I. The Misunderstood Connection between Poverty and Health

Advocates for universal healthcare often cite the connection between health and income as a reason for expanding access to care.^{Reference DeVol and Bedroussian1} According to a common view, good health is a prerequisite to prosperity. And indeed, there is a critical connection between health and income. In the United States, men in the top 1% of the income distribution have a life expectancy 14.5 years longer than men in the bottom 1%.^{Reference Chetty2} The income disparity in life expectancy exists even within small geographic areas. In Cincinnati, Indianapolis, and Louisville, residents in the top 25% of the income distribution live about 8.5 years longer than do residents in the bottom 25%.³

Under the access-to-care view of health and income, if people get sick and lack insurance, they do not receive adequate healthcare. As a result, they stay sick, cannot work, and their income drops. The Affordable Care Act (ACA) addresses the connection between health and income directly. With the ACA's guarantee of coverage, even for pre-existing conditions, when people get sick, they can receive good healthcare, and their health will improve. As a result, they are able to work, and their income will rise. According to this common access-to-care view, healthcare leads to good health, which in turn leads to a good income.

There is important truth to the paradigm of poor health causing poverty. For example, researchers have documented the problem of medical bankruptcy — before the ACA's enactment, most Americans who declared bankruptcy cited illness or medical bills as a cause for their bankruptcy.^{Reference Himmelstein4} Or people who develop a mental illness may no longer be able to work, losing their ability to earn an income.^{Reference Glymour and Berkman5} In underdeveloped economies, untreated infectious diseases such as malaria and hookworm can trap people in a condition of persistent poverty.^{Reference Plucinski6}

But there are two problems with the common understanding that access to healthcare leads to a good income by ensuring good health. First, healthcare has a smaller effect on health than do other factors like genetics, diet, education, and environment.^{Reference Schroeder7} Indeed, the life expectancy of lower-income persons is not correlated with the amount of healthcare that they receive, nor do their mortality rates change when they become eligible for Medicare.⁸ Second, income probably has a greater effect on health than health on income.

To be sure, some important studies find that access to care improves health. For example, researchers studied three states that expanded their Medicaid programs between 2000 and 2005 to cover “childless adults.”^{Reference Orentlicher9} When compared with demographically-similar, neighboring states that did not extend coverage to childless adults, expansion states saw a significant reduction in mortality rates, 6.1%.^{Reference Sommers10}

Similarly, Massachusetts experienced significant health benefits following its 2006 healthcare reform that was similar in design to the Affordable Care Act. Researchers compared rates of death in 14 Massachusetts counties with mortality rates in 513 demographically similar counties in 46 other states. Rates of death from all causes were 2.9% percent lower in the 14 Massachusetts counties. Moreover, those counties saw a 4.5% drop in “health-care amenable mortality,” or mortality that results from illnesses for which good healthcare can prevent death.^{Reference Sommers11}

But other studies do not find that access to care improves health. Just as mortality rates for the poor do not change when they become eligible for Medicare, Medicare's first ten years of operation did not lead to a discernible impact on death rates of seniors.^{Reference Finkelstein and McKnight12}

While the early impact of Medicare might have been limited because it predated important advances in medical care,¹³ such as statins to lower cholesterol, a study on Medicare's later impact came to a similar conclusion. Researchers followed for 18 years the healthcare use and health outcomes of people nationwide who were age 50–61 in 1992. They found that insured individuals used more healthcare resources than uninsured persons, but saw no evidence that being insured lowered the risk of death 14 years into the study and only mild evidence of a mortality benefit 16–20 years later.^{Reference Black14}

Consider too a study that measured the impact of healthcare on health after an Oregon lottery for uninsured persons gave lottery winners access to Medicaid coverage. By using a lottery to decide access, Oregon effectively created a randomized controlled study of the benefits of healthcare coverage. Winners enrolled in Medicaid could be compared with applicants who lost the lottery and remained uninsured. Two years after Medicaid enrollment, lottery winners received more healthcare than their uninsured counterparts, and reported less financial strain, but they experienced no reduction in three key measures: hypertension, high cholesterol, or diabetes.^{Reference Baicker15}

II. The Link from Wealth to Health

While it is not clear how much increasing access to healthcare improves health, a number of studies have found that higher incomes lead to better health. For example, in a study of 24 countries in the Organization for Economic Co-operation and Development, increases in the minimum wage were associated with decreased mortality rates, in part because higher wages are associated with reductions in smoking.^{Reference Lenhart16}

The health benefits of income are especially important for children. Parental income plays a key role in determining a child's health even into adulthood.^{Reference Case17} Consider here the impact of the Mothers' Pension program, the first U.S. government-sponsored welfare program for poor single mothers with dependent children, which ran from 1911 to 1935.^{Reference Aizer18} Because the program preceded Medicaid, housing benefits, and food stamps, its impact can be assessed in isolation rather than in combination with other programs for the poor.¹⁹ When women received Mothers' Pension benefits, their sons had an extra year of life expectancy on average, with boys in the poorest families gaining 1.5 years longevity. The sons also were less likely to be underweight, they had more years of formal education, and they earned higher incomes as adults.²⁰ In terms of the impact on life expectancy, the education and income effects appeared to be more important than the nutritional effects of the maternal pension benefits.²¹

Improvements in socioeconomic status have led to improvements in health status outside the U.S. as well. Case-in-point: Canada's interesting policy experiment during the 1970's. For four years, the province of Manitoba guaranteed a minimum annual income for all residents of Dauphin, a small, rural city.^{Reference Forget22} At any one time, only about a third of the city's residents received payments from the provincial government, and for many of those people, the payments were small. Nevertheless, everyone benefited from the “income security” — everyone knew that whatever happened to their income stream, they could expect a basic annual income. This security was particularly valuable in Dauphin, an agricultural community where incomes depended on harvest sizes and commodity prices that could vary substantially from year to year.²³ Thus, for example, farming families could be more comfortable letting their older children remain in school rather than work at home once their children's labor was not necessary for the family's financial well-being.^24,

With the income security from the Manitoba policy, health status improved significantly. Before implementation of the minimum income program, people in Dauphin were more likely to be hospitalized than residents of other rural Manitoba communities; that gap in hospitalization rates disappeared by the program's end.²⁵ This decline largely occurred for the kinds of hospitalizations that tend to be sensitive to levels of income security.^{Reference Forget26} For example, accidents and injuries are more common for people who worry about earning an adequate income.²⁷ Income insecurity makes people more likely to continue working when fatigued, to increase alcohol intake, or not to hire caretakers who can protect their children from accidents. Just as overall hospitalization rates were higher in Dauphin before the minimum income program, so were hospitalizations for accidents and injury — and both gaps disappeared during the program.²⁸ The same patterns were seen with hospitalizations for anxiety disorders, depression, and other mental health conditions.²⁹ In sum, providing income security to Dauphiners made them less likely to need hospital care.

Note that the reduction in hospitalizations cannot be attributed to better access to healthcare insurance. One might wonder whether providing a minimum income made it more possible for poor residents of Dauphin to purchase healthcare coverage and therefore to obtain medical care early in their illnesses when outpatient care alone could treat their conditions. However, Manitoba had implemented a program of universal health insurance before the minimum income experiment,³⁰ so even the poor had good access to health care. And for accidents and injuries, having healthcare insurance either would not affect rates of hospitalization or should increase rather than decrease hospital admissions.

Income supplement programs have been successful in the United States and other countries at improving health, too. In addition to the Mothers' Pension program, a good U.S. example is the earned income tax credit (EITC), an income supplement for low-wage working parents. Studies have found that the federal and state EITCs reduce infant mortality^{Reference Arno31} and increase the health of newborns. In a study of the New York State and New York City EITCs, for example, researchers found that the income supplements led to a significant drop in low-birthweight newborns.^{Reference Wicks-Lim and Arno32} Another study of the federal EITC also found that it led to a drop in low birthweight, particularly among African-American children.^{Reference Haynes33} The EITC provides other benefits for health. For example, increases in the EITC have led to decreases in maternal smoking.^{Reference Averett and Wang34}

Other data also suggest that the connection between income and health is more important than the connection between healthcare and health. When poor elderly persons in Mexico's Yucatan were given an income supplement of $67 per month (31% of minimum wage), their improvements in health measures gave them health measures comparable to neighbors who were 5-10 years younger.^{Reference Aguila35}

Researchers have studied health outcomes in England under that country's universal National Health Service (NHS). Evidence indicates that socioeconomic status is much more important than access to healthcare for one's health. For example, the higher the socioeconomic status of a person, the lower the mortality rate. People in the highest civil service grade for government employees have a mortality rate about half that of people in the lowest civil service grade, even though they all have good access to healthcare.^{Reference Marmot, Berkman and Kawachi36} Moreover, the gap in mortality rates among men in England by socioeconomic status has actually widened since the introduction of the NHS in 1948.^{Reference Sreenivasan, Black, Townsend, Whitehead and Townsend37}

Neglecting the connection between income and health can cause us to blame the victims rather than trying to fix their problems. For example, as childhood obesity has risen, observers often point to decreases in children's physical activity and increases in “screen time.” In one view, if children would exercise more, they would be more physically fit.^{Reference Baidal and Taveras38} Yet studies looking at children who walk or bicycle to school instead of being driven find that the walkers and bikers have the same or higher rates of obesity as those who are driven. Why does this counterintuitive relationship exist? Poor children are more likely to walk or bike to school than children in higher-income families, and while many factors affect health, poverty is the strongest predictor of ill health.^{Reference Chaufan39}

III. Addressing Poverty Rather Than Medicalizing It

If income is more important than healthcare for health, this tells us that poverty is a major public health problem. Just as the medical profession has recognized gun violence as a public health problem,^{Reference Mendenhall40} it needs to recognize poverty as a public health problem. If society wants to promote better health for the poor, it would have a bigger impact through a greater commitment to anti-poverty programs than through health insurance programs such as Medicaid or the ACA. A few policies seem particularly warranted:

1. 1. Adopt a guaranteed minimum income. As Manitoba found when it adopted a minimum income policy for Dauphin, the income guarantee improved the health of the town's residents even though they already had access to healthcare coverage. A guaranteed minimum income would provide other benefits. For example, income supplements have been found to improve educational achievement of children,^{Reference Dahl and Lochner41} and providing a decent minimum income would allow for a single government program for the needy rather than a combination of food, housing, and income support programs. Public health insurance programs still would be needed — as recognized by the ACA, even many middle-income Americans cannot afford health care coverage without subsidies.

   Many people worry that a guaranteed minimum income would discourage some recipients from working, but the experience with income support programs suggest that this would not be a significant problem. Experiments with guaranteed minimum incomes in the United States in the 1960s and 1970s found that to the extent the policies reduced workforce participation, they did so primarily by encouraging mothers to spend more time at home with their infant children or encouraging high school students to extend their education before entering the workforce.⁴²

   A guaranteed minimum income would be the most straightforward reform, and probably the most effective. The next two reforms would be less valuable, but potentially more achievable politically since they build on current anti-poverty policies.

2. 2. Increase the minimum wage. As mentioned earlier, when countries increase their minimum wage, people's health improves. While the United States has had a federal minimum wage for decades, it has not kept up with inflation. For example, the $1.60 minimum wage in 1968 is worth $11.64 today,⁴³ but the minimum wage today sits at only $7.25. Raising the minimum wage to at least $11.64 and indexing it for inflation would be good for the public's health.

3. 3. Improve the EITC. As noted, the EITC has resulted in better health in the United States. However, eligibility is limited to working adults with children. Accordingly, many policy experts advocate for an expansion of the EITC's eligibility criteria to include childless adults.

   In addition to extending the EITC to more of the poor, it would be important to raise the amount of the EITC. That directly results in higher incomes for the poor. Further, combining an EITC increase with an increase in the minimum wage addresses an important concern with minimum wage increases — the possibility of job loss. While economists disagree on the extent to which increases in the minimum wage lead to job loss, a higher minimum wage will, at some point, cause some employers to reduce the number of employees to compensate for their increased labor costs. With increases in the EITC, the income of the poor can be augmented without undermining their job opportunities.

   There also is an element of fairness from using both the EITC and the minimum wage to help the poor. While employers have a responsibility to pay a decent wage, the public as a whole has a responsibility to address the problem of poverty. Because the EITC is funded by all taxpayers, increases in the EITC are much more effective than increases in the minimum wage at spreading widely the financing costs of anti-poverty policy.

4. 4. Provide housing in higher-income neighborhoods for low-income families. As discussed, the health of children suffers when their families are poor. Childhood health also suffers when kids live in neighborhoods that have high rates of poverty and other indicia of disadvantage, such as low-performing schools, high crime rates, and environments with high levels of lead and other harmful substances.⁴⁴ But housing programs have traditionally provided homes for poor families in lower-income communities. Providing homes for poor families in higher-income neighborhoods instead would do much to promote the health of children — lower-income children have longer life expectancies if they live in communities with highly educated populations, high incomes, and high levels of government expenditures.^{Reference Orentlicher45} This likely reflects the fact that poor children have brighter economic prospects if they reside in more affluent neighborhoods.^{Reference Chetty and Hendren46} Indeed, the economic status of a child's neighborhood is more important for the child's economic future than is the economic status of the child's family.^{Reference Sharkey and Graham47}