Uninsured individuals receive fewer healthcare services for at least three reasons: responsibility for the entire bill, higher prices, and potential provider reductions for concern of nonpayment. I isolate reductions when uninsured patients are solely financially responsible by capitalizing on Maryland’s highly regulated health care system. Prices are set by the state, are uniform across all patients, and hospitals are compensated for free care and bad debt. I use a unique feature of the data, multiple readmissions for patients who gain or lose insurance between visits, to isolate the reductions in quantity demanded when individuals are faced with paying the full price without an insurance contribution. A Blinder–Oaxaca decomposition estimates uninsured individuals receive 6% fewer services after accounting for differences in patient, illness, and hospital characteristics than when these same individuals are insured.
This is a preview of subscription content, access via your institution.
Buy single article
Instant access to the full article PDF.
Price excludes VAT (USA)
Tax calculation will be finalised during checkout.
This is a fee for service environment, but is unusual in that hospital- rate center- specific prices are set by the HSCRC. Additionally, there are limits to the total number of services a hospital can provide in a year based on their aggregate patient mix.
A patient is categorized into a listing of diagnosis or procedure. This initial categorization is the root Diagnosis Related Group (DRG). There are 314 root DRGs in Maryland. Once the root DRG has been established, secondary diagnosis and other factors (obesity, other illnesses) are then considered and the patient is assigned a severity class (1–4). For severity, the 1–4 scale runs low to high. The combination of the root DRG and the level of severity form the APRDRG (Systems 2013).
For more information on the rate setting system in MD, please see https://hscrc.state.md.us/Documents/pdr/GeneralInformation/MarylandAll-PayorHospitalSystem.pdf.
For government payers there is, at most, a 6% difference between Medicare/Medicaid and non-Medicare/Medicaid patients. This doesn’t create an incentive to provide fewer services to government payers, however, because when a hospital’s base rate is established, the fraction of government patients is considered and built into the base rate. If a hospital serves more government payers, its base rate increases the next year, if it serves fewer, it declines. As such, a hospital has no financial incentive to provide a differential service mix to different payers. Details of how the base rate is established can be found in Appendix 1.
Some individuals may chose to be uninsured if they believe their private health signal is very good, believing they will not get sick or injured. One might be concerned that individuals with good private health information may be healthier than average and more likely to be uninsured. However, in the data, I restrict my sample to inpatient admissions. Figure 3 illustrates that there is not a clear relationship between APRDRG weight (a measure of severity) and services. Furthermore, Long et al. (1998) find that individuals are unable to anticipate changes to insurance status to take advantage of periods of relatively generous insurance. This suggests that reductions to services observed in this paper are not ‘consumption smoothing’ on the part of the uninsured. While some young people improperly estimate their risks of illness, and have been termed the ‘young invincibles’ (Smith 2014), the remainder of this analysis will focus on an income constraint as being the driving motivation for remaining uninsured.
I only observe if a patient returns to the same hospital for an additional visit. I can not observe I they go to Hospital A and then Hospital B. This is a limitation of the data, and of the study.
This assumes that their job is n the formal sector and is paying payroll taxes on their behalf.
Baicker, K., Taubman, S. L., Allen, H. L., Bernstein, M., Gruber, J. H., Newhouse, J. P., et al. (2013). The Oregon experiment—Effects of Medicaid on clinical outcomes. New England Journal of Medicine, 368(18), 1713–1722.
Bruen, B. K., Ku, L., Lu, X., & Shin, P. (2013). No evidence that primary care physicians offer less care to medicaid, community health center, or uninsured patients. Health Affairs, 32(9), 1624–1630.
Buchmueller, T. C., Grumbach, K., Kronick, R., & Kahn, J. G. (2005). Book review: The effect of health insurance on medical care utilization and implications for insurance expansion: A review of the literature. Medical Care Research and Review, 62(1), 3–30.
Cooper, Z., Craig, S. V., Gaynor, M., & Van Reenen, J. (2019). The price ain’t right? Hospital prices and health spending on the privately insured. The Quarterly Journal of Economics, 134(1), 51–107.
Currie, J., & Gruber, J. (2001). Public health insurance and medical treatment: the equalizing impact of the medicaid expansions. Journal of Public Economics, 82(1), 63–89.
Doyle, J. J. (2005). Health insurance, treatment and outcomes: Using auto accidents as health shocks. The Review of Economics and Statistics, 87(2), 256–270.
HSCRC. (2010). Disclosure of hospital financial and statistical data. https://nam02.safelinks.protection.outlook.com/?url=https%3A%2F%2Fhscrc.maryland.gov%2FDocuments%2Fpdr%2Far%2FDisclosure%2FHSCRC_DisclosureReport-FY09.pdf&data=02%7C01%7Caccook%40bgsu.edu%7C6206de70c6c5429916f908d7c0315230%7Ccdcb729d51064d7cb75ba30c455d5b0a%7C1%7C1%7C637189190746587382&sdata=z%2BUvH1CSv7ocMALrd0tZHoOhGT4o13oOoCKat1KTmyo%3D&reserved=0.
HSCRC. (December 4, 2013). Report on results of uncompensated care policy and final recommendation to suspend the formula for calculating the hospital specific results.
Keeler, E. B., & Rolph, J. E. (1988). The demand for episodes of treatment in the health insurance experiment. Journal of Health Economics, 7(4), 337–367.
Lefgren, L., & McIntyre, F. (2009). Explaining the puzzle of crossstate differences in bankruptcy rates. The Journal of Law & Economics, 52(2), 367–393.
Long, S. H., Marquis, M. S., & Rodgers, J. (1998). Do people shift their use of health services over time to take advantage of insurance? Journal of Health Economics, 17(1), 105–115.
Mahoney, N. (2015). Bankruptcy as implicit health insurance. American Economic Review, 105(2), 710–46.
Manning, W. G., Newhouse, J. P., Duan, N., Keeler, E. B., & Leibowitz, A. (1987). Health insurance and the demand for medical care: Evidence from a randomized experiment. The American Economic Review, 77, 251–277.
Marquis, M. S., & Long, S. H. (1994–1995). The uninsured access gap: Narrowing the estimates. Inquiry, 31(4), 405–414.
Murray, R. (2009). Setting hospital rates to control costs and boost quality: The maryland experience. Health Affairs, 28(5), 1395–1405.
Reinhardt, U. E. (2011). The many different prices paid to providers and the flawed theory of cost shifting: Is it time for a more rational all-payer system? Health Affairs, 30(11), 2125–2133.
Smith, S. (2014). Young invincibles’ ok with risk of no insurance. CNN. Retrieved February 11, 2014 from http://edition.cnn.com/2009/HEALTH/02/17/cdc.young.people/.
Systems, H. I. (July, 2013). 3m apr drg classification system. https://news.3m.com/press-release/company/state-maryland-adopts-3m-apr-drg-classification-system-payment-healthcare-serv.
Weissman, J. S., & Epstein, A. M. (1993). The insurance gap: Does it make a difference? Annual Review of Public Health, 14(1), 243–270.
White, C., Bond, A. M., Reschovsky, J. D., & LTHPOLICY, P. C. (2013). High and varying prices for privately insured patients underscore hospital market power. Center for Studying Health System Change Research Brief No.
White, F. A., French, D., Zwemer, F. L., & Fairbanks, R. J. (2007). Care without coverage: Is there a relationship between insurance and ed care? The Journal of Emergency Medicine, 32(2), 159–165.
Thanks to Steve Martin, Jack Barron, and James Bland for their input and suggestions, to the HSCRC for proving me with data, to Dr. Rich Parker and Dr. Dean Tully for medical consultation, to the SHaPE seminar at Purdue and the AERUS conference for feedback and suggestions. All errors are my own.
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Appendix 1: Details of base rate adjustments
As hospitals have different underlying costs, they are not paid uniformly for an APRDRG weight of 1. The hospital-specific payment for a DRG of weight 1 is called a hospital’s base rate. To derive a hospital’s base rate, one takes the ratio of allowable charges to the hospital’s case mix adjusted discharges (CMAD). CMAD is the sum of all APRDRG weights for patients treated in the hospital in that year. Allowable charges differ from total charges in a few important ways, not all charges from the hospital are included in the approved charges calculation. Specifically, patients who stay for only one night are excluded from allowable charges, as are some patients that are deemed to be outliers. Additionally, there are a few additional adjustments. The first adjustment is for inflation and productivity using Global Insight’s index, which measures the effect of general inflation in the economy on health care expenses. The second adjustment is a volume adjustment, which relies upon a historically made assumption about variable costs. The adjustment works as follows. If a hospital’s service volume increased 10% in a year, in calculating the following years allowable revenue, only 85 percent of the increase would be considered as admissible charges. This is assumed to be the variable cost portion of the increased revenue of the hospital. The remaining adjustments are small in comparison to the two above, but for the sake of completeness are detailed below.The third adjustment considers any changes in the proportion of Medicare/Medicaid patients to the hospital. This adjustment exists because Medicare and Medicaid only pay 94% of charges. If there is growth in Medicare/Medicaid then it could be disadvantageous to a hospital’s bottom line. For illustration, consider the following example of a hospital with 200 million in charges and 100 million of those charges belonging to Medicare and Medicaid patients. The discount from the hospital to the government is 6 million: \((1-.94)*100\) million. The existing $6 million discount is taken into account in the charge structure. An adjustment would be made only on the basis of growth in the amount of Medicare and Medicaid patients. If the total charges of the hospital remained at 200 million, but now 150 million was attributable to Medicare and Medicaid, this growth in government-insured individuals induces a volume adjustment to this hospital. With 150 million in charges attribute to governmental insurance, this hospital would have a 9 million dollar shortfall, but 6 million was already accounted for in the initial cost structure. The volume adjustment covers this 3 million of additional discounts to the government. When half of the hospital’s charges were governmental, the hospital was getting 3 million of revenue. To not disadvantage the hospital when its governmental charges rise to 150, Maryland’s adjustment takes the 3 million ‘discount’ that previously was revenue, and simply increases the its initial charges by this amount. Now this hospital would be credited with 203 million in aggregate charges, its initial 200 million plus the 3 million discount for the government. The fourth adjustment pertains to capital investment, certificate-of-need project adjustments. If a hospital has major renovations, a portion of those costs can be covered by an increased base rate. The fifth adjustment is an adjustment for medical education. There are two types of medical education expenses, direct medical education expenses (DME) and indirect medical education expenses (IME). Both of these expenses are built into the base rate in levels. Thus, any adjustment to the base rate would be because of growth in DME or IME. For example, if a program added two more residents, their salaries would be an increase in DME. Finally, there is an adjustment for changes in the rate of bad debt (uncompensated care). To calculate future uncompensated care, an average is taken of previous rates of uncompensated care and a prediction of uncompensated care based on a regression. Again, only changes in uncompensated care require an adjustment; levels are built into the base rate.
Appendix 2: Comparisons to medicaid with high and low risk illnesses
See Fig. 7
Appendix 3: Variation in Charges per Weight by insurance status and hospital
See Table 6.
Appendix 4: Replication excluding charity care
See Table 9.
See Table 10.
Rights and permissions
About this article
Cite this article
Cook, A. Do the uninsured demand less care? Evidence from Maryland’s hospitals. Int J Health Econ Manag. 20, 251–276 (2020). https://doi.org/10.1007/s10754-020-09280-4
- Health insurance
- Demand for health care
- Health and inequality