40	ANNALS, AAPSS, 678, July 2018
DOI: 10.1177/0002716218763128
A Brief History
of EvidenceBased
Policy
By
Jon Baron
763128ANN THE ANNALS OF THE AMERICAN ACADEMYA BRIEF HISTORY OF EVIDENCE-BASED POLICY
research-article2018
This article provides a brief history of evidence-based
policy, which it defines as encompassing (1) the application
of rigorous research methods, particularly randomized
controlled trials (RCTs), to build credible
evidence about “what works” to improve the human
condition; and (2) the use of such evidence to focus
public and private resources on effective interventions.
Evidence-based policy emerged first in medicine after
World War II, and has made tremendous contributions
to human health. In social policy, a few RCTs were
conducted before 1980, but the number grew rapidly in
U.S. welfare and employment programs during the
1980s and 1990s and had an important impact on government
policy. Since 2000, evidence-based policy has
seen a major expansion in other social policy areas,
including education and international development
assistance. A recent milestone is the U.S. enactment of
“tiered evidence” social programs in which rigorous
evidence is the defining principle in awarding government
funding for interventions.
Keywords:	 evidence-based policy; evidence-based
medicine; social policy; randomized controlled
trial; welfare reform; tiered evi-
dence
Evidence-based policy, as I define it in this
article, encompasses two core elements: the
application of rigorous research methods, particularly
randomized controlled trials (RCTs), to
build credible evidence about “what works” to
improve the human condition; and the use of
such evidence to focus public and private
resources on programs, practices, and treatments
(“interventions”) shown to be effective.
Jon Baron is vice president of evidence-based policy at
the Laura and John Arnold Foundation, a nonprofit
philanthropic foundation. He is also the founder and
former president of the nonprofit Coalition for
Evidence-Based Policy, which worked with federal
officials from 2001 to 2015 to advance evidence-based
reforms in government programs.
Correspondence: jbaron@arnoldfoundation.org
A BRIEF HISTORY OF EVIDENCE-BASED POLICY	 41
Over most of modern world history, humanity did not need evidence-based
policy to make important progress in improving health and life around the globe.
For example, in the case of medical or public health interventions, such as penicillin
for the treatment of bacterial infections, insulin for the treatment of diabetes,
urban sanitation, and refrigeration of perishable foods, the treatment effects
were so large that they could be detected using methods that were not rigorous,
such as simply observing people’s health before and after the intervention. The
same is true of certain social and political interventions, such as a money economy
as opposed to barter, rule of law as opposed to autocracy or anarchy, and
schooling of children. Here, too, the intervention effects were so large that they
could be directly observed without a rigorous evaluation.
But in the twenty-first century, many countries have already exploited the
kinds of interventions that have blockbuster effects. To continue to make progress,
evaluation methods need to detect effects that are more modest but still
quite important—for example, a 10 percent increase in the survival rate (for a
medical treatment), or a 20 percent reduction in the school dropout rate (for an
education intervention). In these cases, simply observing people before and after
cannot establish whether any change in their condition is due to the intervention
or to other, confounding factors, such as the body’s own immune defenses (in the
case of a medical treatment) or demographic changes in the student population
(in the case of an education intervention). To determine whether the intervention
caused the outcomes we observe, we need an evaluation method that controls for
confounding factors—such as an RCT that randomizes a sizable number of
patients or students.
This article traces the history of evidence-based policy from its inception, after
the end of World War II, through the present. My goal is to provide an overview,
leaving it to other authors in this volume to fill in the specifics in particular time
periods or fields of policy.
Evidence-Based Medicine in the Postwar Era, and Its
Amazing Contribution to Human Health
Before World War II, medicine in the United States and other advanced countries
was based mostly on anecdote and unscientific evidence. Sir Richard Doll,
the British epidemiologist who became a leading figure in evidence-based medicine,
recounts the state of medicine when he completed his training in the 1930s:
New treatments were almost always introduced on the grounds that in the hands of
professor A or in the hands of a consultant at one of the leading teaching hospitals, the
results in a small series of patients . . . had been superior to those recorded by professor
B (or some other consultant) or by the same investigator previously. Under these conditions
variability of outcome, chance, and the unconscious (leave alone the conscious) in
the selection of patients brought about apparently important differences in the results
obtained; consequently, there were many competing new treatments. … [W]hen
I began to investigate peptic ulcers, I was soon able to prepare a list of treatments
42	 THE ANNALS OF THE AMERICAN ACADEMY
beginning with each letter of the alphabet. Standard treatments, for their part, tended
to be passed from one textbook to another without ever being adequately evaluated.
(Doll 1998, 1217)
The 1940s and 1950s saw the first RCTs published in medicine, including the
1946 UK trial of streptomycin for treating pulmonary tuberculosis and the U.S. Salk
polio vaccine field trials of 1954. These studies made an enormous contribution to
public health. The Salk vaccine trials, for instance, demonstrated the effectiveness of
a medical intervention that would, in the years that followed, eradicate paralytic
polio—a devastating disease that, in the early 1950s, afflicted an average of sixteen
thousand people in the United States each year, causing an average of nineteen hundred
deaths (Centers for Disease Control and Prevention 1999).
Despite the success of the Salk vaccine and streptomycin studies, RCTs initially
spread slowly and not without opposition until policy-makers embraced
them as a requirement for licensing new drugs (Doll 1998, 1219). In 1962,
Congress enacted legislation that, as implemented by the Food and Drug
Administration (FDA), required that well-conducted RCTs demonstrate the
effectiveness of any new pharmaceutical drug (and, as later amended, medical
devices) before the FDA would approve it for marketing (FDA 2017). The FDA
requirement directly embodied the second component of evidence-based policy
described above—using rigorous evidence to focus public and private resources
(in this case, an FDA license, and a potentially large and lucrative market) on
effective interventions. The FDA policy also created a powerful new incentive for
pharmaceutical companies to conduct RCTs aimed at building the number of
proven-effective interventions—the first component of evidence-based policy.
The FDA policy, along with parallel support for clinical trials by the National
Institutes of Health (NIH), transformed the RCT in medicine from a rare and
controversial method that had first appeared in the medical literature only 15
years earlier (Medical Research Council 1948) into the widely used gold standard
for assessing the effectiveness of all new drugs and medical devices. Between
1966 and 1995, the number of clinical research articles based on RCTs surged
from about one hundred to ten thousand annually (Chassin 1998, 574).
Since the early 1960s, RCTs required by the FDA or funded by the NIH and
other agencies have produced the conclusive evidence of effectiveness behind most
major medical advances, including vaccines for measles, hepatitis B, and rubella;
interventions for hypertension and high cholesterol, which helped to cut the incidence
of coronary heart disease and stroke by more than 50 percent over the past
half-century; and cancer treatments that have dramatically improved survival rates
from leukemia, Hodgkin’s disease, breast cancer, and many other cancers. Such
advances have profoundly improved life and health in America (Gifford 1996).
Early RCTs in Social Policy, 1930s–1970s
The emergence of RCTs in social policy was far more gradual than it was in medicine.
Several of the early social policy RCTs are discussed in more detail in other
A BRIEF HISTORY OF EVIDENCE-BASED POLICY	 43
articles; I summarize them briefly here to place them in historical context. Early
RCTs included:
•• the Cambridge-Somerville Youth Study, a relatively small RCT initiated in
the 1930s to evaluate a program that provided counseling and group recreational
activities, including summer camp, to low-income adolescent boys at
risk for delinquency (Dishion, McCord, and Poulin 1999);
•• the Manhattan Bail Bond project, an RCT initiated in 1961 to test the
effects of releasing certain defendants without bail before trial (Botein
1965);
•• the Perry Preschool study, a small RCT initiated in 1962 to evaluate a program
that provided high-quality preschool education to three- and fouryear-old
children from low socioeconomic backgrounds (Schweinhart 2004);
•• the Abecedarian study, a small RCT initiated in 1972 to evaluate a program
that provided educational child care and high-quality preschool from birth
to age 5 for children from disadvantaged backgrounds (Ramey, Sparling,
and Ramey 2012);
•• the income maintenance experiments initiated in the late 1960s and 1970s,
which comprised four large-scale RCTs of “negative income tax” systems
that provided a guaranteed payment to people with no earned income,
gradually eliminated the payment as income rose, and converted into a
graduated positive tax system as income rose further (Munnell 1986);
•• the National Supported Work Demonstration, a large-scale RCT launched
in 1974 to evaluate a program that offered subsidized jobs to hard-toemploy
people, followed by assistance in finding an unsubsidized job
(Manpower Demonstration Research Corporation 1980);
•• three large-scale RCTs conducted in the 1970s to test the effects of providing
housing vouchers to low-income households (Merrill and Joseph 1980);
and
•• the RAND Health Insurance Experiment, a large-scale RCT initiated in
1971 to test how different levels of cost-sharing (i.e., the portion of the bill
the patient pays) affected health-care expenditures and patient health
(Manning et al. 1987).
The policy impact of these early RCTs varied. At one end, the CambridgeSomerville
Youth Study generated significant interest among scholars because it
found sizable adverse effects on participants’ life outcomes, but remained largely
unknown in the policy community. The findings from the income maintenance
experiments—modest adverse effects on participants’ employment and earnings—
likely dampened the initial enthusiasm in the policy community for a negative
income tax (a version of the idea had been proposed by the Nixon administration
and passed by the House of Representatives in 1969). Similarly, the disappointing
findings of the National Supported Work Demonstration on the employment and
earnings of three of the four targeted populations—ex-offenders, ex-addicts, and
youth (there were positive impacts for welfare mothers)—helped to diminish
interest in supported work as a tool to address poverty. On the other hand, the
44	 THE ANNALS OF THE AMERICAN ACADEMY
findings of the Perry Preschool and Abecedarian studies—large, long-term positive
effects on outcomes such as educational attainment, employment, and criminal
activity—have had an important influence on policy, inspiring many federal,
state, and local initiatives to expand preschool, particularly for low-income children.
That occurred despite the limitations of both studies, which were small
demonstration projects with key departures from random assignment that somewhat
reduce confidence in their findings.1
The most important contribution that these early RCTs made to the history of
evidence-based policy, however, may be unrelated to their findings. Judy Gueron
and Howard Rolston (2013) suggest that it was to demonstrate that RCTs—
including large, multisite trials—were indeed feasible in diverse areas of social
policy. These early studies also built a community of social policy researchers at
MDRC (which was created by the Ford Foundation to administer the National
Supported Work Demonstration), Mathematica Policy Research, Abt Associates,
and elsewhere with the expertise to carry out large-scale RCTs. And by 1980, a
group of funders—most notably the Ford Foundation and U.S. Department of
Labor (DOL), but also the U.S. Departments of Health and Human Services
(HHS) and Housing and Urban Development (HUD)—had developed an interest
and experience in sponsoring RCTs.
Welfare and Employment RCTs of the 1980s
and 1990s, and Their Impact on Policy
A key development in the early 1980s was MDRC’s launch of the Work/Welfare
Demonstration, with core funding from the Ford Foundation. This initiative,
comprising eight RCTs of different state-level welfare-to-work programs, showed
for the first time that large-scale RCTs could be successfully integrated into normal
state agency operations to evaluate programs developed and administered by
those agencies. It was a novel approach at the time; the earlier large-scale RCTs
described above had evaluated programs designed by researchers, foundations,
and/or federal officials for purposes of the study (Gueron and Rolston 2013).
The Work/Welfare Demonstration’s success encouraged HHS, in the mid-
1980s, to begin funding large-scale RCTs of welfare-to-work and other employment,
income supplementation, and related programs for the poor. These efforts
had strong support from Congress, which, in the Family Support Act of 1988,
required HHS to use RCTs to evaluate various welfare reforms. Support also
came from the White House. During the George H. W. Bush and Bill Clinton
administrations, HHS—with White House support—implemented a “demonstration
waiver” policy, under which HHS allowed states to waive certain provisions
of federal welfare law, using authority given to the HHS Secretary by
section 1115 of the Social Security Act, so that they could test new welfare-towork
programs and other reforms. HHS approved the waivers only if states
agreed to rigorously evaluate their reforms using RCTs wherever feasible (HHS
1994).
A BRIEF HISTORY OF EVIDENCE-BASED POLICY	 45
HHS thus directly funded or facilitated (e.g., through demonstration waivers)
more than eighty-five RCTs in the 1980s and 1990s, building a sizable body of
credible evidence with clear policy relevance. For example, these studies convincingly
demonstrated that work-focused welfare reform programs that emphasized
short-term job-search assistance and training, and encouraged participants
to find work quickly, had larger effects on employment, earnings, and welfare
dependence than reform programs emphasizing remedial education. The workfocused
models were also much less costly to operate (Bloom, Hill, and Riccio
2003). Three major work-focused reform models (two in California and one in
Oregon) were found to be especially effective. Each of the three increased participants’
employment and earnings by 20 to 50 percent over a follow-up period
of several years, and generated net government savings (e.g., from reduced welfare
and food stamps payments) of $2,500 to $7,500 per person in 2017 dollars
(Freedman et al. 1996, 2000; Hamilton et al. 2001).
The studies also found that when programs combined mandatory participation
in employment-focused services (e.g., job search assistance or job training) with
earnings supplements for participants who did find work, they could raise overall
income and move many people out of poverty. For example, the Minnesota
Family Investment Program, using such a strategy, not only produced sizable
gains of 20 to 40 percent in employment and earnings for single-parent, longterm
welfare recipients, but also reduced the proportion with overall income
below the poverty line from 85 percent to 75 percent (Gennetian, Miller, and
Smith 2005).
These findings had an important impact on policy and practice. According to
federal officials and others involved in welfare reform, they helped to build political
consensus for the strong work requirements in the Personal Responsibility
and Work Opportunity Reconciliation Act of 1996 (i.e., federal welfare reform)
and shape many of the work-first state-level reforms that followed. The findings’
scientific rigor was critical to their policy impact (Haskins 2007). The sort of
work-first and job club models that these studies found to be cost-effective are
now the mainstay of welfare systems in the United States, Canada, the United
Kingdom, Australia, and much of Europe.
Evidence-Based Policy Expands in Other Areas:
2000–Present
In early 2000s, three developments accelerated the expansion of evidence-based
policy in areas of social policy beyond welfare and employment. Two of them
primarily advanced the first component of evidence-based policy, that is, building
credible evidence about what works through the use of rigorous research methods.
The first was Congress’s enactment of the Education Sciences Reform Act
of 2002, which established the Institute of Education Sciences (IES) as a largely
independent research organization in the U.S. Department of Education and
46	 THE ANNALS OF THE AMERICAN ACADEMY
directed IES-commissioned evaluations of educational programs to use RCTs
wherever feasible to measure program impact. As discussed in Whitehurst (this
volume), over the following years and up to the present, IES has funded many
large RCTs and other rigorous impact evaluations assessing the effectiveness of a
wide range of education interventions—including classroom curricula, teacher
professional development programs, school choice programs, educational software
products, and many others.
The second development was the 2003 launch of the Jameel Poverty Action
Lab (J-PAL) at the Massachusetts Institute of Technology, discussed in Chupein
and Glennerster (this volume), which began conducting large-scale RCTs in
developing countries to determine what works to reduce poverty and improve
education, health, and other outcomes. Such studies were a major departure
from prior, less rigorous research methods in the field of international development
assistance, and they fundamentally transformed research in that field over
the following decade. Between 2003 and as of this writing, J-PAL RCTs have
evaluated a vast array of interventions in developing countries, including educational
tutoring in primary schools, child immunization campaigns, financial
incentives to reduce child marriage and teen parenting, and the promotion of
self-employment among the very poor, to name just a few.
The third development was the advent of the Coalition for Evidence-Based
Policy, a nonprofit, nonpartisan organization that I founded in 2001, which
became an effective partner with Congress and the Executive Branch in advancing
both components of evidence-based policy—the development and the use of
rigorous evidence (Wallace 2011). According to an independent assessment in
2009, the coalition “successfully influenced legislative language, increased funding
for evidence-based evaluations and programs … and raised the level of
debate in the policy process regarding standards of evidence” (Herk 2009, 1).
Perhaps its most important contribution was to enlist and support the Office of
Management and Budget (OMB) as a key proponent of evidence-based policy
across the federal government, through a coalition-OMB partnership that
spanned the administrations of George W. Bush and Barack Obama. For example,
in early 2004, the coalition worked closely with OMB to develop guidance on
evidence-based approaches for federal agencies, in which OMB for the first time
identified well-conducted RCTs as the strongest method for evaluating program
effectiveness (OMB 2004). In 2007, the coalition worked closely with OMB and
the House and Senate appropriations committees to gain congressional passage
of a $10 million pilot program to fund evidence-based home visitation interventions
that “have been shown, in well-designed randomized controlled trials, to
produce sizable, sustained effects on important child outcomes such as abuse and
neglect” (Haskins and Margolis 2015, 32–34). In January 2009, the coalition proposed
a three-tiered approach to evidence-based social investment that was a
basis for the Obama administration’s tiered-evidence grant programs, discussed
below, and the coalition worked closely with OMB to design some of the programs
(Coalition for Evidence-Based Policy 2009).
Between 2009 and 2011, Congress enacted six tiered-evidence social programs
in areas such as K–12 education, early childhood home visiting, and teen
A BRIEF HISTORY OF EVIDENCE-BASED POLICY	 47
pregnancy prevention. This marked a major milestone in the history of evidencebased
policy, in that rigorous evidence was the defining principle for awarding
funding for interventions. The programs, which are discussed in detail in Haskins
and Margolis (2015), award their largest grants—those in the top tier—to interventions
that have strong or highly promising evidence of effectiveness. These
grants pay for expanding the intervention and testing (through a required replication
RCT) whether the effects found in prior studies can be reproduced on a
larger scale. The tiered-evidence programs also award smaller grants (in a second
and, in some programs, a third tier) to implement interventions with moderate or
preliminary evidence of effectiveness, and to rigorously test, usually through an
RCT, whether they produce the hoped-for effects. If they do, they can move into
the top tier and qualify for more funding; if not, the funds are directed to other,
more promising efforts.
Some of the tiered-evidence programs fell victim to congressional budget cuts
starting in 2013, but three of them attracted bipartisan support and continue to
this day: the Department of Education’s Education Innovation and Research
(EIR) program (funded at $100 million in Fiscal Year [FY] 2017); HHS’s
Maternal, Infant, and Early Childhood Home Visiting Program (funded at $400
million in FY2017); and HHS’s Teen Pregnancy Prevention (TPP) Program
(funded at $101 million in FY2017).
Findings from the initial set of tiered-evidence grants (awarded in 2010–2013)
are now being reported. As expected, a number of the funded interventions were
found to produce small or no positive effects, as is true whenever rigorous evaluations
are carried out. But some were major successes; that is, they were found
in the required RCT evaluations to produce sizable, sustained effects on important
life outcomes. These include the following:
•• Per Scholas Job Training, a program that provides training and employment
services in the information technology sector to low-income workers (in the
third year after random assignment, workers in the Per Scholas group
earned an average of $4,800, or 27 percent, more than workers in the control
group) (Schaberg 2017; Hendra et al. 2016);
•• Knowledge Is Power Program (KIPP) elementary and middle schools, a
national network of public charter schools whose mission is to help underserved
students enroll in and graduate from college (two to three years after
random assignment, students in KIPP schools scored 5 to 10 percentile
points higher in reading and math than students in the control group)
(Tuttle et al. 2015); and
•• Teen Options to Prevent Pregnancy (TOPP), an intervention for lowincome
adolescent mothers that aims to reduce rapid repeat pregnancy and
promote healthy birth spacing (20 months after random assignment, 21
percent of the TOPP group had experienced a repeat pregnancy, compared
to 38 percent of the control group) (Stevens et al. 2017).
48	 THE ANNALS OF THE AMERICAN ACADEMY
Concluding Observations
The first and most clearly successful example of evidence-based policy in action
occurred in the field of medicine—particularly pharmaceutical medicine, which
uses rigorous evidence of effectiveness to determine which drugs can be marketed,
and which has thereby generated tremendous advances in human health
over the past half century. Evidence-based policy has also scored some notable
successes in social policy. Space constraints prevent me from discussing the challenges
along the way in both evidence-based medicine and social policy, but as
illustrative examples these have included opposition from the American Medical
Association and pharmaceutical industry to the 1962 legislation requiring rigorous
evidence of drug effectiveness prior to marketing (the legislation was nevertheless
enacted because Congress felt urgency to respond to the thalidomide
tragedy);2 opposition from certain policy officials and academics to the primacy
of RCTs for establishing effectiveness in medicine (e.g., Frieden 2017) and in
social programs (e.g., Schorr 2012; Deaton 2010); and the dilution of the standards
for rigorous evidence of effectiveness in some of the tiered-evidence social
programs (e.g., Coalition for Evidence-Based Policy 2010; Baron 2014).
Despite important recent advances, tiered-evidence programs, rigorous evaluation
requirements, and other evidence-based approaches have so far gained only
a foothold in social policy; the majority of social spending is still allocated with
little regard to rigorous evidence about what works. But the narrative of this
article suggests that history is on the side of evidence. As Robert Slavin (2017)
has observed, evidence-based social policy may now be in a position similar to the
automobile industry in 1909: there were few cars and lots of engineering problems,
but anyone with eyes to see knew that the automobile was the future. So
may it be with evidence.
Notes
1. See http://evidencebasedprograms.org/1366-2/65-2.
2. Thalidomide is a pharmaceutical drug that, in the late 1950s and early 1960s, was marketed worldwide
to alleviate sleeplessness and, for pregnant women, morning sickness. Described by its manufacturer
as completely safe, it caused severe birth defects that affected thousands of children.
References
Baron, Jon. 2 April 2014. In Support of the Reauthorization of HHS’s Maternal, Infant, and Early
Childhood Home Visiting Program, with Recommendations for Improvement. Testimony before the
House Ways and Means Subcommittee on Human Resources. Available from http://coalition4evi
dence.org/wp-content/uploads/2014/08/Coalition- testimony-4.2.2014.pdf.
Bloom, Howard S., Carolyn J. Hill, and James A. Riccio. 2003. Linking program implementation and
effectiveness: Lessons from a pooled sample of welfare-to-work experiments. Journal of Policy Analysis
and Management 22 (4): 551–75.
A BRIEF HISTORY OF EVIDENCE-BASED POLICY	 49
Botein, B. 1965. The Manhattan Bail Project: Its impact in criminology and the criminal law process. Texas
Law Review 43:319–31.
Centers for Disease Control and Prevention. 2 April 1999. Achievements in public health, 1900–1999:
Impact of vaccines universally recommended for children–United States, 1990–1998. Morbidity and
Mortality Weekly Report 48 (12): 243–48.
Chassin, Mark R. 1998. Is health care ready for six sigma quality? Milbank Quarterly 76 (4): 565–91.
Chupein, Thomas, and Rachel Glennerster. 2018. Evidence-informed policy from an international perspective.
The ANNALS of the American Academy of Political and Social Science (this volume).
Coalition for Evidence-Based Policy. 2009. Suggestions for the administration’s new social entrepreneurship
initiative: Focus on building a body of research-proven programs, shown to produce major gains
in education, poverty reduction, crime prevention, and other areas. Washington, DC: Coalition for
Evidence-Based Policy. Available from http://coalition4evidence.org/wp-content/uploads/2009/06/
ideas-for-social- entrepreneurship-initiative-12309.pdf.
Coalition for Evidence-Based Policy. 2010. HHS’s Evidence-Based Teen Pregnancy Prevention Program:
Excellent first step, but only 2 of 28 approved models have strong evidence of effectiveness. Washington,
DC: Coalition for Evidence-Based Policy. Available from http://coalition4evidence.org/wp-content/
uploads/2010/05/Coalition-comments-HHS- Teen-Pregnancy-Prevention-May-2010.pdf.
Deaton, Angus. 2010. Instruments, randomization, and learning about development. Journal of Economic
Literature 48:424–55.
Dishion, Thomas J., Joan McCord, and Francois Poulin. 1999. When interventions harm: Peer groups and
problem behavior. American Psychologist 54 (9): 755–64.
Doll, Richard. 1998. Controlled trials: The 1948 watershed. British Medical Journal 317:1217–20.
Freedman, Stephen, Daniel Friedlander, Winston Lin, and Amanda Schwede. 1996. The GAIN evaluation:
Five-year impacts on employment, earnings, and AFDC receipt. Working Paper 96.1, MDRC,
Washington, DC.
Freedman, Stephen, Jean Tansey Knab, Lisa A. Gennetian, and David Navarro. 2000. The Los Angeles
jobs-first GAIN evaluation: Final report on a work first program in a major urban center. Washington,
DC: MDRC.
Frieden, Thomas R. 2017. Evidence for health decision making—Beyond randomized, controlled trials.
New England Journal of Medicine 377:465–75.
Gennetian, Lisa A., Cynthia Miller, and Jared Smith. 2005. Turning welfare into a work support: Six-year
impacts on parents and children from the Minnesota Family Investment Program. Washington, DC:
MDRC.
Gifford, Ray W. 1996. FDR and hypertension: If we’d only known then what we know now. Geriatrics 51
(1): 29–32.
Gueron, Judith M., and Howard Rolston. 2013. Fighting for reliable evidence. New York, NY: Russell Sage
Foundation.
Hamilton, Gayle, Stephen Freedman, Lisa Gennetian, Charles Michalopoulos, Johanna Walter, Diana
Adams-Ciardullo, Anna Gassman-Pines, Sharon McGroder, Martha Zaslow, Jennifer Brooks, et al.
2001. National evaluation of welfare-to-work strategies: How effective are different welfare-to-work
approaches? Five-year adult and child impacts for eleven programs. Washington, DC: MDRC and
Child Trends.
Haskins, Ron. 2007. Work over welfare: The inside story of the 1996 welfare reform law. Washington, DC:
Brookings Institution Press.
Haskins, Ron, and Greg Margolis. 2015. Show me the evidence: Obama’s fight for rigor and results in social
policy. Washington, DC: Brookings Institution Press.
Hendra, Richard, David H. Greenberg, Gayle Hamilton, Ari Oppenheim, Alexandra Pennington, Kelsey
Schaberg, and Betsy L. Tessler. 2016. Encouraging evidence on a sector-focused advancement strategy:
Two-year impacts from the WorkAdvance demonstration. Washington, DC: MDRC.
Herk, Monica. 2009. The Coalition for Evidence-Based Policy: Its role in advancing evidence- based
reform, 2004–2009. Available from http://coalition4evidence.org/wp-content/uploads/Indep-
assessment-of-Coalitions-work-2009.pdf.
Manning, Willard G., Joseph P. Newhouse, Naihua Duan, Emmett B. Keeler, Arleen Leibowitz, and M.
Susan Marquis. 1987. Health insurance and the demand for medical care: Evidence from a randomized
experiment. American Economic Review 77 (3): 251–77.
50	 THE ANNALS OF THE AMERICAN ACADEMY
Manpower Demonstration Research Corporation. 1980. Summary and findings of the National Supported
Work Demonstration. Pensacola, FL: Ballinger Publishing Company.
Medical Research Council Streptomycin in Tuberculosis Trials Committee. 1948. Streptomycin treatment
for pulmonary tuberculosis. British Medical Journal 2 (4582): 769–82.
Merrill, Sally R., and Catherine A. Joseph. 1980. Housing improvement and upgrading in the housing
allowance demand experiment. Cambridge, MA: Abt Associates Inc.
Munnell, Alicia, ed. 1986. Lessons from the Income Maintenance Experiments: Proceedings of a conference.
Boston, MA: Federal Reserve Bank of Boston, Conference Series 30.
Office of Management and Budget. 2004. What constitutes strong evidence of a program’s effectiveness.
Washington, DC: Office of Management and Budget.
Ramey, Craig T., Joseph J. Sparling, and Sharon Landesman Ramey. 2012. Abecedarian: The ideas, the
approach, and the findings. Los Altos, CA: Sociometrics Corporation.
Schaberg, Kelsey. 2017. Can sector strategies promote longer-term effects? Three-year impacts from the
WorkAdvance demonstration. Washington, DC: MDRC.
Schorr, Lisbeth B. 2012. Broader evidence for bigger impact. Stanford Social Innovation Review 10 (4):
50–55.
Schweinhart, Lawrence J. 2004. The High/Scope Perry Preschool Study through age 40: Summary, conclusions,
and frequently asked questions. Ypsilanti, MI: High/Scope Press.
Slavin, Robert. 15 June 2017. The age of evidence. Huffington Post. Available from http://www.huffing
tonpost.com.
Stevens, Jack, Robyn Lutz, Ngozi Osuagwu, Dana Rotz, and Brian Goesling. 2017. A randomized trial of
motivational interviewing and facilitated contraceptive access to prevent rapid repeat pregnancy
among adolescent mothers. American Journal of Obstetrics and Gynecology 217 (4): 423.e1–423.e9.
Tuttle, Christina Clark, Philip Gleason, Virginia Knechtel, Ira Nichols-Barrer, Kevin Booker, Gregory
Chojnacki, Thomas Coen, and Lisbeth Goble Tuttle. 2015. Understanding the effect of KIPP as it
scales, vol. I, Impacts on achievement and other outcomes final report of KIPP’s Investing in Innovation
Grant evaluation. Cambridge, MA: Mathematica Policy Research.
U.S. Department of Health and Human Services. 27 September 1994. Demonstration proposals pursuant
to section 1115(a) of the Social Security Act; Policies and Procedures. Federal Register 59 (186).
U.S. Food and Drug Administration. 1 April 2017. Adequate and well-controlled studies. 21 Code of
Federal Regulations §314.126.
Wallace, John W. 2011. Review of the Coalition for Evidence-Based Policy. Washington, DC: Coalition for
Evidence-Based Policy. Available from http://coalition4evidence.org/wp-content/uploads/Report-on-
the-Coalition-for-Evidence-Based-Policy-March-2011.pdf.
Whitehurst, Grover J. (Russ). 2018. The Institute of Education Sciences: A model for federal research
offices. The ANNALS of the American Academy of Political and Social Science (this volume).