No measure of individual market structure had a statistically significant impact on the probability of coverage. However, the pattern of estimated coefficients is of some interest, in that it is in general consistent with the same potential diseconomies of small scale that we observed in the small-group market. That is, a larger number of insurers in the individual market, and therefore presumably greater competition, did not raise the probability of private coverage. Instead, in states with a greater number of insurers and lower market concentration, individuals in worker families without either employer or public coverage were generally less likely (in all specifications of the model) to be privately insured — although these effects were not statistically significant.
When we control for some, but not all measures of market structure (and for all types of regulation), we found that the average individual-product loss ratio among commercial insurers was negatively related to coverage, although the statistical significance of this relationship was weak. The negative sign of the coefficient is unexpected, but might be explained in terms of an intervening, unobserved relationship between insurer loss ratios and insurance prices. That is, in markets where price levels are high (and coverage is low), insurers may accept higher loss ratios rather than raise prices still further; in these markets, insurers may be especially cautious about triggering adverse selection and extinguishing demand.