inclusive rule applying, as described in this report’s Methods section) for 2011 and 2012, unlike the Trends study. Some fluctuations in the number of filings for individual states may be attributable to the different sampling rules for the Trends study and “State Market Reforms.” As a result, the number of fillings sometimes fluctuates dramatically from year to year for a given state, but differences in sampling methods only explain some of the results. For example, data collection efforts for Pennsylvania in the individual market from the Trends study resulted in 16 plan filings in 2008, 30 in 2009, 24 in 2010, and 35 in 2011, with 22 different insurance carriers represented. In comparison, for the current study there were 15 carriers in Pennsylvania’s individual market included in the sample, yielding 10 filings in 2011, and 32 in 2012.
Finally, it is important to note that state procedures for posting filings in their public portal and their process for reviewing filings vary, even among states that have the same regulatory authority (file and use or prior approval). For example as noted previously, in some states files on proposed rate increases that are rejected by the regulator are kept open until a compromise rate increase can be arrived at while in other states in response to a rejection from the regulator the carrier may re-file a new rate at a at a later date under a separate tracking number. Although use of the SERFF portal and the SERFF file template did improve the consistency of the information presented in filings, in some cases sections of the template were left blank or could only be found in the correspondence attached to the filing. As such, while the completeness of the filing documentation submitted by carriers has improved since the beginning of the Trends study, the data presented in this report is subject to the limitations of its sources.