New Census Report Proves Again That Anti-Poverty Programs Work

September 19, 2016

Only so much number crunching a lone blogger like me can do. So I’m behind the curve on the Census Bureau’s Supplemental Poverty Measure report, issued the same day as the report using the official measure.

As in the past, the SPM shifts poverty rates up and down. The overall poverty rate, for example, is higher — 14.3%, as compared to the official 13.5%. The child poverty rate drops from 19.7% to 16.1%, while the senior poverty rate rises from 8.8% to 14.3%.

These differences, as well as others derive from numerous differences between the measures. For example, the SPM includes the children who aren’t part of the official measure’s poverty universe.*

This is relatively minor, compared to other differences — thresholds among them. Instead of those I’ve nattered about, the SPM begins with consumer spending on four basic needs, plus a small additional for others.

It then deducts for work-related expenses, e.g., transportation, child care, and for child support payments and medical costs that individuals themselves must pay. (Those medical out-of-pockets largely explain the higher senior poverty rate.)

The annual threshold adjustments differ too — and in a way that may make yearly changes in the SPM poverty rates “confusing,” the Center on Budget and Policy Priorities says. It specifically cautions against comparing the new SPM figures to last year’s.

I’ll confine myself then to what we can glean from another major difference. For the official measure, only pre-tax cash income counts in determining whether a household and the members it recognizes were poor.

The SPM deducts for taxes. It also includes income derived from the refundable tax credits and the cash-equivalent value of a some safety net benefits that the federal government funds either entirely or in combination with states.

What we can see, because of an analysis the Bureau provides, is what poverty rates would have been without one or another of the safety net benefits — both cash and cash-equivalent. It folds in Social Security retirement and disability benefits, though they’re not for low-income people only.

As always, Social Security proves the most effective anti-poverty measure we’ve got. Without the benefits it provided, about 26.6 million more people would have been part of the poverty rate, boosting it to over 22.6%.

Well over a third of all seniors would have been poor — nearly triple the rate with those benefits factored in.

The Earned Income Tax Credit and Child Tax Credit again come in second. They lifted about 9.2 million workers and their dependents over the poverty threshold, including 4.8 million children. Their already-high poverty rate would have been 22.6% without the credits.

Not all low-income workers benefited, however. Current law denies the federal EITC to both young and elderly workers. And the credit is very small for age-eligible adults who don’t have children — or who do, but not in their homes for more than half the year.

In short, an anti-poverty measure that works, but could work better. One could say the same for other safety net benefits the SPM report accounts for.

SNAP (the food stamp program), for example, as I’ve often said. Yet even with its current limits, it lifted roughly 4.6 million people, including nearly 2 million children out of poverty last year.

Results for Temporary Assistance for Needy Families, which I’ve also often gone on about, were pathetic — a 0.2% nick in the poverty rate.

An even more pathetic impact from the Low Income Home Energy Assistance Program, which neither the administration nor Congress seems much interested in funding.

We don’t see a large boost over the poverty thresholds from federally-funded housing subsidies either — a generously rounded up 2.5 million fewer poor because of it. In this case, we do have an actively interested administration — and what seems moderate support from majorities in Congress.

But the SPM thresholds take account of what people must spend for housing. And, as everyone knows, housing costs have been rising virtually everywhere.

So federal budgets would need to do more to keep those costs from driving up poverty rates than merely ensure that as many households have vouchers as they do now — or the same chance to live in public housing.

The Census Bureau reports each of the anti-poverty programs separately. So we can’t see how many people were lifted out of poverty by, say, SNAP plus a housing subsidy.

The Center on Budget rolls all the programs together and concludes that they cut the poverty rate almost in half last year.

It also notes, as have other analysts, that households surveyed tend to under-report the benefits they’ve received — mainly just because it’s hard to recall exactly how much one gets, perhaps from multiple sources and usually over some period of time.

At a minimum then, the safety net benefits, plus Social Security lifted about 38.1 million people over their poverty threshold — more seniors than younger people, but still about 7.9 million children.

CBPP warns that cuts in the programs would plunge more people into poverty. That’s, I think, what any fair-minded person would conclude from the SPM analysis.

It shows, with hard numbers, that our major anti-poverty programs work, notwithstanding the constant drumbeat from the right about how they’ve failed. It also shows they could work better — some perhaps if just more amply funded, some surely if also reformed.

* The SPM report adjusts the official rates to include the missing children. So one finds a higher overall poverty rate and higher child poverty rate there.


Subsidized Housing Cuts Poverty Rate More Than Census Measure Shows

October 13, 2015

My post on the dismal status of the Housing Choice voucher program led off with a call for larger investments in anti-poverty measures that have proved effective. But I didn’t wrap back around to how the vouchers fit in. Just too much, I felt, to cram into a single post.

So here’s the missing piece.

The Coalition on Human Needs, the source of my jumping-off point, draws on the Census Bureau’s Supplemental Poverty Measure for proofs of effectiveness. Many other progressive analysts and advocates do so as well.

As I’ve said before, the Bureau shows the impacts of major social insurance and safety net programs by recalculating poverty rates without their cash value.

The impacts of federally-funded housing subsidies — vouchers and public housing rolled together — seem relatively small, though not negligible when analyzed this way.

Without them, the SPM poverty rate would have been 0.9% higher in 2014. So it would seem they lifted roughly 2.8 million people over the poverty threshold — about 7 million fewer than the refundable tax credits, the top-ranked program one might classify as safety net.

The impact of the housing subsidy programs on the poverty rate has remained about the same ever since the Bureau began issuing its SPM reports in 2010. Small variations, but within the statistical margin of error.

The under-funding I’ve been going on about is one reason the impacts aren’t greater. But it’s not the only reason. Eligibility for housing assistance is another.

An individual or family doesn’t have to be hovering near the poverty threshold to qualify. The cut-off instead is 30% of the median income for the area they live in. That’s often, if not always considerably higher than the applicable poverty threshold.

In the District of Columbia, for example, the cut-off for a couple with with two children last year was about $9,000 higher than the maximum the family could have and still be counted as poor, if renters.

True, the DC Housing Authority exercises preferences when awarding new housing vouchers and public housing units, including one for homeless people, who are likely to be poor. But how many people in poverty actually benefited from such assistance last year is an open question.

More to the point, we can’t assume that what DCHA does reflects housing authority policies generally — except probably its current focus on veterans.

The impact figure is relatively small for a third, important reason. The dollar value of a housing subsidy doesn’t capture its full anti-poverty impact.

The value the SPM attributes to it is the different between the market-rate rent for an apartment, including basic utilities and what the household actually paid. But safe, stable housing is a platform of sorts for rising out of poverty. Or looked at another way, not having it makes rising more difficult.

Shifting around from place to place — or even worse, living in a shelter — makes finding and keeping a job unusually challenging. No permanent address. In some cases, no ready, regular access to a shower or a washing machine and dryer. Limited, if any access to a computer. Negative effects on both physical and mental health.

And if nothing else, time and energy that must be diverted to negotiating yet another temporary housing arrangement, packing and unpacking, figuring out new transit routes to work or training, other services, school and daycare for the kids, if any, etc.

For children, housing instability often has long-term consequences that make poverty in adulthood more likely — these better documented by research than consequences for adults.

Children whose families move around a lot, even if not in and out of a shelter, face higher risks of mental health problems — some manifesting themselves as what experts refer to as behavioral problems.

These help account for academic difficulties, as measured by standardized test scores and grades. So do frequent shifts from one school to another.

The end result is a relatively high dropout rate. And we know what employment and earnings prospects are for folks without at least a high school diploma.

Such effects are hard to convert to dollars that adults, both current and future, would have if stably housed, thanks to vouchers or apartments in public housing.

But I’m sure as can be that enabling low-income people to live in decent housing they can afford reduces poverty more than the SPM shows.


Better Poverty Measure Changes Rates, Strengthens Case for Safety Net

September 21, 2015

As I noted last week, the Census Bureau published the results of its latest Supplemental Poverty Measure analysis at the same time as its official poverty measure rates for 2014.

As in the past, the SPM produces a somewhat higher nationwide poverty rate — 15.3%. Though a tad lower than the comparable rate last year, slides from the Bureau say it’s not enough to pass the statistical test.

We also see different rates, both higher and lower, for the major population groups the Bureau breaks out. For example, the child poverty rate is 4.8% lower — 3.5 million or so fewer children. At the same time, the senior poverty rate rises by nearly as much.

We see shifts among major race/ethnicity groups as well. The largest are for blacks (3% lower) and for Asians (4.8% higher).

All these shifts and others reflect four major ways the SPM differs from the official measure — the base it proceeds from, adjustments it makes for certain basic living and other “necessary” costs, whom it includes as part of a family and what it counts as income.

This last gives us a glimpse — imperfect, but the best we’ve got — of how well some of our major federal anti-poverty measures work. And once again, we get reliable hard data proving that they do work, right-wing canards notwithstanding.

For example, we generally see lower deep poverty rates, i.e., the percent of the population overall or of a particular group that lived on incomes no greater than half the applicable poverty threshold — about $9,535 for a parent with two children.

The overall deep poverty rate is 1.6% lower than what the official measure produces. The deep poverty rate for children drops more markedly — from 9.7% to 4.3%.

The Census Bureau attributes the lower deep poverty rates to non-cash benefits targeted to low-income people — a type of income the SPM captures, while the official measure doesn’t. Seniors are the exception here, it notes.

Their deep poverty rate goes up to 5.1%, making it the same as the rate for the population as a whole. This is mainly because both the official measure and the SPM count Social Security benefits as income, but only the latter adjusts for medical out-of-pocket costs, along with others deducted from the base.

It’s nevertheless still the case that Social Security proves the single most effective anti-poverty program we’ve got. Without Social Security benefits, half of all people 65 and over would fall below the poverty threshold.

The Census Bureau shows this and the effects of other benefits — mostly parts of the safety net — by deducting their value and displaying the new poverty rate.

So we learn, for example, that not counting the refundable Earned Income Tax Credit and Child Tax Credit would make the SPM poverty rate 3.1% higher. Little back-of-the-envelope math tells us that the tax credits effectively lifted about 9.8 million people out of poverty, including more than 5.2 million children.

SNAP (food stamp) benefits rank third among the anti-poverty impacts. They account for about 4.7 million fewer poor people, almost half of them children.

On the other hand, LIHEAP (Low Income Home Energy Assistance Program) benefits lifted only about 316,000 people above the poverty threshold — and so few in the working-age (18-64) group as to make no nick in their poverty rate whatever.

Now, the analysis doesn’t reflect the way benefits work in the real world. Most families that receive federally-funded help with their heating bills probably also get SNAP benefits, for example. Likewise low-income working families that get an annual budget boost from the refundable tax credits.

We don’t yet have an analysis that rolls all such safety net benefits together, though we do have one for 2012 that shows they cut the SPM poverty rate by nearly half and the child poverty rate by even more.

Do we nonetheless have policy lessons here? Well, of course, we do. Don’t want to try your patience, followers, but can’t restrain myself from flagging (flogging?) a few.

LIHEAP has become a pitiful thing, partly because it got whacked by the 2013 across-the-board cuts, partly because this came on top of earlier cuts and partly because, in case you hadn’t noticed, home heating costs have increased.

So fewer households are getting such help as LIHEAP provides and they’re getting less — so much less that the average grant didn’t cover even two months of heating during the 2014-15 winter season.

Not going to see much improvement, if any so long as the Congressional Republican majority insists on keeping appropriations for non-defense programs below the caps set by the Budget Control Act. The House Appropriations Committee has, in fact, approved a $25 million cut for LIHEAP.

Changes in the refundable tax credits that help account for the effectiveness the SPM analysis indicates will expire at the end of 2017. And what seems a bipartisan sentiment in favor of expanding the EITC for childless workers is thus far little more than that — and not all that bipartisan, if we judge by cosponsors of bills pending in Congress.

Though SNAP clearly lifts people of all ages out of poverty, it doesn’t prevent a goodly number from going hungry at least some of the time. More about this in an upcoming post — and more perhaps about other issues one can tease out of the new SPM report.

 


Better Poverty Measure, Worse Poverty Rate, But Not for Everybody

October 16, 2014

As in the past, the Census Bureau’s Supplemental Poverty Measure yields a higher poverty rate than the official measure that was the basis for the reports the Bureau issued last month. According to the just-released SPM report, the rate last year was 15.5%, rather than 14.5%.*

This means that about 2.9 million more people — roughly 48.7 million in all — were living in poverty. At the same time, 1.3% fewer people lived in deep poverty, i.e., at or below 50% of the income threshold that determines who’s counted as poor.

These differences as well as the many others reflect the fact that the SPM is constructed differently from the official measure. There’s a brief explanation of how it’s built in the last section below.

Other Shifts in Poverty Rates

We see shifts up and down for state-level rates. For example, the rate for the District of Columbia rises from 19.9% to 22.4%. Rates fall in 26 states and rise in 13. (These reflect three-year averages to compensate for the relatively small sample sizes.)

As in the past, rates also shift for major race/ethnicity groups. Most of the shifts are relatively small. An exception here for Asians, whose poverty rate was 5.9% higher, and for blacks, whose deep poverty rate was 4.6% lower.

The most marked shifts are again for the young and the old.

  • The child poverty rate drops from 20.4% to 16.4%, reducing the number of poor children by about 2.9 million.
  • The deep poverty rate for children is less than half the official rate — 4.4%, as compared to 9.3%.
  • By contrast, the poverty rate for people 65 and older rises from 9.5% to 14.6%.
  • And the deep poverty rate for seniors ticks up from 2.7% to 4.8%.

Poverty Rates Without Key Federal Benefits

The changes for seniors largely reflect the fact that the SPM factors in medical out-of-pocket costs. But the SPM report also tells us that the senior poverty rate would have been 52.6% without Social Security payments. In other words, Social Security protected about 23.4 million seniors from poverty last year — more than three and a half times as many as were poor.

This is only one of the policy-relevant figures the SPM report provides in a section that shows how poverty rates would change if some particular benefit weren’t counted as income. Some examples Census has helpfully translated into raw numbers:

  • The refundable Earned Income Tax Credit and Child Tax Credit lifted 8.8 million people out of poverty.
  • But for SNAP (food stamp) benefits, about 4.8 million more people would have fallen below the poverty threshold.
  • Unemployment insurance benefits lifted 2 million people over the threshold.

So we see that the much-maligned safety net programs work. But we also see that policy choices have impaired the impacts some of the biggies formerly had.

For example, SNAP benefits lifted about 5 million people out of poverty in 2012, before the across-the-board cuts became effective. We’ve yet to see the effects of the further, targeted whack at benefits that’s part of the new Farm Bill.

The anti-poverty impacts of UI benefits shrunk further — a trend dating back to 2010, according to the Center on Budget and Policy Priorities. The number of people the benefits lifted out of poverty last year was nearly half a million fewer than in 2012.

And that was before Congress let the Emergency Unemployment Compensation program die at the end of last year. The new UI figure almost surely reflects reductions it made when it last renewed the program, however.

SPM 101

As I’ve explained before, the SPM is a more complex — and generally viewed as better — poverty measure than the one that’s used for official purposes, e.g., as the basis for the federal poverty guidelines that help determine eligibility for many safety net and other means-tested programs.

The Bureau begins by setting initial thresholds based on what the roughly 33rd percentile of households with two children spend on four basic needs — food, shelter, clothing and utilities.

It then bumps the amount up a bit to account for some other needs, e.g., household supplies, transportation that’s not work-related. It also makes some housing cost adjustments based on differences between major geographic areas and whether households rent or own — and in the latter case, with or without a mortgage.

Next, it deducts for certain other necessary expenses, e.g., work-related expenses, out-of-pocket costs for health care. And, as income, it adds the value of some non-cash benefits that households can use for the four basic needs. It also, for the same reason, folds in the refundable tax credits.

The report I link to at the beginning of this post provides a fuller — and considerably more wonkish — explanation.

* This is the same rate the Census Bureau reported last month. However, most of the official rates in the SPM report differ somewhat because the Bureau has included children under 15 who are unrelated to anyone they’re living with, e.g., foster children. The official measure doesn’t include them as part of a family unit.

I’m using the adjusted rates so we can have apples-to-apples comparisons. But the rates reported last month are those that should be used for other purposes.

UPDATE: The Center on Budget and Policy Priorities reports that the refundable tax credits lifted 9.4 million people out of poverty. This figure, it says, is based on its analysis of the SPM data. I don’t know why it’s higher than the Census Bureau figure I linked to.