Cookies Notification

This site uses cookies. By continuing to browse the site you are agreeing to our use of cookies. Find out more.


Children's Health

What We Say And What We Do: Why US Investments In Children’s Health Are Falling Short

  1. Janet Currie ([email protected]) is the Henry Putnam Professor of Economics and Public Affairs, Princeton School of Public and International Affairs, Princeton University, in Princeton, New Jersey.
PUBLISHED:Free Accesshttps://doi.org/10.1377/hlthaff.2020.00633


This article explores the gap between what Americans say we believe with respect to spending on child health and what we actually do, which falls short of norms in other developed countries. Three possible reasons are identified: a lack of information about the effectiveness of government programs affecting child health, the idea that “investments” in child health should pay for themselves, and ideological preferences that prioritize other goals and that may themselves be rooted in this country’s racial history. These factors are not mutually exclusive, and all may be at play simultaneously.


In 1990 more than 90 percent of Americans polled in the General Social Survey said that the government should maintain or increase spending on programs providing health care and nutrition to poor children.1 This percentage has changed little over time: In 2017, 87 percent of respondents to a Kaiser Family Foundation poll said that they wanted spending on Medicaid, the public health insurance program that covers poor women and children, among others, to stay the same or increase.2

This favorable view is justified by the evidence.3 Government programs for children improve their chances of growing up to be healthy, productive adults, as discussed here. Yet US spending on families with children continues to lag that of other wealthy countries.4

This article explores some possible reasons for this perplexing and discouraging situation. First, perhaps there is an information gap—maybe, despite the evidence, the US public still believes that government is wasteful and ineffective.

Second, spending on children is often framed in terms of a narrow financial investment paradigm that asks whether a program will generate future cost savings, in the form of lower welfare payments, health care costs, and crime, or higher future tax revenues. This narrow financial investment paradigm may have backfired by suggesting that programs for children should pay for themselves. This yardstick is not applied to expensive social programs such as Medicare and Social Security.

Growing political polarization rooted in ideological differences and the nation’s racial history may be a third reason. Many commentators have discussed the rising partisan divide in roll-call voting in Congress, especially since the 1980s.5 The ideological differences leading to this divide may make it difficult to pass redistributive policies. In turn, the roots of opposition to redistributive policies from voters and legislators may lie in the nation’s racial history. In their book on redistributive policies, Alberto Alesina and Edward Glaeser examine many possible explanations for the lack of redistributive policies in the US, including different levels of openness and social mobility. They conclude that attitudes toward race are among the most important: White voters often perceive that redistribution disproportionately benefits non-White families, and this tends to undermine support for these programs.6

These reasons are not mutually exclusive, and they could be complementary. For example, ideological polarization may tend to reduce the amount of information about opposing views that people receive and accept.7

A Comparative Perspective

Poverty, measured in terms of market income from earnings, capital, and private transfers, is actually lower in the US than in the UK and is about the same as in France. What is really different about other rich countries is the extent to which their governments compensate for low market income through public transfers. The UK and France reduce child poverty to about 12 percent through transfers; in the US, child poverty is about 20 percent when measured in terms of disposable income.4 Public transfers include both cash and in-kind supports such as public housing. (The value of health insurance is difficult to determine and is not included in official poverty measures.) Consideration of these supports is relevant because poverty has a negative causal effect on children’s health.3

In addition to cash transfers, other rich countries support children’s health through policies such as paid parental leave and nurse home visits. The US is an outlier, as it is one of only two countries worldwide without a national paid maternity leave policy (the other is Papua New Guinea).8 Unlike other rich countries, the US also has no national paid paternity, parental, or family leave policies.8

By supporting mothers around the critical periods pre- and postbirth, paid maternity leave reduces low birthweight and prematurity.9 This improves the health of children, as in addition to higher infant mortality, common conditions such as asthma,10 attention deficit hyperactivity disorder,11 and autism11 have been linked to low birthweight. Maternity leave programs have also been shown to have positive long-term effects on children’s future education and wages.12

Even job-protected unpaid leave is beneficial. The Family and Medical Leave Act of 1993 reduced infant mortality by 10 percent in eligible families even though not all families could afford to take the leave.13 However, only 60 percent of US mothers are eligible for unpaid leave under the law because small businesses are exempt and some workers in nonexempt businesses are also not covered.14

Evaluations of nurse home visit programs before and after birth show that the programs reduce low birthweight and child maltreatment and improve immunizations and age-appropriate child development.15 However, although many states have significant nurse home visit initiatives, many families lack access.16 In contrast, several European countries provide free nurse home visits as part of a comprehensive maternal and child health system. For example, Denmark has had nurse home visits since 1937, and France provides universal pre- and postnatal visits by midwives and nurses.17

Why Doesn’t The US Do More?

The Information Gap

Given the evidence, why doesn’t the US do more? One possibility is that Americans are skeptical of government. Exhibits 1 and 2 show what Americans say about their confidence in the executive and legislative branches of the US government. Trends for the judicial branch are similar. In recent years, few Americans say that they have “a great deal” of confidence in either the executive or the legislative branch, and the share who say that they have “hardly any” confidence has risen.

Exhibit 1 Americans’ confidence in the executive branch of the federal government, 1973–2018

Exhibit 1
SOURCE Data from the General Social Survey, GSS variables: A) confed. NOTES Data are available from the General Social Survey Data Explorer from NORC at the University of Chicago. The question is: “As far as the people running [the executive branch of the federal government] are concerned, would you say you have a great deal of confidence, only some confidence, or hardly any confidence at all in them?” Trend lines are based on a nonparametric model.

Exhibit 2 Americans’ confidence in Congress, 1973–2018

Exhibit 2
SOURCE Data from the General Social Survey, GSS variables: B) conlegis. NOTES Data are available from the General Social Survey Data Explorer from NORC at the University of Chicago. The question is: “As far as the people running [Congress] are concerned, would you say you have a great deal of confidence, only some confidence, or hardly any confidence at all in them?” Trend lines are based on a nonparametric model.

Perhaps Americans want a safety net for children similar to the one that exists in other countries but have no confidence that their government can provide it. This skepticism is fed by public figures who openly deride public programs, making claims such as “Medicaid is worse than no insurance coverage at all.”18

This skepticism is at odds with the government’s track record of improving children’s outcomes via public programs. The National Academies of Sciences, Engineering, and Medicine report that increases in income support under the Earned Income Tax Credit have improved children’s educational and health outcomes and that the Supplemental Nutrition Assistance Program (SNAP, formerly food stamps) has improved birth outcomes and child and adult health.3

Improvements in children’s access to health care are a particularly compelling success story. Beginning in the late 1980s federal legislation first encouraged and then required states to extend public health insurance coverage to low-income pregnant women and children under Medicaid. Now, over 40 percent of all births in the US are paid for by the Medicaid program, up from about 12 percent in 1979.19 The Children’s Health Insurance Program, which began in 1998, further expanded health insurance coverage for low-income children. In 1986, 15–20 percent of US children were eligible for public health insurance,20 compared to the 40 percent of children covered in 2018.21 These gains are now threatened by changes in regulations, such as the implementation of work requirements, and by the fiscal crises that states face as a result of coronavirus disease 2019 (COVID-19).

The coverage expansions had immediate positive effects on child health.22 Among pregnant women, the 30-percentage-point increase in eligibility for public health insurance reduced infant mortality by 8.5 percent.23 The 15.1-percentage-point increase in eligibility for public health insurance among children reduced the probability that a child went without at least one annual doctor’s visit by 9.6 percentage points and reduced preventable hospitalizations.24 As many as six million children received improved access to basic preventive care.22 One way to assess the success of these measures is to look at child mortality. For example, Canadian child mortality has been lower than US child mortality for decades, but the improvements in the US have been so dramatic since the Medicaid expansions that they helped reduce the US child mortality rate toward the lower Canadian child mortality rate between 1990 and 2010.25

Expansions of public health insurance for pregnant women and children also have had long-term effects. Children born in the late 1980s and early 1990s are now young adults. Studies that follow these cohorts and compare them with children born a few years earlier who were ineligible for public health insurance or who received fewer years of coverage show that children eligible for public health insurance through maternal coverage in utero or in early childhood had higher achievement test scores and were more likely to have ever attended college than children who were not eligible.26 Eligible children are also less likely to have been hospitalized with chronic conditions27 and continued to have lower mortality rates into adulthood.28 As young adults, they work more, have higher earnings, and pay more taxes.29 These expansions have had intergenerational effects, as women who were eligible for public health insurance as young children are now having children of their own and are themselves delivering healthier babies.30

By considering the initial introduction of Medicaid in the 1960s, it is possible to study even longer-term effects. Studies of the initial Medicaid rollout show that people who were covered as children had lower rates of disability, high blood pressure, heart disease, and obesity and were more likely to be working up to fifty years later.31

Nor is Medicaid an isolated success story. The Special Supplemental Nutrition Program for Women, Infants, and Children (WIC) is one of the largest US programs aimed at improving health and nutrition in pregnancy and children younger than age five, and it has been extensively evaluated.32 WIC increases pregnancy weight gain and reduces the probability of infant prematurity and low birthweight. The gains are especially large in the most disadvantaged households.32 Children whose mothers received WIC while pregnant are also less likely to have attention deficit hyperactivity disorder and other mental health disorders that are commonly diagnosed in childhood.33

These well-documented successes have not prevented cutbacks. Between 2016 and 2018, 400,000 children lost Medicaid coverage, largely because of policies making it more difficult to stay covered, and the number of children lacking health insurance rose.34 First Focus for Children documents a decline of 3.1 percent in the share of federal spending on children’s health from 2015 to 2019.35 Because most Medicaid dollars go to the elderly and disabled even though children make up the bulk of the caseload, spending on children can fall even as overall Medicaid spending rises.22

There have been waves of WIC clinic closures in some states recently, which have harmed infant health. In Texas, having a WIC clinic in the ZIP code of residence increased birthweight for all women and especially for those with a high school education or less.36 SNAP has also seen changes in the rules for adults that have made children ineligible. And because SNAP eligibility also determines eligibility for school meals, it is estimated that 1.2 million children may lose access to SNAP and 500,000 children may lose access to school meals as a result.37

If it is primarily lack of information about the success of government programs that has undermined support for these programs, then the remedy would be straightforward. However, prior beliefs such as the ideologies discussed below can play a strong role in determining whether or not new information is accepted.38

The Investment Paradigm: Is It Paying Off?

Programs for children are often described as investments that will both reduce future public spending and increase the wages and taxes that will be paid when today’s children become adults.39 This language is strikingly absent in public discussions of other big-ticket government programs such as Medicare. We do not expect large public expenditures on Medicare to pay for themselves; instead, we make these expenditures because we value the health and well-being of the elderly.

To illustrate the difference between these two ways of thinking, exhibit 3 summarizes evidence of the effects of the expansions of public health insurance to pregnant women. The exhibit is based on work by Nathaniel Hendren and Ben Sprung-Keyser, who summarize evidence about the proven costs and benefits of a wide range of government programs (including those affecting children of all ages) in a common framework.40

Exhibit 3 Costs, benefits, and willingness to pay for Medicaid expansions for pregnant women, per birth

Cost/value categoriesValue
Cost to government (total)$ 4,033
 Initial medical costs surrounding birth2,968
 Reduced taxes from lower maternal labor supply642
 Future child college costs to government423
Benefit to government (total)12,023
 Higher taxes paid by children11,419
 Cost savings from reduced child hospitalizations604
Willingness to pay (total)81,965
 Children’s higher after-tax income53,997
 Value of child lives saved25,989
 Value of health insurance for pregnant women1,979

SOURCE Based on the welfare analysis of Nathaniel Hendren and Ben Sprung-Keyser (see note 40 in text) and the author’s calculations. NOTES All values are in 2019 dollars, and both costs and benefits are discounted to the child’s year of birth, using a discount rate of 3 percent. I used a value of a statistical life of $9.4 million (in line with the Environmental Protection Agency’s guidance, which has become standard in the health economics literature) rather than $1 million, as in the analysis by Hendren and Sprung-Keyser.

The first two sections of exhibit 3 focus on the narrow financial investment paradigm by illustrating the costs and benefits of the expansions to government. Medical expenditures are the largest cost, but they are partly offset by reductions in uncompensated care costs. The cost side also includes reductions in taxes from mothers who reduce their labor supply in the birth year because of public health insurance, as well as future public expenditures on the children’s college costs (as children whose mothers are covered are more likely to attend college).26 The largest benefits to government are the higher future taxes paid by the children of covered mothers and the children’s lower future health care expenditures.

Judged by this investment metric, the expansions of health insurance to pregnant women were a success, with financial benefits to government of about three times the cost of $4,033 per birth. However, this calculation ignores the social value of benefits that accrue to individual citizens rather than to government coffers. The third section of exhibit 3 incorporates the value of lives saved, the value of reducing uncertainty resulting from health risks, and the value that children would place on their own higher future earnings (since children are citizens, too). None of these factors is included in the narrow “costs and benefits to government” financial investments paradigm.

Clearly, the investment paradigm greatly undervalues the benefits of the program from the point of view of its value to citizens. Incorporating these benefits yields a total of $81,965, which is more than twenty times greater than the cost. This calculation of what people would be willing to pay still excludes many intangibles such as the consumption value of better health. One of Hendren and Sprung-Keyser’s conclusions is that the value of programs for children is generally much greater than a narrow financial “costs and benefits to government” calculation would suggest and is much higher than for most of the nonchild programs they reviewed.40

It may be time to move away from the narrow financial investment paradigm when weighing programs for children against other social priorities.

Even so, it may be time to move away from the narrow financial investment paradigm when weighing programs for children against other social priorities and instead consider programs on the basis of their social value to all citizens, including the children themselves.

Ideology, Polarization, Immigration, And Race

Perhaps voters and policy makers understand that programs for children pass even a narrow “benefits to government must exceed costs to government” test but still do not support them for ideological reasons. Given the growing political polarization the US has been experiencing since 1980, ideological differences may represent ever-bigger obstacles to redistributive policies over time.41

At least some of the ideologies apparent in today’s political discourse have a past connected with the nation’s racial history. For example, some leaders of the Confederacy asserted that they were fighting to maintain states’ rights rather than to defend slavery.42

Today, states’ rights are often invoked as a reason to dismantle federal antipoverty programs and return the money to the states in the form of block grants. Proposals to block grant the Medicaid program have been around for decades and have recently been put forward again by the Trump administration. States that embrace this model would have their federal Medicaid funding capped but would have increased flexibility to set standards for coverage and benefits.43

Because health care costs have been rising faster than inflation, however, even an indexed block grant would likely lead to large shortfalls in federal funding over time compared with what states receive from the federal government under the current system.22 In turn, experience shows that states facing budget shortfalls often retrench in terms of lowering Medicaid income cutoffs for women and children.22 The main argument against the “states’ rights are paramount” view underlying Medicaid block grants is that all Americans should have access to basic health care regardless of which state they live in.

Emersonian self-reliance44 is a second ideology strongly associated with the US but interpreted differently by the left and the right. On the left, commentators have focused on “making work pay” so that families can become self-reliant.45 These commentators point to the fact that most beneficiaries of programs for children such as the Earned Income Tax Credit, SNAP, and Medicaid live in families where the parents work but do not make enough to cover all necessities.3 On the right, proposals to encourage self-sufficiency involve imposing work requirements on participants in social programs. The Trump administration has tightened work requirements for SNAP and allowed states to do the same with Medicaid. Although the Medicaid rules do not target children, they do affect parents, and research has shown that when parents lose coverage, so do their children.46 Work requirements can have a dramatic effect: In the first six months of Arkansas’s work requirement program, 17,000 of about 65,000 people who were targeted lost coverage. Many people who lost Medicaid in Arkansas were actually complying with the work requirements but did not meet the reporting requirements, which involved, for example, contacting the welfare office during usual work hours.47

The value placed on the concept of self-reliance in the US is closely related to two other ideas: the idea that there are “deserving” and “undeserving” poor people, and the idea that parents are responsible for their own children, rather than children being the responsibility of the state. Robert Moffitt argues that the idea that some poor people are more deserving than others may underlie the shift of redistributive dollars toward elderly and disabled people and away from families with children during the past several decades.48 The “hands off” view of US policy makers toward the family is reflected in the fact that the US is the only country in the world never to have ratified the United Nations Convention on the Rights of the Child, in part because of concern about “parents’ rights.”49

The ideology of self-reliance also has a racial aspect. As of 2017, 49 percent of Americans endorsed the statement that “blacks who cannot get ahead in this country are mostly responsible for their own condition.”41 This is one of the most polarizing issues in the US, with 78 percent of Republicans or those who lean Republican endorsing the statement, compared to 28 percent of Democrats or those who lean Democratic.41 Americans remain polarized even in the face of current events. For example, 82 percent of Republicans or those who lean Republican felt that “some people taking advantage of the situation to engage in criminal behavior” played a large role in the protests following the death of George Floyd in May 2020, compared to 39 percent of Democrats and those who lean Democratic.50

Americans have become similarly polarized when it comes to immigration. In 2017, 84 percent of people who identified as Democrat/lean Democrat said that immigrants “strengthen[ed] the country with their hard work and talents,” up from 55 percent in 2011, compared with 42 percent of people who identified as Republican/lean Republican, up from 33 percent in 2011.41

The Trump administration has tightened immigration enforcement and tried to change rules governing when immigrants are deemed to be a “public charge.” Being named a public charge can threaten an immigrant’s path to citizenship and lead to deportation. In 2019 the Trump administration reversed previous government policy so that use of Medicaid could, in some circumstances, make someone a public charge. However, for the time being, this change has been blocked by federal courts.51

Roughly 7.6 million US citizen children have undocumented family members, and more live in families with documented family members who may want to apply for US citizenship.52 Although the changes in policy do not directly apply to US citizen children, fear that using Medicaid will result in deportation or denial of citizenship of other family members is depressing child Medicaid caseloads: The Kaiser Family Foundation has found that 28 percent of health centers report that immigrant patients have been disenrolling their children from Medicaid over the past year.53 Moreover, US citizens in families where even one member is undocumented and uses an Individual Taxpayer Identification Number (ITIN) are ineligible for relief under the Coronavirus Aid, Relief and Economic Security Act.54

The ideological divide over race and immigration disproportionately affects children because children in the US are increasingly non-White.

The ideological divide over race and immigration disproportionately affects children because, as shown in exhibit 4, children in the US are increasingly non-White. In 1980, 74 percent of children were non-Hispanic White, but by 2040, that fraction is projected to be only 42 percent. In contrast, adults, and especially older voters, will remain majority White for decades.55 This “racial generation gap” may be one of the most important reasons why proposals to devote more of the nation’s resources to supporting children and families have stalled.56

Exhibit 4 Historical and projected race/ethnicity of US children, 1980–2040

Exhibit 4
SOURCE Data drawn from National Academies of Sciences, Engineering, and Medicine. A roadmap to reducing child poverty, appendix Figure D2-1 (see note 3 in text). NOTES “Other” is primarily multiracial children. Fewer than 1 percent of children are Native American or Alaskan Native in each year. The data for 1980 to 2009 are intercensal estimates and incorporate the 1980, 1990, 2000, and 2010 censuses as benchmarks. The data for 2010 to 2016 are based on the population estimates released for July 1, 2016. Data beyond 2016 are derived from the national population projections released in December 2014.


When it comes to supporting the health of children and families, what Americans say is not what we do. This article explores three possible reasons for this state of affairs. First, it is possible that voters and policy makers simply do not believe that government can be effective in this sphere. Educating them about the tremendous success of policies such as public health insurance and WIC may increase their enthusiasm for action, although entrenched beliefs may make the job of education harder.

Second, perhaps policy makers know that these programs can be effective but are applying a narrow financial investment paradigm in which the government must save money over the long-term for a program to be worthwhile. Many policies for children, such as the Medicaid expansions, pass even this test, especially once the long-term as well as the short-term benefits are considered. However, adopting an alternative “willingness to pay” paradigm that includes the value to people as well as any savings to the Treasury is arguably more appropriate and makes the case for spending on children’s programs even stronger.

Third, perhaps lack of support for children’s programs primarily reflects ideological stances that put more value on “states’ rights,” self-reliance, and restricting immigration than on children’s health. As discussed earlier, these priorities may also reflect a racial generation gap, as older White voters might not see the importance of paying to support programs for a rising generation of non-White children. Confronting this reality may be an essential first step in garnering greater support for programs supporting child health and well-being.


Janet Currie receives financial support as a program director at the National Bureau of Economic Research; as an adviser for BRIQ in Bonn, Germany; as an affiliated faculty member of the Norwegian School of Economics; and as an adviser for the Minneapolis Federal Reserve. The author thanks Iris Arbogast for truly excellent research assistance.


Loading Comments...