The effect of state health insurance benefit mandates on premiums and employee contributions

ABSTRACT

The average US state has 40 benefit mandates, laws requiring health insurance to cover particular conditions, treatments, providers or people. We investigate the extent to which these mandates increase the health insurance premiums paid by employers, and the extent to which these higher premiums are passed on to employees in the form of higher employee contributions. We use state-level data on premiums and employee contributions to health insurance from the insurance component of the 1996–2011 Medical Expenditure Panel Survey. Our main analysis is a fixed effects regression that controls for age, race, income, union membership and the presence of state mandate waivers. We find robust evidence that the average mandate increases premiums by approximately 0.6%, and that mandates lead to similar increases in employee contributions for single-coverage health insurance plans. Alternative specifications using an AR(1) error structure estimate a larger effect of mandates, while those using generalized estimating equations estimate smaller effects. We find that mandates requiring insurers to cover a specific benefit, as opposed to a specific type of provider or person, lead to the largest increases in employee contributions.

KEYWORDS:

• Health insurance
• benefit mandates
• employee contributions
• premiums

JEL CLASSIFICATION:

• I18
• J32
• J38

Acknowledgements

The authors wish to thank James Kelly and Eleanor Lewin for helpful comments.

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

¹ Kowalski, Congdon and Showalter (2008) use 2004 data on high-deductible plans from ehealthinsurance.com and find that the average mandate increases premiums by 0.26%. Gohmann and McCrickard (2009) use 2006 data from ehealthinsurance.com and find that different mandates have wildly different effects, ranging from greatly increasing premiums, to insignificance, to decreasing premiums. LaPierre et al. (2009) use 1997–2003 data from the Community Tracking Survey and find no significant effects of mandates on premiums.

² This name can be confusing because it is used to refer to both the general category of mandates that specify what insurers must cover (as opposed to, say, individual mandates or employer mandates) and the specific subset of these insurer mandates that target conditions and treatments (as opposed to people or providers). It may be clearer to refer to these specific treatment/condition mandates as treatment mandates.

³ Medical Expenditure Panel Survey’s Insurance Component (MEPS-IC) did not conduct their employer survey in 2007, and data are missing from some states prior to 2003; some of these early years have data from only 40 states. Our estimation strategy is adapted to be relatively robust to missing observations.

⁴ See Baltagi and Wu (1999) for further explanation of the properties of fixed effects models with AR(1) disturbances in the presence of missing observations.

⁵ Gruber (1994) found that a maternity benefit mandate was passed on entirely in the form of lower wages for 20- to 40-year-old women. Lahey (2012) found that infertility mandates reduce wages and employment for 28- to 42-year-old women. Bailey (2013) found that diabetes mandates reduced the wages of obese workers.