By JASON DePARLE
The New York Times
July 27, 1998

WASHINGTON, July 24 -- As the welfare rolls continue to plunge, white recipients are leaving the system much faster than black and Hispanic recipients, pushing the minority share of the caseload to the highest level on record.

White, black and Hispanic recipients are all leaving welfare at unprecedented rates. But the disproportionately large exodus of whites has altered the racial balance in a program long rife with racial conflict and stereotypes, according to figures that were compiled in an analysis of recent state data by The New York Times.

The legacy of those stereotypes makes the discussion of race and welfare an unusually sensitive one. In the past, advocates and scholars have taken pains to note there were more white families on welfare than black. But that is no longer the case.

Blacks now outnumber whites. The Hispanic share of the rolls is growing fastest. And black and Hispanic recipients combined outnumber whites by about 2 to 1. In addition, the remaining caseload is increasingly concentrated in large cities.

Some analysts warn that the growing racial and urban imbalance could erode political support for welfare, especially when times turn tight. More immediately, the changing demographics suggest that states may need new strategies as they serve those left behind, like recipients who do not speak English.

Consider the changing nature of the New York City caseload, which is larger than that of every state but California. Since the city's rolls peaked in March 1995, the number of whites on welfare has fallen 57 percent. That is nearly twice the 30 percent rate of decline for blacks. And it is nearly eight times the decline for Hispanic recipients, which is just 7 percent, lagging the declines for blacks and whites as it has nationwide. The city's welfare rolls are now 5 percent white, 33 percent black and 59 percent Hispanic.

Illinois, Ohio, Pennsylvania, Florida, Michigan -- most of the states with large welfare populations like these have seen the number of whites on welfare declining faster than those of minorities. So have other states with significant caseload declines, like Wisconsin, Massachusetts and New Jersey.

The growing minority domination of the rolls is new, little-noticed and as yet largely unexplained. Most officials reacted with surprise when presented with the figures.

"Good grief!" said Representative E. Clay Shaw Jr., the Florida Republican who was the primary author of the 1996 Federal law that imposed time limits and work requirements. "That's definitely something we should look at. We don't want to leave one or two ethnic groups behind."

"I'm stunned," said Representative Robert T. Matsui, a California Democrat who is among the legislators most knowledgeable about the program. Mr. Matsui counts himself among those who have fought "a perception that welfare was a minority program. We tried to show that wasn't the case."

A Greater Distance To Cover for Minorities

There are a number of potential explanations for the changing racial demographics. They include possible discrimination by employers or by landlords in neighborhoods near jobs. In addition, unpublished data from the Census Bureau, prepared for The New York Times, suggest that minority recipients were significantly more disadvantaged than their white counterparts when the rolls peaked in 1994.

On average, they had less education, lower incomes and more children. They were less likely to have ever been married, a statistic that predicts lower rates of child support and lessened chances of leaving the rolls through a subsequent marriage. Perhaps most important, minority recipients were much more likely to live in poor, central city neighborhoods, far from the job growth that rings many cities.

Race is intertwined with place. Only 31 percent of white welfare families lived in city centers, the census data showed. But 63 percent of Hispanic welfare families lived in those job-scarce areas, as did 71 percent of blacks. This may also mean that minorities have faced less pressure from caseworkers to leave the rolls. Most states instituted their tough new rules outside the big cities, in regions with stronger economies and more responsive bureaucracies.

While the minority domination of the rolls could revive negative stereotypes, it comes at a time of unusual good will toward recipients of all races. With caseloads falling at a startling pace -- for minorities as well as whites -- taxpayers seem well-satisfied with the new ethos of time limits and work demands.

Flush with Federal money, states are investing in a variety of new employment services. And facing labor shortages, many corporations are courting a welfare population they once took care to avoid.

Citing that optimistic climate, some minority leaders say they do not expect welfare programs to attract new racial hostility. "If we had had this conversation six or seven years ago, it would have been a real concern," said Mayor Dennis Archer of Detroit, who is black. But as jobs increase and poverty declines, even in the inner cities, Mr. Archer said, a racial backlash is unlikely. "Even those insensitive to minorities aren't willing to just turn their back and withdraw services just because of race," he said.

Others were less sanguine. Representative Donald M. Payne, a Newark Democrat, warned that the growing minority share of the rolls could erode support for welfare spending and reinforce racial bias in general.

"Wedge-issue politicians always use welfare as an issue," said Mr. Payne, a former chairman of the Congressional Black Caucus. "There's no question that stereotyping will expand. Before, it was misinformation. It might even now go to codify the stereotyping."

A Growing Imbalance As Rolls Get Shorter

The most recent national figures on welfare and race are 17 months old. They show that the number of white families receiving Federal cash assistance declined 25 percent after the rolls peaked in 1994. By contrast, the number of black families fell 17 percent and that of Hispanic families 9 percent. But those differences appear to have widened in recent months, as the caseload declines have accelerated.

The New York Times surveyed 15 programs -- 14 states and New York City -- which account for nearly 70 percent of the nation's welfare population. Among them, only California had a sharper decline among blacks than whites: 40 percent for blacks versus 31 percent for whites. The Hispanic decline in California was slower, 22 percent.

In all other programs, the number of whites on welfare declined faster than those of black or Hispanic recipients. And in more than two-thirds of the programs studied, the white rate of decline outpaced both the black and Hispanic rates by at least 10 percentage points. In Wisconsin, where the caseload declines have been most dramatic, an astonishing 96 percent of white recipients have left the rolls. (Black recipients declined 74 percent and Hispanic recipients, 78 percent.)

A result is an added imbalance in a program that already had a disproportionate share of minorities. By early 1997, blacks accounted for 37 percent of the nation's welfare caseload, though they are just 13 percent of the general population. Hispanic families accounted for 22 percent of the welfare rolls, though they are 11 percent of the general population.

Whites, by contrast, accounted for just 35 percent of the rolls, though they are 73 percent of the population. That is the smallest white percentage since the Government began compiling figures in 1973. The vast majority receiving Federal aid are single mothers and their children.

As the rolls grow more dominated by minorities, they are also more concentrated in large cities. Detroit, Miami, St. Louis, Cleveland, Baltimore, Milwaukee and Philadelphia all saw their caseloads fall. But in each of those cities, the declines lagged the state average.

As a result, 48 percent of Pennsylvania's recipients now live in Philadelphia, up from 38 percent four years ago. In Wisconsin, virtually all the state's welfare recipients -- 85 percent -- now live in Milwaukee, up from 39 percent a decade ago.

But the urban lag is not universal. Atlanta, Boston, Los Angeles and Bridgeport, Conn., have all cut their rolls at a pace that matches that of their states as a whole. And the declines in New York City, Chicago and Newark have lagged the state average by only a small percentage.

Hispanic Share Is Growing the Fastest

Among the most striking trends is the growing Hispanic share of the caseload. As recently as 1983, Hispanic recipients accounted for just 12 percent of the nation's caseload, about half their current share. Some of that increase is owing to the growing Hispanic share of the general population, but Hispanic recipients have also been leaving the welfare rolls more slowly.

There are several possible explanations. Hispanic recipients lag blacks and whites in education levels and language skills. In addition, they tend to have larger families than white recipients. And some analysts suggest that Hispanic women face greater cultural pressures to stay at home with their children.

Citing the prevalence of language barriers, some advocates contend that Hispanic recipients need more training, especially in basic language skills. "I would view this as a wake-up call, that the system is not working as it is supposed to," said Charles Kamasaki, vice president of the National Council of La Raza, a Hispanic civil rights group.

But most states emphasize immediate job placements. "Non-English-speaking people have been coming to this country and finding work for years," said Dick Powers, a spokesman for the Massachusetts Department of Transitional Assistance.

The census data cast new light on the obstacles that minority recipients face. The data come from interviews with 5,400 welfare recipients in March 1994. One of the racial difference they reveal is in education. Among Hispanic recipients, 64 percent lacked a high school degree, as did 40 percent of the blacks. By contrast, only 33 percent of the white recipients lacked a high school degree. And those differences may understate the actual disparities in skills, because whites on average attend better schools than minority students.

A second difference is that minority women are less likely to marry. About 61 percent of the black women on welfare had never been married. About 40 percent of the Hispanic women had never been married and 31 percent of the whites.

Researchers are uncertain why black women marry at lower rates, but economics may play a role. William Julius Wilson, a Harvard sociologist, has argued that high rates of unemployment among black men makes them less attractive as potential spouses. Mr. Shaw, the Florida Republican, has proposed spending $2 billion over the next five years to raise the employment and marriage rates of welfare fathers.

A third difference is that minority women have larger families. Just 20 percent of white welfare recipients had more than two children, the census data show. But the figure for black and Hispanic recipients was nearly twice as high, 38 percent. Large families make it harder to find child care. They also reduce the economic rewards of working, since baby-sitting bills are higher.

A fourth explanation for the racial differences is on geography. The census data show 64 percent of black recipients lived in census tracts where at least a fifth of the population was poor. The figure for Hispanic recipients was also very high, 55 percent. But for whites it was just 21 percent. That not only suggests that black and Hispanic people live farther from jobs. It may also mean they have less work experience. "A lot of the people who have been on the rolls for the longest period of time have no one to vouch for them," said Bruce Katz, of the Brookings Institution, a research organization in Washington.

A fifth disadvantage can be seen in the census data. Minority recipients started out poorer. About 74 percent of black and 72 percent of Hispanic recipients spent the entire year in poverty, compared with 63 percent of whites. Given broader income trends, it follows that white recipients may also find it easier to turn to more prosperous relatives for help.

Among the unknowns is whether the racial imbalance will grow or diminish. "The big question is whether this is who leaves first, or who leaves ever," said Christopher Jencks, a Harvard sociologist. Like some others, Mr. Jencks warns that "the more black and Hispanic the program becomes, the more political pressure there is to cut back." Then again, he notes, "most people already thought that it was all black and Hispanic."

"So," he said, "it may not make as much difference as you might at first think."

Copyright © 1998 by The New York Times Co. Reprinted with permission.