The challenge. In 2000, nearly 40 million Americans were uninsured. One of the most perplexing aspects of this problem is that nearly 78 percent of the uninsured were employed workers or their dependents. Who, exactly, were these working uninsured? What were the differences between insured and uninsured low-income workers? Did patterns vary from state-to-state? Were there differences between uninsured workers in larger firms and small firms? Researchers at the Urban Institute wanted answers to these questions to understand insurance market reforms that were designed to expand coverage for the uninsured.
Grantees' background. From 1993 to 1994, health economists Linda J. Blumberg, PhD, and Len M. Nichols, PhD, worked together in President Clinton's Office of Management and Budget (OMB). Blumberg, a health policy adviser with a PhD in economics from the University of Michigan, and Nichols, a senior health policy adviser with a PhD in economics from the University of Illinois, were part of a team working on health care reform that promised basic health care for all Americans—a centerpiece of Clinton's presidential campaign.
After the health care reform effort ended in defeat in the fall of 1994, Blumberg and Nichols left the OMB but continued to work together. Blumberg returned to the Urban Institute, a Washington-based nonpartisan economic and social policy research organization where she had been a research associate; Nichols joined her there as a senior research associate.
They continued to study the working uninsured, their dependents, and insurance reforms designed to address their needs. Said Blumberg: "After Clinton health care reform had collapsed, we were still interested in these issues, in finding a mechanism for expanding coverage for the uninsured." Nichols agreed. "We thought at the time that we would never have a national system, but we also realized people would still need health care and would buy health insurance one way or another," he said. "We wanted the market to work better, and decided that studying variations across states was a device we could use to analyze 'Does policy matter?'"
They saw this emphasis as a way to help inform policy-makers on how to expand coverage and to identify what interventions work best. Said Blumberg, "Identifying categories of uninsured workers within each state should be helpful to policy-makers at the state and federal levels."
The research. In 2000, Blumberg and Nichols received a two-year grant from the Robert Wood Johnson Foundation (RWJF) to carry out their research, "Studies of the Working Uninsured, Their Dependents and Insurance Reform on Their Behalf." The project was funded by the Foundation's Changes in Health Care Financing and Organization (HCFO) initiative. RWJF established HCFO in 1989 to support research and analysis into health care financing changes and their effects on cost, access and quality. (See also the HCFO Program Results Report.)
Blumberg and Nichols had three goals for their study: (1) to identify differences between insured and uninsured low-income workers (workers with incomes at or less than 200% of the federal poverty level); (2) to determine the effects of reform in the small-group health insurance market (typically firms with 50 or fewer employees); and (3) to identify differences between uninsured workers in large firms and those in small firms.
From June 2000 through July 2002, the researchers compared income, health status, and the size of employers of working uninsured people. Blumberg and Nichols relied on data from three major national surveys:
- The 1999 Current Population Survey (CPS), conducted by the Bureau of the Census for the Bureau of Labor Statistics
- The 1999 National Survey of America's Families (NSAF), an Urban Institute study on the well-being of children and nonelderly adults
- The 1999 National Health Interview Survey (NHIS), conducted by the National Center for Health Statistics of the Centers for Disease Control and Prevention
Blumberg and Nichols worked together on the project until Nichols left the Urban Institute in August 2001. At that point, Blumberg directed the project and collaborated with research associates and assistants to complete it.
The findings. Based on their original goals for the study, Blumberg, Nichols, and their collaborators reported these key findings from their analysis:
- Differences between uninsured and insured low-income workers. Nationally, uninsured workers are more likely than insured workers to be Hispanic, noncitizens, in fair or poor health and work in small firms. They are also more likely to have job tenure of less than one year and to work in agriculture, retail, or construction industries. However, this profile varies among states. Though the national uninsurance rate was 35 percent among low-income workers, that rate ranged from 21 percent in Massachusetts to 47 percent in Texas, two of the 13 states in the NSAF study.
- Reforms in the small group insurance market. Between 1993 and 1996, reforms encouraged pooling—where companies "pool" or share the cost of insurance—in hopes that pooling would increase coverage for high-risk workers in small firms. High-risk workers are more likely to have diseases and medical conditions such as diabetes, high cholesterol, and heart disease. The study found that market reforms had some, but not all, of the expected effects. They had a modest, positive effect on coverage for high-risk workers in small firms, and a negative effect on coverage for low-risk workers in small firms. The researchers concluded that a "fundamental failure of the reforms is that they could not affect the higher premiums faced by all small firms relative to large firms. Increased risk pooling across only small firms does not address their higher administrative costs or risks associated with insuring a particular firm."
- Differences between uninsured people working in large and in small firms. Uninsured workers in large firms differ from their counterparts in small firms in several ways. Uninsured workers in large firms are younger and less likely to be Hispanic and noncitizens. They are more likely to be married and be highly educated, and more likely to have employee-sponsored insurance plans available to them. They are also more likely to work part time; have job tenures of less than one year; and work in the retail industry or in technical, sales, or administrative support occupations. The researchers noted that policy-makers focus more on uninsured people working in small firms and run the risk of neglecting uninsured workers in large firms. They noted that "27 percent of all uninsured workers work for firms with more than 1,000 workers."
Blumberg, Nichols, and research associates published two articles and several reports and briefs from their study. These include: "Exploring State Variation in Uninsurance Rates Among Low-Income Workers" and "State Health Insurance Market Reforms and Access to Insurance for High-Risk Employees."
Blumberg's perspective. Blumberg is a senior fellow at the Urban Institute, and as of 2011 continues to work on health economics issues. She recalls that her earlier work through RWJF "gave me a more in-depth understanding in areas that I had been interested in. To feel that you have an opportunity to go in-depth and into the weeds in those issues was useful for thinking about other areas to move into and research."
Blumberg notes that one major difference between the health care reform legislation that passed under President Obama compared to the effort that did not pass under President Clinton is that "the Obama administration let Congress take the lead in designing the legislation. I'm not sure if that was a deciding factor between success and failure, but that certainly is a distinct difference" she says. "Also, the Clinton approach was very focused on universal coverage—everyone would have had health insurance under that approach. While (Obama's) Affordable Care Act will increase coverage significantly, some will still go without coverage."
Blumberg believes that her HCFO project continues to be relevant today. "All of the research that we have done related to the uninsured became part of a tremendous body of work that has contributed to help plan to expand insurance coverage. I feel that it has stood up very well."
Nichols' perspective. Though Nichols moved from health economics into health policy research after leaving the Urban Institute, he believes that his HCFO research "helped build the case that if you want to provide coverage for people, you have to do comprehensive health insurance reform." And he believes the recent reform initiated by the Obama administration is a start. "It signals a commitment that all Americans will have access to insurance and thereby access to health care. And any great nation should have that," said Nichols, a professor of health policy and director of the Center for Health Policy Research and Ethics, George Mason University as of 2011.
RWJF perspective. HCFO initiative supports investigator-initiated research, policy analysis, and evaluation projects that examine major changes in health care financing and organization, as well as their effects on cost, access, and quality. Launched by RWJF in 1989, HCFO strives to bridge the health policy and health services research communities by providing public and private decision-makers with usable and timely information on health care policy, financing, and market developments; and by bringing together the policy and research communities through significant convening, issues identification, research translation, and communication activities.
"HCFO provides the evidence we need to look at the impact of health care policy changes, as well as the impact of how these policies are implemented," said Nancy Barrand, HCFO program officer and special adviser for program development at RWJF. "HCFO projects help us understand whether or not health care policies are effective, whether or not they are cost effective, and whether or not they are doing what they are supposed to do. Through HCFO projects, we can understand what works, we can measure the impact of changes in the health care market, and we can better understand how policies can be implemented and what those policies should be."