Number of kids returning to school this year: 48.1 million, all receiving free meals.

Number of people housed with the help of federal rental assistance: 10.4 million, 23 percent of whom are disabled.

Number of workers who lost their unemployment benefits on Labor Day: more than 8 million.

To help people, we have to know how many people are in need, how many people receive benefits, and what the gap is between those two numbers. For the past 15 years, the American Community Survey (ACS), conducted annually by the Census Bureau, has been one source of such data. But the pandemic made that data collection impossible.

Get TalkPoverty In Your Inbox

The American Community Survey tracks how Americans are doing on a granular level annually, not just once a decade: It measures the highest level of education people have completed, how many people experience poverty each year, and how people commute to work. Any community with at least 65,000 people has ACS data that anyone can view. That data is used to allocate resources for more than 130 programs, many of which fight poverty, including SNAP, Head Start, Section 8 vouchers, Unemployment Insurance, and the Census itself.

The Census Bureau collects the majority of its information through Internet, telephone, and mail-in surveys, then follows up with some of the people who do not respond with phone calls and in-person visits. But in 2020, the stay at home orders at the beginning of the pandemic interrupted the data collection process. Mail operations were canceled for April, May, and June of 2020, as were in-person interviews. So, while the Census did still get some respondents, it was not able to collect sufficient data among key groups who tend to be less responsive: People with lower incomes, lower educational attainment, and those who do not own their own homes. As a result, they decided they couldn’t offer their usual ACS data release.

The people we are missing data on are the exact group of people TalkPoverty focuses on: The same people who were hit hardest by the pandemic, and for whom accurate data is most important in developing a response.  The American Community Survey is how we know, for example, the number of people who have health insurance, what their household income is (and how much of it depends on public benefits), and how many people have dipped below the poverty line by state and congressional district. That’s especially important because it lets us track geographic inequities over time.

The Census will still be able to provide useful data from a related Census product, the Current Population Survey. That data, released September 14^th, will include national poverty rates, health insurance coverage data, and income data that lets us calculate the gender wage gap. We’ll have a broad sense of how Americans were faring overall in 2020, and how effective federal aid programs such as expanded Unemployment Insurance and SNAP were over the year. However, there will not be any state or local breakdowns of that data. So, while we will have the official annual poverty estimates, we will not have detailed data that would show if certain groups of people, such as Black women in Michigan or Latinas in Texas, were more likely to experience poverty in 2020.

Later this fall, the Census is releasing what it’s calling “experimental” ACS data on “a limited number of data tables for limited geographies.” It’s unclear right now what exactly that means – we do not know which data points or locations it will cover. It is unclear how the many agencies that rely on this data to calculate necessary funding for benefits will be doing their math, even as we need data more than ever to reflect changed economic circumstances for millions of Americans. What data there is will provide us with important information about a year when so many communities that rely on the safety net were in turmoil — from grocery store clerks to elementary school kids.

To understand how this proposal would work, it’s worth starting with how the Census defines poverty in the first place. The OPM was first created in the 1960s by Mollie Orshansky, an economist working for the Social Security Administration, who proposed poverty thresholds related to the cost of food: Any family earning less than three times the USDA estimate for the subsistence food budget is considered poor. Those thresholds have remained in place over the last half century, virtually unchanged other than by cost-of-living adjustments. Right now, according to the OPM, a family of three (with two adults and one child) counts as poor if their income amounts to less than $20,212 per year.

Get TalkPoverty In Your Inbox

The current OPM has faced criticism on a number of fronts, including failing to consider geographic variation (cost of living in Missoula is very different than Manhattan), account for any of the costs families face (such as medical costs or transit costs), or provide a full picture of families’ income (criticisms the Census has attempted to address with the creation of its much more robust Supplemental Poverty Measure (SPM)). In short, the OPM is too low for most people’s understanding of poverty. It’s hard to reconcile a measurement that says only 12.3 percent of people in the U.S. are poor when more than 4 in 10 adults would struggle to come up with $400 in an emergency and 70 percent of voters have confronted a serious financial hardship in the last year.

The Trump administration’s proposal would drive the OPM even lower.

Right now, the poverty thresholds are adjusted annually by the CPI-U, or Consumer Price Index for Urban Consumers, which measures inflation by tracking the change in price for a set group of consumer goods across a range of geographic areas. Trump is proposing switching to a different measure of inflation. The proposal isn’t specific on which measure they would use, but it twice mentions the possibility of using the “chained CPI” (or the C-CPI-U) as an alternative. The chained CPI differs from the CPI-U in that it allows for substitution across similar kinds of items. For example, the chained CPI takes into account the idea that if the price of chicken is high, perhaps you switch to pork, whereas the CPI-U only accounts for substitutions within categories, such as picking a Granny Smith apple over a Red Delicious. By allowing for these additional substitutions, chained CPI shows a slower rate of inflation, but for many families who are already choosing between paying the rent and buying food, they are already living as frugally as possible.

In addition to being impractical, the chained CPI would push poverty thresholds further and further down over time. The chained CPI has grown more slowly than the CPI-U by about 0.25 percentage points annually. That means that every year, chained CPI will redraw the thresholds lower than the CPI-U would have, in a way that compounds over time. For example, say that inflation under the CPI-U increased by 2.4 percent a year over the next decade. That means that 2018’s $20,212 poverty threshold for two adults and one child of today would be $25,021 in 2027 (in 2018 dollars). If the chained CPI grew 0.25 percentage points less each year the poverty threshold for this family would be about $544 lower at $24,447. By 2037 the difference would grow to $1,439 less.

This change to the OPM matters way beyond just measuring who is poor and who isn’t — changing it would have a ripple effect for millions of people. The poverty thresholds the Census publishes are the basis of the poverty guidelines issued by HHS. And HHS’s poverty guidelines are the basis of eligibility for a whole range of programs that help people meet their basic needs. Making this change would have long-term echo effects, chipping away at eligibility over the years. For example, school kids qualify for free meals and milk if they are at less than 130 percent of the federal poverty guidelines, and for reduced-price meals and milk if they are at less than 185 percent of the guidelines. Children up to age 5 can are eligible to attend Head Start or Early Head Start if they are poor (as defined by the guidelines). Low-income taxpayers can receive assistance resolving disputes with the IRS if they are at or under 250 percent of the poverty guidelines. As these definitions capture fewer and fewer people due to reduced inflation increases, there will be fewer and fewer people who receive the services they need.

We have estimates of how rapidly these changes can increase from the moment earlier this decade when policymakers (including some Democrats) supported changing the cost of living adjustment for Social Security benefits to the chained CPI. Research Joan Entmacher and I did at that time revealed that while the initial impact of this cut might seem small — averaging only about 0.3 percent cut to benefits per year — overtime these cuts would compound to create real harm. For example, if the chained CPI had been substituted for a single woman with a monthly benefit of $1,100, her benefit would be cut by $56 per month at age 80. By the time she reached 95, that cut would be more than $100 a month. If she received a benefit for 30 years, her cumulative losses would be nearly $20,000.

Benefits already fall short all too often. SNAP benefits typically run out before the end of the month, and kids go hungry in the summer without the benefit of school meals. Yet rather than provide families the support they need to meet their basic needs, Trump is proposing to artificially reduce poverty while actually cutting benefits — creating the illusion of a win from his perspective that is actually a heavy loss for low-income families.

As antipoverty advocates, we’ve become so accustomed to poverty staying stagnant—or getting worse—since the turn of this century that we barely know how to celebrate this kind of progress properly.  So, let’s take a minute to appreciate it.

Get TalkPoverty In Your Inbox

The U.S. poverty rate saw one of the largest single-year declines in almost 50 years. After years of stagnant incomes the typical American household got a raise for the first time in nearly a decade, with median household income rising more in one year than ever recorded. Meanwhile, the share of Americans without health insurance dropped to a record low, making 2015 the first time we’ve seen simultaneous improvements on all three fronts—poverty, median household income, and health insurance coverage—since 1999.

The improvements we’re marking this week didn’t happen by accident—they’re the result of policies that work. And with the wrong policy choices, we could see these important gains erased.

Take the minimum wage. State and local minimum wage increases likely played a big role in the decline in poverty and increase in income that we saw in 2015—as well as the faster wage growth among African American and Hispanic workers, and women (all of whom are disproportionately likely to be low-wage workers). They also likely helped the gender wage gap hit 80 cents for the first time ever, and led to even stronger progress on the wage gap for African American women. And in states that enacted minimum wage increases, low-wage workers saw faster wage growth than workers in states where minimum wages stayed flat.

On the health insurance front, the Affordable Care Act remains the gift that keeps on giving. The tremendous gains in coverage that we saw in 2014 and 2015 brought the nation to a record low uninsured rate of 9.1%. But here too, variation across states demonstrates the difference that policy choices make. The average share of uninsured individuals in states that hadn’t expanded Medicaid by January 2015 was 12.3%, compared to just 7.2% in states that had adopted Medicaid expansion. It is clear that the national gains would have been even stronger if 19 states weren’t still refusing to expand their Medicaid programs.

And let’s not forget the safety net. Census data released yesterday using the Supplemental Poverty Measure (SPM), an alternative measure that takes into account critical poverty-reducing investments such as nutrition assistance and tax credits for working families, reminds us how important these and other programs are to cutting poverty and mitigating hardship. According to the SPM, without Social Security nearly 27 million more Americans would have been poor in 2015—that would be an increase of more than 58% in the share of poor Americans. The Supplemental Nutrition Assistance Program protected 4.6 million Americans from poverty. And the Earned Income Tax Credit and low-income portion of the Child Tax Credit together accounted for 9.2 million fewer individuals living in poverty last year.

We’ve got to keep moving the needle in the right direction, particularly as we have yet to return to pre-recession levels of poverty and incomes. And despite strong progress on the national poverty rate, poverty among children, female-headed families, African-Americans, Hispanics, Native Americans, and people with disabilities remains dramatically higher than the overall national rate.

But the big takeaway from yesterday’s data that we should all be shouting about is that we know we can make great progress in cutting poverty—if we make the right choices. Now is the time to double down on the policies we know make a difference, such as making job-creating investments in infrastructure, research, and education; raising the minimum wage and further strengthening tax credits for working families; ensuring paid leave, childcare, and fair schedules, as well as closing the gender wage gap; protecting and strengthening vital safety net programs; and removing barriers to employment for people with criminal records and their families, and workers with disabilities.

Today we should all celebrate the progress we are finally making. But tomorrow, we must use this news to inform the work we do together in the months and years ahead. Now is the time to recommit to making the right choices so that we don’t turn back.

Get TalkPoverty In Your Inbox

If your initial reaction to that is “whoa, that’s waaay too low for a person to lead a minimally decent life on in the USA,” then you’re in good company. In a recent survey conducted by the conservative American Enterprise Institute (AEI) and the Los Angeles Times, Americans were asked, “[What is the] highest annual income a family of four can have and still be considered poor by the federal government.” The average response was $32,293—an amount 34 percent higher than the current federal poverty measure.

In short, conservatives did a poll on how much income it takes to avoid poverty, and the answer they got back was more than $8,000 above the federal poverty line.

The wonks reading this might be thinking “well, if the federal government says a married couple with their two kids only needs $24,000 to live a minimally decent life, then they must have good reasons to think this is enough.”  I’m a bit wonkish myself and generally trust official government statistics—but the federal poverty measure is a big exception.

The main reason I don’t trust this approach to measuring poverty is shown in the figure below. In 1963, the poverty line for a family of four was 50% below median family income—or one-half of the income of the typical four-person family in America. Today, however, the poverty line for a family of four is 66% below median family income.

That means to be officially counted as poor today, a family has to be much poorer compared to the typical American family than it had to be in 1963. In fact, if the federal poverty line today was set at the same place relative to median income as it was in 1963 it would be about $33,000, rather than $24,000.

The AEI survey results are not a fluke. We know from decades of evidence that the public’s understanding of the income needed to avoid poverty increases over time at a rate faster than inflation, and closer to the increases in mainstream incomes and living standards.

So why hasn’t the official poverty line been adjusted over time in a way that reflects the public’s more accurate understanding?

The reasons for this are largely political.

In the early 1960s—around the time when the Beatles were just becoming famous here—Mollie Orshansky, an employee in the Social Security Administration (SSA), developed working estimates of what it meant to be poor at that time.

The data available to Orshansky wasn’t particularly sophisticated, or even timely. For example, she based her estimates in part on a food consumption survey conducted in 1955. When the federal government started using her calculation of the poverty line in the mid-1960s, Orshansky and federal officials understood that it would need to be adjusted over the long-term for increases in mainstream living standards. The SSA “made a tentative decision early in 1968 to adjust the poverty thresholds for the higher general standard of living.”

But then two things happened that year. First, officials in the Johnson administration prohibited the SSA from making this kind of adjustment, likely in part due to concern that the updated figures would show an increase in poverty. Second, Richard Nixon was elected President.  After he took office, his budget office issued a directive making the Orshansky thresholds the “official” poverty measure, and specifying that they would be adjusted for inflation only.

In 1970, Orshansky said this decision would likely “freeze the poverty line despite changes in buying habits and changes in acceptable living standards.” There have been repeated recommendations to reform the poverty measure since then, but no President has been willing to revise the Nixon directive.

The Census Bureau has an alternative measure of poverty, the Supplemental Poverty Measure, which improves the current federal poverty measure in many respects. But even this approach puts the poverty line at about $25,000 to $26,000 for a family of four—and that’s still too low.

Here’s a better approach: dump the current official poverty measure and replace it with two different measures. One measure would be anchored to half of the typical (median) American family’s income in 2016 and then adjusted for inflation over time; the other would be set at the same level initially, but adjusted annually using the median income over a 5-year period. This way, the poverty line won’t drift away from mainstream living standards of living over time.

It’s 2016: We need a poverty measure that reflects what the public thinks is required to meet basic needs in Beyoncé America.  But instead we’re stuck in Beatlemania America—and that needs to change.

Two Census reports—the Current Population Survey and the American Community Survey—are critical resources for advocates, researchers, journalists, and policymakers alike. They provide rich information on issues that impact people’s health and economic wellbeing, ranging from their living situations, to public benefits usage, to how much money they earn.

Get TalkPoverty In Your Inbox

These data, which are for 2015, inform us about what is working to cut poverty and reduce inequality, and how we might do better from a public policy perspective. Here are four key trends wonks will be examining closely:

Incomes are rising—likely for minimum wage workers, too

Some researchers are forecasting that real median household income might see the largest one-year jump in more than a decade. Low-wage workers should see a rise, too—especially in states that raised their minimum wage. This increase is particularly important for women, who make up nearly two-thirds of all minimum wage workers.

Rising wages, particularly for low-wage workers, could mean that this is the year we learn that the gender wage gap among full-time workers—which stood at women earning 79 cents for every dollar earned by their male counterparts—finally broke the 80 cent-barrier. (Not quite shattering this particular glass ceiling, but moving a step closer!)

Anti-poverty advocates will also examine these data to see if the income gains will reach families in deep poverty—those who have incomes of less than half the poverty line (approximately $12,000 annually, for a family of four).

Don’t kid yourself—racial and gender inequalities are alive and well

Any improvement in the poverty rate and the gender wage gap is critical, but we’re a long way from widespread economic equality. For example, the wage gap for women of color is severe: Last year, African American women typically earned only 60 cents, Native American women 59 cents, and Hispanic women 55 cents, for every dollar earned by their white Non-Hispanic male counterparts.

These gender and racial disparities apply to poverty rates, too. Hispanics and African Americans experienced poverty rates about 2.5 times higher last year than white Non-Hispanics. Women are also more likely to face poverty, as are individuals born in a foreign country, persons with disabilities, and single-parent families.

There is a hidden story in these data about who is more likely to be poor and paid unfairly that wonks and others need to shine a light on.

The data are seriously flawed—especially for LGBTQ people

Every year this Census release sparks conversation about how the stats themselves could be improved. Two topics come up repeatedly: The flawed way that we measure poverty and the shocking lack of data about LGBTQ people.

There is widespread agreement that the federal poverty line—$24,300 for a family of four in 2016—is far too low, which means many more Americans are experiencing serious economic hardship than are deemed officially “poor.” This disconnect isn’t surprising, considering the Official Poverty Measure (OPM) was developed more than half a century ago. A lot of things have improved since then—cars, phones, computers, Americans’ appreciation of soccer—but the OPM hasn’t, even as families’ needs and spending patterns have changed dramatically.

The OPM also fails to account for numerous public policies that relieve hardship. This is one reason why many wonks are fans of the Census Bureau’s alternative Supplemental Poverty Measure (SPM).

The SPM includes income households receive through assistance programs like the Supplemental Nutrition Assistance Program (SNAP), which lifted 4.7 million people above the poverty line in 2014; as well as tax credits such as the Earned Income Tax Credit and Child Tax Credit, which together lifted 9.8 million people out of poverty in 2014. The SPM also incorporates some of households’ necessary expenditures, such as clothing and utilities, and geographic variation in housing costs.

While these poverty measurements are less than ideal, they are far better than having almost no data at all—as is the case for members of the LGBTQ community. The lack of sexual orientation and gender identity data in these data sets is glaring, given that the limited data we do have demonstrate that LGBTQ individuals face higher poverty rates than many other communities. Since funding for anti-poverty initiatives often depends upon being able to show that economic need exists, this dearth of data can prevent the LGBTQ community from receiving help even when there is a clear need.

We must redouble our efforts to ensure we collect much-needed data on the LGBTQ community while also working to reform the way we are collecting and measuring poverty data.

Public policy choices reduce or exacerbate poverty, inequality, and hardship

Last year the Census data demonstrated the huge impact the Affordable Care Act had on Americans’ health care coverage, as uninsured rates fell to a historic low with declines in all 50 states and the District of Columbia. This year wonks anticipate a new low in the uninsurance rate—perhaps even below 9 percent—though it would be even lower if more states expanded Medicaid coverage.

Next week’s release will also show how other smart social programs are effectively reducing poverty. For example, last year’s SPM data revealed that without Social Security fully half of American seniors would have been poor, and that without refundable tax credits, nearly 1 in 4 children would have fallen below the poverty line as well. This evidence has fueled increasing calls from advocates and policymakers to strengthen and expand Social Security, refundable tax credits, and other key safety net programs.

Wonks look forward to continuing to assess our public policy choices based on this year’s data.  We already know a lot about what to do to reduce hardship, boost economic mobility, and increase opportunity.  The new Census data can help us move in the right direction—if we ask the right questions, and look for the answers.