This Article Abstract Figures Only Reprint (PDF) Erratum (v20,p263) Submit a response to this article Alert me when this article is cited Alert me when eLetters 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 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 HighWire Citing Articles via Google Scholar Google Scholar Articles by Gabel, J. R. Articles by Schoen, C. Search for Related Content PubMed PubMed Citation Articles by Gabel, J. R. Articles by Schoen, C. Related Collections Insurance - Employer-Based System

DataWatch

Embraceable You: How Employers Influence Health Plan Enrollment

Abstract

 
Based on data from a 1999 national survey of 1,939 randomly selected employers, this paper examines the policies that affect the percentage of workers eligible for and enrolled in a firm’s health plan. In 1994, 14 percent of employees worked for a firm offering cash-back payments, but fewer than 1 percent worked for a firm with income-related premiums or deductibles. The strongest determinants of eligibility rates are the waiting time for new employees before they are deemed eligible, and eligibility standards for part-time workers. The primary determinants of the take-up rate are lowest monthly employee contribution for single coverage, and the percentage of the workforce earning less than $20,000 per year.

Researchers have analyzed data from both employer and household surveys and reached the same conclusion about the erosion of job-based health insurance. Today a lower percentage of workers receive coverage from their employer than was the case ten years ago, not because fewer employers offer coverage, but because fewer employees accept coverage that is offered. Data from the National Medical and Expenditure Survey (NMES) of 1987 and the Medical Expenditure Panel Survey (MEPS) of 1996 show that the percentage of workers offered coverage from their employer grew from 72 percent in 1987 to 75 percent in 1996. However, the percentage of workers taking up coverage declined from 88 percent to 80 percent.⁴ Hence, to better understand why a robust economy, tight labor markets, and low health care inflation have not greatly lowered the percentage of workers without coverage, one needs to understand how employers encourage and discourage enrollment in their health plans.

This paper shows how eligibility and take-up rates vary by employer policies and characteristics, such as firm size, employee income, industry, and region. Previous studies have analyzed the factors that predict which workers are likely to have job-based coverage.⁵ In contrast, we are unaware of prior analyses that have documented employers’ specific policies that influence plan enrollment and then, using multivariate techniques, analyzed how these policies affect eligibility and take-up rates.

Data Sources And Methods

Top Data Sources And Methods Descriptive Results Multivariate Findings Shoring Up Employer-Based... NOTES

 
Study data are from a special supplement to the Henry J. Kaiser Family Foundation/Health Research and Educational Trust (KFF/HRET) 1999 Employer Health Benefit Survey. Working with the Commonwealth Fund, the HRET developed a supplement that addressed policies that might encourage or discourage workers to enroll in their employers’ health plan(s). Among the issues addressed were (1) company philosophy, (2) eligibility criteria, (3) cash-back payments for declining coverage, (4) underwriting policies for employees who miss open enrollment, (5) income-based premiums and deductibles, and (6) policies for hourly workers.

The KFF/HRET survey (conducted by KPMG Peat Marwick from 1991 to 1998) is a telephone survey with employee benefit managers for a random stratified sample of private and public firms with three or more workers. We drew the sample from a listing of firms by Dun and Bradstreet and stratified the sample by firm size (number of workers) and industry. National Research conducted the interviews in the spring of 1999, asking as many as 400 questions about the companies’ health plans. We completed interviews with 1,939 employers, with an overall response rate of 60 percent. Of these employers, 1,825 offered health benefits. When using the full sample, the statistical error is 3 percent, meaning we can be 95 percent confident that the sample results are within three percentage points of the true mean. Because KFF/HRET randomly selects firms, statistical weights can be used to extrapolate to national (as well as regional, industry, and firm-size) averages. These weights allow KFF/HRET to present findings based on the number of workers covered by a health plan, the number of total workers, and the number of firms.

Because of the complex sample design, we have used the statistical program SUDAAN to calculate standard errors and conduct statistical tests. In the descriptive tables, we test whether the difference among types of firms is statistically different from the overall average. Two-way descriptive analyses, however, do not control for possible intervening variables. To isolate the effect of the above factors on coverage and take-up rates, we conducted a multivariate analysis.⁶ The dependent variables in these analyses are the eligibility rate and the take-up rate.

Descriptive Results

Top Data Sources And Methods Descriptive Results Multivariate Findings Shoring Up Employer-Based... NOTES

 
Coverage, eligibility, and take-up rates. Not all firms offer health benefits. Among small employers (fewer than 200 workers) only 60 percent of firms offered health benefits in 1999. Virtually all large firms (99 percent) offered health benefits to their workers.⁷

Among those firms that offered health benefits, two-thirds of workers were enrolled in their employer’s health plans (Exhibit 1). Firms having low percentages of low-earning workers had significantly higher coverage rates than did firms with relatively more low-earning workers.⁸ While 90 percent of state and local government workers were enrolled in their health plan, employers covered only 40 percent of workers in the retail sector.⁹

Coverage strategies of employers. In interviews with employee benefits managers, we asked whether they encouraged their workers to enroll in the firm’s plan, whether they encouraged their workers to enroll in the spouse’s plan, or if they were indifferent as to whether their workforce received company coverage. Few indicated that they encouraged workers to accept coverage in their spouse’s plan. Only 4 percent of covered employees worked in firms that encouraged enrollment in their spouse’s health plan (among firms offering health benefits). Further analysis revealed a contradiction in eligibility policies and stated philosophies.¹⁰ Consequently, we omitted this question from the multivariate analysis.

Coverage for part-time and temporary workers. Part-time workers constitute 22 percent of U.S. workers, while temporary workers account for 2 percent.¹¹ Hence, the standards an employer sets for eligibility of these workers will profoundly influence the percentage of the workforce eligible for and enrolled in the firm’s health plan. In fact, 41 percent of part-time workers were eligible for benefits in 1999, but only 3 percent of temporary workers were (Exhibit 2).¹² The mean number of weekly hours necessary to qualify for health benefits for part-time workers is twenty-two; for full-time workers, thirty-three.¹³ Firms with fewer than 1,000 workers are more stringent than larger firms are in covering part-time workers. Part-time workers in the health care, finance, and state and local government sectors are most likely to be eligible for health benefits.

Policies affecting eligibility and take-up rates. Among firms offering health coverage, nearly 14 percent of employees worked for firms that offer cash-back payments when an employee declines coverage (Exhibit 3). Firms employing 5,000 or more workers were nearly twice as likely to offer cash-back payments as small firms (3–199 workers) were. Firms with a high concentration of low-earning workers were one-third as likely to offer cash-back payments as were firms in which 10 percent or less of the workforce were low-earning workers. About 30 percent of workers in the finance sector worked for a firm providing cash-back payments, whereas fewer than 3 percent of employees in the retail sector did.

Some employers subject employees to medical underwriting when they decline coverage during open enrollment. These underwriting practices are designed to prevent adverse selection, where employees pay no premiums until they become sick and enroll in the company plan. A harsher policy is to refuse enrollment when the employee declines coverage during open enrollment. Overall, one of three employees worked for a firm that medically underwrites such employees.

Plan design. As with other goods and services, workers are more likely to take up coverage when the price is low. The average monthly employee contribution for single coverage for the plan with the lowest monthly employee contribution was $29 per month in 1999 (Exhibit 4).¹⁴ This figure was highest in the retail sector ($46) and lowest in the state/local government sector ($7).

The richer the benefit package, the more attractive the health plan is to employees, other factors held constant. Exhibit 4 shows the average number of benefits offered to employees among a package of twelve benefits.¹⁶ Employees in small firms, firms in the Midwest, and firms in the mining/construction/wholesale and health care industries had the fewest benefits covered.

Preexisting condition clauses diminish the value of health benefits, leaving new employees more vulnerable to the cost of illness. Insurers view these clauses as a way to reduce adverse selection. In 1999, 43 percent of employees worked in firms with preexisting condition clauses in all non–health maintenance organization (HMO) plans. Small firms were most likely to use such clauses.

Multivariate Findings

Top Data Sources And Methods Descriptive Results Multivariate Findings Shoring Up Employer-Based... NOTES

 
Our aim in the multivariate analyses is to measure the effect of employers’ basic design and coverage decisions on workers’ eligibility and take-up rates. For the eligibility rate regression, key independent variables are (1) waiting time for new employees before they qualify for coverage; (2) eligibility standards for part-time workers; and (3) whether temporary employees are eligible for benefits. Others are size of firm, region, industry sector, and percentage of the workforce who are low-earning workers. These variables are specified as a series of binary variables.

In explaining the take-up rate, the following are key independent variables: (1) Does the firm offer cash-back payments if an employee declines coverage? (2) Does the firm vary the contribution requirement according to the income of the worker? (3) Does the firm offer more than one plan? (4) Do all plans have preexisting condition exclusions? (5) Is there dependent coverage? (6) What is the lowest monthly contribution level for single coverage for employees? (7) Is the worker subject to medical underwriting if the worker enrolls outside of the open season? Other independent variables in the take up regression included firm size, region, and industry sector.¹⁷

The monthly contribution level represents the cost of health insurance to the worker. The regression for the take-up rate is structured so that we can observe whether firms with many, some, or few low-earning workers respond differently to required contribution levels. Our hypothesis was that high contribution requirements discourage low-earning workers considerably more than they do middle-and high-earning workers from taking up coverage.

Our regressions were able to explain substantial variation in eligibility and take-up rates.¹⁸ The subsequent discussion first highlights findings about factors that explain why some firms have high eligibility rates while other firms have low ones, and then repeats the analysis for take-up rates. Unless explicitly noted, we have restricted the discussion to variables significant at the 95 percent or higher confidence level.

Eligibility rate. The most significant factors determining the eligibility rate are (1) the waiting period before an employee is eligible for benefits; and (2) the eligibility criteria for part-time workers. Exhibit 5 displays the effect of these factors on the percentage of the workforce eligible for enrollment. These estimates apply to a standard set of workers possessing characteristics of the average firm in the study sample. Thus, the average firm has an eligibility rate of 77 percent. For this standard firm, if there is no waiting period before eligibility, the expected eligibility rate is 82 percent.

View larger version (12K): [in this window] [in a new window] EXHIBIT 5 Expected Eligibility Rates For A Standardized Firm: Effect Of Waiting Period And Required Weekly Hours On Eligibility

Take-up rate. Multivariate analysis indicates that most policies and benefit design characteristics had statistically insignificant effects on the take-up rate. Among these policies and characteristics are (1) use of preexisting condition clauses; (2) use of income-related premium contributions and deductibles; (3) offering cash-back payments to employees who decline coverage; and (4) use of medical underwriting when an employee misses the open enrollment.

Two factors significantly affected the percentage of workers opting to enroll in the company health plan. First was the minimum monthly employee contribution for single coverage. Second was the percentage of the workforce earning less than $20,000 per year, interacting with this minimum monthly contribution figure for single coverage. This variable measures how workers with low earnings respond differently to price than higher-earning workers do.

Exhibit 6 shows the take-up rate for a firm with standardized characteristics except for the percentage of workers earning less than $20,000 per year. If the single monthly contribution for the lowest-cost plan is $50, then the predicted take-up rate for a firm with many low-income workers (35 percent or more of the workforce) is 72 percent. A firm with few low-income workers (fewer than 10 percent of the workforce) and facing a $50 monthly contribution will have a take-up rate of 84 percent. If employees must contribute $100, the take-up rate for the firm with many low-income workers is 55 percent, while the take-up rate for a firm with few low-income workers is 79 percent. When no monthly contribution is required, the take-up rate for both firms is 89 percent. By implication, the above results indicate that low-earning workers are the most sensitive to the price of health insurance, represented by the monthly contribution for premiums.

View larger version (14K): [in this window] [in a new window] EXHIBIT 6 Expected Take-Up Rates, By Worker Contribution Levels For Firms With Few, Some, And Many Low-Income Workers

Top Data Sources And Methods Descriptive Results Multivariate Findings Shoring Up Employer-Based... NOTES

What public policies might increase eligibility rates? Some might argue that if the federal government can mandate coverage for former employees under the Consolidated Omnibus Budget Reconciliation Act (COBRA), then Congress could improve the portability of group coverage by limiting the waiting period before new employees become eligible for benefits.²⁰ The counterargument is that the mandate would raise the cost of providing health benefits. If the workforce is paid according to its productivity, some employers may react by no longer offering health benefits, while others might reduce the richness of their coverage, increase monthly contributions, or reduce wages. Only a real-world experiment through the laboratory of state regulation can determine the actual results of such a mandate.

Employers can seemingly influence the take-up rate through a host of policies and benefit design decisions. Yet our multivariate analysis suggests that the principal decision employers make that affects the take-up rate is the price they set for participating in the company plan—measured as the lowest monthly contribution rate for single coverage among plans offered. Workers in firms with many low-earning workers are far more price-sensitive than are workers in firms with few low-earning workers, as Exhibit 6 shows.

Therefore, substantial monthly contributions for low-earning workers may be the Achilles’ heel of market-oriented strategies to control health care costs. An impressive body of research concludes that employees will readily switch from one plan to another according to differences in what they must contribute toward monthly premiums.²¹ One study found that a mere $10 increase in the relative monthly employee contribution for a plan increased the probability of an employee’s switching plans from 5 percent to 25 percent.²² Other research shows that firms that offer a choice of plans and that do not subsidize higher-cost plans experience lower inflation rates in their premiums.²³ However, the dark side of price-sensitivity on the part of employees is that low-earning workers will not take up benefits when they face large monthly premium contributions.²⁴ Hence, to ensure widespread participation among such workers, employers must make certain that at least one plan requires little or no employee contribution. An ideal contribution formula would set the employer’s contribution at the price of the lowest-cost plan and contribute that fixed amount to each plan offered, regardless of plan cost. Only about one-fifth of employees work for employers that use such a formula.²⁵ Employers, however, do not build a benefits strategy around gaining the participation of low-earning workers. On the contrary, it is the highly skilled and highly paid workers that most firms strive to retain.

After five years of low inflation, the cost of health insurance is now rising at the most rapid rate since 1993. The strongest economy and lowest unemployment rate is still shielding workers from the rising cost of health insurance.²⁶ However, the slowing of the economy in early 2001 suggests that employee contributions are likely to rise dramatically. The good news is that rising employee contributions will encourage employees to move to lower-cost plans. The bad news is that higher employee contributions will lead to lower take-up and coverage rates, particularly for low-earning workers. This fundamental relationship is one more major impediment for the employer-based system to overcome if it is to truly serve all working families.

Editor's Notes

 
Jon Gabel, Jeremy Pickreign, and Heidi Whitmore are with the Health Research and Educational Trust (HRET) in Washington D.C. Cathy Schoen is with the Commonwealth Fund in New York City.

The authors gratefully thank the Commonwealth Fund for its financial support.

NOTES

Top Data Sources And Methods Descriptive Results Multivariate Findings Shoring Up Employer-Based... NOTES

 

1. U.S. Bureau of Labor Statistics, <146.142.4.24/cgi-bin/surveymost> (12 January 2000).
2. P. Fronstin, Sources of Health Insurance Coverage and Characteristics of the Uninsured: Analysis of the March 1999 Current Population Survey, EBRI Issue Brief no. 217 (Washington: Employee Benefit Research Institute, January 2000).
3. U.S. Census Bureau, "Health Insurance Coverage: 1999," 10 October 2000, <www.census.gov/hhes/hlthins/hlthin99/hi99ta.html > (12 April 2001).
4. P.F. Cooper and B.S. Schone, "More Offers, Fewer Takers for Employment-Based Health Insurance: 1987 and 1996," Health Affairs (Nov/Dec 1997): 142–149.
5. For example, see S.H. Long and M.S. Marquis, "Gaps in Employer Coverage: Lack of Supply or Lack of Demand?" Health Affairs (Supplement 1993): 282–293; and K.E. Thorpe and C.S. Florence, "Why Are Workers Uninsured? Employer-Sponsored Health Insurance in 1997," Health Affairs (Mar/Apr 1999): 213–218.
6. Because the eligibility rate may affect the take-up rate, we used Heckman’s Lambda to control for possible selection effects in this multivariate analysis. See Note 17.
7. See L. Leavitt et al., Employer Health Benefits 1999 (Menlo Park, Calif.: Kaiser Family Foundation, and Chicago: Health Research and Educational Trust, 1999), 30–32. The full sample for calculating the percentage of firms offering coverage is based on 1,939 firms that completed the questionnaire and 755 that answered one question, "Does your firm provide health benefits to your employees?"
8. Low-earning firms are defined as firms in which 35 percent or more of the workers earn less than $20,000 per year. High-earning firms are firms in which 20 percent or more of workers earn $75,000 or more a year.
9. The survey did not collect data on the percentage of the workforce belonging to unions, one potential influence on eligibility and take-up rates.
10. For example, firms that said they "encouraged" participation in their health plan on average required higher contributions for premiums than did firms that said they "discouraged" participation.
11. Bureau of Labor Statistics, Report on the American Workforce, Table 11, <www.bls.gov/opub/rtaw/pdf/table11.pdf> (12 April 2001).
12. Presumably, employers did not consider contract workers as temporary workers. The question was not clear in distinguishing the two.
13. The median for part-time workers is twenty, which indicates that a few firms have much higher requirements.
14. We use the lowest monthly contribution for single coverage because we believe that this is the statistic that potential "decliners" examine. Those who are sensitive to the contribution rate may accept single rather than family coverage.
15. T. Rice, K. Desmond, and N. Pour at, "Dark Clouds in Pleasantville: Trends in Job-Based Health Insurance, 1996–1998" (Presentation at the Association of Health Services Research annual meeting, Chicago, June 1999).
16. Benefits include adult periodic physical exams, well-baby care, all reversible contraceptives, prenatal care, mammography screening, abortion services, oral contraceptives, outpatient mental health care, inpatient mental health care, chiropractic care, prescription drugs, and infertility treatments.
17. We used the Heckman’s Lambda technique to control for potential selection effects. Our reasoning was that eligibility criteria may affect the take-up rate. Firms with more stringent eligibility standards may exclude employees who are less likely to participate in the health plan. Heckman’s Lambda is an attempt to enter the effect of eligibility requirements as a control variable in the take-up equation.
18. The multiple r squared for the eligibility equation was 0.33 and for the take-up equation was 0.27. The overall regressions were statistically significant at an alpha of .000. Details of the multivariate anlaysis are available from Jon Gabel, <jgabel{at}aha.org>.
19. Our explanation for this counterintuitive result is that firms that do not cover part-time workers hire fewer part-time workers.
20. Former workers pay 102 percent of the monthly cost of premiums under COBRA. However, numerous studies have found that COBRA enrollees incur claims expenses that are considerably greater than the premium contribution. Hence, employers are in fact contributing to the coverage of COBRA enrollees.
21. M. Morrisey, Price Sensitivity in Health Care: Implications for Health Care Policy (Washington: NFIB Foundation, 1992), 40–45; R. Feldman et al., "The Demand for Employment-Based Health Insurance Plans," Journal of Human Resources (Winter 1989): 115–142; and P. Short and A. Taylor, "Premiums, Benefits, and Employee Choice of Health Options," Journal of Health Economics (December 1989): 293–312.
22. T. Buchmueller and P. Feldstein, "The Effect of Price on Switching among Health Plans," Journal of Health Economics (April 1997): 231–247.
23. K.A. Hunt et al., "Paying More Twice: When Employers Subsidize Higher-Cost Health Plans," Health Affairs (Nov/Dec 1997): 150–156.
24. Critics of competition-based approaches to cost control identify risk selection as a major flaw with this strategy.
25. Leavitt et al., Employer Health Benefits 1999.
26. J. Gabel et al., "Job-Based Health Insurance in 2000: Premiums Rise Sharply while Coverage Grows," Health Affairs (Sep/Oct 2000): 144–151.

                      What's this?

This article has been cited by other articles:

				H. S. Luft Universal Health Care Coverage: A Potential Hybrid Solution JAMA, March 14, 2007; 297(10): 1115 - 1118. [Full Text] [PDF]