The myriad, changing faces of poverty

This is a story about poverty, but maybe not the one you're used to.

Picture a child, Janessa, 3 years old, living in a hardscrabble neighborhood in Minneapolis with her single mother, who is just 19 and whose education consists of a high school diploma, working part-time as a hotel maid on the night shift. Grandma helps out because of the schedule. The refrigerator is mostly empty; thank heavens for the local food shelf.

Then there's Helen and Edgar, elderly folks with health problems and no pension, scratching to get by on Social Security and mostly failing. And there's Juanita, along with Mohammed, seeking but not necessarily finding a better life in America. Finally, there's Mike, the manufacturing worker whose job just got sent overseas.

All of them are today's poor. And in this instance, all of them are fictitious people. Actually, it's more accurate to call them stereotypes, and they're familiar because you regularly read individual profiles of the poor. They make for good reading because they offer real-life drama—the trials and tribulations, successes and (more often) disappointments in life. Maybe the stories are compelling because you've had similar experiences. Some research estimates that as many as half of all people are poor at some point in their lives.

But these shopworn stories are often misleading. No single story, or even group of stories, can encompass the many changing faces of the poor, the attributes of this population and the diverse circumstances that lead to poverty. Nor is all poverty created equal. Poverty tends to run high among college students and single-female households. Their respective prospects for getting out of poverty, however, are very different.

Neat and tidy stereotypes—the faces of today's poverty—belie the complexity of the issue. That's due in part to society's growing understanding of who is poor and why. Maybe you were hoping for a definitive answer about whether poverty is getting better or worse. You'd better look elsewhere, because the only appropriate one-word answer is "depends."

About the only thing you can say for sure about poor people is that they don't have much money. Most other conclusions might best be called generalizations, with exceptions so numerous that it's hard to consider them exceptions. Women are more likely than men to be poor; children more likely than working adults and the elderly; minorities more likely than white people. Single-parent families are more likely to be poor than married-couple families. The more education you have, the less likely you are to be poor. Native-born people are less likely to be poor than immigrants.

But start digging below the surface of those generalizations, and things get statistically sticky, and quickly. What about white, single-parent households? Blacks with high school or college degrees? Immigrants in North Dakota versus those in the Twin Cities? Women working part time versus full time? How does poverty compare to last year? How about two decades ago? How long are people poor? Has that changed much?

Broadly defined, poverty has trended up since about 2000, when the national economy went into recession. Some might say we're losing the battle on poverty, but that's probably too simplistic. Trying to categorize and depict poverty is like surveying options at the salad bar: There are a multitude of attributes or variables to consider and virtually no end to the possible combinations to investigate. Is poverty improving? The answer depends on which attributes, time frame, geography and comparison group you choose.

Because of this complexity, the sum of these articles amounts to this: Poverty is a multihued, multifaced, multifaceted phenomenon. It's about a confluence of factors that collide in different ways for different people. It's about race. It's about education. It's about family structure, workforce participation, gender, age, nativity, the economy, geography, socialization and more. It's even about arcane policy matters like how poverty is measured. What the headlines miss is that poverty is often getting better and worse at the same time, but for different groups or over different time frames.

This is not to downplay or dismiss the very real hardship that poverty inflicts upon tens of millions of people across the country, including many in the Ninth District. Rather, this and other fedgazette articles attempt to offer some context for thinking about the multidirectional trajectory of poverty.

We the poor

Poverty thresholds in the United States are defined rather simply in terms of income and household size. For a single person under 65 years old, poverty is anything below $10,160. For a family of three (any combination of kids and adults), the threshold is $15,577; a family of six, $26,683.

Now step back and take a historical look. As recently as 1949, the poverty rate was 40 percent. The rate plunged over the next couple of decades, particularly in the 1960s, dropping to a low of about 11 percent in 1973. Since then, poverty has bounced as high as 15 percent after recessions in the early 1980s and early 1990s. After the latter recession, poverty took a steady downward course and got as low as 11.3 percent in 2000—with some states like Minnesota and Wisconsin moving into the single digits.

But poverty has crept back up, hitting 12.7 percent in 2004, before declining by one-tenth of a percentage point last year, according to Census data released at the end of this summer. The story was mostly similar in Ninth District states (see chart).

Not surprisingly, the biggest factor in the general uptick in poverty is the economy. A 2005 study by Hilary Hoynes, Marianne Page and Ann Stevens (all at the University of California-Davis) shows that a "strong cyclical component in the poverty rate—with relatively higher poverty rates in high unemployment periods" closely mirrors the nation's recessions over the past 40 years.

But each state has unique demographic factors and economic circumstances. For example, Wisconsin is proportionally more dependent than other states on manufacturing, a sector that was particularly hard hit in the last recession. As a result, its poverty rate shot up from 7.9 percent in 2001 to 12.4 percent in 2004 before finally falling considerably last year. There were several other possible reasons for the initial steep increase in state poverty, according to Dan Veroff, a demographic specialist and director of the Applied Population Laboratory at the University of Wisconsin-Madison. For one, the state has seen higher immigration of late, and "recent arrivals are either underemployed or underpaid until they become more established."

Veroff also stressed that annual Census figures require careful reading, and the Bureau itself warns against overanalyzing state poverty rates in the short term. That's because Census figures are estimates, based on samples and then extrapolated to the broader population. As a result, the margin of error on state-level poverty estimates (in econ-speak, the standard deviations necessary for a high level of confidence in these estimates) covers a sizable range, but the average person only reads or hears about the final estimate.

For example, Wisconsin's poverty estimate in 2000 was 9.3 percent, but factor in the standard deviation (0.83), and the Census is basically saying that it's 90 percent sure that the state's poverty rate was somewhere between 8.5 percent and 10.1 percent, and its best estimate is right in the middle. In 2005, the state's poverty rate was 10.2 percent, with a standard deviation of 0.99, meaning that there is a statistical chance that last year's poverty rate was actually lower than in 2000. The point here is that poverty trends, particularly at the state level, are not quite as concrete as they are portrayed.

The big picture

While short-term poverty trends are hard to ignore, and even feel anxious about, it's the longer, larger, slow-moving trends that ultimately offer a better window on any progress toward tackling poverty.

At first glance, the big picture doesn't look so good either. Sure, poverty has fallen considerably since the 1950s, but it hasn't improved much in absolute terms over the past 30 years despite a steady increase in the country's gross domestic product. For children, the lack of progress is starker. From about 1969 through 1974, the child poverty rate was at or below 15 percent. It's been higher ever since.

The one clear winner over the past three decades has been the senior population, whose poverty rate declined significantly from about 35 percent in the mid-1960s to 10 percent in 2004 (see the snapshot of poverty by age group). But even here, alternative poverty measures that take into consideration medical costs for seniors paint a less rosy picture.

Such statistics make it easy to think we're losing the War on Poverty that President Lyndon Johnson famously declared more than four decades ago. But there's a lot happening below the surface of these broad poverty measures. Poverty rates for subgroups and those with certain characteristics have been changing, sometimes rapidly. While we cannot say that absolute poverty is getting better, especially in the short term, it is improving in a relative sense, especially for certain kinds of households that have historically been most at risk.

For example, few indicators are more closely correlated with poverty than family structure. The poverty rate for married-couple families in 2005 was 5.1 percent, while the rate for single-parent families headed by females was almost 29 percent. For the much smaller number of father-only families, the poverty rate is much lower—about 13 percent—mostly because fathers are more likely to be employed. Over time, you could say that family poverty is getting both better and worse: better because poverty in female-headed households was 36 percent as recently as 1993; worse because it had inched below 26 percent in 2000.

But this population has seen dramatic changes: The number of single-parent families (male- and female-headed) rose threefold from 1970 to 2005, while the number of married-couple households rose just 30 percent. In all, the proportion of single-parent households doubled to 26 percent over this period. While divorce has contributed to the single-parent trend, a larger contributor is births to unmarried mothers. From 1960 to 1999, births to single females went from about 5 percent of all births to about one-third. What does all of this mean? Single-parent households are at significantly higher risk for poverty compared with two-parent households, and the number of single-parent households is increasing rapidly. Yet over the long term, the poverty rate of this group is trending downward.

All other things held equal, Hoynes, Page and Stevens found that changes in family structure alone would have pushed poverty rates from 13.3 percent in 1967 to 17 percent in 2003, or about 5 percentage points higher than the actual rate.

One countervailing force that has helped balance the increase in single-parent households is steadily rising workforce participation by women over the past four decades. Indeed, if all other factors held constant, this trend would have chopped poverty rates significantly across the board.

The point is that poverty trends don't develop in a vacuum, but in consort with other trends. Today, it's more socially acceptable to be a single parent than in decades past; probably more important, rising wages and strong job growth over time mean that it's economically more feasible to be a single parent today than in the past. Is that good or bad? That depends on your social and political philosophy. From an economics standpoint, it's neither; trends merely represent the choices people make.

Job ladders, mobility and income

Also widely blamed for the lack of progress against poverty, particularly of late, is lagging median income. Advocates of the working poor accurately point out that wages for the lowest earners have risen slowly, especially when compared to higher-income earners.

According to a joint 2006 analysis by the Center on Budget and Policy Priorities and the Economic Policy Institute, income for the bottom 20 percent of families from 1980-82 to 2001-03 (using smoothed three-year averages) rose about 19 percent nationwide (inflation adjusted). But the top 20 percent of families saw incomes rise by 58 percent over this period. Income trends for district states were scattered around this national average, but the general trend remained intact—all district states saw slower average income growth in the bottom quintile, although income grew strongly and evenly at both ends in South Dakota, and North Dakota was close to parity (see chart).

Income growth has been particularly paltry for the very-low income. Using three-year averages for 1998-2000 and 2003-05, and adjusting for inflation, Census figures show that wages in the bottom decile rose less than 2 percent nationwide over this period; no district state saw more than a 3 percent increase, and Wisconsin actually dropped by four cents (to $7.90 an hour).

But the connection between median wages and poverty is complicated. Most income measures for individuals and families don't show a clear picture of actual income for individual workers or specific families. In other words, the bottom 20 percent of households in 2000 is not the same 20 percent of households in 2005. Every year, income goes up and down for individual workers and households; some households end up leaving the bottom income bracket, and others in higher brackets lose income and fall to lower brackets.

When income figures are released every year, the ensuing public discussion tends to imply that yesterday's poor are today's poor and that their incomes aren't rising. Certainly, that's the case for some. But poverty is mostly a transient, episodic experience: That is, it affects many people, but the vast majority of poor people are not poor their whole lives. Annual earnings have been found to be very volatile, particularly for those with low incomes. Research also shows that incomes of specific individuals and households (followed longitudinally) rise faster than the incomes of generic groups. (See more discussion on income mobility.)

The conversation about median wages and income also tends to undervalue the effect of work intensity on poverty. Despite what critics suggest is a nothing-but-crumbs job market, the majority of people who work are not poor. Although the poverty rate for those working at least 27 weeks in a year has been creeping up, rising from a low of 4.7 percent in 2000 to 5.6 percent in 2004, it remains well below 1993 levels of 6.7 percent. (See the sidebar on the working poor.)

No one answer

Such nuances make it hard to paint poverty trends in black-and-white tones. "It depends on your perspective," said Veroff of the UW Applied Population Lab. "On the ground, if you talk to service providers, they are seeing much bigger numbers." But this supposed influx of needy "is hard to square with the data" in a proportional sense, he said. "It's a very complicated task to evaluate poverty over the short term."

That's because there are exceptions to every supposed trend, even short-term ones. Take North Dakota, for example. The state's poverty rate zigzagged its way from 15 percent in 1998 down to 9.7 percent in 2004 before creeping back up last year. And despite comparatively low median incomes, child poverty is fairly low in the state. But it's not good news everywhere in the state. Poverty levels are still very high on Indian reservations, averaging about 25 percent, according to a 2006 report by the North Dakota State Data Center.

Richard Rathge, director of the center, said it's useful to view the incidence and severity of poverty in terms of three separate but interconnected dynamics. The first is the result of circumstance—the loss of a job or a health crisis. Things are going well, Rathge said, "and then all of the sudden" something happens to upset that economic balance. Such episodes are typically temporary. At the other end of the spectrum is persistent poverty, which can be endemic within certain populations or locations, such as Native American reservations in the state, Rathge said.

The third dynamic is "everything in the middle"—mostly characteristics that influence or indicate the likelihood of poverty, such as household structure, education, race and workforce participation, but also external, overarching variables like geography and the economy. North Dakota offers a good demonstration of how these variables mix and can even run counter to the broader trends.

Children, for example, typically account for a big proportion of the poor. Rathge said one of the reasons for falling poverty rates in North Dakota, particularly among children, is the departure of thousands of young people and families from the state. In 1980, almost 12,000 babies were born in the state. In the 1990s, the annual average number of births was about 8,500, dropping to 7,700 in 2000, according to the Data Center.

"As the number of families with children declines, you see an interesting correlate with the decline in poverty," said Rathge. But other factors are in play as well. The state has benefited from the energy crunch of the past few years. As the ninth-ranked state in oil production and a big coal producer, North Dakota is doing well. Most of the energy-based economic activity is occurring in the western part of the state, and coincidentally "that's the exact area of the state that's been struggling" with poverty, Rathge said.

A strong economy, very low unemployment and a small population of younger people also affects other populations that typically have higher-than-average poverty rates. Although North Dakota has relatively few immigrants, Rathge said their poverty rates tend to be lower than elsewhere because there are good work opportunities. Unlike the Hmong migration into the Twin Cities last year from refugee camps in Thailand, immigrants to North Dakota tend to come expressly to work and "don't stand still" very long if better opportunities beckon elsewhere, Rathge said. He noted that when a food processing plant shut down recently in West Fargo, many of its immigrant workers simply hopped the border to a turkey processing plant in Minnesota.

Poverty doesn't happen for one reason, but for many reasons, some of them temporary, like a job loss, some of them long-lasting, like the decision to drop out of high school. In the end, the ability to make inroads against poverty means getting a critical number of these factors—personal, political, social, economic, demographic and geographic—flowing in the right direction.

"There is never one answer, but a combination of answers" regarding trends in poverty, Rathge said. "We try to lump everyone together ... but poverty isn't homogeneous."