Health Affairs, 23, no. 4 (2004): 167-175 doi: 10.1377/hlthaff.23.4.167 © 2004 by Project HOPE	New Online   Obama's Success & Next Steps   3 Health Policy Options for President Obama   A Hybrid Proposal   Bipartisan Compromise

This Article Abstract Figures Only Reprint (PDF) Submit a response to this article Alert me when this article is cited Alert me when Comments are posted Alert me if a correction is posted Services E-mail this article to a friend Similar articles in this journal Similar articles in ISI Web of Science Similar articles in PubMed Alert me to new issues of the journal Add to My Personal Archive Download to Citation Manager Reprints & Permissions Citing Articles Citing Articles via ISI Web of Science (2) Citing Articles via Google Scholar Google Scholar Articles by Monheit, A. C. Articles by Fox, K. S. Search for Related Content PubMed PubMed Citation Articles by Monheit, A. C. Articles by Fox, K. S. Related Collections Access To Care Health Reform Insurance Coverage Insurance - Employer-Based System Legal/Regulatory Issues Business Of Health Safety-Net Systems State/Local Issues Consumer Issues

Health Tracking

MARKETWATCH

Community Rating And Sustainable Individual Health Insurance Markets In New Jersey

Abstract

 
The New Jersey Individual Health Coverage Program (IHCP) was implemented in 1993; key provisions included pure community rating and guaranteed issue/renewal of coverage. Despite positive early evaluations, the IHCP appears to be heading for collapse. Using unique administrative and survey data, we examined trends in IHCP enrollment and premiums. We found the stability of the IHCP to be fragile in light of improving opportunities for job-related health insurance. We also found that it is retaining high-risk enrollees. Institutional realities and the difficulty of identifying a control group preclude attributing causality to the plan’s pure community rating and open enrollment provisions.

Previous research and the current context. In 1999/2000 evaluations, Katherine Swartz and Deborah Garnick concluded that initially the IHCP achieved its stated goals, and they found no evidence of adverse selection by enrollees.³ While they noted that a decline in IHCP enrollment began in 1996 and was accompanied by rising premiums, they attributed the latter to perverse incentives inherent in the IHCP loss assessment system rather than adverse selection.

In contrast, the IHCP’s current situation points to a market that is heading for collapse. Enrollment has declined from a peak of 186,130 lives at the end of 1995 to 84,968 at the end of 2001. In addition, premiums have increased two- to threefold above their early levels. These changes have raised concerns as to whether a comprehensive regulatory effort such as the IHCP can yield a sustainable health insurance market.

In this paper we examine trends in IHCP enrollment and premiums and consider whether the IHCP regulatory provisions may be associated with some unintended consequences for insurance market stability and access to coverage. We also consider whether the institutional and economic realities that confronted the IHCP contributed to these trends. We find evidence suggesting that the stability of the IHCP market may be sensitive to changing economic circumstances and opportunities to obtain employment-related health insurance. The trends that we observe also suggest that the IHCP may be retaining adverse health risks.

Community rating, adverse selection, and market stability. IHCP provisions such as pure community rating and guaranteed issue were intended to expand access to coverage by limiting insurers’ risk selection and medical underwriting practices. However, such requirements can also have a destabilizing effect on specific health plans within an insurance market and, ultimately, upon the entire market.

Since pure community rating imposes the same premium on low- and high-risk people, the premiums of low risks exceed their actuarially fair level, while those of high risks are lower than their fair level. A sustainable market equilibrium may be tenuous under such a requirement.⁴ In a market with choice among community-rated health plans, low risks will seek entry to cheaper, more restrictive health plans that are unattractive to high risks, leaving the latter in the more generous and expensive plans. As low-risk people leave the more generous health plans, higher risks will dominate such plans, and their premiums will rise.

Such a "separating market equilibrium" can be sustained only if low risks find the more restrictive plans to be of value and remain in the market. Should low-risk people defect from the market, total plan enrollment will decline. The market risk profile will become increasingly dominated by high risks, and a market-wide adverse-selection death spiral may ensue as cycles of rising premiums spur further defections of lower-risk enrollees. As we discuss below, the decline in IHCP enrollment and rise in premiums has been accompanied by retention of potentially adverse health risks. However, establishing a causal link to the presence of pure community rating remains difficult.

Data And Methods

Top Data And Methods Findings Conclusions Editor's Notes NOTES

 
Data on IHCP aggregate enrollment, plan-specific enrollment, and premiums were obtained from administrative records of the IHCP board.⁵ For our analysis, we use premiums for single coverage, the predominant form of IHCP contracts.

Our tabulations are based on enrollment data from the first quarter of 1994 to the fourth quarter of 2001 for representative IHCP plans: Plans B, C, and D (all indemnity) and all HMO plans. Trends in IHCP plan premiums are based upon data from March 1996 to December 2001. For each of the following plan types—Plans B and C (40 percent and 30 percent coinsurance, respectively, and both with $1,000 deductibles); Plan D (20 percent co-insurance and $500 deductible, the most generous plan offered based on these provisions); and an HMO plan ($15 copayment)—we constructed a plan-specific composite premium. This measure is based upon premiums for each plan from the top four carriers (based on enrollment) weighted by their enrollment shares.⁶ We also compare trends in the composite premiums to those for selected small carriers.

We supplement these tabulations with household data from the Current Population Survey (CPS) on the health insurance status of New Jersey residents, published data on residents’ employment, data on premiums for employment-based coverage in New Jersey from the Medical Expenditure Panel Survey—Insurance Component (MEPS-IC), and data on premiums in the New Jersey small-employer group market. Finally, we include data from the 2001 New Jersey Family Household Survey (NJFHS), a statewide probability sample of 2,265 families conducted by the Center for State Health Policy, Rutgers University, and its 2002 supplement of 601 families with IHCP subscribers.

Findings

Top Data And Methods Findings Conclusions Editor's Notes NOTES

 
Trends in IHCP enrollment and premiums. The IHCP initially experienced a sharp increase in enrollment, from 51,648 lives in the first quarter of 1994 to 186,130 lives in the fourth quarter of 1995. However, since then IHCP enrollment has declined dramatically (Exhibit 1).

View larger version (21K): [in this window] [in a new window]   EXHIBIT 1 Enrollment In The New Jersey Individual Health Coverage Program (IHCP), First Quarter 1994 Through Fourth Quarter 2001

View larger version (30K): [in this window] [in a new window]   EXHIBIT 2 Enrollment In The New Jersey Individual Health Coverage Program (IHCP), By Plan Type, First Quarter 1994 Through Fourth Quarter 2001

View larger version (27K): [in this window] [in a new window]   EXHIBIT 3 Trends In Premiums For Selected New Jersey Individual Health Coverage Program (IHCP) Plans, By Plan Type, First Quarter 1996 Through Fourth Quarter 2001

Potential factors contributing to IHCP market instability. Carrier loss assessment mechanism. As Swartz and Garnick noted, the loss assessment mechanism (which allowed reimbursement for payouts in excess of 75 percent of net premium income) encouraged small carriers to sell coverage in the IHCP and to charge very low premiums during their initial years, to expand their market share.⁸ However, this strategy was not successful. By charging low premiums, these small carriers incurred larger-than-expected losses, which led them to raise premiums and, later, led several small carriers to defect from the market. This contributed a certain degree of "market chaos," as many enrollees shifted carriers or dropped coverage. Swartz and Garnick have argued that the rise in IHCP premiums between 1995 and early 1998 reflected the behavior of these small carriers rather than adverse selection against certain carriers or the IHCP.⁹

We find evidence that is consistent with this assertion. Compared with the trend in the composite premium over this period, the premiums of such small carriers increased markedly. For example, between the first quarters of 1996 and 1998, Plan B premiums for Manhattan National Life Insurance increased by 415 percent; for Metropolitan Life Insurance, 86 percent; and for Time Insurance Company, 110 percent. In contrast, the composite premium for Plan B increased by only 20 percent. As small carriers raised premiums, their market shares declined precipitously from 27 percent of insured lives in 1996 to less than 1 percent by 1998. Eight small carriers accounted for roughly half of enrollment losses over this period. By 2000 only one of the carriers (Metropolitan) remained in the market.

Subsequently, the state legislature changed the loss assessment mechanism to a two-year retrospective period effective 1 January 1998 and required reimbursable losses to exceed 115 percent of carriers’ income. State officials believe that the latter provision greatly reduced incentives for carriers to "game" the system.¹⁰ Small carriers continued to raise premiums and lose enrollment; after 1998 several additional small carriers withdrew from the market.

Access to employer-sponsored health insurance. Over our study period, a lower-cost substitute for IHCP coverage—employer-sponsored insurance—became more widely available. This was the result of two factors. First, New Jersey experienced sizable increases in economic activity and employment opportunities, which provided increased access to employer coverage. For example, between January 1994 and January 2000 seasonally adjusted resident employment increased from 3.702 million to 4.035 million.¹¹ Correspondingly, CPS data for New Jersey reveal that enrollment in employer coverage increased from 4.7 million in 1996 to 5.3 million in 2000.¹²

Next, coincident with the implementation of the IHCP, New Jersey created the Small Employer Health Benefits Program (SEHBP) in January 1994, which provided small employers (those with 2–50 full-time employees) with open enrollment into standardized health plans at modified community rates (adjusted for employees’ age, sex, and family status and for business location). The SEHBP experienced a rapid increase in enrollment, from 694,312 in the fourth quarter of 1994 to a peak of 937,784 in the third quarter of 1999 (declining to 884,104 in the third quarter of 2001). Finally, IHCP data also indicate a decline in contracts issued to employed relative to non-employed subscribers, from just over 2:1 in June 1994 to just over 1:1 in January 2001, consistent with the shift to employer-sponsored coverage.

Exhibit 4 suggests that differential premiums provided ample incentives for IHCP enrollees to obtain employer-sponsored coverage. The data indicate that by 2000, IHCP premiums for our sampled individual coverage plans exceeded and (with the exception of HMO coverage) rose faster than employer coverage premiums. Moreover, if people perceive the cost of employer coverage as their own contribution rather than the full premium, then the differential in out-of-pocket premium costs between employer and IHCP plans is substantial. As well, modified community rating in the SEHBP may have exacerbated the impact of growing availability of employer coverage on IHCP enrollment and premiums, as the cost advantage of employer coverage would be greatest for lower-risk (that is, younger) workers.¹³

Elimination of the Health Access subsidy program. After the beginning of 1996, New Jersey eliminated Health Access, the state’s subsidy program for IHCP enrollees with incomes below 250 percent of the federal poverty level. The program began in May 1995; by March 1996 it subsidized as many as 20,000 enrollees. Fiscal constraints ended new enrollments in the program after 31 December 1995. This precluded replacement of low-income enrollees whose circumstances no longer warranted use of the access program.¹⁴

Implementation of the New Jersey SCHIP program. As part of the State Children’s Health Insurance Program (SCHIP), the state enacted NJ KidCare in December 1997 for children under age nineteen in families earning less than 350 percent of poverty. The program was expanded to adults with incomes less than 200 percent of poverty through NJ FamilyCare in July 2000. It is unlikely that these programs resulted in sizable crowding out of IHCP enrollees. Enrollment in KidCare lagged initially, and much of the subsequent enrollment growth was in plans for children with family incomes below 150 percent of poverty (who were very unlikely to have enrolled in IHCP coverage). In addition, FamilyCare was implemented toward the very end of our study period and was not a factor for most of the period of declining IHCP enrollment.¹⁵

Further considerations. Carrier loss assessment, growth in employer coverage, and elimination of the Health Access program may explain a substantial part of the decline in IHCP enrollment and rise in premiums before 1998. Although the robustness of the employer coverage market through 2000 may have continued to fuel IHCP trends, its effect was diminishing as economic activity slowed down. Indeed, since 1998 IHCP enrollment has continued to decline at a rate of about 3 percent per quarter, and premiums have continued to rise. Thus, it is reasonable to ask whether these continuing trends are unique to the IHCP, reflecting an adverse-selection death spiral, or simply consistent with national trends in enrollment and premiums.

Changing age composition of the IHCP. Comparisons between data from the NJFHS and published 1996 tabulations provide evidence that the IHCP has shifted markedly toward older and thus potentially higher-cost enrollees.¹⁶ In 1996, 44.6 percent of new adult enrollees to the IHCP (those ages 18–64 enrolled less than a year) were ages 45–64, with a mean age of 41.9 years; by contrast, in 2002, 66.3 percent of new adult enrollees were between those ages, and their average age rose to 48.4 years (both figures significantly different from 1996, p < .01). As noted below, the change in these data for IHCP enrollees during this period also exceeded that for people with employer coverage. Finally, we also find some evidence that the percentage of new adult enrollees reporting fair or poor health rose between 1996 and 2002 (4.3 percent compared with 8.1 percent, p < .10), although this increase was not statistically different from that for people with employer coverage.

Following Swartz and Garnick, we also compared the age composition of new adult IHCP enrollees with all New Jersey adults having employer coverage. We found that the former were older (48.4 years compared with 40.7 years, p < .01) and more likely to be ages 45–64 years (66.3 percent compared with 39.5 percent, p < .01). These differences have widened considerably compared with those reported for 1996 (data not shown). When we compare all adult IHCP enrollees with adults having employer coverage, we observe that the former are older, more likely to be female, and more likely to be in fair/poor health (data not shown). Thus, the IHCP may be attracting and retaining more potentially costly enrollees.

Comparison with national trends. The decline in IHCP enrollment appears to be far more severe than national declines in individual coverage. While IHCP enrollment declined by 41 percent between 1996 and 2000, national CPS estimates of individual coverage indicate only a 4 percent decline.¹⁷ MEPS data between 1996 and 1999 reveal an 18 percent decline in non-group coverage—well below the 34 percent decline in IHCP enrollment during this period. Comparisons with national survey data on employer-sponsored coverage also suggest that the IHCP sustained a more pronounced shift in enrollment from indemnity to HMO coverage.¹⁸ Correspondingly, IHCP premiums rose more than employer coverage premiums did. These comparisons suggest that the trends for New Jersey may reflect the unique circumstances of the IHCP rather than national trends.

A caveat regarding interpretation. While the trends in IHCP enrollment and age composition may reflect necessary conditions for an adverse-selection death spiral in a community-rated insurance market, we cannot establish that they are sufficient. This reflects at least two important limitations. First, the unique nature of our administrative and household data on IHCP enrollment limits comparisons with potential control states that lack the IHCP regulatory provisions. Thus, we cannot infer whether community rating in the IHCP played a causal role in the trends we observe. Next, while household data suggest that the age composition of new IHCP enrollees may have changed over time, further analysis is required to discern whether such health-related characteristics have also changed across IHCP plan types. Thus, while the IHCP trends that we observe may be consistent with adverse selection and the retention of poor health risks, we cannot definitively assign causality to pure community rating.

Conclusions

Top Data And Methods Findings Conclusions Editor's Notes NOTES

 
Our analysis strongly suggests that the IHCP is in the midst of an enrollment crisis that threatens its market stability and ability to fulfill its stated goals. In assessing the New Jersey experience, it is important to note that enrollment in individual health insurance is fragile to begin with, representing a small minority of all nonelderly, privately insured people. A certain amount of churning and disenrollment is also expected in this market, since enrollment spells are typically short and frequently serve to bridge spells without employer coverage.¹⁹ One must further recognize that the IHCP is a voluntary and unsubsidized insurance market.

Critics will be quick to attribute problems in the IHCP to community rating and open enrollment. However, other factors we have identified may have played a key role in the decline in covered lives and rise in premiums. These trends exceed those observed nationally and have been accompanied by a deterioration in the IHCP’s risk composition. While the IHCP’s market rules may have contributed to an adverse-selection death spiral, we cannot definitively identify this market dynamic. More importantly, it is not clear whether the decline of this market would have continued absent the buoyant employer coverage market. In this regard, the New Jersey experience raises the critical policy issue of whether the individual market can serve as "private insurer of last resort" when it is inherently unstable and encounters alternative private coverage with considerably lower out-of-pocket premium costs. Market stability may be further threatened should pure community rating exacerbate disparities in coverage costs between these markets for low-risk enrollees.

In this context, existing research points to the difficulty of sustaining the individual market, especially when small-group market reform is present and robust economic circumstances provide better access to lower-cost employer coverage.²⁰ The interrelationship between the individual and employment-based markets, therefore, is an unavoidable reality confronting policymakers in the design of individual market reform.

Finally, the behavior of the individual insurance market in New Jersey raises some hard choices for policymakers. Given the state’s current fiscal difficulties and empirical evidence suggesting that large subsidies would be required in the face of low demand responsiveness, efforts to stabilize the IHCP through subsidized premiums would seem prohibitive.²¹ Alternatives such as moving to modified community rating might make individual coverage more attractive to low-risk enrollees. However, such a strategy might further segment the market to the disadvantage of higher risks, the very group whose access to coverage the IHCP is seeking to ensure. While pooling of IHCP together with the SEHBP was earlier evaluated and dismissed because of the potential for a sizable increase in SEHBP premiums, such pooling may have less of an impact now, as IHCP enrollees now represent less than a tenth of total enrollment in the SEHBP market (compared with nearly a quarter when blending was first evaluated).²²

In sum, should IHCP enrollment continue to decline and premiums continue to increase, New Jersey policymakers may face the dilemma of seeing the goals of the IHCP vitiated, plan choice diminished, and the market segmented between plans enrolling high- and low-risk people. Should this be the case, history may repeat itself as insurers experience unsustainable losses and the availability of even high-cost individual coverage diminishes.

Editor's Notes

Top Data And Methods Findings Conclusions Editor's Notes NOTES

 
Alan Monheit (monheiac{at}umdnj.edu) is a professor in the Health Systems and Policy Division, School of Public Health, University of Medicine and Dentistry of New Jersey, in Piscataway. Joel Cantor directs the Center for State Health Policy, Rutgers University, in New Brunswick, where Margaret Koller is associate director for planning and program initiatives and Kimberley Fox, a senior policy analyst.

Financial support from the Commonwealth Fund and the Changes in Health Care Financing and Organization (HCFO) initiative of the Robert Wood Johnson Foundation is gratefully acknowledged. The authors are indebted to Ellen F. DeRosa, New Jersey Department of Banking and Insurance (NJDOBI), for the data used in this analysis. They are grateful to DeRosa, Wardell Sanders, and Vicki Mangiaracina of the NJDOBI for helpful discussions regarding the interpretation of trends in IHCP enrollment and premiums. They also thank Kathy Swartz and two anonymous referees for useful and insightful comments, and Piu Banerjee for excellent research assistance.

NOTES

Top Data And Methods Findings Conclusions Editor's Notes NOTES

 

1. For a discussion of circumstances leading to the IHCP’s formation, see K. Swartz and D. Garnick, "Lessons from New Jersey," Journal of Health Politics, Policy and Law 25, no. 1 (2000): 45–70.[CrossRef][ISI][Medline]
2. The loss assessment system served to reduce the costs to carriers seeking to enter the market and to provide subsidies to sustain a market with potential adverse selection. See ibid.; and K. Swartz and D. Garnick, "Regulating Markets: Lessons from New Jersey’s Individual Health Insurance Coverage Program" (Unpublished manuscript, Harvard University, October 1998).
3. Swartz and Garnick, "Lessons from New Jersey"; and K. Swartz and D. Garnick, "Can Adverse Selection Be Avoided in a Market for Individual Insurance?" Medical Care Research and Review 56, no. 3 (1999): 373–388.[Abstract/Free Full Text]
4. M. Rothschild and J. Stiglitz, "Equilibrium in Competitive Insurance Markets: An Essay on the Economics of Imperfect Information," Quarterly Journal of Economics 90, no. 4 (1976): 630–649; and T. Buchmueller and J. DiNardo, "Did Community Rating Induce an Adverse Selection Death Spiral? Evidence from New York, Pennsylvania, and Connecticut," American Economic Review 92, no. 1 (2002): 280–293.[CrossRef][ISI]
5. For the most recent IHCP plan characteristics and monthly premiums, see "New Jersey Individual Health Coverage Program Rates," 4 March 2004, www.nj.gov/dobi/ihcrates.htm (13 April 2004).
6. Carriers’ market share changed over time, so the composite premium will include different carriers.
7. We expect that network and referral restrictions would lead most enrollees to consider HMOs to be less generous than indemnity plans. See Buchmueller and DiNardo, "Did Community Rating Induce an Adverse Selection Death Spiral?"; and R. Feldman and B. Dowd, "Risk Segmentation: Goal or Problem?" Journal of Health Economics 19, no. 4 (2000): 499–512.[CrossRef][ISI][Medline]
8. The discussion follows Swartz and Garnick, "Lessons from New Jersey"; Swartz and Garnick, "Regulating Markets"; and K. Swartz and D. Garnick, "Unintended but Predictable Outcomes of Regulations: The Case of New Jersey’s Individual Health Care Program" (Unpublished manuscript, Harvard University, 12 January 1998).
9. Swartz and Garnick, "Lessons from New Jersey"; and Swartz and Garnick, "Regulating Markets." The small carriers were Manhattan National Life Insurance, Metropolitan Life Insurance, National Casualty, Protective Life Insurance, Time Insurance, TMG Life Insurance, Travelers Insurance, and Washington National Insurance.
10. Based on discussions with representatives from the New Jersey Department of Banking and Insurance, Fall 2003.
11. New Jersey Department of Labor, "New Jersey Economic Indicators: Historical Data Series, 1976–2001, Resident Employment, Series 2," www.wnjpin.net/OneStopCareerCenter/LaborMarketInformation/lmi07/series2.htm (13 April 2004).
12. Data are from P. Fronstin, Sources of Health Insurance and Characteristics of the Uninsured: An Analysis of the March 2001 Current Population Survey, EBRI Issue Brief no. 240 (Washington: Employee Benefit Research Institute, December 2001). Following Fronstin, we have adjusted the estimates to reflect the inclusion of an insurance status verification question in the CPS in 1998.
13. See New Jersey Small Employer Health Benefits Program Board, "The Effects on the Individual and Small Employer Health Coverage Markets of Permitting Individuals to Purchase Small Employer Health Benefits Plan: A Report to the New Jersey Legislature" (Unpublished report, 25 September 1996), for a comparison of monthly premiums for a small employer group insured through specific IHCP and SEHBP plans.
14. Swartz and Garnick, "Regulating Markets."
15. S. Silow-Carroll et al., Assessing State Strategies for Health Coverage Expansions: Case Studies of Oregon, Rhode Island, New Jersey, and Georgia (New York: Commonwealth Fund, November 2002).
16. Swartz and Garnick, "Lessons from New Jersey."
17. Fronstin, Sources of Health Insurance.
18. Data are from Henry J. Kaiser Family Foundation/Health Research and Educational Trust, Employer Health Benefits: 2002 Annual Survey (Washington: Kaiser Family Foundation, September 2002).
19. T. McBride et al., "The Dynamics of Individual Insurance Coverage in the U.S." (Paper presented at the Twentieth Annual Research Meeting of AcademyHealth, Nashville, Tennessee, 27–29 June 2003). Median spell length for people with individual coverage is five months. On average, older people, especially those nearing retirement, remain in the market longer than younger people.
20. See M. Hall, "An Evaluation of New York’s Reform Law," Journal of Health Politics, Policy and Law 24 no. 1 (2000): 71–100; K. Thorpe, "Who Purchases Individual Insurance? A Comparision of New York State, Regional, and National Patterns, 1994–1997" (Unpublished manuscript, Tulane University Medical Center, March 1999); and D. Chollet, "Changes in the Individual Insurance Market" (Paper presented at the Annual Research Meeting of AcademyHealth, Washington D.C., 23 June 2002).
21. M.S. Marquis and S.H. Long, "Worker Demand for Health Insurance in the Non-Group Market," Journal of Health Economics 14, no. 1 (1994): 47–63.
22. See New Jersey SEHBP Board, "The Effects on the Individual and Small Employer Health Coverage Markets." Figures in the text are the authors’ computations; details are available on request by e-mailing Alan Monheit, monheiac{at}umdnj.edu.

                      What's this?