ARTICLE IN PRESS

Teaching and Teacher Education 24 (2008) 610–622

www.elsevier.com/locate/tate

Raising standards in American schools? Problems with

improving teacher quality
Emma Smith
School of Education, University of Birmingham, Birmingham, B15 2TT, UK
Received 12 July 2006; received in revised form 9 May 2007; accepted 26 September 2007

Abstract

The quality of the teacher workforce is a subject of perennial concern in many developed countries. In the United States,
through the No Child Left Behind (NCLB) Act, the federal government has a mandate for reform of teacher education that
is unprecedented in its scale. Essentially the Act demands that every teacher of core academic subjects must be deemed to
be highly qualified in every subject they teach by the end of the 2005–2006 academic year. This paper considers the impact
that NCLB’s teacher quality mandate is having in US schools. By drawing upon school-level data for the state of
California, we examine the progress that this state is making towards meeting NCLB’s mandate and also the role that
teacher quality can play as a determinant of school success. The findings suggest that overall California has a well-qualified
and highly experienced teacher workforce which is relatively equitably distributed among the states’ institutions. On the
other hand, the distribution of California’s students appears to be less fair, with students from poorer homes and certain
ethnic backgrounds being disproportionately represented in the state’s least wealthy and least successful schools. In
addition, the finding that it is student background factors rather than teacher quality characteristics that are the key
determinants of school success, also brings into question the extent to which requiring teachers to improve their subject
content skills will really help close the achievement gaps in California’s schools.
r 2007 Elsevier Ltd. All rights reserved.

Keywords: Teacher quality; No Child Left Behind; Educational standards

1. Introduction performance that exists today, we might well

Over 20 years ago, the US government published
a searing indictment of the state of the American
public school system. The document was called
A nation at risk and it forcefully condemned the
‘rising tide of mediocrity’ that was eroding the
nation’s schools:
If an unfriendly foreign power had attempted to
impose on America the mediocre educational
E-mail address: e.smith@bham.ac.uk
have viewed it as an act of war (NCEE, 1983).
Although the ideology and evidence underpinning
A nation at risk have proven to be controversial
(Berliner & Biddle, 1995), it was to become the most
important US education reform document of the
twentieth century (Ravitch, 2003). Along with falling
or stagnating levels of pupil performance in Amer-
ican public schools, it was the quality of the teaching
profession that excited the most attention. The
authors of the report were particularly concerned

0742-051X/$ - see front matter r 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tate.2007.09.013
 ARTICLE IN PRESS
E. Smith / Teaching and Teacher Education 24 (2008) 610–622 611

that teachers were being drawn disproportionately Through an examination of the early impact of
from the lowest quartile of graduating high school this piece of legislation and an exploration of one
and college students and that in certain shortage state’s efforts to fully comply with this new federal
subjects, such as mathematics, science and English, mandate, this paper considers how close the
teachers were simply not qualified to teach (NCEE, American public school system will come to its goal
1983). Among their recommendations was a call for of having every child taught by a highly qualified
strengthening the teaching profession by raising its professional by the end of the 2005–2006 school
standards for training, entry and professional devel- year. A key tenet behind this new piece of legislation
opment, one aspect of this being the requirement that is that teacher quality is central to school success.
prospective teachers take fewer courses in education Using publicly available data for every school in
and more in the subjects they expect to teach. California, this paper will test this assumption and
According to some commentators, in the two consider the role that ‘teacher quality’, as defined
decades following the publication of A nation at risk under the aegis of NCLB, has on school outcomes.
little has changed (Koret Task Force on K-12 But first we begin by describing the key features of
Education, 2003). Teachers are still being drawn NCLB.
from the bottom ranks of college graduates, and
while teacher salaries may have risen in real terms
2. No Child Left Behind in practice
since the early 1980s, they still lag behind that of
other professions (Hanushek & Rivkin, 2004; Hoxby,
According to NCLB, every teacher of core
2003; Manzo, 2005). Indeed the question of ‘how the
academic subjects1 must be highly qualified by the
nation’s teachers are recruited, prepared and trained
end of the 2005–2006 academic year. By the term
has become the hottest topics in the public and
‘highly qualified’, the legislation stipulates that the
academic discourse regarding education’ (Cochrane-
teacher must have obtained ‘full state certification
Smith, 2005b, p. 3).
as a teacher or passed the State teacher licensing
In 2002, the US government passed into law No
examination, and hold a licence to teach in such a
Child Left Behind (NCLB): a piece of legislation
State’ (US Department of Education, 2002, p. 4).
that has provided the federal government with a
For new teachers (those employed after the start of
mandate for education reform on an unprecedented
the 2002–2003 academic year) this means that they
scale. For the nation’s public school teachers,
must posses at least a bachelor’s degree and pass
NCLB means complex systems of performance
state academic tests or must successfully complete,
and accountability measures aimed at addressing
for each of their teaching areas, an academic major,
concerns over teacher quality and increasing the
or coursework equivalent or a graduate degree.
number of highly qualified teachers in America’s
Veteran teachers must meet the same criteria as
schools. NCLB is arguably the most important piece
newly qualified teachers or must demonstrate
of US educational legislation of the past 35 years.
competence in all the subjects that they teach
In broad terms, the Act links government funding
according to a High Objective Uniform State
to strict improvement and accountability policies
Standard Evaluation, or HOUSSE (US Department
for America’s public schools. Much of the attention
of Education, 2002). Each State is free to develop
already given to NCLB has focused on its mandate
their own HOUSSE criteria but must address
to raise the achievement levels of all students
teachers’ skills in both subject-matter knowledge
(for example, Abedi, 2004; Linn, 2003; Smith,
and teaching practice as well as provide coherent
2005). However, it also legislates for reform in
information about the teacher’s attainment of
the way in which teachers are trained and recruited.
content knowledge in the subjects that they teach
As it applies to teacher quality, NCLB has two
(Centre on Educational Policy, 2005).
key objectives, the first is to ensure that every
In addition to improving teacher quality, NCLB
teacher is highly qualified in the subjects they
mandates for improvement in the basic skills of all
teach and the second is to reduce the barriers to
students. In order to determine whether a school is
becoming a teacher by ‘retooling’ traditional
meeting the requirements of NCLB for student
teacher education programmes and opening up
alternative routes into the profession (US Depart- 1
Core subjects are English, reading or language arts, mathe-
ment of Education, 2004). Both have proven to be matics, science, foreign languages, civics and government,
controversial. economics, arts, history and geography.
 ARTICLE IN PRESS
612 E. Smith / Teaching and Teacher Education 24 (2008) 610–622
achievement, the Act requires that by 2005, states
assess performance annually in grades 3–82 in
English Language Arts and Mathematics, and in
Science by 2007 (US Department of Education,
2002). If students reach the required proficiency
targets in these tests then the school is considered to
have made Adequate Yearly Progress (AYP). If a
school fails to make AYP, a series of sanctions can
be administered by the school district, a situation
akin to being on ‘special measures’ in England. The
form of these sanctions ranges from district-level
monitoring through to giving parents the option to
transfer their children out of ‘failing’ schools and
providing students who remain in the school with
additional tutoring. In more extreme cases, where a
school fails to make AYP for 4 or more consecutive
years, that school can be faced with having to
replace staff, aspects of the curriculum or, at the
extreme, be restructured as a Charter school or one
run by a private company. The AYP status of a
school is therefore crucial in determining the
organisation, funding and operation of a school.
That the federal government has sought to play
such a large role in education through NCLB is
unprecedented. Traditionally the US government,
and Republican administrations in particular, have
played only a limited part in education policy
(Rudalevige, 2003). This level of federal interven-
tion in an area that has long been the domain of the
states has raised concerns about whether NCLB is in
fact constitutional (McColl, 2005). Indeed, from the
perspective of many state and local educators, the
NCLB provisions constitute an ‘unprecedented
federal intrusion into their authority to establish
policies and programmes appropriate to their
respective communities’ (McDermott & Jensen,
2005, p. 45). In the US, federal funds only
contribute around 7–8% of total public school
funding; the remainder is split between the state
(typically around 48%) and local taxes (around
45%), although the proportions do vary (Education
Trust, 2002). With NCLB it is, according to
Margaret Dayton, the Chairman of the Utah
Education Committee, a case of ‘7% of money
dictating 100% of policy’ (Dayton, 2005). Of
course, in theory, states could forgo their federal
funding and so be exempt from many of NCLB’s
mandates. While there is a suggestion that some
states and school districts are considering doing
this, it is likely that any shortfalls in this funding will
2
National curriculum years 4–9.
have a disproportionate impact on disadvantaged
students—the main recipients of these funds
(McDermott & Jensen, 2005).
3. Producing high-quality teachers
No one denies the need for a good teacher. Some
estimations of the value of having a good teacher
can be quite staggering. For example, having 3
consecutive years of good teachers (those placed in
the 85th percentile or above in terms of quality)
could overcome the average achievement deficit
between low-income students and others (Hanush-
ek, 2002; Hanushek & Rivkin, 2004). Value-added
studies in Tennessee have estimated that, at the
extreme, 5th grade students who had experienced
highly ineffective teachers for 3 years were likely to
score around 50 percentile points below their peers
of comparable previous achievement who had been
taught to be highly effective teachers for the same
period (Rivers & Sanders, 2002). However, the key
issue is how to obtain good teachers, especially
when research does not find any ‘systematic link
between teacher characteristics and student out-
comes’ (Hanushek & Rivkin, 2004, p. 7).
Uncertainty over the most effective way to train
the teacher workforce underlies much of the
controversy that has surrounded NCLB’s Highly
Qualified teacher requirements, and no other area
seems to have excited such debate as the issue of
teacher certification. This key area has divided
opinion between those who favour a move towards
a decentralised teacher licensing system governed by
choice and market forces where prospective teachers
can enter the profession by a variety of pathways,
and those who advocate that teacher preparation
should include training in teaching methods and
pedagogy and should remain largely within the
schools of education.
Proponents of the decentralised system favour
pathways into teaching that are free from the
education coursework requirements of many of
the traditional routes (Walsh, 2001) and view the
current system whereby prospective teachers with-
out undergraduate coursework in education have to
cover additional education programmes before
becoming eligible for state certification acts as a
deterrent to highly able individuals who might
otherwise enter teaching (US Department of Educa-
tion, 2002). They place the emphasis on teachers
having high verbal skills and subject expertise rather
than the pedagogic skills taught in schools of
 ARTICLE IN PRESS
E. Smith / Teaching and Teacher Education 24 (2008) 610–622
education, and criticise traditional routes into
teaching for failing to provide teachers with
sufficient subject knowledge to enable them to teach
their subject effectively (Finn, 2003; Walsh, 2001).
On the other hand, those that advocate the existing
centralised system contend that the current certifi-
cation process already assesses prospective teachers
for their skills in content knowledge and verbal
ability, as well as their teaching ability, and that
state tests are already effective in ensuring that only
high-quality teachers are admitted to the profession
(Darling-Hammond, 2000; Darling-Hammond &
Younge, 2002). According to them, the policy
objectives of NCLB for improving teacher quality
essentially mean the ‘dismantling of teacher educa-
tion systems and the redefinition of teacher quali-
fications to include little preparation for teaching’
(Darling-Hammond & Younge, 2002, p. 13). That
NCLB’s focus is almost exclusively on subject-
matter knowledge while appearing to sideline
pedagogy and other professional knowledge and
skills excites particular concern: ‘we should not
confuse a highly qualified taker of tests about
teaching with a highly qualified classroom teacher’
(Berliner, 2005, p. 208, also Cochran-Smith, 2005a;
Kaplan & Owings, 2003). Both of these perspectives
cite a research base that provides the evidence for
their preferred approach and both criticise each
other for the lack of scientific rigour in the studies
they use to support their claims (Darling-Hammond
& Younge, 2002; Walsh, 2001). However, overall
the evidence-base on which both sides base their
claims does remain unclear (Archer, 2002). The lack
of experimental studies and the use of different
outcome measures and definitions of certification,
coupled with selective citation of research findings
can make evidence that supports links between
students’ achievement and teacher certification or
alternative routes into the profession difficult to
determine (Hanushek & Rivkin, 2004).
Despite clear differences over how teachers
should be trained and certified, what both perspec-
tives do share is the belief that teacher character-
istics are one of the key determinants of school
success. Indeed, it is clear that with NCLB, what
was once an academic debate now has a national
focus. This use of the term ‘teacher quality’ is not
unproblematic. Under NCLB, teacher quality is
measured not by observing practicing teachers, or
by measuring the progress of their students; instead
teachers are expected to demonstrate their compe-
tency through their subject knowledge skills, rather
613
than their pedagogic skills. Arguably this might
make teacher quality easier to measure, as what we
are assessing here are some of the measurable
characteristics of teachers, such as whether they
possess a Masters degree and numbers of years of
service, rather than the somewhat more problematic
and arguably more subjective measure of what it is
to be a good teacher. So in this sense, what we are
looking at are proxy measures of teacher quality.
However, under NCLB, these are the characteristics
that count, and it is on this basis, with appropriate
caveats about the meaning of the term teacher
quality, that we proceed with the analysis.
The following section will explore some of the
tensions and consequences of this new piece of
legislation by describing the progress that one case
study state, California, is making towards full
compliance with NCLB’s Highly Qualified Teacher
mandates.
4. No Child Left Behind in California
California is the most populous state in the US, it
has approximately 9000 public schools with around
300,000 teachers teaching over 6 million students,
almost a quarter of whom are English language
learners (NCES, 2005). With its rising living costs
contrasted against blighted urban schools and state-
wide budget deficits—it is has been described as a
‘poster state’ for many of the problems facing those
who wish to train, recruit and retain good teachers
(Quality Counts, 2003). A recent report by the
RAND Co-operation charts the slow decline of the
state’s educational fortunes. Three decades ago
what was once regarded as one of the best education
systems in the US now has among the lowest
performing students, the poorest school facilities
and teacher salaries which have not increased in real
terms since 1969 (Carroll, Krop, Arkes, Morrison,
& Flanagan, 2005). The current struggle to support
education at the minimum levels mandated by the
state’s constitution points to a state requiring long-
term solutions and ‘serious school finance reform’
(Slack, 2005, p. 23).
This section begins by describing some of the key
features of California’s school accountability system
before providing a brief profile of the characteristics
of the state’s school teachers and students. It will
conclude with an examination of the role that
teacher characteristics, and in particular those that
relate to ‘teacher quality’, play in determining the
successful outcomes of schooling. The data used to
 ARTICLE IN PRESS
614 E. Smith / Teaching and Teacher Education 24 (2008) 610–622

undertake this analysis was retrieved from the 2005b). In 2003–2004, the consequences of several
publicly accessible databases of the California consecutive years of failure to meet AYP require-
Department of Education which provide school- ments resulted in over 1 million students being
level data on the achievement and characteristics of eligible to relocate to another school in their district,
all California’s public schools (California Depart- although only 3609 actually chose to do so
ment of Education, 2005c). (California Department of Education, 2005a). As
we can see in Table 1, it is the schools in the poorest
4.1. State testing and school accountability school districts3 that are failing to make AYP. For
example, 45% of the schools in the highest poverty
Since 1999, California has adopted a growth school districts did not make AYP in 2004,
model for monitoring school improvement. The compared with only 18% of schools in low-poverty
Academic Performance Index (API) measures an- districts (Table 1).
nual improvement and differs fundamentally from
NCLB’s status bar model, which assumes that all
4.2. The characteristics of California’s teachers
schools must meet common minimum academic
levels regardless of their starting point (California
In 1999 only 46% of California’s school districts
Department of Education, 2005b). However, school
required teachers to hold full standard state
performance on both measures does differ signifi-
certification in their subjects, and only 30%
cantly. For example, in California in 2004, around
required a subject major or minor (Carroll et al.,
90% of schools met state API performance targets,
2005). However, under NCLB, to be termed ‘highly
while only around 65% of schools made AYP under
qualified’ in California, a teacher of core academic
NCLB (California Department of Education,
subjects must have a bachelor’s degree; a state
2005c). Student progress in Californian schools is
credential or have held an intern certificate/creden-
measured using a series of standardised tests that
tial for no more than 3 years and demonstrated core
are clustered together under the Standardised
academic subject-matter competence (California
Testing and Reporting Programme (STAR). These
Department of Education, 2004). In 1996, Califor-
tests include the California Standards Tests which,
nia implemented an ambitious $1.6 billion a year
for the purpose of NCLB, are used to assess
programme to reduce class sizes from kindergarten
students annually in English Language Arts in
through to the third grade from 30 students to 20.
grades 2–8 and in mathematics in grades 2–7. Since
One consequence of the class size reduction
2005, students have also been tested in science in
programme was a dramatic increase in the recruit-
grades 8 and 10 (California Department of Educa-
ment of inexperienced and uncertified teachers,
tion, 2005a).
resulting in large gaps in the distribution of qualified
In order for a school to make AYP in California,
teachers between schools attended by low-income
it must meet each of the following requirements: at
and non-white students and others (Carroll et al.,
least 95% of the student cohort must participate in
2005; Jepsen & Rivkin, 2002). As a result, in 2000
the STAR programme tests, all students must reach
only 50% of grade 7–12 mathematics teachers had
certain minimum competency levels on these tests,
both a major in their field and state certification
high schools must achieve minimum graduation
(Blank, Langesen, Laird, Toye, & de Mello, 2004).
rates and all schools must demonstrate improve-
In order to meet the mandates of NCLB, California
ment or achieve minimum levels on the API
has invested large sums in programmes to recruit
(California Department of Education, 2005a).
and train teachers. In addition to a well-funded
Schools that fail to make AYP for 2 consecutive
teacher induction programme, the state has in-
years in the same areas are identified for pro-
creased the amount it spends on programmes to
grammes of improvement. For example, a school
recruit teachers from around $14 million in 1998 to
that failed to meet proficiency levels in English
$160 million 2 years later (Carroll et al., 2005).
Language Arts in Year 1 and in Year 2 would be
In Spring 2004, California estimated that around
identified for improvement. On the other hand, a
52% of classes in core academic subjects were
school that failed to meet or exceed proficiency
levels in mathematics in Year 1, but in English 3
District poverty quartile levels are determined by the
Language Arts in Year 2 would not be identified for proportion of students attending schools in the district who are
improvement (California Department of Education, eligible for free school meals.
 ARTICLE IN PRESS
E. Smith / Teaching and Teacher Education 24 (2008) 610–622 615

Table 1
School performance under NCLB, according to school district poverty quartile

Schools making AYP Schools not making AYP

N % N %

First poverty quartile (least poor) 1450 82 320 18

Second poverty quartile 1514 70 644 30
Third poverty quartile 1341 61 870 39
Fourth poverty quartile (most poor) 1233 55 1004 45
All schools 5538 66 2838 34

Source: California Department of Education (2005c).

Table 2 AYP is similar to that in schools that were

Number of classes in the core academic subjects being taught by
‘highly qualified’ teachers
School type Total number of Percentage of core
core academic academic classes
classes taught by ‘highly
qualified’ teachers
All schools 630,647 52
High-poverty 153,922 40
schools
Low-poverty 165,591 60
schools
Source: California Department of Education (2005a, p. 37).
taught by ‘highly qualified’ teachers, and gave a
figure of 40% for classes in ‘high poverty’ schools
(Table 2). Estimates such as these represent a
relatively low starting point from which the state
has to meet its targets for compliance with NCLB’s
Highly Qualified Teacher requirements by
2005–2006.
Although only just over half of core academic
classes across the state were being taught by ‘highly
qualified’ teachers in 2004, this figure does vary
slightly between schools with differing levels of
success with regard to making AYP. For example,
47% of classes in schools that failed to make AYP
were being taught by teachers who were NCLB
compliant, compared with 54% of classes in schools
that made AYP (Table 3). Even so, and NCLB’s
teacher quality targets notwithstanding, California’s
teachers appear to be well-trained and experienced
professionals. Around 90% of teachers have teach-
ing credentials, while over a third posses a Masters’
degree or higher (Table 3). The proportion of
teachers who are teaching on emergency certificates
also remains quite small. Overall, the distribution of
qualified teachers among schools that do not make
successful on this indicator. However, as we have
seen earlier, just because teachers have a bachelor’s
degree and certification to teach, it does not mean
that they are considered to be ‘highly qualified’
under NCLB—that designation depends on them
additionally demonstrating sufficient content
knowledge for their subject areas.
A concern among many commentators is that it is
teachers with less experience and with fewer
teaching credentials who are disproportionately
teaching in California’s poorest and non-white
schools (Quality Counts, 2003). However, the
evidence available to us from the state’s publicly
available database only offers limited support for
this contention. Despite around 15% of California’s
workforce being newly employed (Carroll et al.,
2005), their age and experience profiles appear to be
evenly distributed across the different types of
schools (Table 4). The only indicator on which
there are clear differences in ‘teacher quality’
between schools that do and do not make AYP is
in the number of classes taught by NCLB compliant
teachers (there is no disaggregated data readily
available that cites the numbers of teachers who are
NCLB compliant). The reason for this difference is
unclear, perhaps schools that are meeting AYP have
been able to invest in recent professional training
programmes in order to assist their staff in
becoming NCLB compliant, perhaps other concerns
in schools who are failing to make AYP mean that
professional training is not given the same priority
as in other schools.
In addition, teacher qualifications in California’s
most and least wealthy school districts are also
relatively evenly distributed. Although slightly more
teachers in schools in the most wealthy school
districts have a Masters’ degree or higher, it is still
the case that around a third of all California’s
 ARTICLE IN PRESS
616 E. Smith / Teaching and Teacher Education 24 (2008) 610–622

Table 3
Teacher qualifications in California, according to school type

Schools making AYP Schools not making AYP All schools

N Mean N Mean N Mean

Classes taught by NCLB compliant teachers (%) 5872 54 3048 47 8920 52

Completed a teacher preparation programme (%) 5969 94 3149 89 9118 92
Teaching with emergency credentials (%) 5969 4 3149 6 9118 4
Teachers with an MA degree or higher (%) 5974 35 3155 34 9129 35

Source: California Department of Education (2005c).

Table 4
Teacher experience in California, according to school type

Schools making AYP Schools not making AYP All schools (N ¼ 9163)
(N ¼ 5999) (years) (N ¼ 3164) (years) (years)

Mean teaching experience (district) 11 10 11

Mean teaching experience (total) 14 13 14
Mean age of teachers 44 44 44

Source: California Department of Education (2005c).

70
60
50
40
30
20
10
most wealthy least wealthy all schools
0
1994 1995 1996 1997 1998 1999 2000 2001 2002 2003

Fig. 1. Teachers with Masters’ degrees or higher, according to the relative wealth of school districts. Source: California Department of
Education (2005b).

teachers possess this qualification, regardless of
whether they teach in high- or low-poverty-school
districts—a pattern which has remained relatively
constant over the past decade (Fig. 1).
There is, however, some evidence to suggest that
schools that are located in the richest school
districts do have fewer teachers with emergency
certificates than schools in the least wealthy
districts. For example, in 2003, 58% of schools in
the most wealthy districts did not have any teachers
with emergency certificates, compared with 42% of
schools in the least wealthy districts. Even so, the
proportion of teachers with emergency credentials
has fallen steadily since 1999, when around 11% of
teachers had this certification, so that by 2004 only
around 3% of California’s 353,097 teachers were
teaching on emergency certificates (California De-
partment of Education, 2005c). Some researchers
attribute this ‘appearance of a decrease’ (Goe, 2002)
in the number of emergency credentialed teachers to
the addition of a new category of ‘pre-intern’
teachers. Pre-internship is the next step up the
teacher credentialing ladder and encouraging tea-
chers to enrol on preparation programmes as
teacher interns thereby lifts them out of the
‘emergency teacher’ category. However, the propor-
tion of teachers registered as ‘pre-interns’ has, at
around 2% of the total teacher population remained
relatively stable since 1999. The same is also true for
the proportions of teachers registered on other
 ARTICLE IN PRESS
E. Smith / Teaching and Teacher Education 24 (2008) 610–622 617

credentialing programmes, either with universities
or school districts, where the numbers have stayed
at around 1–2% of the teacher population since
1999 (California Department of Education, 2005c).
4.3. The characteristics of California’s students
According to the California Department of
Education, approximately 45% of children enrolled
in kindergarten to 12th grade are Hispanic or
Latino, 34% are white, 11% are Asian, Filipino or
Pacific Islander, 8% are African American, and
almost 1% are Native American. Together, these
students speak more than 56 different languages
(California Department of Education, 2005c). In-
deed it is students with limited English skills (here
termed English learners, or EL), along with those
from the African-American and Hispanic commu-
nities who are more likely to be taught in schools in
the least wealthy school districts (Table 5). For
example, 43% of students enrolled in schools in the
least wealthy school districts are designated as being
EL, compared with only 15% of learners in the
most wealthy districts. Perhaps unsurprisingly
students from economically disadvantaged homes
also overwhelmingly attend schools in the lowest
poverty quartiles. On this indicator, students
designated as having Special Educational Needs
(SEN) are more evenly distributed among the
school districts.
Similarly, schools that made AYP tested higher
proportions of students from white and Asian
backgrounds and fewer students from economically
disadvantaged and Hispanic homes, compare with
schools that failed this indicator. As we can see in
Table 6, 66% of the learners in schools that did not
make AYP came from economically disadvantaged
homes, compared with 46% of the learners in
schools that were successful on this indicator.
Of course, testing higher proportions of certain
groups of students in ‘failing’ or ‘high poverty’
schools is only really an issue if these students
actually achieve lower results. Using the English
Language Arts test as an example, Table 7 shows
the proportions of students from the different
student groups who achieved or surpassed profi-
ciency levels in 2004. Students from the African-
American and Hispanic communities were less

Table 5
Mean percentage of students enrolled for English Language Arts tests, according to poverty quartile

First quartile (most wealthy) Second quartile Third quartile Fourth quartile (least wealthy)
N ¼ 1752 (%) N ¼ 2120 (%) N ¼ 2176 (%) N ¼ 2203 (%)

African-American 4 7 10 10
Asian 12 7 5 5
Hispanic 21 36 45 58
White 56 43 33 21
Economically 21 44 60 77
disadvantaged
English learners 15 23 32 43
SEN 10 11 11 11

Source: California Department of Education (2005c).

Table 6
Mean percentage of students enrolled for English Language Arts tests

Schools making AYP Schools not making AYP All schools (N ¼ 9060) (%)
(N ¼ 5916) (%) (N ¼ 3144) (%)

African-American 7 11 8
Asian 8 4 7
Hispanic 34 54 41
White 44 25 37
Economically disadvantaged 46 66 53
English learners 24 37 28
SEN 10 12 11

Source: California Department of Education (2005c).

 ARTICLE IN PRESS
618 E. Smith / Teaching and Teacher Education 24 (2008) 610–622
Table 7
Mean percentage of students making AYP (English Language
Arts test)
Number of Students making AYP
schools (%)
African-American 4314 30
Asian 4194 58
Hispanic 7420 29
White 6900 51
Economically 7599 27
disadvantaged
English learners 6790 23
SEN 6999 16
All students 8340 39
Source: California Department of Education (2005c).
likely to reach minimum proficiency levels on the
state literacy assessment than students from White
or Asian homes. EL, students with SEN and those
from economically disadvantaged homes also
achieved relatively low proficiency rates on this test.
Thus it seems that on the one hand California has
a highly educated and experienced teacher work-
force where, regardless of whether the school is
considered to be ‘failing’ under NCLB or is located
in a high-poverty-school district, around a third of
teachers have at least a Masters’ degree and all have
similar levels of teaching experience. On the other
hand, higher proportions of teachers in the least
wealthy school districts do not have full teacher
certification, although the numbers of teachers in
this category continues to fall across the state.
However, and more crucially as far as the NCLB
mandate for improvement is concerned, only
around half of all classes are being taught by
teachers who are considered to be ‘highly qualified’.
Although the number of these classes is lowest in
high-poverty-school districts, at around 42%, it is
also an issue for the most wealthy districts—where
around 58% of classes were taught by NCLB
compliant staff. To date, whether the state will
meet the requirement of having 100% of classes
taught by highly qualified teachers by the end of the
2005–2006 academic year does seem unlikely
(Keller, 2006). Indeed, although California reduced
the percentage of under-qualified teachers by an
average annual rate of 9% between 2000 and 2003.
In order to comply with NCLB, the state would
have to continue to reduce the numbers of these
teachers at a rate of 60% a year (Futernick, 2003).
While, to some extent, the distribution of
California’s teachers appears quite equitable, the
distribution of certain subgroups of students does
not. For example, students from the Hispanic
community—the majority ethnic group in Califor-
nia—are overwhelmingly attending schools in the
poorest school districts which, in turn, are failing to
reach minimum standards under NCLB.
4.4. ‘Teacher quality’ and student success
The Teaching Commission believes that quality
teachers are the critical factor in helping people
overcome the damaging effects of poverty, lack
of parental guidance, and other challenges (The
Teaching Commission, 2004, p. 14).
In California, high-quality school-level data that
has been collected over several years allows for a
closer examination of at least one of the determi-
nants of school success: performance in state-wide
tests. By considering the influence of both teacher-
and student-level variables on school outcomes, we
can examine the extent to which teacher character-
istics, many of which are taken as proxies of teacher
quality under NCLB, can influence school outcomes
at the aggregate level. The outcome measure used
was the proportion of students who demonstrated
that they were proficient or higher on state tests in
English Language Arts, although similar trends
were found for the mathematics assessments. Once
again, the data used in the analysis was retrieved
from publicly accessible datasets on the California
Department of Education Data and Statistics
website (California Department of Education,
2005c).
A number of linear regression models were
carried out. Each considered the amount of variance
in a school’s test outcome that could be explained
by student and teacher characteristics. Variables
were selected for inclusion in the models based on
the results of the descriptive analysis described
earlier. Attempts were also made to include
variables which contributed to the most parsimo-
nious model. The most powerful model related the
proportion of students who achieved proficiency
levels or higher in the English Language Art tests to
a range of student background factors, such as the
proportion of students who received free school
meals as well as the students’ minority background
and language status. The most powerful teacher-
level variables were those linked to teacher char-
acteristics, namely the proportion of female teachers
and African–American teachers in a school, and
variables linked to teacher quality, such as the
 ARTICLE IN PRESS
E. Smith / Teaching and Teacher Education 24 (2008) 610–622 619

number of classes taught by NCLB compliant accounted for by student-level variables, with the
teachers, the number of teachers who were certified teacher-level variables only explaining 3% of the
and the number who held a Masters’ qualification variance. The standardised b coefficients give us an
or higher. In the final model the student- and idea of the strength of the relationship between
teacher-level variables were entered as separate these variables and the test outcome. For example,
blocks. As our interest here was with students’ test the percentage of students in a school who receive
outcomes, it makes sense that the block containing free school meals has a strong negative relationship
student-level variables was entered into the model with test outcomes. This relationship is so strong
first, followed by a second block of teacher-level that it suggests that if every other variable in this
variables. The amount of variance in school model remained the same, for every 1% increase in
examination outcomes that can be explained by the number of students with free school meals, the
the student and teacher variables and the model proportion of students achieving proficiency levels
coefficients for the multiple linear regression analy- or higher in the ELA test would go down by 0.6%.
sis are given in Table 8. Similarly, for a school with a 10% increase in free
Table 8 shows that 69% of the variance in the school meal students, proficiency levels would fall
English Language Arts test outcomes can be by 6% and so on. It is also interesting to note that
the teacher-level variables that have the strongest
relationship with test outcomes are related to
teacher characteristics, the proportion of female or
Table 8 African-American teachers in a school, and not to
Standardised beta coefficients for proficiency in English Lan-
guage Arts testsa
teacher quality, for example, the proportion of
teachers with Masters’ degrees. Variables relating to
Standardized t Sig. the experience of teachers and the numbers teaching
beta coefficients on emergency certificates had no impact on the
Block 1: Student characteristics model and so were excluded from the analysis. It is
Constant 222.2 0.000 important to remember that in this analysis several
% Free school meal 0.6 66.2 0.000 different models combining a range of teacher- and
students in school student-level characteristics were carried out. In
% English learners 0.1 9.9 0.000
each model the results consistently provided no
Number of students in 0.08 12.9 0.000
school evidence to support the contention that student
% Hispanic students 0.3 28.1 0.000 characteristics are ‘less influential in predicting
achievement levels than variables assessing the
Block 2: Teacher characteristics
Constant 38.5 0.000 quality of the teaching force’ (Darling-Hammond,
% free school meal 0.6 56.9 0.000 2000). What the model does suggest is that teacher
students in school characteristics, many of which are considered to be
% English learners 0.08 6.7 0.000 proxies for teacher quality under NCLB, are less
Number of students in 0.1 21.6 0.000
influential than student characteristics in predicting
school
% Hispanic students 0.3 29.9 0.000 student academic outcomes.
% African-American 0.1 14.0 0.000
teachers 5. Discussion
% Female teachers 0.1 21.1 0.000
% Teachers with MA 0.05 7.3 0.000
The NCLB Act is certainly equitable in its intent
or higher
% Completed a 0.01 2.0 0.048 of ensuring that every student is taught by a ‘highly
teacher preparation qualified’ teacher. However, the definition of ‘highly
programme qualified’ is linked only to an individual’s knowl-
% Classes taught by 0.02 2.7 0.007 edge of the subjects that they teach as defined by
NCLB compliant
content tests that can be overly complicated and
teachers
confusing (Berliner, 2005). Of course, ‘teacher
Block 1: Student variables only R2 ¼ 0.69 quality’ is an elusive measure and there is plenty
Block 2: Student and teacher variables R2 ¼ 0.72
of scope for arguing that holding a Masters’ degree
a
Descriptive statistics related to this model are presented in does not necessarily confer quality. However, in the
Appendix A. context of NCLB, these are the types of indicators
 ARTICLE IN PRESS
620 E. Smith / Teaching and Teacher Education 24 (2008) 610–622
that count. There are no provisions under NCLB’s
highly qualified teacher criteria to link teacher
performance directly to student achievement, or to
assess classroom performance. Indeed some alter-
native routes into teaching do not require that
prospective teachers spend any time in teaching
practice at all (US Department of Education, 2004).
More fundamentally, the evidence base under-
pinning many of the key aims of NCLB’s highly
qualified teacher provisions is unclear. It is not clear
what qualities make a good teacher. High verbal
skills and strong subject knowledge may be among
them, but any list is unlikely to be exhaustive. The
research evidence is also unclear about the value of
a teaching certificate, or the effectiveness of alter-
native routes into teaching.
In California, the teacher workforce largely
comprises well-qualified and highly experienced
professionals. Where differences do exist, they
indicate that schools that fail to make AYP have
slightly more teachers who have not completed
teacher preparation programmes. But more worry-
ingly perhaps for all schools, the number of core
academic classes being taught by NCLB compliant
‘highly qualified’ teachers, appears to fall short of its
100% target. On other potential indicators of
quality, such as the number of experienced teachers
or those with Masters’ degrees, the distribution of
teachers among the state’s schools seems reasonably
equitable. In contrast, the distribution of students
among the states’ ‘failing’ or high-poverty schools
does appear to be less fair. Students from economic-
ally disadvantaged homes along with those from the
Hispanic community are disproportionately attend-
ing schools that are located in the highest poverty
school districts or are failing to reach minimum
competency levels on state tests. This later finding
will, of course, come as no surprise to educators in
the US where the historical, economic, social and
moral consequences of the achievement gap are of
paramount concern (for example, Ladson-Billings,
2006).
With regard to the determinants of school
success, here taken to be performance in high-stakes
tests, it is clear that the factors most closely linked
to school outcomes are not related to the quality of
teachers at all. Rather, it is the nature of the school
student population that is most likely to determine
how a school performs on the state’s accountability
tests and in turn makes AYP. This is not to say that
teachers cannot make a difference to individual
students; however, as this analysis has shown, at
aggregate level, there is little evidence of a ‘teacher
effect’.
In her Presidential Address to the 2005 meeting of
the American Educational Research Association,
Marilyn Cochrane-Smith drew attention on the
emergence of a ‘new teacher education’ (Cochrane-
Smith, 2005b), in which teacher education was being
constructed as a policy problem, based on research
and evidence and driven by outcomes. Within each
of these three elements there are mixed opportu-
nities and challenges for those involved in the
training and development of practitioners. Within
the policy context, she argues that teacher education
is considered by policymakers to be the key link to
teacher quality. However, as we have seen in the
discussion above, there is little consensus and
limited research evidence to tell us what constitutes
teacher quality and how to capture it when we find
it. Certainly under NCLB, the key element of
teacher quality is subject competency. However, as
we have seen in California, around one third of
teachers possess a Masters’ degree qualification and
these teachers can be found equally among the
state’s schools—both high and low poverty.
Concerns about the quality of teaching profession
are not confined to the United States. In England
and Wales teacher training is now largely a
postgraduate qualification that is undertaken fol-
lowing a subject-specific 3 or 4-year undergraduate
degree programme. Of course, possessing a high
level of subject knowledge is in itself no guarantee of
teaching quality and concerns about the quality of
teacher trainees do persist. One consequence of this
is that Initial Teacher Training programmes and
teacher trainees in England are subjected to regular
quality inspections by the schools inspectorate,
Ofsted. However, even with such a closely regulated
system, there is still uncertainty that our teachers
are of the ‘highest quality’. For example, in England
students on Initial Teacher Training courses that
have been awarded the lowest grades for teaching
quality by Ofsted Inspectors are just as likely as
those on higher performing courses to gain Quali-
fied Teacher Status and a subsequent teaching post
(Ref 2007).
Cochrane-Smith (2005b) goes on to argue that
prior to the mid-1990s, the emphasis in teacher
education was mainly about the pedagogical process
and the inputs of education, whereas more recently
the focus has shifted and is much more about
outcomes, and test outcomes in particular. The
relationship between teacher quality and test
 ARTICLE IN PRESS
E. Smith / Teaching and Teacher Education 24 (2008) 610–622
outcomes is problematic for several reasons. First, it
assumes that test outcomes are the most tangible
outcome of schooling, it pays little regard to the
school’s and its teachers’ role in helping form
socially and morally responsible citizens. Secondly
and perhaps most crucially, it ignores that fact
that teachers alone cannot solve all the problems
in a nation’ schools. The relationship between social
exclusion, poverty, inadequate housing, poor
health, education under-funding and schooling is
perhaps overlooked. As we have seen in California,
it is the pupils not the teachers who are inequitably
distributed among the state’s schools and it is
the pupil characteristics and not the teachers’
that are most closely linked to the one crude
outcome of success: test performance. However,
while the early results from NCLB’s Highly
Qualified Teacher mandate do give much to be
concerned about, there are also plenty of positives.
For example, states now have, or are close to
having, detailed records about the qualification and
certification of their teachers and schools of educa-
tion now have much more information than ever
before on the preparation, development and career
progression of their teaching graduates (Cochrane-
Smith, 2005b).
Appendix A
See Table A1 for further details.
Table A1
Linear regression model, descriptive statistics
Mean
% students proficient plus in 39.2
English Language Arts
% free school meal students in 51.1
school
% English learners 30.1
Number of students in school 765.3
% Hispanic students 42.4
% African-American teachers 4.2
% Female teachers 75.9
% Teachers with MA or higher 34.1
% Completed a teacher 92.4
preparation programme
% Classes taught by NCLB 52.9
compliant teachers
Number of schools ¼ 7888
621
References
Abedi, J. (2004). The No Child Left Behind Act and English
language learners: Assessment and accountability issues.
Educational Researcher, 33(1), 4–14.
Archer, J. (2002). Research: Focusing in on teachers. Education
Week, 21(29), 36–39.
Berliner, D. C. (2005). The near impossibility of testing for
teacher quality. Journal of Teacher Education, 56(3), 205–213.
Berliner, D. C., & Biddle, B. J. (1995). The manufactured crisis:
Myths, fraud and the attack on America’s Public Schools.
Massachusetts: Perseus Books.
Blank, R. K., Langesen, D., Laird, E., Toye, C., & de Mello, V.
B. (2004). Meeting NCLB goals for highly qualified teachers:
Estimates by state from survey data. Education Policy
Analysis Archives, 12(70).
California Department of Education. (2004). NCLB teacher
requirements: A resource guide. Retrieved from /www.cde.ca.-
gov/nclb/sr/tq/documents/nclbresourceguide.pdfS, July 2005.
California Department of Education. (2005a). California’s con-
solidated state performance report: Part I. Retrieved from
/www.cde.ca.gov/nclb/sr/rt/documents/usdecsprjanfinal.docS,
July 2006.
California Department of Education. (2005b). 2005 Accountability
progress report: information guide. Retrieved from /www.cde.ca.
gov/ta/ac/ay/documents/infoguide05.pdfS, October 2005.
California Department of Education. (2005c). California basic
education data system. Retrieved from /www.cde.ca.gov/ds/
sd/cb/S, June 2006.
Carroll, S. J., Krop, C., Arkes, J., Morrison, P. A., & Flanagan,
A. (2005). California’s K-12 public schools: How are they
doing? Santa Monica: The RAND Corporation; Retrieved
from /www.rand.org/S, June 2005.
Centre on Education Policy. (2005). From the capital to the
classroom: Year 3 of the No Child Left Behind Act. Washington,
DC: Centre of Education Policy; Available at /www.ctredpol.
org/pubs/nclby3/press/cep-nclby3_21Mar2005.pdfS.
Cochran-Smith, M. (2005a). NCLB: Three years and counting.
Journal of Teacher Education, 56(2), 99–103.
Cochran-Smith, M. (2005b). The new teacher education: For
better or for worse? Educational Researcher, 34(7), 3–17.
Darling-Hammond, L. (2000). Teacher quality and student
achievement: A review of state policy evidence. Education
Std. Dev. Policy Analysis Archives, 8(1).
Darling-Hammond, L., & Younge, P. (2002). Defining ‘highly-
20.0 qualified’ teachers: What does ‘scientifically-based’ research
actually tell us? Educational Researcher, December, 13–25.
29.9 Dayton, M. (2005). The future of the No Child Left Behind Act.
Speech given at the Cato Institute Policy Forum, 31 May
24.4 2005. Available at /www.cato.org/events/050531pf.htmlS,
627.6 July 2005.
29.0 Education Trust. (2002). The funding gap: Low-income and
9.4 minority students receive fewer dollars. The Education Trust,
17.3 August 2002 /www2.edtrust.org/EdTrust/Product+Catalog/
16.9 special+reports.htm#2003S, accessed May 2004.
11.1 Finn, C. E. (2003). Teacher reform gone astray. In P. E. Peterson
(Ed.), Our schools and our futureyAre we still at risk?.
32.8 California: Hoover Institution Press.
Futernick, K. (2003), California and the Federal No Child Left Behind
Left Behind Act. Education Commission for the States; Retrieved
from /www.edfordemocracy.org/TQI/index.htmS, June 2005.
 ARTICLE IN PRESS
622 E. Smith / Teaching and Teacher Education 24 (2008) 610–622
Goe, L. (2002). Legislating equity: The distribution of emergency
permit teachers in California. Education Policy Analysis
Archives, 10(42); Retrieved November 2005 from /http://
www.epaa.asu.edu/epaa/v10n42/S.
Hanushek, E. A. (2002). Teacher quality. In L. T. Izumi, & W.
M. Evers (Eds.), Teacher quality. Hoover Institution Press
Publication No. 505.
Hanushek, E. A., & Rivkin, S. G. (2004). How to improve the
supply of high-quality teachers. In D. Ravitch (Ed.),
Brookings papers on education. Washington, DC: Brookings
Institution Press.
Hoxby, C. M. (2003). What has changed and what has not? In P.
Peterson (Ed.), Our schools and our futureyAre we still at
risk?. California: Hoover Institution Press.
Jepsen, C., & Rivkin, S. (2002). Class size reduction, teacher
quality and academic achievement in California Public Ele-
mentary Schools. San Francisco: The Public Policy Institute of
California.
Kaplan, L. S., & Owings, W. A. (2003). NCLB: The politics of
teacher quality. Phi Delta Kappan, 687–692.
Keller, B. (2006). No state meeting teacher provision of ‘no child’
law. Education Week, 24th May 2006.
Koret Task Force on K-12 Education. (2003). ‘y are we still at
risk?’, Findings and recommendations. In P. Peterson (Ed.),
Our schools and our futureyAre we still at risk?. California:
Hoover Institution Press.
Ladson-Billings, G. (2006). From the achievement gap to the
education debt: Understanding achievement in US Schools,
2006 Presidential Address to the American Educational
Research Association. Educational Researcher, 35(7), 3–12.
Linn, R. L. (2003). Accountability: Responsibility and reasonable
expectations. Educational Researcher, 32(7), 3–13.
Manzo, K. K. (2005). Study finds teachers are losing ground on
salary front. Education Week, 1 September 2005, p. 12.
McColl, A. (2005). Tough call: Is NCLB constitutional? Phi
Delta Kappan, 86, 604–610.
McDermott, K. A., & Jensen, L. S. (2005). Dubious sovereignty:
Federal conditions of aid and the NCLB act. Peabody Journal
of Education, 80(2), 39–56.
National Centre for Educational Statistics. (2005). The nation’s
report card: State profiles. Retrieved from /http://www.nce-
s.ed.gov/nationsreportcard/states/S, July 2005.
National Commission on Excellence in Education (NCEE).
(1983). A nation at risk. Washington, DC: US Department
of Education, US Printing Office.
Quality Counts. (2003). Quality counts: California. Education
Week, 22(17), 115–144.
Ravitch, D. (2003). A historic document. In P. E. Peterson (Ed.),
Our schools and our futureyAre we still at risk?. California:
Hoover Institution Press.
Rivers, J. C., & Sanders, W. L. (2002). Teacher quality and equity
in educational opportunity: Findings and policy implications.
In L. T. Izumi, & W. M. Evers (Eds.), Teacher quality.
Hoover Institution Press Publication No. 505.
Rudalevige, A. (2003). No Child Left Behind: Forging a
constitutional compromise. In P. Peterson, & M. West
(Eds.), No Child Left Behind? The politics and practice of
school accountability. Washington, DC: The Brookings
Institute.
Slack, J. L. (2005). New report details not-so-golden state
of California Education. Education Week, 12 January 2005,
pp. 1, 23.
Smith, E. (2005). Raising standards in American schools: The
case of No Child Left Behind. Journal of Education Policy,
20(4), 507–524.
The Teaching Commission. (2004). Teaching at risk: A call to action,
The Teaching Commission; Retrieved July 2006 from
/www.theteachingcommission.org/press/FINAL_Report.pdfS.
US Department of Education. (2002), No Child Left Behind:
Executive summary. /www.ed.gov/nclb/overview/intro/presi-
dentplan/page_pg3.htmlS, accessed June 2004.
US Department of Education. (2004). Meeting the highly qualified
teachers challenge: The secretary’s second annual report on
teacher quality, Washington, DC.
Walsh, K. (2001). Teacher certification reconsidered: Stumbling
for quality. Maryland: The Abell Foundation Retrieved June
2006, from /www.educationnext.org/unabridged/20021/
walsh.pdfS.