What Impact HIPAA? State Regulation and Private Health Insurance Coverage Among Adults.. 6. Summary and Conclusions


The literature describing the impact of state regulation on health private health insurance is fraught with methodological difficulties and contradictory findings. This study offers several improvements on this literature: (1) We estimate coverage among only adults in worker families, minimizing potential error in the measurement of key variables. (2) We introduce continuous measures of regulations that govern how insurers price coverage and how they set waiting periods for coverage of preexisting conditions. (3) We estimate the impact of regulations individually, rather than in “bundles.” (4) We introduce measures of market structure to control for supply-side variables that may influence the price of insurance (which is unobserved) and therefore the rate of private coverage.

Our results related to market structure are new to the literature. We find that states with more concentrated group insurance markets enjoyed higher rates of small employer coverage, all else being equal. This finding is consistent with diseconomies of small scale in states with more fragmented insurance markets, but also with potential differences in regulatory oversight that may be endogenous to market structure.

With respect to regulatory impacts, we find that very few forms of regulation affected small employer coverage independent of their primary impacts on market structure. However, all-product guaranteed issue (as HIPAA requires) significantly raised the probability of coverage among adults in worker families. Our earlier research indicated that all-product guaranteed issue also improved competition in the group market, increasing the number of insurers that participated.

We find that the impact of rate regulation on coverage was significant and relatively complex in the group market, but insignificant in the individual market. In states with a narrow composite rate band (or lacking a composite band, narrow bands on both age and health rating), small employer coverage was significantly higher among adults associated with workers in firms with fewer than 100 employees (although we found no impact on workers in firms with fewer than 25 employees). However, when both narrow composite rate bands and guaranteed issue were imposed, their combined effect on coverage was negative: employer coverage was significantly lower among adults associated with workers in firms with fewer than 100 workers.

We could discern no significant impact of rate regulation on coverage among workers in the smallest firms. This pattern is consistent with the conventional economic theory of price discrimination: if insurers enjoy some degree of monopoly pricing power and “larger” small firms demonstrate lower price elasticity of demand for insurance compared to the smallest firms, insurers may raise prices more for larger small firms than for the smallest firms for which average premium levels generally are higher.

Most significant for anticipating the possible results that HIPAA may have on coverage, we observed no negative impact on employer coverage either from guaranteed issue or shorter waiting periods on coverage for preexisting conditions. Indeed, it is likely that HIPAA’s provision requiring all-product guaranteed issue in the small-group market has helped to extend insurance coverage to a greater number of employees and their dependents, all else being equal.

In the individual market, some states have implemented regulation that is much more extensive than HIPAA’s very modest provisions. Our analysis produced no evidence that most of the states’ access reforms in the individual market have affected the general rate of individual coverage. Specifically, neither constraints on rating nor shorter preexisting condition exclusions had any apparent impact on coverage among adults in worker families without employer or public coverage.

In contrast, regulation requiring guaranteed issue of all products in the individual market appears to have substantially reduced individual coverage, although the statistical significance of this effect was weak. The availability of a high-risk pool — an alternative to guaranteed issue to ensure access in the individual market — had no significant impact on individual coverage.

Despite its statistical weakness, the impact of all-product guaranteed issue on individual coverage is of some interest in considering the potential impact of HIPAA’s more modest provision for guaranteed issue to HIPAA-eligible workers and dependents. Others (Pollitz et al., 2000) have observed that, in states where the absence or looseness of rate regulation permits, the premiums charged to HIPAA-eligibles can be extremely high, circumventing the coverage protection that HIPAA intended. Our results suggest that insurers may in fact substantially raise price in the individual market when constrained by guaranteed issue, producing lower rates of individual coverage.

Finally, some mention of the need for further research in this area is in order, both with respect to the impacts of regulation and with respect to the potential importance of insurance market structure. Both our the results and those of all other empirical studies of health insurance regulation to date rely on observation of very few changes in state regulation prior to 1997. Thus, in most cases, only a few observations of change drive all of the estimates of regulatory impact. It is essential that more recent years (after the states implemented regulations to comply with HIPAA) be studied to corroborate the conclusions reached here or in other studies of regulatory impacts.

Also, our results regarding the impact of market structure on coverage suggest that diseconomies of small scale in the supply of health insurance may be a problem. In the small-group market, we observed a positive effect of greater market concentration on employer coverage. In the individual market, we observed no impact of statistical significance, but the signs of key market structure variables were consistent with diseconomies of small scale. Further investigation of insurance supply and its impact on coverage is warranted — including specifically the extent of competition and scale economies in the production of health insurance, and differences in the administration of regulations that may be endogenous to market structure.