Skip to content

Hearing on Evaluating Efforts to Help Families Support their Children and Escape Poverty

July 17, 2013

Hearing on Evaluating Efforts to Help Families Support their Children and Escape Poverty










July 17, 2013


Printed for the use of the Committee on Ways and Means


DAVE CAMP, Michigan,Chairman

PAUL RYAN, Wisconsin
DEVIN NUNES, California
JIM GERLACH, Pennsylvania
TOM PRICE, Georgia
DIANE BLACK, Tennessee
TOM REED, New York
MIKE KELLY, Pennsylvania

RICHARD E. NEAL, Massachusetts
JOHN B. LARSON, Connecticut
RON KIND, Wisconsin

JENNIFER M. SAFAVIAN, Staff Director and General Counsel
JANICE MAYS, Minority Chief Counsel


DAVID G. REICHERT, Washington ,Chairman

MIKE KELLY, Pennsylvania
TOM REED, New York







Advisory of July 17, 2013 announcing the hearing


Jon Baron
President, Coalition for Evidence-Based Policy
Kristen Cox
Executive Director, Utah Governor’s Office of Management and Budget
Steve Aos, Director
Washington State Institute for Public Policy
David B. Muhlhausen, Ph.D.
Research Fellow, Empirical Policy Analysis, The Heritage Foundation
Tara Smith
Research Associate, Ray Marshall Center, Lyndon B. Johnson School of Public Affairs, The University of Texas



Wednesday, July 17, 2013
U.S. House of Representatives, 
Committee on Ways and Means, 
Washington, D.C. 


The subcommittee met, pursuant to call, at 4:06 p.m., in Room 1100, Longworth House Office Building, Hon. Dave Reichert [chairman of the subcommittee] presiding.

[The advisory of the hearing follows:]


Chairman Reichert.  Good afternoon, the committee will come to order.  This is the second in our series of three hearings on welfare reform.  In our first hearing, we learned that programs designed to help low‑income families often don’t do enough to help recipients go to work and get ahead.  Today we will explore what we know about the effectiveness of such programs, how we can hold more programs accountable for their performance, and how we can ensure they provide real help so recipients can support their families and move up the economic ladder.  Over one‑third of American households receive low‑income benefits today, and Federal spending alone on these programs equals $15,000 per individual below the poverty line each year.  Yet few programs can show that they improve outcomes for those in need. 

What we will hear today is that in many cases, these programs are either untested or have not been proven to work.  According to program evaluation experts, “Since 1990, there have been 10 instances in which an entire Federal social program has been evaluated using the scientific gold standard method of randomly assigning individuals to a program or control group.  Nine of the evaluations found weak or no positive effects.” 

In another example, a review of 13 rigorous studies on employment and training programs showed three‑quarters of them had weak or no positive effects on those that they were supposed to be helping.  All of this comes at a cost.  The programs in question continued to spend literally billions of dollars every year without delivering the results promised to those in need. 

We know many social programs lack meaningful outcomes, but some programs go further and can even be harmful.  For example, Scared Straight ‑‑ which I am familiar with as a former sheriff ‑‑ organized visits to prisons by juvenile delinquents with the goal of deterring them from future offending.  However, instead of reducing crime, these programs actually increased the odds that participants will find themselves in trouble in the future.  In fact, a comprehensive review of research by Washington State Institute for Public Policy estimated that every dollar spent on the program actually creates $76 in additional cost for taxpayers, crime victims, and the participants themselves because the youth who go through these programs are more likely to commit crimes in the future. 

This all suggests that more programs, including those in our jurisdiction, should be evaluated to ensure the families are receiving real help.  Ultimately, Congress and the administration should fund what works so we can deliver better results to those in need.  This is an issue that can and should be bipartisan as it is all about doing right by recipients and taxpayers alike. 

Last week the Obama administration hosted a full‑day conference on funding what works, highlighting how the private sector is willing to work with government to ensure that programs are really making a difference.  Especially given our current fiscal climate, it is important to ensure our resources are focused on efforts that have the greatest impact on those in need, and I am proud to say that my home State of Washington is a leader in this regard, as Steve Aos of the Washington State Institute for Public Policy will shortly tell us.  We will also hear from experts from Utah and Texas, as well as national leaders, about what is being done and what more can be done to ensure that these programs are held accountable for producing real results. I look forward to all of your testimony today. 

Mr. Doggett, would you care to make an opening statement? 

Mr. Doggett.  Thank you very much, Mr. Chairman, and thanks to all of our witnesses.  I welcome this opportunity to discuss the programs and strategies that have proven to be most successful in helping our families escape poverty.  Federal initiatives help raised 40 million Americans above the poverty line in 2011 under a comprehensive measure that counts all assistance known as the supplemental poverty measure. 

Taken as a whole, public policy is having an immense impact on the well‑being of many of our least fortunate neighbors.  This, however, still leaves the question of which specific approaches are most effective in achieving our objectives, and as we contemplate that question, I believe that the focus of this subcommittee ought to be on the one program, the Temporary Assistance for Needy Families, that is within our jurisdiction.  That ought to be our primary focus, especially since the TANF program is set to expire on September the 30th.  We have very few legislative days prior to that time, with the Congress being out most of August and the beginning of September, and I would suggest we get about the work on that specific piece of legislation. 

I voted for the 1996 welfare reform law myself because I believe that helping people find a job is the best strategy to reducing poverty. But this premise hinges on two very important principles.  First, assistance has to be available when jobs are scarce, as they have been until very recently; and, second, a real effort has to be made to help people find, maintain, and advance in employment.  Any fair reading of the last decade of the TANF program finds it lacking on both counts.  The percentage of poor single mothers who are working has been dropping almost consistently for the past 12 years, after having made significant progress in the mid and late 1990s. 

Even more troubling, the percentage of poor mothers who are neither working nor receiving any assistance from TANF is more than twice as high as it was when TANF was established in 1996. 

Some of our colleagues often complain that our Federal programs are allowed to drift on autopilot.  That seems to me to be accurate as it relates to TANF.  This program is in real need of a significant reevaluation rather than this stop‑start for brief periods approach that has been taken in recent years.  Instead of working toward that goal, we spent most of the last year in this subcommittee debating whether the administration was giving the States too much flexibility in the TANF program. 

For those who think that work requirements, stricter work requirements constitute a panacea on this issue, it is noteworthy that a number of States, including those that have Republican Governors, have complained that the current TANF work participation requirements really don’t measure success.  Rather than continue the same tired old arguments, our committee can actively advance the debate on this issue by reviewing evidence on specific strategies that might help TANF recipients get and retain jobs.  One promising approach is boosting both employment and earnings through sectoral training programs that target high‑demand occupations and provide training and job search assistance to low‑income individuals. 

Unlike some past training programs, these efforts are squarely aimed at preparing folks for job opportunities that exist in their communities.  I look forward especially to having a native from Austin, Tara Smith, with the Ray Marshall Center at the University of Texas offer comments about the success that is reflected in Capital IDEA in Austin and Project QUEST in San Antonio that have shown real promise in helping people find not only jobs, but lasting careers.  The Alamo Academies in San Antonio have taken this same successful sectoral employment approach and have partnered with high schools, community colleges, aerospace companies at Port San Antonio to provide specialized advanced manufacturing training. 

I have been out to meet with some of those students.  They are impressive.  They are high school students who complete the program and receive valuable credentials along with their high school diploma when graduating, and some are averaging a starting pay of over $30,000 out of school each year. 

Mr. Chairman, I look forward to a productive discussion about how these and other proven strategies might help us to improve outcomes for TANF recipients and other struggling Americans.  Let’s find a path forward toward our common goal of increasing employment and reducing poverty.  Thank you very much. 

Chairman Reichert.  Thank you, Mr. Doggett.  Without objection, each member will have the opportunity to submit a written statement and have it included in the record at this point. 

I want to remind our witnesses, please, to limit their oral statements to 5 minutes.  However, without objection, all of the written testimony will be made a part of the permanent record. 

On our panel this afternoon we will be hearing from Jon Baron, president, Coalition for Evidence‑Based Policy; Kristen Cox, executive director, Utah’s Governor’s Office of Management and Budget; Steve Aos, director of Washington State Institute for Public Policy; David Muhlhausen, Ph.D. research fellow, Empirical Policy Analysis, The Heritage Foundation; and Tara Smith, research associate, Ray Marshall Center, Lyndon B. Johnson School of Public Affairs at The University of Texas. 

Welcome to all of you.  Thank you for taking the time to be with us today.  I will just let you know that you see three members in front of you.  Others are on the floor speaking on a bill, which I just came back from.  That is why we started a little bit late.  So we will have some other members joining us here shortly. 

Mr. Baron, please proceed with your testimony. 

Mr. Baron.  Thank you, Mr. Chairman.  Chairman Reichert, Ranking Member Doggett, and Congressman Davis, I appreciate the opportunity to testify before the subcommittee on behalf of the nonpartisan, nonprofit Coalition for Evidence‑Based Policy.  My testimony will address how evidence‑based program reforms can greatly increase the effectiveness of government social spending in improving people’s lives. 

It is often assumed that the only way to increase government’s impact on social problems such as poverty and educational failure is to spend more money, an assumption that conflicts with the current national interest in reducing the deficit.  Largely overlooked, however, are clear examples from welfare and other areas where rigorous randomized trials, which as you mentioned, are widely considered the strongest method of evaluating program effectiveness, have identified program reforms that produced important improvements in people’s lives while simultaneously reducing government spending. 

As an illustrative example in the 1980s and 1990s, government and foundations sponsored a large number of randomized trials of State and local welfare reforms.  Three major reforms ‑‑ two in California, one in Oregon ‑‑ were found especially effective.  They focused on moving welfare recipients quickly into the workforce through short‑term job search, assistance, and training, and were found to produce gains in participants’ employment and earnings of 20 to 50 percent sustained over several years. 

Importantly, they also produced net savings to the government in reduced welfare and food stamps of between $1700 and $6,000 per person.  These findings helped build the political consensus for the strong work requirements in the 1996 Welfare Reform Act. 

A second example is in foster care where in the late 1990s, HHS granted Illinois a waiver from Federal law to implement subsidized guardianship, which is an alternative to foster care in which the State pays a subsidy to the child’s relative or foster parent to serve as their subsidized guardian, as their legal guardian. 

Illinois evaluated subsidized guardianship in a large randomized trial which, over a 9‑year period, found that the program increased children’s placement in a permanent home by 8 percent, reduced their days in foster care by 16 percent, and produced net savings to the foster care system of about $2300 per child.  Based on those findings, CBO scored savings of $800 million for Federal legislation that was enacted in 2008 to expand subsidized guardianship nationally.  To identify enough of these reforms to generate broad‑based improvement in government effectiveness will require strategic trial and error.  In other words, rigorously testing many promising reforms to identify the few that are effective.  The instances of effectiveness that I just described are exceptions that have emerged from testing a much larger pool. 

More generally, most innovations, typically 80 to 90 percent, are found to produce weak or no positive effects when rigorously evaluated, a pattern that occurs not just in social spending, but in other fields where randomized trials are done, including medicine and business. 

In my testimony, I offer concrete suggestions for the subcommittee’s consideration to greatly accelerate the rate of program innovation and rigorous testing in social spending so as to grow the number of proven cost saving reforms like those I discussed.  I suggest, for example, authorizing greater use of Federal waivers to stimulate State and local innovation and evidence building, which was a tool deployed with great success in welfare reform under both Republican and Democratic administrations.  I also suggest steps this subcommittee can take to facilitate greater use of low cost randomized control trials, such as the subsidized guardianship trial that I described earlier, which cost just $100,000 to conduct, yet identified an innovation that CBO scored as saving $800 million.  These suggestions are designed to catalyze evidence‑driven improvements in a social spending system that, in many cases, has fallen well short of its intended goals. 

The American poverty rate, for example, now at 15 percent, has shown little change, whether by official or the supplemental measures, the National Academy measures since the 1970s.  In K‑12 education, reading and math achievement of 17‑year‑olds, who are the end product of our K‑12 system, is virtually unchanged over the past 40 years according to official measures, even though there has been a 90 percent increase in public spending per student, adjusted for inflation, since that time.  Evidence‑based policy offers a demonstrated path to more effective, less expensive government.  Thank you. 

Chairman Reichert.  Thank you, Mr. Baron. 

[The statement of Mr. Baron follows:]

Chairman Reichert.  Ms. Cox, please. 


Ms. Cox.  Thanks for having me here, Chairman and Ranking Member Doggett.  It is an honor to be here.  I am the executive director of the Governor’s Office of Management and Budget, and prior to this position, I was the executive director of the Department of Workforce Services.  We oversaw the implementation and administration of over 90 different Federal programs, which included everything from TANF and food stamps to child care to housing, a plethora of services that impact low‑income individuals. I come to this discussion with a real on‑the‑ground perspective in how you actually operationalize evidence‑based practices and penetrate into the day‑to‑day work of our folks.

Let me go through some of what we do and some things for you to think about.  First of all, our goal in Utah is to improve the operations of all of our systems in Utah State government by 25 percent over the next 3‑1/2 years, the remainder of Governor Herbert’s administration.  It is a bold initiative, but we think there is ample capacity in all of our systems to do better, and in social services, it really resides on integrating evidence‑based practices into the day‑to‑day work of our employees, so they are spending more and more of their time doing what works, and less time doing the things that don’t or are wrapped up in compliance initiatives, which is part of the Federal bureaucracy. 

A few things, just observations we have before we get into the evidence‑based practices.  One, there is significant goal disalignment against ‑‑ across services that serve low‑income individuals, so the ability to assess if a program has been successful or not is so dependent on the point and policy objective of the program that they are all over the place.  For example, the Food Stamp Act of 1964’s intended policy objectives was twofold, to promote the agricultural economy, and to give nutritional sustenance to low‑income individuals.  With amendments, the ABAP program, employment and training services were offered, but really more as an eligibility criteria than a true strategy to move people to work.  Add housing initiatives, TANF, Medicaid, across the board, the policy objectives are different, so when we really want to talk about the impact to low‑income individuals, we need to be clear on what our intended purpose is, and we don’t have that right now. 

Second piece of concern is the ability of measurement.  We have too many contradictory and conflicting measures out there in the public service arena. As an administrator of 90 different programs, trying to get clarity on how well I am doing is a challenge.  We have created a very simple ratio, quality throughput divided by operating expense, which will baseline and drive all of our performance in State government in Utah.  It is not so simple in the Federal navigation system of measurements, even in common core.  In Utah, we were very aggressive about understanding how important evidence‑based practices was.  When I was executive director of Workforce Services, we set up an evidence‑based arm of our agency specifically to do randomized sampling, propensity scoring, everything we needed to know to analyze and assess TANF participation, job training, does it work or not, but what we found is what the national studies say on a very universal level aren’t necessarily true for the unique demographics in Utah.  Even within Utah, we saw variations from region to region. 

So while evidence‑based practices at the national scale are important to help direct Federal policy, the States need the ability, flexibility, and resources to create that same ability at the local level to really fine‑tune and penetrate those evidence‑based practices into our system. 

The next piece is penetrating evidence‑based practices into our operations.  It is great to have theory, it is great to have tons of data, you can Google and you can find hundreds of social service evidence‑based practice reports, but why aren’t they penetrating our system?  Part of it is our folks are so heavily focused on compliance activities that they don’t have the time to step back and think about what should we be doing.  It is not an excuse, but it is a reality for people.  Operationalizing takes the ability to translate global goals into the day‑to‑day work of our employees, it requires that we have clear policy objectives for them, it really requires that we stop doing the stuff that doesn’t work and start doing what does. My hope is for every new policy initiative Congress puts out, they eliminate another one.  There is only 8 hours in a day, it is the employee’s biggest constraint.  If you want them to really spend 80 percent, 90 percent of the time that makes the biggest difference, then we have got to strip away the 70 percent of the stuff that is junk. 

A few suggestions or hopes or recommendations.  A lot of demonstration projects are going on at the national level.  Fantastic.  Push some of those resources to the State level so we can customize our own practices that we need to make it relevant for us.  Align the policy objectives, which is simple.  Not simple, but critical.  Simplify the measures, as you said so well, give States more flexibility in the ability to innovate, hold us accountable, but give us the ability to be innovative.  Thank you very much. 

Chairman Reichert.  Thank you, Ms. Cox.

[The statement of Ms. Cox follows:]

Chairman Reichert.  Mr. Aos.


Mr. Aos.  Mr. Chairman, members of the committee, my name is Steve Aos, and I am the director of the Washington State Institute for Public Policy.  You asked me to come today and provide testimony on how the Washington State legislature, another legislative body, is using the Institute to try to reform public policies in the State of Washington, so I am going to give a little bit of an overview of the Institute and how the legislature uses it and then talk about some of the specific applications of the approach in Washington. 

The Institute was created 30 years ago in 1983.  The legislature passed a joint resolution in that session and said we want to have an institute to help the legislative branch of government figure some things out.  The key aspect of the Institute is that all of the assignments to us, including the ones I will talk about today, come to us because the legislature passes a study bill, goes through both Houses, passes both Houses, and is signed by the Governor.  We don’t respond to an individual member’s request, but a bill has to go through and pass for it to have a study undertaken.  And in recent years what the legislature has been asking us to do, in about the last 15 years, is to find out what works and what doesn’t work to achieve particular public policy outcomes.  So we will get a direction that will say how can Washington public policy affect the crime rate in Washington State, or the rate of child abuse and neglect, or how can we get greater high school graduation levels in Washington State, and it will say to the Institute, tell us what works and what doesn’t to achieve those outcomes. 

It is a nonpartisan group, an equal number of Republicans and Democrats are on our board of directors, and they are cochaired by one Republican and one Democrat at all times, so it was set up to be a nonpartisan group, and that is how we operate. 

What works and what doesn’t work has been something that the legislature has been finding particularly attractive in learning how to do the numbers on the one hand and the legislative process has been learning how to use the numbers to actually craft budgets and policies in an increasing number of important areas.  When we do this work, when we get that assignment to say what works, we are playing the role of the investment adviser, if you will.  We make buy and sell decisions to our legislature, and we look all around the country and all around the world at all of the most rigorous evaluations on a given topic.  We throw out evaluations that we think aren’t rigorous enough to warrant any further consideration, and then we assess all that in a systematic way, and we do a cost‑benefit analysis, and we come back to our legislature saying this thing looks like it is a winner, this thing is maybe iffy, and this thing looks like a loser in terms of benefit cost. 

You mentioned the Scared Straight program in your opening remarks.  We have got lots of losers.  We love to find losers because then if we are already funding those kinds of programs in Washington, they have become things that we can then cut in terms of programs that are ineffective.  So that is the role that the legislature has had us do consistently over time.  The hallmark of our work has been to take not only what works but actually to do a benefit‑cost analysis of each of those of that effort. 

Crime policy has been the area where we have moved it the furthest, but we are moving ahead in K‑12 education and child welfare and some of the topics that are directly before this committee right now.  We now actually can point to lower crime rates in the State of Washington and the reduced level of taxpayer spending in the State of Washington as a result of all that work and all the previous budgetary decisions that have been made from that work. 

The latest approach in Washington, passed unanimously in the last two sessions, has been to take the Institute’s list of what works and what doesn’t work and send a message to the executive agencies saying align everything that you do and tell us what approximates the Institute’s list and what isn’t on the Institute’s list, and those reports will then come back to the legislature from the executive agencies in about five or six different areas of public policy.  So this is the legislature’s attempt to try to take the information we have been doing and actually craft budgets around it by giving the executive branch a time to respond, saying are you doing things that the Institute has found to work or not work. 

Mr. Chairman, I just wanted to give an overview today about how the legislature back in your home State has been using this information.  By the way, it is 68 degrees back in your home State today, I just checked.  I am happy to get back there in a few hours.  It is real progress.  Session by session, we get better and better at doing the work, the legislature gets used to asking the question and taking the information back, and it is a nonpartisan effort, that is the thing that is perhaps most encouraging.  Thank you for allowing me to testify. 

Chairman Reichert.  Thank you.  I think 68 is the humidity level here in D.C. if I am not mistaken.  At least. 

Thank you for your testimony, Mr. Aos.

[The statement of Mr. Aos follows:]

Chairman Reichert.  Thank you, Mr. Muhlhausen, for being here, and you are up for 5 minutes.


Mr. Muhlhausen.  Thank you.  My name is David Muhlhausen, I am a research fellow in empirical policy analysis in the Center for Data Analysis at The Heritage Foundation.  I thank Chairman Reichert, Ranking Member Doggett, and the rest of the committee for the opportunity to testify today on the need to evaluate Federal social programs.  The views I express in this testimony are my own and should not be construed to represent any official position of The Heritage Foundation.  My testimony is based on my recently published book, Do Federal Social Programs Work? 

My spoken testimony will focus on three points:  First, the best method for assessing the effectiveness of Federal social programs is large‑scale, multi‑site experimental impact evaluations that use random assignment.  Unfortunately, these scientific rigorous assessments are rarely done.  From my count, only 20 large‑scale multi‑site experimental impact evaluations assessing the effectiveness of 21 Federal programs have been published since 1990. 

The consequence of so few Federal social programs being rigorously assessed means that Congress has no credible information on the performance of the majority of social programs.  To solve this problem, Congress should specifically mandate the multi‑site experimental evaluation of these programs.  When Congress creates social programs, the funding activities are intended to be spread out across the Nation.  For this reason, Federal social programs should be assessed for national effectiveness.  While an individual program operating at a single site may undergo an experimental evaluation, the small scale single site evaluation would not inform Federal policymakers of the general effectiveness of the broader national program. 

The success of a single program that serves a particular jurisdiction or a population does not necessarily mean that the same program will achieve similar success in other jurisdictions or among different populations, thus small‑scale evaluations are poor substitutes for large‑scale multi‑site evaluations.  A multi‑site experimental evaluation that uses the performance of a program operating in numerous and diverse settings will produce results that are more informative to policymakers. 

Second, the Federal Government does not have a good record of replicating successful social programs on a national scale.  Policymakers and advocates often assume that a social program that is effective in one setting will automatically produce the same result in other settings.  This is a faulty assumption.  For example, for the Center for Employment Training replication, the Federal Government attempted to replicate the successful outcomes of a youth job training program in San Jose, California, in 12 locations throughout the United States.  A multi‑site experimental evaluation found that the Federal government was unable to replicate the successful outcomes in these other settings.  Just because an innovative program appears to have worked in one location does not mean that the program can be effectively implemented on a larger scale. 

Third, policymakers should be mindful that Federal social programs do occasionally produce harmful impacts on recipients.  However, social program advocates too frequently ignore these findings.  Nevertheless, Congress should be aware of these harmful impacts.  Here are two examples.  From the 3‑year‑old cohort of the Head Start Impact Study, kindergarten teachers report that the math abilities of children given access to Head Start were worse than similar children not given access to Head Start. 

Students participating in educational, after‑school educational activities under the 21st Century Learning Centers Program, were more likely to have disciplinary and behavioral problems such as getting suspended from school.  Further, they were less likely to achieve at high levels in class and were less likely to put effort into English classes.  With the Federal debt reaching staggering heights, the best method for assessing the effectiveness of social programs and making sure that money is spent wisely are large‑scale multi‑site experimental evaluations, yet to date, this method has been used in only a handful of Federal programs.  Congress needs to reverse this trend of not rigorously evaluating Federal social programs.  Thank you. 

Chairman Reichert.  Thank you, Mr. Muhlhausen.

[The statement of Mr. Muhlhausen follows:]

Chairman Reichert.  Ms. Smith, you are recognized for 5 minutes. 


Ms. Smith.  Thank you.  Good afternoon, Chairman Reichert, Ranking Member Doggett, and members of the committee.  My name is Tara Smith, I am with the Ray Marshall Center for the Study of Human Resources at the University of Texas at Austin’s LBJ School of Public Affairs.  Thank you for inviting me today. 

This hearing’s focus on families coincides with a growing body of research on two‑generation programs designed to link services for children and parents so that families as a whole can build the human capital they need to succeed in school and the labor market.  Today, I would like to share findings and lessons learned from evaluations of two such programs. 

Capital IDEA is a sectoral‑based training program in Austin, Texas, that was built on a model pioneered by Project QUEST in San Antonio for employer‑driven workforce development.  Capital IDEA provides training primarily in health care for low‑income and disadvantaged adults.  The evaluation tracks participants from 2003 forward and includes outcome, impact, and ROI analyses.  CareerAdvance is a career pathways program in Tulsa, Oklahoma, which provides parents of Head Start and early Head Start students with training for health care occupations.  This program launched in 2009, and the evaluation includes an implementation study as well as outcome and impact studies focused on parents and children. 

The impact evaluations for both programs use quasi‑experimental research methods based on a carefully matched comparison group.  There are five key points I would like to emphasize about these evaluations.  First, rigorous quasi‑experimental methods have been found to produce impact estimates similar to those found in random control trials.  Quasi‑experimental methods also address issues such as the localized nature of programs which limits the pool of prospective and eligible applicants needed to support a random control trial. 

Second, the use of administrative records and propensity score matching techniques helps to keep evaluation costs reasonable.  In both evaluations State UI records provide consistent, comprehensive, and inexpensive data on employment and earnings.  For Capital IDEA, the comparison group is drawn from individuals who receive job search assistance at a local one‑stop career center and who closely resemble participants along 18 characteristics, including demographics and prior employment and earnings history.  For CareerAdvance, the comparison group is drawn from other Head Start parents matched along multiple characteristics, including a documented interest in pursuing further education and training.  The rigor of the comparison group matching design undergirds our confidence in the evaluation findings. 

Third, sectoral and career pathway models have demonstrated effectiveness in a number of industries and labor markets by connecting low‑income and low‑skilled adults with the training they need to enter higher paying careers.  If you have my written testimony in front of you, the chart on page 3 shows our most recent findings from Capital IDEA.  We find that on average the earnings of participants continue to grow over time while comparison group members who receive only job search assistance or other basic workforce services tend to have relatively flat earnings. 

My fourth point is that two‑generation strategies which look to build on sectoral or career pathway programs by linking those adult education and workforce training services with high quality educational opportunities for children show promise.  Wrap‑around and family support services including child care, transportation assistance, counseling, and other resources ensure that participants in both generations receive the help they need to achieve at a high level. 

In CareerAdvance this support can include a monthly financial incentive for performance and attendance to help offset the costs of participation and provide some financial stability for the family. 

Finally, because social programs rarely involve cookie cutter approaches, it is important to consider a broad base of evidence when evaluating program effectiveness.  Implementation and process evaluations provide important context for understanding how programs operate and identifying which services may lead to better outcomes and impacts over time.  This is particularly true for new and emerging programs and program replication efforts. 

In conclusion, strategies that focus on basic skills which provide counseling and other support services, which increase opportunities to earn and learn so that parents can support their families while in training and target skill development at high wage, high demand occupations in the local labor market all appear to have significant rigorous evaluation support and could be promoted in Federal programs. 

By investing in proven approaches and promising strategies, such as two‑generation initiatives, and supporting a broad range of research and evaluation efforts on those investments, the Federal Government can play an integral role in building the knowledge base needed to expand and improve efforts to move families out of poverty.  Thank you. 

Chairman Reichert.  Thank you, Ms. Smith.

[The statement of Ms. Smith follows:]

Chairman Reichert.  And for your information, panelists, now is the question‑and‑answer phase, so I am sure the members on the committee would like to ask a few questions.  I will start with Mr. Baron.  In your testimony, you note the important groundwork laid for welfare reform when high quality experiments were conducted in the 1980s and 1990s to find the best way to help people move from welfare to work.  In fact, this research helped shape the successful 1996 reforms which created TANF. 

Beyond TANF, however, the subcommittee oversees other programs that haven’t benefited from the type of experimentation and high quality evaluation that led to welfare reform.  Given where we are now, in your opinion, what should we be focused on first?  What would be our number one priority, and is this just the first thing we need to focus on and are there any other priorities that kind of fall off of that?  I would be interested in hearing that. 

Mr. Baron.  Thank you.  One of the reasons why there is such a large body of strong, in many cases, replicated evidence from randomized trials in Welfare‑to‑Work is because the Federal Government had in place for many years, starting at the end of the Reagan administration through the first Bush administration, and then into the Clinton administration, a waiver evaluation policy, meaning the Federal Government said to the States, we will allow you to do your own welfare reform demonstrations. We, the Feds, will waive provisions of law and regulation to allow you to do those reforms if, and here was the quid pro quo, you do a rigorous evaluation, usually a randomized evaluation, to determine whether it works or not. That policy, that waiver evaluation policy resulted in more than 20 large‑scale randomized control trials that contributed to the important body of knowledge, the evidence that helped inform the work‑focused 1999 Welfare Reform Act. 

That kind of waiver-evaluation approach could be used in many other programs.  The same general concept, it would have to be adapted, it could be used in unemployment insurance, in foster care, in SSI, and disability insurance, and other areas to allow State and local innovation, open the door fairly wide, coupled with a requirement for rigorous evaluation to determine which of those innovations really work and which do not. 

That is something that ‑‑ and also importantly, as in welfare, with a requirement for cost neutrality, so that you are testing innovations that are designed to improve people’s lives while not adding to the deficit or that are cost saving while also improving people’s lives or not causing any harm. 

Chairman Reichert.  Appreciate that.  Thank you.

Ms. Cox, you describe how Utah has a specific division in the Department of Workforce Services focused on building evidence about the effectiveness of programs.  What did you find most difficult about measuring effectiveness?  I am going to guess one of the things was the disalignment piece that you spoke about. 

Ms. Cox.  Uh‑huh.

Chairman Reichert.  How did you use this information to make decisions about which programs were managed and funded? 

Ms. Cox.  Well, you know, I have the same question I think you raised earlier, do job training programs work or not? I had heard a lot of the literature, just like you had, but we wanted to test it in Utah.  So we did a really rigorous assessment longitudinal, we did the whole randomization, and looked at ‑‑ this is just one of the studies, for example, in job training, did it make a difference or not.  The bigger question was sometimes, and what the type of job services that were delivered and when. 

So we found, for example, unpaid internships really didn’t work, paid internships didn’t work in all parts of the State except for one place, so it was replicating why that worked.  We found that occupational training tied to an employer did work. 

So there are things we found that did and didn’t work, but the interesting thing we found, for example, is that when they completed the training was a big variable.  We had too many people, 50 percent of our folks coming into the job training program, starting it, stopping halfway through or a third of the way through.  The taxpayers lost the up-front cost, and the person doesn’t get the benefit, so it really forced us to look at new strategies on completion.  Same thing with TANF participation.  What things do people tend to participate in without us having to chase paperwork.  They tended to be employment related and things that related to their lives and then which of those ended up helping them become self‑sufficient and moving to jobs.  We were able to more narrowly tailor what kind of participation activities we focus on.  The challenges are, there is not a budget, there is not an appropriation in these Federal programs.  Like I said, 80 percent of our budget at Workforce Services was Federal.  It wasn’t a set‑aside amount of money saying, here, do evidence‑based practices, it is something we had to kind of internally create and cobble together some funds for that to happen, and there is not an appropriation directly for that.  That funding piece is a challenge because while we talk at this level, and you guys have to make those decisions because you have such a huge impact on the Nation, once you get into the operational level, people kind of may not take it as seriously, and there is not a requirement or mandate with some cases like this for that to really happen.  As we go now, a look at all the other services in State government will be doing the same thing. 

Chairman Reichert.  Thank you for that answer. My time has expired, but I am going to ask Mr. Aos a quick question because I know he has to leave and catch an airplane here in a few minutes.  Could what you do in Washington State be replicated at the Federal level, and what are some of the key challenges we might have to overcome here in the Capital to do so? 

Mr. Aos.  Mr. Chairman, I have worked in the State capital of Washington for 36 years, so I know that place pretty well.  I don’t know this place very well, so I am not going to be the one to give you advice on can what be done in Washington State be transferred to Washington, D.C.  I think the principle is that it works so well in Washington State is that the request for this information is bipartisan.  We rarely get a demand for a study that only comes from one party or the other.  It is almost unanimous votes, that they want to find out what reduces crime or what gets more high school graduates. 

The other thing that we have done then is that the rigor with which we as the people that draw the information go through to assess the evidence fairly and to use return on investment analysis to rank options because you can find things that work that cost an awful lot more money than the benefits they derive, so it is that aspect. 

And, then, finally, I would just add that we use that evidence to cut programs in addition to add programs that work, and I think that that message that the legislature is using evidence to change budgeting up and down has resonated around and causes actions and responsiveness to the notion so that evidence‑based doesn’t mean just a code word for spend more money. 

Chairman Reichert.  Thank you for your answer.  Mr. Doggett, you are recognized. 

Mr. Doggett.  Thank you, Mr. Chairman.  Mr. Aos, I notice in your evaluation it looks like as far as child welfare is concerned that the Nurse Family Partnership, the visiting nurse program for low‑income families, is way out on top as being the most cost‑effective program. 

Mr. Aos.  Yes, it is right near the top of our list.  I think it is also a buy recommendation from Jon Baron’s group as well. 

Mr. Doggett.  Exactly.  Thank you.  I wanted to ask him also, what is it about this program that seems to have the most benefit? 

Mr. Baron.  The program is very well designed in the following sense:  It is for women who are poor, pregnant with their first child, and most of them are single.  They are visited by a nurse, many of them ‑‑ you know, they are pregnant with their first child, they are concerned about their health, so they are particularly receptive to advice from a nurse.  The nurse teaches them basic parenting, nutrition, not to smoke or drink during pregnancy.  If they are interested in practicing birth control, how to do it effectively. One of the reasons this program is on the top of Steve’s list and the top of our top tier panel’s list is that it has been evaluated in three different randomized control trials, in different cities, different ethnic groups, actually different decades. 

In all three cases, it was found to produce large improvements in life outcomes including, for example, a 40 to 50 percent reduction in incidents of child abuse and neglect and hospitalization. 

Mr. Doggett.  You are aware that the Federal funding for that program expires in little more than a year, the Federal Home Visiting Program.  Do you favor its extension? 

Mr. Baron.  Definitely.  In many Federal social programs, evidence plays little role in how funds get allocated, whether they are formula programs or even most competitive grant programs.  The evidence‑based home visiting program is an important exception to that.  Evidence, especially for the largest grants plays a central role in determining what gets funded. 

Mr. Doggett.  Thank you very much. 

Ms. Smith, let me talk to you about another innovative program that you focused on with Capital IDEA and Project QUEST.  This is not only about just securing any job that is there, but as you mentioned a career pathway so that a person has hopes of not only getting a job but getting a job that will help them support their family at a livable wage. As I understand the program, again, it is not just about how you have become a radiology technician or someone who works in semiconductors, but it is about getting some counseling to go along with that training to be sure that you are able to fulfill all the responsibilities.  Can you elaborate a little on how those programs work? 

Ms. Smith.  Yes.  The Capital IDEA model provides the occupational training as a connection with the associate degree program or community college program that builds that occupational skill, but they also work on building the soft skills that are important in the workplace and make someone a successful employee, and so through weekly sessions with a career coach, participants go through and talk about issues like time management, communication and interpersonal relationship skills, and work on building kind of some self‑confidence that they can take into the workplace and make sure they are going to be a valuable employee.

Mr. Doggett.  And how might those programs interface with TANF?  Is there the potential to assist more TANF recipients, but to help them achieve some of the same success that Capital IDEA and Project QUEST are already achieving? 

Ms. Smith.  Certainly.  Actually both of those programs, as well as the Career Advance program in Tulsa, serve TANF recipients already.  They are part of that low income and disadvantaged group that these programs are explicitly trying to move forward in the workplace. 

Mr. Doggett.  Are there other recommendations that you have that we should consider as we are renewing and reauthorizing the TANF program to assure that more economically disadvantaged people actually move into living wage jobs? 

Ms. Smith.  Yes, I think reconsidering the work requirements to allow individuals to engage in that longer term intensive skill development that has been shown to lead to higher paying careers that actually move people out of poverty would be an important change to consider.  The emphasis on work first with a very short term emphasis on job achieving skills hasn’t been shown to be effective in the same way that building an occupational credential that employers value has. 

Mr. Doggett.  And, Ms. Cox, isn’t that also the finding of an analysis in Utah that was made last year, that work first is not necessarily as important as the training activities? 

Ms. Cox.  Actually, I don’t think it is either/or.  I think it is short‑term occupational training that is connected to an employer, so I think it is bridging both of those worlds.  We know that connection to the labor force over time is an important indicator based on what we saw, and I can’t speak for other States, but for long‑term — after 4 years there were retained earnings and increased earnings. 

So there is a balance.  For men, for example, their struggles seem to be a little bit different.  There is not a lot of men in TANF, but our population has gone from 6 percent to 13 percent.  Men in TANF often have criminal background issues.  Sometimes they need to get attached to the labor market quickly so they can reengage.  Most of our TANF recipients get off — 70 percent are off between 2 and 9 months — so this long‑term thing isn’t as critical as maybe for the folks in the 30 percent who don’t have a high school diploma. 

So I am always really cautious of this one‑size‑fits‑all and the need to really let States give us the policy objectives, the goal of the outcome.  Do the evidence‑based practices at the national level, but States need to customize it for their unique demographics. 

Mr. Doggett.  Thank you all. 

Chairman Reichert.  Thank you, Mr. Doggett.  Mr. Renacci, you are recognized. 

Mr. Renacci.  Thank you, Mr. Chairman.  I want to thank the witnesses for testifying today.  In my home State of Ohio, an estimated 1.8 million Ohioans are living below the poverty line.  Poverty in my home State of Ohio has increased by approximately 58 percent over the last decade despite a stagnant population and a whole host of Federal programs created to end the cycle of poverty.  So I am glad that we are here today to discuss policies that work because, frankly, the people of Ohio and the Nation cannot continue on this path.  We must find ways to address our struggling economy, improve our education system and workforce training programs, and connect individuals to temporary resources they need to succeed. 

With that in mind, and I would like to start with you first, Mrs. Cox, what should the Federal Government’s role be in social welfare?  And to add on to that, should the Federal Government incentivize States to focus on outcome‑based programs and should funding to States be tied to performance?  I know you talked a lot about your State and maybe funding towards the State. 

Ms. Cox.  Well, again, you know, our goal is, even within the State of Utah, 25 percent improvement over the next 3 1/2 years, and some State agencies are saying that is impossible, but when we get into the guts of the systems we are seeing there is a lot of capacity there.  Having said that, I am all for outcomes and results.  It is taxpayer dollars, and we need to be accountable.  My preference is that we are held accountable and then given the flexibility to design the solutions that work.  We spend a significant amount of time and energy on demonstration projects at the national level and pilot projects, and we at the State just need the flexibility.  We have people at our doorstep today.  I don’t have 5 years to do a demonstration project.  You guys can and give me what I need.  I need the outcomes and the flexibility to get the results today for the people standing on my doorstep. 

So we need that, and then we also really need to emphasize with States that they, too, need to be held accountable for results, and in many cases, they need to pay more attention to evidence‑based practices and not just do what feels good. 

Mr. Renacci.  So, in general, you believe that States should be given the dollars but there should potentially be incentives tied to those outcomes, and it should be outcome based? 

Ms. Cox.  I would be open ‑‑ the devil is in the details, or God, depending on which way you say it, so it would depend on the program and how it was specifically designed, but it is something I wouldn’t be scared of. 

Mr. Renacci.  Anyone else want to take a stab at that, what the Federal government’s role should be?  Mr. Baron? 

Mr. Baron.  Yes, I think one of the challenges here with holding States accountable and so on is that at this point we, meaning the country really, and researchers and policy officials really don’t have a whole lot of strong evidence, replicated evidence, as David referred to, about what works.  A lot of times.  So there is not ‑‑

Mr. Renacci.  But if, in Ohio, we have approximately 58 percent over the last decade has increased, you know, something is not working. 

Mr. Baron.  Yes.  The system is not working, meaning over time, and it is true nationally, even since the early 1970s, the poverty rate across the United States by various measures has not changed a whole lot, despite a whole lot of innovation, a whole lot of things going on.  What is lacking, I would suggest, are interventions like the Nurse Family Partnership we discussed before, where the strong evidence that has been replicated across different sites that they work.  So as a first step doing the kind of innovation and coupled with evaluation designed to grow the number of proven programs might be paramount.  In the few cases where there are proven approaches like the Nurse Family Partnership, just try to scale those up more widely. 

Mr. Renacci.  Let me change the pace to outcome because I know as a small business owner before I got here, I set programs up and then I looked at the outcomes and decided if they weren’t working we would change those programs.  So we have programs that aren’t working.  What are some of the consequences of leaving in place government programs that do not work?  I mean, why are we leaving these in here?  Are you saying that we don’t have the outcomes yet?  I mean, there are certain things that aren’t working.  What are some of the consequences of leaving those programs still intact?  Mr. Muhlhausen? 

Mr. Muhlhausen.  Well, one of the consequences is that we waste billions of dollars and we leave people with no hope, and so I think we need to ‑‑ when we look at programs, especially programs that are trying to lift people out of poverty, are the comprehensive effect of various programs, are they trapping people in poverty?  So if someone who is receiving TANF benefits, food stamps, and also a housing subsidy, if they get a job and increase their earnings so they get a chance to work more hours, will their income increase cause them to lose their housing subsidy?  In that case, that is an incentive not to gain the additional experience, not to gain the additional income through your own labors.  So in some sense, the combined effect of our entire welfare system can create a trap for individuals. 

Mr. Renacci.  Thank you, Mr. Chairman, I yield back.

Chairman Reichert.  Mr. Davis. 

Mr. Davis.  Thank you very much, Mr. Chairman.  I want to thank all of the witnesses for coming. 

I know that the ranking member has raised a question relative to the home visiting program, the nurse visiting in program.  And I couldn’t help but smile because I recall that when we were working on the Affordable Care Act that one of the provisions that I supported very strongly, and actually secured a woman to come and testify from the Near North Health Corporation which is a community health center in Chicago.  And I run into her quite frequently.  And I always tell her whenever I do that she was very instrumental in helping us to include that program in the Affordable Care Act.  And so I was very pleased to hear your analysis and the impact of it. 

Let me ask, when programs like that, for whatever the reason, are not reauthorized, are not refunded, does that take away or detract from progress that is being made relative to not only moving people out of poverty, but also in helping them improve the health status of people and communities that they benefit? 

Mr. Baron.  Congressman, funding of the Nurse Family Partnership, which incidentally was launched as a pilot program under the Bush administration ‑‑ proposed by the Bush administration, and then scaled up by the Obama administration ‑‑ that funding, there is strong replicated evidence that it improves people’s lives.  So defunding it presumably would do the opposite.  But I would also note that the detail is in the specific type of home visiting program. 

The Nurse Family Partnership has been shown effective, but there are many other types of home visiting programs that have been shown not to work.  There was a federally sponsored evaluation, the Comprehensive Child Development Program at HHS in the 1990s, which was a paraprofessional home visiting program.  There was a large randomized demonstration that found no impacts. 

So one of the unique things about the program that was enacted is that it had a high evidence standard so that specific home visiting models, like the Nurse Family Partnership, received priority for funding.  That is unique in Federal social spending. 

Mr. Davis.  Thank you very much.  I really appreciate that kind of work, having been engaged not only in health care but also in aspects of dealing with poverty and poverty‑stricken people and communities for a long period of time.  I just find that to be incredible work. 

Dr. Muhlhausen, let me ask you, if I could ‑‑ I go through the list of different kinds of programs.  And I just looked at Supporting Healthy Marriages.  Do you have any revelations on the impact of a program like that, or that specific program? 

Mr. Muhlhausen.  Well, there are two programs, Building Strong Families and Supporting Healthy Marriages.  And I think the goal is noble.  But both of these evaluations show that both programs failed to affect the rate of marriage.  So that, considered that in that sense, the program is a failure.  In the case of the Building Strong Families case, the program actually had some negative impacts.  But you have to balance that between, if you look at the site‑by‑site analysis, on the local level in some cities, the program had consistent negative effects; but in Oklahoma, while it didn’t boost marriage, it actually found some positive impacts on the marital relationships of individuals participating. 

So you have to learn what happened in Oklahoma but systematically, when you look at the entire program, there is not much success. 

Mr. Davis.  And I think that is so unfortunate because I think that marriage does play a significant role in the ultimate organization and development of our society.  And many of my friends and many people that I interact with don’t have much faith in it, I believe.  Thank you very much. 

Chairman Reichert.  Thank you, Mr. Davis.  Mr. Reed. 

Mr. Reed.  Thank you, Mr. Chairman.  Thank you to our witnesses. 

Ms. Cox, I wanted to start with you and move to a couple of other people on the panel.  I am very interested in us starting our conversation as we go down this path of TANF reform and other reforms as to coming to a common understanding as to what the definition of “it works” is.  And I want to have a clear understanding from the panel as to ‑‑ especially you, Ms. Cox ‑‑ out in the field, on the front line, in the States.  How is it presently defined to be “a success” under these programs?  I have heard things such as getting people out of poverty.  It is as easy as defining people who get the benefits, who actually receive a check.  So I just want to get clarification from you on that.  And then if you have any guidance or recommendations as to what is a good working definition of defining success. 

Ms. Cox.  It is a really great question, because we talk about it in such broad terms, self‑sufficiency and moving people out of poverty.  But what do we mean by that? TANF ‑‑ in Utah, is 50 percent of poverty level.  Food stamps has a different eligibility criteria.  Medicaid now we know with the ACA reform is going to take you to up 130 percent plus.  So at what point does the Federal Government mean “out of poverty” because there are so many different contradictory definitions of what that means.  In Utah, again, with over 90‑plus programs for TANF, our definition for quality is throughput divided by operating expense so we can get a quality cost per case.  It is the number of positive closures we have, the percentage of those that are placement and employment, right?  Because sometimes it is so easy for a State to say, oh, they got married, they got Social Security.  We want to focus on employment because we know in the long term, that 4 or 5, 6 years down the road, that gives them a better chance at self‑sufficiency divided by those costs.  We can hit that.  That is step one.  But if our entire caseload of people who are on low‑income services ‑‑ our TANF caseload is a drop in the bucket.  We spent a lot of time on TANF.  But in our State, it is 6,000 to 10,000 of cases ‑‑ that is not individuals ‑‑ and our entire caseload is more like 200,000. 

So the broader question is what is the policy objective for Medicaid, for food stamps, for child care?  Because if we don’t drive those policy objectives, TANF isn’t going to move the whole system.  TANF has been a success.  We know in the last 16 years, caseloads have declined almost by 50 percent.  We can continue to improve it.  We know that.  I can get people off of TANF.  But moving them off food stamps and Medicaid into true self‑sufficiency, that is a much broader public policy agenda that has yet to be defined.  So for us, it is benefit.  Their case is closed.  They are off benefits.  And they have a job.  That is the ideal scenario for us. 

Mr. Reed.  That is the ideal.  Okay.  Mr. Muhlhausen. 

Mr. Muhlhausen.  I think one thing we need to think about is policy significance versus statistical significance.  We like to come here, and especially me, we like to talk about statistical significance, a particular program, boost the wages up let’s say a head of the household by $1,000 every year.  This finding was statistically significant.  Well, what is the policy significance of that?  That additional $1,000, does it necessarily raise that family above, say, the poverty level? 

So sometimes these programs we are talking about are actually ‑‑ while they do have a positive impact, and I can say statistically significant, meaning we believe the results actually occurred and were reliable.  But sometimes the size of the effect is actually not that meaningful as far as changing the individual’s life.  So I think we need to think about not only statistical significance, but also policy significance.  Is a program, let’s say, moving somebody above the poverty threshold?  And are they on a trajectory where they are not going to be dependent on receiving future government services? 

Mr. Reed.  Go ahead, Ms. Smith.  Please. 

Ms. Smith.  I think a standard way of looking at whether you are making a difference for families is, are they in stable employment with rising earnings?  That is what lifts families out of poverty.  And it has been shown to have a really positive impact on the children of those families as well.  They do better when they have that sort of financial security. 

Mr. Reed.  I am getting ‑‑ because I am running out of time and I don’t mean to cut you off. 

I am getting consensus from the panel that having a simple definition of, we are going to have x number of dollars or x number of benefits in the hands of a recipient, is probably not the best definitional program or definition of success.  Am I misinterpreting anything anyone is saying there?  With all the nods of the head, it sounds like there is agreement there.  So I appreciate that because a lot of times, I have conversations with members up here and they are just as simple as, Well, if we get x number of dollars in the hands of a recipient, that is a win.  That is a success.  And clearly it is much broader than that.  My intention in doing the work here is to improve lives, not just give benefits to people.  So I appreciate that.  With that, I yield back. 

Chairman Reichert.  Thank you.  Mr. Kelly. 

Mr. Kelly.  Thank you, Mr. Chairman.  Again, thank you all for being here. 

One of the things I am trying to understand ‑‑ and it is not for a lack of investment, is it, on the part of the government?  I am looking at the investment that we make each year.  And if the numbers I am hearing are right, it is about $600 billion a year that goes into trying to lift people out of poverty or support the most vulnerable in our society.  That is a lot of money.  But I think our concern is, what is the return on that investment?  What are investing these dollars in?  And the idea was to lift people out of whatever conditions they were in. 

So, Ms. Cox, I heard you say that part of the problem is, sometimes there is a negative incentive because once you get out of one level, then you go into another.  And all of a sudden, it is like, well, this doesn’t work in my best interest. 

So best practices, are you able to share those with each other?  I know there are a lot of programs.  Your State is doing some things that maybe other States should do. 

Do you have the ability to communicate back and forth? 

Ms. Cox.  Yes.  There are associations and forums.  But you know, when you are in the trenches ‑‑ and especially during the recession, our caseloads increased by 63 percent.  So we were just treading water to get through that and get people back to work and contain costs.  We were able actually to reduce our costs by 33 percent while our caseloads increased by 63 percent and improve our timeliness.  But it is really difficult.  It is nice to have the luxury, with all due respect, to analyze this stuff for 4 or 5 years.  But that isn’t a reality when you are on the ground. 

So we need people ‑‑ these brilliant people here to inform us, to educate us.  We are committed to evidence‑based practices in Utah.  It is the only way we will hit our 25 percent improvement in some of our agencies.  But sometimes we need States just to be able to innovate to get to the clear results.  Because if we don’t have time to do the evidence‑based practice, and we are not allowed to innovate, we are stagnant, and we can never make progress.  So that is the bind States are in.  We need evidence‑based practices.  But when it is not there, we need the flexibility to innovate to get the results for you guys and for the taxpayers.

Mr. Kelly.  And you mentioned innovation.  My friend David Bradley has talked to me many times.  He is with the National Community Action Foundation.  And he talks about the ability to look at innovation, to look at the performance data, and then also local control of these dollars.  And again, I have a friend in Sharon, Pennsylvania, by the name of Ron Errett who runs a program up there.  It is the Community Action Partnership of Mercer County.  So I have seen locally in the district that I represent a lot of programs that work really well.  And I think we have got to be careful that we don’t paint everybody with the same brush and say, we are wasting this money.  Nothing is going the right way.  I don’t think that is true.  And we referenced the Nurse Family Partnership, how much that has worked. 

But it does come down to, how do we get that?  How do we get the innovation message out?  How do we share those practices?  Mr. Baron, you made some comments about that when you talked about the ability.  And the results speak for themselves.  When you see something good, how do you get that out?  Because I have got to tell you, in my district, I was able to look up and down northwestern Pennsylvania.  The poverty level is probably somewhere between 25 and 30 percent.  Not really mattering what town you go into, big cities, little towns, it is about the same.  This poverty thing is something that is really troubling that we spend all this money, but we haven’t gotten any results and we don’t see that happen.  I know part of it is the economy not bouncing back.  And maybe we are spending a lot of time criticizing programs and not coming up with leadership programs or strategies that lifts everybody. 

Ms. Cox.  Can I make one more point on that just to be kind of bossy? 

Mr. Kelly.  Sure. 

Ms. Cox.  There are associations, administrators, that are always connecting and going to conferences and talking about best practices.  But part of the challenge is, we have the evidence‑based practice.  But if you were to look and do our mapping and look at an employee’s time.  They have 8 hours a day.  And let’s say we even know what the evidence‑based practice is.  We know that they should be doing X, Y, Z every day with their customers to get the impact.  If you were to map out how much time they actually spend doing that, in some cases, you will find 10, 20, 30, 40 percent of their time is actually spent on the evidence‑based practice.  The rest of the time it is spent on compliance, recording, paperwork, a lot of other stuff.  So can you imagine the capacity to impact low‑income individuals?  If we could just double the time on the ground in your operations and in your systems design of what people do ‑‑

Mr. Kelly.  I am going to agree with you because I have got to tell you.  I run a private business.  We do the same thing with our business trying to be in compliance.  If I could just do what we are designed to do every day and not worry about being in compliance with the Feds, the local, and the State, we would probably get a lot more done.  Mr. Baron you were going to say something? 

Mr. Baron.  Yes.  One of the challenges in sort of identifying and sharing best practices, things that really work is that a lot of programs, almost every program claims to be evidence‑based and backed by strong evidence and effective.  And the truth is, while some of them are, when most programs ‑‑ even those that are backed by pretty good evidence — are subjected to a definitive evaluation, many of the promising findings are not reproduced.  Sometimes they are.  So you do have some examples of effectiveness, but many times they are not.

Steve Aos’ organization, the Washington State Institute, does a valuable service by trying to distill what is really backed by strong evidence from others that are not.  But there are some instances.  There was a program that the reemployment and eligibility assessment program at the Department of Labor which has been shown very effective in a four‑State randomized control trial with large effects on employment outcomes and reductions in spending. 

Mr. Kelly.  Thank you. 

Chairman Reichert.  Thank you.  Mr. Crowley. 

Mr. Crowley.  Thank you, Mr. Chairman.  Thank you for yielding me the time. 

I think we are all good stewards.  I think we all want to be good stewards of our taxpayer dollars and certainly we want to put money where we think it best works for the American people. 

Mr. Baron, I just want to follow up my comments of Mr. Doggett as well as Mr. Davis as it pertains to the Nurse Family Partnership is the kind of program that has been proven to get results.  As you know, I know ‑‑ I am not so sure my colleagues know ‑‑ that the program started in my home State of New York ‑‑ in fact, in Elmira, in my good friend, Mr. Reed’s district.  And I have long followed the very impressive work that they have been involved in in replicating this model and achieving very significant positive outcomes.  Reductions in child abuse and neglect, better educational outcomes for children, and greater likelihood of economic stability for the mother, these are just some of the results that actually save the government money in the long term. 

Mr. Baron, do you think that is correct, that it has an effect in terms of saving taxpayer dollars in the long run? 

Mr. Baron.  In this case, I think the answer is yes, especially with the Nurse Family Partnership.  One of the trials that was done in Memphis, Tennessee, measured not only the impacts you are describing, like child hospitalizations and educational outcomes for the children, it also measured participants’ use of government assistance ‑‑ Medicaid, food stamps, welfare over a 12‑year period.  And this was not a projection.  They measured use of government assistance in the treatment versus control group.  So this was a credible finding.  And it found that the program produced savings that more than offset the cost of the program. 

Mr. Crowley.  Mr. Baron, would you also agree that programs like this particular one are an example of the importance of looking at successes not just in the short term, but in the greater awards to our society down the road?  And not just immediate. 

Mr. Baron.  Yes.  Because some of the effects are longer term.  Some of the effects were short term.  There were immediate effects on reductions in child maltreatment and hospitalizations.  But there were also longer term.

Mr. Crowley.  So leading to a much longer and productive life for the child in the long term.

Mr. Baron.  That is correct.  But a slight nuance on that is that some programs ‑‑ especially in workforce development ‑‑ they produce short‑term effects, which dissipate over time.  And so for those kinds of programs, it is important not only to measure the short‑term effects which are sometimes large, but whether it produces sustained effects on the amount of ‑‑

Mr. Crowley.  I appreciate your comments. 

Better access to health care coverage, nutrition programs in early childhood, we believe leads to greater health and reduced medical costs later in life.  And that is what we are talking about at this point in time as it pertains to the Nurse Family Partnership. 

Strong education, mentoring, and family supportive programs reduce incidents of criminal activity and school dropout rates as well.  I think too often, we have a tendency and I think is a willingness to cut a program for ideological purposes not because it is what is in the best interest of our country.  We see it in the Affordable Care Act.  We have seen it in the farm bill nutrition programs.  We have seen it as well in the social services block grants and others.  It seems like the lesson here today is that we need to carefully invest in the programs that work and really put a lot of thought and study into our budgeting process. 

So I would think the budget approach we have recently seen with policies, like sequestration, is the exact opposite of what we ought to be doing.  Blunt across‑the‑board cuts and not replacing them with a thoughtful plan that grows the economy certainly doesn’t fit with the evidence‑based approach that seems to be the recommendation of the witnesses here today.  I would hope that my colleagues not just on this committee but on the Budget and Appropriations Committees as well and every other committee draw lessons from this hearing.  And I look forward to more a constructive conversation and hearings like this in the future. 

With that, I yield back the balance of my time. 

Chairman Reichert.  I can assure my good friend that I am very interested in evidence‑based results as an old time cop.  So with 33 years in that field, I am looking for evidence.  Mr. Young. 

Mr. Young.  I want to thank the chairman for holding this hearing on what really works.  I think it is incredibly important.  It may be mundane to some people.  It may be boring to others.  Metrics and data and all these other things.  But let me begin by defining the challenge and perhaps identifying the opportunity or opportunities as I see them here.  My interest in this topic actually emerged the second I found out I was going to get on this committee.  I suspected it wasn’t unlikely I was going to be on the Human Resources Subcommittee, and that is a good thing.  We deal on this committee with what some regard as unsolvable social pathologies.  And I sort of refused to believe that. 

So I know a number of pilot programs over the years have taken place across our 50 States.  So I directed members, associates on our team to try and identify the results of those different pilot programs and find some central repository where we could find out what really works and under what circumstances.  It was incredibly tedious work; and frankly, there was no such repository.  There was not a navigable Web site.  There wasn’t a particular organization that seemed to have answers about what really works and what data you could look at and what works under different circumstances. 

I found this frustrating as a policymaker.  And I think at the State level, local level, not for profits, academics, think tankers, and so on, could all benefit from more clarity here and a more robust collection of data. 

What are we left with without this sort of repository of accessible data?  Well, we make policy based on ideology, on politics, on analogy, sometimes anecdotes, the news of the day; but we don’t really make decisions based on the hard data.  So I started doing a bit of reading.  I discovered that in 1988, under the AFDC Reform Act of that year ‑‑ not often talked about ‑‑ there was a requirement that data be collected on the recipient population of AFDC.  That data years later ‑‑ oftentimes it does take years for this information to be teased out ‑‑ that data established the intellectual groundwork for a bipartisan reform of the AFDC program, now the TANF program. 

We need to make similar efforts in other areas.  We need to do more evidence‑based policymaking. 

Mr. Baron, thank you.  It is so great to see you here today.  You are really a gift to this conversation.  I thank everyone else as well.  And we have convened a group of people to discuss this topic.  It is my hope that we can come up with a more systematic way of collecting data in a number of different areas and promulgate and disseminate that data to others for the purpose of research and also for the purpose of evaluation so that when innovation occurs at the State level, we will know if, in fact, it is working and then share.  It is an iterative process.  Share what is learned with others.  And that will enable us to do very creative things, like social impact funds, pay for performance, pay for success in some of these social areas, the same sort of thing we do in, say, the transportation sector, performance‑based contracting. 

Mr. Baron, you said Congress could take steps in your statement towards what I have envisioned, I believe, by authorizing and encouraging agencies to allow greater research access to administrative data with appropriate privacy protections so as to facilitate low‑cost rigorous evaluation. 

I have two questions for you.  First, what data is the Federal Government failing to collect that we ought to be collecting about beneficiaries of government programs?  And second ‑‑ and I see my time is running down, so you can submit this in writing.  But second, I did want to get it on the record, what data is already being collected by the Federal Government, such as receipt of government assistance, employment status, earning status that we should release to the public for research purposes so it is not just our bureaucrats who are armed with all the information. 

Mr. Baron. I would like to submit a response in writing.  But also a very quick answer to the second part of your question.  This subcommittee took a major step forward, we believe.  We were very supportive of the subcommittee’s action to increase researcher access with appropriate privacy protections to the National Directory of New Hires, the NDNH data, which has, at the Federal level, the employment and earnings records.  You can use it to measure employment and earning records ‑‑ earning outcomes for participants in any study.  It should be more widely available.  It would reduce the cost of some of these rigorous studies by a factor of 10 or more. 

Mr. Young.  Thank you. 

Mr. Reichert.  Thank you, Mr. Young.  Thank you to all of the witnesses for being here today and taking the time to be with us.  And we have finished this almost in record time, just in time for votes.  I thank the members for being here too.  We look forward to working with you and reaching out to you and asking you questions, more questions that will come, I am sure, as we struggle with trying to find solutions here that are evidence‑based, where we hold programs accountable, make sure that we are really helping those people who need the help and ensure that they are moving up that economic ladder as we all hope that they do. 

So if members have additional questions for the witnesses, they will submit them to you in writing.  And we would appreciate receiving your responses for the record within 2 weeks.  The committee stands adjourned.  Thank you.

[Whereupon, at 5:30 p.m., the subcommittee was adjourned.]

Public Submissions For The Record

Family Equality Council

American Evaluation Association
Urban Institute
Zero to Three
Capital IDEA Board of Directors
PEW Charitable Trusts
Chautauqua Opportunities