Why Employers Are Likely to Drop Health Insurance – A Simplified View

eytan
Print Friendly, PDF & Email

Richard Evans / Scott Hinds

203.901.1631 /.1632

richard@ / hinds@ssrllc.com

July 11, 2011

Why Employers Are Likely to Drop Health Insurance – A Simplified View

  • All else equal, employers ‘compete’ for employees with compensation packages of wage + benefits, the primary benefit typically being health coverage
  • Logically, employers will seek to produce compensation packages of net wage and health coverage as efficiently as possible; and, will choose between ESI- and HIE-based approaches based largely on which approach delivers a given level of wage and health coverage at the lowest cost
  • For employees with household incomes in the range eligible for subsides on the HIEs (133% – 400% FPL), employers’ total cost (including tax effects, penalties, and any added wage given to employees to cover their added premium costs) of delivering family and ‘single + one’ health coverage is lowered (relative to traditional ESI) by shifting employees to HIEs
  • The exception is the market for single coverage, where ESI-based compensation packages are more efficient across the majority of subsidy-eligible income ranges
  • We conclude that HIE-based packages are the most efficient choice for employers with subsidy-eligible employees that desire family coverage, ‘single + one’ coverage, or no coverage. These groups encompass roughly 62% of employees, 53% of enrollees, and 73% of premiums paid
  • The preceding assumes that employees are indifferent to HIE v. ESI at a given level of net (of taxes and health premiums) wage and actuarial value (AV) of health coverage; in fact we believe that all else equal employees should prefer HIE-based compensation packages, as these packages offer a far greater variety of coverage options (in terms of both the source and generosity of coverage). Younger / healthier employees in particular should prefer HIE-based packages, as these groups are likely to place greater relative value on the opportunity to purchase less generous (or even no) coverage than is typical of ESI-based compensation packages, freeing net wage for other uses
  • We further conclude that, absent substantial (and presently unforeseen) changes to either PPACA or the federal tax code, employers with large populations of subsidy-eligible employees will shift these employees onto the HIEs
  • A large-scale ESI to HIE shift implies lower overall levels of coverage (as former ESI recipients choose less generous or no coverage on the HIE) than would be expected if ESI were ‘stable’. 2014+ expectations for system-wide volume gains may not be met; and, 2014 may bring less relief than expected to hospitals’ uncompensated care burden. Federal budgets for HIE subsidies are too low
  • We favor HMOs on the belief that near-term increases in MLRs are unlikely; though we recognize that an ESI to HIE shift in 2014 may mean fewer contracts and lower average contract sizes than expected, as well as operating margin pressure as enrollee mix shifts from large groups under ESI to individuals on the HIEs

To the Employer, What’s the Most Cost-Effective Means of Delivering a Given Level of Compensation?

We frame the question of whether employers will continue to offer employer-sponsored health insurance (ESI) after 2014, or shift employees onto health insurance exchanges (HIEs), as a simple matter of relative efficiency

Employers compete for labor with compensation packages consisting of wage plus benefits, with the most valuable benefit typically being health coverage. Ignoring non-health benefits for simplicity, we believe, all else equal, that employees seek and accept employment in expectation of a given level of wage and health coverage

At any given level of net wage and actuarial value (AV) of health coverage, we assume that employees are indifferent as to whether their health coverage comes from ESI or an HIE. For example, consider a head of household choosing between a job with an employer (Employer A) that offers ESI, and one that does not (Employer B). Employer A offers a pre-tax salary of $71,250; and, family health coverage with an AV of 0.82 that requires the employee to cover 30% of premium costs. In 2014 dollars her health insurance premium cost is $5,182, which reduces her pre-tax income to $66,068, leaving her with after-tax / after-health cost purchasing power of $53,370. Employer B offers a higher pre-tax salary of $75,362, but does not offer ESI. However the exact same type and value of health coverage offered by Employer A is offered on her state’s HIE, and she can buy this coverage for an annual premium contribution of $6,769 (which is not tax-deductible). After paying her taxes and health premiums under Employer B’s offer, her wage is $53,370 (Exhibit 1). In both cases she has the exact same health coverage and net purchasing power after taxes and health premiums; thus in terms of net wage and actuarial value of health coverage, the two offers are equal

Given that these two offers are equal to the employee[1], the matter of determining which offer is likely to be more common depends on which offer costs the employer less – the one that offers a lower salary and ESI (Employer A), or the one that offers a higher salary and no ESI (Employer B). I.e. a rational employer, seeking to produce a given compensation package of net wage + health benefits as efficiently as possible, should choose between ESI and HIE on the basis of which venue allows the employer to deliver the target compensation package at the lowest possible cost. Employer B’s offer is less expensive, by roughly $2,666 (Exhibit 2)

For beneficiaries seeking family coverage, we find that eliminating ESI and offering additional wage sufficient to cover employees’ added costs of purchasing health coverage on an HIE is the most efficient means of compensating an employee with a household income level that falls within the eligibility standards for HIE premium subsidies (138% – 400% FPL)

If the employer chooses ESI, the firm pays its share of the total insurance premium with pre-tax dollars, thus the employer’s cost of providing health coverage is simply the employer’s share of the total insurance premium, adjusted for tax (tc = effective corporate tax rate):

If the employer chooses not to offer ESI, but chooses instead to compensate employees for any added cost of purchasing equivalent coverage on an HIE, the employer’s cost is the penalty the employer pays for not offering coverage (or for offering insufficient coverage), plus the tax-effected cost of providing additional cash compensation to employees whose premium costs on the HIE are higher than under ESI (tp = personal tax rate):

In comparing these two alternatives, we assume:

  • average family health costs in 2014 will be $17,906[2]
  • employers seek to give their employees access to a health plan having an actuarial value of 0.82 as efficiently as possible. We estimate annual premiums of $18,354 for this coverage on the HIE, and $17,274 under ESI[3]
  • a 2014 Federal Poverty Level (FPL) of $20,135 for a family of 3.19 persons[4]
  • that employees enrolled in family coverage under ESI pay 30% of applicable premiums[5]
  • a penalty paid by the employer of $3,000 per subsidy eligible employee purchasing coverage on the HIE[6]; and,
  • that the average effective US corporate tax rate is 29%

Exhibit 3 provides the comparison; costs to the employer of providing family health insurance under ESI, or of having employees purchase equivalent coverage on the HIE, are shown for all subsidy-eligible income levels. Under our assumptions, for all subsidy-eligible income levels (138% – 400% FPL) the HIEs are a more efficient setting than ESI for purchasing a given level of family health insurance. A detailed table of results is provided in Appendix I(b)

 

This basic conclusion is relatively insensitive to changes in our assumptions, the exception being relative rates of health cost growth and general inflation. We assume 7.5% annual health cost inflation (the 5-year trailing average is 8.2%) and 1.5% CPI growth (consistent with CBO’s long-term model); if we lower health cost growth to the rate of CPI, then HIEs are more efficient ‘only’ up to a level of 380% FPL

Family coverage is a bit more than one-half of the ESI premium market; we estimate that single and ‘single + one’ policies are roughly one-quarter and one-fifth of the ESI premium market, respectively (Exhibit 4)

Premiums in the single + one and single markets obviously are lower than in the market for family coverage. Lower premiums mean lower costs to employers for ESI, and smaller federal subsidies on the HIEs (since lower premiums are less likely to go very far beyond the percent-of-income limits). As a result, in the single + one market HIE is a more efficient choice for employers up to about 350% FPL assuming the $3,000 employer penalty applies, and throughout the entire subsidy-eligible income range if we assume the $2,000 penalty applies[7] (Exhibit 5).

In the single market, ESI (not HIE) is the most efficient choice across almost the entire range of subsidy-eligible incomes assuming the $3,000 employer penalty applies; if we assume the $2,000 penalty applies ESI is still the most efficient option across more than half of the subsidy-eligible income range[8] (Exhibit 6)

Thus eliminating (or re-designing, or restricting) ESI, and increasing cash wage sufficiently to offset any additional health premium costs employees might face for an HIE, is the most efficient means of compensating employees in subsidy-eligible income ranges (133% – 400% FPL) who want either family coverage, single + one coverage[9], or no coverage[10]. These groups are the majority, representing roughly 62% of employees, 53% of enrollees, and 73% of premiums paid (Exhibit 4, again). The exception is the market for single coverage, where ESI is more efficient than an HIE across more than half of the subsidy-eligible income range

Thus far we’ve focused narrowly on the question of which venue allows employers to most efficiently create a given level of compensation in the form of net wage plus health coverage. In a relatively loose labor market, we believe this simple construct is highly reflective of likely employer behavior. However in a tighter labor market, employers will compete to meet employees’ rising wage demands, and this strengthens the case for the shift of subsidy-eligible employees to HIEs. In Exhibit 2, we showed that two employers making offers of equivalent net wage and health coverage value spend different amounts on these offers: Employer A spent $59,173, and Employer B spent $56,507, which is $2,666 or 4.5 percent less. Under the assumption of a loose labor market, Employer B is likely to keep the savings; however under the assumption of a tight labor market, Employer B is likely to transfer this value to the employee in the form of a higher wage. So, by raising her spending to the same total amount as Employer A, Employer B can raise her cash offer to the employee by $3,755[11]. Thus at the same cost as Employer A, Employer B produces a much more attractive offer (to the employee, B’s offer is worth 5.1% more than A’s after-tax), and all else equal ‘wins’ the employee. The inference is clear: as the labor market tightens, A’s motive to switch to an HIE-based strategy becomes even stronger

To the Employee, Dollar Values Equal, is There Reason to Prefer Either HIE or ESI?

Returning to our base case where the offers from Employers A and B had equal values, our employee may nevertheless have reason to prefer B. On the HIE, she has more health insurance options to choose from; and, her ‘choice-set’ includes anything from purchasing less coverage to declining health insurance entirely, using some or all of the added wage offered by Employer B for other purposes. Whether she wants insurance or not, the HIE-based package (Employer B’s) offers her greater optionality, thus in pure economic terms B’s offer is ‘worth’ more, despite having an equivalent after-tax dollar value to the more restrictive offer from Employer A

All else equal[12], employees should prefer HIEs to ESI at identical after-tax dollar values, simply because HIEs offer greater optionality. And, the younger and healthier the employee, the stronger the preference for employers ‘offering’ HIE. The option to not purchase insurance, or to purchase less insurance than typically offered under ESI, frees added wage for other purposes. Plainly both options are more attractive to (generally) younger and (specifically) healthier employees. For right or wrong, all else equal, we suspect younger and healthier employees are more attractive to employers, thus we see this selection bias as a further motive for employers to shift to an HIE-based compensation scheme for subsidy-eligible employees

Mechanically, How Might Employers Move Toward HIEs?

As we interpret the law, employers wanting to move employees from ESI to HIEs have three options:

  1. Eliminate ESI entirely and pay non-subsidy eligible workers’ added incremental costs of acquiring coverage on an HIE. For employers with employees above the subsidy-eligible income ranges this is likely to be the least efficient option, for two reasons. Penalties for not offering ESI are $2,000 per employee, and the sum of employees used to calculate the penalty includes employees with incomes above the subsidy-eligible range. And, employees that are not subsidy-eligible would likely find these employers relatively unattractive, as non-subsidy-eligible employees would have neither HIE-based federal subsidies nor tax subsidies for purchasing health coverage. In theory, and in practice, these employers could use savings from the shift of subsidy-eligible workers to HIE to pay the added after-tax health premium burden of non-subsidy-eligible employees; but this quickly reduces total savings to the employer, mooting the economic motive for the shift
  1. Offer the same ESI package to all employees, but set employee premium shares to a level at which employees with household incomes < 400% FPL have premium costs that are > 9.8% of income. Employees with household incomes below the 400% FPL threshold would become subsidy eligible, and could purchase coverage efficiently (with wages grossed up by the employer for those facing higher HIE than ESI premiums) on the HIEs. This option retains ESI, and its tax advantages, for non-subsidy eligible employees. Employers would pay a $3,000 penalty for each employee that received subsidized coverage on an HIE; but unlike the $2,000 penalty for not offering coverage, this penalty is not multiplied across the entire employee base
  1. Create separate companies for subsidy-eligible and non-subsidy eligible employees. Offer ESI only to non-subsidy eligible employees, and raise cash wage to subsidy-eligible employees who face higher premium costs on the HIEs. This reduces the number of employees to whom the $2,000 penalty for not offering coverage applies and preserves ESI for non-subsidy-eligible employees. Ignoring transaction costs associated with creating separate companies, we believe this is the most efficient option for a majority of employers. Total penalties are lowest under this option, and employers can offer ESI to non-subsidy-eligible employees at more attractive rates of premium sharing

What Might Change the Relative Efficiency of HIE over ESI?

Employers’ willingness to supplant ESI with HIE must in part depend on the belief that the HIEs will: 1) come into being and continue to exist; and, 2) continue to offer a more efficient source of actuarial value for subsidy-eligible employees than ESI

The most apparent risk to the existence and relative efficiency of the HIEs is adverse selection, which we’ve covered extensively elsewhere[13]. We believe the HIEs are so susceptible to adverse selection as to face the risk of outright failure without structural changes to the HIEs. As long as the HIEs appear at risk of failure, employers should be less willing to transfer employees out of ESI, despite the potential savings. This is hardly good news for the reform effort, as having the HIEs fail as a result of adverse selection is a far worse outcome than paying more subsidies than budgeted because of a greater than anticipated ESI to HIE shift

Assuming the HIEs are made stable with respect to adverse selection, the relative efficiency of HIEs over ESI for subsidy-eligible employees can be narrowed or eliminated by lowering beneficiaries’ subsidies, raising employers’ penalties, and/or eliminating the tax-protected status of health premiums for non-subsidy-eligible employees in organizations whose subsidy-eligible employees are directed to HIEs. Lowering beneficiaries’ subsidies is a political non-starter; and, this option compounds already substantial adverse-selection pressures. Raising employer penalties is more politically feasible than lowering beneficiaries’ subsidies, but is nevertheless still extremely challenging. Bear in mind that the proposed penalties were lowered as a result of political pressure placed on a very popular Democratic president with bicameral legislative majorities; and, at the time the penalties were lowered, the Democratic advantage still included a 60-seat Senate majority. This level of Democratic legislative power is comparatively rare and has since weakened; in contrast the lobbying power of (particularly small) employers remains strong. And, either lowering beneficiaries’ subsidies or raising employers’ penalties requires ‘re-opening’ PPACA[14], which in and of itself is an extreme risk, as it’s practically impossible to open the law to modification and make only narrow revisions. The third option, eliminating or reducing the tax-protected status of health premiums paid on behalf of non-subsidy-eligible employees in firms where subsidy-eligible employees are directed to the HIEs, is the most likely. This could be done either through legislation modifying the existing tax code – also politically difficult, as it’s difficult to control the boundaries of a legislative effort to modify tax code – or it may be possible for IRS to eliminate or reduce these tax advantages via rule-making, without the need for legislative changes. To be clear, we haven’t attempted to determine whether IRS can in fact make effective modifications without new legislation; nevertheless, either by new legislation or new rule-making, we view changes to the tax advantaged status of health premiums as the most effective and accessible option for reducing the relative efficiency of HIEs over ESI for subsidy-eligible employees

Investment Relevance

A large-scale ESI to HIE shift – which we plainly anticipate — implies lower overall levels of coverage (as former ESI recipients choose less generous or no coverage on the HIEs) than would be expected if ESI were ‘stable’. 2014+ expectations for system-wide volume gains may not be met; and, 2014 may bring less relief than expected to hospitals’ uncompensated care burden[15]. Federal budgets for HIE subsidies are too low in the event of an ESI to HIE shift[16]; either the budget cap on subsidies will have to be raised, the relative efficiency (to employers) of HIE over ESI will have to be reduced, or these added subsidies will meaningfully compound health cost-related fiscal pressures

We favor HMOs on the belief that near-term increases in MLRs are unlikely[17]; though we recognize that an ESI to HIE shift in 2014 may mean fewer contracts and lower average contract sizes than expected, as well as operating margin pressures as enrollee mix shifts from large groups under ESI to individuals on the HIEs

  1. We’ve established that the two offers are equal in terms of net wage and actuarial value; however we acknowledge that employees may have other reasons to prefer one source of health coverage over another, and this is explored later in this note
  2. This figure is the result of adjusting the 2010 Milliman Medical Index (MMI) figure of $18,074 for a family of four down proportionally to the average US household size of 3.19, and bringing the result forward at an assumed annual health cost inflation rate of 7.5%
  3. 0.82 is the current average AV for ESI coverage. ; on the HIE: ; under ESI:
  4. The 2011 level adjusted for family size and inflated forward at the rate of CPI growth assumed in CBO’s long-term model
  5. Kaiser/HRET 2010 Employer Health Benefits Survey measures a 30% employee share of total premiums; the 2010 Milliman Medical Index measures a 29% share
  6. Employers that do not offer coverage face a penalty of $2,000 times the total number of firm employees (minus 50) if a single employee receives subsidized coverage on an HIE. Employers that offer coverage that does not meet standards for affordability (employee share of premiums < 9.8% of household income) or generosity (≥ 0.60 AV) face a penalty of $3,000 for each employee that receives subsidized coverage on an HIE (rather than all employees in the firm)
  7. We assume an employee premium share of 25% for ESI in the single + one market, vs. 30% in the family coverage market. We calculate the FPL range on the assumption of a 2-person household
  8. We assume an employee premium share of 19% for ESI in the single market. We calculate the FPL range on the assumption of a 1-person household
  9. In the preceding text and exhibits we show HIE to be the most efficient option for single + one coverage only up to 350% FPL. However, if we assume employers configure in such a way as to lower the applicable penalty to $2,000, that employees choose less generous coverage on the HIEs, and/or that plans with MLRs above the 80% level assumed are available on the HIEs, then it is reasonable to believe that all or nearly all single + one policies can be purchased more efficiently on the HIEs
  10. If I don’t want coverage, then plainly I prefer the employer that offers the higher salary, which in our example is the one that does not offer ESI
  11. , where tc = 0.29
  12. Ignoring in particular the risk that the HIEs are unstable, and/or any added search costs associated with purchasing coverage on the HIE instead of thru ESI
  13. “Reform Prospects, Declining Risks to Insurers, and Adverse Selection on the Exchanges” Sector & Sovereign Research LLC, December 16, 2009; and “Post-2014 Reform-Related Volume Gains are Modest” Sector & Sovereign Research LLC, March 2, 2011
  14. Patient Protection and Affordable Care Act
  15. “Post-2014 Reform-Related Volume Gains are Modest” Sector & Sovereign Research LLC, March 2, 2011
  16. PPACA includes a failsafe maximum percentage of GDP spent on HIE subsidies of 0.504
  17. “What Next for the MLR Cycle?” Sector & Sovereign Research LLC, June 10, 2011
Print Friendly, PDF & Email