San Franciscofs ePay Or Playf Employer Mandate Expanded Private Coverage By Local Firms And A Public Care Program

¹Carrie H. Colla (carrie.h.colla{at}dartmouth.edu) is an assistant professor at the Dartmouth Institute for Health Policy and Clinical Practice, in Lebanon, New Hampshire.
²William H. Dow is the Henry J. Kaiser Professor of Health Economics and head of the Division of Health Policy and Management at the School of Public Health, University of California, Berkeley.
³Arindrajit Dube is an assistant professor in the Department of Economics at the University of Massachusetts, in Amherst.

Abstract

In 2008 San Francisco implemented a pay-or-play employer mandate that required firms operating in the city to provide health insurance coverage for employees or contribute to the cityfs gpublic optionh health access program, Healthy San Francisco. Using data from our Bay Area Employer Health Benefits Survey, we found that in the first two years after implementation, more employers offered insurance and provided employee health benefit coverage relative to employers outside San Francisco not subject to the mandate. Sixty-seven percent reported in 2009 that they had expanded benefits since 2007. Although 22percentt of firms responding to the survey reported contributing to Healthy San Francisco for some employees, we observed no crowd-out of private insurance. Premium changes between 2007 and 2009 were similar in San Francisco and surrounding areas, but more of the burden of premium contributions in San Francisco shifted from workers to employers. Overall, 64percent of firms respondingg to the survey supported the employer mandate. San Franciscofs experience indicates that such a mandate is feasible, increases access, and is acceptable to many employers, which bodes well for the national employer mandate that will take effect under the Affordable Care Act in 2014.

San Francisco took the first steps toward achieving universal health care coverage in the city by implementing the Health Care Security Ordinance in January2008. This ordinance included a pay-or-play employer health spending mandate, requiringg all private-sector firms employing twenty or more workers to contribute a minimum amount to health benefits per worker-hour. Employers can fulfill the requirement by providing private insurance, paying into a reimbursement account, or contributing to Healthy San Francisco.

Healthy San Francisco is a low-cost health access plan created by the ordinance and administered by the cityfs Department of Public Health. The plan covers primary care, specialty care, urgent and emergency care, laboratory tests, inpatient hospitalization, radiology, and pharmaceuticals.

The plan strengthened safety-net services and provided a gpublic optionh as an alternative to private health insurance for employers to choose to satisfy the mandate. Healthy San Francisco is not formally insurance, but it offers benefits tantamount to a generous Medicaid policy, with the caveat that it is restricted to a limited network of providers located within San Francisco, mainly composed of community health centers.¹ The plan is not licensed as insurance and is not regulated by state insurance agencies.

Research on Healthy San Francisco has found high satisfaction rates among enrollees and providers as well as improved continuity of care for patients. The plan has been cited as a potential model for improving health access in other localities. However, the employer mandate portion of the ordinance has been criticized by industry groups, such as the San Francisconci Chamber of Commerce and the Golden Gate Restaurant Association.

The mandate was also challenged in an unsuccessful lawsuit that the Supreme Court declined to review. The plaintiffs claimed that the mandate was overly burdensome, but it is unknown whether these concerns are shared broadly by the employer community. Previous work has shown that although 76percent of employers had to make a change to health benefitseither by expandingin eligibility or coverage to more workers or by increasing the generosity of benefitsin order to comply, there was early employere support for the mandate while it was being phased in.⁵ This article extends that work to assess a broader range of employer responses two years into implementation.

The Affordable Care Act calls for an employer mandate of a similar magnitude to be implemented nationwide in 2014. During 2008 in San Francisco, the mandated contribution was $1.76 per worker-hour, or $3,632 annually, for each full-time worker at firms with a hundred or more employees. Firms with twenty to ninety-nine employees were required to contribute at least $1.17 per worker-hour. For-profit firms with fewer than twenty employees and nonprofit firms with fewer than fifty employees were exempt.

These minimum contribution amounts were slightly less than the 2008 national average employer health insurance contribution per employee, $3,983.⁶ The Affordable Care Act requires employers with at least fifty workers to pay a penalty of $2,000$3,000 per uncovered workere for not providing coverage.⁷

There are many questions as to how an employer mandate will perform in practice, and little evidence to inform projections. The Congressional Budget Office estimated that about one million people will lose employer-sponsored coverage under the Affordable Care Act and will instead receive federally subsidized coverage through health insurance exchanges,⁸ while the Urban Institute and RAND reported that the law will have little effect on overall employer coverage.⁹^,¹⁰ However, a report from the consulting firm McKinsey indicated that up to a third of employers would definitely or probably drop coverage.¹¹ The conflicting estimates reflect the fact that, much like San Franciscofs Health Care Security Ordinance, the Affordable Care Act will change many variables in the way care is financed and delivered.

Local employer mandates are currently in effect in two states, Massachusetts and Hawaii, yet these mandates are very different in character and magnitude from the national employer mandate. In Massachusetts the penalty for not providing coverage is only $295 annually per worker, and the employer mandate is coupled with an individual mandate. Hawaii simply requires employers to provide coverage. Research on reforms in both states found increases in the proportion of employers offering coverage, and in Massachusetts there was no evidence of dropped coverage or restricted eligibility.

The San Francisco mandate has a minimum contribution more similar in magnitude to the Affordable Care Actfs requirement. Thus, developments in San Francisco provide a unique opportunity to generate new evidence on the potential impact of the Affordable Care Act employer mandate.

Specifically, we investigated the following questions: When given a pay-or-play choice, do employers comply with a mandate by expanding private insurance, by offering a mini-medical plan (a limited benefit plan with a very low maximum annual payout, often $10,000 or less), or by dropping private insurance? Do employer mandates increase the number of employees with health benefits? How does the mandate affect employer and employee health insurance premiums? How supportive are employers of a pay-or-play mandate?

We addressed these questions by studying trends in San Francisco two years after implementation, compared to trends among firms in surrounding counties that were not subject to the law.

Study Data And Methods

Data

The primary data source for our analysis was the Bay Area Employer Health Benefits Survey, a telephone survey that was conducted in 2008 and 2009 by the authors at the University of California, Berkeley. The survey was conducted by interviewing representatives at firms who were responsible for company benefit decisions in San Francisco and surrounding counties. This article analyzes for-profit firms with 20200 employees, which is the group we expected to be most affected by the mandate.¹⁷

In the Bay Area Employer Health Benefits Survey, we interviewed 349 and 329 San Francisco firms in 2008 and 2009, respectively. We also interviewed 290 and 229 comparison firms in 2008 and 2009, respectively, from four surrounding counties not subject to the mandate.

The sample frame included for-profit firms with twenty or more employees as listed in the Dun and Bradstreet Database and provided by Survey Sampling, a private firm that provides sampling and data collection solutions for survey research. The sample included the population of San Francisco firms and a random sample of firms drawn from surrounding counties.

The response rate for the survey was 20percent, creating the potential for nonresponse bias. However, we tested for suchh bias using data from the City of San Francisco; the Dun and Bradstreet Database; the California HealthCare Foundationfs California Employer Health Benefits Survey;¹⁸ and responses to a question on insurance offering asked of all firms, including survey nonresponders. We found that nonresponse bias was minimal: The firms that responded to the 2008 Bay Area Employer Health Benefits Survey were similar in industry, firm size, and geographic distribution to those that refused to participate.

The 2008 survey collected data on 2007 benefit offerings. The 2009 survey, conducted in late 2009, collected current benefit information.

We chose Alameda, Contra Costa, Marin, and Santa Clara Counties for the survey because these counties are geographically adjacent and contain metropolitan areas (Oakland and San Jose), and the firms there are similar in size, worker population, and industry composition to those in our San Francisco sample (see Appendix1).¹⁹ The 2007 pre-mandate offer rate in San Francisco and surrounding counties was similar. According to the American Communityy Survey,²⁰ these counties were also similar to San Francisco in household size and median household income during 200610.0

Furthermore, an additional 639 firms in San Francisco and 371 firms in comparison counties refused to participate in the survey but answered the following question: gDoes your firm currently offer health benefits to at least some of your employees?h Ninety-three percent of the firms that completed the survey offered health insurance, compared to 91percent of the firmss that refused to participate. Thus, although the low response rate was concerning, evidence suggested that systematic nonresponse differences in health insurance offerings were small.

When we used our weighting scheme and limited the sample to firms with more than twenty employees in the Bay Area, our insurance offer rate was consistent with the 95percent offer rate in the 2007 California Employer Health Benefits Survey,¹⁸ further corroborating the representativeness of our sample..

Finally, we compared our estimates of the percentage of firms contributing to Healthy San Francisco with administrative reports from the City of San Francisco. In 2008, 980 employers had elected to pay into Healthy San Francisco, the public option, out of what we estimatedbased on County Business Pattern data²¹to be approximately 4,500 total covered employers. This yielded administrative estimates of public option demand that equaledal our survey-based estimate of 22percent.²²

The Bay Area Employer Health Benefits Survey asked questions on insurance offering, plan types, premiums, changes to health and other types of benefits, and demographic and socioeconomic characteristics of firm employees.⁵ The survey also included questions for San Francisco firms only on employer burden and support for the ordinance. For example, it asked whether firms found the record-keeping and other administrative requirements of the health spending mandate to be very burdensome, somewhat burdensome, not too burdensome, or not burdensome at all.

Methods

To investigate the extent to which health benefit changes were probably caused by the ordinance rather than reflecting broader trends, we analyzed health benefit changes in San Francisco firms compared to changes in similar firms in surrounding Bay Area counties not subject to the mandate. Our analysis began by evaluating health benefit offer patterns in San Francisco and the comparison sample of firms in 2007 (before the ordinance was implemented) and 2009, including whether the firm offered coverage, the proportion of employees eligible for coverage, and the proportion of employees taking up coverage among those who were eligible.

Using this information, we calculated the mean proportion of employees who had health benefit coverage from their employer, including Healthy San Francisco benefits.²³ We regression-adjusted these figures to determine the difference in offering associated with the ordinance, controlling for industry and firm size. We calculated changes in single coverage premiums between 2007 and 2009 using the Bay Area Employer Health Benefits Survey as well as statewide trends from the California Employer Health Benefits Survey.¹⁸^,²⁴

We calculated changes made to health benefit packages, including types of benefits, and created summary indicators for any expansions and contractions in coverage. For all analyses, we both tested for changes in outcomes within San Francisco and tested whether these changes were different from corresponding changes outside San Francisco.

We further explored the extent of these changes among two subsets of firms with low pre-mandate health spending. The first subset was what we called a high-impact group, composed of firms that had to increase health benefit spending by 50centss or more per worker-hour on average according to 2007 spending patterns. The second subset consisted of firms that did not offer any health benefits in 2007. For each of these groups, we created parallel comparison samples of firms outside of San Francisco based on 2007 benefits, and again we tested for differences in pre-post trends in San Francisco versus comparison firms from adjacent counties.

All results reported were weighted to reflect the population distribution of firm characteristics in the surveyed counties. Appendix1¹⁹ shows descriptive characteristics of each subsample used for analysis..

Study Results

Absent wage reductions, these differences indicate that an outcome of the employer mandate was to shift the health coverage costs burden toward the employer. Economic theory and some evidence suggests that workers will ultimately bear that burden through smaller wage increases, although preliminary findings from San Francisco do not show such changes.²⁶

We found no evidence of crowd-out of private health plans by Healthy San Francisco. In both San Francisco and surrounding counties, less than 1percent of firms in the 2009 survey indicated that they had stopped offering health insurance sincee 2007. In the high-impact group, 1.4percent of San Francisco firms stopped offering coverage to their workers, a figure nott statistically different from zero; the 95percent confidence interval rules out values larger than 4percent.t.

Consistent with the changes in health benefits observed in Exhibit3, 33percent of San Francisco firms and 55percent of high-impact San Francisco firms responded that they had to increasease spending in response to the Health Care Security Ordinance (Exhibit4). Firms in the high-impact group were significantly more likely to find it difficult to comply with the ordinance and too find the law burdensome, compared to firms overall.

San Francisco employers were well informed about the ordinance, and support for it was strong overall.²⁷ Although still substantial, support was significantly lower among the high-impact firms.

Discussion

In the two years following the implementation of a pay-or-play mandate in San Francisco, we found that the mandate was associated with increases in the proportion of firms offering coverage and the proportion of workers with health benefits. We also found no evidence of crowd-out of private employer-sponsored insurance by Healthy San Francisco, which met high demand. And we found that coverage increases were concentrated in firms that had offered few or no health benefits prior to the mandate.

Although premium changes between 2007 and 2009 were similar in San Francisco and surrounding areas, the composition of the premium contributions in San Francisco shifted from workers to employers. Despite being burdened with compliance, a majority of San Francisco employers were supportive of the policy.²⁸

There have been concerns of excess cost growth in Massachusetts under health reform.²⁹ Similarly, an expansion in the generosity of health benefits in San Francisco could have led to an increase in total premiums, but we found no evidence of this effect. Instead, employers appear to have responded by taking on a greater share of the premium increases, thus slowing growth in employee contributions.

As noted above, it is important to understand that higher employer contributions to health benefits may often be funded out of lower wage growth in the long run. Although the evidence from the San Francisco mandate in the first two years after the ordinancefs implementation does not indicate job loss or reduction in wage growth,²⁶ this phenomenon has been found in other settings.³⁰^,³¹ Furthermore, research on the San Francisco mandate has found that in restaurants, an industry greatly affected by the mandate, firms have raised prices to consumers.²⁶

San Francisco has achieved high levels of employer health benefit coverage, indicating that the city has met one of the overarching goals of the ordinance. We found that compliance was high, and we estimated that only 5percent of workers were still withoutt coverage.

Our estimates suggest that absent the ordinance, another 10percent of workers would be without employer-paid health benefits.. The high compliance rate and successful expansion of coverage has positive implications for the implementation of the national employer mandate that will take effect under the Affordable Care Act in 2014.

Some observers have expressed concerns that employers will drop private insurance in response to pay-or-play mandates.¹¹ We found no evidence of private insurance crowd-out in San Francisco associated with the ordinance, although we might expect less crowd-out in this setting because Healthy San Francisco is not formal insurance. Healthy San Francisco and private health insurance are not likely to be perceived as substitutes because of the limited provider network, the stigma associated with public programs, and the requirement that enrollees must be uninsured for ninety days to be eligible.

More surprising is the fact that we found employers initiating insurance offerings after the mandate. Models simulating the likely impact of the Affordable Care Act on health benefit coverage have indicated little or negative changes in employer offering of coverage. Perhaps these models should be revisited in light of increases in the number of employers offering coverage in San Franciscois and Massachusetts.³³

These results provide evidence that an employer pay-or-play mandate can increase health benefit coverage for workers in for-profit firms. However, San Franciscofs mandate was not designed to improve coverage levels for small business employees, dependents, and others out of the labor force. There were also many exclusions to the employer coverage mandate: Based on our calculations from the Dun and Bradstreet Database, about 25percent of all San Francisco workers at for-profit firms were employed at firmss that were exempt because of size.

San Francisco has attempted to address these groups through the creation of Healthy San Francisco. Future research will be needed to assess the impact of the ordinance on health benefit coverage and access in the overall population.

Although San Francisco has many unique characteristics that make conclusions difficult to generalize, such as high initial rates of coverage and high levels of income, it is the closest evidence we have to an employer mandate similar in magnitude to that in the Affordable Care Act. In addition, our conclusions have the usual limitations of self-reported survey data and are not able to fully reflect differences in employee characteristics or richness of benefit packages.

San Franciscofs Health Care Security Ordinance has made strides toward its goal of increasing coverage for and access to health care, partially funded with expanded financing contributions by employers. Although most firms found the ordinance administratively burdensome, and a substantial subset of them found it difficult to comply with, it has achieved support from the majority of San Francisco businesses. This finding of broad support is perhaps surprising, given the vocal opposition by employer groups such as the Golden Gate Restaurant Association, which sued to block implementation of the policy.

It will be important to monitor closely the implementation of the federal employer mandate to assess how the Affordable Care Act is received, not only by industry lobby groups but by the broader national business community. Overall, San Franciscofs experience indicates that a strong pay-or-play mandate can be effective, feasible, and acceptable to employers.

Health Affairs, January 2013