You are here

Progress on Income and Poverty Continued Into 2017 But Abruptly Halted on Health Coverage, Census Data Show

October 4, 2018

Income rose and poverty declined for the third straight year in 2017 but progress on health coverage halted and apparently began to reverse, according to Census data released September 12. One Census survey found no significant change in the share of Americans without health insurance in 2017, but a larger Census survey that is more sensitive to small changes showed the uninsured rate edged up slightly.

The failure of health coverage to improve after three straight years of progress is likely due at least in part to Trump Administration actions throughout 2017 to hinder the Affordable Care Act (ACA).[1] Ongoing efforts by the Trump Administration and some in Congress to weaken various aspects of the ACA pose the risk of continued erosion of coverage gains in the years ahead.

Uninsured Rate Rises in Larger Census Survey

The share of Americans without health insurance remained unchanged at 8.8 percent in 2017, according to the Census Bureau’s main national survey, the Current Population Survey (CPS). Results from Census’ American Community Survey (ACS), however, indicate that the uninsured rate edged up last year.

The failure of health coverage to improve after three straight years of progress is likely due at least in part to Trump Administration actions throughout 2017 to hinder the Affordable Care Act.

Both surveys have strengths. The CPS is longer, has more detailed health insurance questions, and is likely to measure the uninsurance rate more accurately in any one year. At the same time, the ACS’ much larger sample size (3 million homes rather than 100,000) makes it better able to reliably detect small year-to-year changes.[2]

The stalling of progress on the uninsured rate in the 2017 CPS — and the small increase in the uninsured rate in the ACS — contrast with the three preceding years of historic declines in the uninsured rate following full implementation of the ACA in 2014. When the ACA’s eligibility rules were fully in place, millions of uninsured Americans became eligible to enroll, and most states conducted outreach to this population. The uninsured rate dropped from 13.3 percent in 2013 to a record low of 8.8 percent in 2016, and the number of uninsured fell by 13.7 million people over the same period. (Figures on insurance, income, and poverty throughout this analysis are from the CPS unless otherwise noted.)

Conditions were favorable for additional insurance gains in 2017: the number of people with full-time, year-round employment rose by 2.4 million in 2017; full-time jobs are a major route to insurance coverage. Moreover, several million Americans remained uninsured but eligible for coverage under the ACA or Medicaid, and substantial progress in reaching them had been made in the preceding years, as the number of Americans without coverage fell by 4.0 million in 2015 and by more than 900,000 in 2016.

But insurance gains halted in 2017, likely due at least in part to efforts by the Trump Administration to hinder various operations of the ACA. President Trump’s actions began in January 2017 with an executive order, issued on his first day in office, that instructed federal agencies to waive and delay ACA provisions “to the maximum extent permitted by law.”[3] Efforts to weaken the ACA continued throughout the year, with the cancellation of most federal outreach efforts toward the end of the ACA marketplace open-enrollment season for 2017. Meanwhile, year-long congressional and Administration attempts to repeal the ACA may have created some public confusion about the availability of coverage.[4]

Data Suggest Importance of State Efforts on Health Insurance

The data also underscore notable differences between states that did and didn’t expand Medicaid coverage under the ACA to all non-elderly adults with incomes up to 138 percent of the poverty line. According to ACS figures:[5]

Figure 1
Uninsured Rate Gap Between Medicaid Expansion States and Others Widening
  • For people in the 32 states (including the District of Columbia) that expanded Medicaid by January 2017, the uninsured rate was unchanged at 6.6 percent.
  • For states that did not expand Medicaid, the uninsured rate rose in 2017, from 11.8 percent to 12.2 percent.
  • The gap in uninsured rates between expansion states and non-expansion states has now grown for four straight years, from a gap of 4.0 percentage points in 2013 to one of 5.6 percentage points in 2017. (See Figure 1.) If the uninsured rate had fallen in non-expansion states at the same rate since 2013 as it did in expansion states, another 4.5 million uninsured Americans would have had coverage last year.
  • Since 2013, the likelihood of being uninsured fell by nearly half (by 49 percent) in expansion states, compared with a decline of just over one-fourth (28 percent) in states that did not expand.

In Louisiana — the sole state to expand Medicaid during 2016 (in July) — the uninsured rate fell by one-sixth from 2016 to 2017 (from 10.3 percent to 8.4 percent), far more than in any other state. Between 2015, the year before the expansion was instituted, and 2017, Louisiana’s uninsurance rate fell by nearly one-third.

Three-Year Gains in Income and Poverty Are Largest in Decades But Rates of Improvement Slowed

Median household income rose by $1,063 or 1.8 percent in 2017, to $61,372, the data show. The poverty rate fell to 12.3 percent, from 12.7 percent in 2016. Coming on the heels of strong gains in the two previous years, 2017 contributed to the largest improvements in median income and poverty of any three-year period in decades, although the rates of improvement in both income and poverty slowed from 2015 to 2017.

Between 2014 and 2017, the typical household’s income rose by 10.4 percent after inflation, the largest such three-year gain on record, with data back to 1967. Over the same years, the poverty rate fell by 2.5 percentage points (leaving 7.0 million fewer people in poverty), the largest three-year decline in the poverty rate since the 1960s (1966 through 1969).

Although median household income grew in all three years — 2015, 2016, and 2017 — the rate of growth slowed, with increases of 5.1 percent, 3.1 percent, and 1.8 percent in those years, respectively.

Similarly, the official poverty rate fell to 12.3 percent in 2017, down from 12.7 percent the year before, but the decline (0.4 percentage points) was smaller than in 2015 or 2016 (when the declines were 1.2 percentage points and 0.8 percentage points, respectively).

The continued but slowing progress reflects patterns in the broader economy, with continuing job growth but slow wage growth. The economy added more than 2 million jobs in each of the last three years (2.8 million, 2.5 million, and 2.3 million in 2015, 2016, and 2017), Labor Department data show. Real average weekly earnings grew 0.5 percent in 2017, down from 1.9 percent and 0.8 percent in 2015 and 2016.

Median Income About Where It Was in 2000

Median income reached $61,372 in 2017, a $1,063 increase over 2016, after adjusting for inflation. At first glance this marks the highest such figure recorded for any year in the Census Bureau’s historical tables, but Census cautions that it redesigned the CPS in 2013 to measure income more accurately. Census reports that after accounting for the changes made in 2013 to its survey questionnaire, the median income level in the CPS for 2017 is statistically indistinguishable from the level for various previous years, including 1998, 1999, 2000, and 2007.[6]

Median incomes dipped in the 2001 downturn, never regained their 2000 level during the subsequent economic expansion through 2007, fell deeply in the Great Recession, and have only now climbed back to approximately their previous high.

Put differently, median income is about where it was two decades ago. In most of the intervening years, it was lower.

Wide Gaps Remain Among Racial and Ethnic Groups

African Americans and Latinos continue to experience the highest risk of poverty by a wide margin. The Census figures show that in 2017:

  • Some 21.2 percent of African Americans lived below the poverty line. This was 2.4 times the poverty rate for people identifying as non-Hispanic white.
  • About 18.3 percent of Hispanics lived below the poverty line — about 2.1 times the rate for non-Hispanic whites.[7]

Though non-Hispanic whites had a relatively modest poverty rate of 8.7 percent in 2017, they continue to constitute the single largest racial or ethnic group in poverty, due to their large numbers in the overall population. In 2017, about 17 million non-Hispanic whites were poor, compared with 10.8 million Hispanics and 9 million African Americans.

Poverty Rates Higher for Children Than Other Age Groups

Poverty among children declined again in 2017, though children remain the age group most likely to be poor. Some 17.5 percent of children lived below the official poverty line in 2017, a higher share than for adults aged 18 to 64 (11.2 percent) or seniors aged 65 and older (9.2 percent).

Poverty rates were substantially higher for African American (29.0 percent) and Hispanic (25.0 percent) children than for non-Hispanic white children (10.9 percent).

Poverty among children is of particular concern because children who experience poverty frequently show lasting deficits in education and health relative to similar non-poor children. Conversely, participation in anti-poverty programs that raise family resources is associated with improvements in children’s health, educational achievement, and long-term earning power, studies have found.[8]

Millions More Would Be Poor Without Economic Security Programs

The Census data show that poverty and economic hardship would be far more widespread if not for key federal programs. Using the Supplemental Poverty Measure (SPM) — which, unlike the official poverty measure, accounts for taxes and non-cash benefits — Census reported that government programs lifted millions of people above the SPM poverty line in 2017. The data show that:

  • SNAP, formerly known as food stamps, lifted 3.4 million people out of poverty;
  • Refundable tax credits (the Earned Income Tax Credit and the low-income portion of the Child Tax Credit) lifted out 8.3 million people;
  • The Supplemental Security Income program, which serves low-income people who are elderly or have serious disabilities, lifted out 3.2 million;
  • Low-income housing assistance lifted out 2.9 million people; and
  • The Low-Income Home Energy Assistance Program lifted out of poverty 185,000 people.

These programs also reduced the severity of poverty for tens of millions more people.

Moreover, Census data tend to understate the degree to which these programs lift people out of poverty or closer to the poverty line.[9] The Census figures cited here use the SPM, which is favored by most analysts because it measures income more comprehensively than the “official” measure does. But the SPM figures do not correct for the underreporting of government benefits in the Census surveys, which results in some income that households receive from these programs not being reflected in these data.

Supplemental Measure Shows Government Assistance Driving Long-Term Progress Against Poverty

Some policymakers may be tempted to use the new poverty data to argue that federal anti-poverty programs have failed, since the official poverty rate has changed little since the late 1960s. Most analysts would reject such a conclusion, however, because comparing poverty rates between the 1960s and today by using the official poverty measure yields misleading results. The official poverty measure doesn’t accurately portray the overall financial well-being of low-income Americans or how that has changed over time, because it counts only cash incomes and omits the impact of key government income-assistance programs such as SNAP, the Earned Income Tax Credit, and rental subsidies.

The SPM does count these benefits and makes other improvements in poverty measurement, such as netting out income and payroll taxes. The Census’s SPM tables start in 2009, so they can’t be used to answer various historical questions, such as how much the poverty rate has changed since the 1960s or whether it has returned to its level in 2007, before the Great Recession. However, Columbia University researchers have developed a set of historical SPM estimates known as the “anchored SPM,” which shows poverty trends and the impact of anti-poverty efforts over the long term.[10]

Our update of the Columbia researchers’ anchored SPM series, which uses the recently released Census data and a modernized 2017 poverty line, shows that the poverty rate fell by nearly half (45 percent) between 1967 and 2017 — from 25.1 percent to 13.9 percent over this period.[11]

Policymakers Should Build on Progress, Not Undermine It

A goal for policymakers should be to sustain recent progress in income and poverty and restore progress on health insurance.

Continued progress on poverty and income is far from certain. Some proposals from the President and some congressional leaders would risk increasing poverty and hardship for low- and modest-income households, including those with low-wage workers. Both the President’s 2019 budget[12] and the budget plan that the House Budget Committee adopted on a party-line vote in June[13] call for deep cuts in Medicaid, SNAP, and income assistance, as well as in non-defense discretionary programs such as rental assistance. The latest Census data suggest that cutting those and other basic assistance programs would cause poverty and the share of people without health insurance to rise.

Alternative approaches would build on recent progress rather than jeopardize it. And the Federal Reserve, when setting interest rates, should pay heed to the positive consequences of tight labor markets, given continued moderate inflation. The President and Congress should enact a long-overdue increase in the minimum wage, which has remained unchanged for eight years and has shrunk considerably in purchasing power over recent decades. They also should strengthen refundable tax credits for low-income workers and their families and increase low-income families’ access to affordable housing and child care (the latter of which steps would enable more parents to work or work more hours). Together with steps to stabilize and strengthen health insurance markets, such approaches would make it more likely that the gains of a growing economy are more broadly shared.

On health insurance, federal policymakers should first do no harm. Unfortunately, efforts to weaken health coverage continue. The Trump Administration, for example, is implementing rules promoting substandard health plans — outside the ACA marketplaces — that lack important consumer protections.[14] These plans likely will cause health insurance premiums within the marketplaces to rise, and consequently could cause the number of people without insurance to climb.

States, too, have ways to impact health coverage. As one example, states that haven’t adopted the Medicaid expansion can do so, and states that have adopted it can support vigorous outreach to bolster enrollment in Medicaid and ACA marketplace programs. Indeed, efforts to implement the Medicaid expansion continue in Maine, where voters approved the expansion through a ballot measure last November. And voters in Idaho, Nebraska, and Utah will vote on similar ballot measures to expand Medicaid coverage this fall.

End Notes

[1] For details on Administration efforts to weaken the ACA, see Center on Budget and Policy Priorities, “Sabotage Watch: Tracking Efforts to Undermine the ACA,” last updated September 12, 2018, https://www.cbpp.org/sabotage-watch-tracking-efforts-to-undermine-the-aca.

[2] The large size of the ACS also makes it suitable for local and state estimates, its main purpose. Nationwide, the ACS uninsured rate inched up from 8.6 percent in 2016 to 8.7 percent in 2017, a statistically significant change. CPS and ACS rates for 2017 are both found in Edward R. Berchick, Emily Hood, and Jessica C. Barnett, “Health Insurance Coverage in the United States: 2017,” U.S. Census Bureau, September 12, 2018, https://www.census.gov/content/dam/Census/library/publications/2018/demo/p60-264.pdf.

[3] Timothy Jost, “Trump Executive Order on ACA: What It Won’t Do, What It Might Do, and When,” Health Affairs, January 21, 2017, https://www.healthaffairs.org/do/10.1377/hblog20170121.058405/full/.

[4] Sara R. Collins et al., “Americans’ Views on Health Insurance at the End of a Turbulent Year,” Commonwealth Fund, March 2018, https://www.commonwealthfund.org/sites/default/files/documents/___media_files_publications_issue_brief_2018_mar_collins_views_hlt_insurance_turbulent_year_aca_tracking_survey.pdf.

[5] Matt Broaddus, “Census: States Not Expanding Medicaid Lagging on Health Coverage,” Center on Budget and Policy Priorities, September 12, 2018, https://www.cbpp.org/blog/census-states-not-expanding-medicaid-lagging-on-health-coverage.

[6] Jonathan L. Rothbaum, “Redesigned Questions May Contribute to Increase,” Census Bureau, September 12, 2018, https://census.gov/library/stories/2018/09/highest-median-household-income-on-record.html. Income data from the Census Bureau’s much larger American Community Survey (ACS), released September 13, 2018, suggest that median income was slightly higher in 2017 than in 2007 (by $201 or 0.3 percent); that difference is statistically significant in part because of the ACS’ large size of approximately 3 million addresses. The ACS does not provide directly comparable household data for years before 2005.

[7] The Census data also show that 10 percent of Asian Americans lived below the poverty line in 2017, just above the non-Hispanic white rate.

[8] Arloc Sherman and Tazra Mitchell, “Economic Security Programs Help Low-Income Children Succeed Over Long Term, Many Studies Find,” Center on Budget and Policy Priorities, July 17, 2017, https://www.cbpp.org/research/poverty-and-inequality/economic-security-programs-help-low-income-children-succeed-over.

[10] The anchored SPM uses a modernized poverty threshold that reflects current living standards. The threshold is adjusted downward for previous years to reflect the effects of inflation so that the threshold grows from year to year at the same pace as inflation. Some analysts prefer the anchored SPM to the standard SPM, under which the poverty threshold grows slightly faster than inflation as living standards rise across decades. Christopher Wimer et al., “Trends in Poverty with an Anchored Supplemental Poverty Measure,” Columbia Population Research Center Working Paper No. 13-01, December 2013, http://cupop.columbia.edu/publications/2013.

[11] Danilo Trisi, “Economic Security Programs Cut Poverty Nearly in Half Over Last 50 Years, New Data Show,” Center on Budget and Policy Priorities, September 14, 2018, https://www.cbpp.org/research/poverty-and-inequality/economic-security-programs-cut-poverty-nearly-in-half-over-last-50.

[12] Center on Budget and Policy Priorities, “The 2019 Trump Budget: Hurts Struggling Families, Shortchanges National Needs,” February 20, 2018, https://www.cbpp.org/research/federal-budget/the-2019-trump-budget-hurts-struggling-families-shortchanges-national-needs.

[13] Joel Friedman and Richard Kogan, “House GOP Budget Retains Tax Cuts for the Wealthy, Proposes Deep Program Cuts for Millions of Americans,” Center on Budget and Policy Priorities, June 28, 2018, https://www.cbpp.org/research/federal-budget/house-gop-budget-retains-tax-cuts-for-the-wealthy-proposes-deep-program-cuts.

[14] Sara Lueck, “Key Flaws of Short-Term Health Plans Pose Risks to Consumers,” Center on Budget and Policy Priorities, September 20, 2018, https://www.cbpp.org/research/health/key-flaws-of-short-term-health-plans-pose-risks-to-consumers.