Study

NCHC Writer
Default Image

Down A Dangerous Path:
The Erosion of Health Insurance
Coverage in the United States

Steven Findlay, M.P.H.
Joel Miller, M.S.Ed.

National Coalition on Health Care

May 1999

TABLE OF CONTENTS

EXECUTIVE SUMMARY
INTRODUCTION
DECLINING COVERAGE ON THE JOB
HEALTH INSURANCE, MINORITY GROUPS, AND WELFARE REFORM HEALTH INSURANCE AND UNEMPLOYMENT
INCOME, POVERTY, AGE, AND HEALTH INSURANCE
PROJECTIONS
CONCLUSION
NOTES

EXECUTIVE SUMMARY

Understanding the rapidly changing social, economic, and market dynamics of health insurance coverage in the United States is vital to designing solutions to the problem of the uninsured. In 1997, 43.4 million Americans, 16 percent of the population, had no health insurance.1 Over a three-year period, about 30 percent of the population – 81 million people – can expect to experience a gap in their coverage lasting at least one month.2 And 44 percent of workers who lose their jobs can expect to experience a gap in health insurance coverage lasting a month or longer.3 Even if the rosy economic conditions prevalent since 1992 prevail for another decade, a projected 52 to 54 million non-elderly Americans – one in five – will be uninsured in 2009. If an economic downturn occurs, we project that as many as 61.4 million non-elderly Americans – one in four – could be uninsured in 2009.

This paper describes the major forces undermining health insurance coverage in the United States today, all of which threaten to further erode coverage in the future. It surveys the latest research and analysis in the field. This research all points in the same direction: there is a confluence of social, cultural, demographic, economic, and political forces that is causing a deep and persistent decline in health insurance coverage.Who Is Most Affected?

Our analysis concludes that the decline in coverage disproportionately affects:

  • Low- and middle-income families
  • Young adults ages 18 to 24
  • The near-elderly (ages 55-64)
  • Minority and immigrant populations
  • People who work in small businesses
  • People who are self-employed
  • People who have alternative work arrangements such as day-labor jobs, temporary, or part-time jobs.

Major Forces

The following 10 factors are the chief culprits causing the erosion in health insurance coverage:

  • Employment-based coverage is declining despite exceptional economic good times; fewer people have coverage through the workplace and their jobs.
  • Health insurance is too expensive for many small businesses and unaffordable for many low-income and middle-income Americans. The price of health insurance has increased more rapidly than wages and family income for the past 20 years.
  • Individually purchased health insurance is becoming prohibitively expensive for people who are self-employed, unemployed, or lack access to employer-based coverage. The price varies widely by age and health status.
  • Underlying health care costs continue to rise. Health insurers and managed care plans hike premiums in anticipation of such increases, which are driven primarily by the growth in technology and volume of costly medical procedures.
  • Faced with rising premiums, businesses have shifted costs to workers; that in turn has led more and more workers to decline coverage for themselves and/or their families.
  • Service-sector jobs are increasing faster than manufacturing jobs. Employers in the service sector are less likely to offer health insurance.
  • The number of “contingent” workers is growing. Part-time, temporary or contract employees, and the self-employed are less likely to be offered or able to afford health insurance.
  • Minority and immigrant population groups make up an increasing proportion of the population. They are more likely to be unemployed, work at jobs that do not offer health insurance and have low-wage jobs that prevent them from buying coverage either because they cannot afford it or do not see it as a necessity.
  • Welfare reform has undermined Medicaid coverage among the poor and near-poor.
  • Employers are scaling back retiree coverage to cut costs and lower liability. Rules enacted in 1995 required companies to report such future costs.

Effects of a Recession

The forces undermining health insurance coverage today will be exacerbated in an economic downturn or recession. Our projections underscore other recent analyses.4,5,6 They indicate that a recession or economic downturn, coupled with health care costs expected to rise at two to three times the rate of general inflation over the next few years, could add another 16 to 18 million Americans to the ranks of the uninsured by 2009. This means as many as 61.4 million Americans could be without health insurance in 2009 — one of every four non-elderly people in the United States in that year.Action Needed Now

Steps can and should be taken, beginning in 1999, to ensure that the number of uninsured Americans does not increase and is reduced over time. Those steps should be consistent with a plan to move toward a health system in which every American has health insurance coverage. They also should be consistent with the need to restrain the growth in health care costs and to improve the quality of care.

The four major problems plaguing our health system today – mounting costs, the growing number of uninsured, poor quality care, and the expanding gap between care available to affluent versus lower-income Americans – are deeply interrelated. Efforts to expand health insurance coverage are made much more difficult by rising health care prices. Thus, our nation will have a better chance of expanding coverage if costs can be contained. And we will have a better chance of reducing health costs if all Americans have health insurance. Reduced access to care among the uninsured increases health care costs in the long run because the uninsured do not get the routine preventive care they need and they tend to seek medical attention when they are sicker. They are also more likely to seek care in more expensive settings such as hospital emergency rooms and emergency clinics. Likewise, we can never have the highest quality of care possible in a system in which tens of millions of Americans lack appropriate care even as others receive too much or the wrong care.

It is thus a vicious cycle. A decade of research and real-life experience now compellingly show us that a continually rising number of people without health insurance will further destabilize an already strained and faltering health care system. We continue to ignore this problem at our peril.

INTRODUCTION

The United States has enjoyed an unprecedented economic boom since 1992. Low inflation and interest rates combined with rising wages and salaries has produced strong consumer demand for goods and services. That has propelled rapid business growth and rising employment. The unemployment rate has dipped to the lowest level in almost 30 years, 4.4 percent. Wages have risen and businesses have expanded their fringe benefits over the last decade. More large- and mid-sized companies offer retirement plans, flexible spending accounts, child care, and various forms of insurance. In addition, the price of health insurance rose only modestly between 1994 and 1998. Yet during this boom period, the number of Americans with no health insurance has risen sharply.

In 1990, 35.6 million Americans, 16 percent of the non-elderly population, lacked health insurance. By 1997, the number of uninsured below age 65 had risen to 43.1 million, 18.3 percent of the non-elderly population. (See figure 1.) From 1996 to 1997, the number of people without health insurance increased by 1.7 million, the largest annual increase since 1992. The basic facts in 1997 were these:7

  • 57% of people ages 18 to 64 who had no health insurance worked either full-time or part- time.
  • 18% did not work.
  • 25% (10.8 million) were minors below age 18.
  • 42% of the uninsured lived in households with an annual income below $25,000; 24% of the uninsured had family incomes above $50,000; 34% had incomes between $25,000 and $50,000.

In many urban areas, the situation has grown particularly grim. In 21 of the nation’s largest metropolitan areas, 20 percent or more of the non-elderly population lack health insurance today. In El Paso, Texas, 39 percent of residents lack health insurance; in Los Angeles, 31 percent; in Houston and Tucson, 29 percent; in Miami, Florida, 28 percent; in New York City, 25 percent; and in New Orleans, 22 percent.8 Wide geographical variation in the number of uninsured occurs at the state level, too. In 16 states, the number of uninsured people exceeds the national average of 16 percent of all residents. In six states (Arizona, Arkansas, California, Mississippi, New Mexico, and Texas), more than one in five non-elderly residents lack health insurance.9

The percentage without health insurance skyrockets when you look closely at low- and middle-income working families. In a recent analysis of U.S. Bureau of the Census data, Jocelyn Guyer and Cindy Mann found that in 31 states the proportion of families with incomes below $32,000 (200 percent of poverty for a family of four) that lacked health insurance exceeded 40 percent in 1997. In a few states (Arizona, Texas, and Arkansas), almost two-thirds of such families were uninsured. The average for all states was 33.7 percent. The proportion of uninsured under 200% of poverty ranged from a low of 8 percent in Hawaii to 63 percent in Texas.10FIGURE 1

Source: Employee Benefits Research Institute.

DECLINING COVERAGE ON THE JOB

Most Americans (61 percent) receive health insurance coverage through their employer. But this proportion has been declining over the last decade. In 1987, 69.2 percent of the non-elderly population had employment-based coverage. By 1997, that proportion had declined to 64.2 percent.11 Loss of coverage in the private sector is responsible for most of the decrease. The number of federal, state, and local government workers without health insurance increased from 6.5 percent in 1987 to 8.5 percent in 1997. But the number of private-sector workers and their family members without coverage increased from 15.4 percent in 1987 to 18.6 percent in 1997. Put another way, 33.5 million private-sector employees and family members in 1997 had no health insurance (of 180 million private-sector workers and dependants). That is up from 24.7 million uninsured private-sector employees in 1987 (out of 160.8 million workers and dependents).12

There are several major reasons for this decline. Chief among them is that private-sector employers, faced with the rising costs of health insurance since 1987, have shifted part of these costs to employees. This change has occurred especially in small businesses. Workers are now asked to pay a greater proportion of the premiums and higher deductibles as well. One recent study found that workers at small businesses (fewer than 200 employees) in 1998 paid an average 44 percent of the premium for family coverage, up from 34 percent in 1988. By comparison, employees in larger firms (200 or more employees) paid an average 28 percent of premium costs for family coverage.13 The difference can be significant. Employees of small firms can wind up paying $500 to $1,000 more per year out of their own pockets for family health insurance coverage.

Faced with higher out-of-pocket costs for health insurance, a growing number of workers are choosing not to take health insurance. An estimated 6 million fewer people took health insurance from their employers in 1996 than in 1987.14 This decline occurred despite the fact that more small businesses have provided health insurance to their workers over the last decade. In 1998, for example, 54 percent of firms with fewer than 200 workers offered health insurance, up from 46 percent in 1989. (See figure 2.) Notably, however, a steady increase between 1989 and 1996 in the number of small businesses willing to provide health insurance appears to be reversing. In 1996, for example, 59 percent of firms with fewer than 200 workers provided health insurance, 5 percentage points higher than in 1998. Ominously, the drop in coverage was largest for the smallest, and most vulnerable, firms. In 1998, 49 percent of firms with three to nine workers provided health insurance, down from 53 percent in 1996. And 71 percent of businesses with 10-24 workers provided health insurance, down from 78 percent in 1996. These statistics come from a survey of 1,581 randomly selected firms with 199 or fewer workers. The survey was conducted in the summer of 1998.15FIGURE 2

Source: Gabel et al., Health Benefits in 1998 for Small Employers. A Report to the Henry J. Kaiser Family Foundation (February 1999).

FIGURE 3

Source: Gabel et al., Health Benefits in 1998 for Small Employers. A Report to the Henry J. Kaiser Family Foundation (February 1999).

Many employees are now being asked to pay not just a larger share of health insurance, but the entire cost. Millions of Americans who are counted as having employer-provided coverage actually get no help from an employer in paying for that coverage, according to a recent study. The study, by Dr. Olveen Carrasquillo and colleagues, found that 9.1 million Americans who had coverage through an employer-based plan paid the full cost themselves in 1996.16 That is 8 percent of workers and their families with private employer-based coverage.

The same study also demonstrates that the decline in private-sector employer-based coverage has led to a system in which an increasing number of Americans get coverage from the government – federal, state, and local. Overall, government provided health insurance to 34.2 percent of the population in 1996 – 91.4 million people. The government provides insurance as the nation’s largest employer. Millions more participate in public programs such as Medicare, Medicaid, and the Children’s Health Insurance Program (CHIP). Twenty-two million had health insurance because they were either employed by local, county, state, or federal government or were dependents of such workers. In contrast to the commonly held belief that private employers provide the bulk of health insurance in our nation, the study concludes that only 43 percent of the population (114.9 million of a population of 267.2 million) had health insurance paid for by private-sector employers in 1996. Just over 7 percent (18.7 million) purchased coverage on their own, the study found.17

Structural changes in the economy are the second major reason for the decline in employment-based coverage. After World War II, labor unions began to demand health insurance coverage as part of collective bargaining agreements. Unions were particularly pervasive in the burgeoning heavy industries that generated the postwar economic boom. By the 1960s and ’70s, almost all workers in heavy industry and large companies had health insurance, as they do today. In addition, most unionized workers had health insurance. In 1995, 5.9 percent of union members lacked health insurance compared to 16.8 percent of nonunion workers.18 But labor unions represent a smaller percentage of workers today, and fewer Americans work in industries that are heavily unionized. A decrease in industrial and manufacturing jobs and a corresponding rise in employment in the service and white-collar business sectors of the economy has contributed to a decline in affordable health insurance. Businesses in the service sector – especially retail trade, business services industries, and construction – are, in particular, less likely to offer health insurance to workers. (See Figure 4.)FIGURE 4

Percentage of Workers Uninsured By Industry
Industries with a Rising Percentage
Of Uninsured Workers		19901995Percentage
Point Change
Business Services21.9%26.1%4.2%
Mining12.5%16.7%4.2%
Self-Employed21.7%24.6%2.9%
Manufacturing10.6%13.5%2.9%
Retail Trade22.7%24.9%2.2%
Transportation9.9%12.0%2.1%
Wholesale Trade13.6%15.2%1.6%
Construction28.4%29.9%1.5%
Professional9.6%10.2%0.6%
Finance8.2%8.1%-0.1%

Source: Kenneth Thorpe, Analysis of Tabulations from the March 1991-96, Census Bureau Current Population Surveys. (See Note 19.)

At the same time, the nature of the labor force is changing. More and more Americans are working part-time, are self-employed, or work on a contractual, “contingent,” temporary, or day-labor basis. Independent contractors and the self-employed are the most common alternative arrangement. Today, about 15 percent of the workforce – some 22 million workers – are employed in such “non-traditional” jobs. Another 10 percent or so – some 14 million people – work part-time. Notably, people ages 50 to 64 – the near-elderly – are the most likely to work outside the traditional wage and salary structure. Some 20 percent of people in this age bracket who are working are employed in alternative arrangements, and demographic trends among the baby-boom generation indicate that an increasing number of 50-64-year-olds may seek such alternative career choices in the future. (See Figure 5.)FIGURE 5

Distribution of Work Arrangement By Age of Worker, 1997
 Worker Age
Work Arrangement16-3435-4950- 64Total
Temporary Worker1.4%0.8%0.7%1.0%
Contract Worker0.7%0.6%0.5%0.5%
On-Call Worker1.8%1.3%1.5%1.6%
Independent Contractor3.7%7.8%9.1%6.5%
Self-Employed2.4%5.5%8.1%5.5%
Wage/Salary Worker90%84%80.1%84.9%
Total100%100%100%100%

Source: Kenneth Thorpe, Analysis of February 1997 Supplement to the Current Population Survey.

People who have either chosen or been compelled to earn a living outside a traditional job are much less likely to have health insurance. (See Figure 6.) For example, 53.3 percent of temporary workers were uninsured in 1997 as were 27 percent of independent contractors and a third of “on-call” workers. The data in Figure 6 also show the variation of insurance coverage by age. While 8.4 percent of wage and salary workers ages 50 to 64 were uninsured, 21 percent of older workers employed as independent contractors were uninsured. Nearly one in four workers ages 16 to 34 was uninsured, including more than 60 percent of temporary workers in this age category. Figure 6 shows as well that health insurance coverage for workers in alternative work arrangements was far lower compared to regular full-time workers across all age categories.FIGURE 6

Distribution of Uninsured Workers By Worker Age and Work Arrangement, 1997
 Worker Age
Work Arrangement16-3435-4950- 64Total
Temporary Worker60.9%49.3%32.8%53.3%
Contract Worker15%20.1%17.7%17.1%
On-Call Worker35.9%33%25.4%33.1%
Independent Contractor35.9%27.1%21.4%27%
Self-employed32.4%18.1%14.8%19.9%
Wage and Salary Worker21.8%11%8.4%15.1%
Total23.3%13.3%10.5%16.8%

Source: Kenneth Thorpe, Analysis of the February 1997 Supplement to Current Population Survey.

The chief reason nontraditional workers do not have coverage is that they do not work enough weeks per year or hours per week to qualify under their employers’ health benefit plans. In an analysis of Census Bureau data, Kenneth Thorpe found that 53.3 percent of workers in 1997 who were not eligible for coverage at a job where health insurance was offered either worked part-time or only worked a portion of the year.20 The analysis also found that only about half of part-time employees (fewer than 20 hours a week) worked at firms that offered health insurance. And 73 percent of them were not eligible for coverage.

Overall, Thorpe calculates, 66.7 million (54.2 percent) of the 123 million American workers employed on any given day in 1997 had coverage through their employer; 10.1 million workers (8.2 percent) were not eligible for coverage; 11.4 million (9.3 percent) were eligible but declined coverage; and 34.8 million (28.3 percent) worked for firms that did not offer coverage. Twenty-one million (17 percent) of the 123 million workers in this tally were uninsured. Of the 11.4 million eligible workers who declined coverage at their own jobs, 7.5 million picked up coverage through a spouse’s job; 800,000 purchased it individually; 600,000 had some kind of public health insurance; and 2.5 million were uninsured.21FIGURE 7

Distribution of Workers’ Access to Health Insurance, 1997
% DistributionAge of Worker
 16-3435-4950-64Total
Through own employer48.6%60.1%58.2%54%
Through spouse or parent25.3%22.4%25%24%
Purchased individually2.8%4.2%6.3%5%
Uninsured23.3%13.3%10.5%17%

Source: Kenneth Thorpe, Analysis of February 1998 Supplement to the Current Population Survey.
Note: Figure presents data for workers and families, not entire non-elderly population.

The fact that so many Americans now get health insurance through a spouse’s or parent’s health plan – sometimes declining coverage at their own job – is an indication of the extensive cost shifting in the health care system. (See Figure 7.) The main reason people do this is because it is cheaper to join the spouse’s plan. For example, an employee at a small firm that requires workers to pay 50 percent of the premium is highly likely to opt out and join a spouse’s plan at a larger company that requires a lower (say, 20%) premium contribution. As discussed above, it is more likely (but not always) the case in today’s system that the small businesses will require the larger premium contribution.

Employers are well aware of the phenomenon. Some now bar spouses from joining their health benefit plan if they are offered coverage at their own job. But the vast majority of employers have responded instead by asking workers to pay a larger portion of the cost for spousal and dependent care coverage than they do for single or worker-only coverage.22 This trend could portend a continued decrease in employer-based coverage as premiums rise, since the amount of the family contributions tend to rise as the premiums increase. The low- and middle-income worker facing the prospect of a bigger slice of his or her earnings being taken out of their paycheck for family health insurance is more likely to decline the coverage. And, denied coverage through a spouse’s cheaper plan, employees facing higher premium contributions at their own jobs are also more likely to decline coverage.

The downward trend in coverage through one’s own job is exacerbated by the increased use of “waiting time” – the period new employees must wait before they become eligible for coverage. During this period, millions of Americans choose to “go bare” (have no health insurance) for periods up to six months even if they are offered COBRA (Consolidated Omnibus Budget Reconciliation Act of 1985) coverage through their previous job. That is because they must pay the full premium for COBRA coverage. Only a fraction (less than 5 percent) of those eligible for COBRA opt for it because of the high cost. In 1997, 27 percent (2.7 million) of the 10.1 million people who were ineligible for health insurance at a job were in this “waiting period” predicament because they had not worked at the job long enough.23

Many Americans were led to believe that a 1996 federal law addressed such problems. It did not. The Health Insurance Portability and Accountability Act (HIPAA) makes it easier for some people who switch jobs to retain coverage or for those who lose employer-sponsored health insurance to obtain coverage on their own. But it does not eliminate waiting times to qualify for coverage. Moreover, it requires people to exhaust 18 months of COBRA benefits before qualifying for guaranteed individual coverage. Overall, there is little evidence HIPAA has been much help in extending coverage to workers who have lost employer-based coverage. Insurers are allowed under the law to charge such individuals substantially higher premiums. This has limited access to HIPPA-qualified coverage.24

HEALTH INSURANCE, MINORITY GROUPS, AND WELFARE REFORM

Demographic changes, too, are adding to the decline in health insurance coverage. Together minorities comprised 46 percent of the uninsured, although these groups represent 24 percent of the U.S. population.25 In a recent study of urban areas, researchers at the UCLA Center for Health Policy Research found that the American cities with the highest uninsured rates also had the largest minority and immigrant populations.26 The rapid growth in the proportion of the population that is Black, Hispanic, or of Asian extraction – combined with the increasing number of new immigrants – present serious challenges, especially in urban areas.

The reason for the decline of coverage among these population groups is not complicated. Unemployment is on average higher – sometimes three to four times what it is for the non- minority population. In addition, minority and immigrant population groups are much more likely to have low-wage jobs that do not offer health insurance. They have fewer family assets as well. Thus people in these population groups do not have the income or reserve assets to purchase health insurance on their own if they are between jobs or unemployed. They also are more likely to have non traditional, part-time or temporary jobs. In 1997, 7.4 million African Americans (21.5 percent), 10.5 million Hispanic Americans (34 percent), and 2.2 million Asian and Pacific Islanders (20.7 percent) had no health insurance. By comparison, 15 percent of White Americans had no health insurance.

But the problem is not isolated to low-income minorities. A study by Allyson Hall and Karen Scott Collins of the Commonwealth Fund and Sherry Glied of Columbia University found that minority workers are more likely to be uninsured even after income has been factored out. Put another way, members of minorities with jobs and incomes comparable to those of White Americans are less likely to have employment-based coverage. The study found, for example, that among workers with employment-based coverage in their own name in 1997, 69 percent of White Americans had coverage compared to 52 percent of Black Americans and 44 percent of Hispanic Americans. Even high income minority individuals were less likely to have insurance. For example, 70 percent of Hispanics in executive positions had employment-based coverage compared to 83 percent of Whites. Overall, the authors calculated, Blacks and Hispanics are 21 percent less likely than Whites to have coverage through their employers. The authors say the reasons for the disparity are not well understood. But they surmise that the following factors play a role: (1) minority workers are more likely to decline coverage if they have to pay a large portion of the premium; (2) minority workers are more likely to have jobs where the employer requires them to pay a larger share of the premium; (3) some minority workers may not value health insurance and thus decline it at work.27

A projected sharp increase in the proportion of the population that will be minorities in the future is likely to lower health insurance coverage rates in the United States. The Census Bureau reported in January 1999 that the number of immigrants living in the United States tripled between 1970 and 1998, rising from 9.6 million to 26.3 million. Most are members of minority populations. The influx is expected to continue although at a less brisk pace. Overall, the Bureau projects that 30 percent of the population will be members of minority groups in the year 2005, rising to 35.8 percent in 2020. The proportion is 24 percent today.

Welfare reform also appears to be exacerbating the decline in health insurance coverage. When Congress passed landmark welfare reform legislation in 1996, it delinked welfare and Medicaid, the federal-state health insurance program for low-income people. Those dropping off welfare cash assistance roles were guaranteed continuing Medicaid coverage if they qualified. Even those who got jobs and made too much money to qualify for Medicaid were guaranteed transitional coverage for six months. At the time, the Congressional Budget Office (CBO) projected no net loss of coverage because it assumed many welfare recipients would pick up coverage as they joined the workforce while most others would retain Medicaid.

But there are strong early signs that many former welfare recipients are not gaining coverage at new jobs and that those dropping off the welfare roles are losing Medicaid coverage. Data are not yet available for the first full year (1998) of the federal welfare overhaul. But several recent studies and news reports document the trend, by using primarily data from states that led the shift to welfare reform between 1994 and 1997. Mark Greenberg of the Center for Law and Social Policy cites data showing that 400,000 children below age 15 lost Medicaid coverage nationwide between 1995 and 1996. Total enrollment fell from 15.5 million in 1995 to 15.1 million in 1996. The number of adults ages 21 to 44 with Medicaid dropped from 8.6 million to 8.2 million over the same period, he reports. Based on a review of more recent state data (1997 and 1998), Greenberg concludes that roughly one-third of children and “a majority” of adults (ranging from 60 percent to 80 percent) who go off welfare also drop off the Medicaid roles.28 In another study, Marilyn R. Ellwood and her colleagues at the Mathematica Policy Research in Cambridge, Massachusetts, concluded that non-elderly Medicaid enrollment declined 2 percent nationwide between 1995 and 1996.29 A recent study in New York finds that, between 1995 and 1998, the number of Medicaid enrollees in the state declined by 300,000, or 10 percent. During the same period, the number of uninsured in New York increased by about 450,000, or 17 percent.30 Finally, preliminary data from the Congressional Budget Office (CBO) indicates that as many as 2.5 million people lost Medicaid coverage between 1997 and 1998, with coverage declining from 12 percent to 11 percent of the U.S. population.31

These preliminary studies and reports in the media32 give an early indication of what is happening. The majority of previous welfare (and Medicaid) recipients are getting jobs where they are either not offered health insurance or where the employee costs are high enough to discourage low-wage workers from opting for coverage. In addition, many former welfare recipients do not know that they can stay on Medicaid when they become employed. Still others appear to join the workforce only briefly or switch jobs often. In both cases, they may not satisfy initial employers’ waiting periods to qualify for health insurance. Some may also get jobs that offers coverage, take them, and then leave or lose those jobs – losing the coverage at the same time. They then may be reluctant to go back on Medicaid, or not know that they can.

HEALTH INSURANCE AND UNEMPLOYMENT

While most people who lack health insurance are employed, 18 percent of the uninsured between the ages of 18 and 64 in 1997 – 7.8 million Americans – did not work during the year. That is an increase from 15.5 percent (6.2 million) in 1996.33

The link between the unemployment rate and the number of uninsured is more complex than generally appreciated. The sharp rise in the number of uninsured in recent years has occurred despite a prevailing low unemployment rate. But there is still a strong relationship between the two. A recent Urban Institute study found in Wisconsin, for example, that the strongest factor affecting health insurance coverage at the county level was the unemployment rate, as opposed to, say, the number of small businesses that did not offer coverage, immigrant population levels, or income level.34

National data also show that at times the unemployment rate strongly affects fluctuations in the number of uninsured. The largest rise in the number of uninsured over the last decade occurred between 1991 and 1992, a year in which a recession drove the unemployment rate to 6.5 percent. Two million people were added to the ranks of the uninsured that year, with a rise from 36.3 million to 38.3 million.

A recent Census Bureau analysis also underscores the relationship. It finds that during the period from 1993 to 1995, 44 percent of persons who lost their jobs also reported a loss of health insurance coverage lasting a month or more.35

In another analysis of data from the Census Bureau, Kenneth Thorpe calculates that each 0.5 percentage point increase in the unemployment rate can in theory – other factors being held constant – add approximately one million Americans to the ranks of the uninsured over a period of a year. (See Figure 8.) In 1997, for example, the unemployment rate was 5 percent in a workforce of 139 million (employed at some point during the year). Thus some 7 million Americans who wanted to work were unemployed. If the unemployment rate had climbed to 5.5 percent, an additional 695,000 people would have been out of work. About a quarter of these workers, Thorpe calculates, would be able to retain coverage by joining a spouse’s plan or through COBRA. That leaves 522,000 million workers with roughly half a million spouses and/or dependents who would lose coverage for some period of time.FIGURE 8

The Relationship Between Employment and Health Insurance
Workers Ages 18 to 64 and Their Dependents
Unemployment RatePercentage UninsuredNumber
(millions)
5.0%19.6%32.3
5.5%20.2%33.3
6.0%20.8%34.3
6.5%21.4%35.3

Source: Kenneth Thorpe, Derived from a regression analysis based on March 1991 through March 1998, Current Population Survey.

INCOME, POVERTY, AGE, AND HEALTH INSURANCE

It has been an established fact for more than a decade that income level correlates closely with health insurance coverage. Low-income Americans are much more likely to lack health insurance than affluent Americans. But that relationship, too, has become more complex over time. Recent data indicate that more and more middle-income people are at risk of becoming uninsured. This is primarily because the cost of health insurance has risen so sharply over the last 15 years that more and more middle-income people find it unaffordable.

More than a third of all Americans living in poverty (36.8 percent) were uninsured in 1997. (See Figure 11.) But 29 percent of non-elderly individuals and families with incomes up to 200 percent of the federal poverty level — $32,000 a year annual income for a family of four — are also uninsured. That is a solid “middle-class” income in many areas of the country. The median household income in the United States in 1997, for example, was $37,000. (The poverty level in 1997 was $13,330 for a family of three and $16,700 for a family of four.)

Consider this illustrative example of the problem: In the late 1980s, health insurance for a healthy family of four (purchased individually) cost around $2,500 a year, $208.33 a month. A family making an average income then of $25,000 a year or $2,083 a month would have had some trouble paying for health insurance; it would have cost them roughly 12 percent of their after-tax income. But today they would have to pay an even larger portion of their income for that coverage. Health insurance for a family of four now costs from $5,000 to $7,000 a year, or $416 to $583 a month. A family with an income of $32,000 would have take-home pay of roughly $2,300 a month. Health insurance would cost 18 percent to 25 percent of their monthly income.

As we know, of course, most people do not buy health insurance on their own but get it through a job, and their employer usually pays part of the cost. Our hypothetical family above would still be stretched today if their employer required them to pay a sizable portion of the cost. If they had to pay 50 percent of the premium, for instance, their share of the premium would be $208 to $291 a month – 9 percent to 12.6 percent of their income. And that does not take into account deductibles and other out-of-pocket health care expenses. Many health economists agree that health care expenditures which exceed 10 percent of a family’s income constitute a burden – and lead to a drop off of coverage.36

A recent study explores this terrain in some depth.37 In a complex analysis of health insurance coverage and health care cost data for the years 1979 to 1995, Richard Kronick and Todd Gilmer conclude that the lion’s share of the decline in health insurance coverage in the last 20 years is attributable to the fact that health care and health insurance prices have increased far more rapidly than income. This, they say, has priced many workers out of coverage. The researchers calculate, for example, that in 1995 an average middle-income worker spent 7.3 percent of income for health care and health insurance, up from 4.5 percent of income in 1979. (The calculation is based on a per capita or per person share of national health expenditures.) That translates into a 60 percent rise in per capita health spending over the 20-year period, adjusted for inflation, while real income hardly rose at all.

The effect on health coverage is clear. Among workers spending less than 5 percent of their income on health care, less than 10 percent are uninsured. Among workers spending 5 percent to 10 percent of income on health care, approximately 15 percent are uninsured. When the proportion of income spent on health care rises above 10 percent, the uninsured rate soars – to 32 percent for those spending between 10 percent and 15 percent of income and to 50 percent for those spending from 15 percent to 25 percent of income.38

Looking over the 16-year time period, Kronick and Gilmer calculated that in 1979-83, per capita expenditures on health care (including insurance) were less than 5 percent of income for about half of all workers and more than 10 percent for about one-fifth of workers. By 1992-95, however, per capita expenditures on health care were less than 5 percent for only 26 percent of workers and more than 10 percent for a third of all workers. Again, the result: millions of workers priced out of coverage. The researchers further calculate that of those earning $10,000 to $15,000 (in adjusted 1995 dollars) during the period 1979-83, 27 percent were uninsured. By 1992-95, 43 percent were uninsured. The decline occurred for middle-income workers as well, although the effects were not as pronounced.

Kenneth Thorpe also has recently tabulated family spending on health care. As Figure 9 shows, families with annual incomes below $15,000 in 1995 spent an average 12 percent of their incomes on health care premiums (their share of employer-provided coverage) and direct out-of-pocket costs. Data from numerous sources show that between 30 percent and 45 percent of this group are uninsured (depending on whether you are looking at individuals, families, or households). Families with incomes of $15,000 to $30,000 were spending between 8 percent and 10 percent of incomes on health care. In contrast, families with incomes between $50,000 and $70,000 spend 4.4 percent of their incomes on health.FIGURE 9

Health Care Expenditures By Family Income, 1995
Annual IncomeYearly Health SpendingHealth Spending By Percentage Total
Below $10,000$1,46512.1%
$10,000 – $14,999$1,87512.4%
$15,000 – $19,999$1,65510.2%
$20,000 – $29,999$1,8057.9%
$30,000 – $49,999$1,8136.9%
$50,000 – $69,999$1,8364.4%

Source: Kenneth Thorpe, Analysis of Consumer Expenditure Survey, 1995, U.S. Department of Labor.

Another recent study documents that individually purchased health insurance is increasingly unaffordable and thus not a viable alternative for most families – including those with fairly high incomes – who lose employer-based coverage.39 Deborah Chollet and Adele Kirk at The Alpha Center in Washington, D.C., found that even in families with incomes above 400 percent of poverty, only 52 percent of adults and 59 percent of children with neither employer-based nor public (Medicaid) coverage reported having an individual health insurance policy at least part of the year of the study (1996). Only 29 percent of people in families with incomes between 200 percent and 300 percent of poverty purchased individual health insurance when they did not have insurance from an employer or Medicaid. In addition, the study found that 30 percent to 40 percent of people with individually purchased insurance in 1996 held their policy for only part of the year and that policies differed widely in the medical services they covered. (Somewhere between 16 million and 20 million people in 1998 held individually purchased health insurance for at least part of the year.)

Age makes a difference, too. (See Figures 10 and 11). Forty-seven percent of households headed by a person ages 19 to 34, and with annual incomes below the poverty level, lack insurance coverage. A recent study bolsters this point, especially for those below age 25. The study found that while nearly half of workers between ages 18 and 24 had health insurance at their jobs, the health insurance “acceptance rate” declined 13 percent between 1987 and 1996. As a result, 38 percent of 18- to 24-year-olds lacked health insurance in 1996. Although they make up one- tenth of the population, this age group represents one-fifth of the uninsured. Young people cited price as the chief reason they did not opt for coverage at work or buy it on their own. This age group also suffers sometimes from abrupt loss of coverage under a parent’s insurance plan when they leave college – even if they do not yet have full-time jobs. Shorter job tenure also was another factor in the lower coverage rate among 18- to 24-year-olds. Not surprisingly, the study found that only 34% of 18- to 24-year-olds reported having a doctor or regular place where they can get care. 40FIGURE 10

Family Spending on Health Care, By Age of Head of Household and Family Income, 1995
Age$15,000 – $19,000$20,000 – $29,999$30,000 – $39,999
 $%$%$%
19-34$5352.7%$5002.5%$1,4205.5%
35-49$1,2656%$1,3906.8%$1,2624.4%
50-64$1,3957.6%$2,70410%$2,0536.8%

Source: Kenneth Thorpe, Analysis of Consumer Expenditure Survey, 1995. U.S. Department of Labor.

But it is the near elderly (people ages 55 to 64) who are perhaps the most worrisome age group with respect to health insurance. Low- and moderate-income people ages 50 to 64 naturally spend more (7 percent to 10 percent of annual income) for health care than younger people. They have greater health needs and spend more for health insurance and on medical goods and services not covered by insurance. Only food (15 percent) and housing (23 percent) account for a higher share of annual expenditures than health care among families headed by a 55- to 64-year-old.

More to the point, many in this age cohort are in a transition period, moving from full-time work to early retirement, contractual or consultant work, or part-time employment. A recent study by Sherry Glied and Mark Stabile at Columbia University looks closely at the older end (61- to 64-year olds) of this population group. The good news is they find that, overall, almost 84 percent of 61- to 64-year olds had some form of health insurance in 1997 – 52.2 percent from employers; 12 percent from Medicare (as disabled persons); 9.6 percent privately purchased; and 10 percent from either Medicaid or CHAMPUS (retired military). Sixteen percent of 61- to 64-year olds are uninsured, up from 14 percent in 1989 but essentially unchanged since 1993.41

The bad news, however, is that Glied and Stabile’s data show a sharp and ominous drop-off of employer-based and privately purchased coverage for 61- to 64-year olds relative to people younger than 61. Slightly more than 67 percent of people ages 45 to 60 have employer-sponsored coverage, for example, compared to 52.2 percent of 61- to 64-year-olds. And while almost 70 percent of 57- to 60-year-olds have health insurance, the proportion drops to 62.4 percent for 61- to 64-year-olds – leaving 37.6 percent uninsured. The percentage of 61- to 64-year-olds with some form of private coverage (employment based or individually purchased) has declined from 68 percent in 1989.

Health insurance coverage among the near elderly in the future, Glied and Stabile assert, will depend in part on coverage trends among younger workers. That is, as employer-based coverage erodes among people aged 40 to 55, it will also erode among 55 to 64 year olds. But future coverage among the near-elderly will also depend to a substantial degree on their employment and retirement patterns. The planned increase in the Social Security retirement eligibility age and proposals to increase the Medicare eligibility age, now under discussion in Congress, will be addition factors. Less than a third (31.4 percent) of people ages 61 to 64 now work full time. That is up slightly from 30 percent in 1993.42 But labor force participation among 55- to 64-year-olds has generally declined over the past 40 years, from 85 percent in 1960 to approximately 65 percent now.43 In 1995, 58 percent of 62-year-olds elected to begin receiving early Social Security benefits, up from 40 percent in 1980.44

The question for the future is whether the baby-boom generation – fast approaching this age cohort (the oldest baby boomers, born in 1946, are turning 53 in 1999) – will work longer or continue the trend to early retirement. No one knows for sure. There are cogent arguments both ways. They may choose to work longer because they are living longer and are generally a healthier population than past generations.45 In addition, this generation may have new options for innovative work arrangements that could include health insurance – options unavailable to previous generations. For example, valued and experienced employees who work at companies for many years may be offered the option of continued work out of their homes or when they reach their late 50s and early 60s. Part of that bargain may be health insurance. But, as discussed above, these alternative work arrangements may just as easily not carry the option of health insurance. Indeed, some companies may specifically choose to move older workers into such arrangements so they do not have to provide them with benefits, including health insurance. The higher the price of insurance, the more likely that is to occur.

The planned increase in the Social Security eligibility age – and the proposed increase in the Medicare eligibility age – also may affect the age of retirement. The Social Security eligibility age, by law, will rise slowly to 67 over the next 28 years. In addition, the reduction in benefits for early retirement (before age 67) will ultimately rise to 30 percent from the current 20 percent. Proposals have been advanced to raise the eligibility age for Medicare to 67 as well.

These shifts could compel future 60- to 67-year-olds to work longer. That could mean more would retain health insurance coverage. On the other hand, the shifts could produce more uninsured older Americans. That would occur if baby boomers maintain or even accelerate the trend to earlier retirement. And the decline in coverage could be especially sharp if the Medicare eligibility age is raised to 67 without an adequate program in place to help retired and unemployed 65- and 66-year-olds obtain private coverage or buy into Medicare.

Glied and Stabile project health insurance coverage for 61- to 64-year-olds in 2005 under four scenarios. They conclude that it is more likely that fewer people in this age cohort will have health insurance in the future, though the proportion depends on factors mentioned above. Under the worst case scenario, 1.25 million more 61- to 64-year-olds would be uninsured in 2005 than predicted today, and the percentage of 61- to 64-year olds without private coverage (employer-based or individual) would rise from 37.6 percent to 42.1 percent. Under the most optimistic scenario, which includes a limited Medicare buy-in for people ages 61 to 64, coverage levels would remain the same as in 1997. “Both the number and the percentage of older Americans who lack private insurance coverage are likely to rise over the next decade,” Glied and Stabile conclude.46

The steady decline in the proportion of large employers who offer health insurance coverage to early retirees also clouds the outlook for future 55- to 64-year-olds. A recent survey of large employers shows the percentage of companies with 500 or more employees providing coverage to retirees not yet eligible for Medicare fell from 46 percent in 1993 to 36 percent in 1998. Companies providing coverage to Medicare-eligible retirees fell from 40 percent to 30 percent during the same period.47 The U.S. General Accounting Office (GAO) recently concluded that employer-sponsored retiree coverage was highly likely to decline further. And where such coverage is preserved, retirees will likely pay a larger share of the costs for the benefit.48

Data gathered by Chollet and Kirk underscore that individually purchased insurance is not currently an affordable option for the majority of the near-elderly. They found that a healthy 60-year-old man in 1996 would have paid as much as $700 a month ($8,400 a year) for health insurance if he was living in a high-cost area. The study, which probed the individual health insurance market in 10 states in 1996, found that premiums ranged widely for the near-elderly, from $135 a month to $700. A 60-year-old with a medical condition, such as high blood pressure or diabetes or a past history of cancer, would incur additional costs and could pay as much as $1,000 a month for individually purchased health insurance.49

Whether young or late middle-aged, the number of uninsured below the poverty level (See Figure 11) is particularly disturbing because a sizable proportion of this population is eligible for Medicaid – but not enrolled. Medicaid coverage varies from state to state. But in general, adults and children are covered if family income is below 133 percent of the poverty level. Since 1996, a growing number of states have expanded Medicaid to cover pregnant women and children in families with income up to 185 percent of the federal poverty level. The bottom line is this: Medicaid covered 11 percent to 12 percent of the total population in 1997 (approximately 32 million people) but that represented only 45 percent of those with incomes below the poverty level.

Medicaid is a highly valuable (and valued) source of health insurance coverage and care for tens of millions of Americans, but it has consistently left out many who qualify and need help – now more than ever. Eligibility criteria leave out millions of low-income Americans who can no longer afford health insurance on their own because the price is too high.FIGURE 11

Number and Percentage of Non-Elderly Without Health Insurance, 1997
(numbers in millions)

INCOME AS PERCENT OF POVERTY
 < 100101 – 200201 – 300301 +Total
Age Category#%#%#%#%#%
0-185.727.3%3.221.9%1.410.9%1.34.9%11.615.3%
19-346.447.0%5.037.7%2.422.7%2.19.2%15.926.2%
35-493.844.5%2.831.5%1.616.6%2.15.7%10.316.3%
50-641.836.0%1.324.5%0.816.0%1.46.4%5.314.2%
Total17.836.8%12.229.2%6.116.4%7.06.4%43.118.2%

Source: Kenneth Thorpe, Analysis of Tabulations from March 1998, Current Population Survey. Families defined as health insurance units, i.e. those family members eligible for insurance through a single policy.

The effect of millions of low-income people of all ages being without health insurance is enormous, and by now well known. Numerous studies show unequivocally that people with low incomes who are uninsured often do not obtain needed or adequate medical care.50 This failure occurs both because they cannot afford to go to the doctor and because easily accessible medical care may not be available where they live. The low-income uninsured rely on getting care at doctors’ offices, hospitals, and clinics, which often provide care for free or at reduced fees. But this system varies considerably across the country and is driven largely by state and local policies that affect the organization and financing of the health care safety net. These problems are at risk of becoming greater in the near future. Doctors, hospitals, and clinics in many parts of the country are finding it increasingly difficult to justify unlimited and open access to care for uninsured people for the following reasons:

  1. The steady growth in the number of uninsured persons
  2. The increased fiscal pressure imposed on doctors and health facilities by managed care organizations
  3. The decline in public subsidies (direct and indirect) for indigent and charity care
  4. The increasing difficulty health facilities face shifting costs for the medically indigent onto other payers (businesses, unions and consumers)
  5. A reduced focus on the uninsured because of the increased competition for paying patients.

One recent study documents this trend. Peter Cunningham and his colleagues at the Center for Studying Health System Change in Washington D.C., asked 10,881 doctors how much free or charity care they delivered in 1996 and 1997. Doctors who got 85 percent or more of their income from managed care plans spent, on average, 5.2 hours a month caring for indigent patients. In contrast, doctors deriving no income from managed care spent on average 10 hours a month providing care to indigent patients. In addition, the researchers found that, collectively, doctors in communities with high managed care penetration provided about 25 percent less charity care than doctors in communities with low managed care penetration.51

PROJECTIONS

As intimated throughout this paper, health insurance coverage is likely to erode further in the absence of government action. A recession or economic downturn is virtually certain to exacerbate the decline. In addition to possible layoffs and “downsizing” on the part of both large and small businesses, some small businesses could choose to drop health insurance in order to maintain wages, increase profit margins, or just to stay in business as sales decline. Most businesses would be more likely, however, not to drop coverage but to pass on more of the costs of health insurance to employees. As described above, that would compel some employees to make a rational economic choice to decline coverage. Still other small businesses (and perhaps some large ones) may choose to hire fewer full-time workers and to use part-time, temporary, or contract workers instead.

All that said, the precise effect of any future recession or economic downturn on health insurance coverage is impossible to predict. The magnitude and scope of the impact depends on too many factors – not the least of which is the depth of any recession. Nevertheless, the experience of the past decade, as discussed above, indicates that certain things would happen. In what follows, we make a set of assumptions about medical and health insurance costs and dynamics, and the impact of a recession or significant economic downturn. We then calculate the possible effect on the number of uninsured over the period 1999 to 2008 if the economy were to undergo some temporary turmoil, or longer lasting slowdown. The assumptions are these:

  1. The underlying erosion rate in health insurance prevails, with an average of 1.1 million people losing coverage every year between 1999 and 2008.52
  2. A recession or economic downturn/slowdown causes an increase in the unemployment rate of between 0.5 and 2 percentage points.
  3. The price of health insurance continues to rise at two to four times the rate of inflation over the period 1999 to 2008, due to dynamics in the marketplace.
  4. The trend of employers shifting costs to workers and using more temporary and part- time workers continues over the period (exacerbated by a possible recession).Taking the unemployment rate first, we use Thorpe’s calculations and logic (presented previously) to predict that an additional 1 million to 4 million Americans could lose coverage in the event of a recession or economic downturn due to the “unemployment effect” alone. Clearly, a 10-year recession is highly unlikely (though we’d point out that in 1988 few people would have predicted a 10-year sustained economic boom). But two to four years of downturn or stagnant economic growth could erode coverage among the most vulnerable populations for relatively long periods as unemployment remained elevated. We estimate that half those who lose coverage could go without insurance for years – essentially raising the underlying rate of health insurance erosion from 1.1 million a year to between 1,150,000 and 1,300,000 a year from 1999 to 2008.The effect of the rising price of health insurance is more complex. Premium hikes are unlikely to be slowed by a recession or downturn in the economy. Demand for health care does not abate. And while competition in the marketplace is much greater today than in years past, it is unlikely to produce significantly lower underlying costs for medical care even in a recession. Over the long term, such costs are the primary driver of premiums.The Health Care Financing Administration recently issued projections for the rise in health insurance premiums, pegging private sector increases at an average 7 percent to 8 percent per year over the next decade.53 That is two to three times the rate at which premiums rose between 1993 and 1998, a period in which the number of uninsured increased by 4.1 million from 39.3 million to 43.4 million. Recent surveys indicate that premiums are increasing 5 percent to 10 percent from 1998 levels for the 1999 coverage year. Many businesses are expecting similar increases in 1999 for the 2000 coverage year.Studies in the past have attempted to gauge the impact of price increases on health insurance coverage, but this is far from an exact science. In periods when health insurance prices have risen sharply, such as in the late 1980s, employment-based coverage eroded. But the magnitude of the decline has not always tracked the magnitude of the increase in the price of health insurance. That said, many experts believe that for every 1 percent increase in average and aggregate health insurance premiums nationwide, there is a theoretical 0.1 percent to 0.4 percent reduction in the number of employees (and their dependents) with health insurance coverage.54Thus in a population of 165 million with employer-based coverage (1997), from 165,000 to 660,000 people could lose coverage for every 1 percent increase in the aggregate price of health insurance. This theoretical effect is more likely to occur when premiums rise well above the general inflation rate. That is because most employers and employees tolerate some increase in their premium rate at or near the level of general inflation. But premium increases are expected to exceed the inflation rate over the next decade, as cited above. Even if the price effect is at the low end of what the theory predicts (i.e., 0.1 percent loss of coverage for every 1 percent increase in prices), the toll could mount significantly over time in a period of sustained premium increases above the level of inflation. If the predicted average 7 percent per year increase in premiums prevails between 1999 and 2007, we calculate that an additional 377,000 people each year could lose coverage. The calculation assumes a general inflation rate averaging 3 percent a year and that therefore 0.4 percent of people with employment-based and individually purchased insurance coverage (188.5 million people in 1997) would lose coverage. The calculation also conservatively assumes that half would find alternative coverage and that half would remain uninsured. (Note that the calculation does not adjust for an increase in the population from 1999 to 2008.)Finally, a combined effect of an economic downturn and rising prices sustained over many years would have a third (and independent or additive) effect. We project that 10 percent of small businesses (fewer than 20 workers) that currently offer health insurance would choose to either drop coverage altogether, shift costs to employees, or use more part-time or temporary workers. If so, an estimated 400,000 to 1.3 million additional people could lose their health insurance over the period 1999 to 2008.This calculation is based on U.S. Bureau of the Census and Department of Labor data and several studies referenced in this paper. Roughly 4,572,000 firms in the United States have fewer than 20 employees. They employ 27 million people, an average of 6 per firm. About half the firms offer coverage and an estimated 42 percent of the 27 million workers are covered, or some 11.3 million people. If 5 percent of firms opted to drop coverage altogether, 685,800 would lose coverage. If another 5 percent shifted costs to workers, an estimated 25 percent of workers would be at risk of declining coverage, or 171,450 workers. Total: 857,250 people losing coverage. The range of the effect, depending on the mix of changes, is 400,000 to 1.3 million people.To summarize our projections of the combined effect of a period of elevated inflation in health insurance premiums and an economic recession or downturn occurring at some point in the next decade:
    • An underlying increase in the number of uninsured averaging 1.1 million per year
    • A 0.5 to 2 percentage point increase in the unemployment rate producing from 1 to 4 million additional insured persons – with half that number remaining uninsured for the period 1999 to 2008. This would raise the underlying health insurance erosion rate from 1.1 million a year to between 1,150,000 and 1,300,000 a year from 1999 to 2008.
    • 377,000 each year losing coverage due to a “price effect” alone
    • 400,000 to 1.3 million losing coverage between 1999 and 2008 as small businesses drop coverage, shift costs to workers, or turn to contingent or part-time workers (an average of 40,000 to 130,000 a year).
    Trending these projections over the next decade, an estimated average of 1.6 million to 1.8 million people per year could lose health insurance coverage over the next decade in the event of an economic downturn and a period of sustained premium increases. That would add 16 million to 18 million people to the ranks of the uninsured by the year 2009 – bringing the number of uninsured to between 59.4 million and 61.4 million in that year. The latter figure would constitute between 22.5 percent and 25 percent of non-elderly people in 2009. (The calculation is based on Census Bureau projections of between 244 and 272 million non-elderly people in the United States in the year 2010.)Three recent studies project similar scenarios. One, a study funded by the AFL-CIO and conducted by The Lewin Group, concluded that the trend to higher employee contributions and “other economic factors” could result in 8.1 million people losing employer-based coverage by 2002. That would yield 51.5 million uninsured people in 2002, on a par with our own calculations. The economic factors cited in the Lewin study include a projected rise in health spending averaging twice the rate of general inflation. In addition, the study assumes that premium contributions by employees will rise at about 10 percent annually over the next several years.55The second study, funded by the Health Insurance Association of America (HIAA) and conducted by William S. Custer, estimates that the number of uninsured could reach 60 million by 2007 if “the nation experiences an economic downturn and higher-than-predicted health care cost inflation.” The study’s methodology is not detailed.56 A third study, by Richard Kronick and Todd Gilmer, estimates that if health spending per capita (including the price of health insurance) were to increase at an average 5.5 percent a year over the next 10 years, the percentage of uninsured workers will rise by 0.4 percentage points a year, or 4 percentage points over 10 years.57 Applying that calculation to the private-sector workforce, the number of uninsured workers and their dependents would increase from 33.5 million in 1997 (18.6 percent of 180 million private-sector workers and their family members) to 45.2 million in 2008 (22.6 percent of an estimated 200 million private-sector workers and their family members). Applying the calculation to the total population would mean 1.8 million people a year become uninsured, or 18 million by 2009 – yielding 61.4 million uninsured people by 2009.But what will happen if there is no recession and/or if the predicted rise in health care prices fails to materialize? The status of health insurance coverage in the United States will still not be good; the number of people without coverage will continue to increase over the next decade, albeit more slowly. A best case scenario is that the 1.1 million average annual increase in the number of uninsured falls by 20 percent to 30 percent. If it falls 20 percent, 8.8 million more Americans will be uninsured in the year 2009 than are uninsured today. In that case, 52.2 million people would lack health insurance in the United States in 2009. If the 1.1 million average rate is sustained over the next decade, 54.4 million will be uninsured in 2009 – 22.3 percent of the non-elderly population.



    CONCLUSION

    The accelerating decline in health insurance coverage in the United States is a serious problem, affecting the financial security and health of millions of Americans every day. Rapid changes in the economy and in the health care and health insurance marketplaces are increasing the risks that low- and middle-income working families will become uninsured or suffer lapses in health insurance coverage. Despite strong economic growth and low unemployment, employer-sponsored health insurance coverage has continued to erode throughout the past decade. The reasons are many and the conclusion is clear: Employment and health insurance is increasingly a connection at risk.An economic downturn or recession in the next few years would very likely exacerbate the decline in employer-based coverage. Our projections indicate that from 16 million to 18 million people could be added to the ranks of the uninsured by the year 2009. In that year, as many as 61.4 million Americans could be without the security of health insurance, 25 percent of all Americans under 65.In the absence of government intervention, the erosion of employer-based health insurance coverage will likely continue irrespective of macroeconomic conditions. Steps can and should be taken, beginning in 1999, to ensure that the number of uninsured Americans does not increase and is reduced over time. Those steps should be consistent with a plan to move towards a health system in which every American has health insurance coverage. They also should be consistent with the need to restrain the growth in health care costs and improve the quality of care.The four major problems plaguing our health system today – mounting costs, the growing number of uninsured, poor quality care, and the expanding gap between care available to affluent versus lower-income Americans – are deeply interrelated. Efforts to expand health insurance coverage are made much more difficult by rising health care prices. Thus our nation will have a better chance of expanding coverage if costs can be contained. And we will have a better chance of reducing health costs if all Americans have health insurance. That is because reduced access to care among the uninsured increases health care costs in the long run; the uninsured do not get the routine preventive care they need, and they tend to seek medical attention when they are sicker. Likewise, we can never have the highest quality of care possible if we have a system in which tens of millions of people lack coverage and are increasingly unable to afford medical care if they do not have health insurance.It is thus a vicious circle. A decade of research and experience now compellingly show us that a continually rising number of uninsured people will further destabilize an already strained and faltering health care system. We continue to ignore this problem at our peril.The authors gratefully acknowledge the assistance of Kenneth E. Thorpe, Ph.D., Director of the Institute for Health Services Research at Tulane University Medical Center, in the preparation of this paper.



    NOTES
    1. U.S. Bureau of the Census. Health Insurance Coverage: 1997. (September 1998). Data from the March 1998 Current Population Survey.
    2. U.S. Bureau of the Census. Dynamics of Economic Well Being: Health Insurance 1993 to 1995, Who Loses Coverage and for How Long. (August 1998). Data from the Survey of Income and Program Participation. The calculation here is made on the basis of the current U.S. population of 270 million.
    3. U.S. Bureau of the Census as cited in note 2. Data are for 1993-1995, but the statistic is likely to be applicable to more recent years as well. Also see: Kuttner, Robert. “The American Health Care System: Health Insurance Coverage.” New England Journal of Medicine 340:2 (January 14, 1999).
    4. The Lewin Group. Exploring the Determinants of Employer Health Insurance Coverage. Report issued by the AFL-CIO, (February, 1998).
    5. Custer, W.S. Health Insurance Coverage and the Uninsured. A report issued by the Health Insurance Association of America. (December 1998).
    6. Kronick, R., and Gilmer, T. “Explaining the Decline in Health Insurance Coverage, 1979-1995.” Health Affairs (March/April 1999).
    7. U.S. Bureau of the Census as cited in note 1.
    8. Levan, R., Brown, E.R., Lara, L, and Wyn, R. Nearly One-Fifth of Urban Americans Lack Health Insurance. UCLA Center for Health Policy Research. (December 1998).
    9. U.S. Bureau of the Census as cited in note 1.
    10. Guyer, J., and Mann, C. Employed But Not Insured: A State-by-State Analysis of the Number of Low-Income Working Parents Who Lack Health Insurance. Center on Budget and Policy Priorities. Washington, D.C. (March 1999).
    11. Employee Benefits Research Institute (EBRI). Sources of Health Insurance and Characteristics of the Uninsured. (December 1998). Data are from the Census Bureau’s March 1998 Current Population Survey. Data gathered before 1995 are generally acknowledged to have undercounted the number of people with employment-based coverage, because of the way the survey was conducted and the questions asked. Thus the magnitude of the decline in such coverage from the late 1980s to the late 1990s may in reality be larger than reflected in these numbers. See also: (a) Fronstin, P. and Synder, S., “An Examination of the Decline in Employer-Based Health Insurance.” Inquiry (Winter 1996-97); (b) Employer-Based Health Insurance: Cost Increase and Family Coverage Decreases. Report to Congress from the U.S. General Accounting Office, (February 1997) (c) Kuttner, R., “The American Health Care System: Health Insurance Coverage.” New England Journal of Medicine 340:2 (January 14, 1999). (d) Rice, T. Gabel, J. et al., Trends in Job-Based Health Insurance Coverage, Report from the Henry J. Kaiser Family Foundation. ( June 1998).
    12. Employee Benefits Research Institute; Data obtained on request.
    13. Gabel, J. et al. Health Benefits in 1998 for Small Employers. A report to the Henry J. Kaiser Family Foundation (February 1999). See also: Gabel, J., Ginsburg, P. and Hunt, K. A. “Small Employers and Their Health Benefits, 1988-1996, An Awkward Adolescence.” Health Affairs (September/October 1997).
    14. Cooper, P.F. and Schone, B.S. “More Offers, Fewer Takers for Employment-Based Health Insurance: 1987 and 1996.” Health Affairs (November/December 1997).
    15. Gabel, J. et al. Health Benefits in 1998 for Small Employers. A report from the Henry J. Kaiser Family Foundation (February 1999). See also: Ginsburg, Paul B., Gabel, J. and Hunt, Kelly A. “Tracking Small-Firm Coverage 1989-1996.” Health Affairs (January/February 1998).
    16. Carrasquillo, O., Himmelstein, D.U., Woolhandler, S., and Bor, D.H. “A Reappraisal of Private Employers’ Role in Providing Health Insurance.” New England Journal of Medicine 340:2 (January 14, 1999).
    17. Carrasquillo, O., as cited in note 16. Some experts have questioned the analysis in this paper. They suggest it lumps coverage through government as an employer with government as a provider of public insurance programs (Medicare and Medicaid). They note the two are very different. When government provides coverage as an employer, it generally behaves in the same manner as private employers and provides comparable benefits. The authors of the study concur that the two types of coverage exist in very different frameworks. But they note that government employers, whether federal, state, or local, are not as free to drop health insurance coverage as private sector employers.
    18. Thorpe, K. The Rising Number of Uninsured Workers: An Approaching Crisis in Health Care Financing. A report issued by The National Coalition on Health Care, Washington, D.C. (October 1997).
    19. Figures 4-11 were prepared by Kenneth E. Thorpe, Ph.D., Director of the Institute for Health Services Research at Tulane University Medical Center. They are derived from Profiles of the Uninsured in the Workplace, an unpublished paper prepared for the National Coalition on Health Care in 1998.
    20. Thorpe, K. “Why Are Workers Uninsured? Employer-Sponsored Health Insurance in 1997.” Health Affairs (March/April 1999).
    21. Thorpe, K. as cited in note 20.
    22. Meyer, J. A., and Naughton, D.H. Assessing Business Attitudes on Health Care. Report issued by the Economic and Social Research Institute (October 1996); see also Gabel, J. et al as cited in note 13.
    23. Thorpe, K. as cited in note 20.
    24. General Accounting Office (March 1998) Health Insurance Standards: New Federal Law Creates Challenges for Consumers, Insurers, Regulators. See also: Kuttner, R., “The American Health Care System: Health Insurance Coverage.” New England Journal of Medicine 340:2 (January 14, 1999).
    25. U.S. Bureau of the Census as cited in note 1.
    26. Levan et al. as cited in note 8.
    27. Hall, A., Collins, K.S., and Glied. S. Employer-Sponsored Health Insurance: Implications for Minority Workers. A Commonwealth Fund report. (February 1999).
    28. Greenberg, M. Participation in Welfare and Medicaid Enrollment. Report issued by the Kaiser Commission on Medicaid and the Uninsured. (September 1998).
    29. Ellwood, M.R., and Ku, L. “Welfare and Immigration Reforms: Unintended Side Effects for Medicaid.” Health Affairs (May/June 1998).
    30. Tallon, J.R. The Uninsured in New York. Report issued by the United Hospital Fund Conference on the Uninsured in New York State (June 1998).
    31. As cited by June O’Neil, former director of the Congressional Budget Office in an address before the American Enterprise Institute (January 7, 1999).
    32. Pear, R. “Poor Workers Lose Medicaid Coverage Despite Eligibility.” New York Times (April 11, 1999):1. The article gives the following rundown of the decline in the number of people on Medicaid since the 1996 welfare law was enacted: New Jersey (8 percent); Oregon (11 percent); Texas (15 percent); Pennsylvania (13 percent); California (8 percent); Florida (4 percent). The percentages do not separate out coverage decreases due to favorable economic conditions or factors other than the welfare overhaul.
    33. U.S. Bureau of the Census as cited in note 1. See also EBRI as cited in note 11.
    34. Marsteller, J., Nichols, L. and Zuckerman, S. Variations in the Uninsured: State and County Level Analyses. Washington, D.C. Urban Institute (June 1998).
    35. U.S. Bureau of the Census. Dynamics of Economic Well Being: Health Insurance, 1993-1995. (August 1998).
    36. Shearer, G. Hidden from View: The Growing Burden of Health Care Costs. A Consumer’s Union report. (January 1998).
    37. Kronick, R. and Gilmer, T. “Explaining the Decline in Health Insurance Coverage, 1979-1995.” Health Affairs (March/April 1999).
    38. Kronick and Gilmer as cited in note 37.
    39. Chollet, D. J., and Kirk, A. M. Understanding Individual Health Insurance Markets. Report for The Henry J. Kaiser Family Foundation. (March 1998). See also: (a) Gabel, J., Hunt, K. and Kim, J.K. The Financial Burden of Self-Paid Health Insurance on the Poor and Near-Poor. Report to The Commonwealth Fund (March 1998). (b) Private Health Insurance: Millions Relying on Individual Market Face Cost and Coverage Trade-Offs. Report to Congress from the U.S. General Accounting Office. (November 1996).
    40. Wellner, A. “Young Uninsured.” American Demographics (February 1999). This article is based in part on research by Cooper and Shone as cited in note 14.
    41. Glied, S., and Stabile, M. “Covering Older Americans: Forecast for the Next Decade.” Health Affairs (January/February 1999).
    42. Quinn, J. Retirement Patterns and Bridge Jobs in the 1990s, as cited in Demography Is Not Destiny. A report by the National Academy on An Aging Society. (January 1999).
    43. 1998 “Green Book,” Committee on Ways and Means, U.S. House of Representatives. (May 1998).
    44. National Academy on An Aging Society. Demography is Not Destiny (January 1999).
    45. Demography is Not Destiny as cited in note 44.
    46. Glied and Stabile as cited in note 41.
    47. Mercer/Foster Higgins National Survey of Employer-sponsored Health Plans. William M. Mercer, Inc. (January 1999).
    48. U.S. General Accounting Office. Private Health Insurance: Employer Coverage Trends Signal Possible Decline in Access for 55- to 64-Year-Olds. Testimony before Congress. (June 25, 1998).
    49. Chollet and Kirk as cited in note 39.
    50. The Kaiser Commission on Medicaid and the Uninsured. Uninsured in America Chart Book. (June 1998). Based on The Kaiser/Commonwealth 1997 National Survey of Health Insurance. (December 1997); See also: (a) Schoen, C. and Puleo, E. “Low Income Working Families at Risk: Uninsured and Underserved.” Journal of Urban Health (March 1998); (b) Cunningham, P.J., and Kemper, P. “Ability to Obtain Medical Care for the Uninsured.” Journal of the American Medical Association (1998), 310: 921-927.
    51. Cunningham, P.J. et al. “Managed Care and Physician’s Provision of Charity Care.” Journal of the American Medical Association 281:12 (March 24/31, 1999).
    52. The average of 1.1 million a year is derived from EBRI data for 1987 to 1997. See Figure 1, page 2.
    53. Smith, S. et al. “Health Spending, The Next Decade.” Health Affairs (September/October 1998). See also: (a) Thorpe, K. E. Changes in the Growth in Health Care Spending: Implications for Consumers. A report by the National Coalition on Health Care (April 1997) and (b) Iglehart, J. K. “The American Health Care System: Expenditures.” New England Journal of Medicine 340:1 (January 7, 1999).
    54. There are references to this model in the literature. But it has also been a central focus of the recent debate over proposed federal legislation to regulate managed care. Both advocates and opponents of such legislation agree that requiring health insurers and managed care plans to provide additional coverage or carry out additional administrative tasks will likely increase the price of health insurance over time, and that this does have the potential to undermine employer-based coverage. But there is broad and deep disagreement about the magnitude of such an affect, and conflicting studies on the issue. See: Exploring the Determinants of Employer Health Insurance Coverage. Report issued by the AFL-CIO. (February 1998). Also used as a source: Letter to the American Association of Health Plans from John F. Sheils, The Lewin Group (November 17, 1997).
    55. Exploring the Determinants of Employer Health Insurance Coverage. Report issued by the AFL-CIO. (February 1998). As cited in note 4.
    56. Custer, W.S. Health Insurance Coverage and the Uninsured. Report issued by the Health Insurance Association of America. (December 1998). This study also discusses the impact on health premiums and insurance coverage of state-imposed benefit mandates and insurance reforms. It concludes that such reforms (especially small-group community rating combined with a guaranteed issue requirement) increase the probability that an individual will be uninsured by 15.8 percent.
    57. Kronick and Gilmer as cited in note 37.

More from releases