Herald News (Passaic County, NJ), February 19, 2007, Monday

Copyright 2007 North Jersey Media Group Inc.,

All Rights Reserved

Herald News (Passaic County, NJ)

February 19, 2007 Monday

All Editions

SECTION: WORKING; Pg. D01

HEADLINE: Government's statistical snapshot gets another life

BYLINE: By ANDREA GURWITT, Herald News, North Jersey Media Group

BODY:

The SIPP is saved. For this year, at least.

The Survey of Income and Program Participation , a snapshot of statistics about people's lives, made it over the breach along with its Census Bureau comrades on Feb. 14, when the Senate passed a resolution to continue to fund government programs at 2006 levels.

But its last-minute rescue does not leave the survey to wander free. Next year will be its last, after which it will be replaced by what the Census Bureau touts as an improved and cheaper model. Researchers worry the new survey will not have been properly road-tested before it is put into use.

There's no box-office potential in the tragic life of a government data set, especially one with a name so dry as to seem parched, but if any survey may be said to have excitement potential, it's the Survey of Income and Program Participation. The SIPP is the beat cop of surveys, doing gritty, pound-the-pavement kind of work while its colleagues put in their 9-to-5's behind a desk.

The SIPP is a 23-year-old effort to track the paths of people's lives, from sources of income to housing and hunger, to the way their lives intersect with federal programs like welfare and health care. Participants are interviewed three times a year and are tracked from two to four years. By the end, an intimate, month-to-month, good-fortune-and-bad picture of their lives emerges.

Federal officials and researchers use the results to figure out if government programs are effective, how and when people use them, and when and why they stop.

The Department of Health and Human Services, for example, looks at SIPP numbers to understand the impact of welfare reform. The Heritage Foundation, a conservative think tank in Washington, D.C., uses the data to track how long people remain in poverty. The Poverty Research Institute at Legal Services of New Jersey looks at the national trends in hunger found in the SIPP to point to important issues to be tracked by the state.

The SIPP was once such a rising star, the National Academy of Sciences suggested it become the official source of income and poverty estimates.

The plan last year was to cut the survey funding from $33 million to $9 million, with only $3.6 million going for data collection -- a figure so inadequate, researchers said, as to be of real importance to the nation's poor -- to make room to pay for the 2010 Census. Data collection was planned to stop last September. The president's budget did not give the bureau enough money to run the two, and the bureau was forced to make a decision.

But assuming the president signs off on the joint budget resolution passed last week, the Census Bureau will have about $15 million to spend on the SIPP this year, said David Johnson, division chief for the bureau's housing and household economics statistics division, which oversees the SIPP. That means the SIPP lives, just in diminished capacity. Interviewers may be able to work through September, but will talk to half the number of people in their current batch, Johnson said.

But as the SIPP fades to black, the Census Bureau is grooming its new Dynamics of Economic Well-Being System (DEWS) for its 2009 debut. Sleeker, more efficient, less time-consuming, the DEWS is estimated to cost about $25 million a year. The bureau hopes fewer people will drop out of the new survey than the SIPP, whose attrition rate at one point hit 35 percent.

The biggest change is that the DEWS will question participants only once a year, instead of three times.

DEWS interviewers will use something called an event history calendar to jog people's memories of the past 12 months. They may ask something like, "Did you lose a job this past year? If so which month? Did you go on food stamps after and for how many months? You got married? Did you move out of subsidized housing before the wedding or after?" By remembering big events, people supposedly will be able to remember the smaller ones sprinkled around them.

This approach concerns Serena Rice, managing director of the Poverty Research Institute at Legal Services of New Jersey in Edison. People remember eviction or job loss better than other events like short periods of hunger, which may have a memory shelf-life of only four months, she said.

"There is a possibility there will be some underreporting (of hunger), given the 12-month time frame," she said.

Johnson could not say if the Census Bureau had looked at the question of hunger recall, but, he said, it hoped to work with researchers at the University of Michigan who use the event history calendar in their Panel Study of Income Dynamics "to evaluate some of these issues."

The DEWS questionnaire will also be shorter, and will include only core questions. The SIPP used core questions and "modules," additional, in-depth questions on different topics.

And so disability statistics may also be affected by the new, shorter survey, said Andrew Houtenville, senior research associate at the Employment and Disability Institute at Cornell University.

The SIPP had a disability module that included more than 50 questions that asked about the degree of severity of a disability, how long the disability lasted, if it affected someone's ability to do basic activities and if a disabled participant used personal assistance services, Houtenville said. These questions helped develop budget projections for government programs, and helped determine the number of disabled in the U.S., he said.

DEWS will include disability questions, but perhaps as few as five or six. They will be broad questions, and because they are broad and few, they may reduce the estimate of the number of disabled people nationwide.

"When you use fewer questions, you capture fewer people," Houtenville said.

The Census Bureau has not settled yet on the questions to include in the DEWS, Johnson said. And so researchers wait warily to see if the new survey will be as thorough and reliable as the SIPP.

"It's hard for me as a researcher to think this is a better method, but this is what their hope is," says economist Heather Boushey, with the Center for Economic Policy and Research, a liberal economic and social policy think tank in Washington.

"I think that's an open question -- the reliability of the information collected with the event history calendar," said Ralph Rector, senior research fellow and project manager at the Heritage Foundation.

"Measurement in and of itself is not too glamorous, but it does matter when it comes to developing programs to address issues faced by people with disabilities," Houtenville said.

The president's budget for 2008 does not include money for the SIPP. The DEWS interviewers in 2009 will use the event history calendar to ask participants about the previous year, Johnson said.

Reach Andrea Gurwitt at 973-569-7159 or gurwitt@northjersey.com.