task 10

American educational policy was rapidly transformed between 1980 and
2001. Accountability was introduced into a sphere that had long been
loosely coupled, both major political parties reevaluated longstanding posi-
tions, and significant institutional control over the schooling shifted to the
federal government for the first time in the nation’s history. These changes
cannot be explained by conventional theories such as interest groups, ratio-
nal choice, and historical institutionalism. Drawing on extensive archival
research and more than 80 interviews, this article argues that this transfor-
mation can be explained by a changed policy paradigm which restructured
the political landscape around education reform. More generally, while pre-
vious scholars have observed that ‘‘policies create politics,’’ it should also be
recognized that ‘‘paradigms create politics.’’

KEYWORDS: A Nation at Risk, accountability, assessment, No Child Left
Behind, paradigm, politics, school reform

Over the course of only one generation, American educational policy hasbeen substantially transformed into the system we have today. As
recently as 1980, states and local districts were primarily in charge of school-
ing; a Republican president was calling for the abolition of the Department
of Education; and the most influential scholarly lens for understanding
schools depicted them as ‘‘loosely coupled systems’’ in which myth and cer-
emony mattered more than academic outputs (Meyer & Rowan, 1977).
Today, under the No Child Left Behind Act of 2001, the federal government
has assumed a greater degree of control over schooling than at any previous
point in the nation’s history. A Republican president led the charge for this
expanded federal role, and demands for accountability for results are so

JAL MEHTA is an assistant professor at the Harvard Graduate School of Education, 447
Gutman Library, 6 Appian Way, Cambridge, MA 02138, USA, e-mail: mehtaja@gse
.harvard.edu. His research focuses on the policy and politics of creating high quality
schooling at scale, with a particular interest in the professionalization of teaching.

American Educational Research Journal
April 2013, Vol. 50, No. 2, pp. 285–324

DOI: 10.3102/0002831212471417
! 2013 AERA. http://aerj.aera.net

at UNIV CALIFORNIA BERKELEY LIB on July 23, 2013http://aerj.aera.netDownloaded from

http://aerj.aera.net

ubiquitous that even one of the leading theorists of loosely coupled systems
has argued that that framework no longer applies (Rowan, 2006).

What explains this transformation? Traditional approaches that privilege
interest groups, rational choice calculations, or path dependent processes
are unable to explain key features of these changes. The most powerful
interest groups in the domain, teachers unions, have frequently been
opposed to the movement towards accountability.1 Rational choice explan-
ations can explain why politicians emphasize testing and accountability in
a context in which the public favors them, but they cannot explain how
or why that context shifted between 1980 and 2001. Path dependent explan-
ations can explain the persistence of long-standing norms against federal
involvement in schooling, but are particularly unsuited to explaining why
there recently has been such a significant departure from this well-worn
path.

This article suggests that the impetus for the transformation was the cre-
ation of a powerful educational paradigm, which crystallized in the famous A
Nation at Risk report. This paradigm, which emerged in the early 1980s and
is still dominant today, holds that educational success is central to national,
state, and individual economic success; that American schools across the
board are substantially underperforming and in need of reform; that schools
rather than social forces should be held responsible for academic outcomes;
and that success should be measured by externally verifiable tests. This par-
adigm has directed the school reform movement over the last 25 years, pro-
ducing a variety of policy efforts that are consistent with its tenets, including
charter schools, public school choice, vouchers, and the subject of this arti-
cle, the growth of state and federal efforts to impose standards and introduce
accountability. These assumptions not only have redirected the policy goals
around schooling; they have restructured the politics of education.
Specifically, under the reign of the A Nation at Risk paradigm, more power-
ful political actors have entered the domain; interest groups have shifted to
embrace the new paradigm; critics out of step with the paradigm have been
rhetorically marginalized; and the venue in which education policy is dis-
cussed has shifted upwards, as the new paradigm has legitimized the claims
of federal and state government to assert increasing control over what had
previously been the province of local districts.2

A Nation at Risk has not been ignored in previous accounts of American
educational history; it is often cited as a critical document in American
school reform (Boyd & Kerchner, 1988; DeBray-Pelot & McGuinn, 2009;
Graham & Gordon, 2003; Guthrie & Springer, 2004; McDermott, 2011).
This article seeks to build on and extend this literature by drawing on
new state-level evidence to explore exactly why A Nation at Risk resonated
so powerfully with such diverse constituencies and how A Nation at Risk re-
shaped state politics. I also look deeper into the past, finding a more diverse
set of antecedents than is usually identified; and further into the future,

Mehta

286
at UNIV CALIFORNIA BERKELEY LIB on July 23, 2013http://aerj.aera.netDownloaded from

http://aerj.aera.net

seeking to specify more precisely how A Nation at Risk affected subsequent
reform efforts.

The article also makes a theoretical contribution by showing the mech-
anisms through which paradigms can shape politics. In its emphasis on para-
digms, this article joins a growing literature that emphasizes the role that
ideational factors play in affecting political outcomes (Beland & Cox, 2010;
Berman, 2001; Blyth, 2002; Somers & Block, 2005; Steensland, 2006; see
Mehta, 2010, for a review). This piece builds upon this growing literature
to illustrate the ways in which a powerful paradigm can restructure the polit-
ical landscape. Paradigms can shift the direction and boundaries of debate,
which actors are involved, and ultimately can provide the impetus for insti-
tutional transformation. In so doing, ideas provide an important complement
to more traditional interest group, rational choice, or institutional explana-
tions. Implications for understanding the role of paradigms in social and
political life are discussed in conclusion.

Changes to Explain

No Child Left Behind is the culmination of nearly two decades of changes
that have transformed American education. Changes are evident in policy, in
institutional responsibility, and in politics. At the policy level, an array of re-
forms has proliferated at the state level since the 1980s, including charter
schools, public school choice, vouchers, and standards-based reform. All of
these reforms have challenged traditional public schools. The most ubiquitous
of these reforms, and the one that became the template for federal policy, is
standards-based reform. Standards-based reform brought together three ele-
ments that its proponents hoped would create systemic change: setting stand-
ards for what students should be expected to do, establishing assessments to
measure progress, and holding schools accountable for progress towards
these goals. Standards-based reform spread through the states beginning in
the early 1990s, was encouraged by the federal passage of Goals 2000 and
the Improving America’s Schools Act in 1994, and became a federal require-
ment under No Child Left Behind.3 Given the pluralism that has traditionally
characterized American schooling—10,000 Democracies, as one prominent
book about local districts labeled it—this movement marks a considerable
shift towards a particular policy vision of school reform.

Not only has a particular vision of schooling become dominant; there is
increasing evidence that the policy emphasis on accountability is penetrating
the technical core of actual school practice, a significant departure from many
previous policy efforts. Historically, educational policy has been seen as pen-
dulum-swinging cycles of policy reform that have done little to alter the
underlying grammar of schooling (Tyack & Cuban, 1995). This view is consis-
tent with Meyer and Rowan’s (1977) famous description of schools as ‘‘loosely
coupled systems’’ that conform to public will in their outward appearance but

How Paradigms Create Politics

287
at UNIV CALIFORNIA BERKELEY LIB on July 23, 2013http://aerj.aera.netDownloaded from

http://aerj.aera.net

not their internal practice. Recent work on the implementation of test-based
accountability suggests that loosely coupled systems are giving way to more
tightly coupled ones, as the attaching of significant stakes to testing has caused
schools and teachers to direct their efforts towards improving student perfor-
mance on those tests (Fuhrman, 1999, 2001). While a significant debate rages
about whether these changes have been good (Peterson & West, 2003), bad
(Meier, 2002), or mixed (Elmore, 2004), what matters for these purposes is
that they clearly have been consequential in how the school system functions,
so much so that even Rowan (2006) now argues that loosely coupled school
systems are a thing of the past.

These changes have also marked a considerable shift in who has institu-
tional control over schooling. Since the nation’s inception, schooling in
America has been controlled by local school districts, with states playing an
important but secondary role. The initial break with state and local control
came with the creation of the Elementary and Secondary Education Act
(ESEA) in 1965, which established the precedent of federal involvement in
schooling and created an ongoing federal funding stream for education.
However, these funds were directed primarily to high-poverty schools, did
not ask for accountability for results, and left almost all decisions about the
governance of schools to localities and states. What is notable about the recent
changes encapsulated in No Child Left Behind is the way in which the federal
government has significantly extended its reach: while still footing less than 10
percent of the bill, it has now extended its reach from high poverty schools to
all public schools, specifying the grades in which students need to be tested,
the pace at which schools need to improve, and the series of escalating con-
sequences if schools do not improve. While schooling remains a function
shared across levels of government, the federal government has greatly
increased its role in shaping the day to day life of all public schools
(McGuinn, 2006; Peterson & West, 2003).

Finally, these changes have been possible only because of considerable
political shifts in the positions of the major parties on educational issues.
Democrats retreated from their long-standing position that the role of the
party was to provide greater funding for high-poverty students to a vision
of school reform that emphasizes accountability as much as spending. The
shift was even greater for Republicans, who moved from President Ronald
Reagan’s pledge to abolish the Department of Education in 1980 to the great-
est expansion of the federal role in the nation’s history when President
George W. Bush signed the No Child Left Behind Act. This article seeks to
explain these policy, institutional, and political changes.

Puzzles Not Fully Explained by Other Approaches

These developments pose a number of problems for the leading theo-
ries in political sociology. While the debate between competing approaches

Mehta

288
at UNIV CALIFORNIA BERKELEY LIB on July 23, 2013http://aerj.aera.netDownloaded from

http://aerj.aera.net

on education policy is not as fully developed as the debate about the welfare
state, the leading explanations for these changes can be grouped under three
headings: interest groups, rational choice, and historical institutionalism.

Interest group explanations emphasize the role that business groups
have played in demanding standards and accountability (Goldberg &
Traiman, 2001; Murphy, 1990; Odden & Marsh, 1990). Previously a nonfactor
in educational politics, business groups are said to have provided powerful
voices supporting standards and accountability at both the state and federal
levels (Ginsberg & Wimpelberg, 1987; Goldberg & Traiman, 2001). There are
two problems with this view. First, a business-centered approach does not
explain why business groups, which historically have been opposed to
school finance reforms in an effort to keep their taxes down (Mazzoni,
1995), have come in recent years to see their interests as lying with school
reform. Explaining this change is critical to understanding their role in this
transformation. Second, a business-driven view of the changes fails to
explain why business was able to impose its will in an increasingly crowded
interest group landscape. Detailed studies of education politics have sug-
gested that since 1980 the number of groups involved in educational reform
has multiplied, with business being only one group among many that have
sought to make a greater claim over education (Mawhinney & Lugg, 2001;
Mazzoni, 1995). The most important of these groups are the teachers unions:
One study suggested that in 43 of 50 states they were the most powerful ac-
tors in educational politics (Thomas & Herbenar, 1991). The larger of these
unions, the National Education Association (NEA), has consistently opposed
efforts to introduce educational accountability. Yet despite its considerable
financial resources and political power, it failed to block movements for
accountability.4 Any interest group explanation has to explain why account-
ability triumphed (even among Democrats) despite objections from the
strongest interest groups in school politics.

A second explanation draws on the rational choice tradition, particularly
the median voter theorem. McGuinn (2006) has done some of the most
detailed reconstruction of the federal politics of this transformation, and
while he is not a rational choice theorist himself, a number of his observa-
tions are consistent with the idea that strategic imperatives have led both
parties to increasingly emphasize education reform. As education rose on
the agenda in recent years, first state and later federal politicians acted stra-
tegically to offer plans for education reform in order to win voters to their
cause. There is considerable merit to this strategic rational choice view. It ex-
plains why both parties have pursued the education issue, why Democrats
were willing to buck the NEA, and why Republicans were willing to sacrifice
long-standing principles against federal control in favor of short-term elec-
toral advantage. But it is also limited in that it assumes much of what needs
to be explained, offering no account of how the context was created within
which these choices became rational.5 Key aspects of the context that need

How Paradigms Create Politics

289
at UNIV CALIFORNIA BERKELEY LIB on July 23, 2013http://aerj.aera.netDownloaded from

http://aerj.aera.net

to be explained include the following: Why is education now a high salience
issue for states and the federal government when that was not the case in the
past? What has primed the public to see the problem in terms that make
standards and accountability the logical solution? Overall, while rational
choice approaches can be important in explaining the strategic behavior
of political actors in pursuing education reform, they do not explain how
the context was created within which these strategic actions were carried
out.

If rational choice is overly focused on a short time horizon, historical in-
stitutionalism is often the preferred approach for those whose questions are
focused on change over longer periods (Orren & Skowronek, 2004; Thelen
& Steinmo, 1992). But while historical institutionalism can explain the base-
line from which the recent changes depart—decentralized federalism that
had long inhibited national action on schools—its long-standing emphasis
on path dependence would seem to foreclose the institutional changes
that are so remarkable about the case under study. In partial answer to
this objection, an important book by Manna (2006) has argued that while
historical institutionalism has traditionally focused on how America’s system
of federalism impedes major policy development, in this case having multi-
ple venues provided opportunities for ‘‘borrowing,’’ with state and federal
developments feeding off one another. Of course, this begs a further ques-
tion: Given that American education had embraced local control of school-
ing since its inception and that the arrangements of federalism had always
permitted borrowing between levels of government, what prompted
increased state and federal involvement over the past 20 years? As is often
true of historical institutionalist approaches, the mechanism for change is
not specified (Clemens & Cook, 1999).

Finally, any theory of these changes that gives causal primacy to any
single set of actors is likely to fall short, since the existing literature on
standards-based reform suggests that many different actors were responsible
for initiating the policy. Standards-based reform was driven by business
groups in Texas, a court decision in Kentucky, state legislators in Utah, the
governor’s office in Michigan, and a state superintendent of schools in
Maryland, to name just five.6 Any workable theory would need to explain
how different actors came to advocate similar policies.

To briefly summarize the puzzles unexplained by these approaches:

1. Why did the agenda status of education rise, encouraging ‘‘rational’’ politicians
to make it a central issue?

2. Why did both parties come to support the reforms, despite the long-standing
historical differences on education between the parties?

3. What explains why different actors—courts, legislators, governors, state bu-
reaucrats, business groups—each initiated similar reforms?

Mehta

290
at UNIV CALIFORNIA BERKELEY LIB on July 23, 2013http://aerj.aera.netDownloaded from

http://aerj.aera.net

4. Why did standards and accountability triumph despite the opposition of the
strongest interest group in the field?

5. Given predictions of institutional ‘‘lock-in’’ and ‘‘path dependence,’’ why was
there such a significant shift away from the institutional patterns that had char-
acterized American schooling since its inception?

The Power of Ideas: How Paradigms Create Politics

Recognizing the limitations of these approaches, a quickly growing liter-
ature on ‘‘ideas and politics’’ has sprung up over the past 15 years (Beland,
2005; Beland & Cox, 2010; Berman, 1998; Campbell, 2002; Hall, 1993; Mehta,
2010; Steensland 2006). Using a variety of different frameworks and lan-
guage, scholars have invoked paradigms (Hall, 1993; Kuhn, 1962), roadmaps
or worldviews (Goldstein & Keohane, 1993), or simply ideas (Berman, 1998)
as ways to explain actors’ commitments to their chosen ends. In areas as
diverse as human rights policy (Sikkink, 1993), airline and trucking deregu-
lation (Derthick & Quirk, 1985), industrial policy (Dobbin, 1994), and the
welfare state (Berman, 1998), research suggests that ideas provide important
templates that guide policy action. Responding to earlier materialist conten-
tions that ideas are largely epiphenomenal, much of this work has sought to
contrast interest-based and ideational approaches, showing that ideas were
important in creating policy even when interest groups were arrayed against
their triumph (Derthick & Quirk, 1985). This work has had considerable
impact in the field, and even those who were at one time skeptical about
the causal role of ideas have begun to incorporate them into their work.7

If ideationally inclined scholars have succeeded in creating a place at
a table previously dominated by Marxist, pluralist, state-centered, and ratio-
nal choice approaches, they have only just begun to build a more sophisti-
cated set of conceptual tools that would specify how ideas matter and how
they interact with other forces (such as interests, institutions, and policy en-
trepreneurs) to affect policy selection and change. In recent years, a few ide-
ationally oriented scholars have started to specify pathways through which
ideas affect politics, but they remain few and far between (Blyth, 2002;
Lieberman, 2002; Steensland, 2006).

In this work, I explore the salience of one particular kind of idea:
a ‘‘problem definition’’ or ‘‘policy paradigm.’’ A problem definition is a partic-
ular way of understanding a complex reality. For example, homelessness can
be seen as the product of a housing shortage, high unemployment, or a lack
of individual gumption. Problem definitions resist efforts to separate the nor-
mative and the empirical, as they generally evoke both normative and
empirical descriptions in ways that are mutually reinforcing.8 The way
a problem is framed has significant implications for the types of policy sol-
utions that will seem desirable, and hence much of political argument is

How Paradigms Create Politics

291
at UNIV CALIFORNIA BERKELEY LIB on July 23, 2013http://aerj.aera.netDownloaded from

http://aerj.aera.net

fought at the level of problem definition. During the conflict stage, ‘‘problem
definition’’ is the appropriate term; when one has triumphed and assumes
the status of a master narrative, it can be called a ‘‘policy paradigm,’’ follow-
ing Kuhn’s view of paradigms as dominant views that preclude significant
dissent.

Here I will focus on paradigms at a middle level of generality—that of
defining a problem within an issue area, in this case education. There are
also shifts in the climate of ideas at the broader level of public philoso-
phies—such as whether the government is seen as the solution or the prob-
lem—that are both informed by and shape the way more specific issues are
regarded (Mehta, 2010). While I focus primarily on the story of how the edu-
cational problem definition was reshaped, I also discuss in passing the
broader shifts in public dialogue around government and social policy as
they are relevant for fully understanding the story.

In particular, this article uses the educational case to investigate one
strand of how ideas matter; more specifically, how paradigms can create pol-
itics.9 While some of the most cited writing about paradigms has focused on
how they shape and reshape the cognitive maps held by key policymakers
(Hall, 1993; Legro, 2000; Weir, 1992), less attention has been devoted to how
paradigms can reshape the political environment around an issue (see also
Baumgartner & Jones, 1993). Once a changed definition of a problem comes
to the fore, I will argue, it has the potential to reshape virtually every aspect
of the politics governing the issue.

One impact of a new problem definition is that it changes the nature of
the debate. A dominant problem definition serves to bound the potential
possibilities of what can be advocated, giving it a powerful agenda-setting
function. Policy entrepreneurs who offer solutions that are consistent with
the broader agenda are elevated, while those whose solutions do not fit
the dominant problem definition are marginalized. The dominant problem
definition also affects who has standing to speak: If the problem is that
schools are not as efficient as for-profits, business leaders become embold-
ened; if the problem is unleashing students’ creativity, then artists and teach-
ers are empowered. Problem definitions not only provide a template for
their proponents; they also can constrain the positions that their opponents
can take.

Another effect of a new problem definition is that it changes the constel-
lation of actors. When new problem definitions come to the fore, new actors
become involved and new cleavages are created.10 New problem definitions
can motivate the formation of new groups, which in turn can have a signifi-
cant effect on subsequent debate. Precisely because these new groups
accept the dominant conception of the problem, they are welcomed by
the broader political environment and can play a critical role in shaping pol-
icy alternatives. New problem definitions can also create opportunities for

Mehta

292
at UNIV CALIFORNIA BERKELEY LIB on July 23, 2013http://aerj.aera.netDownloaded from

http://aerj.aera.net

policy entrepreneurs and experts since existing groups seek guidance on
how to position themselves in a new environment.

A new problem definition can also create an opportunity for major insti-
tutional change. Which actors are motivated and legitimated to act in an area
is dependent in part upon how the area is defined. As we will see, when
education became more heavily defined as an economic issue, state and fed-
eral actors who had previously seen education as largely a local function
were motivated and enabled to act because the issue was now seen as falling
within their jurisdiction. While institutional theories tend to emphasize stabil-
ity, idea-oriented theories can provide an account of major shifts in institu-
tional responsibility.

In sum, once crystallized, a new paradigm not only delimits policy options
to conform to that paradigm (Hall, 1993; Weir, 1992), but it can restructure the
political landscape around an issue and change the agenda status of the issue,
the players involved, their standing to speak, and the venue in which the issue
is debated. In recent years, scholars in American political development
(Campbell, 2003; Pierson, 2003) have seized on Schattschneider’s (1935) obser-
vation that ‘‘policies create politics.’’ Scholars of paradigm changes should rec-
ognize that ‘‘paradigms create politics’’ as well.11

Such a view complements the other explanations above in ways that al-
lows for a fuller explanation of these changes. Interests still matter, but para-
digms can help us understand why actors come to assume the positions that
they did. Rational calculations are still useful for understanding why politi-
cians take the positions they do in the short run, but paradigms help to
explain how the context was created in which those positions came to
seem rational. Historical institutional explanations help us understand the
potency of norms against federal action, while ideas can help us understand
how and why federal actors were able to expand their purview.

Data and Methods

There have been three critical events in the transformation of schooling
since 1980. The first was the publication of the blockbuster 1983 report A
Nation at Risk, a catalytic document that crystallized a new paradigm that
would spark an avalanche of efforts to reform American schools. The second
was the states’ adoption of standards-based reform in the 1990s. The third
was the federal move towards standards-based reform, which built upon
the state reforms and culminated in No Child Left Behind. Each of these
events is examined using careful process tracing (Mahoney, 1999), drawing
on documentary and interview evidence that allows for a detailed recon-
struction of who advocated what and for what reason. Taken at any given
moment in time, process tracing allows us to see which actors, interests, in-
stitutions, and ideas were important in producing a paradigm or a policy out-
come. Taken over two decades, it permits an analysis of how actors’

How Paradigms Create Politics

293
at UNIV CALIFORNIA BERKELEY LIB on July 23, 2013http://aerj.aera.netDownloaded from

http://aerj.aera.net

positions changed over time, a perspective that is critical for an examination
that seeks to understand not only who pushed for what but how they came
to know what to push for.

While there has been previous work on the changing federal politics of
education (Cross 2004; DeBray, 2006; Kosar, 2005; Manna, 2006; McGuinn,
2006), there has been far less that has sought to integrate empirical research
on states into a broader account of the national movement towards
stan

task 10

Task 10