Uninsured Rate Rose in 2019; Income and Poverty Data Overtaken by Pandemic Recession
September 15, 2020
Income rose and poverty declined in 2019, as would be expected in the peak year of a decade-long economic recovery, but the number of Americans without health insurance increased for the third consecutive year despite a growing economy.
"The number of Americans without health insurance increased for the third consecutive year despite a growing economy."The rise in the share of Americans who were uninsured at the time of the Census Bureau survey — from 8.9 percent (28.6 million people) in 2018 to 9.2 percent (29.6 million people) in 2019 — likely reflects continued Administration actions that have reduced access to health insurance coverage. The data show a 0.7 percentage-point decline in health insurance among Hispanic individuals, driven by a large, 1.4 percentage-point drop in the share of Hispanic individuals with public coverage.
These health coverage data come from Census’ American Community Survey (ACS), which was conducted before the pandemic hit. The income and poverty data, by contrast, come from the Current Population Survey (CPS), and their reliability is compromised because Census chiefly surveyed households for the CPS in March, as the pandemic was hitting. As Census noted in a paper released today, the pandemic and associated safety concerns forced it to stop in-person interviews and make other changes in collecting data, which sharply reduced response rates in the survey — especially among lower-income households. Census concluded that the increased non-response artificially boosted the income figures for 2019 and artificially lowered the poverty rate.
Accordingly, Census issued adjusted figures for some limited points, including household median income and the overall poverty rate, to account for the biases the pandemic caused in the survey results. The adjusted Census numbers, which are much more reliable than the unadjusted numbers, show median household income climbing 4.1 percent in 2019 after adjusting for inflation, from $64,180 in 2018 to $66,790 in 2019.
The adjusted poverty numbers show the poverty rate falling from 11.9 percent in 2018 to 11.1 percent in 2019.
These data, however, bear little resemblance to conditions today, as the pandemic and recession have fundamentally altered the economic landscape. Households’ circumstances differ sharply now from those reflected in the 2019 data.
Much more current Census Bureau data — from Census’ Household Pulse survey, which since April has collected data on a weekly or bi-weekly basis — show that as of late August, more than 22 million adults were in households that lacked sufficient food the previous week and millions of renters had fallen behind on rent. And with millions of workers having lost their jobs and hence their employer-based health coverage, the ranks of the uninsured will likely increase.
These hardship data underscore the critical need for the nation’s leaders to work out an agreement on strong new stimulus and relief measures before Congress adjourns later this month. It’s particularly urgent to enact measures strengthening food and rental assistance, public health measures, state and local fiscal relief, and unemployment insurance, among other areas. Without such measures, poverty and hardship will likely worsen further in the months ahead.
Health Coverage in 2019
The uninsured rate rose from 8.9 percent in 2018 to 9.2 percent in 2019, according to the ACS. The ACS data were collected before the pandemic and so are unaffected by the March drop in response rates. Census noted in its written documents issued today (and explained on its press call) that these data give a more reliable picture of how health coverage changed in 2019 than the CPS. The ACS is also a larger survey and generally more suitable for measuring recent year-to-year changes in health insurance.
The increases in the uninsured rate in 2017, 2018, and 2019 followed six consecutive years of decline, the ACS data show — from 15.5 percent in 2010 to a historic low of 8.6 percent in 2016. But because of the increases after 2016, 2.3 million more Americans were uninsured in 2019 than in 2016, including more than 720,000 children.
The regression in the last three years came despite steady job growth that should have improved health coverage, and it likely reflects, in large part, the effects of Trump Administration policies that reduce access to coverage. For example, the Administration’s policies toward immigrants, particularly its so-called “public charge” rule, have created a climate of fear among immigrants and their family members that discourages eligible people from enrolling in Medicaid or marketplace coverage. The new Census data show that groups disproportionately affected by these policies, including Hispanic adults, Hispanic children, and children who were not born in the United States, all saw much larger-than-average increases in uninsured rates.
To be sure, the uninsured rate in 2019 was still far below its level before enactment of the Affordable Care Act (ACA). But the ACA’s remaining coverage gains will be in jeopardy if the Administration succeeds in its efforts to undo the law through the courts or legislation. Just last month the Administration filed its latest brief with the Supreme Court in its effort to overturn all of the ACA. That would add approximately 20 million people to the ranks of the uninsured, the Urban Institute estimated before the pandemic, and it would likely cause even larger coverage losses today since millions of people are turning to ACA coverage programs during the recession.
Income and Poverty in 2019
Median household income stood at $66,790 in 2019, about 4.1 percent above the inflation-adjusted 2018 level of $64,180, and 11.1 percent of Americans lived below the official poverty line, compared to 11.9 percent in 2018 — according to the Census data that correct for the impact of the pandemic on data collection in the CPS.
These improvements shouldn’t be surprising, as 2019 was the tenth (and final) full year of a record-long economic recovery. As in most past “peak” recovery years, median income appears to have reached a new high in 2019. Median household income set new highs in 15 of the last 50 years, including more than 40 percent of non-recession years. The poverty rate fell in 2019 for the fifth straight year.
Although incomes clearly grew and poverty declined in 2019, a better assessment of changes in these areas in 2019 will require additional data, including detailed income and poverty findings from the ACS that are due out later this week. The ACS not only has a larger sample size than the CPS but also collected its 2019 data before the pandemic began. We expect its income and poverty data for 2019 to be the more reliable, in particular for the non-elderly population. (The 2019 ACS includes some wording changes that especially affect the elderly.)
2019 Economy Bears Little Resemblance to That of Today
Since the declaration on March 13 of a national coronavirus emergency, the nation has endured the deepest recession since the 1930s, fundamentally altering the economic landscape and making the 2019 findings an inaccurate gauge of current circumstances. The official unemployment rate, 3.5 percent in February, peaked above 14 percent in April and still exceeded 8 percent in August. Some 35 million people were either officially unemployed or lived with an unemployed family member in August, including 9 million children. This figure rises to 61 million workers and family members when one includes those not officially unemployed but nonetheless sidelined from the labor market, including workers with an unpaid absence from their job due to illness, caretaking responsibilities, or other reasons, as well as those who want a job but aren’t actively seeking work given the state of the labor market and the risks the pandemic poses.
Other data show sharp increases in various measures of hardship. Multiple studies have found that the number of households having difficulty affording food has soared during the crisis. A July Brookings study found that “young children are experiencing food insecurity to an extent unprecedented in modern times,” and more than 22 million adults — or more than 1 in 10 — reported in August that their household had “not enough to eat” sometimes or often in the last seven days. By comparison, fewer than 4 percent of adults reported experiencing this problem at any point in 2019, according to a CBPP analysis of a separate Census survey.
In addition, millions of renters are behind on rent, and data suggest the number climbed higher over the summer.
Adding to these concerns, the economic crisis and the pandemic appear to be hitting hardest those who already were struggling most to make ends meet. They include workers with no college education, who have lost their jobs at twice the national rate, and workers in low-paid industries, which have shed jobs at nearly three times the rate in high-paid industries. They also include Black and Latino individuals, many of whom work in low-paying jobs and entered the pandemic with limited household resources due to factors such as inadequate educational opportunities, employment discrimination, and other forms of racism.
Today’s Census data show that in 2019, before the recession, Black and Hispanic individuals were more than twice as likely to be poor as non-Hispanic whites: 18.8 and 15.7 percent, respectively, versus 7.3 percent. (These data are not adjusted for the reporting bias described above, but similar disparities are also found in prior years.) And during the pandemic, employment levels have fallen more among Black and Latino workers than among non-Hispanic white workers. Data that Census collected in August as part of its Household Pulse survey show Black and Latino adults also are more likely to live in households that lack sufficient food.
 Jonathan Rothbaum, “How Does the Pandemic Affect Survey Response: Using Administrative Data to Evaluate Nonresponse in the Current Population Survey Annual Social and Economic Supplement,” U.S. Census Bureau, September 15, 2020, https://www.census.gov/newsroom/blogs/research-matters/2020/09/pandemic-affect-survey-response.html.
 CBPP analysis of the August 2020 basic monthly Current Population Survey public-use file.
 Food insecurity may have doubled overall — and tripled among households with children — in April and May 2020, according to estimates using data from the COVID Impact Survey, Current Population Survey, and other sources. Diane Schanzenbach and Abigail Pitts, “How Much Has Food Insecurity Risen? Evidence from the Census Household Pulse Survey,” Northwestern University Institute for Policy Research, June 10, 2020, https://www.ipr.northwestern.edu/documents/reports/ipr-rapid-research-reports-pulse-hh-data-10-june-2020.pdf; Diane Schanzenbach and Abigail Pitts, “Food Insecurity Triples for Families With Children During COVID-19 Pandemic,” Northwestern University Institute for Policy Research, May 13, 2020, Ahttps://www.ipr.northwestern.edu/news/2020/food-insecurity-triples-for-families-during-covid.html.
 Lauren Bauer, “About 14 million children in the US are not getting enough to eat,” Brookings Institution, July 9, 2020, https://www.brookings.edu/blog/up-front/2020/07/09/about-14-million-children-in-the-us-are-not-getting-enough-to-eat.
 Census Bureau Household Pulse Survey data, Food Table 2b. Food Sufficiency for Households, in the Last 7 Days, by Selected Characteristics: United States, data collected August 19-31, 2020, https://www.census.gov/data/tables/2020/demo/hhp/hhp13.html; and CBPP analysis of the December 2019 Current Population Survey Food Security Supplement. Recent methodological changes to the survey mean that current Pulse data on food sufficiency are not comparable with estimates for July or earlier months.
 See, in particular, Figure 8 in CBPP, “Tracking the COVID-19 Recession’s Effects on Food, Housing, and Employment Hardships,” updated September 11, 2020, https://www.cbpp.org/research/poverty-and-inequality/tracking-the-covid-19-recessions-effects-on-food-housing-and; and National Multifamily Housing Council Rent Payment Tracker. Note that the Pulse survey changed its question regarding rent payments in August. These and other methodological changes mean that the survey’s hardship data are not comparable with Pulse estimates from earlier weeks. Prior to the change, the Pulse survey showed rent problems growing from 18.9 percent in late June through 21.4 percent in late July.